diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,89340 @@ +{ + "best_metric": 0.029851149767637253, + "best_model_checkpoint": "final_models/lait_bur_phi/checkpoint-12750", + "epoch": 6.0, + "eval_steps": 500, + "global_step": 12750, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.00047058823529411766, + "grad_norm": 286.8828638287195, + "learning_rate": 1.5e-07, + "loss": 20.7391, + "step": 1 + }, + { + "epoch": 0.0009411764705882353, + "grad_norm": 737.308100666324, + "learning_rate": 3e-07, + "loss": 21.0523, + "step": 2 + }, + { + "epoch": 0.001411764705882353, + "grad_norm": 343.28950077761675, + "learning_rate": 4.5e-07, + "loss": 20.5542, + "step": 3 + }, + { + "epoch": 0.0018823529411764706, + "grad_norm": 753.7143641533416, + "learning_rate": 6e-07, + "loss": 21.0367, + "step": 4 + }, + { + "epoch": 0.002352941176470588, + "grad_norm": 642.0536043207344, + "learning_rate": 7.499999999999999e-07, + "loss": 20.5499, + "step": 5 + }, + { + "epoch": 0.002823529411764706, + "grad_norm": 263.5292251439027, + "learning_rate": 9e-07, + "loss": 20.576, + "step": 6 + }, + { + "epoch": 0.0032941176470588237, + "grad_norm": 308.3263662371733, + "learning_rate": 1.05e-06, + "loss": 19.9334, + "step": 7 + }, + { + "epoch": 0.0037647058823529413, + "grad_norm": 362.75864530768496, + "learning_rate": 1.2e-06, + "loss": 21.1152, + "step": 8 + }, + { + "epoch": 0.0042352941176470585, + "grad_norm": 343.304528736484, + "learning_rate": 1.3499999999999998e-06, + "loss": 20.2133, + "step": 9 + }, + { + "epoch": 0.004705882352941176, + "grad_norm": 253.94132113145895, + "learning_rate": 1.4999999999999998e-06, + "loss": 19.047, + "step": 10 + }, + { + "epoch": 0.0051764705882352945, + "grad_norm": 258.88391587475076, + "learning_rate": 1.6499999999999997e-06, + "loss": 19.173, + "step": 11 + }, + { + "epoch": 0.005647058823529412, + "grad_norm": 376.3654713625469, + "learning_rate": 1.8e-06, + "loss": 20.2885, + "step": 12 + }, + { + "epoch": 0.00611764705882353, + "grad_norm": 364.4827629345546, + "learning_rate": 1.9499999999999995e-06, + "loss": 18.8647, + "step": 13 + }, + { + "epoch": 0.006588235294117647, + "grad_norm": 627.0602190200464, + "learning_rate": 2.1e-06, + "loss": 18.3534, + "step": 14 + }, + { + "epoch": 0.007058823529411765, + "grad_norm": 251.37560440407293, + "learning_rate": 2.2499999999999996e-06, + "loss": 18.5301, + "step": 15 + }, + { + "epoch": 0.0075294117647058826, + "grad_norm": 293.88481491831857, + "learning_rate": 2.4e-06, + "loss": 18.2008, + "step": 16 + }, + { + "epoch": 0.008, + "grad_norm": 1099.1714710658032, + "learning_rate": 2.55e-06, + "loss": 16.2574, + "step": 17 + }, + { + "epoch": 0.008470588235294117, + "grad_norm": 346.69094533632824, + "learning_rate": 2.6999999999999996e-06, + "loss": 16.7245, + "step": 18 + }, + { + "epoch": 0.008941176470588235, + "grad_norm": 181.98937811235362, + "learning_rate": 2.85e-06, + "loss": 16.5907, + "step": 19 + }, + { + "epoch": 0.009411764705882352, + "grad_norm": 526.658876320406, + "learning_rate": 2.9999999999999997e-06, + "loss": 15.5959, + "step": 20 + }, + { + "epoch": 0.00988235294117647, + "grad_norm": 261.85581634588704, + "learning_rate": 3.15e-06, + "loss": 14.5648, + "step": 21 + }, + { + "epoch": 0.010352941176470589, + "grad_norm": 119.85496152582756, + "learning_rate": 3.2999999999999993e-06, + "loss": 14.0929, + "step": 22 + }, + { + "epoch": 0.010823529411764706, + "grad_norm": 715.3326929763317, + "learning_rate": 3.4499999999999996e-06, + "loss": 12.686, + "step": 23 + }, + { + "epoch": 0.011294117647058824, + "grad_norm": 115.80378529539759, + "learning_rate": 3.6e-06, + "loss": 12.6727, + "step": 24 + }, + { + "epoch": 0.011764705882352941, + "grad_norm": 63.19461638956368, + "learning_rate": 3.7499999999999997e-06, + "loss": 12.0643, + "step": 25 + }, + { + "epoch": 0.01223529411764706, + "grad_norm": 57.521661291267904, + "learning_rate": 3.899999999999999e-06, + "loss": 11.5598, + "step": 26 + }, + { + "epoch": 0.012705882352941176, + "grad_norm": 60.25678487559171, + "learning_rate": 4.049999999999999e-06, + "loss": 12.0764, + "step": 27 + }, + { + "epoch": 0.013176470588235295, + "grad_norm": 79.39259804385469, + "learning_rate": 4.2e-06, + "loss": 11.7143, + "step": 28 + }, + { + "epoch": 0.013647058823529411, + "grad_norm": 62.652980375387294, + "learning_rate": 4.35e-06, + "loss": 11.7431, + "step": 29 + }, + { + "epoch": 0.01411764705882353, + "grad_norm": 130.44015196059746, + "learning_rate": 4.499999999999999e-06, + "loss": 12.0069, + "step": 30 + }, + { + "epoch": 0.014588235294117647, + "grad_norm": 81.68679625055253, + "learning_rate": 4.6499999999999995e-06, + "loss": 11.4409, + "step": 31 + }, + { + "epoch": 0.015058823529411765, + "grad_norm": 63.52821775910404, + "learning_rate": 4.8e-06, + "loss": 10.8523, + "step": 32 + }, + { + "epoch": 0.015529411764705882, + "grad_norm": 59.68909024142909, + "learning_rate": 4.95e-06, + "loss": 10.6451, + "step": 33 + }, + { + "epoch": 0.016, + "grad_norm": 58.00476386706095, + "learning_rate": 5.1e-06, + "loss": 10.1353, + "step": 34 + }, + { + "epoch": 0.01647058823529412, + "grad_norm": 52.76230257872452, + "learning_rate": 5.25e-06, + "loss": 10.0673, + "step": 35 + }, + { + "epoch": 0.016941176470588234, + "grad_norm": 50.65899680971922, + "learning_rate": 5.399999999999999e-06, + "loss": 9.7426, + "step": 36 + }, + { + "epoch": 0.017411764705882352, + "grad_norm": 49.26101794026779, + "learning_rate": 5.549999999999999e-06, + "loss": 9.5078, + "step": 37 + }, + { + "epoch": 0.01788235294117647, + "grad_norm": 40.333789615589914, + "learning_rate": 5.7e-06, + "loss": 9.6629, + "step": 38 + }, + { + "epoch": 0.01835294117647059, + "grad_norm": 43.284856816655086, + "learning_rate": 5.85e-06, + "loss": 9.489, + "step": 39 + }, + { + "epoch": 0.018823529411764704, + "grad_norm": 47.26329304691993, + "learning_rate": 5.999999999999999e-06, + "loss": 8.9261, + "step": 40 + }, + { + "epoch": 0.019294117647058823, + "grad_norm": 48.73545902054059, + "learning_rate": 6.1499999999999996e-06, + "loss": 9.0843, + "step": 41 + }, + { + "epoch": 0.01976470588235294, + "grad_norm": 44.547189360711194, + "learning_rate": 6.3e-06, + "loss": 8.8694, + "step": 42 + }, + { + "epoch": 0.02023529411764706, + "grad_norm": 44.184505209286534, + "learning_rate": 6.449999999999999e-06, + "loss": 8.728, + "step": 43 + }, + { + "epoch": 0.020705882352941178, + "grad_norm": 46.26305466865864, + "learning_rate": 6.599999999999999e-06, + "loss": 8.5772, + "step": 44 + }, + { + "epoch": 0.021176470588235293, + "grad_norm": 54.736231991988795, + "learning_rate": 6.749999999999999e-06, + "loss": 8.3753, + "step": 45 + }, + { + "epoch": 0.02164705882352941, + "grad_norm": 44.469044615500145, + "learning_rate": 6.899999999999999e-06, + "loss": 8.3847, + "step": 46 + }, + { + "epoch": 0.02211764705882353, + "grad_norm": 44.42353475096991, + "learning_rate": 7.049999999999999e-06, + "loss": 8.4258, + "step": 47 + }, + { + "epoch": 0.02258823529411765, + "grad_norm": 44.87232528366972, + "learning_rate": 7.2e-06, + "loss": 7.8546, + "step": 48 + }, + { + "epoch": 0.023058823529411764, + "grad_norm": 42.18004170199655, + "learning_rate": 7.35e-06, + "loss": 7.9067, + "step": 49 + }, + { + "epoch": 0.023529411764705882, + "grad_norm": 44.94705325020646, + "learning_rate": 7.499999999999999e-06, + "loss": 7.6046, + "step": 50 + }, + { + "epoch": 0.024, + "grad_norm": 44.70078924922282, + "learning_rate": 7.65e-06, + "loss": 7.4338, + "step": 51 + }, + { + "epoch": 0.02447058823529412, + "grad_norm": 44.330196545962664, + "learning_rate": 7.799999999999998e-06, + "loss": 7.3343, + "step": 52 + }, + { + "epoch": 0.024941176470588234, + "grad_norm": 47.42977145081085, + "learning_rate": 7.949999999999998e-06, + "loss": 7.3829, + "step": 53 + }, + { + "epoch": 0.025411764705882352, + "grad_norm": 52.835043461823574, + "learning_rate": 8.099999999999999e-06, + "loss": 7.1025, + "step": 54 + }, + { + "epoch": 0.02588235294117647, + "grad_norm": 50.60685853650408, + "learning_rate": 8.249999999999999e-06, + "loss": 6.7425, + "step": 55 + }, + { + "epoch": 0.02635294117647059, + "grad_norm": 50.34656818203375, + "learning_rate": 8.4e-06, + "loss": 6.5626, + "step": 56 + }, + { + "epoch": 0.026823529411764704, + "grad_norm": 46.559066218413555, + "learning_rate": 8.55e-06, + "loss": 6.9098, + "step": 57 + }, + { + "epoch": 0.027294117647058823, + "grad_norm": 55.14308310606401, + "learning_rate": 8.7e-06, + "loss": 6.0873, + "step": 58 + }, + { + "epoch": 0.02776470588235294, + "grad_norm": 55.28932546715435, + "learning_rate": 8.849999999999998e-06, + "loss": 5.9444, + "step": 59 + }, + { + "epoch": 0.02823529411764706, + "grad_norm": 55.211319168688725, + "learning_rate": 8.999999999999999e-06, + "loss": 5.8704, + "step": 60 + }, + { + "epoch": 0.028705882352941175, + "grad_norm": 55.87875132854646, + "learning_rate": 9.149999999999999e-06, + "loss": 5.6672, + "step": 61 + }, + { + "epoch": 0.029176470588235293, + "grad_norm": 60.99838382431281, + "learning_rate": 9.299999999999999e-06, + "loss": 5.4104, + "step": 62 + }, + { + "epoch": 0.029647058823529412, + "grad_norm": 59.21478378578891, + "learning_rate": 9.45e-06, + "loss": 5.3763, + "step": 63 + }, + { + "epoch": 0.03011764705882353, + "grad_norm": 55.20725004831107, + "learning_rate": 9.6e-06, + "loss": 5.7369, + "step": 64 + }, + { + "epoch": 0.03058823529411765, + "grad_norm": 61.427827298613174, + "learning_rate": 9.75e-06, + "loss": 4.9828, + "step": 65 + }, + { + "epoch": 0.031058823529411764, + "grad_norm": 61.19790946274717, + "learning_rate": 9.9e-06, + "loss": 4.9618, + "step": 66 + }, + { + "epoch": 0.03152941176470588, + "grad_norm": 62.039601061665984, + "learning_rate": 1.005e-05, + "loss": 4.9071, + "step": 67 + }, + { + "epoch": 0.032, + "grad_norm": 60.73112362386961, + "learning_rate": 1.02e-05, + "loss": 4.9608, + "step": 68 + }, + { + "epoch": 0.03247058823529412, + "grad_norm": 58.09116828315848, + "learning_rate": 1.035e-05, + "loss": 5.1502, + "step": 69 + }, + { + "epoch": 0.03294117647058824, + "grad_norm": 61.34085992882963, + "learning_rate": 1.05e-05, + "loss": 4.884, + "step": 70 + }, + { + "epoch": 0.033411764705882356, + "grad_norm": 59.59504119076543, + "learning_rate": 1.0649999999999998e-05, + "loss": 4.636, + "step": 71 + }, + { + "epoch": 0.03388235294117647, + "grad_norm": 57.57308784882959, + "learning_rate": 1.0799999999999998e-05, + "loss": 4.8318, + "step": 72 + }, + { + "epoch": 0.034352941176470586, + "grad_norm": 65.85540692841242, + "learning_rate": 1.0949999999999998e-05, + "loss": 4.1696, + "step": 73 + }, + { + "epoch": 0.034823529411764705, + "grad_norm": 60.70745950991346, + "learning_rate": 1.1099999999999999e-05, + "loss": 4.394, + "step": 74 + }, + { + "epoch": 0.03529411764705882, + "grad_norm": 61.9611673695319, + "learning_rate": 1.1249999999999999e-05, + "loss": 4.1678, + "step": 75 + }, + { + "epoch": 0.03576470588235294, + "grad_norm": 63.58027376297199, + "learning_rate": 1.14e-05, + "loss": 4.0304, + "step": 76 + }, + { + "epoch": 0.03623529411764706, + "grad_norm": 62.46592015022313, + "learning_rate": 1.155e-05, + "loss": 4.0062, + "step": 77 + }, + { + "epoch": 0.03670588235294118, + "grad_norm": 58.148940973847814, + "learning_rate": 1.17e-05, + "loss": 4.1689, + "step": 78 + }, + { + "epoch": 0.0371764705882353, + "grad_norm": 68.5139002019894, + "learning_rate": 1.1849999999999998e-05, + "loss": 3.846, + "step": 79 + }, + { + "epoch": 0.03764705882352941, + "grad_norm": 62.10026116520012, + "learning_rate": 1.1999999999999999e-05, + "loss": 4.0779, + "step": 80 + }, + { + "epoch": 0.03811764705882353, + "grad_norm": 66.06418764677115, + "learning_rate": 1.2149999999999999e-05, + "loss": 3.695, + "step": 81 + }, + { + "epoch": 0.038588235294117645, + "grad_norm": 61.199342478447555, + "learning_rate": 1.2299999999999999e-05, + "loss": 3.666, + "step": 82 + }, + { + "epoch": 0.039058823529411764, + "grad_norm": 66.02301953706473, + "learning_rate": 1.245e-05, + "loss": 3.5172, + "step": 83 + }, + { + "epoch": 0.03952941176470588, + "grad_norm": 65.23306856168236, + "learning_rate": 1.26e-05, + "loss": 3.4272, + "step": 84 + }, + { + "epoch": 0.04, + "grad_norm": 66.20773476362366, + "learning_rate": 1.275e-05, + "loss": 3.3371, + "step": 85 + }, + { + "epoch": 0.04047058823529412, + "grad_norm": 57.91583694430885, + "learning_rate": 1.2899999999999998e-05, + "loss": 3.7854, + "step": 86 + }, + { + "epoch": 0.04094117647058824, + "grad_norm": 64.41843424928416, + "learning_rate": 1.3049999999999999e-05, + "loss": 3.3019, + "step": 87 + }, + { + "epoch": 0.041411764705882356, + "grad_norm": 61.88982829305557, + "learning_rate": 1.3199999999999997e-05, + "loss": 3.2908, + "step": 88 + }, + { + "epoch": 0.04188235294117647, + "grad_norm": 66.25284455446722, + "learning_rate": 1.3349999999999998e-05, + "loss": 3.0361, + "step": 89 + }, + { + "epoch": 0.042352941176470586, + "grad_norm": 66.48604414153876, + "learning_rate": 1.3499999999999998e-05, + "loss": 2.9511, + "step": 90 + }, + { + "epoch": 0.042823529411764705, + "grad_norm": 66.33815193795937, + "learning_rate": 1.3649999999999998e-05, + "loss": 2.88, + "step": 91 + }, + { + "epoch": 0.04329411764705882, + "grad_norm": 60.83388487794292, + "learning_rate": 1.3799999999999998e-05, + "loss": 3.0893, + "step": 92 + }, + { + "epoch": 0.04376470588235294, + "grad_norm": 62.36567377527326, + "learning_rate": 1.3949999999999999e-05, + "loss": 2.8095, + "step": 93 + }, + { + "epoch": 0.04423529411764706, + "grad_norm": 65.50124306725256, + "learning_rate": 1.4099999999999999e-05, + "loss": 2.6265, + "step": 94 + }, + { + "epoch": 0.04470588235294118, + "grad_norm": 59.7922765792744, + "learning_rate": 1.4249999999999999e-05, + "loss": 2.7419, + "step": 95 + }, + { + "epoch": 0.0451764705882353, + "grad_norm": 65.49391394129486, + "learning_rate": 1.44e-05, + "loss": 2.4942, + "step": 96 + }, + { + "epoch": 0.04564705882352941, + "grad_norm": 58.398456539224895, + "learning_rate": 1.455e-05, + "loss": 2.8396, + "step": 97 + }, + { + "epoch": 0.04611764705882353, + "grad_norm": 60.67404529929337, + "learning_rate": 1.47e-05, + "loss": 2.6322, + "step": 98 + }, + { + "epoch": 0.046588235294117646, + "grad_norm": 61.81189261738568, + "learning_rate": 1.485e-05, + "loss": 2.2785, + "step": 99 + }, + { + "epoch": 0.047058823529411764, + "grad_norm": 56.32962056761837, + "learning_rate": 1.4999999999999999e-05, + "loss": 2.4319, + "step": 100 + }, + { + "epoch": 0.04752941176470588, + "grad_norm": 62.71105305512622, + "learning_rate": 1.5149999999999999e-05, + "loss": 2.09, + "step": 101 + }, + { + "epoch": 0.048, + "grad_norm": 58.344264799198086, + "learning_rate": 1.53e-05, + "loss": 2.1437, + "step": 102 + }, + { + "epoch": 0.04847058823529412, + "grad_norm": 58.30346393689416, + "learning_rate": 1.545e-05, + "loss": 1.9839, + "step": 103 + }, + { + "epoch": 0.04894117647058824, + "grad_norm": 59.78774187681486, + "learning_rate": 1.5599999999999996e-05, + "loss": 1.8058, + "step": 104 + }, + { + "epoch": 0.04941176470588235, + "grad_norm": 56.724281977487486, + "learning_rate": 1.5749999999999997e-05, + "loss": 1.8518, + "step": 105 + }, + { + "epoch": 0.04988235294117647, + "grad_norm": 57.91914236283996, + "learning_rate": 1.5899999999999997e-05, + "loss": 1.6528, + "step": 106 + }, + { + "epoch": 0.050352941176470586, + "grad_norm": 55.44786760110433, + "learning_rate": 1.6049999999999997e-05, + "loss": 1.9989, + "step": 107 + }, + { + "epoch": 0.050823529411764705, + "grad_norm": 51.89307767056903, + "learning_rate": 1.6199999999999997e-05, + "loss": 1.7438, + "step": 108 + }, + { + "epoch": 0.05129411764705882, + "grad_norm": 46.872967509197366, + "learning_rate": 1.6349999999999998e-05, + "loss": 1.892, + "step": 109 + }, + { + "epoch": 0.05176470588235294, + "grad_norm": 52.55599509353864, + "learning_rate": 1.6499999999999998e-05, + "loss": 1.673, + "step": 110 + }, + { + "epoch": 0.05223529411764706, + "grad_norm": 57.260407438653154, + "learning_rate": 1.6649999999999998e-05, + "loss": 1.5539, + "step": 111 + }, + { + "epoch": 0.05270588235294118, + "grad_norm": 51.51411449848559, + "learning_rate": 1.68e-05, + "loss": 1.4724, + "step": 112 + }, + { + "epoch": 0.0531764705882353, + "grad_norm": 50.225153537569845, + "learning_rate": 1.695e-05, + "loss": 1.1789, + "step": 113 + }, + { + "epoch": 0.05364705882352941, + "grad_norm": 45.263282167237115, + "learning_rate": 1.71e-05, + "loss": 1.3165, + "step": 114 + }, + { + "epoch": 0.05411764705882353, + "grad_norm": 46.708599410139584, + "learning_rate": 1.725e-05, + "loss": 1.0393, + "step": 115 + }, + { + "epoch": 0.054588235294117646, + "grad_norm": 45.015570184206474, + "learning_rate": 1.74e-05, + "loss": 0.9837, + "step": 116 + }, + { + "epoch": 0.055058823529411764, + "grad_norm": 40.206424280405706, + "learning_rate": 1.755e-05, + "loss": 1.2798, + "step": 117 + }, + { + "epoch": 0.05552941176470588, + "grad_norm": 37.008269110969174, + "learning_rate": 1.7699999999999997e-05, + "loss": 1.399, + "step": 118 + }, + { + "epoch": 0.056, + "grad_norm": 38.721862893591116, + "learning_rate": 1.7849999999999997e-05, + "loss": 1.2214, + "step": 119 + }, + { + "epoch": 0.05647058823529412, + "grad_norm": 37.73932656133891, + "learning_rate": 1.7999999999999997e-05, + "loss": 1.0503, + "step": 120 + }, + { + "epoch": 0.05694117647058824, + "grad_norm": 30.23924055372066, + "learning_rate": 1.8149999999999997e-05, + "loss": 1.244, + "step": 121 + }, + { + "epoch": 0.05741176470588235, + "grad_norm": 31.23169741940704, + "learning_rate": 1.8299999999999998e-05, + "loss": 0.9657, + "step": 122 + }, + { + "epoch": 0.05788235294117647, + "grad_norm": 37.40897539537868, + "learning_rate": 1.8449999999999998e-05, + "loss": 0.9664, + "step": 123 + }, + { + "epoch": 0.05835294117647059, + "grad_norm": 36.44527883782951, + "learning_rate": 1.8599999999999998e-05, + "loss": 0.8221, + "step": 124 + }, + { + "epoch": 0.058823529411764705, + "grad_norm": 33.3642213228657, + "learning_rate": 1.875e-05, + "loss": 0.7614, + "step": 125 + }, + { + "epoch": 0.059294117647058824, + "grad_norm": 31.06600439519094, + "learning_rate": 1.89e-05, + "loss": 0.5654, + "step": 126 + }, + { + "epoch": 0.05976470588235294, + "grad_norm": 28.641360769720848, + "learning_rate": 1.905e-05, + "loss": 0.6369, + "step": 127 + }, + { + "epoch": 0.06023529411764706, + "grad_norm": 24.60682004342332, + "learning_rate": 1.92e-05, + "loss": 0.7168, + "step": 128 + }, + { + "epoch": 0.06070588235294118, + "grad_norm": 19.86459118764046, + "learning_rate": 1.935e-05, + "loss": 0.8687, + "step": 129 + }, + { + "epoch": 0.0611764705882353, + "grad_norm": 19.876778623299383, + "learning_rate": 1.95e-05, + "loss": 0.823, + "step": 130 + }, + { + "epoch": 0.06164705882352941, + "grad_norm": 17.520383410952498, + "learning_rate": 1.965e-05, + "loss": 0.6439, + "step": 131 + }, + { + "epoch": 0.06211764705882353, + "grad_norm": 16.896873761421947, + "learning_rate": 1.98e-05, + "loss": 0.6216, + "step": 132 + }, + { + "epoch": 0.06258823529411765, + "grad_norm": 13.901196258415084, + "learning_rate": 1.995e-05, + "loss": 0.7216, + "step": 133 + }, + { + "epoch": 0.06305882352941176, + "grad_norm": 20.496578267650914, + "learning_rate": 2.01e-05, + "loss": 0.5105, + "step": 134 + }, + { + "epoch": 0.06352941176470588, + "grad_norm": 21.164676965222046, + "learning_rate": 2.025e-05, + "loss": 0.5774, + "step": 135 + }, + { + "epoch": 0.064, + "grad_norm": 21.2372879581247, + "learning_rate": 2.04e-05, + "loss": 0.5984, + "step": 136 + }, + { + "epoch": 0.06447058823529411, + "grad_norm": 17.963498229253382, + "learning_rate": 2.055e-05, + "loss": 0.4103, + "step": 137 + }, + { + "epoch": 0.06494117647058824, + "grad_norm": 16.645527513394924, + "learning_rate": 2.07e-05, + "loss": 0.3664, + "step": 138 + }, + { + "epoch": 0.06541176470588235, + "grad_norm": 12.709384042990301, + "learning_rate": 2.085e-05, + "loss": 0.3668, + "step": 139 + }, + { + "epoch": 0.06588235294117648, + "grad_norm": 14.460235480033123, + "learning_rate": 2.1e-05, + "loss": 0.431, + "step": 140 + }, + { + "epoch": 0.06635294117647059, + "grad_norm": 11.247092343190806, + "learning_rate": 2.1149999999999996e-05, + "loss": 0.578, + "step": 141 + }, + { + "epoch": 0.06682352941176471, + "grad_norm": 11.755051826290272, + "learning_rate": 2.1299999999999996e-05, + "loss": 0.1981, + "step": 142 + }, + { + "epoch": 0.06729411764705882, + "grad_norm": 10.357593147760547, + "learning_rate": 2.1449999999999996e-05, + "loss": 0.2765, + "step": 143 + }, + { + "epoch": 0.06776470588235294, + "grad_norm": 10.085121671099964, + "learning_rate": 2.1599999999999996e-05, + "loss": 0.2263, + "step": 144 + }, + { + "epoch": 0.06823529411764706, + "grad_norm": 9.7426852081744, + "learning_rate": 2.1749999999999997e-05, + "loss": 0.157, + "step": 145 + }, + { + "epoch": 0.06870588235294117, + "grad_norm": 10.28272437045224, + "learning_rate": 2.1899999999999997e-05, + "loss": 0.2108, + "step": 146 + }, + { + "epoch": 0.0691764705882353, + "grad_norm": 8.586632911473531, + "learning_rate": 2.2049999999999997e-05, + "loss": 0.1815, + "step": 147 + }, + { + "epoch": 0.06964705882352941, + "grad_norm": 8.2949314756054, + "learning_rate": 2.2199999999999998e-05, + "loss": 0.1537, + "step": 148 + }, + { + "epoch": 0.07011764705882353, + "grad_norm": 7.93566599592368, + "learning_rate": 2.2349999999999998e-05, + "loss": 0.1204, + "step": 149 + }, + { + "epoch": 0.07058823529411765, + "grad_norm": 7.398456247249101, + "learning_rate": 2.2499999999999998e-05, + "loss": 0.2414, + "step": 150 + }, + { + "epoch": 0.07105882352941177, + "grad_norm": 7.1773807218111125, + "learning_rate": 2.2649999999999998e-05, + "loss": 0.1419, + "step": 151 + }, + { + "epoch": 0.07152941176470588, + "grad_norm": 6.93398897800265, + "learning_rate": 2.28e-05, + "loss": 0.104, + "step": 152 + }, + { + "epoch": 0.072, + "grad_norm": 6.097291813836074, + "learning_rate": 2.295e-05, + "loss": 0.1355, + "step": 153 + }, + { + "epoch": 0.07247058823529412, + "grad_norm": 6.091534380742603, + "learning_rate": 2.31e-05, + "loss": 0.0896, + "step": 154 + }, + { + "epoch": 0.07294117647058823, + "grad_norm": 5.981848835193809, + "learning_rate": 2.325e-05, + "loss": 0.1871, + "step": 155 + }, + { + "epoch": 0.07341176470588236, + "grad_norm": 5.468481835297579, + "learning_rate": 2.34e-05, + "loss": 0.1116, + "step": 156 + }, + { + "epoch": 0.07388235294117647, + "grad_norm": 5.051428567728681, + "learning_rate": 2.3549999999999996e-05, + "loss": 0.102, + "step": 157 + }, + { + "epoch": 0.0743529411764706, + "grad_norm": 4.815974946484687, + "learning_rate": 2.3699999999999997e-05, + "loss": 0.0913, + "step": 158 + }, + { + "epoch": 0.0748235294117647, + "grad_norm": 4.65801997678475, + "learning_rate": 2.3849999999999997e-05, + "loss": 0.274, + "step": 159 + }, + { + "epoch": 0.07529411764705882, + "grad_norm": 4.697882552992091, + "learning_rate": 2.3999999999999997e-05, + "loss": 0.0904, + "step": 160 + }, + { + "epoch": 0.07576470588235294, + "grad_norm": 4.5785708828374165, + "learning_rate": 2.4149999999999997e-05, + "loss": 0.1554, + "step": 161 + }, + { + "epoch": 0.07623529411764705, + "grad_norm": 3.818163309100854, + "learning_rate": 2.4299999999999998e-05, + "loss": 0.107, + "step": 162 + }, + { + "epoch": 0.07670588235294118, + "grad_norm": 4.016587394533516, + "learning_rate": 2.4449999999999998e-05, + "loss": 0.1513, + "step": 163 + }, + { + "epoch": 0.07717647058823529, + "grad_norm": 3.7658930246027817, + "learning_rate": 2.4599999999999998e-05, + "loss": 0.103, + "step": 164 + }, + { + "epoch": 0.07764705882352942, + "grad_norm": 3.3975655774155262, + "learning_rate": 2.475e-05, + "loss": 0.0637, + "step": 165 + }, + { + "epoch": 0.07811764705882353, + "grad_norm": 4.719930033017316, + "learning_rate": 2.49e-05, + "loss": 0.201, + "step": 166 + }, + { + "epoch": 0.07858823529411765, + "grad_norm": 3.587563471168977, + "learning_rate": 2.505e-05, + "loss": 0.1407, + "step": 167 + }, + { + "epoch": 0.07905882352941176, + "grad_norm": 3.562248128779394, + "learning_rate": 2.52e-05, + "loss": 0.1592, + "step": 168 + }, + { + "epoch": 0.07952941176470588, + "grad_norm": 2.984740092460642, + "learning_rate": 2.535e-05, + "loss": 0.1111, + "step": 169 + }, + { + "epoch": 0.08, + "grad_norm": 3.0260115607839873, + "learning_rate": 2.55e-05, + "loss": 0.0884, + "step": 170 + }, + { + "epoch": 0.08047058823529411, + "grad_norm": 3.697808281252457, + "learning_rate": 2.565e-05, + "loss": 0.0694, + "step": 171 + }, + { + "epoch": 0.08094117647058824, + "grad_norm": 4.393940049335206, + "learning_rate": 2.5799999999999997e-05, + "loss": 0.2528, + "step": 172 + }, + { + "epoch": 0.08141176470588235, + "grad_norm": 3.340822069367983, + "learning_rate": 2.5949999999999997e-05, + "loss": 0.1263, + "step": 173 + }, + { + "epoch": 0.08188235294117648, + "grad_norm": 2.357845459200362, + "learning_rate": 2.6099999999999997e-05, + "loss": 0.0338, + "step": 174 + }, + { + "epoch": 0.08235294117647059, + "grad_norm": 2.8864268200269674, + "learning_rate": 2.6249999999999998e-05, + "loss": 0.0526, + "step": 175 + }, + { + "epoch": 0.08282352941176471, + "grad_norm": 3.023016014523894, + "learning_rate": 2.6399999999999995e-05, + "loss": 0.0935, + "step": 176 + }, + { + "epoch": 0.08329411764705882, + "grad_norm": 2.0985028030143673, + "learning_rate": 2.6549999999999995e-05, + "loss": 0.0299, + "step": 177 + }, + { + "epoch": 0.08376470588235294, + "grad_norm": 2.8804792737703204, + "learning_rate": 2.6699999999999995e-05, + "loss": 0.0681, + "step": 178 + }, + { + "epoch": 0.08423529411764706, + "grad_norm": 2.3652183126240116, + "learning_rate": 2.6849999999999995e-05, + "loss": 0.0397, + "step": 179 + }, + { + "epoch": 0.08470588235294117, + "grad_norm": 3.0600538913261355, + "learning_rate": 2.6999999999999996e-05, + "loss": 0.1659, + "step": 180 + }, + { + "epoch": 0.0851764705882353, + "grad_norm": 3.7350977316184095, + "learning_rate": 2.7149999999999996e-05, + "loss": 0.0665, + "step": 181 + }, + { + "epoch": 0.08564705882352941, + "grad_norm": 5.741566793401181, + "learning_rate": 2.7299999999999996e-05, + "loss": 0.1588, + "step": 182 + }, + { + "epoch": 0.08611764705882353, + "grad_norm": 2.3136773490294447, + "learning_rate": 2.7449999999999996e-05, + "loss": 0.0873, + "step": 183 + }, + { + "epoch": 0.08658823529411765, + "grad_norm": 5.214842214221372, + "learning_rate": 2.7599999999999997e-05, + "loss": 0.1285, + "step": 184 + }, + { + "epoch": 0.08705882352941176, + "grad_norm": 1.6166191419894762, + "learning_rate": 2.7749999999999997e-05, + "loss": 0.0226, + "step": 185 + }, + { + "epoch": 0.08752941176470588, + "grad_norm": 2.0173140025523857, + "learning_rate": 2.7899999999999997e-05, + "loss": 0.0552, + "step": 186 + }, + { + "epoch": 0.088, + "grad_norm": 1.5313886444344893, + "learning_rate": 2.8049999999999997e-05, + "loss": 0.0229, + "step": 187 + }, + { + "epoch": 0.08847058823529412, + "grad_norm": 2.628484317287644, + "learning_rate": 2.8199999999999998e-05, + "loss": 0.1026, + "step": 188 + }, + { + "epoch": 0.08894117647058823, + "grad_norm": 4.669777934123752, + "learning_rate": 2.8349999999999998e-05, + "loss": 0.1452, + "step": 189 + }, + { + "epoch": 0.08941176470588236, + "grad_norm": 2.13571578662631, + "learning_rate": 2.8499999999999998e-05, + "loss": 0.0581, + "step": 190 + }, + { + "epoch": 0.08988235294117647, + "grad_norm": 6.403195639114379, + "learning_rate": 2.865e-05, + "loss": 0.088, + "step": 191 + }, + { + "epoch": 0.0903529411764706, + "grad_norm": 1.8321349679551928, + "learning_rate": 2.88e-05, + "loss": 0.0581, + "step": 192 + }, + { + "epoch": 0.0908235294117647, + "grad_norm": 1.4757597519556962, + "learning_rate": 2.895e-05, + "loss": 0.0207, + "step": 193 + }, + { + "epoch": 0.09129411764705882, + "grad_norm": 1.7947493051670043, + "learning_rate": 2.91e-05, + "loss": 0.028, + "step": 194 + }, + { + "epoch": 0.09176470588235294, + "grad_norm": 4.297032987331882, + "learning_rate": 2.925e-05, + "loss": 0.1542, + "step": 195 + }, + { + "epoch": 0.09223529411764705, + "grad_norm": 2.852813974975233, + "learning_rate": 2.94e-05, + "loss": 0.1353, + "step": 196 + }, + { + "epoch": 0.09270588235294118, + "grad_norm": 1.8365575075201666, + "learning_rate": 2.955e-05, + "loss": 0.0246, + "step": 197 + }, + { + "epoch": 0.09317647058823529, + "grad_norm": 1.75593818655513, + "learning_rate": 2.97e-05, + "loss": 0.0642, + "step": 198 + }, + { + "epoch": 0.09364705882352942, + "grad_norm": 3.0297999451839153, + "learning_rate": 2.985e-05, + "loss": 0.1159, + "step": 199 + }, + { + "epoch": 0.09411764705882353, + "grad_norm": 1.9891651730363866, + "learning_rate": 2.9999999999999997e-05, + "loss": 0.0701, + "step": 200 + }, + { + "epoch": 0.09458823529411765, + "grad_norm": 1.7707772639685808, + "learning_rate": 3.0149999999999998e-05, + "loss": 0.0746, + "step": 201 + }, + { + "epoch": 0.09505882352941177, + "grad_norm": 1.6133150209265876, + "learning_rate": 3.0299999999999998e-05, + "loss": 0.0616, + "step": 202 + }, + { + "epoch": 0.09552941176470588, + "grad_norm": 1.9482241456185516, + "learning_rate": 3.0449999999999998e-05, + "loss": 0.0861, + "step": 203 + }, + { + "epoch": 0.096, + "grad_norm": 1.068905419792845, + "learning_rate": 3.06e-05, + "loss": 0.0147, + "step": 204 + }, + { + "epoch": 0.09647058823529411, + "grad_norm": 2.0087684172066664, + "learning_rate": 3.0749999999999995e-05, + "loss": 0.1638, + "step": 205 + }, + { + "epoch": 0.09694117647058824, + "grad_norm": 1.1398522660576191, + "learning_rate": 3.09e-05, + "loss": 0.0285, + "step": 206 + }, + { + "epoch": 0.09741176470588235, + "grad_norm": 1.4627423476174977, + "learning_rate": 3.1049999999999996e-05, + "loss": 0.1175, + "step": 207 + }, + { + "epoch": 0.09788235294117648, + "grad_norm": 1.3071889716009422, + "learning_rate": 3.119999999999999e-05, + "loss": 0.0484, + "step": 208 + }, + { + "epoch": 0.09835294117647059, + "grad_norm": 1.0414733857102902, + "learning_rate": 3.1349999999999996e-05, + "loss": 0.0313, + "step": 209 + }, + { + "epoch": 0.0988235294117647, + "grad_norm": 1.2670151468281095, + "learning_rate": 3.149999999999999e-05, + "loss": 0.0455, + "step": 210 + }, + { + "epoch": 0.09929411764705882, + "grad_norm": 1.301637962206306, + "learning_rate": 3.165e-05, + "loss": 0.0617, + "step": 211 + }, + { + "epoch": 0.09976470588235294, + "grad_norm": 0.9399825359853629, + "learning_rate": 3.1799999999999994e-05, + "loss": 0.0285, + "step": 212 + }, + { + "epoch": 0.10023529411764706, + "grad_norm": 1.02334877111067, + "learning_rate": 3.195e-05, + "loss": 0.0343, + "step": 213 + }, + { + "epoch": 0.10070588235294117, + "grad_norm": 0.8150627792320192, + "learning_rate": 3.2099999999999994e-05, + "loss": 0.0115, + "step": 214 + }, + { + "epoch": 0.1011764705882353, + "grad_norm": 1.098347535031592, + "learning_rate": 3.225e-05, + "loss": 0.0659, + "step": 215 + }, + { + "epoch": 0.10164705882352941, + "grad_norm": 0.7902892291724234, + "learning_rate": 3.2399999999999995e-05, + "loss": 0.028, + "step": 216 + }, + { + "epoch": 0.10211764705882354, + "grad_norm": 1.4303709110710998, + "learning_rate": 3.255e-05, + "loss": 0.0948, + "step": 217 + }, + { + "epoch": 0.10258823529411765, + "grad_norm": 0.9911606054219618, + "learning_rate": 3.2699999999999995e-05, + "loss": 0.053, + "step": 218 + }, + { + "epoch": 0.10305882352941176, + "grad_norm": 0.7908275846451995, + "learning_rate": 3.285e-05, + "loss": 0.0155, + "step": 219 + }, + { + "epoch": 0.10352941176470588, + "grad_norm": 0.5894938877121815, + "learning_rate": 3.2999999999999996e-05, + "loss": 0.0082, + "step": 220 + }, + { + "epoch": 0.104, + "grad_norm": 1.4440705908073121, + "learning_rate": 3.315e-05, + "loss": 0.1023, + "step": 221 + }, + { + "epoch": 0.10447058823529412, + "grad_norm": 0.8909877534612225, + "learning_rate": 3.3299999999999996e-05, + "loss": 0.0542, + "step": 222 + }, + { + "epoch": 0.10494117647058823, + "grad_norm": 1.2996573610830293, + "learning_rate": 3.345e-05, + "loss": 0.1153, + "step": 223 + }, + { + "epoch": 0.10541176470588236, + "grad_norm": 0.7039812463816616, + "learning_rate": 3.36e-05, + "loss": 0.032, + "step": 224 + }, + { + "epoch": 0.10588235294117647, + "grad_norm": 0.727316377929752, + "learning_rate": 3.375e-05, + "loss": 0.1194, + "step": 225 + }, + { + "epoch": 0.1063529411764706, + "grad_norm": 0.6858543927216185, + "learning_rate": 3.39e-05, + "loss": 0.0302, + "step": 226 + }, + { + "epoch": 0.1068235294117647, + "grad_norm": 1.0689296721427441, + "learning_rate": 3.405e-05, + "loss": 0.1215, + "step": 227 + }, + { + "epoch": 0.10729411764705882, + "grad_norm": 0.8671450174680458, + "learning_rate": 3.42e-05, + "loss": 0.0962, + "step": 228 + }, + { + "epoch": 0.10776470588235294, + "grad_norm": 0.6218373736321802, + "learning_rate": 3.435e-05, + "loss": 0.0391, + "step": 229 + }, + { + "epoch": 0.10823529411764705, + "grad_norm": 0.47842886962094805, + "learning_rate": 3.45e-05, + "loss": 0.0068, + "step": 230 + }, + { + "epoch": 0.10870588235294118, + "grad_norm": 1.858350693155272, + "learning_rate": 3.465e-05, + "loss": 0.201, + "step": 231 + }, + { + "epoch": 0.10917647058823529, + "grad_norm": 0.4288705314781912, + "learning_rate": 3.48e-05, + "loss": 0.0059, + "step": 232 + }, + { + "epoch": 0.10964705882352942, + "grad_norm": 0.7402879440684037, + "learning_rate": 3.4949999999999996e-05, + "loss": 0.1076, + "step": 233 + }, + { + "epoch": 0.11011764705882353, + "grad_norm": 0.524138421079942, + "learning_rate": 3.51e-05, + "loss": 0.0358, + "step": 234 + }, + { + "epoch": 0.11058823529411765, + "grad_norm": 1.1037577331944475, + "learning_rate": 3.5249999999999996e-05, + "loss": 0.0882, + "step": 235 + }, + { + "epoch": 0.11105882352941177, + "grad_norm": 0.4299159911409175, + "learning_rate": 3.539999999999999e-05, + "loss": 0.017, + "step": 236 + }, + { + "epoch": 0.11152941176470588, + "grad_norm": 0.5277921387996534, + "learning_rate": 3.555e-05, + "loss": 0.0552, + "step": 237 + }, + { + "epoch": 0.112, + "grad_norm": 0.6234558392791811, + "learning_rate": 3.5699999999999994e-05, + "loss": 0.0327, + "step": 238 + }, + { + "epoch": 0.11247058823529411, + "grad_norm": 0.7102846375994675, + "learning_rate": 3.585e-05, + "loss": 0.0522, + "step": 239 + }, + { + "epoch": 0.11294117647058824, + "grad_norm": 1.5379114315625948, + "learning_rate": 3.5999999999999994e-05, + "loss": 0.1717, + "step": 240 + }, + { + "epoch": 0.11341176470588235, + "grad_norm": 0.334550078499663, + "learning_rate": 3.615e-05, + "loss": 0.0062, + "step": 241 + }, + { + "epoch": 0.11388235294117648, + "grad_norm": 0.3038837691000333, + "learning_rate": 3.6299999999999995e-05, + "loss": 0.0042, + "step": 242 + }, + { + "epoch": 0.11435294117647059, + "grad_norm": 0.29669825851025317, + "learning_rate": 3.645e-05, + "loss": 0.0041, + "step": 243 + }, + { + "epoch": 0.1148235294117647, + "grad_norm": 0.6990890824215724, + "learning_rate": 3.6599999999999995e-05, + "loss": 0.0554, + "step": 244 + }, + { + "epoch": 0.11529411764705882, + "grad_norm": 1.6001577160675546, + "learning_rate": 3.675e-05, + "loss": 0.1721, + "step": 245 + }, + { + "epoch": 0.11576470588235294, + "grad_norm": 1.215405401663401, + "learning_rate": 3.6899999999999996e-05, + "loss": 0.1328, + "step": 246 + }, + { + "epoch": 0.11623529411764706, + "grad_norm": 0.6292815461610967, + "learning_rate": 3.705e-05, + "loss": 0.0381, + "step": 247 + }, + { + "epoch": 0.11670588235294117, + "grad_norm": 0.8995884941567189, + "learning_rate": 3.7199999999999996e-05, + "loss": 0.0941, + "step": 248 + }, + { + "epoch": 0.1171764705882353, + "grad_norm": 0.34555005991567467, + "learning_rate": 3.735e-05, + "loss": 0.0168, + "step": 249 + }, + { + "epoch": 0.11764705882352941, + "grad_norm": 0.9143744287927662, + "learning_rate": 3.75e-05, + "loss": 0.0524, + "step": 250 + }, + { + "epoch": 0.11811764705882354, + "grad_norm": 0.9786392189498541, + "learning_rate": 3.7649999999999994e-05, + "loss": 0.1096, + "step": 251 + }, + { + "epoch": 0.11858823529411765, + "grad_norm": 0.2562429035736632, + "learning_rate": 3.78e-05, + "loss": 0.0036, + "step": 252 + }, + { + "epoch": 0.11905882352941176, + "grad_norm": 0.5987136371775817, + "learning_rate": 3.7949999999999994e-05, + "loss": 0.0427, + "step": 253 + }, + { + "epoch": 0.11952941176470588, + "grad_norm": 0.25109651931233523, + "learning_rate": 3.81e-05, + "loss": 0.0035, + "step": 254 + }, + { + "epoch": 0.12, + "grad_norm": 1.1107218141165618, + "learning_rate": 3.8249999999999995e-05, + "loss": 0.162, + "step": 255 + }, + { + "epoch": 0.12047058823529412, + "grad_norm": 0.29443294503564826, + "learning_rate": 3.84e-05, + "loss": 0.0144, + "step": 256 + }, + { + "epoch": 0.12094117647058823, + "grad_norm": 0.25542131595971934, + "learning_rate": 3.8549999999999995e-05, + "loss": 0.0037, + "step": 257 + }, + { + "epoch": 0.12141176470588236, + "grad_norm": 0.9137506088379062, + "learning_rate": 3.87e-05, + "loss": 0.1332, + "step": 258 + }, + { + "epoch": 0.12188235294117647, + "grad_norm": 13.56950335085045, + "learning_rate": 3.8849999999999996e-05, + "loss": 0.0732, + "step": 259 + }, + { + "epoch": 0.1223529411764706, + "grad_norm": 0.24082813853860466, + "learning_rate": 3.9e-05, + "loss": 0.0034, + "step": 260 + }, + { + "epoch": 0.1228235294117647, + "grad_norm": 5.675980738876112, + "learning_rate": 3.9149999999999996e-05, + "loss": 0.1415, + "step": 261 + }, + { + "epoch": 0.12329411764705882, + "grad_norm": 0.7640339815341638, + "learning_rate": 3.93e-05, + "loss": 0.0897, + "step": 262 + }, + { + "epoch": 0.12376470588235294, + "grad_norm": 0.2793508276913099, + "learning_rate": 3.945e-05, + "loss": 0.0143, + "step": 263 + }, + { + "epoch": 0.12423529411764705, + "grad_norm": 1.631644984220751, + "learning_rate": 3.96e-05, + "loss": 0.097, + "step": 264 + }, + { + "epoch": 0.12470588235294118, + "grad_norm": 0.23359502370250315, + "learning_rate": 3.975e-05, + "loss": 0.0032, + "step": 265 + }, + { + "epoch": 0.1251764705882353, + "grad_norm": 0.479215306870661, + "learning_rate": 3.99e-05, + "loss": 0.0447, + "step": 266 + }, + { + "epoch": 0.12564705882352942, + "grad_norm": 0.3611104364511379, + "learning_rate": 4.005e-05, + "loss": 0.0393, + "step": 267 + }, + { + "epoch": 0.12611764705882353, + "grad_norm": 0.22351930071620021, + "learning_rate": 4.02e-05, + "loss": 0.0031, + "step": 268 + }, + { + "epoch": 0.12658823529411764, + "grad_norm": 0.5191537948157681, + "learning_rate": 4.035e-05, + "loss": 0.078, + "step": 269 + }, + { + "epoch": 0.12705882352941175, + "grad_norm": 0.45702558068326554, + "learning_rate": 4.05e-05, + "loss": 0.1052, + "step": 270 + }, + { + "epoch": 0.1275294117647059, + "grad_norm": 0.44997610700521745, + "learning_rate": 4.065e-05, + "loss": 0.103, + "step": 271 + }, + { + "epoch": 0.128, + "grad_norm": 0.4716155601029948, + "learning_rate": 4.08e-05, + "loss": 0.0804, + "step": 272 + }, + { + "epoch": 0.12847058823529411, + "grad_norm": 0.6295649752177405, + "learning_rate": 4.095e-05, + "loss": 0.2301, + "step": 273 + }, + { + "epoch": 0.12894117647058823, + "grad_norm": 0.36818579815939567, + "learning_rate": 4.11e-05, + "loss": 0.0428, + "step": 274 + }, + { + "epoch": 0.12941176470588237, + "grad_norm": 0.3283478633547613, + "learning_rate": 4.125e-05, + "loss": 0.0404, + "step": 275 + }, + { + "epoch": 0.12988235294117648, + "grad_norm": 0.4833049014276938, + "learning_rate": 4.14e-05, + "loss": 0.0879, + "step": 276 + }, + { + "epoch": 0.1303529411764706, + "grad_norm": 0.45190504082367033, + "learning_rate": 4.155e-05, + "loss": 0.0663, + "step": 277 + }, + { + "epoch": 0.1308235294117647, + "grad_norm": 0.5598828613796948, + "learning_rate": 4.17e-05, + "loss": 0.0747, + "step": 278 + }, + { + "epoch": 0.1312941176470588, + "grad_norm": 0.26589287208200774, + "learning_rate": 4.185e-05, + "loss": 0.0105, + "step": 279 + }, + { + "epoch": 0.13176470588235295, + "grad_norm": 0.5410152877657892, + "learning_rate": 4.2e-05, + "loss": 0.0389, + "step": 280 + }, + { + "epoch": 0.13223529411764706, + "grad_norm": 0.3580084399381269, + "learning_rate": 4.215e-05, + "loss": 0.0274, + "step": 281 + }, + { + "epoch": 0.13270588235294117, + "grad_norm": 0.5366052943015804, + "learning_rate": 4.229999999999999e-05, + "loss": 0.0488, + "step": 282 + }, + { + "epoch": 0.13317647058823529, + "grad_norm": 1.2508527834613499, + "learning_rate": 4.2449999999999995e-05, + "loss": 0.1455, + "step": 283 + }, + { + "epoch": 0.13364705882352942, + "grad_norm": 0.4594872167540642, + "learning_rate": 4.259999999999999e-05, + "loss": 0.0655, + "step": 284 + }, + { + "epoch": 0.13411764705882354, + "grad_norm": 0.62619079748314, + "learning_rate": 4.2749999999999996e-05, + "loss": 0.0659, + "step": 285 + }, + { + "epoch": 0.13458823529411765, + "grad_norm": 0.4904326006311203, + "learning_rate": 4.289999999999999e-05, + "loss": 0.0424, + "step": 286 + }, + { + "epoch": 0.13505882352941176, + "grad_norm": 0.8601091264609083, + "learning_rate": 4.3049999999999996e-05, + "loss": 0.1104, + "step": 287 + }, + { + "epoch": 0.13552941176470587, + "grad_norm": 0.5211063118972272, + "learning_rate": 4.319999999999999e-05, + "loss": 0.0789, + "step": 288 + }, + { + "epoch": 0.136, + "grad_norm": 0.5470233018207051, + "learning_rate": 4.334999999999999e-05, + "loss": 0.007, + "step": 289 + }, + { + "epoch": 0.13647058823529412, + "grad_norm": 0.5223676218249417, + "learning_rate": 4.3499999999999993e-05, + "loss": 0.0068, + "step": 290 + }, + { + "epoch": 0.13694117647058823, + "grad_norm": 3.2075244640625935, + "learning_rate": 4.364999999999999e-05, + "loss": 0.2502, + "step": 291 + }, + { + "epoch": 0.13741176470588234, + "grad_norm": 0.6209088761074408, + "learning_rate": 4.3799999999999994e-05, + "loss": 0.0602, + "step": 292 + }, + { + "epoch": 0.13788235294117648, + "grad_norm": 3.325411094725444, + "learning_rate": 4.394999999999999e-05, + "loss": 0.2293, + "step": 293 + }, + { + "epoch": 0.1383529411764706, + "grad_norm": 1.392537647468027, + "learning_rate": 4.4099999999999995e-05, + "loss": 0.0597, + "step": 294 + }, + { + "epoch": 0.1388235294117647, + "grad_norm": 0.17157739568880837, + "learning_rate": 4.424999999999999e-05, + "loss": 0.0023, + "step": 295 + }, + { + "epoch": 0.13929411764705882, + "grad_norm": 0.45206335300018474, + "learning_rate": 4.4399999999999995e-05, + "loss": 0.0455, + "step": 296 + }, + { + "epoch": 0.13976470588235293, + "grad_norm": 0.7377247862543327, + "learning_rate": 4.454999999999999e-05, + "loss": 0.0926, + "step": 297 + }, + { + "epoch": 0.14023529411764707, + "grad_norm": 0.4141599137115635, + "learning_rate": 4.4699999999999996e-05, + "loss": 0.0506, + "step": 298 + }, + { + "epoch": 0.14070588235294118, + "grad_norm": 0.25759177004337247, + "learning_rate": 4.484999999999999e-05, + "loss": 0.0148, + "step": 299 + }, + { + "epoch": 0.1411764705882353, + "grad_norm": 0.5518999898171434, + "learning_rate": 4.4999999999999996e-05, + "loss": 0.1274, + "step": 300 + }, + { + "epoch": 0.1416470588235294, + "grad_norm": 0.7318504817797742, + "learning_rate": 4.514999999999999e-05, + "loss": 0.11, + "step": 301 + }, + { + "epoch": 0.14211764705882354, + "grad_norm": 0.6584800425269607, + "learning_rate": 4.5299999999999997e-05, + "loss": 0.1711, + "step": 302 + }, + { + "epoch": 0.14258823529411765, + "grad_norm": 0.5864477162379889, + "learning_rate": 4.5449999999999993e-05, + "loss": 0.0655, + "step": 303 + }, + { + "epoch": 0.14305882352941177, + "grad_norm": 0.4873696403325203, + "learning_rate": 4.56e-05, + "loss": 0.0881, + "step": 304 + }, + { + "epoch": 0.14352941176470588, + "grad_norm": 0.7932491951726061, + "learning_rate": 4.5749999999999994e-05, + "loss": 0.159, + "step": 305 + }, + { + "epoch": 0.144, + "grad_norm": 0.626719910869147, + "learning_rate": 4.59e-05, + "loss": 0.1168, + "step": 306 + }, + { + "epoch": 0.14447058823529413, + "grad_norm": 0.42308060870884995, + "learning_rate": 4.6049999999999994e-05, + "loss": 0.0802, + "step": 307 + }, + { + "epoch": 0.14494117647058824, + "grad_norm": 0.4359215970661929, + "learning_rate": 4.62e-05, + "loss": 0.0906, + "step": 308 + }, + { + "epoch": 0.14541176470588235, + "grad_norm": 0.6005793767328336, + "learning_rate": 4.6349999999999995e-05, + "loss": 0.1064, + "step": 309 + }, + { + "epoch": 0.14588235294117646, + "grad_norm": 0.21572351298125764, + "learning_rate": 4.65e-05, + "loss": 0.0281, + "step": 310 + }, + { + "epoch": 0.14635294117647057, + "grad_norm": 0.3587326753882433, + "learning_rate": 4.6649999999999996e-05, + "loss": 0.0321, + "step": 311 + }, + { + "epoch": 0.1468235294117647, + "grad_norm": 0.4325915669170649, + "learning_rate": 4.68e-05, + "loss": 0.0462, + "step": 312 + }, + { + "epoch": 0.14729411764705883, + "grad_norm": 0.44391210736873177, + "learning_rate": 4.6949999999999996e-05, + "loss": 0.0789, + "step": 313 + }, + { + "epoch": 0.14776470588235294, + "grad_norm": 0.06686247188713114, + "learning_rate": 4.709999999999999e-05, + "loss": 0.0009, + "step": 314 + }, + { + "epoch": 0.14823529411764705, + "grad_norm": 0.06577720117177746, + "learning_rate": 4.7249999999999997e-05, + "loss": 0.0009, + "step": 315 + }, + { + "epoch": 0.1487058823529412, + "grad_norm": 0.055385147436607586, + "learning_rate": 4.7399999999999993e-05, + "loss": 0.0008, + "step": 316 + }, + { + "epoch": 0.1491764705882353, + "grad_norm": 0.2953113703622541, + "learning_rate": 4.755e-05, + "loss": 0.052, + "step": 317 + }, + { + "epoch": 0.1496470588235294, + "grad_norm": 0.31885911307735604, + "learning_rate": 4.7699999999999994e-05, + "loss": 0.0627, + "step": 318 + }, + { + "epoch": 0.15011764705882352, + "grad_norm": 0.04314731107666473, + "learning_rate": 4.785e-05, + "loss": 0.0006, + "step": 319 + }, + { + "epoch": 0.15058823529411763, + "grad_norm": 0.7462285910561057, + "learning_rate": 4.7999999999999994e-05, + "loss": 0.1769, + "step": 320 + }, + { + "epoch": 0.15105882352941177, + "grad_norm": 0.036014190558841724, + "learning_rate": 4.815e-05, + "loss": 0.0005, + "step": 321 + }, + { + "epoch": 0.15152941176470588, + "grad_norm": 0.03499421819309583, + "learning_rate": 4.8299999999999995e-05, + "loss": 0.0005, + "step": 322 + }, + { + "epoch": 0.152, + "grad_norm": 0.36347053828688997, + "learning_rate": 4.845e-05, + "loss": 0.0531, + "step": 323 + }, + { + "epoch": 0.1524705882352941, + "grad_norm": 0.6177408793706001, + "learning_rate": 4.8599999999999995e-05, + "loss": 0.1117, + "step": 324 + }, + { + "epoch": 0.15294117647058825, + "grad_norm": 0.30135211268266826, + "learning_rate": 4.875e-05, + "loss": 0.0514, + "step": 325 + }, + { + "epoch": 0.15341176470588236, + "grad_norm": 0.27225379825222884, + "learning_rate": 4.8899999999999996e-05, + "loss": 0.0255, + "step": 326 + }, + { + "epoch": 0.15388235294117647, + "grad_norm": 0.24454422255275915, + "learning_rate": 4.905e-05, + "loss": 0.0325, + "step": 327 + }, + { + "epoch": 0.15435294117647058, + "grad_norm": 0.025709010794471954, + "learning_rate": 4.9199999999999997e-05, + "loss": 0.0004, + "step": 328 + }, + { + "epoch": 0.1548235294117647, + "grad_norm": 0.6967479752336637, + "learning_rate": 4.935e-05, + "loss": 0.102, + "step": 329 + }, + { + "epoch": 0.15529411764705883, + "grad_norm": 0.5386931170822591, + "learning_rate": 4.95e-05, + "loss": 0.1221, + "step": 330 + }, + { + "epoch": 0.15576470588235294, + "grad_norm": 0.3531125137582149, + "learning_rate": 4.965e-05, + "loss": 0.0782, + "step": 331 + }, + { + "epoch": 0.15623529411764706, + "grad_norm": 0.2839625842875903, + "learning_rate": 4.98e-05, + "loss": 0.0421, + "step": 332 + }, + { + "epoch": 0.15670588235294117, + "grad_norm": 0.02211394033707424, + "learning_rate": 4.995e-05, + "loss": 0.0003, + "step": 333 + }, + { + "epoch": 0.1571764705882353, + "grad_norm": 0.024651152261976162, + "learning_rate": 5.01e-05, + "loss": 0.0004, + "step": 334 + }, + { + "epoch": 0.15764705882352942, + "grad_norm": 0.33923988946576983, + "learning_rate": 5.025e-05, + "loss": 0.0297, + "step": 335 + }, + { + "epoch": 0.15811764705882353, + "grad_norm": 0.6256660351654192, + "learning_rate": 5.04e-05, + "loss": 0.1549, + "step": 336 + }, + { + "epoch": 0.15858823529411764, + "grad_norm": 0.3192535845136378, + "learning_rate": 5.055e-05, + "loss": 0.0675, + "step": 337 + }, + { + "epoch": 0.15905882352941175, + "grad_norm": 0.020941088867125816, + "learning_rate": 5.07e-05, + "loss": 0.0003, + "step": 338 + }, + { + "epoch": 0.1595294117647059, + "grad_norm": 0.3206522200152038, + "learning_rate": 5.0849999999999996e-05, + "loss": 0.0447, + "step": 339 + }, + { + "epoch": 0.16, + "grad_norm": 0.5030016412354531, + "learning_rate": 5.1e-05, + "loss": 0.0823, + "step": 340 + }, + { + "epoch": 0.16047058823529411, + "grad_norm": 0.555375192944742, + "learning_rate": 5.1149999999999996e-05, + "loss": 0.0924, + "step": 341 + }, + { + "epoch": 0.16094117647058823, + "grad_norm": 0.01881701481999482, + "learning_rate": 5.13e-05, + "loss": 0.0003, + "step": 342 + }, + { + "epoch": 0.16141176470588237, + "grad_norm": 0.04737577603140962, + "learning_rate": 5.145e-05, + "loss": 0.0022, + "step": 343 + }, + { + "epoch": 0.16188235294117648, + "grad_norm": 0.6300903144432766, + "learning_rate": 5.1599999999999994e-05, + "loss": 0.1151, + "step": 344 + }, + { + "epoch": 0.1623529411764706, + "grad_norm": 0.5236997774881683, + "learning_rate": 5.174999999999999e-05, + "loss": 0.0431, + "step": 345 + }, + { + "epoch": 0.1628235294117647, + "grad_norm": 0.03814438791103473, + "learning_rate": 5.1899999999999994e-05, + "loss": 0.0019, + "step": 346 + }, + { + "epoch": 0.1632941176470588, + "grad_norm": 0.4401465129540934, + "learning_rate": 5.204999999999999e-05, + "loss": 0.0986, + "step": 347 + }, + { + "epoch": 0.16376470588235295, + "grad_norm": 0.17285660931116414, + "learning_rate": 5.2199999999999995e-05, + "loss": 0.0057, + "step": 348 + }, + { + "epoch": 0.16423529411764706, + "grad_norm": 0.013101461301262586, + "learning_rate": 5.234999999999999e-05, + "loss": 0.0002, + "step": 349 + }, + { + "epoch": 0.16470588235294117, + "grad_norm": 0.7306988476660407, + "learning_rate": 5.2499999999999995e-05, + "loss": 0.0899, + "step": 350 + }, + { + "epoch": 0.16517647058823529, + "grad_norm": 0.01234888611379341, + "learning_rate": 5.264999999999999e-05, + "loss": 0.0002, + "step": 351 + }, + { + "epoch": 0.16564705882352943, + "grad_norm": 0.012544871052527752, + "learning_rate": 5.279999999999999e-05, + "loss": 0.0002, + "step": 352 + }, + { + "epoch": 0.16611764705882354, + "grad_norm": 0.49944166438238596, + "learning_rate": 5.294999999999999e-05, + "loss": 0.0459, + "step": 353 + }, + { + "epoch": 0.16658823529411765, + "grad_norm": 0.426500725082891, + "learning_rate": 5.309999999999999e-05, + "loss": 0.0588, + "step": 354 + }, + { + "epoch": 0.16705882352941176, + "grad_norm": 0.15561357638359646, + "learning_rate": 5.324999999999999e-05, + "loss": 0.0213, + "step": 355 + }, + { + "epoch": 0.16752941176470587, + "grad_norm": 0.18587576080707038, + "learning_rate": 5.339999999999999e-05, + "loss": 0.017, + "step": 356 + }, + { + "epoch": 0.168, + "grad_norm": 0.16335059258331877, + "learning_rate": 5.3549999999999994e-05, + "loss": 0.0179, + "step": 357 + }, + { + "epoch": 0.16847058823529412, + "grad_norm": 0.15237999079783252, + "learning_rate": 5.369999999999999e-05, + "loss": 0.0242, + "step": 358 + }, + { + "epoch": 0.16894117647058823, + "grad_norm": 0.37047251987108953, + "learning_rate": 5.3849999999999994e-05, + "loss": 0.0563, + "step": 359 + }, + { + "epoch": 0.16941176470588235, + "grad_norm": 0.41813434885834, + "learning_rate": 5.399999999999999e-05, + "loss": 0.0362, + "step": 360 + }, + { + "epoch": 0.16988235294117648, + "grad_norm": 0.46137308102874103, + "learning_rate": 5.4149999999999995e-05, + "loss": 0.1043, + "step": 361 + }, + { + "epoch": 0.1703529411764706, + "grad_norm": 0.01937793778756034, + "learning_rate": 5.429999999999999e-05, + "loss": 0.0003, + "step": 362 + }, + { + "epoch": 0.1708235294117647, + "grad_norm": 0.16714967858926563, + "learning_rate": 5.4449999999999995e-05, + "loss": 0.0159, + "step": 363 + }, + { + "epoch": 0.17129411764705882, + "grad_norm": 0.4616127591040911, + "learning_rate": 5.459999999999999e-05, + "loss": 0.1047, + "step": 364 + }, + { + "epoch": 0.17176470588235293, + "grad_norm": 0.29546315466391393, + "learning_rate": 5.4749999999999996e-05, + "loss": 0.048, + "step": 365 + }, + { + "epoch": 0.17223529411764707, + "grad_norm": 0.43785548176894845, + "learning_rate": 5.489999999999999e-05, + "loss": 0.0702, + "step": 366 + }, + { + "epoch": 0.17270588235294118, + "grad_norm": 0.7825563650435774, + "learning_rate": 5.5049999999999996e-05, + "loss": 0.1294, + "step": 367 + }, + { + "epoch": 0.1731764705882353, + "grad_norm": 0.1407389013186677, + "learning_rate": 5.519999999999999e-05, + "loss": 0.0152, + "step": 368 + }, + { + "epoch": 0.1736470588235294, + "grad_norm": 0.3736367203813344, + "learning_rate": 5.535e-05, + "loss": 0.0759, + "step": 369 + }, + { + "epoch": 0.17411764705882352, + "grad_norm": 0.5585858320716944, + "learning_rate": 5.5499999999999994e-05, + "loss": 0.0579, + "step": 370 + }, + { + "epoch": 0.17458823529411766, + "grad_norm": 0.20758358890571685, + "learning_rate": 5.565e-05, + "loss": 0.0097, + "step": 371 + }, + { + "epoch": 0.17505882352941177, + "grad_norm": 0.4696705002453773, + "learning_rate": 5.5799999999999994e-05, + "loss": 0.0483, + "step": 372 + }, + { + "epoch": 0.17552941176470588, + "grad_norm": 0.28981838581149627, + "learning_rate": 5.595e-05, + "loss": 0.0392, + "step": 373 + }, + { + "epoch": 0.176, + "grad_norm": 0.15923078807667726, + "learning_rate": 5.6099999999999995e-05, + "loss": 0.0094, + "step": 374 + }, + { + "epoch": 0.17647058823529413, + "grad_norm": 0.47454023441100707, + "learning_rate": 5.625e-05, + "loss": 0.0971, + "step": 375 + }, + { + "epoch": 0.17694117647058824, + "grad_norm": 0.4506446895033084, + "learning_rate": 5.6399999999999995e-05, + "loss": 0.1204, + "step": 376 + }, + { + "epoch": 0.17741176470588235, + "grad_norm": 0.13464650039682144, + "learning_rate": 5.654999999999999e-05, + "loss": 0.0177, + "step": 377 + }, + { + "epoch": 0.17788235294117646, + "grad_norm": 0.37065936964201046, + "learning_rate": 5.6699999999999996e-05, + "loss": 0.0524, + "step": 378 + }, + { + "epoch": 0.17835294117647058, + "grad_norm": 0.2756698552410824, + "learning_rate": 5.684999999999999e-05, + "loss": 0.0443, + "step": 379 + }, + { + "epoch": 0.17882352941176471, + "grad_norm": 0.12465915775748541, + "learning_rate": 5.6999999999999996e-05, + "loss": 0.0161, + "step": 380 + }, + { + "epoch": 0.17929411764705883, + "grad_norm": 0.01611334139786362, + "learning_rate": 5.714999999999999e-05, + "loss": 0.0002, + "step": 381 + }, + { + "epoch": 0.17976470588235294, + "grad_norm": 0.6091275350645304, + "learning_rate": 5.73e-05, + "loss": 0.1741, + "step": 382 + }, + { + "epoch": 0.18023529411764705, + "grad_norm": 0.14927112055839303, + "learning_rate": 5.7449999999999994e-05, + "loss": 0.024, + "step": 383 + }, + { + "epoch": 0.1807058823529412, + "grad_norm": 0.21636969614578544, + "learning_rate": 5.76e-05, + "loss": 0.0389, + "step": 384 + }, + { + "epoch": 0.1811764705882353, + "grad_norm": 0.41828549429144407, + "learning_rate": 5.7749999999999994e-05, + "loss": 0.0773, + "step": 385 + }, + { + "epoch": 0.1816470588235294, + "grad_norm": 0.9131023309829566, + "learning_rate": 5.79e-05, + "loss": 0.0351, + "step": 386 + }, + { + "epoch": 0.18211764705882352, + "grad_norm": 0.24350752169158302, + "learning_rate": 5.8049999999999995e-05, + "loss": 0.0592, + "step": 387 + }, + { + "epoch": 0.18258823529411763, + "grad_norm": 0.46923899541196884, + "learning_rate": 5.82e-05, + "loss": 0.1481, + "step": 388 + }, + { + "epoch": 0.18305882352941177, + "grad_norm": 0.13411119473443434, + "learning_rate": 5.8349999999999995e-05, + "loss": 0.0237, + "step": 389 + }, + { + "epoch": 0.18352941176470589, + "grad_norm": 0.28535325071570367, + "learning_rate": 5.85e-05, + "loss": 0.0633, + "step": 390 + }, + { + "epoch": 0.184, + "grad_norm": 0.030589645497125336, + "learning_rate": 5.8649999999999996e-05, + "loss": 0.0005, + "step": 391 + }, + { + "epoch": 0.1844705882352941, + "grad_norm": 0.01848166192237216, + "learning_rate": 5.88e-05, + "loss": 0.0003, + "step": 392 + }, + { + "epoch": 0.18494117647058825, + "grad_norm": 0.05913326286012348, + "learning_rate": 5.8949999999999996e-05, + "loss": 0.0067, + "step": 393 + }, + { + "epoch": 0.18541176470588236, + "grad_norm": 0.19709484933840674, + "learning_rate": 5.91e-05, + "loss": 0.0343, + "step": 394 + }, + { + "epoch": 0.18588235294117647, + "grad_norm": 0.0745372664811943, + "learning_rate": 5.925e-05, + "loss": 0.0076, + "step": 395 + }, + { + "epoch": 0.18635294117647058, + "grad_norm": 0.5985962150316175, + "learning_rate": 5.94e-05, + "loss": 0.1332, + "step": 396 + }, + { + "epoch": 0.1868235294117647, + "grad_norm": 0.10179167775212421, + "learning_rate": 5.955e-05, + "loss": 0.0186, + "step": 397 + }, + { + "epoch": 0.18729411764705883, + "grad_norm": 0.03796067733651243, + "learning_rate": 5.97e-05, + "loss": 0.0008, + "step": 398 + }, + { + "epoch": 0.18776470588235294, + "grad_norm": 0.4311564567719575, + "learning_rate": 5.985e-05, + "loss": 0.1187, + "step": 399 + }, + { + "epoch": 0.18823529411764706, + "grad_norm": 0.1623044967748219, + "learning_rate": 5.9999999999999995e-05, + "loss": 0.0048, + "step": 400 + }, + { + "epoch": 0.18870588235294117, + "grad_norm": 0.3005749886284318, + "learning_rate": 6.015e-05, + "loss": 0.0512, + "step": 401 + }, + { + "epoch": 0.1891764705882353, + "grad_norm": 0.17390501510199885, + "learning_rate": 6.0299999999999995e-05, + "loss": 0.0252, + "step": 402 + }, + { + "epoch": 0.18964705882352942, + "grad_norm": 0.27996332614297836, + "learning_rate": 6.045e-05, + "loss": 0.0774, + "step": 403 + }, + { + "epoch": 0.19011764705882353, + "grad_norm": 0.15209107981282272, + "learning_rate": 6.0599999999999996e-05, + "loss": 0.0314, + "step": 404 + }, + { + "epoch": 0.19058823529411764, + "grad_norm": 0.11377273749780598, + "learning_rate": 6.075e-05, + "loss": 0.0122, + "step": 405 + }, + { + "epoch": 0.19105882352941175, + "grad_norm": 0.27227584449108977, + "learning_rate": 6.0899999999999996e-05, + "loss": 0.0781, + "step": 406 + }, + { + "epoch": 0.1915294117647059, + "grad_norm": 0.2649620011793759, + "learning_rate": 6.104999999999999e-05, + "loss": 0.0559, + "step": 407 + }, + { + "epoch": 0.192, + "grad_norm": 0.2655874993544058, + "learning_rate": 6.12e-05, + "loss": 0.0576, + "step": 408 + }, + { + "epoch": 0.19247058823529412, + "grad_norm": 0.00781763536030228, + "learning_rate": 6.134999999999999e-05, + "loss": 0.0001, + "step": 409 + }, + { + "epoch": 0.19294117647058823, + "grad_norm": 0.34090257847214783, + "learning_rate": 6.149999999999999e-05, + "loss": 0.1077, + "step": 410 + }, + { + "epoch": 0.19341176470588237, + "grad_norm": 0.40860534928204023, + "learning_rate": 6.165e-05, + "loss": 0.1032, + "step": 411 + }, + { + "epoch": 0.19388235294117648, + "grad_norm": 0.24034703382905884, + "learning_rate": 6.18e-05, + "loss": 0.0359, + "step": 412 + }, + { + "epoch": 0.1943529411764706, + "grad_norm": 0.1307010882457204, + "learning_rate": 6.194999999999999e-05, + "loss": 0.0156, + "step": 413 + }, + { + "epoch": 0.1948235294117647, + "grad_norm": 0.3469452080851705, + "learning_rate": 6.209999999999999e-05, + "loss": 0.072, + "step": 414 + }, + { + "epoch": 0.1952941176470588, + "grad_norm": 0.17006316906484034, + "learning_rate": 6.225e-05, + "loss": 0.0315, + "step": 415 + }, + { + "epoch": 0.19576470588235295, + "grad_norm": 0.3624082095624842, + "learning_rate": 6.239999999999999e-05, + "loss": 0.0709, + "step": 416 + }, + { + "epoch": 0.19623529411764706, + "grad_norm": 0.5263828333550717, + "learning_rate": 6.254999999999999e-05, + "loss": 0.0798, + "step": 417 + }, + { + "epoch": 0.19670588235294117, + "grad_norm": 0.26533074616196956, + "learning_rate": 6.269999999999999e-05, + "loss": 0.0703, + "step": 418 + }, + { + "epoch": 0.1971764705882353, + "grad_norm": 0.007373305279375864, + "learning_rate": 6.285e-05, + "loss": 0.0001, + "step": 419 + }, + { + "epoch": 0.1976470588235294, + "grad_norm": 0.006175986653150406, + "learning_rate": 6.299999999999999e-05, + "loss": 0.0001, + "step": 420 + }, + { + "epoch": 0.19811764705882354, + "grad_norm": 0.04706352339746888, + "learning_rate": 6.314999999999999e-05, + "loss": 0.0028, + "step": 421 + }, + { + "epoch": 0.19858823529411765, + "grad_norm": 0.265966208642225, + "learning_rate": 6.33e-05, + "loss": 0.0461, + "step": 422 + }, + { + "epoch": 0.19905882352941176, + "grad_norm": 0.014476907323079175, + "learning_rate": 6.345e-05, + "loss": 0.0002, + "step": 423 + }, + { + "epoch": 0.19952941176470587, + "grad_norm": 0.48849934919863774, + "learning_rate": 6.359999999999999e-05, + "loss": 0.0544, + "step": 424 + }, + { + "epoch": 0.2, + "grad_norm": 0.21674232645699942, + "learning_rate": 6.374999999999999e-05, + "loss": 0.0268, + "step": 425 + }, + { + "epoch": 0.20047058823529412, + "grad_norm": 0.7237355365465177, + "learning_rate": 6.39e-05, + "loss": 0.1665, + "step": 426 + }, + { + "epoch": 0.20094117647058823, + "grad_norm": 0.034126960599915414, + "learning_rate": 6.405e-05, + "loss": 0.0003, + "step": 427 + }, + { + "epoch": 0.20141176470588235, + "grad_norm": 0.3812462317138448, + "learning_rate": 6.419999999999999e-05, + "loss": 0.0786, + "step": 428 + }, + { + "epoch": 0.20188235294117646, + "grad_norm": 0.0037239580543833952, + "learning_rate": 6.434999999999999e-05, + "loss": 0.0, + "step": 429 + }, + { + "epoch": 0.2023529411764706, + "grad_norm": 0.35566984226825454, + "learning_rate": 6.45e-05, + "loss": 0.0418, + "step": 430 + }, + { + "epoch": 0.2028235294117647, + "grad_norm": 0.20682483489090372, + "learning_rate": 6.465e-05, + "loss": 0.0157, + "step": 431 + }, + { + "epoch": 0.20329411764705882, + "grad_norm": 0.27918847920234974, + "learning_rate": 6.479999999999999e-05, + "loss": 0.0489, + "step": 432 + }, + { + "epoch": 0.20376470588235293, + "grad_norm": 0.0028328267563786265, + "learning_rate": 6.494999999999999e-05, + "loss": 0.0, + "step": 433 + }, + { + "epoch": 0.20423529411764707, + "grad_norm": 0.2526645022429795, + "learning_rate": 6.51e-05, + "loss": 0.0646, + "step": 434 + }, + { + "epoch": 0.20470588235294118, + "grad_norm": 0.4307884232747386, + "learning_rate": 6.525e-05, + "loss": 0.1365, + "step": 435 + }, + { + "epoch": 0.2051764705882353, + "grad_norm": 0.3460256544764087, + "learning_rate": 6.539999999999999e-05, + "loss": 0.1075, + "step": 436 + }, + { + "epoch": 0.2056470588235294, + "grad_norm": 0.18761924622368847, + "learning_rate": 6.555e-05, + "loss": 0.0528, + "step": 437 + }, + { + "epoch": 0.20611764705882352, + "grad_norm": 0.0022464421289071852, + "learning_rate": 6.57e-05, + "loss": 0.0, + "step": 438 + }, + { + "epoch": 0.20658823529411766, + "grad_norm": 0.23767267245635282, + "learning_rate": 6.584999999999999e-05, + "loss": 0.0346, + "step": 439 + }, + { + "epoch": 0.20705882352941177, + "grad_norm": 0.22550408484086323, + "learning_rate": 6.599999999999999e-05, + "loss": 0.0338, + "step": 440 + }, + { + "epoch": 0.20752941176470588, + "grad_norm": 0.015915362974893425, + "learning_rate": 6.615e-05, + "loss": 0.0002, + "step": 441 + }, + { + "epoch": 0.208, + "grad_norm": 0.578559702349361, + "learning_rate": 6.63e-05, + "loss": 0.1071, + "step": 442 + }, + { + "epoch": 0.20847058823529413, + "grad_norm": 0.348413870268156, + "learning_rate": 6.644999999999999e-05, + "loss": 0.103, + "step": 443 + }, + { + "epoch": 0.20894117647058824, + "grad_norm": 0.30746380163117754, + "learning_rate": 6.659999999999999e-05, + "loss": 0.0805, + "step": 444 + }, + { + "epoch": 0.20941176470588235, + "grad_norm": 0.10163249343953455, + "learning_rate": 6.675e-05, + "loss": 0.0151, + "step": 445 + }, + { + "epoch": 0.20988235294117646, + "grad_norm": 0.16260749974669247, + "learning_rate": 6.69e-05, + "loss": 0.0274, + "step": 446 + }, + { + "epoch": 0.21035294117647058, + "grad_norm": 0.29340068634075867, + "learning_rate": 6.704999999999999e-05, + "loss": 0.0771, + "step": 447 + }, + { + "epoch": 0.21082352941176472, + "grad_norm": 0.2020522563147642, + "learning_rate": 6.72e-05, + "loss": 0.0525, + "step": 448 + }, + { + "epoch": 0.21129411764705883, + "grad_norm": 0.1348553997475705, + "learning_rate": 6.735e-05, + "loss": 0.0232, + "step": 449 + }, + { + "epoch": 0.21176470588235294, + "grad_norm": 0.2644408327968712, + "learning_rate": 6.75e-05, + "loss": 0.018, + "step": 450 + }, + { + "epoch": 0.21223529411764705, + "grad_norm": 0.3521374876867396, + "learning_rate": 6.764999999999999e-05, + "loss": 0.1076, + "step": 451 + }, + { + "epoch": 0.2127058823529412, + "grad_norm": 0.41894438016656915, + "learning_rate": 6.78e-05, + "loss": 0.1245, + "step": 452 + }, + { + "epoch": 0.2131764705882353, + "grad_norm": 0.13937755888158732, + "learning_rate": 6.795e-05, + "loss": 0.0154, + "step": 453 + }, + { + "epoch": 0.2136470588235294, + "grad_norm": 0.0082451281848511, + "learning_rate": 6.81e-05, + "loss": 0.0001, + "step": 454 + }, + { + "epoch": 0.21411764705882352, + "grad_norm": 0.164821885703681, + "learning_rate": 6.824999999999999e-05, + "loss": 0.0131, + "step": 455 + }, + { + "epoch": 0.21458823529411764, + "grad_norm": 0.2357480748126528, + "learning_rate": 6.84e-05, + "loss": 0.0473, + "step": 456 + }, + { + "epoch": 0.21505882352941177, + "grad_norm": 0.3508983066469006, + "learning_rate": 6.855e-05, + "loss": 0.0793, + "step": 457 + }, + { + "epoch": 0.21552941176470589, + "grad_norm": 0.2367459680782197, + "learning_rate": 6.87e-05, + "loss": 0.039, + "step": 458 + }, + { + "epoch": 0.216, + "grad_norm": 0.5009939662541293, + "learning_rate": 6.884999999999999e-05, + "loss": 0.0999, + "step": 459 + }, + { + "epoch": 0.2164705882352941, + "grad_norm": 0.46612901633688775, + "learning_rate": 6.9e-05, + "loss": 0.1325, + "step": 460 + }, + { + "epoch": 0.21694117647058825, + "grad_norm": 0.24186526115147589, + "learning_rate": 6.915e-05, + "loss": 0.0807, + "step": 461 + }, + { + "epoch": 0.21741176470588236, + "grad_norm": 0.4151982445182383, + "learning_rate": 6.93e-05, + "loss": 0.0744, + "step": 462 + }, + { + "epoch": 0.21788235294117647, + "grad_norm": 0.0013271067100703768, + "learning_rate": 6.945e-05, + "loss": 0.0, + "step": 463 + }, + { + "epoch": 0.21835294117647058, + "grad_norm": 0.5846143305963367, + "learning_rate": 6.96e-05, + "loss": 0.1801, + "step": 464 + }, + { + "epoch": 0.2188235294117647, + "grad_norm": 0.1519000560447201, + "learning_rate": 6.975e-05, + "loss": 0.025, + "step": 465 + }, + { + "epoch": 0.21929411764705883, + "grad_norm": 0.1286045885388513, + "learning_rate": 6.989999999999999e-05, + "loss": 0.0242, + "step": 466 + }, + { + "epoch": 0.21976470588235295, + "grad_norm": 0.4577145471816805, + "learning_rate": 7.005e-05, + "loss": 0.1303, + "step": 467 + }, + { + "epoch": 0.22023529411764706, + "grad_norm": 0.20996711213282448, + "learning_rate": 7.02e-05, + "loss": 0.0588, + "step": 468 + }, + { + "epoch": 0.22070588235294117, + "grad_norm": 0.06544532795467337, + "learning_rate": 7.034999999999999e-05, + "loss": 0.0108, + "step": 469 + }, + { + "epoch": 0.2211764705882353, + "grad_norm": 0.04457771660869315, + "learning_rate": 7.049999999999999e-05, + "loss": 0.0039, + "step": 470 + }, + { + "epoch": 0.22164705882352942, + "grad_norm": 0.36196040249291, + "learning_rate": 7.065e-05, + "loss": 0.0732, + "step": 471 + }, + { + "epoch": 0.22211764705882353, + "grad_norm": 0.3032172272731772, + "learning_rate": 7.079999999999999e-05, + "loss": 0.0437, + "step": 472 + }, + { + "epoch": 0.22258823529411764, + "grad_norm": 0.27546767559930174, + "learning_rate": 7.094999999999999e-05, + "loss": 0.0749, + "step": 473 + }, + { + "epoch": 0.22305882352941175, + "grad_norm": 0.41542090463652076, + "learning_rate": 7.11e-05, + "loss": 0.0447, + "step": 474 + }, + { + "epoch": 0.2235294117647059, + "grad_norm": 0.1860767085285441, + "learning_rate": 7.125e-05, + "loss": 0.0453, + "step": 475 + }, + { + "epoch": 0.224, + "grad_norm": 0.24055920923106397, + "learning_rate": 7.139999999999999e-05, + "loss": 0.0269, + "step": 476 + }, + { + "epoch": 0.22447058823529412, + "grad_norm": 0.42193273255952657, + "learning_rate": 7.154999999999999e-05, + "loss": 0.0727, + "step": 477 + }, + { + "epoch": 0.22494117647058823, + "grad_norm": 0.3683861633301569, + "learning_rate": 7.17e-05, + "loss": 0.0725, + "step": 478 + }, + { + "epoch": 0.22541176470588234, + "grad_norm": 0.05536144775472802, + "learning_rate": 7.184999999999998e-05, + "loss": 0.006, + "step": 479 + }, + { + "epoch": 0.22588235294117648, + "grad_norm": 0.1386007477968992, + "learning_rate": 7.199999999999999e-05, + "loss": 0.026, + "step": 480 + }, + { + "epoch": 0.2263529411764706, + "grad_norm": 0.39889417086340834, + "learning_rate": 7.214999999999999e-05, + "loss": 0.0813, + "step": 481 + }, + { + "epoch": 0.2268235294117647, + "grad_norm": 0.20486223069904722, + "learning_rate": 7.23e-05, + "loss": 0.0315, + "step": 482 + }, + { + "epoch": 0.2272941176470588, + "grad_norm": 0.3030175510948233, + "learning_rate": 7.244999999999999e-05, + "loss": 0.0611, + "step": 483 + }, + { + "epoch": 0.22776470588235295, + "grad_norm": 0.1246974366457326, + "learning_rate": 7.259999999999999e-05, + "loss": 0.0081, + "step": 484 + }, + { + "epoch": 0.22823529411764706, + "grad_norm": 0.2625317551132651, + "learning_rate": 7.274999999999999e-05, + "loss": 0.0605, + "step": 485 + }, + { + "epoch": 0.22870588235294118, + "grad_norm": 0.20527785408034754, + "learning_rate": 7.29e-05, + "loss": 0.04, + "step": 486 + }, + { + "epoch": 0.2291764705882353, + "grad_norm": 0.23173210013667972, + "learning_rate": 7.304999999999999e-05, + "loss": 0.0409, + "step": 487 + }, + { + "epoch": 0.2296470588235294, + "grad_norm": 0.7256444606951424, + "learning_rate": 7.319999999999999e-05, + "loss": 0.0239, + "step": 488 + }, + { + "epoch": 0.23011764705882354, + "grad_norm": 0.0008336449692627138, + "learning_rate": 7.335e-05, + "loss": 0.0, + "step": 489 + }, + { + "epoch": 0.23058823529411765, + "grad_norm": 0.24923096453740257, + "learning_rate": 7.35e-05, + "loss": 0.0275, + "step": 490 + }, + { + "epoch": 0.23105882352941176, + "grad_norm": 0.19705981776640585, + "learning_rate": 7.364999999999999e-05, + "loss": 0.0327, + "step": 491 + }, + { + "epoch": 0.23152941176470587, + "grad_norm": 0.0007174657343052605, + "learning_rate": 7.379999999999999e-05, + "loss": 0.0, + "step": 492 + }, + { + "epoch": 0.232, + "grad_norm": 0.7147258297884287, + "learning_rate": 7.395e-05, + "loss": 0.1646, + "step": 493 + }, + { + "epoch": 0.23247058823529412, + "grad_norm": 0.19613991150809063, + "learning_rate": 7.41e-05, + "loss": 0.034, + "step": 494 + }, + { + "epoch": 0.23294117647058823, + "grad_norm": 0.22649699657310787, + "learning_rate": 7.424999999999999e-05, + "loss": 0.0227, + "step": 495 + }, + { + "epoch": 0.23341176470588235, + "grad_norm": 0.3758109460176854, + "learning_rate": 7.439999999999999e-05, + "loss": 0.088, + "step": 496 + }, + { + "epoch": 0.23388235294117646, + "grad_norm": 0.47161801044849866, + "learning_rate": 7.455e-05, + "loss": 0.0787, + "step": 497 + }, + { + "epoch": 0.2343529411764706, + "grad_norm": 0.0005749290570322706, + "learning_rate": 7.47e-05, + "loss": 0.0, + "step": 498 + }, + { + "epoch": 0.2348235294117647, + "grad_norm": 0.2195591859907323, + "learning_rate": 7.484999999999999e-05, + "loss": 0.0364, + "step": 499 + }, + { + "epoch": 0.23529411764705882, + "grad_norm": 0.4041953864401068, + "learning_rate": 7.5e-05, + "loss": 0.0997, + "step": 500 + }, + { + "epoch": 0.23576470588235293, + "grad_norm": 0.0013887279635641512, + "learning_rate": 7.515e-05, + "loss": 0.0, + "step": 501 + }, + { + "epoch": 0.23623529411764707, + "grad_norm": 0.0013387698973815292, + "learning_rate": 7.529999999999999e-05, + "loss": 0.0, + "step": 502 + }, + { + "epoch": 0.23670588235294118, + "grad_norm": 0.0011774855817452544, + "learning_rate": 7.544999999999999e-05, + "loss": 0.0, + "step": 503 + }, + { + "epoch": 0.2371764705882353, + "grad_norm": 0.2695953326985913, + "learning_rate": 7.56e-05, + "loss": 0.0522, + "step": 504 + }, + { + "epoch": 0.2376470588235294, + "grad_norm": 0.2608736214839984, + "learning_rate": 7.575e-05, + "loss": 0.0644, + "step": 505 + }, + { + "epoch": 0.23811764705882352, + "grad_norm": 0.4715056930104455, + "learning_rate": 7.589999999999999e-05, + "loss": 0.0839, + "step": 506 + }, + { + "epoch": 0.23858823529411766, + "grad_norm": 0.5789937371207308, + "learning_rate": 7.604999999999999e-05, + "loss": 0.0957, + "step": 507 + }, + { + "epoch": 0.23905882352941177, + "grad_norm": 0.36336799855510726, + "learning_rate": 7.62e-05, + "loss": 0.0427, + "step": 508 + }, + { + "epoch": 0.23952941176470588, + "grad_norm": 0.28746183493064414, + "learning_rate": 7.635e-05, + "loss": 0.0406, + "step": 509 + }, + { + "epoch": 0.24, + "grad_norm": 0.27964231554113966, + "learning_rate": 7.649999999999999e-05, + "loss": 0.0163, + "step": 510 + }, + { + "epoch": 0.24047058823529413, + "grad_norm": 0.5438259319141486, + "learning_rate": 7.664999999999999e-05, + "loss": 0.1744, + "step": 511 + }, + { + "epoch": 0.24094117647058824, + "grad_norm": 0.17429957899282875, + "learning_rate": 7.68e-05, + "loss": 0.0374, + "step": 512 + }, + { + "epoch": 0.24141176470588235, + "grad_norm": 0.3285684855158139, + "learning_rate": 7.695e-05, + "loss": 0.0732, + "step": 513 + }, + { + "epoch": 0.24188235294117647, + "grad_norm": 0.1556741594222514, + "learning_rate": 7.709999999999999e-05, + "loss": 0.0282, + "step": 514 + }, + { + "epoch": 0.24235294117647058, + "grad_norm": 0.2167178676977119, + "learning_rate": 7.725e-05, + "loss": 0.0506, + "step": 515 + }, + { + "epoch": 0.24282352941176472, + "grad_norm": 0.002202673736922909, + "learning_rate": 7.74e-05, + "loss": 0.0, + "step": 516 + }, + { + "epoch": 0.24329411764705883, + "grad_norm": 0.1958859935400213, + "learning_rate": 7.755e-05, + "loss": 0.0319, + "step": 517 + }, + { + "epoch": 0.24376470588235294, + "grad_norm": 0.15700828476174258, + "learning_rate": 7.769999999999999e-05, + "loss": 0.0376, + "step": 518 + }, + { + "epoch": 0.24423529411764705, + "grad_norm": 0.17511306571046586, + "learning_rate": 7.785e-05, + "loss": 0.0245, + "step": 519 + }, + { + "epoch": 0.2447058823529412, + "grad_norm": 0.14851190044972798, + "learning_rate": 7.8e-05, + "loss": 0.0106, + "step": 520 + }, + { + "epoch": 0.2451764705882353, + "grad_norm": 0.009828749251883259, + "learning_rate": 7.815e-05, + "loss": 0.0001, + "step": 521 + }, + { + "epoch": 0.2456470588235294, + "grad_norm": 0.22225443856857463, + "learning_rate": 7.829999999999999e-05, + "loss": 0.0223, + "step": 522 + }, + { + "epoch": 0.24611764705882352, + "grad_norm": 0.10655791171821113, + "learning_rate": 7.845e-05, + "loss": 0.0097, + "step": 523 + }, + { + "epoch": 0.24658823529411764, + "grad_norm": 0.15424995782042425, + "learning_rate": 7.86e-05, + "loss": 0.0297, + "step": 524 + }, + { + "epoch": 0.24705882352941178, + "grad_norm": 0.13330792588709633, + "learning_rate": 7.874999999999999e-05, + "loss": 0.0192, + "step": 525 + }, + { + "epoch": 0.2475294117647059, + "grad_norm": 0.399403354246287, + "learning_rate": 7.89e-05, + "loss": 0.0856, + "step": 526 + }, + { + "epoch": 0.248, + "grad_norm": 1.4015689771748856, + "learning_rate": 7.905e-05, + "loss": 0.0554, + "step": 527 + }, + { + "epoch": 0.2484705882352941, + "grad_norm": 0.3507541383561029, + "learning_rate": 7.92e-05, + "loss": 0.0593, + "step": 528 + }, + { + "epoch": 0.24894117647058825, + "grad_norm": 0.1787794614174664, + "learning_rate": 7.934999999999999e-05, + "loss": 0.0413, + "step": 529 + }, + { + "epoch": 0.24941176470588236, + "grad_norm": 0.46770151557217193, + "learning_rate": 7.95e-05, + "loss": 0.1365, + "step": 530 + }, + { + "epoch": 0.24988235294117647, + "grad_norm": 0.17675206603049062, + "learning_rate": 7.965e-05, + "loss": 0.0401, + "step": 531 + }, + { + "epoch": 0.2503529411764706, + "grad_norm": 0.10335140017418261, + "learning_rate": 7.98e-05, + "loss": 0.0164, + "step": 532 + }, + { + "epoch": 0.2508235294117647, + "grad_norm": 1.1080347506213797, + "learning_rate": 7.994999999999999e-05, + "loss": 0.0945, + "step": 533 + }, + { + "epoch": 0.25129411764705883, + "grad_norm": 0.0009235662320929226, + "learning_rate": 8.01e-05, + "loss": 0.0, + "step": 534 + }, + { + "epoch": 0.25176470588235295, + "grad_norm": 0.17544966401836462, + "learning_rate": 8.025e-05, + "loss": 0.0392, + "step": 535 + }, + { + "epoch": 0.25223529411764706, + "grad_norm": 0.21981418291867727, + "learning_rate": 8.04e-05, + "loss": 0.0321, + "step": 536 + }, + { + "epoch": 0.25270588235294117, + "grad_norm": 0.0019976171517339064, + "learning_rate": 8.054999999999999e-05, + "loss": 0.0, + "step": 537 + }, + { + "epoch": 0.2531764705882353, + "grad_norm": 0.20401438234390012, + "learning_rate": 8.07e-05, + "loss": 0.0327, + "step": 538 + }, + { + "epoch": 0.2536470588235294, + "grad_norm": 0.008100779862890142, + "learning_rate": 8.085e-05, + "loss": 0.0001, + "step": 539 + }, + { + "epoch": 0.2541176470588235, + "grad_norm": 0.1608746716489686, + "learning_rate": 8.1e-05, + "loss": 0.035, + "step": 540 + }, + { + "epoch": 0.25458823529411767, + "grad_norm": 0.06465844165451562, + "learning_rate": 8.115e-05, + "loss": 0.0034, + "step": 541 + }, + { + "epoch": 0.2550588235294118, + "grad_norm": 0.24787735717840415, + "learning_rate": 8.13e-05, + "loss": 0.0431, + "step": 542 + }, + { + "epoch": 0.2555294117647059, + "grad_norm": 0.08810597840005781, + "learning_rate": 8.145e-05, + "loss": 0.0123, + "step": 543 + }, + { + "epoch": 0.256, + "grad_norm": 0.0028509003984469104, + "learning_rate": 8.16e-05, + "loss": 0.0, + "step": 544 + }, + { + "epoch": 0.2564705882352941, + "grad_norm": 0.37878276082625306, + "learning_rate": 8.175e-05, + "loss": 0.1304, + "step": 545 + }, + { + "epoch": 0.25694117647058823, + "grad_norm": 0.18136778217939795, + "learning_rate": 8.19e-05, + "loss": 0.047, + "step": 546 + }, + { + "epoch": 0.25741176470588234, + "grad_norm": 0.00017138679103306673, + "learning_rate": 8.205e-05, + "loss": 0.0, + "step": 547 + }, + { + "epoch": 0.25788235294117645, + "grad_norm": 0.00013183079050716249, + "learning_rate": 8.22e-05, + "loss": 0.0, + "step": 548 + }, + { + "epoch": 0.25835294117647056, + "grad_norm": 0.01948575615325054, + "learning_rate": 8.235e-05, + "loss": 0.0009, + "step": 549 + }, + { + "epoch": 0.25882352941176473, + "grad_norm": 0.06578565686317037, + "learning_rate": 8.25e-05, + "loss": 0.0092, + "step": 550 + }, + { + "epoch": 0.25929411764705884, + "grad_norm": 0.14722345667389403, + "learning_rate": 8.265e-05, + "loss": 0.0212, + "step": 551 + }, + { + "epoch": 0.25976470588235295, + "grad_norm": 0.0037708778610393325, + "learning_rate": 8.28e-05, + "loss": 0.0001, + "step": 552 + }, + { + "epoch": 0.26023529411764706, + "grad_norm": 6.189115485367203e-05, + "learning_rate": 8.295e-05, + "loss": 0.0, + "step": 553 + }, + { + "epoch": 0.2607058823529412, + "grad_norm": 0.7565460097459729, + "learning_rate": 8.31e-05, + "loss": 0.1966, + "step": 554 + }, + { + "epoch": 0.2611764705882353, + "grad_norm": 0.17872109121897864, + "learning_rate": 8.325e-05, + "loss": 0.0242, + "step": 555 + }, + { + "epoch": 0.2616470588235294, + "grad_norm": 0.20358377044217457, + "learning_rate": 8.34e-05, + "loss": 0.0486, + "step": 556 + }, + { + "epoch": 0.2621176470588235, + "grad_norm": 0.1915027601658896, + "learning_rate": 8.355e-05, + "loss": 0.0337, + "step": 557 + }, + { + "epoch": 0.2625882352941176, + "grad_norm": 0.07887344183666428, + "learning_rate": 8.37e-05, + "loss": 0.0112, + "step": 558 + }, + { + "epoch": 0.2630588235294118, + "grad_norm": 0.28451954916034905, + "learning_rate": 8.385e-05, + "loss": 0.07, + "step": 559 + }, + { + "epoch": 0.2635294117647059, + "grad_norm": 0.23462990221959684, + "learning_rate": 8.4e-05, + "loss": 0.0423, + "step": 560 + }, + { + "epoch": 0.264, + "grad_norm": 0.1968205299883704, + "learning_rate": 8.415e-05, + "loss": 0.0229, + "step": 561 + }, + { + "epoch": 0.2644705882352941, + "grad_norm": 0.21018790491102204, + "learning_rate": 8.43e-05, + "loss": 0.0017, + "step": 562 + }, + { + "epoch": 0.26494117647058824, + "grad_norm": 0.011252587959197216, + "learning_rate": 8.444999999999998e-05, + "loss": 0.0001, + "step": 563 + }, + { + "epoch": 0.26541176470588235, + "grad_norm": 0.18961108749924127, + "learning_rate": 8.459999999999998e-05, + "loss": 0.0332, + "step": 564 + }, + { + "epoch": 0.26588235294117646, + "grad_norm": 0.5788958509213495, + "learning_rate": 8.474999999999999e-05, + "loss": 0.0972, + "step": 565 + }, + { + "epoch": 0.26635294117647057, + "grad_norm": 0.7264057459350098, + "learning_rate": 8.489999999999999e-05, + "loss": 0.1505, + "step": 566 + }, + { + "epoch": 0.2668235294117647, + "grad_norm": 0.3045553473400984, + "learning_rate": 8.504999999999998e-05, + "loss": 0.0981, + "step": 567 + }, + { + "epoch": 0.26729411764705885, + "grad_norm": 0.3679335278929406, + "learning_rate": 8.519999999999998e-05, + "loss": 0.0804, + "step": 568 + }, + { + "epoch": 0.26776470588235296, + "grad_norm": 0.1883424114932698, + "learning_rate": 8.534999999999999e-05, + "loss": 0.0236, + "step": 569 + }, + { + "epoch": 0.26823529411764707, + "grad_norm": 0.003010130098858239, + "learning_rate": 8.549999999999999e-05, + "loss": 0.0, + "step": 570 + }, + { + "epoch": 0.2687058823529412, + "grad_norm": 0.051792295244216004, + "learning_rate": 8.564999999999998e-05, + "loss": 0.0068, + "step": 571 + }, + { + "epoch": 0.2691764705882353, + "grad_norm": 1.7369627278897792, + "learning_rate": 8.579999999999998e-05, + "loss": 0.2609, + "step": 572 + }, + { + "epoch": 0.2696470588235294, + "grad_norm": 0.1278486867197014, + "learning_rate": 8.594999999999999e-05, + "loss": 0.0318, + "step": 573 + }, + { + "epoch": 0.2701176470588235, + "grad_norm": 0.27111412406752655, + "learning_rate": 8.609999999999999e-05, + "loss": 0.0891, + "step": 574 + }, + { + "epoch": 0.27058823529411763, + "grad_norm": 0.043327666093907764, + "learning_rate": 8.624999999999998e-05, + "loss": 0.005, + "step": 575 + }, + { + "epoch": 0.27105882352941174, + "grad_norm": 0.012993576670791531, + "learning_rate": 8.639999999999999e-05, + "loss": 0.0001, + "step": 576 + }, + { + "epoch": 0.2715294117647059, + "grad_norm": 0.3952572526626677, + "learning_rate": 8.654999999999999e-05, + "loss": 0.1128, + "step": 577 + }, + { + "epoch": 0.272, + "grad_norm": 0.48799067135297824, + "learning_rate": 8.669999999999998e-05, + "loss": 0.0595, + "step": 578 + }, + { + "epoch": 0.27247058823529413, + "grad_norm": 0.15061679408890474, + "learning_rate": 8.684999999999998e-05, + "loss": 0.0403, + "step": 579 + }, + { + "epoch": 0.27294117647058824, + "grad_norm": 0.13016133989104262, + "learning_rate": 8.699999999999999e-05, + "loss": 0.0357, + "step": 580 + }, + { + "epoch": 0.27341176470588235, + "grad_norm": 0.10732667622890883, + "learning_rate": 8.714999999999999e-05, + "loss": 0.0129, + "step": 581 + }, + { + "epoch": 0.27388235294117647, + "grad_norm": 0.2654123833998399, + "learning_rate": 8.729999999999998e-05, + "loss": 0.0333, + "step": 582 + }, + { + "epoch": 0.2743529411764706, + "grad_norm": 0.0001124368848209731, + "learning_rate": 8.744999999999998e-05, + "loss": 0.0, + "step": 583 + }, + { + "epoch": 0.2748235294117647, + "grad_norm": 0.566914305778177, + "learning_rate": 8.759999999999999e-05, + "loss": 0.0707, + "step": 584 + }, + { + "epoch": 0.2752941176470588, + "grad_norm": 0.19912315975235784, + "learning_rate": 8.774999999999999e-05, + "loss": 0.041, + "step": 585 + }, + { + "epoch": 0.27576470588235297, + "grad_norm": 0.0004853934552710639, + "learning_rate": 8.789999999999998e-05, + "loss": 0.0, + "step": 586 + }, + { + "epoch": 0.2762352941176471, + "grad_norm": 0.16005219081726652, + "learning_rate": 8.804999999999999e-05, + "loss": 0.044, + "step": 587 + }, + { + "epoch": 0.2767058823529412, + "grad_norm": 0.2154518517596966, + "learning_rate": 8.819999999999999e-05, + "loss": 0.0311, + "step": 588 + }, + { + "epoch": 0.2771764705882353, + "grad_norm": 0.3748816483782177, + "learning_rate": 8.834999999999999e-05, + "loss": 0.1053, + "step": 589 + }, + { + "epoch": 0.2776470588235294, + "grad_norm": 0.2144613190503374, + "learning_rate": 8.849999999999998e-05, + "loss": 0.0523, + "step": 590 + }, + { + "epoch": 0.2781176470588235, + "grad_norm": 0.013767722914324713, + "learning_rate": 8.864999999999999e-05, + "loss": 0.0001, + "step": 591 + }, + { + "epoch": 0.27858823529411764, + "grad_norm": 0.15486627855948787, + "learning_rate": 8.879999999999999e-05, + "loss": 0.0148, + "step": 592 + }, + { + "epoch": 0.27905882352941175, + "grad_norm": 0.25440227258149056, + "learning_rate": 8.895e-05, + "loss": 0.0373, + "step": 593 + }, + { + "epoch": 0.27952941176470586, + "grad_norm": 0.6174306825879542, + "learning_rate": 8.909999999999998e-05, + "loss": 0.1636, + "step": 594 + }, + { + "epoch": 0.28, + "grad_norm": 0.3695564671712912, + "learning_rate": 8.924999999999999e-05, + "loss": 0.1218, + "step": 595 + }, + { + "epoch": 0.28047058823529414, + "grad_norm": 0.15453340248271738, + "learning_rate": 8.939999999999999e-05, + "loss": 0.0332, + "step": 596 + }, + { + "epoch": 0.28094117647058825, + "grad_norm": 0.7516566593048928, + "learning_rate": 8.955e-05, + "loss": 0.1298, + "step": 597 + }, + { + "epoch": 0.28141176470588236, + "grad_norm": 0.14028934125314418, + "learning_rate": 8.969999999999998e-05, + "loss": 0.0307, + "step": 598 + }, + { + "epoch": 0.2818823529411765, + "grad_norm": 0.3269798548893962, + "learning_rate": 8.984999999999999e-05, + "loss": 0.1069, + "step": 599 + }, + { + "epoch": 0.2823529411764706, + "grad_norm": 0.14634746861553805, + "learning_rate": 8.999999999999999e-05, + "loss": 0.026, + "step": 600 + }, + { + "epoch": 0.2828235294117647, + "grad_norm": 0.35921027861203103, + "learning_rate": 9.014999999999998e-05, + "loss": 0.0819, + "step": 601 + }, + { + "epoch": 0.2832941176470588, + "grad_norm": 0.17668699189710416, + "learning_rate": 9.029999999999999e-05, + "loss": 0.0491, + "step": 602 + }, + { + "epoch": 0.2837647058823529, + "grad_norm": 0.09110533381636325, + "learning_rate": 9.044999999999999e-05, + "loss": 0.0178, + "step": 603 + }, + { + "epoch": 0.2842352941176471, + "grad_norm": 0.04451787620536029, + "learning_rate": 9.059999999999999e-05, + "loss": 0.0046, + "step": 604 + }, + { + "epoch": 0.2847058823529412, + "grad_norm": 0.0029141576822639278, + "learning_rate": 9.074999999999998e-05, + "loss": 0.0, + "step": 605 + }, + { + "epoch": 0.2851764705882353, + "grad_norm": 0.09211020697070675, + "learning_rate": 9.089999999999999e-05, + "loss": 0.0163, + "step": 606 + }, + { + "epoch": 0.2856470588235294, + "grad_norm": 0.11608653187696404, + "learning_rate": 9.104999999999999e-05, + "loss": 0.0195, + "step": 607 + }, + { + "epoch": 0.28611764705882353, + "grad_norm": 0.11057671961101896, + "learning_rate": 9.12e-05, + "loss": 0.014, + "step": 608 + }, + { + "epoch": 0.28658823529411764, + "grad_norm": 0.4314128749812535, + "learning_rate": 9.134999999999998e-05, + "loss": 0.0806, + "step": 609 + }, + { + "epoch": 0.28705882352941176, + "grad_norm": 0.14711380080516442, + "learning_rate": 9.149999999999999e-05, + "loss": 0.0271, + "step": 610 + }, + { + "epoch": 0.28752941176470587, + "grad_norm": 0.23172454273783608, + "learning_rate": 9.164999999999999e-05, + "loss": 0.0817, + "step": 611 + }, + { + "epoch": 0.288, + "grad_norm": 0.31586808110713954, + "learning_rate": 9.18e-05, + "loss": 0.07, + "step": 612 + }, + { + "epoch": 0.28847058823529415, + "grad_norm": 0.2476788370784004, + "learning_rate": 9.194999999999999e-05, + "loss": 0.093, + "step": 613 + }, + { + "epoch": 0.28894117647058826, + "grad_norm": 0.15786863463450881, + "learning_rate": 9.209999999999999e-05, + "loss": 0.0359, + "step": 614 + }, + { + "epoch": 0.28941176470588237, + "grad_norm": 0.3221593255117168, + "learning_rate": 9.224999999999999e-05, + "loss": 0.0689, + "step": 615 + }, + { + "epoch": 0.2898823529411765, + "grad_norm": 0.18023193155270636, + "learning_rate": 9.24e-05, + "loss": 0.0487, + "step": 616 + }, + { + "epoch": 0.2903529411764706, + "grad_norm": 0.19095273421369555, + "learning_rate": 9.254999999999999e-05, + "loss": 0.0452, + "step": 617 + }, + { + "epoch": 0.2908235294117647, + "grad_norm": 0.00017765880079392143, + "learning_rate": 9.269999999999999e-05, + "loss": 0.0, + "step": 618 + }, + { + "epoch": 0.2912941176470588, + "grad_norm": 0.27899286312128707, + "learning_rate": 9.285e-05, + "loss": 0.0915, + "step": 619 + }, + { + "epoch": 0.2917647058823529, + "grad_norm": 0.12360751731087471, + "learning_rate": 9.3e-05, + "loss": 0.0167, + "step": 620 + }, + { + "epoch": 0.29223529411764704, + "grad_norm": 0.15762157087380976, + "learning_rate": 9.314999999999999e-05, + "loss": 0.0439, + "step": 621 + }, + { + "epoch": 0.29270588235294115, + "grad_norm": 0.18128771611328576, + "learning_rate": 9.329999999999999e-05, + "loss": 0.0388, + "step": 622 + }, + { + "epoch": 0.2931764705882353, + "grad_norm": 0.1455990187005961, + "learning_rate": 9.345e-05, + "loss": 0.0193, + "step": 623 + }, + { + "epoch": 0.2936470588235294, + "grad_norm": 0.0890965796904104, + "learning_rate": 9.36e-05, + "loss": 0.0179, + "step": 624 + }, + { + "epoch": 0.29411764705882354, + "grad_norm": 0.24773620857017406, + "learning_rate": 9.374999999999999e-05, + "loss": 0.044, + "step": 625 + }, + { + "epoch": 0.29458823529411765, + "grad_norm": 0.178767940916411, + "learning_rate": 9.389999999999999e-05, + "loss": 0.047, + "step": 626 + }, + { + "epoch": 0.29505882352941176, + "grad_norm": 0.15542289625270797, + "learning_rate": 9.405e-05, + "loss": 0.0398, + "step": 627 + }, + { + "epoch": 0.2955294117647059, + "grad_norm": 0.2380103559403114, + "learning_rate": 9.419999999999999e-05, + "loss": 0.0397, + "step": 628 + }, + { + "epoch": 0.296, + "grad_norm": 6.078181958865306e-05, + "learning_rate": 9.434999999999999e-05, + "loss": 0.0, + "step": 629 + }, + { + "epoch": 0.2964705882352941, + "grad_norm": 0.13678895988363166, + "learning_rate": 9.449999999999999e-05, + "loss": 0.0252, + "step": 630 + }, + { + "epoch": 0.2969411764705882, + "grad_norm": 0.40617282785085806, + "learning_rate": 9.465e-05, + "loss": 0.1191, + "step": 631 + }, + { + "epoch": 0.2974117647058824, + "grad_norm": 3.24679608113886e-05, + "learning_rate": 9.479999999999999e-05, + "loss": 0.0, + "step": 632 + }, + { + "epoch": 0.2978823529411765, + "grad_norm": 0.13053963812477767, + "learning_rate": 9.494999999999999e-05, + "loss": 0.0116, + "step": 633 + }, + { + "epoch": 0.2983529411764706, + "grad_norm": 0.14923744819329, + "learning_rate": 9.51e-05, + "loss": 0.0276, + "step": 634 + }, + { + "epoch": 0.2988235294117647, + "grad_norm": 0.000296957583993352, + "learning_rate": 9.525e-05, + "loss": 0.0, + "step": 635 + }, + { + "epoch": 0.2992941176470588, + "grad_norm": 0.35687227350729733, + "learning_rate": 9.539999999999999e-05, + "loss": 0.086, + "step": 636 + }, + { + "epoch": 0.29976470588235293, + "grad_norm": 0.48292228755665045, + "learning_rate": 9.554999999999999e-05, + "loss": 0.174, + "step": 637 + }, + { + "epoch": 0.30023529411764704, + "grad_norm": 0.03738786716734579, + "learning_rate": 9.57e-05, + "loss": 0.0034, + "step": 638 + }, + { + "epoch": 0.30070588235294116, + "grad_norm": 0.00041122632083788956, + "learning_rate": 9.585e-05, + "loss": 0.0, + "step": 639 + }, + { + "epoch": 0.30117647058823527, + "grad_norm": 0.10934288985248147, + "learning_rate": 9.599999999999999e-05, + "loss": 0.0107, + "step": 640 + }, + { + "epoch": 0.30164705882352943, + "grad_norm": 0.2414219532078688, + "learning_rate": 9.614999999999999e-05, + "loss": 0.0538, + "step": 641 + }, + { + "epoch": 0.30211764705882355, + "grad_norm": 0.19020389363806936, + "learning_rate": 9.63e-05, + "loss": 0.0229, + "step": 642 + }, + { + "epoch": 0.30258823529411766, + "grad_norm": 0.000878858823241256, + "learning_rate": 9.645e-05, + "loss": 0.0, + "step": 643 + }, + { + "epoch": 0.30305882352941177, + "grad_norm": 2.6126528890121487e-05, + "learning_rate": 9.659999999999999e-05, + "loss": 0.0, + "step": 644 + }, + { + "epoch": 0.3035294117647059, + "grad_norm": 0.26947071765055963, + "learning_rate": 9.675e-05, + "loss": 0.0735, + "step": 645 + }, + { + "epoch": 0.304, + "grad_norm": 0.00041956194277015013, + "learning_rate": 9.69e-05, + "loss": 0.0, + "step": 646 + }, + { + "epoch": 0.3044705882352941, + "grad_norm": 0.08093339785385541, + "learning_rate": 9.705e-05, + "loss": 0.0112, + "step": 647 + }, + { + "epoch": 0.3049411764705882, + "grad_norm": 0.39023032048139816, + "learning_rate": 9.719999999999999e-05, + "loss": 0.105, + "step": 648 + }, + { + "epoch": 0.3054117647058823, + "grad_norm": 1.0391693869195415, + "learning_rate": 9.735e-05, + "loss": 0.0537, + "step": 649 + }, + { + "epoch": 0.3058823529411765, + "grad_norm": 0.00011548960164770617, + "learning_rate": 9.75e-05, + "loss": 0.0, + "step": 650 + }, + { + "epoch": 0.3063529411764706, + "grad_norm": 0.006429405683118584, + "learning_rate": 9.764999999999999e-05, + "loss": 0.0, + "step": 651 + }, + { + "epoch": 0.3068235294117647, + "grad_norm": 0.36603525159076, + "learning_rate": 9.779999999999999e-05, + "loss": 0.0771, + "step": 652 + }, + { + "epoch": 0.30729411764705883, + "grad_norm": 0.22056783599816746, + "learning_rate": 9.795e-05, + "loss": 0.0642, + "step": 653 + }, + { + "epoch": 0.30776470588235294, + "grad_norm": 0.12567045676202923, + "learning_rate": 9.81e-05, + "loss": 0.0112, + "step": 654 + }, + { + "epoch": 0.30823529411764705, + "grad_norm": 0.10912612331775874, + "learning_rate": 9.824999999999999e-05, + "loss": 0.0293, + "step": 655 + }, + { + "epoch": 0.30870588235294116, + "grad_norm": 0.2591908145593766, + "learning_rate": 9.839999999999999e-05, + "loss": 0.0842, + "step": 656 + }, + { + "epoch": 0.3091764705882353, + "grad_norm": 0.26990922762681524, + "learning_rate": 9.855e-05, + "loss": 0.0649, + "step": 657 + }, + { + "epoch": 0.3096470588235294, + "grad_norm": 0.15796517143765346, + "learning_rate": 9.87e-05, + "loss": 0.0264, + "step": 658 + }, + { + "epoch": 0.31011764705882355, + "grad_norm": 0.16404876434649898, + "learning_rate": 9.884999999999999e-05, + "loss": 0.0447, + "step": 659 + }, + { + "epoch": 0.31058823529411766, + "grad_norm": 0.19489087814514444, + "learning_rate": 9.9e-05, + "loss": 0.0333, + "step": 660 + }, + { + "epoch": 0.3110588235294118, + "grad_norm": 0.7761554060558974, + "learning_rate": 9.915e-05, + "loss": 0.2242, + "step": 661 + }, + { + "epoch": 0.3115294117647059, + "grad_norm": 0.14582074761868355, + "learning_rate": 9.93e-05, + "loss": 0.0299, + "step": 662 + }, + { + "epoch": 0.312, + "grad_norm": 2.3589978240071376e-05, + "learning_rate": 9.944999999999999e-05, + "loss": 0.0, + "step": 663 + }, + { + "epoch": 0.3124705882352941, + "grad_norm": 0.21144660622670344, + "learning_rate": 9.96e-05, + "loss": 0.0285, + "step": 664 + }, + { + "epoch": 0.3129411764705882, + "grad_norm": 0.6077553231734316, + "learning_rate": 9.975e-05, + "loss": 0.0511, + "step": 665 + }, + { + "epoch": 0.31341176470588233, + "grad_norm": 0.22794216079517718, + "learning_rate": 9.99e-05, + "loss": 0.0504, + "step": 666 + }, + { + "epoch": 0.31388235294117645, + "grad_norm": 0.5965194916967707, + "learning_rate": 0.00010004999999999999, + "loss": 0.1281, + "step": 667 + }, + { + "epoch": 0.3143529411764706, + "grad_norm": 0.3177765621974166, + "learning_rate": 0.0001002, + "loss": 0.0885, + "step": 668 + }, + { + "epoch": 0.3148235294117647, + "grad_norm": 0.22775146632289778, + "learning_rate": 0.00010035, + "loss": 0.0633, + "step": 669 + }, + { + "epoch": 0.31529411764705884, + "grad_norm": 0.12017682791115857, + "learning_rate": 0.0001005, + "loss": 0.0161, + "step": 670 + }, + { + "epoch": 0.31576470588235295, + "grad_norm": 0.18879245105988512, + "learning_rate": 0.00010065, + "loss": 0.0589, + "step": 671 + }, + { + "epoch": 0.31623529411764706, + "grad_norm": 0.14496973512746122, + "learning_rate": 0.0001008, + "loss": 0.0341, + "step": 672 + }, + { + "epoch": 0.31670588235294117, + "grad_norm": 0.5145628616343331, + "learning_rate": 0.00010095, + "loss": 0.1001, + "step": 673 + }, + { + "epoch": 0.3171764705882353, + "grad_norm": 0.174931744573718, + "learning_rate": 0.0001011, + "loss": 0.0373, + "step": 674 + }, + { + "epoch": 0.3176470588235294, + "grad_norm": 0.2180875718475856, + "learning_rate": 0.00010125, + "loss": 0.0261, + "step": 675 + }, + { + "epoch": 0.3181176470588235, + "grad_norm": 0.16618823083803166, + "learning_rate": 0.0001014, + "loss": 0.0444, + "step": 676 + }, + { + "epoch": 0.31858823529411767, + "grad_norm": 0.2978509383339829, + "learning_rate": 0.00010155, + "loss": 0.1084, + "step": 677 + }, + { + "epoch": 0.3190588235294118, + "grad_norm": 0.27750292658624426, + "learning_rate": 0.00010169999999999999, + "loss": 0.0697, + "step": 678 + }, + { + "epoch": 0.3195294117647059, + "grad_norm": 0.27740978822814, + "learning_rate": 0.00010185, + "loss": 0.0714, + "step": 679 + }, + { + "epoch": 0.32, + "grad_norm": 0.27658174920599626, + "learning_rate": 0.000102, + "loss": 0.0825, + "step": 680 + }, + { + "epoch": 0.3204705882352941, + "grad_norm": 0.013721334416525965, + "learning_rate": 0.00010215, + "loss": 0.0002, + "step": 681 + }, + { + "epoch": 0.32094117647058823, + "grad_norm": 0.12084135104179872, + "learning_rate": 0.00010229999999999999, + "loss": 0.0212, + "step": 682 + }, + { + "epoch": 0.32141176470588234, + "grad_norm": 0.7164618001209927, + "learning_rate": 0.00010245, + "loss": 0.115, + "step": 683 + }, + { + "epoch": 0.32188235294117645, + "grad_norm": 0.00010885752278305467, + "learning_rate": 0.0001026, + "loss": 0.0, + "step": 684 + }, + { + "epoch": 0.32235294117647056, + "grad_norm": 0.11420939092083371, + "learning_rate": 0.00010275, + "loss": 0.0126, + "step": 685 + }, + { + "epoch": 0.32282352941176473, + "grad_norm": 0.26234778392809754, + "learning_rate": 0.0001029, + "loss": 0.0555, + "step": 686 + }, + { + "epoch": 0.32329411764705884, + "grad_norm": 0.00012484948202162545, + "learning_rate": 0.00010305, + "loss": 0.0, + "step": 687 + }, + { + "epoch": 0.32376470588235295, + "grad_norm": 0.1834629138470622, + "learning_rate": 0.00010319999999999999, + "loss": 0.0506, + "step": 688 + }, + { + "epoch": 0.32423529411764707, + "grad_norm": 0.1856682395718354, + "learning_rate": 0.00010334999999999998, + "loss": 0.0597, + "step": 689 + }, + { + "epoch": 0.3247058823529412, + "grad_norm": 0.12435091064904057, + "learning_rate": 0.00010349999999999998, + "loss": 0.0316, + "step": 690 + }, + { + "epoch": 0.3251764705882353, + "grad_norm": 0.32261165963925736, + "learning_rate": 0.00010364999999999999, + "loss": 0.1136, + "step": 691 + }, + { + "epoch": 0.3256470588235294, + "grad_norm": 0.21694404673106615, + "learning_rate": 0.00010379999999999999, + "loss": 0.069, + "step": 692 + }, + { + "epoch": 0.3261176470588235, + "grad_norm": 0.07422303881781875, + "learning_rate": 0.00010394999999999998, + "loss": 0.0099, + "step": 693 + }, + { + "epoch": 0.3265882352941176, + "grad_norm": 0.003137409897449416, + "learning_rate": 0.00010409999999999998, + "loss": 0.0, + "step": 694 + }, + { + "epoch": 0.3270588235294118, + "grad_norm": 0.016257969224889444, + "learning_rate": 0.00010424999999999999, + "loss": 0.0001, + "step": 695 + }, + { + "epoch": 0.3275294117647059, + "grad_norm": 0.147101873576588, + "learning_rate": 0.00010439999999999999, + "loss": 0.0421, + "step": 696 + }, + { + "epoch": 0.328, + "grad_norm": 0.19836189677077581, + "learning_rate": 0.00010454999999999998, + "loss": 0.0519, + "step": 697 + }, + { + "epoch": 0.3284705882352941, + "grad_norm": 0.09860851165771622, + "learning_rate": 0.00010469999999999998, + "loss": 0.0159, + "step": 698 + }, + { + "epoch": 0.32894117647058824, + "grad_norm": 0.17072270177557472, + "learning_rate": 0.00010484999999999999, + "loss": 0.0417, + "step": 699 + }, + { + "epoch": 0.32941176470588235, + "grad_norm": 0.21891972781148913, + "learning_rate": 0.00010499999999999999, + "loss": 0.0783, + "step": 700 + }, + { + "epoch": 0.32988235294117646, + "grad_norm": 0.20806881102870803, + "learning_rate": 0.00010514999999999998, + "loss": 0.0711, + "step": 701 + }, + { + "epoch": 0.33035294117647057, + "grad_norm": 0.23628520452124233, + "learning_rate": 0.00010529999999999998, + "loss": 0.0629, + "step": 702 + }, + { + "epoch": 0.3308235294117647, + "grad_norm": 0.46263089355725956, + "learning_rate": 0.00010544999999999999, + "loss": 0.1493, + "step": 703 + }, + { + "epoch": 0.33129411764705885, + "grad_norm": 0.25931152525676543, + "learning_rate": 0.00010559999999999998, + "loss": 0.0517, + "step": 704 + }, + { + "epoch": 0.33176470588235296, + "grad_norm": 0.30266529777612344, + "learning_rate": 0.00010574999999999998, + "loss": 0.0579, + "step": 705 + }, + { + "epoch": 0.3322352941176471, + "grad_norm": 0.14370006581517497, + "learning_rate": 0.00010589999999999999, + "loss": 0.0271, + "step": 706 + }, + { + "epoch": 0.3327058823529412, + "grad_norm": 0.0746628914357129, + "learning_rate": 0.00010604999999999999, + "loss": 0.0079, + "step": 707 + }, + { + "epoch": 0.3331764705882353, + "grad_norm": 0.11104468426215068, + "learning_rate": 0.00010619999999999998, + "loss": 0.0235, + "step": 708 + }, + { + "epoch": 0.3336470588235294, + "grad_norm": 0.1866684493372899, + "learning_rate": 0.00010634999999999998, + "loss": 0.0219, + "step": 709 + }, + { + "epoch": 0.3341176470588235, + "grad_norm": 0.4357320227050833, + "learning_rate": 0.00010649999999999999, + "loss": 0.1329, + "step": 710 + }, + { + "epoch": 0.33458823529411763, + "grad_norm": 0.0006021272766637867, + "learning_rate": 0.00010664999999999999, + "loss": 0.0, + "step": 711 + }, + { + "epoch": 0.33505882352941174, + "grad_norm": 0.10634221719374608, + "learning_rate": 0.00010679999999999998, + "loss": 0.0139, + "step": 712 + }, + { + "epoch": 0.3355294117647059, + "grad_norm": 0.0977860508589045, + "learning_rate": 0.00010694999999999998, + "loss": 0.0131, + "step": 713 + }, + { + "epoch": 0.336, + "grad_norm": 0.0249546439168468, + "learning_rate": 0.00010709999999999999, + "loss": 0.0001, + "step": 714 + }, + { + "epoch": 0.33647058823529413, + "grad_norm": 0.26587211316812587, + "learning_rate": 0.00010724999999999999, + "loss": 0.0677, + "step": 715 + }, + { + "epoch": 0.33694117647058824, + "grad_norm": 0.27445499354788316, + "learning_rate": 0.00010739999999999998, + "loss": 0.0384, + "step": 716 + }, + { + "epoch": 0.33741176470588236, + "grad_norm": 0.1788182421068827, + "learning_rate": 0.00010754999999999999, + "loss": 0.0132, + "step": 717 + }, + { + "epoch": 0.33788235294117647, + "grad_norm": 0.18590782825825047, + "learning_rate": 0.00010769999999999999, + "loss": 0.0393, + "step": 718 + }, + { + "epoch": 0.3383529411764706, + "grad_norm": 0.0018457991822116051, + "learning_rate": 0.00010784999999999999, + "loss": 0.0, + "step": 719 + }, + { + "epoch": 0.3388235294117647, + "grad_norm": 0.000583334342366598, + "learning_rate": 0.00010799999999999998, + "loss": 0.0, + "step": 720 + }, + { + "epoch": 0.3392941176470588, + "grad_norm": 0.1069248511113544, + "learning_rate": 0.00010814999999999999, + "loss": 0.0158, + "step": 721 + }, + { + "epoch": 0.33976470588235297, + "grad_norm": 0.14312796218315493, + "learning_rate": 0.00010829999999999999, + "loss": 0.0201, + "step": 722 + }, + { + "epoch": 0.3402352941176471, + "grad_norm": 0.15252281622794905, + "learning_rate": 0.00010845, + "loss": 0.0214, + "step": 723 + }, + { + "epoch": 0.3407058823529412, + "grad_norm": 0.32353004265270313, + "learning_rate": 0.00010859999999999998, + "loss": 0.1054, + "step": 724 + }, + { + "epoch": 0.3411764705882353, + "grad_norm": 0.25503497973015943, + "learning_rate": 0.00010874999999999999, + "loss": 0.0957, + "step": 725 + }, + { + "epoch": 0.3416470588235294, + "grad_norm": 0.11270741082674776, + "learning_rate": 0.00010889999999999999, + "loss": 0.0209, + "step": 726 + }, + { + "epoch": 0.3421176470588235, + "grad_norm": 0.330005988662385, + "learning_rate": 0.00010904999999999998, + "loss": 0.0815, + "step": 727 + }, + { + "epoch": 0.34258823529411764, + "grad_norm": 0.19125679640818716, + "learning_rate": 0.00010919999999999998, + "loss": 0.0318, + "step": 728 + }, + { + "epoch": 0.34305882352941175, + "grad_norm": 0.15923525110732173, + "learning_rate": 0.00010934999999999999, + "loss": 0.0294, + "step": 729 + }, + { + "epoch": 0.34352941176470586, + "grad_norm": 0.03228194715837051, + "learning_rate": 0.00010949999999999999, + "loss": 0.0003, + "step": 730 + }, + { + "epoch": 0.344, + "grad_norm": 0.21741923769991386, + "learning_rate": 0.00010964999999999998, + "loss": 0.09, + "step": 731 + }, + { + "epoch": 0.34447058823529414, + "grad_norm": 0.14655299303679958, + "learning_rate": 0.00010979999999999999, + "loss": 0.0443, + "step": 732 + }, + { + "epoch": 0.34494117647058825, + "grad_norm": 0.23149529681366895, + "learning_rate": 0.00010994999999999999, + "loss": 0.0595, + "step": 733 + }, + { + "epoch": 0.34541176470588236, + "grad_norm": 0.00020078777687270282, + "learning_rate": 0.00011009999999999999, + "loss": 0.0, + "step": 734 + }, + { + "epoch": 0.3458823529411765, + "grad_norm": 0.06340150392885374, + "learning_rate": 0.00011024999999999998, + "loss": 0.0113, + "step": 735 + }, + { + "epoch": 0.3463529411764706, + "grad_norm": 0.00945239835574964, + "learning_rate": 0.00011039999999999999, + "loss": 0.0001, + "step": 736 + }, + { + "epoch": 0.3468235294117647, + "grad_norm": 4.5565332294669105e-05, + "learning_rate": 0.00011054999999999999, + "loss": 0.0, + "step": 737 + }, + { + "epoch": 0.3472941176470588, + "grad_norm": 1.707591364577702e-05, + "learning_rate": 0.0001107, + "loss": 0.0, + "step": 738 + }, + { + "epoch": 0.3477647058823529, + "grad_norm": 0.14375064171329968, + "learning_rate": 0.00011084999999999998, + "loss": 0.0293, + "step": 739 + }, + { + "epoch": 0.34823529411764703, + "grad_norm": 0.36479355578054146, + "learning_rate": 0.00011099999999999999, + "loss": 0.1347, + "step": 740 + }, + { + "epoch": 0.3487058823529412, + "grad_norm": 1.1579888860273027e-05, + "learning_rate": 0.00011114999999999999, + "loss": 0.0, + "step": 741 + }, + { + "epoch": 0.3491764705882353, + "grad_norm": 0.17813467975382852, + "learning_rate": 0.0001113, + "loss": 0.0459, + "step": 742 + }, + { + "epoch": 0.3496470588235294, + "grad_norm": 0.10250282686725308, + "learning_rate": 0.00011144999999999998, + "loss": 0.0179, + "step": 743 + }, + { + "epoch": 0.35011764705882353, + "grad_norm": 0.22949683462276257, + "learning_rate": 0.00011159999999999999, + "loss": 0.0856, + "step": 744 + }, + { + "epoch": 0.35058823529411764, + "grad_norm": 0.5402638542406405, + "learning_rate": 0.00011174999999999999, + "loss": 0.1239, + "step": 745 + }, + { + "epoch": 0.35105882352941176, + "grad_norm": 0.09521438737572492, + "learning_rate": 0.0001119, + "loss": 0.011, + "step": 746 + }, + { + "epoch": 0.35152941176470587, + "grad_norm": 0.1471630194562237, + "learning_rate": 0.00011204999999999999, + "loss": 0.0344, + "step": 747 + }, + { + "epoch": 0.352, + "grad_norm": 0.1643193288408653, + "learning_rate": 0.00011219999999999999, + "loss": 0.0304, + "step": 748 + }, + { + "epoch": 0.3524705882352941, + "grad_norm": 0.18758033623858897, + "learning_rate": 0.00011235, + "loss": 0.0391, + "step": 749 + }, + { + "epoch": 0.35294117647058826, + "grad_norm": 0.002626716253851376, + "learning_rate": 0.0001125, + "loss": 0.0, + "step": 750 + }, + { + "epoch": 0.35341176470588237, + "grad_norm": 0.8610032626856396, + "learning_rate": 0.00011264999999999999, + "loss": 0.1716, + "step": 751 + }, + { + "epoch": 0.3538823529411765, + "grad_norm": 8.619225481514665, + "learning_rate": 0.00011279999999999999, + "loss": 0.1654, + "step": 752 + }, + { + "epoch": 0.3543529411764706, + "grad_norm": 9.790293269645528e-06, + "learning_rate": 0.00011295, + "loss": 0.0, + "step": 753 + }, + { + "epoch": 0.3548235294117647, + "grad_norm": 0.7145088280427662, + "learning_rate": 0.00011309999999999998, + "loss": 0.0337, + "step": 754 + }, + { + "epoch": 0.3552941176470588, + "grad_norm": 8.678564968376989, + "learning_rate": 0.00011324999999999999, + "loss": 1.3796, + "step": 755 + }, + { + "epoch": 0.3557647058823529, + "grad_norm": 1.927780116725811e-05, + "learning_rate": 0.00011339999999999999, + "loss": 0.0, + "step": 756 + }, + { + "epoch": 0.35623529411764704, + "grad_norm": 1.288215917242588, + "learning_rate": 0.00011355, + "loss": 0.1184, + "step": 757 + }, + { + "epoch": 0.35670588235294115, + "grad_norm": 1.9077696797128225, + "learning_rate": 0.00011369999999999999, + "loss": 0.1692, + "step": 758 + }, + { + "epoch": 0.3571764705882353, + "grad_norm": 0.2731871201851395, + "learning_rate": 0.00011384999999999999, + "loss": 0.0462, + "step": 759 + }, + { + "epoch": 0.35764705882352943, + "grad_norm": 0.008841653863063572, + "learning_rate": 0.00011399999999999999, + "loss": 0.0001, + "step": 760 + }, + { + "epoch": 0.35811764705882354, + "grad_norm": 2.310193190860215, + "learning_rate": 0.00011415, + "loss": 0.4394, + "step": 761 + }, + { + "epoch": 0.35858823529411765, + "grad_norm": 0.8482355607309079, + "learning_rate": 0.00011429999999999999, + "loss": 0.14, + "step": 762 + }, + { + "epoch": 0.35905882352941176, + "grad_norm": 1.4065105889742526, + "learning_rate": 0.00011444999999999999, + "loss": 0.1983, + "step": 763 + }, + { + "epoch": 0.3595294117647059, + "grad_norm": 3.9991188844622574, + "learning_rate": 0.0001146, + "loss": 0.0406, + "step": 764 + }, + { + "epoch": 0.36, + "grad_norm": 0.4810482377949575, + "learning_rate": 0.00011475, + "loss": 0.1532, + "step": 765 + }, + { + "epoch": 0.3604705882352941, + "grad_norm": 0.267898635624307, + "learning_rate": 0.00011489999999999999, + "loss": 0.0448, + "step": 766 + }, + { + "epoch": 0.3609411764705882, + "grad_norm": 0.24426864629820252, + "learning_rate": 0.00011504999999999999, + "loss": 0.0438, + "step": 767 + }, + { + "epoch": 0.3614117647058824, + "grad_norm": 0.11501268158577765, + "learning_rate": 0.0001152, + "loss": 0.0194, + "step": 768 + }, + { + "epoch": 0.3618823529411765, + "grad_norm": 0.14498804958885103, + "learning_rate": 0.00011535, + "loss": 0.0151, + "step": 769 + }, + { + "epoch": 0.3623529411764706, + "grad_norm": 0.0012194007340571582, + "learning_rate": 0.00011549999999999999, + "loss": 0.0, + "step": 770 + }, + { + "epoch": 0.3628235294117647, + "grad_norm": 0.979681227992404, + "learning_rate": 0.00011564999999999999, + "loss": 0.1693, + "step": 771 + }, + { + "epoch": 0.3632941176470588, + "grad_norm": 0.3786803514693947, + "learning_rate": 0.0001158, + "loss": 0.0712, + "step": 772 + }, + { + "epoch": 0.36376470588235293, + "grad_norm": 0.35762968611420876, + "learning_rate": 0.00011595, + "loss": 0.1151, + "step": 773 + }, + { + "epoch": 0.36423529411764705, + "grad_norm": 0.13719260756482876, + "learning_rate": 0.00011609999999999999, + "loss": 0.0511, + "step": 774 + }, + { + "epoch": 0.36470588235294116, + "grad_norm": 0.23020289066467686, + "learning_rate": 0.00011624999999999999, + "loss": 0.0951, + "step": 775 + }, + { + "epoch": 0.36517647058823527, + "grad_norm": 0.13030320602313675, + "learning_rate": 0.0001164, + "loss": 0.0373, + "step": 776 + }, + { + "epoch": 0.36564705882352944, + "grad_norm": 0.3039175275927562, + "learning_rate": 0.00011654999999999999, + "loss": 0.1062, + "step": 777 + }, + { + "epoch": 0.36611764705882355, + "grad_norm": 0.022613581111934157, + "learning_rate": 0.00011669999999999999, + "loss": 0.0017, + "step": 778 + }, + { + "epoch": 0.36658823529411766, + "grad_norm": 0.1946188469195553, + "learning_rate": 0.00011685, + "loss": 0.0569, + "step": 779 + }, + { + "epoch": 0.36705882352941177, + "grad_norm": 0.19639630283317683, + "learning_rate": 0.000117, + "loss": 0.0452, + "step": 780 + }, + { + "epoch": 0.3675294117647059, + "grad_norm": 0.0013078546739187223, + "learning_rate": 0.00011714999999999999, + "loss": 0.0, + "step": 781 + }, + { + "epoch": 0.368, + "grad_norm": 0.20414129025225367, + "learning_rate": 0.00011729999999999999, + "loss": 0.0494, + "step": 782 + }, + { + "epoch": 0.3684705882352941, + "grad_norm": 0.1345090181855797, + "learning_rate": 0.00011745, + "loss": 0.0289, + "step": 783 + }, + { + "epoch": 0.3689411764705882, + "grad_norm": 0.24774001451120028, + "learning_rate": 0.0001176, + "loss": 0.0821, + "step": 784 + }, + { + "epoch": 0.36941176470588233, + "grad_norm": 0.002169989563055113, + "learning_rate": 0.00011774999999999999, + "loss": 0.0, + "step": 785 + }, + { + "epoch": 0.3698823529411765, + "grad_norm": 0.0011342556540466138, + "learning_rate": 0.00011789999999999999, + "loss": 0.0, + "step": 786 + }, + { + "epoch": 0.3703529411764706, + "grad_norm": 0.36249804132805835, + "learning_rate": 0.00011805, + "loss": 0.1308, + "step": 787 + }, + { + "epoch": 0.3708235294117647, + "grad_norm": 0.48867737946728584, + "learning_rate": 0.0001182, + "loss": 0.2193, + "step": 788 + }, + { + "epoch": 0.37129411764705883, + "grad_norm": 0.08952249170447715, + "learning_rate": 0.00011834999999999999, + "loss": 0.0212, + "step": 789 + }, + { + "epoch": 0.37176470588235294, + "grad_norm": 0.23339809264707173, + "learning_rate": 0.0001185, + "loss": 0.0476, + "step": 790 + }, + { + "epoch": 0.37223529411764705, + "grad_norm": 0.2727800448077631, + "learning_rate": 0.00011865, + "loss": 0.0928, + "step": 791 + }, + { + "epoch": 0.37270588235294116, + "grad_norm": 0.1765688797839621, + "learning_rate": 0.0001188, + "loss": 0.056, + "step": 792 + }, + { + "epoch": 0.3731764705882353, + "grad_norm": 0.002000626533727875, + "learning_rate": 0.00011894999999999999, + "loss": 0.0, + "step": 793 + }, + { + "epoch": 0.3736470588235294, + "grad_norm": 0.3058784075464332, + "learning_rate": 0.0001191, + "loss": 0.1754, + "step": 794 + }, + { + "epoch": 0.37411764705882355, + "grad_norm": 0.17711848207534864, + "learning_rate": 0.00011925, + "loss": 0.0415, + "step": 795 + }, + { + "epoch": 0.37458823529411767, + "grad_norm": 0.001959396005488369, + "learning_rate": 0.0001194, + "loss": 0.0, + "step": 796 + }, + { + "epoch": 0.3750588235294118, + "grad_norm": 0.2328294932959992, + "learning_rate": 0.00011954999999999999, + "loss": 0.0759, + "step": 797 + }, + { + "epoch": 0.3755294117647059, + "grad_norm": 0.001336067588935342, + "learning_rate": 0.0001197, + "loss": 0.0, + "step": 798 + }, + { + "epoch": 0.376, + "grad_norm": 0.15352088611345907, + "learning_rate": 0.00011985, + "loss": 0.0484, + "step": 799 + }, + { + "epoch": 0.3764705882352941, + "grad_norm": 0.4039214667373786, + "learning_rate": 0.00011999999999999999, + "loss": 0.1843, + "step": 800 + }, + { + "epoch": 0.3769411764705882, + "grad_norm": 0.14335162779105817, + "learning_rate": 0.00012014999999999999, + "loss": 0.0373, + "step": 801 + }, + { + "epoch": 0.37741176470588234, + "grad_norm": 0.09540398605095436, + "learning_rate": 0.0001203, + "loss": 0.0202, + "step": 802 + }, + { + "epoch": 0.37788235294117645, + "grad_norm": 0.06514696319763294, + "learning_rate": 0.00012045, + "loss": 0.0139, + "step": 803 + }, + { + "epoch": 0.3783529411764706, + "grad_norm": 0.17091787876743147, + "learning_rate": 0.00012059999999999999, + "loss": 0.0509, + "step": 804 + }, + { + "epoch": 0.3788235294117647, + "grad_norm": 0.0018087674556099356, + "learning_rate": 0.00012075, + "loss": 0.0, + "step": 805 + }, + { + "epoch": 0.37929411764705884, + "grad_norm": 0.26024012172655997, + "learning_rate": 0.0001209, + "loss": 0.0758, + "step": 806 + }, + { + "epoch": 0.37976470588235295, + "grad_norm": 0.13098859140975527, + "learning_rate": 0.00012105, + "loss": 0.042, + "step": 807 + }, + { + "epoch": 0.38023529411764706, + "grad_norm": 0.12534878045519293, + "learning_rate": 0.00012119999999999999, + "loss": 0.0382, + "step": 808 + }, + { + "epoch": 0.38070588235294117, + "grad_norm": 0.0027197164317621213, + "learning_rate": 0.00012135, + "loss": 0.0, + "step": 809 + }, + { + "epoch": 0.3811764705882353, + "grad_norm": 0.001189039091311747, + "learning_rate": 0.0001215, + "loss": 0.0, + "step": 810 + }, + { + "epoch": 0.3816470588235294, + "grad_norm": 0.07707406782094225, + "learning_rate": 0.00012165, + "loss": 0.0181, + "step": 811 + }, + { + "epoch": 0.3821176470588235, + "grad_norm": 0.2662219619379841, + "learning_rate": 0.00012179999999999999, + "loss": 0.049, + "step": 812 + }, + { + "epoch": 0.3825882352941177, + "grad_norm": 0.15905420782431412, + "learning_rate": 0.00012194999999999998, + "loss": 0.0365, + "step": 813 + }, + { + "epoch": 0.3830588235294118, + "grad_norm": 0.09305154260774567, + "learning_rate": 0.00012209999999999999, + "loss": 0.0198, + "step": 814 + }, + { + "epoch": 0.3835294117647059, + "grad_norm": 0.000987793121270184, + "learning_rate": 0.00012225, + "loss": 0.0, + "step": 815 + }, + { + "epoch": 0.384, + "grad_norm": 0.1323755961734063, + "learning_rate": 0.0001224, + "loss": 0.0396, + "step": 816 + }, + { + "epoch": 0.3844705882352941, + "grad_norm": 0.0008688674828032656, + "learning_rate": 0.00012254999999999997, + "loss": 0.0, + "step": 817 + }, + { + "epoch": 0.38494117647058823, + "grad_norm": 0.15756799562525275, + "learning_rate": 0.00012269999999999997, + "loss": 0.0696, + "step": 818 + }, + { + "epoch": 0.38541176470588234, + "grad_norm": 0.18963819867214562, + "learning_rate": 0.00012284999999999998, + "loss": 0.073, + "step": 819 + }, + { + "epoch": 0.38588235294117645, + "grad_norm": 0.0009315822995929629, + "learning_rate": 0.00012299999999999998, + "loss": 0.0, + "step": 820 + }, + { + "epoch": 0.38635294117647057, + "grad_norm": 0.1446241451167008, + "learning_rate": 0.00012314999999999998, + "loss": 0.0477, + "step": 821 + }, + { + "epoch": 0.38682352941176473, + "grad_norm": 0.11318745952379487, + "learning_rate": 0.0001233, + "loss": 0.0445, + "step": 822 + }, + { + "epoch": 0.38729411764705884, + "grad_norm": 0.0009400579744799482, + "learning_rate": 0.00012345, + "loss": 0.0, + "step": 823 + }, + { + "epoch": 0.38776470588235296, + "grad_norm": 0.3443648377938207, + "learning_rate": 0.0001236, + "loss": 0.1384, + "step": 824 + }, + { + "epoch": 0.38823529411764707, + "grad_norm": 0.0009325264314644436, + "learning_rate": 0.00012374999999999997, + "loss": 0.0, + "step": 825 + }, + { + "epoch": 0.3887058823529412, + "grad_norm": 0.09946243851799845, + "learning_rate": 0.00012389999999999998, + "loss": 0.0378, + "step": 826 + }, + { + "epoch": 0.3891764705882353, + "grad_norm": 0.11184160195875231, + "learning_rate": 0.00012404999999999998, + "loss": 0.0323, + "step": 827 + }, + { + "epoch": 0.3896470588235294, + "grad_norm": 0.24343317926143024, + "learning_rate": 0.00012419999999999998, + "loss": 0.1145, + "step": 828 + }, + { + "epoch": 0.3901176470588235, + "grad_norm": 0.14229897572652614, + "learning_rate": 0.00012435, + "loss": 0.0548, + "step": 829 + }, + { + "epoch": 0.3905882352941176, + "grad_norm": 0.23561164660317166, + "learning_rate": 0.0001245, + "loss": 0.0964, + "step": 830 + }, + { + "epoch": 0.3910588235294118, + "grad_norm": 0.18449147938416452, + "learning_rate": 0.00012465, + "loss": 0.0486, + "step": 831 + }, + { + "epoch": 0.3915294117647059, + "grad_norm": 0.0006067080415679645, + "learning_rate": 0.00012479999999999997, + "loss": 0.0, + "step": 832 + }, + { + "epoch": 0.392, + "grad_norm": 0.0005734812706658427, + "learning_rate": 0.00012494999999999997, + "loss": 0.0, + "step": 833 + }, + { + "epoch": 0.3924705882352941, + "grad_norm": 0.1852997919717565, + "learning_rate": 0.00012509999999999998, + "loss": 0.0868, + "step": 834 + }, + { + "epoch": 0.39294117647058824, + "grad_norm": 0.15435452041171924, + "learning_rate": 0.00012524999999999998, + "loss": 0.0441, + "step": 835 + }, + { + "epoch": 0.39341176470588235, + "grad_norm": 0.006935549602994425, + "learning_rate": 0.00012539999999999999, + "loss": 0.0, + "step": 836 + }, + { + "epoch": 0.39388235294117646, + "grad_norm": 0.03875374354587596, + "learning_rate": 0.00012555, + "loss": 0.0025, + "step": 837 + }, + { + "epoch": 0.3943529411764706, + "grad_norm": 0.403634053516392, + "learning_rate": 0.0001257, + "loss": 0.0764, + "step": 838 + }, + { + "epoch": 0.3948235294117647, + "grad_norm": 0.13255257882870314, + "learning_rate": 0.00012585, + "loss": 0.0414, + "step": 839 + }, + { + "epoch": 0.3952941176470588, + "grad_norm": 0.9008696662805523, + "learning_rate": 0.00012599999999999997, + "loss": 0.3239, + "step": 840 + }, + { + "epoch": 0.39576470588235296, + "grad_norm": 0.17791624262438763, + "learning_rate": 0.00012614999999999998, + "loss": 0.0226, + "step": 841 + }, + { + "epoch": 0.3962352941176471, + "grad_norm": 0.15437365038006415, + "learning_rate": 0.00012629999999999998, + "loss": 0.0524, + "step": 842 + }, + { + "epoch": 0.3967058823529412, + "grad_norm": 0.25161130554523736, + "learning_rate": 0.00012644999999999998, + "loss": 0.0771, + "step": 843 + }, + { + "epoch": 0.3971764705882353, + "grad_norm": 0.18397555109075658, + "learning_rate": 0.0001266, + "loss": 0.0359, + "step": 844 + }, + { + "epoch": 0.3976470588235294, + "grad_norm": 0.2305406894592903, + "learning_rate": 0.00012675, + "loss": 0.0722, + "step": 845 + }, + { + "epoch": 0.3981176470588235, + "grad_norm": 0.000800644328789443, + "learning_rate": 0.0001269, + "loss": 0.0, + "step": 846 + }, + { + "epoch": 0.39858823529411763, + "grad_norm": 0.05347624660809854, + "learning_rate": 0.00012705, + "loss": 0.0036, + "step": 847 + }, + { + "epoch": 0.39905882352941174, + "grad_norm": 0.13458589134495633, + "learning_rate": 0.00012719999999999997, + "loss": 0.0062, + "step": 848 + }, + { + "epoch": 0.39952941176470586, + "grad_norm": 0.10323638278681282, + "learning_rate": 0.00012734999999999998, + "loss": 0.0077, + "step": 849 + }, + { + "epoch": 0.4, + "grad_norm": 0.25254273395917515, + "learning_rate": 0.00012749999999999998, + "loss": 0.0767, + "step": 850 + }, + { + "epoch": 0.40047058823529413, + "grad_norm": 0.2600685176438858, + "learning_rate": 0.00012764999999999999, + "loss": 0.0671, + "step": 851 + }, + { + "epoch": 0.40094117647058825, + "grad_norm": 0.29505181295104116, + "learning_rate": 0.0001278, + "loss": 0.1104, + "step": 852 + }, + { + "epoch": 0.40141176470588236, + "grad_norm": 0.07620652283399855, + "learning_rate": 0.00012795, + "loss": 0.0216, + "step": 853 + }, + { + "epoch": 0.40188235294117647, + "grad_norm": 0.08081113445418302, + "learning_rate": 0.0001281, + "loss": 0.02, + "step": 854 + }, + { + "epoch": 0.4023529411764706, + "grad_norm": 0.16168303367237502, + "learning_rate": 0.00012824999999999997, + "loss": 0.0441, + "step": 855 + }, + { + "epoch": 0.4028235294117647, + "grad_norm": 0.001633630194174716, + "learning_rate": 0.00012839999999999998, + "loss": 0.0, + "step": 856 + }, + { + "epoch": 0.4032941176470588, + "grad_norm": 1.1535423945071916, + "learning_rate": 0.00012854999999999998, + "loss": 0.1725, + "step": 857 + }, + { + "epoch": 0.4037647058823529, + "grad_norm": 0.4913249861014143, + "learning_rate": 0.00012869999999999998, + "loss": 0.1696, + "step": 858 + }, + { + "epoch": 0.4042352941176471, + "grad_norm": 0.26163171234432947, + "learning_rate": 0.00012885, + "loss": 0.1234, + "step": 859 + }, + { + "epoch": 0.4047058823529412, + "grad_norm": 0.29514923383833086, + "learning_rate": 0.000129, + "loss": 0.1007, + "step": 860 + }, + { + "epoch": 0.4051764705882353, + "grad_norm": 0.028171747947588387, + "learning_rate": 0.00012915, + "loss": 0.001, + "step": 861 + }, + { + "epoch": 0.4056470588235294, + "grad_norm": 0.09893818713839897, + "learning_rate": 0.0001293, + "loss": 0.034, + "step": 862 + }, + { + "epoch": 0.4061176470588235, + "grad_norm": 0.141359366832634, + "learning_rate": 0.00012944999999999998, + "loss": 0.0464, + "step": 863 + }, + { + "epoch": 0.40658823529411764, + "grad_norm": 0.3367080830118533, + "learning_rate": 0.00012959999999999998, + "loss": 0.1015, + "step": 864 + }, + { + "epoch": 0.40705882352941175, + "grad_norm": 0.0045732808694067964, + "learning_rate": 0.00012974999999999998, + "loss": 0.0001, + "step": 865 + }, + { + "epoch": 0.40752941176470586, + "grad_norm": 0.08768718394382789, + "learning_rate": 0.00012989999999999999, + "loss": 0.0307, + "step": 866 + }, + { + "epoch": 0.408, + "grad_norm": 0.02678919027993363, + "learning_rate": 0.00013005, + "loss": 0.0028, + "step": 867 + }, + { + "epoch": 0.40847058823529414, + "grad_norm": 0.09338194769790692, + "learning_rate": 0.0001302, + "loss": 0.0331, + "step": 868 + }, + { + "epoch": 0.40894117647058825, + "grad_norm": 0.05965468699899704, + "learning_rate": 0.00013035, + "loss": 0.0104, + "step": 869 + }, + { + "epoch": 0.40941176470588236, + "grad_norm": 0.08204406742243245, + "learning_rate": 0.0001305, + "loss": 0.0216, + "step": 870 + }, + { + "epoch": 0.4098823529411765, + "grad_norm": 0.4248992284091426, + "learning_rate": 0.00013064999999999998, + "loss": 0.2463, + "step": 871 + }, + { + "epoch": 0.4103529411764706, + "grad_norm": 0.1019933269010442, + "learning_rate": 0.00013079999999999998, + "loss": 0.0329, + "step": 872 + }, + { + "epoch": 0.4108235294117647, + "grad_norm": 0.1437865277983647, + "learning_rate": 0.00013094999999999998, + "loss": 0.0573, + "step": 873 + }, + { + "epoch": 0.4112941176470588, + "grad_norm": 0.1210868359846615, + "learning_rate": 0.0001311, + "loss": 0.0361, + "step": 874 + }, + { + "epoch": 0.4117647058823529, + "grad_norm": 0.05851716313924378, + "learning_rate": 0.00013125, + "loss": 0.0099, + "step": 875 + }, + { + "epoch": 0.41223529411764703, + "grad_norm": 0.08359279887028188, + "learning_rate": 0.0001314, + "loss": 0.0132, + "step": 876 + }, + { + "epoch": 0.4127058823529412, + "grad_norm": 0.17145052530024943, + "learning_rate": 0.00013155, + "loss": 0.0448, + "step": 877 + }, + { + "epoch": 0.4131764705882353, + "grad_norm": 0.16332262467965264, + "learning_rate": 0.00013169999999999998, + "loss": 0.0385, + "step": 878 + }, + { + "epoch": 0.4136470588235294, + "grad_norm": 0.4575325632219064, + "learning_rate": 0.00013184999999999998, + "loss": 0.1339, + "step": 879 + }, + { + "epoch": 0.41411764705882353, + "grad_norm": 0.051742528416582306, + "learning_rate": 0.00013199999999999998, + "loss": 0.0073, + "step": 880 + }, + { + "epoch": 0.41458823529411765, + "grad_norm": 0.11254399840525751, + "learning_rate": 0.00013215, + "loss": 0.0376, + "step": 881 + }, + { + "epoch": 0.41505882352941176, + "grad_norm": 0.2254543351011952, + "learning_rate": 0.0001323, + "loss": 0.1201, + "step": 882 + }, + { + "epoch": 0.41552941176470587, + "grad_norm": 0.1948785634086715, + "learning_rate": 0.00013245, + "loss": 0.0811, + "step": 883 + }, + { + "epoch": 0.416, + "grad_norm": 0.123886163347168, + "learning_rate": 0.0001326, + "loss": 0.0419, + "step": 884 + }, + { + "epoch": 0.4164705882352941, + "grad_norm": 0.19811516190676082, + "learning_rate": 0.00013275, + "loss": 0.0777, + "step": 885 + }, + { + "epoch": 0.41694117647058826, + "grad_norm": 0.0005988286108253943, + "learning_rate": 0.00013289999999999998, + "loss": 0.0, + "step": 886 + }, + { + "epoch": 0.41741176470588237, + "grad_norm": 0.2667575417388602, + "learning_rate": 0.00013304999999999998, + "loss": 0.0885, + "step": 887 + }, + { + "epoch": 0.4178823529411765, + "grad_norm": 0.00047041575563252613, + "learning_rate": 0.00013319999999999999, + "loss": 0.0, + "step": 888 + }, + { + "epoch": 0.4183529411764706, + "grad_norm": 0.11674440632781087, + "learning_rate": 0.00013335, + "loss": 0.0427, + "step": 889 + }, + { + "epoch": 0.4188235294117647, + "grad_norm": 0.12143362336925338, + "learning_rate": 0.0001335, + "loss": 0.0429, + "step": 890 + }, + { + "epoch": 0.4192941176470588, + "grad_norm": 0.056190223107676594, + "learning_rate": 0.00013365, + "loss": 0.0173, + "step": 891 + }, + { + "epoch": 0.41976470588235293, + "grad_norm": 0.15917056655402248, + "learning_rate": 0.0001338, + "loss": 0.061, + "step": 892 + }, + { + "epoch": 0.42023529411764704, + "grad_norm": 0.11319085480914544, + "learning_rate": 0.00013395, + "loss": 0.0426, + "step": 893 + }, + { + "epoch": 0.42070588235294115, + "grad_norm": 0.1089364240943215, + "learning_rate": 0.00013409999999999998, + "loss": 0.0139, + "step": 894 + }, + { + "epoch": 0.4211764705882353, + "grad_norm": 0.4529491003439922, + "learning_rate": 0.00013424999999999998, + "loss": 0.0787, + "step": 895 + }, + { + "epoch": 0.42164705882352943, + "grad_norm": 0.30921122220516956, + "learning_rate": 0.0001344, + "loss": 0.0848, + "step": 896 + }, + { + "epoch": 0.42211764705882354, + "grad_norm": 0.15765389743740693, + "learning_rate": 0.00013455, + "loss": 0.0978, + "step": 897 + }, + { + "epoch": 0.42258823529411765, + "grad_norm": 0.1470455114013861, + "learning_rate": 0.0001347, + "loss": 0.0553, + "step": 898 + }, + { + "epoch": 0.42305882352941176, + "grad_norm": 0.21544786783588696, + "learning_rate": 0.00013485, + "loss": 0.0822, + "step": 899 + }, + { + "epoch": 0.4235294117647059, + "grad_norm": 0.09879559595219177, + "learning_rate": 0.000135, + "loss": 0.0261, + "step": 900 + }, + { + "epoch": 0.424, + "grad_norm": 0.0557237749123759, + "learning_rate": 0.00013514999999999998, + "loss": 0.0153, + "step": 901 + }, + { + "epoch": 0.4244705882352941, + "grad_norm": 0.17003475400677476, + "learning_rate": 0.00013529999999999998, + "loss": 0.078, + "step": 902 + }, + { + "epoch": 0.4249411764705882, + "grad_norm": 0.06275419232124853, + "learning_rate": 0.00013544999999999999, + "loss": 0.0169, + "step": 903 + }, + { + "epoch": 0.4254117647058824, + "grad_norm": 0.20318835420443004, + "learning_rate": 0.0001356, + "loss": 0.0496, + "step": 904 + }, + { + "epoch": 0.4258823529411765, + "grad_norm": 0.1237803466449436, + "learning_rate": 0.00013575, + "loss": 0.0476, + "step": 905 + }, + { + "epoch": 0.4263529411764706, + "grad_norm": 0.15557610123275334, + "learning_rate": 0.0001359, + "loss": 0.0663, + "step": 906 + }, + { + "epoch": 0.4268235294117647, + "grad_norm": 0.10220787967553188, + "learning_rate": 0.00013605, + "loss": 0.0295, + "step": 907 + }, + { + "epoch": 0.4272941176470588, + "grad_norm": 0.11758996851965259, + "learning_rate": 0.0001362, + "loss": 0.0435, + "step": 908 + }, + { + "epoch": 0.42776470588235294, + "grad_norm": 0.10773483184326732, + "learning_rate": 0.00013634999999999998, + "loss": 0.0348, + "step": 909 + }, + { + "epoch": 0.42823529411764705, + "grad_norm": 0.00019608183506781044, + "learning_rate": 0.00013649999999999998, + "loss": 0.0, + "step": 910 + }, + { + "epoch": 0.42870588235294116, + "grad_norm": 0.3619287803814491, + "learning_rate": 0.00013665, + "loss": 0.1107, + "step": 911 + }, + { + "epoch": 0.42917647058823527, + "grad_norm": 0.1165063072606651, + "learning_rate": 0.0001368, + "loss": 0.0475, + "step": 912 + }, + { + "epoch": 0.42964705882352944, + "grad_norm": 0.041033390086671354, + "learning_rate": 0.00013695, + "loss": 0.0055, + "step": 913 + }, + { + "epoch": 0.43011764705882355, + "grad_norm": 0.07840916483930467, + "learning_rate": 0.0001371, + "loss": 0.015, + "step": 914 + }, + { + "epoch": 0.43058823529411766, + "grad_norm": 0.12524048319397826, + "learning_rate": 0.00013725, + "loss": 0.0502, + "step": 915 + }, + { + "epoch": 0.43105882352941177, + "grad_norm": 0.2316609481540973, + "learning_rate": 0.0001374, + "loss": 0.1134, + "step": 916 + }, + { + "epoch": 0.4315294117647059, + "grad_norm": 0.25153832946515026, + "learning_rate": 0.00013754999999999998, + "loss": 0.1469, + "step": 917 + }, + { + "epoch": 0.432, + "grad_norm": 0.3184498604178992, + "learning_rate": 0.00013769999999999999, + "loss": 0.1133, + "step": 918 + }, + { + "epoch": 0.4324705882352941, + "grad_norm": 0.08411772808272493, + "learning_rate": 0.00013785, + "loss": 0.0308, + "step": 919 + }, + { + "epoch": 0.4329411764705882, + "grad_norm": 0.16222167261595846, + "learning_rate": 0.000138, + "loss": 0.0968, + "step": 920 + }, + { + "epoch": 0.43341176470588233, + "grad_norm": 0.10957385300749718, + "learning_rate": 0.00013815, + "loss": 0.0338, + "step": 921 + }, + { + "epoch": 0.4338823529411765, + "grad_norm": 0.00011817218775184991, + "learning_rate": 0.0001383, + "loss": 0.0, + "step": 922 + }, + { + "epoch": 0.4343529411764706, + "grad_norm": 0.20153023386986574, + "learning_rate": 0.00013845, + "loss": 0.087, + "step": 923 + }, + { + "epoch": 0.4348235294117647, + "grad_norm": 0.2118680317666335, + "learning_rate": 0.0001386, + "loss": 0.0903, + "step": 924 + }, + { + "epoch": 0.43529411764705883, + "grad_norm": 0.14117260928584774, + "learning_rate": 0.00013874999999999998, + "loss": 0.0619, + "step": 925 + }, + { + "epoch": 0.43576470588235294, + "grad_norm": 0.13167913484466728, + "learning_rate": 0.0001389, + "loss": 0.0798, + "step": 926 + }, + { + "epoch": 0.43623529411764705, + "grad_norm": 0.13927874309053806, + "learning_rate": 0.00013905, + "loss": 0.0825, + "step": 927 + }, + { + "epoch": 0.43670588235294117, + "grad_norm": 0.17855573279759526, + "learning_rate": 0.0001392, + "loss": 0.1043, + "step": 928 + }, + { + "epoch": 0.4371764705882353, + "grad_norm": 0.08000862334433242, + "learning_rate": 0.00013935, + "loss": 0.0266, + "step": 929 + }, + { + "epoch": 0.4376470588235294, + "grad_norm": 0.1348204977777772, + "learning_rate": 0.0001395, + "loss": 0.0254, + "step": 930 + }, + { + "epoch": 0.43811764705882356, + "grad_norm": 0.00011194447537777837, + "learning_rate": 0.00013965, + "loss": 0.0, + "step": 931 + }, + { + "epoch": 0.43858823529411767, + "grad_norm": 0.2853405735132526, + "learning_rate": 0.00013979999999999998, + "loss": 0.1119, + "step": 932 + }, + { + "epoch": 0.4390588235294118, + "grad_norm": 0.1680694571442778, + "learning_rate": 0.00013995, + "loss": 0.0545, + "step": 933 + }, + { + "epoch": 0.4395294117647059, + "grad_norm": 7.10471615677838e-05, + "learning_rate": 0.0001401, + "loss": 0.0, + "step": 934 + }, + { + "epoch": 0.44, + "grad_norm": 0.09626920957790071, + "learning_rate": 0.00014025, + "loss": 0.04, + "step": 935 + }, + { + "epoch": 0.4404705882352941, + "grad_norm": 0.2584114346745528, + "learning_rate": 0.0001404, + "loss": 0.1057, + "step": 936 + }, + { + "epoch": 0.4409411764705882, + "grad_norm": 0.14931599956017771, + "learning_rate": 0.00014055, + "loss": 0.0407, + "step": 937 + }, + { + "epoch": 0.44141176470588234, + "grad_norm": 0.00030352002904889984, + "learning_rate": 0.00014069999999999998, + "loss": 0.0, + "step": 938 + }, + { + "epoch": 0.44188235294117645, + "grad_norm": 0.17416766134554768, + "learning_rate": 0.00014084999999999998, + "loss": 0.0778, + "step": 939 + }, + { + "epoch": 0.4423529411764706, + "grad_norm": 0.13081046801796387, + "learning_rate": 0.00014099999999999998, + "loss": 0.0346, + "step": 940 + }, + { + "epoch": 0.4428235294117647, + "grad_norm": 0.08931512297188197, + "learning_rate": 0.00014115, + "loss": 0.0326, + "step": 941 + }, + { + "epoch": 0.44329411764705884, + "grad_norm": 0.10599201823244656, + "learning_rate": 0.0001413, + "loss": 0.0378, + "step": 942 + }, + { + "epoch": 0.44376470588235295, + "grad_norm": 0.06750652098593231, + "learning_rate": 0.00014144999999999997, + "loss": 0.0261, + "step": 943 + }, + { + "epoch": 0.44423529411764706, + "grad_norm": 0.00017202688555590184, + "learning_rate": 0.00014159999999999997, + "loss": 0.0, + "step": 944 + }, + { + "epoch": 0.4447058823529412, + "grad_norm": 0.049878339049676056, + "learning_rate": 0.00014174999999999998, + "loss": 0.0091, + "step": 945 + }, + { + "epoch": 0.4451764705882353, + "grad_norm": 0.41579028541497665, + "learning_rate": 0.00014189999999999998, + "loss": 0.1374, + "step": 946 + }, + { + "epoch": 0.4456470588235294, + "grad_norm": 0.38290153297790885, + "learning_rate": 0.00014204999999999998, + "loss": 0.2236, + "step": 947 + }, + { + "epoch": 0.4461176470588235, + "grad_norm": 0.1964521848393474, + "learning_rate": 0.0001422, + "loss": 0.0842, + "step": 948 + }, + { + "epoch": 0.4465882352941177, + "grad_norm": 0.1661526276450583, + "learning_rate": 0.00014235, + "loss": 0.1004, + "step": 949 + }, + { + "epoch": 0.4470588235294118, + "grad_norm": 0.09652790169264189, + "learning_rate": 0.0001425, + "loss": 0.0348, + "step": 950 + }, + { + "epoch": 0.4475294117647059, + "grad_norm": 0.215336040366408, + "learning_rate": 0.00014264999999999997, + "loss": 0.1141, + "step": 951 + }, + { + "epoch": 0.448, + "grad_norm": 0.214440384062931, + "learning_rate": 0.00014279999999999997, + "loss": 0.1138, + "step": 952 + }, + { + "epoch": 0.4484705882352941, + "grad_norm": 0.2760328422944764, + "learning_rate": 0.00014294999999999998, + "loss": 0.0904, + "step": 953 + }, + { + "epoch": 0.44894117647058823, + "grad_norm": 0.11261111655918363, + "learning_rate": 0.00014309999999999998, + "loss": 0.0398, + "step": 954 + }, + { + "epoch": 0.44941176470588234, + "grad_norm": 0.20384502728276155, + "learning_rate": 0.00014324999999999999, + "loss": 0.0711, + "step": 955 + }, + { + "epoch": 0.44988235294117646, + "grad_norm": 0.7611124731459109, + "learning_rate": 0.0001434, + "loss": 0.1244, + "step": 956 + }, + { + "epoch": 0.45035294117647057, + "grad_norm": 0.16755951065269106, + "learning_rate": 0.00014355, + "loss": 0.0806, + "step": 957 + }, + { + "epoch": 0.4508235294117647, + "grad_norm": 0.20422244961550642, + "learning_rate": 0.00014369999999999997, + "loss": 0.0691, + "step": 958 + }, + { + "epoch": 0.45129411764705885, + "grad_norm": 0.11994129413779082, + "learning_rate": 0.00014384999999999997, + "loss": 0.0191, + "step": 959 + }, + { + "epoch": 0.45176470588235296, + "grad_norm": 0.0002853110340592059, + "learning_rate": 0.00014399999999999998, + "loss": 0.0, + "step": 960 + }, + { + "epoch": 0.45223529411764707, + "grad_norm": 0.05287458001270689, + "learning_rate": 0.00014414999999999998, + "loss": 0.0153, + "step": 961 + }, + { + "epoch": 0.4527058823529412, + "grad_norm": 0.17485923122052274, + "learning_rate": 0.00014429999999999998, + "loss": 0.0707, + "step": 962 + }, + { + "epoch": 0.4531764705882353, + "grad_norm": 0.00013208920652787008, + "learning_rate": 0.00014445, + "loss": 0.0, + "step": 963 + }, + { + "epoch": 0.4536470588235294, + "grad_norm": 0.09476422088530234, + "learning_rate": 0.0001446, + "loss": 0.0306, + "step": 964 + }, + { + "epoch": 0.4541176470588235, + "grad_norm": 0.24694229471387072, + "learning_rate": 0.00014475, + "loss": 0.0989, + "step": 965 + }, + { + "epoch": 0.4545882352941176, + "grad_norm": 0.10376628021450773, + "learning_rate": 0.00014489999999999997, + "loss": 0.0064, + "step": 966 + }, + { + "epoch": 0.45505882352941174, + "grad_norm": 0.2465818096763799, + "learning_rate": 0.00014504999999999997, + "loss": 0.0991, + "step": 967 + }, + { + "epoch": 0.4555294117647059, + "grad_norm": 0.3913119939238422, + "learning_rate": 0.00014519999999999998, + "loss": 0.1258, + "step": 968 + }, + { + "epoch": 0.456, + "grad_norm": 0.17977023238986567, + "learning_rate": 0.00014534999999999998, + "loss": 0.0724, + "step": 969 + }, + { + "epoch": 0.45647058823529413, + "grad_norm": 0.10389262661894651, + "learning_rate": 0.00014549999999999999, + "loss": 0.0328, + "step": 970 + }, + { + "epoch": 0.45694117647058824, + "grad_norm": 0.058769970243634, + "learning_rate": 0.00014565, + "loss": 0.0148, + "step": 971 + }, + { + "epoch": 0.45741176470588235, + "grad_norm": 0.14814172103686166, + "learning_rate": 0.0001458, + "loss": 0.0318, + "step": 972 + }, + { + "epoch": 0.45788235294117646, + "grad_norm": 0.749242412685977, + "learning_rate": 0.00014595, + "loss": 0.0923, + "step": 973 + }, + { + "epoch": 0.4583529411764706, + "grad_norm": 0.0690554232965291, + "learning_rate": 0.00014609999999999997, + "loss": 0.0172, + "step": 974 + }, + { + "epoch": 0.4588235294117647, + "grad_norm": 0.07426074444908444, + "learning_rate": 0.00014624999999999998, + "loss": 0.0204, + "step": 975 + }, + { + "epoch": 0.4592941176470588, + "grad_norm": 0.2291451509734054, + "learning_rate": 0.00014639999999999998, + "loss": 0.093, + "step": 976 + }, + { + "epoch": 0.45976470588235296, + "grad_norm": 0.4175709317684765, + "learning_rate": 0.00014654999999999998, + "loss": 0.1379, + "step": 977 + }, + { + "epoch": 0.4602352941176471, + "grad_norm": 0.07477097873383305, + "learning_rate": 0.0001467, + "loss": 0.0339, + "step": 978 + }, + { + "epoch": 0.4607058823529412, + "grad_norm": 0.0321279039090547, + "learning_rate": 0.00014685, + "loss": 0.0005, + "step": 979 + }, + { + "epoch": 0.4611764705882353, + "grad_norm": 0.1398866264627896, + "learning_rate": 0.000147, + "loss": 0.0496, + "step": 980 + }, + { + "epoch": 0.4616470588235294, + "grad_norm": 0.14972228893278708, + "learning_rate": 0.00014714999999999997, + "loss": 0.0613, + "step": 981 + }, + { + "epoch": 0.4621176470588235, + "grad_norm": 0.006553429693085149, + "learning_rate": 0.00014729999999999998, + "loss": 0.0001, + "step": 982 + }, + { + "epoch": 0.46258823529411763, + "grad_norm": 0.16759586188689493, + "learning_rate": 0.00014744999999999998, + "loss": 0.0716, + "step": 983 + }, + { + "epoch": 0.46305882352941174, + "grad_norm": 0.044757720634719986, + "learning_rate": 0.00014759999999999998, + "loss": 0.0119, + "step": 984 + }, + { + "epoch": 0.46352941176470586, + "grad_norm": 0.0013258306738971697, + "learning_rate": 0.00014774999999999999, + "loss": 0.0, + "step": 985 + }, + { + "epoch": 0.464, + "grad_norm": 0.1413617074982139, + "learning_rate": 0.0001479, + "loss": 0.061, + "step": 986 + }, + { + "epoch": 0.46447058823529414, + "grad_norm": 0.15392784067484963, + "learning_rate": 0.00014805, + "loss": 0.0704, + "step": 987 + }, + { + "epoch": 0.46494117647058825, + "grad_norm": 0.0008125200273716852, + "learning_rate": 0.0001482, + "loss": 0.0, + "step": 988 + }, + { + "epoch": 0.46541176470588236, + "grad_norm": 0.0004133655324837666, + "learning_rate": 0.00014834999999999997, + "loss": 0.0, + "step": 989 + }, + { + "epoch": 0.46588235294117647, + "grad_norm": 0.14495899758745023, + "learning_rate": 0.00014849999999999998, + "loss": 0.0572, + "step": 990 + }, + { + "epoch": 0.4663529411764706, + "grad_norm": 0.12055320719626694, + "learning_rate": 0.00014864999999999998, + "loss": 0.0422, + "step": 991 + }, + { + "epoch": 0.4668235294117647, + "grad_norm": 0.16905842480315997, + "learning_rate": 0.00014879999999999998, + "loss": 0.0809, + "step": 992 + }, + { + "epoch": 0.4672941176470588, + "grad_norm": 0.20932795082636738, + "learning_rate": 0.00014895, + "loss": 0.0926, + "step": 993 + }, + { + "epoch": 0.4677647058823529, + "grad_norm": 0.00020109640753572446, + "learning_rate": 0.0001491, + "loss": 0.0, + "step": 994 + }, + { + "epoch": 0.4682352941176471, + "grad_norm": 0.188963646199739, + "learning_rate": 0.00014925, + "loss": 0.0835, + "step": 995 + }, + { + "epoch": 0.4687058823529412, + "grad_norm": 0.0001791282203985687, + "learning_rate": 0.0001494, + "loss": 0.0, + "step": 996 + }, + { + "epoch": 0.4691764705882353, + "grad_norm": 0.30658801924605356, + "learning_rate": 0.00014954999999999998, + "loss": 0.0556, + "step": 997 + }, + { + "epoch": 0.4696470588235294, + "grad_norm": 0.3129675394829724, + "learning_rate": 0.00014969999999999998, + "loss": 0.1154, + "step": 998 + }, + { + "epoch": 0.47011764705882353, + "grad_norm": 0.13145196664749503, + "learning_rate": 0.00014984999999999998, + "loss": 0.0532, + "step": 999 + }, + { + "epoch": 0.47058823529411764, + "grad_norm": 0.27464799381751787, + "learning_rate": 0.00015, + "loss": 0.1821, + "step": 1000 + }, + { + "epoch": 0.47105882352941175, + "grad_norm": 0.10457792153621126, + "learning_rate": 0.00015014999999999996, + "loss": 0.0329, + "step": 1001 + }, + { + "epoch": 0.47152941176470586, + "grad_norm": 0.19316218685052794, + "learning_rate": 0.0001503, + "loss": 0.0952, + "step": 1002 + }, + { + "epoch": 0.472, + "grad_norm": 0.07316558215395876, + "learning_rate": 0.00015044999999999997, + "loss": 0.0251, + "step": 1003 + }, + { + "epoch": 0.47247058823529414, + "grad_norm": 0.40685956783786614, + "learning_rate": 0.00015059999999999997, + "loss": 0.0445, + "step": 1004 + }, + { + "epoch": 0.47294117647058825, + "grad_norm": 0.18370144810557942, + "learning_rate": 0.00015074999999999998, + "loss": 0.0681, + "step": 1005 + }, + { + "epoch": 0.47341176470588237, + "grad_norm": 0.20247202145912982, + "learning_rate": 0.00015089999999999998, + "loss": 0.0753, + "step": 1006 + }, + { + "epoch": 0.4738823529411765, + "grad_norm": 0.050244751961215875, + "learning_rate": 0.00015104999999999996, + "loss": 0.0085, + "step": 1007 + }, + { + "epoch": 0.4743529411764706, + "grad_norm": 0.11154022924736705, + "learning_rate": 0.0001512, + "loss": 0.0358, + "step": 1008 + }, + { + "epoch": 0.4748235294117647, + "grad_norm": 0.2851219473208837, + "learning_rate": 0.00015134999999999997, + "loss": 0.0986, + "step": 1009 + }, + { + "epoch": 0.4752941176470588, + "grad_norm": 0.11466486823760777, + "learning_rate": 0.0001515, + "loss": 0.0542, + "step": 1010 + }, + { + "epoch": 0.4757647058823529, + "grad_norm": 0.11727184727982846, + "learning_rate": 0.00015164999999999997, + "loss": 0.0332, + "step": 1011 + }, + { + "epoch": 0.47623529411764703, + "grad_norm": 0.11999174363583695, + "learning_rate": 0.00015179999999999998, + "loss": 0.0468, + "step": 1012 + }, + { + "epoch": 0.4767058823529412, + "grad_norm": 0.15345589397372775, + "learning_rate": 0.00015194999999999998, + "loss": 0.0462, + "step": 1013 + }, + { + "epoch": 0.4771764705882353, + "grad_norm": 0.1430493006419013, + "learning_rate": 0.00015209999999999998, + "loss": 0.0435, + "step": 1014 + }, + { + "epoch": 0.4776470588235294, + "grad_norm": 8.366001861815061e-05, + "learning_rate": 0.00015224999999999996, + "loss": 0.0, + "step": 1015 + }, + { + "epoch": 0.47811764705882354, + "grad_norm": 0.10512973278826197, + "learning_rate": 0.0001524, + "loss": 0.032, + "step": 1016 + }, + { + "epoch": 0.47858823529411765, + "grad_norm": 1.1582176803663322, + "learning_rate": 0.00015254999999999997, + "loss": 0.0954, + "step": 1017 + }, + { + "epoch": 0.47905882352941176, + "grad_norm": 1.9875921034903852, + "learning_rate": 0.0001527, + "loss": 0.2046, + "step": 1018 + }, + { + "epoch": 0.47952941176470587, + "grad_norm": 0.21425789889601252, + "learning_rate": 0.00015284999999999997, + "loss": 0.0634, + "step": 1019 + }, + { + "epoch": 0.48, + "grad_norm": 0.12069515238374341, + "learning_rate": 0.00015299999999999998, + "loss": 0.0483, + "step": 1020 + }, + { + "epoch": 0.4804705882352941, + "grad_norm": 0.37941437504898895, + "learning_rate": 0.00015314999999999998, + "loss": 0.1102, + "step": 1021 + }, + { + "epoch": 0.48094117647058826, + "grad_norm": 0.29406053468458604, + "learning_rate": 0.00015329999999999999, + "loss": 0.1121, + "step": 1022 + }, + { + "epoch": 0.4814117647058824, + "grad_norm": 0.10220768920446412, + "learning_rate": 0.00015344999999999996, + "loss": 0.0074, + "step": 1023 + }, + { + "epoch": 0.4818823529411765, + "grad_norm": 0.20292580872702595, + "learning_rate": 0.0001536, + "loss": 0.06, + "step": 1024 + }, + { + "epoch": 0.4823529411764706, + "grad_norm": 0.16630284987384164, + "learning_rate": 0.00015374999999999997, + "loss": 0.0147, + "step": 1025 + }, + { + "epoch": 0.4828235294117647, + "grad_norm": 0.1921010730528137, + "learning_rate": 0.0001539, + "loss": 0.0974, + "step": 1026 + }, + { + "epoch": 0.4832941176470588, + "grad_norm": 0.3507218937435455, + "learning_rate": 0.00015404999999999998, + "loss": 0.0378, + "step": 1027 + }, + { + "epoch": 0.48376470588235293, + "grad_norm": 0.0029602428280424463, + "learning_rate": 0.00015419999999999998, + "loss": 0.0001, + "step": 1028 + }, + { + "epoch": 0.48423529411764704, + "grad_norm": 0.07007471805264322, + "learning_rate": 0.00015434999999999998, + "loss": 0.0125, + "step": 1029 + }, + { + "epoch": 0.48470588235294115, + "grad_norm": 0.33509683018965225, + "learning_rate": 0.0001545, + "loss": 0.1145, + "step": 1030 + }, + { + "epoch": 0.4851764705882353, + "grad_norm": 0.3661904574148443, + "learning_rate": 0.00015464999999999996, + "loss": 0.1261, + "step": 1031 + }, + { + "epoch": 0.48564705882352943, + "grad_norm": 0.05517155926131357, + "learning_rate": 0.0001548, + "loss": 0.0099, + "step": 1032 + }, + { + "epoch": 0.48611764705882354, + "grad_norm": 0.15113921202502487, + "learning_rate": 0.00015494999999999997, + "loss": 0.0542, + "step": 1033 + }, + { + "epoch": 0.48658823529411765, + "grad_norm": 0.3257990969512233, + "learning_rate": 0.0001551, + "loss": 0.1448, + "step": 1034 + }, + { + "epoch": 0.48705882352941177, + "grad_norm": 0.1842462868997423, + "learning_rate": 0.00015524999999999998, + "loss": 0.0808, + "step": 1035 + }, + { + "epoch": 0.4875294117647059, + "grad_norm": 0.13856512298595908, + "learning_rate": 0.00015539999999999998, + "loss": 0.0209, + "step": 1036 + }, + { + "epoch": 0.488, + "grad_norm": 0.015059743612730284, + "learning_rate": 0.00015554999999999999, + "loss": 0.0015, + "step": 1037 + }, + { + "epoch": 0.4884705882352941, + "grad_norm": 0.2665700978311753, + "learning_rate": 0.0001557, + "loss": 0.0926, + "step": 1038 + }, + { + "epoch": 0.4889411764705882, + "grad_norm": 0.17456457357105448, + "learning_rate": 0.00015584999999999997, + "loss": 0.0453, + "step": 1039 + }, + { + "epoch": 0.4894117647058824, + "grad_norm": 0.1887498475984473, + "learning_rate": 0.000156, + "loss": 0.0729, + "step": 1040 + }, + { + "epoch": 0.4898823529411765, + "grad_norm": 0.04303188138054942, + "learning_rate": 0.00015614999999999997, + "loss": 0.0061, + "step": 1041 + }, + { + "epoch": 0.4903529411764706, + "grad_norm": 4.0699113578387985e-05, + "learning_rate": 0.0001563, + "loss": 0.0, + "step": 1042 + }, + { + "epoch": 0.4908235294117647, + "grad_norm": 0.04189225594205606, + "learning_rate": 0.00015644999999999998, + "loss": 0.0051, + "step": 1043 + }, + { + "epoch": 0.4912941176470588, + "grad_norm": 0.3223470757128681, + "learning_rate": 0.00015659999999999998, + "loss": 0.1071, + "step": 1044 + }, + { + "epoch": 0.49176470588235294, + "grad_norm": 0.18719518545337016, + "learning_rate": 0.00015675, + "loss": 0.0095, + "step": 1045 + }, + { + "epoch": 0.49223529411764705, + "grad_norm": 0.04558404757768107, + "learning_rate": 0.0001569, + "loss": 0.0069, + "step": 1046 + }, + { + "epoch": 0.49270588235294116, + "grad_norm": 0.09032904301869749, + "learning_rate": 0.00015704999999999997, + "loss": 0.0268, + "step": 1047 + }, + { + "epoch": 0.49317647058823527, + "grad_norm": 0.11388599599116697, + "learning_rate": 0.0001572, + "loss": 0.0287, + "step": 1048 + }, + { + "epoch": 0.49364705882352944, + "grad_norm": 3.7927344116580674e-05, + "learning_rate": 0.00015734999999999998, + "loss": 0.0, + "step": 1049 + }, + { + "epoch": 0.49411764705882355, + "grad_norm": 0.08787724110465804, + "learning_rate": 0.00015749999999999998, + "loss": 0.0099, + "step": 1050 + }, + { + "epoch": 0.49458823529411766, + "grad_norm": 0.20555550677441756, + "learning_rate": 0.00015764999999999998, + "loss": 0.0605, + "step": 1051 + }, + { + "epoch": 0.4950588235294118, + "grad_norm": 0.07499696583937826, + "learning_rate": 0.0001578, + "loss": 0.0242, + "step": 1052 + }, + { + "epoch": 0.4955294117647059, + "grad_norm": 0.06436276660628246, + "learning_rate": 0.00015794999999999996, + "loss": 0.0194, + "step": 1053 + }, + { + "epoch": 0.496, + "grad_norm": 4.481637162584442e-05, + "learning_rate": 0.0001581, + "loss": 0.0, + "step": 1054 + }, + { + "epoch": 0.4964705882352941, + "grad_norm": 0.10041256226420299, + "learning_rate": 0.00015824999999999997, + "loss": 0.0197, + "step": 1055 + }, + { + "epoch": 0.4969411764705882, + "grad_norm": 0.3647499062223184, + "learning_rate": 0.0001584, + "loss": 0.1755, + "step": 1056 + }, + { + "epoch": 0.49741176470588233, + "grad_norm": 0.19284833091531658, + "learning_rate": 0.00015854999999999998, + "loss": 0.0894, + "step": 1057 + }, + { + "epoch": 0.4978823529411765, + "grad_norm": 0.1902289487926045, + "learning_rate": 0.00015869999999999998, + "loss": 0.0699, + "step": 1058 + }, + { + "epoch": 0.4983529411764706, + "grad_norm": 0.00040547431383935626, + "learning_rate": 0.00015884999999999999, + "loss": 0.0, + "step": 1059 + }, + { + "epoch": 0.4988235294117647, + "grad_norm": 0.07580258100907744, + "learning_rate": 0.000159, + "loss": 0.0137, + "step": 1060 + }, + { + "epoch": 0.49929411764705883, + "grad_norm": 0.16808539527584243, + "learning_rate": 0.00015914999999999997, + "loss": 0.055, + "step": 1061 + }, + { + "epoch": 0.49976470588235294, + "grad_norm": 0.027787879674526353, + "learning_rate": 0.0001593, + "loss": 0.002, + "step": 1062 + }, + { + "epoch": 0.5002352941176471, + "grad_norm": 0.13399583121767264, + "learning_rate": 0.00015944999999999997, + "loss": 0.0279, + "step": 1063 + }, + { + "epoch": 0.5007058823529412, + "grad_norm": 0.1895879801625132, + "learning_rate": 0.0001596, + "loss": 0.0825, + "step": 1064 + }, + { + "epoch": 0.5011764705882353, + "grad_norm": 0.1331930374432588, + "learning_rate": 0.00015974999999999998, + "loss": 0.0343, + "step": 1065 + }, + { + "epoch": 0.5016470588235294, + "grad_norm": 0.20543539462780078, + "learning_rate": 0.00015989999999999998, + "loss": 0.0935, + "step": 1066 + }, + { + "epoch": 0.5021176470588236, + "grad_norm": 0.06081049148366827, + "learning_rate": 0.00016005, + "loss": 0.0127, + "step": 1067 + }, + { + "epoch": 0.5025882352941177, + "grad_norm": 0.29757706022394514, + "learning_rate": 0.0001602, + "loss": 0.1319, + "step": 1068 + }, + { + "epoch": 0.5030588235294118, + "grad_norm": 0.13731781621890365, + "learning_rate": 0.00016034999999999997, + "loss": 0.0467, + "step": 1069 + }, + { + "epoch": 0.5035294117647059, + "grad_norm": 0.11771660483685509, + "learning_rate": 0.0001605, + "loss": 0.0172, + "step": 1070 + }, + { + "epoch": 0.504, + "grad_norm": 0.005620015368943685, + "learning_rate": 0.00016064999999999997, + "loss": 0.0002, + "step": 1071 + }, + { + "epoch": 0.5044705882352941, + "grad_norm": 0.17894066390265262, + "learning_rate": 0.0001608, + "loss": 0.0558, + "step": 1072 + }, + { + "epoch": 0.5049411764705882, + "grad_norm": 0.000511938110265801, + "learning_rate": 0.00016094999999999998, + "loss": 0.0, + "step": 1073 + }, + { + "epoch": 0.5054117647058823, + "grad_norm": 0.0007309895197349286, + "learning_rate": 0.00016109999999999999, + "loss": 0.0, + "step": 1074 + }, + { + "epoch": 0.5058823529411764, + "grad_norm": 0.07829392932181345, + "learning_rate": 0.00016125, + "loss": 0.0134, + "step": 1075 + }, + { + "epoch": 0.5063529411764706, + "grad_norm": 0.1447092031469999, + "learning_rate": 0.0001614, + "loss": 0.0237, + "step": 1076 + }, + { + "epoch": 0.5068235294117647, + "grad_norm": 0.0001908489796233581, + "learning_rate": 0.00016154999999999997, + "loss": 0.0, + "step": 1077 + }, + { + "epoch": 0.5072941176470588, + "grad_norm": 0.00016581787016791582, + "learning_rate": 0.0001617, + "loss": 0.0, + "step": 1078 + }, + { + "epoch": 0.5077647058823529, + "grad_norm": 0.5355392958496219, + "learning_rate": 0.00016184999999999998, + "loss": 0.0752, + "step": 1079 + }, + { + "epoch": 0.508235294117647, + "grad_norm": 0.13696910299882298, + "learning_rate": 0.000162, + "loss": 0.0226, + "step": 1080 + }, + { + "epoch": 0.5087058823529412, + "grad_norm": 0.3524329093272123, + "learning_rate": 0.00016214999999999998, + "loss": 0.1323, + "step": 1081 + }, + { + "epoch": 0.5091764705882353, + "grad_norm": 0.0626092715733951, + "learning_rate": 0.0001623, + "loss": 0.0071, + "step": 1082 + }, + { + "epoch": 0.5096470588235295, + "grad_norm": 0.2595897928954412, + "learning_rate": 0.00016245, + "loss": 0.073, + "step": 1083 + }, + { + "epoch": 0.5101176470588236, + "grad_norm": 0.28261094096133066, + "learning_rate": 0.0001626, + "loss": 0.1258, + "step": 1084 + }, + { + "epoch": 0.5105882352941177, + "grad_norm": 0.04654933585341538, + "learning_rate": 0.00016274999999999997, + "loss": 0.0048, + "step": 1085 + }, + { + "epoch": 0.5110588235294118, + "grad_norm": 0.22556868306348582, + "learning_rate": 0.0001629, + "loss": 0.0637, + "step": 1086 + }, + { + "epoch": 0.5115294117647059, + "grad_norm": 0.011790598615959828, + "learning_rate": 0.00016304999999999998, + "loss": 0.0001, + "step": 1087 + }, + { + "epoch": 0.512, + "grad_norm": 0.1546724037537526, + "learning_rate": 0.0001632, + "loss": 0.0304, + "step": 1088 + }, + { + "epoch": 0.5124705882352941, + "grad_norm": 0.20560057705367796, + "learning_rate": 0.00016334999999999999, + "loss": 0.0818, + "step": 1089 + }, + { + "epoch": 0.5129411764705882, + "grad_norm": 0.08701738709493566, + "learning_rate": 0.0001635, + "loss": 0.0188, + "step": 1090 + }, + { + "epoch": 0.5134117647058823, + "grad_norm": 0.1537005631142685, + "learning_rate": 0.00016365, + "loss": 0.0377, + "step": 1091 + }, + { + "epoch": 0.5138823529411765, + "grad_norm": 0.0018679802173699933, + "learning_rate": 0.0001638, + "loss": 0.0, + "step": 1092 + }, + { + "epoch": 0.5143529411764706, + "grad_norm": 0.09398753859834684, + "learning_rate": 0.00016394999999999997, + "loss": 0.0157, + "step": 1093 + }, + { + "epoch": 0.5148235294117647, + "grad_norm": 0.23971307207095818, + "learning_rate": 0.0001641, + "loss": 0.0861, + "step": 1094 + }, + { + "epoch": 0.5152941176470588, + "grad_norm": 0.12540995533649385, + "learning_rate": 0.00016424999999999998, + "loss": 0.026, + "step": 1095 + }, + { + "epoch": 0.5157647058823529, + "grad_norm": 0.26909752607473053, + "learning_rate": 0.0001644, + "loss": 0.0427, + "step": 1096 + }, + { + "epoch": 0.516235294117647, + "grad_norm": 0.11015746131668666, + "learning_rate": 0.00016455, + "loss": 0.0271, + "step": 1097 + }, + { + "epoch": 0.5167058823529411, + "grad_norm": 0.14418151141546712, + "learning_rate": 0.0001647, + "loss": 0.0287, + "step": 1098 + }, + { + "epoch": 0.5171764705882353, + "grad_norm": 0.13170044733152225, + "learning_rate": 0.00016485, + "loss": 0.0418, + "step": 1099 + }, + { + "epoch": 0.5176470588235295, + "grad_norm": 0.1169196052535333, + "learning_rate": 0.000165, + "loss": 0.0329, + "step": 1100 + }, + { + "epoch": 0.5181176470588236, + "grad_norm": 0.26220314743297696, + "learning_rate": 0.00016514999999999998, + "loss": 0.0973, + "step": 1101 + }, + { + "epoch": 0.5185882352941177, + "grad_norm": 0.19163340303183496, + "learning_rate": 0.0001653, + "loss": 0.0494, + "step": 1102 + }, + { + "epoch": 0.5190588235294118, + "grad_norm": 0.1391025981737119, + "learning_rate": 0.00016544999999999998, + "loss": 0.0303, + "step": 1103 + }, + { + "epoch": 0.5195294117647059, + "grad_norm": 0.26966463091624476, + "learning_rate": 0.0001656, + "loss": 0.0812, + "step": 1104 + }, + { + "epoch": 0.52, + "grad_norm": 0.19854941409994825, + "learning_rate": 0.00016575, + "loss": 0.0729, + "step": 1105 + }, + { + "epoch": 0.5204705882352941, + "grad_norm": 0.00012769441759537866, + "learning_rate": 0.0001659, + "loss": 0.0, + "step": 1106 + }, + { + "epoch": 0.5209411764705882, + "grad_norm": 0.22915693010443405, + "learning_rate": 0.00016604999999999997, + "loss": 0.0404, + "step": 1107 + }, + { + "epoch": 0.5214117647058824, + "grad_norm": 0.1267720831461711, + "learning_rate": 0.0001662, + "loss": 0.0304, + "step": 1108 + }, + { + "epoch": 0.5218823529411765, + "grad_norm": 0.182882158180967, + "learning_rate": 0.00016634999999999998, + "loss": 0.0569, + "step": 1109 + }, + { + "epoch": 0.5223529411764706, + "grad_norm": 0.08100339355387615, + "learning_rate": 0.0001665, + "loss": 0.0163, + "step": 1110 + }, + { + "epoch": 0.5228235294117647, + "grad_norm": 0.00042111557210932916, + "learning_rate": 0.00016664999999999998, + "loss": 0.0, + "step": 1111 + }, + { + "epoch": 0.5232941176470588, + "grad_norm": 0.17574664409463925, + "learning_rate": 0.0001668, + "loss": 0.057, + "step": 1112 + }, + { + "epoch": 0.5237647058823529, + "grad_norm": 0.18100432277846107, + "learning_rate": 0.00016695, + "loss": 0.0338, + "step": 1113 + }, + { + "epoch": 0.524235294117647, + "grad_norm": 0.0007300057213221582, + "learning_rate": 0.0001671, + "loss": 0.0, + "step": 1114 + }, + { + "epoch": 0.5247058823529411, + "grad_norm": 1.4586647131203205, + "learning_rate": 0.00016724999999999997, + "loss": 0.0467, + "step": 1115 + }, + { + "epoch": 0.5251764705882352, + "grad_norm": 0.21086752710789183, + "learning_rate": 0.0001674, + "loss": 0.0672, + "step": 1116 + }, + { + "epoch": 0.5256470588235295, + "grad_norm": 0.19851738257560284, + "learning_rate": 0.00016754999999999998, + "loss": 0.0281, + "step": 1117 + }, + { + "epoch": 0.5261176470588236, + "grad_norm": 0.0906508363826366, + "learning_rate": 0.0001677, + "loss": 0.018, + "step": 1118 + }, + { + "epoch": 0.5265882352941177, + "grad_norm": 6.180670466370176e-07, + "learning_rate": 0.00016785, + "loss": 0.0, + "step": 1119 + }, + { + "epoch": 0.5270588235294118, + "grad_norm": 0.23681447409260825, + "learning_rate": 0.000168, + "loss": 0.0654, + "step": 1120 + }, + { + "epoch": 0.5275294117647059, + "grad_norm": 0.2854892473229347, + "learning_rate": 0.00016815, + "loss": 0.0467, + "step": 1121 + }, + { + "epoch": 0.528, + "grad_norm": 0.18308152170587946, + "learning_rate": 0.0001683, + "loss": 0.033, + "step": 1122 + }, + { + "epoch": 0.5284705882352941, + "grad_norm": 0.07930901003665634, + "learning_rate": 0.00016844999999999997, + "loss": 0.0083, + "step": 1123 + }, + { + "epoch": 0.5289411764705882, + "grad_norm": 0.21817791968088932, + "learning_rate": 0.0001686, + "loss": 0.0708, + "step": 1124 + }, + { + "epoch": 0.5294117647058824, + "grad_norm": 0.7686950483176034, + "learning_rate": 0.00016874999999999998, + "loss": 0.1517, + "step": 1125 + }, + { + "epoch": 0.5298823529411765, + "grad_norm": 0.115958082676955, + "learning_rate": 0.00016889999999999996, + "loss": 0.0371, + "step": 1126 + }, + { + "epoch": 0.5303529411764706, + "grad_norm": 0.27867001920837703, + "learning_rate": 0.00016905, + "loss": 0.0522, + "step": 1127 + }, + { + "epoch": 0.5308235294117647, + "grad_norm": 0.19839607986819732, + "learning_rate": 0.00016919999999999997, + "loss": 0.0128, + "step": 1128 + }, + { + "epoch": 0.5312941176470588, + "grad_norm": 0.2088840317805112, + "learning_rate": 0.00016935, + "loss": 0.0813, + "step": 1129 + }, + { + "epoch": 0.5317647058823529, + "grad_norm": 0.13783067822069975, + "learning_rate": 0.00016949999999999997, + "loss": 0.0355, + "step": 1130 + }, + { + "epoch": 0.532235294117647, + "grad_norm": 0.18642068134359366, + "learning_rate": 0.00016964999999999998, + "loss": 0.0825, + "step": 1131 + }, + { + "epoch": 0.5327058823529411, + "grad_norm": 0.07631538602006775, + "learning_rate": 0.00016979999999999998, + "loss": 0.0125, + "step": 1132 + }, + { + "epoch": 0.5331764705882353, + "grad_norm": 0.2911541868308559, + "learning_rate": 0.00016994999999999998, + "loss": 0.1592, + "step": 1133 + }, + { + "epoch": 0.5336470588235294, + "grad_norm": 0.12188340331812605, + "learning_rate": 0.00017009999999999996, + "loss": 0.0387, + "step": 1134 + }, + { + "epoch": 0.5341176470588235, + "grad_norm": 0.0002778491556328095, + "learning_rate": 0.00017025, + "loss": 0.0, + "step": 1135 + }, + { + "epoch": 0.5345882352941177, + "grad_norm": 0.0743646639326081, + "learning_rate": 0.00017039999999999997, + "loss": 0.0144, + "step": 1136 + }, + { + "epoch": 0.5350588235294118, + "grad_norm": 0.35850623193511094, + "learning_rate": 0.00017055, + "loss": 0.1425, + "step": 1137 + }, + { + "epoch": 0.5355294117647059, + "grad_norm": 0.1525492371373276, + "learning_rate": 0.00017069999999999998, + "loss": 0.0429, + "step": 1138 + }, + { + "epoch": 0.536, + "grad_norm": 0.0005390470263133024, + "learning_rate": 0.00017084999999999998, + "loss": 0.0, + "step": 1139 + }, + { + "epoch": 0.5364705882352941, + "grad_norm": 0.06103397372047682, + "learning_rate": 0.00017099999999999998, + "loss": 0.011, + "step": 1140 + }, + { + "epoch": 0.5369411764705883, + "grad_norm": 0.008349868366281436, + "learning_rate": 0.00017114999999999999, + "loss": 0.0002, + "step": 1141 + }, + { + "epoch": 0.5374117647058824, + "grad_norm": 0.22791776670570332, + "learning_rate": 0.00017129999999999996, + "loss": 0.0944, + "step": 1142 + }, + { + "epoch": 0.5378823529411765, + "grad_norm": 0.1926171831559406, + "learning_rate": 0.00017145, + "loss": 0.0871, + "step": 1143 + }, + { + "epoch": 0.5383529411764706, + "grad_norm": 0.2002432019358004, + "learning_rate": 0.00017159999999999997, + "loss": 0.1002, + "step": 1144 + }, + { + "epoch": 0.5388235294117647, + "grad_norm": 0.15147622869003713, + "learning_rate": 0.00017175, + "loss": 0.0441, + "step": 1145 + }, + { + "epoch": 0.5392941176470588, + "grad_norm": 0.05646113368100192, + "learning_rate": 0.00017189999999999998, + "loss": 0.0107, + "step": 1146 + }, + { + "epoch": 0.5397647058823529, + "grad_norm": 0.006013391554297216, + "learning_rate": 0.00017204999999999998, + "loss": 0.0001, + "step": 1147 + }, + { + "epoch": 0.540235294117647, + "grad_norm": 0.1485856907714036, + "learning_rate": 0.00017219999999999998, + "loss": 0.0473, + "step": 1148 + }, + { + "epoch": 0.5407058823529411, + "grad_norm": 0.47512896269587873, + "learning_rate": 0.00017235, + "loss": 0.1855, + "step": 1149 + }, + { + "epoch": 0.5411764705882353, + "grad_norm": 0.20941192599889422, + "learning_rate": 0.00017249999999999996, + "loss": 0.0785, + "step": 1150 + }, + { + "epoch": 0.5416470588235294, + "grad_norm": 0.0041408253345394256, + "learning_rate": 0.00017265, + "loss": 0.0, + "step": 1151 + }, + { + "epoch": 0.5421176470588235, + "grad_norm": 0.12031354281629894, + "learning_rate": 0.00017279999999999997, + "loss": 0.0323, + "step": 1152 + }, + { + "epoch": 0.5425882352941176, + "grad_norm": 0.08908682860127425, + "learning_rate": 0.00017294999999999998, + "loss": 0.0249, + "step": 1153 + }, + { + "epoch": 0.5430588235294118, + "grad_norm": 0.12732352351998372, + "learning_rate": 0.00017309999999999998, + "loss": 0.0422, + "step": 1154 + }, + { + "epoch": 0.5435294117647059, + "grad_norm": 0.08683823648753505, + "learning_rate": 0.00017324999999999998, + "loss": 0.0169, + "step": 1155 + }, + { + "epoch": 0.544, + "grad_norm": 0.0007239589103149187, + "learning_rate": 0.00017339999999999996, + "loss": 0.0, + "step": 1156 + }, + { + "epoch": 0.5444705882352942, + "grad_norm": 0.000245544787314119, + "learning_rate": 0.00017355, + "loss": 0.0, + "step": 1157 + }, + { + "epoch": 0.5449411764705883, + "grad_norm": 0.23091726021198336, + "learning_rate": 0.00017369999999999997, + "loss": 0.0806, + "step": 1158 + }, + { + "epoch": 0.5454117647058824, + "grad_norm": 0.19856237609972457, + "learning_rate": 0.00017385, + "loss": 0.0762, + "step": 1159 + }, + { + "epoch": 0.5458823529411765, + "grad_norm": 0.10242173266850386, + "learning_rate": 0.00017399999999999997, + "loss": 0.0209, + "step": 1160 + }, + { + "epoch": 0.5463529411764706, + "grad_norm": 0.20254907538754524, + "learning_rate": 0.00017414999999999998, + "loss": 0.0994, + "step": 1161 + }, + { + "epoch": 0.5468235294117647, + "grad_norm": 0.08063188979482795, + "learning_rate": 0.00017429999999999998, + "loss": 0.0221, + "step": 1162 + }, + { + "epoch": 0.5472941176470588, + "grad_norm": 0.0978223596096978, + "learning_rate": 0.00017444999999999998, + "loss": 0.0198, + "step": 1163 + }, + { + "epoch": 0.5477647058823529, + "grad_norm": 4.1315884099149006e-05, + "learning_rate": 0.00017459999999999996, + "loss": 0.0, + "step": 1164 + }, + { + "epoch": 0.548235294117647, + "grad_norm": 0.17064428131098133, + "learning_rate": 0.00017475, + "loss": 0.0576, + "step": 1165 + }, + { + "epoch": 0.5487058823529412, + "grad_norm": 0.15322054840848587, + "learning_rate": 0.00017489999999999997, + "loss": 0.0537, + "step": 1166 + }, + { + "epoch": 0.5491764705882353, + "grad_norm": 7.648484025955528e-06, + "learning_rate": 0.00017505, + "loss": 0.0, + "step": 1167 + }, + { + "epoch": 0.5496470588235294, + "grad_norm": 0.046832425284161115, + "learning_rate": 0.00017519999999999998, + "loss": 0.0082, + "step": 1168 + }, + { + "epoch": 0.5501176470588235, + "grad_norm": 0.12269605496157386, + "learning_rate": 0.00017534999999999998, + "loss": 0.0372, + "step": 1169 + }, + { + "epoch": 0.5505882352941176, + "grad_norm": 8.029219600346333e-05, + "learning_rate": 0.00017549999999999998, + "loss": 0.0, + "step": 1170 + }, + { + "epoch": 0.5510588235294117, + "grad_norm": 0.29059540066757567, + "learning_rate": 0.00017565, + "loss": 0.0575, + "step": 1171 + }, + { + "epoch": 0.5515294117647059, + "grad_norm": 0.07202019364907085, + "learning_rate": 0.00017579999999999996, + "loss": 0.0106, + "step": 1172 + }, + { + "epoch": 0.552, + "grad_norm": 0.43564474723825436, + "learning_rate": 0.00017595, + "loss": 0.0319, + "step": 1173 + }, + { + "epoch": 0.5524705882352942, + "grad_norm": 0.10552834035068559, + "learning_rate": 0.00017609999999999997, + "loss": 0.0195, + "step": 1174 + }, + { + "epoch": 0.5529411764705883, + "grad_norm": 0.17493084686910773, + "learning_rate": 0.00017625, + "loss": 0.0691, + "step": 1175 + }, + { + "epoch": 0.5534117647058824, + "grad_norm": 5.272488748904654e-05, + "learning_rate": 0.00017639999999999998, + "loss": 0.0, + "step": 1176 + }, + { + "epoch": 0.5538823529411765, + "grad_norm": 0.09011820693944794, + "learning_rate": 0.00017654999999999998, + "loss": 0.0131, + "step": 1177 + }, + { + "epoch": 0.5543529411764706, + "grad_norm": 0.07185663278508808, + "learning_rate": 0.00017669999999999999, + "loss": 0.0058, + "step": 1178 + }, + { + "epoch": 0.5548235294117647, + "grad_norm": 0.17741827149287187, + "learning_rate": 0.00017685, + "loss": 0.036, + "step": 1179 + }, + { + "epoch": 0.5552941176470588, + "grad_norm": 0.07517457461279967, + "learning_rate": 0.00017699999999999997, + "loss": 0.0173, + "step": 1180 + }, + { + "epoch": 0.5557647058823529, + "grad_norm": 0.03843614977582696, + "learning_rate": 0.00017715, + "loss": 0.0058, + "step": 1181 + }, + { + "epoch": 0.556235294117647, + "grad_norm": 0.2124836861770988, + "learning_rate": 0.00017729999999999997, + "loss": 0.0845, + "step": 1182 + }, + { + "epoch": 0.5567058823529412, + "grad_norm": 0.0006612196638560092, + "learning_rate": 0.00017745, + "loss": 0.0, + "step": 1183 + }, + { + "epoch": 0.5571764705882353, + "grad_norm": 0.13934207180823147, + "learning_rate": 0.00017759999999999998, + "loss": 0.0292, + "step": 1184 + }, + { + "epoch": 0.5576470588235294, + "grad_norm": 0.11859035765988186, + "learning_rate": 0.00017774999999999998, + "loss": 0.0243, + "step": 1185 + }, + { + "epoch": 0.5581176470588235, + "grad_norm": 0.1336208782173958, + "learning_rate": 0.0001779, + "loss": 0.0295, + "step": 1186 + }, + { + "epoch": 0.5585882352941176, + "grad_norm": 0.002282927376701705, + "learning_rate": 0.00017805, + "loss": 0.0, + "step": 1187 + }, + { + "epoch": 0.5590588235294117, + "grad_norm": 0.28981413507719894, + "learning_rate": 0.00017819999999999997, + "loss": 0.1333, + "step": 1188 + }, + { + "epoch": 0.5595294117647058, + "grad_norm": 0.14330234037602868, + "learning_rate": 0.00017835, + "loss": 0.0371, + "step": 1189 + }, + { + "epoch": 0.56, + "grad_norm": 0.029918895761329375, + "learning_rate": 0.00017849999999999997, + "loss": 0.0021, + "step": 1190 + }, + { + "epoch": 0.5604705882352942, + "grad_norm": 0.252154499701993, + "learning_rate": 0.00017865, + "loss": 0.0833, + "step": 1191 + }, + { + "epoch": 0.5609411764705883, + "grad_norm": 0.15549468720715906, + "learning_rate": 0.00017879999999999998, + "loss": 0.056, + "step": 1192 + }, + { + "epoch": 0.5614117647058824, + "grad_norm": 0.13640261368523016, + "learning_rate": 0.00017894999999999999, + "loss": 0.0353, + "step": 1193 + }, + { + "epoch": 0.5618823529411765, + "grad_norm": 0.09749022362425279, + "learning_rate": 0.0001791, + "loss": 0.0212, + "step": 1194 + }, + { + "epoch": 0.5623529411764706, + "grad_norm": 0.062012364696979544, + "learning_rate": 0.00017925, + "loss": 0.0085, + "step": 1195 + }, + { + "epoch": 0.5628235294117647, + "grad_norm": 0.0006563301265901764, + "learning_rate": 0.00017939999999999997, + "loss": 0.0, + "step": 1196 + }, + { + "epoch": 0.5632941176470588, + "grad_norm": 0.21743436103776825, + "learning_rate": 0.00017955, + "loss": 0.0455, + "step": 1197 + }, + { + "epoch": 0.563764705882353, + "grad_norm": 0.2187324065071588, + "learning_rate": 0.00017969999999999998, + "loss": 0.1031, + "step": 1198 + }, + { + "epoch": 0.5642352941176471, + "grad_norm": 0.2192643439478881, + "learning_rate": 0.00017984999999999998, + "loss": 0.1185, + "step": 1199 + }, + { + "epoch": 0.5647058823529412, + "grad_norm": 0.027036299945655595, + "learning_rate": 0.00017999999999999998, + "loss": 0.002, + "step": 1200 + }, + { + "epoch": 0.5651764705882353, + "grad_norm": 0.11926489413663842, + "learning_rate": 0.00018015, + "loss": 0.0357, + "step": 1201 + }, + { + "epoch": 0.5656470588235294, + "grad_norm": 0.11469309006991454, + "learning_rate": 0.00018029999999999996, + "loss": 0.0376, + "step": 1202 + }, + { + "epoch": 0.5661176470588235, + "grad_norm": 0.05580302394974989, + "learning_rate": 0.00018045, + "loss": 0.0084, + "step": 1203 + }, + { + "epoch": 0.5665882352941176, + "grad_norm": 0.19138285455529197, + "learning_rate": 0.00018059999999999997, + "loss": 0.0721, + "step": 1204 + }, + { + "epoch": 0.5670588235294117, + "grad_norm": 0.058335195443375544, + "learning_rate": 0.00018075, + "loss": 0.0069, + "step": 1205 + }, + { + "epoch": 0.5675294117647058, + "grad_norm": 0.26867754528496085, + "learning_rate": 0.00018089999999999998, + "loss": 0.0544, + "step": 1206 + }, + { + "epoch": 0.568, + "grad_norm": 0.031929768667321314, + "learning_rate": 0.00018104999999999998, + "loss": 0.0032, + "step": 1207 + }, + { + "epoch": 0.5684705882352942, + "grad_norm": 0.09705963366436123, + "learning_rate": 0.00018119999999999999, + "loss": 0.0275, + "step": 1208 + }, + { + "epoch": 0.5689411764705883, + "grad_norm": 0.0327468810068151, + "learning_rate": 0.00018135, + "loss": 0.0028, + "step": 1209 + }, + { + "epoch": 0.5694117647058824, + "grad_norm": 0.08788008890689399, + "learning_rate": 0.00018149999999999997, + "loss": 0.0145, + "step": 1210 + }, + { + "epoch": 0.5698823529411765, + "grad_norm": 0.1067696229849387, + "learning_rate": 0.00018165, + "loss": 0.0264, + "step": 1211 + }, + { + "epoch": 0.5703529411764706, + "grad_norm": 0.2799648205279318, + "learning_rate": 0.00018179999999999997, + "loss": 0.0615, + "step": 1212 + }, + { + "epoch": 0.5708235294117647, + "grad_norm": 0.0003016374161014499, + "learning_rate": 0.00018195, + "loss": 0.0, + "step": 1213 + }, + { + "epoch": 0.5712941176470588, + "grad_norm": 0.07152368265490773, + "learning_rate": 0.00018209999999999998, + "loss": 0.0092, + "step": 1214 + }, + { + "epoch": 0.571764705882353, + "grad_norm": 0.00013457134968014942, + "learning_rate": 0.00018224999999999998, + "loss": 0.0, + "step": 1215 + }, + { + "epoch": 0.5722352941176471, + "grad_norm": 0.07408441282017346, + "learning_rate": 0.0001824, + "loss": 0.0109, + "step": 1216 + }, + { + "epoch": 0.5727058823529412, + "grad_norm": 0.14412844634082958, + "learning_rate": 0.00018255, + "loss": 0.0536, + "step": 1217 + }, + { + "epoch": 0.5731764705882353, + "grad_norm": 0.0988459157112718, + "learning_rate": 0.00018269999999999997, + "loss": 0.0158, + "step": 1218 + }, + { + "epoch": 0.5736470588235294, + "grad_norm": 0.20235441354616632, + "learning_rate": 0.00018285, + "loss": 0.0591, + "step": 1219 + }, + { + "epoch": 0.5741176470588235, + "grad_norm": 0.1116258324172071, + "learning_rate": 0.00018299999999999998, + "loss": 0.0424, + "step": 1220 + }, + { + "epoch": 0.5745882352941176, + "grad_norm": 0.0001005693546282167, + "learning_rate": 0.00018315, + "loss": 0.0, + "step": 1221 + }, + { + "epoch": 0.5750588235294117, + "grad_norm": 0.3637379563854363, + "learning_rate": 0.00018329999999999998, + "loss": 0.1502, + "step": 1222 + }, + { + "epoch": 0.5755294117647058, + "grad_norm": 0.0002371713167794552, + "learning_rate": 0.00018345, + "loss": 0.0, + "step": 1223 + }, + { + "epoch": 0.576, + "grad_norm": 0.0003741495447157671, + "learning_rate": 0.0001836, + "loss": 0.0, + "step": 1224 + }, + { + "epoch": 0.5764705882352941, + "grad_norm": 0.000986457178446213, + "learning_rate": 0.00018375, + "loss": 0.0, + "step": 1225 + }, + { + "epoch": 0.5769411764705883, + "grad_norm": 0.17087800532633496, + "learning_rate": 0.00018389999999999997, + "loss": 0.0358, + "step": 1226 + }, + { + "epoch": 0.5774117647058824, + "grad_norm": 0.17160693204175909, + "learning_rate": 0.00018405, + "loss": 0.0505, + "step": 1227 + }, + { + "epoch": 0.5778823529411765, + "grad_norm": 0.09553897174446185, + "learning_rate": 0.00018419999999999998, + "loss": 0.0336, + "step": 1228 + }, + { + "epoch": 0.5783529411764706, + "grad_norm": 0.11437755651573196, + "learning_rate": 0.00018435, + "loss": 0.0247, + "step": 1229 + }, + { + "epoch": 0.5788235294117647, + "grad_norm": 0.2294463204678339, + "learning_rate": 0.00018449999999999999, + "loss": 0.0774, + "step": 1230 + }, + { + "epoch": 0.5792941176470588, + "grad_norm": 0.09205039429764318, + "learning_rate": 0.00018465, + "loss": 0.026, + "step": 1231 + }, + { + "epoch": 0.579764705882353, + "grad_norm": 0.029722201566555213, + "learning_rate": 0.0001848, + "loss": 0.0034, + "step": 1232 + }, + { + "epoch": 0.5802352941176471, + "grad_norm": 0.26243981457957233, + "learning_rate": 0.00018495, + "loss": 0.0871, + "step": 1233 + }, + { + "epoch": 0.5807058823529412, + "grad_norm": 0.19439898142680836, + "learning_rate": 0.00018509999999999997, + "loss": 0.0679, + "step": 1234 + }, + { + "epoch": 0.5811764705882353, + "grad_norm": 0.08970355177643104, + "learning_rate": 0.00018525, + "loss": 0.0191, + "step": 1235 + }, + { + "epoch": 0.5816470588235294, + "grad_norm": 0.03582566605073923, + "learning_rate": 0.00018539999999999998, + "loss": 0.0048, + "step": 1236 + }, + { + "epoch": 0.5821176470588235, + "grad_norm": 0.03199948876025304, + "learning_rate": 0.00018555, + "loss": 0.0031, + "step": 1237 + }, + { + "epoch": 0.5825882352941176, + "grad_norm": 0.07967280840147543, + "learning_rate": 0.0001857, + "loss": 0.0142, + "step": 1238 + }, + { + "epoch": 0.5830588235294117, + "grad_norm": 0.26916436397205307, + "learning_rate": 0.00018585, + "loss": 0.0442, + "step": 1239 + }, + { + "epoch": 0.5835294117647059, + "grad_norm": 0.006961023894991987, + "learning_rate": 0.000186, + "loss": 0.0002, + "step": 1240 + }, + { + "epoch": 0.584, + "grad_norm": 0.17048605822861287, + "learning_rate": 0.00018615, + "loss": 0.0521, + "step": 1241 + }, + { + "epoch": 0.5844705882352941, + "grad_norm": 0.10780311034728055, + "learning_rate": 0.00018629999999999997, + "loss": 0.0306, + "step": 1242 + }, + { + "epoch": 0.5849411764705882, + "grad_norm": 0.0006768665744816679, + "learning_rate": 0.00018645, + "loss": 0.0, + "step": 1243 + }, + { + "epoch": 0.5854117647058823, + "grad_norm": 0.2809153414268674, + "learning_rate": 0.00018659999999999998, + "loss": 0.1184, + "step": 1244 + }, + { + "epoch": 0.5858823529411765, + "grad_norm": 0.02813060769031972, + "learning_rate": 0.00018675, + "loss": 0.0041, + "step": 1245 + }, + { + "epoch": 0.5863529411764706, + "grad_norm": 0.02897968457702694, + "learning_rate": 0.0001869, + "loss": 0.0019, + "step": 1246 + }, + { + "epoch": 0.5868235294117647, + "grad_norm": 0.0005723953892143997, + "learning_rate": 0.00018705, + "loss": 0.0, + "step": 1247 + }, + { + "epoch": 0.5872941176470589, + "grad_norm": 0.13726018825089428, + "learning_rate": 0.0001872, + "loss": 0.0453, + "step": 1248 + }, + { + "epoch": 0.587764705882353, + "grad_norm": 0.11086604373536178, + "learning_rate": 0.00018735, + "loss": 0.0266, + "step": 1249 + }, + { + "epoch": 0.5882352941176471, + "grad_norm": 0.02938057248737364, + "learning_rate": 0.00018749999999999998, + "loss": 0.002, + "step": 1250 + }, + { + "epoch": 0.5887058823529412, + "grad_norm": 0.6210753242804036, + "learning_rate": 0.00018764999999999998, + "loss": 0.0545, + "step": 1251 + }, + { + "epoch": 0.5891764705882353, + "grad_norm": 0.0001702553996665859, + "learning_rate": 0.00018779999999999998, + "loss": 0.0, + "step": 1252 + }, + { + "epoch": 0.5896470588235294, + "grad_norm": 0.1314323666848825, + "learning_rate": 0.00018794999999999996, + "loss": 0.0445, + "step": 1253 + }, + { + "epoch": 0.5901176470588235, + "grad_norm": 0.04570197335983622, + "learning_rate": 0.0001881, + "loss": 0.0051, + "step": 1254 + }, + { + "epoch": 0.5905882352941176, + "grad_norm": 0.03551073323938706, + "learning_rate": 0.00018824999999999997, + "loss": 0.0052, + "step": 1255 + }, + { + "epoch": 0.5910588235294117, + "grad_norm": 0.0668646919736275, + "learning_rate": 0.00018839999999999997, + "loss": 0.0105, + "step": 1256 + }, + { + "epoch": 0.5915294117647059, + "grad_norm": 0.00021813361035349961, + "learning_rate": 0.00018854999999999998, + "loss": 0.0, + "step": 1257 + }, + { + "epoch": 0.592, + "grad_norm": 0.0021992382651155394, + "learning_rate": 0.00018869999999999998, + "loss": 0.0, + "step": 1258 + }, + { + "epoch": 0.5924705882352941, + "grad_norm": 0.1245892965707693, + "learning_rate": 0.00018884999999999996, + "loss": 0.0423, + "step": 1259 + }, + { + "epoch": 0.5929411764705882, + "grad_norm": 0.12217533855173521, + "learning_rate": 0.00018899999999999999, + "loss": 0.0302, + "step": 1260 + }, + { + "epoch": 0.5934117647058823, + "grad_norm": 0.20755780236097826, + "learning_rate": 0.00018914999999999996, + "loss": 0.0629, + "step": 1261 + }, + { + "epoch": 0.5938823529411764, + "grad_norm": 0.09699923694449561, + "learning_rate": 0.0001893, + "loss": 0.032, + "step": 1262 + }, + { + "epoch": 0.5943529411764706, + "grad_norm": 0.31092023646788847, + "learning_rate": 0.00018944999999999997, + "loss": 0.0821, + "step": 1263 + }, + { + "epoch": 0.5948235294117648, + "grad_norm": 0.0007037471744447029, + "learning_rate": 0.00018959999999999997, + "loss": 0.0, + "step": 1264 + }, + { + "epoch": 0.5952941176470589, + "grad_norm": 0.1580497461458683, + "learning_rate": 0.00018974999999999998, + "loss": 0.078, + "step": 1265 + }, + { + "epoch": 0.595764705882353, + "grad_norm": 0.0009859973406532892, + "learning_rate": 0.00018989999999999998, + "loss": 0.0, + "step": 1266 + }, + { + "epoch": 0.5962352941176471, + "grad_norm": 0.062194963961441224, + "learning_rate": 0.00019004999999999996, + "loss": 0.0174, + "step": 1267 + }, + { + "epoch": 0.5967058823529412, + "grad_norm": 0.14236552209337952, + "learning_rate": 0.0001902, + "loss": 0.0572, + "step": 1268 + }, + { + "epoch": 0.5971764705882353, + "grad_norm": 0.06760019003737448, + "learning_rate": 0.00019034999999999996, + "loss": 0.0187, + "step": 1269 + }, + { + "epoch": 0.5976470588235294, + "grad_norm": 0.0009696714728813455, + "learning_rate": 0.0001905, + "loss": 0.0, + "step": 1270 + }, + { + "epoch": 0.5981176470588235, + "grad_norm": 0.10455406771445903, + "learning_rate": 0.00019064999999999997, + "loss": 0.0231, + "step": 1271 + }, + { + "epoch": 0.5985882352941176, + "grad_norm": 0.1382842203393118, + "learning_rate": 0.00019079999999999998, + "loss": 0.0554, + "step": 1272 + }, + { + "epoch": 0.5990588235294118, + "grad_norm": 0.14796623510852394, + "learning_rate": 0.00019094999999999998, + "loss": 0.0642, + "step": 1273 + }, + { + "epoch": 0.5995294117647059, + "grad_norm": 0.06642306217685727, + "learning_rate": 0.00019109999999999998, + "loss": 0.0149, + "step": 1274 + }, + { + "epoch": 0.6, + "grad_norm": 0.10317017748784084, + "learning_rate": 0.00019124999999999996, + "loss": 0.0226, + "step": 1275 + }, + { + "epoch": 0.6004705882352941, + "grad_norm": 0.11973403563985634, + "learning_rate": 0.0001914, + "loss": 0.0383, + "step": 1276 + }, + { + "epoch": 0.6009411764705882, + "grad_norm": 0.0013160728213015933, + "learning_rate": 0.00019154999999999997, + "loss": 0.0, + "step": 1277 + }, + { + "epoch": 0.6014117647058823, + "grad_norm": 0.12536378956318917, + "learning_rate": 0.0001917, + "loss": 0.0478, + "step": 1278 + }, + { + "epoch": 0.6018823529411764, + "grad_norm": 0.23320884931450797, + "learning_rate": 0.00019184999999999997, + "loss": 0.1164, + "step": 1279 + }, + { + "epoch": 0.6023529411764705, + "grad_norm": 0.001071648008990148, + "learning_rate": 0.00019199999999999998, + "loss": 0.0, + "step": 1280 + }, + { + "epoch": 0.6028235294117648, + "grad_norm": 0.23953874137627812, + "learning_rate": 0.00019214999999999998, + "loss": 0.088, + "step": 1281 + }, + { + "epoch": 0.6032941176470589, + "grad_norm": 0.11504104080426791, + "learning_rate": 0.00019229999999999999, + "loss": 0.0232, + "step": 1282 + }, + { + "epoch": 0.603764705882353, + "grad_norm": 0.45883781322063877, + "learning_rate": 0.00019244999999999996, + "loss": 0.2069, + "step": 1283 + }, + { + "epoch": 0.6042352941176471, + "grad_norm": 0.10523609667741592, + "learning_rate": 0.0001926, + "loss": 0.0317, + "step": 1284 + }, + { + "epoch": 0.6047058823529412, + "grad_norm": 0.11986672253808844, + "learning_rate": 0.00019274999999999997, + "loss": 0.0327, + "step": 1285 + }, + { + "epoch": 0.6051764705882353, + "grad_norm": 1.6661445487933186e-05, + "learning_rate": 0.0001929, + "loss": 0.0, + "step": 1286 + }, + { + "epoch": 0.6056470588235294, + "grad_norm": 0.12793781552917746, + "learning_rate": 0.00019304999999999998, + "loss": 0.0512, + "step": 1287 + }, + { + "epoch": 0.6061176470588235, + "grad_norm": 2.0124883025465908e-05, + "learning_rate": 0.00019319999999999998, + "loss": 0.0, + "step": 1288 + }, + { + "epoch": 0.6065882352941177, + "grad_norm": 0.13580220553343575, + "learning_rate": 0.00019334999999999998, + "loss": 0.0585, + "step": 1289 + }, + { + "epoch": 0.6070588235294118, + "grad_norm": 0.1462876264333955, + "learning_rate": 0.0001935, + "loss": 0.052, + "step": 1290 + }, + { + "epoch": 0.6075294117647059, + "grad_norm": 1.4560569470964607e-05, + "learning_rate": 0.00019364999999999996, + "loss": 0.0, + "step": 1291 + }, + { + "epoch": 0.608, + "grad_norm": 0.07641271155589734, + "learning_rate": 0.0001938, + "loss": 0.0066, + "step": 1292 + }, + { + "epoch": 0.6084705882352941, + "grad_norm": 0.11636394723811797, + "learning_rate": 0.00019394999999999997, + "loss": 0.048, + "step": 1293 + }, + { + "epoch": 0.6089411764705882, + "grad_norm": 8.793394827272046e-07, + "learning_rate": 0.0001941, + "loss": 0.0, + "step": 1294 + }, + { + "epoch": 0.6094117647058823, + "grad_norm": 0.0966107764020975, + "learning_rate": 0.00019424999999999998, + "loss": 0.0243, + "step": 1295 + }, + { + "epoch": 0.6098823529411764, + "grad_norm": 0.09874428408372664, + "learning_rate": 0.00019439999999999998, + "loss": 0.0465, + "step": 1296 + }, + { + "epoch": 0.6103529411764705, + "grad_norm": 0.24594750615736533, + "learning_rate": 0.00019454999999999999, + "loss": 0.0728, + "step": 1297 + }, + { + "epoch": 0.6108235294117647, + "grad_norm": 0.22496097612235616, + "learning_rate": 0.0001947, + "loss": 0.0943, + "step": 1298 + }, + { + "epoch": 0.6112941176470589, + "grad_norm": 0.04328077954286754, + "learning_rate": 0.00019484999999999997, + "loss": 0.0077, + "step": 1299 + }, + { + "epoch": 0.611764705882353, + "grad_norm": 0.1748801432446382, + "learning_rate": 0.000195, + "loss": 0.0452, + "step": 1300 + }, + { + "epoch": 0.6122352941176471, + "grad_norm": 0.161640560742925, + "learning_rate": 0.00019514999999999997, + "loss": 0.0748, + "step": 1301 + }, + { + "epoch": 0.6127058823529412, + "grad_norm": 0.07320364394003268, + "learning_rate": 0.00019529999999999998, + "loss": 0.0079, + "step": 1302 + }, + { + "epoch": 0.6131764705882353, + "grad_norm": 0.21080345826953734, + "learning_rate": 0.00019544999999999998, + "loss": 0.0956, + "step": 1303 + }, + { + "epoch": 0.6136470588235294, + "grad_norm": 0.22355581282981163, + "learning_rate": 0.00019559999999999998, + "loss": 0.1045, + "step": 1304 + }, + { + "epoch": 0.6141176470588235, + "grad_norm": 0.10826551253835054, + "learning_rate": 0.00019574999999999996, + "loss": 0.0367, + "step": 1305 + }, + { + "epoch": 0.6145882352941177, + "grad_norm": 0.024520107214846456, + "learning_rate": 0.0001959, + "loss": 0.0002, + "step": 1306 + }, + { + "epoch": 0.6150588235294118, + "grad_norm": 0.04033900535505578, + "learning_rate": 0.00019604999999999997, + "loss": 0.0037, + "step": 1307 + }, + { + "epoch": 0.6155294117647059, + "grad_norm": 0.127192514410351, + "learning_rate": 0.0001962, + "loss": 0.0217, + "step": 1308 + }, + { + "epoch": 0.616, + "grad_norm": 0.001047930846622143, + "learning_rate": 0.00019634999999999998, + "loss": 0.0, + "step": 1309 + }, + { + "epoch": 0.6164705882352941, + "grad_norm": 0.39609049707401456, + "learning_rate": 0.00019649999999999998, + "loss": 0.0975, + "step": 1310 + }, + { + "epoch": 0.6169411764705882, + "grad_norm": 0.27753187730820117, + "learning_rate": 0.00019664999999999998, + "loss": 0.1064, + "step": 1311 + }, + { + "epoch": 0.6174117647058823, + "grad_norm": 0.12098860435305839, + "learning_rate": 0.00019679999999999999, + "loss": 0.0218, + "step": 1312 + }, + { + "epoch": 0.6178823529411764, + "grad_norm": 0.21129831408389146, + "learning_rate": 0.00019694999999999996, + "loss": 0.0466, + "step": 1313 + }, + { + "epoch": 0.6183529411764705, + "grad_norm": 0.021080254576759258, + "learning_rate": 0.0001971, + "loss": 0.0025, + "step": 1314 + }, + { + "epoch": 0.6188235294117647, + "grad_norm": 0.34885624933734327, + "learning_rate": 0.00019724999999999997, + "loss": 0.0352, + "step": 1315 + }, + { + "epoch": 0.6192941176470588, + "grad_norm": 0.16566356674080518, + "learning_rate": 0.0001974, + "loss": 0.0302, + "step": 1316 + }, + { + "epoch": 0.619764705882353, + "grad_norm": 0.15002111527779863, + "learning_rate": 0.00019754999999999998, + "loss": 0.018, + "step": 1317 + }, + { + "epoch": 0.6202352941176471, + "grad_norm": 0.206240835530169, + "learning_rate": 0.00019769999999999998, + "loss": 0.0925, + "step": 1318 + }, + { + "epoch": 0.6207058823529412, + "grad_norm": 0.16654989007631774, + "learning_rate": 0.00019784999999999998, + "loss": 0.0558, + "step": 1319 + }, + { + "epoch": 0.6211764705882353, + "grad_norm": 0.049396443270268625, + "learning_rate": 0.000198, + "loss": 0.0078, + "step": 1320 + }, + { + "epoch": 0.6216470588235294, + "grad_norm": 0.05808066109494974, + "learning_rate": 0.00019814999999999996, + "loss": 0.0069, + "step": 1321 + }, + { + "epoch": 0.6221176470588236, + "grad_norm": 0.16933110200296111, + "learning_rate": 0.0001983, + "loss": 0.0308, + "step": 1322 + }, + { + "epoch": 0.6225882352941177, + "grad_norm": 0.11302364521154883, + "learning_rate": 0.00019844999999999997, + "loss": 0.0104, + "step": 1323 + }, + { + "epoch": 0.6230588235294118, + "grad_norm": 0.129561573291103, + "learning_rate": 0.0001986, + "loss": 0.0366, + "step": 1324 + }, + { + "epoch": 0.6235294117647059, + "grad_norm": 0.2191713285117613, + "learning_rate": 0.00019874999999999998, + "loss": 0.0622, + "step": 1325 + }, + { + "epoch": 0.624, + "grad_norm": 0.0004409367767769569, + "learning_rate": 0.00019889999999999998, + "loss": 0.0, + "step": 1326 + }, + { + "epoch": 0.6244705882352941, + "grad_norm": 0.24192212623593373, + "learning_rate": 0.00019905, + "loss": 0.1028, + "step": 1327 + }, + { + "epoch": 0.6249411764705882, + "grad_norm": 0.08993727062038717, + "learning_rate": 0.0001992, + "loss": 0.01, + "step": 1328 + }, + { + "epoch": 0.6254117647058823, + "grad_norm": 0.30091425303933195, + "learning_rate": 0.00019934999999999997, + "loss": 0.088, + "step": 1329 + }, + { + "epoch": 0.6258823529411764, + "grad_norm": 0.06137814281861265, + "learning_rate": 0.0001995, + "loss": 0.0228, + "step": 1330 + }, + { + "epoch": 0.6263529411764706, + "grad_norm": 0.01732143890437609, + "learning_rate": 0.00019964999999999997, + "loss": 0.0027, + "step": 1331 + }, + { + "epoch": 0.6268235294117647, + "grad_norm": 0.000472257163503522, + "learning_rate": 0.0001998, + "loss": 0.0, + "step": 1332 + }, + { + "epoch": 0.6272941176470588, + "grad_norm": 0.023599061060827344, + "learning_rate": 0.00019994999999999998, + "loss": 0.0017, + "step": 1333 + }, + { + "epoch": 0.6277647058823529, + "grad_norm": 0.24875821174141316, + "learning_rate": 0.00020009999999999998, + "loss": 0.1408, + "step": 1334 + }, + { + "epoch": 0.6282352941176471, + "grad_norm": 0.1320143844180369, + "learning_rate": 0.00020025, + "loss": 0.0665, + "step": 1335 + }, + { + "epoch": 0.6287058823529412, + "grad_norm": 0.07590328031692882, + "learning_rate": 0.0002004, + "loss": 0.0199, + "step": 1336 + }, + { + "epoch": 0.6291764705882353, + "grad_norm": 0.14748330222039413, + "learning_rate": 0.00020054999999999997, + "loss": 0.0425, + "step": 1337 + }, + { + "epoch": 0.6296470588235294, + "grad_norm": 0.05273605482091126, + "learning_rate": 0.0002007, + "loss": 0.0079, + "step": 1338 + }, + { + "epoch": 0.6301176470588236, + "grad_norm": 0.16966002727068052, + "learning_rate": 0.00020084999999999998, + "loss": 0.0693, + "step": 1339 + }, + { + "epoch": 0.6305882352941177, + "grad_norm": 0.08457355582774521, + "learning_rate": 0.000201, + "loss": 0.0227, + "step": 1340 + }, + { + "epoch": 0.6310588235294118, + "grad_norm": 0.099377180824405, + "learning_rate": 0.00020114999999999998, + "loss": 0.0272, + "step": 1341 + }, + { + "epoch": 0.6315294117647059, + "grad_norm": 0.002931764005734067, + "learning_rate": 0.0002013, + "loss": 0.0, + "step": 1342 + }, + { + "epoch": 0.632, + "grad_norm": 0.1329042802746641, + "learning_rate": 0.00020145, + "loss": 0.0327, + "step": 1343 + }, + { + "epoch": 0.6324705882352941, + "grad_norm": 0.002199102696773995, + "learning_rate": 0.0002016, + "loss": 0.0, + "step": 1344 + }, + { + "epoch": 0.6329411764705882, + "grad_norm": 0.12602234052673233, + "learning_rate": 0.00020174999999999997, + "loss": 0.0362, + "step": 1345 + }, + { + "epoch": 0.6334117647058823, + "grad_norm": 0.13679693728534933, + "learning_rate": 0.0002019, + "loss": 0.0411, + "step": 1346 + }, + { + "epoch": 0.6338823529411765, + "grad_norm": 0.10346191560139166, + "learning_rate": 0.00020204999999999998, + "loss": 0.0306, + "step": 1347 + }, + { + "epoch": 0.6343529411764706, + "grad_norm": 0.03285428769000204, + "learning_rate": 0.0002022, + "loss": 0.0051, + "step": 1348 + }, + { + "epoch": 0.6348235294117647, + "grad_norm": 0.006312782657507422, + "learning_rate": 0.00020234999999999999, + "loss": 0.0001, + "step": 1349 + }, + { + "epoch": 0.6352941176470588, + "grad_norm": 0.1543109031434328, + "learning_rate": 0.0002025, + "loss": 0.0522, + "step": 1350 + }, + { + "epoch": 0.6357647058823529, + "grad_norm": 0.24926590907358617, + "learning_rate": 0.00020264999999999997, + "loss": 0.1373, + "step": 1351 + }, + { + "epoch": 0.636235294117647, + "grad_norm": 0.0921840978172648, + "learning_rate": 0.0002028, + "loss": 0.0184, + "step": 1352 + }, + { + "epoch": 0.6367058823529411, + "grad_norm": 0.09051956672215357, + "learning_rate": 0.00020294999999999997, + "loss": 0.0259, + "step": 1353 + }, + { + "epoch": 0.6371764705882353, + "grad_norm": 0.0005626940392773808, + "learning_rate": 0.0002031, + "loss": 0.0, + "step": 1354 + }, + { + "epoch": 0.6376470588235295, + "grad_norm": 0.0282146347976368, + "learning_rate": 0.00020324999999999998, + "loss": 0.0031, + "step": 1355 + }, + { + "epoch": 0.6381176470588236, + "grad_norm": 0.350868838264281, + "learning_rate": 0.00020339999999999998, + "loss": 0.0755, + "step": 1356 + }, + { + "epoch": 0.6385882352941177, + "grad_norm": 0.0018254983571818239, + "learning_rate": 0.00020355, + "loss": 0.0, + "step": 1357 + }, + { + "epoch": 0.6390588235294118, + "grad_norm": 0.10517310788885606, + "learning_rate": 0.0002037, + "loss": 0.0325, + "step": 1358 + }, + { + "epoch": 0.6395294117647059, + "grad_norm": 0.12635114095702182, + "learning_rate": 0.00020384999999999997, + "loss": 0.0412, + "step": 1359 + }, + { + "epoch": 0.64, + "grad_norm": 0.06696217140879014, + "learning_rate": 0.000204, + "loss": 0.016, + "step": 1360 + }, + { + "epoch": 0.6404705882352941, + "grad_norm": 0.013414973659285841, + "learning_rate": 0.00020414999999999997, + "loss": 0.0001, + "step": 1361 + }, + { + "epoch": 0.6409411764705882, + "grad_norm": 0.06453072748915933, + "learning_rate": 0.0002043, + "loss": 0.015, + "step": 1362 + }, + { + "epoch": 0.6414117647058823, + "grad_norm": 0.14409925833923534, + "learning_rate": 0.00020444999999999998, + "loss": 0.0697, + "step": 1363 + }, + { + "epoch": 0.6418823529411765, + "grad_norm": 0.18317506463190733, + "learning_rate": 0.00020459999999999999, + "loss": 0.0755, + "step": 1364 + }, + { + "epoch": 0.6423529411764706, + "grad_norm": 0.24501019844351574, + "learning_rate": 0.00020475, + "loss": 0.1165, + "step": 1365 + }, + { + "epoch": 0.6428235294117647, + "grad_norm": 0.10849743783875117, + "learning_rate": 0.0002049, + "loss": 0.0279, + "step": 1366 + }, + { + "epoch": 0.6432941176470588, + "grad_norm": 0.12340114184865575, + "learning_rate": 0.00020504999999999997, + "loss": 0.0485, + "step": 1367 + }, + { + "epoch": 0.6437647058823529, + "grad_norm": 0.1114461139836742, + "learning_rate": 0.0002052, + "loss": 0.0559, + "step": 1368 + }, + { + "epoch": 0.644235294117647, + "grad_norm": 0.17255252073136867, + "learning_rate": 0.00020534999999999998, + "loss": 0.0954, + "step": 1369 + }, + { + "epoch": 0.6447058823529411, + "grad_norm": 0.0002113067273011334, + "learning_rate": 0.0002055, + "loss": 0.0, + "step": 1370 + }, + { + "epoch": 0.6451764705882352, + "grad_norm": 0.11955559953234549, + "learning_rate": 0.00020564999999999998, + "loss": 0.0421, + "step": 1371 + }, + { + "epoch": 0.6456470588235295, + "grad_norm": 0.1051594381179474, + "learning_rate": 0.0002058, + "loss": 0.0406, + "step": 1372 + }, + { + "epoch": 0.6461176470588236, + "grad_norm": 0.07274535589000908, + "learning_rate": 0.00020595, + "loss": 0.0245, + "step": 1373 + }, + { + "epoch": 0.6465882352941177, + "grad_norm": 0.00010089033859252272, + "learning_rate": 0.0002061, + "loss": 0.0, + "step": 1374 + }, + { + "epoch": 0.6470588235294118, + "grad_norm": 4.581815352863915e-05, + "learning_rate": 0.00020624999999999997, + "loss": 0.0, + "step": 1375 + }, + { + "epoch": 0.6475294117647059, + "grad_norm": 0.20234499083653776, + "learning_rate": 0.00020639999999999998, + "loss": 0.0984, + "step": 1376 + }, + { + "epoch": 0.648, + "grad_norm": 0.20045210811313782, + "learning_rate": 0.00020654999999999998, + "loss": 0.1165, + "step": 1377 + }, + { + "epoch": 0.6484705882352941, + "grad_norm": 0.09205071068864366, + "learning_rate": 0.00020669999999999996, + "loss": 0.0201, + "step": 1378 + }, + { + "epoch": 0.6489411764705882, + "grad_norm": 0.05239455055051824, + "learning_rate": 0.00020684999999999999, + "loss": 0.0139, + "step": 1379 + }, + { + "epoch": 0.6494117647058824, + "grad_norm": 7.020515204150633e-05, + "learning_rate": 0.00020699999999999996, + "loss": 0.0, + "step": 1380 + }, + { + "epoch": 0.6498823529411765, + "grad_norm": 3.554124018614268e-05, + "learning_rate": 0.00020715, + "loss": 0.0, + "step": 1381 + }, + { + "epoch": 0.6503529411764706, + "grad_norm": 0.15962888611583598, + "learning_rate": 0.00020729999999999997, + "loss": 0.0618, + "step": 1382 + }, + { + "epoch": 0.6508235294117647, + "grad_norm": 0.14027029644705852, + "learning_rate": 0.00020744999999999997, + "loss": 0.0364, + "step": 1383 + }, + { + "epoch": 0.6512941176470588, + "grad_norm": 0.0616292036666949, + "learning_rate": 0.00020759999999999998, + "loss": 0.0122, + "step": 1384 + }, + { + "epoch": 0.6517647058823529, + "grad_norm": 0.07039486913183424, + "learning_rate": 0.00020774999999999998, + "loss": 0.0157, + "step": 1385 + }, + { + "epoch": 0.652235294117647, + "grad_norm": 0.30958047350126194, + "learning_rate": 0.00020789999999999996, + "loss": 0.1506, + "step": 1386 + }, + { + "epoch": 0.6527058823529411, + "grad_norm": 0.06445251837171687, + "learning_rate": 0.00020805, + "loss": 0.0167, + "step": 1387 + }, + { + "epoch": 0.6531764705882352, + "grad_norm": 0.0003137619722552894, + "learning_rate": 0.00020819999999999996, + "loss": 0.0, + "step": 1388 + }, + { + "epoch": 0.6536470588235294, + "grad_norm": 0.23082233500209462, + "learning_rate": 0.00020835, + "loss": 0.1057, + "step": 1389 + }, + { + "epoch": 0.6541176470588236, + "grad_norm": 0.07529794388072791, + "learning_rate": 0.00020849999999999997, + "loss": 0.0237, + "step": 1390 + }, + { + "epoch": 0.6545882352941177, + "grad_norm": 0.0002806465047300114, + "learning_rate": 0.00020864999999999998, + "loss": 0.0, + "step": 1391 + }, + { + "epoch": 0.6550588235294118, + "grad_norm": 0.23886259639341154, + "learning_rate": 0.00020879999999999998, + "loss": 0.1319, + "step": 1392 + }, + { + "epoch": 0.6555294117647059, + "grad_norm": 0.03967271039019128, + "learning_rate": 0.00020894999999999998, + "loss": 0.0054, + "step": 1393 + }, + { + "epoch": 0.656, + "grad_norm": 0.21075758547592946, + "learning_rate": 0.00020909999999999996, + "loss": 0.0574, + "step": 1394 + }, + { + "epoch": 0.6564705882352941, + "grad_norm": 0.024072109912282823, + "learning_rate": 0.00020925, + "loss": 0.0007, + "step": 1395 + }, + { + "epoch": 0.6569411764705883, + "grad_norm": 0.16140390983120864, + "learning_rate": 0.00020939999999999997, + "loss": 0.0621, + "step": 1396 + }, + { + "epoch": 0.6574117647058824, + "grad_norm": 0.11237347666057504, + "learning_rate": 0.00020955, + "loss": 0.0309, + "step": 1397 + }, + { + "epoch": 0.6578823529411765, + "grad_norm": 0.11700500620023965, + "learning_rate": 0.00020969999999999997, + "loss": 0.0154, + "step": 1398 + }, + { + "epoch": 0.6583529411764706, + "grad_norm": 0.23589818355052364, + "learning_rate": 0.00020984999999999998, + "loss": 0.0253, + "step": 1399 + }, + { + "epoch": 0.6588235294117647, + "grad_norm": 0.12480466379236772, + "learning_rate": 0.00020999999999999998, + "loss": 0.0447, + "step": 1400 + }, + { + "epoch": 0.6592941176470588, + "grad_norm": 0.0811968464998344, + "learning_rate": 0.00021014999999999999, + "loss": 0.0186, + "step": 1401 + }, + { + "epoch": 0.6597647058823529, + "grad_norm": 0.31527503910022514, + "learning_rate": 0.00021029999999999996, + "loss": 0.1565, + "step": 1402 + }, + { + "epoch": 0.660235294117647, + "grad_norm": 7.682047702085944e-05, + "learning_rate": 0.00021045, + "loss": 0.0, + "step": 1403 + }, + { + "epoch": 0.6607058823529411, + "grad_norm": 0.14374139713755946, + "learning_rate": 0.00021059999999999997, + "loss": 0.0394, + "step": 1404 + }, + { + "epoch": 0.6611764705882353, + "grad_norm": 0.2046226059427851, + "learning_rate": 0.00021074999999999997, + "loss": 0.0684, + "step": 1405 + }, + { + "epoch": 0.6616470588235294, + "grad_norm": 2.176479437542506e-05, + "learning_rate": 0.00021089999999999998, + "loss": 0.0, + "step": 1406 + }, + { + "epoch": 0.6621176470588235, + "grad_norm": 7.888769622974879e-06, + "learning_rate": 0.00021104999999999998, + "loss": 0.0, + "step": 1407 + }, + { + "epoch": 0.6625882352941177, + "grad_norm": 0.00044026955340373627, + "learning_rate": 0.00021119999999999996, + "loss": 0.0, + "step": 1408 + }, + { + "epoch": 0.6630588235294118, + "grad_norm": 8.853136522115703e-06, + "learning_rate": 0.00021135, + "loss": 0.0, + "step": 1409 + }, + { + "epoch": 0.6635294117647059, + "grad_norm": 5.0183577881594614e-06, + "learning_rate": 0.00021149999999999996, + "loss": 0.0, + "step": 1410 + }, + { + "epoch": 0.664, + "grad_norm": 2.175265440034365e-06, + "learning_rate": 0.00021165, + "loss": 0.0, + "step": 1411 + }, + { + "epoch": 0.6644705882352941, + "grad_norm": 0.3382707790791311, + "learning_rate": 0.00021179999999999997, + "loss": 0.072, + "step": 1412 + }, + { + "epoch": 0.6649411764705883, + "grad_norm": 0.6163029199173325, + "learning_rate": 0.00021194999999999997, + "loss": 0.1092, + "step": 1413 + }, + { + "epoch": 0.6654117647058824, + "grad_norm": 5.2230194801708645e-06, + "learning_rate": 0.00021209999999999998, + "loss": 0.0, + "step": 1414 + }, + { + "epoch": 0.6658823529411765, + "grad_norm": 1.0570486257999104e-05, + "learning_rate": 0.00021224999999999998, + "loss": 0.0, + "step": 1415 + }, + { + "epoch": 0.6663529411764706, + "grad_norm": 0.09610792956877183, + "learning_rate": 0.00021239999999999996, + "loss": 0.0182, + "step": 1416 + }, + { + "epoch": 0.6668235294117647, + "grad_norm": 0.07263503537235129, + "learning_rate": 0.00021255, + "loss": 0.0121, + "step": 1417 + }, + { + "epoch": 0.6672941176470588, + "grad_norm": 0.2852308867377539, + "learning_rate": 0.00021269999999999997, + "loss": 0.0972, + "step": 1418 + }, + { + "epoch": 0.6677647058823529, + "grad_norm": 0.19399453300175626, + "learning_rate": 0.00021285, + "loss": 0.0601, + "step": 1419 + }, + { + "epoch": 0.668235294117647, + "grad_norm": 0.003197706464118584, + "learning_rate": 0.00021299999999999997, + "loss": 0.0001, + "step": 1420 + }, + { + "epoch": 0.6687058823529411, + "grad_norm": 0.09927895361313985, + "learning_rate": 0.00021314999999999998, + "loss": 0.0307, + "step": 1421 + }, + { + "epoch": 0.6691764705882353, + "grad_norm": 0.38902650364596997, + "learning_rate": 0.00021329999999999998, + "loss": 0.1695, + "step": 1422 + }, + { + "epoch": 0.6696470588235294, + "grad_norm": 0.0009300074405148168, + "learning_rate": 0.00021344999999999998, + "loss": 0.0, + "step": 1423 + }, + { + "epoch": 0.6701176470588235, + "grad_norm": 0.2238644530127365, + "learning_rate": 0.00021359999999999996, + "loss": 0.0804, + "step": 1424 + }, + { + "epoch": 0.6705882352941176, + "grad_norm": 0.09810833142001404, + "learning_rate": 0.00021375, + "loss": 0.0251, + "step": 1425 + }, + { + "epoch": 0.6710588235294118, + "grad_norm": 0.08066497002507049, + "learning_rate": 0.00021389999999999997, + "loss": 0.0205, + "step": 1426 + }, + { + "epoch": 0.6715294117647059, + "grad_norm": 0.02097116866386786, + "learning_rate": 0.00021405, + "loss": 0.0021, + "step": 1427 + }, + { + "epoch": 0.672, + "grad_norm": 0.13361664788343436, + "learning_rate": 0.00021419999999999998, + "loss": 0.0542, + "step": 1428 + }, + { + "epoch": 0.6724705882352942, + "grad_norm": 0.148930346970496, + "learning_rate": 0.00021434999999999998, + "loss": 0.0719, + "step": 1429 + }, + { + "epoch": 0.6729411764705883, + "grad_norm": 0.12518242009826472, + "learning_rate": 0.00021449999999999998, + "loss": 0.0476, + "step": 1430 + }, + { + "epoch": 0.6734117647058824, + "grad_norm": 0.14529524456285942, + "learning_rate": 0.00021464999999999999, + "loss": 0.0616, + "step": 1431 + }, + { + "epoch": 0.6738823529411765, + "grad_norm": 0.10813448292768804, + "learning_rate": 0.00021479999999999996, + "loss": 0.042, + "step": 1432 + }, + { + "epoch": 0.6743529411764706, + "grad_norm": 0.13103679427398426, + "learning_rate": 0.00021495, + "loss": 0.0609, + "step": 1433 + }, + { + "epoch": 0.6748235294117647, + "grad_norm": 0.08494313295535734, + "learning_rate": 0.00021509999999999997, + "loss": 0.0217, + "step": 1434 + }, + { + "epoch": 0.6752941176470588, + "grad_norm": 0.1799211146620175, + "learning_rate": 0.00021525, + "loss": 0.1023, + "step": 1435 + }, + { + "epoch": 0.6757647058823529, + "grad_norm": 7.238512639981366e-06, + "learning_rate": 0.00021539999999999998, + "loss": 0.0, + "step": 1436 + }, + { + "epoch": 0.676235294117647, + "grad_norm": 0.16083044609033584, + "learning_rate": 0.00021554999999999998, + "loss": 0.0469, + "step": 1437 + }, + { + "epoch": 0.6767058823529412, + "grad_norm": 0.06384501831140896, + "learning_rate": 0.00021569999999999998, + "loss": 0.0117, + "step": 1438 + }, + { + "epoch": 0.6771764705882353, + "grad_norm": 0.18605481747830516, + "learning_rate": 0.00021585, + "loss": 0.0761, + "step": 1439 + }, + { + "epoch": 0.6776470588235294, + "grad_norm": 0.18784164092177924, + "learning_rate": 0.00021599999999999996, + "loss": 0.0406, + "step": 1440 + }, + { + "epoch": 0.6781176470588235, + "grad_norm": 0.241504700554367, + "learning_rate": 0.00021615, + "loss": 0.0951, + "step": 1441 + }, + { + "epoch": 0.6785882352941176, + "grad_norm": 0.10089152574381087, + "learning_rate": 0.00021629999999999997, + "loss": 0.0308, + "step": 1442 + }, + { + "epoch": 0.6790588235294117, + "grad_norm": 1.4405428166615083e-06, + "learning_rate": 0.00021645, + "loss": 0.0, + "step": 1443 + }, + { + "epoch": 0.6795294117647059, + "grad_norm": 0.05697159025845361, + "learning_rate": 0.00021659999999999998, + "loss": 0.0086, + "step": 1444 + }, + { + "epoch": 0.68, + "grad_norm": 0.2414405071324109, + "learning_rate": 0.00021674999999999998, + "loss": 0.1138, + "step": 1445 + }, + { + "epoch": 0.6804705882352942, + "grad_norm": 0.13620894234193082, + "learning_rate": 0.0002169, + "loss": 0.0368, + "step": 1446 + }, + { + "epoch": 0.6809411764705883, + "grad_norm": 0.0589682199763565, + "learning_rate": 0.00021705, + "loss": 0.0126, + "step": 1447 + }, + { + "epoch": 0.6814117647058824, + "grad_norm": 0.00021763186409042064, + "learning_rate": 0.00021719999999999997, + "loss": 0.0, + "step": 1448 + }, + { + "epoch": 0.6818823529411765, + "grad_norm": 0.1490721577520266, + "learning_rate": 0.00021735, + "loss": 0.0394, + "step": 1449 + }, + { + "epoch": 0.6823529411764706, + "grad_norm": 0.00015486522060798144, + "learning_rate": 0.00021749999999999997, + "loss": 0.0, + "step": 1450 + }, + { + "epoch": 0.6828235294117647, + "grad_norm": 7.133872309384504e-07, + "learning_rate": 0.00021764999999999998, + "loss": 0.0, + "step": 1451 + }, + { + "epoch": 0.6832941176470588, + "grad_norm": 0.20811493260199532, + "learning_rate": 0.00021779999999999998, + "loss": 0.0742, + "step": 1452 + }, + { + "epoch": 0.6837647058823529, + "grad_norm": 0.11248573933576395, + "learning_rate": 0.00021794999999999999, + "loss": 0.0485, + "step": 1453 + }, + { + "epoch": 0.684235294117647, + "grad_norm": 0.13079261663880953, + "learning_rate": 0.00021809999999999996, + "loss": 0.0559, + "step": 1454 + }, + { + "epoch": 0.6847058823529412, + "grad_norm": 0.17605280616365315, + "learning_rate": 0.00021825, + "loss": 0.0631, + "step": 1455 + }, + { + "epoch": 0.6851764705882353, + "grad_norm": 0.08046884361926764, + "learning_rate": 0.00021839999999999997, + "loss": 0.0293, + "step": 1456 + }, + { + "epoch": 0.6856470588235294, + "grad_norm": 0.11119147954145579, + "learning_rate": 0.00021855, + "loss": 0.0421, + "step": 1457 + }, + { + "epoch": 0.6861176470588235, + "grad_norm": 0.14113677913801181, + "learning_rate": 0.00021869999999999998, + "loss": 0.0532, + "step": 1458 + }, + { + "epoch": 0.6865882352941176, + "grad_norm": 0.11860129371632253, + "learning_rate": 0.00021884999999999998, + "loss": 0.044, + "step": 1459 + }, + { + "epoch": 0.6870588235294117, + "grad_norm": 8.380646729753888e-06, + "learning_rate": 0.00021899999999999998, + "loss": 0.0, + "step": 1460 + }, + { + "epoch": 0.6875294117647058, + "grad_norm": 0.05289614341597529, + "learning_rate": 0.00021915, + "loss": 0.0081, + "step": 1461 + }, + { + "epoch": 0.688, + "grad_norm": 0.12649527093856097, + "learning_rate": 0.00021929999999999996, + "loss": 0.0466, + "step": 1462 + }, + { + "epoch": 0.6884705882352942, + "grad_norm": 0.23162966835667828, + "learning_rate": 0.00021945, + "loss": 0.106, + "step": 1463 + }, + { + "epoch": 0.6889411764705883, + "grad_norm": 8.55709419830328e-06, + "learning_rate": 0.00021959999999999997, + "loss": 0.0, + "step": 1464 + }, + { + "epoch": 0.6894117647058824, + "grad_norm": 0.20040411539300265, + "learning_rate": 0.00021975, + "loss": 0.0852, + "step": 1465 + }, + { + "epoch": 0.6898823529411765, + "grad_norm": 0.04102744645853634, + "learning_rate": 0.00021989999999999998, + "loss": 0.0075, + "step": 1466 + }, + { + "epoch": 0.6903529411764706, + "grad_norm": 0.0743586396013501, + "learning_rate": 0.00022004999999999998, + "loss": 0.023, + "step": 1467 + }, + { + "epoch": 0.6908235294117647, + "grad_norm": 0.00033718885957864577, + "learning_rate": 0.00022019999999999999, + "loss": 0.0, + "step": 1468 + }, + { + "epoch": 0.6912941176470588, + "grad_norm": 0.04785558633494389, + "learning_rate": 0.00022035, + "loss": 0.0101, + "step": 1469 + }, + { + "epoch": 0.691764705882353, + "grad_norm": 0.06925611497375418, + "learning_rate": 0.00022049999999999997, + "loss": 0.0144, + "step": 1470 + }, + { + "epoch": 0.6922352941176471, + "grad_norm": 0.22584188082445317, + "learning_rate": 0.00022065, + "loss": 0.0756, + "step": 1471 + }, + { + "epoch": 0.6927058823529412, + "grad_norm": 0.05806146551970872, + "learning_rate": 0.00022079999999999997, + "loss": 0.0172, + "step": 1472 + }, + { + "epoch": 0.6931764705882353, + "grad_norm": 0.1498528912194676, + "learning_rate": 0.00022095, + "loss": 0.0723, + "step": 1473 + }, + { + "epoch": 0.6936470588235294, + "grad_norm": 6.299562077503457e-05, + "learning_rate": 0.00022109999999999998, + "loss": 0.0, + "step": 1474 + }, + { + "epoch": 0.6941176470588235, + "grad_norm": 0.24141240337308983, + "learning_rate": 0.00022124999999999998, + "loss": 0.1392, + "step": 1475 + }, + { + "epoch": 0.6945882352941176, + "grad_norm": 0.1012280921494038, + "learning_rate": 0.0002214, + "loss": 0.0354, + "step": 1476 + }, + { + "epoch": 0.6950588235294117, + "grad_norm": 0.08659474803689388, + "learning_rate": 0.00022155, + "loss": 0.0166, + "step": 1477 + }, + { + "epoch": 0.6955294117647058, + "grad_norm": 0.09616216287622786, + "learning_rate": 0.00022169999999999997, + "loss": 0.0297, + "step": 1478 + }, + { + "epoch": 0.696, + "grad_norm": 0.07657509554536059, + "learning_rate": 0.00022185, + "loss": 0.0161, + "step": 1479 + }, + { + "epoch": 0.6964705882352941, + "grad_norm": 0.04925769955090059, + "learning_rate": 0.00022199999999999998, + "loss": 0.0128, + "step": 1480 + }, + { + "epoch": 0.6969411764705883, + "grad_norm": 0.049547849719350265, + "learning_rate": 0.00022215, + "loss": 0.0095, + "step": 1481 + }, + { + "epoch": 0.6974117647058824, + "grad_norm": 0.09710334604833032, + "learning_rate": 0.00022229999999999998, + "loss": 0.0179, + "step": 1482 + }, + { + "epoch": 0.6978823529411765, + "grad_norm": 1.8812080456142754e-05, + "learning_rate": 0.00022244999999999999, + "loss": 0.0, + "step": 1483 + }, + { + "epoch": 0.6983529411764706, + "grad_norm": 0.13306763508047983, + "learning_rate": 0.0002226, + "loss": 0.0529, + "step": 1484 + }, + { + "epoch": 0.6988235294117647, + "grad_norm": 0.07413981414483023, + "learning_rate": 0.00022275, + "loss": 0.0191, + "step": 1485 + }, + { + "epoch": 0.6992941176470588, + "grad_norm": 0.2061120415344631, + "learning_rate": 0.00022289999999999997, + "loss": 0.086, + "step": 1486 + }, + { + "epoch": 0.699764705882353, + "grad_norm": 0.12759350396856842, + "learning_rate": 0.00022305, + "loss": 0.0402, + "step": 1487 + }, + { + "epoch": 0.7002352941176471, + "grad_norm": 0.1832223964608256, + "learning_rate": 0.00022319999999999998, + "loss": 0.0861, + "step": 1488 + }, + { + "epoch": 0.7007058823529412, + "grad_norm": 0.056814908762696396, + "learning_rate": 0.00022335, + "loss": 0.0193, + "step": 1489 + }, + { + "epoch": 0.7011764705882353, + "grad_norm": 0.01667190295673713, + "learning_rate": 0.00022349999999999998, + "loss": 0.0009, + "step": 1490 + }, + { + "epoch": 0.7016470588235294, + "grad_norm": 0.06267099379019328, + "learning_rate": 0.00022365, + "loss": 0.0077, + "step": 1491 + }, + { + "epoch": 0.7021176470588235, + "grad_norm": 0.07158220352803588, + "learning_rate": 0.0002238, + "loss": 0.0117, + "step": 1492 + }, + { + "epoch": 0.7025882352941176, + "grad_norm": 0.10552275621517034, + "learning_rate": 0.00022395, + "loss": 0.0551, + "step": 1493 + }, + { + "epoch": 0.7030588235294117, + "grad_norm": 0.21237293870861793, + "learning_rate": 0.00022409999999999997, + "loss": 0.0975, + "step": 1494 + }, + { + "epoch": 0.7035294117647058, + "grad_norm": 0.07591204124895463, + "learning_rate": 0.00022425, + "loss": 0.0082, + "step": 1495 + }, + { + "epoch": 0.704, + "grad_norm": 0.43155247448270084, + "learning_rate": 0.00022439999999999998, + "loss": 0.0495, + "step": 1496 + }, + { + "epoch": 0.7044705882352941, + "grad_norm": 0.20906831669435724, + "learning_rate": 0.00022455, + "loss": 0.0274, + "step": 1497 + }, + { + "epoch": 0.7049411764705882, + "grad_norm": 0.18831128303776343, + "learning_rate": 0.0002247, + "loss": 0.0849, + "step": 1498 + }, + { + "epoch": 0.7054117647058824, + "grad_norm": 0.06918337600922445, + "learning_rate": 0.00022485, + "loss": 0.0173, + "step": 1499 + }, + { + "epoch": 0.7058823529411765, + "grad_norm": 0.2494435830154463, + "learning_rate": 0.000225, + "loss": 0.1321, + "step": 1500 + }, + { + "epoch": 0.7063529411764706, + "grad_norm": 0.09794737080193516, + "learning_rate": 0.00022514999999999997, + "loss": 0.0434, + "step": 1501 + }, + { + "epoch": 0.7068235294117647, + "grad_norm": 4.8709394473309416e-05, + "learning_rate": 0.00022529999999999997, + "loss": 0.0, + "step": 1502 + }, + { + "epoch": 0.7072941176470589, + "grad_norm": 0.29516186848658393, + "learning_rate": 0.00022544999999999995, + "loss": 0.1282, + "step": 1503 + }, + { + "epoch": 0.707764705882353, + "grad_norm": 0.09946149958402628, + "learning_rate": 0.00022559999999999998, + "loss": 0.0277, + "step": 1504 + }, + { + "epoch": 0.7082352941176471, + "grad_norm": 0.07676468806066213, + "learning_rate": 0.00022574999999999996, + "loss": 0.0061, + "step": 1505 + }, + { + "epoch": 0.7087058823529412, + "grad_norm": 0.02654506130810427, + "learning_rate": 0.0002259, + "loss": 0.0037, + "step": 1506 + }, + { + "epoch": 0.7091764705882353, + "grad_norm": 0.06542347901474772, + "learning_rate": 0.00022604999999999997, + "loss": 0.0149, + "step": 1507 + }, + { + "epoch": 0.7096470588235294, + "grad_norm": 0.06447790556800906, + "learning_rate": 0.00022619999999999997, + "loss": 0.0129, + "step": 1508 + }, + { + "epoch": 0.7101176470588235, + "grad_norm": 0.06915436754640227, + "learning_rate": 0.00022634999999999997, + "loss": 0.0183, + "step": 1509 + }, + { + "epoch": 0.7105882352941176, + "grad_norm": 0.17865557691608658, + "learning_rate": 0.00022649999999999998, + "loss": 0.094, + "step": 1510 + }, + { + "epoch": 0.7110588235294117, + "grad_norm": 1.544806447366367e-05, + "learning_rate": 0.00022664999999999995, + "loss": 0.0, + "step": 1511 + }, + { + "epoch": 0.7115294117647059, + "grad_norm": 0.18739701186321234, + "learning_rate": 0.00022679999999999998, + "loss": 0.0731, + "step": 1512 + }, + { + "epoch": 0.712, + "grad_norm": 0.48931723124111914, + "learning_rate": 0.00022694999999999996, + "loss": 0.1329, + "step": 1513 + }, + { + "epoch": 0.7124705882352941, + "grad_norm": 0.17228270661011602, + "learning_rate": 0.0002271, + "loss": 0.0886, + "step": 1514 + }, + { + "epoch": 0.7129411764705882, + "grad_norm": 0.11799324922824128, + "learning_rate": 0.00022724999999999997, + "loss": 0.0393, + "step": 1515 + }, + { + "epoch": 0.7134117647058823, + "grad_norm": 0.0864311863651394, + "learning_rate": 0.00022739999999999997, + "loss": 0.0278, + "step": 1516 + }, + { + "epoch": 0.7138823529411765, + "grad_norm": 0.00044149653737338704, + "learning_rate": 0.00022754999999999997, + "loss": 0.0, + "step": 1517 + }, + { + "epoch": 0.7143529411764706, + "grad_norm": 0.0008130666592499984, + "learning_rate": 0.00022769999999999998, + "loss": 0.0, + "step": 1518 + }, + { + "epoch": 0.7148235294117647, + "grad_norm": 0.04413074630603585, + "learning_rate": 0.00022784999999999995, + "loss": 0.0057, + "step": 1519 + }, + { + "epoch": 0.7152941176470589, + "grad_norm": 0.11251891418654557, + "learning_rate": 0.00022799999999999999, + "loss": 0.0402, + "step": 1520 + }, + { + "epoch": 0.715764705882353, + "grad_norm": 0.30876342310807237, + "learning_rate": 0.00022814999999999996, + "loss": 0.003, + "step": 1521 + }, + { + "epoch": 0.7162352941176471, + "grad_norm": 0.1630309291094905, + "learning_rate": 0.0002283, + "loss": 0.0558, + "step": 1522 + }, + { + "epoch": 0.7167058823529412, + "grad_norm": 0.10988826834042703, + "learning_rate": 0.00022844999999999997, + "loss": 0.0171, + "step": 1523 + }, + { + "epoch": 0.7171764705882353, + "grad_norm": 0.00017398816639517168, + "learning_rate": 0.00022859999999999997, + "loss": 0.0, + "step": 1524 + }, + { + "epoch": 0.7176470588235294, + "grad_norm": 0.17519237948307614, + "learning_rate": 0.00022874999999999998, + "loss": 0.0619, + "step": 1525 + }, + { + "epoch": 0.7181176470588235, + "grad_norm": 0.06499030063307314, + "learning_rate": 0.00022889999999999998, + "loss": 0.0123, + "step": 1526 + }, + { + "epoch": 0.7185882352941176, + "grad_norm": 0.09794678936349069, + "learning_rate": 0.00022904999999999996, + "loss": 0.0241, + "step": 1527 + }, + { + "epoch": 0.7190588235294118, + "grad_norm": 0.09503989691564818, + "learning_rate": 0.0002292, + "loss": 0.0226, + "step": 1528 + }, + { + "epoch": 0.7195294117647059, + "grad_norm": 0.12545699422080506, + "learning_rate": 0.00022934999999999996, + "loss": 0.0362, + "step": 1529 + }, + { + "epoch": 0.72, + "grad_norm": 0.056021114797964025, + "learning_rate": 0.0002295, + "loss": 0.0083, + "step": 1530 + }, + { + "epoch": 0.7204705882352941, + "grad_norm": 0.12884126198190124, + "learning_rate": 0.00022964999999999997, + "loss": 0.0347, + "step": 1531 + }, + { + "epoch": 0.7209411764705882, + "grad_norm": 0.11391620627256206, + "learning_rate": 0.00022979999999999997, + "loss": 0.0233, + "step": 1532 + }, + { + "epoch": 0.7214117647058823, + "grad_norm": 6.775942817827348e-06, + "learning_rate": 0.00022994999999999998, + "loss": 0.0, + "step": 1533 + }, + { + "epoch": 0.7218823529411764, + "grad_norm": 0.00022144844322159746, + "learning_rate": 0.00023009999999999998, + "loss": 0.0, + "step": 1534 + }, + { + "epoch": 0.7223529411764706, + "grad_norm": 0.16640771559749773, + "learning_rate": 0.00023024999999999996, + "loss": 0.0235, + "step": 1535 + }, + { + "epoch": 0.7228235294117648, + "grad_norm": 0.09987410218892209, + "learning_rate": 0.0002304, + "loss": 0.016, + "step": 1536 + }, + { + "epoch": 0.7232941176470589, + "grad_norm": 0.2505816586210855, + "learning_rate": 0.00023054999999999997, + "loss": 0.0776, + "step": 1537 + }, + { + "epoch": 0.723764705882353, + "grad_norm": 0.12016236386302828, + "learning_rate": 0.0002307, + "loss": 0.0206, + "step": 1538 + }, + { + "epoch": 0.7242352941176471, + "grad_norm": 0.0709190530407887, + "learning_rate": 0.00023084999999999997, + "loss": 0.0058, + "step": 1539 + }, + { + "epoch": 0.7247058823529412, + "grad_norm": 0.13293692710622682, + "learning_rate": 0.00023099999999999998, + "loss": 0.0506, + "step": 1540 + }, + { + "epoch": 0.7251764705882353, + "grad_norm": 0.07209688241358865, + "learning_rate": 0.00023114999999999998, + "loss": 0.0109, + "step": 1541 + }, + { + "epoch": 0.7256470588235294, + "grad_norm": 0.049439839589434285, + "learning_rate": 0.00023129999999999998, + "loss": 0.0048, + "step": 1542 + }, + { + "epoch": 0.7261176470588235, + "grad_norm": 0.0713122018926105, + "learning_rate": 0.00023144999999999996, + "loss": 0.0091, + "step": 1543 + }, + { + "epoch": 0.7265882352941176, + "grad_norm": 0.048951836996677434, + "learning_rate": 0.0002316, + "loss": 0.0135, + "step": 1544 + }, + { + "epoch": 0.7270588235294118, + "grad_norm": 0.21330790762326046, + "learning_rate": 0.00023174999999999997, + "loss": 0.0866, + "step": 1545 + }, + { + "epoch": 0.7275294117647059, + "grad_norm": 0.20159589868440428, + "learning_rate": 0.0002319, + "loss": 0.1044, + "step": 1546 + }, + { + "epoch": 0.728, + "grad_norm": 6.80762832974374e-07, + "learning_rate": 0.00023204999999999998, + "loss": 0.0, + "step": 1547 + }, + { + "epoch": 0.7284705882352941, + "grad_norm": 2.505540981571707e-05, + "learning_rate": 0.00023219999999999998, + "loss": 0.0, + "step": 1548 + }, + { + "epoch": 0.7289411764705882, + "grad_norm": 0.08017248480394178, + "learning_rate": 0.00023234999999999998, + "loss": 0.0151, + "step": 1549 + }, + { + "epoch": 0.7294117647058823, + "grad_norm": 0.08320687396806604, + "learning_rate": 0.00023249999999999999, + "loss": 0.0187, + "step": 1550 + }, + { + "epoch": 0.7298823529411764, + "grad_norm": 0.12934036992721612, + "learning_rate": 0.00023264999999999996, + "loss": 0.0453, + "step": 1551 + }, + { + "epoch": 0.7303529411764705, + "grad_norm": 0.08279892467228812, + "learning_rate": 0.0002328, + "loss": 0.0152, + "step": 1552 + }, + { + "epoch": 0.7308235294117648, + "grad_norm": 0.06424672969567695, + "learning_rate": 0.00023294999999999997, + "loss": 0.0153, + "step": 1553 + }, + { + "epoch": 0.7312941176470589, + "grad_norm": 2.861097220435327e-06, + "learning_rate": 0.00023309999999999997, + "loss": 0.0, + "step": 1554 + }, + { + "epoch": 0.731764705882353, + "grad_norm": 2.0736173894803084e-05, + "learning_rate": 0.00023324999999999998, + "loss": 0.0, + "step": 1555 + }, + { + "epoch": 0.7322352941176471, + "grad_norm": 0.1362417124348537, + "learning_rate": 0.00023339999999999998, + "loss": 0.0245, + "step": 1556 + }, + { + "epoch": 0.7327058823529412, + "grad_norm": 0.2339561168540705, + "learning_rate": 0.00023354999999999996, + "loss": 0.1104, + "step": 1557 + }, + { + "epoch": 0.7331764705882353, + "grad_norm": 0.14214719292593578, + "learning_rate": 0.0002337, + "loss": 0.0346, + "step": 1558 + }, + { + "epoch": 0.7336470588235294, + "grad_norm": 0.00019600941019585615, + "learning_rate": 0.00023384999999999997, + "loss": 0.0, + "step": 1559 + }, + { + "epoch": 0.7341176470588235, + "grad_norm": 0.058822544220407905, + "learning_rate": 0.000234, + "loss": 0.0083, + "step": 1560 + }, + { + "epoch": 0.7345882352941177, + "grad_norm": 5.803896012372221e-06, + "learning_rate": 0.00023414999999999997, + "loss": 0.0, + "step": 1561 + }, + { + "epoch": 0.7350588235294118, + "grad_norm": 0.18221042943894505, + "learning_rate": 0.00023429999999999998, + "loss": 0.056, + "step": 1562 + }, + { + "epoch": 0.7355294117647059, + "grad_norm": 1.5151981274153052e-05, + "learning_rate": 0.00023444999999999998, + "loss": 0.0, + "step": 1563 + }, + { + "epoch": 0.736, + "grad_norm": 0.06791627282206578, + "learning_rate": 0.00023459999999999998, + "loss": 0.0145, + "step": 1564 + }, + { + "epoch": 0.7364705882352941, + "grad_norm": 0.1060131351245245, + "learning_rate": 0.00023474999999999996, + "loss": 0.0128, + "step": 1565 + }, + { + "epoch": 0.7369411764705882, + "grad_norm": 5.536129371662512e-05, + "learning_rate": 0.0002349, + "loss": 0.0, + "step": 1566 + }, + { + "epoch": 0.7374117647058823, + "grad_norm": 0.06529522027486744, + "learning_rate": 0.00023504999999999997, + "loss": 0.0067, + "step": 1567 + }, + { + "epoch": 0.7378823529411764, + "grad_norm": 0.3515761602905019, + "learning_rate": 0.0002352, + "loss": 0.1372, + "step": 1568 + }, + { + "epoch": 0.7383529411764705, + "grad_norm": 0.15707789846757259, + "learning_rate": 0.00023534999999999997, + "loss": 0.0322, + "step": 1569 + }, + { + "epoch": 0.7388235294117647, + "grad_norm": 0.20256112165777193, + "learning_rate": 0.00023549999999999998, + "loss": 0.0966, + "step": 1570 + }, + { + "epoch": 0.7392941176470588, + "grad_norm": 0.04390318945391368, + "learning_rate": 0.00023564999999999998, + "loss": 0.0106, + "step": 1571 + }, + { + "epoch": 0.739764705882353, + "grad_norm": 0.08726628864308598, + "learning_rate": 0.00023579999999999999, + "loss": 0.0213, + "step": 1572 + }, + { + "epoch": 0.7402352941176471, + "grad_norm": 0.15560778748161477, + "learning_rate": 0.00023594999999999996, + "loss": 0.0476, + "step": 1573 + }, + { + "epoch": 0.7407058823529412, + "grad_norm": 0.04520475805695719, + "learning_rate": 0.0002361, + "loss": 0.002, + "step": 1574 + }, + { + "epoch": 0.7411764705882353, + "grad_norm": 0.16316866479836545, + "learning_rate": 0.00023624999999999997, + "loss": 0.0757, + "step": 1575 + }, + { + "epoch": 0.7416470588235294, + "grad_norm": 0.18870475562320912, + "learning_rate": 0.0002364, + "loss": 0.1026, + "step": 1576 + }, + { + "epoch": 0.7421176470588235, + "grad_norm": 0.06452751882017066, + "learning_rate": 0.00023654999999999998, + "loss": 0.0151, + "step": 1577 + }, + { + "epoch": 0.7425882352941177, + "grad_norm": 0.12025979987542047, + "learning_rate": 0.00023669999999999998, + "loss": 0.0803, + "step": 1578 + }, + { + "epoch": 0.7430588235294118, + "grad_norm": 0.3228684474474286, + "learning_rate": 0.00023684999999999998, + "loss": 0.1639, + "step": 1579 + }, + { + "epoch": 0.7435294117647059, + "grad_norm": 8.362460422070641e-05, + "learning_rate": 0.000237, + "loss": 0.0, + "step": 1580 + }, + { + "epoch": 0.744, + "grad_norm": 0.0448037727025677, + "learning_rate": 0.00023714999999999996, + "loss": 0.0071, + "step": 1581 + }, + { + "epoch": 0.7444705882352941, + "grad_norm": 7.915955282221005e-05, + "learning_rate": 0.0002373, + "loss": 0.0, + "step": 1582 + }, + { + "epoch": 0.7449411764705882, + "grad_norm": 0.16364030875038746, + "learning_rate": 0.00023744999999999997, + "loss": 0.0802, + "step": 1583 + }, + { + "epoch": 0.7454117647058823, + "grad_norm": 0.09298421893571432, + "learning_rate": 0.0002376, + "loss": 0.0332, + "step": 1584 + }, + { + "epoch": 0.7458823529411764, + "grad_norm": 0.13447451703130694, + "learning_rate": 0.00023774999999999998, + "loss": 0.0497, + "step": 1585 + }, + { + "epoch": 0.7463529411764706, + "grad_norm": 0.10613194897712953, + "learning_rate": 0.00023789999999999998, + "loss": 0.0405, + "step": 1586 + }, + { + "epoch": 0.7468235294117647, + "grad_norm": 0.03029057993469312, + "learning_rate": 0.00023804999999999999, + "loss": 0.0053, + "step": 1587 + }, + { + "epoch": 0.7472941176470588, + "grad_norm": 0.08749942545285004, + "learning_rate": 0.0002382, + "loss": 0.0336, + "step": 1588 + }, + { + "epoch": 0.7477647058823529, + "grad_norm": 0.054427085112119876, + "learning_rate": 0.00023834999999999997, + "loss": 0.0109, + "step": 1589 + }, + { + "epoch": 0.7482352941176471, + "grad_norm": 0.10565014245511035, + "learning_rate": 0.0002385, + "loss": 0.0512, + "step": 1590 + }, + { + "epoch": 0.7487058823529412, + "grad_norm": 4.273044015115511e-05, + "learning_rate": 0.00023864999999999997, + "loss": 0.0, + "step": 1591 + }, + { + "epoch": 0.7491764705882353, + "grad_norm": 0.00011696844829116109, + "learning_rate": 0.0002388, + "loss": 0.0, + "step": 1592 + }, + { + "epoch": 0.7496470588235294, + "grad_norm": 0.1113783961763305, + "learning_rate": 0.00023894999999999998, + "loss": 0.0452, + "step": 1593 + }, + { + "epoch": 0.7501176470588236, + "grad_norm": 0.00011814622504424889, + "learning_rate": 0.00023909999999999998, + "loss": 0.0, + "step": 1594 + }, + { + "epoch": 0.7505882352941177, + "grad_norm": 0.0008811581697106151, + "learning_rate": 0.00023925, + "loss": 0.0, + "step": 1595 + }, + { + "epoch": 0.7510588235294118, + "grad_norm": 0.22358208526505113, + "learning_rate": 0.0002394, + "loss": 0.1131, + "step": 1596 + }, + { + "epoch": 0.7515294117647059, + "grad_norm": 0.013971228666550978, + "learning_rate": 0.00023954999999999997, + "loss": 0.0008, + "step": 1597 + }, + { + "epoch": 0.752, + "grad_norm": 0.14118332317568763, + "learning_rate": 0.0002397, + "loss": 0.0684, + "step": 1598 + }, + { + "epoch": 0.7524705882352941, + "grad_norm": 0.20377685043852314, + "learning_rate": 0.00023984999999999998, + "loss": 0.099, + "step": 1599 + }, + { + "epoch": 0.7529411764705882, + "grad_norm": 0.0007602797904045065, + "learning_rate": 0.00023999999999999998, + "loss": 0.0, + "step": 1600 + }, + { + "epoch": 0.7534117647058823, + "grad_norm": 0.7961378774274471, + "learning_rate": 0.00024014999999999998, + "loss": 0.1397, + "step": 1601 + }, + { + "epoch": 0.7538823529411764, + "grad_norm": 0.10696966886913595, + "learning_rate": 0.00024029999999999999, + "loss": 0.0227, + "step": 1602 + }, + { + "epoch": 0.7543529411764706, + "grad_norm": 0.1687968742158424, + "learning_rate": 0.00024044999999999996, + "loss": 0.0534, + "step": 1603 + }, + { + "epoch": 0.7548235294117647, + "grad_norm": 0.2607967195147905, + "learning_rate": 0.0002406, + "loss": 0.1086, + "step": 1604 + }, + { + "epoch": 0.7552941176470588, + "grad_norm": 1.1777538205822646e-06, + "learning_rate": 0.00024074999999999997, + "loss": 0.0, + "step": 1605 + }, + { + "epoch": 0.7557647058823529, + "grad_norm": 3.2892649524406563, + "learning_rate": 0.0002409, + "loss": 0.3146, + "step": 1606 + }, + { + "epoch": 0.756235294117647, + "grad_norm": 0.1310769582238832, + "learning_rate": 0.00024104999999999998, + "loss": 0.0315, + "step": 1607 + }, + { + "epoch": 0.7567058823529412, + "grad_norm": 69.803480198251, + "learning_rate": 0.00024119999999999998, + "loss": 3.5565, + "step": 1608 + }, + { + "epoch": 0.7571764705882353, + "grad_norm": 0.08667990756441975, + "learning_rate": 0.00024134999999999998, + "loss": 0.0169, + "step": 1609 + }, + { + "epoch": 0.7576470588235295, + "grad_norm": 1.0695528989966008, + "learning_rate": 0.0002415, + "loss": 0.1673, + "step": 1610 + }, + { + "epoch": 0.7581176470588236, + "grad_norm": 0.5933458249782753, + "learning_rate": 0.00024164999999999996, + "loss": 0.0777, + "step": 1611 + }, + { + "epoch": 0.7585882352941177, + "grad_norm": 2.9785716019692665, + "learning_rate": 0.0002418, + "loss": 0.2763, + "step": 1612 + }, + { + "epoch": 0.7590588235294118, + "grad_norm": 0.30988934626967546, + "learning_rate": 0.00024194999999999997, + "loss": 0.0497, + "step": 1613 + }, + { + "epoch": 0.7595294117647059, + "grad_norm": 0.38735499203836676, + "learning_rate": 0.0002421, + "loss": 0.0709, + "step": 1614 + }, + { + "epoch": 0.76, + "grad_norm": 2.1005335285595413e-09, + "learning_rate": 0.00024224999999999998, + "loss": 0.0, + "step": 1615 + }, + { + "epoch": 0.7604705882352941, + "grad_norm": 0.138444274784216, + "learning_rate": 0.00024239999999999998, + "loss": 0.0441, + "step": 1616 + }, + { + "epoch": 0.7609411764705882, + "grad_norm": 0.16276582372758486, + "learning_rate": 0.00024255, + "loss": 0.0822, + "step": 1617 + }, + { + "epoch": 0.7614117647058823, + "grad_norm": 0.12647437346322704, + "learning_rate": 0.0002427, + "loss": 0.0455, + "step": 1618 + }, + { + "epoch": 0.7618823529411765, + "grad_norm": 0.25791370880122044, + "learning_rate": 0.00024284999999999997, + "loss": 0.0427, + "step": 1619 + }, + { + "epoch": 0.7623529411764706, + "grad_norm": 5.443865612311048, + "learning_rate": 0.000243, + "loss": 0.0334, + "step": 1620 + }, + { + "epoch": 0.7628235294117647, + "grad_norm": 0.39580068797911977, + "learning_rate": 0.00024314999999999997, + "loss": 0.0917, + "step": 1621 + }, + { + "epoch": 0.7632941176470588, + "grad_norm": 0.16989532516774114, + "learning_rate": 0.0002433, + "loss": 0.0297, + "step": 1622 + }, + { + "epoch": 0.7637647058823529, + "grad_norm": 0.11300012914126771, + "learning_rate": 0.00024344999999999998, + "loss": 0.0441, + "step": 1623 + }, + { + "epoch": 0.764235294117647, + "grad_norm": 0.3754044185584836, + "learning_rate": 0.00024359999999999999, + "loss": 0.1075, + "step": 1624 + }, + { + "epoch": 0.7647058823529411, + "grad_norm": 0.06799422658300885, + "learning_rate": 0.00024375, + "loss": 0.0092, + "step": 1625 + }, + { + "epoch": 0.7651764705882353, + "grad_norm": 0.2208828763831871, + "learning_rate": 0.00024389999999999997, + "loss": 0.0457, + "step": 1626 + }, + { + "epoch": 0.7656470588235295, + "grad_norm": 0.06700185279837625, + "learning_rate": 0.00024404999999999997, + "loss": 0.0274, + "step": 1627 + }, + { + "epoch": 0.7661176470588236, + "grad_norm": 0.174698812967978, + "learning_rate": 0.00024419999999999997, + "loss": 0.1055, + "step": 1628 + }, + { + "epoch": 0.7665882352941177, + "grad_norm": 0.20780096424106503, + "learning_rate": 0.00024435, + "loss": 0.1165, + "step": 1629 + }, + { + "epoch": 0.7670588235294118, + "grad_norm": 0.11304078949896373, + "learning_rate": 0.0002445, + "loss": 0.0029, + "step": 1630 + }, + { + "epoch": 0.7675294117647059, + "grad_norm": 0.16057639753097813, + "learning_rate": 0.00024464999999999996, + "loss": 0.0446, + "step": 1631 + }, + { + "epoch": 0.768, + "grad_norm": 0.3330631669747782, + "learning_rate": 0.0002448, + "loss": 0.1182, + "step": 1632 + }, + { + "epoch": 0.7684705882352941, + "grad_norm": 0.2032560053414731, + "learning_rate": 0.00024494999999999996, + "loss": 0.1246, + "step": 1633 + }, + { + "epoch": 0.7689411764705882, + "grad_norm": 0.16865902253226223, + "learning_rate": 0.00024509999999999994, + "loss": 0.0813, + "step": 1634 + }, + { + "epoch": 0.7694117647058824, + "grad_norm": 4.280411961325423e-08, + "learning_rate": 0.00024524999999999997, + "loss": 0.0, + "step": 1635 + }, + { + "epoch": 0.7698823529411765, + "grad_norm": 0.13528202444886847, + "learning_rate": 0.00024539999999999995, + "loss": 0.0421, + "step": 1636 + }, + { + "epoch": 0.7703529411764706, + "grad_norm": 1.208559791528829e-09, + "learning_rate": 0.00024555, + "loss": 0.0, + "step": 1637 + }, + { + "epoch": 0.7708235294117647, + "grad_norm": 1.7021431771700211e-06, + "learning_rate": 0.00024569999999999995, + "loss": 0.0, + "step": 1638 + }, + { + "epoch": 0.7712941176470588, + "grad_norm": 0.33336519775473894, + "learning_rate": 0.00024585, + "loss": 0.11, + "step": 1639 + }, + { + "epoch": 0.7717647058823529, + "grad_norm": 0.0734926414186909, + "learning_rate": 0.00024599999999999996, + "loss": 0.0139, + "step": 1640 + }, + { + "epoch": 0.772235294117647, + "grad_norm": 0.1809872781033329, + "learning_rate": 0.00024615, + "loss": 0.0341, + "step": 1641 + }, + { + "epoch": 0.7727058823529411, + "grad_norm": 0.09051044659090574, + "learning_rate": 0.00024629999999999997, + "loss": 0.0404, + "step": 1642 + }, + { + "epoch": 0.7731764705882352, + "grad_norm": 0.6392266407575713, + "learning_rate": 0.00024645, + "loss": 0.2715, + "step": 1643 + }, + { + "epoch": 0.7736470588235295, + "grad_norm": 0.11458753393121437, + "learning_rate": 0.0002466, + "loss": 0.0592, + "step": 1644 + }, + { + "epoch": 0.7741176470588236, + "grad_norm": 0.40496173042407935, + "learning_rate": 0.00024675, + "loss": 0.1212, + "step": 1645 + }, + { + "epoch": 0.7745882352941177, + "grad_norm": 0.2404083617166275, + "learning_rate": 0.0002469, + "loss": 0.0682, + "step": 1646 + }, + { + "epoch": 0.7750588235294118, + "grad_norm": 0.3334671250067424, + "learning_rate": 0.00024704999999999996, + "loss": 0.1652, + "step": 1647 + }, + { + "epoch": 0.7755294117647059, + "grad_norm": 6.926319033827803e-07, + "learning_rate": 0.0002472, + "loss": 0.0, + "step": 1648 + }, + { + "epoch": 0.776, + "grad_norm": 1.1953686532245934e-07, + "learning_rate": 0.00024734999999999997, + "loss": 0.0, + "step": 1649 + }, + { + "epoch": 0.7764705882352941, + "grad_norm": 0.18071903173488404, + "learning_rate": 0.00024749999999999994, + "loss": 0.0641, + "step": 1650 + }, + { + "epoch": 0.7769411764705882, + "grad_norm": 0.3122978541772392, + "learning_rate": 0.00024765, + "loss": 0.0831, + "step": 1651 + }, + { + "epoch": 0.7774117647058824, + "grad_norm": 0.2019324198904025, + "learning_rate": 0.00024779999999999995, + "loss": 0.0941, + "step": 1652 + }, + { + "epoch": 0.7778823529411765, + "grad_norm": 0.16986781700011871, + "learning_rate": 0.00024795, + "loss": 0.0857, + "step": 1653 + }, + { + "epoch": 0.7783529411764706, + "grad_norm": 4.954925710003757e-06, + "learning_rate": 0.00024809999999999996, + "loss": 0.0, + "step": 1654 + }, + { + "epoch": 0.7788235294117647, + "grad_norm": 0.45461994392233224, + "learning_rate": 0.00024825, + "loss": 0.2719, + "step": 1655 + }, + { + "epoch": 0.7792941176470588, + "grad_norm": 0.3164994729928536, + "learning_rate": 0.00024839999999999997, + "loss": 0.0996, + "step": 1656 + }, + { + "epoch": 0.7797647058823529, + "grad_norm": 0.1184718835730229, + "learning_rate": 0.00024855, + "loss": 0.0451, + "step": 1657 + }, + { + "epoch": 0.780235294117647, + "grad_norm": 0.4476343303915561, + "learning_rate": 0.0002487, + "loss": 0.1752, + "step": 1658 + }, + { + "epoch": 0.7807058823529411, + "grad_norm": 40.566096545671265, + "learning_rate": 0.00024885, + "loss": 0.417, + "step": 1659 + }, + { + "epoch": 0.7811764705882352, + "grad_norm": 1.2686734554476196, + "learning_rate": 0.000249, + "loss": 0.6436, + "step": 1660 + }, + { + "epoch": 0.7816470588235294, + "grad_norm": 1.9097874483871848e-10, + "learning_rate": 0.00024914999999999996, + "loss": 0.0, + "step": 1661 + }, + { + "epoch": 0.7821176470588236, + "grad_norm": 2.8509303188384854, + "learning_rate": 0.0002493, + "loss": 1.7557, + "step": 1662 + }, + { + "epoch": 0.7825882352941177, + "grad_norm": 1.6749056342356987, + "learning_rate": 0.00024944999999999996, + "loss": 0.9612, + "step": 1663 + }, + { + "epoch": 0.7830588235294118, + "grad_norm": 0.24401330564005375, + "learning_rate": 0.00024959999999999994, + "loss": 0.1272, + "step": 1664 + }, + { + "epoch": 0.7835294117647059, + "grad_norm": 6.291621238044489, + "learning_rate": 0.00024974999999999997, + "loss": 2.6493, + "step": 1665 + }, + { + "epoch": 0.784, + "grad_norm": 2.8923290108914217, + "learning_rate": 0.00024989999999999995, + "loss": 0.6967, + "step": 1666 + }, + { + "epoch": 0.7844705882352941, + "grad_norm": 55.71975298418378, + "learning_rate": 0.00025005, + "loss": 6.9677, + "step": 1667 + }, + { + "epoch": 0.7849411764705883, + "grad_norm": 0.007631855041375853, + "learning_rate": 0.00025019999999999996, + "loss": 0.0002, + "step": 1668 + }, + { + "epoch": 0.7854117647058824, + "grad_norm": 1.3108133471139396, + "learning_rate": 0.00025035, + "loss": 0.6078, + "step": 1669 + }, + { + "epoch": 0.7858823529411765, + "grad_norm": 0.08895285142870112, + "learning_rate": 0.00025049999999999996, + "loss": 0.0391, + "step": 1670 + }, + { + "epoch": 0.7863529411764706, + "grad_norm": 1.9717683614693543e-08, + "learning_rate": 0.00025065, + "loss": 0.0, + "step": 1671 + }, + { + "epoch": 0.7868235294117647, + "grad_norm": 0.17390342979108545, + "learning_rate": 0.00025079999999999997, + "loss": 0.0889, + "step": 1672 + }, + { + "epoch": 0.7872941176470588, + "grad_norm": 1.798079325459703, + "learning_rate": 0.00025095, + "loss": 0.9908, + "step": 1673 + }, + { + "epoch": 0.7877647058823529, + "grad_norm": 1.8001521499523312, + "learning_rate": 0.0002511, + "loss": 1.003, + "step": 1674 + }, + { + "epoch": 0.788235294117647, + "grad_norm": 2.7449704103424406e-09, + "learning_rate": 0.00025125, + "loss": 0.0, + "step": 1675 + }, + { + "epoch": 0.7887058823529411, + "grad_norm": 4.1860476124889745, + "learning_rate": 0.0002514, + "loss": 2.3598, + "step": 1676 + }, + { + "epoch": 0.7891764705882353, + "grad_norm": 4.220786601300536, + "learning_rate": 0.00025154999999999996, + "loss": 2.3272, + "step": 1677 + }, + { + "epoch": 0.7896470588235294, + "grad_norm": 3.1439551730974773, + "learning_rate": 0.0002517, + "loss": 1.7259, + "step": 1678 + }, + { + "epoch": 0.7901176470588235, + "grad_norm": 1.2206128190612064, + "learning_rate": 0.00025184999999999997, + "loss": 0.6233, + "step": 1679 + }, + { + "epoch": 0.7905882352941176, + "grad_norm": 1.4372841827566416, + "learning_rate": 0.00025199999999999995, + "loss": 0.714, + "step": 1680 + }, + { + "epoch": 0.7910588235294118, + "grad_norm": 6.124090880555993, + "learning_rate": 0.00025215, + "loss": 2.9008, + "step": 1681 + }, + { + "epoch": 0.7915294117647059, + "grad_norm": 4.907760241479119, + "learning_rate": 0.00025229999999999995, + "loss": 2.1401, + "step": 1682 + }, + { + "epoch": 0.792, + "grad_norm": 3.1298166344660547, + "learning_rate": 0.00025245, + "loss": 1.251, + "step": 1683 + }, + { + "epoch": 0.7924705882352941, + "grad_norm": 0.3403778832958722, + "learning_rate": 0.00025259999999999996, + "loss": 0.1231, + "step": 1684 + }, + { + "epoch": 0.7929411764705883, + "grad_norm": 3.7004713616021823, + "learning_rate": 0.00025275, + "loss": 1.1593, + "step": 1685 + }, + { + "epoch": 0.7934117647058824, + "grad_norm": 1.667544714734065, + "learning_rate": 0.00025289999999999997, + "loss": 0.4509, + "step": 1686 + }, + { + "epoch": 0.7938823529411765, + "grad_norm": 0.9755887093499961, + "learning_rate": 0.00025305, + "loss": 0.3266, + "step": 1687 + }, + { + "epoch": 0.7943529411764706, + "grad_norm": 4.043522958916226, + "learning_rate": 0.0002532, + "loss": 0.0826, + "step": 1688 + }, + { + "epoch": 0.7948235294117647, + "grad_norm": 3.5304920417291163, + "learning_rate": 0.00025335, + "loss": 0.4827, + "step": 1689 + }, + { + "epoch": 0.7952941176470588, + "grad_norm": 5.514482102349381, + "learning_rate": 0.0002535, + "loss": 0.1274, + "step": 1690 + }, + { + "epoch": 0.7957647058823529, + "grad_norm": 1.9573820471420966, + "learning_rate": 0.00025365, + "loss": 0.396, + "step": 1691 + }, + { + "epoch": 0.796235294117647, + "grad_norm": 1.2276235393224655, + "learning_rate": 0.0002538, + "loss": 0.4957, + "step": 1692 + }, + { + "epoch": 0.7967058823529412, + "grad_norm": 2.8417668599461052, + "learning_rate": 0.00025394999999999997, + "loss": 0.2421, + "step": 1693 + }, + { + "epoch": 0.7971764705882353, + "grad_norm": 4.132591612936327, + "learning_rate": 0.0002541, + "loss": 0.1229, + "step": 1694 + }, + { + "epoch": 0.7976470588235294, + "grad_norm": 2.3040397928191143, + "learning_rate": 0.00025425, + "loss": 0.0413, + "step": 1695 + }, + { + "epoch": 0.7981176470588235, + "grad_norm": 0.9892157472375221, + "learning_rate": 0.00025439999999999995, + "loss": 0.2751, + "step": 1696 + }, + { + "epoch": 0.7985882352941176, + "grad_norm": 0.10980302514222494, + "learning_rate": 0.00025455, + "loss": 0.0301, + "step": 1697 + }, + { + "epoch": 0.7990588235294117, + "grad_norm": 1.015527718300017, + "learning_rate": 0.00025469999999999996, + "loss": 0.193, + "step": 1698 + }, + { + "epoch": 0.7995294117647059, + "grad_norm": 1.0884852259283324, + "learning_rate": 0.00025485, + "loss": 0.203, + "step": 1699 + }, + { + "epoch": 0.8, + "grad_norm": 1.0565570884277853, + "learning_rate": 0.00025499999999999996, + "loss": 0.202, + "step": 1700 + }, + { + "epoch": 0.8004705882352942, + "grad_norm": 0.866005333708952, + "learning_rate": 0.00025515, + "loss": 0.177, + "step": 1701 + }, + { + "epoch": 0.8009411764705883, + "grad_norm": 4.886437669214568, + "learning_rate": 0.00025529999999999997, + "loss": 0.9527, + "step": 1702 + }, + { + "epoch": 0.8014117647058824, + "grad_norm": 3.6433640408435277, + "learning_rate": 0.00025545, + "loss": 0.6911, + "step": 1703 + }, + { + "epoch": 0.8018823529411765, + "grad_norm": 0.9254912203503699, + "learning_rate": 0.0002556, + "loss": 0.4127, + "step": 1704 + }, + { + "epoch": 0.8023529411764706, + "grad_norm": 3.8087779418141126, + "learning_rate": 0.00025575, + "loss": 0.6904, + "step": 1705 + }, + { + "epoch": 0.8028235294117647, + "grad_norm": 1.5182813563407835, + "learning_rate": 0.0002559, + "loss": 0.4576, + "step": 1706 + }, + { + "epoch": 0.8032941176470588, + "grad_norm": 6.891293099019175, + "learning_rate": 0.00025604999999999996, + "loss": 0.1726, + "step": 1707 + }, + { + "epoch": 0.8037647058823529, + "grad_norm": 3.207374965895995, + "learning_rate": 0.0002562, + "loss": 0.5457, + "step": 1708 + }, + { + "epoch": 0.804235294117647, + "grad_norm": 1.6693958224704781, + "learning_rate": 0.00025634999999999997, + "loss": 0.4674, + "step": 1709 + }, + { + "epoch": 0.8047058823529412, + "grad_norm": 2.147010554001132, + "learning_rate": 0.00025649999999999995, + "loss": 0.1714, + "step": 1710 + }, + { + "epoch": 0.8051764705882353, + "grad_norm": 2.440824736416024, + "learning_rate": 0.00025665, + "loss": 0.5107, + "step": 1711 + }, + { + "epoch": 0.8056470588235294, + "grad_norm": 0.37104894713426245, + "learning_rate": 0.00025679999999999995, + "loss": 0.1716, + "step": 1712 + }, + { + "epoch": 0.8061176470588235, + "grad_norm": 0.4337573123060512, + "learning_rate": 0.00025695, + "loss": 0.187, + "step": 1713 + }, + { + "epoch": 0.8065882352941176, + "grad_norm": 0.4909650897139153, + "learning_rate": 0.00025709999999999996, + "loss": 0.2017, + "step": 1714 + }, + { + "epoch": 0.8070588235294117, + "grad_norm": 0.5791221379168588, + "learning_rate": 0.00025725, + "loss": 0.1137, + "step": 1715 + }, + { + "epoch": 0.8075294117647058, + "grad_norm": 1.053605440931456, + "learning_rate": 0.00025739999999999997, + "loss": 0.026, + "step": 1716 + }, + { + "epoch": 0.808, + "grad_norm": 2.839316464674025, + "learning_rate": 0.00025755, + "loss": 0.5482, + "step": 1717 + }, + { + "epoch": 0.8084705882352942, + "grad_norm": 0.9240702074473705, + "learning_rate": 0.0002577, + "loss": 0.0229, + "step": 1718 + }, + { + "epoch": 0.8089411764705883, + "grad_norm": 1.392937469699732, + "learning_rate": 0.00025785, + "loss": 0.3203, + "step": 1719 + }, + { + "epoch": 0.8094117647058824, + "grad_norm": 0.4771680904196432, + "learning_rate": 0.000258, + "loss": 0.0579, + "step": 1720 + }, + { + "epoch": 0.8098823529411765, + "grad_norm": 0.8893072003908544, + "learning_rate": 0.00025815, + "loss": 0.0462, + "step": 1721 + }, + { + "epoch": 0.8103529411764706, + "grad_norm": 0.941307786398059, + "learning_rate": 0.0002583, + "loss": 0.035, + "step": 1722 + }, + { + "epoch": 0.8108235294117647, + "grad_norm": 0.4541944748344663, + "learning_rate": 0.00025844999999999997, + "loss": 0.0598, + "step": 1723 + }, + { + "epoch": 0.8112941176470588, + "grad_norm": 0.2295051179050811, + "learning_rate": 0.0002586, + "loss": 0.0962, + "step": 1724 + }, + { + "epoch": 0.8117647058823529, + "grad_norm": 0.5120956860206897, + "learning_rate": 0.00025875, + "loss": 0.0269, + "step": 1725 + }, + { + "epoch": 0.812235294117647, + "grad_norm": 2.2484109935127643, + "learning_rate": 0.00025889999999999995, + "loss": 0.3528, + "step": 1726 + }, + { + "epoch": 0.8127058823529412, + "grad_norm": 0.7185007716951346, + "learning_rate": 0.00025905, + "loss": 0.0163, + "step": 1727 + }, + { + "epoch": 0.8131764705882353, + "grad_norm": 2.0812050112331715, + "learning_rate": 0.00025919999999999996, + "loss": 0.3275, + "step": 1728 + }, + { + "epoch": 0.8136470588235294, + "grad_norm": 1.309811343457451, + "learning_rate": 0.00025935, + "loss": 0.0293, + "step": 1729 + }, + { + "epoch": 0.8141176470588235, + "grad_norm": 0.20114473784682774, + "learning_rate": 0.00025949999999999997, + "loss": 0.1453, + "step": 1730 + }, + { + "epoch": 0.8145882352941176, + "grad_norm": 0.5456924836167574, + "learning_rate": 0.00025965, + "loss": 0.2273, + "step": 1731 + }, + { + "epoch": 0.8150588235294117, + "grad_norm": 0.41811604282408277, + "learning_rate": 0.00025979999999999997, + "loss": 0.1546, + "step": 1732 + }, + { + "epoch": 0.8155294117647058, + "grad_norm": 0.8925770835811226, + "learning_rate": 0.00025995, + "loss": 0.2775, + "step": 1733 + }, + { + "epoch": 0.816, + "grad_norm": 1.6153894776711752, + "learning_rate": 0.0002601, + "loss": 0.1005, + "step": 1734 + }, + { + "epoch": 0.8164705882352942, + "grad_norm": 2.8966891913103496, + "learning_rate": 0.00026025, + "loss": 0.4569, + "step": 1735 + }, + { + "epoch": 0.8169411764705883, + "grad_norm": 1.575674051078874, + "learning_rate": 0.0002604, + "loss": 0.1418, + "step": 1736 + }, + { + "epoch": 0.8174117647058824, + "grad_norm": 2.566003820543012, + "learning_rate": 0.00026055, + "loss": 0.0987, + "step": 1737 + }, + { + "epoch": 0.8178823529411765, + "grad_norm": 0.8138954866493936, + "learning_rate": 0.0002607, + "loss": 0.2739, + "step": 1738 + }, + { + "epoch": 0.8183529411764706, + "grad_norm": 0.7080067305916206, + "learning_rate": 0.00026084999999999997, + "loss": 0.1844, + "step": 1739 + }, + { + "epoch": 0.8188235294117647, + "grad_norm": 0.5335448466304389, + "learning_rate": 0.000261, + "loss": 0.2339, + "step": 1740 + }, + { + "epoch": 0.8192941176470588, + "grad_norm": 0.982095979762015, + "learning_rate": 0.00026115, + "loss": 0.2386, + "step": 1741 + }, + { + "epoch": 0.819764705882353, + "grad_norm": 0.46130704173550946, + "learning_rate": 0.00026129999999999995, + "loss": 0.1692, + "step": 1742 + }, + { + "epoch": 0.8202352941176471, + "grad_norm": 1.9332829577708865, + "learning_rate": 0.00026145, + "loss": 0.114, + "step": 1743 + }, + { + "epoch": 0.8207058823529412, + "grad_norm": 1.21018712615669, + "learning_rate": 0.00026159999999999996, + "loss": 0.0539, + "step": 1744 + }, + { + "epoch": 0.8211764705882353, + "grad_norm": 1.5658686177624959, + "learning_rate": 0.00026175, + "loss": 0.2342, + "step": 1745 + }, + { + "epoch": 0.8216470588235294, + "grad_norm": 0.8054046696043342, + "learning_rate": 0.00026189999999999997, + "loss": 0.2177, + "step": 1746 + }, + { + "epoch": 0.8221176470588235, + "grad_norm": 2.836168403353714, + "learning_rate": 0.00026205, + "loss": 0.3224, + "step": 1747 + }, + { + "epoch": 0.8225882352941176, + "grad_norm": 2.4542879596358915, + "learning_rate": 0.0002622, + "loss": 0.0602, + "step": 1748 + }, + { + "epoch": 0.8230588235294117, + "grad_norm": 2.528126088662286, + "learning_rate": 0.00026235, + "loss": 0.0618, + "step": 1749 + }, + { + "epoch": 0.8235294117647058, + "grad_norm": 0.7689599157355071, + "learning_rate": 0.0002625, + "loss": 0.2256, + "step": 1750 + }, + { + "epoch": 0.824, + "grad_norm": 1.3441096132882604, + "learning_rate": 0.00026264999999999996, + "loss": 0.2274, + "step": 1751 + }, + { + "epoch": 0.8244705882352941, + "grad_norm": 0.655462360643153, + "learning_rate": 0.0002628, + "loss": 0.0251, + "step": 1752 + }, + { + "epoch": 0.8249411764705883, + "grad_norm": 0.5863304055873337, + "learning_rate": 0.00026294999999999997, + "loss": 0.0241, + "step": 1753 + }, + { + "epoch": 0.8254117647058824, + "grad_norm": 43.302072342639725, + "learning_rate": 0.0002631, + "loss": 0.4229, + "step": 1754 + }, + { + "epoch": 0.8258823529411765, + "grad_norm": 1.2466197122260128, + "learning_rate": 0.00026325, + "loss": 0.2232, + "step": 1755 + }, + { + "epoch": 0.8263529411764706, + "grad_norm": 0.17853557535250517, + "learning_rate": 0.00026339999999999995, + "loss": 0.019, + "step": 1756 + }, + { + "epoch": 0.8268235294117647, + "grad_norm": 1.2747741389697704, + "learning_rate": 0.00026355, + "loss": 0.2504, + "step": 1757 + }, + { + "epoch": 0.8272941176470588, + "grad_norm": 0.1890106080032327, + "learning_rate": 0.00026369999999999996, + "loss": 0.0046, + "step": 1758 + }, + { + "epoch": 0.827764705882353, + "grad_norm": 0.18164314889069624, + "learning_rate": 0.00026384999999999994, + "loss": 0.0303, + "step": 1759 + }, + { + "epoch": 0.8282352941176471, + "grad_norm": 22.103281599411474, + "learning_rate": 0.00026399999999999997, + "loss": 0.3189, + "step": 1760 + }, + { + "epoch": 0.8287058823529412, + "grad_norm": 0.5495166992410326, + "learning_rate": 0.00026414999999999994, + "loss": 0.0949, + "step": 1761 + }, + { + "epoch": 0.8291764705882353, + "grad_norm": 1.0282266458690181, + "learning_rate": 0.0002643, + "loss": 0.192, + "step": 1762 + }, + { + "epoch": 0.8296470588235294, + "grad_norm": 1.0476925797782977, + "learning_rate": 0.00026444999999999995, + "loss": 0.0274, + "step": 1763 + }, + { + "epoch": 0.8301176470588235, + "grad_norm": 0.9551522871054404, + "learning_rate": 0.0002646, + "loss": 0.1163, + "step": 1764 + }, + { + "epoch": 0.8305882352941176, + "grad_norm": 0.6541352675038686, + "learning_rate": 0.00026474999999999996, + "loss": 0.0168, + "step": 1765 + }, + { + "epoch": 0.8310588235294117, + "grad_norm": 0.4747876735753005, + "learning_rate": 0.0002649, + "loss": 0.0122, + "step": 1766 + }, + { + "epoch": 0.8315294117647059, + "grad_norm": 0.24249884380859688, + "learning_rate": 0.00026504999999999996, + "loss": 0.0083, + "step": 1767 + }, + { + "epoch": 0.832, + "grad_norm": 0.22670092947083748, + "learning_rate": 0.0002652, + "loss": 0.0406, + "step": 1768 + }, + { + "epoch": 0.8324705882352941, + "grad_norm": 0.35142926136940456, + "learning_rate": 0.00026534999999999997, + "loss": 0.1296, + "step": 1769 + }, + { + "epoch": 0.8329411764705882, + "grad_norm": 0.15391870484826006, + "learning_rate": 0.0002655, + "loss": 0.0474, + "step": 1770 + }, + { + "epoch": 0.8334117647058824, + "grad_norm": 0.2381477882094401, + "learning_rate": 0.00026565, + "loss": 0.1304, + "step": 1771 + }, + { + "epoch": 0.8338823529411765, + "grad_norm": 0.16369047676835932, + "learning_rate": 0.00026579999999999996, + "loss": 0.0844, + "step": 1772 + }, + { + "epoch": 0.8343529411764706, + "grad_norm": 0.10650004833868379, + "learning_rate": 0.00026595, + "loss": 0.0462, + "step": 1773 + }, + { + "epoch": 0.8348235294117647, + "grad_norm": 0.23317960095149787, + "learning_rate": 0.00026609999999999996, + "loss": 0.0729, + "step": 1774 + }, + { + "epoch": 0.8352941176470589, + "grad_norm": 0.6338095750184588, + "learning_rate": 0.00026624999999999994, + "loss": 0.2801, + "step": 1775 + }, + { + "epoch": 0.835764705882353, + "grad_norm": 0.007422227297472254, + "learning_rate": 0.00026639999999999997, + "loss": 0.0002, + "step": 1776 + }, + { + "epoch": 0.8362352941176471, + "grad_norm": 0.42775216039624925, + "learning_rate": 0.00026654999999999995, + "loss": 0.1389, + "step": 1777 + }, + { + "epoch": 0.8367058823529412, + "grad_norm": 0.18509308232049, + "learning_rate": 0.0002667, + "loss": 0.1164, + "step": 1778 + }, + { + "epoch": 0.8371764705882353, + "grad_norm": 0.003785328213933265, + "learning_rate": 0.00026684999999999995, + "loss": 0.0001, + "step": 1779 + }, + { + "epoch": 0.8376470588235294, + "grad_norm": 1.0972790375802783, + "learning_rate": 0.000267, + "loss": 0.3675, + "step": 1780 + }, + { + "epoch": 0.8381176470588235, + "grad_norm": 0.1673750888103066, + "learning_rate": 0.00026714999999999996, + "loss": 0.0378, + "step": 1781 + }, + { + "epoch": 0.8385882352941176, + "grad_norm": 0.0035461799709240456, + "learning_rate": 0.0002673, + "loss": 0.0001, + "step": 1782 + }, + { + "epoch": 0.8390588235294117, + "grad_norm": 1.286779363126863, + "learning_rate": 0.00026744999999999997, + "loss": 0.2803, + "step": 1783 + }, + { + "epoch": 0.8395294117647059, + "grad_norm": 0.04329100977854106, + "learning_rate": 0.0002676, + "loss": 0.0063, + "step": 1784 + }, + { + "epoch": 0.84, + "grad_norm": 0.09289383861895022, + "learning_rate": 0.00026775, + "loss": 0.0273, + "step": 1785 + }, + { + "epoch": 0.8404705882352941, + "grad_norm": 0.0024725884878083735, + "learning_rate": 0.0002679, + "loss": 0.0001, + "step": 1786 + }, + { + "epoch": 0.8409411764705882, + "grad_norm": 0.21708327283204884, + "learning_rate": 0.00026805, + "loss": 0.0756, + "step": 1787 + }, + { + "epoch": 0.8414117647058823, + "grad_norm": 0.12510688545115878, + "learning_rate": 0.00026819999999999996, + "loss": 0.0248, + "step": 1788 + }, + { + "epoch": 0.8418823529411765, + "grad_norm": 0.52495568865928, + "learning_rate": 0.00026835, + "loss": 0.153, + "step": 1789 + }, + { + "epoch": 0.8423529411764706, + "grad_norm": 0.1069863452819155, + "learning_rate": 0.00026849999999999997, + "loss": 0.0298, + "step": 1790 + }, + { + "epoch": 0.8428235294117647, + "grad_norm": 0.3442281313101642, + "learning_rate": 0.00026864999999999994, + "loss": 0.1818, + "step": 1791 + }, + { + "epoch": 0.8432941176470589, + "grad_norm": 0.15351467087514145, + "learning_rate": 0.0002688, + "loss": 0.063, + "step": 1792 + }, + { + "epoch": 0.843764705882353, + "grad_norm": 0.21110973893090157, + "learning_rate": 0.00026894999999999995, + "loss": 0.0876, + "step": 1793 + }, + { + "epoch": 0.8442352941176471, + "grad_norm": 0.04809619663370406, + "learning_rate": 0.0002691, + "loss": 0.0091, + "step": 1794 + }, + { + "epoch": 0.8447058823529412, + "grad_norm": 0.27294877306069465, + "learning_rate": 0.00026924999999999996, + "loss": 0.1154, + "step": 1795 + }, + { + "epoch": 0.8451764705882353, + "grad_norm": 0.12528618268897748, + "learning_rate": 0.0002694, + "loss": 0.0639, + "step": 1796 + }, + { + "epoch": 0.8456470588235294, + "grad_norm": 0.240943113197725, + "learning_rate": 0.00026954999999999997, + "loss": 0.148, + "step": 1797 + }, + { + "epoch": 0.8461176470588235, + "grad_norm": 0.11245273957697133, + "learning_rate": 0.0002697, + "loss": 0.0469, + "step": 1798 + }, + { + "epoch": 0.8465882352941176, + "grad_norm": 0.11463628343591455, + "learning_rate": 0.00026984999999999997, + "loss": 0.0511, + "step": 1799 + }, + { + "epoch": 0.8470588235294118, + "grad_norm": 0.09980762328930602, + "learning_rate": 0.00027, + "loss": 0.0342, + "step": 1800 + }, + { + "epoch": 0.8475294117647059, + "grad_norm": 0.047657317044188706, + "learning_rate": 0.00027015, + "loss": 0.0149, + "step": 1801 + }, + { + "epoch": 0.848, + "grad_norm": 0.1554573628471381, + "learning_rate": 0.00027029999999999996, + "loss": 0.0491, + "step": 1802 + }, + { + "epoch": 0.8484705882352941, + "grad_norm": 0.39600649121344245, + "learning_rate": 0.00027045, + "loss": 0.2081, + "step": 1803 + }, + { + "epoch": 0.8489411764705882, + "grad_norm": 0.0014554328767558112, + "learning_rate": 0.00027059999999999996, + "loss": 0.0, + "step": 1804 + }, + { + "epoch": 0.8494117647058823, + "grad_norm": 0.29429320456297753, + "learning_rate": 0.00027074999999999994, + "loss": 0.1752, + "step": 1805 + }, + { + "epoch": 0.8498823529411764, + "grad_norm": 0.11600870249554816, + "learning_rate": 0.00027089999999999997, + "loss": 0.0576, + "step": 1806 + }, + { + "epoch": 0.8503529411764705, + "grad_norm": 0.12206979792766472, + "learning_rate": 0.00027104999999999995, + "loss": 0.0524, + "step": 1807 + }, + { + "epoch": 0.8508235294117648, + "grad_norm": 0.0711268483790433, + "learning_rate": 0.0002712, + "loss": 0.0148, + "step": 1808 + }, + { + "epoch": 0.8512941176470589, + "grad_norm": 0.2558321910569146, + "learning_rate": 0.00027134999999999995, + "loss": 0.0868, + "step": 1809 + }, + { + "epoch": 0.851764705882353, + "grad_norm": 0.0016044914768455361, + "learning_rate": 0.0002715, + "loss": 0.0, + "step": 1810 + }, + { + "epoch": 0.8522352941176471, + "grad_norm": 0.35928799541685286, + "learning_rate": 0.00027164999999999996, + "loss": 0.1781, + "step": 1811 + }, + { + "epoch": 0.8527058823529412, + "grad_norm": 0.049084552537475436, + "learning_rate": 0.0002718, + "loss": 0.0068, + "step": 1812 + }, + { + "epoch": 0.8531764705882353, + "grad_norm": 0.10852275730146298, + "learning_rate": 0.00027194999999999997, + "loss": 0.0514, + "step": 1813 + }, + { + "epoch": 0.8536470588235294, + "grad_norm": 0.2725395728429933, + "learning_rate": 0.0002721, + "loss": 0.1437, + "step": 1814 + }, + { + "epoch": 0.8541176470588235, + "grad_norm": 0.19814215802020727, + "learning_rate": 0.00027225, + "loss": 0.0854, + "step": 1815 + }, + { + "epoch": 0.8545882352941176, + "grad_norm": 0.04364496537320186, + "learning_rate": 0.0002724, + "loss": 0.0059, + "step": 1816 + }, + { + "epoch": 0.8550588235294118, + "grad_norm": 0.08611794709254326, + "learning_rate": 0.00027255, + "loss": 0.0401, + "step": 1817 + }, + { + "epoch": 0.8555294117647059, + "grad_norm": 0.23113850668103153, + "learning_rate": 0.00027269999999999996, + "loss": 0.1645, + "step": 1818 + }, + { + "epoch": 0.856, + "grad_norm": 0.08626575420301374, + "learning_rate": 0.00027285, + "loss": 0.0449, + "step": 1819 + }, + { + "epoch": 0.8564705882352941, + "grad_norm": 0.08133863293286873, + "learning_rate": 0.00027299999999999997, + "loss": 0.0324, + "step": 1820 + }, + { + "epoch": 0.8569411764705882, + "grad_norm": 0.5240995386057642, + "learning_rate": 0.00027314999999999994, + "loss": 0.098, + "step": 1821 + }, + { + "epoch": 0.8574117647058823, + "grad_norm": 0.22809253247346298, + "learning_rate": 0.0002733, + "loss": 0.0915, + "step": 1822 + }, + { + "epoch": 0.8578823529411764, + "grad_norm": 0.13140829238301438, + "learning_rate": 0.00027344999999999995, + "loss": 0.0704, + "step": 1823 + }, + { + "epoch": 0.8583529411764705, + "grad_norm": 0.060239222831875126, + "learning_rate": 0.0002736, + "loss": 0.0232, + "step": 1824 + }, + { + "epoch": 0.8588235294117647, + "grad_norm": 0.0013502665496556367, + "learning_rate": 0.00027374999999999996, + "loss": 0.0, + "step": 1825 + }, + { + "epoch": 0.8592941176470589, + "grad_norm": 0.1888509189153781, + "learning_rate": 0.0002739, + "loss": 0.0782, + "step": 1826 + }, + { + "epoch": 0.859764705882353, + "grad_norm": 0.10579526399970862, + "learning_rate": 0.00027404999999999997, + "loss": 0.0506, + "step": 1827 + }, + { + "epoch": 0.8602352941176471, + "grad_norm": 0.15407396023787007, + "learning_rate": 0.0002742, + "loss": 0.0443, + "step": 1828 + }, + { + "epoch": 0.8607058823529412, + "grad_norm": 0.12271863201892685, + "learning_rate": 0.00027435, + "loss": 0.0522, + "step": 1829 + }, + { + "epoch": 0.8611764705882353, + "grad_norm": 0.0012770124824962397, + "learning_rate": 0.0002745, + "loss": 0.0, + "step": 1830 + }, + { + "epoch": 0.8616470588235294, + "grad_norm": 0.35915780903160704, + "learning_rate": 0.00027465, + "loss": 0.2645, + "step": 1831 + }, + { + "epoch": 0.8621176470588235, + "grad_norm": 0.11631152791244162, + "learning_rate": 0.0002748, + "loss": 0.0534, + "step": 1832 + }, + { + "epoch": 0.8625882352941177, + "grad_norm": 0.20968610246636107, + "learning_rate": 0.00027495, + "loss": 0.1426, + "step": 1833 + }, + { + "epoch": 0.8630588235294118, + "grad_norm": 0.09090466049352967, + "learning_rate": 0.00027509999999999996, + "loss": 0.0408, + "step": 1834 + }, + { + "epoch": 0.8635294117647059, + "grad_norm": 0.0467186891976625, + "learning_rate": 0.00027525, + "loss": 0.0099, + "step": 1835 + }, + { + "epoch": 0.864, + "grad_norm": 0.16420076476670606, + "learning_rate": 0.00027539999999999997, + "loss": 0.0705, + "step": 1836 + }, + { + "epoch": 0.8644705882352941, + "grad_norm": 0.16144496913256778, + "learning_rate": 0.00027554999999999995, + "loss": 0.0807, + "step": 1837 + }, + { + "epoch": 0.8649411764705882, + "grad_norm": 0.18222163380065992, + "learning_rate": 0.0002757, + "loss": 0.0947, + "step": 1838 + }, + { + "epoch": 0.8654117647058823, + "grad_norm": 0.10912159102212177, + "learning_rate": 0.00027584999999999996, + "loss": 0.0638, + "step": 1839 + }, + { + "epoch": 0.8658823529411764, + "grad_norm": 0.1949345842869015, + "learning_rate": 0.000276, + "loss": 0.0859, + "step": 1840 + }, + { + "epoch": 0.8663529411764705, + "grad_norm": 0.001157885227780314, + "learning_rate": 0.00027614999999999996, + "loss": 0.0, + "step": 1841 + }, + { + "epoch": 0.8668235294117647, + "grad_norm": 0.32371758247754534, + "learning_rate": 0.0002763, + "loss": 0.2189, + "step": 1842 + }, + { + "epoch": 0.8672941176470588, + "grad_norm": 0.0009941253219243455, + "learning_rate": 0.00027644999999999997, + "loss": 0.0, + "step": 1843 + }, + { + "epoch": 0.867764705882353, + "grad_norm": 0.08929394286427352, + "learning_rate": 0.0002766, + "loss": 0.0294, + "step": 1844 + }, + { + "epoch": 0.8682352941176471, + "grad_norm": 0.13605168256925407, + "learning_rate": 0.00027675, + "loss": 0.0683, + "step": 1845 + }, + { + "epoch": 0.8687058823529412, + "grad_norm": 0.0010430034307301861, + "learning_rate": 0.0002769, + "loss": 0.0, + "step": 1846 + }, + { + "epoch": 0.8691764705882353, + "grad_norm": 0.3799209056483094, + "learning_rate": 0.00027705, + "loss": 0.0967, + "step": 1847 + }, + { + "epoch": 0.8696470588235294, + "grad_norm": 0.06133549900460582, + "learning_rate": 0.0002772, + "loss": 0.0195, + "step": 1848 + }, + { + "epoch": 0.8701176470588236, + "grad_norm": 0.14849798791574567, + "learning_rate": 0.00027735, + "loss": 0.0416, + "step": 1849 + }, + { + "epoch": 0.8705882352941177, + "grad_norm": 0.3005372215840502, + "learning_rate": 0.00027749999999999997, + "loss": 0.1085, + "step": 1850 + }, + { + "epoch": 0.8710588235294118, + "grad_norm": 0.0011362143799177886, + "learning_rate": 0.00027764999999999995, + "loss": 0.0, + "step": 1851 + }, + { + "epoch": 0.8715294117647059, + "grad_norm": 0.11186765641097413, + "learning_rate": 0.0002778, + "loss": 0.0467, + "step": 1852 + }, + { + "epoch": 0.872, + "grad_norm": 0.15073031740584314, + "learning_rate": 0.00027794999999999995, + "loss": 0.0534, + "step": 1853 + }, + { + "epoch": 0.8724705882352941, + "grad_norm": 0.28689632116544517, + "learning_rate": 0.0002781, + "loss": 0.1074, + "step": 1854 + }, + { + "epoch": 0.8729411764705882, + "grad_norm": 0.2896089696010033, + "learning_rate": 0.00027824999999999996, + "loss": 0.1784, + "step": 1855 + }, + { + "epoch": 0.8734117647058823, + "grad_norm": 0.16416523310588607, + "learning_rate": 0.0002784, + "loss": 0.0817, + "step": 1856 + }, + { + "epoch": 0.8738823529411764, + "grad_norm": 0.0012167189694716866, + "learning_rate": 0.00027854999999999997, + "loss": 0.0, + "step": 1857 + }, + { + "epoch": 0.8743529411764706, + "grad_norm": 0.0701281921034765, + "learning_rate": 0.0002787, + "loss": 0.0199, + "step": 1858 + }, + { + "epoch": 0.8748235294117647, + "grad_norm": 0.15010092382462045, + "learning_rate": 0.00027885, + "loss": 0.0717, + "step": 1859 + }, + { + "epoch": 0.8752941176470588, + "grad_norm": 0.17797104618770648, + "learning_rate": 0.000279, + "loss": 0.096, + "step": 1860 + }, + { + "epoch": 0.8757647058823529, + "grad_norm": 0.15332987296392808, + "learning_rate": 0.00027915, + "loss": 0.0774, + "step": 1861 + }, + { + "epoch": 0.8762352941176471, + "grad_norm": 0.08933845008560415, + "learning_rate": 0.0002793, + "loss": 0.0348, + "step": 1862 + }, + { + "epoch": 0.8767058823529412, + "grad_norm": 0.08939464755397693, + "learning_rate": 0.00027945, + "loss": 0.0482, + "step": 1863 + }, + { + "epoch": 0.8771764705882353, + "grad_norm": 0.16125748200919374, + "learning_rate": 0.00027959999999999997, + "loss": 0.0699, + "step": 1864 + }, + { + "epoch": 0.8776470588235294, + "grad_norm": 0.11067934373098424, + "learning_rate": 0.00027975, + "loss": 0.0549, + "step": 1865 + }, + { + "epoch": 0.8781176470588236, + "grad_norm": 0.2165816034500059, + "learning_rate": 0.0002799, + "loss": 0.1113, + "step": 1866 + }, + { + "epoch": 0.8785882352941177, + "grad_norm": 0.0013034107547834522, + "learning_rate": 0.00028004999999999995, + "loss": 0.0, + "step": 1867 + }, + { + "epoch": 0.8790588235294118, + "grad_norm": 0.0810237619609323, + "learning_rate": 0.0002802, + "loss": 0.0488, + "step": 1868 + }, + { + "epoch": 0.8795294117647059, + "grad_norm": 0.001149497973649319, + "learning_rate": 0.00028034999999999996, + "loss": 0.0, + "step": 1869 + }, + { + "epoch": 0.88, + "grad_norm": 0.0009493103735902026, + "learning_rate": 0.0002805, + "loss": 0.0, + "step": 1870 + }, + { + "epoch": 0.8804705882352941, + "grad_norm": 0.2898898330382849, + "learning_rate": 0.00028064999999999996, + "loss": 0.1232, + "step": 1871 + }, + { + "epoch": 0.8809411764705882, + "grad_norm": 0.30399294327949977, + "learning_rate": 0.0002808, + "loss": 0.1295, + "step": 1872 + }, + { + "epoch": 0.8814117647058823, + "grad_norm": 0.0009049402575937405, + "learning_rate": 0.00028094999999999997, + "loss": 0.0, + "step": 1873 + }, + { + "epoch": 0.8818823529411765, + "grad_norm": 0.09144539166964487, + "learning_rate": 0.0002811, + "loss": 0.0294, + "step": 1874 + }, + { + "epoch": 0.8823529411764706, + "grad_norm": 0.14294672847885562, + "learning_rate": 0.00028125, + "loss": 0.0554, + "step": 1875 + }, + { + "epoch": 0.8828235294117647, + "grad_norm": 0.12228923767062416, + "learning_rate": 0.00028139999999999996, + "loss": 0.0598, + "step": 1876 + }, + { + "epoch": 0.8832941176470588, + "grad_norm": 0.19274677956930358, + "learning_rate": 0.00028155, + "loss": 0.0884, + "step": 1877 + }, + { + "epoch": 0.8837647058823529, + "grad_norm": 0.13943889459125294, + "learning_rate": 0.00028169999999999996, + "loss": 0.0558, + "step": 1878 + }, + { + "epoch": 0.884235294117647, + "grad_norm": 0.1302743732037569, + "learning_rate": 0.00028185, + "loss": 0.0624, + "step": 1879 + }, + { + "epoch": 0.8847058823529412, + "grad_norm": 0.17691549183814312, + "learning_rate": 0.00028199999999999997, + "loss": 0.0643, + "step": 1880 + }, + { + "epoch": 0.8851764705882353, + "grad_norm": 0.21333878218558966, + "learning_rate": 0.00028215, + "loss": 0.0955, + "step": 1881 + }, + { + "epoch": 0.8856470588235295, + "grad_norm": 0.2158105095139265, + "learning_rate": 0.0002823, + "loss": 0.1014, + "step": 1882 + }, + { + "epoch": 0.8861176470588236, + "grad_norm": 0.18816778247154312, + "learning_rate": 0.00028244999999999995, + "loss": 0.1044, + "step": 1883 + }, + { + "epoch": 0.8865882352941177, + "grad_norm": 0.12610289690701307, + "learning_rate": 0.0002826, + "loss": 0.0601, + "step": 1884 + }, + { + "epoch": 0.8870588235294118, + "grad_norm": 0.21028647170619838, + "learning_rate": 0.00028274999999999996, + "loss": 0.0835, + "step": 1885 + }, + { + "epoch": 0.8875294117647059, + "grad_norm": 0.0008939053356576731, + "learning_rate": 0.00028289999999999994, + "loss": 0.0, + "step": 1886 + }, + { + "epoch": 0.888, + "grad_norm": 0.1716255605227123, + "learning_rate": 0.00028304999999999997, + "loss": 0.0477, + "step": 1887 + }, + { + "epoch": 0.8884705882352941, + "grad_norm": 0.001032820936623738, + "learning_rate": 0.00028319999999999994, + "loss": 0.0, + "step": 1888 + }, + { + "epoch": 0.8889411764705882, + "grad_norm": 0.36984192813841443, + "learning_rate": 0.00028335, + "loss": 0.141, + "step": 1889 + }, + { + "epoch": 0.8894117647058823, + "grad_norm": 0.16269285740100484, + "learning_rate": 0.00028349999999999995, + "loss": 0.0676, + "step": 1890 + }, + { + "epoch": 0.8898823529411765, + "grad_norm": 0.2676717520042121, + "learning_rate": 0.00028365, + "loss": 0.1293, + "step": 1891 + }, + { + "epoch": 0.8903529411764706, + "grad_norm": 0.13919299585393125, + "learning_rate": 0.00028379999999999996, + "loss": 0.0576, + "step": 1892 + }, + { + "epoch": 0.8908235294117647, + "grad_norm": 0.0009712158506608997, + "learning_rate": 0.00028395, + "loss": 0.0, + "step": 1893 + }, + { + "epoch": 0.8912941176470588, + "grad_norm": 0.0010639163675293677, + "learning_rate": 0.00028409999999999997, + "loss": 0.0, + "step": 1894 + }, + { + "epoch": 0.8917647058823529, + "grad_norm": 0.08516901697779461, + "learning_rate": 0.00028425, + "loss": 0.0303, + "step": 1895 + }, + { + "epoch": 0.892235294117647, + "grad_norm": 0.06856983658567271, + "learning_rate": 0.0002844, + "loss": 0.0263, + "step": 1896 + }, + { + "epoch": 0.8927058823529411, + "grad_norm": 0.13273391967607315, + "learning_rate": 0.00028455, + "loss": 0.0406, + "step": 1897 + }, + { + "epoch": 0.8931764705882353, + "grad_norm": 0.32764405811735214, + "learning_rate": 0.0002847, + "loss": 0.1551, + "step": 1898 + }, + { + "epoch": 0.8936470588235295, + "grad_norm": 0.10551981056852623, + "learning_rate": 0.00028484999999999996, + "loss": 0.0434, + "step": 1899 + }, + { + "epoch": 0.8941176470588236, + "grad_norm": 0.0009449342088025554, + "learning_rate": 0.000285, + "loss": 0.0, + "step": 1900 + }, + { + "epoch": 0.8945882352941177, + "grad_norm": 0.17037846824691374, + "learning_rate": 0.00028514999999999997, + "loss": 0.0455, + "step": 1901 + }, + { + "epoch": 0.8950588235294118, + "grad_norm": 0.06419388452799485, + "learning_rate": 0.00028529999999999994, + "loss": 0.021, + "step": 1902 + }, + { + "epoch": 0.8955294117647059, + "grad_norm": 0.043301356415271636, + "learning_rate": 0.00028544999999999997, + "loss": 0.0092, + "step": 1903 + }, + { + "epoch": 0.896, + "grad_norm": 0.15915689957274773, + "learning_rate": 0.00028559999999999995, + "loss": 0.0635, + "step": 1904 + }, + { + "epoch": 0.8964705882352941, + "grad_norm": 0.06093184623323497, + "learning_rate": 0.00028575, + "loss": 0.0201, + "step": 1905 + }, + { + "epoch": 0.8969411764705882, + "grad_norm": 0.26426812095124536, + "learning_rate": 0.00028589999999999996, + "loss": 0.0821, + "step": 1906 + }, + { + "epoch": 0.8974117647058824, + "grad_norm": 0.1309122335503928, + "learning_rate": 0.00028605, + "loss": 0.0647, + "step": 1907 + }, + { + "epoch": 0.8978823529411765, + "grad_norm": 0.12578457704544244, + "learning_rate": 0.00028619999999999996, + "loss": 0.0784, + "step": 1908 + }, + { + "epoch": 0.8983529411764706, + "grad_norm": 0.0012716048193199216, + "learning_rate": 0.00028635, + "loss": 0.0, + "step": 1909 + }, + { + "epoch": 0.8988235294117647, + "grad_norm": 0.11708656200912565, + "learning_rate": 0.00028649999999999997, + "loss": 0.0491, + "step": 1910 + }, + { + "epoch": 0.8992941176470588, + "grad_norm": 0.0014319591461550985, + "learning_rate": 0.00028665, + "loss": 0.0, + "step": 1911 + }, + { + "epoch": 0.8997647058823529, + "grad_norm": 0.0680321992359343, + "learning_rate": 0.0002868, + "loss": 0.0215, + "step": 1912 + }, + { + "epoch": 0.900235294117647, + "grad_norm": 0.31743220805392125, + "learning_rate": 0.00028694999999999995, + "loss": 0.0879, + "step": 1913 + }, + { + "epoch": 0.9007058823529411, + "grad_norm": 0.10945461889906816, + "learning_rate": 0.0002871, + "loss": 0.0387, + "step": 1914 + }, + { + "epoch": 0.9011764705882352, + "grad_norm": 0.1939282245853889, + "learning_rate": 0.00028724999999999996, + "loss": 0.0754, + "step": 1915 + }, + { + "epoch": 0.9016470588235294, + "grad_norm": 0.0031411324066471856, + "learning_rate": 0.00028739999999999994, + "loss": 0.0001, + "step": 1916 + }, + { + "epoch": 0.9021176470588236, + "grad_norm": 0.24040826993103842, + "learning_rate": 0.00028754999999999997, + "loss": 0.0666, + "step": 1917 + }, + { + "epoch": 0.9025882352941177, + "grad_norm": 0.003674117491091287, + "learning_rate": 0.00028769999999999995, + "loss": 0.0001, + "step": 1918 + }, + { + "epoch": 0.9030588235294118, + "grad_norm": 0.04268389940300542, + "learning_rate": 0.00028785, + "loss": 0.0067, + "step": 1919 + }, + { + "epoch": 0.9035294117647059, + "grad_norm": 0.3148347698512027, + "learning_rate": 0.00028799999999999995, + "loss": 0.114, + "step": 1920 + }, + { + "epoch": 0.904, + "grad_norm": 0.03443722175193799, + "learning_rate": 0.00028815, + "loss": 0.0053, + "step": 1921 + }, + { + "epoch": 0.9044705882352941, + "grad_norm": 0.11710965515569664, + "learning_rate": 0.00028829999999999996, + "loss": 0.0825, + "step": 1922 + }, + { + "epoch": 0.9049411764705882, + "grad_norm": 0.109090008775221, + "learning_rate": 0.00028845, + "loss": 0.0527, + "step": 1923 + }, + { + "epoch": 0.9054117647058824, + "grad_norm": 0.1138805338790004, + "learning_rate": 0.00028859999999999997, + "loss": 0.0691, + "step": 1924 + }, + { + "epoch": 0.9058823529411765, + "grad_norm": 0.17637670057493485, + "learning_rate": 0.00028875, + "loss": 0.0928, + "step": 1925 + }, + { + "epoch": 0.9063529411764706, + "grad_norm": 0.0024284259585497735, + "learning_rate": 0.0002889, + "loss": 0.0001, + "step": 1926 + }, + { + "epoch": 0.9068235294117647, + "grad_norm": 0.3140841654909315, + "learning_rate": 0.00028905, + "loss": 0.1462, + "step": 1927 + }, + { + "epoch": 0.9072941176470588, + "grad_norm": 0.30027748104630547, + "learning_rate": 0.0002892, + "loss": 0.164, + "step": 1928 + }, + { + "epoch": 0.9077647058823529, + "grad_norm": 0.15038761710800302, + "learning_rate": 0.00028934999999999996, + "loss": 0.1094, + "step": 1929 + }, + { + "epoch": 0.908235294117647, + "grad_norm": 0.09275033240026874, + "learning_rate": 0.0002895, + "loss": 0.0295, + "step": 1930 + }, + { + "epoch": 0.9087058823529411, + "grad_norm": 0.15331590531204986, + "learning_rate": 0.00028964999999999997, + "loss": 0.0589, + "step": 1931 + }, + { + "epoch": 0.9091764705882353, + "grad_norm": 0.2010740978569698, + "learning_rate": 0.00028979999999999994, + "loss": 0.0446, + "step": 1932 + }, + { + "epoch": 0.9096470588235294, + "grad_norm": 0.001813300770869064, + "learning_rate": 0.00028995, + "loss": 0.0, + "step": 1933 + }, + { + "epoch": 0.9101176470588235, + "grad_norm": 0.2627215824045066, + "learning_rate": 0.00029009999999999995, + "loss": 0.09, + "step": 1934 + }, + { + "epoch": 0.9105882352941177, + "grad_norm": 0.18528045286428768, + "learning_rate": 0.00029025, + "loss": 0.1193, + "step": 1935 + }, + { + "epoch": 0.9110588235294118, + "grad_norm": 0.45132685421946456, + "learning_rate": 0.00029039999999999996, + "loss": 0.1039, + "step": 1936 + }, + { + "epoch": 0.9115294117647059, + "grad_norm": 0.2622720039231139, + "learning_rate": 0.00029055, + "loss": 0.0243, + "step": 1937 + }, + { + "epoch": 0.912, + "grad_norm": 1.610944856247646, + "learning_rate": 0.00029069999999999996, + "loss": 0.2204, + "step": 1938 + }, + { + "epoch": 0.9124705882352941, + "grad_norm": 0.037114370010731146, + "learning_rate": 0.00029085, + "loss": 0.0105, + "step": 1939 + }, + { + "epoch": 0.9129411764705883, + "grad_norm": 0.34011973488368985, + "learning_rate": 0.00029099999999999997, + "loss": 0.2895, + "step": 1940 + }, + { + "epoch": 0.9134117647058824, + "grad_norm": 0.22468732213812487, + "learning_rate": 0.00029115, + "loss": 0.0807, + "step": 1941 + }, + { + "epoch": 0.9138823529411765, + "grad_norm": 0.1130638599919798, + "learning_rate": 0.0002913, + "loss": 0.0616, + "step": 1942 + }, + { + "epoch": 0.9143529411764706, + "grad_norm": 0.006075991327392897, + "learning_rate": 0.00029145, + "loss": 0.0002, + "step": 1943 + }, + { + "epoch": 0.9148235294117647, + "grad_norm": 0.0639510403791329, + "learning_rate": 0.0002916, + "loss": 0.018, + "step": 1944 + }, + { + "epoch": 0.9152941176470588, + "grad_norm": 0.5499149911084891, + "learning_rate": 0.00029174999999999996, + "loss": 0.2196, + "step": 1945 + }, + { + "epoch": 0.9157647058823529, + "grad_norm": 0.05273102694337401, + "learning_rate": 0.0002919, + "loss": 0.0326, + "step": 1946 + }, + { + "epoch": 0.916235294117647, + "grad_norm": 0.0786686928654614, + "learning_rate": 0.00029204999999999997, + "loss": 0.0303, + "step": 1947 + }, + { + "epoch": 0.9167058823529411, + "grad_norm": 0.015908182839696503, + "learning_rate": 0.00029219999999999995, + "loss": 0.0004, + "step": 1948 + }, + { + "epoch": 0.9171764705882353, + "grad_norm": 0.20758822271049548, + "learning_rate": 0.00029235, + "loss": 0.1331, + "step": 1949 + }, + { + "epoch": 0.9176470588235294, + "grad_norm": 0.011074586096933209, + "learning_rate": 0.00029249999999999995, + "loss": 0.0003, + "step": 1950 + }, + { + "epoch": 0.9181176470588235, + "grad_norm": 0.14208311865181109, + "learning_rate": 0.00029265, + "loss": 0.0578, + "step": 1951 + }, + { + "epoch": 0.9185882352941176, + "grad_norm": 0.015843392992954387, + "learning_rate": 0.00029279999999999996, + "loss": 0.0002, + "step": 1952 + }, + { + "epoch": 0.9190588235294118, + "grad_norm": 0.22878052353348785, + "learning_rate": 0.00029295, + "loss": 0.1102, + "step": 1953 + }, + { + "epoch": 0.9195294117647059, + "grad_norm": 0.06534986766176695, + "learning_rate": 0.00029309999999999997, + "loss": 0.0431, + "step": 1954 + }, + { + "epoch": 0.92, + "grad_norm": 0.03764418610089165, + "learning_rate": 0.00029325, + "loss": 0.0123, + "step": 1955 + }, + { + "epoch": 0.9204705882352942, + "grad_norm": 0.040992762931612414, + "learning_rate": 0.0002934, + "loss": 0.0121, + "step": 1956 + }, + { + "epoch": 0.9209411764705883, + "grad_norm": 0.06132897953602661, + "learning_rate": 0.00029355, + "loss": 0.0148, + "step": 1957 + }, + { + "epoch": 0.9214117647058824, + "grad_norm": 0.1361081680583677, + "learning_rate": 0.0002937, + "loss": 0.0563, + "step": 1958 + }, + { + "epoch": 0.9218823529411765, + "grad_norm": 0.10044852278851142, + "learning_rate": 0.00029384999999999996, + "loss": 0.038, + "step": 1959 + }, + { + "epoch": 0.9223529411764706, + "grad_norm": 0.024280422050685927, + "learning_rate": 0.000294, + "loss": 0.004, + "step": 1960 + }, + { + "epoch": 0.9228235294117647, + "grad_norm": 0.12742434628811702, + "learning_rate": 0.00029414999999999997, + "loss": 0.098, + "step": 1961 + }, + { + "epoch": 0.9232941176470588, + "grad_norm": 0.30365953351473624, + "learning_rate": 0.00029429999999999994, + "loss": 0.207, + "step": 1962 + }, + { + "epoch": 0.9237647058823529, + "grad_norm": 0.13684531357722562, + "learning_rate": 0.00029445, + "loss": 0.082, + "step": 1963 + }, + { + "epoch": 0.924235294117647, + "grad_norm": 0.0787557741431526, + "learning_rate": 0.00029459999999999995, + "loss": 0.0355, + "step": 1964 + }, + { + "epoch": 0.9247058823529412, + "grad_norm": 0.021179545315787418, + "learning_rate": 0.00029475, + "loss": 0.0043, + "step": 1965 + }, + { + "epoch": 0.9251764705882353, + "grad_norm": 0.10615357069674039, + "learning_rate": 0.00029489999999999996, + "loss": 0.1061, + "step": 1966 + }, + { + "epoch": 0.9256470588235294, + "grad_norm": 0.18587554318483307, + "learning_rate": 0.00029505, + "loss": 0.1249, + "step": 1967 + }, + { + "epoch": 0.9261176470588235, + "grad_norm": 0.10159823560145598, + "learning_rate": 0.00029519999999999997, + "loss": 0.0559, + "step": 1968 + }, + { + "epoch": 0.9265882352941176, + "grad_norm": 0.11035494963974922, + "learning_rate": 0.00029535, + "loss": 0.0656, + "step": 1969 + }, + { + "epoch": 0.9270588235294117, + "grad_norm": 0.13564615684596124, + "learning_rate": 0.00029549999999999997, + "loss": 0.0831, + "step": 1970 + }, + { + "epoch": 0.9275294117647059, + "grad_norm": 0.020819967142821867, + "learning_rate": 0.00029565, + "loss": 0.0035, + "step": 1971 + }, + { + "epoch": 0.928, + "grad_norm": 0.07363872760897187, + "learning_rate": 0.0002958, + "loss": 0.0436, + "step": 1972 + }, + { + "epoch": 0.9284705882352942, + "grad_norm": 0.19523614273636347, + "learning_rate": 0.00029595, + "loss": 0.1317, + "step": 1973 + }, + { + "epoch": 0.9289411764705883, + "grad_norm": 0.10026611119293696, + "learning_rate": 0.0002961, + "loss": 0.0582, + "step": 1974 + }, + { + "epoch": 0.9294117647058824, + "grad_norm": 0.016130475288252205, + "learning_rate": 0.00029624999999999996, + "loss": 0.0013, + "step": 1975 + }, + { + "epoch": 0.9298823529411765, + "grad_norm": 0.049722208101314586, + "learning_rate": 0.0002964, + "loss": 0.0234, + "step": 1976 + }, + { + "epoch": 0.9303529411764706, + "grad_norm": 0.1614368485018345, + "learning_rate": 0.00029654999999999997, + "loss": 0.1043, + "step": 1977 + }, + { + "epoch": 0.9308235294117647, + "grad_norm": 0.0712374617167679, + "learning_rate": 0.00029669999999999995, + "loss": 0.0357, + "step": 1978 + }, + { + "epoch": 0.9312941176470588, + "grad_norm": 0.10586877220419214, + "learning_rate": 0.00029685, + "loss": 0.0509, + "step": 1979 + }, + { + "epoch": 0.9317647058823529, + "grad_norm": 0.0031379488940866244, + "learning_rate": 0.00029699999999999996, + "loss": 0.0, + "step": 1980 + }, + { + "epoch": 0.932235294117647, + "grad_norm": 0.00839439566547431, + "learning_rate": 0.00029715, + "loss": 0.0001, + "step": 1981 + }, + { + "epoch": 0.9327058823529412, + "grad_norm": 0.2090806156053248, + "learning_rate": 0.00029729999999999996, + "loss": 0.1029, + "step": 1982 + }, + { + "epoch": 0.9331764705882353, + "grad_norm": 0.14847678790604008, + "learning_rate": 0.00029745, + "loss": 0.0565, + "step": 1983 + }, + { + "epoch": 0.9336470588235294, + "grad_norm": 0.0020167673798485693, + "learning_rate": 0.00029759999999999997, + "loss": 0.0, + "step": 1984 + }, + { + "epoch": 0.9341176470588235, + "grad_norm": 0.07986330020555178, + "learning_rate": 0.00029775, + "loss": 0.0349, + "step": 1985 + }, + { + "epoch": 0.9345882352941176, + "grad_norm": 0.17114719244753643, + "learning_rate": 0.0002979, + "loss": 0.092, + "step": 1986 + }, + { + "epoch": 0.9350588235294117, + "grad_norm": 0.13694385923995933, + "learning_rate": 0.00029805, + "loss": 0.062, + "step": 1987 + }, + { + "epoch": 0.9355294117647058, + "grad_norm": 0.16552936005950727, + "learning_rate": 0.0002982, + "loss": 0.1225, + "step": 1988 + }, + { + "epoch": 0.936, + "grad_norm": 0.16406165573478979, + "learning_rate": 0.00029835, + "loss": 0.1367, + "step": 1989 + }, + { + "epoch": 0.9364705882352942, + "grad_norm": 0.08835637967186448, + "learning_rate": 0.0002985, + "loss": 0.0562, + "step": 1990 + }, + { + "epoch": 0.9369411764705883, + "grad_norm": 0.10047552479322666, + "learning_rate": 0.00029864999999999997, + "loss": 0.0484, + "step": 1991 + }, + { + "epoch": 0.9374117647058824, + "grad_norm": 0.1295823082882088, + "learning_rate": 0.0002988, + "loss": 0.0793, + "step": 1992 + }, + { + "epoch": 0.9378823529411765, + "grad_norm": 0.00017680745968608552, + "learning_rate": 0.00029895, + "loss": 0.0, + "step": 1993 + }, + { + "epoch": 0.9383529411764706, + "grad_norm": 0.10378482051953221, + "learning_rate": 0.00029909999999999995, + "loss": 0.0658, + "step": 1994 + }, + { + "epoch": 0.9388235294117647, + "grad_norm": 0.1105841112862019, + "learning_rate": 0.00029925, + "loss": 0.0648, + "step": 1995 + }, + { + "epoch": 0.9392941176470588, + "grad_norm": 0.000155715160264628, + "learning_rate": 0.00029939999999999996, + "loss": 0.0, + "step": 1996 + }, + { + "epoch": 0.939764705882353, + "grad_norm": 0.1379038971276799, + "learning_rate": 0.00029955, + "loss": 0.0858, + "step": 1997 + }, + { + "epoch": 0.9402352941176471, + "grad_norm": 0.08458008522927433, + "learning_rate": 0.00029969999999999997, + "loss": 0.0355, + "step": 1998 + }, + { + "epoch": 0.9407058823529412, + "grad_norm": 0.07235100576556883, + "learning_rate": 0.00029985, + "loss": 0.0308, + "step": 1999 + }, + { + "epoch": 0.9411764705882353, + "grad_norm": 0.22382094492428278, + "learning_rate": 0.0003, + "loss": 0.1218, + "step": 2000 + }, + { + "epoch": 0.9416470588235294, + "grad_norm": 0.2202228330138317, + "learning_rate": 0.0002999999935946321, + "loss": 0.0966, + "step": 2001 + }, + { + "epoch": 0.9421176470588235, + "grad_norm": 0.05620804230342866, + "learning_rate": 0.00029999997437852905, + "loss": 0.0192, + "step": 2002 + }, + { + "epoch": 0.9425882352941176, + "grad_norm": 0.12244759224987045, + "learning_rate": 0.0002999999423516924, + "loss": 0.1201, + "step": 2003 + }, + { + "epoch": 0.9430588235294117, + "grad_norm": 0.10846278186876683, + "learning_rate": 0.00029999989751412494, + "loss": 0.0381, + "step": 2004 + }, + { + "epoch": 0.9435294117647058, + "grad_norm": 0.590868992562441, + "learning_rate": 0.00029999983986583054, + "loss": 0.1208, + "step": 2005 + }, + { + "epoch": 0.944, + "grad_norm": 0.00010592077411697861, + "learning_rate": 0.000299999769406814, + "loss": 0.0, + "step": 2006 + }, + { + "epoch": 0.9444705882352942, + "grad_norm": 0.08056153016333226, + "learning_rate": 0.0002999996861370814, + "loss": 0.041, + "step": 2007 + }, + { + "epoch": 0.9449411764705883, + "grad_norm": 0.058254295776222584, + "learning_rate": 0.0002999995900566399, + "loss": 0.0204, + "step": 2008 + }, + { + "epoch": 0.9454117647058824, + "grad_norm": 0.19714081625401092, + "learning_rate": 0.0002999994811654977, + "loss": 0.0997, + "step": 2009 + }, + { + "epoch": 0.9458823529411765, + "grad_norm": 0.20052001431439007, + "learning_rate": 0.000299999359463664, + "loss": 0.1181, + "step": 2010 + }, + { + "epoch": 0.9463529411764706, + "grad_norm": 0.0914987068585103, + "learning_rate": 0.0002999992249511493, + "loss": 0.0437, + "step": 2011 + }, + { + "epoch": 0.9468235294117647, + "grad_norm": 0.10866897915924152, + "learning_rate": 0.00029999907762796505, + "loss": 0.0502, + "step": 2012 + }, + { + "epoch": 0.9472941176470588, + "grad_norm": 0.00012660636558328226, + "learning_rate": 0.00029999891749412375, + "loss": 0.0, + "step": 2013 + }, + { + "epoch": 0.947764705882353, + "grad_norm": 0.13821138912588377, + "learning_rate": 0.0002999987445496392, + "loss": 0.0813, + "step": 2014 + }, + { + "epoch": 0.9482352941176471, + "grad_norm": 0.13047322701467293, + "learning_rate": 0.00029999855879452616, + "loss": 0.0822, + "step": 2015 + }, + { + "epoch": 0.9487058823529412, + "grad_norm": 0.09942919487595117, + "learning_rate": 0.00029999836022880045, + "loss": 0.0605, + "step": 2016 + }, + { + "epoch": 0.9491764705882353, + "grad_norm": 0.00012302443185633764, + "learning_rate": 0.00029999814885247906, + "loss": 0.0, + "step": 2017 + }, + { + "epoch": 0.9496470588235294, + "grad_norm": 0.15577463798737753, + "learning_rate": 0.00029999792466558, + "loss": 0.102, + "step": 2018 + }, + { + "epoch": 0.9501176470588235, + "grad_norm": 0.08067612721421187, + "learning_rate": 0.0002999976876681224, + "loss": 0.0657, + "step": 2019 + }, + { + "epoch": 0.9505882352941176, + "grad_norm": 0.1258754072059517, + "learning_rate": 0.00029999743786012656, + "loss": 0.0686, + "step": 2020 + }, + { + "epoch": 0.9510588235294117, + "grad_norm": 0.19370886597176207, + "learning_rate": 0.0002999971752416138, + "loss": 0.1555, + "step": 2021 + }, + { + "epoch": 0.9515294117647058, + "grad_norm": 0.11190798458177832, + "learning_rate": 0.00029999689981260655, + "loss": 0.0826, + "step": 2022 + }, + { + "epoch": 0.952, + "grad_norm": 0.06041367571126895, + "learning_rate": 0.0002999966115731283, + "loss": 0.0247, + "step": 2023 + }, + { + "epoch": 0.9524705882352941, + "grad_norm": 0.19153003497364546, + "learning_rate": 0.00029999631052320374, + "loss": 0.1113, + "step": 2024 + }, + { + "epoch": 0.9529411764705882, + "grad_norm": 0.13102799548735103, + "learning_rate": 0.0002999959966628585, + "loss": 0.074, + "step": 2025 + }, + { + "epoch": 0.9534117647058824, + "grad_norm": 0.00011564273442302102, + "learning_rate": 0.0002999956699921194, + "loss": 0.0, + "step": 2026 + }, + { + "epoch": 0.9538823529411765, + "grad_norm": 0.0645374718444627, + "learning_rate": 0.0002999953305110144, + "loss": 0.0176, + "step": 2027 + }, + { + "epoch": 0.9543529411764706, + "grad_norm": 0.18101967338821676, + "learning_rate": 0.0002999949782195724, + "loss": 0.0719, + "step": 2028 + }, + { + "epoch": 0.9548235294117647, + "grad_norm": 0.3100093020849019, + "learning_rate": 0.0002999946131178236, + "loss": 0.1158, + "step": 2029 + }, + { + "epoch": 0.9552941176470588, + "grad_norm": 0.2591283253412002, + "learning_rate": 0.0002999942352057991, + "loss": 0.1769, + "step": 2030 + }, + { + "epoch": 0.955764705882353, + "grad_norm": 0.00011803552523296613, + "learning_rate": 0.00029999384448353114, + "loss": 0.0, + "step": 2031 + }, + { + "epoch": 0.9562352941176471, + "grad_norm": 0.000122214667376361, + "learning_rate": 0.00029999344095105326, + "loss": 0.0, + "step": 2032 + }, + { + "epoch": 0.9567058823529412, + "grad_norm": 0.09858985880445661, + "learning_rate": 0.0002999930246083997, + "loss": 0.0413, + "step": 2033 + }, + { + "epoch": 0.9571764705882353, + "grad_norm": 0.15360206444708155, + "learning_rate": 0.0002999925954556062, + "loss": 0.0792, + "step": 2034 + }, + { + "epoch": 0.9576470588235294, + "grad_norm": 0.06743900285555247, + "learning_rate": 0.0002999921534927093, + "loss": 0.0277, + "step": 2035 + }, + { + "epoch": 0.9581176470588235, + "grad_norm": 0.00014029958594306981, + "learning_rate": 0.00029999169871974677, + "loss": 0.0, + "step": 2036 + }, + { + "epoch": 0.9585882352941176, + "grad_norm": 0.00013029848628353342, + "learning_rate": 0.0002999912311367575, + "loss": 0.0, + "step": 2037 + }, + { + "epoch": 0.9590588235294117, + "grad_norm": 0.1634653195182575, + "learning_rate": 0.00029999075074378136, + "loss": 0.0802, + "step": 2038 + }, + { + "epoch": 0.9595294117647059, + "grad_norm": 0.03706110280597206, + "learning_rate": 0.00029999025754085944, + "loss": 0.0073, + "step": 2039 + }, + { + "epoch": 0.96, + "grad_norm": 0.1285297778567277, + "learning_rate": 0.00029998975152803383, + "loss": 0.0739, + "step": 2040 + }, + { + "epoch": 0.9604705882352941, + "grad_norm": 0.09543302922270215, + "learning_rate": 0.0002999892327053477, + "loss": 0.0373, + "step": 2041 + }, + { + "epoch": 0.9609411764705882, + "grad_norm": 0.00022765358791364889, + "learning_rate": 0.0002999887010728454, + "loss": 0.0, + "step": 2042 + }, + { + "epoch": 0.9614117647058823, + "grad_norm": 0.10721248060985178, + "learning_rate": 0.0002999881566305724, + "loss": 0.0764, + "step": 2043 + }, + { + "epoch": 0.9618823529411765, + "grad_norm": 0.10343535783135376, + "learning_rate": 0.0002999875993785751, + "loss": 0.0644, + "step": 2044 + }, + { + "epoch": 0.9623529411764706, + "grad_norm": 0.06358888429094892, + "learning_rate": 0.0002999870293169011, + "loss": 0.0254, + "step": 2045 + }, + { + "epoch": 0.9628235294117647, + "grad_norm": 0.17221573043601873, + "learning_rate": 0.00029998644644559916, + "loss": 0.1873, + "step": 2046 + }, + { + "epoch": 0.9632941176470589, + "grad_norm": 0.11769391424957105, + "learning_rate": 0.000299985850764719, + "loss": 0.092, + "step": 2047 + }, + { + "epoch": 0.963764705882353, + "grad_norm": 0.14249082860123677, + "learning_rate": 0.00029998524227431146, + "loss": 0.1147, + "step": 2048 + }, + { + "epoch": 0.9642352941176471, + "grad_norm": 0.00018827506270685778, + "learning_rate": 0.00029998462097442866, + "loss": 0.0, + "step": 2049 + }, + { + "epoch": 0.9647058823529412, + "grad_norm": 0.0003279469901907598, + "learning_rate": 0.0002999839868651235, + "loss": 0.0, + "step": 2050 + }, + { + "epoch": 0.9651764705882353, + "grad_norm": 0.09415341132544329, + "learning_rate": 0.0002999833399464502, + "loss": 0.0383, + "step": 2051 + }, + { + "epoch": 0.9656470588235294, + "grad_norm": 0.16223148664052636, + "learning_rate": 0.000299982680218464, + "loss": 0.1255, + "step": 2052 + }, + { + "epoch": 0.9661176470588235, + "grad_norm": 0.000270716241265367, + "learning_rate": 0.00029998200768122123, + "loss": 0.0, + "step": 2053 + }, + { + "epoch": 0.9665882352941176, + "grad_norm": 0.23171138258622961, + "learning_rate": 0.00029998132233477936, + "loss": 0.1469, + "step": 2054 + }, + { + "epoch": 0.9670588235294117, + "grad_norm": 0.00021932021564105637, + "learning_rate": 0.0002999806241791969, + "loss": 0.0, + "step": 2055 + }, + { + "epoch": 0.9675294117647059, + "grad_norm": 0.22251492523025385, + "learning_rate": 0.0002999799132145335, + "loss": 0.1336, + "step": 2056 + }, + { + "epoch": 0.968, + "grad_norm": 0.0654492499826287, + "learning_rate": 0.00029997918944084986, + "loss": 0.0248, + "step": 2057 + }, + { + "epoch": 0.9684705882352941, + "grad_norm": 0.10911909554877004, + "learning_rate": 0.0002999784528582078, + "loss": 0.0643, + "step": 2058 + }, + { + "epoch": 0.9689411764705882, + "grad_norm": 0.19607673003179404, + "learning_rate": 0.0002999777034666702, + "loss": 0.1637, + "step": 2059 + }, + { + "epoch": 0.9694117647058823, + "grad_norm": 0.07378024886790242, + "learning_rate": 0.00029997694126630115, + "loss": 0.0318, + "step": 2060 + }, + { + "epoch": 0.9698823529411764, + "grad_norm": 0.039777225382759886, + "learning_rate": 0.00029997616625716567, + "loss": 0.0108, + "step": 2061 + }, + { + "epoch": 0.9703529411764706, + "grad_norm": 0.00018657180908272577, + "learning_rate": 0.0002999753784393299, + "loss": 0.0, + "step": 2062 + }, + { + "epoch": 0.9708235294117648, + "grad_norm": 0.08519260986875232, + "learning_rate": 0.00029997457781286125, + "loss": 0.0302, + "step": 2063 + }, + { + "epoch": 0.9712941176470589, + "grad_norm": 0.09167250246910823, + "learning_rate": 0.000299973764377828, + "loss": 0.0402, + "step": 2064 + }, + { + "epoch": 0.971764705882353, + "grad_norm": 0.0001933551367370677, + "learning_rate": 0.00029997293813429965, + "loss": 0.0, + "step": 2065 + }, + { + "epoch": 0.9722352941176471, + "grad_norm": 0.1335378598757244, + "learning_rate": 0.0002999720990823468, + "loss": 0.0776, + "step": 2066 + }, + { + "epoch": 0.9727058823529412, + "grad_norm": 0.10373890337589504, + "learning_rate": 0.0002999712472220411, + "loss": 0.079, + "step": 2067 + }, + { + "epoch": 0.9731764705882353, + "grad_norm": 0.11839948874989216, + "learning_rate": 0.0002999703825534552, + "loss": 0.0665, + "step": 2068 + }, + { + "epoch": 0.9736470588235294, + "grad_norm": 0.12043094650918129, + "learning_rate": 0.0002999695050766631, + "loss": 0.0563, + "step": 2069 + }, + { + "epoch": 0.9741176470588235, + "grad_norm": 0.11942866442109176, + "learning_rate": 0.0002999686147917396, + "loss": 0.0566, + "step": 2070 + }, + { + "epoch": 0.9745882352941176, + "grad_norm": 0.00023623098855551534, + "learning_rate": 0.0002999677116987609, + "loss": 0.0, + "step": 2071 + }, + { + "epoch": 0.9750588235294118, + "grad_norm": 0.11205473963294034, + "learning_rate": 0.00029996679579780394, + "loss": 0.0974, + "step": 2072 + }, + { + "epoch": 0.9755294117647059, + "grad_norm": 0.061757140857848905, + "learning_rate": 0.0002999658670889471, + "loss": 0.0264, + "step": 2073 + }, + { + "epoch": 0.976, + "grad_norm": 0.1332075781524337, + "learning_rate": 0.0002999649255722696, + "loss": 0.0702, + "step": 2074 + }, + { + "epoch": 0.9764705882352941, + "grad_norm": 0.18216557590815138, + "learning_rate": 0.0002999639712478519, + "loss": 0.1238, + "step": 2075 + }, + { + "epoch": 0.9769411764705882, + "grad_norm": 0.00014355884737658498, + "learning_rate": 0.00029996300411577547, + "loss": 0.0, + "step": 2076 + }, + { + "epoch": 0.9774117647058823, + "grad_norm": 0.11074854541579617, + "learning_rate": 0.00029996202417612294, + "loss": 0.0764, + "step": 2077 + }, + { + "epoch": 0.9778823529411764, + "grad_norm": 0.1396391946563065, + "learning_rate": 0.000299961031428978, + "loss": 0.0721, + "step": 2078 + }, + { + "epoch": 0.9783529411764705, + "grad_norm": 0.09939543494215121, + "learning_rate": 0.00029996002587442537, + "loss": 0.051, + "step": 2079 + }, + { + "epoch": 0.9788235294117648, + "grad_norm": 0.047253266478283734, + "learning_rate": 0.000299959007512551, + "loss": 0.0141, + "step": 2080 + }, + { + "epoch": 0.9792941176470589, + "grad_norm": 0.16247512359219024, + "learning_rate": 0.00029995797634344186, + "loss": 0.0929, + "step": 2081 + }, + { + "epoch": 0.979764705882353, + "grad_norm": 0.00012603486582819714, + "learning_rate": 0.000299956932367186, + "loss": 0.0, + "step": 2082 + }, + { + "epoch": 0.9802352941176471, + "grad_norm": 0.0001270316653031606, + "learning_rate": 0.0002999558755838726, + "loss": 0.0, + "step": 2083 + }, + { + "epoch": 0.9807058823529412, + "grad_norm": 0.08327626520189689, + "learning_rate": 0.00029995480599359186, + "loss": 0.0212, + "step": 2084 + }, + { + "epoch": 0.9811764705882353, + "grad_norm": 0.00020271450200689365, + "learning_rate": 0.00029995372359643515, + "loss": 0.0, + "step": 2085 + }, + { + "epoch": 0.9816470588235294, + "grad_norm": 0.11759943509532549, + "learning_rate": 0.00029995262839249497, + "loss": 0.0459, + "step": 2086 + }, + { + "epoch": 0.9821176470588235, + "grad_norm": 0.10614127821890466, + "learning_rate": 0.0002999515203818648, + "loss": 0.1055, + "step": 2087 + }, + { + "epoch": 0.9825882352941177, + "grad_norm": 0.0001345332378504843, + "learning_rate": 0.0002999503995646393, + "loss": 0.0, + "step": 2088 + }, + { + "epoch": 0.9830588235294118, + "grad_norm": 0.06562947956682112, + "learning_rate": 0.00029994926594091413, + "loss": 0.0306, + "step": 2089 + }, + { + "epoch": 0.9835294117647059, + "grad_norm": 0.057532421477068066, + "learning_rate": 0.00029994811951078616, + "loss": 0.0261, + "step": 2090 + }, + { + "epoch": 0.984, + "grad_norm": 0.06507184225221731, + "learning_rate": 0.0002999469602743533, + "loss": 0.0242, + "step": 2091 + }, + { + "epoch": 0.9844705882352941, + "grad_norm": 0.12866308431749204, + "learning_rate": 0.0002999457882317146, + "loss": 0.1194, + "step": 2092 + }, + { + "epoch": 0.9849411764705882, + "grad_norm": 0.06664376170747709, + "learning_rate": 0.00029994460338297, + "loss": 0.0308, + "step": 2093 + }, + { + "epoch": 0.9854117647058823, + "grad_norm": 0.05167234926650462, + "learning_rate": 0.0002999434057282209, + "loss": 0.0244, + "step": 2094 + }, + { + "epoch": 0.9858823529411764, + "grad_norm": 0.09031737343918976, + "learning_rate": 0.0002999421952675694, + "loss": 0.0653, + "step": 2095 + }, + { + "epoch": 0.9863529411764705, + "grad_norm": 0.00012038223889597523, + "learning_rate": 0.00029994097200111903, + "loss": 0.0, + "step": 2096 + }, + { + "epoch": 0.9868235294117647, + "grad_norm": 7.845028939797628e-05, + "learning_rate": 0.0002999397359289741, + "loss": 0.0, + "step": 2097 + }, + { + "epoch": 0.9872941176470589, + "grad_norm": 0.17677346837851812, + "learning_rate": 0.00029993848705124035, + "loss": 0.0836, + "step": 2098 + }, + { + "epoch": 0.987764705882353, + "grad_norm": 0.13018997248653308, + "learning_rate": 0.0002999372253680244, + "loss": 0.0859, + "step": 2099 + }, + { + "epoch": 0.9882352941176471, + "grad_norm": 0.29062083473043554, + "learning_rate": 0.0002999359508794339, + "loss": 0.0376, + "step": 2100 + }, + { + "epoch": 0.9887058823529412, + "grad_norm": 0.0728595019972137, + "learning_rate": 0.0002999346635855778, + "loss": 0.0262, + "step": 2101 + }, + { + "epoch": 0.9891764705882353, + "grad_norm": 0.13523891393053747, + "learning_rate": 0.00029993336348656594, + "loss": 0.1089, + "step": 2102 + }, + { + "epoch": 0.9896470588235294, + "grad_norm": 0.023639017313644317, + "learning_rate": 0.00029993205058250947, + "loss": 0.0102, + "step": 2103 + }, + { + "epoch": 0.9901176470588235, + "grad_norm": 0.07078609530492622, + "learning_rate": 0.00029993072487352043, + "loss": 0.0294, + "step": 2104 + }, + { + "epoch": 0.9905882352941177, + "grad_norm": 0.13003495294383044, + "learning_rate": 0.0002999293863597121, + "loss": 0.0506, + "step": 2105 + }, + { + "epoch": 0.9910588235294118, + "grad_norm": 0.1572191712078713, + "learning_rate": 0.00029992803504119876, + "loss": 0.0983, + "step": 2106 + }, + { + "epoch": 0.9915294117647059, + "grad_norm": 0.10932912362289982, + "learning_rate": 0.0002999266709180959, + "loss": 0.07, + "step": 2107 + }, + { + "epoch": 0.992, + "grad_norm": 0.030168850350492167, + "learning_rate": 0.0002999252939905199, + "loss": 0.0047, + "step": 2108 + }, + { + "epoch": 0.9924705882352941, + "grad_norm": 0.13909921922825794, + "learning_rate": 0.0002999239042585884, + "loss": 0.0786, + "step": 2109 + }, + { + "epoch": 0.9929411764705882, + "grad_norm": 0.00018869193852937206, + "learning_rate": 0.0002999225017224201, + "loss": 0.0, + "step": 2110 + }, + { + "epoch": 0.9934117647058823, + "grad_norm": 0.3549463970458745, + "learning_rate": 0.0002999210863821349, + "loss": 0.094, + "step": 2111 + }, + { + "epoch": 0.9938823529411764, + "grad_norm": 0.21109843871913364, + "learning_rate": 0.0002999196582378535, + "loss": 0.1134, + "step": 2112 + }, + { + "epoch": 0.9943529411764706, + "grad_norm": 0.11622960957763297, + "learning_rate": 0.0002999182172896979, + "loss": 0.0457, + "step": 2113 + }, + { + "epoch": 0.9948235294117647, + "grad_norm": 0.08044910527935599, + "learning_rate": 0.0002999167635377912, + "loss": 0.02, + "step": 2114 + }, + { + "epoch": 0.9952941176470588, + "grad_norm": 0.00011794925128179109, + "learning_rate": 0.00029991529698225757, + "loss": 0.0, + "step": 2115 + }, + { + "epoch": 0.995764705882353, + "grad_norm": 0.03102783709308967, + "learning_rate": 0.0002999138176232223, + "loss": 0.0061, + "step": 2116 + }, + { + "epoch": 0.9962352941176471, + "grad_norm": 0.05494995121410092, + "learning_rate": 0.00029991232546081156, + "loss": 0.0162, + "step": 2117 + }, + { + "epoch": 0.9967058823529412, + "grad_norm": 0.10285152732585154, + "learning_rate": 0.000299910820495153, + "loss": 0.0258, + "step": 2118 + }, + { + "epoch": 0.9971764705882353, + "grad_norm": 0.13670032390529516, + "learning_rate": 0.00029990930272637506, + "loss": 0.0737, + "step": 2119 + }, + { + "epoch": 0.9976470588235294, + "grad_norm": 0.11268196620765708, + "learning_rate": 0.00029990777215460737, + "loss": 0.0515, + "step": 2120 + }, + { + "epoch": 0.9981176470588236, + "grad_norm": 0.22816659604147108, + "learning_rate": 0.0002999062287799806, + "loss": 0.1377, + "step": 2121 + }, + { + "epoch": 0.9985882352941177, + "grad_norm": 0.054734165390183755, + "learning_rate": 0.0002999046726026266, + "loss": 0.0155, + "step": 2122 + }, + { + "epoch": 0.9990588235294118, + "grad_norm": 3.997680365943417e-05, + "learning_rate": 0.00029990310362267833, + "loss": 0.0, + "step": 2123 + }, + { + "epoch": 0.9995294117647059, + "grad_norm": 0.24588925980965734, + "learning_rate": 0.0002999015218402697, + "loss": 0.1534, + "step": 2124 + }, + { + "epoch": 1.0, + "grad_norm": 0.14281550330855808, + "learning_rate": 0.0002998999272555359, + "loss": 0.0633, + "step": 2125 + }, + { + "epoch": 1.0, + "eval_loss": 0.059286605566740036, + "eval_runtime": 59.2411, + "eval_samples_per_second": 12.66, + "eval_steps_per_second": 6.33, + "step": 2125 + }, + { + "epoch": 1.0004705882352942, + "grad_norm": 0.07781138737495952, + "learning_rate": 0.000299898319868613, + "loss": 0.0328, + "step": 2126 + }, + { + "epoch": 1.0009411764705882, + "grad_norm": 0.14514789413062812, + "learning_rate": 0.00029989669967963835, + "loss": 0.0611, + "step": 2127 + }, + { + "epoch": 1.0014117647058824, + "grad_norm": 0.05643264289583865, + "learning_rate": 0.00029989506668875033, + "loss": 0.0206, + "step": 2128 + }, + { + "epoch": 1.0018823529411764, + "grad_norm": 0.10034631215417256, + "learning_rate": 0.00029989342089608835, + "loss": 0.0519, + "step": 2129 + }, + { + "epoch": 1.0023529411764707, + "grad_norm": 0.0001094903046460335, + "learning_rate": 0.000299891762301793, + "loss": 0.0, + "step": 2130 + }, + { + "epoch": 1.0028235294117647, + "grad_norm": 0.051573219998651665, + "learning_rate": 0.00029989009090600595, + "loss": 0.0198, + "step": 2131 + }, + { + "epoch": 1.003294117647059, + "grad_norm": 0.1295431364831027, + "learning_rate": 0.00029988840670887, + "loss": 0.0709, + "step": 2132 + }, + { + "epoch": 1.003764705882353, + "grad_norm": 0.32674610374105506, + "learning_rate": 0.0002998867097105288, + "loss": 0.2098, + "step": 2133 + }, + { + "epoch": 1.0042352941176471, + "grad_norm": 0.09317613905781476, + "learning_rate": 0.00029988499991112747, + "loss": 0.0388, + "step": 2134 + }, + { + "epoch": 1.0047058823529411, + "grad_norm": 0.07247106200523831, + "learning_rate": 0.000299883277310812, + "loss": 0.0288, + "step": 2135 + }, + { + "epoch": 1.0051764705882353, + "grad_norm": 0.16401638587060083, + "learning_rate": 0.0002998815419097294, + "loss": 0.0354, + "step": 2136 + }, + { + "epoch": 1.0056470588235293, + "grad_norm": 0.10356976685777747, + "learning_rate": 0.0002998797937080279, + "loss": 0.0514, + "step": 2137 + }, + { + "epoch": 1.0061176470588236, + "grad_norm": 0.07876814051108724, + "learning_rate": 0.000299878032705857, + "loss": 0.0338, + "step": 2138 + }, + { + "epoch": 1.0065882352941176, + "grad_norm": 0.09557936655862138, + "learning_rate": 0.0002998762589033669, + "loss": 0.0542, + "step": 2139 + }, + { + "epoch": 1.0070588235294118, + "grad_norm": 0.24931279310455937, + "learning_rate": 0.00029987447230070917, + "loss": 0.124, + "step": 2140 + }, + { + "epoch": 1.0075294117647058, + "grad_norm": 0.07050313384247804, + "learning_rate": 0.00029987267289803635, + "loss": 0.0277, + "step": 2141 + }, + { + "epoch": 1.008, + "grad_norm": 3.9678005472713755e-05, + "learning_rate": 0.00029987086069550216, + "loss": 0.0, + "step": 2142 + }, + { + "epoch": 1.008470588235294, + "grad_norm": 4.659293314338473e-05, + "learning_rate": 0.0002998690356932613, + "loss": 0.0, + "step": 2143 + }, + { + "epoch": 1.0089411764705882, + "grad_norm": 4.292322018203778e-05, + "learning_rate": 0.0002998671978914698, + "loss": 0.0, + "step": 2144 + }, + { + "epoch": 1.0094117647058825, + "grad_norm": 0.0639651793174012, + "learning_rate": 0.00029986534729028444, + "loss": 0.0435, + "step": 2145 + }, + { + "epoch": 1.0098823529411765, + "grad_norm": 0.1002809164989747, + "learning_rate": 0.00029986348388986333, + "loss": 0.0443, + "step": 2146 + }, + { + "epoch": 1.0103529411764707, + "grad_norm": 0.1287768323953373, + "learning_rate": 0.00029986160769036565, + "loss": 0.0983, + "step": 2147 + }, + { + "epoch": 1.0108235294117647, + "grad_norm": 0.14596492954994197, + "learning_rate": 0.0002998597186919515, + "loss": 0.0801, + "step": 2148 + }, + { + "epoch": 1.011294117647059, + "grad_norm": 5.4585885458315915e-05, + "learning_rate": 0.0002998578168947824, + "loss": 0.0, + "step": 2149 + }, + { + "epoch": 1.011764705882353, + "grad_norm": 0.0878671017645504, + "learning_rate": 0.00029985590229902073, + "loss": 0.0469, + "step": 2150 + }, + { + "epoch": 1.0122352941176471, + "grad_norm": 0.11253571480379816, + "learning_rate": 0.0002998539749048299, + "loss": 0.0875, + "step": 2151 + }, + { + "epoch": 1.0127058823529411, + "grad_norm": 0.09779736312258339, + "learning_rate": 0.0002998520347123746, + "loss": 0.1156, + "step": 2152 + }, + { + "epoch": 1.0131764705882353, + "grad_norm": 0.03641130023941014, + "learning_rate": 0.00029985008172182055, + "loss": 0.0106, + "step": 2153 + }, + { + "epoch": 1.0136470588235293, + "grad_norm": 0.7685380306960489, + "learning_rate": 0.00029984811593333446, + "loss": 0.1584, + "step": 2154 + }, + { + "epoch": 1.0141176470588236, + "grad_norm": 0.2166356914132082, + "learning_rate": 0.00029984613734708427, + "loss": 0.1407, + "step": 2155 + }, + { + "epoch": 1.0145882352941176, + "grad_norm": 0.13192451725419646, + "learning_rate": 0.00029984414596323896, + "loss": 0.0307, + "step": 2156 + }, + { + "epoch": 1.0150588235294118, + "grad_norm": 0.20000712775991605, + "learning_rate": 0.00029984214178196863, + "loss": 0.0403, + "step": 2157 + }, + { + "epoch": 1.0155294117647058, + "grad_norm": 0.00962862919185714, + "learning_rate": 0.00029984012480344444, + "loss": 0.0001, + "step": 2158 + }, + { + "epoch": 1.016, + "grad_norm": 0.2834476757825672, + "learning_rate": 0.0002998380950278386, + "loss": 0.0707, + "step": 2159 + }, + { + "epoch": 1.016470588235294, + "grad_norm": 0.1591962608717463, + "learning_rate": 0.00029983605245532453, + "loss": 0.0425, + "step": 2160 + }, + { + "epoch": 1.0169411764705882, + "grad_norm": 0.008447376818585985, + "learning_rate": 0.0002998339970860766, + "loss": 0.0002, + "step": 2161 + }, + { + "epoch": 1.0174117647058825, + "grad_norm": 0.19376655543533933, + "learning_rate": 0.00029983192892027037, + "loss": 0.1211, + "step": 2162 + }, + { + "epoch": 1.0178823529411765, + "grad_norm": 0.10208614458844621, + "learning_rate": 0.0002998298479580825, + "loss": 0.0433, + "step": 2163 + }, + { + "epoch": 1.0183529411764707, + "grad_norm": 0.1547379006383667, + "learning_rate": 0.0002998277541996907, + "loss": 0.1117, + "step": 2164 + }, + { + "epoch": 1.0188235294117647, + "grad_norm": 0.15888528800781995, + "learning_rate": 0.0002998256476452738, + "loss": 0.0758, + "step": 2165 + }, + { + "epoch": 1.019294117647059, + "grad_norm": 0.14058644318838892, + "learning_rate": 0.0002998235282950117, + "loss": 0.1034, + "step": 2166 + }, + { + "epoch": 1.019764705882353, + "grad_norm": 0.039577590731786676, + "learning_rate": 0.0002998213961490854, + "loss": 0.0133, + "step": 2167 + }, + { + "epoch": 1.0202352941176471, + "grad_norm": 0.12590337375650798, + "learning_rate": 0.00029981925120767705, + "loss": 0.0306, + "step": 2168 + }, + { + "epoch": 1.0207058823529411, + "grad_norm": 0.13158201641585346, + "learning_rate": 0.00029981709347096975, + "loss": 0.0672, + "step": 2169 + }, + { + "epoch": 1.0211764705882354, + "grad_norm": 0.10213890975762423, + "learning_rate": 0.00029981492293914777, + "loss": 0.0432, + "step": 2170 + }, + { + "epoch": 1.0216470588235294, + "grad_norm": 0.03336617265806572, + "learning_rate": 0.0002998127396123966, + "loss": 0.0075, + "step": 2171 + }, + { + "epoch": 1.0221176470588236, + "grad_norm": 0.4427774668198752, + "learning_rate": 0.00029981054349090264, + "loss": 0.0236, + "step": 2172 + }, + { + "epoch": 1.0225882352941176, + "grad_norm": 0.39522767597411385, + "learning_rate": 0.0002998083345748534, + "loss": 0.0967, + "step": 2173 + }, + { + "epoch": 1.0230588235294118, + "grad_norm": 0.06260951420662474, + "learning_rate": 0.0002998061128644376, + "loss": 0.0221, + "step": 2174 + }, + { + "epoch": 1.0235294117647058, + "grad_norm": 0.15971353087809684, + "learning_rate": 0.00029980387835984494, + "loss": 0.0534, + "step": 2175 + }, + { + "epoch": 1.024, + "grad_norm": 0.06822238636149659, + "learning_rate": 0.0002998016310612663, + "loss": 0.0238, + "step": 2176 + }, + { + "epoch": 1.024470588235294, + "grad_norm": 0.21292483066109083, + "learning_rate": 0.0002997993709688937, + "loss": 0.026, + "step": 2177 + }, + { + "epoch": 1.0249411764705882, + "grad_norm": 0.00012092787956343565, + "learning_rate": 0.0002997970980829199, + "loss": 0.0, + "step": 2178 + }, + { + "epoch": 1.0254117647058822, + "grad_norm": 0.14377953355467962, + "learning_rate": 0.0002997948124035393, + "loss": 0.0671, + "step": 2179 + }, + { + "epoch": 1.0258823529411765, + "grad_norm": 0.07671629571954044, + "learning_rate": 0.00029979251393094685, + "loss": 0.0527, + "step": 2180 + }, + { + "epoch": 1.0263529411764707, + "grad_norm": 0.263349729379691, + "learning_rate": 0.00029979020266533913, + "loss": 0.0173, + "step": 2181 + }, + { + "epoch": 1.0268235294117647, + "grad_norm": 0.07948414219792428, + "learning_rate": 0.0002997878786069133, + "loss": 0.0223, + "step": 2182 + }, + { + "epoch": 1.027294117647059, + "grad_norm": 0.33341128779317236, + "learning_rate": 0.0002997855417558679, + "loss": 0.0995, + "step": 2183 + }, + { + "epoch": 1.027764705882353, + "grad_norm": 0.08687892028721754, + "learning_rate": 0.0002997831921124026, + "loss": 0.0315, + "step": 2184 + }, + { + "epoch": 1.0282352941176471, + "grad_norm": 0.00023511419598438348, + "learning_rate": 0.000299780829676718, + "loss": 0.0, + "step": 2185 + }, + { + "epoch": 1.0287058823529411, + "grad_norm": 0.0001412187260443634, + "learning_rate": 0.00029977845444901586, + "loss": 0.0, + "step": 2186 + }, + { + "epoch": 1.0291764705882354, + "grad_norm": 0.1344277867192039, + "learning_rate": 0.0002997760664294991, + "loss": 0.0635, + "step": 2187 + }, + { + "epoch": 1.0296470588235294, + "grad_norm": 0.03895748295402818, + "learning_rate": 0.00029977366561837154, + "loss": 0.0097, + "step": 2188 + }, + { + "epoch": 1.0301176470588236, + "grad_norm": 0.05915624785195011, + "learning_rate": 0.0002997712520158383, + "loss": 0.0291, + "step": 2189 + }, + { + "epoch": 1.0305882352941176, + "grad_norm": 0.0864520290747099, + "learning_rate": 0.00029976882562210563, + "loss": 0.0692, + "step": 2190 + }, + { + "epoch": 1.0310588235294118, + "grad_norm": 0.0951247648093249, + "learning_rate": 0.00029976638643738055, + "loss": 0.0566, + "step": 2191 + }, + { + "epoch": 1.0315294117647058, + "grad_norm": 0.15367665004300096, + "learning_rate": 0.0002997639344618714, + "loss": 0.1177, + "step": 2192 + }, + { + "epoch": 1.032, + "grad_norm": 0.0001703402868174726, + "learning_rate": 0.0002997614696957878, + "loss": 0.0, + "step": 2193 + }, + { + "epoch": 1.032470588235294, + "grad_norm": 0.07363040496975241, + "learning_rate": 0.00029975899213934004, + "loss": 0.0389, + "step": 2194 + }, + { + "epoch": 1.0329411764705883, + "grad_norm": 0.07925008482584517, + "learning_rate": 0.0002997565017927398, + "loss": 0.065, + "step": 2195 + }, + { + "epoch": 1.0334117647058823, + "grad_norm": 0.06859593163205693, + "learning_rate": 0.00029975399865619977, + "loss": 0.0142, + "step": 2196 + }, + { + "epoch": 1.0338823529411765, + "grad_norm": 0.00021122608382334273, + "learning_rate": 0.00029975148272993373, + "loss": 0.0, + "step": 2197 + }, + { + "epoch": 1.0343529411764707, + "grad_norm": 0.00020848432128625925, + "learning_rate": 0.0002997489540141565, + "loss": 0.0, + "step": 2198 + }, + { + "epoch": 1.0348235294117647, + "grad_norm": 0.08841753186122608, + "learning_rate": 0.00029974641250908406, + "loss": 0.064, + "step": 2199 + }, + { + "epoch": 1.035294117647059, + "grad_norm": 0.19705181775008987, + "learning_rate": 0.0002997438582149335, + "loss": 0.1338, + "step": 2200 + }, + { + "epoch": 1.035764705882353, + "grad_norm": 0.11901902119929081, + "learning_rate": 0.000299741291131923, + "loss": 0.0888, + "step": 2201 + }, + { + "epoch": 1.0362352941176471, + "grad_norm": 0.06662101278734629, + "learning_rate": 0.00029973871126027173, + "loss": 0.0357, + "step": 2202 + }, + { + "epoch": 1.0367058823529411, + "grad_norm": 0.05971781401580144, + "learning_rate": 0.00029973611860020013, + "loss": 0.0266, + "step": 2203 + }, + { + "epoch": 1.0371764705882354, + "grad_norm": 0.00015716464521473817, + "learning_rate": 0.0002997335131519294, + "loss": 0.0, + "step": 2204 + }, + { + "epoch": 1.0376470588235294, + "grad_norm": 0.20821449418737817, + "learning_rate": 0.0002997308949156823, + "loss": 0.1627, + "step": 2205 + }, + { + "epoch": 1.0381176470588236, + "grad_norm": 0.0846894292320037, + "learning_rate": 0.00029972826389168235, + "loss": 0.0314, + "step": 2206 + }, + { + "epoch": 1.0385882352941176, + "grad_norm": 0.08192041744802686, + "learning_rate": 0.00029972562008015425, + "loss": 0.0437, + "step": 2207 + }, + { + "epoch": 1.0390588235294118, + "grad_norm": 0.16050304210714533, + "learning_rate": 0.0002997229634813238, + "loss": 0.0795, + "step": 2208 + }, + { + "epoch": 1.0395294117647058, + "grad_norm": 0.03528097453256743, + "learning_rate": 0.0002997202940954179, + "loss": 0.0107, + "step": 2209 + }, + { + "epoch": 1.04, + "grad_norm": 0.062188862805997044, + "learning_rate": 0.00029971761192266447, + "loss": 0.0274, + "step": 2210 + }, + { + "epoch": 1.040470588235294, + "grad_norm": 0.1274651364455585, + "learning_rate": 0.0002997149169632926, + "loss": 0.0682, + "step": 2211 + }, + { + "epoch": 1.0409411764705883, + "grad_norm": 0.0873375915888108, + "learning_rate": 0.0002997122092175325, + "loss": 0.0187, + "step": 2212 + }, + { + "epoch": 1.0414117647058823, + "grad_norm": 0.0002929999970531052, + "learning_rate": 0.00029970948868561547, + "loss": 0.0, + "step": 2213 + }, + { + "epoch": 1.0418823529411765, + "grad_norm": 0.07891584494095542, + "learning_rate": 0.00029970675536777365, + "loss": 0.0486, + "step": 2214 + }, + { + "epoch": 1.0423529411764705, + "grad_norm": 0.043523395862442506, + "learning_rate": 0.0002997040092642407, + "loss": 0.0224, + "step": 2215 + }, + { + "epoch": 1.0428235294117647, + "grad_norm": 9.705306605868976e-05, + "learning_rate": 0.00029970125037525106, + "loss": 0.0, + "step": 2216 + }, + { + "epoch": 1.043294117647059, + "grad_norm": 0.12314488242709776, + "learning_rate": 0.00029969847870104033, + "loss": 0.0424, + "step": 2217 + }, + { + "epoch": 1.043764705882353, + "grad_norm": 0.12652648087661011, + "learning_rate": 0.00029969569424184527, + "loss": 0.086, + "step": 2218 + }, + { + "epoch": 1.0442352941176472, + "grad_norm": 0.11025250022006902, + "learning_rate": 0.0002996928969979037, + "loss": 0.0707, + "step": 2219 + }, + { + "epoch": 1.0447058823529412, + "grad_norm": 0.00031701638324980444, + "learning_rate": 0.00029969008696945445, + "loss": 0.0, + "step": 2220 + }, + { + "epoch": 1.0451764705882354, + "grad_norm": 0.15241974875961306, + "learning_rate": 0.00029968726415673753, + "loss": 0.1486, + "step": 2221 + }, + { + "epoch": 1.0456470588235294, + "grad_norm": 0.052316947738376626, + "learning_rate": 0.00029968442855999406, + "loss": 0.0147, + "step": 2222 + }, + { + "epoch": 1.0461176470588236, + "grad_norm": 0.15477687460108724, + "learning_rate": 0.0002996815801794662, + "loss": 0.0963, + "step": 2223 + }, + { + "epoch": 1.0465882352941176, + "grad_norm": 0.039667673540531065, + "learning_rate": 0.0002996787190153972, + "loss": 0.0128, + "step": 2224 + }, + { + "epoch": 1.0470588235294118, + "grad_norm": 0.08503060125532058, + "learning_rate": 0.0002996758450680314, + "loss": 0.038, + "step": 2225 + }, + { + "epoch": 1.0475294117647058, + "grad_norm": 0.13001705927990537, + "learning_rate": 0.0002996729583376143, + "loss": 0.0826, + "step": 2226 + }, + { + "epoch": 1.048, + "grad_norm": 0.09458747677215211, + "learning_rate": 0.00029967005882439244, + "loss": 0.0612, + "step": 2227 + }, + { + "epoch": 1.048470588235294, + "grad_norm": 5.820569826519992e-05, + "learning_rate": 0.0002996671465286134, + "loss": 0.0, + "step": 2228 + }, + { + "epoch": 1.0489411764705883, + "grad_norm": 0.06850897286614135, + "learning_rate": 0.00029966422145052594, + "loss": 0.0268, + "step": 2229 + }, + { + "epoch": 1.0494117647058823, + "grad_norm": 0.00013702930969256588, + "learning_rate": 0.00029966128359037986, + "loss": 0.0, + "step": 2230 + }, + { + "epoch": 1.0498823529411765, + "grad_norm": 4.780129664639251e-05, + "learning_rate": 0.0002996583329484261, + "loss": 0.0, + "step": 2231 + }, + { + "epoch": 1.0503529411764705, + "grad_norm": 0.05067371318752623, + "learning_rate": 0.00029965536952491665, + "loss": 0.0232, + "step": 2232 + }, + { + "epoch": 1.0508235294117647, + "grad_norm": 0.20489359878054064, + "learning_rate": 0.00029965239332010454, + "loss": 0.1434, + "step": 2233 + }, + { + "epoch": 1.0512941176470587, + "grad_norm": 0.09351525365281409, + "learning_rate": 0.000299649404334244, + "loss": 0.0405, + "step": 2234 + }, + { + "epoch": 1.051764705882353, + "grad_norm": 0.024582481106256784, + "learning_rate": 0.0002996464025675903, + "loss": 0.0056, + "step": 2235 + }, + { + "epoch": 1.0522352941176472, + "grad_norm": 0.05374213387047464, + "learning_rate": 0.00029964338802039987, + "loss": 0.0205, + "step": 2236 + }, + { + "epoch": 1.0527058823529412, + "grad_norm": 0.08183400124873208, + "learning_rate": 0.00029964036069293006, + "loss": 0.0423, + "step": 2237 + }, + { + "epoch": 1.0531764705882354, + "grad_norm": 0.10311228659087397, + "learning_rate": 0.0002996373205854395, + "loss": 0.0069, + "step": 2238 + }, + { + "epoch": 1.0536470588235294, + "grad_norm": 0.1510905218041195, + "learning_rate": 0.00029963426769818774, + "loss": 0.0961, + "step": 2239 + }, + { + "epoch": 1.0541176470588236, + "grad_norm": 0.15427995845208126, + "learning_rate": 0.0002996312020314356, + "loss": 0.0891, + "step": 2240 + }, + { + "epoch": 1.0545882352941176, + "grad_norm": 0.14157422717874457, + "learning_rate": 0.00029962812358544485, + "loss": 0.0962, + "step": 2241 + }, + { + "epoch": 1.0550588235294118, + "grad_norm": 0.19708501175249568, + "learning_rate": 0.0002996250323604785, + "loss": 0.1428, + "step": 2242 + }, + { + "epoch": 1.0555294117647058, + "grad_norm": 2.6829971470049028e-05, + "learning_rate": 0.00029962192835680037, + "loss": 0.0, + "step": 2243 + }, + { + "epoch": 1.056, + "grad_norm": 0.12915878918249024, + "learning_rate": 0.0002996188115746757, + "loss": 0.1088, + "step": 2244 + }, + { + "epoch": 1.056470588235294, + "grad_norm": 0.07740512040112607, + "learning_rate": 0.0002996156820143707, + "loss": 0.0389, + "step": 2245 + }, + { + "epoch": 1.0569411764705883, + "grad_norm": 0.07723838645231898, + "learning_rate": 0.00029961253967615255, + "loss": 0.0274, + "step": 2246 + }, + { + "epoch": 1.0574117647058823, + "grad_norm": 0.20836844347399458, + "learning_rate": 0.0002996093845602897, + "loss": 0.1051, + "step": 2247 + }, + { + "epoch": 1.0578823529411765, + "grad_norm": 0.2638202957047167, + "learning_rate": 0.00029960621666705155, + "loss": 0.0472, + "step": 2248 + }, + { + "epoch": 1.0583529411764705, + "grad_norm": 0.06423129428351533, + "learning_rate": 0.00029960303599670866, + "loss": 0.024, + "step": 2249 + }, + { + "epoch": 1.0588235294117647, + "grad_norm": 0.14125492207163115, + "learning_rate": 0.0002995998425495327, + "loss": 0.1097, + "step": 2250 + }, + { + "epoch": 1.0592941176470587, + "grad_norm": 0.0015306269898773216, + "learning_rate": 0.0002995966363257964, + "loss": 0.0, + "step": 2251 + }, + { + "epoch": 1.059764705882353, + "grad_norm": 0.4218447375635817, + "learning_rate": 0.0002995934173257736, + "loss": 0.1485, + "step": 2252 + }, + { + "epoch": 1.0602352941176472, + "grad_norm": 0.00015696274649325362, + "learning_rate": 0.0002995901855497393, + "loss": 0.0, + "step": 2253 + }, + { + "epoch": 1.0607058823529412, + "grad_norm": 0.0012604402724780315, + "learning_rate": 0.00029958694099796934, + "loss": 0.0, + "step": 2254 + }, + { + "epoch": 1.0611764705882354, + "grad_norm": 0.18584920468313101, + "learning_rate": 0.0002995836836707409, + "loss": 0.092, + "step": 2255 + }, + { + "epoch": 1.0616470588235294, + "grad_norm": 0.0797707675505317, + "learning_rate": 0.00029958041356833217, + "loss": 0.0353, + "step": 2256 + }, + { + "epoch": 1.0621176470588236, + "grad_norm": 0.17901957556839346, + "learning_rate": 0.00029957713069102244, + "loss": 0.0527, + "step": 2257 + }, + { + "epoch": 1.0625882352941176, + "grad_norm": 0.05022847899121942, + "learning_rate": 0.00029957383503909204, + "loss": 0.0216, + "step": 2258 + }, + { + "epoch": 1.0630588235294118, + "grad_norm": 0.22037070469351805, + "learning_rate": 0.00029957052661282256, + "loss": 0.0523, + "step": 2259 + }, + { + "epoch": 1.0635294117647058, + "grad_norm": 0.5011645556992886, + "learning_rate": 0.0002995672054124964, + "loss": 0.102, + "step": 2260 + }, + { + "epoch": 1.064, + "grad_norm": 0.07203595851408802, + "learning_rate": 0.0002995638714383973, + "loss": 0.0202, + "step": 2261 + }, + { + "epoch": 1.064470588235294, + "grad_norm": 0.37780942882952334, + "learning_rate": 0.00029956052469080994, + "loss": 0.0632, + "step": 2262 + }, + { + "epoch": 1.0649411764705883, + "grad_norm": 2.55216665891596e-05, + "learning_rate": 0.0002995571651700202, + "loss": 0.0, + "step": 2263 + }, + { + "epoch": 1.0654117647058823, + "grad_norm": 1.478636414261241e-05, + "learning_rate": 0.000299553792876315, + "loss": 0.0, + "step": 2264 + }, + { + "epoch": 1.0658823529411765, + "grad_norm": 0.031004698082896624, + "learning_rate": 0.0002995504078099823, + "loss": 0.0103, + "step": 2265 + }, + { + "epoch": 1.0663529411764705, + "grad_norm": 0.14487902977314948, + "learning_rate": 0.00029954700997131126, + "loss": 0.0799, + "step": 2266 + }, + { + "epoch": 1.0668235294117647, + "grad_norm": 0.11020759802285748, + "learning_rate": 0.000299543599360592, + "loss": 0.0756, + "step": 2267 + }, + { + "epoch": 1.0672941176470587, + "grad_norm": 0.05022654286302082, + "learning_rate": 0.0002995401759781159, + "loss": 0.02, + "step": 2268 + }, + { + "epoch": 1.067764705882353, + "grad_norm": 0.056852795447879904, + "learning_rate": 0.00029953673982417527, + "loss": 0.0154, + "step": 2269 + }, + { + "epoch": 1.0682352941176472, + "grad_norm": 0.12067422275311031, + "learning_rate": 0.00029953329089906357, + "loss": 0.071, + "step": 2270 + }, + { + "epoch": 1.0687058823529412, + "grad_norm": 0.4663195096100154, + "learning_rate": 0.0002995298292030754, + "loss": 0.2017, + "step": 2271 + }, + { + "epoch": 1.0691764705882354, + "grad_norm": 0.07724737753278463, + "learning_rate": 0.00029952635473650633, + "loss": 0.0491, + "step": 2272 + }, + { + "epoch": 1.0696470588235294, + "grad_norm": 2.647996259198188e-05, + "learning_rate": 0.00029952286749965315, + "loss": 0.0, + "step": 2273 + }, + { + "epoch": 1.0701176470588236, + "grad_norm": 0.04287790751809607, + "learning_rate": 0.0002995193674928137, + "loss": 0.0131, + "step": 2274 + }, + { + "epoch": 1.0705882352941176, + "grad_norm": 0.09310348191994589, + "learning_rate": 0.0002995158547162869, + "loss": 0.0786, + "step": 2275 + }, + { + "epoch": 1.0710588235294118, + "grad_norm": 0.09728935284697095, + "learning_rate": 0.00029951232917037264, + "loss": 0.0541, + "step": 2276 + }, + { + "epoch": 1.0715294117647058, + "grad_norm": 0.07334891211177062, + "learning_rate": 0.0002995087908553722, + "loss": 0.0301, + "step": 2277 + }, + { + "epoch": 1.072, + "grad_norm": 0.07544938772201687, + "learning_rate": 0.00029950523977158766, + "loss": 0.0262, + "step": 2278 + }, + { + "epoch": 1.072470588235294, + "grad_norm": 0.051888399438448626, + "learning_rate": 0.00029950167591932233, + "loss": 0.016, + "step": 2279 + }, + { + "epoch": 1.0729411764705883, + "grad_norm": 0.21689097518959022, + "learning_rate": 0.0002994980992988806, + "loss": 0.1472, + "step": 2280 + }, + { + "epoch": 1.0734117647058823, + "grad_norm": 0.08413055389905069, + "learning_rate": 0.0002994945099105679, + "loss": 0.0382, + "step": 2281 + }, + { + "epoch": 1.0738823529411765, + "grad_norm": 0.09705756120016094, + "learning_rate": 0.00029949090775469077, + "loss": 0.0714, + "step": 2282 + }, + { + "epoch": 1.0743529411764705, + "grad_norm": 0.16496072597146627, + "learning_rate": 0.00029948729283155683, + "loss": 0.1129, + "step": 2283 + }, + { + "epoch": 1.0748235294117647, + "grad_norm": 0.06121524645680676, + "learning_rate": 0.0002994836651414749, + "loss": 0.0261, + "step": 2284 + }, + { + "epoch": 1.0752941176470587, + "grad_norm": 0.061146533304652, + "learning_rate": 0.00029948002468475475, + "loss": 0.029, + "step": 2285 + }, + { + "epoch": 1.075764705882353, + "grad_norm": 0.458768783513096, + "learning_rate": 0.0002994763714617073, + "loss": 0.0701, + "step": 2286 + }, + { + "epoch": 1.076235294117647, + "grad_norm": 0.16988402352407977, + "learning_rate": 0.00029947270547264456, + "loss": 0.1136, + "step": 2287 + }, + { + "epoch": 1.0767058823529412, + "grad_norm": 0.27967283325865255, + "learning_rate": 0.00029946902671787955, + "loss": 0.1263, + "step": 2288 + }, + { + "epoch": 1.0771764705882352, + "grad_norm": 0.17228332471491792, + "learning_rate": 0.00029946533519772657, + "loss": 0.0725, + "step": 2289 + }, + { + "epoch": 1.0776470588235294, + "grad_norm": 0.0001263014536940672, + "learning_rate": 0.0002994616309125008, + "loss": 0.0, + "step": 2290 + }, + { + "epoch": 1.0781176470588236, + "grad_norm": 0.49592945398265453, + "learning_rate": 0.0002994579138625187, + "loss": 0.1351, + "step": 2291 + }, + { + "epoch": 1.0785882352941176, + "grad_norm": 0.10331724316037595, + "learning_rate": 0.00029945418404809767, + "loss": 0.0153, + "step": 2292 + }, + { + "epoch": 1.0790588235294118, + "grad_norm": 0.052207297412918865, + "learning_rate": 0.0002994504414695562, + "loss": 0.0182, + "step": 2293 + }, + { + "epoch": 1.0795294117647058, + "grad_norm": 0.00014465126985276755, + "learning_rate": 0.000299446686127214, + "loss": 0.0, + "step": 2294 + }, + { + "epoch": 1.08, + "grad_norm": 0.0002689067233396105, + "learning_rate": 0.00029944291802139184, + "loss": 0.0, + "step": 2295 + }, + { + "epoch": 1.080470588235294, + "grad_norm": 0.08821310636876367, + "learning_rate": 0.00029943913715241137, + "loss": 0.0249, + "step": 2296 + }, + { + "epoch": 1.0809411764705883, + "grad_norm": 0.1448533245884381, + "learning_rate": 0.0002994353435205956, + "loss": 0.0599, + "step": 2297 + }, + { + "epoch": 1.0814117647058823, + "grad_norm": 0.10248724742066338, + "learning_rate": 0.00029943153712626855, + "loss": 0.0468, + "step": 2298 + }, + { + "epoch": 1.0818823529411765, + "grad_norm": 0.08621096029702734, + "learning_rate": 0.0002994277179697553, + "loss": 0.0384, + "step": 2299 + }, + { + "epoch": 1.0823529411764705, + "grad_norm": 0.19910360975908742, + "learning_rate": 0.000299423886051382, + "loss": 0.0688, + "step": 2300 + }, + { + "epoch": 1.0828235294117647, + "grad_norm": 0.2649074554230692, + "learning_rate": 0.00029942004137147586, + "loss": 0.1543, + "step": 2301 + }, + { + "epoch": 1.0832941176470587, + "grad_norm": 0.03835871155421774, + "learning_rate": 0.0002994161839303653, + "loss": 0.0165, + "step": 2302 + }, + { + "epoch": 1.083764705882353, + "grad_norm": 0.1877989924515068, + "learning_rate": 0.00029941231372837974, + "loss": 0.095, + "step": 2303 + }, + { + "epoch": 1.084235294117647, + "grad_norm": 0.00032574292404106763, + "learning_rate": 0.0002994084307658497, + "loss": 0.0, + "step": 2304 + }, + { + "epoch": 1.0847058823529412, + "grad_norm": 0.08243641650179605, + "learning_rate": 0.0002994045350431069, + "loss": 0.0522, + "step": 2305 + }, + { + "epoch": 1.0851764705882352, + "grad_norm": 0.0005192682791212073, + "learning_rate": 0.00029940062656048396, + "loss": 0.0, + "step": 2306 + }, + { + "epoch": 1.0856470588235294, + "grad_norm": 0.001372180501897422, + "learning_rate": 0.0002993967053183147, + "loss": 0.0, + "step": 2307 + }, + { + "epoch": 1.0861176470588236, + "grad_norm": 0.05125895511490301, + "learning_rate": 0.000299392771316934, + "loss": 0.0167, + "step": 2308 + }, + { + "epoch": 1.0865882352941176, + "grad_norm": 0.11933622195737614, + "learning_rate": 0.00029938882455667786, + "loss": 0.0577, + "step": 2309 + }, + { + "epoch": 1.0870588235294119, + "grad_norm": 0.0010142572865975091, + "learning_rate": 0.00029938486503788337, + "loss": 0.0, + "step": 2310 + }, + { + "epoch": 1.0875294117647059, + "grad_norm": 0.0012307290043791909, + "learning_rate": 0.0002993808927608887, + "loss": 0.0, + "step": 2311 + }, + { + "epoch": 1.088, + "grad_norm": 0.1171269119002656, + "learning_rate": 0.000299376907726033, + "loss": 0.0513, + "step": 2312 + }, + { + "epoch": 1.088470588235294, + "grad_norm": 0.004547435858599383, + "learning_rate": 0.0002993729099336568, + "loss": 0.0001, + "step": 2313 + }, + { + "epoch": 1.0889411764705883, + "grad_norm": 0.8387887370669702, + "learning_rate": 0.00029936889938410137, + "loss": 0.0684, + "step": 2314 + }, + { + "epoch": 1.0894117647058823, + "grad_norm": 0.011036630003289858, + "learning_rate": 0.0002993648760777092, + "loss": 0.0, + "step": 2315 + }, + { + "epoch": 1.0898823529411765, + "grad_norm": 0.11570534348567761, + "learning_rate": 0.0002993608400148241, + "loss": 0.05, + "step": 2316 + }, + { + "epoch": 1.0903529411764705, + "grad_norm": 0.09733936585354254, + "learning_rate": 0.00029935679119579066, + "loss": 0.0482, + "step": 2317 + }, + { + "epoch": 1.0908235294117647, + "grad_norm": 0.024689632647570854, + "learning_rate": 0.0002993527296209546, + "loss": 0.0058, + "step": 2318 + }, + { + "epoch": 1.0912941176470587, + "grad_norm": 0.1370942847841445, + "learning_rate": 0.0002993486552906629, + "loss": 0.0877, + "step": 2319 + }, + { + "epoch": 1.091764705882353, + "grad_norm": 0.0006202464157086339, + "learning_rate": 0.00029934456820526344, + "loss": 0.0, + "step": 2320 + }, + { + "epoch": 1.092235294117647, + "grad_norm": 0.004495565504568113, + "learning_rate": 0.00029934046836510537, + "loss": 0.0001, + "step": 2321 + }, + { + "epoch": 1.0927058823529412, + "grad_norm": 0.16675090867378822, + "learning_rate": 0.00029933635577053877, + "loss": 0.0845, + "step": 2322 + }, + { + "epoch": 1.0931764705882352, + "grad_norm": 0.14261287115253027, + "learning_rate": 0.00029933223042191487, + "loss": 0.1084, + "step": 2323 + }, + { + "epoch": 1.0936470588235294, + "grad_norm": 0.11827998145191358, + "learning_rate": 0.000299328092319586, + "loss": 0.113, + "step": 2324 + }, + { + "epoch": 1.0941176470588236, + "grad_norm": 0.0009478342068102063, + "learning_rate": 0.0002993239414639057, + "loss": 0.0, + "step": 2325 + }, + { + "epoch": 1.0945882352941176, + "grad_norm": 0.12513844776133057, + "learning_rate": 0.0002993197778552283, + "loss": 0.0747, + "step": 2326 + }, + { + "epoch": 1.0950588235294119, + "grad_norm": 0.11860158929693825, + "learning_rate": 0.0002993156014939095, + "loss": 0.0909, + "step": 2327 + }, + { + "epoch": 1.0955294117647059, + "grad_norm": 0.114467547384511, + "learning_rate": 0.00029931141238030587, + "loss": 0.0632, + "step": 2328 + }, + { + "epoch": 1.096, + "grad_norm": 0.0600113427752681, + "learning_rate": 0.00029930721051477533, + "loss": 0.0277, + "step": 2329 + }, + { + "epoch": 1.096470588235294, + "grad_norm": 0.21117186920996103, + "learning_rate": 0.0002993029958976766, + "loss": 0.1425, + "step": 2330 + }, + { + "epoch": 1.0969411764705883, + "grad_norm": 0.1357492591959589, + "learning_rate": 0.00029929876852936973, + "loss": 0.0684, + "step": 2331 + }, + { + "epoch": 1.0974117647058823, + "grad_norm": 0.14026740299239535, + "learning_rate": 0.00029929452841021573, + "loss": 0.1019, + "step": 2332 + }, + { + "epoch": 1.0978823529411765, + "grad_norm": 0.12655645455646555, + "learning_rate": 0.00029929027554057675, + "loss": 0.0695, + "step": 2333 + }, + { + "epoch": 1.0983529411764705, + "grad_norm": 0.12644806369698328, + "learning_rate": 0.0002992860099208159, + "loss": 0.0672, + "step": 2334 + }, + { + "epoch": 1.0988235294117648, + "grad_norm": 9.927834069913429e-05, + "learning_rate": 0.0002992817315512976, + "loss": 0.0, + "step": 2335 + }, + { + "epoch": 1.0992941176470588, + "grad_norm": 0.15552062866718827, + "learning_rate": 0.0002992774404323872, + "loss": 0.0813, + "step": 2336 + }, + { + "epoch": 1.099764705882353, + "grad_norm": 0.01942066190656366, + "learning_rate": 0.0002992731365644512, + "loss": 0.0044, + "step": 2337 + }, + { + "epoch": 1.100235294117647, + "grad_norm": 0.08640118005071946, + "learning_rate": 0.0002992688199478571, + "loss": 0.044, + "step": 2338 + }, + { + "epoch": 1.1007058823529412, + "grad_norm": 0.21316851196784165, + "learning_rate": 0.00029926449058297364, + "loss": 0.1557, + "step": 2339 + }, + { + "epoch": 1.1011764705882352, + "grad_norm": 0.0001047827700733144, + "learning_rate": 0.0002992601484701706, + "loss": 0.0, + "step": 2340 + }, + { + "epoch": 1.1016470588235294, + "grad_norm": 7.254061601639532e-05, + "learning_rate": 0.0002992557936098187, + "loss": 0.0, + "step": 2341 + }, + { + "epoch": 1.1021176470588236, + "grad_norm": 0.1384776066483524, + "learning_rate": 0.00029925142600229, + "loss": 0.1041, + "step": 2342 + }, + { + "epoch": 1.1025882352941176, + "grad_norm": 0.08753725393316734, + "learning_rate": 0.00029924704564795735, + "loss": 0.026, + "step": 2343 + }, + { + "epoch": 1.1030588235294116, + "grad_norm": 0.00014592224232268177, + "learning_rate": 0.000299242652547195, + "loss": 0.0, + "step": 2344 + }, + { + "epoch": 1.1035294117647059, + "grad_norm": 6.457252952880864e-05, + "learning_rate": 0.0002992382467003781, + "loss": 0.0, + "step": 2345 + }, + { + "epoch": 1.104, + "grad_norm": 0.1312880901958213, + "learning_rate": 0.00029923382810788287, + "loss": 0.0948, + "step": 2346 + }, + { + "epoch": 1.104470588235294, + "grad_norm": 0.1457925856051436, + "learning_rate": 0.0002992293967700868, + "loss": 0.0829, + "step": 2347 + }, + { + "epoch": 1.1049411764705883, + "grad_norm": 0.11926328770939411, + "learning_rate": 0.00029922495268736824, + "loss": 0.0667, + "step": 2348 + }, + { + "epoch": 1.1054117647058823, + "grad_norm": 0.07570136372640324, + "learning_rate": 0.0002992204958601068, + "loss": 0.0328, + "step": 2349 + }, + { + "epoch": 1.1058823529411765, + "grad_norm": 0.01751793986202705, + "learning_rate": 0.0002992160262886831, + "loss": 0.0036, + "step": 2350 + }, + { + "epoch": 1.1063529411764705, + "grad_norm": 4.674788856889534e-05, + "learning_rate": 0.0002992115439734788, + "loss": 0.0, + "step": 2351 + }, + { + "epoch": 1.1068235294117648, + "grad_norm": 0.07924733006558345, + "learning_rate": 0.0002992070489148768, + "loss": 0.0572, + "step": 2352 + }, + { + "epoch": 1.1072941176470588, + "grad_norm": 0.049633394340110264, + "learning_rate": 0.000299202541113261, + "loss": 0.0158, + "step": 2353 + }, + { + "epoch": 1.107764705882353, + "grad_norm": 0.08975341873010481, + "learning_rate": 0.0002991980205690163, + "loss": 0.0499, + "step": 2354 + }, + { + "epoch": 1.108235294117647, + "grad_norm": 3.2185595036894904e-05, + "learning_rate": 0.0002991934872825289, + "loss": 0.0, + "step": 2355 + }, + { + "epoch": 1.1087058823529412, + "grad_norm": 0.11736291052169252, + "learning_rate": 0.0002991889412541858, + "loss": 0.0535, + "step": 2356 + }, + { + "epoch": 1.1091764705882352, + "grad_norm": 0.1953796572287713, + "learning_rate": 0.0002991843824843754, + "loss": 0.1052, + "step": 2357 + }, + { + "epoch": 1.1096470588235294, + "grad_norm": 0.08871821513264759, + "learning_rate": 0.000299179810973487, + "loss": 0.0403, + "step": 2358 + }, + { + "epoch": 1.1101176470588234, + "grad_norm": 0.11350714855797149, + "learning_rate": 0.000299175226721911, + "loss": 0.0661, + "step": 2359 + }, + { + "epoch": 1.1105882352941177, + "grad_norm": 0.10847428156501306, + "learning_rate": 0.00029917062973003897, + "loss": 0.0818, + "step": 2360 + }, + { + "epoch": 1.1110588235294117, + "grad_norm": 0.0697496649195065, + "learning_rate": 0.0002991660199982634, + "loss": 0.0306, + "step": 2361 + }, + { + "epoch": 1.1115294117647059, + "grad_norm": 0.1270955577998242, + "learning_rate": 0.0002991613975269782, + "loss": 0.0754, + "step": 2362 + }, + { + "epoch": 1.112, + "grad_norm": 0.08909743950307657, + "learning_rate": 0.0002991567623165779, + "loss": 0.0245, + "step": 2363 + }, + { + "epoch": 1.112470588235294, + "grad_norm": 0.10958853124821424, + "learning_rate": 0.00029915211436745855, + "loss": 0.0687, + "step": 2364 + }, + { + "epoch": 1.1129411764705883, + "grad_norm": 0.11361331437011991, + "learning_rate": 0.00029914745368001704, + "loss": 0.0615, + "step": 2365 + }, + { + "epoch": 1.1134117647058823, + "grad_norm": 0.06463023252268497, + "learning_rate": 0.0002991427802546514, + "loss": 0.0283, + "step": 2366 + }, + { + "epoch": 1.1138823529411765, + "grad_norm": 0.1313124587058758, + "learning_rate": 0.0002991380940917608, + "loss": 0.0785, + "step": 2367 + }, + { + "epoch": 1.1143529411764705, + "grad_norm": 0.18192126326066255, + "learning_rate": 0.0002991333951917454, + "loss": 0.108, + "step": 2368 + }, + { + "epoch": 1.1148235294117648, + "grad_norm": 0.10084339472254712, + "learning_rate": 0.00029912868355500666, + "loss": 0.0448, + "step": 2369 + }, + { + "epoch": 1.1152941176470588, + "grad_norm": 1.318605010033143e-05, + "learning_rate": 0.0002991239591819468, + "loss": 0.0, + "step": 2370 + }, + { + "epoch": 1.115764705882353, + "grad_norm": 0.06834550843215628, + "learning_rate": 0.00029911922207296937, + "loss": 0.03, + "step": 2371 + }, + { + "epoch": 1.116235294117647, + "grad_norm": 0.09252459362531133, + "learning_rate": 0.00029911447222847893, + "loss": 0.0956, + "step": 2372 + }, + { + "epoch": 1.1167058823529412, + "grad_norm": 0.11134816165010553, + "learning_rate": 0.00029910970964888126, + "loss": 0.0627, + "step": 2373 + }, + { + "epoch": 1.1171764705882352, + "grad_norm": 7.588908399926721e-06, + "learning_rate": 0.0002991049343345829, + "loss": 0.0, + "step": 2374 + }, + { + "epoch": 1.1176470588235294, + "grad_norm": 0.05706932857800119, + "learning_rate": 0.00029910014628599184, + "loss": 0.0242, + "step": 2375 + }, + { + "epoch": 1.1181176470588234, + "grad_norm": 0.09424212778672075, + "learning_rate": 0.00029909534550351695, + "loss": 0.0479, + "step": 2376 + }, + { + "epoch": 1.1185882352941177, + "grad_norm": 0.1163135527838731, + "learning_rate": 0.00029909053198756827, + "loss": 0.0724, + "step": 2377 + }, + { + "epoch": 1.1190588235294117, + "grad_norm": 0.1759620739419542, + "learning_rate": 0.00029908570573855683, + "loss": 0.1055, + "step": 2378 + }, + { + "epoch": 1.1195294117647059, + "grad_norm": 6.749703311053271e-06, + "learning_rate": 0.0002990808667568949, + "loss": 0.0, + "step": 2379 + }, + { + "epoch": 1.12, + "grad_norm": 0.17350769312286368, + "learning_rate": 0.0002990760150429957, + "loss": 0.181, + "step": 2380 + }, + { + "epoch": 1.120470588235294, + "grad_norm": 0.0004963788398155692, + "learning_rate": 0.0002990711505972736, + "loss": 0.0, + "step": 2381 + }, + { + "epoch": 1.1209411764705883, + "grad_norm": 0.10145998758682352, + "learning_rate": 0.00029906627342014407, + "loss": 0.0442, + "step": 2382 + }, + { + "epoch": 1.1214117647058823, + "grad_norm": 0.12174974450035722, + "learning_rate": 0.00029906138351202356, + "loss": 0.063, + "step": 2383 + }, + { + "epoch": 1.1218823529411766, + "grad_norm": 0.046279813895301704, + "learning_rate": 0.0002990564808733298, + "loss": 0.0144, + "step": 2384 + }, + { + "epoch": 1.1223529411764706, + "grad_norm": 0.03626653661916099, + "learning_rate": 0.00029905156550448145, + "loss": 0.0109, + "step": 2385 + }, + { + "epoch": 1.1228235294117648, + "grad_norm": 0.12004472904990475, + "learning_rate": 0.00029904663740589827, + "loss": 0.0836, + "step": 2386 + }, + { + "epoch": 1.1232941176470588, + "grad_norm": 0.10025726492827446, + "learning_rate": 0.0002990416965780012, + "loss": 0.0533, + "step": 2387 + }, + { + "epoch": 1.123764705882353, + "grad_norm": 0.08932910482111416, + "learning_rate": 0.0002990367430212123, + "loss": 0.0614, + "step": 2388 + }, + { + "epoch": 1.124235294117647, + "grad_norm": 0.14999142011600203, + "learning_rate": 0.0002990317767359544, + "loss": 0.0943, + "step": 2389 + }, + { + "epoch": 1.1247058823529412, + "grad_norm": 0.03953001385923055, + "learning_rate": 0.0002990267977226518, + "loss": 0.015, + "step": 2390 + }, + { + "epoch": 1.1251764705882352, + "grad_norm": 0.09465807346884579, + "learning_rate": 0.00029902180598172977, + "loss": 0.0299, + "step": 2391 + }, + { + "epoch": 1.1256470588235294, + "grad_norm": 0.18874948096583102, + "learning_rate": 0.00029901680151361447, + "loss": 0.1374, + "step": 2392 + }, + { + "epoch": 1.1261176470588234, + "grad_norm": 0.017665582735347227, + "learning_rate": 0.00029901178431873346, + "loss": 0.0027, + "step": 2393 + }, + { + "epoch": 1.1265882352941177, + "grad_norm": 0.0620287818843603, + "learning_rate": 0.0002990067543975151, + "loss": 0.0151, + "step": 2394 + }, + { + "epoch": 1.1270588235294117, + "grad_norm": 0.06383964646171192, + "learning_rate": 0.0002990017117503891, + "loss": 0.0336, + "step": 2395 + }, + { + "epoch": 1.127529411764706, + "grad_norm": 0.23677776196133535, + "learning_rate": 0.0002989966563777861, + "loss": 0.1408, + "step": 2396 + }, + { + "epoch": 1.1280000000000001, + "grad_norm": 1.4143059615699106e-05, + "learning_rate": 0.00029899158828013774, + "loss": 0.0, + "step": 2397 + }, + { + "epoch": 1.1284705882352941, + "grad_norm": 1.078018299175261, + "learning_rate": 0.000298986507457877, + "loss": 0.0355, + "step": 2398 + }, + { + "epoch": 1.1289411764705881, + "grad_norm": 0.13690614114384372, + "learning_rate": 0.0002989814139114377, + "loss": 0.0669, + "step": 2399 + }, + { + "epoch": 1.1294117647058823, + "grad_norm": 0.2168437871038141, + "learning_rate": 0.0002989763076412549, + "loss": 0.023, + "step": 2400 + }, + { + "epoch": 1.1298823529411766, + "grad_norm": 0.19971430338420137, + "learning_rate": 0.00029897118864776476, + "loss": 0.1731, + "step": 2401 + }, + { + "epoch": 1.1303529411764706, + "grad_norm": 0.09164613555580312, + "learning_rate": 0.00029896605693140437, + "loss": 0.0432, + "step": 2402 + }, + { + "epoch": 1.1308235294117648, + "grad_norm": 0.0824315385585514, + "learning_rate": 0.0002989609124926121, + "loss": 0.0472, + "step": 2403 + }, + { + "epoch": 1.1312941176470588, + "grad_norm": 0.01897724227884664, + "learning_rate": 0.0002989557553318272, + "loss": 0.0029, + "step": 2404 + }, + { + "epoch": 1.131764705882353, + "grad_norm": 0.13849269901850444, + "learning_rate": 0.0002989505854494902, + "loss": 0.0616, + "step": 2405 + }, + { + "epoch": 1.132235294117647, + "grad_norm": 0.10726390458071897, + "learning_rate": 0.00029894540284604255, + "loss": 0.0808, + "step": 2406 + }, + { + "epoch": 1.1327058823529412, + "grad_norm": 0.14713200175338473, + "learning_rate": 0.000298940207521927, + "loss": 0.1114, + "step": 2407 + }, + { + "epoch": 1.1331764705882352, + "grad_norm": 0.12126627496039508, + "learning_rate": 0.00029893499947758706, + "loss": 0.0997, + "step": 2408 + }, + { + "epoch": 1.1336470588235295, + "grad_norm": 0.1016486303423417, + "learning_rate": 0.00029892977871346776, + "loss": 0.0564, + "step": 2409 + }, + { + "epoch": 1.1341176470588235, + "grad_norm": 0.09655861004539712, + "learning_rate": 0.00029892454523001484, + "loss": 0.0429, + "step": 2410 + }, + { + "epoch": 1.1345882352941177, + "grad_norm": 0.06161792537272614, + "learning_rate": 0.00029891929902767526, + "loss": 0.0337, + "step": 2411 + }, + { + "epoch": 1.1350588235294117, + "grad_norm": 5.262967866612265e-06, + "learning_rate": 0.0002989140401068971, + "loss": 0.0, + "step": 2412 + }, + { + "epoch": 1.135529411764706, + "grad_norm": 0.12539691553267823, + "learning_rate": 0.0002989087684681295, + "loss": 0.089, + "step": 2413 + }, + { + "epoch": 1.1360000000000001, + "grad_norm": 0.05926369213392218, + "learning_rate": 0.00029890348411182267, + "loss": 0.0399, + "step": 2414 + }, + { + "epoch": 1.1364705882352941, + "grad_norm": 0.021673384062622834, + "learning_rate": 0.00029889818703842796, + "loss": 0.0034, + "step": 2415 + }, + { + "epoch": 1.1369411764705881, + "grad_norm": 0.17560019495397958, + "learning_rate": 0.0002988928772483977, + "loss": 0.076, + "step": 2416 + }, + { + "epoch": 1.1374117647058823, + "grad_norm": 0.10497499529410656, + "learning_rate": 0.0002988875547421854, + "loss": 0.0778, + "step": 2417 + }, + { + "epoch": 1.1378823529411766, + "grad_norm": 0.08663174903971262, + "learning_rate": 0.00029888221952024567, + "loss": 0.0159, + "step": 2418 + }, + { + "epoch": 1.1383529411764706, + "grad_norm": 0.09591645406328964, + "learning_rate": 0.00029887687158303416, + "loss": 0.0566, + "step": 2419 + }, + { + "epoch": 1.1388235294117648, + "grad_norm": 0.13010994425832412, + "learning_rate": 0.00029887151093100755, + "loss": 0.0627, + "step": 2420 + }, + { + "epoch": 1.1392941176470588, + "grad_norm": 0.02101008519288559, + "learning_rate": 0.00029886613756462365, + "loss": 0.0032, + "step": 2421 + }, + { + "epoch": 1.139764705882353, + "grad_norm": 0.0971937654471964, + "learning_rate": 0.00029886075148434146, + "loss": 0.1039, + "step": 2422 + }, + { + "epoch": 1.140235294117647, + "grad_norm": 7.567563827339887e-06, + "learning_rate": 0.0002988553526906209, + "loss": 0.0, + "step": 2423 + }, + { + "epoch": 1.1407058823529412, + "grad_norm": 0.05065129299825054, + "learning_rate": 0.0002988499411839232, + "loss": 0.0214, + "step": 2424 + }, + { + "epoch": 1.1411764705882352, + "grad_norm": 0.13231580217948952, + "learning_rate": 0.0002988445169647103, + "loss": 0.0965, + "step": 2425 + }, + { + "epoch": 1.1416470588235295, + "grad_norm": 0.027563900761687278, + "learning_rate": 0.00029883908003344554, + "loss": 0.0138, + "step": 2426 + }, + { + "epoch": 1.1421176470588235, + "grad_norm": 8.036054501467769e-06, + "learning_rate": 0.00029883363039059343, + "loss": 0.0, + "step": 2427 + }, + { + "epoch": 1.1425882352941177, + "grad_norm": 0.040785599489995, + "learning_rate": 0.0002988281680366192, + "loss": 0.0064, + "step": 2428 + }, + { + "epoch": 1.1430588235294117, + "grad_norm": 6.676687554550253e-06, + "learning_rate": 0.0002988226929719894, + "loss": 0.0, + "step": 2429 + }, + { + "epoch": 1.143529411764706, + "grad_norm": 9.213482723610599e-06, + "learning_rate": 0.0002988172051971717, + "loss": 0.0, + "step": 2430 + }, + { + "epoch": 1.144, + "grad_norm": 0.039246172370109765, + "learning_rate": 0.0002988117047126346, + "loss": 0.0135, + "step": 2431 + }, + { + "epoch": 1.1444705882352941, + "grad_norm": 8.64448255989689e-06, + "learning_rate": 0.00029880619151884813, + "loss": 0.0, + "step": 2432 + }, + { + "epoch": 1.1449411764705881, + "grad_norm": 1.554401118474235e-05, + "learning_rate": 0.000298800665616283, + "loss": 0.0, + "step": 2433 + }, + { + "epoch": 1.1454117647058824, + "grad_norm": 0.24031526815661136, + "learning_rate": 0.00029879512700541115, + "loss": 0.2282, + "step": 2434 + }, + { + "epoch": 1.1458823529411766, + "grad_norm": 0.05114319795182605, + "learning_rate": 0.0002987895756867056, + "loss": 0.0222, + "step": 2435 + }, + { + "epoch": 1.1463529411764706, + "grad_norm": 0.0189978753121614, + "learning_rate": 0.00029878401166064046, + "loss": 0.0057, + "step": 2436 + }, + { + "epoch": 1.1468235294117648, + "grad_norm": 0.19602699006690213, + "learning_rate": 0.000298778434927691, + "loss": 0.1046, + "step": 2437 + }, + { + "epoch": 1.1472941176470588, + "grad_norm": 0.028288694290242968, + "learning_rate": 0.0002987728454883334, + "loss": 0.0072, + "step": 2438 + }, + { + "epoch": 1.147764705882353, + "grad_norm": 0.09249665229973428, + "learning_rate": 0.00029876724334304506, + "loss": 0.1064, + "step": 2439 + }, + { + "epoch": 1.148235294117647, + "grad_norm": 0.1255978511943085, + "learning_rate": 0.00029876162849230446, + "loss": 0.1115, + "step": 2440 + }, + { + "epoch": 1.1487058823529412, + "grad_norm": 0.089529899178455, + "learning_rate": 0.00029875600093659115, + "loss": 0.046, + "step": 2441 + }, + { + "epoch": 1.1491764705882352, + "grad_norm": 4.784164033309992e-06, + "learning_rate": 0.00029875036067638565, + "loss": 0.0, + "step": 2442 + }, + { + "epoch": 1.1496470588235295, + "grad_norm": 0.1417840971286799, + "learning_rate": 0.00029874470771216983, + "loss": 0.0934, + "step": 2443 + }, + { + "epoch": 1.1501176470588235, + "grad_norm": 0.11486880540267759, + "learning_rate": 0.0002987390420444262, + "loss": 0.0208, + "step": 2444 + }, + { + "epoch": 1.1505882352941177, + "grad_norm": 0.0015808239572483887, + "learning_rate": 0.00029873336367363903, + "loss": 0.0, + "step": 2445 + }, + { + "epoch": 1.1510588235294117, + "grad_norm": 2.377390463873954e-06, + "learning_rate": 0.00029872767260029294, + "loss": 0.0, + "step": 2446 + }, + { + "epoch": 1.151529411764706, + "grad_norm": 0.05076273246994109, + "learning_rate": 0.00029872196882487417, + "loss": 0.0159, + "step": 2447 + }, + { + "epoch": 1.152, + "grad_norm": 0.048180626023792646, + "learning_rate": 0.00029871625234786974, + "loss": 0.0157, + "step": 2448 + }, + { + "epoch": 1.1524705882352941, + "grad_norm": 2.8492400523317652e-06, + "learning_rate": 0.0002987105231697679, + "loss": 0.0, + "step": 2449 + }, + { + "epoch": 1.1529411764705881, + "grad_norm": 0.08801005945669256, + "learning_rate": 0.000298704781291058, + "loss": 0.0303, + "step": 2450 + }, + { + "epoch": 1.1534117647058824, + "grad_norm": 0.08503060588734943, + "learning_rate": 0.0002986990267122304, + "loss": 0.0344, + "step": 2451 + }, + { + "epoch": 1.1538823529411766, + "grad_norm": 0.14289673143283718, + "learning_rate": 0.0002986932594337765, + "loss": 0.058, + "step": 2452 + }, + { + "epoch": 1.1543529411764706, + "grad_norm": 0.05400402232113157, + "learning_rate": 0.0002986874794561889, + "loss": 0.0278, + "step": 2453 + }, + { + "epoch": 1.1548235294117646, + "grad_norm": 6.3917711248355915e-06, + "learning_rate": 0.0002986816867799613, + "loss": 0.0, + "step": 2454 + }, + { + "epoch": 1.1552941176470588, + "grad_norm": 7.0457814631984545e-06, + "learning_rate": 0.0002986758814055883, + "loss": 0.0, + "step": 2455 + }, + { + "epoch": 1.155764705882353, + "grad_norm": 0.09735375145423172, + "learning_rate": 0.0002986700633335659, + "loss": 0.0515, + "step": 2456 + }, + { + "epoch": 1.156235294117647, + "grad_norm": 0.16283724276407932, + "learning_rate": 0.0002986642325643908, + "loss": 0.0808, + "step": 2457 + }, + { + "epoch": 1.1567058823529413, + "grad_norm": 0.24150566167796195, + "learning_rate": 0.00029865838909856103, + "loss": 0.0947, + "step": 2458 + }, + { + "epoch": 1.1571764705882353, + "grad_norm": 0.12492952143059079, + "learning_rate": 0.0002986525329365757, + "loss": 0.066, + "step": 2459 + }, + { + "epoch": 1.1576470588235295, + "grad_norm": 0.05086204749606407, + "learning_rate": 0.00029864666407893487, + "loss": 0.0194, + "step": 2460 + }, + { + "epoch": 1.1581176470588235, + "grad_norm": 0.0879607801633555, + "learning_rate": 0.00029864078252613983, + "loss": 0.056, + "step": 2461 + }, + { + "epoch": 1.1585882352941177, + "grad_norm": 0.12845870764875522, + "learning_rate": 0.0002986348882786929, + "loss": 0.0891, + "step": 2462 + }, + { + "epoch": 1.1590588235294117, + "grad_norm": 0.057638508018848914, + "learning_rate": 0.00029862898133709747, + "loss": 0.0349, + "step": 2463 + }, + { + "epoch": 1.159529411764706, + "grad_norm": 0.11582212039064116, + "learning_rate": 0.000298623061701858, + "loss": 0.0705, + "step": 2464 + }, + { + "epoch": 1.16, + "grad_norm": 0.12370679399363112, + "learning_rate": 0.00029861712937348007, + "loss": 0.0683, + "step": 2465 + }, + { + "epoch": 1.1604705882352941, + "grad_norm": 0.09461870370471208, + "learning_rate": 0.0002986111843524703, + "loss": 0.0548, + "step": 2466 + }, + { + "epoch": 1.1609411764705881, + "grad_norm": 0.12852749450198467, + "learning_rate": 0.00029860522663933655, + "loss": 0.056, + "step": 2467 + }, + { + "epoch": 1.1614117647058824, + "grad_norm": 0.09537574256248323, + "learning_rate": 0.00029859925623458745, + "loss": 0.0407, + "step": 2468 + }, + { + "epoch": 1.1618823529411766, + "grad_norm": 0.13373233162479625, + "learning_rate": 0.000298593273138733, + "loss": 0.0554, + "step": 2469 + }, + { + "epoch": 1.1623529411764706, + "grad_norm": 0.0010562401545239585, + "learning_rate": 0.00029858727735228416, + "loss": 0.0, + "step": 2470 + }, + { + "epoch": 1.1628235294117646, + "grad_norm": 0.0003687877158147573, + "learning_rate": 0.00029858126887575306, + "loss": 0.0, + "step": 2471 + }, + { + "epoch": 1.1632941176470588, + "grad_norm": 0.15797419201523336, + "learning_rate": 0.00029857524770965285, + "loss": 0.1212, + "step": 2472 + }, + { + "epoch": 1.163764705882353, + "grad_norm": 0.12415094946900639, + "learning_rate": 0.00029856921385449763, + "loss": 0.1016, + "step": 2473 + }, + { + "epoch": 1.164235294117647, + "grad_norm": 0.42038927463998543, + "learning_rate": 0.0002985631673108029, + "loss": 0.0693, + "step": 2474 + }, + { + "epoch": 1.1647058823529413, + "grad_norm": 0.00042720283041908756, + "learning_rate": 0.00029855710807908493, + "loss": 0.0, + "step": 2475 + }, + { + "epoch": 1.1651764705882353, + "grad_norm": 0.09431730341652149, + "learning_rate": 0.00029855103615986133, + "loss": 0.0497, + "step": 2476 + }, + { + "epoch": 1.1656470588235295, + "grad_norm": 0.10646559417391643, + "learning_rate": 0.0002985449515536505, + "loss": 0.0525, + "step": 2477 + }, + { + "epoch": 1.1661176470588235, + "grad_norm": 0.08867965708679766, + "learning_rate": 0.0002985388542609723, + "loss": 0.0491, + "step": 2478 + }, + { + "epoch": 1.1665882352941177, + "grad_norm": 0.08189849050550704, + "learning_rate": 0.0002985327442823473, + "loss": 0.0458, + "step": 2479 + }, + { + "epoch": 1.1670588235294117, + "grad_norm": 0.12997351634613502, + "learning_rate": 0.00029852662161829744, + "loss": 0.0869, + "step": 2480 + }, + { + "epoch": 1.167529411764706, + "grad_norm": 0.17206698750788715, + "learning_rate": 0.0002985204862693455, + "loss": 0.1343, + "step": 2481 + }, + { + "epoch": 1.168, + "grad_norm": 0.00012941600631589579, + "learning_rate": 0.00029851433823601563, + "loss": 0.0, + "step": 2482 + }, + { + "epoch": 1.1684705882352941, + "grad_norm": 0.00015201042653145247, + "learning_rate": 0.00029850817751883276, + "loss": 0.0, + "step": 2483 + }, + { + "epoch": 1.1689411764705882, + "grad_norm": 0.012257975996925123, + "learning_rate": 0.0002985020041183231, + "loss": 0.0019, + "step": 2484 + }, + { + "epoch": 1.1694117647058824, + "grad_norm": 0.06600001266763798, + "learning_rate": 0.00029849581803501393, + "loss": 0.0468, + "step": 2485 + }, + { + "epoch": 1.1698823529411766, + "grad_norm": 0.03903980799628922, + "learning_rate": 0.0002984896192694335, + "loss": 0.0124, + "step": 2486 + }, + { + "epoch": 1.1703529411764706, + "grad_norm": 0.0802783488570201, + "learning_rate": 0.00029848340782211125, + "loss": 0.0572, + "step": 2487 + }, + { + "epoch": 1.1708235294117646, + "grad_norm": 0.0896157708490795, + "learning_rate": 0.0002984771836935777, + "loss": 0.0742, + "step": 2488 + }, + { + "epoch": 1.1712941176470588, + "grad_norm": 0.052864331916026304, + "learning_rate": 0.0002984709468843644, + "loss": 0.0267, + "step": 2489 + }, + { + "epoch": 1.171764705882353, + "grad_norm": 0.0003065773939868567, + "learning_rate": 0.0002984646973950039, + "loss": 0.0, + "step": 2490 + }, + { + "epoch": 1.172235294117647, + "grad_norm": 0.10401685148465757, + "learning_rate": 0.00029845843522603006, + "loss": 0.0437, + "step": 2491 + }, + { + "epoch": 1.1727058823529413, + "grad_norm": 0.0579127672055136, + "learning_rate": 0.0002984521603779777, + "loss": 0.0278, + "step": 2492 + }, + { + "epoch": 1.1731764705882353, + "grad_norm": 0.2545430644692609, + "learning_rate": 0.0002984458728513827, + "loss": 0.3334, + "step": 2493 + }, + { + "epoch": 1.1736470588235295, + "grad_norm": 0.1552319307645211, + "learning_rate": 0.000298439572646782, + "loss": 0.1379, + "step": 2494 + }, + { + "epoch": 1.1741176470588235, + "grad_norm": 0.05946778713790466, + "learning_rate": 0.0002984332597647137, + "loss": 0.0369, + "step": 2495 + }, + { + "epoch": 1.1745882352941177, + "grad_norm": 0.05833332274742833, + "learning_rate": 0.000298426934205717, + "loss": 0.024, + "step": 2496 + }, + { + "epoch": 1.1750588235294117, + "grad_norm": 0.049908126917531485, + "learning_rate": 0.00029842059597033204, + "loss": 0.016, + "step": 2497 + }, + { + "epoch": 1.175529411764706, + "grad_norm": 0.030012903249097507, + "learning_rate": 0.0002984142450591002, + "loss": 0.0103, + "step": 2498 + }, + { + "epoch": 1.176, + "grad_norm": 0.050128692126929594, + "learning_rate": 0.0002984078814725639, + "loss": 0.0149, + "step": 2499 + }, + { + "epoch": 1.1764705882352942, + "grad_norm": 0.0730202158647451, + "learning_rate": 0.0002984015052112665, + "loss": 0.0312, + "step": 2500 + }, + { + "epoch": 1.1769411764705882, + "grad_norm": 0.08855130120870505, + "learning_rate": 0.0002983951162757527, + "loss": 0.0419, + "step": 2501 + }, + { + "epoch": 1.1774117647058824, + "grad_norm": 0.0832091649259298, + "learning_rate": 0.0002983887146665681, + "loss": 0.052, + "step": 2502 + }, + { + "epoch": 1.1778823529411764, + "grad_norm": 0.09586451081925708, + "learning_rate": 0.00029838230038425944, + "loss": 0.0476, + "step": 2503 + }, + { + "epoch": 1.1783529411764706, + "grad_norm": 0.09918651677301926, + "learning_rate": 0.0002983758734293745, + "loss": 0.0537, + "step": 2504 + }, + { + "epoch": 1.1788235294117646, + "grad_norm": 0.12342277787065783, + "learning_rate": 0.00029836943380246215, + "loss": 0.0797, + "step": 2505 + }, + { + "epoch": 1.1792941176470588, + "grad_norm": 0.16180329255975862, + "learning_rate": 0.00029836298150407245, + "loss": 0.1242, + "step": 2506 + }, + { + "epoch": 1.179764705882353, + "grad_norm": 0.10217704479853484, + "learning_rate": 0.00029835651653475643, + "loss": 0.0652, + "step": 2507 + }, + { + "epoch": 1.180235294117647, + "grad_norm": 0.016009415546245277, + "learning_rate": 0.00029835003889506616, + "loss": 0.0026, + "step": 2508 + }, + { + "epoch": 1.1807058823529413, + "grad_norm": 2.6548439200851134e-05, + "learning_rate": 0.000298343548585555, + "loss": 0.0, + "step": 2509 + }, + { + "epoch": 1.1811764705882353, + "grad_norm": 0.11058218712949955, + "learning_rate": 0.0002983370456067771, + "loss": 0.0495, + "step": 2510 + }, + { + "epoch": 1.1816470588235295, + "grad_norm": 0.09250804171305893, + "learning_rate": 0.0002983305299592879, + "loss": 0.0452, + "step": 2511 + }, + { + "epoch": 1.1821176470588235, + "grad_norm": 0.13101700294192214, + "learning_rate": 0.0002983240016436439, + "loss": 0.1076, + "step": 2512 + }, + { + "epoch": 1.1825882352941177, + "grad_norm": 0.09224535692342509, + "learning_rate": 0.0002983174606604027, + "loss": 0.0365, + "step": 2513 + }, + { + "epoch": 1.1830588235294117, + "grad_norm": 0.22996071355055991, + "learning_rate": 0.0002983109070101228, + "loss": 0.1337, + "step": 2514 + }, + { + "epoch": 1.183529411764706, + "grad_norm": 0.030132325541398226, + "learning_rate": 0.00029830434069336403, + "loss": 0.0091, + "step": 2515 + }, + { + "epoch": 1.184, + "grad_norm": 0.0736974571067485, + "learning_rate": 0.0002982977617106871, + "loss": 0.0416, + "step": 2516 + }, + { + "epoch": 1.1844705882352942, + "grad_norm": 0.07128434775462505, + "learning_rate": 0.0002982911700626539, + "loss": 0.0433, + "step": 2517 + }, + { + "epoch": 1.1849411764705882, + "grad_norm": 2.462760530571339e-05, + "learning_rate": 0.0002982845657498274, + "loss": 0.0, + "step": 2518 + }, + { + "epoch": 1.1854117647058824, + "grad_norm": 0.10087169099231307, + "learning_rate": 0.0002982779487727718, + "loss": 0.0636, + "step": 2519 + }, + { + "epoch": 1.1858823529411764, + "grad_norm": 0.04084900300773342, + "learning_rate": 0.0002982713191320519, + "loss": 0.014, + "step": 2520 + }, + { + "epoch": 1.1863529411764706, + "grad_norm": 0.15444174255647472, + "learning_rate": 0.00029826467682823415, + "loss": 0.0946, + "step": 2521 + }, + { + "epoch": 1.1868235294117646, + "grad_norm": 0.00019120428451690688, + "learning_rate": 0.0002982580218618858, + "loss": 0.0, + "step": 2522 + }, + { + "epoch": 1.1872941176470588, + "grad_norm": 0.07404535345002954, + "learning_rate": 0.0002982513542335751, + "loss": 0.0541, + "step": 2523 + }, + { + "epoch": 1.187764705882353, + "grad_norm": 2.059589713496645e-05, + "learning_rate": 0.00029824467394387166, + "loss": 0.0, + "step": 2524 + }, + { + "epoch": 1.188235294117647, + "grad_norm": 3.799462656440028e-05, + "learning_rate": 0.0002982379809933459, + "loss": 0.0, + "step": 2525 + }, + { + "epoch": 1.188705882352941, + "grad_norm": 0.0747316100626205, + "learning_rate": 0.00029823127538256943, + "loss": 0.0302, + "step": 2526 + }, + { + "epoch": 1.1891764705882353, + "grad_norm": 0.0854000079881979, + "learning_rate": 0.000298224557112115, + "loss": 0.0492, + "step": 2527 + }, + { + "epoch": 1.1896470588235295, + "grad_norm": 0.0750853342399964, + "learning_rate": 0.0002982178261825563, + "loss": 0.0474, + "step": 2528 + }, + { + "epoch": 1.1901176470588235, + "grad_norm": 0.1264941925729342, + "learning_rate": 0.0002982110825944683, + "loss": 0.0949, + "step": 2529 + }, + { + "epoch": 1.1905882352941177, + "grad_norm": 0.059102766821503186, + "learning_rate": 0.0002982043263484269, + "loss": 0.0369, + "step": 2530 + }, + { + "epoch": 1.1910588235294117, + "grad_norm": 0.06216423543505771, + "learning_rate": 0.00029819755744500904, + "loss": 0.0272, + "step": 2531 + }, + { + "epoch": 1.191529411764706, + "grad_norm": 0.23689408687977995, + "learning_rate": 0.0002981907758847929, + "loss": 0.1008, + "step": 2532 + }, + { + "epoch": 1.192, + "grad_norm": 0.06159968476060956, + "learning_rate": 0.0002981839816683576, + "loss": 0.0249, + "step": 2533 + }, + { + "epoch": 1.1924705882352942, + "grad_norm": 5.350210901395624e-05, + "learning_rate": 0.00029817717479628346, + "loss": 0.0, + "step": 2534 + }, + { + "epoch": 1.1929411764705882, + "grad_norm": 0.07458991864756909, + "learning_rate": 0.00029817035526915177, + "loss": 0.0203, + "step": 2535 + }, + { + "epoch": 1.1934117647058824, + "grad_norm": 0.11135008873015288, + "learning_rate": 0.00029816352308754496, + "loss": 0.0723, + "step": 2536 + }, + { + "epoch": 1.1938823529411764, + "grad_norm": 1.8418986693267366e-05, + "learning_rate": 0.00029815667825204654, + "loss": 0.0, + "step": 2537 + }, + { + "epoch": 1.1943529411764706, + "grad_norm": 0.05703555010811047, + "learning_rate": 0.0002981498207632411, + "loss": 0.024, + "step": 2538 + }, + { + "epoch": 1.1948235294117646, + "grad_norm": 0.11016044847241857, + "learning_rate": 0.0002981429506217143, + "loss": 0.051, + "step": 2539 + }, + { + "epoch": 1.1952941176470588, + "grad_norm": 0.05069397088792932, + "learning_rate": 0.0002981360678280529, + "loss": 0.0182, + "step": 2540 + }, + { + "epoch": 1.195764705882353, + "grad_norm": 0.12117498010134957, + "learning_rate": 0.00029812917238284466, + "loss": 0.0655, + "step": 2541 + }, + { + "epoch": 1.196235294117647, + "grad_norm": 0.08499059314682977, + "learning_rate": 0.00029812226428667854, + "loss": 0.0347, + "step": 2542 + }, + { + "epoch": 1.196705882352941, + "grad_norm": 0.20315595853623306, + "learning_rate": 0.0002981153435401445, + "loss": 0.1659, + "step": 2543 + }, + { + "epoch": 1.1971764705882353, + "grad_norm": 0.03443744620873586, + "learning_rate": 0.00029810841014383366, + "loss": 0.014, + "step": 2544 + }, + { + "epoch": 1.1976470588235295, + "grad_norm": 0.07276002225246522, + "learning_rate": 0.00029810146409833815, + "loss": 0.042, + "step": 2545 + }, + { + "epoch": 1.1981176470588235, + "grad_norm": 0.00010335444559661477, + "learning_rate": 0.0002980945054042511, + "loss": 0.0, + "step": 2546 + }, + { + "epoch": 1.1985882352941177, + "grad_norm": 0.08626549605033973, + "learning_rate": 0.00029808753406216693, + "loss": 0.0239, + "step": 2547 + }, + { + "epoch": 1.1990588235294117, + "grad_norm": 9.519841700342073e-05, + "learning_rate": 0.000298080550072681, + "loss": 0.0, + "step": 2548 + }, + { + "epoch": 1.199529411764706, + "grad_norm": 0.08309759867031788, + "learning_rate": 0.0002980735534363897, + "loss": 0.0376, + "step": 2549 + }, + { + "epoch": 1.2, + "grad_norm": 0.17901628297943484, + "learning_rate": 0.0002980665441538907, + "loss": 0.0797, + "step": 2550 + }, + { + "epoch": 1.2004705882352942, + "grad_norm": 0.022928350320813463, + "learning_rate": 0.00029805952222578257, + "loss": 0.0044, + "step": 2551 + }, + { + "epoch": 1.2009411764705882, + "grad_norm": 0.09143508428406093, + "learning_rate": 0.000298052487652665, + "loss": 0.0662, + "step": 2552 + }, + { + "epoch": 1.2014117647058824, + "grad_norm": 0.022492760489877632, + "learning_rate": 0.0002980454404351388, + "loss": 0.0037, + "step": 2553 + }, + { + "epoch": 1.2018823529411764, + "grad_norm": 0.1529776600846864, + "learning_rate": 0.0002980383805738058, + "loss": 0.1151, + "step": 2554 + }, + { + "epoch": 1.2023529411764706, + "grad_norm": 0.07087176082338628, + "learning_rate": 0.00029803130806926897, + "loss": 0.0358, + "step": 2555 + }, + { + "epoch": 1.2028235294117646, + "grad_norm": 0.08474222002275664, + "learning_rate": 0.0002980242229221324, + "loss": 0.0415, + "step": 2556 + }, + { + "epoch": 1.2032941176470588, + "grad_norm": 0.05389919418859092, + "learning_rate": 0.00029801712513300105, + "loss": 0.0195, + "step": 2557 + }, + { + "epoch": 1.2037647058823528, + "grad_norm": 0.10634558412666428, + "learning_rate": 0.00029801001470248123, + "loss": 0.0647, + "step": 2558 + }, + { + "epoch": 1.204235294117647, + "grad_norm": 0.04211219024152901, + "learning_rate": 0.00029800289163118014, + "loss": 0.0145, + "step": 2559 + }, + { + "epoch": 1.204705882352941, + "grad_norm": 0.10492139862072106, + "learning_rate": 0.00029799575591970623, + "loss": 0.0505, + "step": 2560 + }, + { + "epoch": 1.2051764705882353, + "grad_norm": 0.07582156030129997, + "learning_rate": 0.00029798860756866883, + "loss": 0.039, + "step": 2561 + }, + { + "epoch": 1.2056470588235295, + "grad_norm": 0.06323794263435357, + "learning_rate": 0.00029798144657867845, + "loss": 0.0336, + "step": 2562 + }, + { + "epoch": 1.2061176470588235, + "grad_norm": 0.2419408652988807, + "learning_rate": 0.0002979742729503467, + "loss": 0.0889, + "step": 2563 + }, + { + "epoch": 1.2065882352941177, + "grad_norm": 0.08041231230880971, + "learning_rate": 0.0002979670866842862, + "loss": 0.0222, + "step": 2564 + }, + { + "epoch": 1.2070588235294117, + "grad_norm": 0.07657558941314324, + "learning_rate": 0.0002979598877811107, + "loss": 0.0511, + "step": 2565 + }, + { + "epoch": 1.207529411764706, + "grad_norm": 0.06970887055143174, + "learning_rate": 0.0002979526762414351, + "loss": 0.0434, + "step": 2566 + }, + { + "epoch": 1.208, + "grad_norm": 0.056803062213636395, + "learning_rate": 0.0002979454520658752, + "loss": 0.0226, + "step": 2567 + }, + { + "epoch": 1.2084705882352942, + "grad_norm": 0.05924168090895405, + "learning_rate": 0.0002979382152550481, + "loss": 0.0155, + "step": 2568 + }, + { + "epoch": 1.2089411764705882, + "grad_norm": 1.5767533016442632e-05, + "learning_rate": 0.0002979309658095717, + "loss": 0.0, + "step": 2569 + }, + { + "epoch": 1.2094117647058824, + "grad_norm": 0.11966749127099917, + "learning_rate": 0.0002979237037300653, + "loss": 0.068, + "step": 2570 + }, + { + "epoch": 1.2098823529411764, + "grad_norm": 0.24595509273089566, + "learning_rate": 0.00029791642901714897, + "loss": 0.0819, + "step": 2571 + }, + { + "epoch": 1.2103529411764706, + "grad_norm": 0.16406997472429002, + "learning_rate": 0.0002979091416714441, + "loss": 0.0829, + "step": 2572 + }, + { + "epoch": 1.2108235294117646, + "grad_norm": 0.26143293685396995, + "learning_rate": 0.00029790184169357307, + "loss": 0.1546, + "step": 2573 + }, + { + "epoch": 1.2112941176470589, + "grad_norm": 0.07934002728045647, + "learning_rate": 0.00029789452908415923, + "loss": 0.0293, + "step": 2574 + }, + { + "epoch": 1.2117647058823529, + "grad_norm": 0.02583222572050141, + "learning_rate": 0.0002978872038438273, + "loss": 0.004, + "step": 2575 + }, + { + "epoch": 1.212235294117647, + "grad_norm": 0.08290016512976693, + "learning_rate": 0.0002978798659732027, + "loss": 0.0327, + "step": 2576 + }, + { + "epoch": 1.212705882352941, + "grad_norm": 0.5043211143622518, + "learning_rate": 0.00029787251547291215, + "loss": 0.0489, + "step": 2577 + }, + { + "epoch": 1.2131764705882353, + "grad_norm": 0.1127959411701829, + "learning_rate": 0.00029786515234358357, + "loss": 0.0981, + "step": 2578 + }, + { + "epoch": 1.2136470588235295, + "grad_norm": 0.035929256779948356, + "learning_rate": 0.0002978577765858457, + "loss": 0.0102, + "step": 2579 + }, + { + "epoch": 1.2141176470588235, + "grad_norm": 7.87116941426348e-06, + "learning_rate": 0.00029785038820032844, + "loss": 0.0, + "step": 2580 + }, + { + "epoch": 1.2145882352941175, + "grad_norm": 0.1787823760878887, + "learning_rate": 0.0002978429871876628, + "loss": 0.1184, + "step": 2581 + }, + { + "epoch": 1.2150588235294117, + "grad_norm": 0.09718289760386742, + "learning_rate": 0.00029783557354848094, + "loss": 0.0587, + "step": 2582 + }, + { + "epoch": 1.215529411764706, + "grad_norm": 8.046777211455568e-06, + "learning_rate": 0.0002978281472834159, + "loss": 0.0, + "step": 2583 + }, + { + "epoch": 1.216, + "grad_norm": 0.159804606494231, + "learning_rate": 0.00029782070839310206, + "loss": 0.1125, + "step": 2584 + }, + { + "epoch": 1.2164705882352942, + "grad_norm": 1.0245477642560663e-05, + "learning_rate": 0.0002978132568781746, + "loss": 0.0, + "step": 2585 + }, + { + "epoch": 1.2169411764705882, + "grad_norm": 0.12588195168993388, + "learning_rate": 0.00029780579273927006, + "loss": 0.0713, + "step": 2586 + }, + { + "epoch": 1.2174117647058824, + "grad_norm": 0.10568022921167756, + "learning_rate": 0.0002977983159770258, + "loss": 0.0524, + "step": 2587 + }, + { + "epoch": 1.2178823529411764, + "grad_norm": 0.10293594090791466, + "learning_rate": 0.0002977908265920804, + "loss": 0.0725, + "step": 2588 + }, + { + "epoch": 1.2183529411764706, + "grad_norm": 0.12385775805620235, + "learning_rate": 0.00029778332458507353, + "loss": 0.0347, + "step": 2589 + }, + { + "epoch": 1.2188235294117646, + "grad_norm": 0.07623751268311987, + "learning_rate": 0.0002977758099566458, + "loss": 0.0531, + "step": 2590 + }, + { + "epoch": 1.2192941176470589, + "grad_norm": 1.223983008438758e-05, + "learning_rate": 0.0002977682827074391, + "loss": 0.0, + "step": 2591 + }, + { + "epoch": 1.2197647058823529, + "grad_norm": 0.047028982353425335, + "learning_rate": 0.0002977607428380963, + "loss": 0.0187, + "step": 2592 + }, + { + "epoch": 1.220235294117647, + "grad_norm": 0.032787706911137346, + "learning_rate": 0.00029775319034926125, + "loss": 0.0082, + "step": 2593 + }, + { + "epoch": 1.220705882352941, + "grad_norm": 1.6751089108957412e-05, + "learning_rate": 0.000297745625241579, + "loss": 0.0, + "step": 2594 + }, + { + "epoch": 1.2211764705882353, + "grad_norm": 0.1601840296836375, + "learning_rate": 0.0002977380475156958, + "loss": 0.1587, + "step": 2595 + }, + { + "epoch": 1.2216470588235295, + "grad_norm": 0.10114023168402621, + "learning_rate": 0.00029773045717225856, + "loss": 0.0875, + "step": 2596 + }, + { + "epoch": 1.2221176470588235, + "grad_norm": 0.07418587567933947, + "learning_rate": 0.0002977228542119157, + "loss": 0.0308, + "step": 2597 + }, + { + "epoch": 1.2225882352941175, + "grad_norm": 0.036238075101853645, + "learning_rate": 0.0002977152386353165, + "loss": 0.0136, + "step": 2598 + }, + { + "epoch": 1.2230588235294118, + "grad_norm": 0.04051896356441298, + "learning_rate": 0.0002977076104431114, + "loss": 0.0266, + "step": 2599 + }, + { + "epoch": 1.223529411764706, + "grad_norm": 0.11193945592583518, + "learning_rate": 0.00029769996963595184, + "loss": 0.0841, + "step": 2600 + }, + { + "epoch": 1.224, + "grad_norm": 0.06847462349824332, + "learning_rate": 0.0002976923162144904, + "loss": 0.0351, + "step": 2601 + }, + { + "epoch": 1.2244705882352942, + "grad_norm": 0.12107332259500131, + "learning_rate": 0.0002976846501793808, + "loss": 0.0893, + "step": 2602 + }, + { + "epoch": 1.2249411764705882, + "grad_norm": 5.257866267714421e-05, + "learning_rate": 0.00029767697153127766, + "loss": 0.0, + "step": 2603 + }, + { + "epoch": 1.2254117647058824, + "grad_norm": 1.527161833199992e-05, + "learning_rate": 0.0002976692802708368, + "loss": 0.0, + "step": 2604 + }, + { + "epoch": 1.2258823529411764, + "grad_norm": 0.05091889059917715, + "learning_rate": 0.000297661576398715, + "loss": 0.0257, + "step": 2605 + }, + { + "epoch": 1.2263529411764706, + "grad_norm": 0.07282319611232758, + "learning_rate": 0.0002976538599155704, + "loss": 0.0347, + "step": 2606 + }, + { + "epoch": 1.2268235294117646, + "grad_norm": 2.0872372487964175e-05, + "learning_rate": 0.00029764613082206183, + "loss": 0.0, + "step": 2607 + }, + { + "epoch": 1.2272941176470589, + "grad_norm": 2.2324204731014692e-05, + "learning_rate": 0.0002976383891188496, + "loss": 0.0, + "step": 2608 + }, + { + "epoch": 1.2277647058823529, + "grad_norm": 0.11587136417380123, + "learning_rate": 0.0002976306348065947, + "loss": 0.1002, + "step": 2609 + }, + { + "epoch": 1.228235294117647, + "grad_norm": 0.0581025294275699, + "learning_rate": 0.00029762286788595954, + "loss": 0.0299, + "step": 2610 + }, + { + "epoch": 1.228705882352941, + "grad_norm": 1.5453347517579064e-05, + "learning_rate": 0.0002976150883576073, + "loss": 0.0, + "step": 2611 + }, + { + "epoch": 1.2291764705882353, + "grad_norm": 1.5979346477504514e-05, + "learning_rate": 0.0002976072962222025, + "loss": 0.0, + "step": 2612 + }, + { + "epoch": 1.2296470588235293, + "grad_norm": 0.13044904560411624, + "learning_rate": 0.0002975994914804106, + "loss": 0.1171, + "step": 2613 + }, + { + "epoch": 1.2301176470588235, + "grad_norm": 0.04276799681971224, + "learning_rate": 0.00029759167413289815, + "loss": 0.0125, + "step": 2614 + }, + { + "epoch": 1.2305882352941175, + "grad_norm": 1.6785655024484285e-05, + "learning_rate": 0.00029758384418033276, + "loss": 0.0, + "step": 2615 + }, + { + "epoch": 1.2310588235294118, + "grad_norm": 0.03499167721376724, + "learning_rate": 0.00029757600162338326, + "loss": 0.0099, + "step": 2616 + }, + { + "epoch": 1.231529411764706, + "grad_norm": 0.20481670337365454, + "learning_rate": 0.0002975681464627193, + "loss": 0.1511, + "step": 2617 + }, + { + "epoch": 1.232, + "grad_norm": 0.08358161122462911, + "learning_rate": 0.00029756027869901183, + "loss": 0.0393, + "step": 2618 + }, + { + "epoch": 1.2324705882352942, + "grad_norm": 0.055575386567324976, + "learning_rate": 0.0002975523983329328, + "loss": 0.0153, + "step": 2619 + }, + { + "epoch": 1.2329411764705882, + "grad_norm": 0.1323008639651822, + "learning_rate": 0.00029754450536515517, + "loss": 0.0718, + "step": 2620 + }, + { + "epoch": 1.2334117647058824, + "grad_norm": 0.05030502780615344, + "learning_rate": 0.00029753659979635315, + "loss": 0.0267, + "step": 2621 + }, + { + "epoch": 1.2338823529411764, + "grad_norm": 0.1535292331950826, + "learning_rate": 0.00029752868162720176, + "loss": 0.0961, + "step": 2622 + }, + { + "epoch": 1.2343529411764707, + "grad_norm": 1.932597696284362e-05, + "learning_rate": 0.0002975207508583774, + "loss": 0.0, + "step": 2623 + }, + { + "epoch": 1.2348235294117647, + "grad_norm": 0.15126184824202824, + "learning_rate": 0.00029751280749055727, + "loss": 0.1127, + "step": 2624 + }, + { + "epoch": 1.2352941176470589, + "grad_norm": 0.08420021339988057, + "learning_rate": 0.0002975048515244199, + "loss": 0.0409, + "step": 2625 + }, + { + "epoch": 1.2357647058823529, + "grad_norm": 0.051046220147299974, + "learning_rate": 0.0002974968829606447, + "loss": 0.0218, + "step": 2626 + }, + { + "epoch": 1.236235294117647, + "grad_norm": 0.18147984751001683, + "learning_rate": 0.0002974889017999122, + "loss": 0.0913, + "step": 2627 + }, + { + "epoch": 1.236705882352941, + "grad_norm": 0.06760036024230523, + "learning_rate": 0.00029748090804290403, + "loss": 0.0376, + "step": 2628 + }, + { + "epoch": 1.2371764705882353, + "grad_norm": 0.05330113113524196, + "learning_rate": 0.00029747290169030295, + "loss": 0.0255, + "step": 2629 + }, + { + "epoch": 1.2376470588235293, + "grad_norm": 0.028066380401747495, + "learning_rate": 0.0002974648827427927, + "loss": 0.0074, + "step": 2630 + }, + { + "epoch": 1.2381176470588235, + "grad_norm": 0.06910422664948522, + "learning_rate": 0.0002974568512010582, + "loss": 0.0369, + "step": 2631 + }, + { + "epoch": 1.2385882352941175, + "grad_norm": 0.06373956362219016, + "learning_rate": 0.00029744880706578526, + "loss": 0.0238, + "step": 2632 + }, + { + "epoch": 1.2390588235294118, + "grad_norm": 0.06628156189013783, + "learning_rate": 0.00029744075033766103, + "loss": 0.0254, + "step": 2633 + }, + { + "epoch": 1.239529411764706, + "grad_norm": 0.06993844724157705, + "learning_rate": 0.0002974326810173735, + "loss": 0.0381, + "step": 2634 + }, + { + "epoch": 1.24, + "grad_norm": 0.04781983171878787, + "learning_rate": 0.00029742459910561187, + "loss": 0.0165, + "step": 2635 + }, + { + "epoch": 1.2404705882352942, + "grad_norm": 0.06326975631338148, + "learning_rate": 0.00029741650460306636, + "loss": 0.0222, + "step": 2636 + }, + { + "epoch": 1.2409411764705882, + "grad_norm": 0.09200025381674411, + "learning_rate": 0.0002974083975104283, + "loss": 0.0314, + "step": 2637 + }, + { + "epoch": 1.2414117647058824, + "grad_norm": 0.05795454773562192, + "learning_rate": 0.00029740027782839003, + "loss": 0.021, + "step": 2638 + }, + { + "epoch": 1.2418823529411764, + "grad_norm": 0.07937282323849623, + "learning_rate": 0.00029739214555764504, + "loss": 0.0296, + "step": 2639 + }, + { + "epoch": 1.2423529411764707, + "grad_norm": 0.03517751334323287, + "learning_rate": 0.00029738400069888787, + "loss": 0.0115, + "step": 2640 + }, + { + "epoch": 1.2428235294117647, + "grad_norm": 0.05848615557646132, + "learning_rate": 0.00029737584325281415, + "loss": 0.0479, + "step": 2641 + }, + { + "epoch": 1.2432941176470589, + "grad_norm": 0.05186319384964818, + "learning_rate": 0.0002973676732201205, + "loss": 0.0244, + "step": 2642 + }, + { + "epoch": 1.2437647058823529, + "grad_norm": 0.03607098323411429, + "learning_rate": 0.00029735949060150476, + "loss": 0.0079, + "step": 2643 + }, + { + "epoch": 1.244235294117647, + "grad_norm": 0.07046484897990904, + "learning_rate": 0.0002973512953976657, + "loss": 0.0374, + "step": 2644 + }, + { + "epoch": 1.244705882352941, + "grad_norm": 0.11933940445188514, + "learning_rate": 0.0002973430876093033, + "loss": 0.0729, + "step": 2645 + }, + { + "epoch": 1.2451764705882353, + "grad_norm": 0.06311344295739155, + "learning_rate": 0.00029733486723711845, + "loss": 0.0278, + "step": 2646 + }, + { + "epoch": 1.2456470588235293, + "grad_norm": 0.076384459942782, + "learning_rate": 0.0002973266342818133, + "loss": 0.0531, + "step": 2647 + }, + { + "epoch": 1.2461176470588236, + "grad_norm": 0.07148436867939248, + "learning_rate": 0.00029731838874409093, + "loss": 0.0391, + "step": 2648 + }, + { + "epoch": 1.2465882352941176, + "grad_norm": 0.1428639555380925, + "learning_rate": 0.0002973101306246556, + "loss": 0.1196, + "step": 2649 + }, + { + "epoch": 1.2470588235294118, + "grad_norm": 0.06476576846318477, + "learning_rate": 0.0002973018599242125, + "loss": 0.0265, + "step": 2650 + }, + { + "epoch": 1.247529411764706, + "grad_norm": 0.11220127735500302, + "learning_rate": 0.0002972935766434681, + "loss": 0.068, + "step": 2651 + }, + { + "epoch": 1.248, + "grad_norm": 0.11014501332031741, + "learning_rate": 0.00029728528078312976, + "loss": 0.074, + "step": 2652 + }, + { + "epoch": 1.248470588235294, + "grad_norm": 0.05201010673206281, + "learning_rate": 0.00029727697234390604, + "loss": 0.0252, + "step": 2653 + }, + { + "epoch": 1.2489411764705882, + "grad_norm": 0.11228801678072, + "learning_rate": 0.0002972686513265065, + "loss": 0.0699, + "step": 2654 + }, + { + "epoch": 1.2494117647058824, + "grad_norm": 0.11700830484690769, + "learning_rate": 0.00029726031773164176, + "loss": 0.0808, + "step": 2655 + }, + { + "epoch": 1.2498823529411764, + "grad_norm": 0.09906647355183432, + "learning_rate": 0.00029725197156002357, + "loss": 0.0532, + "step": 2656 + }, + { + "epoch": 1.2503529411764707, + "grad_norm": 0.08347464353069471, + "learning_rate": 0.00029724361281236475, + "loss": 0.0623, + "step": 2657 + }, + { + "epoch": 1.2508235294117647, + "grad_norm": 0.10632142646852955, + "learning_rate": 0.0002972352414893792, + "loss": 0.0925, + "step": 2658 + }, + { + "epoch": 1.251294117647059, + "grad_norm": 0.05229924365495816, + "learning_rate": 0.00029722685759178187, + "loss": 0.022, + "step": 2659 + }, + { + "epoch": 1.251764705882353, + "grad_norm": 0.1181588544090161, + "learning_rate": 0.0002972184611202886, + "loss": 0.1378, + "step": 2660 + }, + { + "epoch": 1.2522352941176471, + "grad_norm": 0.14296053163621683, + "learning_rate": 0.00029721005207561683, + "loss": 0.0959, + "step": 2661 + }, + { + "epoch": 1.2527058823529411, + "grad_norm": 8.817154871629713e-06, + "learning_rate": 0.00029720163045848444, + "loss": 0.0, + "step": 2662 + }, + { + "epoch": 1.2531764705882353, + "grad_norm": 0.12743881851569808, + "learning_rate": 0.0002971931962696109, + "loss": 0.0609, + "step": 2663 + }, + { + "epoch": 1.2536470588235293, + "grad_norm": 0.05788581661443565, + "learning_rate": 0.00029718474950971626, + "loss": 0.0186, + "step": 2664 + }, + { + "epoch": 1.2541176470588236, + "grad_norm": 0.033801653685517144, + "learning_rate": 0.0002971762901795222, + "loss": 0.0193, + "step": 2665 + }, + { + "epoch": 1.2545882352941176, + "grad_norm": 0.10570145521202715, + "learning_rate": 0.000297167818279751, + "loss": 0.0728, + "step": 2666 + }, + { + "epoch": 1.2550588235294118, + "grad_norm": 0.01615633959001101, + "learning_rate": 0.0002971593338111263, + "loss": 0.0018, + "step": 2667 + }, + { + "epoch": 1.255529411764706, + "grad_norm": 0.06369339836347397, + "learning_rate": 0.0002971508367743727, + "loss": 0.0349, + "step": 2668 + }, + { + "epoch": 1.256, + "grad_norm": 0.10127530478163294, + "learning_rate": 0.0002971423271702158, + "loss": 0.0481, + "step": 2669 + }, + { + "epoch": 1.256470588235294, + "grad_norm": 0.03932365435150761, + "learning_rate": 0.00029713380499938246, + "loss": 0.0138, + "step": 2670 + }, + { + "epoch": 1.2569411764705882, + "grad_norm": 0.1267789462502722, + "learning_rate": 0.00029712527026260047, + "loss": 0.0911, + "step": 2671 + }, + { + "epoch": 1.2574117647058825, + "grad_norm": 0.0876563449549422, + "learning_rate": 0.0002971167229605988, + "loss": 0.0436, + "step": 2672 + }, + { + "epoch": 1.2578823529411765, + "grad_norm": 0.0752398902674229, + "learning_rate": 0.00029710816309410737, + "loss": 0.0419, + "step": 2673 + }, + { + "epoch": 1.2583529411764705, + "grad_norm": 0.10270488012345581, + "learning_rate": 0.00029709959066385723, + "loss": 0.054, + "step": 2674 + }, + { + "epoch": 1.2588235294117647, + "grad_norm": 0.040094975038003904, + "learning_rate": 0.0002970910056705806, + "loss": 0.0105, + "step": 2675 + }, + { + "epoch": 1.259294117647059, + "grad_norm": 0.1382104284952026, + "learning_rate": 0.0002970824081150105, + "loss": 0.1134, + "step": 2676 + }, + { + "epoch": 1.259764705882353, + "grad_norm": 0.029645966648999433, + "learning_rate": 0.00029707379799788134, + "loss": 0.0104, + "step": 2677 + }, + { + "epoch": 1.2602352941176471, + "grad_norm": 7.607860749237966e-06, + "learning_rate": 0.0002970651753199285, + "loss": 0.0, + "step": 2678 + }, + { + "epoch": 1.2607058823529411, + "grad_norm": 0.09517094685744797, + "learning_rate": 0.0002970565400818883, + "loss": 0.0435, + "step": 2679 + }, + { + "epoch": 1.2611764705882353, + "grad_norm": 0.44106920166841235, + "learning_rate": 0.0002970478922844983, + "loss": 0.0965, + "step": 2680 + }, + { + "epoch": 1.2616470588235293, + "grad_norm": 0.0948819540412527, + "learning_rate": 0.00029703923192849694, + "loss": 0.05, + "step": 2681 + }, + { + "epoch": 1.2621176470588236, + "grad_norm": 0.07878834683649785, + "learning_rate": 0.000297030559014624, + "loss": 0.0427, + "step": 2682 + }, + { + "epoch": 1.2625882352941176, + "grad_norm": 0.04439773507754488, + "learning_rate": 0.00029702187354362016, + "loss": 0.0102, + "step": 2683 + }, + { + "epoch": 1.2630588235294118, + "grad_norm": 0.0644422887946759, + "learning_rate": 0.0002970131755162272, + "loss": 0.0323, + "step": 2684 + }, + { + "epoch": 1.263529411764706, + "grad_norm": 0.06845331940359774, + "learning_rate": 0.0002970044649331879, + "loss": 0.033, + "step": 2685 + }, + { + "epoch": 1.264, + "grad_norm": 0.03324754338053994, + "learning_rate": 0.0002969957417952463, + "loss": 0.0091, + "step": 2686 + }, + { + "epoch": 1.264470588235294, + "grad_norm": 0.1757746002522142, + "learning_rate": 0.00029698700610314725, + "loss": 0.0747, + "step": 2687 + }, + { + "epoch": 1.2649411764705882, + "grad_norm": 0.06925932970011223, + "learning_rate": 0.000296978257857637, + "loss": 0.0431, + "step": 2688 + }, + { + "epoch": 1.2654117647058825, + "grad_norm": 0.12453315386561714, + "learning_rate": 0.00029696949705946255, + "loss": 0.0766, + "step": 2689 + }, + { + "epoch": 1.2658823529411765, + "grad_norm": 0.29798950375463046, + "learning_rate": 0.0002969607237093722, + "loss": 0.1044, + "step": 2690 + }, + { + "epoch": 1.2663529411764705, + "grad_norm": 0.13599304141146976, + "learning_rate": 0.00029695193780811523, + "loss": 0.0814, + "step": 2691 + }, + { + "epoch": 1.2668235294117647, + "grad_norm": 0.10562382912314922, + "learning_rate": 0.0002969431393564419, + "loss": 0.0671, + "step": 2692 + }, + { + "epoch": 1.267294117647059, + "grad_norm": 0.08153841675385703, + "learning_rate": 0.0002969343283551038, + "loss": 0.0143, + "step": 2693 + }, + { + "epoch": 1.267764705882353, + "grad_norm": 0.13165599174684514, + "learning_rate": 0.0002969255048048533, + "loss": 0.1087, + "step": 2694 + }, + { + "epoch": 1.2682352941176471, + "grad_norm": 1.5089380956444913e-05, + "learning_rate": 0.00029691666870644405, + "loss": 0.0, + "step": 2695 + }, + { + "epoch": 1.2687058823529411, + "grad_norm": 4.589100843080756e-05, + "learning_rate": 0.0002969078200606307, + "loss": 0.0, + "step": 2696 + }, + { + "epoch": 1.2691764705882354, + "grad_norm": 0.11796711640295207, + "learning_rate": 0.0002968989588681688, + "loss": 0.0619, + "step": 2697 + }, + { + "epoch": 1.2696470588235294, + "grad_norm": 0.060419072631476715, + "learning_rate": 0.0002968900851298154, + "loss": 0.0227, + "step": 2698 + }, + { + "epoch": 1.2701176470588236, + "grad_norm": 0.07538253341472459, + "learning_rate": 0.00029688119884632827, + "loss": 0.0378, + "step": 2699 + }, + { + "epoch": 1.2705882352941176, + "grad_norm": 5.46976718870009e-05, + "learning_rate": 0.0002968723000184662, + "loss": 0.0, + "step": 2700 + }, + { + "epoch": 1.2710588235294118, + "grad_norm": 0.10029053073906422, + "learning_rate": 0.0002968633886469894, + "loss": 0.0631, + "step": 2701 + }, + { + "epoch": 1.271529411764706, + "grad_norm": 0.13653052603952143, + "learning_rate": 0.00029685446473265876, + "loss": 0.0985, + "step": 2702 + }, + { + "epoch": 1.272, + "grad_norm": 0.04300127616624249, + "learning_rate": 0.00029684552827623657, + "loss": 0.0134, + "step": 2703 + }, + { + "epoch": 1.272470588235294, + "grad_norm": 0.04332321119630446, + "learning_rate": 0.000296836579278486, + "loss": 0.0184, + "step": 2704 + }, + { + "epoch": 1.2729411764705882, + "grad_norm": 0.0854668663522192, + "learning_rate": 0.0002968276177401713, + "loss": 0.0093, + "step": 2705 + }, + { + "epoch": 1.2734117647058825, + "grad_norm": 0.0813077340027717, + "learning_rate": 0.00029681864366205786, + "loss": 0.0444, + "step": 2706 + }, + { + "epoch": 1.2738823529411765, + "grad_norm": 0.09350512408464992, + "learning_rate": 0.0002968096570449121, + "loss": 0.0664, + "step": 2707 + }, + { + "epoch": 1.2743529411764705, + "grad_norm": 0.0662417671180788, + "learning_rate": 0.0002968006578895016, + "loss": 0.0307, + "step": 2708 + }, + { + "epoch": 1.2748235294117647, + "grad_norm": 0.08866692841969945, + "learning_rate": 0.0002967916461965948, + "loss": 0.0392, + "step": 2709 + }, + { + "epoch": 1.275294117647059, + "grad_norm": 0.10228997947259814, + "learning_rate": 0.00029678262196696143, + "loss": 0.0561, + "step": 2710 + }, + { + "epoch": 1.275764705882353, + "grad_norm": 0.09594582436436741, + "learning_rate": 0.0002967735852013722, + "loss": 0.0636, + "step": 2711 + }, + { + "epoch": 1.2762352941176471, + "grad_norm": 0.08527473037518563, + "learning_rate": 0.0002967645359005988, + "loss": 0.0683, + "step": 2712 + }, + { + "epoch": 1.2767058823529411, + "grad_norm": 3.275720087458997e-05, + "learning_rate": 0.0002967554740654142, + "loss": 0.0, + "step": 2713 + }, + { + "epoch": 1.2771764705882354, + "grad_norm": 0.04941671537645501, + "learning_rate": 0.0002967463996965923, + "loss": 0.021, + "step": 2714 + }, + { + "epoch": 1.2776470588235294, + "grad_norm": 0.16385070889328468, + "learning_rate": 0.0002967373127949081, + "loss": 0.1045, + "step": 2715 + }, + { + "epoch": 1.2781176470588236, + "grad_norm": 0.04512831876118571, + "learning_rate": 0.0002967282133611375, + "loss": 0.0231, + "step": 2716 + }, + { + "epoch": 1.2785882352941176, + "grad_norm": 0.07469543349850939, + "learning_rate": 0.0002967191013960579, + "loss": 0.0428, + "step": 2717 + }, + { + "epoch": 1.2790588235294118, + "grad_norm": 0.057252895202467895, + "learning_rate": 0.00029670997690044745, + "loss": 0.0251, + "step": 2718 + }, + { + "epoch": 1.2795294117647058, + "grad_norm": 0.011721624988305908, + "learning_rate": 0.0002967008398750853, + "loss": 0.0017, + "step": 2719 + }, + { + "epoch": 1.28, + "grad_norm": 0.04941328839966066, + "learning_rate": 0.00029669169032075186, + "loss": 0.0154, + "step": 2720 + }, + { + "epoch": 1.280470588235294, + "grad_norm": 0.03527728283796433, + "learning_rate": 0.0002966825282382285, + "loss": 0.0112, + "step": 2721 + }, + { + "epoch": 1.2809411764705882, + "grad_norm": 0.06831455435959033, + "learning_rate": 0.0002966733536282978, + "loss": 0.0323, + "step": 2722 + }, + { + "epoch": 1.2814117647058825, + "grad_norm": 0.04573289545498092, + "learning_rate": 0.0002966641664917433, + "loss": 0.0087, + "step": 2723 + }, + { + "epoch": 1.2818823529411765, + "grad_norm": 0.05846102657492262, + "learning_rate": 0.0002966549668293495, + "loss": 0.0215, + "step": 2724 + }, + { + "epoch": 1.2823529411764705, + "grad_norm": 0.10832213379195, + "learning_rate": 0.0002966457546419023, + "loss": 0.0694, + "step": 2725 + }, + { + "epoch": 1.2828235294117647, + "grad_norm": 2.5299178641769356e-05, + "learning_rate": 0.0002966365299301883, + "loss": 0.0, + "step": 2726 + }, + { + "epoch": 1.283294117647059, + "grad_norm": 0.03746862400894933, + "learning_rate": 0.00029662729269499546, + "loss": 0.0115, + "step": 2727 + }, + { + "epoch": 1.283764705882353, + "grad_norm": 0.1069542047591076, + "learning_rate": 0.00029661804293711254, + "loss": 0.0877, + "step": 2728 + }, + { + "epoch": 1.2842352941176471, + "grad_norm": 0.1448827473135949, + "learning_rate": 0.00029660878065732965, + "loss": 0.117, + "step": 2729 + }, + { + "epoch": 1.2847058823529411, + "grad_norm": 0.04423650344704903, + "learning_rate": 0.0002965995058564377, + "loss": 0.0165, + "step": 2730 + }, + { + "epoch": 1.2851764705882354, + "grad_norm": 2.4399483248613687e-05, + "learning_rate": 0.000296590218535229, + "loss": 0.0, + "step": 2731 + }, + { + "epoch": 1.2856470588235294, + "grad_norm": 0.00020244101559432055, + "learning_rate": 0.00029658091869449657, + "loss": 0.0, + "step": 2732 + }, + { + "epoch": 1.2861176470588236, + "grad_norm": 0.09530232852682624, + "learning_rate": 0.0002965716063350347, + "loss": 0.0698, + "step": 2733 + }, + { + "epoch": 1.2865882352941176, + "grad_norm": 0.06073826188227115, + "learning_rate": 0.0002965622814576387, + "loss": 0.0313, + "step": 2734 + }, + { + "epoch": 1.2870588235294118, + "grad_norm": 0.013701054333767209, + "learning_rate": 0.000296552944063105, + "loss": 0.0014, + "step": 2735 + }, + { + "epoch": 1.2875294117647058, + "grad_norm": 0.12882188316794851, + "learning_rate": 0.0002965435941522311, + "loss": 0.0571, + "step": 2736 + }, + { + "epoch": 1.288, + "grad_norm": 0.09347134099677823, + "learning_rate": 0.00029653423172581534, + "loss": 0.0868, + "step": 2737 + }, + { + "epoch": 1.288470588235294, + "grad_norm": 0.11426673064965655, + "learning_rate": 0.0002965248567846575, + "loss": 0.0799, + "step": 2738 + }, + { + "epoch": 1.2889411764705883, + "grad_norm": 0.04628731770940934, + "learning_rate": 0.0002965154693295582, + "loss": 0.0138, + "step": 2739 + }, + { + "epoch": 1.2894117647058825, + "grad_norm": 0.028385037599366992, + "learning_rate": 0.00029650606936131913, + "loss": 0.0092, + "step": 2740 + }, + { + "epoch": 1.2898823529411765, + "grad_norm": 0.024691074226499753, + "learning_rate": 0.0002964966568807432, + "loss": 0.005, + "step": 2741 + }, + { + "epoch": 1.2903529411764705, + "grad_norm": 0.0664158860748136, + "learning_rate": 0.00029648723188863417, + "loss": 0.0406, + "step": 2742 + }, + { + "epoch": 1.2908235294117647, + "grad_norm": 0.05149006183491765, + "learning_rate": 0.000296477794385797, + "loss": 0.0261, + "step": 2743 + }, + { + "epoch": 1.291294117647059, + "grad_norm": 0.13556172570917482, + "learning_rate": 0.00029646834437303777, + "loss": 0.0996, + "step": 2744 + }, + { + "epoch": 1.291764705882353, + "grad_norm": 0.07712952256682207, + "learning_rate": 0.00029645888185116353, + "loss": 0.0338, + "step": 2745 + }, + { + "epoch": 1.292235294117647, + "grad_norm": 0.1330485630001692, + "learning_rate": 0.00029644940682098234, + "loss": 0.0878, + "step": 2746 + }, + { + "epoch": 1.2927058823529411, + "grad_norm": 0.019883049421069342, + "learning_rate": 0.00029643991928330346, + "loss": 0.0041, + "step": 2747 + }, + { + "epoch": 1.2931764705882354, + "grad_norm": 0.07082880794914802, + "learning_rate": 0.0002964304192389372, + "loss": 0.0373, + "step": 2748 + }, + { + "epoch": 1.2936470588235294, + "grad_norm": 0.040658002503641634, + "learning_rate": 0.000296420906688695, + "loss": 0.0154, + "step": 2749 + }, + { + "epoch": 1.2941176470588236, + "grad_norm": 0.04767728772466225, + "learning_rate": 0.00029641138163338907, + "loss": 0.0116, + "step": 2750 + }, + { + "epoch": 1.2945882352941176, + "grad_norm": 0.17353462524687085, + "learning_rate": 0.00029640184407383303, + "loss": 0.1112, + "step": 2751 + }, + { + "epoch": 1.2950588235294118, + "grad_norm": 0.1452202623825693, + "learning_rate": 0.00029639229401084137, + "loss": 0.0647, + "step": 2752 + }, + { + "epoch": 1.2955294117647058, + "grad_norm": 1.652698698725526e-05, + "learning_rate": 0.00029638273144522983, + "loss": 0.0, + "step": 2753 + }, + { + "epoch": 1.296, + "grad_norm": 0.11853988246022212, + "learning_rate": 0.00029637315637781494, + "loss": 0.0982, + "step": 2754 + }, + { + "epoch": 1.296470588235294, + "grad_norm": 7.119828740678697e-06, + "learning_rate": 0.0002963635688094146, + "loss": 0.0, + "step": 2755 + }, + { + "epoch": 1.2969411764705883, + "grad_norm": 9.736328935366827e-06, + "learning_rate": 0.0002963539687408476, + "loss": 0.0, + "step": 2756 + }, + { + "epoch": 1.2974117647058825, + "grad_norm": 1.3194428063028743e-05, + "learning_rate": 0.00029634435617293374, + "loss": 0.0, + "step": 2757 + }, + { + "epoch": 1.2978823529411765, + "grad_norm": 0.01257585287184361, + "learning_rate": 0.00029633473110649407, + "loss": 0.0016, + "step": 2758 + }, + { + "epoch": 1.2983529411764705, + "grad_norm": 0.10156252121033388, + "learning_rate": 0.00029632509354235063, + "loss": 0.0527, + "step": 2759 + }, + { + "epoch": 1.2988235294117647, + "grad_norm": 0.11078881583162141, + "learning_rate": 0.00029631544348132643, + "loss": 0.0671, + "step": 2760 + }, + { + "epoch": 1.299294117647059, + "grad_norm": 0.08109697698589156, + "learning_rate": 0.0002963057809242457, + "loss": 0.0395, + "step": 2761 + }, + { + "epoch": 1.299764705882353, + "grad_norm": 5.7325562074091776e-05, + "learning_rate": 0.00029629610587193365, + "loss": 0.0, + "step": 2762 + }, + { + "epoch": 1.300235294117647, + "grad_norm": 0.09898135481509587, + "learning_rate": 0.00029628641832521663, + "loss": 0.0723, + "step": 2763 + }, + { + "epoch": 1.3007058823529412, + "grad_norm": 0.15643317081946947, + "learning_rate": 0.0002962767182849219, + "loss": 0.1164, + "step": 2764 + }, + { + "epoch": 1.3011764705882354, + "grad_norm": 6.174360472950066e-06, + "learning_rate": 0.00029626700575187794, + "loss": 0.0, + "step": 2765 + }, + { + "epoch": 1.3016470588235294, + "grad_norm": 0.2391110233301646, + "learning_rate": 0.0002962572807269143, + "loss": 0.1821, + "step": 2766 + }, + { + "epoch": 1.3021176470588236, + "grad_norm": 0.09118009733737885, + "learning_rate": 0.00029624754321086146, + "loss": 0.0648, + "step": 2767 + }, + { + "epoch": 1.3025882352941176, + "grad_norm": 0.07043734753222466, + "learning_rate": 0.00029623779320455106, + "loss": 0.0524, + "step": 2768 + }, + { + "epoch": 1.3030588235294118, + "grad_norm": 5.356914070815326e-05, + "learning_rate": 0.00029622803070881586, + "loss": 0.0, + "step": 2769 + }, + { + "epoch": 1.3035294117647058, + "grad_norm": 5.537316035689823e-05, + "learning_rate": 0.00029621825572448965, + "loss": 0.0, + "step": 2770 + }, + { + "epoch": 1.304, + "grad_norm": 0.08812496609011383, + "learning_rate": 0.0002962084682524071, + "loss": 0.0472, + "step": 2771 + }, + { + "epoch": 1.304470588235294, + "grad_norm": 0.07631987234778383, + "learning_rate": 0.0002961986682934043, + "loss": 0.0398, + "step": 2772 + }, + { + "epoch": 1.3049411764705883, + "grad_norm": 1.1422760377885066e-05, + "learning_rate": 0.0002961888558483181, + "loss": 0.0, + "step": 2773 + }, + { + "epoch": 1.3054117647058823, + "grad_norm": 0.14686170468443213, + "learning_rate": 0.0002961790309179866, + "loss": 0.1015, + "step": 2774 + }, + { + "epoch": 1.3058823529411765, + "grad_norm": 2.2186163824320826e-05, + "learning_rate": 0.00029616919350324873, + "loss": 0.0, + "step": 2775 + }, + { + "epoch": 1.3063529411764705, + "grad_norm": 0.1521129882885456, + "learning_rate": 0.0002961593436049449, + "loss": 0.1042, + "step": 2776 + }, + { + "epoch": 1.3068235294117647, + "grad_norm": 0.07826838158069835, + "learning_rate": 0.0002961494812239162, + "loss": 0.0551, + "step": 2777 + }, + { + "epoch": 1.307294117647059, + "grad_norm": 0.050819302491129224, + "learning_rate": 0.00029613960636100487, + "loss": 0.0199, + "step": 2778 + }, + { + "epoch": 1.307764705882353, + "grad_norm": 0.09683568194120871, + "learning_rate": 0.0002961297190170544, + "loss": 0.0587, + "step": 2779 + }, + { + "epoch": 1.308235294117647, + "grad_norm": 0.012073434475918554, + "learning_rate": 0.00029611981919290916, + "loss": 0.0022, + "step": 2780 + }, + { + "epoch": 1.3087058823529412, + "grad_norm": 0.05344491819145629, + "learning_rate": 0.0002961099068894146, + "loss": 0.0205, + "step": 2781 + }, + { + "epoch": 1.3091764705882354, + "grad_norm": 0.1255165912575236, + "learning_rate": 0.00029609998210741744, + "loss": 0.0719, + "step": 2782 + }, + { + "epoch": 1.3096470588235294, + "grad_norm": 0.03366742247518661, + "learning_rate": 0.0002960900448477651, + "loss": 0.0083, + "step": 2783 + }, + { + "epoch": 1.3101176470588236, + "grad_norm": 0.16530220464633577, + "learning_rate": 0.0002960800951113064, + "loss": 0.0987, + "step": 2784 + }, + { + "epoch": 1.3105882352941176, + "grad_norm": 0.073782205586991, + "learning_rate": 0.00029607013289889106, + "loss": 0.054, + "step": 2785 + }, + { + "epoch": 1.3110588235294118, + "grad_norm": 0.03530081221884047, + "learning_rate": 0.0002960601582113699, + "loss": 0.0133, + "step": 2786 + }, + { + "epoch": 1.3115294117647058, + "grad_norm": 0.8242804910223728, + "learning_rate": 0.00029605017104959475, + "loss": 0.281, + "step": 2787 + }, + { + "epoch": 1.312, + "grad_norm": 0.028876699847351483, + "learning_rate": 0.00029604017141441865, + "loss": 0.0064, + "step": 2788 + }, + { + "epoch": 1.312470588235294, + "grad_norm": 0.08481502241232028, + "learning_rate": 0.0002960301593066956, + "loss": 0.0501, + "step": 2789 + }, + { + "epoch": 1.3129411764705883, + "grad_norm": 3.1995452539584786e-05, + "learning_rate": 0.0002960201347272807, + "loss": 0.0, + "step": 2790 + }, + { + "epoch": 1.3134117647058823, + "grad_norm": 0.07613391681781301, + "learning_rate": 0.00029601009767703007, + "loss": 0.015, + "step": 2791 + }, + { + "epoch": 1.3138823529411765, + "grad_norm": 0.0007142277338666301, + "learning_rate": 0.0002960000481568009, + "loss": 0.0, + "step": 2792 + }, + { + "epoch": 1.3143529411764705, + "grad_norm": 0.0410960865026435, + "learning_rate": 0.0002959899861674515, + "loss": 0.0064, + "step": 2793 + }, + { + "epoch": 1.3148235294117647, + "grad_norm": 0.20601735186304548, + "learning_rate": 0.0002959799117098412, + "loss": 0.0824, + "step": 2794 + }, + { + "epoch": 1.315294117647059, + "grad_norm": 9.888537506015922, + "learning_rate": 0.00029596982478483037, + "loss": 0.0625, + "step": 2795 + }, + { + "epoch": 1.315764705882353, + "grad_norm": 0.17843395500405484, + "learning_rate": 0.00029595972539328053, + "loss": 0.0829, + "step": 2796 + }, + { + "epoch": 1.316235294117647, + "grad_norm": 1.2939835010863265, + "learning_rate": 0.00029594961353605426, + "loss": 0.1581, + "step": 2797 + }, + { + "epoch": 1.3167058823529412, + "grad_norm": 0.02294205079552966, + "learning_rate": 0.0002959394892140151, + "loss": 0.0041, + "step": 2798 + }, + { + "epoch": 1.3171764705882354, + "grad_norm": 0.13686830253556584, + "learning_rate": 0.00029592935242802774, + "loss": 0.0702, + "step": 2799 + }, + { + "epoch": 1.3176470588235294, + "grad_norm": 0.07340960734213033, + "learning_rate": 0.0002959192031789579, + "loss": 0.022, + "step": 2800 + }, + { + "epoch": 1.3181176470588234, + "grad_norm": 0.24802758925724663, + "learning_rate": 0.00029590904146767236, + "loss": 0.1342, + "step": 2801 + }, + { + "epoch": 1.3185882352941176, + "grad_norm": 2.8682004806323143, + "learning_rate": 0.000295898867295039, + "loss": 0.1193, + "step": 2802 + }, + { + "epoch": 1.3190588235294118, + "grad_norm": 0.1736442236090424, + "learning_rate": 0.0002958886806619268, + "loss": 0.1176, + "step": 2803 + }, + { + "epoch": 1.3195294117647058, + "grad_norm": 0.05403869839355191, + "learning_rate": 0.00029587848156920564, + "loss": 0.0115, + "step": 2804 + }, + { + "epoch": 1.32, + "grad_norm": 0.09612997954236063, + "learning_rate": 0.0002958682700177466, + "loss": 0.0393, + "step": 2805 + }, + { + "epoch": 1.320470588235294, + "grad_norm": 0.19244058411639778, + "learning_rate": 0.00029585804600842187, + "loss": 0.1163, + "step": 2806 + }, + { + "epoch": 1.3209411764705883, + "grad_norm": 0.23711340564495614, + "learning_rate": 0.00029584780954210454, + "loss": 0.1441, + "step": 2807 + }, + { + "epoch": 1.3214117647058823, + "grad_norm": 8.358674337318718, + "learning_rate": 0.0002958375606196689, + "loss": 0.0198, + "step": 2808 + }, + { + "epoch": 1.3218823529411765, + "grad_norm": 0.30333515498330266, + "learning_rate": 0.0002958272992419903, + "loss": 0.1007, + "step": 2809 + }, + { + "epoch": 1.3223529411764705, + "grad_norm": 0.1609686385427299, + "learning_rate": 0.000295817025409945, + "loss": 0.0739, + "step": 2810 + }, + { + "epoch": 1.3228235294117647, + "grad_norm": 0.19606482734987069, + "learning_rate": 0.00029580673912441057, + "loss": 0.0875, + "step": 2811 + }, + { + "epoch": 1.323294117647059, + "grad_norm": 0.11144989428014802, + "learning_rate": 0.00029579644038626546, + "loss": 0.0529, + "step": 2812 + }, + { + "epoch": 1.323764705882353, + "grad_norm": 0.14224739850040657, + "learning_rate": 0.00029578612919638913, + "loss": 0.0803, + "step": 2813 + }, + { + "epoch": 1.324235294117647, + "grad_norm": 0.016826317386845754, + "learning_rate": 0.0002957758055556623, + "loss": 0.0003, + "step": 2814 + }, + { + "epoch": 1.3247058823529412, + "grad_norm": 0.5287168345552562, + "learning_rate": 0.0002957654694649667, + "loss": 0.2312, + "step": 2815 + }, + { + "epoch": 1.3251764705882354, + "grad_norm": 0.18120888781890251, + "learning_rate": 0.000295755120925185, + "loss": 0.0627, + "step": 2816 + }, + { + "epoch": 1.3256470588235294, + "grad_norm": 0.21700235534464424, + "learning_rate": 0.0002957447599372011, + "loss": 0.1354, + "step": 2817 + }, + { + "epoch": 1.3261176470588234, + "grad_norm": 0.14760069389827846, + "learning_rate": 0.00029573438650189974, + "loss": 0.1338, + "step": 2818 + }, + { + "epoch": 1.3265882352941176, + "grad_norm": 0.20513970566028064, + "learning_rate": 0.000295724000620167, + "loss": 0.1829, + "step": 2819 + }, + { + "epoch": 1.3270588235294118, + "grad_norm": 0.04981694015001662, + "learning_rate": 0.00029571360229288984, + "loss": 0.0129, + "step": 2820 + }, + { + "epoch": 1.3275294117647058, + "grad_norm": 0.1212985583200148, + "learning_rate": 0.0002957031915209563, + "loss": 0.0875, + "step": 2821 + }, + { + "epoch": 1.328, + "grad_norm": 0.10174083027494606, + "learning_rate": 0.00029569276830525553, + "loss": 0.0682, + "step": 2822 + }, + { + "epoch": 1.328470588235294, + "grad_norm": 0.0012331844686254187, + "learning_rate": 0.00029568233264667775, + "loss": 0.0, + "step": 2823 + }, + { + "epoch": 1.3289411764705883, + "grad_norm": 0.0015936471638384017, + "learning_rate": 0.00029567188454611415, + "loss": 0.0, + "step": 2824 + }, + { + "epoch": 1.3294117647058823, + "grad_norm": 0.17132013108180205, + "learning_rate": 0.00029566142400445716, + "loss": 0.045, + "step": 2825 + }, + { + "epoch": 1.3298823529411765, + "grad_norm": 0.03345761583480316, + "learning_rate": 0.0002956509510226001, + "loss": 0.0094, + "step": 2826 + }, + { + "epoch": 1.3303529411764705, + "grad_norm": 0.10070957962454817, + "learning_rate": 0.00029564046560143735, + "loss": 0.0738, + "step": 2827 + }, + { + "epoch": 1.3308235294117647, + "grad_norm": 0.1323646458341415, + "learning_rate": 0.0002956299677418645, + "loss": 0.0601, + "step": 2828 + }, + { + "epoch": 1.331294117647059, + "grad_norm": 0.00038479303460909817, + "learning_rate": 0.000295619457444778, + "loss": 0.0, + "step": 2829 + }, + { + "epoch": 1.331764705882353, + "grad_norm": 0.05720699559227261, + "learning_rate": 0.00029560893471107573, + "loss": 0.0206, + "step": 2830 + }, + { + "epoch": 1.332235294117647, + "grad_norm": 0.1570060608127341, + "learning_rate": 0.00029559839954165616, + "loss": 0.0251, + "step": 2831 + }, + { + "epoch": 1.3327058823529412, + "grad_norm": 0.12075395184692506, + "learning_rate": 0.00029558785193741914, + "loss": 0.0742, + "step": 2832 + }, + { + "epoch": 1.3331764705882354, + "grad_norm": 0.10124723849008058, + "learning_rate": 0.0002955772918992654, + "loss": 0.0414, + "step": 2833 + }, + { + "epoch": 1.3336470588235294, + "grad_norm": 0.03804076960454692, + "learning_rate": 0.00029556671942809694, + "loss": 0.0092, + "step": 2834 + }, + { + "epoch": 1.3341176470588234, + "grad_norm": 0.061154841122718163, + "learning_rate": 0.00029555613452481657, + "loss": 0.0262, + "step": 2835 + }, + { + "epoch": 1.3345882352941176, + "grad_norm": 0.0002000528154231874, + "learning_rate": 0.00029554553719032844, + "loss": 0.0, + "step": 2836 + }, + { + "epoch": 1.3350588235294119, + "grad_norm": 0.03400248628830289, + "learning_rate": 0.00029553492742553754, + "loss": 0.006, + "step": 2837 + }, + { + "epoch": 1.3355294117647059, + "grad_norm": 0.00014546206594272395, + "learning_rate": 0.00029552430523134993, + "loss": 0.0, + "step": 2838 + }, + { + "epoch": 1.336, + "grad_norm": 0.03045551824057359, + "learning_rate": 0.0002955136706086729, + "loss": 0.0049, + "step": 2839 + }, + { + "epoch": 1.336470588235294, + "grad_norm": 7.500623454225777e-05, + "learning_rate": 0.0002955030235584146, + "loss": 0.0, + "step": 2840 + }, + { + "epoch": 1.3369411764705883, + "grad_norm": 0.04816055806993561, + "learning_rate": 0.00029549236408148453, + "loss": 0.0107, + "step": 2841 + }, + { + "epoch": 1.3374117647058823, + "grad_norm": 0.19213977563560386, + "learning_rate": 0.00029548169217879285, + "loss": 0.0942, + "step": 2842 + }, + { + "epoch": 1.3378823529411765, + "grad_norm": 0.05177367701451383, + "learning_rate": 0.000295471007851251, + "loss": 0.0205, + "step": 2843 + }, + { + "epoch": 1.3383529411764705, + "grad_norm": 0.07606318031822606, + "learning_rate": 0.00029546031109977164, + "loss": 0.0332, + "step": 2844 + }, + { + "epoch": 1.3388235294117647, + "grad_norm": 0.11114367342975041, + "learning_rate": 0.00029544960192526824, + "loss": 0.056, + "step": 2845 + }, + { + "epoch": 1.3392941176470587, + "grad_norm": 0.06233738510775247, + "learning_rate": 0.0002954388803286554, + "loss": 0.0448, + "step": 2846 + }, + { + "epoch": 1.339764705882353, + "grad_norm": 0.27492081362311344, + "learning_rate": 0.0002954281463108488, + "loss": 0.1623, + "step": 2847 + }, + { + "epoch": 1.340235294117647, + "grad_norm": 0.1980598966824866, + "learning_rate": 0.0002954173998727652, + "loss": 0.1355, + "step": 2848 + }, + { + "epoch": 1.3407058823529412, + "grad_norm": 0.085128356027909, + "learning_rate": 0.00029540664101532227, + "loss": 0.0353, + "step": 2849 + }, + { + "epoch": 1.3411764705882354, + "grad_norm": 0.16657470445944042, + "learning_rate": 0.0002953958697394391, + "loss": 0.1131, + "step": 2850 + }, + { + "epoch": 1.3416470588235294, + "grad_norm": 0.06140069382759848, + "learning_rate": 0.0002953850860460355, + "loss": 0.03, + "step": 2851 + }, + { + "epoch": 1.3421176470588234, + "grad_norm": 0.1062736856336364, + "learning_rate": 0.00029537428993603237, + "loss": 0.088, + "step": 2852 + }, + { + "epoch": 1.3425882352941176, + "grad_norm": 9.891301605323584e-05, + "learning_rate": 0.00029536348141035183, + "loss": 0.0, + "step": 2853 + }, + { + "epoch": 1.3430588235294119, + "grad_norm": 0.15330870199530133, + "learning_rate": 0.000295352660469917, + "loss": 0.1298, + "step": 2854 + }, + { + "epoch": 1.3435294117647059, + "grad_norm": 5.58079530304228e-05, + "learning_rate": 0.00029534182711565203, + "loss": 0.0, + "step": 2855 + }, + { + "epoch": 1.3439999999999999, + "grad_norm": 0.17469921114600181, + "learning_rate": 0.0002953309813484821, + "loss": 0.098, + "step": 2856 + }, + { + "epoch": 1.344470588235294, + "grad_norm": 0.05806494243656373, + "learning_rate": 0.0002953201231693336, + "loss": 0.0236, + "step": 2857 + }, + { + "epoch": 1.3449411764705883, + "grad_norm": 7.486528560095255e-05, + "learning_rate": 0.00029530925257913366, + "loss": 0.0, + "step": 2858 + }, + { + "epoch": 1.3454117647058823, + "grad_norm": 0.07379475503782021, + "learning_rate": 0.0002952983695788109, + "loss": 0.0267, + "step": 2859 + }, + { + "epoch": 1.3458823529411765, + "grad_norm": 0.09699908853578722, + "learning_rate": 0.00029528747416929463, + "loss": 0.0477, + "step": 2860 + }, + { + "epoch": 1.3463529411764705, + "grad_norm": 4.773744025641305e-05, + "learning_rate": 0.0002952765663515155, + "loss": 0.0, + "step": 2861 + }, + { + "epoch": 1.3468235294117648, + "grad_norm": 0.07048669601215629, + "learning_rate": 0.000295265646126405, + "loss": 0.03, + "step": 2862 + }, + { + "epoch": 1.3472941176470588, + "grad_norm": 0.00020649168892954, + "learning_rate": 0.00029525471349489576, + "loss": 0.0, + "step": 2863 + }, + { + "epoch": 1.347764705882353, + "grad_norm": 0.17327149051074647, + "learning_rate": 0.00029524376845792153, + "loss": 0.098, + "step": 2864 + }, + { + "epoch": 1.348235294117647, + "grad_norm": 0.030533968269225803, + "learning_rate": 0.0002952328110164171, + "loss": 0.0178, + "step": 2865 + }, + { + "epoch": 1.3487058823529412, + "grad_norm": 0.06609929932377563, + "learning_rate": 0.0002952218411713182, + "loss": 0.0268, + "step": 2866 + }, + { + "epoch": 1.3491764705882354, + "grad_norm": 0.12018530329240658, + "learning_rate": 0.0002952108589235618, + "loss": 0.0508, + "step": 2867 + }, + { + "epoch": 1.3496470588235294, + "grad_norm": 0.0969471399384097, + "learning_rate": 0.0002951998642740858, + "loss": 0.0515, + "step": 2868 + }, + { + "epoch": 1.3501176470588234, + "grad_norm": 0.05763773993122284, + "learning_rate": 0.00029518885722382913, + "loss": 0.0238, + "step": 2869 + }, + { + "epoch": 1.3505882352941176, + "grad_norm": 0.13016573225360725, + "learning_rate": 0.00029517783777373194, + "loss": 0.1572, + "step": 2870 + }, + { + "epoch": 1.3510588235294119, + "grad_norm": 0.09937796903886034, + "learning_rate": 0.0002951668059247353, + "loss": 0.0802, + "step": 2871 + }, + { + "epoch": 1.3515294117647059, + "grad_norm": 0.1102821928013641, + "learning_rate": 0.0002951557616777814, + "loss": 0.0826, + "step": 2872 + }, + { + "epoch": 1.3519999999999999, + "grad_norm": 0.04596930148241357, + "learning_rate": 0.0002951447050338135, + "loss": 0.0179, + "step": 2873 + }, + { + "epoch": 1.352470588235294, + "grad_norm": 0.14243005155320845, + "learning_rate": 0.00029513363599377587, + "loss": 0.1056, + "step": 2874 + }, + { + "epoch": 1.3529411764705883, + "grad_norm": 0.133840391987507, + "learning_rate": 0.00029512255455861375, + "loss": 0.0993, + "step": 2875 + }, + { + "epoch": 1.3534117647058823, + "grad_norm": 0.06326804940594152, + "learning_rate": 0.0002951114607292737, + "loss": 0.034, + "step": 2876 + }, + { + "epoch": 1.3538823529411765, + "grad_norm": 0.031149707144414094, + "learning_rate": 0.0002951003545067032, + "loss": 0.008, + "step": 2877 + }, + { + "epoch": 1.3543529411764705, + "grad_norm": 0.09189986086518367, + "learning_rate": 0.0002950892358918506, + "loss": 0.0481, + "step": 2878 + }, + { + "epoch": 1.3548235294117648, + "grad_norm": 0.12683218748501712, + "learning_rate": 0.00029507810488566563, + "loss": 0.0846, + "step": 2879 + }, + { + "epoch": 1.3552941176470588, + "grad_norm": 6.596283962907325e-05, + "learning_rate": 0.000295066961489099, + "loss": 0.0, + "step": 2880 + }, + { + "epoch": 1.355764705882353, + "grad_norm": 0.0527215326768972, + "learning_rate": 0.0002950558057031022, + "loss": 0.0166, + "step": 2881 + }, + { + "epoch": 1.356235294117647, + "grad_norm": 0.09601814067807712, + "learning_rate": 0.0002950446375286281, + "loss": 0.0632, + "step": 2882 + }, + { + "epoch": 1.3567058823529412, + "grad_norm": 0.12778078930380382, + "learning_rate": 0.00029503345696663054, + "loss": 0.0727, + "step": 2883 + }, + { + "epoch": 1.3571764705882354, + "grad_norm": 0.09306017202965298, + "learning_rate": 0.0002950222640180644, + "loss": 0.054, + "step": 2884 + }, + { + "epoch": 1.3576470588235294, + "grad_norm": 0.04637326045022564, + "learning_rate": 0.0002950110586838855, + "loss": 0.02, + "step": 2885 + }, + { + "epoch": 1.3581176470588234, + "grad_norm": 3.6758199230224816e-05, + "learning_rate": 0.00029499984096505107, + "loss": 0.0, + "step": 2886 + }, + { + "epoch": 1.3585882352941177, + "grad_norm": 0.035253472632312835, + "learning_rate": 0.00029498861086251887, + "loss": 0.0102, + "step": 2887 + }, + { + "epoch": 1.3590588235294119, + "grad_norm": 3.771786289495913e-05, + "learning_rate": 0.00029497736837724817, + "loss": 0.0, + "step": 2888 + }, + { + "epoch": 1.3595294117647059, + "grad_norm": 3.624487186375947e-05, + "learning_rate": 0.0002949661135101991, + "loss": 0.0, + "step": 2889 + }, + { + "epoch": 1.3599999999999999, + "grad_norm": 5.502891598415512e-05, + "learning_rate": 0.0002949548462623329, + "loss": 0.0, + "step": 2890 + }, + { + "epoch": 1.360470588235294, + "grad_norm": 0.21705183956742555, + "learning_rate": 0.0002949435666346118, + "loss": 0.1791, + "step": 2891 + }, + { + "epoch": 1.3609411764705883, + "grad_norm": 0.09477511910478498, + "learning_rate": 0.00029493227462799914, + "loss": 0.0478, + "step": 2892 + }, + { + "epoch": 1.3614117647058823, + "grad_norm": 0.07735174786432432, + "learning_rate": 0.00029492097024345936, + "loss": 0.0356, + "step": 2893 + }, + { + "epoch": 1.3618823529411765, + "grad_norm": 0.08833144393409947, + "learning_rate": 0.00029490965348195794, + "loss": 0.0606, + "step": 2894 + }, + { + "epoch": 1.3623529411764705, + "grad_norm": 0.04140164402374996, + "learning_rate": 0.00029489832434446126, + "loss": 0.0151, + "step": 2895 + }, + { + "epoch": 1.3628235294117648, + "grad_norm": 5.035638833717385e-05, + "learning_rate": 0.000294886982831937, + "loss": 0.0, + "step": 2896 + }, + { + "epoch": 1.3632941176470588, + "grad_norm": 3.5795020257550275e-05, + "learning_rate": 0.00029487562894535375, + "loss": 0.0, + "step": 2897 + }, + { + "epoch": 1.363764705882353, + "grad_norm": 0.13120117653802035, + "learning_rate": 0.0002948642626856812, + "loss": 0.0862, + "step": 2898 + }, + { + "epoch": 1.364235294117647, + "grad_norm": 0.10321324972413293, + "learning_rate": 0.00029485288405389, + "loss": 0.0115, + "step": 2899 + }, + { + "epoch": 1.3647058823529412, + "grad_norm": 0.042409086146513134, + "learning_rate": 0.000294841493050952, + "loss": 0.0211, + "step": 2900 + }, + { + "epoch": 1.3651764705882352, + "grad_norm": 6.365715243648835e-05, + "learning_rate": 0.00029483008967784007, + "loss": 0.0, + "step": 2901 + }, + { + "epoch": 1.3656470588235294, + "grad_norm": 0.09442671293396322, + "learning_rate": 0.0002948186739355281, + "loss": 0.0498, + "step": 2902 + }, + { + "epoch": 1.3661176470588234, + "grad_norm": 0.06576992790564157, + "learning_rate": 0.00029480724582499107, + "loss": 0.0275, + "step": 2903 + }, + { + "epoch": 1.3665882352941177, + "grad_norm": 0.08365756381019616, + "learning_rate": 0.0002947958053472049, + "loss": 0.0467, + "step": 2904 + }, + { + "epoch": 1.3670588235294119, + "grad_norm": 0.050583074948135937, + "learning_rate": 0.0002947843525031468, + "loss": 0.0309, + "step": 2905 + }, + { + "epoch": 1.3675294117647059, + "grad_norm": 0.06302815314362267, + "learning_rate": 0.0002947728872937948, + "loss": 0.0192, + "step": 2906 + }, + { + "epoch": 1.3679999999999999, + "grad_norm": 0.06650793566788496, + "learning_rate": 0.00029476140972012815, + "loss": 0.0362, + "step": 2907 + }, + { + "epoch": 1.368470588235294, + "grad_norm": 7.726486149987301e-05, + "learning_rate": 0.000294749919783127, + "loss": 0.0, + "step": 2908 + }, + { + "epoch": 1.3689411764705883, + "grad_norm": 0.11091755784393549, + "learning_rate": 0.00029473841748377277, + "loss": 0.0787, + "step": 2909 + }, + { + "epoch": 1.3694117647058823, + "grad_norm": 0.03855036027732881, + "learning_rate": 0.0002947269028230477, + "loss": 0.0132, + "step": 2910 + }, + { + "epoch": 1.3698823529411766, + "grad_norm": 0.05535603076357395, + "learning_rate": 0.0002947153758019353, + "loss": 0.0175, + "step": 2911 + }, + { + "epoch": 1.3703529411764706, + "grad_norm": 5.431989750386153e-05, + "learning_rate": 0.00029470383642142, + "loss": 0.0, + "step": 2912 + }, + { + "epoch": 1.3708235294117648, + "grad_norm": 0.18353930246092354, + "learning_rate": 0.0002946922846824872, + "loss": 0.0986, + "step": 2913 + }, + { + "epoch": 1.3712941176470588, + "grad_norm": 0.13766844218439253, + "learning_rate": 0.00029468072058612367, + "loss": 0.116, + "step": 2914 + }, + { + "epoch": 1.371764705882353, + "grad_norm": 0.45980483066509936, + "learning_rate": 0.0002946691441333169, + "loss": 0.1542, + "step": 2915 + }, + { + "epoch": 1.372235294117647, + "grad_norm": 0.1326836668863154, + "learning_rate": 0.0002946575553250556, + "loss": 0.1197, + "step": 2916 + }, + { + "epoch": 1.3727058823529412, + "grad_norm": 0.07497782724660294, + "learning_rate": 0.00029464595416232956, + "loss": 0.0421, + "step": 2917 + }, + { + "epoch": 1.3731764705882352, + "grad_norm": 5.1508731137664973e-05, + "learning_rate": 0.0002946343406461296, + "loss": 0.0, + "step": 2918 + }, + { + "epoch": 1.3736470588235294, + "grad_norm": 0.05844366882600012, + "learning_rate": 0.0002946227147774475, + "loss": 0.0253, + "step": 2919 + }, + { + "epoch": 1.3741176470588234, + "grad_norm": 0.06943378501504373, + "learning_rate": 0.0002946110765572761, + "loss": 0.0256, + "step": 2920 + }, + { + "epoch": 1.3745882352941177, + "grad_norm": 0.1352609833586989, + "learning_rate": 0.00029459942598660953, + "loss": 0.0881, + "step": 2921 + }, + { + "epoch": 1.3750588235294119, + "grad_norm": 0.07772674236848637, + "learning_rate": 0.00029458776306644273, + "loss": 0.0488, + "step": 2922 + }, + { + "epoch": 1.375529411764706, + "grad_norm": 0.00027144757700131464, + "learning_rate": 0.0002945760877977717, + "loss": 0.0, + "step": 2923 + }, + { + "epoch": 1.376, + "grad_norm": 0.1361964630402768, + "learning_rate": 0.0002945644001815937, + "loss": 0.0525, + "step": 2924 + }, + { + "epoch": 1.3764705882352941, + "grad_norm": 0.11540026063549452, + "learning_rate": 0.0002945527002189068, + "loss": 0.091, + "step": 2925 + }, + { + "epoch": 1.3769411764705883, + "grad_norm": 0.5280245258630258, + "learning_rate": 0.0002945409879107103, + "loss": 0.4763, + "step": 2926 + }, + { + "epoch": 1.3774117647058823, + "grad_norm": 0.07386525501334686, + "learning_rate": 0.00029452926325800445, + "loss": 0.039, + "step": 2927 + }, + { + "epoch": 1.3778823529411763, + "grad_norm": 0.07156288319027534, + "learning_rate": 0.0002945175262617906, + "loss": 0.0339, + "step": 2928 + }, + { + "epoch": 1.3783529411764706, + "grad_norm": 0.07457193317327113, + "learning_rate": 0.00029450577692307115, + "loss": 0.0345, + "step": 2929 + }, + { + "epoch": 1.3788235294117648, + "grad_norm": 0.10444777245919057, + "learning_rate": 0.0002944940152428496, + "loss": 0.0532, + "step": 2930 + }, + { + "epoch": 1.3792941176470588, + "grad_norm": 0.10282942971013673, + "learning_rate": 0.0002944822412221303, + "loss": 0.0684, + "step": 2931 + }, + { + "epoch": 1.379764705882353, + "grad_norm": 0.0004595076641329231, + "learning_rate": 0.00029447045486191904, + "loss": 0.0, + "step": 2932 + }, + { + "epoch": 1.380235294117647, + "grad_norm": 0.0005735486343386189, + "learning_rate": 0.0002944586561632223, + "loss": 0.0, + "step": 2933 + }, + { + "epoch": 1.3807058823529412, + "grad_norm": 0.03989336604938802, + "learning_rate": 0.0002944468451270477, + "loss": 0.0154, + "step": 2934 + }, + { + "epoch": 1.3811764705882352, + "grad_norm": 0.18969858231776246, + "learning_rate": 0.000294435021754404, + "loss": 0.0771, + "step": 2935 + }, + { + "epoch": 1.3816470588235295, + "grad_norm": 0.1161482152045141, + "learning_rate": 0.000294423186046301, + "loss": 0.0833, + "step": 2936 + }, + { + "epoch": 1.3821176470588235, + "grad_norm": 0.05378506611326651, + "learning_rate": 0.00029441133800374957, + "loss": 0.026, + "step": 2937 + }, + { + "epoch": 1.3825882352941177, + "grad_norm": 0.0015797137064809466, + "learning_rate": 0.00029439947762776154, + "loss": 0.0, + "step": 2938 + }, + { + "epoch": 1.383058823529412, + "grad_norm": 0.06188840041220541, + "learning_rate": 0.00029438760491934987, + "loss": 0.026, + "step": 2939 + }, + { + "epoch": 1.383529411764706, + "grad_norm": 0.10433912512911392, + "learning_rate": 0.0002943757198795284, + "loss": 0.0229, + "step": 2940 + }, + { + "epoch": 1.384, + "grad_norm": 0.0020360134810942848, + "learning_rate": 0.0002943638225093124, + "loss": 0.0, + "step": 2941 + }, + { + "epoch": 1.3844705882352941, + "grad_norm": 3.990731563560176e-05, + "learning_rate": 0.00029435191280971783, + "loss": 0.0, + "step": 2942 + }, + { + "epoch": 1.3849411764705883, + "grad_norm": 0.07883329277719102, + "learning_rate": 0.0002943399907817619, + "loss": 0.0417, + "step": 2943 + }, + { + "epoch": 1.3854117647058823, + "grad_norm": 3.803691475696528e-05, + "learning_rate": 0.0002943280564264627, + "loss": 0.0, + "step": 2944 + }, + { + "epoch": 1.3858823529411763, + "grad_norm": 0.08118507902135562, + "learning_rate": 0.0002943161097448396, + "loss": 0.0142, + "step": 2945 + }, + { + "epoch": 1.3863529411764706, + "grad_norm": 3.029834158319891e-05, + "learning_rate": 0.0002943041507379128, + "loss": 0.0, + "step": 2946 + }, + { + "epoch": 1.3868235294117648, + "grad_norm": 0.030328017972995648, + "learning_rate": 0.00029429217940670383, + "loss": 0.0081, + "step": 2947 + }, + { + "epoch": 1.3872941176470588, + "grad_norm": 0.16033452603546036, + "learning_rate": 0.0002942801957522349, + "loss": 0.1024, + "step": 2948 + }, + { + "epoch": 1.387764705882353, + "grad_norm": 0.3478890175167038, + "learning_rate": 0.00029426819977552965, + "loss": 0.135, + "step": 2949 + }, + { + "epoch": 1.388235294117647, + "grad_norm": 0.04597883929939182, + "learning_rate": 0.0002942561914776124, + "loss": 0.0271, + "step": 2950 + }, + { + "epoch": 1.3887058823529412, + "grad_norm": 2.751572729912482e-05, + "learning_rate": 0.0002942441708595089, + "loss": 0.0, + "step": 2951 + }, + { + "epoch": 1.3891764705882352, + "grad_norm": 0.012800640332380703, + "learning_rate": 0.0002942321379222457, + "loss": 0.0011, + "step": 2952 + }, + { + "epoch": 1.3896470588235295, + "grad_norm": 0.020014074674942532, + "learning_rate": 0.0002942200926668504, + "loss": 0.0046, + "step": 2953 + }, + { + "epoch": 1.3901176470588235, + "grad_norm": 0.06772590251544586, + "learning_rate": 0.00029420803509435186, + "loss": 0.0213, + "step": 2954 + }, + { + "epoch": 1.3905882352941177, + "grad_norm": 0.0813624393078359, + "learning_rate": 0.00029419596520577975, + "loss": 0.05, + "step": 2955 + }, + { + "epoch": 1.391058823529412, + "grad_norm": 0.08353586673878573, + "learning_rate": 0.00029418388300216493, + "loss": 0.0401, + "step": 2956 + }, + { + "epoch": 1.391529411764706, + "grad_norm": 0.11161889757182887, + "learning_rate": 0.0002941717884845393, + "loss": 0.0646, + "step": 2957 + }, + { + "epoch": 1.392, + "grad_norm": 0.03641580622669307, + "learning_rate": 0.00029415968165393577, + "loss": 0.0131, + "step": 2958 + }, + { + "epoch": 1.3924705882352941, + "grad_norm": 0.05842416373914457, + "learning_rate": 0.00029414756251138835, + "loss": 0.0345, + "step": 2959 + }, + { + "epoch": 1.3929411764705883, + "grad_norm": 0.05975854765190929, + "learning_rate": 0.000294135431057932, + "loss": 0.0336, + "step": 2960 + }, + { + "epoch": 1.3934117647058823, + "grad_norm": 0.19493991035128713, + "learning_rate": 0.0002941232872946029, + "loss": 0.2181, + "step": 2961 + }, + { + "epoch": 1.3938823529411764, + "grad_norm": 0.044176303238132764, + "learning_rate": 0.00029411113122243814, + "loss": 0.0139, + "step": 2962 + }, + { + "epoch": 1.3943529411764706, + "grad_norm": 0.1310132359194158, + "learning_rate": 0.0002940989628424759, + "loss": 0.0676, + "step": 2963 + }, + { + "epoch": 1.3948235294117648, + "grad_norm": 0.08018853999444131, + "learning_rate": 0.0002940867821557554, + "loss": 0.0357, + "step": 2964 + }, + { + "epoch": 1.3952941176470588, + "grad_norm": 0.06263280286357144, + "learning_rate": 0.00029407458916331705, + "loss": 0.0303, + "step": 2965 + }, + { + "epoch": 1.395764705882353, + "grad_norm": 0.0654033888906878, + "learning_rate": 0.00029406238386620205, + "loss": 0.0413, + "step": 2966 + }, + { + "epoch": 1.396235294117647, + "grad_norm": 0.09684376292721769, + "learning_rate": 0.0002940501662654528, + "loss": 0.055, + "step": 2967 + }, + { + "epoch": 1.3967058823529412, + "grad_norm": 3.966631845684468e-05, + "learning_rate": 0.0002940379363621129, + "loss": 0.0, + "step": 2968 + }, + { + "epoch": 1.3971764705882352, + "grad_norm": 0.1159152856814401, + "learning_rate": 0.00029402569415722666, + "loss": 0.0616, + "step": 2969 + }, + { + "epoch": 1.3976470588235295, + "grad_norm": 0.10721359036109455, + "learning_rate": 0.00029401343965183965, + "loss": 0.0858, + "step": 2970 + }, + { + "epoch": 1.3981176470588235, + "grad_norm": 0.10407231056588527, + "learning_rate": 0.0002940011728469986, + "loss": 0.0585, + "step": 2971 + }, + { + "epoch": 1.3985882352941177, + "grad_norm": 3.1773714353136705e-05, + "learning_rate": 0.000293988893743751, + "loss": 0.0, + "step": 2972 + }, + { + "epoch": 1.3990588235294117, + "grad_norm": 0.1287311436929873, + "learning_rate": 0.0002939766023431457, + "loss": 0.0688, + "step": 2973 + }, + { + "epoch": 1.399529411764706, + "grad_norm": 4.443460538672517e-05, + "learning_rate": 0.0002939642986462323, + "loss": 0.0, + "step": 2974 + }, + { + "epoch": 1.4, + "grad_norm": 0.09212774174025683, + "learning_rate": 0.0002939519826540616, + "loss": 0.0619, + "step": 2975 + }, + { + "epoch": 1.4004705882352941, + "grad_norm": 0.04059583348675733, + "learning_rate": 0.0002939396543676856, + "loss": 0.0107, + "step": 2976 + }, + { + "epoch": 1.4009411764705884, + "grad_norm": 0.1263178237611921, + "learning_rate": 0.00029392731378815706, + "loss": 0.0864, + "step": 2977 + }, + { + "epoch": 1.4014117647058824, + "grad_norm": 0.041487908353856955, + "learning_rate": 0.00029391496091652993, + "loss": 0.0139, + "step": 2978 + }, + { + "epoch": 1.4018823529411764, + "grad_norm": 0.13244413441086222, + "learning_rate": 0.0002939025957538593, + "loss": 0.0731, + "step": 2979 + }, + { + "epoch": 1.4023529411764706, + "grad_norm": 0.05598485551115231, + "learning_rate": 0.0002938902183012011, + "loss": 0.0308, + "step": 2980 + }, + { + "epoch": 1.4028235294117648, + "grad_norm": 0.08712428750167689, + "learning_rate": 0.00029387782855961253, + "loss": 0.0371, + "step": 2981 + }, + { + "epoch": 1.4032941176470588, + "grad_norm": 0.13400518099711684, + "learning_rate": 0.00029386542653015167, + "loss": 0.1169, + "step": 2982 + }, + { + "epoch": 1.4037647058823528, + "grad_norm": 0.04043645372654276, + "learning_rate": 0.0002938530122138777, + "loss": 0.0159, + "step": 2983 + }, + { + "epoch": 1.404235294117647, + "grad_norm": 0.12920018779154452, + "learning_rate": 0.0002938405856118509, + "loss": 0.0438, + "step": 2984 + }, + { + "epoch": 1.4047058823529412, + "grad_norm": 0.13719911645623045, + "learning_rate": 0.0002938281467251326, + "loss": 0.0459, + "step": 2985 + }, + { + "epoch": 1.4051764705882352, + "grad_norm": 2.477977883424465e-05, + "learning_rate": 0.00029381569555478505, + "loss": 0.0, + "step": 2986 + }, + { + "epoch": 1.4056470588235295, + "grad_norm": 0.15407253838409335, + "learning_rate": 0.00029380323210187166, + "loss": 0.0759, + "step": 2987 + }, + { + "epoch": 1.4061176470588235, + "grad_norm": 0.08754400644480467, + "learning_rate": 0.00029379075636745694, + "loss": 0.0151, + "step": 2988 + }, + { + "epoch": 1.4065882352941177, + "grad_norm": 0.07384139095667458, + "learning_rate": 0.0002937782683526063, + "loss": 0.0389, + "step": 2989 + }, + { + "epoch": 1.4070588235294117, + "grad_norm": 0.17249424861939397, + "learning_rate": 0.0002937657680583864, + "loss": 0.0368, + "step": 2990 + }, + { + "epoch": 1.407529411764706, + "grad_norm": 0.03471125073773411, + "learning_rate": 0.0002937532554858647, + "loss": 0.0122, + "step": 2991 + }, + { + "epoch": 1.408, + "grad_norm": 0.04822463083632749, + "learning_rate": 0.0002937407306361099, + "loss": 0.0145, + "step": 2992 + }, + { + "epoch": 1.4084705882352941, + "grad_norm": 0.06979408293100459, + "learning_rate": 0.0002937281935101916, + "loss": 0.0357, + "step": 2993 + }, + { + "epoch": 1.4089411764705884, + "grad_norm": 4.685274660026446e-05, + "learning_rate": 0.0002937156441091806, + "loss": 0.0, + "step": 2994 + }, + { + "epoch": 1.4094117647058824, + "grad_norm": 0.07959729112383342, + "learning_rate": 0.0002937030824341487, + "loss": 0.0436, + "step": 2995 + }, + { + "epoch": 1.4098823529411764, + "grad_norm": 0.04293990891481864, + "learning_rate": 0.00029369050848616863, + "loss": 0.0098, + "step": 2996 + }, + { + "epoch": 1.4103529411764706, + "grad_norm": 0.13205192848484845, + "learning_rate": 0.0002936779222663144, + "loss": 0.0665, + "step": 2997 + }, + { + "epoch": 1.4108235294117648, + "grad_norm": 0.027875281604623284, + "learning_rate": 0.0002936653237756609, + "loss": 0.0057, + "step": 2998 + }, + { + "epoch": 1.4112941176470588, + "grad_norm": 0.09893634653998812, + "learning_rate": 0.00029365271301528406, + "loss": 0.0473, + "step": 2999 + }, + { + "epoch": 1.4117647058823528, + "grad_norm": 0.07943363996720984, + "learning_rate": 0.00029364008998626086, + "loss": 0.0351, + "step": 3000 + }, + { + "epoch": 1.412235294117647, + "grad_norm": 0.07314831749890216, + "learning_rate": 0.00029362745468966946, + "loss": 0.0393, + "step": 3001 + }, + { + "epoch": 1.4127058823529413, + "grad_norm": 0.053135582743435494, + "learning_rate": 0.0002936148071265889, + "loss": 0.0279, + "step": 3002 + }, + { + "epoch": 1.4131764705882353, + "grad_norm": 0.09716379348880566, + "learning_rate": 0.00029360214729809946, + "loss": 0.0477, + "step": 3003 + }, + { + "epoch": 1.4136470588235295, + "grad_norm": 0.13574476057959042, + "learning_rate": 0.00029358947520528227, + "loss": 0.0917, + "step": 3004 + }, + { + "epoch": 1.4141176470588235, + "grad_norm": 0.1872743695149637, + "learning_rate": 0.0002935767908492196, + "loss": 0.1494, + "step": 3005 + }, + { + "epoch": 1.4145882352941177, + "grad_norm": 0.13134320753711506, + "learning_rate": 0.0002935640942309947, + "loss": 0.1073, + "step": 3006 + }, + { + "epoch": 1.4150588235294117, + "grad_norm": 0.09678849456314316, + "learning_rate": 0.000293551385351692, + "loss": 0.0512, + "step": 3007 + }, + { + "epoch": 1.415529411764706, + "grad_norm": 0.03150964716211879, + "learning_rate": 0.0002935386642123968, + "loss": 0.0115, + "step": 3008 + }, + { + "epoch": 1.416, + "grad_norm": 0.05133775903765434, + "learning_rate": 0.00029352593081419573, + "loss": 0.0224, + "step": 3009 + }, + { + "epoch": 1.4164705882352941, + "grad_norm": 0.1320760458501456, + "learning_rate": 0.0002935131851581762, + "loss": 0.0905, + "step": 3010 + }, + { + "epoch": 1.4169411764705884, + "grad_norm": 0.10221602153857552, + "learning_rate": 0.00029350042724542657, + "loss": 0.0363, + "step": 3011 + }, + { + "epoch": 1.4174117647058824, + "grad_norm": 0.07341829552054473, + "learning_rate": 0.0002934876570770367, + "loss": 0.0453, + "step": 3012 + }, + { + "epoch": 1.4178823529411764, + "grad_norm": 0.08153679527577079, + "learning_rate": 0.00029347487465409714, + "loss": 0.0201, + "step": 3013 + }, + { + "epoch": 1.4183529411764706, + "grad_norm": 3.009815190243232e-05, + "learning_rate": 0.0002934620799776995, + "loss": 0.0, + "step": 3014 + }, + { + "epoch": 1.4188235294117648, + "grad_norm": 0.16810772936475024, + "learning_rate": 0.00029344927304893653, + "loss": 0.1082, + "step": 3015 + }, + { + "epoch": 1.4192941176470588, + "grad_norm": 0.1199744133778607, + "learning_rate": 0.00029343645386890204, + "loss": 0.0876, + "step": 3016 + }, + { + "epoch": 1.4197647058823528, + "grad_norm": 0.08291637521809171, + "learning_rate": 0.00029342362243869093, + "loss": 0.0347, + "step": 3017 + }, + { + "epoch": 1.420235294117647, + "grad_norm": 0.22200176484131975, + "learning_rate": 0.00029341077875939884, + "loss": 0.0985, + "step": 3018 + }, + { + "epoch": 1.4207058823529413, + "grad_norm": 0.10757012296663175, + "learning_rate": 0.0002933979228321229, + "loss": 0.0963, + "step": 3019 + }, + { + "epoch": 1.4211764705882353, + "grad_norm": 0.04858009970074415, + "learning_rate": 0.00029338505465796095, + "loss": 0.0189, + "step": 3020 + }, + { + "epoch": 1.4216470588235295, + "grad_norm": 0.01235199976811833, + "learning_rate": 0.00029337217423801206, + "loss": 0.0014, + "step": 3021 + }, + { + "epoch": 1.4221176470588235, + "grad_norm": 2.9106688443421413e-05, + "learning_rate": 0.00029335928157337625, + "loss": 0.0, + "step": 3022 + }, + { + "epoch": 1.4225882352941177, + "grad_norm": 0.06705345755276466, + "learning_rate": 0.00029334637666515457, + "loss": 0.033, + "step": 3023 + }, + { + "epoch": 1.4230588235294117, + "grad_norm": 0.17589344099925883, + "learning_rate": 0.00029333345951444923, + "loss": 0.1355, + "step": 3024 + }, + { + "epoch": 1.423529411764706, + "grad_norm": 0.10760918809175486, + "learning_rate": 0.00029332053012236345, + "loss": 0.0698, + "step": 3025 + }, + { + "epoch": 1.424, + "grad_norm": 2.4938601853104823e-05, + "learning_rate": 0.0002933075884900014, + "loss": 0.0, + "step": 3026 + }, + { + "epoch": 1.4244705882352942, + "grad_norm": 0.05586672390461264, + "learning_rate": 0.00029329463461846833, + "loss": 0.0326, + "step": 3027 + }, + { + "epoch": 1.4249411764705882, + "grad_norm": 0.05909206125678394, + "learning_rate": 0.00029328166850887066, + "loss": 0.0278, + "step": 3028 + }, + { + "epoch": 1.4254117647058824, + "grad_norm": 0.09875923239460074, + "learning_rate": 0.00029326869016231566, + "loss": 0.0443, + "step": 3029 + }, + { + "epoch": 1.4258823529411764, + "grad_norm": 0.10552876855195611, + "learning_rate": 0.0002932556995799118, + "loss": 0.0512, + "step": 3030 + }, + { + "epoch": 1.4263529411764706, + "grad_norm": 0.16145967134897307, + "learning_rate": 0.00029324269676276856, + "loss": 0.1005, + "step": 3031 + }, + { + "epoch": 1.4268235294117648, + "grad_norm": 0.03970687868085143, + "learning_rate": 0.0002932296817119964, + "loss": 0.0133, + "step": 3032 + }, + { + "epoch": 1.4272941176470588, + "grad_norm": 0.07439307099598091, + "learning_rate": 0.0002932166544287069, + "loss": 0.0331, + "step": 3033 + }, + { + "epoch": 1.4277647058823528, + "grad_norm": 2.1432264300272576e-05, + "learning_rate": 0.00029320361491401263, + "loss": 0.0, + "step": 3034 + }, + { + "epoch": 1.428235294117647, + "grad_norm": 0.06444261211966586, + "learning_rate": 0.00029319056316902723, + "loss": 0.0227, + "step": 3035 + }, + { + "epoch": 1.4287058823529413, + "grad_norm": 6.139587232155451e-05, + "learning_rate": 0.0002931774991948654, + "loss": 0.0, + "step": 3036 + }, + { + "epoch": 1.4291764705882353, + "grad_norm": 0.0344749590226846, + "learning_rate": 0.0002931644229926428, + "loss": 0.0095, + "step": 3037 + }, + { + "epoch": 1.4296470588235295, + "grad_norm": 0.15323549538063375, + "learning_rate": 0.00029315133456347633, + "loss": 0.0856, + "step": 3038 + }, + { + "epoch": 1.4301176470588235, + "grad_norm": 1.903931111995475e-05, + "learning_rate": 0.0002931382339084837, + "loss": 0.0, + "step": 3039 + }, + { + "epoch": 1.4305882352941177, + "grad_norm": 0.08455905496943684, + "learning_rate": 0.0002931251210287838, + "loss": 0.0443, + "step": 3040 + }, + { + "epoch": 1.4310588235294117, + "grad_norm": 0.12403215079278085, + "learning_rate": 0.00029311199592549656, + "loss": 0.0868, + "step": 3041 + }, + { + "epoch": 1.431529411764706, + "grad_norm": 0.03126608806709261, + "learning_rate": 0.00029309885859974294, + "loss": 0.007, + "step": 3042 + }, + { + "epoch": 1.432, + "grad_norm": 0.2097698425965339, + "learning_rate": 0.00029308570905264484, + "loss": 0.1566, + "step": 3043 + }, + { + "epoch": 1.4324705882352942, + "grad_norm": 0.39448839806244856, + "learning_rate": 0.00029307254728532536, + "loss": 0.3693, + "step": 3044 + }, + { + "epoch": 1.4329411764705882, + "grad_norm": 0.0826723616405963, + "learning_rate": 0.00029305937329890863, + "loss": 0.0485, + "step": 3045 + }, + { + "epoch": 1.4334117647058824, + "grad_norm": 0.14753574952467774, + "learning_rate": 0.00029304618709451966, + "loss": 0.1408, + "step": 3046 + }, + { + "epoch": 1.4338823529411764, + "grad_norm": 0.0729044967087333, + "learning_rate": 0.0002930329886732847, + "loss": 0.025, + "step": 3047 + }, + { + "epoch": 1.4343529411764706, + "grad_norm": 0.1174139446437571, + "learning_rate": 0.00029301977803633094, + "loss": 0.0591, + "step": 3048 + }, + { + "epoch": 1.4348235294117648, + "grad_norm": 0.12756453060510342, + "learning_rate": 0.0002930065551847866, + "loss": 0.0751, + "step": 3049 + }, + { + "epoch": 1.4352941176470588, + "grad_norm": 0.1291581586118145, + "learning_rate": 0.00029299332011978107, + "loss": 0.0545, + "step": 3050 + }, + { + "epoch": 1.4357647058823528, + "grad_norm": 0.00014847360798140674, + "learning_rate": 0.00029298007284244456, + "loss": 0.0, + "step": 3051 + }, + { + "epoch": 1.436235294117647, + "grad_norm": 0.055342599339805454, + "learning_rate": 0.0002929668133539085, + "loss": 0.0234, + "step": 3052 + }, + { + "epoch": 1.4367058823529413, + "grad_norm": 0.03663934841828154, + "learning_rate": 0.0002929535416553054, + "loss": 0.0098, + "step": 3053 + }, + { + "epoch": 1.4371764705882353, + "grad_norm": 0.0050239189457539106, + "learning_rate": 0.0002929402577477686, + "loss": 0.0001, + "step": 3054 + }, + { + "epoch": 1.4376470588235293, + "grad_norm": 0.19359928943131502, + "learning_rate": 0.0002929269616324327, + "loss": 0.0462, + "step": 3055 + }, + { + "epoch": 1.4381176470588235, + "grad_norm": 0.0020357155510757068, + "learning_rate": 0.0002929136533104332, + "loss": 0.0, + "step": 3056 + }, + { + "epoch": 1.4385882352941177, + "grad_norm": 0.20961687506998938, + "learning_rate": 0.0002929003327829067, + "loss": 0.1182, + "step": 3057 + }, + { + "epoch": 1.4390588235294117, + "grad_norm": 0.03023922195373599, + "learning_rate": 0.0002928870000509909, + "loss": 0.009, + "step": 3058 + }, + { + "epoch": 1.439529411764706, + "grad_norm": 0.12216931942816031, + "learning_rate": 0.0002928736551158244, + "loss": 0.0579, + "step": 3059 + }, + { + "epoch": 1.44, + "grad_norm": 0.040451163030972785, + "learning_rate": 0.00029286029797854704, + "loss": 0.0104, + "step": 3060 + }, + { + "epoch": 1.4404705882352942, + "grad_norm": 0.029874924542455276, + "learning_rate": 0.00029284692864029946, + "loss": 0.0081, + "step": 3061 + }, + { + "epoch": 1.4409411764705882, + "grad_norm": 0.034757217952983824, + "learning_rate": 0.0002928335471022235, + "loss": 0.0073, + "step": 3062 + }, + { + "epoch": 1.4414117647058824, + "grad_norm": 0.08834917150263637, + "learning_rate": 0.00029282015336546205, + "loss": 0.0407, + "step": 3063 + }, + { + "epoch": 1.4418823529411764, + "grad_norm": 0.11206373008333907, + "learning_rate": 0.00029280674743115896, + "loss": 0.0552, + "step": 3064 + }, + { + "epoch": 1.4423529411764706, + "grad_norm": 0.08028068384022442, + "learning_rate": 0.0002927933293004592, + "loss": 0.0399, + "step": 3065 + }, + { + "epoch": 1.4428235294117648, + "grad_norm": 0.1002839661100192, + "learning_rate": 0.00029277989897450867, + "loss": 0.0588, + "step": 3066 + }, + { + "epoch": 1.4432941176470588, + "grad_norm": 0.11895395973625666, + "learning_rate": 0.0002927664564544545, + "loss": 0.0925, + "step": 3067 + }, + { + "epoch": 1.4437647058823528, + "grad_norm": 0.11349169853782712, + "learning_rate": 0.00029275300174144464, + "loss": 0.0851, + "step": 3068 + }, + { + "epoch": 1.444235294117647, + "grad_norm": 0.06860869718168047, + "learning_rate": 0.0002927395348366282, + "loss": 0.0303, + "step": 3069 + }, + { + "epoch": 1.4447058823529413, + "grad_norm": 0.028934435585767568, + "learning_rate": 0.00029272605574115544, + "loss": 0.0055, + "step": 3070 + }, + { + "epoch": 1.4451764705882353, + "grad_norm": 0.12580767643008675, + "learning_rate": 0.00029271256445617737, + "loss": 0.064, + "step": 3071 + }, + { + "epoch": 1.4456470588235293, + "grad_norm": 0.00023390938838506312, + "learning_rate": 0.0002926990609828463, + "loss": 0.0, + "step": 3072 + }, + { + "epoch": 1.4461176470588235, + "grad_norm": 0.09887059152087035, + "learning_rate": 0.00029268554532231556, + "loss": 0.0616, + "step": 3073 + }, + { + "epoch": 1.4465882352941177, + "grad_norm": 0.06054579621912136, + "learning_rate": 0.00029267201747573934, + "loss": 0.0211, + "step": 3074 + }, + { + "epoch": 1.4470588235294117, + "grad_norm": 0.41421933486071727, + "learning_rate": 0.00029265847744427303, + "loss": 0.0461, + "step": 3075 + }, + { + "epoch": 1.447529411764706, + "grad_norm": 0.11325145160360074, + "learning_rate": 0.000292644925229073, + "loss": 0.0688, + "step": 3076 + }, + { + "epoch": 1.448, + "grad_norm": 0.1832944531753582, + "learning_rate": 0.00029263136083129666, + "loss": 0.0815, + "step": 3077 + }, + { + "epoch": 1.4484705882352942, + "grad_norm": 0.08225021730406391, + "learning_rate": 0.00029261778425210256, + "loss": 0.0498, + "step": 3078 + }, + { + "epoch": 1.4489411764705882, + "grad_norm": 0.13868922846319792, + "learning_rate": 0.0002926041954926501, + "loss": 0.0967, + "step": 3079 + }, + { + "epoch": 1.4494117647058824, + "grad_norm": 0.0724735035613423, + "learning_rate": 0.00029259059455409987, + "loss": 0.0593, + "step": 3080 + }, + { + "epoch": 1.4498823529411764, + "grad_norm": 0.052590275539449595, + "learning_rate": 0.00029257698143761354, + "loss": 0.0086, + "step": 3081 + }, + { + "epoch": 1.4503529411764706, + "grad_norm": 5.46420151859419e-05, + "learning_rate": 0.00029256335614435356, + "loss": 0.0, + "step": 3082 + }, + { + "epoch": 1.4508235294117646, + "grad_norm": 0.14068188664122014, + "learning_rate": 0.00029254971867548373, + "loss": 0.0632, + "step": 3083 + }, + { + "epoch": 1.4512941176470588, + "grad_norm": 5.344249217018871e-05, + "learning_rate": 0.00029253606903216874, + "loss": 0.0, + "step": 3084 + }, + { + "epoch": 1.4517647058823528, + "grad_norm": 0.18696566619987004, + "learning_rate": 0.00029252240721557433, + "loss": 0.0832, + "step": 3085 + }, + { + "epoch": 1.452235294117647, + "grad_norm": 0.06011957014360054, + "learning_rate": 0.00029250873322686727, + "loss": 0.0164, + "step": 3086 + }, + { + "epoch": 1.4527058823529413, + "grad_norm": 0.053866592264698546, + "learning_rate": 0.00029249504706721534, + "loss": 0.0283, + "step": 3087 + }, + { + "epoch": 1.4531764705882353, + "grad_norm": 0.04318803679738429, + "learning_rate": 0.00029248134873778754, + "loss": 0.0176, + "step": 3088 + }, + { + "epoch": 1.4536470588235293, + "grad_norm": 3.6120109546503345e-05, + "learning_rate": 0.0002924676382397536, + "loss": 0.0, + "step": 3089 + }, + { + "epoch": 1.4541176470588235, + "grad_norm": 0.07972434150602944, + "learning_rate": 0.0002924539155742846, + "loss": 0.0439, + "step": 3090 + }, + { + "epoch": 1.4545882352941177, + "grad_norm": 0.10530863950390928, + "learning_rate": 0.0002924401807425525, + "loss": 0.0614, + "step": 3091 + }, + { + "epoch": 1.4550588235294117, + "grad_norm": 0.10553966999030935, + "learning_rate": 0.0002924264337457303, + "loss": 0.0507, + "step": 3092 + }, + { + "epoch": 1.455529411764706, + "grad_norm": 0.09190180491629672, + "learning_rate": 0.000292412674584992, + "loss": 0.0216, + "step": 3093 + }, + { + "epoch": 1.456, + "grad_norm": 0.14704373715207408, + "learning_rate": 0.0002923989032615128, + "loss": 0.139, + "step": 3094 + }, + { + "epoch": 1.4564705882352942, + "grad_norm": 0.07190635503223167, + "learning_rate": 0.0002923851197764688, + "loss": 0.007, + "step": 3095 + }, + { + "epoch": 1.4569411764705882, + "grad_norm": 0.07949036962378883, + "learning_rate": 0.00029237132413103715, + "loss": 0.0552, + "step": 3096 + }, + { + "epoch": 1.4574117647058824, + "grad_norm": 0.03513311528812378, + "learning_rate": 0.00029235751632639617, + "loss": 0.014, + "step": 3097 + }, + { + "epoch": 1.4578823529411764, + "grad_norm": 0.09090314027999231, + "learning_rate": 0.00029234369636372496, + "loss": 0.063, + "step": 3098 + }, + { + "epoch": 1.4583529411764706, + "grad_norm": 0.04874029428318201, + "learning_rate": 0.0002923298642442039, + "loss": 0.0214, + "step": 3099 + }, + { + "epoch": 1.4588235294117646, + "grad_norm": 0.11779575041380629, + "learning_rate": 0.00029231601996901433, + "loss": 0.0817, + "step": 3100 + }, + { + "epoch": 1.4592941176470589, + "grad_norm": 0.00010070907264421587, + "learning_rate": 0.0002923021635393386, + "loss": 0.0, + "step": 3101 + }, + { + "epoch": 1.4597647058823529, + "grad_norm": 0.05073019764482686, + "learning_rate": 0.0002922882949563601, + "loss": 0.0227, + "step": 3102 + }, + { + "epoch": 1.460235294117647, + "grad_norm": 0.06640018160611111, + "learning_rate": 0.0002922744142212633, + "loss": 0.039, + "step": 3103 + }, + { + "epoch": 1.4607058823529413, + "grad_norm": 0.06327710121091734, + "learning_rate": 0.00029226052133523363, + "loss": 0.0292, + "step": 3104 + }, + { + "epoch": 1.4611764705882353, + "grad_norm": 0.16314539129158181, + "learning_rate": 0.0002922466162994577, + "loss": 0.1204, + "step": 3105 + }, + { + "epoch": 1.4616470588235293, + "grad_norm": 0.06584631386219045, + "learning_rate": 0.000292232699115123, + "loss": 0.0274, + "step": 3106 + }, + { + "epoch": 1.4621176470588235, + "grad_norm": 0.08879387314698112, + "learning_rate": 0.0002922187697834182, + "loss": 0.0709, + "step": 3107 + }, + { + "epoch": 1.4625882352941177, + "grad_norm": 0.05797619951243465, + "learning_rate": 0.00029220482830553284, + "loss": 0.0255, + "step": 3108 + }, + { + "epoch": 1.4630588235294117, + "grad_norm": 0.09385081512895789, + "learning_rate": 0.0002921908746826576, + "loss": 0.0841, + "step": 3109 + }, + { + "epoch": 1.4635294117647057, + "grad_norm": 0.035113326939818015, + "learning_rate": 0.0002921769089159843, + "loss": 0.0141, + "step": 3110 + }, + { + "epoch": 1.464, + "grad_norm": 0.10148268953454985, + "learning_rate": 0.00029216293100670554, + "loss": 0.0613, + "step": 3111 + }, + { + "epoch": 1.4644705882352942, + "grad_norm": 0.09895348414542418, + "learning_rate": 0.0002921489409560152, + "loss": 0.0777, + "step": 3112 + }, + { + "epoch": 1.4649411764705882, + "grad_norm": 0.00010681133581222649, + "learning_rate": 0.00029213493876510807, + "loss": 0.0, + "step": 3113 + }, + { + "epoch": 1.4654117647058824, + "grad_norm": 0.09836095030479476, + "learning_rate": 0.0002921209244351799, + "loss": 0.0676, + "step": 3114 + }, + { + "epoch": 1.4658823529411764, + "grad_norm": 0.0673769892796237, + "learning_rate": 0.00029210689796742783, + "loss": 0.0313, + "step": 3115 + }, + { + "epoch": 1.4663529411764706, + "grad_norm": 0.0912449438503712, + "learning_rate": 0.00029209285936304956, + "loss": 0.0521, + "step": 3116 + }, + { + "epoch": 1.4668235294117646, + "grad_norm": 0.08484784306681109, + "learning_rate": 0.0002920788086232442, + "loss": 0.0606, + "step": 3117 + }, + { + "epoch": 1.4672941176470589, + "grad_norm": 0.14182761808390107, + "learning_rate": 0.00029206474574921165, + "loss": 0.085, + "step": 3118 + }, + { + "epoch": 1.4677647058823529, + "grad_norm": 5.689779790520155e-05, + "learning_rate": 0.0002920506707421529, + "loss": 0.0, + "step": 3119 + }, + { + "epoch": 1.468235294117647, + "grad_norm": 0.04926912661090044, + "learning_rate": 0.0002920365836032703, + "loss": 0.0257, + "step": 3120 + }, + { + "epoch": 1.4687058823529413, + "grad_norm": 4.813600382969248e-05, + "learning_rate": 0.00029202248433376663, + "loss": 0.0, + "step": 3121 + }, + { + "epoch": 1.4691764705882353, + "grad_norm": 6.512196298239687e-05, + "learning_rate": 0.00029200837293484627, + "loss": 0.0, + "step": 3122 + }, + { + "epoch": 1.4696470588235293, + "grad_norm": 5.556931240200135e-05, + "learning_rate": 0.0002919942494077142, + "loss": 0.0, + "step": 3123 + }, + { + "epoch": 1.4701176470588235, + "grad_norm": 4.5357879144400324e-05, + "learning_rate": 0.0002919801137535769, + "loss": 0.0, + "step": 3124 + }, + { + "epoch": 1.4705882352941178, + "grad_norm": 0.04878333508519357, + "learning_rate": 0.0002919659659736414, + "loss": 0.0165, + "step": 3125 + }, + { + "epoch": 1.4710588235294118, + "grad_norm": 6.32095812262443e-05, + "learning_rate": 0.00029195180606911606, + "loss": 0.0, + "step": 3126 + }, + { + "epoch": 1.4715294117647058, + "grad_norm": 0.18297720664106562, + "learning_rate": 0.0002919376340412102, + "loss": 0.1049, + "step": 3127 + }, + { + "epoch": 1.472, + "grad_norm": 0.06737872333175245, + "learning_rate": 0.00029192344989113425, + "loss": 0.0354, + "step": 3128 + }, + { + "epoch": 1.4724705882352942, + "grad_norm": 0.09205695083813314, + "learning_rate": 0.0002919092536200995, + "loss": 0.0516, + "step": 3129 + }, + { + "epoch": 1.4729411764705882, + "grad_norm": 0.03428192148102358, + "learning_rate": 0.00029189504522931847, + "loss": 0.0113, + "step": 3130 + }, + { + "epoch": 1.4734117647058824, + "grad_norm": 0.18432270637874504, + "learning_rate": 0.00029188082472000453, + "loss": 0.1463, + "step": 3131 + }, + { + "epoch": 1.4738823529411764, + "grad_norm": 0.04404840529071025, + "learning_rate": 0.0002918665920933722, + "loss": 0.021, + "step": 3132 + }, + { + "epoch": 1.4743529411764706, + "grad_norm": 3.348666923668149e-05, + "learning_rate": 0.0002918523473506372, + "loss": 0.0, + "step": 3133 + }, + { + "epoch": 1.4748235294117646, + "grad_norm": 0.0002847943641672941, + "learning_rate": 0.00029183809049301585, + "loss": 0.0, + "step": 3134 + }, + { + "epoch": 1.4752941176470589, + "grad_norm": 0.13179454750515, + "learning_rate": 0.0002918238215217259, + "loss": 0.0827, + "step": 3135 + }, + { + "epoch": 1.4757647058823529, + "grad_norm": 0.07253774125940436, + "learning_rate": 0.00029180954043798595, + "loss": 0.0368, + "step": 3136 + }, + { + "epoch": 1.476235294117647, + "grad_norm": 0.04980098458527335, + "learning_rate": 0.00029179524724301563, + "loss": 0.017, + "step": 3137 + }, + { + "epoch": 1.4767058823529413, + "grad_norm": 0.05893201368833602, + "learning_rate": 0.00029178094193803576, + "loss": 0.0203, + "step": 3138 + }, + { + "epoch": 1.4771764705882353, + "grad_norm": 0.045258319987323194, + "learning_rate": 0.00029176662452426796, + "loss": 0.0317, + "step": 3139 + }, + { + "epoch": 1.4776470588235293, + "grad_norm": 0.1105645940703335, + "learning_rate": 0.0002917522950029351, + "loss": 0.09, + "step": 3140 + }, + { + "epoch": 1.4781176470588235, + "grad_norm": 2.4345684141055962e-05, + "learning_rate": 0.00029173795337526093, + "loss": 0.0, + "step": 3141 + }, + { + "epoch": 1.4785882352941178, + "grad_norm": 0.12837796313419036, + "learning_rate": 0.0002917235996424703, + "loss": 0.1172, + "step": 3142 + }, + { + "epoch": 1.4790588235294118, + "grad_norm": 0.03517779523938057, + "learning_rate": 0.0002917092338057892, + "loss": 0.0103, + "step": 3143 + }, + { + "epoch": 1.4795294117647058, + "grad_norm": 0.060117342022769395, + "learning_rate": 0.00029169485586644434, + "loss": 0.0229, + "step": 3144 + }, + { + "epoch": 1.48, + "grad_norm": 0.06865770578728315, + "learning_rate": 0.00029168046582566385, + "loss": 0.0363, + "step": 3145 + }, + { + "epoch": 1.4804705882352942, + "grad_norm": 0.090767602768168, + "learning_rate": 0.00029166606368467667, + "loss": 0.0644, + "step": 3146 + }, + { + "epoch": 1.4809411764705882, + "grad_norm": 0.1271439166955909, + "learning_rate": 0.0002916516494447127, + "loss": 0.1517, + "step": 3147 + }, + { + "epoch": 1.4814117647058824, + "grad_norm": 0.1327377463668867, + "learning_rate": 0.0002916372231070031, + "loss": 0.1037, + "step": 3148 + }, + { + "epoch": 1.4818823529411764, + "grad_norm": 0.053838641347208456, + "learning_rate": 0.00029162278467277994, + "loss": 0.0327, + "step": 3149 + }, + { + "epoch": 1.4823529411764707, + "grad_norm": 0.11543939361199351, + "learning_rate": 0.0002916083341432763, + "loss": 0.123, + "step": 3150 + }, + { + "epoch": 1.4828235294117647, + "grad_norm": 0.036873437800516586, + "learning_rate": 0.0002915938715197264, + "loss": 0.0083, + "step": 3151 + }, + { + "epoch": 1.4832941176470589, + "grad_norm": 0.08013891671935797, + "learning_rate": 0.00029157939680336527, + "loss": 0.058, + "step": 3152 + }, + { + "epoch": 1.4837647058823529, + "grad_norm": 1.7673754493796316e-05, + "learning_rate": 0.00029156490999542923, + "loss": 0.0, + "step": 3153 + }, + { + "epoch": 1.484235294117647, + "grad_norm": 0.14535261163966212, + "learning_rate": 0.0002915504110971555, + "loss": 0.1073, + "step": 3154 + }, + { + "epoch": 1.484705882352941, + "grad_norm": 0.08333084817986028, + "learning_rate": 0.0002915359001097824, + "loss": 0.0495, + "step": 3155 + }, + { + "epoch": 1.4851764705882353, + "grad_norm": 0.178275560601605, + "learning_rate": 0.0002915213770345492, + "loss": 0.0986, + "step": 3156 + }, + { + "epoch": 1.4856470588235293, + "grad_norm": 0.0559679306362114, + "learning_rate": 0.0002915068418726962, + "loss": 0.0246, + "step": 3157 + }, + { + "epoch": 1.4861176470588235, + "grad_norm": 0.09190989973775454, + "learning_rate": 0.00029149229462546483, + "loss": 0.0455, + "step": 3158 + }, + { + "epoch": 1.4865882352941178, + "grad_norm": 0.14056574806769692, + "learning_rate": 0.0002914777352940975, + "loss": 0.1006, + "step": 3159 + }, + { + "epoch": 1.4870588235294118, + "grad_norm": 0.09098474712969822, + "learning_rate": 0.0002914631638798376, + "loss": 0.0887, + "step": 3160 + }, + { + "epoch": 1.4875294117647058, + "grad_norm": 1.0093258427262757e-05, + "learning_rate": 0.00029144858038392964, + "loss": 0.0, + "step": 3161 + }, + { + "epoch": 1.488, + "grad_norm": 0.1109749375939031, + "learning_rate": 0.00029143398480761915, + "loss": 0.0623, + "step": 3162 + }, + { + "epoch": 1.4884705882352942, + "grad_norm": 0.030387161837105336, + "learning_rate": 0.00029141937715215256, + "loss": 0.0138, + "step": 3163 + }, + { + "epoch": 1.4889411764705882, + "grad_norm": 0.048991118510364015, + "learning_rate": 0.0002914047574187776, + "loss": 0.0328, + "step": 3164 + }, + { + "epoch": 1.4894117647058824, + "grad_norm": 0.04799719706428221, + "learning_rate": 0.0002913901256087427, + "loss": 0.0193, + "step": 3165 + }, + { + "epoch": 1.4898823529411764, + "grad_norm": 0.1009909020238574, + "learning_rate": 0.00029137548172329756, + "loss": 0.0641, + "step": 3166 + }, + { + "epoch": 1.4903529411764707, + "grad_norm": 0.07037857422268852, + "learning_rate": 0.0002913608257636929, + "loss": 0.027, + "step": 3167 + }, + { + "epoch": 1.4908235294117647, + "grad_norm": 0.027552385085636615, + "learning_rate": 0.0002913461577311803, + "loss": 0.0071, + "step": 3168 + }, + { + "epoch": 1.4912941176470589, + "grad_norm": 0.14028604093400643, + "learning_rate": 0.00029133147762701246, + "loss": 0.1049, + "step": 3169 + }, + { + "epoch": 1.4917647058823529, + "grad_norm": 0.04207470302803286, + "learning_rate": 0.00029131678545244333, + "loss": 0.0178, + "step": 3170 + }, + { + "epoch": 1.492235294117647, + "grad_norm": 0.12122489010972302, + "learning_rate": 0.00029130208120872747, + "loss": 0.0977, + "step": 3171 + }, + { + "epoch": 1.492705882352941, + "grad_norm": 0.06515967330738967, + "learning_rate": 0.00029128736489712087, + "loss": 0.0423, + "step": 3172 + }, + { + "epoch": 1.4931764705882353, + "grad_norm": 0.04625719228058767, + "learning_rate": 0.0002912726365188803, + "loss": 0.0163, + "step": 3173 + }, + { + "epoch": 1.4936470588235293, + "grad_norm": 0.04614659985894375, + "learning_rate": 0.0002912578960752636, + "loss": 0.0244, + "step": 3174 + }, + { + "epoch": 1.4941176470588236, + "grad_norm": 0.07095817127285238, + "learning_rate": 0.00029124314356752967, + "loss": 0.0326, + "step": 3175 + }, + { + "epoch": 1.4945882352941178, + "grad_norm": 0.08395768283215302, + "learning_rate": 0.00029122837899693847, + "loss": 0.0449, + "step": 3176 + }, + { + "epoch": 1.4950588235294118, + "grad_norm": 0.0944870018715626, + "learning_rate": 0.0002912136023647511, + "loss": 0.0374, + "step": 3177 + }, + { + "epoch": 1.4955294117647058, + "grad_norm": 0.024197253944405524, + "learning_rate": 0.0002911988136722293, + "loss": 0.0041, + "step": 3178 + }, + { + "epoch": 1.496, + "grad_norm": 1.0576979145542398e-05, + "learning_rate": 0.00029118401292063634, + "loss": 0.0, + "step": 3179 + }, + { + "epoch": 1.4964705882352942, + "grad_norm": 0.0783291503222867, + "learning_rate": 0.00029116920011123615, + "loss": 0.0441, + "step": 3180 + }, + { + "epoch": 1.4969411764705882, + "grad_norm": 0.012154707837419551, + "learning_rate": 0.00029115437524529383, + "loss": 0.0016, + "step": 3181 + }, + { + "epoch": 1.4974117647058822, + "grad_norm": 5.384626705268384e-06, + "learning_rate": 0.0002911395383240755, + "loss": 0.0, + "step": 3182 + }, + { + "epoch": 1.4978823529411764, + "grad_norm": 0.07590913843307492, + "learning_rate": 0.0002911246893488483, + "loss": 0.0416, + "step": 3183 + }, + { + "epoch": 1.4983529411764707, + "grad_norm": 0.08326067797199796, + "learning_rate": 0.0002911098283208805, + "loss": 0.0402, + "step": 3184 + }, + { + "epoch": 1.4988235294117647, + "grad_norm": 1.5293949710476833e-05, + "learning_rate": 0.0002910949552414411, + "loss": 0.0, + "step": 3185 + }, + { + "epoch": 1.4992941176470589, + "grad_norm": 0.06788213631466905, + "learning_rate": 0.0002910800701118005, + "loss": 0.0217, + "step": 3186 + }, + { + "epoch": 1.4997647058823529, + "grad_norm": 0.13133204621518774, + "learning_rate": 0.00029106517293323, + "loss": 0.0993, + "step": 3187 + }, + { + "epoch": 1.5002352941176471, + "grad_norm": 0.0603239582594229, + "learning_rate": 0.0002910502637070017, + "loss": 0.0326, + "step": 3188 + }, + { + "epoch": 1.5007058823529413, + "grad_norm": 0.05356253717128205, + "learning_rate": 0.0002910353424343891, + "loss": 0.0351, + "step": 3189 + }, + { + "epoch": 1.5011764705882353, + "grad_norm": 8.87905456273821e-06, + "learning_rate": 0.00029102040911666645, + "loss": 0.0, + "step": 3190 + }, + { + "epoch": 1.5016470588235293, + "grad_norm": 0.06907892582222594, + "learning_rate": 0.0002910054637551092, + "loss": 0.0457, + "step": 3191 + }, + { + "epoch": 1.5021176470588236, + "grad_norm": 0.09629290091583674, + "learning_rate": 0.0002909905063509936, + "loss": 0.0675, + "step": 3192 + }, + { + "epoch": 1.5025882352941178, + "grad_norm": 0.13207304547072782, + "learning_rate": 0.0002909755369055973, + "loss": 0.0714, + "step": 3193 + }, + { + "epoch": 1.5030588235294118, + "grad_norm": 0.09415501548271173, + "learning_rate": 0.00029096055542019865, + "loss": 0.0689, + "step": 3194 + }, + { + "epoch": 1.5035294117647058, + "grad_norm": 0.025753732192004383, + "learning_rate": 0.0002909455618960772, + "loss": 0.0058, + "step": 3195 + }, + { + "epoch": 1.504, + "grad_norm": 0.05626495196900941, + "learning_rate": 0.0002909305563345134, + "loss": 0.0274, + "step": 3196 + }, + { + "epoch": 1.5044705882352942, + "grad_norm": 9.237916959934452e-06, + "learning_rate": 0.0002909155387367888, + "loss": 0.0, + "step": 3197 + }, + { + "epoch": 1.5049411764705882, + "grad_norm": 0.0368170651190938, + "learning_rate": 0.0002909005091041861, + "loss": 0.0123, + "step": 3198 + }, + { + "epoch": 1.5054117647058822, + "grad_norm": 0.1496549082047674, + "learning_rate": 0.00029088546743798866, + "loss": 0.146, + "step": 3199 + }, + { + "epoch": 1.5058823529411764, + "grad_norm": 0.0695937895273681, + "learning_rate": 0.00029087041373948135, + "loss": 0.0334, + "step": 3200 + }, + { + "epoch": 1.5063529411764707, + "grad_norm": 0.057365109871207895, + "learning_rate": 0.00029085534800994976, + "loss": 0.0145, + "step": 3201 + }, + { + "epoch": 1.5068235294117647, + "grad_norm": 0.08085143865814266, + "learning_rate": 0.0002908402702506805, + "loss": 0.0138, + "step": 3202 + }, + { + "epoch": 1.5072941176470587, + "grad_norm": 0.13872398754191304, + "learning_rate": 0.0002908251804629614, + "loss": 0.0755, + "step": 3203 + }, + { + "epoch": 1.507764705882353, + "grad_norm": 9.44140751326166e-06, + "learning_rate": 0.0002908100786480811, + "loss": 0.0, + "step": 3204 + }, + { + "epoch": 1.5082352941176471, + "grad_norm": 0.08805008186146412, + "learning_rate": 0.0002907949648073294, + "loss": 0.0659, + "step": 3205 + }, + { + "epoch": 1.5087058823529413, + "grad_norm": 0.05258988970287802, + "learning_rate": 0.00029077983894199716, + "loss": 0.0201, + "step": 3206 + }, + { + "epoch": 1.5091764705882353, + "grad_norm": 0.105946704336445, + "learning_rate": 0.00029076470105337615, + "loss": 0.0706, + "step": 3207 + }, + { + "epoch": 1.5096470588235293, + "grad_norm": 0.08074573094021423, + "learning_rate": 0.0002907495511427592, + "loss": 0.0453, + "step": 3208 + }, + { + "epoch": 1.5101176470588236, + "grad_norm": 0.06334566273488397, + "learning_rate": 0.0002907343892114402, + "loss": 0.0355, + "step": 3209 + }, + { + "epoch": 1.5105882352941178, + "grad_norm": 0.10886799902887881, + "learning_rate": 0.0002907192152607141, + "loss": 0.0799, + "step": 3210 + }, + { + "epoch": 1.5110588235294118, + "grad_norm": 0.07145921515370093, + "learning_rate": 0.00029070402929187673, + "loss": 0.0347, + "step": 3211 + }, + { + "epoch": 1.5115294117647058, + "grad_norm": 0.2505789865368661, + "learning_rate": 0.0002906888313062252, + "loss": 0.1216, + "step": 3212 + }, + { + "epoch": 1.512, + "grad_norm": 0.05182381576587915, + "learning_rate": 0.00029067362130505734, + "loss": 0.0268, + "step": 3213 + }, + { + "epoch": 1.5124705882352942, + "grad_norm": 0.05647466479483765, + "learning_rate": 0.00029065839928967224, + "loss": 0.0261, + "step": 3214 + }, + { + "epoch": 1.5129411764705882, + "grad_norm": 0.04378138192588346, + "learning_rate": 0.0002906431652613699, + "loss": 0.0187, + "step": 3215 + }, + { + "epoch": 1.5134117647058822, + "grad_norm": 0.15236423511296165, + "learning_rate": 0.0002906279192214514, + "loss": 0.1237, + "step": 3216 + }, + { + "epoch": 1.5138823529411765, + "grad_norm": 0.03951029548414616, + "learning_rate": 0.0002906126611712188, + "loss": 0.0076, + "step": 3217 + }, + { + "epoch": 1.5143529411764707, + "grad_norm": 0.0706405722940916, + "learning_rate": 0.00029059739111197524, + "loss": 0.0479, + "step": 3218 + }, + { + "epoch": 1.5148235294117647, + "grad_norm": 0.04586839165624125, + "learning_rate": 0.0002905821090450249, + "loss": 0.0251, + "step": 3219 + }, + { + "epoch": 1.5152941176470587, + "grad_norm": 0.09755531351687177, + "learning_rate": 0.00029056681497167285, + "loss": 0.0795, + "step": 3220 + }, + { + "epoch": 1.515764705882353, + "grad_norm": 0.011576948867885134, + "learning_rate": 0.00029055150889322536, + "loss": 0.0014, + "step": 3221 + }, + { + "epoch": 1.5162352941176471, + "grad_norm": 1.8934896831572336e-05, + "learning_rate": 0.0002905361908109896, + "loss": 0.0, + "step": 3222 + }, + { + "epoch": 1.5167058823529411, + "grad_norm": 4.702984046395505e-06, + "learning_rate": 0.0002905208607262738, + "loss": 0.0, + "step": 3223 + }, + { + "epoch": 1.5171764705882353, + "grad_norm": 7.5040457360266845e-06, + "learning_rate": 0.0002905055186403873, + "loss": 0.0, + "step": 3224 + }, + { + "epoch": 1.5176470588235293, + "grad_norm": 5.451457433934652e-06, + "learning_rate": 0.0002904901645546403, + "loss": 0.0, + "step": 3225 + }, + { + "epoch": 1.5181176470588236, + "grad_norm": 0.09280266689365997, + "learning_rate": 0.0002904747984703441, + "loss": 0.0528, + "step": 3226 + }, + { + "epoch": 1.5185882352941178, + "grad_norm": 0.05197011628254229, + "learning_rate": 0.00029045942038881113, + "loss": 0.0267, + "step": 3227 + }, + { + "epoch": 1.5190588235294118, + "grad_norm": 0.14834685974340145, + "learning_rate": 0.00029044403031135466, + "loss": 0.107, + "step": 3228 + }, + { + "epoch": 1.5195294117647058, + "grad_norm": 0.06176564183255728, + "learning_rate": 0.0002904286282392892, + "loss": 0.0325, + "step": 3229 + }, + { + "epoch": 1.52, + "grad_norm": 0.27610766384346747, + "learning_rate": 0.00029041321417393006, + "loss": 0.0683, + "step": 3230 + }, + { + "epoch": 1.5204705882352942, + "grad_norm": 0.5404109071593799, + "learning_rate": 0.0002903977881165937, + "loss": 0.4181, + "step": 3231 + }, + { + "epoch": 1.5209411764705882, + "grad_norm": 3.2578665176960876e-06, + "learning_rate": 0.0002903823500685976, + "loss": 0.0, + "step": 3232 + }, + { + "epoch": 1.5214117647058822, + "grad_norm": 0.05035557554186721, + "learning_rate": 0.0002903669000312602, + "loss": 0.0209, + "step": 3233 + }, + { + "epoch": 1.5218823529411765, + "grad_norm": 5.859983031992676e-06, + "learning_rate": 0.0002903514380059011, + "loss": 0.0, + "step": 3234 + }, + { + "epoch": 1.5223529411764707, + "grad_norm": 0.00011343890376777882, + "learning_rate": 0.0002903359639938407, + "loss": 0.0, + "step": 3235 + }, + { + "epoch": 1.5228235294117647, + "grad_norm": 0.04414529637439559, + "learning_rate": 0.0002903204779964007, + "loss": 0.0153, + "step": 3236 + }, + { + "epoch": 1.5232941176470587, + "grad_norm": 0.10940723702177434, + "learning_rate": 0.00029030498001490356, + "loss": 0.0618, + "step": 3237 + }, + { + "epoch": 1.523764705882353, + "grad_norm": 0.08098370034813686, + "learning_rate": 0.00029028947005067297, + "loss": 0.0449, + "step": 3238 + }, + { + "epoch": 1.5242352941176471, + "grad_norm": 0.0015024063480780219, + "learning_rate": 0.00029027394810503346, + "loss": 0.0, + "step": 3239 + }, + { + "epoch": 1.5247058823529411, + "grad_norm": 0.20767788002278859, + "learning_rate": 0.0002902584141793108, + "loss": 0.1505, + "step": 3240 + }, + { + "epoch": 1.5251764705882351, + "grad_norm": 0.102777710552776, + "learning_rate": 0.0002902428682748316, + "loss": 0.0958, + "step": 3241 + }, + { + "epoch": 1.5256470588235294, + "grad_norm": 0.10375913754068944, + "learning_rate": 0.0002902273103929236, + "loss": 0.0666, + "step": 3242 + }, + { + "epoch": 1.5261176470588236, + "grad_norm": 0.0606689925624748, + "learning_rate": 0.00029021174053491547, + "loss": 0.041, + "step": 3243 + }, + { + "epoch": 1.5265882352941178, + "grad_norm": 0.0757934875321023, + "learning_rate": 0.0002901961587021369, + "loss": 0.0555, + "step": 3244 + }, + { + "epoch": 1.5270588235294118, + "grad_norm": 0.07140125749705574, + "learning_rate": 0.00029018056489591874, + "loss": 0.0617, + "step": 3245 + }, + { + "epoch": 1.5275294117647058, + "grad_norm": 0.05276731355419397, + "learning_rate": 0.0002901649591175928, + "loss": 0.022, + "step": 3246 + }, + { + "epoch": 1.528, + "grad_norm": 0.037662972671625466, + "learning_rate": 0.0002901493413684918, + "loss": 0.0153, + "step": 3247 + }, + { + "epoch": 1.5284705882352942, + "grad_norm": 0.00011405811512343652, + "learning_rate": 0.0002901337116499496, + "loss": 0.0, + "step": 3248 + }, + { + "epoch": 1.5289411764705882, + "grad_norm": 0.07969635900470953, + "learning_rate": 0.0002901180699633011, + "loss": 0.027, + "step": 3249 + }, + { + "epoch": 1.5294117647058822, + "grad_norm": 0.11231171734014703, + "learning_rate": 0.00029010241630988217, + "loss": 0.0346, + "step": 3250 + }, + { + "epoch": 1.5298823529411765, + "grad_norm": 1.185893769046897e-05, + "learning_rate": 0.0002900867506910297, + "loss": 0.0, + "step": 3251 + }, + { + "epoch": 1.5303529411764707, + "grad_norm": 0.07221305474335067, + "learning_rate": 0.00029007107310808153, + "loss": 0.0266, + "step": 3252 + }, + { + "epoch": 1.5308235294117647, + "grad_norm": 0.07070088933077914, + "learning_rate": 0.00029005538356237674, + "loss": 0.0296, + "step": 3253 + }, + { + "epoch": 1.5312941176470587, + "grad_norm": 0.21229847268179655, + "learning_rate": 0.00029003968205525523, + "loss": 0.1456, + "step": 3254 + }, + { + "epoch": 1.531764705882353, + "grad_norm": 0.1117402164636735, + "learning_rate": 0.00029002396858805794, + "loss": 0.0633, + "step": 3255 + }, + { + "epoch": 1.5322352941176471, + "grad_norm": 0.027972501151289735, + "learning_rate": 0.0002900082431621269, + "loss": 0.0066, + "step": 3256 + }, + { + "epoch": 1.5327058823529411, + "grad_norm": 0.09337433289120266, + "learning_rate": 0.0002899925057788052, + "loss": 0.0611, + "step": 3257 + }, + { + "epoch": 1.5331764705882351, + "grad_norm": 0.11004222269503945, + "learning_rate": 0.00028997675643943686, + "loss": 0.0512, + "step": 3258 + }, + { + "epoch": 1.5336470588235294, + "grad_norm": 4.7794678839911245e-06, + "learning_rate": 0.0002899609951453669, + "loss": 0.0, + "step": 3259 + }, + { + "epoch": 1.5341176470588236, + "grad_norm": 4.056450956493037e-06, + "learning_rate": 0.0002899452218979415, + "loss": 0.0, + "step": 3260 + }, + { + "epoch": 1.5345882352941178, + "grad_norm": 0.028244159044091482, + "learning_rate": 0.0002899294366985076, + "loss": 0.005, + "step": 3261 + }, + { + "epoch": 1.5350588235294118, + "grad_norm": 0.31775259631345537, + "learning_rate": 0.0002899136395484136, + "loss": 0.2346, + "step": 3262 + }, + { + "epoch": 1.5355294117647058, + "grad_norm": 0.06602133053084036, + "learning_rate": 0.00028989783044900845, + "loss": 0.0137, + "step": 3263 + }, + { + "epoch": 1.536, + "grad_norm": 0.11714244590560971, + "learning_rate": 0.0002898820094016424, + "loss": 0.0369, + "step": 3264 + }, + { + "epoch": 1.5364705882352943, + "grad_norm": 7.929336652220581e-06, + "learning_rate": 0.0002898661764076666, + "loss": 0.0, + "step": 3265 + }, + { + "epoch": 1.5369411764705883, + "grad_norm": 0.11353302417630895, + "learning_rate": 0.00028985033146843335, + "loss": 0.0585, + "step": 3266 + }, + { + "epoch": 1.5374117647058823, + "grad_norm": 3.1192886914862457e-06, + "learning_rate": 0.00028983447458529575, + "loss": 0.0, + "step": 3267 + }, + { + "epoch": 1.5378823529411765, + "grad_norm": 0.029875444864687205, + "learning_rate": 0.0002898186057596082, + "loss": 0.0081, + "step": 3268 + }, + { + "epoch": 1.5383529411764707, + "grad_norm": 0.054453451281284156, + "learning_rate": 0.0002898027249927259, + "loss": 0.0417, + "step": 3269 + }, + { + "epoch": 1.5388235294117647, + "grad_norm": 0.08817598113848743, + "learning_rate": 0.00028978683228600514, + "loss": 0.0485, + "step": 3270 + }, + { + "epoch": 1.5392941176470587, + "grad_norm": 0.07977108767563956, + "learning_rate": 0.0002897709276408032, + "loss": 0.0464, + "step": 3271 + }, + { + "epoch": 1.539764705882353, + "grad_norm": 4.655663645552174e-06, + "learning_rate": 0.00028975501105847856, + "loss": 0.0, + "step": 3272 + }, + { + "epoch": 1.5402352941176471, + "grad_norm": 0.10145829153312493, + "learning_rate": 0.00028973908254039044, + "loss": 0.0413, + "step": 3273 + }, + { + "epoch": 1.5407058823529411, + "grad_norm": 0.16847630423617432, + "learning_rate": 0.00028972314208789926, + "loss": 0.1493, + "step": 3274 + }, + { + "epoch": 1.5411764705882351, + "grad_norm": 0.0789150855800341, + "learning_rate": 0.00028970718970236644, + "loss": 0.0457, + "step": 3275 + }, + { + "epoch": 1.5416470588235294, + "grad_norm": 0.01945296292248251, + "learning_rate": 0.00028969122538515435, + "loss": 0.0057, + "step": 3276 + }, + { + "epoch": 1.5421176470588236, + "grad_norm": 3.6220449601795865e-06, + "learning_rate": 0.00028967524913762636, + "loss": 0.0, + "step": 3277 + }, + { + "epoch": 1.5425882352941176, + "grad_norm": 0.03253284606085496, + "learning_rate": 0.00028965926096114706, + "loss": 0.0097, + "step": 3278 + }, + { + "epoch": 1.5430588235294118, + "grad_norm": 0.166603165753775, + "learning_rate": 0.00028964326085708183, + "loss": 0.113, + "step": 3279 + }, + { + "epoch": 1.5435294117647058, + "grad_norm": 0.08503756003568387, + "learning_rate": 0.00028962724882679715, + "loss": 0.0477, + "step": 3280 + }, + { + "epoch": 1.544, + "grad_norm": 0.04968306515715102, + "learning_rate": 0.00028961122487166063, + "loss": 0.0222, + "step": 3281 + }, + { + "epoch": 1.5444705882352943, + "grad_norm": 0.12946329024337247, + "learning_rate": 0.00028959518899304064, + "loss": 0.1164, + "step": 3282 + }, + { + "epoch": 1.5449411764705883, + "grad_norm": 7.942038093340566e-06, + "learning_rate": 0.00028957914119230686, + "loss": 0.0, + "step": 3283 + }, + { + "epoch": 1.5454117647058823, + "grad_norm": 3.3924284377173734e-06, + "learning_rate": 0.0002895630814708297, + "loss": 0.0, + "step": 3284 + }, + { + "epoch": 1.5458823529411765, + "grad_norm": 0.0625939496680716, + "learning_rate": 0.00028954700982998093, + "loss": 0.0272, + "step": 3285 + }, + { + "epoch": 1.5463529411764707, + "grad_norm": 0.01233978903147532, + "learning_rate": 0.000289530926271133, + "loss": 0.0012, + "step": 3286 + }, + { + "epoch": 1.5468235294117647, + "grad_norm": 0.09215174375360143, + "learning_rate": 0.0002895148307956596, + "loss": 0.0595, + "step": 3287 + }, + { + "epoch": 1.5472941176470587, + "grad_norm": 0.13701370539696003, + "learning_rate": 0.00028949872340493525, + "loss": 0.0799, + "step": 3288 + }, + { + "epoch": 1.547764705882353, + "grad_norm": 0.09733029954886456, + "learning_rate": 0.00028948260410033576, + "loss": 0.048, + "step": 3289 + }, + { + "epoch": 1.5482352941176472, + "grad_norm": 0.13544347968308224, + "learning_rate": 0.00028946647288323766, + "loss": 0.1473, + "step": 3290 + }, + { + "epoch": 1.5487058823529412, + "grad_norm": 0.22361074711823486, + "learning_rate": 0.00028945032975501877, + "loss": 0.2057, + "step": 3291 + }, + { + "epoch": 1.5491764705882352, + "grad_norm": 0.08859067289913304, + "learning_rate": 0.0002894341747170577, + "loss": 0.0711, + "step": 3292 + }, + { + "epoch": 1.5496470588235294, + "grad_norm": 0.0910857179417233, + "learning_rate": 0.00028941800777073414, + "loss": 0.0736, + "step": 3293 + }, + { + "epoch": 1.5501176470588236, + "grad_norm": 0.07232919066908126, + "learning_rate": 0.0002894018289174289, + "loss": 0.0396, + "step": 3294 + }, + { + "epoch": 1.5505882352941176, + "grad_norm": 0.06206719527563368, + "learning_rate": 0.0002893856381585237, + "loss": 0.0311, + "step": 3295 + }, + { + "epoch": 1.5510588235294116, + "grad_norm": 0.1302564989310755, + "learning_rate": 0.0002893694354954014, + "loss": 0.094, + "step": 3296 + }, + { + "epoch": 1.5515294117647058, + "grad_norm": 0.14893497342423814, + "learning_rate": 0.00028935322092944564, + "loss": 0.1025, + "step": 3297 + }, + { + "epoch": 1.552, + "grad_norm": 0.07392098237904708, + "learning_rate": 0.00028933699446204126, + "loss": 0.0306, + "step": 3298 + }, + { + "epoch": 1.5524705882352943, + "grad_norm": 4.2640583970864975e-06, + "learning_rate": 0.0002893207560945742, + "loss": 0.0, + "step": 3299 + }, + { + "epoch": 1.5529411764705883, + "grad_norm": 4.515068019661536e-06, + "learning_rate": 0.0002893045058284311, + "loss": 0.0, + "step": 3300 + }, + { + "epoch": 1.5534117647058823, + "grad_norm": 0.11011325969856645, + "learning_rate": 0.000289288243665, + "loss": 0.0916, + "step": 3301 + }, + { + "epoch": 1.5538823529411765, + "grad_norm": 0.060939827123397204, + "learning_rate": 0.00028927196960566964, + "loss": 0.0254, + "step": 3302 + }, + { + "epoch": 1.5543529411764707, + "grad_norm": 0.07675970563154938, + "learning_rate": 0.00028925568365182995, + "loss": 0.0672, + "step": 3303 + }, + { + "epoch": 1.5548235294117647, + "grad_norm": 0.04618783013211161, + "learning_rate": 0.00028923938580487183, + "loss": 0.0232, + "step": 3304 + }, + { + "epoch": 1.5552941176470587, + "grad_norm": 5.49237343956562e-06, + "learning_rate": 0.0002892230760661872, + "loss": 0.0, + "step": 3305 + }, + { + "epoch": 1.555764705882353, + "grad_norm": 0.06978946698446854, + "learning_rate": 0.000289206754437169, + "loss": 0.0261, + "step": 3306 + }, + { + "epoch": 1.5562352941176472, + "grad_norm": 0.09063966750582116, + "learning_rate": 0.0002891904209192112, + "loss": 0.0406, + "step": 3307 + }, + { + "epoch": 1.5567058823529412, + "grad_norm": 0.10056793338226844, + "learning_rate": 0.00028917407551370863, + "loss": 0.0715, + "step": 3308 + }, + { + "epoch": 1.5571764705882352, + "grad_norm": 0.1952261414568226, + "learning_rate": 0.00028915771822205747, + "loss": 0.1134, + "step": 3309 + }, + { + "epoch": 1.5576470588235294, + "grad_norm": 0.2976327791076222, + "learning_rate": 0.00028914134904565455, + "loss": 0.0326, + "step": 3310 + }, + { + "epoch": 1.5581176470588236, + "grad_norm": 0.02991421378636657, + "learning_rate": 0.00028912496798589796, + "loss": 0.0097, + "step": 3311 + }, + { + "epoch": 1.5585882352941176, + "grad_norm": 0.05304010188230672, + "learning_rate": 0.0002891085750441867, + "loss": 0.0232, + "step": 3312 + }, + { + "epoch": 1.5590588235294116, + "grad_norm": 0.08178555860713525, + "learning_rate": 0.0002890921702219208, + "loss": 0.0352, + "step": 3313 + }, + { + "epoch": 1.5595294117647058, + "grad_norm": 0.029061852666395837, + "learning_rate": 0.00028907575352050133, + "loss": 0.0103, + "step": 3314 + }, + { + "epoch": 1.56, + "grad_norm": 0.09204956109009645, + "learning_rate": 0.00028905932494133034, + "loss": 0.0322, + "step": 3315 + }, + { + "epoch": 1.5604705882352943, + "grad_norm": 0.12006133394777961, + "learning_rate": 0.0002890428844858109, + "loss": 0.0359, + "step": 3316 + }, + { + "epoch": 1.5609411764705883, + "grad_norm": 0.17309649445015915, + "learning_rate": 0.0002890264321553471, + "loss": 0.0921, + "step": 3317 + }, + { + "epoch": 1.5614117647058823, + "grad_norm": 0.056415881037300636, + "learning_rate": 0.00028900996795134414, + "loss": 0.0303, + "step": 3318 + }, + { + "epoch": 1.5618823529411765, + "grad_norm": 0.12066719381446195, + "learning_rate": 0.000288993491875208, + "loss": 0.12, + "step": 3319 + }, + { + "epoch": 1.5623529411764707, + "grad_norm": 0.06992772390360544, + "learning_rate": 0.000288977003928346, + "loss": 0.0399, + "step": 3320 + }, + { + "epoch": 1.5628235294117647, + "grad_norm": 0.10212551347040187, + "learning_rate": 0.0002889605041121661, + "loss": 0.0734, + "step": 3321 + }, + { + "epoch": 1.5632941176470587, + "grad_norm": 0.07411750738098229, + "learning_rate": 0.0002889439924280776, + "loss": 0.0413, + "step": 3322 + }, + { + "epoch": 1.563764705882353, + "grad_norm": 0.10900933803038172, + "learning_rate": 0.0002889274688774906, + "loss": 0.1096, + "step": 3323 + }, + { + "epoch": 1.5642352941176472, + "grad_norm": 0.07205996416053916, + "learning_rate": 0.00028891093346181633, + "loss": 0.0536, + "step": 3324 + }, + { + "epoch": 1.5647058823529412, + "grad_norm": 0.07368520609241838, + "learning_rate": 0.000288894386182467, + "loss": 0.0556, + "step": 3325 + }, + { + "epoch": 1.5651764705882352, + "grad_norm": 0.17685249755136045, + "learning_rate": 0.0002888778270408558, + "loss": 0.1274, + "step": 3326 + }, + { + "epoch": 1.5656470588235294, + "grad_norm": 0.07163976782804139, + "learning_rate": 0.000288861256038397, + "loss": 0.0385, + "step": 3327 + }, + { + "epoch": 1.5661176470588236, + "grad_norm": 4.455548842643799e-06, + "learning_rate": 0.0002888446731765058, + "loss": 0.0, + "step": 3328 + }, + { + "epoch": 1.5665882352941176, + "grad_norm": 0.012798188717372136, + "learning_rate": 0.00028882807845659846, + "loss": 0.0022, + "step": 3329 + }, + { + "epoch": 1.5670588235294116, + "grad_norm": 0.1824696278854131, + "learning_rate": 0.00028881147188009233, + "loss": 0.1333, + "step": 3330 + }, + { + "epoch": 1.5675294117647058, + "grad_norm": 0.10421278452652537, + "learning_rate": 0.00028879485344840563, + "loss": 0.0353, + "step": 3331 + }, + { + "epoch": 1.568, + "grad_norm": 3.2208954245474042e-06, + "learning_rate": 0.00028877822316295763, + "loss": 0.0, + "step": 3332 + }, + { + "epoch": 1.5684705882352943, + "grad_norm": 0.12383527884251966, + "learning_rate": 0.0002887615810251687, + "loss": 0.1174, + "step": 3333 + }, + { + "epoch": 1.5689411764705883, + "grad_norm": 0.06869775073455593, + "learning_rate": 0.0002887449270364601, + "loss": 0.0472, + "step": 3334 + }, + { + "epoch": 1.5694117647058823, + "grad_norm": 5.923240848372893e-06, + "learning_rate": 0.0002887282611982542, + "loss": 0.0, + "step": 3335 + }, + { + "epoch": 1.5698823529411765, + "grad_norm": 4.267674808101057e-06, + "learning_rate": 0.0002887115835119743, + "loss": 0.0, + "step": 3336 + }, + { + "epoch": 1.5703529411764707, + "grad_norm": 0.016143507195356545, + "learning_rate": 0.00028869489397904486, + "loss": 0.0036, + "step": 3337 + }, + { + "epoch": 1.5708235294117647, + "grad_norm": 4.855677895864225e-06, + "learning_rate": 0.00028867819260089113, + "loss": 0.0, + "step": 3338 + }, + { + "epoch": 1.5712941176470587, + "grad_norm": 0.08718676202954061, + "learning_rate": 0.0002886614793789396, + "loss": 0.0645, + "step": 3339 + }, + { + "epoch": 1.571764705882353, + "grad_norm": 0.0011523066025037518, + "learning_rate": 0.0002886447543146175, + "loss": 0.0, + "step": 3340 + }, + { + "epoch": 1.5722352941176472, + "grad_norm": 0.06914545619003577, + "learning_rate": 0.0002886280174093534, + "loss": 0.0234, + "step": 3341 + }, + { + "epoch": 1.5727058823529412, + "grad_norm": 0.05455414481436048, + "learning_rate": 0.0002886112686645766, + "loss": 0.0162, + "step": 3342 + }, + { + "epoch": 1.5731764705882352, + "grad_norm": 6.1405166711352355e-06, + "learning_rate": 0.00028859450808171763, + "loss": 0.0, + "step": 3343 + }, + { + "epoch": 1.5736470588235294, + "grad_norm": 7.558652044298385e-06, + "learning_rate": 0.00028857773566220784, + "loss": 0.0, + "step": 3344 + }, + { + "epoch": 1.5741176470588236, + "grad_norm": 1.1601113226102117e-05, + "learning_rate": 0.0002885609514074797, + "loss": 0.0, + "step": 3345 + }, + { + "epoch": 1.5745882352941176, + "grad_norm": 0.0824923528910224, + "learning_rate": 0.0002885441553189667, + "loss": 0.0429, + "step": 3346 + }, + { + "epoch": 1.5750588235294116, + "grad_norm": 0.047038556634348784, + "learning_rate": 0.00028852734739810324, + "loss": 0.0203, + "step": 3347 + }, + { + "epoch": 1.5755294117647058, + "grad_norm": 1.0380707161485283e-05, + "learning_rate": 0.00028851052764632487, + "loss": 0.0, + "step": 3348 + }, + { + "epoch": 1.576, + "grad_norm": 0.10901068991975844, + "learning_rate": 0.00028849369606506807, + "loss": 0.1043, + "step": 3349 + }, + { + "epoch": 1.576470588235294, + "grad_norm": 0.06969751275106502, + "learning_rate": 0.0002884768526557703, + "loss": 0.0501, + "step": 3350 + }, + { + "epoch": 1.5769411764705883, + "grad_norm": 0.08113320623448303, + "learning_rate": 0.00028845999741987003, + "loss": 0.0592, + "step": 3351 + }, + { + "epoch": 1.5774117647058823, + "grad_norm": 0.04083180370164303, + "learning_rate": 0.0002884431303588069, + "loss": 0.0123, + "step": 3352 + }, + { + "epoch": 1.5778823529411765, + "grad_norm": 0.08352150436626113, + "learning_rate": 0.00028842625147402134, + "loss": 0.045, + "step": 3353 + }, + { + "epoch": 1.5783529411764707, + "grad_norm": 1.801836654039006e-05, + "learning_rate": 0.00028840936076695493, + "loss": 0.0, + "step": 3354 + }, + { + "epoch": 1.5788235294117647, + "grad_norm": 0.08008473212871936, + "learning_rate": 0.0002883924582390503, + "loss": 0.0356, + "step": 3355 + }, + { + "epoch": 1.5792941176470587, + "grad_norm": 0.0741289947474901, + "learning_rate": 0.0002883755438917508, + "loss": 0.055, + "step": 3356 + }, + { + "epoch": 1.579764705882353, + "grad_norm": 4.319993786124486e-06, + "learning_rate": 0.0002883586177265012, + "loss": 0.0, + "step": 3357 + }, + { + "epoch": 1.5802352941176472, + "grad_norm": 4.284033560039383e-06, + "learning_rate": 0.000288341679744747, + "loss": 0.0, + "step": 3358 + }, + { + "epoch": 1.5807058823529412, + "grad_norm": 0.03361543882166606, + "learning_rate": 0.0002883247299479348, + "loss": 0.0124, + "step": 3359 + }, + { + "epoch": 1.5811764705882352, + "grad_norm": 0.03320359468025521, + "learning_rate": 0.0002883077683375121, + "loss": 0.0109, + "step": 3360 + }, + { + "epoch": 1.5816470588235294, + "grad_norm": 0.051487258712997974, + "learning_rate": 0.0002882907949149277, + "loss": 0.0289, + "step": 3361 + }, + { + "epoch": 1.5821176470588236, + "grad_norm": 0.061372704522499086, + "learning_rate": 0.000288273809681631, + "loss": 0.0353, + "step": 3362 + }, + { + "epoch": 1.5825882352941176, + "grad_norm": 0.0570963359839155, + "learning_rate": 0.00028825681263907276, + "loss": 0.0334, + "step": 3363 + }, + { + "epoch": 1.5830588235294116, + "grad_norm": 0.06554285491211577, + "learning_rate": 0.00028823980378870457, + "loss": 0.038, + "step": 3364 + }, + { + "epoch": 1.5835294117647059, + "grad_norm": 0.06677892168297266, + "learning_rate": 0.0002882227831319791, + "loss": 0.0313, + "step": 3365 + }, + { + "epoch": 1.584, + "grad_norm": 0.04471734913612017, + "learning_rate": 0.00028820575067035, + "loss": 0.0254, + "step": 3366 + }, + { + "epoch": 1.584470588235294, + "grad_norm": 0.06084268903989695, + "learning_rate": 0.00028818870640527183, + "loss": 0.0277, + "step": 3367 + }, + { + "epoch": 1.584941176470588, + "grad_norm": 0.08998039378405838, + "learning_rate": 0.0002881716503382004, + "loss": 0.0518, + "step": 3368 + }, + { + "epoch": 1.5854117647058823, + "grad_norm": 0.09343146724280534, + "learning_rate": 0.0002881545824705922, + "loss": 0.0702, + "step": 3369 + }, + { + "epoch": 1.5858823529411765, + "grad_norm": 0.09111888988700464, + "learning_rate": 0.00028813750280390506, + "loss": 0.0617, + "step": 3370 + }, + { + "epoch": 1.5863529411764707, + "grad_norm": 0.08601480369896246, + "learning_rate": 0.00028812041133959764, + "loss": 0.0591, + "step": 3371 + }, + { + "epoch": 1.5868235294117647, + "grad_norm": 0.10767759735938463, + "learning_rate": 0.0002881033080791296, + "loss": 0.089, + "step": 3372 + }, + { + "epoch": 1.5872941176470587, + "grad_norm": 0.019537626190398717, + "learning_rate": 0.0002880861930239616, + "loss": 0.0043, + "step": 3373 + }, + { + "epoch": 1.587764705882353, + "grad_norm": 0.032110805825407596, + "learning_rate": 0.0002880690661755555, + "loss": 0.0052, + "step": 3374 + }, + { + "epoch": 1.5882352941176472, + "grad_norm": 0.036595562455832724, + "learning_rate": 0.00028805192753537386, + "loss": 0.013, + "step": 3375 + }, + { + "epoch": 1.5887058823529412, + "grad_norm": 0.04509292304837075, + "learning_rate": 0.00028803477710488055, + "loss": 0.0189, + "step": 3376 + }, + { + "epoch": 1.5891764705882352, + "grad_norm": 0.08153305656929465, + "learning_rate": 0.00028801761488554017, + "loss": 0.0359, + "step": 3377 + }, + { + "epoch": 1.5896470588235294, + "grad_norm": 0.1022832044895004, + "learning_rate": 0.0002880004408788185, + "loss": 0.0699, + "step": 3378 + }, + { + "epoch": 1.5901176470588236, + "grad_norm": 0.05354862732026046, + "learning_rate": 0.0002879832550861823, + "loss": 0.0234, + "step": 3379 + }, + { + "epoch": 1.5905882352941176, + "grad_norm": 0.029020534110767784, + "learning_rate": 0.0002879660575090993, + "loss": 0.0064, + "step": 3380 + }, + { + "epoch": 1.5910588235294116, + "grad_norm": 0.03869479084958108, + "learning_rate": 0.0002879488481490382, + "loss": 0.0133, + "step": 3381 + }, + { + "epoch": 1.5915294117647059, + "grad_norm": 0.0529540889123236, + "learning_rate": 0.0002879316270074689, + "loss": 0.0221, + "step": 3382 + }, + { + "epoch": 1.592, + "grad_norm": 0.0651077659565399, + "learning_rate": 0.00028791439408586215, + "loss": 0.042, + "step": 3383 + }, + { + "epoch": 1.592470588235294, + "grad_norm": 0.042103408197086946, + "learning_rate": 0.00028789714938568965, + "loss": 0.0152, + "step": 3384 + }, + { + "epoch": 1.592941176470588, + "grad_norm": 0.16674590918003604, + "learning_rate": 0.0002878798929084242, + "loss": 0.098, + "step": 3385 + }, + { + "epoch": 1.5934117647058823, + "grad_norm": 0.1271062241830723, + "learning_rate": 0.00028786262465553954, + "loss": 0.1481, + "step": 3386 + }, + { + "epoch": 1.5938823529411765, + "grad_norm": 6.039554132921363e-06, + "learning_rate": 0.00028784534462851063, + "loss": 0.0, + "step": 3387 + }, + { + "epoch": 1.5943529411764708, + "grad_norm": 0.034805308782857446, + "learning_rate": 0.0002878280528288131, + "loss": 0.0129, + "step": 3388 + }, + { + "epoch": 1.5948235294117648, + "grad_norm": 0.053482268637332506, + "learning_rate": 0.00028781074925792377, + "loss": 0.02, + "step": 3389 + }, + { + "epoch": 1.5952941176470588, + "grad_norm": 0.031245877057108296, + "learning_rate": 0.00028779343391732053, + "loss": 0.0127, + "step": 3390 + }, + { + "epoch": 1.595764705882353, + "grad_norm": 0.034068302208717136, + "learning_rate": 0.0002877761068084822, + "loss": 0.011, + "step": 3391 + }, + { + "epoch": 1.5962352941176472, + "grad_norm": 0.05700803640118704, + "learning_rate": 0.00028775876793288855, + "loss": 0.0344, + "step": 3392 + }, + { + "epoch": 1.5967058823529412, + "grad_norm": 5.923068353159944e-06, + "learning_rate": 0.0002877414172920204, + "loss": 0.0, + "step": 3393 + }, + { + "epoch": 1.5971764705882352, + "grad_norm": 0.08757916964440306, + "learning_rate": 0.00028772405488735963, + "loss": 0.0572, + "step": 3394 + }, + { + "epoch": 1.5976470588235294, + "grad_norm": 0.10313488933009338, + "learning_rate": 0.000287706680720389, + "loss": 0.0542, + "step": 3395 + }, + { + "epoch": 1.5981176470588236, + "grad_norm": 0.11782104990613707, + "learning_rate": 0.0002876892947925924, + "loss": 0.1092, + "step": 3396 + }, + { + "epoch": 1.5985882352941176, + "grad_norm": 0.05507975985163754, + "learning_rate": 0.0002876718971054547, + "loss": 0.0322, + "step": 3397 + }, + { + "epoch": 1.5990588235294116, + "grad_norm": 0.04997189551795847, + "learning_rate": 0.00028765448766046163, + "loss": 0.0279, + "step": 3398 + }, + { + "epoch": 1.5995294117647059, + "grad_norm": 0.10840170483766559, + "learning_rate": 0.00028763706645910014, + "loss": 0.0686, + "step": 3399 + }, + { + "epoch": 1.6, + "grad_norm": 0.10743596989167162, + "learning_rate": 0.0002876196335028581, + "loss": 0.0529, + "step": 3400 + }, + { + "epoch": 1.600470588235294, + "grad_norm": 0.06792200619112829, + "learning_rate": 0.00028760218879322436, + "loss": 0.0375, + "step": 3401 + }, + { + "epoch": 1.600941176470588, + "grad_norm": 0.09066351529612955, + "learning_rate": 0.00028758473233168874, + "loss": 0.073, + "step": 3402 + }, + { + "epoch": 1.6014117647058823, + "grad_norm": 0.073127902472561, + "learning_rate": 0.0002875672641197421, + "loss": 0.0403, + "step": 3403 + }, + { + "epoch": 1.6018823529411765, + "grad_norm": 0.1082274365516208, + "learning_rate": 0.00028754978415887635, + "loss": 0.0722, + "step": 3404 + }, + { + "epoch": 1.6023529411764705, + "grad_norm": 0.08991281134912774, + "learning_rate": 0.0002875322924505844, + "loss": 0.0482, + "step": 3405 + }, + { + "epoch": 1.6028235294117648, + "grad_norm": 3.3259106160968677e-06, + "learning_rate": 0.00028751478899636, + "loss": 0.0, + "step": 3406 + }, + { + "epoch": 1.6032941176470588, + "grad_norm": 0.04951507233417531, + "learning_rate": 0.00028749727379769817, + "loss": 0.0202, + "step": 3407 + }, + { + "epoch": 1.603764705882353, + "grad_norm": 0.05914829010081802, + "learning_rate": 0.0002874797468560948, + "loss": 0.0542, + "step": 3408 + }, + { + "epoch": 1.6042352941176472, + "grad_norm": 0.11915211823234755, + "learning_rate": 0.0002874622081730466, + "loss": 0.1108, + "step": 3409 + }, + { + "epoch": 1.6047058823529412, + "grad_norm": 0.06647892762826813, + "learning_rate": 0.0002874446577500516, + "loss": 0.0345, + "step": 3410 + }, + { + "epoch": 1.6051764705882352, + "grad_norm": 0.05180117084792197, + "learning_rate": 0.0002874270955886087, + "loss": 0.0324, + "step": 3411 + }, + { + "epoch": 1.6056470588235294, + "grad_norm": 0.11566714144833898, + "learning_rate": 0.0002874095216902178, + "loss": 0.0675, + "step": 3412 + }, + { + "epoch": 1.6061176470588236, + "grad_norm": 0.08735274882042438, + "learning_rate": 0.00028739193605637967, + "loss": 0.0691, + "step": 3413 + }, + { + "epoch": 1.6065882352941177, + "grad_norm": 0.06729628956730183, + "learning_rate": 0.00028737433868859634, + "loss": 0.0387, + "step": 3414 + }, + { + "epoch": 1.6070588235294117, + "grad_norm": 0.05082900370789904, + "learning_rate": 0.0002873567295883707, + "loss": 0.0225, + "step": 3415 + }, + { + "epoch": 1.6075294117647059, + "grad_norm": 0.14413643051162195, + "learning_rate": 0.00028733910875720654, + "loss": 0.1422, + "step": 3416 + }, + { + "epoch": 1.608, + "grad_norm": 0.07537494361411462, + "learning_rate": 0.00028732147619660883, + "loss": 0.0298, + "step": 3417 + }, + { + "epoch": 1.608470588235294, + "grad_norm": 0.08771034584443757, + "learning_rate": 0.00028730383190808364, + "loss": 0.0502, + "step": 3418 + }, + { + "epoch": 1.608941176470588, + "grad_norm": 0.05235076664650421, + "learning_rate": 0.0002872861758931376, + "loss": 0.0221, + "step": 3419 + }, + { + "epoch": 1.6094117647058823, + "grad_norm": 0.0619197428972044, + "learning_rate": 0.0002872685081532788, + "loss": 0.0327, + "step": 3420 + }, + { + "epoch": 1.6098823529411765, + "grad_norm": 0.09191026025902725, + "learning_rate": 0.00028725082869001615, + "loss": 0.0451, + "step": 3421 + }, + { + "epoch": 1.6103529411764705, + "grad_norm": 4.375929953743472e-06, + "learning_rate": 0.00028723313750485946, + "loss": 0.0, + "step": 3422 + }, + { + "epoch": 1.6108235294117645, + "grad_norm": 0.10318559306388937, + "learning_rate": 0.0002872154345993197, + "loss": 0.0252, + "step": 3423 + }, + { + "epoch": 1.6112941176470588, + "grad_norm": 0.07521878041763676, + "learning_rate": 0.0002871977199749088, + "loss": 0.0331, + "step": 3424 + }, + { + "epoch": 1.611764705882353, + "grad_norm": 0.04935478006797422, + "learning_rate": 0.00028717999363313967, + "loss": 0.0172, + "step": 3425 + }, + { + "epoch": 1.6122352941176472, + "grad_norm": 0.0561199701900447, + "learning_rate": 0.0002871622555755262, + "loss": 0.0222, + "step": 3426 + }, + { + "epoch": 1.6127058823529412, + "grad_norm": 0.06676955079284246, + "learning_rate": 0.00028714450580358334, + "loss": 0.0374, + "step": 3427 + }, + { + "epoch": 1.6131764705882352, + "grad_norm": 0.10062179393086818, + "learning_rate": 0.000287126744318827, + "loss": 0.0675, + "step": 3428 + }, + { + "epoch": 1.6136470588235294, + "grad_norm": 3.7592843678486257e-06, + "learning_rate": 0.0002871089711227741, + "loss": 0.0, + "step": 3429 + }, + { + "epoch": 1.6141176470588237, + "grad_norm": 0.0595222474311487, + "learning_rate": 0.0002870911862169426, + "loss": 0.0416, + "step": 3430 + }, + { + "epoch": 1.6145882352941177, + "grad_norm": 0.11011803599520882, + "learning_rate": 0.0002870733896028513, + "loss": 0.0693, + "step": 3431 + }, + { + "epoch": 1.6150588235294117, + "grad_norm": 0.09608380401176907, + "learning_rate": 0.00028705558128202017, + "loss": 0.0283, + "step": 3432 + }, + { + "epoch": 1.6155294117647059, + "grad_norm": 0.13320577509501105, + "learning_rate": 0.0002870377612559702, + "loss": 0.1329, + "step": 3433 + }, + { + "epoch": 1.616, + "grad_norm": 0.05799271858555696, + "learning_rate": 0.0002870199295262232, + "loss": 0.0274, + "step": 3434 + }, + { + "epoch": 1.616470588235294, + "grad_norm": 0.04879973750601557, + "learning_rate": 0.00028700208609430217, + "loss": 0.0206, + "step": 3435 + }, + { + "epoch": 1.616941176470588, + "grad_norm": 5.872528815859382e-06, + "learning_rate": 0.00028698423096173097, + "loss": 0.0, + "step": 3436 + }, + { + "epoch": 1.6174117647058823, + "grad_norm": 0.08692586844430383, + "learning_rate": 0.0002869663641300346, + "loss": 0.0523, + "step": 3437 + }, + { + "epoch": 1.6178823529411765, + "grad_norm": 0.023562609762227835, + "learning_rate": 0.00028694848560073885, + "loss": 0.0055, + "step": 3438 + }, + { + "epoch": 1.6183529411764705, + "grad_norm": 0.017994186343997413, + "learning_rate": 0.00028693059537537077, + "loss": 0.0024, + "step": 3439 + }, + { + "epoch": 1.6188235294117646, + "grad_norm": 0.05412994354435869, + "learning_rate": 0.00028691269345545816, + "loss": 0.0226, + "step": 3440 + }, + { + "epoch": 1.6192941176470588, + "grad_norm": 0.08314763997916243, + "learning_rate": 0.00028689477984252996, + "loss": 0.0392, + "step": 3441 + }, + { + "epoch": 1.619764705882353, + "grad_norm": 0.11253881542848326, + "learning_rate": 0.00028687685453811614, + "loss": 0.0538, + "step": 3442 + }, + { + "epoch": 1.6202352941176472, + "grad_norm": 2.714781794615841e-06, + "learning_rate": 0.00028685891754374755, + "loss": 0.0, + "step": 3443 + }, + { + "epoch": 1.6207058823529412, + "grad_norm": 0.04693707254455385, + "learning_rate": 0.0002868409688609561, + "loss": 0.0094, + "step": 3444 + }, + { + "epoch": 1.6211764705882352, + "grad_norm": 0.09513373914881129, + "learning_rate": 0.00028682300849127465, + "loss": 0.0461, + "step": 3445 + }, + { + "epoch": 1.6216470588235294, + "grad_norm": 0.07354063300694262, + "learning_rate": 0.0002868050364362373, + "loss": 0.0242, + "step": 3446 + }, + { + "epoch": 1.6221176470588237, + "grad_norm": 0.08929572182313655, + "learning_rate": 0.00028678705269737873, + "loss": 0.0536, + "step": 3447 + }, + { + "epoch": 1.6225882352941177, + "grad_norm": 0.08260849339503584, + "learning_rate": 0.0002867690572762349, + "loss": 0.0262, + "step": 3448 + }, + { + "epoch": 1.6230588235294117, + "grad_norm": 0.06913136017094537, + "learning_rate": 0.0002867510501743428, + "loss": 0.0265, + "step": 3449 + }, + { + "epoch": 1.6235294117647059, + "grad_norm": 0.1513530420676068, + "learning_rate": 0.0002867330313932402, + "loss": 0.0985, + "step": 3450 + }, + { + "epoch": 1.624, + "grad_norm": 4.5104517623775535e-06, + "learning_rate": 0.00028671500093446604, + "loss": 0.0, + "step": 3451 + }, + { + "epoch": 1.624470588235294, + "grad_norm": 0.014733932051541282, + "learning_rate": 0.0002866969587995602, + "loss": 0.0016, + "step": 3452 + }, + { + "epoch": 1.6249411764705881, + "grad_norm": 0.008181487272499776, + "learning_rate": 0.00028667890499006364, + "loss": 0.0004, + "step": 3453 + }, + { + "epoch": 1.6254117647058823, + "grad_norm": 2.807036107632744e-06, + "learning_rate": 0.00028666083950751824, + "loss": 0.0, + "step": 3454 + }, + { + "epoch": 1.6258823529411766, + "grad_norm": 0.08479542187476113, + "learning_rate": 0.0002866427623534667, + "loss": 0.0589, + "step": 3455 + }, + { + "epoch": 1.6263529411764706, + "grad_norm": 0.0923835975324468, + "learning_rate": 0.0002866246735294531, + "loss": 0.0882, + "step": 3456 + }, + { + "epoch": 1.6268235294117646, + "grad_norm": 3.873004623534314e-06, + "learning_rate": 0.00028660657303702225, + "loss": 0.0, + "step": 3457 + }, + { + "epoch": 1.6272941176470588, + "grad_norm": 0.027694385145249942, + "learning_rate": 0.00028658846087771994, + "loss": 0.0016, + "step": 3458 + }, + { + "epoch": 1.627764705882353, + "grad_norm": 4.7266298358203325e-06, + "learning_rate": 0.0002865703370530932, + "loss": 0.0, + "step": 3459 + }, + { + "epoch": 1.6282352941176472, + "grad_norm": 0.08837157961925286, + "learning_rate": 0.0002865522015646897, + "loss": 0.0633, + "step": 3460 + }, + { + "epoch": 1.6287058823529412, + "grad_norm": 0.08010747903255391, + "learning_rate": 0.0002865340544140585, + "loss": 0.0314, + "step": 3461 + }, + { + "epoch": 1.6291764705882352, + "grad_norm": 0.10924110182450113, + "learning_rate": 0.00028651589560274934, + "loss": 0.0744, + "step": 3462 + }, + { + "epoch": 1.6296470588235294, + "grad_norm": 0.17240492957282091, + "learning_rate": 0.00028649772513231306, + "loss": 0.1359, + "step": 3463 + }, + { + "epoch": 1.6301176470588237, + "grad_norm": 0.1114875656102418, + "learning_rate": 0.00028647954300430154, + "loss": 0.0818, + "step": 3464 + }, + { + "epoch": 1.6305882352941177, + "grad_norm": 3.5336923902532386e-06, + "learning_rate": 0.0002864613492202676, + "loss": 0.0, + "step": 3465 + }, + { + "epoch": 1.6310588235294117, + "grad_norm": 0.16213237881610873, + "learning_rate": 0.00028644314378176513, + "loss": 0.0818, + "step": 3466 + }, + { + "epoch": 1.631529411764706, + "grad_norm": 0.11409597334756418, + "learning_rate": 0.00028642492669034894, + "loss": 0.0593, + "step": 3467 + }, + { + "epoch": 1.6320000000000001, + "grad_norm": 0.06347519764154892, + "learning_rate": 0.0002864066979475748, + "loss": 0.03, + "step": 3468 + }, + { + "epoch": 1.6324705882352941, + "grad_norm": 0.09160208008609785, + "learning_rate": 0.00028638845755499964, + "loss": 0.0465, + "step": 3469 + }, + { + "epoch": 1.6329411764705881, + "grad_norm": 0.11347491024375111, + "learning_rate": 0.00028637020551418123, + "loss": 0.038, + "step": 3470 + }, + { + "epoch": 1.6334117647058823, + "grad_norm": 0.26536658934488433, + "learning_rate": 0.00028635194182667834, + "loss": 0.2337, + "step": 3471 + }, + { + "epoch": 1.6338823529411766, + "grad_norm": 0.17009203177926374, + "learning_rate": 0.0002863336664940508, + "loss": 0.1398, + "step": 3472 + }, + { + "epoch": 1.6343529411764706, + "grad_norm": 0.18388842863112734, + "learning_rate": 0.00028631537951785945, + "loss": 0.0262, + "step": 3473 + }, + { + "epoch": 1.6348235294117646, + "grad_norm": 0.050420167827349975, + "learning_rate": 0.000286297080899666, + "loss": 0.0248, + "step": 3474 + }, + { + "epoch": 1.6352941176470588, + "grad_norm": 0.10803945942255186, + "learning_rate": 0.0002862787706410334, + "loss": 0.081, + "step": 3475 + }, + { + "epoch": 1.635764705882353, + "grad_norm": 0.09972433611038556, + "learning_rate": 0.00028626044874352534, + "loss": 0.0426, + "step": 3476 + }, + { + "epoch": 1.636235294117647, + "grad_norm": 0.1371014230132716, + "learning_rate": 0.0002862421152087066, + "loss": 0.0431, + "step": 3477 + }, + { + "epoch": 1.636705882352941, + "grad_norm": 0.09777602102083614, + "learning_rate": 0.00028622377003814295, + "loss": 0.0437, + "step": 3478 + }, + { + "epoch": 1.6371764705882352, + "grad_norm": 0.07718774616078242, + "learning_rate": 0.00028620541323340115, + "loss": 0.0316, + "step": 3479 + }, + { + "epoch": 1.6376470588235295, + "grad_norm": 0.12313145604636921, + "learning_rate": 0.00028618704479604895, + "loss": 0.0606, + "step": 3480 + }, + { + "epoch": 1.6381176470588237, + "grad_norm": 0.0572109127422155, + "learning_rate": 0.00028616866472765514, + "loss": 0.0135, + "step": 3481 + }, + { + "epoch": 1.6385882352941177, + "grad_norm": 0.07739964972334606, + "learning_rate": 0.00028615027302978947, + "loss": 0.057, + "step": 3482 + }, + { + "epoch": 1.6390588235294117, + "grad_norm": 0.17926970376470414, + "learning_rate": 0.00028613186970402274, + "loss": 0.1668, + "step": 3483 + }, + { + "epoch": 1.639529411764706, + "grad_norm": 0.11946338617477044, + "learning_rate": 0.0002861134547519265, + "loss": 0.0665, + "step": 3484 + }, + { + "epoch": 1.6400000000000001, + "grad_norm": 6.080705580581166e-06, + "learning_rate": 0.0002860950281750737, + "loss": 0.0, + "step": 3485 + }, + { + "epoch": 1.6404705882352941, + "grad_norm": 0.05394210145132533, + "learning_rate": 0.00028607658997503786, + "loss": 0.0315, + "step": 3486 + }, + { + "epoch": 1.6409411764705881, + "grad_norm": 0.06460292400318011, + "learning_rate": 0.00028605814015339383, + "loss": 0.0201, + "step": 3487 + }, + { + "epoch": 1.6414117647058823, + "grad_norm": 0.07892224065705045, + "learning_rate": 0.0002860396787117173, + "loss": 0.0305, + "step": 3488 + }, + { + "epoch": 1.6418823529411766, + "grad_norm": 0.5762333523280444, + "learning_rate": 0.00028602120565158486, + "loss": 0.1278, + "step": 3489 + }, + { + "epoch": 1.6423529411764706, + "grad_norm": 0.027538649719172074, + "learning_rate": 0.0002860027209745743, + "loss": 0.0059, + "step": 3490 + }, + { + "epoch": 1.6428235294117646, + "grad_norm": 0.045355713461579274, + "learning_rate": 0.00028598422468226425, + "loss": 0.0114, + "step": 3491 + }, + { + "epoch": 1.6432941176470588, + "grad_norm": 0.06354378149886379, + "learning_rate": 0.0002859657167762345, + "loss": 0.0268, + "step": 3492 + }, + { + "epoch": 1.643764705882353, + "grad_norm": 0.05846548346849131, + "learning_rate": 0.00028594719725806556, + "loss": 0.0288, + "step": 3493 + }, + { + "epoch": 1.644235294117647, + "grad_norm": 0.06237348034551396, + "learning_rate": 0.0002859286661293392, + "loss": 0.0368, + "step": 3494 + }, + { + "epoch": 1.644705882352941, + "grad_norm": 0.11190514997421439, + "learning_rate": 0.00028591012339163806, + "loss": 0.049, + "step": 3495 + }, + { + "epoch": 1.6451764705882352, + "grad_norm": 0.10581590961554929, + "learning_rate": 0.00028589156904654565, + "loss": 0.0496, + "step": 3496 + }, + { + "epoch": 1.6456470588235295, + "grad_norm": 0.11297867178048362, + "learning_rate": 0.00028587300309564675, + "loss": 0.0515, + "step": 3497 + }, + { + "epoch": 1.6461176470588237, + "grad_norm": 0.1352747643582626, + "learning_rate": 0.0002858544255405269, + "loss": 0.0522, + "step": 3498 + }, + { + "epoch": 1.6465882352941177, + "grad_norm": 0.0647892220477533, + "learning_rate": 0.0002858358363827728, + "loss": 0.0303, + "step": 3499 + }, + { + "epoch": 1.6470588235294117, + "grad_norm": 0.0835233886319847, + "learning_rate": 0.000285817235623972, + "loss": 0.0546, + "step": 3500 + }, + { + "epoch": 1.647529411764706, + "grad_norm": 0.0931081189585222, + "learning_rate": 0.00028579862326571305, + "loss": 0.0755, + "step": 3501 + }, + { + "epoch": 1.6480000000000001, + "grad_norm": 0.10952921560063919, + "learning_rate": 0.0002857799993095856, + "loss": 0.0709, + "step": 3502 + }, + { + "epoch": 1.6484705882352941, + "grad_norm": 6.375991019430079e-06, + "learning_rate": 0.0002857613637571802, + "loss": 0.0, + "step": 3503 + }, + { + "epoch": 1.6489411764705881, + "grad_norm": 0.031117883246876407, + "learning_rate": 0.00028574271661008845, + "loss": 0.0128, + "step": 3504 + }, + { + "epoch": 1.6494117647058824, + "grad_norm": 0.130157307336796, + "learning_rate": 0.00028572405786990294, + "loss": 0.1223, + "step": 3505 + }, + { + "epoch": 1.6498823529411766, + "grad_norm": 0.17103030715915138, + "learning_rate": 0.00028570538753821704, + "loss": 0.1691, + "step": 3506 + }, + { + "epoch": 1.6503529411764706, + "grad_norm": 0.08282143826764515, + "learning_rate": 0.0002856867056166255, + "loss": 0.0522, + "step": 3507 + }, + { + "epoch": 1.6508235294117646, + "grad_norm": 0.058406821926819424, + "learning_rate": 0.00028566801210672376, + "loss": 0.0233, + "step": 3508 + }, + { + "epoch": 1.6512941176470588, + "grad_norm": 0.09479324690988859, + "learning_rate": 0.0002856493070101084, + "loss": 0.0392, + "step": 3509 + }, + { + "epoch": 1.651764705882353, + "grad_norm": 0.07274872679056851, + "learning_rate": 0.0002856305903283767, + "loss": 0.0357, + "step": 3510 + }, + { + "epoch": 1.652235294117647, + "grad_norm": 0.1023140071759177, + "learning_rate": 0.00028561186206312744, + "loss": 0.0914, + "step": 3511 + }, + { + "epoch": 1.652705882352941, + "grad_norm": 0.07203373449263933, + "learning_rate": 0.00028559312221595996, + "loss": 0.0397, + "step": 3512 + }, + { + "epoch": 1.6531764705882352, + "grad_norm": 0.03951363392975862, + "learning_rate": 0.00028557437078847477, + "loss": 0.018, + "step": 3513 + }, + { + "epoch": 1.6536470588235295, + "grad_norm": 0.06192967940827255, + "learning_rate": 0.0002855556077822733, + "loss": 0.0288, + "step": 3514 + }, + { + "epoch": 1.6541176470588237, + "grad_norm": 0.03029920954170581, + "learning_rate": 0.00028553683319895804, + "loss": 0.0067, + "step": 3515 + }, + { + "epoch": 1.6545882352941177, + "grad_norm": 0.03282750171750897, + "learning_rate": 0.00028551804704013244, + "loss": 0.0121, + "step": 3516 + }, + { + "epoch": 1.6550588235294117, + "grad_norm": 0.0300936172436424, + "learning_rate": 0.0002854992493074009, + "loss": 0.0077, + "step": 3517 + }, + { + "epoch": 1.655529411764706, + "grad_norm": 0.0859601607178454, + "learning_rate": 0.00028548044000236886, + "loss": 0.0457, + "step": 3518 + }, + { + "epoch": 1.6560000000000001, + "grad_norm": 5.576646026962178e-05, + "learning_rate": 0.0002854616191266427, + "loss": 0.0, + "step": 3519 + }, + { + "epoch": 1.6564705882352941, + "grad_norm": 0.05368118877303025, + "learning_rate": 0.0002854427866818298, + "loss": 0.028, + "step": 3520 + }, + { + "epoch": 1.6569411764705881, + "grad_norm": 0.038230024884783574, + "learning_rate": 0.00028542394266953865, + "loss": 0.0167, + "step": 3521 + }, + { + "epoch": 1.6574117647058824, + "grad_norm": 0.05154933024922866, + "learning_rate": 0.00028540508709137846, + "loss": 0.0212, + "step": 3522 + }, + { + "epoch": 1.6578823529411766, + "grad_norm": 0.06537590292778689, + "learning_rate": 0.00028538621994895976, + "loss": 0.0404, + "step": 3523 + }, + { + "epoch": 1.6583529411764706, + "grad_norm": 0.00011273426963205496, + "learning_rate": 0.00028536734124389373, + "loss": 0.0, + "step": 3524 + }, + { + "epoch": 1.6588235294117646, + "grad_norm": 0.22750015098499932, + "learning_rate": 0.0002853484509777928, + "loss": 0.1093, + "step": 3525 + }, + { + "epoch": 1.6592941176470588, + "grad_norm": 3.8798780727998144e-05, + "learning_rate": 0.00028532954915227033, + "loss": 0.0, + "step": 3526 + }, + { + "epoch": 1.659764705882353, + "grad_norm": 8.127597443313115e-05, + "learning_rate": 0.0002853106357689405, + "loss": 0.0, + "step": 3527 + }, + { + "epoch": 1.660235294117647, + "grad_norm": 0.03575755422488583, + "learning_rate": 0.0002852917108294187, + "loss": 0.0132, + "step": 3528 + }, + { + "epoch": 1.660705882352941, + "grad_norm": 2.151237051897658e-05, + "learning_rate": 0.0002852727743353213, + "loss": 0.0, + "step": 3529 + }, + { + "epoch": 1.6611764705882353, + "grad_norm": 0.08512912894690965, + "learning_rate": 0.00028525382628826536, + "loss": 0.0534, + "step": 3530 + }, + { + "epoch": 1.6616470588235295, + "grad_norm": 0.11329985475893814, + "learning_rate": 0.0002852348666898692, + "loss": 0.09, + "step": 3531 + }, + { + "epoch": 1.6621176470588235, + "grad_norm": 0.12155324709121407, + "learning_rate": 0.0002852158955417522, + "loss": 0.0814, + "step": 3532 + }, + { + "epoch": 1.6625882352941177, + "grad_norm": 0.06323752125736976, + "learning_rate": 0.0002851969128455345, + "loss": 0.0455, + "step": 3533 + }, + { + "epoch": 1.6630588235294117, + "grad_norm": 0.05814910152914198, + "learning_rate": 0.0002851779186028372, + "loss": 0.03, + "step": 3534 + }, + { + "epoch": 1.663529411764706, + "grad_norm": 0.07994730323307911, + "learning_rate": 0.0002851589128152827, + "loss": 0.0516, + "step": 3535 + }, + { + "epoch": 1.6640000000000001, + "grad_norm": 0.00014999238881860917, + "learning_rate": 0.000285139895484494, + "loss": 0.0, + "step": 3536 + }, + { + "epoch": 1.6644705882352941, + "grad_norm": 2.836793194369161e-05, + "learning_rate": 0.0002851208666120955, + "loss": 0.0, + "step": 3537 + }, + { + "epoch": 1.6649411764705881, + "grad_norm": 0.06089565305739536, + "learning_rate": 0.0002851018261997121, + "loss": 0.0416, + "step": 3538 + }, + { + "epoch": 1.6654117647058824, + "grad_norm": 4.8517288957645175e-06, + "learning_rate": 0.00028508277424897015, + "loss": 0.0, + "step": 3539 + }, + { + "epoch": 1.6658823529411766, + "grad_norm": 5.090493758104797e-05, + "learning_rate": 0.00028506371076149665, + "loss": 0.0, + "step": 3540 + }, + { + "epoch": 1.6663529411764706, + "grad_norm": 0.04588651306256323, + "learning_rate": 0.0002850446357389198, + "loss": 0.0169, + "step": 3541 + }, + { + "epoch": 1.6668235294117646, + "grad_norm": 0.16238480765652147, + "learning_rate": 0.00028502554918286867, + "loss": 0.1064, + "step": 3542 + }, + { + "epoch": 1.6672941176470588, + "grad_norm": 0.1556384441742541, + "learning_rate": 0.0002850064510949733, + "loss": 0.1178, + "step": 3543 + }, + { + "epoch": 1.667764705882353, + "grad_norm": 2.5256487120742404e-05, + "learning_rate": 0.00028498734147686485, + "loss": 0.0, + "step": 3544 + }, + { + "epoch": 1.668235294117647, + "grad_norm": 0.06194821913535325, + "learning_rate": 0.0002849682203301753, + "loss": 0.033, + "step": 3545 + }, + { + "epoch": 1.668705882352941, + "grad_norm": 0.04022255113574712, + "learning_rate": 0.0002849490876565378, + "loss": 0.0112, + "step": 3546 + }, + { + "epoch": 1.6691764705882353, + "grad_norm": 0.13204872376650903, + "learning_rate": 0.0002849299434575862, + "loss": 0.1097, + "step": 3547 + }, + { + "epoch": 1.6696470588235295, + "grad_norm": 0.0771814665080258, + "learning_rate": 0.00028491078773495564, + "loss": 0.0448, + "step": 3548 + }, + { + "epoch": 1.6701176470588235, + "grad_norm": 0.11141567863035833, + "learning_rate": 0.00028489162049028205, + "loss": 0.0788, + "step": 3549 + }, + { + "epoch": 1.6705882352941175, + "grad_norm": 0.04984475724804974, + "learning_rate": 0.00028487244172520246, + "loss": 0.0296, + "step": 3550 + }, + { + "epoch": 1.6710588235294117, + "grad_norm": 0.08350868906294269, + "learning_rate": 0.0002848532514413548, + "loss": 0.0384, + "step": 3551 + }, + { + "epoch": 1.671529411764706, + "grad_norm": 0.11057133272708958, + "learning_rate": 0.00028483404964037794, + "loss": 0.078, + "step": 3552 + }, + { + "epoch": 1.6720000000000002, + "grad_norm": 0.09573886405266284, + "learning_rate": 0.000284814836323912, + "loss": 0.0792, + "step": 3553 + }, + { + "epoch": 1.6724705882352942, + "grad_norm": 0.11044514100552816, + "learning_rate": 0.0002847956114935977, + "loss": 0.0744, + "step": 3554 + }, + { + "epoch": 1.6729411764705882, + "grad_norm": 0.045441207767843064, + "learning_rate": 0.000284776375151077, + "loss": 0.0229, + "step": 3555 + }, + { + "epoch": 1.6734117647058824, + "grad_norm": 0.11093206269446246, + "learning_rate": 0.00028475712729799285, + "loss": 0.0993, + "step": 3556 + }, + { + "epoch": 1.6738823529411766, + "grad_norm": 0.13608174662354225, + "learning_rate": 0.000284737867935989, + "loss": 0.1328, + "step": 3557 + }, + { + "epoch": 1.6743529411764706, + "grad_norm": 0.1152905821893068, + "learning_rate": 0.0002847185970667104, + "loss": 0.0738, + "step": 3558 + }, + { + "epoch": 1.6748235294117646, + "grad_norm": 0.03239144418921298, + "learning_rate": 0.00028469931469180275, + "loss": 0.0104, + "step": 3559 + }, + { + "epoch": 1.6752941176470588, + "grad_norm": 0.11696158956232892, + "learning_rate": 0.00028468002081291296, + "loss": 0.1299, + "step": 3560 + }, + { + "epoch": 1.675764705882353, + "grad_norm": 0.11479599542849728, + "learning_rate": 0.0002846607154316888, + "loss": 0.0822, + "step": 3561 + }, + { + "epoch": 1.676235294117647, + "grad_norm": 0.09806261177650347, + "learning_rate": 0.000284641398549779, + "loss": 0.0524, + "step": 3562 + }, + { + "epoch": 1.676705882352941, + "grad_norm": 0.11534890314299032, + "learning_rate": 0.00028462207016883336, + "loss": 0.0866, + "step": 3563 + }, + { + "epoch": 1.6771764705882353, + "grad_norm": 0.12116360960365317, + "learning_rate": 0.0002846027302905026, + "loss": 0.1065, + "step": 3564 + }, + { + "epoch": 1.6776470588235295, + "grad_norm": 0.22550396311204343, + "learning_rate": 0.0002845833789164385, + "loss": 0.1748, + "step": 3565 + }, + { + "epoch": 1.6781176470588235, + "grad_norm": 0.04272196377572511, + "learning_rate": 0.00028456401604829367, + "loss": 0.0175, + "step": 3566 + }, + { + "epoch": 1.6785882352941175, + "grad_norm": 0.04579795129620206, + "learning_rate": 0.00028454464168772186, + "loss": 0.0143, + "step": 3567 + }, + { + "epoch": 1.6790588235294117, + "grad_norm": 0.03792270511153563, + "learning_rate": 0.00028452525583637774, + "loss": 0.0134, + "step": 3568 + }, + { + "epoch": 1.679529411764706, + "grad_norm": 2.852970351032915e-06, + "learning_rate": 0.0002845058584959169, + "loss": 0.0, + "step": 3569 + }, + { + "epoch": 1.6800000000000002, + "grad_norm": 0.08671816510819128, + "learning_rate": 0.00028448644966799595, + "loss": 0.051, + "step": 3570 + }, + { + "epoch": 1.6804705882352942, + "grad_norm": 0.10457944984004693, + "learning_rate": 0.0002844670293542726, + "loss": 0.0676, + "step": 3571 + }, + { + "epoch": 1.6809411764705882, + "grad_norm": 0.10805840747580239, + "learning_rate": 0.00028444759755640535, + "loss": 0.0916, + "step": 3572 + }, + { + "epoch": 1.6814117647058824, + "grad_norm": 4.426138827367284e-06, + "learning_rate": 0.0002844281542760539, + "loss": 0.0, + "step": 3573 + }, + { + "epoch": 1.6818823529411766, + "grad_norm": 0.05856095839623471, + "learning_rate": 0.0002844086995148786, + "loss": 0.0225, + "step": 3574 + }, + { + "epoch": 1.6823529411764706, + "grad_norm": 0.041370962127228475, + "learning_rate": 0.00028438923327454114, + "loss": 0.013, + "step": 3575 + }, + { + "epoch": 1.6828235294117646, + "grad_norm": 0.07514865135172005, + "learning_rate": 0.000284369755556704, + "loss": 0.0222, + "step": 3576 + }, + { + "epoch": 1.6832941176470588, + "grad_norm": 0.0968638991013544, + "learning_rate": 0.0002843502663630306, + "loss": 0.0581, + "step": 3577 + }, + { + "epoch": 1.683764705882353, + "grad_norm": 0.12391509875967925, + "learning_rate": 0.0002843307656951855, + "loss": 0.0899, + "step": 3578 + }, + { + "epoch": 1.684235294117647, + "grad_norm": 0.11079945557380118, + "learning_rate": 0.0002843112535548341, + "loss": 0.0616, + "step": 3579 + }, + { + "epoch": 1.684705882352941, + "grad_norm": 0.14744389715974512, + "learning_rate": 0.0002842917299436429, + "loss": 0.1263, + "step": 3580 + }, + { + "epoch": 1.6851764705882353, + "grad_norm": 4.044977043533201e-06, + "learning_rate": 0.00028427219486327923, + "loss": 0.0, + "step": 3581 + }, + { + "epoch": 1.6856470588235295, + "grad_norm": 0.0735030164141917, + "learning_rate": 0.00028425264831541153, + "loss": 0.032, + "step": 3582 + }, + { + "epoch": 1.6861176470588235, + "grad_norm": 0.05518051129400346, + "learning_rate": 0.0002842330903017091, + "loss": 0.0281, + "step": 3583 + }, + { + "epoch": 1.6865882352941175, + "grad_norm": 0.05007025473881421, + "learning_rate": 0.00028421352082384245, + "loss": 0.0161, + "step": 3584 + }, + { + "epoch": 1.6870588235294117, + "grad_norm": 0.05564771389130808, + "learning_rate": 0.00028419393988348275, + "loss": 0.0159, + "step": 3585 + }, + { + "epoch": 1.687529411764706, + "grad_norm": 0.1030185078292677, + "learning_rate": 0.00028417434748230237, + "loss": 0.0825, + "step": 3586 + }, + { + "epoch": 1.688, + "grad_norm": 0.10056477277571545, + "learning_rate": 0.0002841547436219746, + "loss": 0.0559, + "step": 3587 + }, + { + "epoch": 1.6884705882352942, + "grad_norm": 0.06445499562331002, + "learning_rate": 0.0002841351283041737, + "loss": 0.033, + "step": 3588 + }, + { + "epoch": 1.6889411764705882, + "grad_norm": 0.07872747648151908, + "learning_rate": 0.0002841155015305749, + "loss": 0.044, + "step": 3589 + }, + { + "epoch": 1.6894117647058824, + "grad_norm": 0.06478227012476975, + "learning_rate": 0.00028409586330285444, + "loss": 0.0282, + "step": 3590 + }, + { + "epoch": 1.6898823529411766, + "grad_norm": 0.134025415124947, + "learning_rate": 0.00028407621362268955, + "loss": 0.1125, + "step": 3591 + }, + { + "epoch": 1.6903529411764706, + "grad_norm": 0.12098727479631217, + "learning_rate": 0.0002840565524917583, + "loss": 0.0623, + "step": 3592 + }, + { + "epoch": 1.6908235294117646, + "grad_norm": 0.00025051381178587946, + "learning_rate": 0.00028403687991174003, + "loss": 0.0, + "step": 3593 + }, + { + "epoch": 1.6912941176470588, + "grad_norm": 0.09828152197990626, + "learning_rate": 0.0002840171958843147, + "loss": 0.0611, + "step": 3594 + }, + { + "epoch": 1.691764705882353, + "grad_norm": 0.1339518916347007, + "learning_rate": 0.0002839975004111635, + "loss": 0.0898, + "step": 3595 + }, + { + "epoch": 1.692235294117647, + "grad_norm": 1.2594776848440785e-05, + "learning_rate": 0.00028397779349396854, + "loss": 0.0, + "step": 3596 + }, + { + "epoch": 1.692705882352941, + "grad_norm": 0.07132460831806638, + "learning_rate": 0.0002839580751344128, + "loss": 0.0423, + "step": 3597 + }, + { + "epoch": 1.6931764705882353, + "grad_norm": 2.7576686388150547e-06, + "learning_rate": 0.00028393834533418043, + "loss": 0.0, + "step": 3598 + }, + { + "epoch": 1.6936470588235295, + "grad_norm": 2.6749268335888677e-06, + "learning_rate": 0.0002839186040949564, + "loss": 0.0, + "step": 3599 + }, + { + "epoch": 1.6941176470588235, + "grad_norm": 3.7131428747116665e-06, + "learning_rate": 0.0002838988514184267, + "loss": 0.0, + "step": 3600 + }, + { + "epoch": 1.6945882352941175, + "grad_norm": 0.6577955460439142, + "learning_rate": 0.0002838790873062784, + "loss": 0.2019, + "step": 3601 + }, + { + "epoch": 1.6950588235294117, + "grad_norm": 0.22657274082177403, + "learning_rate": 0.0002838593117601992, + "loss": 0.2026, + "step": 3602 + }, + { + "epoch": 1.695529411764706, + "grad_norm": 0.03519391376241419, + "learning_rate": 0.0002838395247818783, + "loss": 0.0132, + "step": 3603 + }, + { + "epoch": 1.696, + "grad_norm": 1.944309587406317e-06, + "learning_rate": 0.0002838197263730055, + "loss": 0.0, + "step": 3604 + }, + { + "epoch": 1.696470588235294, + "grad_norm": 0.029132261419097985, + "learning_rate": 0.0002837999165352717, + "loss": 0.0084, + "step": 3605 + }, + { + "epoch": 1.6969411764705882, + "grad_norm": 0.07090161112385936, + "learning_rate": 0.0002837800952703687, + "loss": 0.0349, + "step": 3606 + }, + { + "epoch": 1.6974117647058824, + "grad_norm": 0.09103679667956736, + "learning_rate": 0.00028376026257998947, + "loss": 0.0467, + "step": 3607 + }, + { + "epoch": 1.6978823529411766, + "grad_norm": 0.049299005842959986, + "learning_rate": 0.0002837404184658276, + "loss": 0.0177, + "step": 3608 + }, + { + "epoch": 1.6983529411764706, + "grad_norm": 0.09222315312814562, + "learning_rate": 0.00028372056292957806, + "loss": 0.0781, + "step": 3609 + }, + { + "epoch": 1.6988235294117646, + "grad_norm": 0.08793387238018108, + "learning_rate": 0.0002837006959729366, + "loss": 0.0585, + "step": 3610 + }, + { + "epoch": 1.6992941176470588, + "grad_norm": 0.15767249825938837, + "learning_rate": 0.00028368081759759983, + "loss": 0.1507, + "step": 3611 + }, + { + "epoch": 1.699764705882353, + "grad_norm": 0.06267631256671387, + "learning_rate": 0.00028366092780526554, + "loss": 0.0232, + "step": 3612 + }, + { + "epoch": 1.700235294117647, + "grad_norm": 0.0708170270741189, + "learning_rate": 0.00028364102659763247, + "loss": 0.0382, + "step": 3613 + }, + { + "epoch": 1.700705882352941, + "grad_norm": 5.4557689715734275e-06, + "learning_rate": 0.00028362111397640016, + "loss": 0.0, + "step": 3614 + }, + { + "epoch": 1.7011764705882353, + "grad_norm": 0.0007049137653518672, + "learning_rate": 0.00028360118994326936, + "loss": 0.0, + "step": 3615 + }, + { + "epoch": 1.7016470588235295, + "grad_norm": 0.05015544484734306, + "learning_rate": 0.00028358125449994156, + "loss": 0.0229, + "step": 3616 + }, + { + "epoch": 1.7021176470588235, + "grad_norm": 0.055262825629453595, + "learning_rate": 0.0002835613076481195, + "loss": 0.0322, + "step": 3617 + }, + { + "epoch": 1.7025882352941175, + "grad_norm": 1.6911354780133223e-06, + "learning_rate": 0.0002835413493895066, + "loss": 0.0, + "step": 3618 + }, + { + "epoch": 1.7030588235294117, + "grad_norm": 0.08731357185430907, + "learning_rate": 0.00028352137972580745, + "loss": 0.0678, + "step": 3619 + }, + { + "epoch": 1.703529411764706, + "grad_norm": 0.17851020068125956, + "learning_rate": 0.0002835013986587276, + "loss": 0.1887, + "step": 3620 + }, + { + "epoch": 1.704, + "grad_norm": 0.12164479807178954, + "learning_rate": 0.0002834814061899734, + "loss": 0.1139, + "step": 3621 + }, + { + "epoch": 1.704470588235294, + "grad_norm": 0.07521671131827956, + "learning_rate": 0.0002834614023212525, + "loss": 0.047, + "step": 3622 + }, + { + "epoch": 1.7049411764705882, + "grad_norm": 0.032556658545836024, + "learning_rate": 0.0002834413870542731, + "loss": 0.0085, + "step": 3623 + }, + { + "epoch": 1.7054117647058824, + "grad_norm": 0.11819115562901325, + "learning_rate": 0.00028342136039074476, + "loss": 0.0861, + "step": 3624 + }, + { + "epoch": 1.7058823529411766, + "grad_norm": 0.05383135613806816, + "learning_rate": 0.0002834013223323778, + "loss": 0.0178, + "step": 3625 + }, + { + "epoch": 1.7063529411764706, + "grad_norm": 0.10600920554355786, + "learning_rate": 0.0002833812728808836, + "loss": 0.0355, + "step": 3626 + }, + { + "epoch": 1.7068235294117646, + "grad_norm": 0.038461082588278066, + "learning_rate": 0.00028336121203797446, + "loss": 0.0154, + "step": 3627 + }, + { + "epoch": 1.7072941176470589, + "grad_norm": 0.030126994868790682, + "learning_rate": 0.0002833411398053637, + "loss": 0.0063, + "step": 3628 + }, + { + "epoch": 1.707764705882353, + "grad_norm": 0.050304550726925634, + "learning_rate": 0.00028332105618476556, + "loss": 0.0195, + "step": 3629 + }, + { + "epoch": 1.708235294117647, + "grad_norm": 0.12073924519081075, + "learning_rate": 0.0002833009611778953, + "loss": 0.0829, + "step": 3630 + }, + { + "epoch": 1.708705882352941, + "grad_norm": 0.11841358777767423, + "learning_rate": 0.0002832808547864691, + "loss": 0.066, + "step": 3631 + }, + { + "epoch": 1.7091764705882353, + "grad_norm": 0.11691651017578217, + "learning_rate": 0.00028326073701220416, + "loss": 0.0746, + "step": 3632 + }, + { + "epoch": 1.7096470588235295, + "grad_norm": 0.02924287892388297, + "learning_rate": 0.0002832406078568186, + "loss": 0.0082, + "step": 3633 + }, + { + "epoch": 1.7101176470588235, + "grad_norm": 0.10035054846505732, + "learning_rate": 0.00028322046732203165, + "loss": 0.0595, + "step": 3634 + }, + { + "epoch": 1.7105882352941175, + "grad_norm": 0.04214068733836579, + "learning_rate": 0.0002832003154095633, + "loss": 0.0134, + "step": 3635 + }, + { + "epoch": 1.7110588235294117, + "grad_norm": 0.000299074644690832, + "learning_rate": 0.0002831801521211347, + "loss": 0.0, + "step": 3636 + }, + { + "epoch": 1.711529411764706, + "grad_norm": 0.07727852735952973, + "learning_rate": 0.0002831599774584679, + "loss": 0.0623, + "step": 3637 + }, + { + "epoch": 1.712, + "grad_norm": 0.0719408747522811, + "learning_rate": 0.0002831397914232858, + "loss": 0.0341, + "step": 3638 + }, + { + "epoch": 1.712470588235294, + "grad_norm": 0.04733118492819853, + "learning_rate": 0.0002831195940173125, + "loss": 0.0206, + "step": 3639 + }, + { + "epoch": 1.7129411764705882, + "grad_norm": 0.07515970088101556, + "learning_rate": 0.00028309938524227296, + "loss": 0.032, + "step": 3640 + }, + { + "epoch": 1.7134117647058824, + "grad_norm": 0.07672501707106137, + "learning_rate": 0.000283079165099893, + "loss": 0.0461, + "step": 3641 + }, + { + "epoch": 1.7138823529411766, + "grad_norm": 0.06934149539894388, + "learning_rate": 0.0002830589335918997, + "loss": 0.0365, + "step": 3642 + }, + { + "epoch": 1.7143529411764706, + "grad_norm": 0.026129251201111755, + "learning_rate": 0.0002830386907200207, + "loss": 0.009, + "step": 3643 + }, + { + "epoch": 1.7148235294117646, + "grad_norm": 2.3739102296055337e-06, + "learning_rate": 0.000283018436485985, + "loss": 0.0, + "step": 3644 + }, + { + "epoch": 1.7152941176470589, + "grad_norm": 0.12769270284062484, + "learning_rate": 0.0002829981708915224, + "loss": 0.0701, + "step": 3645 + }, + { + "epoch": 1.715764705882353, + "grad_norm": 0.04356135104799428, + "learning_rate": 0.0002829778939383636, + "loss": 0.0185, + "step": 3646 + }, + { + "epoch": 1.716235294117647, + "grad_norm": 2.9137532831612768e-06, + "learning_rate": 0.0002829576056282405, + "loss": 0.0, + "step": 3647 + }, + { + "epoch": 1.716705882352941, + "grad_norm": 0.1265411627729199, + "learning_rate": 0.00028293730596288565, + "loss": 0.0879, + "step": 3648 + }, + { + "epoch": 1.7171764705882353, + "grad_norm": 0.14757021062795445, + "learning_rate": 0.00028291699494403284, + "loss": 0.0382, + "step": 3649 + }, + { + "epoch": 1.7176470588235295, + "grad_norm": 0.0956663819788784, + "learning_rate": 0.0002828966725734167, + "loss": 0.095, + "step": 3650 + }, + { + "epoch": 1.7181176470588235, + "grad_norm": 0.15372825203011892, + "learning_rate": 0.0002828763388527729, + "loss": 0.1292, + "step": 3651 + }, + { + "epoch": 1.7185882352941175, + "grad_norm": 0.027251690785049205, + "learning_rate": 0.000282855993783838, + "loss": 0.0068, + "step": 3652 + }, + { + "epoch": 1.7190588235294118, + "grad_norm": 0.08145498840278517, + "learning_rate": 0.0002828356373683495, + "loss": 0.043, + "step": 3653 + }, + { + "epoch": 1.719529411764706, + "grad_norm": 0.07366465921285115, + "learning_rate": 0.0002828152696080461, + "loss": 0.0484, + "step": 3654 + }, + { + "epoch": 1.72, + "grad_norm": 0.05844205175430255, + "learning_rate": 0.0002827948905046672, + "loss": 0.0361, + "step": 3655 + }, + { + "epoch": 1.720470588235294, + "grad_norm": 0.10483516494801846, + "learning_rate": 0.00028277450005995326, + "loss": 0.0557, + "step": 3656 + }, + { + "epoch": 1.7209411764705882, + "grad_norm": 2.3971522259132547e-06, + "learning_rate": 0.0002827540982756458, + "loss": 0.0, + "step": 3657 + }, + { + "epoch": 1.7214117647058824, + "grad_norm": 0.11105111282906903, + "learning_rate": 0.00028273368515348713, + "loss": 0.0895, + "step": 3658 + }, + { + "epoch": 1.7218823529411764, + "grad_norm": 0.12749257159554808, + "learning_rate": 0.00028271326069522076, + "loss": 0.1314, + "step": 3659 + }, + { + "epoch": 1.7223529411764706, + "grad_norm": 3.3018103022332714e-06, + "learning_rate": 0.0002826928249025909, + "loss": 0.0, + "step": 3660 + }, + { + "epoch": 1.7228235294117646, + "grad_norm": 2.1064506943212747e-06, + "learning_rate": 0.000282672377777343, + "loss": 0.0, + "step": 3661 + }, + { + "epoch": 1.7232941176470589, + "grad_norm": 0.0924308868451162, + "learning_rate": 0.0002826519193212233, + "loss": 0.0613, + "step": 3662 + }, + { + "epoch": 1.723764705882353, + "grad_norm": 0.053523068498996906, + "learning_rate": 0.000282631449535979, + "loss": 0.023, + "step": 3663 + }, + { + "epoch": 1.724235294117647, + "grad_norm": 0.09685994965718665, + "learning_rate": 0.00028261096842335836, + "loss": 0.066, + "step": 3664 + }, + { + "epoch": 1.724705882352941, + "grad_norm": 0.05710035163149856, + "learning_rate": 0.00028259047598511053, + "loss": 0.0341, + "step": 3665 + }, + { + "epoch": 1.7251764705882353, + "grad_norm": 0.07564228279362264, + "learning_rate": 0.0002825699722229858, + "loss": 0.0351, + "step": 3666 + }, + { + "epoch": 1.7256470588235295, + "grad_norm": 0.0764487454378244, + "learning_rate": 0.0002825494571387351, + "loss": 0.0293, + "step": 3667 + }, + { + "epoch": 1.7261176470588235, + "grad_norm": 0.06290126214028206, + "learning_rate": 0.00028252893073411065, + "loss": 0.0421, + "step": 3668 + }, + { + "epoch": 1.7265882352941175, + "grad_norm": 0.09925327845931389, + "learning_rate": 0.00028250839301086544, + "loss": 0.0844, + "step": 3669 + }, + { + "epoch": 1.7270588235294118, + "grad_norm": 0.0907857824082925, + "learning_rate": 0.0002824878439707535, + "loss": 0.0515, + "step": 3670 + }, + { + "epoch": 1.727529411764706, + "grad_norm": 0.04838102880703994, + "learning_rate": 0.0002824672836155299, + "loss": 0.015, + "step": 3671 + }, + { + "epoch": 1.728, + "grad_norm": 0.08906241547149724, + "learning_rate": 0.0002824467119469505, + "loss": 0.0701, + "step": 3672 + }, + { + "epoch": 1.728470588235294, + "grad_norm": 3.4105929522203254e-06, + "learning_rate": 0.0002824261289667722, + "loss": 0.0, + "step": 3673 + }, + { + "epoch": 1.7289411764705882, + "grad_norm": 2.2093251135323387e-06, + "learning_rate": 0.000282405534676753, + "loss": 0.0, + "step": 3674 + }, + { + "epoch": 1.7294117647058824, + "grad_norm": 0.04865033449820954, + "learning_rate": 0.0002823849290786517, + "loss": 0.0193, + "step": 3675 + }, + { + "epoch": 1.7298823529411764, + "grad_norm": 0.07493450189424343, + "learning_rate": 0.0002823643121742281, + "loss": 0.0652, + "step": 3676 + }, + { + "epoch": 1.7303529411764704, + "grad_norm": 0.06958832208533232, + "learning_rate": 0.000282343683965243, + "loss": 0.0619, + "step": 3677 + }, + { + "epoch": 1.7308235294117646, + "grad_norm": 1.8791227965085065e-06, + "learning_rate": 0.0002823230444534581, + "loss": 0.0, + "step": 3678 + }, + { + "epoch": 1.7312941176470589, + "grad_norm": 0.050572790978989465, + "learning_rate": 0.0002823023936406362, + "loss": 0.0295, + "step": 3679 + }, + { + "epoch": 1.731764705882353, + "grad_norm": 0.08560623687453806, + "learning_rate": 0.0002822817315285409, + "loss": 0.0664, + "step": 3680 + }, + { + "epoch": 1.732235294117647, + "grad_norm": 0.11059671676339808, + "learning_rate": 0.000282261058118937, + "loss": 0.0933, + "step": 3681 + }, + { + "epoch": 1.732705882352941, + "grad_norm": 0.07517568599156241, + "learning_rate": 0.00028224037341358993, + "loss": 0.0307, + "step": 3682 + }, + { + "epoch": 1.7331764705882353, + "grad_norm": 0.01623890022539191, + "learning_rate": 0.0002822196774142663, + "loss": 0.0033, + "step": 3683 + }, + { + "epoch": 1.7336470588235295, + "grad_norm": 1.4593265187541215e-06, + "learning_rate": 0.0002821989701227337, + "loss": 0.0, + "step": 3684 + }, + { + "epoch": 1.7341176470588235, + "grad_norm": 1.7907005284169745e-06, + "learning_rate": 0.00028217825154076067, + "loss": 0.0, + "step": 3685 + }, + { + "epoch": 1.7345882352941175, + "grad_norm": 0.0672811459538417, + "learning_rate": 0.0002821575216701166, + "loss": 0.0336, + "step": 3686 + }, + { + "epoch": 1.7350588235294118, + "grad_norm": 2.134700010135642e-06, + "learning_rate": 0.00028213678051257194, + "loss": 0.0, + "step": 3687 + }, + { + "epoch": 1.735529411764706, + "grad_norm": 0.15396500424094128, + "learning_rate": 0.00028211602806989814, + "loss": 0.123, + "step": 3688 + }, + { + "epoch": 1.736, + "grad_norm": 0.0818439490681915, + "learning_rate": 0.00028209526434386744, + "loss": 0.0562, + "step": 3689 + }, + { + "epoch": 1.736470588235294, + "grad_norm": 0.10453400204597403, + "learning_rate": 0.0002820744893362533, + "loss": 0.0526, + "step": 3690 + }, + { + "epoch": 1.7369411764705882, + "grad_norm": 0.09352822646068157, + "learning_rate": 0.0002820537030488299, + "loss": 0.0755, + "step": 3691 + }, + { + "epoch": 1.7374117647058824, + "grad_norm": 0.09862458225917553, + "learning_rate": 0.00028203290548337257, + "loss": 0.0652, + "step": 3692 + }, + { + "epoch": 1.7378823529411764, + "grad_norm": 0.1030980885196716, + "learning_rate": 0.0002820120966416575, + "loss": 0.0499, + "step": 3693 + }, + { + "epoch": 1.7383529411764704, + "grad_norm": 0.059332346395487, + "learning_rate": 0.0002819912765254619, + "loss": 0.0337, + "step": 3694 + }, + { + "epoch": 1.7388235294117647, + "grad_norm": 0.060641067538615354, + "learning_rate": 0.0002819704451365638, + "loss": 0.0346, + "step": 3695 + }, + { + "epoch": 1.7392941176470589, + "grad_norm": 0.03183871042885548, + "learning_rate": 0.0002819496024767424, + "loss": 0.0085, + "step": 3696 + }, + { + "epoch": 1.739764705882353, + "grad_norm": 0.06840735594693831, + "learning_rate": 0.0002819287485477778, + "loss": 0.0306, + "step": 3697 + }, + { + "epoch": 1.740235294117647, + "grad_norm": 0.058738571461500105, + "learning_rate": 0.0002819078833514509, + "loss": 0.0287, + "step": 3698 + }, + { + "epoch": 1.740705882352941, + "grad_norm": 0.09472675601412213, + "learning_rate": 0.0002818870068895437, + "loss": 0.0463, + "step": 3699 + }, + { + "epoch": 1.7411764705882353, + "grad_norm": 0.1621481312962005, + "learning_rate": 0.0002818661191638393, + "loss": 0.1414, + "step": 3700 + }, + { + "epoch": 1.7416470588235295, + "grad_norm": 0.09036970922582505, + "learning_rate": 0.0002818452201761215, + "loss": 0.0656, + "step": 3701 + }, + { + "epoch": 1.7421176470588235, + "grad_norm": 0.055735011106954696, + "learning_rate": 0.0002818243099281752, + "loss": 0.0297, + "step": 3702 + }, + { + "epoch": 1.7425882352941175, + "grad_norm": 0.0708386112363894, + "learning_rate": 0.00028180338842178623, + "loss": 0.0517, + "step": 3703 + }, + { + "epoch": 1.7430588235294118, + "grad_norm": 0.06435404049830852, + "learning_rate": 0.0002817824556587414, + "loss": 0.0331, + "step": 3704 + }, + { + "epoch": 1.743529411764706, + "grad_norm": 0.04057699426043614, + "learning_rate": 0.0002817615116408285, + "loss": 0.0149, + "step": 3705 + }, + { + "epoch": 1.744, + "grad_norm": 0.035755224098386586, + "learning_rate": 0.0002817405563698362, + "loss": 0.0031, + "step": 3706 + }, + { + "epoch": 1.744470588235294, + "grad_norm": 1.407095982742067e-06, + "learning_rate": 0.0002817195898475542, + "loss": 0.0, + "step": 3707 + }, + { + "epoch": 1.7449411764705882, + "grad_norm": 2.123115750485215e-06, + "learning_rate": 0.00028169861207577313, + "loss": 0.0, + "step": 3708 + }, + { + "epoch": 1.7454117647058824, + "grad_norm": 2.874163297490229e-06, + "learning_rate": 0.0002816776230562846, + "loss": 0.0, + "step": 3709 + }, + { + "epoch": 1.7458823529411764, + "grad_norm": 0.08130215240346196, + "learning_rate": 0.0002816566227908812, + "loss": 0.0415, + "step": 3710 + }, + { + "epoch": 1.7463529411764704, + "grad_norm": 0.08250406132444603, + "learning_rate": 0.00028163561128135645, + "loss": 0.0578, + "step": 3711 + }, + { + "epoch": 1.7468235294117647, + "grad_norm": 0.08281863534120834, + "learning_rate": 0.0002816145885295048, + "loss": 0.039, + "step": 3712 + }, + { + "epoch": 1.7472941176470589, + "grad_norm": 0.08533469334990001, + "learning_rate": 0.0002815935545371218, + "loss": 0.0345, + "step": 3713 + }, + { + "epoch": 1.7477647058823529, + "grad_norm": 0.07830928433701666, + "learning_rate": 0.00028157250930600374, + "loss": 0.0341, + "step": 3714 + }, + { + "epoch": 1.748235294117647, + "grad_norm": 0.12177275698407591, + "learning_rate": 0.00028155145283794797, + "loss": 0.1031, + "step": 3715 + }, + { + "epoch": 1.748705882352941, + "grad_norm": 0.09754724113922197, + "learning_rate": 0.0002815303851347529, + "loss": 0.0401, + "step": 3716 + }, + { + "epoch": 1.7491764705882353, + "grad_norm": 0.03134542476883126, + "learning_rate": 0.0002815093061982178, + "loss": 0.0087, + "step": 3717 + }, + { + "epoch": 1.7496470588235296, + "grad_norm": 0.15325527756871038, + "learning_rate": 0.0002814882160301429, + "loss": 0.1219, + "step": 3718 + }, + { + "epoch": 1.7501176470588236, + "grad_norm": 0.0876737977695803, + "learning_rate": 0.00028146711463232943, + "loss": 0.0617, + "step": 3719 + }, + { + "epoch": 1.7505882352941176, + "grad_norm": 0.016117273487504736, + "learning_rate": 0.0002814460020065795, + "loss": 0.0027, + "step": 3720 + }, + { + "epoch": 1.7510588235294118, + "grad_norm": 4.139794705982137e-06, + "learning_rate": 0.00028142487815469627, + "loss": 0.0, + "step": 3721 + }, + { + "epoch": 1.751529411764706, + "grad_norm": 0.10427249609980402, + "learning_rate": 0.00028140374307848384, + "loss": 0.065, + "step": 3722 + }, + { + "epoch": 1.752, + "grad_norm": 0.10607404003557332, + "learning_rate": 0.00028138259677974714, + "loss": 0.0627, + "step": 3723 + }, + { + "epoch": 1.752470588235294, + "grad_norm": 0.08105191849329813, + "learning_rate": 0.0002813614392602923, + "loss": 0.039, + "step": 3724 + }, + { + "epoch": 1.7529411764705882, + "grad_norm": 0.04366339333936511, + "learning_rate": 0.00028134027052192624, + "loss": 0.0176, + "step": 3725 + }, + { + "epoch": 1.7534117647058824, + "grad_norm": 0.06451591481019175, + "learning_rate": 0.0002813190905664568, + "loss": 0.0323, + "step": 3726 + }, + { + "epoch": 1.7538823529411764, + "grad_norm": 0.092047445183679, + "learning_rate": 0.0002812978993956929, + "loss": 0.0419, + "step": 3727 + }, + { + "epoch": 1.7543529411764704, + "grad_norm": 0.02877127494339614, + "learning_rate": 0.0002812766970114445, + "loss": 0.0105, + "step": 3728 + }, + { + "epoch": 1.7548235294117647, + "grad_norm": 1.4143615917205045e-06, + "learning_rate": 0.00028125548341552213, + "loss": 0.0, + "step": 3729 + }, + { + "epoch": 1.755294117647059, + "grad_norm": 0.1184477045581902, + "learning_rate": 0.00028123425860973773, + "loss": 0.1071, + "step": 3730 + }, + { + "epoch": 1.755764705882353, + "grad_norm": 2.207927906977265e-06, + "learning_rate": 0.000281213022595904, + "loss": 0.0, + "step": 3731 + }, + { + "epoch": 1.756235294117647, + "grad_norm": 0.09971147450094658, + "learning_rate": 0.0002811917753758345, + "loss": 0.0866, + "step": 3732 + }, + { + "epoch": 1.7567058823529411, + "grad_norm": 3.0497305648697263e-06, + "learning_rate": 0.00028117051695134384, + "loss": 0.0, + "step": 3733 + }, + { + "epoch": 1.7571764705882353, + "grad_norm": 0.1503838066787574, + "learning_rate": 0.0002811492473242477, + "loss": 0.1201, + "step": 3734 + }, + { + "epoch": 1.7576470588235296, + "grad_norm": 1.9697679682958487e-06, + "learning_rate": 0.00028112796649636254, + "loss": 0.0, + "step": 3735 + }, + { + "epoch": 1.7581176470588236, + "grad_norm": 0.08195214369565955, + "learning_rate": 0.0002811066744695059, + "loss": 0.0498, + "step": 3736 + }, + { + "epoch": 1.7585882352941176, + "grad_norm": 0.16045978126775437, + "learning_rate": 0.0002810853712454961, + "loss": 0.1492, + "step": 3737 + }, + { + "epoch": 1.7590588235294118, + "grad_norm": 0.060771449291090275, + "learning_rate": 0.00028106405682615265, + "loss": 0.0307, + "step": 3738 + }, + { + "epoch": 1.759529411764706, + "grad_norm": 0.07405930153889581, + "learning_rate": 0.0002810427312132959, + "loss": 0.0425, + "step": 3739 + }, + { + "epoch": 1.76, + "grad_norm": 0.06475026336880861, + "learning_rate": 0.00028102139440874715, + "loss": 0.0296, + "step": 3740 + }, + { + "epoch": 1.760470588235294, + "grad_norm": 0.029602688318026683, + "learning_rate": 0.0002810000464143287, + "loss": 0.0059, + "step": 3741 + }, + { + "epoch": 1.7609411764705882, + "grad_norm": 8.895988091951759e-06, + "learning_rate": 0.0002809786872318637, + "loss": 0.0, + "step": 3742 + }, + { + "epoch": 1.7614117647058825, + "grad_norm": 3.327521919017722e-06, + "learning_rate": 0.00028095731686317635, + "loss": 0.0, + "step": 3743 + }, + { + "epoch": 1.7618823529411765, + "grad_norm": 0.10564851772788011, + "learning_rate": 0.0002809359353100918, + "loss": 0.0424, + "step": 3744 + }, + { + "epoch": 1.7623529411764705, + "grad_norm": 0.0815662874121778, + "learning_rate": 0.0002809145425744361, + "loss": 0.0315, + "step": 3745 + }, + { + "epoch": 1.7628235294117647, + "grad_norm": 0.16412509313693158, + "learning_rate": 0.00028089313865803644, + "loss": 0.0891, + "step": 3746 + }, + { + "epoch": 1.763294117647059, + "grad_norm": 0.056555391987779806, + "learning_rate": 0.0002808717235627206, + "loss": 0.0276, + "step": 3747 + }, + { + "epoch": 1.763764705882353, + "grad_norm": 0.06095274041882474, + "learning_rate": 0.0002808502972903177, + "loss": 0.0279, + "step": 3748 + }, + { + "epoch": 1.764235294117647, + "grad_norm": 0.1332774175263717, + "learning_rate": 0.0002808288598426576, + "loss": 0.0844, + "step": 3749 + }, + { + "epoch": 1.7647058823529411, + "grad_norm": 0.07693946752428782, + "learning_rate": 0.0002808074112215711, + "loss": 0.0438, + "step": 3750 + }, + { + "epoch": 1.7651764705882353, + "grad_norm": 0.09145220133237707, + "learning_rate": 0.00028078595142889013, + "loss": 0.079, + "step": 3751 + }, + { + "epoch": 1.7656470588235296, + "grad_norm": 0.028762189776962984, + "learning_rate": 0.0002807644804664474, + "loss": 0.0099, + "step": 3752 + }, + { + "epoch": 1.7661176470588236, + "grad_norm": 0.0624155611359708, + "learning_rate": 0.0002807429983360766, + "loss": 0.0353, + "step": 3753 + }, + { + "epoch": 1.7665882352941176, + "grad_norm": 0.05256812420788908, + "learning_rate": 0.0002807215050396125, + "loss": 0.0247, + "step": 3754 + }, + { + "epoch": 1.7670588235294118, + "grad_norm": 0.058435027962748236, + "learning_rate": 0.0002807000005788907, + "loss": 0.0238, + "step": 3755 + }, + { + "epoch": 1.767529411764706, + "grad_norm": 3.119837163031776e-06, + "learning_rate": 0.00028067848495574773, + "loss": 0.0, + "step": 3756 + }, + { + "epoch": 1.768, + "grad_norm": 0.12911660094885624, + "learning_rate": 0.0002806569581720212, + "loss": 0.0705, + "step": 3757 + }, + { + "epoch": 1.768470588235294, + "grad_norm": 2.6621598015152356e-06, + "learning_rate": 0.00028063542022954956, + "loss": 0.0, + "step": 3758 + }, + { + "epoch": 1.7689411764705882, + "grad_norm": 3.5564831595156815e-06, + "learning_rate": 0.00028061387113017226, + "loss": 0.0, + "step": 3759 + }, + { + "epoch": 1.7694117647058825, + "grad_norm": 0.13160945372386965, + "learning_rate": 0.0002805923108757297, + "loss": 0.0939, + "step": 3760 + }, + { + "epoch": 1.7698823529411765, + "grad_norm": 1.5991488160212185e-06, + "learning_rate": 0.00028057073946806323, + "loss": 0.0, + "step": 3761 + }, + { + "epoch": 1.7703529411764705, + "grad_norm": 0.041803156702525, + "learning_rate": 0.00028054915690901524, + "loss": 0.0073, + "step": 3762 + }, + { + "epoch": 1.7708235294117647, + "grad_norm": 0.08933537219212118, + "learning_rate": 0.0002805275632004288, + "loss": 0.0448, + "step": 3763 + }, + { + "epoch": 1.771294117647059, + "grad_norm": 0.09397653937988613, + "learning_rate": 0.00028050595834414835, + "loss": 0.0748, + "step": 3764 + }, + { + "epoch": 1.771764705882353, + "grad_norm": 0.06640261910568289, + "learning_rate": 0.00028048434234201883, + "loss": 0.031, + "step": 3765 + }, + { + "epoch": 1.772235294117647, + "grad_norm": 0.045877164404584225, + "learning_rate": 0.00028046271519588644, + "loss": 0.0276, + "step": 3766 + }, + { + "epoch": 1.7727058823529411, + "grad_norm": 0.08122712072285197, + "learning_rate": 0.0002804410769075984, + "loss": 0.0449, + "step": 3767 + }, + { + "epoch": 1.7731764705882354, + "grad_norm": 0.05586760952158681, + "learning_rate": 0.0002804194274790024, + "loss": 0.0154, + "step": 3768 + }, + { + "epoch": 1.7736470588235296, + "grad_norm": 0.17324543303188408, + "learning_rate": 0.00028039776691194773, + "loss": 0.1362, + "step": 3769 + }, + { + "epoch": 1.7741176470588236, + "grad_norm": 0.07332146670285511, + "learning_rate": 0.0002803760952082841, + "loss": 0.0439, + "step": 3770 + }, + { + "epoch": 1.7745882352941176, + "grad_norm": 0.03313098574866482, + "learning_rate": 0.0002803544123698624, + "loss": 0.0098, + "step": 3771 + }, + { + "epoch": 1.7750588235294118, + "grad_norm": 0.0777041214272742, + "learning_rate": 0.0002803327183985346, + "loss": 0.0423, + "step": 3772 + }, + { + "epoch": 1.775529411764706, + "grad_norm": 0.04813879155056246, + "learning_rate": 0.00028031101329615334, + "loss": 0.0222, + "step": 3773 + }, + { + "epoch": 1.776, + "grad_norm": 0.09689321754940854, + "learning_rate": 0.00028028929706457235, + "loss": 0.0656, + "step": 3774 + }, + { + "epoch": 1.776470588235294, + "grad_norm": 0.05909081146959473, + "learning_rate": 0.0002802675697056463, + "loss": 0.0215, + "step": 3775 + }, + { + "epoch": 1.7769411764705882, + "grad_norm": 1.4615180315066172e-06, + "learning_rate": 0.00028024583122123094, + "loss": 0.0, + "step": 3776 + }, + { + "epoch": 1.7774117647058825, + "grad_norm": 1.6436642041503682e-06, + "learning_rate": 0.00028022408161318266, + "loss": 0.0, + "step": 3777 + }, + { + "epoch": 1.7778823529411765, + "grad_norm": 0.18495454930955937, + "learning_rate": 0.00028020232088335907, + "loss": 0.161, + "step": 3778 + }, + { + "epoch": 1.7783529411764705, + "grad_norm": 0.05923744696266553, + "learning_rate": 0.00028018054903361864, + "loss": 0.0263, + "step": 3779 + }, + { + "epoch": 1.7788235294117647, + "grad_norm": 0.0428023388426644, + "learning_rate": 0.00028015876606582083, + "loss": 0.0151, + "step": 3780 + }, + { + "epoch": 1.779294117647059, + "grad_norm": 0.1125694813899435, + "learning_rate": 0.0002801369719818259, + "loss": 0.0634, + "step": 3781 + }, + { + "epoch": 1.779764705882353, + "grad_norm": 0.03888012871242431, + "learning_rate": 0.0002801151667834953, + "loss": 0.0131, + "step": 3782 + }, + { + "epoch": 1.780235294117647, + "grad_norm": 0.015065418243257071, + "learning_rate": 0.00028009335047269126, + "loss": 0.0022, + "step": 3783 + }, + { + "epoch": 1.7807058823529411, + "grad_norm": 0.0635905378712612, + "learning_rate": 0.00028007152305127694, + "loss": 0.0231, + "step": 3784 + }, + { + "epoch": 1.7811764705882354, + "grad_norm": 0.09313458913171115, + "learning_rate": 0.0002800496845211166, + "loss": 0.0589, + "step": 3785 + }, + { + "epoch": 1.7816470588235294, + "grad_norm": 0.05614854118749223, + "learning_rate": 0.0002800278348840753, + "loss": 0.0266, + "step": 3786 + }, + { + "epoch": 1.7821176470588236, + "grad_norm": 2.0928689124869857e-06, + "learning_rate": 0.00028000597414201906, + "loss": 0.0, + "step": 3787 + }, + { + "epoch": 1.7825882352941176, + "grad_norm": 0.07668752773773138, + "learning_rate": 0.000279984102296815, + "loss": 0.039, + "step": 3788 + }, + { + "epoch": 1.7830588235294118, + "grad_norm": 1.289392307812012e-06, + "learning_rate": 0.000279962219350331, + "loss": 0.0, + "step": 3789 + }, + { + "epoch": 1.783529411764706, + "grad_norm": 1.4270394985788948e-05, + "learning_rate": 0.00027994032530443606, + "loss": 0.0, + "step": 3790 + }, + { + "epoch": 1.784, + "grad_norm": 4.7334376286568024e-05, + "learning_rate": 0.00027991842016099994, + "loss": 0.0, + "step": 3791 + }, + { + "epoch": 1.784470588235294, + "grad_norm": 0.059529193361606565, + "learning_rate": 0.0002798965039218936, + "loss": 0.0206, + "step": 3792 + }, + { + "epoch": 1.7849411764705883, + "grad_norm": 0.029619160670012006, + "learning_rate": 0.0002798745765889886, + "loss": 0.0084, + "step": 3793 + }, + { + "epoch": 1.7854117647058825, + "grad_norm": 0.09922445191438535, + "learning_rate": 0.00027985263816415767, + "loss": 0.0455, + "step": 3794 + }, + { + "epoch": 1.7858823529411765, + "grad_norm": 0.03319709992357642, + "learning_rate": 0.0002798306886492746, + "loss": 0.0079, + "step": 3795 + }, + { + "epoch": 1.7863529411764705, + "grad_norm": 0.15658639542362013, + "learning_rate": 0.00027980872804621385, + "loss": 0.1054, + "step": 3796 + }, + { + "epoch": 1.7868235294117647, + "grad_norm": 0.061663449209841584, + "learning_rate": 0.0002797867563568511, + "loss": 0.0234, + "step": 3797 + }, + { + "epoch": 1.787294117647059, + "grad_norm": 0.19344221830898276, + "learning_rate": 0.0002797647735830627, + "loss": 0.1643, + "step": 3798 + }, + { + "epoch": 1.787764705882353, + "grad_norm": 0.00011970735174896327, + "learning_rate": 0.00027974277972672614, + "loss": 0.0, + "step": 3799 + }, + { + "epoch": 1.788235294117647, + "grad_norm": 0.10389411313553717, + "learning_rate": 0.0002797207747897198, + "loss": 0.0683, + "step": 3800 + }, + { + "epoch": 1.7887058823529411, + "grad_norm": 0.05933322910272824, + "learning_rate": 0.000279698758773923, + "loss": 0.0238, + "step": 3801 + }, + { + "epoch": 1.7891764705882354, + "grad_norm": 0.10889954069938776, + "learning_rate": 0.0002796767316812161, + "loss": 0.0641, + "step": 3802 + }, + { + "epoch": 1.7896470588235294, + "grad_norm": 0.029965322474825393, + "learning_rate": 0.0002796546935134802, + "loss": 0.0079, + "step": 3803 + }, + { + "epoch": 1.7901176470588234, + "grad_norm": 5.11848602382481e-06, + "learning_rate": 0.00027963264427259755, + "loss": 0.0, + "step": 3804 + }, + { + "epoch": 1.7905882352941176, + "grad_norm": 0.10230187290584393, + "learning_rate": 0.00027961058396045124, + "loss": 0.0371, + "step": 3805 + }, + { + "epoch": 1.7910588235294118, + "grad_norm": 0.0003780391411974592, + "learning_rate": 0.00027958851257892526, + "loss": 0.0, + "step": 3806 + }, + { + "epoch": 1.791529411764706, + "grad_norm": 0.059402347951934824, + "learning_rate": 0.0002795664301299047, + "loss": 0.0321, + "step": 3807 + }, + { + "epoch": 1.792, + "grad_norm": 0.0780459443555182, + "learning_rate": 0.00027954433661527554, + "loss": 0.0342, + "step": 3808 + }, + { + "epoch": 1.792470588235294, + "grad_norm": 0.134370801045366, + "learning_rate": 0.0002795222320369246, + "loss": 0.1174, + "step": 3809 + }, + { + "epoch": 1.7929411764705883, + "grad_norm": 0.052169309519613384, + "learning_rate": 0.00027950011639673976, + "loss": 0.0215, + "step": 3810 + }, + { + "epoch": 1.7934117647058825, + "grad_norm": 0.07636701884745942, + "learning_rate": 0.00027947798969660975, + "loss": 0.0327, + "step": 3811 + }, + { + "epoch": 1.7938823529411765, + "grad_norm": 0.10548733736139794, + "learning_rate": 0.00027945585193842433, + "loss": 0.0771, + "step": 3812 + }, + { + "epoch": 1.7943529411764705, + "grad_norm": 0.07179985810613432, + "learning_rate": 0.00027943370312407415, + "loss": 0.0358, + "step": 3813 + }, + { + "epoch": 1.7948235294117647, + "grad_norm": 0.03351145196599308, + "learning_rate": 0.00027941154325545096, + "loss": 0.0091, + "step": 3814 + }, + { + "epoch": 1.795294117647059, + "grad_norm": 2.5920463554085775e-05, + "learning_rate": 0.00027938937233444713, + "loss": 0.0, + "step": 3815 + }, + { + "epoch": 1.795764705882353, + "grad_norm": 0.04950650201376983, + "learning_rate": 0.00027936719036295623, + "loss": 0.0177, + "step": 3816 + }, + { + "epoch": 1.796235294117647, + "grad_norm": 1.949767053480024e-06, + "learning_rate": 0.0002793449973428728, + "loss": 0.0, + "step": 3817 + }, + { + "epoch": 1.7967058823529412, + "grad_norm": 2.2095944863286422e-05, + "learning_rate": 0.0002793227932760921, + "loss": 0.0, + "step": 3818 + }, + { + "epoch": 1.7971764705882354, + "grad_norm": 0.13458251054331596, + "learning_rate": 0.0002793005781645106, + "loss": 0.0804, + "step": 3819 + }, + { + "epoch": 1.7976470588235294, + "grad_norm": 0.073072397411039, + "learning_rate": 0.00027927835201002545, + "loss": 0.0374, + "step": 3820 + }, + { + "epoch": 1.7981176470588234, + "grad_norm": 0.1785625404473734, + "learning_rate": 0.00027925611481453497, + "loss": 0.1195, + "step": 3821 + }, + { + "epoch": 1.7985882352941176, + "grad_norm": 0.10040565906211629, + "learning_rate": 0.0002792338665799383, + "loss": 0.0581, + "step": 3822 + }, + { + "epoch": 1.7990588235294118, + "grad_norm": 0.06184500845867469, + "learning_rate": 0.00027921160730813544, + "loss": 0.0259, + "step": 3823 + }, + { + "epoch": 1.799529411764706, + "grad_norm": 0.17276090895110646, + "learning_rate": 0.0002791893370010276, + "loss": 0.1117, + "step": 3824 + }, + { + "epoch": 1.8, + "grad_norm": 0.11999508972909627, + "learning_rate": 0.00027916705566051676, + "loss": 0.093, + "step": 3825 + }, + { + "epoch": 1.800470588235294, + "grad_norm": 0.1252958429101781, + "learning_rate": 0.00027914476328850575, + "loss": 0.0787, + "step": 3826 + }, + { + "epoch": 1.8009411764705883, + "grad_norm": 1.2357608971035686e-06, + "learning_rate": 0.0002791224598868985, + "loss": 0.0, + "step": 3827 + }, + { + "epoch": 1.8014117647058825, + "grad_norm": 0.13320985750369277, + "learning_rate": 0.00027910014545759986, + "loss": 0.0727, + "step": 3828 + }, + { + "epoch": 1.8018823529411765, + "grad_norm": 0.15492605020423492, + "learning_rate": 0.00027907782000251557, + "loss": 0.1123, + "step": 3829 + }, + { + "epoch": 1.8023529411764705, + "grad_norm": 0.10831438801599283, + "learning_rate": 0.00027905548352355226, + "loss": 0.072, + "step": 3830 + }, + { + "epoch": 1.8028235294117647, + "grad_norm": 0.08102591083320475, + "learning_rate": 0.00027903313602261774, + "loss": 0.0577, + "step": 3831 + }, + { + "epoch": 1.803294117647059, + "grad_norm": 0.06226795669316474, + "learning_rate": 0.00027901077750162043, + "loss": 0.0476, + "step": 3832 + }, + { + "epoch": 1.803764705882353, + "grad_norm": 0.07826012104542875, + "learning_rate": 0.00027898840796246994, + "loss": 0.0584, + "step": 3833 + }, + { + "epoch": 1.804235294117647, + "grad_norm": 0.0817527514726544, + "learning_rate": 0.00027896602740707673, + "loss": 0.0424, + "step": 3834 + }, + { + "epoch": 1.8047058823529412, + "grad_norm": 2.4746732024865635e-06, + "learning_rate": 0.0002789436358373522, + "loss": 0.0, + "step": 3835 + }, + { + "epoch": 1.8051764705882354, + "grad_norm": 0.08494525149173907, + "learning_rate": 0.0002789212332552088, + "loss": 0.0538, + "step": 3836 + }, + { + "epoch": 1.8056470588235294, + "grad_norm": 0.1567345919780738, + "learning_rate": 0.0002788988196625596, + "loss": 0.069, + "step": 3837 + }, + { + "epoch": 1.8061176470588234, + "grad_norm": 0.06172516766479998, + "learning_rate": 0.000278876395061319, + "loss": 0.0297, + "step": 3838 + }, + { + "epoch": 1.8065882352941176, + "grad_norm": 0.038152933404477965, + "learning_rate": 0.0002788539594534021, + "loss": 0.013, + "step": 3839 + }, + { + "epoch": 1.8070588235294118, + "grad_norm": 0.10345769685749466, + "learning_rate": 0.0002788315128407251, + "loss": 0.0492, + "step": 3840 + }, + { + "epoch": 1.8075294117647058, + "grad_norm": 0.07407108546903597, + "learning_rate": 0.0002788090552252049, + "loss": 0.0573, + "step": 3841 + }, + { + "epoch": 1.808, + "grad_norm": 0.08731608321297075, + "learning_rate": 0.00027878658660875967, + "loss": 0.0593, + "step": 3842 + }, + { + "epoch": 1.808470588235294, + "grad_norm": 0.12543599296322652, + "learning_rate": 0.0002787641069933082, + "loss": 0.113, + "step": 3843 + }, + { + "epoch": 1.8089411764705883, + "grad_norm": 0.06510039911962534, + "learning_rate": 0.0002787416163807704, + "loss": 0.0329, + "step": 3844 + }, + { + "epoch": 1.8094117647058825, + "grad_norm": 0.07951385870484068, + "learning_rate": 0.0002787191147730671, + "loss": 0.0689, + "step": 3845 + }, + { + "epoch": 1.8098823529411765, + "grad_norm": 0.0816177600035321, + "learning_rate": 0.00027869660217212005, + "loss": 0.0505, + "step": 3846 + }, + { + "epoch": 1.8103529411764705, + "grad_norm": 0.1776111377922507, + "learning_rate": 0.0002786740785798519, + "loss": 0.1133, + "step": 3847 + }, + { + "epoch": 1.8108235294117647, + "grad_norm": 0.04783896778017388, + "learning_rate": 0.00027865154399818627, + "loss": 0.0226, + "step": 3848 + }, + { + "epoch": 1.811294117647059, + "grad_norm": 0.03944065673419671, + "learning_rate": 0.00027862899842904783, + "loss": 0.01, + "step": 3849 + }, + { + "epoch": 1.811764705882353, + "grad_norm": 0.0838940301653911, + "learning_rate": 0.00027860644187436195, + "loss": 0.0499, + "step": 3850 + }, + { + "epoch": 1.812235294117647, + "grad_norm": 0.0443141017598763, + "learning_rate": 0.0002785838743360551, + "loss": 0.0161, + "step": 3851 + }, + { + "epoch": 1.8127058823529412, + "grad_norm": 0.08146525079939505, + "learning_rate": 0.00027856129581605474, + "loss": 0.0531, + "step": 3852 + }, + { + "epoch": 1.8131764705882354, + "grad_norm": 0.057258137978544225, + "learning_rate": 0.0002785387063162891, + "loss": 0.0262, + "step": 3853 + }, + { + "epoch": 1.8136470588235294, + "grad_norm": 0.07763562265738108, + "learning_rate": 0.00027851610583868745, + "loss": 0.0414, + "step": 3854 + }, + { + "epoch": 1.8141176470588234, + "grad_norm": 0.03137849023559187, + "learning_rate": 0.00027849349438518003, + "loss": 0.0105, + "step": 3855 + }, + { + "epoch": 1.8145882352941176, + "grad_norm": 0.06212998294186622, + "learning_rate": 0.0002784708719576979, + "loss": 0.0367, + "step": 3856 + }, + { + "epoch": 1.8150588235294118, + "grad_norm": 0.09785479990046933, + "learning_rate": 0.00027844823855817315, + "loss": 0.0649, + "step": 3857 + }, + { + "epoch": 1.8155294117647058, + "grad_norm": 0.11123064259594252, + "learning_rate": 0.0002784255941885388, + "loss": 0.0595, + "step": 3858 + }, + { + "epoch": 1.8159999999999998, + "grad_norm": 0.11924226744712603, + "learning_rate": 0.0002784029388507288, + "loss": 0.0925, + "step": 3859 + }, + { + "epoch": 1.816470588235294, + "grad_norm": 0.1078490149911461, + "learning_rate": 0.000278380272546678, + "loss": 0.0471, + "step": 3860 + }, + { + "epoch": 1.8169411764705883, + "grad_norm": 0.1036490367578242, + "learning_rate": 0.00027835759527832226, + "loss": 0.0703, + "step": 3861 + }, + { + "epoch": 1.8174117647058825, + "grad_norm": 2.089042571404356e-06, + "learning_rate": 0.0002783349070475982, + "loss": 0.0, + "step": 3862 + }, + { + "epoch": 1.8178823529411765, + "grad_norm": 0.11293829114406335, + "learning_rate": 0.00027831220785644366, + "loss": 0.1031, + "step": 3863 + }, + { + "epoch": 1.8183529411764705, + "grad_norm": 0.08589670330189976, + "learning_rate": 0.0002782894977067972, + "loss": 0.0473, + "step": 3864 + }, + { + "epoch": 1.8188235294117647, + "grad_norm": 0.10587220096453383, + "learning_rate": 0.0002782667766005984, + "loss": 0.0584, + "step": 3865 + }, + { + "epoch": 1.819294117647059, + "grad_norm": 0.12533440921621933, + "learning_rate": 0.0002782440445397877, + "loss": 0.1154, + "step": 3866 + }, + { + "epoch": 1.819764705882353, + "grad_norm": 0.0180893238332068, + "learning_rate": 0.00027822130152630657, + "loss": 0.0032, + "step": 3867 + }, + { + "epoch": 1.820235294117647, + "grad_norm": 0.044256739159019905, + "learning_rate": 0.00027819854756209734, + "loss": 0.025, + "step": 3868 + }, + { + "epoch": 1.8207058823529412, + "grad_norm": 2.3398377890919956e-06, + "learning_rate": 0.00027817578264910343, + "loss": 0.0, + "step": 3869 + }, + { + "epoch": 1.8211764705882354, + "grad_norm": 0.03404806402065527, + "learning_rate": 0.00027815300678926886, + "loss": 0.0116, + "step": 3870 + }, + { + "epoch": 1.8216470588235294, + "grad_norm": 0.057782967222358865, + "learning_rate": 0.000278130219984539, + "loss": 0.0287, + "step": 3871 + }, + { + "epoch": 1.8221176470588234, + "grad_norm": 0.1236790263754779, + "learning_rate": 0.00027810742223685983, + "loss": 0.1043, + "step": 3872 + }, + { + "epoch": 1.8225882352941176, + "grad_norm": 0.059332276437966916, + "learning_rate": 0.00027808461354817847, + "loss": 0.035, + "step": 3873 + }, + { + "epoch": 1.8230588235294118, + "grad_norm": 0.029315274736101016, + "learning_rate": 0.00027806179392044286, + "loss": 0.0091, + "step": 3874 + }, + { + "epoch": 1.8235294117647058, + "grad_norm": 0.05259785681705308, + "learning_rate": 0.0002780389633556019, + "loss": 0.0171, + "step": 3875 + }, + { + "epoch": 1.8239999999999998, + "grad_norm": 0.06830965748468569, + "learning_rate": 0.00027801612185560554, + "loss": 0.0295, + "step": 3876 + }, + { + "epoch": 1.824470588235294, + "grad_norm": 0.032062378677259175, + "learning_rate": 0.00027799326942240434, + "loss": 0.011, + "step": 3877 + }, + { + "epoch": 1.8249411764705883, + "grad_norm": 3.27113814503085e-06, + "learning_rate": 0.0002779704060579501, + "loss": 0.0, + "step": 3878 + }, + { + "epoch": 1.8254117647058825, + "grad_norm": 1.8465907440303165e-06, + "learning_rate": 0.00027794753176419557, + "loss": 0.0, + "step": 3879 + }, + { + "epoch": 1.8258823529411765, + "grad_norm": 0.11625809442192703, + "learning_rate": 0.00027792464654309427, + "loss": 0.0904, + "step": 3880 + }, + { + "epoch": 1.8263529411764705, + "grad_norm": 4.217074017680206e-05, + "learning_rate": 0.0002779017503966006, + "loss": 0.0, + "step": 3881 + }, + { + "epoch": 1.8268235294117647, + "grad_norm": 0.0774292349919371, + "learning_rate": 0.0002778788433266702, + "loss": 0.0458, + "step": 3882 + }, + { + "epoch": 1.827294117647059, + "grad_norm": 0.12695415787513636, + "learning_rate": 0.00027785592533525924, + "loss": 0.08, + "step": 3883 + }, + { + "epoch": 1.827764705882353, + "grad_norm": 0.12292963898051228, + "learning_rate": 0.00027783299642432515, + "loss": 0.1195, + "step": 3884 + }, + { + "epoch": 1.828235294117647, + "grad_norm": 0.05650532574433524, + "learning_rate": 0.0002778100565958262, + "loss": 0.0227, + "step": 3885 + }, + { + "epoch": 1.8287058823529412, + "grad_norm": 0.047759208710642154, + "learning_rate": 0.0002777871058517215, + "loss": 0.0161, + "step": 3886 + }, + { + "epoch": 1.8291764705882354, + "grad_norm": 0.07974514305198362, + "learning_rate": 0.00027776414419397106, + "loss": 0.0532, + "step": 3887 + }, + { + "epoch": 1.8296470588235294, + "grad_norm": 0.07564996711671887, + "learning_rate": 0.00027774117162453617, + "loss": 0.0522, + "step": 3888 + }, + { + "epoch": 1.8301176470588234, + "grad_norm": 0.11317931567077985, + "learning_rate": 0.00027771818814537855, + "loss": 0.0885, + "step": 3889 + }, + { + "epoch": 1.8305882352941176, + "grad_norm": 2.174586213578824e-06, + "learning_rate": 0.00027769519375846126, + "loss": 0.0, + "step": 3890 + }, + { + "epoch": 1.8310588235294118, + "grad_norm": 8.031290001368371e-06, + "learning_rate": 0.0002776721884657481, + "loss": 0.0, + "step": 3891 + }, + { + "epoch": 1.8315294117647059, + "grad_norm": 0.147033252926329, + "learning_rate": 0.00027764917226920377, + "loss": 0.1036, + "step": 3892 + }, + { + "epoch": 1.8319999999999999, + "grad_norm": 0.01756794394564882, + "learning_rate": 0.000277626145170794, + "loss": 0.003, + "step": 3893 + }, + { + "epoch": 1.832470588235294, + "grad_norm": 0.09660132819874523, + "learning_rate": 0.0002776031071724855, + "loss": 0.0553, + "step": 3894 + }, + { + "epoch": 1.8329411764705883, + "grad_norm": 0.09061855453349606, + "learning_rate": 0.0002775800582762457, + "loss": 0.045, + "step": 3895 + }, + { + "epoch": 1.8334117647058825, + "grad_norm": 5.546057182285137e-06, + "learning_rate": 0.00027755699848404315, + "loss": 0.0, + "step": 3896 + }, + { + "epoch": 1.8338823529411765, + "grad_norm": 2.113996800155019e-06, + "learning_rate": 0.0002775339277978473, + "loss": 0.0, + "step": 3897 + }, + { + "epoch": 1.8343529411764705, + "grad_norm": 0.05769970255601821, + "learning_rate": 0.0002775108462196284, + "loss": 0.0224, + "step": 3898 + }, + { + "epoch": 1.8348235294117647, + "grad_norm": 2.006304840347525e-06, + "learning_rate": 0.0002774877537513579, + "loss": 0.0, + "step": 3899 + }, + { + "epoch": 1.835294117647059, + "grad_norm": 0.08384006114416179, + "learning_rate": 0.0002774646503950078, + "loss": 0.0484, + "step": 3900 + }, + { + "epoch": 1.835764705882353, + "grad_norm": 0.07016289640720379, + "learning_rate": 0.0002774415361525514, + "loss": 0.0388, + "step": 3901 + }, + { + "epoch": 1.836235294117647, + "grad_norm": 0.0690396755046296, + "learning_rate": 0.00027741841102596276, + "loss": 0.0329, + "step": 3902 + }, + { + "epoch": 1.8367058823529412, + "grad_norm": 0.117170566036028, + "learning_rate": 0.0002773952750172168, + "loss": 0.0704, + "step": 3903 + }, + { + "epoch": 1.8371764705882354, + "grad_norm": 0.14649694110498976, + "learning_rate": 0.0002773721281282894, + "loss": 0.0916, + "step": 3904 + }, + { + "epoch": 1.8376470588235294, + "grad_norm": 0.11570780236606934, + "learning_rate": 0.0002773489703611576, + "loss": 0.0772, + "step": 3905 + }, + { + "epoch": 1.8381176470588234, + "grad_norm": 2.151518399752565e-05, + "learning_rate": 0.00027732580171779906, + "loss": 0.0, + "step": 3906 + }, + { + "epoch": 1.8385882352941176, + "grad_norm": 0.143721465613425, + "learning_rate": 0.0002773026222001925, + "loss": 0.0927, + "step": 3907 + }, + { + "epoch": 1.8390588235294119, + "grad_norm": 0.16246356363005657, + "learning_rate": 0.00027727943181031766, + "loss": 0.1122, + "step": 3908 + }, + { + "epoch": 1.8395294117647059, + "grad_norm": 0.07913652364820029, + "learning_rate": 0.00027725623055015495, + "loss": 0.0554, + "step": 3909 + }, + { + "epoch": 1.8399999999999999, + "grad_norm": 0.052797246171968716, + "learning_rate": 0.000277233018421686, + "loss": 0.0224, + "step": 3910 + }, + { + "epoch": 1.840470588235294, + "grad_norm": 0.1597060179958277, + "learning_rate": 0.00027720979542689324, + "loss": 0.09, + "step": 3911 + }, + { + "epoch": 1.8409411764705883, + "grad_norm": 0.060695048046931904, + "learning_rate": 0.0002771865615677599, + "loss": 0.0313, + "step": 3912 + }, + { + "epoch": 1.8414117647058823, + "grad_norm": 0.17085437328151684, + "learning_rate": 0.00027716331684627044, + "loss": 0.1308, + "step": 3913 + }, + { + "epoch": 1.8418823529411765, + "grad_norm": 0.10188104244886512, + "learning_rate": 0.00027714006126441, + "loss": 0.0503, + "step": 3914 + }, + { + "epoch": 1.8423529411764705, + "grad_norm": 0.09835287830998844, + "learning_rate": 0.00027711679482416465, + "loss": 0.0551, + "step": 3915 + }, + { + "epoch": 1.8428235294117647, + "grad_norm": 0.09845519317112796, + "learning_rate": 0.0002770935175275215, + "loss": 0.0629, + "step": 3916 + }, + { + "epoch": 1.843294117647059, + "grad_norm": 0.0601579756941331, + "learning_rate": 0.0002770702293764686, + "loss": 0.0246, + "step": 3917 + }, + { + "epoch": 1.843764705882353, + "grad_norm": 0.09666287558465396, + "learning_rate": 0.0002770469303729949, + "loss": 0.0517, + "step": 3918 + }, + { + "epoch": 1.844235294117647, + "grad_norm": 0.10240423630769348, + "learning_rate": 0.0002770236205190901, + "loss": 0.0548, + "step": 3919 + }, + { + "epoch": 1.8447058823529412, + "grad_norm": 0.05145553033210022, + "learning_rate": 0.00027700029981674506, + "loss": 0.0252, + "step": 3920 + }, + { + "epoch": 1.8451764705882354, + "grad_norm": 0.10431907899030132, + "learning_rate": 0.0002769769682679515, + "loss": 0.0572, + "step": 3921 + }, + { + "epoch": 1.8456470588235294, + "grad_norm": 0.07460175729430224, + "learning_rate": 0.00027695362587470203, + "loss": 0.0469, + "step": 3922 + }, + { + "epoch": 1.8461176470588234, + "grad_norm": 0.09334742608670299, + "learning_rate": 0.0002769302726389902, + "loss": 0.048, + "step": 3923 + }, + { + "epoch": 1.8465882352941176, + "grad_norm": 1.2376467588646775e-06, + "learning_rate": 0.00027690690856281054, + "loss": 0.0, + "step": 3924 + }, + { + "epoch": 1.8470588235294119, + "grad_norm": 0.09540609835232604, + "learning_rate": 0.00027688353364815834, + "loss": 0.0531, + "step": 3925 + }, + { + "epoch": 1.8475294117647059, + "grad_norm": 0.12551938927254755, + "learning_rate": 0.0002768601478970301, + "loss": 0.0864, + "step": 3926 + }, + { + "epoch": 1.8479999999999999, + "grad_norm": 0.15771281920185337, + "learning_rate": 0.00027683675131142294, + "loss": 0.1262, + "step": 3927 + }, + { + "epoch": 1.848470588235294, + "grad_norm": 1.5023531659528401e-06, + "learning_rate": 0.000276813343893335, + "loss": 0.0, + "step": 3928 + }, + { + "epoch": 1.8489411764705883, + "grad_norm": 0.06643242190849727, + "learning_rate": 0.0002767899256447656, + "loss": 0.0292, + "step": 3929 + }, + { + "epoch": 1.8494117647058823, + "grad_norm": 0.12179834240397619, + "learning_rate": 0.0002767664965677146, + "loss": 0.0801, + "step": 3930 + }, + { + "epoch": 1.8498823529411763, + "grad_norm": 0.0907346179814835, + "learning_rate": 0.000276743056664183, + "loss": 0.0459, + "step": 3931 + }, + { + "epoch": 1.8503529411764705, + "grad_norm": 0.11771224857872536, + "learning_rate": 0.0002767196059361727, + "loss": 0.084, + "step": 3932 + }, + { + "epoch": 1.8508235294117648, + "grad_norm": 0.04020926813879026, + "learning_rate": 0.0002766961443856865, + "loss": 0.0158, + "step": 3933 + }, + { + "epoch": 1.851294117647059, + "grad_norm": 0.08460344476176061, + "learning_rate": 0.00027667267201472815, + "loss": 0.0509, + "step": 3934 + }, + { + "epoch": 1.851764705882353, + "grad_norm": 3.4129627893295024e-06, + "learning_rate": 0.00027664918882530225, + "loss": 0.0, + "step": 3935 + }, + { + "epoch": 1.852235294117647, + "grad_norm": 2.8215732100792376e-06, + "learning_rate": 0.0002766256948194144, + "loss": 0.0, + "step": 3936 + }, + { + "epoch": 1.8527058823529412, + "grad_norm": 0.1253177995156808, + "learning_rate": 0.0002766021899990712, + "loss": 0.0751, + "step": 3937 + }, + { + "epoch": 1.8531764705882354, + "grad_norm": 0.14602605833281754, + "learning_rate": 0.0002765786743662799, + "loss": 0.1056, + "step": 3938 + }, + { + "epoch": 1.8536470588235294, + "grad_norm": 0.03828637288442391, + "learning_rate": 0.0002765551479230489, + "loss": 0.0178, + "step": 3939 + }, + { + "epoch": 1.8541176470588234, + "grad_norm": 0.13110567826900069, + "learning_rate": 0.00027653161067138764, + "loss": 0.114, + "step": 3940 + }, + { + "epoch": 1.8545882352941176, + "grad_norm": 0.04305242510677845, + "learning_rate": 0.00027650806261330616, + "loss": 0.0155, + "step": 3941 + }, + { + "epoch": 1.8550588235294119, + "grad_norm": 0.09255460270731419, + "learning_rate": 0.0002764845037508156, + "loss": 0.0571, + "step": 3942 + }, + { + "epoch": 1.8555294117647059, + "grad_norm": 0.13169740797719037, + "learning_rate": 0.00027646093408592804, + "loss": 0.0816, + "step": 3943 + }, + { + "epoch": 1.8559999999999999, + "grad_norm": 0.0986308152935146, + "learning_rate": 0.0002764373536206564, + "loss": 0.0075, + "step": 3944 + }, + { + "epoch": 1.856470588235294, + "grad_norm": 0.17621556398734975, + "learning_rate": 0.0002764137623570146, + "loss": 0.1165, + "step": 3945 + }, + { + "epoch": 1.8569411764705883, + "grad_norm": 0.07610561960726114, + "learning_rate": 0.0002763901602970174, + "loss": 0.0507, + "step": 3946 + }, + { + "epoch": 1.8574117647058823, + "grad_norm": 7.089264363068591e-06, + "learning_rate": 0.0002763665474426806, + "loss": 0.0, + "step": 3947 + }, + { + "epoch": 1.8578823529411763, + "grad_norm": 0.08016069309639871, + "learning_rate": 0.00027634292379602086, + "loss": 0.0376, + "step": 3948 + }, + { + "epoch": 1.8583529411764705, + "grad_norm": 2.021538633147072e-06, + "learning_rate": 0.00027631928935905567, + "loss": 0.0, + "step": 3949 + }, + { + "epoch": 1.8588235294117648, + "grad_norm": 0.08325525776588306, + "learning_rate": 0.0002762956441338036, + "loss": 0.0532, + "step": 3950 + }, + { + "epoch": 1.859294117647059, + "grad_norm": 0.03411699359372118, + "learning_rate": 0.00027627198812228405, + "loss": 0.007, + "step": 3951 + }, + { + "epoch": 1.859764705882353, + "grad_norm": 0.07872314975119249, + "learning_rate": 0.0002762483213265173, + "loss": 0.0614, + "step": 3952 + }, + { + "epoch": 1.860235294117647, + "grad_norm": 0.10541894001084667, + "learning_rate": 0.0002762246437485247, + "loss": 0.0546, + "step": 3953 + }, + { + "epoch": 1.8607058823529412, + "grad_norm": 0.07425948251872903, + "learning_rate": 0.0002762009553903284, + "loss": 0.0454, + "step": 3954 + }, + { + "epoch": 1.8611764705882354, + "grad_norm": 0.07955512430423525, + "learning_rate": 0.0002761772562539515, + "loss": 0.039, + "step": 3955 + }, + { + "epoch": 1.8616470588235294, + "grad_norm": 0.01394563438765351, + "learning_rate": 0.0002761535463414181, + "loss": 0.0024, + "step": 3956 + }, + { + "epoch": 1.8621176470588234, + "grad_norm": 0.03247479863353677, + "learning_rate": 0.0002761298256547529, + "loss": 0.0091, + "step": 3957 + }, + { + "epoch": 1.8625882352941177, + "grad_norm": 7.106769702554578e-06, + "learning_rate": 0.00027610609419598204, + "loss": 0.0, + "step": 3958 + }, + { + "epoch": 1.8630588235294119, + "grad_norm": 0.10972622748985557, + "learning_rate": 0.0002760823519671321, + "loss": 0.0832, + "step": 3959 + }, + { + "epoch": 1.8635294117647059, + "grad_norm": 2.255674488528612e-06, + "learning_rate": 0.0002760585989702309, + "loss": 0.0, + "step": 3960 + }, + { + "epoch": 1.8639999999999999, + "grad_norm": 2.6726587426032805e-06, + "learning_rate": 0.0002760348352073071, + "loss": 0.0, + "step": 3961 + }, + { + "epoch": 1.864470588235294, + "grad_norm": 0.09903459107809746, + "learning_rate": 0.0002760110606803901, + "loss": 0.0357, + "step": 3962 + }, + { + "epoch": 1.8649411764705883, + "grad_norm": 0.05985257433734973, + "learning_rate": 0.00027598727539151054, + "loss": 0.0222, + "step": 3963 + }, + { + "epoch": 1.8654117647058823, + "grad_norm": 0.07238805052305138, + "learning_rate": 0.0002759634793426996, + "loss": 0.0577, + "step": 3964 + }, + { + "epoch": 1.8658823529411763, + "grad_norm": 0.046040220595651306, + "learning_rate": 0.00027593967253598974, + "loss": 0.0181, + "step": 3965 + }, + { + "epoch": 1.8663529411764705, + "grad_norm": 4.206427893151835e-05, + "learning_rate": 0.00027591585497341406, + "loss": 0.0, + "step": 3966 + }, + { + "epoch": 1.8668235294117648, + "grad_norm": 0.02614471530256489, + "learning_rate": 0.00027589202665700676, + "loss": 0.0047, + "step": 3967 + }, + { + "epoch": 1.8672941176470588, + "grad_norm": 0.14214826920550638, + "learning_rate": 0.0002758681875888029, + "loss": 0.112, + "step": 3968 + }, + { + "epoch": 1.867764705882353, + "grad_norm": 0.12880077153689787, + "learning_rate": 0.00027584433777083845, + "loss": 0.0763, + "step": 3969 + }, + { + "epoch": 1.868235294117647, + "grad_norm": 0.18290389027608472, + "learning_rate": 0.0002758204772051503, + "loss": 0.1397, + "step": 3970 + }, + { + "epoch": 1.8687058823529412, + "grad_norm": 0.1004637516955774, + "learning_rate": 0.00027579660589377626, + "loss": 0.0556, + "step": 3971 + }, + { + "epoch": 1.8691764705882354, + "grad_norm": 0.056615734581234894, + "learning_rate": 0.00027577272383875497, + "loss": 0.0284, + "step": 3972 + }, + { + "epoch": 1.8696470588235294, + "grad_norm": 0.06828882585715346, + "learning_rate": 0.0002757488310421262, + "loss": 0.0336, + "step": 3973 + }, + { + "epoch": 1.8701176470588234, + "grad_norm": 0.05537832302564941, + "learning_rate": 0.0002757249275059305, + "loss": 0.0259, + "step": 3974 + }, + { + "epoch": 1.8705882352941177, + "grad_norm": 0.10075483540094181, + "learning_rate": 0.00027570101323220927, + "loss": 0.0597, + "step": 3975 + }, + { + "epoch": 1.8710588235294119, + "grad_norm": 0.15188369822386846, + "learning_rate": 0.00027567708822300496, + "loss": 0.0892, + "step": 3976 + }, + { + "epoch": 1.8715294117647059, + "grad_norm": 0.0780344206622903, + "learning_rate": 0.0002756531524803609, + "loss": 0.0424, + "step": 3977 + }, + { + "epoch": 1.8719999999999999, + "grad_norm": 0.11126085921665661, + "learning_rate": 0.00027562920600632127, + "loss": 0.0786, + "step": 3978 + }, + { + "epoch": 1.872470588235294, + "grad_norm": 0.0978726354794972, + "learning_rate": 0.0002756052488029312, + "loss": 0.0451, + "step": 3979 + }, + { + "epoch": 1.8729411764705883, + "grad_norm": 0.07791565798860338, + "learning_rate": 0.0002755812808722368, + "loss": 0.0354, + "step": 3980 + }, + { + "epoch": 1.8734117647058823, + "grad_norm": 0.1665609605064052, + "learning_rate": 0.0002755573022162851, + "loss": 0.0663, + "step": 3981 + }, + { + "epoch": 1.8738823529411763, + "grad_norm": 0.14379819417101697, + "learning_rate": 0.0002755333128371239, + "loss": 0.1064, + "step": 3982 + }, + { + "epoch": 1.8743529411764706, + "grad_norm": 0.07457948038735666, + "learning_rate": 0.00027550931273680207, + "loss": 0.0368, + "step": 3983 + }, + { + "epoch": 1.8748235294117648, + "grad_norm": 0.13515024910224901, + "learning_rate": 0.00027548530191736926, + "loss": 0.1033, + "step": 3984 + }, + { + "epoch": 1.8752941176470588, + "grad_norm": 0.10730606605594305, + "learning_rate": 0.0002754612803808762, + "loss": 0.0745, + "step": 3985 + }, + { + "epoch": 1.8757647058823528, + "grad_norm": 0.00015859797549689004, + "learning_rate": 0.00027543724812937443, + "loss": 0.0, + "step": 3986 + }, + { + "epoch": 1.876235294117647, + "grad_norm": 0.06257007357319629, + "learning_rate": 0.0002754132051649164, + "loss": 0.0235, + "step": 3987 + }, + { + "epoch": 1.8767058823529412, + "grad_norm": 0.07234077462710818, + "learning_rate": 0.00027538915148955545, + "loss": 0.0314, + "step": 3988 + }, + { + "epoch": 1.8771764705882354, + "grad_norm": 0.04541468363376763, + "learning_rate": 0.000275365087105346, + "loss": 0.0197, + "step": 3989 + }, + { + "epoch": 1.8776470588235294, + "grad_norm": 0.021719170996343136, + "learning_rate": 0.00027534101201434315, + "loss": 0.0046, + "step": 3990 + }, + { + "epoch": 1.8781176470588234, + "grad_norm": 0.09562922913809978, + "learning_rate": 0.00027531692621860307, + "loss": 0.0779, + "step": 3991 + }, + { + "epoch": 1.8785882352941177, + "grad_norm": 0.10708343629735083, + "learning_rate": 0.00027529282972018287, + "loss": 0.0486, + "step": 3992 + }, + { + "epoch": 1.879058823529412, + "grad_norm": 0.03977567296934077, + "learning_rate": 0.0002752687225211404, + "loss": 0.0135, + "step": 3993 + }, + { + "epoch": 1.879529411764706, + "grad_norm": 0.06792010411055883, + "learning_rate": 0.0002752446046235346, + "loss": 0.0322, + "step": 3994 + }, + { + "epoch": 1.88, + "grad_norm": 0.0538029783045924, + "learning_rate": 0.00027522047602942525, + "loss": 0.018, + "step": 3995 + }, + { + "epoch": 1.8804705882352941, + "grad_norm": 0.09459680197588147, + "learning_rate": 0.00027519633674087305, + "loss": 0.07, + "step": 3996 + }, + { + "epoch": 1.8809411764705883, + "grad_norm": 0.13375297039008946, + "learning_rate": 0.0002751721867599396, + "loss": 0.0912, + "step": 3997 + }, + { + "epoch": 1.8814117647058823, + "grad_norm": 0.00025194453057347777, + "learning_rate": 0.00027514802608868743, + "loss": 0.0, + "step": 3998 + }, + { + "epoch": 1.8818823529411763, + "grad_norm": 0.09548697979521048, + "learning_rate": 0.00027512385472918, + "loss": 0.0851, + "step": 3999 + }, + { + "epoch": 1.8823529411764706, + "grad_norm": 0.11091484129584078, + "learning_rate": 0.0002750996726834817, + "loss": 0.0651, + "step": 4000 + }, + { + "epoch": 1.8828235294117648, + "grad_norm": 0.06518495807803483, + "learning_rate": 0.00027507547995365766, + "loss": 0.0291, + "step": 4001 + }, + { + "epoch": 1.8832941176470588, + "grad_norm": 0.07261668153158779, + "learning_rate": 0.00027505127654177413, + "loss": 0.047, + "step": 4002 + }, + { + "epoch": 1.8837647058823528, + "grad_norm": 3.877630032718125e-06, + "learning_rate": 0.0002750270624498983, + "loss": 0.0, + "step": 4003 + }, + { + "epoch": 1.884235294117647, + "grad_norm": 0.039518337709716245, + "learning_rate": 0.000275002837680098, + "loss": 0.0069, + "step": 4004 + }, + { + "epoch": 1.8847058823529412, + "grad_norm": 0.09412010704306596, + "learning_rate": 0.0002749786022344422, + "loss": 0.0504, + "step": 4005 + }, + { + "epoch": 1.8851764705882355, + "grad_norm": 0.16060113094918052, + "learning_rate": 0.0002749543561150009, + "loss": 0.1072, + "step": 4006 + }, + { + "epoch": 1.8856470588235295, + "grad_norm": 0.14012035936859477, + "learning_rate": 0.00027493009932384456, + "loss": 0.1088, + "step": 4007 + }, + { + "epoch": 1.8861176470588235, + "grad_norm": 0.07826722963782903, + "learning_rate": 0.000274905831863045, + "loss": 0.0366, + "step": 4008 + }, + { + "epoch": 1.8865882352941177, + "grad_norm": 8.078355921176316e-05, + "learning_rate": 0.00027488155373467474, + "loss": 0.0, + "step": 4009 + }, + { + "epoch": 1.887058823529412, + "grad_norm": 1.4592261762845639e-05, + "learning_rate": 0.00027485726494080724, + "loss": 0.0, + "step": 4010 + }, + { + "epoch": 1.887529411764706, + "grad_norm": 0.05093138681034409, + "learning_rate": 0.0002748329654835169, + "loss": 0.0173, + "step": 4011 + }, + { + "epoch": 1.888, + "grad_norm": 0.07965371799275703, + "learning_rate": 0.000274808655364879, + "loss": 0.0372, + "step": 4012 + }, + { + "epoch": 1.8884705882352941, + "grad_norm": 0.10863441423075379, + "learning_rate": 0.00027478433458696976, + "loss": 0.0705, + "step": 4013 + }, + { + "epoch": 1.8889411764705883, + "grad_norm": 0.05792710043555197, + "learning_rate": 0.0002747600031518663, + "loss": 0.0251, + "step": 4014 + }, + { + "epoch": 1.8894117647058823, + "grad_norm": 0.09022317966448012, + "learning_rate": 0.0002747356610616466, + "loss": 0.0341, + "step": 4015 + }, + { + "epoch": 1.8898823529411763, + "grad_norm": 0.0003786088970539503, + "learning_rate": 0.0002747113083183896, + "loss": 0.0, + "step": 4016 + }, + { + "epoch": 1.8903529411764706, + "grad_norm": 0.12077749744107957, + "learning_rate": 0.00027468694492417515, + "loss": 0.0778, + "step": 4017 + }, + { + "epoch": 1.8908235294117648, + "grad_norm": 0.12778049946407768, + "learning_rate": 0.0002746625708810841, + "loss": 0.0848, + "step": 4018 + }, + { + "epoch": 1.8912941176470588, + "grad_norm": 0.0675710633187627, + "learning_rate": 0.00027463818619119794, + "loss": 0.0295, + "step": 4019 + }, + { + "epoch": 1.8917647058823528, + "grad_norm": 0.09090034617048917, + "learning_rate": 0.0002746137908565994, + "loss": 0.0735, + "step": 4020 + }, + { + "epoch": 1.892235294117647, + "grad_norm": 0.030171044710602633, + "learning_rate": 0.00027458938487937186, + "loss": 0.0089, + "step": 4021 + }, + { + "epoch": 1.8927058823529412, + "grad_norm": 0.054970512752420764, + "learning_rate": 0.0002745649682615998, + "loss": 0.0209, + "step": 4022 + }, + { + "epoch": 1.8931764705882355, + "grad_norm": 0.0002159375008915589, + "learning_rate": 0.00027454054100536836, + "loss": 0.0, + "step": 4023 + }, + { + "epoch": 1.8936470588235295, + "grad_norm": 2.5866767787955697e-05, + "learning_rate": 0.00027451610311276397, + "loss": 0.0, + "step": 4024 + }, + { + "epoch": 1.8941176470588235, + "grad_norm": 0.10031754878062546, + "learning_rate": 0.00027449165458587354, + "loss": 0.0608, + "step": 4025 + }, + { + "epoch": 1.8945882352941177, + "grad_norm": 0.03757973968800027, + "learning_rate": 0.0002744671954267852, + "loss": 0.0177, + "step": 4026 + }, + { + "epoch": 1.895058823529412, + "grad_norm": 0.0982409358918762, + "learning_rate": 0.00027444272563758784, + "loss": 0.0672, + "step": 4027 + }, + { + "epoch": 1.895529411764706, + "grad_norm": 0.03752468306925772, + "learning_rate": 0.00027441824522037137, + "loss": 0.0138, + "step": 4028 + }, + { + "epoch": 1.896, + "grad_norm": 0.07956523846428086, + "learning_rate": 0.0002743937541772265, + "loss": 0.0403, + "step": 4029 + }, + { + "epoch": 1.8964705882352941, + "grad_norm": 0.10385873848779073, + "learning_rate": 0.0002743692525102448, + "loss": 0.0641, + "step": 4030 + }, + { + "epoch": 1.8969411764705884, + "grad_norm": 0.06912105268318139, + "learning_rate": 0.000274344740221519, + "loss": 0.035, + "step": 4031 + }, + { + "epoch": 1.8974117647058824, + "grad_norm": 0.195870474148882, + "learning_rate": 0.0002743202173131424, + "loss": 0.1164, + "step": 4032 + }, + { + "epoch": 1.8978823529411764, + "grad_norm": 0.0581979188986378, + "learning_rate": 0.00027429568378720945, + "loss": 0.0321, + "step": 4033 + }, + { + "epoch": 1.8983529411764706, + "grad_norm": 0.09510482531755063, + "learning_rate": 0.0002742711396458155, + "loss": 0.0564, + "step": 4034 + }, + { + "epoch": 1.8988235294117648, + "grad_norm": 2.1663695334006355e-05, + "learning_rate": 0.00027424658489105667, + "loss": 0.0, + "step": 4035 + }, + { + "epoch": 1.8992941176470588, + "grad_norm": 1.4538918991559935e-06, + "learning_rate": 0.00027422201952503005, + "loss": 0.0, + "step": 4036 + }, + { + "epoch": 1.8997647058823528, + "grad_norm": 0.08715706376497137, + "learning_rate": 0.00027419744354983364, + "loss": 0.0451, + "step": 4037 + }, + { + "epoch": 1.900235294117647, + "grad_norm": 0.0905570190194165, + "learning_rate": 0.00027417285696756634, + "loss": 0.0507, + "step": 4038 + }, + { + "epoch": 1.9007058823529412, + "grad_norm": 0.0690327684877824, + "learning_rate": 0.00027414825978032805, + "loss": 0.0422, + "step": 4039 + }, + { + "epoch": 1.9011764705882352, + "grad_norm": 1.0299554481591882e-06, + "learning_rate": 0.0002741236519902194, + "loss": 0.0, + "step": 4040 + }, + { + "epoch": 1.9016470588235292, + "grad_norm": 0.1556707343939922, + "learning_rate": 0.0002740990335993421, + "loss": 0.1348, + "step": 4041 + }, + { + "epoch": 1.9021176470588235, + "grad_norm": 0.06792141702554998, + "learning_rate": 0.00027407440460979854, + "loss": 0.0324, + "step": 4042 + }, + { + "epoch": 1.9025882352941177, + "grad_norm": 0.13412747330842123, + "learning_rate": 0.0002740497650236923, + "loss": 0.1283, + "step": 4043 + }, + { + "epoch": 1.903058823529412, + "grad_norm": 8.463459057215534e-07, + "learning_rate": 0.00027402511484312764, + "loss": 0.0, + "step": 4044 + }, + { + "epoch": 1.903529411764706, + "grad_norm": 0.03132139976186395, + "learning_rate": 0.00027400045407020986, + "loss": 0.0042, + "step": 4045 + }, + { + "epoch": 1.904, + "grad_norm": 4.424425149477292e-06, + "learning_rate": 0.00027397578270704504, + "loss": 0.0, + "step": 4046 + }, + { + "epoch": 1.9044705882352941, + "grad_norm": 0.029607155192371417, + "learning_rate": 0.0002739511007557403, + "loss": 0.0082, + "step": 4047 + }, + { + "epoch": 1.9049411764705884, + "grad_norm": 0.05080804655417857, + "learning_rate": 0.00027392640821840365, + "loss": 0.0198, + "step": 4048 + }, + { + "epoch": 1.9054117647058824, + "grad_norm": 0.09228964055339803, + "learning_rate": 0.0002739017050971438, + "loss": 0.0331, + "step": 4049 + }, + { + "epoch": 1.9058823529411764, + "grad_norm": 3.1634401070116823e-06, + "learning_rate": 0.0002738769913940706, + "loss": 0.0, + "step": 4050 + }, + { + "epoch": 1.9063529411764706, + "grad_norm": 0.11568307776702799, + "learning_rate": 0.00027385226711129477, + "loss": 0.0706, + "step": 4051 + }, + { + "epoch": 1.9068235294117648, + "grad_norm": 0.12855319620449385, + "learning_rate": 0.00027382753225092777, + "loss": 0.0226, + "step": 4052 + }, + { + "epoch": 1.9072941176470588, + "grad_norm": 0.24571452880059835, + "learning_rate": 0.0002738027868150822, + "loss": 0.1979, + "step": 4053 + }, + { + "epoch": 1.9077647058823528, + "grad_norm": 7.982221319044516e-06, + "learning_rate": 0.0002737780308058714, + "loss": 0.0, + "step": 4054 + }, + { + "epoch": 1.908235294117647, + "grad_norm": 0.07142176775363851, + "learning_rate": 0.00027375326422540957, + "loss": 0.0526, + "step": 4055 + }, + { + "epoch": 1.9087058823529413, + "grad_norm": 0.09765178265024893, + "learning_rate": 0.000273728487075812, + "loss": 0.0849, + "step": 4056 + }, + { + "epoch": 1.9091764705882353, + "grad_norm": 2.099525611797127e-06, + "learning_rate": 0.0002737036993591948, + "loss": 0.0, + "step": 4057 + }, + { + "epoch": 1.9096470588235293, + "grad_norm": 0.09663740274243043, + "learning_rate": 0.0002736789010776749, + "loss": 0.0651, + "step": 4058 + }, + { + "epoch": 1.9101176470588235, + "grad_norm": 0.08794649714738187, + "learning_rate": 0.00027365409223337016, + "loss": 0.0542, + "step": 4059 + }, + { + "epoch": 1.9105882352941177, + "grad_norm": 5.987416407364896e-07, + "learning_rate": 0.0002736292728283994, + "loss": 0.0, + "step": 4060 + }, + { + "epoch": 1.911058823529412, + "grad_norm": 0.14512850572706756, + "learning_rate": 0.0002736044428648824, + "loss": 0.1176, + "step": 4061 + }, + { + "epoch": 1.911529411764706, + "grad_norm": 5.376550354346728e-06, + "learning_rate": 0.0002735796023449397, + "loss": 0.0, + "step": 4062 + }, + { + "epoch": 1.912, + "grad_norm": 2.8252280484017454e-06, + "learning_rate": 0.00027355475127069275, + "loss": 0.0, + "step": 4063 + }, + { + "epoch": 1.9124705882352941, + "grad_norm": 0.08349309544347643, + "learning_rate": 0.0002735298896442641, + "loss": 0.0615, + "step": 4064 + }, + { + "epoch": 1.9129411764705884, + "grad_norm": 0.06224667024858869, + "learning_rate": 0.0002735050174677769, + "loss": 0.0364, + "step": 4065 + }, + { + "epoch": 1.9134117647058824, + "grad_norm": 0.1057705784387504, + "learning_rate": 0.00027348013474335546, + "loss": 0.0797, + "step": 4066 + }, + { + "epoch": 1.9138823529411764, + "grad_norm": 0.05652243037540723, + "learning_rate": 0.0002734552414731248, + "loss": 0.0261, + "step": 4067 + }, + { + "epoch": 1.9143529411764706, + "grad_norm": 0.2088864722953024, + "learning_rate": 0.000273430337659211, + "loss": 0.1512, + "step": 4068 + }, + { + "epoch": 1.9148235294117648, + "grad_norm": 0.09309814026484811, + "learning_rate": 0.0002734054233037409, + "loss": 0.0616, + "step": 4069 + }, + { + "epoch": 1.9152941176470588, + "grad_norm": 0.11141519812702425, + "learning_rate": 0.00027338049840884245, + "loss": 0.0588, + "step": 4070 + }, + { + "epoch": 1.9157647058823528, + "grad_norm": 0.14346243478677329, + "learning_rate": 0.00027335556297664417, + "loss": 0.1015, + "step": 4071 + }, + { + "epoch": 1.916235294117647, + "grad_norm": 0.00010600862558282859, + "learning_rate": 0.00027333061700927577, + "loss": 0.0, + "step": 4072 + }, + { + "epoch": 1.9167058823529413, + "grad_norm": 2.7248920687838056e-05, + "learning_rate": 0.00027330566050886776, + "loss": 0.0, + "step": 4073 + }, + { + "epoch": 1.9171764705882353, + "grad_norm": 0.08709179821868518, + "learning_rate": 0.0002732806934775515, + "loss": 0.0632, + "step": 4074 + }, + { + "epoch": 1.9176470588235293, + "grad_norm": 0.07189290367591049, + "learning_rate": 0.0002732557159174594, + "loss": 0.047, + "step": 4075 + }, + { + "epoch": 1.9181176470588235, + "grad_norm": 0.07607530849456678, + "learning_rate": 0.0002732307278307246, + "loss": 0.043, + "step": 4076 + }, + { + "epoch": 1.9185882352941177, + "grad_norm": 0.06493342782002494, + "learning_rate": 0.0002732057292194811, + "loss": 0.04, + "step": 4077 + }, + { + "epoch": 1.919058823529412, + "grad_norm": 0.06199511516132389, + "learning_rate": 0.000273180720085864, + "loss": 0.0276, + "step": 4078 + }, + { + "epoch": 1.919529411764706, + "grad_norm": 0.04042399545380617, + "learning_rate": 0.0002731557004320093, + "loss": 0.0173, + "step": 4079 + }, + { + "epoch": 1.92, + "grad_norm": 0.07199724185878083, + "learning_rate": 0.0002731306702600537, + "loss": 0.0425, + "step": 4080 + }, + { + "epoch": 1.9204705882352942, + "grad_norm": 0.08982592639890867, + "learning_rate": 0.00027310562957213487, + "loss": 0.0637, + "step": 4081 + }, + { + "epoch": 1.9209411764705884, + "grad_norm": 0.2321111095309261, + "learning_rate": 0.00027308057837039144, + "loss": 0.1892, + "step": 4082 + }, + { + "epoch": 1.9214117647058824, + "grad_norm": 0.036727764139528676, + "learning_rate": 0.0002730555166569629, + "loss": 0.0112, + "step": 4083 + }, + { + "epoch": 1.9218823529411764, + "grad_norm": 0.031919266056962006, + "learning_rate": 0.00027303044443398967, + "loss": 0.0112, + "step": 4084 + }, + { + "epoch": 1.9223529411764706, + "grad_norm": 2.2199823217337004e-06, + "learning_rate": 0.00027300536170361307, + "loss": 0.0, + "step": 4085 + }, + { + "epoch": 1.9228235294117648, + "grad_norm": 0.09881922091698768, + "learning_rate": 0.00027298026846797515, + "loss": 0.0509, + "step": 4086 + }, + { + "epoch": 1.9232941176470588, + "grad_norm": 0.11000088480148383, + "learning_rate": 0.0002729551647292191, + "loss": 0.0793, + "step": 4087 + }, + { + "epoch": 1.9237647058823528, + "grad_norm": 1.690144285591e-05, + "learning_rate": 0.000272930050489489, + "loss": 0.0, + "step": 4088 + }, + { + "epoch": 1.924235294117647, + "grad_norm": 0.028581850871994396, + "learning_rate": 0.0002729049257509295, + "loss": 0.0069, + "step": 4089 + }, + { + "epoch": 1.9247058823529413, + "grad_norm": 0.03452134231717912, + "learning_rate": 0.0002728797905156866, + "loss": 0.009, + "step": 4090 + }, + { + "epoch": 1.9251764705882353, + "grad_norm": 0.0720040540165714, + "learning_rate": 0.0002728546447859068, + "loss": 0.0408, + "step": 4091 + }, + { + "epoch": 1.9256470588235293, + "grad_norm": 1.8067101862622915e-05, + "learning_rate": 0.00027282948856373774, + "loss": 0.0, + "step": 4092 + }, + { + "epoch": 1.9261176470588235, + "grad_norm": 0.1764790532374325, + "learning_rate": 0.0002728043218513279, + "loss": 0.1531, + "step": 4093 + }, + { + "epoch": 1.9265882352941177, + "grad_norm": 0.046914613983480784, + "learning_rate": 0.0002727791446508266, + "loss": 0.0214, + "step": 4094 + }, + { + "epoch": 1.9270588235294117, + "grad_norm": 0.05384367065718315, + "learning_rate": 0.00027275395696438413, + "loss": 0.0248, + "step": 4095 + }, + { + "epoch": 1.927529411764706, + "grad_norm": 0.1156669723888778, + "learning_rate": 0.0002727287587941516, + "loss": 0.0256, + "step": 4096 + }, + { + "epoch": 1.928, + "grad_norm": 0.024145338723388615, + "learning_rate": 0.00027270355014228113, + "loss": 0.0055, + "step": 4097 + }, + { + "epoch": 1.9284705882352942, + "grad_norm": 0.02604842882105875, + "learning_rate": 0.00027267833101092564, + "loss": 0.008, + "step": 4098 + }, + { + "epoch": 1.9289411764705884, + "grad_norm": 1.8716885929381374e-06, + "learning_rate": 0.0002726531014022389, + "loss": 0.0, + "step": 4099 + }, + { + "epoch": 1.9294117647058824, + "grad_norm": 2.011706915353439e-05, + "learning_rate": 0.00027262786131837573, + "loss": 0.0, + "step": 4100 + }, + { + "epoch": 1.9298823529411764, + "grad_norm": 1.2029623457612608e-05, + "learning_rate": 0.00027260261076149166, + "loss": 0.0, + "step": 4101 + }, + { + "epoch": 1.9303529411764706, + "grad_norm": 0.0434657090045174, + "learning_rate": 0.00027257734973374335, + "loss": 0.0146, + "step": 4102 + }, + { + "epoch": 1.9308235294117648, + "grad_norm": 0.07229550439798474, + "learning_rate": 0.0002725520782372881, + "loss": 0.0318, + "step": 4103 + }, + { + "epoch": 1.9312941176470588, + "grad_norm": 0.038117007658797816, + "learning_rate": 0.00027252679627428423, + "loss": 0.0168, + "step": 4104 + }, + { + "epoch": 1.9317647058823528, + "grad_norm": 0.07651265477368747, + "learning_rate": 0.00027250150384689094, + "loss": 0.0383, + "step": 4105 + }, + { + "epoch": 1.932235294117647, + "grad_norm": 0.15878492301389088, + "learning_rate": 0.0002724762009572684, + "loss": 0.0482, + "step": 4106 + }, + { + "epoch": 1.9327058823529413, + "grad_norm": 4.3905698831476515e-06, + "learning_rate": 0.0002724508876075776, + "loss": 0.0, + "step": 4107 + }, + { + "epoch": 1.9331764705882353, + "grad_norm": 0.03951758297325688, + "learning_rate": 0.0002724255637999803, + "loss": 0.0131, + "step": 4108 + }, + { + "epoch": 1.9336470588235293, + "grad_norm": 0.05893324018134923, + "learning_rate": 0.0002724002295366394, + "loss": 0.0249, + "step": 4109 + }, + { + "epoch": 1.9341176470588235, + "grad_norm": 2.5444164381469495e-06, + "learning_rate": 0.0002723748848197185, + "loss": 0.0, + "step": 4110 + }, + { + "epoch": 1.9345882352941177, + "grad_norm": 9.923498204537427e-06, + "learning_rate": 0.00027234952965138223, + "loss": 0.0, + "step": 4111 + }, + { + "epoch": 1.9350588235294117, + "grad_norm": 0.06620884728543144, + "learning_rate": 0.000272324164033796, + "loss": 0.0364, + "step": 4112 + }, + { + "epoch": 1.9355294117647057, + "grad_norm": 0.14567085875214159, + "learning_rate": 0.0002722987879691261, + "loss": 0.0734, + "step": 4113 + }, + { + "epoch": 1.936, + "grad_norm": 0.12995502218816293, + "learning_rate": 0.00027227340145953994, + "loss": 0.0999, + "step": 4114 + }, + { + "epoch": 1.9364705882352942, + "grad_norm": 0.1279959311336575, + "learning_rate": 0.0002722480045072055, + "loss": 0.0627, + "step": 4115 + }, + { + "epoch": 1.9369411764705884, + "grad_norm": 0.04656130783075648, + "learning_rate": 0.0002722225971142918, + "loss": 0.0188, + "step": 4116 + }, + { + "epoch": 1.9374117647058824, + "grad_norm": 0.05997326983742388, + "learning_rate": 0.00027219717928296885, + "loss": 0.0141, + "step": 4117 + }, + { + "epoch": 1.9378823529411764, + "grad_norm": 0.060652741282763274, + "learning_rate": 0.0002721717510154074, + "loss": 0.0238, + "step": 4118 + }, + { + "epoch": 1.9383529411764706, + "grad_norm": 0.0641562193078379, + "learning_rate": 0.0002721463123137792, + "loss": 0.0267, + "step": 4119 + }, + { + "epoch": 1.9388235294117648, + "grad_norm": 0.13142091915617235, + "learning_rate": 0.0002721208631802568, + "loss": 0.0837, + "step": 4120 + }, + { + "epoch": 1.9392941176470588, + "grad_norm": 0.08294583106774009, + "learning_rate": 0.00027209540361701365, + "loss": 0.0435, + "step": 4121 + }, + { + "epoch": 1.9397647058823528, + "grad_norm": 0.011477089243772002, + "learning_rate": 0.0002720699336262242, + "loss": 0.0014, + "step": 4122 + }, + { + "epoch": 1.940235294117647, + "grad_norm": 0.14243847576117183, + "learning_rate": 0.00027204445321006365, + "loss": 0.1333, + "step": 4123 + }, + { + "epoch": 1.9407058823529413, + "grad_norm": 0.06167201984928205, + "learning_rate": 0.00027201896237070815, + "loss": 0.0397, + "step": 4124 + }, + { + "epoch": 1.9411764705882353, + "grad_norm": 0.02144847067870265, + "learning_rate": 0.0002719934611103348, + "loss": 0.0047, + "step": 4125 + }, + { + "epoch": 1.9416470588235293, + "grad_norm": 0.025742492767148597, + "learning_rate": 0.00027196794943112147, + "loss": 0.006, + "step": 4126 + }, + { + "epoch": 1.9421176470588235, + "grad_norm": 0.02522919994653867, + "learning_rate": 0.000271942427335247, + "loss": 0.0046, + "step": 4127 + }, + { + "epoch": 1.9425882352941177, + "grad_norm": 0.03588576040287146, + "learning_rate": 0.00027191689482489114, + "loss": 0.0102, + "step": 4128 + }, + { + "epoch": 1.9430588235294117, + "grad_norm": 7.893976242613643e-06, + "learning_rate": 0.0002718913519022344, + "loss": 0.0, + "step": 4129 + }, + { + "epoch": 1.9435294117647057, + "grad_norm": 0.00018752200741330125, + "learning_rate": 0.00027186579856945843, + "loss": 0.0, + "step": 4130 + }, + { + "epoch": 1.944, + "grad_norm": 3.2398893844832595e-06, + "learning_rate": 0.00027184023482874545, + "loss": 0.0, + "step": 4131 + }, + { + "epoch": 1.9444705882352942, + "grad_norm": 0.07662914926894063, + "learning_rate": 0.00027181466068227885, + "loss": 0.0326, + "step": 4132 + }, + { + "epoch": 1.9449411764705884, + "grad_norm": 0.09215184707399249, + "learning_rate": 0.00027178907613224264, + "loss": 0.0305, + "step": 4133 + }, + { + "epoch": 1.9454117647058824, + "grad_norm": 0.10320850678465888, + "learning_rate": 0.00027176348118082203, + "loss": 0.0558, + "step": 4134 + }, + { + "epoch": 1.9458823529411764, + "grad_norm": 0.11117877888699257, + "learning_rate": 0.0002717378758302029, + "loss": 0.0639, + "step": 4135 + }, + { + "epoch": 1.9463529411764706, + "grad_norm": 0.017855800291933156, + "learning_rate": 0.000271712260082572, + "loss": 0.001, + "step": 4136 + }, + { + "epoch": 1.9468235294117648, + "grad_norm": 2.8183262938798603e-06, + "learning_rate": 0.0002716866339401171, + "loss": 0.0, + "step": 4137 + }, + { + "epoch": 1.9472941176470588, + "grad_norm": 0.06294893680049135, + "learning_rate": 0.0002716609974050268, + "loss": 0.0282, + "step": 4138 + }, + { + "epoch": 1.9477647058823528, + "grad_norm": 0.1962407459945283, + "learning_rate": 0.0002716353504794906, + "loss": 0.2205, + "step": 4139 + }, + { + "epoch": 1.948235294117647, + "grad_norm": 0.0813704328823802, + "learning_rate": 0.0002716096931656988, + "loss": 0.0179, + "step": 4140 + }, + { + "epoch": 1.9487058823529413, + "grad_norm": 0.10062950125670718, + "learning_rate": 0.0002715840254658427, + "loss": 0.0468, + "step": 4141 + }, + { + "epoch": 1.9491764705882353, + "grad_norm": 0.0866408510636993, + "learning_rate": 0.0002715583473821145, + "loss": 0.0691, + "step": 4142 + }, + { + "epoch": 1.9496470588235293, + "grad_norm": 0.054370335887002755, + "learning_rate": 0.0002715326589167072, + "loss": 0.019, + "step": 4143 + }, + { + "epoch": 1.9501176470588235, + "grad_norm": 0.09181194261037139, + "learning_rate": 0.0002715069600718147, + "loss": 0.0399, + "step": 4144 + }, + { + "epoch": 1.9505882352941177, + "grad_norm": 0.09336296972618595, + "learning_rate": 0.0002714812508496319, + "loss": 0.0606, + "step": 4145 + }, + { + "epoch": 1.9510588235294117, + "grad_norm": 0.06885188029266373, + "learning_rate": 0.0002714555312523543, + "loss": 0.0452, + "step": 4146 + }, + { + "epoch": 1.9515294117647057, + "grad_norm": 0.05032561728521263, + "learning_rate": 0.00027142980128217864, + "loss": 0.0244, + "step": 4147 + }, + { + "epoch": 1.952, + "grad_norm": 0.053226126452066334, + "learning_rate": 0.00027140406094130234, + "loss": 0.0266, + "step": 4148 + }, + { + "epoch": 1.9524705882352942, + "grad_norm": 0.02406801209285963, + "learning_rate": 0.0002713783102319237, + "loss": 0.0057, + "step": 4149 + }, + { + "epoch": 1.9529411764705882, + "grad_norm": 0.08657000536966501, + "learning_rate": 0.0002713525491562421, + "loss": 0.049, + "step": 4150 + }, + { + "epoch": 1.9534117647058824, + "grad_norm": 8.154179769290653e-07, + "learning_rate": 0.0002713267777164575, + "loss": 0.0, + "step": 4151 + }, + { + "epoch": 1.9538823529411764, + "grad_norm": 1.4275450437407057e-06, + "learning_rate": 0.00027130099591477104, + "loss": 0.0, + "step": 4152 + }, + { + "epoch": 1.9543529411764706, + "grad_norm": 0.1372581629670268, + "learning_rate": 0.0002712752037533845, + "loss": 0.0741, + "step": 4153 + }, + { + "epoch": 1.9548235294117648, + "grad_norm": 0.09831441039222147, + "learning_rate": 0.00027124940123450075, + "loss": 0.0617, + "step": 4154 + }, + { + "epoch": 1.9552941176470588, + "grad_norm": 0.0984168003320493, + "learning_rate": 0.00027122358836032337, + "loss": 0.0634, + "step": 4155 + }, + { + "epoch": 1.9557647058823528, + "grad_norm": 0.03014135729387039, + "learning_rate": 0.000271197765133057, + "loss": 0.0107, + "step": 4156 + }, + { + "epoch": 1.956235294117647, + "grad_norm": 0.07340161880831295, + "learning_rate": 0.00027117193155490695, + "loss": 0.0483, + "step": 4157 + }, + { + "epoch": 1.9567058823529413, + "grad_norm": 0.029942123776897842, + "learning_rate": 0.00027114608762807965, + "loss": 0.0088, + "step": 4158 + }, + { + "epoch": 1.9571764705882353, + "grad_norm": 0.035172573615179295, + "learning_rate": 0.0002711202333547822, + "loss": 0.0119, + "step": 4159 + }, + { + "epoch": 1.9576470588235293, + "grad_norm": 0.06831022459586919, + "learning_rate": 0.00027109436873722277, + "loss": 0.0334, + "step": 4160 + }, + { + "epoch": 1.9581176470588235, + "grad_norm": 0.07859381194707211, + "learning_rate": 0.00027106849377761023, + "loss": 0.0415, + "step": 4161 + }, + { + "epoch": 1.9585882352941177, + "grad_norm": 0.0321997318078329, + "learning_rate": 0.00027104260847815457, + "loss": 0.0045, + "step": 4162 + }, + { + "epoch": 1.9590588235294117, + "grad_norm": 0.0404027691826458, + "learning_rate": 0.0002710167128410664, + "loss": 0.0177, + "step": 4163 + }, + { + "epoch": 1.9595294117647057, + "grad_norm": 0.0806366957216166, + "learning_rate": 0.00027099080686855735, + "loss": 0.0463, + "step": 4164 + }, + { + "epoch": 1.96, + "grad_norm": 0.19336015958417177, + "learning_rate": 0.00027096489056283995, + "loss": 0.1021, + "step": 4165 + }, + { + "epoch": 1.9604705882352942, + "grad_norm": 0.03533956686764737, + "learning_rate": 0.00027093896392612753, + "loss": 0.0111, + "step": 4166 + }, + { + "epoch": 1.9609411764705882, + "grad_norm": 0.062478730272988635, + "learning_rate": 0.0002709130269606345, + "loss": 0.0371, + "step": 4167 + }, + { + "epoch": 1.9614117647058822, + "grad_norm": 0.06277375597819984, + "learning_rate": 0.00027088707966857576, + "loss": 0.0251, + "step": 4168 + }, + { + "epoch": 1.9618823529411764, + "grad_norm": 0.05213964497780119, + "learning_rate": 0.0002708611220521675, + "loss": 0.0259, + "step": 4169 + }, + { + "epoch": 1.9623529411764706, + "grad_norm": 0.06070906817177444, + "learning_rate": 0.0002708351541136267, + "loss": 0.0258, + "step": 4170 + }, + { + "epoch": 1.9628235294117649, + "grad_norm": 0.11491211934182434, + "learning_rate": 0.0002708091758551709, + "loss": 0.0962, + "step": 4171 + }, + { + "epoch": 1.9632941176470589, + "grad_norm": 0.09977567889693605, + "learning_rate": 0.000270783187279019, + "loss": 0.0418, + "step": 4172 + }, + { + "epoch": 1.9637647058823529, + "grad_norm": 3.189832822827813e-06, + "learning_rate": 0.00027075718838739043, + "loss": 0.0, + "step": 4173 + }, + { + "epoch": 1.964235294117647, + "grad_norm": 0.04243399052102157, + "learning_rate": 0.0002707311791825057, + "loss": 0.0185, + "step": 4174 + }, + { + "epoch": 1.9647058823529413, + "grad_norm": 0.05366905465806487, + "learning_rate": 0.00027070515966658604, + "loss": 0.0197, + "step": 4175 + }, + { + "epoch": 1.9651764705882353, + "grad_norm": 0.05428658247671484, + "learning_rate": 0.00027067912984185376, + "loss": 0.0194, + "step": 4176 + }, + { + "epoch": 1.9656470588235293, + "grad_norm": 0.03884558972008431, + "learning_rate": 0.0002706530897105318, + "loss": 0.0157, + "step": 4177 + }, + { + "epoch": 1.9661176470588235, + "grad_norm": 0.10821754998252327, + "learning_rate": 0.0002706270392748442, + "loss": 0.0704, + "step": 4178 + }, + { + "epoch": 1.9665882352941177, + "grad_norm": 2.9417319179015338e-06, + "learning_rate": 0.00027060097853701576, + "loss": 0.0, + "step": 4179 + }, + { + "epoch": 1.9670588235294117, + "grad_norm": 5.1058518786247485e-06, + "learning_rate": 0.0002705749074992722, + "loss": 0.0, + "step": 4180 + }, + { + "epoch": 1.9675294117647057, + "grad_norm": 5.923165994122632e-06, + "learning_rate": 0.00027054882616384015, + "loss": 0.0, + "step": 4181 + }, + { + "epoch": 1.968, + "grad_norm": 0.24398187753496783, + "learning_rate": 0.0002705227345329471, + "loss": 0.1487, + "step": 4182 + }, + { + "epoch": 1.9684705882352942, + "grad_norm": 0.07371851290391535, + "learning_rate": 0.00027049663260882124, + "loss": 0.0474, + "step": 4183 + }, + { + "epoch": 1.9689411764705882, + "grad_norm": 0.025046972802530834, + "learning_rate": 0.00027047052039369204, + "loss": 0.004, + "step": 4184 + }, + { + "epoch": 1.9694117647058822, + "grad_norm": 0.06286749356478326, + "learning_rate": 0.0002704443978897894, + "loss": 0.0319, + "step": 4185 + }, + { + "epoch": 1.9698823529411764, + "grad_norm": 0.09900906963513907, + "learning_rate": 0.0002704182650993445, + "loss": 0.0496, + "step": 4186 + }, + { + "epoch": 1.9703529411764706, + "grad_norm": 4.076347467076877e-06, + "learning_rate": 0.00027039212202458905, + "loss": 0.0, + "step": 4187 + }, + { + "epoch": 1.9708235294117649, + "grad_norm": 0.13264080289263325, + "learning_rate": 0.0002703659686677559, + "loss": 0.0487, + "step": 4188 + }, + { + "epoch": 1.9712941176470589, + "grad_norm": 0.0689746687869878, + "learning_rate": 0.00027033980503107856, + "loss": 0.0373, + "step": 4189 + }, + { + "epoch": 1.9717647058823529, + "grad_norm": 0.21421941584575813, + "learning_rate": 0.00027031363111679164, + "loss": 0.1175, + "step": 4190 + }, + { + "epoch": 1.972235294117647, + "grad_norm": 0.05826868864243311, + "learning_rate": 0.0002702874469271305, + "loss": 0.0167, + "step": 4191 + }, + { + "epoch": 1.9727058823529413, + "grad_norm": 0.09893953678962907, + "learning_rate": 0.00027026125246433144, + "loss": 0.0424, + "step": 4192 + }, + { + "epoch": 1.9731764705882353, + "grad_norm": 0.06389742624683434, + "learning_rate": 0.0002702350477306315, + "loss": 0.0348, + "step": 4193 + }, + { + "epoch": 1.9736470588235293, + "grad_norm": 0.06521231659217697, + "learning_rate": 0.0002702088327282687, + "loss": 0.0344, + "step": 4194 + }, + { + "epoch": 1.9741176470588235, + "grad_norm": 1.1350616397794199e-05, + "learning_rate": 0.00027018260745948196, + "loss": 0.0, + "step": 4195 + }, + { + "epoch": 1.9745882352941178, + "grad_norm": 0.1179497578067731, + "learning_rate": 0.0002701563719265111, + "loss": 0.0591, + "step": 4196 + }, + { + "epoch": 1.9750588235294118, + "grad_norm": 0.18327857854575674, + "learning_rate": 0.0002701301261315967, + "loss": 0.0652, + "step": 4197 + }, + { + "epoch": 1.9755294117647058, + "grad_norm": 0.10095459070588038, + "learning_rate": 0.0002701038700769803, + "loss": 0.0393, + "step": 4198 + }, + { + "epoch": 1.976, + "grad_norm": 0.06328030482578434, + "learning_rate": 0.00027007760376490425, + "loss": 0.0445, + "step": 4199 + }, + { + "epoch": 1.9764705882352942, + "grad_norm": 0.10264425885706918, + "learning_rate": 0.0002700513271976119, + "loss": 0.0962, + "step": 4200 + }, + { + "epoch": 1.9769411764705882, + "grad_norm": 0.054721496103803696, + "learning_rate": 0.0002700250403773473, + "loss": 0.0251, + "step": 4201 + }, + { + "epoch": 1.9774117647058822, + "grad_norm": 0.056915593161197875, + "learning_rate": 0.0002699987433063556, + "loss": 0.0201, + "step": 4202 + }, + { + "epoch": 1.9778823529411764, + "grad_norm": 0.1339784548563976, + "learning_rate": 0.0002699724359868826, + "loss": 0.1128, + "step": 4203 + }, + { + "epoch": 1.9783529411764706, + "grad_norm": 0.04216687119176406, + "learning_rate": 0.0002699461184211751, + "loss": 0.0194, + "step": 4204 + }, + { + "epoch": 1.9788235294117649, + "grad_norm": 0.09134667548056759, + "learning_rate": 0.00026991979061148077, + "loss": 0.0564, + "step": 4205 + }, + { + "epoch": 1.9792941176470589, + "grad_norm": 0.07019519659332599, + "learning_rate": 0.00026989345256004814, + "loss": 0.038, + "step": 4206 + }, + { + "epoch": 1.9797647058823529, + "grad_norm": 0.0657508914354131, + "learning_rate": 0.00026986710426912656, + "loss": 0.0232, + "step": 4207 + }, + { + "epoch": 1.980235294117647, + "grad_norm": 0.03579860152608468, + "learning_rate": 0.0002698407457409663, + "loss": 0.0122, + "step": 4208 + }, + { + "epoch": 1.9807058823529413, + "grad_norm": 0.11598521075525711, + "learning_rate": 0.00026981437697781854, + "loss": 0.0418, + "step": 4209 + }, + { + "epoch": 1.9811764705882353, + "grad_norm": 0.15318298740113448, + "learning_rate": 0.0002697879979819354, + "loss": 0.1183, + "step": 4210 + }, + { + "epoch": 1.9816470588235293, + "grad_norm": 0.027892306006369463, + "learning_rate": 0.0002697616087555696, + "loss": 0.0058, + "step": 4211 + }, + { + "epoch": 1.9821176470588235, + "grad_norm": 0.018649616755542366, + "learning_rate": 0.00026973520930097494, + "loss": 0.0017, + "step": 4212 + }, + { + "epoch": 1.9825882352941178, + "grad_norm": 0.05032002728639621, + "learning_rate": 0.0002697087996204062, + "loss": 0.025, + "step": 4213 + }, + { + "epoch": 1.9830588235294118, + "grad_norm": 0.05405376431687241, + "learning_rate": 0.0002696823797161187, + "loss": 0.0204, + "step": 4214 + }, + { + "epoch": 1.9835294117647058, + "grad_norm": 0.12073121846435274, + "learning_rate": 0.00026965594959036906, + "loss": 0.1184, + "step": 4215 + }, + { + "epoch": 1.984, + "grad_norm": 0.05845932690436106, + "learning_rate": 0.00026962950924541435, + "loss": 0.0284, + "step": 4216 + }, + { + "epoch": 1.9844705882352942, + "grad_norm": 0.034956967164636865, + "learning_rate": 0.0002696030586835128, + "loss": 0.0081, + "step": 4217 + }, + { + "epoch": 1.9849411764705882, + "grad_norm": 0.08411722798018198, + "learning_rate": 0.00026957659790692335, + "loss": 0.0434, + "step": 4218 + }, + { + "epoch": 1.9854117647058822, + "grad_norm": 0.05995489548770339, + "learning_rate": 0.0002695501269179059, + "loss": 0.0393, + "step": 4219 + }, + { + "epoch": 1.9858823529411764, + "grad_norm": 0.08786558418094749, + "learning_rate": 0.00026952364571872125, + "loss": 0.0572, + "step": 4220 + }, + { + "epoch": 1.9863529411764707, + "grad_norm": 0.02627315907534494, + "learning_rate": 0.000269497154311631, + "loss": 0.0066, + "step": 4221 + }, + { + "epoch": 1.9868235294117647, + "grad_norm": 0.04750418154050746, + "learning_rate": 0.00026947065269889765, + "loss": 0.0235, + "step": 4222 + }, + { + "epoch": 1.9872941176470589, + "grad_norm": 0.07747645345980012, + "learning_rate": 0.00026944414088278453, + "loss": 0.0368, + "step": 4223 + }, + { + "epoch": 1.9877647058823529, + "grad_norm": 0.03504059804616987, + "learning_rate": 0.00026941761886555595, + "loss": 0.0093, + "step": 4224 + }, + { + "epoch": 1.988235294117647, + "grad_norm": 0.11704224273290253, + "learning_rate": 0.00026939108664947694, + "loss": 0.0947, + "step": 4225 + }, + { + "epoch": 1.9887058823529413, + "grad_norm": 0.12006232489830515, + "learning_rate": 0.0002693645442368135, + "loss": 0.0775, + "step": 4226 + }, + { + "epoch": 1.9891764705882353, + "grad_norm": 0.09228315025091342, + "learning_rate": 0.0002693379916298326, + "loss": 0.0559, + "step": 4227 + }, + { + "epoch": 1.9896470588235293, + "grad_norm": 0.057458993738962696, + "learning_rate": 0.00026931142883080176, + "loss": 0.0202, + "step": 4228 + }, + { + "epoch": 1.9901176470588235, + "grad_norm": 0.04933022328983036, + "learning_rate": 0.00026928485584198973, + "loss": 0.0178, + "step": 4229 + }, + { + "epoch": 1.9905882352941178, + "grad_norm": 0.0894979577058911, + "learning_rate": 0.00026925827266566595, + "loss": 0.056, + "step": 4230 + }, + { + "epoch": 1.9910588235294118, + "grad_norm": 1.1645104602102389e-05, + "learning_rate": 0.0002692316793041007, + "loss": 0.0, + "step": 4231 + }, + { + "epoch": 1.9915294117647058, + "grad_norm": 0.07858240566759088, + "learning_rate": 0.0002692050757595652, + "loss": 0.0384, + "step": 4232 + }, + { + "epoch": 1.992, + "grad_norm": 0.09054970088774639, + "learning_rate": 0.00026917846203433155, + "loss": 0.0417, + "step": 4233 + }, + { + "epoch": 1.9924705882352942, + "grad_norm": 6.581621242234297e-06, + "learning_rate": 0.00026915183813067264, + "loss": 0.0, + "step": 4234 + }, + { + "epoch": 1.9929411764705882, + "grad_norm": 2.5216265826342638e-05, + "learning_rate": 0.0002691252040508623, + "loss": 0.0, + "step": 4235 + }, + { + "epoch": 1.9934117647058822, + "grad_norm": 0.11348274908896262, + "learning_rate": 0.0002690985597971753, + "loss": 0.0916, + "step": 4236 + }, + { + "epoch": 1.9938823529411764, + "grad_norm": 0.048963203307769855, + "learning_rate": 0.00026907190537188707, + "loss": 0.0204, + "step": 4237 + }, + { + "epoch": 1.9943529411764707, + "grad_norm": 0.08652645339216634, + "learning_rate": 0.0002690452407772741, + "loss": 0.0541, + "step": 4238 + }, + { + "epoch": 1.9948235294117647, + "grad_norm": 1.8183967203180375e-05, + "learning_rate": 0.0002690185660156136, + "loss": 0.0, + "step": 4239 + }, + { + "epoch": 1.9952941176470587, + "grad_norm": 0.11234192638400381, + "learning_rate": 0.00026899188108918387, + "loss": 0.095, + "step": 4240 + }, + { + "epoch": 1.9957647058823529, + "grad_norm": 0.06713234230938156, + "learning_rate": 0.00026896518600026377, + "loss": 0.039, + "step": 4241 + }, + { + "epoch": 1.996235294117647, + "grad_norm": 0.129271316275825, + "learning_rate": 0.00026893848075113327, + "loss": 0.0693, + "step": 4242 + }, + { + "epoch": 1.9967058823529413, + "grad_norm": 0.09005363073725922, + "learning_rate": 0.00026891176534407315, + "loss": 0.0525, + "step": 4243 + }, + { + "epoch": 1.9971764705882353, + "grad_norm": 0.09661161106978447, + "learning_rate": 0.00026888503978136506, + "loss": 0.0637, + "step": 4244 + }, + { + "epoch": 1.9976470588235293, + "grad_norm": 0.030087191345088332, + "learning_rate": 0.00026885830406529136, + "loss": 0.0051, + "step": 4245 + }, + { + "epoch": 1.9981176470588236, + "grad_norm": 0.09217517208244617, + "learning_rate": 0.0002688315581981355, + "loss": 0.048, + "step": 4246 + }, + { + "epoch": 1.9985882352941178, + "grad_norm": 0.06895339286617275, + "learning_rate": 0.0002688048021821817, + "loss": 0.0522, + "step": 4247 + }, + { + "epoch": 1.9990588235294118, + "grad_norm": 0.07854843761701509, + "learning_rate": 0.00026877803601971516, + "loss": 0.0328, + "step": 4248 + }, + { + "epoch": 1.9995294117647058, + "grad_norm": 0.10223507086107178, + "learning_rate": 0.00026875125971302167, + "loss": 0.0755, + "step": 4249 + }, + { + "epoch": 2.0, + "grad_norm": 0.1213149818609629, + "learning_rate": 0.0002687244732643881, + "loss": 0.0677, + "step": 4250 + }, + { + "epoch": 2.0, + "eval_loss": 0.04692387953400612, + "eval_runtime": 58.8339, + "eval_samples_per_second": 12.748, + "eval_steps_per_second": 6.374, + "step": 4250 + }, + { + "epoch": 2.0004705882352942, + "grad_norm": 0.07231951935212876, + "learning_rate": 0.0002686976766761022, + "loss": 0.0464, + "step": 4251 + }, + { + "epoch": 2.0009411764705884, + "grad_norm": 0.036381553001238444, + "learning_rate": 0.0002686708699504525, + "loss": 0.0149, + "step": 4252 + }, + { + "epoch": 2.0014117647058822, + "grad_norm": 0.08667289682854228, + "learning_rate": 0.0002686440530897284, + "loss": 0.0484, + "step": 4253 + }, + { + "epoch": 2.0018823529411764, + "grad_norm": 0.08127977177389402, + "learning_rate": 0.00026861722609622027, + "loss": 0.0586, + "step": 4254 + }, + { + "epoch": 2.0023529411764707, + "grad_norm": 0.023359028624510537, + "learning_rate": 0.0002685903889722192, + "loss": 0.0055, + "step": 4255 + }, + { + "epoch": 2.002823529411765, + "grad_norm": 6.35760419021832e-06, + "learning_rate": 0.00026856354172001724, + "loss": 0.0, + "step": 4256 + }, + { + "epoch": 2.0032941176470587, + "grad_norm": 0.05928051214599725, + "learning_rate": 0.0002685366843419073, + "loss": 0.0213, + "step": 4257 + }, + { + "epoch": 2.003764705882353, + "grad_norm": 0.1079992399838013, + "learning_rate": 0.000268509816840183, + "loss": 0.0841, + "step": 4258 + }, + { + "epoch": 2.004235294117647, + "grad_norm": 0.03694733230494985, + "learning_rate": 0.00026848293921713913, + "loss": 0.0117, + "step": 4259 + }, + { + "epoch": 2.0047058823529413, + "grad_norm": 0.09510899428931782, + "learning_rate": 0.0002684560514750711, + "loss": 0.0559, + "step": 4260 + }, + { + "epoch": 2.005176470588235, + "grad_norm": 3.433074388622679e-05, + "learning_rate": 0.00026842915361627523, + "loss": 0.0, + "step": 4261 + }, + { + "epoch": 2.0056470588235293, + "grad_norm": 0.044235057164324104, + "learning_rate": 0.0002684022456430487, + "loss": 0.0178, + "step": 4262 + }, + { + "epoch": 2.0061176470588236, + "grad_norm": 0.028212093362742, + "learning_rate": 0.0002683753275576897, + "loss": 0.0091, + "step": 4263 + }, + { + "epoch": 2.006588235294118, + "grad_norm": 1.1386729793921149e-05, + "learning_rate": 0.0002683483993624971, + "loss": 0.0, + "step": 4264 + }, + { + "epoch": 2.0070588235294116, + "grad_norm": 0.06019996883279864, + "learning_rate": 0.0002683214610597706, + "loss": 0.0228, + "step": 4265 + }, + { + "epoch": 2.007529411764706, + "grad_norm": 0.1029154163002759, + "learning_rate": 0.0002682945126518111, + "loss": 0.0611, + "step": 4266 + }, + { + "epoch": 2.008, + "grad_norm": 0.11370823436927013, + "learning_rate": 0.00026826755414091993, + "loss": 0.0908, + "step": 4267 + }, + { + "epoch": 2.0084705882352942, + "grad_norm": 0.10039747012961543, + "learning_rate": 0.00026824058552939954, + "loss": 0.0621, + "step": 4268 + }, + { + "epoch": 2.0089411764705885, + "grad_norm": 2.498750291755428e-05, + "learning_rate": 0.00026821360681955316, + "loss": 0.0, + "step": 4269 + }, + { + "epoch": 2.0094117647058822, + "grad_norm": 0.10812681653835499, + "learning_rate": 0.00026818661801368494, + "loss": 0.0765, + "step": 4270 + }, + { + "epoch": 2.0098823529411765, + "grad_norm": 0.05581974133549062, + "learning_rate": 0.0002681596191140998, + "loss": 0.0199, + "step": 4271 + }, + { + "epoch": 2.0103529411764707, + "grad_norm": 0.09645573473248567, + "learning_rate": 0.00026813261012310365, + "loss": 0.0386, + "step": 4272 + }, + { + "epoch": 2.010823529411765, + "grad_norm": 0.04926939609825981, + "learning_rate": 0.00026810559104300315, + "loss": 0.0142, + "step": 4273 + }, + { + "epoch": 2.0112941176470587, + "grad_norm": 8.945563170608527e-06, + "learning_rate": 0.0002680785618761059, + "loss": 0.0, + "step": 4274 + }, + { + "epoch": 2.011764705882353, + "grad_norm": 0.07763441162595287, + "learning_rate": 0.00026805152262472025, + "loss": 0.0457, + "step": 4275 + }, + { + "epoch": 2.012235294117647, + "grad_norm": 0.1818670222575831, + "learning_rate": 0.00026802447329115557, + "loss": 0.085, + "step": 4276 + }, + { + "epoch": 2.0127058823529413, + "grad_norm": 0.04601775782925704, + "learning_rate": 0.00026799741387772193, + "loss": 0.0181, + "step": 4277 + }, + { + "epoch": 2.013176470588235, + "grad_norm": 0.06558941583220683, + "learning_rate": 0.0002679703443867304, + "loss": 0.026, + "step": 4278 + }, + { + "epoch": 2.0136470588235293, + "grad_norm": 0.13155770232042924, + "learning_rate": 0.0002679432648204928, + "loss": 0.0832, + "step": 4279 + }, + { + "epoch": 2.0141176470588236, + "grad_norm": 0.027656048887619092, + "learning_rate": 0.00026791617518132186, + "loss": 0.0067, + "step": 4280 + }, + { + "epoch": 2.014588235294118, + "grad_norm": 0.09040366160165257, + "learning_rate": 0.00026788907547153126, + "loss": 0.0473, + "step": 4281 + }, + { + "epoch": 2.0150588235294116, + "grad_norm": 0.06356460184492216, + "learning_rate": 0.00026786196569343525, + "loss": 0.0401, + "step": 4282 + }, + { + "epoch": 2.015529411764706, + "grad_norm": 0.04648445697107479, + "learning_rate": 0.0002678348458493494, + "loss": 0.017, + "step": 4283 + }, + { + "epoch": 2.016, + "grad_norm": 0.04413007354048945, + "learning_rate": 0.00026780771594158966, + "loss": 0.0195, + "step": 4284 + }, + { + "epoch": 2.0164705882352942, + "grad_norm": 0.055737227124507294, + "learning_rate": 0.0002677805759724731, + "loss": 0.0275, + "step": 4285 + }, + { + "epoch": 2.016941176470588, + "grad_norm": 0.04126190475469072, + "learning_rate": 0.0002677534259443177, + "loss": 0.0102, + "step": 4286 + }, + { + "epoch": 2.0174117647058822, + "grad_norm": 0.09020952855028168, + "learning_rate": 0.00026772626585944213, + "loss": 0.0556, + "step": 4287 + }, + { + "epoch": 2.0178823529411765, + "grad_norm": 0.07068319313933444, + "learning_rate": 0.00026769909572016605, + "loss": 0.0415, + "step": 4288 + }, + { + "epoch": 2.0183529411764707, + "grad_norm": 0.07654658365136761, + "learning_rate": 0.0002676719155288098, + "loss": 0.0389, + "step": 4289 + }, + { + "epoch": 2.018823529411765, + "grad_norm": 0.171637218243804, + "learning_rate": 0.00026764472528769485, + "loss": 0.066, + "step": 4290 + }, + { + "epoch": 2.0192941176470587, + "grad_norm": 0.061691943318144774, + "learning_rate": 0.0002676175249991433, + "loss": 0.0236, + "step": 4291 + }, + { + "epoch": 2.019764705882353, + "grad_norm": 0.046234531140742774, + "learning_rate": 0.0002675903146654782, + "loss": 0.0183, + "step": 4292 + }, + { + "epoch": 2.020235294117647, + "grad_norm": 0.11808610270935363, + "learning_rate": 0.00026756309428902343, + "loss": 0.0491, + "step": 4293 + }, + { + "epoch": 2.0207058823529414, + "grad_norm": 0.101027151681517, + "learning_rate": 0.0002675358638721038, + "loss": 0.0596, + "step": 4294 + }, + { + "epoch": 2.021176470588235, + "grad_norm": 0.09842733406057343, + "learning_rate": 0.0002675086234170449, + "loss": 0.0739, + "step": 4295 + }, + { + "epoch": 2.0216470588235294, + "grad_norm": 0.017284118490379488, + "learning_rate": 0.0002674813729261731, + "loss": 0.0028, + "step": 4296 + }, + { + "epoch": 2.0221176470588236, + "grad_norm": 1.580916681289143e-05, + "learning_rate": 0.0002674541124018159, + "loss": 0.0, + "step": 4297 + }, + { + "epoch": 2.022588235294118, + "grad_norm": 0.061920592324056445, + "learning_rate": 0.00026742684184630135, + "loss": 0.0126, + "step": 4298 + }, + { + "epoch": 2.0230588235294116, + "grad_norm": 2.0649573193292863e-05, + "learning_rate": 0.00026739956126195855, + "loss": 0.0, + "step": 4299 + }, + { + "epoch": 2.023529411764706, + "grad_norm": 0.11393316830201954, + "learning_rate": 0.0002673722706511174, + "loss": 0.0534, + "step": 4300 + }, + { + "epoch": 2.024, + "grad_norm": 0.05417143231580649, + "learning_rate": 0.00026734497001610854, + "loss": 0.0139, + "step": 4301 + }, + { + "epoch": 2.0244705882352942, + "grad_norm": 0.03706907443728322, + "learning_rate": 0.00026731765935926374, + "loss": 0.0051, + "step": 4302 + }, + { + "epoch": 2.024941176470588, + "grad_norm": 0.08218360520560271, + "learning_rate": 0.00026729033868291537, + "loss": 0.0541, + "step": 4303 + }, + { + "epoch": 2.0254117647058822, + "grad_norm": 0.13648464387842846, + "learning_rate": 0.00026726300798939675, + "loss": 0.0742, + "step": 4304 + }, + { + "epoch": 2.0258823529411765, + "grad_norm": 0.055768931313765216, + "learning_rate": 0.0002672356672810421, + "loss": 0.0163, + "step": 4305 + }, + { + "epoch": 2.0263529411764707, + "grad_norm": 3.0036775855712755e-06, + "learning_rate": 0.00026720831656018644, + "loss": 0.0, + "step": 4306 + }, + { + "epoch": 2.026823529411765, + "grad_norm": 0.058621299534029836, + "learning_rate": 0.0002671809558291656, + "loss": 0.0223, + "step": 4307 + }, + { + "epoch": 2.0272941176470587, + "grad_norm": 0.06495048183344493, + "learning_rate": 0.00026715358509031643, + "loss": 0.0247, + "step": 4308 + }, + { + "epoch": 2.027764705882353, + "grad_norm": 0.0313687629020427, + "learning_rate": 0.00026712620434597636, + "loss": 0.0055, + "step": 4309 + }, + { + "epoch": 2.028235294117647, + "grad_norm": 3.22462820132792e-06, + "learning_rate": 0.00026709881359848406, + "loss": 0.0, + "step": 4310 + }, + { + "epoch": 2.0287058823529414, + "grad_norm": 0.16482386832041598, + "learning_rate": 0.0002670714128501786, + "loss": 0.1286, + "step": 4311 + }, + { + "epoch": 2.029176470588235, + "grad_norm": 0.09567359472248632, + "learning_rate": 0.0002670440021034003, + "loss": 0.0549, + "step": 4312 + }, + { + "epoch": 2.0296470588235294, + "grad_norm": 0.10061266963840564, + "learning_rate": 0.00026701658136049015, + "loss": 0.0694, + "step": 4313 + }, + { + "epoch": 2.0301176470588236, + "grad_norm": 0.05503834743618064, + "learning_rate": 0.00026698915062378994, + "loss": 0.021, + "step": 4314 + }, + { + "epoch": 2.030588235294118, + "grad_norm": 0.05235220437386775, + "learning_rate": 0.00026696170989564245, + "loss": 0.026, + "step": 4315 + }, + { + "epoch": 2.0310588235294116, + "grad_norm": 0.05585760827818094, + "learning_rate": 0.0002669342591783913, + "loss": 0.0195, + "step": 4316 + }, + { + "epoch": 2.031529411764706, + "grad_norm": 0.07920585157141809, + "learning_rate": 0.00026690679847438077, + "loss": 0.0508, + "step": 4317 + }, + { + "epoch": 2.032, + "grad_norm": 0.05210983497843751, + "learning_rate": 0.0002668793277859563, + "loss": 0.0251, + "step": 4318 + }, + { + "epoch": 2.0324705882352943, + "grad_norm": 0.054314770645397595, + "learning_rate": 0.0002668518471154639, + "loss": 0.0301, + "step": 4319 + }, + { + "epoch": 2.032941176470588, + "grad_norm": 0.11658285227965212, + "learning_rate": 0.0002668243564652506, + "loss": 0.0728, + "step": 4320 + }, + { + "epoch": 2.0334117647058823, + "grad_norm": 0.1197957641332053, + "learning_rate": 0.0002667968558376642, + "loss": 0.0882, + "step": 4321 + }, + { + "epoch": 2.0338823529411765, + "grad_norm": 0.07317051055787677, + "learning_rate": 0.0002667693452350535, + "loss": 0.0315, + "step": 4322 + }, + { + "epoch": 2.0343529411764707, + "grad_norm": 4.743120753433983e-06, + "learning_rate": 0.0002667418246597679, + "loss": 0.0, + "step": 4323 + }, + { + "epoch": 2.034823529411765, + "grad_norm": 0.03287217348695407, + "learning_rate": 0.0002667142941141579, + "loss": 0.0065, + "step": 4324 + }, + { + "epoch": 2.0352941176470587, + "grad_norm": 0.10772436141227111, + "learning_rate": 0.0002666867536005747, + "loss": 0.0659, + "step": 4325 + }, + { + "epoch": 2.035764705882353, + "grad_norm": 0.07017780627029385, + "learning_rate": 0.00026665920312137045, + "loss": 0.0289, + "step": 4326 + }, + { + "epoch": 2.036235294117647, + "grad_norm": 0.04610023926993795, + "learning_rate": 0.000266631642678898, + "loss": 0.016, + "step": 4327 + }, + { + "epoch": 2.0367058823529414, + "grad_norm": 0.04992434661512294, + "learning_rate": 0.00026660407227551113, + "loss": 0.0266, + "step": 4328 + }, + { + "epoch": 2.037176470588235, + "grad_norm": 0.04067633435626863, + "learning_rate": 0.0002665764919135646, + "loss": 0.0157, + "step": 4329 + }, + { + "epoch": 2.0376470588235294, + "grad_norm": 0.07888096636137143, + "learning_rate": 0.00026654890159541386, + "loss": 0.0439, + "step": 4330 + }, + { + "epoch": 2.0381176470588236, + "grad_norm": 0.14064454630176632, + "learning_rate": 0.0002665213013234153, + "loss": 0.0823, + "step": 4331 + }, + { + "epoch": 2.038588235294118, + "grad_norm": 0.0938034491225128, + "learning_rate": 0.00026649369109992596, + "loss": 0.0423, + "step": 4332 + }, + { + "epoch": 2.0390588235294116, + "grad_norm": 0.09043816982509291, + "learning_rate": 0.0002664660709273041, + "loss": 0.0479, + "step": 4333 + }, + { + "epoch": 2.039529411764706, + "grad_norm": 0.06006350146767809, + "learning_rate": 0.0002664384408079085, + "loss": 0.0213, + "step": 4334 + }, + { + "epoch": 2.04, + "grad_norm": 0.08281013091211988, + "learning_rate": 0.0002664108007440989, + "loss": 0.037, + "step": 4335 + }, + { + "epoch": 2.0404705882352943, + "grad_norm": 0.06417953116528283, + "learning_rate": 0.00026638315073823594, + "loss": 0.0276, + "step": 4336 + }, + { + "epoch": 2.040941176470588, + "grad_norm": 0.0613623233319625, + "learning_rate": 0.00026635549079268104, + "loss": 0.0236, + "step": 4337 + }, + { + "epoch": 2.0414117647058823, + "grad_norm": 0.08448210781593146, + "learning_rate": 0.00026632782090979653, + "loss": 0.0583, + "step": 4338 + }, + { + "epoch": 2.0418823529411765, + "grad_norm": 0.03431615732485712, + "learning_rate": 0.0002663001410919455, + "loss": 0.0091, + "step": 4339 + }, + { + "epoch": 2.0423529411764707, + "grad_norm": 0.04413108224309556, + "learning_rate": 0.000266272451341492, + "loss": 0.0181, + "step": 4340 + }, + { + "epoch": 2.0428235294117645, + "grad_norm": 0.30058260197197234, + "learning_rate": 0.0002662447516608008, + "loss": 0.0354, + "step": 4341 + }, + { + "epoch": 2.0432941176470587, + "grad_norm": 0.15171163039566066, + "learning_rate": 0.0002662170420522377, + "loss": 0.1083, + "step": 4342 + }, + { + "epoch": 2.043764705882353, + "grad_norm": 1.3264895277221483e-06, + "learning_rate": 0.0002661893225181691, + "loss": 0.0, + "step": 4343 + }, + { + "epoch": 2.044235294117647, + "grad_norm": 0.0930747293487921, + "learning_rate": 0.0002661615930609625, + "loss": 0.0443, + "step": 4344 + }, + { + "epoch": 2.0447058823529414, + "grad_norm": 0.021034355460603817, + "learning_rate": 0.000266133853682986, + "loss": 0.0032, + "step": 4345 + }, + { + "epoch": 2.045176470588235, + "grad_norm": 0.05239067070076499, + "learning_rate": 0.0002661061043866089, + "loss": 0.0222, + "step": 4346 + }, + { + "epoch": 2.0456470588235294, + "grad_norm": 0.049847391597538745, + "learning_rate": 0.00026607834517420085, + "loss": 0.0203, + "step": 4347 + }, + { + "epoch": 2.0461176470588236, + "grad_norm": 0.09829749600269012, + "learning_rate": 0.00026605057604813285, + "loss": 0.0569, + "step": 4348 + }, + { + "epoch": 2.046588235294118, + "grad_norm": 0.06543586961298899, + "learning_rate": 0.00026602279701077637, + "loss": 0.0387, + "step": 4349 + }, + { + "epoch": 2.0470588235294116, + "grad_norm": 1.1766783224833655e-06, + "learning_rate": 0.000265995008064504, + "loss": 0.0, + "step": 4350 + }, + { + "epoch": 2.047529411764706, + "grad_norm": 0.06747904441032976, + "learning_rate": 0.00026596720921168897, + "loss": 0.0302, + "step": 4351 + }, + { + "epoch": 2.048, + "grad_norm": 0.2841458443400488, + "learning_rate": 0.00026593940045470545, + "loss": 0.2054, + "step": 4352 + }, + { + "epoch": 2.0484705882352943, + "grad_norm": 0.07921246146947505, + "learning_rate": 0.00026591158179592846, + "loss": 0.0465, + "step": 4353 + }, + { + "epoch": 2.048941176470588, + "grad_norm": 0.06475852418321304, + "learning_rate": 0.0002658837532377338, + "loss": 0.0264, + "step": 4354 + }, + { + "epoch": 2.0494117647058823, + "grad_norm": 0.034601834308596875, + "learning_rate": 0.00026585591478249824, + "loss": 0.0056, + "step": 4355 + }, + { + "epoch": 2.0498823529411765, + "grad_norm": 0.11293731537685023, + "learning_rate": 0.0002658280664325993, + "loss": 0.068, + "step": 4356 + }, + { + "epoch": 2.0503529411764707, + "grad_norm": 0.03588591348772522, + "learning_rate": 0.00026580020819041533, + "loss": 0.009, + "step": 4357 + }, + { + "epoch": 2.0508235294117645, + "grad_norm": 3.2983402935805125e-06, + "learning_rate": 0.0002657723400583256, + "loss": 0.0, + "step": 4358 + }, + { + "epoch": 2.0512941176470587, + "grad_norm": 0.060504096481100146, + "learning_rate": 0.00026574446203871013, + "loss": 0.0192, + "step": 4359 + }, + { + "epoch": 2.051764705882353, + "grad_norm": 0.14599828319612396, + "learning_rate": 0.0002657165741339499, + "loss": 0.0892, + "step": 4360 + }, + { + "epoch": 2.052235294117647, + "grad_norm": 0.11489415822475874, + "learning_rate": 0.00026568867634642665, + "loss": 0.0752, + "step": 4361 + }, + { + "epoch": 2.0527058823529414, + "grad_norm": 0.07475710518778816, + "learning_rate": 0.00026566076867852304, + "loss": 0.0226, + "step": 4362 + }, + { + "epoch": 2.053176470588235, + "grad_norm": 0.012095368917058091, + "learning_rate": 0.0002656328511326224, + "loss": 0.0015, + "step": 4363 + }, + { + "epoch": 2.0536470588235294, + "grad_norm": 0.029249365584386785, + "learning_rate": 0.00026560492371110913, + "loss": 0.0067, + "step": 4364 + }, + { + "epoch": 2.0541176470588236, + "grad_norm": 2.481359097229509e-07, + "learning_rate": 0.00026557698641636835, + "loss": 0.0, + "step": 4365 + }, + { + "epoch": 2.054588235294118, + "grad_norm": 6.231839397822439e-06, + "learning_rate": 0.000265549039250786, + "loss": 0.0, + "step": 4366 + }, + { + "epoch": 2.0550588235294116, + "grad_norm": 0.06350671943858702, + "learning_rate": 0.000265521082216749, + "loss": 0.0353, + "step": 4367 + }, + { + "epoch": 2.055529411764706, + "grad_norm": 0.07778702942694744, + "learning_rate": 0.0002654931153166449, + "loss": 0.043, + "step": 4368 + }, + { + "epoch": 2.056, + "grad_norm": 0.07344495069600254, + "learning_rate": 0.0002654651385528622, + "loss": 0.0368, + "step": 4369 + }, + { + "epoch": 2.0564705882352943, + "grad_norm": 0.09977423602633514, + "learning_rate": 0.00026543715192779043, + "loss": 0.0545, + "step": 4370 + }, + { + "epoch": 2.056941176470588, + "grad_norm": 3.6994887932559905e-07, + "learning_rate": 0.0002654091554438196, + "loss": 0.0, + "step": 4371 + }, + { + "epoch": 2.0574117647058823, + "grad_norm": 0.4222065724220411, + "learning_rate": 0.0002653811491033408, + "loss": 0.079, + "step": 4372 + }, + { + "epoch": 2.0578823529411765, + "grad_norm": 2.3247137679322714e-07, + "learning_rate": 0.00026535313290874593, + "loss": 0.0, + "step": 4373 + }, + { + "epoch": 2.0583529411764707, + "grad_norm": 0.040700155178209844, + "learning_rate": 0.0002653251068624278, + "loss": 0.0088, + "step": 4374 + }, + { + "epoch": 2.0588235294117645, + "grad_norm": 1.7097361677714184e-06, + "learning_rate": 0.00026529707096677977, + "loss": 0.0, + "step": 4375 + }, + { + "epoch": 2.0592941176470587, + "grad_norm": 0.11945402904077285, + "learning_rate": 0.00026526902522419637, + "loss": 0.0452, + "step": 4376 + }, + { + "epoch": 2.059764705882353, + "grad_norm": 0.11302470905006912, + "learning_rate": 0.0002652409696370729, + "loss": 0.0479, + "step": 4377 + }, + { + "epoch": 2.060235294117647, + "grad_norm": 0.053443556714430275, + "learning_rate": 0.0002652129042078053, + "loss": 0.0235, + "step": 4378 + }, + { + "epoch": 2.0607058823529414, + "grad_norm": 1.215580434272339e-06, + "learning_rate": 0.00026518482893879056, + "loss": 0.0, + "step": 4379 + }, + { + "epoch": 2.061176470588235, + "grad_norm": 0.06777594558242964, + "learning_rate": 0.0002651567438324265, + "loss": 0.0395, + "step": 4380 + }, + { + "epoch": 2.0616470588235294, + "grad_norm": 0.5746999328384099, + "learning_rate": 0.00026512864889111163, + "loss": 0.1152, + "step": 4381 + }, + { + "epoch": 2.0621176470588236, + "grad_norm": 0.13972124983106685, + "learning_rate": 0.00026510054411724545, + "loss": 0.0532, + "step": 4382 + }, + { + "epoch": 2.062588235294118, + "grad_norm": 0.08992394829609006, + "learning_rate": 0.00026507242951322825, + "loss": 0.0504, + "step": 4383 + }, + { + "epoch": 2.0630588235294116, + "grad_norm": 0.05923755343959378, + "learning_rate": 0.00026504430508146116, + "loss": 0.0281, + "step": 4384 + }, + { + "epoch": 2.063529411764706, + "grad_norm": 0.04181962531461976, + "learning_rate": 0.00026501617082434607, + "loss": 0.0149, + "step": 4385 + }, + { + "epoch": 2.064, + "grad_norm": 2.3635989432161606e-05, + "learning_rate": 0.0002649880267442859, + "loss": 0.0, + "step": 4386 + }, + { + "epoch": 2.0644705882352943, + "grad_norm": 0.07307950610797187, + "learning_rate": 0.00026495987284368415, + "loss": 0.0294, + "step": 4387 + }, + { + "epoch": 2.064941176470588, + "grad_norm": 0.09308822893722384, + "learning_rate": 0.0002649317091249455, + "loss": 0.049, + "step": 4388 + }, + { + "epoch": 2.0654117647058823, + "grad_norm": 0.10247663251179248, + "learning_rate": 0.00026490353559047505, + "loss": 0.0816, + "step": 4389 + }, + { + "epoch": 2.0658823529411765, + "grad_norm": 0.09565499929301423, + "learning_rate": 0.0002648753522426791, + "loss": 0.0434, + "step": 4390 + }, + { + "epoch": 2.0663529411764707, + "grad_norm": 0.11603320032980625, + "learning_rate": 0.0002648471590839646, + "loss": 0.0744, + "step": 4391 + }, + { + "epoch": 2.0668235294117645, + "grad_norm": 5.455290582872181e-07, + "learning_rate": 0.00026481895611673944, + "loss": 0.0, + "step": 4392 + }, + { + "epoch": 2.0672941176470587, + "grad_norm": 0.3737539509743645, + "learning_rate": 0.0002647907433434122, + "loss": 0.3144, + "step": 4393 + }, + { + "epoch": 2.067764705882353, + "grad_norm": 0.06509134523395622, + "learning_rate": 0.00026476252076639245, + "loss": 0.0365, + "step": 4394 + }, + { + "epoch": 2.068235294117647, + "grad_norm": 0.06436671577012351, + "learning_rate": 0.0002647342883880905, + "loss": 0.0227, + "step": 4395 + }, + { + "epoch": 2.0687058823529414, + "grad_norm": 0.039988044985963006, + "learning_rate": 0.0002647060462109176, + "loss": 0.0195, + "step": 4396 + }, + { + "epoch": 2.069176470588235, + "grad_norm": 0.11821234635525428, + "learning_rate": 0.00026467779423728564, + "loss": 0.075, + "step": 4397 + }, + { + "epoch": 2.0696470588235294, + "grad_norm": 0.1079784328048469, + "learning_rate": 0.0002646495324696076, + "loss": 0.0444, + "step": 4398 + }, + { + "epoch": 2.0701176470588236, + "grad_norm": 0.06658332040435734, + "learning_rate": 0.0002646212609102972, + "loss": 0.0211, + "step": 4399 + }, + { + "epoch": 2.070588235294118, + "grad_norm": 0.014963074112794812, + "learning_rate": 0.00026459297956176885, + "loss": 0.0018, + "step": 4400 + }, + { + "epoch": 2.0710588235294116, + "grad_norm": 0.1318497684294005, + "learning_rate": 0.000264564688426438, + "loss": 0.0878, + "step": 4401 + }, + { + "epoch": 2.071529411764706, + "grad_norm": 0.0009687445513764686, + "learning_rate": 0.0002645363875067208, + "loss": 0.0, + "step": 4402 + }, + { + "epoch": 2.072, + "grad_norm": 0.00946077897597067, + "learning_rate": 0.0002645080768050342, + "loss": 0.0001, + "step": 4403 + }, + { + "epoch": 2.0724705882352943, + "grad_norm": 0.02593496401377522, + "learning_rate": 0.00026447975632379637, + "loss": 0.0003, + "step": 4404 + }, + { + "epoch": 2.072941176470588, + "grad_norm": 0.0998541378877906, + "learning_rate": 0.0002644514260654257, + "loss": 0.0503, + "step": 4405 + }, + { + "epoch": 2.0734117647058823, + "grad_norm": 0.13895021771200625, + "learning_rate": 0.0002644230860323419, + "loss": 0.0806, + "step": 4406 + }, + { + "epoch": 2.0738823529411765, + "grad_norm": 0.07756731894287563, + "learning_rate": 0.00026439473622696534, + "loss": 0.0015, + "step": 4407 + }, + { + "epoch": 2.0743529411764707, + "grad_norm": 0.0694290050271367, + "learning_rate": 0.0002643663766517172, + "loss": 0.0316, + "step": 4408 + }, + { + "epoch": 2.0748235294117645, + "grad_norm": 0.052637607507802955, + "learning_rate": 0.00026433800730901947, + "loss": 0.0205, + "step": 4409 + }, + { + "epoch": 2.0752941176470587, + "grad_norm": 6.030116017011612e-06, + "learning_rate": 0.00026430962820129513, + "loss": 0.0, + "step": 4410 + }, + { + "epoch": 2.075764705882353, + "grad_norm": 0.07904825902703796, + "learning_rate": 0.00026428123933096786, + "loss": 0.0585, + "step": 4411 + }, + { + "epoch": 2.076235294117647, + "grad_norm": 0.060038819806610966, + "learning_rate": 0.0002642528407004622, + "loss": 0.0257, + "step": 4412 + }, + { + "epoch": 2.0767058823529414, + "grad_norm": 0.03503761644357082, + "learning_rate": 0.00026422443231220353, + "loss": 0.0088, + "step": 4413 + }, + { + "epoch": 2.077176470588235, + "grad_norm": 0.1714507320822795, + "learning_rate": 0.0002641960141686181, + "loss": 0.1271, + "step": 4414 + }, + { + "epoch": 2.0776470588235294, + "grad_norm": 0.09382656734670608, + "learning_rate": 0.0002641675862721329, + "loss": 0.0847, + "step": 4415 + }, + { + "epoch": 2.0781176470588236, + "grad_norm": 0.09284834001055244, + "learning_rate": 0.00026413914862517585, + "loss": 0.0667, + "step": 4416 + }, + { + "epoch": 2.078588235294118, + "grad_norm": 0.2084878499651227, + "learning_rate": 0.00026411070123017567, + "loss": 0.017, + "step": 4417 + }, + { + "epoch": 2.0790588235294116, + "grad_norm": 0.0195683283683931, + "learning_rate": 0.0002640822440895619, + "loss": 0.0034, + "step": 4418 + }, + { + "epoch": 2.079529411764706, + "grad_norm": 0.09664285551809657, + "learning_rate": 0.0002640537772057649, + "loss": 0.039, + "step": 4419 + }, + { + "epoch": 2.08, + "grad_norm": 0.011296147205886559, + "learning_rate": 0.0002640253005812159, + "loss": 0.0013, + "step": 4420 + }, + { + "epoch": 2.0804705882352943, + "grad_norm": 0.1602235485536568, + "learning_rate": 0.00026399681421834694, + "loss": 0.1111, + "step": 4421 + }, + { + "epoch": 2.080941176470588, + "grad_norm": 0.05541910928963491, + "learning_rate": 0.0002639683181195909, + "loss": 0.0276, + "step": 4422 + }, + { + "epoch": 2.0814117647058823, + "grad_norm": 0.08059486085782874, + "learning_rate": 0.00026393981228738144, + "loss": 0.0443, + "step": 4423 + }, + { + "epoch": 2.0818823529411765, + "grad_norm": 0.02407308396493884, + "learning_rate": 0.0002639112967241532, + "loss": 0.0056, + "step": 4424 + }, + { + "epoch": 2.0823529411764707, + "grad_norm": 0.058587616745577914, + "learning_rate": 0.00026388277143234146, + "loss": 0.0162, + "step": 4425 + }, + { + "epoch": 2.0828235294117645, + "grad_norm": 0.052344530868076354, + "learning_rate": 0.0002638542364143825, + "loss": 0.0311, + "step": 4426 + }, + { + "epoch": 2.0832941176470587, + "grad_norm": 0.12388821041810999, + "learning_rate": 0.00026382569167271317, + "loss": 0.125, + "step": 4427 + }, + { + "epoch": 2.083764705882353, + "grad_norm": 0.04895041317289911, + "learning_rate": 0.00026379713720977156, + "loss": 0.0097, + "step": 4428 + }, + { + "epoch": 2.084235294117647, + "grad_norm": 0.016898823589420423, + "learning_rate": 0.0002637685730279962, + "loss": 0.003, + "step": 4429 + }, + { + "epoch": 2.084705882352941, + "grad_norm": 0.02124185754666531, + "learning_rate": 0.00026373999912982676, + "loss": 0.0045, + "step": 4430 + }, + { + "epoch": 2.085176470588235, + "grad_norm": 0.13026216255531303, + "learning_rate": 0.00026371141551770344, + "loss": 0.0744, + "step": 4431 + }, + { + "epoch": 2.0856470588235294, + "grad_norm": 0.056098623153564495, + "learning_rate": 0.0002636828221940675, + "loss": 0.0274, + "step": 4432 + }, + { + "epoch": 2.0861176470588236, + "grad_norm": 0.10868931632063018, + "learning_rate": 0.00026365421916136093, + "loss": 0.0934, + "step": 4433 + }, + { + "epoch": 2.086588235294118, + "grad_norm": 0.09595098481717246, + "learning_rate": 0.00026362560642202655, + "loss": 0.0756, + "step": 4434 + }, + { + "epoch": 2.0870588235294116, + "grad_norm": 0.0469608082242632, + "learning_rate": 0.0002635969839785081, + "loss": 0.0245, + "step": 4435 + }, + { + "epoch": 2.087529411764706, + "grad_norm": 1.7384178077822203e-07, + "learning_rate": 0.00026356835183325, + "loss": 0.0, + "step": 4436 + }, + { + "epoch": 2.088, + "grad_norm": 5.942908145607621e-07, + "learning_rate": 0.00026353970998869763, + "loss": 0.0, + "step": 4437 + }, + { + "epoch": 2.0884705882352943, + "grad_norm": 0.08183791044975165, + "learning_rate": 0.0002635110584472971, + "loss": 0.0523, + "step": 4438 + }, + { + "epoch": 2.088941176470588, + "grad_norm": 0.05389757538410136, + "learning_rate": 0.00026348239721149535, + "loss": 0.0217, + "step": 4439 + }, + { + "epoch": 2.0894117647058823, + "grad_norm": 0.09633500996232627, + "learning_rate": 0.0002634537262837403, + "loss": 0.0519, + "step": 4440 + }, + { + "epoch": 2.0898823529411765, + "grad_norm": 0.09055881927593713, + "learning_rate": 0.00026342504566648055, + "loss": 0.0548, + "step": 4441 + }, + { + "epoch": 2.0903529411764707, + "grad_norm": 0.025897041825400792, + "learning_rate": 0.0002633963553621655, + "loss": 0.0049, + "step": 4442 + }, + { + "epoch": 2.0908235294117645, + "grad_norm": 1.303951548451649e-07, + "learning_rate": 0.0002633676553732456, + "loss": 0.0, + "step": 4443 + }, + { + "epoch": 2.0912941176470587, + "grad_norm": 0.02437368125106766, + "learning_rate": 0.0002633389457021718, + "loss": 0.0057, + "step": 4444 + }, + { + "epoch": 2.091764705882353, + "grad_norm": 0.19831159213159058, + "learning_rate": 0.0002633102263513961, + "loss": 0.102, + "step": 4445 + }, + { + "epoch": 2.092235294117647, + "grad_norm": 0.11580619099626484, + "learning_rate": 0.0002632814973233713, + "loss": 0.0839, + "step": 4446 + }, + { + "epoch": 2.092705882352941, + "grad_norm": 0.11031314191847297, + "learning_rate": 0.00026325275862055104, + "loss": 0.071, + "step": 4447 + }, + { + "epoch": 2.093176470588235, + "grad_norm": 0.08740730976295466, + "learning_rate": 0.0002632240102453896, + "loss": 0.0358, + "step": 4448 + }, + { + "epoch": 2.0936470588235294, + "grad_norm": 7.029084346773934e-08, + "learning_rate": 0.0002631952522003424, + "loss": 0.0, + "step": 4449 + }, + { + "epoch": 2.0941176470588236, + "grad_norm": 0.08223173790401425, + "learning_rate": 0.00026316648448786536, + "loss": 0.0374, + "step": 4450 + }, + { + "epoch": 2.094588235294118, + "grad_norm": 0.056595832745311166, + "learning_rate": 0.00026313770711041556, + "loss": 0.0346, + "step": 4451 + }, + { + "epoch": 2.0950588235294116, + "grad_norm": 0.40557896141003086, + "learning_rate": 0.0002631089200704506, + "loss": 0.037, + "step": 4452 + }, + { + "epoch": 2.095529411764706, + "grad_norm": 0.057882862457425605, + "learning_rate": 0.00026308012337042906, + "loss": 0.0327, + "step": 4453 + }, + { + "epoch": 2.096, + "grad_norm": 3.2381664685714765e-08, + "learning_rate": 0.0002630513170128103, + "loss": 0.0, + "step": 4454 + }, + { + "epoch": 2.0964705882352943, + "grad_norm": 0.0516792772727353, + "learning_rate": 0.0002630225010000546, + "loss": 0.0246, + "step": 4455 + }, + { + "epoch": 2.096941176470588, + "grad_norm": 0.06437566659572182, + "learning_rate": 0.0002629936753346229, + "loss": 0.0323, + "step": 4456 + }, + { + "epoch": 2.0974117647058823, + "grad_norm": 0.06897346057599796, + "learning_rate": 0.00026296484001897715, + "loss": 0.0319, + "step": 4457 + }, + { + "epoch": 2.0978823529411765, + "grad_norm": 0.04024313402112725, + "learning_rate": 0.00026293599505558, + "loss": 0.0167, + "step": 4458 + }, + { + "epoch": 2.0983529411764708, + "grad_norm": 0.13649419473899682, + "learning_rate": 0.00026290714044689483, + "loss": 0.0773, + "step": 4459 + }, + { + "epoch": 2.0988235294117645, + "grad_norm": 0.04368342782841081, + "learning_rate": 0.0002628782761953861, + "loss": 0.0106, + "step": 4460 + }, + { + "epoch": 2.0992941176470588, + "grad_norm": 0.09240405107251924, + "learning_rate": 0.0002628494023035189, + "loss": 0.0663, + "step": 4461 + }, + { + "epoch": 2.099764705882353, + "grad_norm": 6.299244329841698e-08, + "learning_rate": 0.00026282051877375923, + "loss": 0.0, + "step": 4462 + }, + { + "epoch": 2.100235294117647, + "grad_norm": 3.6212115192997854e-08, + "learning_rate": 0.0002627916256085739, + "loss": 0.0, + "step": 4463 + }, + { + "epoch": 2.100705882352941, + "grad_norm": 0.07535344109265517, + "learning_rate": 0.0002627627228104305, + "loss": 0.0298, + "step": 4464 + }, + { + "epoch": 2.101176470588235, + "grad_norm": 0.08764705744675082, + "learning_rate": 0.00026273381038179744, + "loss": 0.038, + "step": 4465 + }, + { + "epoch": 2.1016470588235294, + "grad_norm": 0.12755828037914382, + "learning_rate": 0.000262704888325144, + "loss": 0.1123, + "step": 4466 + }, + { + "epoch": 2.1021176470588236, + "grad_norm": 0.07655808450326648, + "learning_rate": 0.0002626759566429404, + "loss": 0.0385, + "step": 4467 + }, + { + "epoch": 2.1025882352941174, + "grad_norm": 0.09070009686406906, + "learning_rate": 0.0002626470153376573, + "loss": 0.0616, + "step": 4468 + }, + { + "epoch": 2.1030588235294116, + "grad_norm": 0.05184832212055325, + "learning_rate": 0.00026261806441176666, + "loss": 0.0135, + "step": 4469 + }, + { + "epoch": 2.103529411764706, + "grad_norm": 0.08971779159023996, + "learning_rate": 0.0002625891038677408, + "loss": 0.0472, + "step": 4470 + }, + { + "epoch": 2.104, + "grad_norm": 0.03487258930713617, + "learning_rate": 0.0002625601337080533, + "loss": 0.008, + "step": 4471 + }, + { + "epoch": 2.1044705882352943, + "grad_norm": 0.047434621823974434, + "learning_rate": 0.00026253115393517834, + "loss": 0.016, + "step": 4472 + }, + { + "epoch": 2.104941176470588, + "grad_norm": 0.05257487945518012, + "learning_rate": 0.0002625021645515908, + "loss": 0.0231, + "step": 4473 + }, + { + "epoch": 2.1054117647058823, + "grad_norm": 5.5923384734738633e-08, + "learning_rate": 0.00026247316555976657, + "loss": 0.0, + "step": 4474 + }, + { + "epoch": 2.1058823529411765, + "grad_norm": 5.5689484267988074e-08, + "learning_rate": 0.00026244415696218237, + "loss": 0.0, + "step": 4475 + }, + { + "epoch": 2.1063529411764708, + "grad_norm": 0.06471376256512543, + "learning_rate": 0.00026241513876131564, + "loss": 0.0219, + "step": 4476 + }, + { + "epoch": 2.1068235294117645, + "grad_norm": 0.07083718664237747, + "learning_rate": 0.00026238611095964466, + "loss": 0.0244, + "step": 4477 + }, + { + "epoch": 2.1072941176470588, + "grad_norm": 0.11049475862211028, + "learning_rate": 0.0002623570735596485, + "loss": 0.0441, + "step": 4478 + }, + { + "epoch": 2.107764705882353, + "grad_norm": 0.0246048139840697, + "learning_rate": 0.0002623280265638072, + "loss": 0.0036, + "step": 4479 + }, + { + "epoch": 2.108235294117647, + "grad_norm": 0.058364357148789285, + "learning_rate": 0.0002622989699746014, + "loss": 0.0229, + "step": 4480 + }, + { + "epoch": 2.108705882352941, + "grad_norm": 0.0724939434306462, + "learning_rate": 0.00026226990379451284, + "loss": 0.0308, + "step": 4481 + }, + { + "epoch": 2.109176470588235, + "grad_norm": 0.059476944423933704, + "learning_rate": 0.00026224082802602373, + "loss": 0.0214, + "step": 4482 + }, + { + "epoch": 2.1096470588235294, + "grad_norm": 0.06911141080428797, + "learning_rate": 0.00026221174267161744, + "loss": 0.0255, + "step": 4483 + }, + { + "epoch": 2.1101176470588237, + "grad_norm": 0.049402204902470194, + "learning_rate": 0.0002621826477337779, + "loss": 0.0207, + "step": 4484 + }, + { + "epoch": 2.1105882352941174, + "grad_norm": 0.08143591463633937, + "learning_rate": 0.00026215354321498997, + "loss": 0.0467, + "step": 4485 + }, + { + "epoch": 2.1110588235294117, + "grad_norm": 0.08707931797929754, + "learning_rate": 0.00026212442911773936, + "loss": 0.0333, + "step": 4486 + }, + { + "epoch": 2.111529411764706, + "grad_norm": 0.12695998717115625, + "learning_rate": 0.0002620953054445125, + "loss": 0.0841, + "step": 4487 + }, + { + "epoch": 2.112, + "grad_norm": 0.08561610307090796, + "learning_rate": 0.0002620661721977968, + "loss": 0.0481, + "step": 4488 + }, + { + "epoch": 2.1124705882352943, + "grad_norm": 0.14972417422905954, + "learning_rate": 0.00026203702938008025, + "loss": 0.1032, + "step": 4489 + }, + { + "epoch": 2.112941176470588, + "grad_norm": 0.2059974584796886, + "learning_rate": 0.00026200787699385186, + "loss": 0.1185, + "step": 4490 + }, + { + "epoch": 2.1134117647058823, + "grad_norm": 0.03123719256996157, + "learning_rate": 0.0002619787150416014, + "loss": 0.0081, + "step": 4491 + }, + { + "epoch": 2.1138823529411765, + "grad_norm": 3.1423043608011976e-07, + "learning_rate": 0.00026194954352581944, + "loss": 0.0, + "step": 4492 + }, + { + "epoch": 2.1143529411764708, + "grad_norm": 0.049230066653219356, + "learning_rate": 0.0002619203624489973, + "loss": 0.0291, + "step": 4493 + }, + { + "epoch": 2.1148235294117645, + "grad_norm": 0.09633087718651205, + "learning_rate": 0.00026189117181362733, + "loss": 0.0559, + "step": 4494 + }, + { + "epoch": 2.1152941176470588, + "grad_norm": 0.13085841045722074, + "learning_rate": 0.0002618619716222024, + "loss": 0.0834, + "step": 4495 + }, + { + "epoch": 2.115764705882353, + "grad_norm": 0.13268939465140436, + "learning_rate": 0.00026183276187721635, + "loss": 0.0884, + "step": 4496 + }, + { + "epoch": 2.116235294117647, + "grad_norm": 0.07700462410983741, + "learning_rate": 0.000261803542581164, + "loss": 0.0484, + "step": 4497 + }, + { + "epoch": 2.116705882352941, + "grad_norm": 1.783949192280422e-07, + "learning_rate": 0.0002617743137365407, + "loss": 0.0, + "step": 4498 + }, + { + "epoch": 2.117176470588235, + "grad_norm": 0.06873881889047548, + "learning_rate": 0.0002617450753458427, + "loss": 0.0407, + "step": 4499 + }, + { + "epoch": 2.1176470588235294, + "grad_norm": 0.2950189184543061, + "learning_rate": 0.00026171582741156725, + "loss": 0.1265, + "step": 4500 + }, + { + "epoch": 2.1181176470588237, + "grad_norm": 0.12850286326816346, + "learning_rate": 0.0002616865699362121, + "loss": 0.0958, + "step": 4501 + }, + { + "epoch": 2.1185882352941174, + "grad_norm": 0.04664889047900745, + "learning_rate": 0.0002616573029222761, + "loss": 0.0138, + "step": 4502 + }, + { + "epoch": 2.1190588235294117, + "grad_norm": 6.425865847704312e-08, + "learning_rate": 0.0002616280263722588, + "loss": 0.0, + "step": 4503 + }, + { + "epoch": 2.119529411764706, + "grad_norm": 0.048245257773200254, + "learning_rate": 0.00026159874028866044, + "loss": 0.0169, + "step": 4504 + }, + { + "epoch": 2.12, + "grad_norm": 0.09451431966401962, + "learning_rate": 0.0002615694446739823, + "loss": 0.0684, + "step": 4505 + }, + { + "epoch": 2.1204705882352943, + "grad_norm": 0.07224312490806789, + "learning_rate": 0.0002615401395307263, + "loss": 0.0066, + "step": 4506 + }, + { + "epoch": 2.120941176470588, + "grad_norm": 5.808414503887473e-07, + "learning_rate": 0.00026151082486139534, + "loss": 0.0, + "step": 4507 + }, + { + "epoch": 2.1214117647058823, + "grad_norm": 0.07139266361532288, + "learning_rate": 0.0002614815006684929, + "loss": 0.0277, + "step": 4508 + }, + { + "epoch": 2.1218823529411766, + "grad_norm": 0.08050293457339563, + "learning_rate": 0.0002614521669545236, + "loss": 0.035, + "step": 4509 + }, + { + "epoch": 2.1223529411764708, + "grad_norm": 0.052718763583820154, + "learning_rate": 0.0002614228237219925, + "loss": 0.0181, + "step": 4510 + }, + { + "epoch": 2.1228235294117646, + "grad_norm": 1.1191744943871812e-07, + "learning_rate": 0.00026139347097340574, + "loss": 0.0, + "step": 4511 + }, + { + "epoch": 2.1232941176470588, + "grad_norm": 1.4559162373913602e-07, + "learning_rate": 0.00026136410871127017, + "loss": 0.0, + "step": 4512 + }, + { + "epoch": 2.123764705882353, + "grad_norm": 1.8215689979305954e-07, + "learning_rate": 0.0002613347369380935, + "loss": 0.0, + "step": 4513 + }, + { + "epoch": 2.124235294117647, + "grad_norm": 3.0688547346921565e-07, + "learning_rate": 0.00026130535565638415, + "loss": 0.0, + "step": 4514 + }, + { + "epoch": 2.124705882352941, + "grad_norm": 0.10420944518777168, + "learning_rate": 0.00026127596486865155, + "loss": 0.0554, + "step": 4515 + }, + { + "epoch": 2.1251764705882352, + "grad_norm": 2.6054660726374733e-07, + "learning_rate": 0.00026124656457740574, + "loss": 0.0, + "step": 4516 + }, + { + "epoch": 2.1256470588235294, + "grad_norm": 0.10444077954973084, + "learning_rate": 0.00026121715478515757, + "loss": 0.0629, + "step": 4517 + }, + { + "epoch": 2.1261176470588237, + "grad_norm": 0.11665546913911136, + "learning_rate": 0.0002611877354944189, + "loss": 0.089, + "step": 4518 + }, + { + "epoch": 2.1265882352941174, + "grad_norm": 0.07204177964601581, + "learning_rate": 0.00026115830670770225, + "loss": 0.0261, + "step": 4519 + }, + { + "epoch": 2.1270588235294117, + "grad_norm": 0.20526379649128718, + "learning_rate": 0.00026112886842752104, + "loss": 0.0903, + "step": 4520 + }, + { + "epoch": 2.127529411764706, + "grad_norm": 0.040221887600603724, + "learning_rate": 0.0002610994206563893, + "loss": 0.0158, + "step": 4521 + }, + { + "epoch": 2.128, + "grad_norm": 0.0940570250035376, + "learning_rate": 0.0002610699633968222, + "loss": 0.031, + "step": 4522 + }, + { + "epoch": 2.1284705882352943, + "grad_norm": 0.07628968172537512, + "learning_rate": 0.00026104049665133533, + "loss": 0.0421, + "step": 4523 + }, + { + "epoch": 2.128941176470588, + "grad_norm": 1.8050373059867716e-07, + "learning_rate": 0.00026101102042244543, + "loss": 0.0, + "step": 4524 + }, + { + "epoch": 2.1294117647058823, + "grad_norm": 1.6922437005635112e-07, + "learning_rate": 0.0002609815347126699, + "loss": 0.0, + "step": 4525 + }, + { + "epoch": 2.1298823529411766, + "grad_norm": 0.031081850784765094, + "learning_rate": 0.00026095203952452694, + "loss": 0.0082, + "step": 4526 + }, + { + "epoch": 2.130352941176471, + "grad_norm": 0.059755747280895985, + "learning_rate": 0.00026092253486053555, + "loss": 0.0236, + "step": 4527 + }, + { + "epoch": 2.1308235294117646, + "grad_norm": 0.11151278176914244, + "learning_rate": 0.0002608930207232156, + "loss": 0.0599, + "step": 4528 + }, + { + "epoch": 2.131294117647059, + "grad_norm": 0.06965075096605003, + "learning_rate": 0.0002608634971150878, + "loss": 0.0305, + "step": 4529 + }, + { + "epoch": 2.131764705882353, + "grad_norm": 0.12887777109206164, + "learning_rate": 0.0002608339640386735, + "loss": 0.0709, + "step": 4530 + }, + { + "epoch": 2.1322352941176472, + "grad_norm": 0.11449715529273267, + "learning_rate": 0.00026080442149649516, + "loss": 0.0821, + "step": 4531 + }, + { + "epoch": 2.132705882352941, + "grad_norm": 2.529887629493867e-07, + "learning_rate": 0.0002607748694910756, + "loss": 0.0, + "step": 4532 + }, + { + "epoch": 2.1331764705882352, + "grad_norm": 2.9398382209526133e-07, + "learning_rate": 0.0002607453080249389, + "loss": 0.0, + "step": 4533 + }, + { + "epoch": 2.1336470588235295, + "grad_norm": 4.1925721129193234e-07, + "learning_rate": 0.0002607157371006097, + "loss": 0.0, + "step": 4534 + }, + { + "epoch": 2.1341176470588237, + "grad_norm": 0.0968749054794039, + "learning_rate": 0.00026068615672061353, + "loss": 0.07, + "step": 4535 + }, + { + "epoch": 2.1345882352941175, + "grad_norm": 0.08866043134794764, + "learning_rate": 0.0002606565668874766, + "loss": 0.0526, + "step": 4536 + }, + { + "epoch": 2.1350588235294117, + "grad_norm": 0.06121113814308381, + "learning_rate": 0.00026062696760372606, + "loss": 0.0261, + "step": 4537 + }, + { + "epoch": 2.135529411764706, + "grad_norm": 0.11671201448122227, + "learning_rate": 0.0002605973588718899, + "loss": 0.0939, + "step": 4538 + }, + { + "epoch": 2.136, + "grad_norm": 7.971756218929463e-07, + "learning_rate": 0.00026056774069449685, + "loss": 0.0, + "step": 4539 + }, + { + "epoch": 2.1364705882352943, + "grad_norm": 0.05043313730271342, + "learning_rate": 0.00026053811307407635, + "loss": 0.0174, + "step": 4540 + }, + { + "epoch": 2.136941176470588, + "grad_norm": 0.11773423042967744, + "learning_rate": 0.00026050847601315886, + "loss": 0.0608, + "step": 4541 + }, + { + "epoch": 2.1374117647058823, + "grad_norm": 0.10457463509579874, + "learning_rate": 0.00026047882951427543, + "loss": 0.0591, + "step": 4542 + }, + { + "epoch": 2.1378823529411766, + "grad_norm": 0.12126240690472743, + "learning_rate": 0.00026044917357995813, + "loss": 0.0726, + "step": 4543 + }, + { + "epoch": 2.138352941176471, + "grad_norm": 3.1332724681622475e-07, + "learning_rate": 0.00026041950821273965, + "loss": 0.0, + "step": 4544 + }, + { + "epoch": 2.1388235294117646, + "grad_norm": 3.3351717664456937e-07, + "learning_rate": 0.0002603898334151535, + "loss": 0.0, + "step": 4545 + }, + { + "epoch": 2.139294117647059, + "grad_norm": 0.1258919762068974, + "learning_rate": 0.00026036014918973414, + "loss": 0.0828, + "step": 4546 + }, + { + "epoch": 2.139764705882353, + "grad_norm": 8.224180401721551e-07, + "learning_rate": 0.00026033045553901674, + "loss": 0.0, + "step": 4547 + }, + { + "epoch": 2.1402352941176472, + "grad_norm": 0.0920674686389332, + "learning_rate": 0.0002603007524655373, + "loss": 0.0339, + "step": 4548 + }, + { + "epoch": 2.140705882352941, + "grad_norm": 0.052648302947022156, + "learning_rate": 0.0002602710399718325, + "loss": 0.0196, + "step": 4549 + }, + { + "epoch": 2.1411764705882352, + "grad_norm": 0.03730980029617056, + "learning_rate": 0.0002602413180604401, + "loss": 0.0123, + "step": 4550 + }, + { + "epoch": 2.1416470588235295, + "grad_norm": 0.022479209377170642, + "learning_rate": 0.0002602115867338983, + "loss": 0.0049, + "step": 4551 + }, + { + "epoch": 2.1421176470588237, + "grad_norm": 0.059144383463552115, + "learning_rate": 0.0002601818459947465, + "loss": 0.0324, + "step": 4552 + }, + { + "epoch": 2.1425882352941175, + "grad_norm": 0.11495624659706892, + "learning_rate": 0.0002601520958455246, + "loss": 0.0756, + "step": 4553 + }, + { + "epoch": 2.1430588235294117, + "grad_norm": 0.0538771968376287, + "learning_rate": 0.00026012233628877336, + "loss": 0.0194, + "step": 4554 + }, + { + "epoch": 2.143529411764706, + "grad_norm": 0.06818403843876021, + "learning_rate": 0.0002600925673270345, + "loss": 0.0287, + "step": 4555 + }, + { + "epoch": 2.144, + "grad_norm": 0.10128470547759391, + "learning_rate": 0.0002600627889628504, + "loss": 0.0506, + "step": 4556 + }, + { + "epoch": 2.1444705882352944, + "grad_norm": 0.10137457021369156, + "learning_rate": 0.00026003300119876424, + "loss": 0.0483, + "step": 4557 + }, + { + "epoch": 2.144941176470588, + "grad_norm": 0.07502907266239262, + "learning_rate": 0.0002600032040373201, + "loss": 0.0401, + "step": 4558 + }, + { + "epoch": 2.1454117647058824, + "grad_norm": 7.392744967756985e-07, + "learning_rate": 0.00025997339748106275, + "loss": 0.0, + "step": 4559 + }, + { + "epoch": 2.1458823529411766, + "grad_norm": 0.10277224789166794, + "learning_rate": 0.00025994358153253784, + "loss": 0.0759, + "step": 4560 + }, + { + "epoch": 2.1463529411764704, + "grad_norm": 0.1723424314408979, + "learning_rate": 0.0002599137561942918, + "loss": 0.0927, + "step": 4561 + }, + { + "epoch": 2.1468235294117646, + "grad_norm": 3.975334288704133e-07, + "learning_rate": 0.00025988392146887183, + "loss": 0.0, + "step": 4562 + }, + { + "epoch": 2.147294117647059, + "grad_norm": 1.4512416790439105e-06, + "learning_rate": 0.00025985407735882606, + "loss": 0.0, + "step": 4563 + }, + { + "epoch": 2.147764705882353, + "grad_norm": 3.290960391571486e-07, + "learning_rate": 0.0002598242238667032, + "loss": 0.0, + "step": 4564 + }, + { + "epoch": 2.1482352941176472, + "grad_norm": 0.08449720361094923, + "learning_rate": 0.00025979436099505296, + "loss": 0.0424, + "step": 4565 + }, + { + "epoch": 2.148705882352941, + "grad_norm": 0.08484834545621563, + "learning_rate": 0.0002597644887464257, + "loss": 0.0558, + "step": 4566 + }, + { + "epoch": 2.1491764705882352, + "grad_norm": 0.0662469649697176, + "learning_rate": 0.00025973460712337275, + "loss": 0.0345, + "step": 4567 + }, + { + "epoch": 2.1496470588235295, + "grad_norm": 0.1600041107090878, + "learning_rate": 0.0002597047161284461, + "loss": 0.0936, + "step": 4568 + }, + { + "epoch": 2.1501176470588237, + "grad_norm": 0.0698669326599608, + "learning_rate": 0.00025967481576419864, + "loss": 0.0224, + "step": 4569 + }, + { + "epoch": 2.1505882352941175, + "grad_norm": 0.20526024002334345, + "learning_rate": 0.0002596449060331839, + "loss": 0.1848, + "step": 4570 + }, + { + "epoch": 2.1510588235294117, + "grad_norm": 2.731750846526826e-07, + "learning_rate": 0.0002596149869379564, + "loss": 0.0, + "step": 4571 + }, + { + "epoch": 2.151529411764706, + "grad_norm": 0.11828085454998541, + "learning_rate": 0.00025958505848107133, + "loss": 0.0791, + "step": 4572 + }, + { + "epoch": 2.152, + "grad_norm": 0.06305461568695778, + "learning_rate": 0.00025955512066508477, + "loss": 0.0298, + "step": 4573 + }, + { + "epoch": 2.152470588235294, + "grad_norm": 0.08133697637484287, + "learning_rate": 0.0002595251734925536, + "loss": 0.0543, + "step": 4574 + }, + { + "epoch": 2.152941176470588, + "grad_norm": 2.0792711638471654e-07, + "learning_rate": 0.00025949521696603536, + "loss": 0.0, + "step": 4575 + }, + { + "epoch": 2.1534117647058824, + "grad_norm": 0.07111695680508433, + "learning_rate": 0.0002594652510880885, + "loss": 0.042, + "step": 4576 + }, + { + "epoch": 2.1538823529411766, + "grad_norm": 0.07098052690711137, + "learning_rate": 0.0002594352758612723, + "loss": 0.0293, + "step": 4577 + }, + { + "epoch": 2.1543529411764704, + "grad_norm": 0.0019326063465348363, + "learning_rate": 0.0002594052912881467, + "loss": 0.0001, + "step": 4578 + }, + { + "epoch": 2.1548235294117646, + "grad_norm": 0.05705507026265216, + "learning_rate": 0.00025937529737127263, + "loss": 0.0237, + "step": 4579 + }, + { + "epoch": 2.155294117647059, + "grad_norm": 0.11051575205305145, + "learning_rate": 0.0002593452941132117, + "loss": 0.0523, + "step": 4580 + }, + { + "epoch": 2.155764705882353, + "grad_norm": 0.12137557696319821, + "learning_rate": 0.0002593152815165263, + "loss": 0.0658, + "step": 4581 + }, + { + "epoch": 2.1562352941176472, + "grad_norm": 0.11539770440230393, + "learning_rate": 0.0002592852595837797, + "loss": 0.0608, + "step": 4582 + }, + { + "epoch": 2.156705882352941, + "grad_norm": 0.1575129013874146, + "learning_rate": 0.00025925522831753584, + "loss": 0.0583, + "step": 4583 + }, + { + "epoch": 2.1571764705882353, + "grad_norm": 5.450753683388117e-07, + "learning_rate": 0.00025922518772035963, + "loss": 0.0, + "step": 4584 + }, + { + "epoch": 2.1576470588235295, + "grad_norm": 0.10182051124923527, + "learning_rate": 0.0002591951377948166, + "loss": 0.0782, + "step": 4585 + }, + { + "epoch": 2.1581176470588237, + "grad_norm": 0.1241570130759474, + "learning_rate": 0.00025916507854347327, + "loss": 0.0684, + "step": 4586 + }, + { + "epoch": 2.1585882352941175, + "grad_norm": 0.05909283383969064, + "learning_rate": 0.00025913500996889677, + "loss": 0.0229, + "step": 4587 + }, + { + "epoch": 2.1590588235294117, + "grad_norm": 0.08115785753700205, + "learning_rate": 0.00025910493207365504, + "loss": 0.0343, + "step": 4588 + }, + { + "epoch": 2.159529411764706, + "grad_norm": 4.177169394801069e-07, + "learning_rate": 0.000259074844860317, + "loss": 0.0, + "step": 4589 + }, + { + "epoch": 2.16, + "grad_norm": 0.06413950330597823, + "learning_rate": 0.00025904474833145225, + "loss": 0.0317, + "step": 4590 + }, + { + "epoch": 2.160470588235294, + "grad_norm": 0.08539772732769364, + "learning_rate": 0.00025901464248963106, + "loss": 0.0335, + "step": 4591 + }, + { + "epoch": 2.160941176470588, + "grad_norm": 0.2079030572651457, + "learning_rate": 0.00025898452733742473, + "loss": 0.0666, + "step": 4592 + }, + { + "epoch": 2.1614117647058824, + "grad_norm": 0.04394872134040055, + "learning_rate": 0.00025895440287740514, + "loss": 0.0152, + "step": 4593 + }, + { + "epoch": 2.1618823529411766, + "grad_norm": 1.2198438744517245e-07, + "learning_rate": 0.00025892426911214517, + "loss": 0.0, + "step": 4594 + }, + { + "epoch": 2.1623529411764704, + "grad_norm": 3.946218391292183e-07, + "learning_rate": 0.0002588941260442183, + "loss": 0.0, + "step": 4595 + }, + { + "epoch": 2.1628235294117646, + "grad_norm": 2.063219844423567e-07, + "learning_rate": 0.00025886397367619904, + "loss": 0.0, + "step": 4596 + }, + { + "epoch": 2.163294117647059, + "grad_norm": 0.033330943066324926, + "learning_rate": 0.00025883381201066236, + "loss": 0.0098, + "step": 4597 + }, + { + "epoch": 2.163764705882353, + "grad_norm": 0.1358611011563956, + "learning_rate": 0.0002588036410501844, + "loss": 0.0899, + "step": 4598 + }, + { + "epoch": 2.1642352941176473, + "grad_norm": 0.10015984707422707, + "learning_rate": 0.00025877346079734175, + "loss": 0.0717, + "step": 4599 + }, + { + "epoch": 2.164705882352941, + "grad_norm": 0.09671714216412697, + "learning_rate": 0.000258743271254712, + "loss": 0.06, + "step": 4600 + }, + { + "epoch": 2.1651764705882353, + "grad_norm": 0.06186079155484037, + "learning_rate": 0.00025871307242487354, + "loss": 0.0276, + "step": 4601 + }, + { + "epoch": 2.1656470588235295, + "grad_norm": 0.06656596847180463, + "learning_rate": 0.00025868286431040546, + "loss": 0.0395, + "step": 4602 + }, + { + "epoch": 2.1661176470588237, + "grad_norm": 0.06579047900780358, + "learning_rate": 0.0002586526469138877, + "loss": 0.0384, + "step": 4603 + }, + { + "epoch": 2.1665882352941175, + "grad_norm": 0.09314125067481077, + "learning_rate": 0.0002586224202379009, + "loss": 0.0551, + "step": 4604 + }, + { + "epoch": 2.1670588235294117, + "grad_norm": 7.199000605816015e-07, + "learning_rate": 0.00025859218428502667, + "loss": 0.0, + "step": 4605 + }, + { + "epoch": 2.167529411764706, + "grad_norm": 0.03645303825867837, + "learning_rate": 0.00025856193905784725, + "loss": 0.0112, + "step": 4606 + }, + { + "epoch": 2.168, + "grad_norm": 1.9986810761041648e-07, + "learning_rate": 0.00025853168455894566, + "loss": 0.0, + "step": 4607 + }, + { + "epoch": 2.168470588235294, + "grad_norm": 4.338798088481312e-07, + "learning_rate": 0.000258501420790906, + "loss": 0.0, + "step": 4608 + }, + { + "epoch": 2.168941176470588, + "grad_norm": 1.7134766367880175e-07, + "learning_rate": 0.0002584711477563127, + "loss": 0.0, + "step": 4609 + }, + { + "epoch": 2.1694117647058824, + "grad_norm": 0.12898789502979044, + "learning_rate": 0.0002584408654577514, + "loss": 0.0794, + "step": 4610 + }, + { + "epoch": 2.1698823529411766, + "grad_norm": 0.05677313235764243, + "learning_rate": 0.0002584105738978083, + "loss": 0.0199, + "step": 4611 + }, + { + "epoch": 2.1703529411764704, + "grad_norm": 0.13418630553415753, + "learning_rate": 0.00025838027307907037, + "loss": 0.0995, + "step": 4612 + }, + { + "epoch": 2.1708235294117646, + "grad_norm": 0.11580263605921798, + "learning_rate": 0.00025834996300412556, + "loss": 0.0529, + "step": 4613 + }, + { + "epoch": 2.171294117647059, + "grad_norm": 0.057098901752937324, + "learning_rate": 0.0002583196436755624, + "loss": 0.0197, + "step": 4614 + }, + { + "epoch": 2.171764705882353, + "grad_norm": 0.03279465723606891, + "learning_rate": 0.00025828931509597045, + "loss": 0.0084, + "step": 4615 + }, + { + "epoch": 2.1722352941176473, + "grad_norm": 0.024156005080902595, + "learning_rate": 0.0002582589772679398, + "loss": 0.0048, + "step": 4616 + }, + { + "epoch": 2.172705882352941, + "grad_norm": 2.233452472294789e-07, + "learning_rate": 0.0002582286301940615, + "loss": 0.0, + "step": 4617 + }, + { + "epoch": 2.1731764705882353, + "grad_norm": 0.019751749746847343, + "learning_rate": 0.00025819827387692733, + "loss": 0.0032, + "step": 4618 + }, + { + "epoch": 2.1736470588235295, + "grad_norm": 0.06908714853758217, + "learning_rate": 0.0002581679083191298, + "loss": 0.031, + "step": 4619 + }, + { + "epoch": 2.1741176470588237, + "grad_norm": 0.0535481484031988, + "learning_rate": 0.0002581375335232624, + "loss": 0.0266, + "step": 4620 + }, + { + "epoch": 2.1745882352941175, + "grad_norm": 0.09581618049707215, + "learning_rate": 0.00025810714949191916, + "loss": 0.0683, + "step": 4621 + }, + { + "epoch": 2.1750588235294117, + "grad_norm": 0.14581284537918357, + "learning_rate": 0.00025807675622769514, + "loss": 0.1047, + "step": 4622 + }, + { + "epoch": 2.175529411764706, + "grad_norm": 0.07871176007764735, + "learning_rate": 0.00025804635373318603, + "loss": 0.041, + "step": 4623 + }, + { + "epoch": 2.176, + "grad_norm": 0.11789640047781617, + "learning_rate": 0.0002580159420109883, + "loss": 0.092, + "step": 4624 + }, + { + "epoch": 2.176470588235294, + "grad_norm": 0.0819524640030012, + "learning_rate": 0.00025798552106369937, + "loss": 0.0591, + "step": 4625 + }, + { + "epoch": 2.176941176470588, + "grad_norm": 3.187150718782e-07, + "learning_rate": 0.00025795509089391726, + "loss": 0.0, + "step": 4626 + }, + { + "epoch": 2.1774117647058824, + "grad_norm": 1.5726420836047016e-07, + "learning_rate": 0.00025792465150424083, + "loss": 0.0, + "step": 4627 + }, + { + "epoch": 2.1778823529411766, + "grad_norm": 0.08449035565266377, + "learning_rate": 0.00025789420289726983, + "loss": 0.0384, + "step": 4628 + }, + { + "epoch": 2.1783529411764704, + "grad_norm": 1.3306179462022364e-06, + "learning_rate": 0.00025786374507560467, + "loss": 0.0, + "step": 4629 + }, + { + "epoch": 2.1788235294117646, + "grad_norm": 0.1531216799731339, + "learning_rate": 0.0002578332780418466, + "loss": 0.0893, + "step": 4630 + }, + { + "epoch": 2.179294117647059, + "grad_norm": 0.12852209943304155, + "learning_rate": 0.0002578028017985977, + "loss": 0.0587, + "step": 4631 + }, + { + "epoch": 2.179764705882353, + "grad_norm": 0.027136879987502076, + "learning_rate": 0.00025777231634846074, + "loss": 0.0068, + "step": 4632 + }, + { + "epoch": 2.1802352941176473, + "grad_norm": 1.8463073096203398e-07, + "learning_rate": 0.00025774182169403933, + "loss": 0.0, + "step": 4633 + }, + { + "epoch": 2.180705882352941, + "grad_norm": 0.08522155727369063, + "learning_rate": 0.00025771131783793787, + "loss": 0.0584, + "step": 4634 + }, + { + "epoch": 2.1811764705882353, + "grad_norm": 0.03872340214791719, + "learning_rate": 0.00025768080478276163, + "loss": 0.0137, + "step": 4635 + }, + { + "epoch": 2.1816470588235295, + "grad_norm": 0.10984516410752496, + "learning_rate": 0.0002576502825311164, + "loss": 0.0928, + "step": 4636 + }, + { + "epoch": 2.1821176470588237, + "grad_norm": 0.07251879766540813, + "learning_rate": 0.000257619751085609, + "loss": 0.031, + "step": 4637 + }, + { + "epoch": 2.1825882352941175, + "grad_norm": 0.039088410813220134, + "learning_rate": 0.00025758921044884706, + "loss": 0.0096, + "step": 4638 + }, + { + "epoch": 2.1830588235294117, + "grad_norm": 0.06336767634030191, + "learning_rate": 0.0002575586606234388, + "loss": 0.0305, + "step": 4639 + }, + { + "epoch": 2.183529411764706, + "grad_norm": 0.08557131159721866, + "learning_rate": 0.0002575281016119933, + "loss": 0.0572, + "step": 4640 + }, + { + "epoch": 2.184, + "grad_norm": 0.07277056566440467, + "learning_rate": 0.00025749753341712056, + "loss": 0.0438, + "step": 4641 + }, + { + "epoch": 2.184470588235294, + "grad_norm": 1.8559700188590413e-06, + "learning_rate": 0.0002574669560414312, + "loss": 0.0, + "step": 4642 + }, + { + "epoch": 2.184941176470588, + "grad_norm": 0.05194750869036618, + "learning_rate": 0.0002574363694875367, + "loss": 0.0268, + "step": 4643 + }, + { + "epoch": 2.1854117647058824, + "grad_norm": 0.33515140770148344, + "learning_rate": 0.0002574057737580492, + "loss": 0.1656, + "step": 4644 + }, + { + "epoch": 2.1858823529411766, + "grad_norm": 2.681084254486598e-07, + "learning_rate": 0.0002573751688555818, + "loss": 0.0, + "step": 4645 + }, + { + "epoch": 2.1863529411764704, + "grad_norm": 2.1028332679243272e-07, + "learning_rate": 0.0002573445547827484, + "loss": 0.0, + "step": 4646 + }, + { + "epoch": 2.1868235294117646, + "grad_norm": 0.06296077516029233, + "learning_rate": 0.0002573139315421634, + "loss": 0.0337, + "step": 4647 + }, + { + "epoch": 2.187294117647059, + "grad_norm": 0.07591879923776887, + "learning_rate": 0.0002572832991364423, + "loss": 0.0358, + "step": 4648 + }, + { + "epoch": 2.187764705882353, + "grad_norm": 0.13347096306913897, + "learning_rate": 0.00025725265756820125, + "loss": 0.0741, + "step": 4649 + }, + { + "epoch": 2.1882352941176473, + "grad_norm": 0.06134409040180339, + "learning_rate": 0.00025722200684005715, + "loss": 0.0235, + "step": 4650 + }, + { + "epoch": 2.188705882352941, + "grad_norm": 4.1709851245286495e-07, + "learning_rate": 0.0002571913469546277, + "loss": 0.0, + "step": 4651 + }, + { + "epoch": 2.1891764705882353, + "grad_norm": 0.011717120137831334, + "learning_rate": 0.00025716067791453157, + "loss": 0.0014, + "step": 4652 + }, + { + "epoch": 2.1896470588235295, + "grad_norm": 0.06002063788328228, + "learning_rate": 0.0002571299997223878, + "loss": 0.0256, + "step": 4653 + }, + { + "epoch": 2.1901176470588237, + "grad_norm": 0.05647554635065048, + "learning_rate": 0.0002570993123808167, + "loss": 0.028, + "step": 4654 + }, + { + "epoch": 2.1905882352941175, + "grad_norm": 0.037030278637625175, + "learning_rate": 0.00025706861589243894, + "loss": 0.0122, + "step": 4655 + }, + { + "epoch": 2.1910588235294117, + "grad_norm": 0.1767672155496198, + "learning_rate": 0.0002570379102598762, + "loss": 0.0929, + "step": 4656 + }, + { + "epoch": 2.191529411764706, + "grad_norm": 2.3597541374523895e-07, + "learning_rate": 0.000257007195485751, + "loss": 0.0, + "step": 4657 + }, + { + "epoch": 2.192, + "grad_norm": 1.7519588880339158e-07, + "learning_rate": 0.00025697647157268635, + "loss": 0.0, + "step": 4658 + }, + { + "epoch": 2.192470588235294, + "grad_norm": 4.028590066910665e-07, + "learning_rate": 0.0002569457385233063, + "loss": 0.0, + "step": 4659 + }, + { + "epoch": 2.192941176470588, + "grad_norm": 0.1530455769113019, + "learning_rate": 0.00025691499634023564, + "loss": 0.1168, + "step": 4660 + }, + { + "epoch": 2.1934117647058824, + "grad_norm": 0.13033677443648245, + "learning_rate": 0.0002568842450260999, + "loss": 0.0469, + "step": 4661 + }, + { + "epoch": 2.1938823529411766, + "grad_norm": 0.15444241615621113, + "learning_rate": 0.0002568534845835253, + "loss": 0.0235, + "step": 4662 + }, + { + "epoch": 2.1943529411764704, + "grad_norm": 0.036356991670356976, + "learning_rate": 0.0002568227150151391, + "loss": 0.0096, + "step": 4663 + }, + { + "epoch": 2.1948235294117646, + "grad_norm": 4.0600479499401925e-06, + "learning_rate": 0.000256791936323569, + "loss": 0.0, + "step": 4664 + }, + { + "epoch": 2.195294117647059, + "grad_norm": 0.07422942480709468, + "learning_rate": 0.00025676114851144375, + "loss": 0.0456, + "step": 4665 + }, + { + "epoch": 2.195764705882353, + "grad_norm": 0.04817318152173235, + "learning_rate": 0.0002567303515813928, + "loss": 0.0223, + "step": 4666 + }, + { + "epoch": 2.1962352941176473, + "grad_norm": 0.09185028434879647, + "learning_rate": 0.0002566995455360463, + "loss": 0.046, + "step": 4667 + }, + { + "epoch": 2.196705882352941, + "grad_norm": 0.001236016425806199, + "learning_rate": 0.00025666873037803517, + "loss": 0.0, + "step": 4668 + }, + { + "epoch": 2.1971764705882353, + "grad_norm": 0.07305638117961102, + "learning_rate": 0.00025663790610999134, + "loss": 0.0508, + "step": 4669 + }, + { + "epoch": 2.1976470588235295, + "grad_norm": 0.03411446781930397, + "learning_rate": 0.0002566070727345473, + "loss": 0.01, + "step": 4670 + }, + { + "epoch": 2.1981176470588237, + "grad_norm": 3.478758963215935e-07, + "learning_rate": 0.00025657623025433627, + "loss": 0.0, + "step": 4671 + }, + { + "epoch": 2.1985882352941175, + "grad_norm": 2.441201754181305e-07, + "learning_rate": 0.0002565453786719924, + "loss": 0.0, + "step": 4672 + }, + { + "epoch": 2.1990588235294117, + "grad_norm": 0.09103058550625683, + "learning_rate": 0.00025651451799015067, + "loss": 0.0497, + "step": 4673 + }, + { + "epoch": 2.199529411764706, + "grad_norm": 1.2182117035946855e-07, + "learning_rate": 0.0002564836482114466, + "loss": 0.0, + "step": 4674 + }, + { + "epoch": 2.2, + "grad_norm": 0.06187072515775146, + "learning_rate": 0.00025645276933851667, + "loss": 0.036, + "step": 4675 + }, + { + "epoch": 2.200470588235294, + "grad_norm": 0.08537037730203897, + "learning_rate": 0.00025642188137399815, + "loss": 0.0439, + "step": 4676 + }, + { + "epoch": 2.200941176470588, + "grad_norm": 0.0887518913842108, + "learning_rate": 0.00025639098432052895, + "loss": 0.0585, + "step": 4677 + }, + { + "epoch": 2.2014117647058824, + "grad_norm": 0.1032838320134072, + "learning_rate": 0.0002563600781807478, + "loss": 0.0626, + "step": 4678 + }, + { + "epoch": 2.2018823529411766, + "grad_norm": 0.06108770784670343, + "learning_rate": 0.0002563291629572943, + "loss": 0.0281, + "step": 4679 + }, + { + "epoch": 2.2023529411764704, + "grad_norm": 0.057516347893682304, + "learning_rate": 0.00025629823865280874, + "loss": 0.0228, + "step": 4680 + }, + { + "epoch": 2.2028235294117646, + "grad_norm": 0.12006524597935467, + "learning_rate": 0.00025626730526993223, + "loss": 0.0573, + "step": 4681 + }, + { + "epoch": 2.203294117647059, + "grad_norm": 0.12410323979722318, + "learning_rate": 0.00025623636281130657, + "loss": 0.0813, + "step": 4682 + }, + { + "epoch": 2.203764705882353, + "grad_norm": 0.12099779245595632, + "learning_rate": 0.0002562054112795745, + "loss": 0.084, + "step": 4683 + }, + { + "epoch": 2.2042352941176473, + "grad_norm": 0.06884045584939677, + "learning_rate": 0.00025617445067737933, + "loss": 0.0471, + "step": 4684 + }, + { + "epoch": 2.204705882352941, + "grad_norm": 5.494041328367654e-07, + "learning_rate": 0.0002561434810073653, + "loss": 0.0, + "step": 4685 + }, + { + "epoch": 2.2051764705882353, + "grad_norm": 0.06319914381622375, + "learning_rate": 0.00025611250227217737, + "loss": 0.0335, + "step": 4686 + }, + { + "epoch": 2.2056470588235295, + "grad_norm": 0.1668800472842673, + "learning_rate": 0.0002560815144744612, + "loss": 0.1054, + "step": 4687 + }, + { + "epoch": 2.2061176470588233, + "grad_norm": 0.08774561568141914, + "learning_rate": 0.00025605051761686345, + "loss": 0.053, + "step": 4688 + }, + { + "epoch": 2.2065882352941175, + "grad_norm": 0.16912307842381524, + "learning_rate": 0.0002560195117020313, + "loss": 0.1341, + "step": 4689 + }, + { + "epoch": 2.2070588235294117, + "grad_norm": 0.1124764968188232, + "learning_rate": 0.0002559884967326128, + "loss": 0.0624, + "step": 4690 + }, + { + "epoch": 2.207529411764706, + "grad_norm": 0.093406991172652, + "learning_rate": 0.0002559574727112569, + "loss": 0.0583, + "step": 4691 + }, + { + "epoch": 2.208, + "grad_norm": 0.08496208802707803, + "learning_rate": 0.000255926439640613, + "loss": 0.0469, + "step": 4692 + }, + { + "epoch": 2.208470588235294, + "grad_norm": 5.480842556311145e-07, + "learning_rate": 0.00025589539752333165, + "loss": 0.0, + "step": 4693 + }, + { + "epoch": 2.208941176470588, + "grad_norm": 0.0454521798533908, + "learning_rate": 0.0002558643463620639, + "loss": 0.0146, + "step": 4694 + }, + { + "epoch": 2.2094117647058824, + "grad_norm": 2.748094576348274e-07, + "learning_rate": 0.0002558332861594617, + "loss": 0.0, + "step": 4695 + }, + { + "epoch": 2.2098823529411766, + "grad_norm": 0.0617619940689802, + "learning_rate": 0.0002558022169181778, + "loss": 0.0348, + "step": 4696 + }, + { + "epoch": 2.2103529411764704, + "grad_norm": 4.888996808519845e-07, + "learning_rate": 0.00025577113864086557, + "loss": 0.0, + "step": 4697 + }, + { + "epoch": 2.2108235294117646, + "grad_norm": 0.04501990645470748, + "learning_rate": 0.0002557400513301794, + "loss": 0.0136, + "step": 4698 + }, + { + "epoch": 2.211294117647059, + "grad_norm": 0.08749287796742089, + "learning_rate": 0.00025570895498877404, + "loss": 0.0383, + "step": 4699 + }, + { + "epoch": 2.211764705882353, + "grad_norm": 0.17129491845505398, + "learning_rate": 0.00025567784961930546, + "loss": 0.1067, + "step": 4700 + }, + { + "epoch": 2.212235294117647, + "grad_norm": 0.11632307437830221, + "learning_rate": 0.0002556467352244302, + "loss": 0.0169, + "step": 4701 + }, + { + "epoch": 2.212705882352941, + "grad_norm": 0.06421571502014244, + "learning_rate": 0.00025561561180680563, + "loss": 0.0295, + "step": 4702 + }, + { + "epoch": 2.2131764705882353, + "grad_norm": 0.07725079488581066, + "learning_rate": 0.0002555844793690897, + "loss": 0.037, + "step": 4703 + }, + { + "epoch": 2.2136470588235295, + "grad_norm": 0.27528727347789195, + "learning_rate": 0.0002555533379139414, + "loss": 0.2119, + "step": 4704 + }, + { + "epoch": 2.2141176470588233, + "grad_norm": 0.029877118102580952, + "learning_rate": 0.00025552218744402026, + "loss": 0.0023, + "step": 4705 + }, + { + "epoch": 2.2145882352941175, + "grad_norm": 0.09730947232484984, + "learning_rate": 0.00025549102796198673, + "loss": 0.0538, + "step": 4706 + }, + { + "epoch": 2.2150588235294117, + "grad_norm": 9.019752011818703e-07, + "learning_rate": 0.00025545985947050197, + "loss": 0.0, + "step": 4707 + }, + { + "epoch": 2.215529411764706, + "grad_norm": 0.05820626548557734, + "learning_rate": 0.00025542868197222797, + "loss": 0.03, + "step": 4708 + }, + { + "epoch": 2.216, + "grad_norm": 0.12727182927443295, + "learning_rate": 0.00025539749546982736, + "loss": 0.0875, + "step": 4709 + }, + { + "epoch": 2.216470588235294, + "grad_norm": 0.13737616624715718, + "learning_rate": 0.00025536629996596367, + "loss": 0.0821, + "step": 4710 + }, + { + "epoch": 2.216941176470588, + "grad_norm": 0.09303284923105735, + "learning_rate": 0.0002553350954633012, + "loss": 0.0388, + "step": 4711 + }, + { + "epoch": 2.2174117647058824, + "grad_norm": 0.04151191375765518, + "learning_rate": 0.00025530388196450487, + "loss": 0.0071, + "step": 4712 + }, + { + "epoch": 2.2178823529411766, + "grad_norm": 0.10266059465882384, + "learning_rate": 0.0002552726594722405, + "loss": 0.0617, + "step": 4713 + }, + { + "epoch": 2.2183529411764704, + "grad_norm": 0.1265300625101756, + "learning_rate": 0.00025524142798917466, + "loss": 0.0912, + "step": 4714 + }, + { + "epoch": 2.2188235294117646, + "grad_norm": 0.06344912431860711, + "learning_rate": 0.00025521018751797464, + "loss": 0.0233, + "step": 4715 + }, + { + "epoch": 2.219294117647059, + "grad_norm": 0.05658982033786093, + "learning_rate": 0.0002551789380613086, + "loss": 0.0333, + "step": 4716 + }, + { + "epoch": 2.219764705882353, + "grad_norm": 0.06233996807289081, + "learning_rate": 0.00025514767962184535, + "loss": 0.0302, + "step": 4717 + }, + { + "epoch": 2.220235294117647, + "grad_norm": 0.039039314248108685, + "learning_rate": 0.0002551164122022545, + "loss": 0.0096, + "step": 4718 + }, + { + "epoch": 2.220705882352941, + "grad_norm": 0.06615518012440824, + "learning_rate": 0.00025508513580520644, + "loss": 0.03, + "step": 4719 + }, + { + "epoch": 2.2211764705882353, + "grad_norm": 0.07302629488846415, + "learning_rate": 0.00025505385043337236, + "loss": 0.034, + "step": 4720 + }, + { + "epoch": 2.2216470588235295, + "grad_norm": 3.062032366664907e-07, + "learning_rate": 0.0002550225560894241, + "loss": 0.0, + "step": 4721 + }, + { + "epoch": 2.2221176470588233, + "grad_norm": 0.1479174545465128, + "learning_rate": 0.00025499125277603446, + "loss": 0.0675, + "step": 4722 + }, + { + "epoch": 2.2225882352941175, + "grad_norm": 2.2330339926198342e-07, + "learning_rate": 0.0002549599404958769, + "loss": 0.0, + "step": 4723 + }, + { + "epoch": 2.2230588235294118, + "grad_norm": 0.1612182842939222, + "learning_rate": 0.0002549286192516256, + "loss": 0.102, + "step": 4724 + }, + { + "epoch": 2.223529411764706, + "grad_norm": 0.09900896718501359, + "learning_rate": 0.0002548972890459555, + "loss": 0.053, + "step": 4725 + }, + { + "epoch": 2.224, + "grad_norm": 0.044242445061225974, + "learning_rate": 0.00025486594988154235, + "loss": 0.0186, + "step": 4726 + }, + { + "epoch": 2.224470588235294, + "grad_norm": 0.04107819832679731, + "learning_rate": 0.0002548346017610628, + "loss": 0.0181, + "step": 4727 + }, + { + "epoch": 2.224941176470588, + "grad_norm": 0.1638642359905483, + "learning_rate": 0.000254803244687194, + "loss": 0.0994, + "step": 4728 + }, + { + "epoch": 2.2254117647058824, + "grad_norm": 0.110134094218872, + "learning_rate": 0.0002547718786626141, + "loss": 0.0592, + "step": 4729 + }, + { + "epoch": 2.2258823529411766, + "grad_norm": 0.07643687157789111, + "learning_rate": 0.0002547405036900018, + "loss": 0.0315, + "step": 4730 + }, + { + "epoch": 2.2263529411764704, + "grad_norm": 0.07527317076224688, + "learning_rate": 0.0002547091197720368, + "loss": 0.0386, + "step": 4731 + }, + { + "epoch": 2.2268235294117646, + "grad_norm": 0.13480493299539462, + "learning_rate": 0.0002546777269113993, + "loss": 0.0657, + "step": 4732 + }, + { + "epoch": 2.227294117647059, + "grad_norm": 3.3062003086163684e-07, + "learning_rate": 0.0002546463251107705, + "loss": 0.0, + "step": 4733 + }, + { + "epoch": 2.227764705882353, + "grad_norm": 0.05242264161034898, + "learning_rate": 0.0002546149143728323, + "loss": 0.0173, + "step": 4734 + }, + { + "epoch": 2.228235294117647, + "grad_norm": 0.10185627284796862, + "learning_rate": 0.0002545834947002672, + "loss": 0.0423, + "step": 4735 + }, + { + "epoch": 2.228705882352941, + "grad_norm": 0.10424474158444968, + "learning_rate": 0.00025455206609575876, + "loss": 0.0599, + "step": 4736 + }, + { + "epoch": 2.2291764705882353, + "grad_norm": 0.055692262865462624, + "learning_rate": 0.000254520628561991, + "loss": 0.0295, + "step": 4737 + }, + { + "epoch": 2.2296470588235295, + "grad_norm": 5.841947814539555e-07, + "learning_rate": 0.00025448918210164886, + "loss": 0.0, + "step": 4738 + }, + { + "epoch": 2.2301176470588233, + "grad_norm": 0.079315151302635, + "learning_rate": 0.00025445772671741816, + "loss": 0.0607, + "step": 4739 + }, + { + "epoch": 2.2305882352941175, + "grad_norm": 0.07027373768416234, + "learning_rate": 0.00025442626241198513, + "loss": 0.0389, + "step": 4740 + }, + { + "epoch": 2.2310588235294118, + "grad_norm": 0.10073152173761471, + "learning_rate": 0.00025439478918803715, + "loss": 0.0728, + "step": 4741 + }, + { + "epoch": 2.231529411764706, + "grad_norm": 2.239441802452671e-05, + "learning_rate": 0.00025436330704826205, + "loss": 0.0, + "step": 4742 + }, + { + "epoch": 2.232, + "grad_norm": 0.14850381959235742, + "learning_rate": 0.0002543318159953487, + "loss": 0.0318, + "step": 4743 + }, + { + "epoch": 2.232470588235294, + "grad_norm": 0.062387876470829585, + "learning_rate": 0.00025430031603198644, + "loss": 0.0345, + "step": 4744 + }, + { + "epoch": 2.232941176470588, + "grad_norm": 0.04400279144574347, + "learning_rate": 0.0002542688071608657, + "loss": 0.0154, + "step": 4745 + }, + { + "epoch": 2.2334117647058824, + "grad_norm": 0.04748163047593613, + "learning_rate": 0.0002542372893846773, + "loss": 0.0213, + "step": 4746 + }, + { + "epoch": 2.2338823529411767, + "grad_norm": 0.045613166915561464, + "learning_rate": 0.00025420576270611313, + "loss": 0.0196, + "step": 4747 + }, + { + "epoch": 2.2343529411764704, + "grad_norm": 0.05054319987918921, + "learning_rate": 0.00025417422712786575, + "loss": 0.0228, + "step": 4748 + }, + { + "epoch": 2.2348235294117647, + "grad_norm": 0.11319138723956385, + "learning_rate": 0.00025414268265262833, + "loss": 0.0727, + "step": 4749 + }, + { + "epoch": 2.235294117647059, + "grad_norm": 0.05345479611412796, + "learning_rate": 0.0002541111292830951, + "loss": 0.0198, + "step": 4750 + }, + { + "epoch": 2.235764705882353, + "grad_norm": 0.06129048811327526, + "learning_rate": 0.0002540795670219606, + "loss": 0.0295, + "step": 4751 + }, + { + "epoch": 2.236235294117647, + "grad_norm": 0.0368834431603313, + "learning_rate": 0.0002540479958719207, + "loss": 0.0107, + "step": 4752 + }, + { + "epoch": 2.236705882352941, + "grad_norm": 0.09085081088151771, + "learning_rate": 0.0002540164158356716, + "loss": 0.0439, + "step": 4753 + }, + { + "epoch": 2.2371764705882353, + "grad_norm": 0.09790856875530646, + "learning_rate": 0.0002539848269159104, + "loss": 0.0624, + "step": 4754 + }, + { + "epoch": 2.2376470588235295, + "grad_norm": 0.1106159211932644, + "learning_rate": 0.00025395322911533484, + "loss": 0.0796, + "step": 4755 + }, + { + "epoch": 2.2381176470588233, + "grad_norm": 0.060251814678748995, + "learning_rate": 0.0002539216224366437, + "loss": 0.0258, + "step": 4756 + }, + { + "epoch": 2.2385882352941175, + "grad_norm": 7.2945201536392e-07, + "learning_rate": 0.00025389000688253626, + "loss": 0.0, + "step": 4757 + }, + { + "epoch": 2.2390588235294118, + "grad_norm": 1.1201272837415446e-06, + "learning_rate": 0.00025385838245571264, + "loss": 0.0, + "step": 4758 + }, + { + "epoch": 2.239529411764706, + "grad_norm": 0.08255272140665154, + "learning_rate": 0.00025382674915887374, + "loss": 0.06, + "step": 4759 + }, + { + "epoch": 2.24, + "grad_norm": 0.07477225435704078, + "learning_rate": 0.0002537951069947212, + "loss": 0.0345, + "step": 4760 + }, + { + "epoch": 2.240470588235294, + "grad_norm": 0.09114596471232704, + "learning_rate": 0.00025376345596595743, + "loss": 0.0526, + "step": 4761 + }, + { + "epoch": 2.240941176470588, + "grad_norm": 0.129400830508288, + "learning_rate": 0.0002537317960752855, + "loss": 0.0879, + "step": 4762 + }, + { + "epoch": 2.2414117647058824, + "grad_norm": 0.04937557679046339, + "learning_rate": 0.0002537001273254094, + "loss": 0.0196, + "step": 4763 + }, + { + "epoch": 2.2418823529411767, + "grad_norm": 0.050772984644282, + "learning_rate": 0.00025366844971903385, + "loss": 0.023, + "step": 4764 + }, + { + "epoch": 2.2423529411764704, + "grad_norm": 0.06085941359307288, + "learning_rate": 0.00025363676325886416, + "loss": 0.0378, + "step": 4765 + }, + { + "epoch": 2.2428235294117647, + "grad_norm": 0.06441241762712155, + "learning_rate": 0.00025360506794760653, + "loss": 0.0255, + "step": 4766 + }, + { + "epoch": 2.243294117647059, + "grad_norm": 0.07202748540827147, + "learning_rate": 0.00025357336378796795, + "loss": 0.033, + "step": 4767 + }, + { + "epoch": 2.243764705882353, + "grad_norm": 0.07927579442803313, + "learning_rate": 0.0002535416507826561, + "loss": 0.0582, + "step": 4768 + }, + { + "epoch": 2.244235294117647, + "grad_norm": 0.09174512710961119, + "learning_rate": 0.0002535099289343793, + "loss": 0.0588, + "step": 4769 + }, + { + "epoch": 2.244705882352941, + "grad_norm": 0.042664178240771816, + "learning_rate": 0.00025347819824584695, + "loss": 0.0123, + "step": 4770 + }, + { + "epoch": 2.2451764705882353, + "grad_norm": 0.00020464731623219174, + "learning_rate": 0.0002534464587197689, + "loss": 0.0, + "step": 4771 + }, + { + "epoch": 2.2456470588235296, + "grad_norm": 1.880960237774948e-06, + "learning_rate": 0.00025341471035885587, + "loss": 0.0, + "step": 4772 + }, + { + "epoch": 2.2461176470588233, + "grad_norm": 0.15693871068811013, + "learning_rate": 0.0002533829531658193, + "loss": 0.1178, + "step": 4773 + }, + { + "epoch": 2.2465882352941176, + "grad_norm": 0.20681825853974808, + "learning_rate": 0.00025335118714337146, + "loss": 0.1143, + "step": 4774 + }, + { + "epoch": 2.2470588235294118, + "grad_norm": 0.07309660334646122, + "learning_rate": 0.0002533194122942253, + "loss": 0.0336, + "step": 4775 + }, + { + "epoch": 2.247529411764706, + "grad_norm": 0.050995533958579975, + "learning_rate": 0.00025328762862109457, + "loss": 0.0145, + "step": 4776 + }, + { + "epoch": 2.248, + "grad_norm": 0.039448552523268496, + "learning_rate": 0.00025325583612669373, + "loss": 0.0127, + "step": 4777 + }, + { + "epoch": 2.248470588235294, + "grad_norm": 0.047139181123017994, + "learning_rate": 0.00025322403481373797, + "loss": 0.0139, + "step": 4778 + }, + { + "epoch": 2.248941176470588, + "grad_norm": 0.05575889223899424, + "learning_rate": 0.00025319222468494335, + "loss": 0.0195, + "step": 4779 + }, + { + "epoch": 2.2494117647058824, + "grad_norm": 0.04886385462069326, + "learning_rate": 0.00025316040574302663, + "loss": 0.0175, + "step": 4780 + }, + { + "epoch": 2.2498823529411767, + "grad_norm": 7.531987781364049e-07, + "learning_rate": 0.00025312857799070524, + "loss": 0.0, + "step": 4781 + }, + { + "epoch": 2.2503529411764704, + "grad_norm": 0.06753463000045788, + "learning_rate": 0.0002530967414306974, + "loss": 0.0403, + "step": 4782 + }, + { + "epoch": 2.2508235294117647, + "grad_norm": 0.058153154187545505, + "learning_rate": 0.0002530648960657222, + "loss": 0.0217, + "step": 4783 + }, + { + "epoch": 2.251294117647059, + "grad_norm": 7.135709685774578e-07, + "learning_rate": 0.0002530330418984993, + "loss": 0.0, + "step": 4784 + }, + { + "epoch": 2.251764705882353, + "grad_norm": 0.05144109593319875, + "learning_rate": 0.00025300117893174934, + "loss": 0.0264, + "step": 4785 + }, + { + "epoch": 2.252235294117647, + "grad_norm": 0.0975585684033326, + "learning_rate": 0.0002529693071681934, + "loss": 0.0573, + "step": 4786 + }, + { + "epoch": 2.252705882352941, + "grad_norm": 0.04625735102190221, + "learning_rate": 0.00025293742661055364, + "loss": 0.0181, + "step": 4787 + }, + { + "epoch": 2.2531764705882353, + "grad_norm": 0.08300927880719487, + "learning_rate": 0.0002529055372615527, + "loss": 0.0454, + "step": 4788 + }, + { + "epoch": 2.2536470588235296, + "grad_norm": 0.11890739618798453, + "learning_rate": 0.00025287363912391403, + "loss": 0.0826, + "step": 4789 + }, + { + "epoch": 2.2541176470588233, + "grad_norm": 0.070711525022264, + "learning_rate": 0.00025284173220036215, + "loss": 0.0387, + "step": 4790 + }, + { + "epoch": 2.2545882352941176, + "grad_norm": 0.055540084208316534, + "learning_rate": 0.00025280981649362184, + "loss": 0.0229, + "step": 4791 + }, + { + "epoch": 2.255058823529412, + "grad_norm": 0.04954495973648331, + "learning_rate": 0.0002527778920064189, + "loss": 0.0222, + "step": 4792 + }, + { + "epoch": 2.255529411764706, + "grad_norm": 9.571199749775154e-07, + "learning_rate": 0.00025274595874147984, + "loss": 0.0, + "step": 4793 + }, + { + "epoch": 2.2560000000000002, + "grad_norm": 0.05604675674626663, + "learning_rate": 0.000252714016701532, + "loss": 0.0244, + "step": 4794 + }, + { + "epoch": 2.256470588235294, + "grad_norm": 0.13090217765138157, + "learning_rate": 0.0002526820658893033, + "loss": 0.0954, + "step": 4795 + }, + { + "epoch": 2.2569411764705882, + "grad_norm": 0.05701457195766075, + "learning_rate": 0.0002526501063075225, + "loss": 0.0291, + "step": 4796 + }, + { + "epoch": 2.2574117647058825, + "grad_norm": 0.11690857048764361, + "learning_rate": 0.00025261813795891916, + "loss": 0.0859, + "step": 4797 + }, + { + "epoch": 2.2578823529411762, + "grad_norm": 3.227121020049607e-07, + "learning_rate": 0.00025258616084622345, + "loss": 0.0, + "step": 4798 + }, + { + "epoch": 2.2583529411764705, + "grad_norm": 0.11164937855100238, + "learning_rate": 0.00025255417497216647, + "loss": 0.0788, + "step": 4799 + }, + { + "epoch": 2.2588235294117647, + "grad_norm": 0.00012223512592401958, + "learning_rate": 0.00025252218033947993, + "loss": 0.0, + "step": 4800 + }, + { + "epoch": 2.259294117647059, + "grad_norm": 0.04772431994418005, + "learning_rate": 0.0002524901769508963, + "loss": 0.0215, + "step": 4801 + }, + { + "epoch": 2.259764705882353, + "grad_norm": 0.05229010988651111, + "learning_rate": 0.0002524581648091488, + "loss": 0.018, + "step": 4802 + }, + { + "epoch": 2.260235294117647, + "grad_norm": 0.11020937980960961, + "learning_rate": 0.00025242614391697153, + "loss": 0.0662, + "step": 4803 + }, + { + "epoch": 2.260705882352941, + "grad_norm": 1.1155733783502275e-06, + "learning_rate": 0.00025239411427709915, + "loss": 0.0, + "step": 4804 + }, + { + "epoch": 2.2611764705882353, + "grad_norm": 5.647446522163303e-07, + "learning_rate": 0.0002523620758922672, + "loss": 0.0, + "step": 4805 + }, + { + "epoch": 2.2616470588235296, + "grad_norm": 1.2925683857046644e-06, + "learning_rate": 0.00025233002876521183, + "loss": 0.0, + "step": 4806 + }, + { + "epoch": 2.2621176470588233, + "grad_norm": 0.07104840807657857, + "learning_rate": 0.0002522979728986701, + "loss": 0.0375, + "step": 4807 + }, + { + "epoch": 2.2625882352941176, + "grad_norm": 0.06757503181708421, + "learning_rate": 0.00025226590829537977, + "loss": 0.0337, + "step": 4808 + }, + { + "epoch": 2.263058823529412, + "grad_norm": 0.046002269557052035, + "learning_rate": 0.0002522338349580792, + "loss": 0.0216, + "step": 4809 + }, + { + "epoch": 2.263529411764706, + "grad_norm": 6.322076988265763e-07, + "learning_rate": 0.00025220175288950765, + "loss": 0.0, + "step": 4810 + }, + { + "epoch": 2.2640000000000002, + "grad_norm": 0.10545936853315124, + "learning_rate": 0.00025216966209240516, + "loss": 0.0755, + "step": 4811 + }, + { + "epoch": 2.264470588235294, + "grad_norm": 0.04660490562037406, + "learning_rate": 0.00025213756256951234, + "loss": 0.0153, + "step": 4812 + }, + { + "epoch": 2.2649411764705882, + "grad_norm": 0.06768176659893865, + "learning_rate": 0.00025210545432357076, + "loss": 0.0445, + "step": 4813 + }, + { + "epoch": 2.2654117647058825, + "grad_norm": 0.052487860844495575, + "learning_rate": 0.0002520733373573225, + "loss": 0.0204, + "step": 4814 + }, + { + "epoch": 2.2658823529411762, + "grad_norm": 0.026531131263835488, + "learning_rate": 0.0002520412116735106, + "loss": 0.0047, + "step": 4815 + }, + { + "epoch": 2.2663529411764705, + "grad_norm": 0.08310845830391621, + "learning_rate": 0.0002520090772748787, + "loss": 0.0378, + "step": 4816 + }, + { + "epoch": 2.2668235294117647, + "grad_norm": 0.0004081103763447984, + "learning_rate": 0.00025197693416417126, + "loss": 0.0, + "step": 4817 + }, + { + "epoch": 2.267294117647059, + "grad_norm": 0.12391919828100297, + "learning_rate": 0.0002519447823441334, + "loss": 0.0714, + "step": 4818 + }, + { + "epoch": 2.267764705882353, + "grad_norm": 0.053456584828708506, + "learning_rate": 0.00025191262181751113, + "loss": 0.0268, + "step": 4819 + }, + { + "epoch": 2.268235294117647, + "grad_norm": 0.1246597562867549, + "learning_rate": 0.00025188045258705107, + "loss": 0.0773, + "step": 4820 + }, + { + "epoch": 2.268705882352941, + "grad_norm": 0.029336613178908535, + "learning_rate": 0.0002518482746555007, + "loss": 0.0039, + "step": 4821 + }, + { + "epoch": 2.2691764705882354, + "grad_norm": 0.05099085690750381, + "learning_rate": 0.000251816088025608, + "loss": 0.0241, + "step": 4822 + }, + { + "epoch": 2.2696470588235296, + "grad_norm": 1.1232609720300546e-05, + "learning_rate": 0.00025178389270012205, + "loss": 0.0, + "step": 4823 + }, + { + "epoch": 2.2701176470588234, + "grad_norm": 0.10968395989419612, + "learning_rate": 0.0002517516886817924, + "loss": 0.0555, + "step": 4824 + }, + { + "epoch": 2.2705882352941176, + "grad_norm": 0.09180682885891499, + "learning_rate": 0.00025171947597336946, + "loss": 0.0393, + "step": 4825 + }, + { + "epoch": 2.271058823529412, + "grad_norm": 0.07440075666295165, + "learning_rate": 0.00025168725457760434, + "loss": 0.0409, + "step": 4826 + }, + { + "epoch": 2.271529411764706, + "grad_norm": 0.04268114668129457, + "learning_rate": 0.0002516550244972489, + "loss": 0.0181, + "step": 4827 + }, + { + "epoch": 2.2720000000000002, + "grad_norm": 0.07707885351886308, + "learning_rate": 0.00025162278573505575, + "loss": 0.0443, + "step": 4828 + }, + { + "epoch": 2.272470588235294, + "grad_norm": 0.0913675064956404, + "learning_rate": 0.0002515905382937782, + "loss": 0.0547, + "step": 4829 + }, + { + "epoch": 2.2729411764705882, + "grad_norm": 0.021775107435334472, + "learning_rate": 0.0002515582821761704, + "loss": 0.0048, + "step": 4830 + }, + { + "epoch": 2.2734117647058825, + "grad_norm": 0.10106535427124025, + "learning_rate": 0.0002515260173849872, + "loss": 0.0731, + "step": 4831 + }, + { + "epoch": 2.2738823529411762, + "grad_norm": 0.07642496013871133, + "learning_rate": 0.00025149374392298407, + "loss": 0.0448, + "step": 4832 + }, + { + "epoch": 2.2743529411764705, + "grad_norm": 0.06310782959413815, + "learning_rate": 0.00025146146179291746, + "loss": 0.0306, + "step": 4833 + }, + { + "epoch": 2.2748235294117647, + "grad_norm": 0.1479488723036697, + "learning_rate": 0.0002514291709975443, + "loss": 0.1281, + "step": 4834 + }, + { + "epoch": 2.275294117647059, + "grad_norm": 8.04411204176917e-07, + "learning_rate": 0.00025139687153962246, + "loss": 0.0, + "step": 4835 + }, + { + "epoch": 2.275764705882353, + "grad_norm": 0.0957210872066064, + "learning_rate": 0.00025136456342191043, + "loss": 0.0545, + "step": 4836 + }, + { + "epoch": 2.276235294117647, + "grad_norm": 1.0137488445990358e-06, + "learning_rate": 0.0002513322466471675, + "loss": 0.0, + "step": 4837 + }, + { + "epoch": 2.276705882352941, + "grad_norm": 0.048826009799590946, + "learning_rate": 0.00025129992121815365, + "loss": 0.0163, + "step": 4838 + }, + { + "epoch": 2.2771764705882354, + "grad_norm": 1.1655874482448208e-06, + "learning_rate": 0.00025126758713762964, + "loss": 0.0, + "step": 4839 + }, + { + "epoch": 2.2776470588235296, + "grad_norm": 2.339478449727426e-05, + "learning_rate": 0.000251235244408357, + "loss": 0.0, + "step": 4840 + }, + { + "epoch": 2.2781176470588234, + "grad_norm": 0.11008635870060458, + "learning_rate": 0.00025120289303309795, + "loss": 0.0403, + "step": 4841 + }, + { + "epoch": 2.2785882352941176, + "grad_norm": 1.2861167579418326e-06, + "learning_rate": 0.00025117053301461546, + "loss": 0.0, + "step": 4842 + }, + { + "epoch": 2.279058823529412, + "grad_norm": 0.13831505731064306, + "learning_rate": 0.00025113816435567315, + "loss": 0.0873, + "step": 4843 + }, + { + "epoch": 2.279529411764706, + "grad_norm": 8.960082521197599e-07, + "learning_rate": 0.00025110578705903557, + "loss": 0.0, + "step": 4844 + }, + { + "epoch": 2.2800000000000002, + "grad_norm": 0.09168532276963241, + "learning_rate": 0.0002510734011274678, + "loss": 0.0476, + "step": 4845 + }, + { + "epoch": 2.280470588235294, + "grad_norm": 0.07115237934082064, + "learning_rate": 0.00025104100656373585, + "loss": 0.0388, + "step": 4846 + }, + { + "epoch": 2.2809411764705882, + "grad_norm": 0.07055935523323233, + "learning_rate": 0.0002510086033706063, + "loss": 0.0247, + "step": 4847 + }, + { + "epoch": 2.2814117647058825, + "grad_norm": 0.11439711530592016, + "learning_rate": 0.0002509761915508466, + "loss": 0.0519, + "step": 4848 + }, + { + "epoch": 2.2818823529411763, + "grad_norm": 0.03520626711245762, + "learning_rate": 0.0002509437711072249, + "loss": 0.0081, + "step": 4849 + }, + { + "epoch": 2.2823529411764705, + "grad_norm": 0.08487595089335058, + "learning_rate": 0.00025091134204250997, + "loss": 0.0361, + "step": 4850 + }, + { + "epoch": 2.2828235294117647, + "grad_norm": 0.1401929106332142, + "learning_rate": 0.0002508789043594715, + "loss": 0.0372, + "step": 4851 + }, + { + "epoch": 2.283294117647059, + "grad_norm": 8.441583961809185e-06, + "learning_rate": 0.00025084645806087974, + "loss": 0.0, + "step": 4852 + }, + { + "epoch": 2.283764705882353, + "grad_norm": 0.0540368487561744, + "learning_rate": 0.00025081400314950586, + "loss": 0.0137, + "step": 4853 + }, + { + "epoch": 2.284235294117647, + "grad_norm": 0.03594341167297126, + "learning_rate": 0.0002507815396281216, + "loss": 0.0111, + "step": 4854 + }, + { + "epoch": 2.284705882352941, + "grad_norm": 0.06645184925464022, + "learning_rate": 0.0002507490674994995, + "loss": 0.0265, + "step": 4855 + }, + { + "epoch": 2.2851764705882354, + "grad_norm": 0.12070287408981162, + "learning_rate": 0.0002507165867664129, + "loss": 0.0535, + "step": 4856 + }, + { + "epoch": 2.2856470588235296, + "grad_norm": 0.020655138746025953, + "learning_rate": 0.0002506840974316357, + "loss": 0.0023, + "step": 4857 + }, + { + "epoch": 2.2861176470588234, + "grad_norm": 0.07355961879517851, + "learning_rate": 0.00025065159949794283, + "loss": 0.0393, + "step": 4858 + }, + { + "epoch": 2.2865882352941176, + "grad_norm": 2.0233345396403643e-06, + "learning_rate": 0.00025061909296810966, + "loss": 0.0, + "step": 4859 + }, + { + "epoch": 2.287058823529412, + "grad_norm": 2.2603742772986144e-06, + "learning_rate": 0.0002505865778449124, + "loss": 0.0, + "step": 4860 + }, + { + "epoch": 2.287529411764706, + "grad_norm": 0.15059387472022226, + "learning_rate": 0.000250554054131128, + "loss": 0.0869, + "step": 4861 + }, + { + "epoch": 2.288, + "grad_norm": 0.15807753415623174, + "learning_rate": 0.00025052152182953414, + "loss": 0.0986, + "step": 4862 + }, + { + "epoch": 2.288470588235294, + "grad_norm": 0.07914652423890031, + "learning_rate": 0.0002504889809429093, + "loss": 0.0334, + "step": 4863 + }, + { + "epoch": 2.2889411764705883, + "grad_norm": 0.10477854160111792, + "learning_rate": 0.0002504564314740326, + "loss": 0.0539, + "step": 4864 + }, + { + "epoch": 2.2894117647058825, + "grad_norm": 0.038171358595535936, + "learning_rate": 0.0002504238734256839, + "loss": 0.0126, + "step": 4865 + }, + { + "epoch": 2.2898823529411763, + "grad_norm": 0.07116834646710882, + "learning_rate": 0.0002503913068006438, + "loss": 0.0279, + "step": 4866 + }, + { + "epoch": 2.2903529411764705, + "grad_norm": 0.027534474681884354, + "learning_rate": 0.00025035873160169373, + "loss": 0.0038, + "step": 4867 + }, + { + "epoch": 2.2908235294117647, + "grad_norm": 0.13988605204710333, + "learning_rate": 0.0002503261478316157, + "loss": 0.1224, + "step": 4868 + }, + { + "epoch": 2.291294117647059, + "grad_norm": 0.05723898957913498, + "learning_rate": 0.0002502935554931926, + "loss": 0.0181, + "step": 4869 + }, + { + "epoch": 2.291764705882353, + "grad_norm": 0.09710268998629924, + "learning_rate": 0.0002502609545892079, + "loss": 0.077, + "step": 4870 + }, + { + "epoch": 2.292235294117647, + "grad_norm": 0.05306522286070088, + "learning_rate": 0.00025022834512244583, + "loss": 0.0128, + "step": 4871 + }, + { + "epoch": 2.292705882352941, + "grad_norm": 0.13153133403114245, + "learning_rate": 0.00025019572709569153, + "loss": 0.0696, + "step": 4872 + }, + { + "epoch": 2.2931764705882354, + "grad_norm": 0.09684805681023194, + "learning_rate": 0.0002501631005117307, + "loss": 0.0534, + "step": 4873 + }, + { + "epoch": 2.2936470588235296, + "grad_norm": 0.036738832704657456, + "learning_rate": 0.0002501304653733497, + "loss": 0.0141, + "step": 4874 + }, + { + "epoch": 2.2941176470588234, + "grad_norm": 0.09246221308522586, + "learning_rate": 0.0002500978216833359, + "loss": 0.053, + "step": 4875 + }, + { + "epoch": 2.2945882352941176, + "grad_norm": 0.08358945723292095, + "learning_rate": 0.0002500651694444771, + "loss": 0.0532, + "step": 4876 + }, + { + "epoch": 2.295058823529412, + "grad_norm": 0.03110060958783723, + "learning_rate": 0.00025003250865956203, + "loss": 0.0079, + "step": 4877 + }, + { + "epoch": 2.295529411764706, + "grad_norm": 1.2059745796710375e-06, + "learning_rate": 0.0002499998393313801, + "loss": 0.0, + "step": 4878 + }, + { + "epoch": 2.296, + "grad_norm": 0.07157252627574741, + "learning_rate": 0.00024996716146272137, + "loss": 0.0413, + "step": 4879 + }, + { + "epoch": 2.296470588235294, + "grad_norm": 1.0781443773979766e-06, + "learning_rate": 0.0002499344750563767, + "loss": 0.0, + "step": 4880 + }, + { + "epoch": 2.2969411764705883, + "grad_norm": 0.036288779125878544, + "learning_rate": 0.0002499017801151377, + "loss": 0.0085, + "step": 4881 + }, + { + "epoch": 2.2974117647058825, + "grad_norm": 0.031439990586264385, + "learning_rate": 0.0002498690766417967, + "loss": 0.0075, + "step": 4882 + }, + { + "epoch": 2.2978823529411763, + "grad_norm": 0.12213173204890376, + "learning_rate": 0.00024983636463914665, + "loss": 0.0522, + "step": 4883 + }, + { + "epoch": 2.2983529411764705, + "grad_norm": 0.08876433198375737, + "learning_rate": 0.00024980364410998133, + "loss": 0.0352, + "step": 4884 + }, + { + "epoch": 2.2988235294117647, + "grad_norm": 0.01720288622868928, + "learning_rate": 0.0002497709150570953, + "loss": 0.0027, + "step": 4885 + }, + { + "epoch": 2.299294117647059, + "grad_norm": 0.03646317226052815, + "learning_rate": 0.0002497381774832838, + "loss": 0.0118, + "step": 4886 + }, + { + "epoch": 2.299764705882353, + "grad_norm": 0.08609888840567796, + "learning_rate": 0.00024970543139134275, + "loss": 0.0457, + "step": 4887 + }, + { + "epoch": 2.300235294117647, + "grad_norm": 0.05965624289246044, + "learning_rate": 0.0002496726767840687, + "loss": 0.0265, + "step": 4888 + }, + { + "epoch": 2.300705882352941, + "grad_norm": 3.5004641705234456e-06, + "learning_rate": 0.0002496399136642592, + "loss": 0.0, + "step": 4889 + }, + { + "epoch": 2.3011764705882354, + "grad_norm": 0.03123643124476754, + "learning_rate": 0.0002496071420347124, + "loss": 0.0103, + "step": 4890 + }, + { + "epoch": 2.3016470588235296, + "grad_norm": 1.2595474080531121e-06, + "learning_rate": 0.00024957436189822706, + "loss": 0.0, + "step": 4891 + }, + { + "epoch": 2.3021176470588234, + "grad_norm": 0.04346790963161461, + "learning_rate": 0.0002495415732576028, + "loss": 0.0076, + "step": 4892 + }, + { + "epoch": 2.3025882352941176, + "grad_norm": 0.09052949346779599, + "learning_rate": 0.00024950877611564, + "loss": 0.0458, + "step": 4893 + }, + { + "epoch": 2.303058823529412, + "grad_norm": 8.123322455813007e-07, + "learning_rate": 0.0002494759704751395, + "loss": 0.0, + "step": 4894 + }, + { + "epoch": 2.303529411764706, + "grad_norm": 0.030931630268729398, + "learning_rate": 0.00024944315633890327, + "loss": 0.0062, + "step": 4895 + }, + { + "epoch": 2.304, + "grad_norm": 0.09543195112407611, + "learning_rate": 0.0002494103337097337, + "loss": 0.0598, + "step": 4896 + }, + { + "epoch": 2.304470588235294, + "grad_norm": 0.2305237463918705, + "learning_rate": 0.00024937750259043407, + "loss": 0.1369, + "step": 4897 + }, + { + "epoch": 2.3049411764705883, + "grad_norm": 9.632926263540497e-07, + "learning_rate": 0.00024934466298380824, + "loss": 0.0, + "step": 4898 + }, + { + "epoch": 2.3054117647058825, + "grad_norm": 0.09772077150406232, + "learning_rate": 0.0002493118148926609, + "loss": 0.0593, + "step": 4899 + }, + { + "epoch": 2.3058823529411763, + "grad_norm": 0.1354920995033546, + "learning_rate": 0.00024927895831979745, + "loss": 0.0711, + "step": 4900 + }, + { + "epoch": 2.3063529411764705, + "grad_norm": 0.07061446395814039, + "learning_rate": 0.000249246093268024, + "loss": 0.0395, + "step": 4901 + }, + { + "epoch": 2.3068235294117647, + "grad_norm": 0.08333316435750862, + "learning_rate": 0.0002492132197401474, + "loss": 0.0382, + "step": 4902 + }, + { + "epoch": 2.307294117647059, + "grad_norm": 1.0252112228039468e-05, + "learning_rate": 0.0002491803377389752, + "loss": 0.0, + "step": 4903 + }, + { + "epoch": 2.307764705882353, + "grad_norm": 0.06030047118501716, + "learning_rate": 0.0002491474472673157, + "loss": 0.03, + "step": 4904 + }, + { + "epoch": 2.308235294117647, + "grad_norm": 0.056172097562223466, + "learning_rate": 0.00024911454832797784, + "loss": 0.0202, + "step": 4905 + }, + { + "epoch": 2.308705882352941, + "grad_norm": 0.1279741592600397, + "learning_rate": 0.0002490816409237714, + "loss": 0.0347, + "step": 4906 + }, + { + "epoch": 2.3091764705882354, + "grad_norm": 0.2606019289500271, + "learning_rate": 0.0002490487250575069, + "loss": 0.1181, + "step": 4907 + }, + { + "epoch": 2.309647058823529, + "grad_norm": 9.116283109017304e-07, + "learning_rate": 0.00024901580073199536, + "loss": 0.0, + "step": 4908 + }, + { + "epoch": 2.3101176470588234, + "grad_norm": 0.13854351552298996, + "learning_rate": 0.0002489828679500488, + "loss": 0.0912, + "step": 4909 + }, + { + "epoch": 2.3105882352941176, + "grad_norm": 0.12363761461701396, + "learning_rate": 0.00024894992671447983, + "loss": 0.0961, + "step": 4910 + }, + { + "epoch": 2.311058823529412, + "grad_norm": 0.045837891215471424, + "learning_rate": 0.0002489169770281018, + "loss": 0.0141, + "step": 4911 + }, + { + "epoch": 2.311529411764706, + "grad_norm": 0.09306488343527757, + "learning_rate": 0.0002488840188937287, + "loss": 0.0552, + "step": 4912 + }, + { + "epoch": 2.312, + "grad_norm": 0.11809705406224683, + "learning_rate": 0.00024885105231417536, + "loss": 0.0582, + "step": 4913 + }, + { + "epoch": 2.312470588235294, + "grad_norm": 9.850108551607865e-07, + "learning_rate": 0.00024881807729225727, + "loss": 0.0, + "step": 4914 + }, + { + "epoch": 2.3129411764705883, + "grad_norm": 6.73084228524387e-07, + "learning_rate": 0.00024878509383079075, + "loss": 0.0, + "step": 4915 + }, + { + "epoch": 2.3134117647058825, + "grad_norm": 0.028198078035788124, + "learning_rate": 0.0002487521019325926, + "loss": 0.0069, + "step": 4916 + }, + { + "epoch": 2.3138823529411763, + "grad_norm": 0.12404494300195106, + "learning_rate": 0.00024871910160048063, + "loss": 0.0754, + "step": 4917 + }, + { + "epoch": 2.3143529411764705, + "grad_norm": 0.12037599124169736, + "learning_rate": 0.00024868609283727316, + "loss": 0.0566, + "step": 4918 + }, + { + "epoch": 2.3148235294117647, + "grad_norm": 0.037867902945210204, + "learning_rate": 0.00024865307564578927, + "loss": 0.0105, + "step": 4919 + }, + { + "epoch": 2.315294117647059, + "grad_norm": 0.07901812325666209, + "learning_rate": 0.00024862005002884886, + "loss": 0.0241, + "step": 4920 + }, + { + "epoch": 2.315764705882353, + "grad_norm": 0.07079622047317795, + "learning_rate": 0.0002485870159892724, + "loss": 0.0412, + "step": 4921 + }, + { + "epoch": 2.316235294117647, + "grad_norm": 0.0561421446758549, + "learning_rate": 0.00024855397352988127, + "loss": 0.0193, + "step": 4922 + }, + { + "epoch": 2.316705882352941, + "grad_norm": 0.02728865170397544, + "learning_rate": 0.0002485209226534974, + "loss": 0.0037, + "step": 4923 + }, + { + "epoch": 2.3171764705882354, + "grad_norm": 9.547342087422488e-07, + "learning_rate": 0.00024848786336294346, + "loss": 0.0, + "step": 4924 + }, + { + "epoch": 2.317647058823529, + "grad_norm": 2.5537535358906112e-06, + "learning_rate": 0.0002484547956610429, + "loss": 0.0, + "step": 4925 + }, + { + "epoch": 2.3181176470588234, + "grad_norm": 0.06396709791390069, + "learning_rate": 0.0002484217195506199, + "loss": 0.0229, + "step": 4926 + }, + { + "epoch": 2.3185882352941176, + "grad_norm": 0.03245248753669, + "learning_rate": 0.00024838863503449926, + "loss": 0.0083, + "step": 4927 + }, + { + "epoch": 2.319058823529412, + "grad_norm": 1.6594670156454453e-05, + "learning_rate": 0.00024835554211550664, + "loss": 0.0, + "step": 4928 + }, + { + "epoch": 2.319529411764706, + "grad_norm": 0.09560840947530379, + "learning_rate": 0.00024832244079646827, + "loss": 0.0513, + "step": 4929 + }, + { + "epoch": 2.32, + "grad_norm": 5.292215239839022e-07, + "learning_rate": 0.00024828933108021116, + "loss": 0.0, + "step": 4930 + }, + { + "epoch": 2.320470588235294, + "grad_norm": 0.0583952782960295, + "learning_rate": 0.00024825621296956313, + "loss": 0.0223, + "step": 4931 + }, + { + "epoch": 2.3209411764705883, + "grad_norm": 4.15292391885375e-07, + "learning_rate": 0.00024822308646735254, + "loss": 0.0, + "step": 4932 + }, + { + "epoch": 2.3214117647058825, + "grad_norm": 0.027529641677633828, + "learning_rate": 0.00024818995157640855, + "loss": 0.0065, + "step": 4933 + }, + { + "epoch": 2.3218823529411763, + "grad_norm": 0.049496724021847495, + "learning_rate": 0.00024815680829956115, + "loss": 0.0147, + "step": 4934 + }, + { + "epoch": 2.3223529411764705, + "grad_norm": 0.15130185167734497, + "learning_rate": 0.0002481236566396408, + "loss": 0.0488, + "step": 4935 + }, + { + "epoch": 2.3228235294117647, + "grad_norm": 2.67534576800553e-06, + "learning_rate": 0.0002480904965994789, + "loss": 0.0, + "step": 4936 + }, + { + "epoch": 2.323294117647059, + "grad_norm": 0.12584191228942196, + "learning_rate": 0.00024805732818190747, + "loss": 0.0595, + "step": 4937 + }, + { + "epoch": 2.323764705882353, + "grad_norm": 0.0927773663881662, + "learning_rate": 0.00024802415138975924, + "loss": 0.0506, + "step": 4938 + }, + { + "epoch": 2.324235294117647, + "grad_norm": 0.10129454994150706, + "learning_rate": 0.00024799096622586767, + "loss": 0.0531, + "step": 4939 + }, + { + "epoch": 2.324705882352941, + "grad_norm": 0.008487365043214077, + "learning_rate": 0.00024795777269306696, + "loss": 0.0003, + "step": 4940 + }, + { + "epoch": 2.3251764705882354, + "grad_norm": 0.10518194272716552, + "learning_rate": 0.0002479245707941919, + "loss": 0.0555, + "step": 4941 + }, + { + "epoch": 2.325647058823529, + "grad_norm": 0.10801162640722112, + "learning_rate": 0.0002478913605320783, + "loss": 0.0678, + "step": 4942 + }, + { + "epoch": 2.3261176470588234, + "grad_norm": 0.22139092914397082, + "learning_rate": 0.0002478581419095623, + "loss": 0.18, + "step": 4943 + }, + { + "epoch": 2.3265882352941176, + "grad_norm": 0.05867067391187644, + "learning_rate": 0.0002478249149294809, + "loss": 0.0247, + "step": 4944 + }, + { + "epoch": 2.327058823529412, + "grad_norm": 0.03870566910395883, + "learning_rate": 0.00024779167959467207, + "loss": 0.0113, + "step": 4945 + }, + { + "epoch": 2.327529411764706, + "grad_norm": 0.0003745903033738849, + "learning_rate": 0.0002477584359079741, + "loss": 0.0, + "step": 4946 + }, + { + "epoch": 2.328, + "grad_norm": 0.0799894993623411, + "learning_rate": 0.00024772518387222616, + "loss": 0.044, + "step": 4947 + }, + { + "epoch": 2.328470588235294, + "grad_norm": 0.060677947016986866, + "learning_rate": 0.0002476919234902682, + "loss": 0.0316, + "step": 4948 + }, + { + "epoch": 2.3289411764705883, + "grad_norm": 0.07105036803821772, + "learning_rate": 0.0002476586547649408, + "loss": 0.0373, + "step": 4949 + }, + { + "epoch": 2.3294117647058825, + "grad_norm": 0.06090272290064131, + "learning_rate": 0.00024762537769908535, + "loss": 0.028, + "step": 4950 + }, + { + "epoch": 2.3298823529411763, + "grad_norm": 0.21019870440552763, + "learning_rate": 0.0002475920922955437, + "loss": 0.0908, + "step": 4951 + }, + { + "epoch": 2.3303529411764705, + "grad_norm": 0.08768607072821931, + "learning_rate": 0.0002475587985571587, + "loss": 0.0656, + "step": 4952 + }, + { + "epoch": 2.3308235294117647, + "grad_norm": 0.07471337032604554, + "learning_rate": 0.0002475254964867738, + "loss": 0.0338, + "step": 4953 + }, + { + "epoch": 2.331294117647059, + "grad_norm": 0.11658648339234393, + "learning_rate": 0.0002474921860872331, + "loss": 0.0555, + "step": 4954 + }, + { + "epoch": 2.331764705882353, + "grad_norm": 1.7997515884387322e-06, + "learning_rate": 0.00024745886736138154, + "loss": 0.0, + "step": 4955 + }, + { + "epoch": 2.332235294117647, + "grad_norm": 0.06257446889525431, + "learning_rate": 0.0002474255403120647, + "loss": 0.0329, + "step": 4956 + }, + { + "epoch": 2.332705882352941, + "grad_norm": 0.12747538996543845, + "learning_rate": 0.0002473922049421288, + "loss": 0.0626, + "step": 4957 + }, + { + "epoch": 2.3331764705882354, + "grad_norm": 0.04914197431320419, + "learning_rate": 0.0002473588612544209, + "loss": 0.0152, + "step": 4958 + }, + { + "epoch": 2.333647058823529, + "grad_norm": 1.4050726243506906e-06, + "learning_rate": 0.0002473255092517887, + "loss": 0.0, + "step": 4959 + }, + { + "epoch": 2.3341176470588234, + "grad_norm": 7.392089014680989e-07, + "learning_rate": 0.0002472921489370807, + "loss": 0.0, + "step": 4960 + }, + { + "epoch": 2.3345882352941176, + "grad_norm": 1.2898676393230687e-06, + "learning_rate": 0.0002472587803131459, + "loss": 0.0, + "step": 4961 + }, + { + "epoch": 2.335058823529412, + "grad_norm": 0.04492630472637281, + "learning_rate": 0.0002472254033828342, + "loss": 0.0156, + "step": 4962 + }, + { + "epoch": 2.335529411764706, + "grad_norm": 0.058866041597891855, + "learning_rate": 0.00024719201814899617, + "loss": 0.031, + "step": 4963 + }, + { + "epoch": 2.336, + "grad_norm": 0.1370219850586859, + "learning_rate": 0.0002471586246144831, + "loss": 0.0938, + "step": 4964 + }, + { + "epoch": 2.336470588235294, + "grad_norm": 0.0922696214865996, + "learning_rate": 0.0002471252227821469, + "loss": 0.0566, + "step": 4965 + }, + { + "epoch": 2.3369411764705883, + "grad_norm": 6.191388761282768e-07, + "learning_rate": 0.0002470918126548403, + "loss": 0.0, + "step": 4966 + }, + { + "epoch": 2.3374117647058825, + "grad_norm": 0.12349891533325152, + "learning_rate": 0.00024705839423541664, + "loss": 0.072, + "step": 4967 + }, + { + "epoch": 2.3378823529411763, + "grad_norm": 0.1482006652250605, + "learning_rate": 0.00024702496752673005, + "loss": 0.1181, + "step": 4968 + }, + { + "epoch": 2.3383529411764705, + "grad_norm": 0.05667296553035543, + "learning_rate": 0.00024699153253163533, + "loss": 0.0254, + "step": 4969 + }, + { + "epoch": 2.3388235294117647, + "grad_norm": 0.11502715580018626, + "learning_rate": 0.000246958089252988, + "loss": 0.0556, + "step": 4970 + }, + { + "epoch": 2.339294117647059, + "grad_norm": 0.05285911036210585, + "learning_rate": 0.0002469246376936443, + "loss": 0.0188, + "step": 4971 + }, + { + "epoch": 2.339764705882353, + "grad_norm": 1.670639050348696e-06, + "learning_rate": 0.00024689117785646106, + "loss": 0.0, + "step": 4972 + }, + { + "epoch": 2.340235294117647, + "grad_norm": 0.031004664825003487, + "learning_rate": 0.000246857709744296, + "loss": 0.003, + "step": 4973 + }, + { + "epoch": 2.340705882352941, + "grad_norm": 0.09373504346423246, + "learning_rate": 0.00024682423336000746, + "loss": 0.0533, + "step": 4974 + }, + { + "epoch": 2.3411764705882354, + "grad_norm": 0.0356887625026395, + "learning_rate": 0.00024679074870645446, + "loss": 0.0093, + "step": 4975 + }, + { + "epoch": 2.341647058823529, + "grad_norm": 0.0480998037117044, + "learning_rate": 0.0002467572557864968, + "loss": 0.0157, + "step": 4976 + }, + { + "epoch": 2.3421176470588234, + "grad_norm": 1.8610803972029624e-06, + "learning_rate": 0.0002467237546029949, + "loss": 0.0, + "step": 4977 + }, + { + "epoch": 2.3425882352941176, + "grad_norm": 0.04293650958869881, + "learning_rate": 0.0002466902451588099, + "loss": 0.0079, + "step": 4978 + }, + { + "epoch": 2.343058823529412, + "grad_norm": 5.338256613973267e-07, + "learning_rate": 0.0002466567274568037, + "loss": 0.0, + "step": 4979 + }, + { + "epoch": 2.343529411764706, + "grad_norm": 0.07639850653235533, + "learning_rate": 0.0002466232014998389, + "loss": 0.0464, + "step": 4980 + }, + { + "epoch": 2.344, + "grad_norm": 0.06488471784942385, + "learning_rate": 0.00024658966729077875, + "loss": 0.0331, + "step": 4981 + }, + { + "epoch": 2.344470588235294, + "grad_norm": 0.08012822655885185, + "learning_rate": 0.00024655612483248727, + "loss": 0.0454, + "step": 4982 + }, + { + "epoch": 2.3449411764705883, + "grad_norm": 0.029567033456222462, + "learning_rate": 0.00024652257412782904, + "loss": 0.007, + "step": 4983 + }, + { + "epoch": 2.3454117647058825, + "grad_norm": 0.12717090040740017, + "learning_rate": 0.0002464890151796696, + "loss": 0.0815, + "step": 4984 + }, + { + "epoch": 2.3458823529411763, + "grad_norm": 0.07340274699039034, + "learning_rate": 0.000246455447990875, + "loss": 0.0413, + "step": 4985 + }, + { + "epoch": 2.3463529411764705, + "grad_norm": 0.055034519106089544, + "learning_rate": 0.000246421872564312, + "loss": 0.0303, + "step": 4986 + }, + { + "epoch": 2.3468235294117648, + "grad_norm": 1.1588580511101376e-06, + "learning_rate": 0.0002463882889028481, + "loss": 0.0, + "step": 4987 + }, + { + "epoch": 2.347294117647059, + "grad_norm": 0.09825809795025307, + "learning_rate": 0.0002463546970093516, + "loss": 0.0624, + "step": 4988 + }, + { + "epoch": 2.3477647058823528, + "grad_norm": 0.15720198356660156, + "learning_rate": 0.00024632109688669135, + "loss": 0.1199, + "step": 4989 + }, + { + "epoch": 2.348235294117647, + "grad_norm": 0.14872961290241798, + "learning_rate": 0.000246287488537737, + "loss": 0.0905, + "step": 4990 + }, + { + "epoch": 2.348705882352941, + "grad_norm": 1.9758507752192636e-06, + "learning_rate": 0.00024625387196535885, + "loss": 0.0, + "step": 4991 + }, + { + "epoch": 2.3491764705882354, + "grad_norm": 0.2800107378578929, + "learning_rate": 0.0002462202471724278, + "loss": 0.177, + "step": 4992 + }, + { + "epoch": 2.349647058823529, + "grad_norm": 0.0430901700203951, + "learning_rate": 0.0002461866141618158, + "loss": 0.0135, + "step": 4993 + }, + { + "epoch": 2.3501176470588234, + "grad_norm": 8.779192379643942e-06, + "learning_rate": 0.0002461529729363951, + "loss": 0.0, + "step": 4994 + }, + { + "epoch": 2.3505882352941176, + "grad_norm": 0.00043774801364221855, + "learning_rate": 0.0002461193234990389, + "loss": 0.0, + "step": 4995 + }, + { + "epoch": 2.351058823529412, + "grad_norm": 0.12314139430127559, + "learning_rate": 0.00024608566585262103, + "loss": 0.0569, + "step": 4996 + }, + { + "epoch": 2.351529411764706, + "grad_norm": 0.47972469041582955, + "learning_rate": 0.000246052000000016, + "loss": 0.1118, + "step": 4997 + }, + { + "epoch": 2.352, + "grad_norm": 0.1775652831220031, + "learning_rate": 0.00024601832594409905, + "loss": 0.0336, + "step": 4998 + }, + { + "epoch": 2.352470588235294, + "grad_norm": 0.0918333208237285, + "learning_rate": 0.00024598464368774604, + "loss": 0.0293, + "step": 4999 + }, + { + "epoch": 2.3529411764705883, + "grad_norm": 9.819980440124547e-07, + "learning_rate": 0.00024595095323383365, + "loss": 0.0, + "step": 5000 + }, + { + "epoch": 2.3534117647058825, + "grad_norm": 0.1848185645222352, + "learning_rate": 0.0002459172545852393, + "loss": 0.0899, + "step": 5001 + }, + { + "epoch": 2.3538823529411763, + "grad_norm": 0.01498726982423574, + "learning_rate": 0.0002458835477448408, + "loss": 0.0012, + "step": 5002 + }, + { + "epoch": 2.3543529411764705, + "grad_norm": 0.08910452571994651, + "learning_rate": 0.00024584983271551717, + "loss": 0.0509, + "step": 5003 + }, + { + "epoch": 2.3548235294117648, + "grad_norm": 0.06584876733269797, + "learning_rate": 0.0002458161095001476, + "loss": 0.038, + "step": 5004 + }, + { + "epoch": 2.355294117647059, + "grad_norm": 0.057213514704932154, + "learning_rate": 0.0002457823781016123, + "loss": 0.0236, + "step": 5005 + }, + { + "epoch": 2.3557647058823528, + "grad_norm": 0.04702140185443163, + "learning_rate": 0.0002457486385227922, + "loss": 0.0154, + "step": 5006 + }, + { + "epoch": 2.356235294117647, + "grad_norm": 0.03469952282477649, + "learning_rate": 0.00024571489076656865, + "loss": 0.0106, + "step": 5007 + }, + { + "epoch": 2.356705882352941, + "grad_norm": 0.06850523191387821, + "learning_rate": 0.0002456811348358239, + "loss": 0.0323, + "step": 5008 + }, + { + "epoch": 2.3571764705882354, + "grad_norm": 0.05012368165929943, + "learning_rate": 0.000245647370733441, + "loss": 0.0146, + "step": 5009 + }, + { + "epoch": 2.357647058823529, + "grad_norm": 0.012722639419435683, + "learning_rate": 0.0002456135984623034, + "loss": 0.0022, + "step": 5010 + }, + { + "epoch": 2.3581176470588234, + "grad_norm": 0.07858356976189391, + "learning_rate": 0.0002455798180252956, + "loss": 0.0432, + "step": 5011 + }, + { + "epoch": 2.3585882352941177, + "grad_norm": 0.0889156674083979, + "learning_rate": 0.0002455460294253025, + "loss": 0.0506, + "step": 5012 + }, + { + "epoch": 2.359058823529412, + "grad_norm": 0.1027628454666391, + "learning_rate": 0.00024551223266520983, + "loss": 0.0631, + "step": 5013 + }, + { + "epoch": 2.359529411764706, + "grad_norm": 0.03119260529911947, + "learning_rate": 0.00024547842774790403, + "loss": 0.0034, + "step": 5014 + }, + { + "epoch": 2.36, + "grad_norm": 0.07407372978301616, + "learning_rate": 0.0002454446146762721, + "loss": 0.0242, + "step": 5015 + }, + { + "epoch": 2.360470588235294, + "grad_norm": 0.07558272930473638, + "learning_rate": 0.000245410793453202, + "loss": 0.0285, + "step": 5016 + }, + { + "epoch": 2.3609411764705883, + "grad_norm": 0.1057980575576489, + "learning_rate": 0.00024537696408158215, + "loss": 0.0656, + "step": 5017 + }, + { + "epoch": 2.3614117647058825, + "grad_norm": 0.01623846179837793, + "learning_rate": 0.0002453431265643017, + "loss": 0.0027, + "step": 5018 + }, + { + "epoch": 2.3618823529411763, + "grad_norm": 0.04227191935426709, + "learning_rate": 0.0002453092809042507, + "loss": 0.0171, + "step": 5019 + }, + { + "epoch": 2.3623529411764705, + "grad_norm": 0.08998998569638851, + "learning_rate": 0.0002452754271043195, + "loss": 0.0397, + "step": 5020 + }, + { + "epoch": 2.3628235294117648, + "grad_norm": 0.031293642125347, + "learning_rate": 0.0002452415651673995, + "loss": 0.0029, + "step": 5021 + }, + { + "epoch": 2.363294117647059, + "grad_norm": 0.06873616834971413, + "learning_rate": 0.00024520769509638274, + "loss": 0.0289, + "step": 5022 + }, + { + "epoch": 2.3637647058823528, + "grad_norm": 0.12137212111946558, + "learning_rate": 0.0002451738168941618, + "loss": 0.0794, + "step": 5023 + }, + { + "epoch": 2.364235294117647, + "grad_norm": 0.08221275783648285, + "learning_rate": 0.0002451399305636301, + "loss": 0.0466, + "step": 5024 + }, + { + "epoch": 2.364705882352941, + "grad_norm": 0.11338165290833363, + "learning_rate": 0.00024510603610768166, + "loss": 0.0819, + "step": 5025 + }, + { + "epoch": 2.3651764705882354, + "grad_norm": 0.03985655314868165, + "learning_rate": 0.0002450721335292112, + "loss": 0.0097, + "step": 5026 + }, + { + "epoch": 2.365647058823529, + "grad_norm": 0.10212848699416395, + "learning_rate": 0.00024503822283111423, + "loss": 0.0689, + "step": 5027 + }, + { + "epoch": 2.3661176470588234, + "grad_norm": 6.195354194264167e-07, + "learning_rate": 0.0002450043040162869, + "loss": 0.0, + "step": 5028 + }, + { + "epoch": 2.3665882352941177, + "grad_norm": 2.348288038262242e-06, + "learning_rate": 0.000244970377087626, + "loss": 0.0, + "step": 5029 + }, + { + "epoch": 2.367058823529412, + "grad_norm": 0.13040704828117747, + "learning_rate": 0.00024493644204802906, + "loss": 0.0872, + "step": 5030 + }, + { + "epoch": 2.367529411764706, + "grad_norm": 0.025338925040892506, + "learning_rate": 0.0002449024989003943, + "loss": 0.0034, + "step": 5031 + }, + { + "epoch": 2.368, + "grad_norm": 0.15659383786245873, + "learning_rate": 0.0002448685476476207, + "loss": 0.1223, + "step": 5032 + }, + { + "epoch": 2.368470588235294, + "grad_norm": 0.04862432492207835, + "learning_rate": 0.0002448345882926077, + "loss": 0.0176, + "step": 5033 + }, + { + "epoch": 2.3689411764705883, + "grad_norm": 1.4120316147480151e-06, + "learning_rate": 0.0002448006208382558, + "loss": 0.0, + "step": 5034 + }, + { + "epoch": 2.369411764705882, + "grad_norm": 0.05483046854009826, + "learning_rate": 0.0002447666452874658, + "loss": 0.0264, + "step": 5035 + }, + { + "epoch": 2.3698823529411763, + "grad_norm": 1.184404304191398e-06, + "learning_rate": 0.00024473266164313953, + "loss": 0.0, + "step": 5036 + }, + { + "epoch": 2.3703529411764706, + "grad_norm": 1.738036631595324e-06, + "learning_rate": 0.00024469866990817923, + "loss": 0.0, + "step": 5037 + }, + { + "epoch": 2.3708235294117648, + "grad_norm": 0.5203626369801059, + "learning_rate": 0.00024466467008548814, + "loss": 0.2818, + "step": 5038 + }, + { + "epoch": 2.371294117647059, + "grad_norm": 0.1823728912575324, + "learning_rate": 0.0002446306621779698, + "loss": 0.0983, + "step": 5039 + }, + { + "epoch": 2.3717647058823528, + "grad_norm": 0.24617994075233496, + "learning_rate": 0.00024459664618852877, + "loss": 0.0464, + "step": 5040 + }, + { + "epoch": 2.372235294117647, + "grad_norm": 3.1045127981717556e-06, + "learning_rate": 0.0002445626221200702, + "loss": 0.0, + "step": 5041 + }, + { + "epoch": 2.372705882352941, + "grad_norm": 0.0629467678162802, + "learning_rate": 0.0002445285899754998, + "loss": 0.023, + "step": 5042 + }, + { + "epoch": 2.3731764705882354, + "grad_norm": 0.11128656498810427, + "learning_rate": 0.0002444945497577242, + "loss": 0.0731, + "step": 5043 + }, + { + "epoch": 2.373647058823529, + "grad_norm": 0.05947235608154442, + "learning_rate": 0.0002444605014696506, + "loss": 0.0314, + "step": 5044 + }, + { + "epoch": 2.3741176470588234, + "grad_norm": 7.820744350166046e-06, + "learning_rate": 0.0002444264451141868, + "loss": 0.0, + "step": 5045 + }, + { + "epoch": 2.3745882352941177, + "grad_norm": 0.0717438708977383, + "learning_rate": 0.0002443923806942414, + "loss": 0.0286, + "step": 5046 + }, + { + "epoch": 2.375058823529412, + "grad_norm": 0.05137311351775717, + "learning_rate": 0.0002443583082127237, + "loss": 0.017, + "step": 5047 + }, + { + "epoch": 2.375529411764706, + "grad_norm": 0.16070253874770976, + "learning_rate": 0.0002443242276725437, + "loss": 0.0632, + "step": 5048 + }, + { + "epoch": 2.376, + "grad_norm": 0.08219423598063158, + "learning_rate": 0.00024429013907661197, + "loss": 0.0463, + "step": 5049 + }, + { + "epoch": 2.376470588235294, + "grad_norm": 1.995947287715135e-05, + "learning_rate": 0.0002442560424278399, + "loss": 0.0, + "step": 5050 + }, + { + "epoch": 2.3769411764705883, + "grad_norm": 0.0336535281261223, + "learning_rate": 0.0002442219377291395, + "loss": 0.0074, + "step": 5051 + }, + { + "epoch": 2.377411764705882, + "grad_norm": 0.018756535923115845, + "learning_rate": 0.00024418782498342336, + "loss": 0.0037, + "step": 5052 + }, + { + "epoch": 2.3778823529411763, + "grad_norm": 1.8461154993287534e-05, + "learning_rate": 0.00024415370419360506, + "loss": 0.0, + "step": 5053 + }, + { + "epoch": 2.3783529411764706, + "grad_norm": 0.11419909171371403, + "learning_rate": 0.00024411957536259857, + "loss": 0.0711, + "step": 5054 + }, + { + "epoch": 2.378823529411765, + "grad_norm": 0.07912975182578948, + "learning_rate": 0.0002440854384933187, + "loss": 0.0292, + "step": 5055 + }, + { + "epoch": 2.379294117647059, + "grad_norm": 0.04313615087486802, + "learning_rate": 0.0002440512935886809, + "loss": 0.0101, + "step": 5056 + }, + { + "epoch": 2.379764705882353, + "grad_norm": 0.04068363599632849, + "learning_rate": 0.0002440171406516013, + "loss": 0.0126, + "step": 5057 + }, + { + "epoch": 2.380235294117647, + "grad_norm": 0.011369968305445313, + "learning_rate": 0.00024398297968499677, + "loss": 0.001, + "step": 5058 + }, + { + "epoch": 2.3807058823529412, + "grad_norm": 0.13480239490326607, + "learning_rate": 0.00024394881069178477, + "loss": 0.0937, + "step": 5059 + }, + { + "epoch": 2.3811764705882354, + "grad_norm": 0.03510899156034535, + "learning_rate": 0.00024391463367488348, + "loss": 0.0072, + "step": 5060 + }, + { + "epoch": 2.3816470588235292, + "grad_norm": 0.1282041494801134, + "learning_rate": 0.00024388044863721188, + "loss": 0.1028, + "step": 5061 + }, + { + "epoch": 2.3821176470588235, + "grad_norm": 2.4608340903633673e-05, + "learning_rate": 0.00024384625558168948, + "loss": 0.0, + "step": 5062 + }, + { + "epoch": 2.3825882352941177, + "grad_norm": 0.04285860770358214, + "learning_rate": 0.0002438120545112365, + "loss": 0.0128, + "step": 5063 + }, + { + "epoch": 2.383058823529412, + "grad_norm": 0.10945519856934548, + "learning_rate": 0.000243777845428774, + "loss": 0.0619, + "step": 5064 + }, + { + "epoch": 2.383529411764706, + "grad_norm": 1.1343634134361791e-05, + "learning_rate": 0.00024374362833722344, + "loss": 0.0, + "step": 5065 + }, + { + "epoch": 2.384, + "grad_norm": 0.1364074352530012, + "learning_rate": 0.00024370940323950724, + "loss": 0.1104, + "step": 5066 + }, + { + "epoch": 2.384470588235294, + "grad_norm": 9.083388672798306e-06, + "learning_rate": 0.0002436751701385484, + "loss": 0.0, + "step": 5067 + }, + { + "epoch": 2.3849411764705883, + "grad_norm": 0.08808439943781429, + "learning_rate": 0.0002436409290372705, + "loss": 0.0459, + "step": 5068 + }, + { + "epoch": 2.385411764705882, + "grad_norm": 0.10681771647753983, + "learning_rate": 0.000243606679938598, + "loss": 0.0627, + "step": 5069 + }, + { + "epoch": 2.3858823529411763, + "grad_norm": 0.04162391112405428, + "learning_rate": 0.00024357242284545587, + "loss": 0.0099, + "step": 5070 + }, + { + "epoch": 2.3863529411764706, + "grad_norm": 0.18907412186189584, + "learning_rate": 0.00024353815776076985, + "loss": 0.0699, + "step": 5071 + }, + { + "epoch": 2.386823529411765, + "grad_norm": 0.1319619313274212, + "learning_rate": 0.00024350388468746635, + "loss": 0.0948, + "step": 5072 + }, + { + "epoch": 2.387294117647059, + "grad_norm": 0.23082124937701898, + "learning_rate": 0.00024346960362847247, + "loss": 0.1376, + "step": 5073 + }, + { + "epoch": 2.387764705882353, + "grad_norm": 0.084711334636221, + "learning_rate": 0.000243435314586716, + "loss": 0.0421, + "step": 5074 + }, + { + "epoch": 2.388235294117647, + "grad_norm": 0.09187538563564321, + "learning_rate": 0.00024340101756512532, + "loss": 0.0562, + "step": 5075 + }, + { + "epoch": 2.3887058823529412, + "grad_norm": 3.8187881362692625e-05, + "learning_rate": 0.00024336671256662963, + "loss": 0.0, + "step": 5076 + }, + { + "epoch": 2.3891764705882355, + "grad_norm": 8.434251815981764e-06, + "learning_rate": 0.00024333239959415873, + "loss": 0.0, + "step": 5077 + }, + { + "epoch": 2.3896470588235292, + "grad_norm": 0.09409828604568776, + "learning_rate": 0.0002432980786506431, + "loss": 0.0559, + "step": 5078 + }, + { + "epoch": 2.3901176470588235, + "grad_norm": 1.1109423253782572e-05, + "learning_rate": 0.00024326374973901395, + "loss": 0.0, + "step": 5079 + }, + { + "epoch": 2.3905882352941177, + "grad_norm": 0.04801414428778925, + "learning_rate": 0.0002432294128622031, + "loss": 0.019, + "step": 5080 + }, + { + "epoch": 2.391058823529412, + "grad_norm": 0.0530666846013175, + "learning_rate": 0.0002431950680231431, + "loss": 0.0252, + "step": 5081 + }, + { + "epoch": 2.391529411764706, + "grad_norm": 0.13532476238931948, + "learning_rate": 0.00024316071522476718, + "loss": 0.0664, + "step": 5082 + }, + { + "epoch": 2.392, + "grad_norm": 0.0007473727029435707, + "learning_rate": 0.0002431263544700093, + "loss": 0.0, + "step": 5083 + }, + { + "epoch": 2.392470588235294, + "grad_norm": 0.057826798136076094, + "learning_rate": 0.00024309198576180388, + "loss": 0.0214, + "step": 5084 + }, + { + "epoch": 2.3929411764705883, + "grad_norm": 0.12806171675879904, + "learning_rate": 0.00024305760910308627, + "loss": 0.0774, + "step": 5085 + }, + { + "epoch": 2.393411764705882, + "grad_norm": 0.1004472954664245, + "learning_rate": 0.00024302322449679245, + "loss": 0.0471, + "step": 5086 + }, + { + "epoch": 2.3938823529411764, + "grad_norm": 1.8483863753424934e-05, + "learning_rate": 0.0002429888319458589, + "loss": 0.0, + "step": 5087 + }, + { + "epoch": 2.3943529411764706, + "grad_norm": 0.05860014721891746, + "learning_rate": 0.00024295443145322303, + "loss": 0.0192, + "step": 5088 + }, + { + "epoch": 2.394823529411765, + "grad_norm": 0.12390225714999914, + "learning_rate": 0.0002429200230218228, + "loss": 0.0949, + "step": 5089 + }, + { + "epoch": 2.395294117647059, + "grad_norm": 4.699238913039706e-05, + "learning_rate": 0.00024288560665459679, + "loss": 0.0, + "step": 5090 + }, + { + "epoch": 2.395764705882353, + "grad_norm": 0.1329400944364006, + "learning_rate": 0.0002428511823544844, + "loss": 0.0969, + "step": 5091 + }, + { + "epoch": 2.396235294117647, + "grad_norm": 0.18328964588493185, + "learning_rate": 0.00024281675012442563, + "loss": 0.1152, + "step": 5092 + }, + { + "epoch": 2.3967058823529412, + "grad_norm": 0.15082381363669642, + "learning_rate": 0.00024278230996736109, + "loss": 0.0784, + "step": 5093 + }, + { + "epoch": 2.3971764705882355, + "grad_norm": 0.22583612978024992, + "learning_rate": 0.0002427478618862322, + "loss": 0.1185, + "step": 5094 + }, + { + "epoch": 2.3976470588235292, + "grad_norm": 0.10958958549537393, + "learning_rate": 0.00024271340588398097, + "loss": 0.0708, + "step": 5095 + }, + { + "epoch": 2.3981176470588235, + "grad_norm": 9.592489127040295e-06, + "learning_rate": 0.00024267894196355015, + "loss": 0.0, + "step": 5096 + }, + { + "epoch": 2.3985882352941177, + "grad_norm": 0.1079336381380419, + "learning_rate": 0.00024264447012788308, + "loss": 0.05, + "step": 5097 + }, + { + "epoch": 2.399058823529412, + "grad_norm": 0.1731236783159811, + "learning_rate": 0.00024260999037992384, + "loss": 0.103, + "step": 5098 + }, + { + "epoch": 2.3995294117647057, + "grad_norm": 0.11200486594444564, + "learning_rate": 0.00024257550272261714, + "loss": 0.0523, + "step": 5099 + }, + { + "epoch": 2.4, + "grad_norm": 0.1094942896641683, + "learning_rate": 0.00024254100715890846, + "loss": 0.047, + "step": 5100 + }, + { + "epoch": 2.400470588235294, + "grad_norm": 0.1479788686899474, + "learning_rate": 0.00024250650369174386, + "loss": 0.087, + "step": 5101 + }, + { + "epoch": 2.4009411764705884, + "grad_norm": 0.0739547361979303, + "learning_rate": 0.0002424719923240701, + "loss": 0.0311, + "step": 5102 + }, + { + "epoch": 2.401411764705882, + "grad_norm": 0.07687050451837835, + "learning_rate": 0.00024243747305883461, + "loss": 0.0448, + "step": 5103 + }, + { + "epoch": 2.4018823529411764, + "grad_norm": 0.09815495717022633, + "learning_rate": 0.00024240294589898554, + "loss": 0.0622, + "step": 5104 + }, + { + "epoch": 2.4023529411764706, + "grad_norm": 0.11968851796443683, + "learning_rate": 0.00024236841084747165, + "loss": 0.0798, + "step": 5105 + }, + { + "epoch": 2.402823529411765, + "grad_norm": 7.299524816795761e-05, + "learning_rate": 0.00024233386790724242, + "loss": 0.0, + "step": 5106 + }, + { + "epoch": 2.403294117647059, + "grad_norm": 0.08551115375435289, + "learning_rate": 0.000242299317081248, + "loss": 0.0517, + "step": 5107 + }, + { + "epoch": 2.403764705882353, + "grad_norm": 1.4359213347250044e-05, + "learning_rate": 0.00024226475837243908, + "loss": 0.0, + "step": 5108 + }, + { + "epoch": 2.404235294117647, + "grad_norm": 0.05528822812271507, + "learning_rate": 0.00024223019178376732, + "loss": 0.0215, + "step": 5109 + }, + { + "epoch": 2.4047058823529412, + "grad_norm": 0.029214561764492688, + "learning_rate": 0.00024219561731818475, + "loss": 0.0078, + "step": 5110 + }, + { + "epoch": 2.4051764705882355, + "grad_norm": 0.08827415493558721, + "learning_rate": 0.00024216103497864426, + "loss": 0.0572, + "step": 5111 + }, + { + "epoch": 2.4056470588235292, + "grad_norm": 0.1238470876682644, + "learning_rate": 0.00024212644476809934, + "loss": 0.0629, + "step": 5112 + }, + { + "epoch": 2.4061176470588235, + "grad_norm": 0.05345259113064655, + "learning_rate": 0.0002420918466895042, + "loss": 0.0263, + "step": 5113 + }, + { + "epoch": 2.4065882352941177, + "grad_norm": 0.03710970881415922, + "learning_rate": 0.0002420572407458136, + "loss": 0.0129, + "step": 5114 + }, + { + "epoch": 2.407058823529412, + "grad_norm": 0.1588379335281847, + "learning_rate": 0.0002420226269399831, + "loss": 0.0952, + "step": 5115 + }, + { + "epoch": 2.4075294117647057, + "grad_norm": 0.072572218410385, + "learning_rate": 0.00024198800527496889, + "loss": 0.0366, + "step": 5116 + }, + { + "epoch": 2.408, + "grad_norm": 0.08693689314396627, + "learning_rate": 0.00024195337575372788, + "loss": 0.0272, + "step": 5117 + }, + { + "epoch": 2.408470588235294, + "grad_norm": 0.10610739982729092, + "learning_rate": 0.00024191873837921748, + "loss": 0.0527, + "step": 5118 + }, + { + "epoch": 2.4089411764705884, + "grad_norm": 0.10895346342077332, + "learning_rate": 0.000241884093154396, + "loss": 0.0598, + "step": 5119 + }, + { + "epoch": 2.409411764705882, + "grad_norm": 0.06472552777596001, + "learning_rate": 0.00024184944008222227, + "loss": 0.0329, + "step": 5120 + }, + { + "epoch": 2.4098823529411764, + "grad_norm": 0.08978805914935226, + "learning_rate": 0.00024181477916565582, + "loss": 0.0535, + "step": 5121 + }, + { + "epoch": 2.4103529411764706, + "grad_norm": 0.04264263259459109, + "learning_rate": 0.0002417801104076569, + "loss": 0.0104, + "step": 5122 + }, + { + "epoch": 2.410823529411765, + "grad_norm": 0.10955599491225534, + "learning_rate": 0.00024174543381118638, + "loss": 0.0594, + "step": 5123 + }, + { + "epoch": 2.411294117647059, + "grad_norm": 0.051213366147655086, + "learning_rate": 0.00024171074937920579, + "loss": 0.012, + "step": 5124 + }, + { + "epoch": 2.411764705882353, + "grad_norm": 0.020087041949887998, + "learning_rate": 0.00024167605711467738, + "loss": 0.004, + "step": 5125 + }, + { + "epoch": 2.412235294117647, + "grad_norm": 0.06534847504965127, + "learning_rate": 0.000241641357020564, + "loss": 0.0264, + "step": 5126 + }, + { + "epoch": 2.4127058823529413, + "grad_norm": 0.14006569862756935, + "learning_rate": 0.00024160664909982928, + "loss": 0.085, + "step": 5127 + }, + { + "epoch": 2.4131764705882355, + "grad_norm": 0.10800462125434672, + "learning_rate": 0.00024157193335543734, + "loss": 0.0449, + "step": 5128 + }, + { + "epoch": 2.4136470588235293, + "grad_norm": 1.2165904068572507e-05, + "learning_rate": 0.00024153720979035317, + "loss": 0.0, + "step": 5129 + }, + { + "epoch": 2.4141176470588235, + "grad_norm": 0.03655319580015353, + "learning_rate": 0.00024150247840754228, + "loss": 0.0111, + "step": 5130 + }, + { + "epoch": 2.4145882352941177, + "grad_norm": 0.09835569611773438, + "learning_rate": 0.00024146773920997091, + "loss": 0.0529, + "step": 5131 + }, + { + "epoch": 2.415058823529412, + "grad_norm": 0.057083473963563246, + "learning_rate": 0.000241432992200606, + "loss": 0.0271, + "step": 5132 + }, + { + "epoch": 2.4155294117647057, + "grad_norm": 0.042492186857458834, + "learning_rate": 0.000241398237382415, + "loss": 0.0155, + "step": 5133 + }, + { + "epoch": 2.416, + "grad_norm": 3.828201633088068e-05, + "learning_rate": 0.00024136347475836633, + "loss": 0.0, + "step": 5134 + }, + { + "epoch": 2.416470588235294, + "grad_norm": 0.06382707318495655, + "learning_rate": 0.0002413287043314287, + "loss": 0.0267, + "step": 5135 + }, + { + "epoch": 2.4169411764705884, + "grad_norm": 8.514796715543743e-06, + "learning_rate": 0.00024129392610457176, + "loss": 0.0, + "step": 5136 + }, + { + "epoch": 2.417411764705882, + "grad_norm": 0.09170052437239662, + "learning_rate": 0.0002412591400807658, + "loss": 0.0514, + "step": 5137 + }, + { + "epoch": 2.4178823529411764, + "grad_norm": 0.09469983363454366, + "learning_rate": 0.0002412243462629816, + "loss": 0.0445, + "step": 5138 + }, + { + "epoch": 2.4183529411764706, + "grad_norm": 0.07056479369993823, + "learning_rate": 0.0002411895446541908, + "loss": 0.0419, + "step": 5139 + }, + { + "epoch": 2.418823529411765, + "grad_norm": 0.07956502637930671, + "learning_rate": 0.00024115473525736561, + "loss": 0.0444, + "step": 5140 + }, + { + "epoch": 2.419294117647059, + "grad_norm": 0.07360633160943954, + "learning_rate": 0.0002411199180754789, + "loss": 0.04, + "step": 5141 + }, + { + "epoch": 2.419764705882353, + "grad_norm": 0.10992034464323625, + "learning_rate": 0.00024108509311150422, + "loss": 0.0555, + "step": 5142 + }, + { + "epoch": 2.420235294117647, + "grad_norm": 0.15867807404955814, + "learning_rate": 0.00024105026036841582, + "loss": 0.1305, + "step": 5143 + }, + { + "epoch": 2.4207058823529413, + "grad_norm": 0.2728853784594775, + "learning_rate": 0.00024101541984918864, + "loss": 0.179, + "step": 5144 + }, + { + "epoch": 2.4211764705882355, + "grad_norm": 0.07068448618037688, + "learning_rate": 0.00024098057155679813, + "loss": 0.0075, + "step": 5145 + }, + { + "epoch": 2.4216470588235293, + "grad_norm": 5.651045292950585e-06, + "learning_rate": 0.0002409457154942206, + "loss": 0.0, + "step": 5146 + }, + { + "epoch": 2.4221176470588235, + "grad_norm": 0.028610302228186963, + "learning_rate": 0.00024091085166443285, + "loss": 0.004, + "step": 5147 + }, + { + "epoch": 2.4225882352941177, + "grad_norm": 0.09424065690605068, + "learning_rate": 0.00024087598007041243, + "loss": 0.0193, + "step": 5148 + }, + { + "epoch": 2.423058823529412, + "grad_norm": 0.04263232410888114, + "learning_rate": 0.0002408411007151376, + "loss": 0.0129, + "step": 5149 + }, + { + "epoch": 2.4235294117647057, + "grad_norm": 0.06441055474512948, + "learning_rate": 0.00024080621360158717, + "loss": 0.0234, + "step": 5150 + }, + { + "epoch": 2.424, + "grad_norm": 0.10852929473648293, + "learning_rate": 0.00024077131873274072, + "loss": 0.0544, + "step": 5151 + }, + { + "epoch": 2.424470588235294, + "grad_norm": 0.08576640844743236, + "learning_rate": 0.0002407364161115784, + "loss": 0.0565, + "step": 5152 + }, + { + "epoch": 2.4249411764705884, + "grad_norm": 0.16988910755566916, + "learning_rate": 0.00024070150574108113, + "loss": 0.0989, + "step": 5153 + }, + { + "epoch": 2.425411764705882, + "grad_norm": 0.0956402128904288, + "learning_rate": 0.00024066658762423034, + "loss": 0.0643, + "step": 5154 + }, + { + "epoch": 2.4258823529411764, + "grad_norm": 0.10806554161647043, + "learning_rate": 0.00024063166176400825, + "loss": 0.0677, + "step": 5155 + }, + { + "epoch": 2.4263529411764706, + "grad_norm": 0.04078013150824744, + "learning_rate": 0.0002405967281633977, + "loss": 0.0137, + "step": 5156 + }, + { + "epoch": 2.426823529411765, + "grad_norm": 0.09803525400858147, + "learning_rate": 0.00024056178682538222, + "loss": 0.0533, + "step": 5157 + }, + { + "epoch": 2.427294117647059, + "grad_norm": 0.16578620639004793, + "learning_rate": 0.00024052683775294592, + "loss": 0.1322, + "step": 5158 + }, + { + "epoch": 2.427764705882353, + "grad_norm": 0.05434436421246087, + "learning_rate": 0.00024049188094907363, + "loss": 0.0117, + "step": 5159 + }, + { + "epoch": 2.428235294117647, + "grad_norm": 0.05340353199897802, + "learning_rate": 0.00024045691641675086, + "loss": 0.0296, + "step": 5160 + }, + { + "epoch": 2.4287058823529413, + "grad_norm": 0.06546279589597259, + "learning_rate": 0.00024042194415896376, + "loss": 0.037, + "step": 5161 + }, + { + "epoch": 2.429176470588235, + "grad_norm": 0.174406080024288, + "learning_rate": 0.00024038696417869908, + "loss": 0.1455, + "step": 5162 + }, + { + "epoch": 2.4296470588235293, + "grad_norm": 0.09230840439857743, + "learning_rate": 0.00024035197647894434, + "loss": 0.0558, + "step": 5163 + }, + { + "epoch": 2.4301176470588235, + "grad_norm": 0.06606743981636, + "learning_rate": 0.00024031698106268763, + "loss": 0.0298, + "step": 5164 + }, + { + "epoch": 2.4305882352941177, + "grad_norm": 0.12959252159617565, + "learning_rate": 0.0002402819779329177, + "loss": 0.0297, + "step": 5165 + }, + { + "epoch": 2.431058823529412, + "grad_norm": 0.07694798874412875, + "learning_rate": 0.00024024696709262406, + "loss": 0.0441, + "step": 5166 + }, + { + "epoch": 2.4315294117647057, + "grad_norm": 0.04932599976442681, + "learning_rate": 0.00024021194854479674, + "loss": 0.0202, + "step": 5167 + }, + { + "epoch": 2.432, + "grad_norm": 0.01791790440540607, + "learning_rate": 0.00024017692229242655, + "loss": 0.0029, + "step": 5168 + }, + { + "epoch": 2.432470588235294, + "grad_norm": 0.10874622200037518, + "learning_rate": 0.00024014188833850486, + "loss": 0.0673, + "step": 5169 + }, + { + "epoch": 2.4329411764705884, + "grad_norm": 1.003594725929211e-05, + "learning_rate": 0.0002401068466860238, + "loss": 0.0, + "step": 5170 + }, + { + "epoch": 2.433411764705882, + "grad_norm": 0.09699203605238849, + "learning_rate": 0.00024007179733797604, + "loss": 0.0187, + "step": 5171 + }, + { + "epoch": 2.4338823529411764, + "grad_norm": 0.08821494180978716, + "learning_rate": 0.00024003674029735497, + "loss": 0.0473, + "step": 5172 + }, + { + "epoch": 2.4343529411764706, + "grad_norm": 0.1881973064426054, + "learning_rate": 0.00024000167556715468, + "loss": 0.1297, + "step": 5173 + }, + { + "epoch": 2.434823529411765, + "grad_norm": 0.03881011086708426, + "learning_rate": 0.00023996660315036986, + "loss": 0.0152, + "step": 5174 + }, + { + "epoch": 2.435294117647059, + "grad_norm": 0.07492879094808985, + "learning_rate": 0.00023993152304999582, + "loss": 0.0366, + "step": 5175 + }, + { + "epoch": 2.435764705882353, + "grad_norm": 0.10582652425527202, + "learning_rate": 0.0002398964352690286, + "loss": 0.0784, + "step": 5176 + }, + { + "epoch": 2.436235294117647, + "grad_norm": 0.11194987161112409, + "learning_rate": 0.00023986133981046485, + "loss": 0.0637, + "step": 5177 + }, + { + "epoch": 2.4367058823529413, + "grad_norm": 0.10365856826511383, + "learning_rate": 0.00023982623667730195, + "loss": 0.0568, + "step": 5178 + }, + { + "epoch": 2.437176470588235, + "grad_norm": 0.07054513939915934, + "learning_rate": 0.0002397911258725378, + "loss": 0.0375, + "step": 5179 + }, + { + "epoch": 2.4376470588235293, + "grad_norm": 4.151140360470842e-06, + "learning_rate": 0.00023975600739917113, + "loss": 0.0, + "step": 5180 + }, + { + "epoch": 2.4381176470588235, + "grad_norm": 0.09945734417550468, + "learning_rate": 0.00023972088126020114, + "loss": 0.0611, + "step": 5181 + }, + { + "epoch": 2.4385882352941177, + "grad_norm": 1.8491945904319447e-06, + "learning_rate": 0.00023968574745862782, + "loss": 0.0, + "step": 5182 + }, + { + "epoch": 2.439058823529412, + "grad_norm": 0.08371709726595668, + "learning_rate": 0.0002396506059974518, + "loss": 0.0462, + "step": 5183 + }, + { + "epoch": 2.4395294117647057, + "grad_norm": 3.1012452384140576e-07, + "learning_rate": 0.00023961545687967422, + "loss": 0.0, + "step": 5184 + }, + { + "epoch": 2.44, + "grad_norm": 0.10230881343537683, + "learning_rate": 0.0002395803001082971, + "loss": 0.0703, + "step": 5185 + }, + { + "epoch": 2.440470588235294, + "grad_norm": 1.68854055358265e-06, + "learning_rate": 0.00023954513568632296, + "loss": 0.0, + "step": 5186 + }, + { + "epoch": 2.4409411764705884, + "grad_norm": 1.389455886018175e-06, + "learning_rate": 0.000239509963616755, + "loss": 0.0, + "step": 5187 + }, + { + "epoch": 2.441411764705882, + "grad_norm": 0.09692471713137038, + "learning_rate": 0.00023947478390259712, + "loss": 0.0466, + "step": 5188 + }, + { + "epoch": 2.4418823529411764, + "grad_norm": 2.0344623538573046e-05, + "learning_rate": 0.00023943959654685383, + "loss": 0.0, + "step": 5189 + }, + { + "epoch": 2.4423529411764706, + "grad_norm": 0.05572135157155835, + "learning_rate": 0.00023940440155253024, + "loss": 0.0217, + "step": 5190 + }, + { + "epoch": 2.442823529411765, + "grad_norm": 1.1059976872422846e-05, + "learning_rate": 0.0002393691989226323, + "loss": 0.0, + "step": 5191 + }, + { + "epoch": 2.443294117647059, + "grad_norm": 0.09792698168531838, + "learning_rate": 0.00023933398866016642, + "loss": 0.0554, + "step": 5192 + }, + { + "epoch": 2.443764705882353, + "grad_norm": 0.08243031854140072, + "learning_rate": 0.0002392987707681397, + "loss": 0.0399, + "step": 5193 + }, + { + "epoch": 2.444235294117647, + "grad_norm": 0.0799136334057871, + "learning_rate": 0.00023926354524955994, + "loss": 0.0348, + "step": 5194 + }, + { + "epoch": 2.4447058823529413, + "grad_norm": 0.09332101740483444, + "learning_rate": 0.0002392283121074356, + "loss": 0.058, + "step": 5195 + }, + { + "epoch": 2.445176470588235, + "grad_norm": 0.1020193164468728, + "learning_rate": 0.00023919307134477572, + "loss": 0.0479, + "step": 5196 + }, + { + "epoch": 2.4456470588235293, + "grad_norm": 0.06259693907105485, + "learning_rate": 0.00023915782296459007, + "loss": 0.0096, + "step": 5197 + }, + { + "epoch": 2.4461176470588235, + "grad_norm": 0.07033941683179623, + "learning_rate": 0.00023912256696988904, + "loss": 0.0333, + "step": 5198 + }, + { + "epoch": 2.4465882352941177, + "grad_norm": 0.10914045085535484, + "learning_rate": 0.00023908730336368364, + "loss": 0.056, + "step": 5199 + }, + { + "epoch": 2.447058823529412, + "grad_norm": 0.023471976050434284, + "learning_rate": 0.00023905203214898558, + "loss": 0.0059, + "step": 5200 + }, + { + "epoch": 2.4475294117647057, + "grad_norm": 0.06606200529311218, + "learning_rate": 0.00023901675332880715, + "loss": 0.0195, + "step": 5201 + }, + { + "epoch": 2.448, + "grad_norm": 0.04278630397291829, + "learning_rate": 0.00023898146690616134, + "loss": 0.0097, + "step": 5202 + }, + { + "epoch": 2.448470588235294, + "grad_norm": 0.08789315454357169, + "learning_rate": 0.0002389461728840618, + "loss": 0.0516, + "step": 5203 + }, + { + "epoch": 2.4489411764705884, + "grad_norm": 0.05906231468196163, + "learning_rate": 0.00023891087126552292, + "loss": 0.032, + "step": 5204 + }, + { + "epoch": 2.449411764705882, + "grad_norm": 0.07696076806092203, + "learning_rate": 0.00023887556205355945, + "loss": 0.0391, + "step": 5205 + }, + { + "epoch": 2.4498823529411764, + "grad_norm": 1.049794822497002e-05, + "learning_rate": 0.00023884024525118702, + "loss": 0.0, + "step": 5206 + }, + { + "epoch": 2.4503529411764706, + "grad_norm": 1.495310449755894e-06, + "learning_rate": 0.00023880492086142193, + "loss": 0.0, + "step": 5207 + }, + { + "epoch": 2.450823529411765, + "grad_norm": 0.029328882917930264, + "learning_rate": 0.00023876958888728098, + "loss": 0.0041, + "step": 5208 + }, + { + "epoch": 2.451294117647059, + "grad_norm": 0.04834065906512399, + "learning_rate": 0.00023873424933178178, + "loss": 0.0164, + "step": 5209 + }, + { + "epoch": 2.451764705882353, + "grad_norm": 0.06695388783284402, + "learning_rate": 0.00023869890219794238, + "loss": 0.0274, + "step": 5210 + }, + { + "epoch": 2.452235294117647, + "grad_norm": 0.08307089072772571, + "learning_rate": 0.0002386635474887817, + "loss": 0.0397, + "step": 5211 + }, + { + "epoch": 2.4527058823529413, + "grad_norm": 0.13703407343756407, + "learning_rate": 0.0002386281852073192, + "loss": 0.074, + "step": 5212 + }, + { + "epoch": 2.453176470588235, + "grad_norm": 0.06048481749355958, + "learning_rate": 0.00023859281535657496, + "loss": 0.0172, + "step": 5213 + }, + { + "epoch": 2.4536470588235293, + "grad_norm": 0.0014061692587853823, + "learning_rate": 0.00023855743793956967, + "loss": 0.0, + "step": 5214 + }, + { + "epoch": 2.4541176470588235, + "grad_norm": 0.09759152661862368, + "learning_rate": 0.0002385220529593249, + "loss": 0.0658, + "step": 5215 + }, + { + "epoch": 2.4545882352941177, + "grad_norm": 0.10118672871795632, + "learning_rate": 0.00023848666041886258, + "loss": 0.0445, + "step": 5216 + }, + { + "epoch": 2.455058823529412, + "grad_norm": 0.18956079216277277, + "learning_rate": 0.00023845126032120539, + "loss": 0.0686, + "step": 5217 + }, + { + "epoch": 2.4555294117647057, + "grad_norm": 0.08049367685048855, + "learning_rate": 0.00023841585266937677, + "loss": 0.0329, + "step": 5218 + }, + { + "epoch": 2.456, + "grad_norm": 3.254852010968208e-05, + "learning_rate": 0.00023838043746640065, + "loss": 0.0, + "step": 5219 + }, + { + "epoch": 2.456470588235294, + "grad_norm": 0.13676411403851332, + "learning_rate": 0.00023834501471530165, + "loss": 0.0642, + "step": 5220 + }, + { + "epoch": 2.4569411764705884, + "grad_norm": 0.0511088252876636, + "learning_rate": 0.0002383095844191051, + "loss": 0.0153, + "step": 5221 + }, + { + "epoch": 2.457411764705882, + "grad_norm": 1.5099903585141124e-07, + "learning_rate": 0.0002382741465808368, + "loss": 0.0, + "step": 5222 + }, + { + "epoch": 2.4578823529411764, + "grad_norm": 0.13808428170373174, + "learning_rate": 0.0002382387012035235, + "loss": 0.0261, + "step": 5223 + }, + { + "epoch": 2.4583529411764706, + "grad_norm": 0.19082791406582486, + "learning_rate": 0.0002382032482901922, + "loss": 0.0629, + "step": 5224 + }, + { + "epoch": 2.458823529411765, + "grad_norm": 0.06663762066523439, + "learning_rate": 0.00023816778784387094, + "loss": 0.0298, + "step": 5225 + }, + { + "epoch": 2.4592941176470586, + "grad_norm": 0.16281700388843506, + "learning_rate": 0.00023813231986758815, + "loss": 0.0958, + "step": 5226 + }, + { + "epoch": 2.459764705882353, + "grad_norm": 4.429677022749768e-05, + "learning_rate": 0.00023809684436437294, + "loss": 0.0, + "step": 5227 + }, + { + "epoch": 2.460235294117647, + "grad_norm": 0.12143931809015095, + "learning_rate": 0.00023806136133725506, + "loss": 0.0768, + "step": 5228 + }, + { + "epoch": 2.4607058823529413, + "grad_norm": 0.03135633732105356, + "learning_rate": 0.00023802587078926505, + "loss": 0.0084, + "step": 5229 + }, + { + "epoch": 2.461176470588235, + "grad_norm": 0.07616892481619045, + "learning_rate": 0.00023799037272343387, + "loss": 0.0482, + "step": 5230 + }, + { + "epoch": 2.4616470588235293, + "grad_norm": 0.0838054795072248, + "learning_rate": 0.00023795486714279327, + "loss": 0.0358, + "step": 5231 + }, + { + "epoch": 2.4621176470588235, + "grad_norm": 0.05031439270243378, + "learning_rate": 0.0002379193540503756, + "loss": 0.0217, + "step": 5232 + }, + { + "epoch": 2.4625882352941177, + "grad_norm": 0.05196503151281356, + "learning_rate": 0.00023788383344921384, + "loss": 0.0266, + "step": 5233 + }, + { + "epoch": 2.463058823529412, + "grad_norm": 5.265587925800378e-07, + "learning_rate": 0.0002378483053423417, + "loss": 0.0, + "step": 5234 + }, + { + "epoch": 2.4635294117647057, + "grad_norm": 0.037740858827006514, + "learning_rate": 0.00023781276973279335, + "loss": 0.0173, + "step": 5235 + }, + { + "epoch": 2.464, + "grad_norm": 0.05948395780506221, + "learning_rate": 0.0002377772266236037, + "loss": 0.018, + "step": 5236 + }, + { + "epoch": 2.464470588235294, + "grad_norm": 0.05425043499919518, + "learning_rate": 0.0002377416760178084, + "loss": 0.0202, + "step": 5237 + }, + { + "epoch": 2.4649411764705884, + "grad_norm": 0.11721086471540475, + "learning_rate": 0.00023770611791844364, + "loss": 0.0752, + "step": 5238 + }, + { + "epoch": 2.465411764705882, + "grad_norm": 0.055737938394088075, + "learning_rate": 0.00023767055232854615, + "loss": 0.0228, + "step": 5239 + }, + { + "epoch": 2.4658823529411764, + "grad_norm": 0.0848421359419457, + "learning_rate": 0.00023763497925115347, + "loss": 0.0537, + "step": 5240 + }, + { + "epoch": 2.4663529411764706, + "grad_norm": 1.2861933461224592e-06, + "learning_rate": 0.00023759939868930373, + "loss": 0.0, + "step": 5241 + }, + { + "epoch": 2.466823529411765, + "grad_norm": 0.1384184544052983, + "learning_rate": 0.00023756381064603564, + "loss": 0.0884, + "step": 5242 + }, + { + "epoch": 2.4672941176470586, + "grad_norm": 0.013293057327096479, + "learning_rate": 0.0002375282151243887, + "loss": 0.0022, + "step": 5243 + }, + { + "epoch": 2.467764705882353, + "grad_norm": 0.0712772430623255, + "learning_rate": 0.00023749261212740278, + "loss": 0.0321, + "step": 5244 + }, + { + "epoch": 2.468235294117647, + "grad_norm": 0.11073340623220568, + "learning_rate": 0.00023745700165811868, + "loss": 0.0521, + "step": 5245 + }, + { + "epoch": 2.4687058823529413, + "grad_norm": 0.15080853284317175, + "learning_rate": 0.00023742138371957765, + "loss": 0.0994, + "step": 5246 + }, + { + "epoch": 2.469176470588235, + "grad_norm": 0.05836410633251975, + "learning_rate": 0.00023738575831482164, + "loss": 0.0274, + "step": 5247 + }, + { + "epoch": 2.4696470588235293, + "grad_norm": 1.5040201903736584e-07, + "learning_rate": 0.0002373501254468933, + "loss": 0.0, + "step": 5248 + }, + { + "epoch": 2.4701176470588235, + "grad_norm": 0.12335231464671939, + "learning_rate": 0.00023731448511883575, + "loss": 0.0386, + "step": 5249 + }, + { + "epoch": 2.4705882352941178, + "grad_norm": 0.038742876180496315, + "learning_rate": 0.00023727883733369292, + "loss": 0.013, + "step": 5250 + }, + { + "epoch": 2.471058823529412, + "grad_norm": 0.05594662689173837, + "learning_rate": 0.00023724318209450925, + "loss": 0.0243, + "step": 5251 + }, + { + "epoch": 2.4715294117647058, + "grad_norm": 0.04457156734883388, + "learning_rate": 0.00023720751940432992, + "loss": 0.0157, + "step": 5252 + }, + { + "epoch": 2.472, + "grad_norm": 1.9309560513306704e-07, + "learning_rate": 0.0002371718492662007, + "loss": 0.0, + "step": 5253 + }, + { + "epoch": 2.472470588235294, + "grad_norm": 0.036264902324589274, + "learning_rate": 0.00023713617168316797, + "loss": 0.0109, + "step": 5254 + }, + { + "epoch": 2.4729411764705884, + "grad_norm": 0.11229483770751733, + "learning_rate": 0.00023710048665827872, + "loss": 0.0685, + "step": 5255 + }, + { + "epoch": 2.473411764705882, + "grad_norm": 4.868999077126919e-07, + "learning_rate": 0.00023706479419458075, + "loss": 0.0, + "step": 5256 + }, + { + "epoch": 2.4738823529411764, + "grad_norm": 0.04831515788068069, + "learning_rate": 0.00023702909429512228, + "loss": 0.0184, + "step": 5257 + }, + { + "epoch": 2.4743529411764706, + "grad_norm": 0.05668974277697909, + "learning_rate": 0.00023699338696295228, + "loss": 0.0279, + "step": 5258 + }, + { + "epoch": 2.474823529411765, + "grad_norm": 0.08932036102221023, + "learning_rate": 0.00023695767220112033, + "loss": 0.0621, + "step": 5259 + }, + { + "epoch": 2.4752941176470586, + "grad_norm": 3.338756478689793e-05, + "learning_rate": 0.0002369219500126767, + "loss": 0.0, + "step": 5260 + }, + { + "epoch": 2.475764705882353, + "grad_norm": 0.06316983568815908, + "learning_rate": 0.00023688622040067217, + "loss": 0.0255, + "step": 5261 + }, + { + "epoch": 2.476235294117647, + "grad_norm": 0.02864933081531991, + "learning_rate": 0.0002368504833681582, + "loss": 0.0064, + "step": 5262 + }, + { + "epoch": 2.4767058823529413, + "grad_norm": 0.09391536284573425, + "learning_rate": 0.00023681473891818699, + "loss": 0.0489, + "step": 5263 + }, + { + "epoch": 2.477176470588235, + "grad_norm": 0.17976114584635572, + "learning_rate": 0.00023677898705381123, + "loss": 0.0941, + "step": 5264 + }, + { + "epoch": 2.4776470588235293, + "grad_norm": 0.0359927095232551, + "learning_rate": 0.00023674322777808432, + "loss": 0.008, + "step": 5265 + }, + { + "epoch": 2.4781176470588235, + "grad_norm": 0.09990954011224538, + "learning_rate": 0.0002367074610940603, + "loss": 0.0703, + "step": 5266 + }, + { + "epoch": 2.4785882352941178, + "grad_norm": 0.04809389821174978, + "learning_rate": 0.00023667168700479383, + "loss": 0.0097, + "step": 5267 + }, + { + "epoch": 2.479058823529412, + "grad_norm": 0.17438280436606235, + "learning_rate": 0.0002366359055133401, + "loss": 0.0982, + "step": 5268 + }, + { + "epoch": 2.4795294117647058, + "grad_norm": 0.06066300588729174, + "learning_rate": 0.00023660011662275516, + "loss": 0.0252, + "step": 5269 + }, + { + "epoch": 2.48, + "grad_norm": 0.059667902344988824, + "learning_rate": 0.00023656432033609544, + "loss": 0.0088, + "step": 5270 + }, + { + "epoch": 2.480470588235294, + "grad_norm": 1.6127755509469852e-07, + "learning_rate": 0.0002365285166564182, + "loss": 0.0, + "step": 5271 + }, + { + "epoch": 2.4809411764705884, + "grad_norm": 0.08966851789542572, + "learning_rate": 0.00023649270558678123, + "loss": 0.0334, + "step": 5272 + }, + { + "epoch": 2.481411764705882, + "grad_norm": 7.084354822617719e-06, + "learning_rate": 0.00023645688713024295, + "loss": 0.0, + "step": 5273 + }, + { + "epoch": 2.4818823529411764, + "grad_norm": 0.05591944494837811, + "learning_rate": 0.0002364210612898624, + "loss": 0.0131, + "step": 5274 + }, + { + "epoch": 2.4823529411764707, + "grad_norm": 0.0792064670113442, + "learning_rate": 0.00023638522806869935, + "loss": 0.027, + "step": 5275 + }, + { + "epoch": 2.482823529411765, + "grad_norm": 0.1541009867979567, + "learning_rate": 0.0002363493874698141, + "loss": 0.0883, + "step": 5276 + }, + { + "epoch": 2.4832941176470587, + "grad_norm": 0.1480855927031793, + "learning_rate": 0.0002363135394962676, + "loss": 0.0753, + "step": 5277 + }, + { + "epoch": 2.483764705882353, + "grad_norm": 0.13520371237971796, + "learning_rate": 0.00023627768415112154, + "loss": 0.0566, + "step": 5278 + }, + { + "epoch": 2.484235294117647, + "grad_norm": 0.03444895236140899, + "learning_rate": 0.00023624182143743797, + "loss": 0.0094, + "step": 5279 + }, + { + "epoch": 2.4847058823529413, + "grad_norm": 0.10308308752527645, + "learning_rate": 0.00023620595135827985, + "loss": 0.0577, + "step": 5280 + }, + { + "epoch": 2.485176470588235, + "grad_norm": 0.13639869093524898, + "learning_rate": 0.0002361700739167107, + "loss": 0.082, + "step": 5281 + }, + { + "epoch": 2.4856470588235293, + "grad_norm": 0.04305000515769087, + "learning_rate": 0.00023613418911579454, + "loss": 0.0123, + "step": 5282 + }, + { + "epoch": 2.4861176470588235, + "grad_norm": 0.09042933229678152, + "learning_rate": 0.00023609829695859618, + "loss": 0.0507, + "step": 5283 + }, + { + "epoch": 2.4865882352941178, + "grad_norm": 1.240682776303458e-05, + "learning_rate": 0.00023606239744818094, + "loss": 0.0, + "step": 5284 + }, + { + "epoch": 2.487058823529412, + "grad_norm": 0.09372925474994585, + "learning_rate": 0.0002360264905876148, + "loss": 0.0373, + "step": 5285 + }, + { + "epoch": 2.4875294117647058, + "grad_norm": 0.09169286082600396, + "learning_rate": 0.0002359905763799644, + "loss": 0.0353, + "step": 5286 + }, + { + "epoch": 2.488, + "grad_norm": 0.0009284658202023037, + "learning_rate": 0.00023595465482829697, + "loss": 0.0, + "step": 5287 + }, + { + "epoch": 2.488470588235294, + "grad_norm": 0.09013804476085135, + "learning_rate": 0.0002359187259356805, + "loss": 0.0434, + "step": 5288 + }, + { + "epoch": 2.488941176470588, + "grad_norm": 0.040398072474921, + "learning_rate": 0.00023588278970518333, + "loss": 0.0118, + "step": 5289 + }, + { + "epoch": 2.489411764705882, + "grad_norm": 0.0570925848881619, + "learning_rate": 0.00023584684613987472, + "loss": 0.0228, + "step": 5290 + }, + { + "epoch": 2.4898823529411764, + "grad_norm": 0.054140747748199845, + "learning_rate": 0.00023581089524282432, + "loss": 0.0142, + "step": 5291 + }, + { + "epoch": 2.4903529411764707, + "grad_norm": 0.04715024020381154, + "learning_rate": 0.0002357749370171026, + "loss": 0.018, + "step": 5292 + }, + { + "epoch": 2.490823529411765, + "grad_norm": 0.16031238418030191, + "learning_rate": 0.00023573897146578053, + "loss": 0.0902, + "step": 5293 + }, + { + "epoch": 2.4912941176470587, + "grad_norm": 0.05478208745492306, + "learning_rate": 0.00023570299859192976, + "loss": 0.0283, + "step": 5294 + }, + { + "epoch": 2.491764705882353, + "grad_norm": 0.07584262809957207, + "learning_rate": 0.00023566701839862253, + "loss": 0.0313, + "step": 5295 + }, + { + "epoch": 2.492235294117647, + "grad_norm": 0.15265081650611187, + "learning_rate": 0.00023563103088893172, + "loss": 0.0691, + "step": 5296 + }, + { + "epoch": 2.4927058823529413, + "grad_norm": 0.03129695462846656, + "learning_rate": 0.00023559503606593088, + "loss": 0.0069, + "step": 5297 + }, + { + "epoch": 2.493176470588235, + "grad_norm": 0.04639348528197928, + "learning_rate": 0.0002355590339326941, + "loss": 0.0107, + "step": 5298 + }, + { + "epoch": 2.4936470588235293, + "grad_norm": 4.253081148035005e-07, + "learning_rate": 0.00023552302449229612, + "loss": 0.0, + "step": 5299 + }, + { + "epoch": 2.4941176470588236, + "grad_norm": 0.061989731362949836, + "learning_rate": 0.00023548700774781242, + "loss": 0.0214, + "step": 5300 + }, + { + "epoch": 2.4945882352941178, + "grad_norm": 0.10992720107547899, + "learning_rate": 0.00023545098370231893, + "loss": 0.0635, + "step": 5301 + }, + { + "epoch": 2.495058823529412, + "grad_norm": 2.6942779894141024e-06, + "learning_rate": 0.00023541495235889227, + "loss": 0.0, + "step": 5302 + }, + { + "epoch": 2.4955294117647058, + "grad_norm": 0.025282785286920555, + "learning_rate": 0.0002353789137206098, + "loss": 0.0032, + "step": 5303 + }, + { + "epoch": 2.496, + "grad_norm": 0.1896792607074941, + "learning_rate": 0.00023534286779054923, + "loss": 0.1521, + "step": 5304 + }, + { + "epoch": 2.496470588235294, + "grad_norm": 0.06243488260916051, + "learning_rate": 0.00023530681457178916, + "loss": 0.0295, + "step": 5305 + }, + { + "epoch": 2.496941176470588, + "grad_norm": 0.0993957959702008, + "learning_rate": 0.00023527075406740876, + "loss": 0.0632, + "step": 5306 + }, + { + "epoch": 2.497411764705882, + "grad_norm": 0.055387640973491956, + "learning_rate": 0.00023523468628048764, + "loss": 0.0156, + "step": 5307 + }, + { + "epoch": 2.4978823529411764, + "grad_norm": 0.25234233532172146, + "learning_rate": 0.0002351986112141063, + "loss": 0.118, + "step": 5308 + }, + { + "epoch": 2.4983529411764707, + "grad_norm": 0.030380571102974516, + "learning_rate": 0.00023516252887134558, + "loss": 0.0104, + "step": 5309 + }, + { + "epoch": 2.498823529411765, + "grad_norm": 0.04109878528969162, + "learning_rate": 0.00023512643925528725, + "loss": 0.0106, + "step": 5310 + }, + { + "epoch": 2.4992941176470587, + "grad_norm": 0.05152754159290384, + "learning_rate": 0.00023509034236901348, + "loss": 0.0108, + "step": 5311 + }, + { + "epoch": 2.499764705882353, + "grad_norm": 0.06553754742280955, + "learning_rate": 0.00023505423821560705, + "loss": 0.0411, + "step": 5312 + }, + { + "epoch": 2.500235294117647, + "grad_norm": 0.08084286569315656, + "learning_rate": 0.00023501812679815152, + "loss": 0.0306, + "step": 5313 + }, + { + "epoch": 2.5007058823529413, + "grad_norm": 0.10115422003050509, + "learning_rate": 0.00023498200811973092, + "loss": 0.0579, + "step": 5314 + }, + { + "epoch": 2.501176470588235, + "grad_norm": 0.11420883129531018, + "learning_rate": 0.00023494588218343003, + "loss": 0.0744, + "step": 5315 + }, + { + "epoch": 2.5016470588235293, + "grad_norm": 0.09960360995017954, + "learning_rate": 0.0002349097489923341, + "loss": 0.0472, + "step": 5316 + }, + { + "epoch": 2.5021176470588236, + "grad_norm": 0.06776896287968913, + "learning_rate": 0.00023487360854952915, + "loss": 0.0319, + "step": 5317 + }, + { + "epoch": 2.502588235294118, + "grad_norm": 0.06293369738578278, + "learning_rate": 0.00023483746085810176, + "loss": 0.0308, + "step": 5318 + }, + { + "epoch": 2.503058823529412, + "grad_norm": 0.09381844592961504, + "learning_rate": 0.00023480130592113904, + "loss": 0.0569, + "step": 5319 + }, + { + "epoch": 2.503529411764706, + "grad_norm": 0.13927745925469376, + "learning_rate": 0.00023476514374172887, + "loss": 0.0883, + "step": 5320 + }, + { + "epoch": 2.504, + "grad_norm": 0.05774218788943491, + "learning_rate": 0.0002347289743229596, + "loss": 0.0178, + "step": 5321 + }, + { + "epoch": 2.5044705882352942, + "grad_norm": 0.19229273534901378, + "learning_rate": 0.00023469279766792037, + "loss": 0.0827, + "step": 5322 + }, + { + "epoch": 2.504941176470588, + "grad_norm": 0.08307102613918613, + "learning_rate": 0.0002346566137797008, + "loss": 0.0288, + "step": 5323 + }, + { + "epoch": 2.5054117647058822, + "grad_norm": 0.06561697274391082, + "learning_rate": 0.0002346204226613912, + "loss": 0.0357, + "step": 5324 + }, + { + "epoch": 2.5058823529411764, + "grad_norm": 9.960103359067666e-05, + "learning_rate": 0.0002345842243160824, + "loss": 0.0, + "step": 5325 + }, + { + "epoch": 2.5063529411764707, + "grad_norm": 0.10678430023400544, + "learning_rate": 0.00023454801874686596, + "loss": 0.0265, + "step": 5326 + }, + { + "epoch": 2.506823529411765, + "grad_norm": 0.103276904978311, + "learning_rate": 0.00023451180595683409, + "loss": 0.0582, + "step": 5327 + }, + { + "epoch": 2.5072941176470587, + "grad_norm": 1.5070185273504998e-05, + "learning_rate": 0.0002344755859490794, + "loss": 0.0, + "step": 5328 + }, + { + "epoch": 2.507764705882353, + "grad_norm": 0.14174940238883815, + "learning_rate": 0.00023443935872669534, + "loss": 0.0845, + "step": 5329 + }, + { + "epoch": 2.508235294117647, + "grad_norm": 0.07527350566563944, + "learning_rate": 0.00023440312429277583, + "loss": 0.0403, + "step": 5330 + }, + { + "epoch": 2.5087058823529413, + "grad_norm": 0.12546101473248755, + "learning_rate": 0.00023436688265041555, + "loss": 0.0322, + "step": 5331 + }, + { + "epoch": 2.509176470588235, + "grad_norm": 0.04043769134474353, + "learning_rate": 0.0002343306338027097, + "loss": 0.0093, + "step": 5332 + }, + { + "epoch": 2.5096470588235293, + "grad_norm": 0.06829889490974549, + "learning_rate": 0.00023429437775275406, + "loss": 0.0274, + "step": 5333 + }, + { + "epoch": 2.5101176470588236, + "grad_norm": 0.03634429401564376, + "learning_rate": 0.00023425811450364508, + "loss": 0.0116, + "step": 5334 + }, + { + "epoch": 2.510588235294118, + "grad_norm": 0.10908003949440083, + "learning_rate": 0.00023422184405847985, + "loss": 0.0787, + "step": 5335 + }, + { + "epoch": 2.511058823529412, + "grad_norm": 0.10106749636582127, + "learning_rate": 0.00023418556642035602, + "loss": 0.0628, + "step": 5336 + }, + { + "epoch": 2.511529411764706, + "grad_norm": 3.6693417150003096e-06, + "learning_rate": 0.0002341492815923719, + "loss": 0.0, + "step": 5337 + }, + { + "epoch": 2.512, + "grad_norm": 0.06957778987454984, + "learning_rate": 0.00023411298957762636, + "loss": 0.0291, + "step": 5338 + }, + { + "epoch": 2.5124705882352942, + "grad_norm": 0.13386237167360884, + "learning_rate": 0.00023407669037921902, + "loss": 0.0425, + "step": 5339 + }, + { + "epoch": 2.512941176470588, + "grad_norm": 0.08242100506830975, + "learning_rate": 0.00023404038400024984, + "loss": 0.0378, + "step": 5340 + }, + { + "epoch": 2.5134117647058822, + "grad_norm": 0.0795898893428915, + "learning_rate": 0.00023400407044381965, + "loss": 0.0463, + "step": 5341 + }, + { + "epoch": 2.5138823529411765, + "grad_norm": 0.06484064814843638, + "learning_rate": 0.00023396774971302985, + "loss": 0.0224, + "step": 5342 + }, + { + "epoch": 2.5143529411764707, + "grad_norm": 0.09066320196150825, + "learning_rate": 0.00023393142181098235, + "loss": 0.051, + "step": 5343 + }, + { + "epoch": 2.514823529411765, + "grad_norm": 0.08549018966975803, + "learning_rate": 0.00023389508674077973, + "loss": 0.0335, + "step": 5344 + }, + { + "epoch": 2.5152941176470587, + "grad_norm": 0.11176045322875541, + "learning_rate": 0.00023385874450552522, + "loss": 0.0567, + "step": 5345 + }, + { + "epoch": 2.515764705882353, + "grad_norm": 0.023258312320425984, + "learning_rate": 0.00023382239510832262, + "loss": 0.0016, + "step": 5346 + }, + { + "epoch": 2.516235294117647, + "grad_norm": 0.05613558478105409, + "learning_rate": 0.0002337860385522763, + "loss": 0.0209, + "step": 5347 + }, + { + "epoch": 2.516705882352941, + "grad_norm": 0.02566934715144105, + "learning_rate": 0.00023374967484049132, + "loss": 0.0024, + "step": 5348 + }, + { + "epoch": 2.517176470588235, + "grad_norm": 0.00010538245486006573, + "learning_rate": 0.00023371330397607335, + "loss": 0.0, + "step": 5349 + }, + { + "epoch": 2.5176470588235293, + "grad_norm": 0.1109607345048576, + "learning_rate": 0.00023367692596212858, + "loss": 0.0453, + "step": 5350 + }, + { + "epoch": 2.5181176470588236, + "grad_norm": 0.14566327490078798, + "learning_rate": 0.0002336405408017639, + "loss": 0.0889, + "step": 5351 + }, + { + "epoch": 2.518588235294118, + "grad_norm": 0.022059746678712352, + "learning_rate": 0.0002336041484980868, + "loss": 0.003, + "step": 5352 + }, + { + "epoch": 2.519058823529412, + "grad_norm": 0.0929876400416149, + "learning_rate": 0.0002335677490542053, + "loss": 0.0475, + "step": 5353 + }, + { + "epoch": 2.519529411764706, + "grad_norm": 0.035455636742625654, + "learning_rate": 0.0002335313424732282, + "loss": 0.0085, + "step": 5354 + }, + { + "epoch": 2.52, + "grad_norm": 0.1367640623716187, + "learning_rate": 0.0002334949287582647, + "loss": 0.0445, + "step": 5355 + }, + { + "epoch": 2.5204705882352942, + "grad_norm": 6.816205600981446e-06, + "learning_rate": 0.00023345850791242477, + "loss": 0.0, + "step": 5356 + }, + { + "epoch": 2.520941176470588, + "grad_norm": 0.06533350962999769, + "learning_rate": 0.00023342207993881883, + "loss": 0.0232, + "step": 5357 + }, + { + "epoch": 2.5214117647058822, + "grad_norm": 0.10393696256006478, + "learning_rate": 0.0002333856448405581, + "loss": 0.0547, + "step": 5358 + }, + { + "epoch": 2.5218823529411765, + "grad_norm": 0.09361460418689711, + "learning_rate": 0.00023334920262075437, + "loss": 0.0443, + "step": 5359 + }, + { + "epoch": 2.5223529411764707, + "grad_norm": 0.08406431267568608, + "learning_rate": 0.00023331275328251983, + "loss": 0.0449, + "step": 5360 + }, + { + "epoch": 2.522823529411765, + "grad_norm": 0.0703363084693078, + "learning_rate": 0.00023327629682896758, + "loss": 0.0299, + "step": 5361 + }, + { + "epoch": 2.5232941176470587, + "grad_norm": 0.04611620765952731, + "learning_rate": 0.00023323983326321106, + "loss": 0.0168, + "step": 5362 + }, + { + "epoch": 2.523764705882353, + "grad_norm": 0.05485491072878553, + "learning_rate": 0.0002332033625883645, + "loss": 0.0215, + "step": 5363 + }, + { + "epoch": 2.524235294117647, + "grad_norm": 5.68547827553909e-07, + "learning_rate": 0.00023316688480754267, + "loss": 0.0, + "step": 5364 + }, + { + "epoch": 2.524705882352941, + "grad_norm": 0.10218383530302814, + "learning_rate": 0.00023313039992386094, + "loss": 0.0632, + "step": 5365 + }, + { + "epoch": 2.525176470588235, + "grad_norm": 2.2683464458483198e-07, + "learning_rate": 0.00023309390794043532, + "loss": 0.0, + "step": 5366 + }, + { + "epoch": 2.5256470588235294, + "grad_norm": 0.10040471656583465, + "learning_rate": 0.00023305740886038235, + "loss": 0.0723, + "step": 5367 + }, + { + "epoch": 2.5261176470588236, + "grad_norm": 0.07894941706586513, + "learning_rate": 0.00023302090268681932, + "loss": 0.0197, + "step": 5368 + }, + { + "epoch": 2.526588235294118, + "grad_norm": 1.5479317328409006e-06, + "learning_rate": 0.00023298438942286395, + "loss": 0.0, + "step": 5369 + }, + { + "epoch": 2.527058823529412, + "grad_norm": 0.057773607044069177, + "learning_rate": 0.00023294786907163473, + "loss": 0.0232, + "step": 5370 + }, + { + "epoch": 2.527529411764706, + "grad_norm": 0.07588480063298395, + "learning_rate": 0.0002329113416362506, + "loss": 0.0303, + "step": 5371 + }, + { + "epoch": 2.528, + "grad_norm": 0.1566477787973584, + "learning_rate": 0.00023287480711983125, + "loss": 0.0982, + "step": 5372 + }, + { + "epoch": 2.5284705882352942, + "grad_norm": 0.10306912415434537, + "learning_rate": 0.00023283826552549682, + "loss": 0.0487, + "step": 5373 + }, + { + "epoch": 2.528941176470588, + "grad_norm": 0.07381466168199195, + "learning_rate": 0.00023280171685636822, + "loss": 0.0257, + "step": 5374 + }, + { + "epoch": 2.5294117647058822, + "grad_norm": 0.1071509595726639, + "learning_rate": 0.0002327651611155669, + "loss": 0.0764, + "step": 5375 + }, + { + "epoch": 2.5298823529411765, + "grad_norm": 0.09037636733302178, + "learning_rate": 0.0002327285983062148, + "loss": 0.0604, + "step": 5376 + }, + { + "epoch": 2.5303529411764707, + "grad_norm": 0.1512821313160175, + "learning_rate": 0.00023269202843143462, + "loss": 0.0946, + "step": 5377 + }, + { + "epoch": 2.530823529411765, + "grad_norm": 0.0655227999916093, + "learning_rate": 0.00023265545149434968, + "loss": 0.0231, + "step": 5378 + }, + { + "epoch": 2.5312941176470587, + "grad_norm": 0.036816511582373646, + "learning_rate": 0.0002326188674980837, + "loss": 0.0133, + "step": 5379 + }, + { + "epoch": 2.531764705882353, + "grad_norm": 0.07362308571845588, + "learning_rate": 0.0002325822764457612, + "loss": 0.0387, + "step": 5380 + }, + { + "epoch": 2.532235294117647, + "grad_norm": 8.843522346067873e-06, + "learning_rate": 0.0002325456783405072, + "loss": 0.0, + "step": 5381 + }, + { + "epoch": 2.532705882352941, + "grad_norm": 0.10600834707606714, + "learning_rate": 0.00023250907318544744, + "loss": 0.074, + "step": 5382 + }, + { + "epoch": 2.533176470588235, + "grad_norm": 0.022534708226954046, + "learning_rate": 0.0002324724609837081, + "loss": 0.0054, + "step": 5383 + }, + { + "epoch": 2.5336470588235294, + "grad_norm": 0.12728622964181852, + "learning_rate": 0.0002324358417384161, + "loss": 0.0857, + "step": 5384 + }, + { + "epoch": 2.5341176470588236, + "grad_norm": 0.15748117258497876, + "learning_rate": 0.00023239921545269884, + "loss": 0.1083, + "step": 5385 + }, + { + "epoch": 2.534588235294118, + "grad_norm": 5.145725477896981e-05, + "learning_rate": 0.00023236258212968445, + "loss": 0.0, + "step": 5386 + }, + { + "epoch": 2.535058823529412, + "grad_norm": 7.699528216773875e-05, + "learning_rate": 0.00023232594177250148, + "loss": 0.0, + "step": 5387 + }, + { + "epoch": 2.535529411764706, + "grad_norm": 0.11588097480773984, + "learning_rate": 0.00023228929438427932, + "loss": 0.0719, + "step": 5388 + }, + { + "epoch": 2.536, + "grad_norm": 0.060102507185125637, + "learning_rate": 0.0002322526399681478, + "loss": 0.0219, + "step": 5389 + }, + { + "epoch": 2.5364705882352943, + "grad_norm": 0.09936126239930167, + "learning_rate": 0.0002322159785272374, + "loss": 0.0589, + "step": 5390 + }, + { + "epoch": 2.536941176470588, + "grad_norm": 0.11391906818671656, + "learning_rate": 0.00023217931006467912, + "loss": 0.0726, + "step": 5391 + }, + { + "epoch": 2.5374117647058823, + "grad_norm": 0.04642474212994804, + "learning_rate": 0.00023214263458360467, + "loss": 0.0161, + "step": 5392 + }, + { + "epoch": 2.5378823529411765, + "grad_norm": 0.11617097268792317, + "learning_rate": 0.00023210595208714635, + "loss": 0.0737, + "step": 5393 + }, + { + "epoch": 2.5383529411764707, + "grad_norm": 0.05197146693436621, + "learning_rate": 0.00023206926257843702, + "loss": 0.0174, + "step": 5394 + }, + { + "epoch": 2.538823529411765, + "grad_norm": 0.0422276128975452, + "learning_rate": 0.00023203256606061008, + "loss": 0.0201, + "step": 5395 + }, + { + "epoch": 2.5392941176470587, + "grad_norm": 6.029327100877868e-05, + "learning_rate": 0.00023199586253679964, + "loss": 0.0, + "step": 5396 + }, + { + "epoch": 2.539764705882353, + "grad_norm": 0.09002930976821794, + "learning_rate": 0.00023195915201014035, + "loss": 0.0261, + "step": 5397 + }, + { + "epoch": 2.540235294117647, + "grad_norm": 0.07444072816868873, + "learning_rate": 0.00023192243448376745, + "loss": 0.0369, + "step": 5398 + }, + { + "epoch": 2.540705882352941, + "grad_norm": 0.031558337578266614, + "learning_rate": 0.00023188570996081687, + "loss": 0.0077, + "step": 5399 + }, + { + "epoch": 2.541176470588235, + "grad_norm": 0.13523737898485785, + "learning_rate": 0.00023184897844442495, + "loss": 0.0961, + "step": 5400 + }, + { + "epoch": 2.5416470588235294, + "grad_norm": 0.11824308338091802, + "learning_rate": 0.00023181223993772886, + "loss": 0.0607, + "step": 5401 + }, + { + "epoch": 2.5421176470588236, + "grad_norm": 2.4829007620637655e-06, + "learning_rate": 0.00023177549444386618, + "loss": 0.0, + "step": 5402 + }, + { + "epoch": 2.542588235294118, + "grad_norm": 0.05490745821703842, + "learning_rate": 0.00023173874196597515, + "loss": 0.0271, + "step": 5403 + }, + { + "epoch": 2.543058823529412, + "grad_norm": 0.12716358076045894, + "learning_rate": 0.00023170198250719464, + "loss": 0.0848, + "step": 5404 + }, + { + "epoch": 2.543529411764706, + "grad_norm": 0.09523376220677324, + "learning_rate": 0.00023166521607066409, + "loss": 0.0452, + "step": 5405 + }, + { + "epoch": 2.544, + "grad_norm": 0.08334894232214898, + "learning_rate": 0.00023162844265952348, + "loss": 0.0507, + "step": 5406 + }, + { + "epoch": 2.5444705882352943, + "grad_norm": 0.11763167999437839, + "learning_rate": 0.0002315916622769135, + "loss": 0.1016, + "step": 5407 + }, + { + "epoch": 2.544941176470588, + "grad_norm": 0.1384042086419163, + "learning_rate": 0.00023155487492597536, + "loss": 0.0714, + "step": 5408 + }, + { + "epoch": 2.5454117647058823, + "grad_norm": 0.10252414472940213, + "learning_rate": 0.0002315180806098509, + "loss": 0.0606, + "step": 5409 + }, + { + "epoch": 2.5458823529411765, + "grad_norm": 5.894264363498984e-06, + "learning_rate": 0.00023148127933168245, + "loss": 0.0, + "step": 5410 + }, + { + "epoch": 2.5463529411764707, + "grad_norm": 0.029424811162712058, + "learning_rate": 0.00023144447109461314, + "loss": 0.0092, + "step": 5411 + }, + { + "epoch": 2.546823529411765, + "grad_norm": 0.1026764387275856, + "learning_rate": 0.0002314076559017865, + "loss": 0.0578, + "step": 5412 + }, + { + "epoch": 2.5472941176470587, + "grad_norm": 0.04760473151092981, + "learning_rate": 0.00023137083375634676, + "loss": 0.0165, + "step": 5413 + }, + { + "epoch": 2.547764705882353, + "grad_norm": 0.08931162244127715, + "learning_rate": 0.00023133400466143867, + "loss": 0.0636, + "step": 5414 + }, + { + "epoch": 2.548235294117647, + "grad_norm": 4.312705364339178e-06, + "learning_rate": 0.00023129716862020766, + "loss": 0.0, + "step": 5415 + }, + { + "epoch": 2.548705882352941, + "grad_norm": 0.12911425880504604, + "learning_rate": 0.0002312603256357997, + "loss": 0.092, + "step": 5416 + }, + { + "epoch": 2.549176470588235, + "grad_norm": 0.1494382973749052, + "learning_rate": 0.00023122347571136127, + "loss": 0.0639, + "step": 5417 + }, + { + "epoch": 2.5496470588235294, + "grad_norm": 0.12389072008358618, + "learning_rate": 0.00023118661885003969, + "loss": 0.0977, + "step": 5418 + }, + { + "epoch": 2.5501176470588236, + "grad_norm": 0.10843755938682728, + "learning_rate": 0.00023114975505498264, + "loss": 0.0662, + "step": 5419 + }, + { + "epoch": 2.550588235294118, + "grad_norm": 0.13311951546874684, + "learning_rate": 0.00023111288432933847, + "loss": 0.0872, + "step": 5420 + }, + { + "epoch": 2.5510588235294116, + "grad_norm": 0.04901555497421526, + "learning_rate": 0.0002310760066762561, + "loss": 0.0121, + "step": 5421 + }, + { + "epoch": 2.551529411764706, + "grad_norm": 0.07438036033981173, + "learning_rate": 0.00023103912209888512, + "loss": 0.0304, + "step": 5422 + }, + { + "epoch": 2.552, + "grad_norm": 9.773819959644653e-06, + "learning_rate": 0.00023100223060037558, + "loss": 0.0, + "step": 5423 + }, + { + "epoch": 2.5524705882352943, + "grad_norm": 0.33436420879222567, + "learning_rate": 0.00023096533218387827, + "loss": 0.2039, + "step": 5424 + }, + { + "epoch": 2.552941176470588, + "grad_norm": 0.05280597238982805, + "learning_rate": 0.00023092842685254442, + "loss": 0.0227, + "step": 5425 + }, + { + "epoch": 2.5534117647058823, + "grad_norm": 0.06805961467326792, + "learning_rate": 0.000230891514609526, + "loss": 0.0337, + "step": 5426 + }, + { + "epoch": 2.5538823529411765, + "grad_norm": 0.10609288430008286, + "learning_rate": 0.00023085459545797542, + "loss": 0.0538, + "step": 5427 + }, + { + "epoch": 2.5543529411764707, + "grad_norm": 0.12448209517286811, + "learning_rate": 0.00023081766940104585, + "loss": 0.0801, + "step": 5428 + }, + { + "epoch": 2.554823529411765, + "grad_norm": 0.08163794936738955, + "learning_rate": 0.00023078073644189085, + "loss": 0.0434, + "step": 5429 + }, + { + "epoch": 2.5552941176470587, + "grad_norm": 4.3042636577912095e-06, + "learning_rate": 0.00023074379658366478, + "loss": 0.0, + "step": 5430 + }, + { + "epoch": 2.555764705882353, + "grad_norm": 0.12376818345624625, + "learning_rate": 0.0002307068498295224, + "loss": 0.0838, + "step": 5431 + }, + { + "epoch": 2.556235294117647, + "grad_norm": 0.11625206945682176, + "learning_rate": 0.00023066989618261919, + "loss": 0.0743, + "step": 5432 + }, + { + "epoch": 2.556705882352941, + "grad_norm": 0.09968344232741769, + "learning_rate": 0.00023063293564611114, + "loss": 0.0757, + "step": 5433 + }, + { + "epoch": 2.557176470588235, + "grad_norm": 0.10921545832796417, + "learning_rate": 0.00023059596822315491, + "loss": 0.0619, + "step": 5434 + }, + { + "epoch": 2.5576470588235294, + "grad_norm": 0.11641017840985911, + "learning_rate": 0.00023055899391690766, + "loss": 0.0531, + "step": 5435 + }, + { + "epoch": 2.5581176470588236, + "grad_norm": 0.08453180267558096, + "learning_rate": 0.00023052201273052723, + "loss": 0.0468, + "step": 5436 + }, + { + "epoch": 2.558588235294118, + "grad_norm": 2.9278505106518294e-05, + "learning_rate": 0.00023048502466717192, + "loss": 0.0, + "step": 5437 + }, + { + "epoch": 2.5590588235294116, + "grad_norm": 0.08897877841690685, + "learning_rate": 0.00023044802973000073, + "loss": 0.0522, + "step": 5438 + }, + { + "epoch": 2.559529411764706, + "grad_norm": 0.10561605412695559, + "learning_rate": 0.00023041102792217323, + "loss": 0.0407, + "step": 5439 + }, + { + "epoch": 2.56, + "grad_norm": 0.0488274742259282, + "learning_rate": 0.00023037401924684946, + "loss": 0.0206, + "step": 5440 + }, + { + "epoch": 2.5604705882352943, + "grad_norm": 0.047806605245546946, + "learning_rate": 0.00023033700370719026, + "loss": 0.0243, + "step": 5441 + }, + { + "epoch": 2.560941176470588, + "grad_norm": 0.05587779072692818, + "learning_rate": 0.00023029998130635692, + "loss": 0.0171, + "step": 5442 + }, + { + "epoch": 2.5614117647058823, + "grad_norm": 0.3169875078771684, + "learning_rate": 0.00023026295204751126, + "loss": 0.1395, + "step": 5443 + }, + { + "epoch": 2.5618823529411765, + "grad_norm": 0.11558689717366545, + "learning_rate": 0.00023022591593381582, + "loss": 0.0796, + "step": 5444 + }, + { + "epoch": 2.5623529411764707, + "grad_norm": 0.04425085711847103, + "learning_rate": 0.0002301888729684337, + "loss": 0.0153, + "step": 5445 + }, + { + "epoch": 2.562823529411765, + "grad_norm": 0.0943006237198555, + "learning_rate": 0.00023015182315452843, + "loss": 0.041, + "step": 5446 + }, + { + "epoch": 2.5632941176470587, + "grad_norm": 0.10161409264360262, + "learning_rate": 0.00023011476649526438, + "loss": 0.0592, + "step": 5447 + }, + { + "epoch": 2.563764705882353, + "grad_norm": 0.07502656141408832, + "learning_rate": 0.0002300777029938063, + "loss": 0.0223, + "step": 5448 + }, + { + "epoch": 2.564235294117647, + "grad_norm": 0.08165532522148038, + "learning_rate": 0.0002300406326533196, + "loss": 0.0351, + "step": 5449 + }, + { + "epoch": 2.564705882352941, + "grad_norm": 0.08483689077351884, + "learning_rate": 0.00023000355547697027, + "loss": 0.0297, + "step": 5450 + }, + { + "epoch": 2.565176470588235, + "grad_norm": 0.09962010383420053, + "learning_rate": 0.00022996647146792488, + "loss": 0.0535, + "step": 5451 + }, + { + "epoch": 2.5656470588235294, + "grad_norm": 2.8205017660581236e-06, + "learning_rate": 0.0002299293806293506, + "loss": 0.0, + "step": 5452 + }, + { + "epoch": 2.5661176470588236, + "grad_norm": 4.783818017470563e-06, + "learning_rate": 0.00022989228296441515, + "loss": 0.0, + "step": 5453 + }, + { + "epoch": 2.566588235294118, + "grad_norm": 0.04395159526264666, + "learning_rate": 0.00022985517847628686, + "loss": 0.0079, + "step": 5454 + }, + { + "epoch": 2.5670588235294116, + "grad_norm": 0.3007603208757607, + "learning_rate": 0.0002298180671681347, + "loss": 0.0891, + "step": 5455 + }, + { + "epoch": 2.567529411764706, + "grad_norm": 0.06680079264790757, + "learning_rate": 0.00022978094904312803, + "loss": 0.0242, + "step": 5456 + }, + { + "epoch": 2.568, + "grad_norm": 0.08250742742318098, + "learning_rate": 0.000229743824104437, + "loss": 0.0356, + "step": 5457 + }, + { + "epoch": 2.5684705882352943, + "grad_norm": 6.4239140779084705e-06, + "learning_rate": 0.00022970669235523224, + "loss": 0.0, + "step": 5458 + }, + { + "epoch": 2.568941176470588, + "grad_norm": 3.523843061296428e-06, + "learning_rate": 0.00022966955379868502, + "loss": 0.0, + "step": 5459 + }, + { + "epoch": 2.5694117647058823, + "grad_norm": 0.10673077835360967, + "learning_rate": 0.00022963240843796714, + "loss": 0.0478, + "step": 5460 + }, + { + "epoch": 2.5698823529411765, + "grad_norm": 0.12083285780020299, + "learning_rate": 0.000229595256276251, + "loss": 0.0691, + "step": 5461 + }, + { + "epoch": 2.5703529411764707, + "grad_norm": 0.04715653049907431, + "learning_rate": 0.00022955809731670955, + "loss": 0.0191, + "step": 5462 + }, + { + "epoch": 2.570823529411765, + "grad_norm": 0.14037846145396984, + "learning_rate": 0.00022952093156251635, + "loss": 0.102, + "step": 5463 + }, + { + "epoch": 2.5712941176470587, + "grad_norm": 0.061388285428971665, + "learning_rate": 0.0002294837590168455, + "loss": 0.0364, + "step": 5464 + }, + { + "epoch": 2.571764705882353, + "grad_norm": 0.08427283206750376, + "learning_rate": 0.00022944657968287184, + "loss": 0.0401, + "step": 5465 + }, + { + "epoch": 2.572235294117647, + "grad_norm": 1.2959694098773485e-05, + "learning_rate": 0.00022940939356377052, + "loss": 0.0, + "step": 5466 + }, + { + "epoch": 2.572705882352941, + "grad_norm": 4.777730934468231e-06, + "learning_rate": 0.00022937220066271752, + "loss": 0.0, + "step": 5467 + }, + { + "epoch": 2.573176470588235, + "grad_norm": 2.971739201628852e-06, + "learning_rate": 0.0002293350009828893, + "loss": 0.0, + "step": 5468 + }, + { + "epoch": 2.5736470588235294, + "grad_norm": 0.0703315026604167, + "learning_rate": 0.00022929779452746283, + "loss": 0.03, + "step": 5469 + }, + { + "epoch": 2.5741176470588236, + "grad_norm": 4.784838877151e-06, + "learning_rate": 0.0002292605812996157, + "loss": 0.0, + "step": 5470 + }, + { + "epoch": 2.574588235294118, + "grad_norm": 0.0461876511822125, + "learning_rate": 0.00022922336130252622, + "loss": 0.0146, + "step": 5471 + }, + { + "epoch": 2.5750588235294116, + "grad_norm": 0.07233884537399185, + "learning_rate": 0.00022918613453937314, + "loss": 0.0319, + "step": 5472 + }, + { + "epoch": 2.575529411764706, + "grad_norm": 0.037600491344289534, + "learning_rate": 0.00022914890101333573, + "loss": 0.0077, + "step": 5473 + }, + { + "epoch": 2.576, + "grad_norm": 0.10970540152175279, + "learning_rate": 0.00022911166072759386, + "loss": 0.0682, + "step": 5474 + }, + { + "epoch": 2.576470588235294, + "grad_norm": 0.18652008549931418, + "learning_rate": 0.00022907441368532822, + "loss": 0.0715, + "step": 5475 + }, + { + "epoch": 2.576941176470588, + "grad_norm": 0.03917839474625426, + "learning_rate": 0.00022903715988971973, + "loss": 0.0099, + "step": 5476 + }, + { + "epoch": 2.5774117647058823, + "grad_norm": 0.15142872975780713, + "learning_rate": 0.0002289998993439502, + "loss": 0.1121, + "step": 5477 + }, + { + "epoch": 2.5778823529411765, + "grad_norm": 3.9858411566680994e-05, + "learning_rate": 0.00022896263205120167, + "loss": 0.0, + "step": 5478 + }, + { + "epoch": 2.5783529411764707, + "grad_norm": 0.06065024821696951, + "learning_rate": 0.0002289253580146571, + "loss": 0.0349, + "step": 5479 + }, + { + "epoch": 2.578823529411765, + "grad_norm": 0.07727225485716348, + "learning_rate": 0.00022888807723749988, + "loss": 0.0463, + "step": 5480 + }, + { + "epoch": 2.5792941176470587, + "grad_norm": 0.1374683789532204, + "learning_rate": 0.0002288507897229138, + "loss": 0.0189, + "step": 5481 + }, + { + "epoch": 2.579764705882353, + "grad_norm": 2.517505656175105e-05, + "learning_rate": 0.00022881349547408363, + "loss": 0.0, + "step": 5482 + }, + { + "epoch": 2.580235294117647, + "grad_norm": 0.1286836765313498, + "learning_rate": 0.00022877619449419436, + "loss": 0.0867, + "step": 5483 + }, + { + "epoch": 2.580705882352941, + "grad_norm": 0.0744736158102052, + "learning_rate": 0.00022873888678643164, + "loss": 0.0383, + "step": 5484 + }, + { + "epoch": 2.581176470588235, + "grad_norm": 0.07683873899789254, + "learning_rate": 0.0002287015723539818, + "loss": 0.0332, + "step": 5485 + }, + { + "epoch": 2.5816470588235294, + "grad_norm": 0.14230251474912278, + "learning_rate": 0.00022866425120003165, + "loss": 0.0753, + "step": 5486 + }, + { + "epoch": 2.5821176470588236, + "grad_norm": 0.061367423180433874, + "learning_rate": 0.0002286269233277686, + "loss": 0.0225, + "step": 5487 + }, + { + "epoch": 2.582588235294118, + "grad_norm": 0.09202975830921623, + "learning_rate": 0.0002285895887403806, + "loss": 0.0565, + "step": 5488 + }, + { + "epoch": 2.5830588235294116, + "grad_norm": 0.10093690643961162, + "learning_rate": 0.00022855224744105625, + "loss": 0.0473, + "step": 5489 + }, + { + "epoch": 2.583529411764706, + "grad_norm": 0.17546193262408544, + "learning_rate": 0.00022851489943298472, + "loss": 0.1273, + "step": 5490 + }, + { + "epoch": 2.584, + "grad_norm": 7.5948292952578866e-06, + "learning_rate": 0.0002284775447193556, + "loss": 0.0, + "step": 5491 + }, + { + "epoch": 2.584470588235294, + "grad_norm": 7.944193285386418e-06, + "learning_rate": 0.00022844018330335928, + "loss": 0.0, + "step": 5492 + }, + { + "epoch": 2.584941176470588, + "grad_norm": 0.3360393813169937, + "learning_rate": 0.00022840281518818657, + "loss": 0.2696, + "step": 5493 + }, + { + "epoch": 2.5854117647058823, + "grad_norm": 0.06464107233445952, + "learning_rate": 0.00022836544037702883, + "loss": 0.0349, + "step": 5494 + }, + { + "epoch": 2.5858823529411765, + "grad_norm": 0.01877578694408563, + "learning_rate": 0.00022832805887307815, + "loss": 0.0028, + "step": 5495 + }, + { + "epoch": 2.5863529411764707, + "grad_norm": 0.09571624544508262, + "learning_rate": 0.00022829067067952702, + "loss": 0.0509, + "step": 5496 + }, + { + "epoch": 2.586823529411765, + "grad_norm": 0.18075167023451752, + "learning_rate": 0.00022825327579956863, + "loss": 0.0963, + "step": 5497 + }, + { + "epoch": 2.5872941176470587, + "grad_norm": 0.09924635360282529, + "learning_rate": 0.0002282158742363966, + "loss": 0.0458, + "step": 5498 + }, + { + "epoch": 2.587764705882353, + "grad_norm": 0.07864837506824421, + "learning_rate": 0.00022817846599320528, + "loss": 0.0514, + "step": 5499 + }, + { + "epoch": 2.588235294117647, + "grad_norm": 0.08199508584088154, + "learning_rate": 0.00022814105107318952, + "loss": 0.0345, + "step": 5500 + }, + { + "epoch": 2.588705882352941, + "grad_norm": 0.04052585917579874, + "learning_rate": 0.00022810362947954475, + "loss": 0.0148, + "step": 5501 + }, + { + "epoch": 2.589176470588235, + "grad_norm": 0.10427449754083958, + "learning_rate": 0.00022806620121546684, + "loss": 0.0474, + "step": 5502 + }, + { + "epoch": 2.5896470588235294, + "grad_norm": 0.09075428644158529, + "learning_rate": 0.0002280287662841525, + "loss": 0.0422, + "step": 5503 + }, + { + "epoch": 2.5901176470588236, + "grad_norm": 0.05205917371276955, + "learning_rate": 0.00022799132468879878, + "loss": 0.0174, + "step": 5504 + }, + { + "epoch": 2.590588235294118, + "grad_norm": 0.035620638499585744, + "learning_rate": 0.00022795387643260336, + "loss": 0.0053, + "step": 5505 + }, + { + "epoch": 2.5910588235294116, + "grad_norm": 4.205047629522637e-06, + "learning_rate": 0.00022791642151876455, + "loss": 0.0, + "step": 5506 + }, + { + "epoch": 2.591529411764706, + "grad_norm": 0.06280332968553426, + "learning_rate": 0.00022787895995048118, + "loss": 0.0229, + "step": 5507 + }, + { + "epoch": 2.592, + "grad_norm": 0.09916592344645585, + "learning_rate": 0.00022784149173095262, + "loss": 0.0581, + "step": 5508 + }, + { + "epoch": 2.592470588235294, + "grad_norm": 0.07925486074820103, + "learning_rate": 0.00022780401686337885, + "loss": 0.0342, + "step": 5509 + }, + { + "epoch": 2.592941176470588, + "grad_norm": 0.10907373425276903, + "learning_rate": 0.00022776653535096044, + "loss": 0.0553, + "step": 5510 + }, + { + "epoch": 2.5934117647058823, + "grad_norm": 0.09628787991876402, + "learning_rate": 0.00022772904719689844, + "loss": 0.0502, + "step": 5511 + }, + { + "epoch": 2.5938823529411765, + "grad_norm": 0.10093116972806858, + "learning_rate": 0.00022769155240439455, + "loss": 0.0368, + "step": 5512 + }, + { + "epoch": 2.5943529411764708, + "grad_norm": 0.0889598804790976, + "learning_rate": 0.00022765405097665103, + "loss": 0.0369, + "step": 5513 + }, + { + "epoch": 2.594823529411765, + "grad_norm": 1.9173297009880576e-06, + "learning_rate": 0.00022761654291687065, + "loss": 0.0, + "step": 5514 + }, + { + "epoch": 2.5952941176470588, + "grad_norm": 0.13341306396393207, + "learning_rate": 0.00022757902822825683, + "loss": 0.0324, + "step": 5515 + }, + { + "epoch": 2.595764705882353, + "grad_norm": 0.034505381646609996, + "learning_rate": 0.00022754150691401345, + "loss": 0.0042, + "step": 5516 + }, + { + "epoch": 2.596235294117647, + "grad_norm": 0.06065219759083162, + "learning_rate": 0.00022750397897734504, + "loss": 0.0252, + "step": 5517 + }, + { + "epoch": 2.596705882352941, + "grad_norm": 0.12084915662831672, + "learning_rate": 0.00022746644442145665, + "loss": 0.0547, + "step": 5518 + }, + { + "epoch": 2.597176470588235, + "grad_norm": 0.09782538092911232, + "learning_rate": 0.00022742890324955396, + "loss": 0.0608, + "step": 5519 + }, + { + "epoch": 2.5976470588235294, + "grad_norm": 0.0961094446484866, + "learning_rate": 0.00022739135546484315, + "loss": 0.0452, + "step": 5520 + }, + { + "epoch": 2.5981176470588236, + "grad_norm": 0.07654789415687589, + "learning_rate": 0.00022735380107053094, + "loss": 0.0399, + "step": 5521 + }, + { + "epoch": 2.598588235294118, + "grad_norm": 0.07742791426735714, + "learning_rate": 0.00022731624006982473, + "loss": 0.0275, + "step": 5522 + }, + { + "epoch": 2.5990588235294116, + "grad_norm": 0.038042625473043934, + "learning_rate": 0.0002272786724659324, + "loss": 0.0103, + "step": 5523 + }, + { + "epoch": 2.599529411764706, + "grad_norm": 0.0782529404835084, + "learning_rate": 0.00022724109826206235, + "loss": 0.0414, + "step": 5524 + }, + { + "epoch": 2.6, + "grad_norm": 0.13768890159619246, + "learning_rate": 0.00022720351746142366, + "loss": 0.057, + "step": 5525 + }, + { + "epoch": 2.600470588235294, + "grad_norm": 0.02413778076418392, + "learning_rate": 0.00022716593006722594, + "loss": 0.0039, + "step": 5526 + }, + { + "epoch": 2.600941176470588, + "grad_norm": 0.13299742759938715, + "learning_rate": 0.0002271283360826793, + "loss": 0.1246, + "step": 5527 + }, + { + "epoch": 2.6014117647058823, + "grad_norm": 0.10828251465444297, + "learning_rate": 0.0002270907355109944, + "loss": 0.0388, + "step": 5528 + }, + { + "epoch": 2.6018823529411765, + "grad_norm": 0.11580147093389788, + "learning_rate": 0.00022705312835538257, + "loss": 0.0787, + "step": 5529 + }, + { + "epoch": 2.6023529411764708, + "grad_norm": 0.13399841407764193, + "learning_rate": 0.00022701551461905563, + "loss": 0.0667, + "step": 5530 + }, + { + "epoch": 2.602823529411765, + "grad_norm": 0.03971112821394866, + "learning_rate": 0.00022697789430522602, + "loss": 0.0091, + "step": 5531 + }, + { + "epoch": 2.6032941176470588, + "grad_norm": 0.06785057734404469, + "learning_rate": 0.00022694026741710657, + "loss": 0.0326, + "step": 5532 + }, + { + "epoch": 2.603764705882353, + "grad_norm": 0.08938392667705443, + "learning_rate": 0.00022690263395791095, + "loss": 0.0377, + "step": 5533 + }, + { + "epoch": 2.604235294117647, + "grad_norm": 0.08891334641605428, + "learning_rate": 0.00022686499393085322, + "loss": 0.0551, + "step": 5534 + }, + { + "epoch": 2.604705882352941, + "grad_norm": 7.014237847769551e-06, + "learning_rate": 0.00022682734733914792, + "loss": 0.0, + "step": 5535 + }, + { + "epoch": 2.605176470588235, + "grad_norm": 5.774555679877671e-05, + "learning_rate": 0.00022678969418601038, + "loss": 0.0, + "step": 5536 + }, + { + "epoch": 2.6056470588235294, + "grad_norm": 7.069081538188478e-06, + "learning_rate": 0.0002267520344746563, + "loss": 0.0, + "step": 5537 + }, + { + "epoch": 2.6061176470588236, + "grad_norm": 1.1713941295845053e-05, + "learning_rate": 0.000226714368208302, + "loss": 0.0, + "step": 5538 + }, + { + "epoch": 2.606588235294118, + "grad_norm": 0.13913381172007883, + "learning_rate": 0.00022667669539016435, + "loss": 0.092, + "step": 5539 + }, + { + "epoch": 2.6070588235294117, + "grad_norm": 0.08242007014854473, + "learning_rate": 0.00022663901602346082, + "loss": 0.0394, + "step": 5540 + }, + { + "epoch": 2.607529411764706, + "grad_norm": 0.09648853537996843, + "learning_rate": 0.00022660133011140942, + "loss": 0.0392, + "step": 5541 + }, + { + "epoch": 2.608, + "grad_norm": 0.08685414955860825, + "learning_rate": 0.00022656363765722872, + "loss": 0.0343, + "step": 5542 + }, + { + "epoch": 2.608470588235294, + "grad_norm": 0.054315459336860744, + "learning_rate": 0.00022652593866413775, + "loss": 0.0188, + "step": 5543 + }, + { + "epoch": 2.608941176470588, + "grad_norm": 0.21659034823459788, + "learning_rate": 0.00022648823313535635, + "loss": 0.0889, + "step": 5544 + }, + { + "epoch": 2.6094117647058823, + "grad_norm": 0.09413640134284415, + "learning_rate": 0.00022645052107410463, + "loss": 0.0454, + "step": 5545 + }, + { + "epoch": 2.6098823529411765, + "grad_norm": 0.0601970301182321, + "learning_rate": 0.0002264128024836034, + "loss": 0.024, + "step": 5546 + }, + { + "epoch": 2.6103529411764708, + "grad_norm": 5.7095542738913615e-06, + "learning_rate": 0.00022637507736707408, + "loss": 0.0, + "step": 5547 + }, + { + "epoch": 2.6108235294117645, + "grad_norm": 0.09821109755371428, + "learning_rate": 0.00022633734572773852, + "loss": 0.0591, + "step": 5548 + }, + { + "epoch": 2.6112941176470588, + "grad_norm": 0.11443010405790717, + "learning_rate": 0.00022629960756881916, + "loss": 0.0634, + "step": 5549 + }, + { + "epoch": 2.611764705882353, + "grad_norm": 0.05549734706872563, + "learning_rate": 0.00022626186289353913, + "loss": 0.0192, + "step": 5550 + }, + { + "epoch": 2.612235294117647, + "grad_norm": 0.06471286740439515, + "learning_rate": 0.00022622411170512193, + "loss": 0.0272, + "step": 5551 + }, + { + "epoch": 2.612705882352941, + "grad_norm": 0.05432567274970421, + "learning_rate": 0.0002261863540067917, + "loss": 0.0234, + "step": 5552 + }, + { + "epoch": 2.613176470588235, + "grad_norm": 0.1141862059972174, + "learning_rate": 0.00022614858980177314, + "loss": 0.0729, + "step": 5553 + }, + { + "epoch": 2.6136470588235294, + "grad_norm": 0.08174272431711346, + "learning_rate": 0.00022611081909329153, + "loss": 0.0317, + "step": 5554 + }, + { + "epoch": 2.6141176470588237, + "grad_norm": 0.14139072648813117, + "learning_rate": 0.00022607304188457264, + "loss": 0.0904, + "step": 5555 + }, + { + "epoch": 2.614588235294118, + "grad_norm": 0.07598984213621657, + "learning_rate": 0.0002260352581788428, + "loss": 0.0413, + "step": 5556 + }, + { + "epoch": 2.6150588235294117, + "grad_norm": 7.76422107416134e-06, + "learning_rate": 0.000225997467979329, + "loss": 0.0, + "step": 5557 + }, + { + "epoch": 2.615529411764706, + "grad_norm": 0.02914385739894694, + "learning_rate": 0.00022595967128925866, + "loss": 0.0056, + "step": 5558 + }, + { + "epoch": 2.616, + "grad_norm": 0.07335174539475377, + "learning_rate": 0.0002259218681118598, + "loss": 0.0203, + "step": 5559 + }, + { + "epoch": 2.616470588235294, + "grad_norm": 0.17605365391981534, + "learning_rate": 0.000225884058450361, + "loss": 0.1084, + "step": 5560 + }, + { + "epoch": 2.616941176470588, + "grad_norm": 2.7981058263791782e-06, + "learning_rate": 0.00022584624230799144, + "loss": 0.0, + "step": 5561 + }, + { + "epoch": 2.6174117647058823, + "grad_norm": 0.08451749515752591, + "learning_rate": 0.00022580841968798072, + "loss": 0.0378, + "step": 5562 + }, + { + "epoch": 2.6178823529411765, + "grad_norm": 1.8114167764784013e-06, + "learning_rate": 0.00022577059059355914, + "loss": 0.0, + "step": 5563 + }, + { + "epoch": 2.6183529411764708, + "grad_norm": 0.07355663245395727, + "learning_rate": 0.00022573275502795747, + "loss": 0.0306, + "step": 5564 + }, + { + "epoch": 2.6188235294117646, + "grad_norm": 0.021789676813565213, + "learning_rate": 0.00022569491299440704, + "loss": 0.0045, + "step": 5565 + }, + { + "epoch": 2.6192941176470588, + "grad_norm": 0.015793676023945716, + "learning_rate": 0.00022565706449613978, + "loss": 0.0022, + "step": 5566 + }, + { + "epoch": 2.619764705882353, + "grad_norm": 0.1629367848021877, + "learning_rate": 0.0002256192095363881, + "loss": 0.0644, + "step": 5567 + }, + { + "epoch": 2.620235294117647, + "grad_norm": 0.34569855260106697, + "learning_rate": 0.00022558134811838504, + "loss": 0.0832, + "step": 5568 + }, + { + "epoch": 2.620705882352941, + "grad_norm": 0.05371261158652827, + "learning_rate": 0.00022554348024536413, + "loss": 0.0183, + "step": 5569 + }, + { + "epoch": 2.621176470588235, + "grad_norm": 0.14205755011073654, + "learning_rate": 0.00022550560592055942, + "loss": 0.0786, + "step": 5570 + }, + { + "epoch": 2.6216470588235294, + "grad_norm": 0.10558231975596427, + "learning_rate": 0.00022546772514720566, + "loss": 0.041, + "step": 5571 + }, + { + "epoch": 2.6221176470588237, + "grad_norm": 2.4899120088337384e-06, + "learning_rate": 0.00022542983792853797, + "loss": 0.0, + "step": 5572 + }, + { + "epoch": 2.622588235294118, + "grad_norm": 0.07696385374073197, + "learning_rate": 0.00022539194426779216, + "loss": 0.0257, + "step": 5573 + }, + { + "epoch": 2.6230588235294117, + "grad_norm": 0.09986011010083186, + "learning_rate": 0.00022535404416820453, + "loss": 0.0378, + "step": 5574 + }, + { + "epoch": 2.623529411764706, + "grad_norm": 0.11931235396896783, + "learning_rate": 0.0002253161376330119, + "loss": 0.0542, + "step": 5575 + }, + { + "epoch": 2.624, + "grad_norm": 0.032623602488935255, + "learning_rate": 0.00022527822466545164, + "loss": 0.0069, + "step": 5576 + }, + { + "epoch": 2.624470588235294, + "grad_norm": 0.04095050351452975, + "learning_rate": 0.00022524030526876183, + "loss": 0.0088, + "step": 5577 + }, + { + "epoch": 2.624941176470588, + "grad_norm": 0.13591572764790147, + "learning_rate": 0.0002252023794461809, + "loss": 0.0667, + "step": 5578 + }, + { + "epoch": 2.6254117647058823, + "grad_norm": 0.1120298473611248, + "learning_rate": 0.00022516444720094785, + "loss": 0.0639, + "step": 5579 + }, + { + "epoch": 2.6258823529411766, + "grad_norm": 0.08142611305167315, + "learning_rate": 0.00022512650853630234, + "loss": 0.032, + "step": 5580 + }, + { + "epoch": 2.626352941176471, + "grad_norm": 0.09903580700293568, + "learning_rate": 0.00022508856345548455, + "loss": 0.0512, + "step": 5581 + }, + { + "epoch": 2.6268235294117646, + "grad_norm": 5.619859501044087e-05, + "learning_rate": 0.00022505061196173513, + "loss": 0.0, + "step": 5582 + }, + { + "epoch": 2.627294117647059, + "grad_norm": 0.10226064995064953, + "learning_rate": 0.00022501265405829526, + "loss": 0.057, + "step": 5583 + }, + { + "epoch": 2.627764705882353, + "grad_norm": 0.08420819733741944, + "learning_rate": 0.00022497468974840687, + "loss": 0.0514, + "step": 5584 + }, + { + "epoch": 2.6282352941176472, + "grad_norm": 0.04758342400374433, + "learning_rate": 0.00022493671903531218, + "loss": 0.0136, + "step": 5585 + }, + { + "epoch": 2.628705882352941, + "grad_norm": 0.07867857083970241, + "learning_rate": 0.00022489874192225414, + "loss": 0.0345, + "step": 5586 + }, + { + "epoch": 2.6291764705882352, + "grad_norm": 0.12053303336134406, + "learning_rate": 0.00022486075841247617, + "loss": 0.0899, + "step": 5587 + }, + { + "epoch": 2.6296470588235294, + "grad_norm": 0.05189380146823858, + "learning_rate": 0.0002248227685092222, + "loss": 0.0217, + "step": 5588 + }, + { + "epoch": 2.6301176470588237, + "grad_norm": 0.08337575069678133, + "learning_rate": 0.00022478477221573687, + "loss": 0.0465, + "step": 5589 + }, + { + "epoch": 2.630588235294118, + "grad_norm": 0.011962394186774134, + "learning_rate": 0.00022474676953526508, + "loss": 0.0014, + "step": 5590 + }, + { + "epoch": 2.6310588235294117, + "grad_norm": 1.1890112650113216e-05, + "learning_rate": 0.00022470876047105262, + "loss": 0.0, + "step": 5591 + }, + { + "epoch": 2.631529411764706, + "grad_norm": 0.20581088083645957, + "learning_rate": 0.00022467074502634558, + "loss": 0.1216, + "step": 5592 + }, + { + "epoch": 2.632, + "grad_norm": 0.11416310791793155, + "learning_rate": 0.00022463272320439063, + "loss": 0.0673, + "step": 5593 + }, + { + "epoch": 2.632470588235294, + "grad_norm": 0.16178531150632014, + "learning_rate": 0.00022459469500843502, + "loss": 0.103, + "step": 5594 + }, + { + "epoch": 2.632941176470588, + "grad_norm": 0.024270396268433473, + "learning_rate": 0.0002245566604417266, + "loss": 0.0056, + "step": 5595 + }, + { + "epoch": 2.6334117647058823, + "grad_norm": 0.18559848441852606, + "learning_rate": 0.0002245186195075137, + "loss": 0.1366, + "step": 5596 + }, + { + "epoch": 2.6338823529411766, + "grad_norm": 0.13983990315937042, + "learning_rate": 0.0002244805722090451, + "loss": 0.0828, + "step": 5597 + }, + { + "epoch": 2.634352941176471, + "grad_norm": 0.08054490429496425, + "learning_rate": 0.0002244425185495704, + "loss": 0.0446, + "step": 5598 + }, + { + "epoch": 2.6348235294117646, + "grad_norm": 0.14701057089092032, + "learning_rate": 0.00022440445853233945, + "loss": 0.0863, + "step": 5599 + }, + { + "epoch": 2.635294117647059, + "grad_norm": 0.12434940735846504, + "learning_rate": 0.00022436639216060275, + "loss": 0.0645, + "step": 5600 + }, + { + "epoch": 2.635764705882353, + "grad_norm": 0.03934287965712306, + "learning_rate": 0.00022432831943761146, + "loss": 0.0132, + "step": 5601 + }, + { + "epoch": 2.636235294117647, + "grad_norm": 0.15794278314374616, + "learning_rate": 0.00022429024036661708, + "loss": 0.0559, + "step": 5602 + }, + { + "epoch": 2.636705882352941, + "grad_norm": 7.923671295944696e-05, + "learning_rate": 0.00022425215495087177, + "loss": 0.0, + "step": 5603 + }, + { + "epoch": 2.6371764705882352, + "grad_norm": 0.10502069405375032, + "learning_rate": 0.00022421406319362823, + "loss": 0.0275, + "step": 5604 + }, + { + "epoch": 2.6376470588235295, + "grad_norm": 0.026821789838143432, + "learning_rate": 0.00022417596509813968, + "loss": 0.0049, + "step": 5605 + }, + { + "epoch": 2.6381176470588237, + "grad_norm": 0.029296735950000232, + "learning_rate": 0.0002241378606676599, + "loss": 0.0067, + "step": 5606 + }, + { + "epoch": 2.638588235294118, + "grad_norm": 1.100989210554283e-05, + "learning_rate": 0.00022409974990544312, + "loss": 0.0, + "step": 5607 + }, + { + "epoch": 2.6390588235294117, + "grad_norm": 7.507002191311472e-06, + "learning_rate": 0.0002240616328147443, + "loss": 0.0, + "step": 5608 + }, + { + "epoch": 2.639529411764706, + "grad_norm": 0.08418724548111164, + "learning_rate": 0.00022402350939881873, + "loss": 0.0343, + "step": 5609 + }, + { + "epoch": 2.64, + "grad_norm": 0.08731816227980363, + "learning_rate": 0.00022398537966092236, + "loss": 0.0509, + "step": 5610 + }, + { + "epoch": 2.640470588235294, + "grad_norm": 0.13047720140321628, + "learning_rate": 0.00022394724360431167, + "loss": 0.0676, + "step": 5611 + }, + { + "epoch": 2.640941176470588, + "grad_norm": 0.10260095608845546, + "learning_rate": 0.00022390910123224373, + "loss": 0.0416, + "step": 5612 + }, + { + "epoch": 2.6414117647058823, + "grad_norm": 7.96805359877072e-05, + "learning_rate": 0.00022387095254797597, + "loss": 0.0, + "step": 5613 + }, + { + "epoch": 2.6418823529411766, + "grad_norm": 0.032277415462382636, + "learning_rate": 0.00022383279755476651, + "loss": 0.0045, + "step": 5614 + }, + { + "epoch": 2.642352941176471, + "grad_norm": 0.0661052632980214, + "learning_rate": 0.00022379463625587406, + "loss": 0.0383, + "step": 5615 + }, + { + "epoch": 2.6428235294117646, + "grad_norm": 0.1173867028610432, + "learning_rate": 0.00022375646865455766, + "loss": 0.0344, + "step": 5616 + }, + { + "epoch": 2.643294117647059, + "grad_norm": 0.03492460519582077, + "learning_rate": 0.00022371829475407713, + "loss": 0.0082, + "step": 5617 + }, + { + "epoch": 2.643764705882353, + "grad_norm": 0.12200785696361631, + "learning_rate": 0.00022368011455769251, + "loss": 0.0755, + "step": 5618 + }, + { + "epoch": 2.644235294117647, + "grad_norm": 0.08490750864172991, + "learning_rate": 0.00022364192806866481, + "loss": 0.0087, + "step": 5619 + }, + { + "epoch": 2.644705882352941, + "grad_norm": 0.10290993399931865, + "learning_rate": 0.00022360373529025523, + "loss": 0.048, + "step": 5620 + }, + { + "epoch": 2.6451764705882352, + "grad_norm": 0.08825462950803094, + "learning_rate": 0.00022356553622572564, + "loss": 0.0508, + "step": 5621 + }, + { + "epoch": 2.6456470588235295, + "grad_norm": 6.1398732057703625e-06, + "learning_rate": 0.00022352733087833838, + "loss": 0.0, + "step": 5622 + }, + { + "epoch": 2.6461176470588237, + "grad_norm": 0.05237385913787872, + "learning_rate": 0.00022348911925135646, + "loss": 0.0309, + "step": 5623 + }, + { + "epoch": 2.646588235294118, + "grad_norm": 6.079988064570868e-06, + "learning_rate": 0.00022345090134804325, + "loss": 0.0, + "step": 5624 + }, + { + "epoch": 2.6470588235294117, + "grad_norm": 0.09282769175621344, + "learning_rate": 0.0002234126771716628, + "loss": 0.0388, + "step": 5625 + }, + { + "epoch": 2.647529411764706, + "grad_norm": 0.37260658068373254, + "learning_rate": 0.00022337444672547967, + "loss": 0.0995, + "step": 5626 + }, + { + "epoch": 2.648, + "grad_norm": 0.08958212609476358, + "learning_rate": 0.00022333621001275884, + "loss": 0.0622, + "step": 5627 + }, + { + "epoch": 2.648470588235294, + "grad_norm": 0.0975527933216584, + "learning_rate": 0.00022329796703676597, + "loss": 0.0666, + "step": 5628 + }, + { + "epoch": 2.648941176470588, + "grad_norm": 0.12022759216477226, + "learning_rate": 0.00022325971780076721, + "loss": 0.068, + "step": 5629 + }, + { + "epoch": 2.6494117647058824, + "grad_norm": 1.005356018108064e-05, + "learning_rate": 0.00022322146230802924, + "loss": 0.0, + "step": 5630 + }, + { + "epoch": 2.6498823529411766, + "grad_norm": 0.09445869129339243, + "learning_rate": 0.0002231832005618192, + "loss": 0.0779, + "step": 5631 + }, + { + "epoch": 2.650352941176471, + "grad_norm": 1.3613072060480346e-05, + "learning_rate": 0.00022314493256540485, + "loss": 0.0, + "step": 5632 + }, + { + "epoch": 2.6508235294117646, + "grad_norm": 0.06847204510501041, + "learning_rate": 0.0002231066583220545, + "loss": 0.0423, + "step": 5633 + }, + { + "epoch": 2.651294117647059, + "grad_norm": 0.07812438076624677, + "learning_rate": 0.00022306837783503698, + "loss": 0.0299, + "step": 5634 + }, + { + "epoch": 2.651764705882353, + "grad_norm": 0.05188241206538186, + "learning_rate": 0.00022303009110762154, + "loss": 0.0177, + "step": 5635 + }, + { + "epoch": 2.652235294117647, + "grad_norm": 0.0008032522538012305, + "learning_rate": 0.00022299179814307818, + "loss": 0.0, + "step": 5636 + }, + { + "epoch": 2.652705882352941, + "grad_norm": 0.040126785122449306, + "learning_rate": 0.0002229534989446772, + "loss": 0.0101, + "step": 5637 + }, + { + "epoch": 2.6531764705882352, + "grad_norm": 0.11011966268677299, + "learning_rate": 0.00022291519351568955, + "loss": 0.0722, + "step": 5638 + }, + { + "epoch": 2.6536470588235295, + "grad_norm": 5.09651114810174e-06, + "learning_rate": 0.00022287688185938673, + "loss": 0.0, + "step": 5639 + }, + { + "epoch": 2.6541176470588237, + "grad_norm": 0.12498163836382463, + "learning_rate": 0.00022283856397904077, + "loss": 0.1052, + "step": 5640 + }, + { + "epoch": 2.654588235294118, + "grad_norm": 0.0674483205245908, + "learning_rate": 0.00022280023987792416, + "loss": 0.0304, + "step": 5641 + }, + { + "epoch": 2.6550588235294117, + "grad_norm": 0.11382629101984151, + "learning_rate": 0.00022276190955930998, + "loss": 0.0687, + "step": 5642 + }, + { + "epoch": 2.655529411764706, + "grad_norm": 0.1457704794282687, + "learning_rate": 0.00022272357302647187, + "loss": 0.0982, + "step": 5643 + }, + { + "epoch": 2.656, + "grad_norm": 5.555142023725386e-05, + "learning_rate": 0.00022268523028268385, + "loss": 0.0, + "step": 5644 + }, + { + "epoch": 2.656470588235294, + "grad_norm": 0.054906998848161505, + "learning_rate": 0.00022264688133122069, + "loss": 0.0205, + "step": 5645 + }, + { + "epoch": 2.656941176470588, + "grad_norm": 0.13580754293575661, + "learning_rate": 0.00022260852617535752, + "loss": 0.0893, + "step": 5646 + }, + { + "epoch": 2.6574117647058824, + "grad_norm": 0.09647419113622414, + "learning_rate": 0.00022257016481837006, + "loss": 0.0639, + "step": 5647 + }, + { + "epoch": 2.6578823529411766, + "grad_norm": 0.09984777284353641, + "learning_rate": 0.00022253179726353458, + "loss": 0.0735, + "step": 5648 + }, + { + "epoch": 2.658352941176471, + "grad_norm": 0.11285415550172416, + "learning_rate": 0.00022249342351412783, + "loss": 0.0628, + "step": 5649 + }, + { + "epoch": 2.6588235294117646, + "grad_norm": 0.03548905266231542, + "learning_rate": 0.00022245504357342716, + "loss": 0.0081, + "step": 5650 + }, + { + "epoch": 2.659294117647059, + "grad_norm": 0.04143481958573815, + "learning_rate": 0.00022241665744471037, + "loss": 0.0124, + "step": 5651 + }, + { + "epoch": 2.659764705882353, + "grad_norm": 0.06195284990594894, + "learning_rate": 0.00022237826513125584, + "loss": 0.0314, + "step": 5652 + }, + { + "epoch": 2.660235294117647, + "grad_norm": 0.07363331825249714, + "learning_rate": 0.00022233986663634241, + "loss": 0.0483, + "step": 5653 + }, + { + "epoch": 2.660705882352941, + "grad_norm": 0.08156244351775184, + "learning_rate": 0.0002223014619632496, + "loss": 0.0313, + "step": 5654 + }, + { + "epoch": 2.6611764705882353, + "grad_norm": 7.441363101920684e-05, + "learning_rate": 0.00022226305111525726, + "loss": 0.0, + "step": 5655 + }, + { + "epoch": 2.6616470588235295, + "grad_norm": 0.11907401291040216, + "learning_rate": 0.00022222463409564595, + "loss": 0.0776, + "step": 5656 + }, + { + "epoch": 2.6621176470588237, + "grad_norm": 0.13488786957769971, + "learning_rate": 0.0002221862109076966, + "loss": 0.1203, + "step": 5657 + }, + { + "epoch": 2.662588235294118, + "grad_norm": 0.049701722428858584, + "learning_rate": 0.00022214778155469077, + "loss": 0.0153, + "step": 5658 + }, + { + "epoch": 2.6630588235294117, + "grad_norm": 0.07916682161662365, + "learning_rate": 0.0002221093460399105, + "loss": 0.0303, + "step": 5659 + }, + { + "epoch": 2.663529411764706, + "grad_norm": 0.07831924862891937, + "learning_rate": 0.0002220709043666384, + "loss": 0.062, + "step": 5660 + }, + { + "epoch": 2.664, + "grad_norm": 0.029728880085780292, + "learning_rate": 0.0002220324565381576, + "loss": 0.0043, + "step": 5661 + }, + { + "epoch": 2.664470588235294, + "grad_norm": 0.08666715738200124, + "learning_rate": 0.0002219940025577516, + "loss": 0.0468, + "step": 5662 + }, + { + "epoch": 2.664941176470588, + "grad_norm": 0.032350848128447164, + "learning_rate": 0.00022195554242870473, + "loss": 0.0043, + "step": 5663 + }, + { + "epoch": 2.6654117647058824, + "grad_norm": 0.15739831297334694, + "learning_rate": 0.0002219170761543016, + "loss": 0.118, + "step": 5664 + }, + { + "epoch": 2.6658823529411766, + "grad_norm": 0.0620612243231823, + "learning_rate": 0.00022187860373782737, + "loss": 0.0331, + "step": 5665 + }, + { + "epoch": 2.666352941176471, + "grad_norm": 0.1184189173542629, + "learning_rate": 0.00022184012518256785, + "loss": 0.0965, + "step": 5666 + }, + { + "epoch": 2.6668235294117646, + "grad_norm": 0.05834711603627988, + "learning_rate": 0.00022180164049180926, + "loss": 0.0211, + "step": 5667 + }, + { + "epoch": 2.667294117647059, + "grad_norm": 0.07910521502312004, + "learning_rate": 0.00022176314966883837, + "loss": 0.0352, + "step": 5668 + }, + { + "epoch": 2.667764705882353, + "grad_norm": 0.06191399101333882, + "learning_rate": 0.0002217246527169425, + "loss": 0.0196, + "step": 5669 + }, + { + "epoch": 2.668235294117647, + "grad_norm": 0.050351917456128964, + "learning_rate": 0.00022168614963940952, + "loss": 0.0142, + "step": 5670 + }, + { + "epoch": 2.668705882352941, + "grad_norm": 0.09374838445372077, + "learning_rate": 0.00022164764043952772, + "loss": 0.0598, + "step": 5671 + }, + { + "epoch": 2.6691764705882353, + "grad_norm": 0.08266911569605939, + "learning_rate": 0.00022160912512058598, + "loss": 0.0405, + "step": 5672 + }, + { + "epoch": 2.6696470588235295, + "grad_norm": 1.738043987152907e-05, + "learning_rate": 0.0002215706036858737, + "loss": 0.0, + "step": 5673 + }, + { + "epoch": 2.6701176470588237, + "grad_norm": 0.0001777200936098311, + "learning_rate": 0.00022153207613868087, + "loss": 0.0, + "step": 5674 + }, + { + "epoch": 2.6705882352941175, + "grad_norm": 0.0738134997955879, + "learning_rate": 0.00022149354248229784, + "loss": 0.033, + "step": 5675 + }, + { + "epoch": 2.6710588235294117, + "grad_norm": 0.03112620878967297, + "learning_rate": 0.00022145500272001558, + "loss": 0.0041, + "step": 5676 + }, + { + "epoch": 2.671529411764706, + "grad_norm": 0.04836779725415039, + "learning_rate": 0.00022141645685512564, + "loss": 0.018, + "step": 5677 + }, + { + "epoch": 2.672, + "grad_norm": 0.11831142226575787, + "learning_rate": 0.00022137790489091999, + "loss": 0.0934, + "step": 5678 + }, + { + "epoch": 2.672470588235294, + "grad_norm": 0.07839776962128224, + "learning_rate": 0.00022133934683069115, + "loss": 0.0267, + "step": 5679 + }, + { + "epoch": 2.672941176470588, + "grad_norm": 0.032466562668570086, + "learning_rate": 0.00022130078267773218, + "loss": 0.0077, + "step": 5680 + }, + { + "epoch": 2.6734117647058824, + "grad_norm": 0.11625031874986198, + "learning_rate": 0.00022126221243533664, + "loss": 0.0586, + "step": 5681 + }, + { + "epoch": 2.6738823529411766, + "grad_norm": 0.09822160463785695, + "learning_rate": 0.00022122363610679863, + "loss": 0.0603, + "step": 5682 + }, + { + "epoch": 2.674352941176471, + "grad_norm": 0.016875055771443923, + "learning_rate": 0.0002211850536954127, + "loss": 0.0026, + "step": 5683 + }, + { + "epoch": 2.6748235294117646, + "grad_norm": 0.050475328130521614, + "learning_rate": 0.00022114646520447406, + "loss": 0.0109, + "step": 5684 + }, + { + "epoch": 2.675294117647059, + "grad_norm": 0.0807702806439252, + "learning_rate": 0.00022110787063727832, + "loss": 0.0365, + "step": 5685 + }, + { + "epoch": 2.675764705882353, + "grad_norm": 0.06831633584092604, + "learning_rate": 0.00022106926999712165, + "loss": 0.0244, + "step": 5686 + }, + { + "epoch": 2.676235294117647, + "grad_norm": 0.143664928515055, + "learning_rate": 0.0002210306632873007, + "loss": 0.104, + "step": 5687 + }, + { + "epoch": 2.676705882352941, + "grad_norm": 0.07098019957365968, + "learning_rate": 0.00022099205051111272, + "loss": 0.0276, + "step": 5688 + }, + { + "epoch": 2.6771764705882353, + "grad_norm": 0.07735081202937615, + "learning_rate": 0.0002209534316718554, + "loss": 0.0339, + "step": 5689 + }, + { + "epoch": 2.6776470588235295, + "grad_norm": 0.08864681572171518, + "learning_rate": 0.000220914806772827, + "loss": 0.0424, + "step": 5690 + }, + { + "epoch": 2.6781176470588237, + "grad_norm": 0.07726066667617161, + "learning_rate": 0.0002208761758173263, + "loss": 0.0218, + "step": 5691 + }, + { + "epoch": 2.6785882352941175, + "grad_norm": 0.08707029463135416, + "learning_rate": 0.0002208375388086525, + "loss": 0.0566, + "step": 5692 + }, + { + "epoch": 2.6790588235294117, + "grad_norm": 0.06209014350096241, + "learning_rate": 0.00022079889575010543, + "loss": 0.0337, + "step": 5693 + }, + { + "epoch": 2.679529411764706, + "grad_norm": 2.2391539174723443e-05, + "learning_rate": 0.0002207602466449854, + "loss": 0.0, + "step": 5694 + }, + { + "epoch": 2.68, + "grad_norm": 0.044245464775005246, + "learning_rate": 0.00022072159149659327, + "loss": 0.0159, + "step": 5695 + }, + { + "epoch": 2.680470588235294, + "grad_norm": 0.03453214421374258, + "learning_rate": 0.00022068293030823026, + "loss": 0.0084, + "step": 5696 + }, + { + "epoch": 2.680941176470588, + "grad_norm": 0.10272050461207344, + "learning_rate": 0.00022064426308319838, + "loss": 0.0575, + "step": 5697 + }, + { + "epoch": 2.6814117647058824, + "grad_norm": 0.030041608575511542, + "learning_rate": 0.0002206055898247999, + "loss": 0.0017, + "step": 5698 + }, + { + "epoch": 2.6818823529411766, + "grad_norm": 4.383956324847463e-05, + "learning_rate": 0.0002205669105363377, + "loss": 0.0, + "step": 5699 + }, + { + "epoch": 2.682352941176471, + "grad_norm": 4.374526629350151e-06, + "learning_rate": 0.00022052822522111522, + "loss": 0.0, + "step": 5700 + }, + { + "epoch": 2.6828235294117646, + "grad_norm": 0.04647903909678315, + "learning_rate": 0.0002204895338824364, + "loss": 0.0217, + "step": 5701 + }, + { + "epoch": 2.683294117647059, + "grad_norm": 0.03691628394390675, + "learning_rate": 0.0002204508365236056, + "loss": 0.0071, + "step": 5702 + }, + { + "epoch": 2.683764705882353, + "grad_norm": 6.559055045764722e-06, + "learning_rate": 0.0002204121331479278, + "loss": 0.0, + "step": 5703 + }, + { + "epoch": 2.684235294117647, + "grad_norm": 0.0427616304391647, + "learning_rate": 0.0002203734237587085, + "loss": 0.014, + "step": 5704 + }, + { + "epoch": 2.684705882352941, + "grad_norm": 0.06992171483292925, + "learning_rate": 0.00022033470835925361, + "loss": 0.0262, + "step": 5705 + }, + { + "epoch": 2.6851764705882353, + "grad_norm": 0.1018564532855495, + "learning_rate": 0.00022029598695286963, + "loss": 0.0709, + "step": 5706 + }, + { + "epoch": 2.6856470588235295, + "grad_norm": 0.05400170366279313, + "learning_rate": 0.00022025725954286356, + "loss": 0.0237, + "step": 5707 + }, + { + "epoch": 2.6861176470588237, + "grad_norm": 0.1192360051500252, + "learning_rate": 0.00022021852613254297, + "loss": 0.0573, + "step": 5708 + }, + { + "epoch": 2.6865882352941175, + "grad_norm": 3.965117869716654e-06, + "learning_rate": 0.00022017978672521577, + "loss": 0.0, + "step": 5709 + }, + { + "epoch": 2.6870588235294117, + "grad_norm": 0.1243324899699653, + "learning_rate": 0.00022014104132419055, + "loss": 0.0592, + "step": 5710 + }, + { + "epoch": 2.687529411764706, + "grad_norm": 0.0281487952255711, + "learning_rate": 0.0002201022899327764, + "loss": 0.0047, + "step": 5711 + }, + { + "epoch": 2.6879999999999997, + "grad_norm": 0.025213171387407973, + "learning_rate": 0.00022006353255428281, + "loss": 0.0033, + "step": 5712 + }, + { + "epoch": 2.688470588235294, + "grad_norm": 0.10634743097666025, + "learning_rate": 0.00022002476919201987, + "loss": 0.0641, + "step": 5713 + }, + { + "epoch": 2.688941176470588, + "grad_norm": 0.1263509690316451, + "learning_rate": 0.00021998599984929818, + "loss": 0.0484, + "step": 5714 + }, + { + "epoch": 2.6894117647058824, + "grad_norm": 0.03114948833727455, + "learning_rate": 0.00021994722452942886, + "loss": 0.0093, + "step": 5715 + }, + { + "epoch": 2.6898823529411766, + "grad_norm": 0.07339662913844656, + "learning_rate": 0.00021990844323572346, + "loss": 0.0336, + "step": 5716 + }, + { + "epoch": 2.690352941176471, + "grad_norm": 0.07284594795572989, + "learning_rate": 0.0002198696559714941, + "loss": 0.0339, + "step": 5717 + }, + { + "epoch": 2.6908235294117646, + "grad_norm": 0.07307456269844026, + "learning_rate": 0.0002198308627400534, + "loss": 0.032, + "step": 5718 + }, + { + "epoch": 2.691294117647059, + "grad_norm": 0.07521117145535039, + "learning_rate": 0.00021979206354471453, + "loss": 0.0503, + "step": 5719 + }, + { + "epoch": 2.691764705882353, + "grad_norm": 0.11529623996898519, + "learning_rate": 0.00021975325838879107, + "loss": 0.0615, + "step": 5720 + }, + { + "epoch": 2.692235294117647, + "grad_norm": 0.08127367176329722, + "learning_rate": 0.0002197144472755972, + "loss": 0.0235, + "step": 5721 + }, + { + "epoch": 2.692705882352941, + "grad_norm": 1.127974694910909e-05, + "learning_rate": 0.00021967563020844762, + "loss": 0.0, + "step": 5722 + }, + { + "epoch": 2.6931764705882353, + "grad_norm": 0.049210822321451485, + "learning_rate": 0.00021963680719065746, + "loss": 0.0103, + "step": 5723 + }, + { + "epoch": 2.6936470588235295, + "grad_norm": 0.0671829761679756, + "learning_rate": 0.0002195979782255424, + "loss": 0.013, + "step": 5724 + }, + { + "epoch": 2.6941176470588237, + "grad_norm": 0.09538129325940463, + "learning_rate": 0.0002195591433164186, + "loss": 0.0444, + "step": 5725 + }, + { + "epoch": 2.6945882352941175, + "grad_norm": 0.07484083391961152, + "learning_rate": 0.00021952030246660283, + "loss": 0.0299, + "step": 5726 + }, + { + "epoch": 2.6950588235294117, + "grad_norm": 0.06257456697284275, + "learning_rate": 0.00021948145567941217, + "loss": 0.0228, + "step": 5727 + }, + { + "epoch": 2.695529411764706, + "grad_norm": 0.11008909642187302, + "learning_rate": 0.0002194426029581644, + "loss": 0.0602, + "step": 5728 + }, + { + "epoch": 2.6959999999999997, + "grad_norm": 0.3697923020667262, + "learning_rate": 0.00021940374430617777, + "loss": 0.2122, + "step": 5729 + }, + { + "epoch": 2.696470588235294, + "grad_norm": 0.10424627334591718, + "learning_rate": 0.0002193648797267709, + "loss": 0.0626, + "step": 5730 + }, + { + "epoch": 2.696941176470588, + "grad_norm": 0.155617852638088, + "learning_rate": 0.0002193260092232631, + "loss": 0.0934, + "step": 5731 + }, + { + "epoch": 2.6974117647058824, + "grad_norm": 8.44127568011932e-05, + "learning_rate": 0.00021928713279897408, + "loss": 0.0, + "step": 5732 + }, + { + "epoch": 2.6978823529411766, + "grad_norm": 0.06495986589758772, + "learning_rate": 0.00021924825045722404, + "loss": 0.0114, + "step": 5733 + }, + { + "epoch": 2.698352941176471, + "grad_norm": 0.1490141679550143, + "learning_rate": 0.00021920936220133373, + "loss": 0.1009, + "step": 5734 + }, + { + "epoch": 2.6988235294117646, + "grad_norm": 0.10300633015560635, + "learning_rate": 0.0002191704680346245, + "loss": 0.0601, + "step": 5735 + }, + { + "epoch": 2.699294117647059, + "grad_norm": 0.06072117893184566, + "learning_rate": 0.00021913156796041794, + "loss": 0.0269, + "step": 5736 + }, + { + "epoch": 2.699764705882353, + "grad_norm": 0.13041942281938393, + "learning_rate": 0.0002190926619820364, + "loss": 0.094, + "step": 5737 + }, + { + "epoch": 2.700235294117647, + "grad_norm": 0.1330977719433218, + "learning_rate": 0.00021905375010280266, + "loss": 0.0757, + "step": 5738 + }, + { + "epoch": 2.700705882352941, + "grad_norm": 0.054358052826089985, + "learning_rate": 0.00021901483232603995, + "loss": 0.0134, + "step": 5739 + }, + { + "epoch": 2.7011764705882353, + "grad_norm": 0.05922226159086422, + "learning_rate": 0.000218975908655072, + "loss": 0.0139, + "step": 5740 + }, + { + "epoch": 2.7016470588235295, + "grad_norm": 0.10095516621104232, + "learning_rate": 0.0002189369790932232, + "loss": 0.0563, + "step": 5741 + }, + { + "epoch": 2.7021176470588237, + "grad_norm": 0.04208427004298113, + "learning_rate": 0.0002188980436438182, + "loss": 0.0121, + "step": 5742 + }, + { + "epoch": 2.7025882352941175, + "grad_norm": 0.1603177695509702, + "learning_rate": 0.0002188591023101823, + "loss": 0.0962, + "step": 5743 + }, + { + "epoch": 2.7030588235294117, + "grad_norm": 0.09225948243837029, + "learning_rate": 0.00021882015509564132, + "loss": 0.0424, + "step": 5744 + }, + { + "epoch": 2.703529411764706, + "grad_norm": 0.020118023301800163, + "learning_rate": 0.00021878120200352158, + "loss": 0.0042, + "step": 5745 + }, + { + "epoch": 2.7039999999999997, + "grad_norm": 0.04460696672000479, + "learning_rate": 0.00021874224303714983, + "loss": 0.0147, + "step": 5746 + }, + { + "epoch": 2.704470588235294, + "grad_norm": 0.20469279677980598, + "learning_rate": 0.00021870327819985331, + "loss": 0.1198, + "step": 5747 + }, + { + "epoch": 2.704941176470588, + "grad_norm": 1.4636195540004178e-05, + "learning_rate": 0.00021866430749495987, + "loss": 0.0, + "step": 5748 + }, + { + "epoch": 2.7054117647058824, + "grad_norm": 0.1286079038445198, + "learning_rate": 0.00021862533092579774, + "loss": 0.0851, + "step": 5749 + }, + { + "epoch": 2.7058823529411766, + "grad_norm": 0.12556445288792242, + "learning_rate": 0.00021858634849569576, + "loss": 0.0652, + "step": 5750 + }, + { + "epoch": 2.706352941176471, + "grad_norm": 0.12072134237601438, + "learning_rate": 0.00021854736020798319, + "loss": 0.0542, + "step": 5751 + }, + { + "epoch": 2.7068235294117646, + "grad_norm": 0.1420074881937066, + "learning_rate": 0.00021850836606598987, + "loss": 0.0529, + "step": 5752 + }, + { + "epoch": 2.707294117647059, + "grad_norm": 3.479948897935647e-05, + "learning_rate": 0.00021846936607304603, + "loss": 0.0, + "step": 5753 + }, + { + "epoch": 2.707764705882353, + "grad_norm": 0.08347524471812443, + "learning_rate": 0.00021843036023248247, + "loss": 0.0301, + "step": 5754 + }, + { + "epoch": 2.708235294117647, + "grad_norm": 1.2306874000764923e-05, + "learning_rate": 0.00021839134854763055, + "loss": 0.0, + "step": 5755 + }, + { + "epoch": 2.708705882352941, + "grad_norm": 0.07202956800721542, + "learning_rate": 0.00021835233102182199, + "loss": 0.0387, + "step": 5756 + }, + { + "epoch": 2.7091764705882353, + "grad_norm": 0.032821133935074645, + "learning_rate": 0.00021831330765838903, + "loss": 0.0043, + "step": 5757 + }, + { + "epoch": 2.7096470588235295, + "grad_norm": 0.26159713664291134, + "learning_rate": 0.00021827427846066462, + "loss": 0.1348, + "step": 5758 + }, + { + "epoch": 2.7101176470588237, + "grad_norm": 0.03840987588249993, + "learning_rate": 0.0002182352434319819, + "loss": 0.0119, + "step": 5759 + }, + { + "epoch": 2.7105882352941175, + "grad_norm": 0.018902624535656932, + "learning_rate": 0.0002181962025756747, + "loss": 0.0034, + "step": 5760 + }, + { + "epoch": 2.7110588235294117, + "grad_norm": 0.00011557371103671976, + "learning_rate": 0.00021815715589507728, + "loss": 0.0, + "step": 5761 + }, + { + "epoch": 2.711529411764706, + "grad_norm": 0.08535910415827938, + "learning_rate": 0.00021811810339352446, + "loss": 0.0463, + "step": 5762 + }, + { + "epoch": 2.7119999999999997, + "grad_norm": 8.181927055024227e-06, + "learning_rate": 0.0002180790450743515, + "loss": 0.0, + "step": 5763 + }, + { + "epoch": 2.712470588235294, + "grad_norm": 8.243282385392092e-06, + "learning_rate": 0.0002180399809408941, + "loss": 0.0, + "step": 5764 + }, + { + "epoch": 2.712941176470588, + "grad_norm": 0.07107732200615795, + "learning_rate": 0.00021800091099648865, + "loss": 0.0307, + "step": 5765 + }, + { + "epoch": 2.7134117647058824, + "grad_norm": 0.10758481300250501, + "learning_rate": 0.0002179618352444719, + "loss": 0.0641, + "step": 5766 + }, + { + "epoch": 2.7138823529411766, + "grad_norm": 0.11037851758792813, + "learning_rate": 0.00021792275368818102, + "loss": 0.0551, + "step": 5767 + }, + { + "epoch": 2.714352941176471, + "grad_norm": 4.362477768426576e-06, + "learning_rate": 0.00021788366633095375, + "loss": 0.0, + "step": 5768 + }, + { + "epoch": 2.7148235294117646, + "grad_norm": 0.08544259561289515, + "learning_rate": 0.00021784457317612852, + "loss": 0.0486, + "step": 5769 + }, + { + "epoch": 2.715294117647059, + "grad_norm": 0.03143920137478252, + "learning_rate": 0.0002178054742270439, + "loss": 0.0052, + "step": 5770 + }, + { + "epoch": 2.715764705882353, + "grad_norm": 0.00014245948265012118, + "learning_rate": 0.0002177663694870392, + "loss": 0.0, + "step": 5771 + }, + { + "epoch": 2.716235294117647, + "grad_norm": 0.054140437524111434, + "learning_rate": 0.00021772725895945422, + "loss": 0.024, + "step": 5772 + }, + { + "epoch": 2.716705882352941, + "grad_norm": 0.16024149403851598, + "learning_rate": 0.0002176881426476291, + "loss": 0.1271, + "step": 5773 + }, + { + "epoch": 2.7171764705882353, + "grad_norm": 0.08127512783084473, + "learning_rate": 0.00021764902055490456, + "loss": 0.0522, + "step": 5774 + }, + { + "epoch": 2.7176470588235295, + "grad_norm": 0.10634317895067895, + "learning_rate": 0.00021760989268462182, + "loss": 0.055, + "step": 5775 + }, + { + "epoch": 2.7181176470588237, + "grad_norm": 0.05677362219022509, + "learning_rate": 0.00021757075904012266, + "loss": 0.0105, + "step": 5776 + }, + { + "epoch": 2.7185882352941175, + "grad_norm": 0.11517378472374307, + "learning_rate": 0.00021753161962474924, + "loss": 0.0701, + "step": 5777 + }, + { + "epoch": 2.7190588235294118, + "grad_norm": 7.677213377294013e-06, + "learning_rate": 0.00021749247444184422, + "loss": 0.0, + "step": 5778 + }, + { + "epoch": 2.719529411764706, + "grad_norm": 0.08488384091308564, + "learning_rate": 0.00021745332349475092, + "loss": 0.0495, + "step": 5779 + }, + { + "epoch": 2.7199999999999998, + "grad_norm": 0.04783232352056214, + "learning_rate": 0.0002174141667868129, + "loss": 0.0202, + "step": 5780 + }, + { + "epoch": 2.720470588235294, + "grad_norm": 2.2333084290980928e-05, + "learning_rate": 0.00021737500432137435, + "loss": 0.0, + "step": 5781 + }, + { + "epoch": 2.720941176470588, + "grad_norm": 0.12040006275494215, + "learning_rate": 0.00021733583610177997, + "loss": 0.0597, + "step": 5782 + }, + { + "epoch": 2.7214117647058824, + "grad_norm": 0.035956810679021484, + "learning_rate": 0.00021729666213137494, + "loss": 0.0125, + "step": 5783 + }, + { + "epoch": 2.7218823529411766, + "grad_norm": 0.10111336289988368, + "learning_rate": 0.00021725748241350486, + "loss": 0.0477, + "step": 5784 + }, + { + "epoch": 2.722352941176471, + "grad_norm": 0.05351133803200033, + "learning_rate": 0.00021721829695151585, + "loss": 0.0226, + "step": 5785 + }, + { + "epoch": 2.7228235294117646, + "grad_norm": 0.2464446077198068, + "learning_rate": 0.00021717910574875458, + "loss": 0.1858, + "step": 5786 + }, + { + "epoch": 2.723294117647059, + "grad_norm": 0.08111819563866139, + "learning_rate": 0.00021713990880856818, + "loss": 0.0151, + "step": 5787 + }, + { + "epoch": 2.723764705882353, + "grad_norm": 0.08710269508757165, + "learning_rate": 0.0002171007061343043, + "loss": 0.0516, + "step": 5788 + }, + { + "epoch": 2.724235294117647, + "grad_norm": 0.024117580239335266, + "learning_rate": 0.0002170614977293109, + "loss": 0.0034, + "step": 5789 + }, + { + "epoch": 2.724705882352941, + "grad_norm": 0.1523711077267296, + "learning_rate": 0.00021702228359693672, + "loss": 0.0876, + "step": 5790 + }, + { + "epoch": 2.7251764705882353, + "grad_norm": 0.08404959550658192, + "learning_rate": 0.00021698306374053075, + "loss": 0.0411, + "step": 5791 + }, + { + "epoch": 2.7256470588235295, + "grad_norm": 0.04694719858876052, + "learning_rate": 0.00021694383816344258, + "loss": 0.0051, + "step": 5792 + }, + { + "epoch": 2.7261176470588238, + "grad_norm": 8.292895343496208e-06, + "learning_rate": 0.0002169046068690223, + "loss": 0.0, + "step": 5793 + }, + { + "epoch": 2.7265882352941175, + "grad_norm": 0.07017594911725848, + "learning_rate": 0.00021686536986062043, + "loss": 0.035, + "step": 5794 + }, + { + "epoch": 2.7270588235294118, + "grad_norm": 0.08294580889955673, + "learning_rate": 0.000216826127141588, + "loss": 0.0401, + "step": 5795 + }, + { + "epoch": 2.727529411764706, + "grad_norm": 0.07440124879406342, + "learning_rate": 0.0002167868787152765, + "loss": 0.0362, + "step": 5796 + }, + { + "epoch": 2.7279999999999998, + "grad_norm": 0.06471197536091294, + "learning_rate": 0.000216747624585038, + "loss": 0.0224, + "step": 5797 + }, + { + "epoch": 2.728470588235294, + "grad_norm": 0.1958748751528078, + "learning_rate": 0.00021670836475422498, + "loss": 0.0685, + "step": 5798 + }, + { + "epoch": 2.728941176470588, + "grad_norm": 0.056000242421407664, + "learning_rate": 0.00021666909922619037, + "loss": 0.0186, + "step": 5799 + }, + { + "epoch": 2.7294117647058824, + "grad_norm": 0.03893804864751852, + "learning_rate": 0.0002166298280042877, + "loss": 0.0098, + "step": 5800 + }, + { + "epoch": 2.7298823529411766, + "grad_norm": 8.486738914492555e-05, + "learning_rate": 0.00021659055109187088, + "loss": 0.0, + "step": 5801 + }, + { + "epoch": 2.7303529411764704, + "grad_norm": 0.03168765899106277, + "learning_rate": 0.0002165512684922944, + "loss": 0.0092, + "step": 5802 + }, + { + "epoch": 2.7308235294117646, + "grad_norm": 0.046304272709404716, + "learning_rate": 0.0002165119802089131, + "loss": 0.0159, + "step": 5803 + }, + { + "epoch": 2.731294117647059, + "grad_norm": 0.04935070562800206, + "learning_rate": 0.00021647268624508254, + "loss": 0.0229, + "step": 5804 + }, + { + "epoch": 2.731764705882353, + "grad_norm": 0.22276990037121316, + "learning_rate": 0.00021643338660415844, + "loss": 0.1357, + "step": 5805 + }, + { + "epoch": 2.732235294117647, + "grad_norm": 0.14384485193752966, + "learning_rate": 0.0002163940812894973, + "loss": 0.0823, + "step": 5806 + }, + { + "epoch": 2.732705882352941, + "grad_norm": 0.08344446235420921, + "learning_rate": 0.00021635477030445597, + "loss": 0.0534, + "step": 5807 + }, + { + "epoch": 2.7331764705882353, + "grad_norm": 0.07854910710994294, + "learning_rate": 0.0002163154536523918, + "loss": 0.0446, + "step": 5808 + }, + { + "epoch": 2.7336470588235295, + "grad_norm": 0.1613133442734245, + "learning_rate": 0.0002162761313366626, + "loss": 0.0582, + "step": 5809 + }, + { + "epoch": 2.7341176470588238, + "grad_norm": 0.0627114318172618, + "learning_rate": 0.0002162368033606267, + "loss": 0.0139, + "step": 5810 + }, + { + "epoch": 2.7345882352941175, + "grad_norm": 1.7884134889390998e-05, + "learning_rate": 0.00021619746972764288, + "loss": 0.0, + "step": 5811 + }, + { + "epoch": 2.7350588235294118, + "grad_norm": 0.05530995668825153, + "learning_rate": 0.00021615813044107045, + "loss": 0.0201, + "step": 5812 + }, + { + "epoch": 2.735529411764706, + "grad_norm": 4.892628679847065e-05, + "learning_rate": 0.00021611878550426922, + "loss": 0.0, + "step": 5813 + }, + { + "epoch": 2.7359999999999998, + "grad_norm": 0.07423610818198548, + "learning_rate": 0.00021607943492059936, + "loss": 0.0318, + "step": 5814 + }, + { + "epoch": 2.736470588235294, + "grad_norm": 2.9793077396316438e-05, + "learning_rate": 0.00021604007869342168, + "loss": 0.0, + "step": 5815 + }, + { + "epoch": 2.736941176470588, + "grad_norm": 0.08314593730992696, + "learning_rate": 0.00021600071682609733, + "loss": 0.0248, + "step": 5816 + }, + { + "epoch": 2.7374117647058824, + "grad_norm": 1.838665541936678e-05, + "learning_rate": 0.00021596134932198803, + "loss": 0.0, + "step": 5817 + }, + { + "epoch": 2.7378823529411767, + "grad_norm": 0.21799549901962761, + "learning_rate": 0.00021592197618445596, + "loss": 0.0857, + "step": 5818 + }, + { + "epoch": 2.7383529411764704, + "grad_norm": 0.09326289798389249, + "learning_rate": 0.00021588259741686373, + "loss": 0.0471, + "step": 5819 + }, + { + "epoch": 2.7388235294117647, + "grad_norm": 0.12262065738308375, + "learning_rate": 0.00021584321302257464, + "loss": 0.0748, + "step": 5820 + }, + { + "epoch": 2.739294117647059, + "grad_norm": 0.04324915785273087, + "learning_rate": 0.00021580382300495215, + "loss": 0.0093, + "step": 5821 + }, + { + "epoch": 2.739764705882353, + "grad_norm": 0.04107024593711574, + "learning_rate": 0.00021576442736736032, + "loss": 0.0108, + "step": 5822 + }, + { + "epoch": 2.740235294117647, + "grad_norm": 0.098376201059796, + "learning_rate": 0.00021572502611316393, + "loss": 0.0488, + "step": 5823 + }, + { + "epoch": 2.740705882352941, + "grad_norm": 0.08955482068141855, + "learning_rate": 0.0002156856192457279, + "loss": 0.0393, + "step": 5824 + }, + { + "epoch": 2.7411764705882353, + "grad_norm": 0.0006728881776708717, + "learning_rate": 0.00021564620676841782, + "loss": 0.0, + "step": 5825 + }, + { + "epoch": 2.7416470588235295, + "grad_norm": 1.5740240838194086e-05, + "learning_rate": 0.0002156067886845996, + "loss": 0.0, + "step": 5826 + }, + { + "epoch": 2.7421176470588238, + "grad_norm": 0.04076470246610874, + "learning_rate": 0.00021556736499763992, + "loss": 0.0091, + "step": 5827 + }, + { + "epoch": 2.7425882352941175, + "grad_norm": 0.0960586172212733, + "learning_rate": 0.00021552793571090568, + "loss": 0.0653, + "step": 5828 + }, + { + "epoch": 2.7430588235294118, + "grad_norm": 0.06496984520530837, + "learning_rate": 0.0002154885008277642, + "loss": 0.0174, + "step": 5829 + }, + { + "epoch": 2.743529411764706, + "grad_norm": 2.5982929524209945e-05, + "learning_rate": 0.0002154490603515836, + "loss": 0.0, + "step": 5830 + }, + { + "epoch": 2.7439999999999998, + "grad_norm": 0.0982355171712741, + "learning_rate": 0.0002154096142857323, + "loss": 0.0379, + "step": 5831 + }, + { + "epoch": 2.744470588235294, + "grad_norm": 0.08163765592307166, + "learning_rate": 0.000215370162633579, + "loss": 0.0341, + "step": 5832 + }, + { + "epoch": 2.744941176470588, + "grad_norm": 0.027054285282009424, + "learning_rate": 0.00021533070539849316, + "loss": 0.0067, + "step": 5833 + }, + { + "epoch": 2.7454117647058824, + "grad_norm": 0.07925530031964152, + "learning_rate": 0.00021529124258384466, + "loss": 0.0319, + "step": 5834 + }, + { + "epoch": 2.7458823529411767, + "grad_norm": 0.06456876085442717, + "learning_rate": 0.00021525177419300383, + "loss": 0.0135, + "step": 5835 + }, + { + "epoch": 2.7463529411764704, + "grad_norm": 0.054550668197537545, + "learning_rate": 0.0002152123002293414, + "loss": 0.0223, + "step": 5836 + }, + { + "epoch": 2.7468235294117647, + "grad_norm": 0.11281564505001941, + "learning_rate": 0.0002151728206962287, + "loss": 0.0653, + "step": 5837 + }, + { + "epoch": 2.747294117647059, + "grad_norm": 0.09846299965050902, + "learning_rate": 0.00021513333559703738, + "loss": 0.0491, + "step": 5838 + }, + { + "epoch": 2.7477647058823527, + "grad_norm": 0.12117384194308863, + "learning_rate": 0.00021509384493513978, + "loss": 0.0896, + "step": 5839 + }, + { + "epoch": 2.748235294117647, + "grad_norm": 0.10438027032727437, + "learning_rate": 0.00021505434871390848, + "loss": 0.0433, + "step": 5840 + }, + { + "epoch": 2.748705882352941, + "grad_norm": 0.11750133262745446, + "learning_rate": 0.00021501484693671677, + "loss": 0.0726, + "step": 5841 + }, + { + "epoch": 2.7491764705882353, + "grad_norm": 0.11450099316571591, + "learning_rate": 0.00021497533960693823, + "loss": 0.0785, + "step": 5842 + }, + { + "epoch": 2.7496470588235296, + "grad_norm": 0.033575971855080246, + "learning_rate": 0.00021493582672794693, + "loss": 0.0064, + "step": 5843 + }, + { + "epoch": 2.7501176470588238, + "grad_norm": 0.03325882529120329, + "learning_rate": 0.00021489630830311758, + "loss": 0.0084, + "step": 5844 + }, + { + "epoch": 2.7505882352941176, + "grad_norm": 0.1467739473248641, + "learning_rate": 0.00021485678433582516, + "loss": 0.0936, + "step": 5845 + }, + { + "epoch": 2.751058823529412, + "grad_norm": 1.6837169850844708e-05, + "learning_rate": 0.00021481725482944518, + "loss": 0.0, + "step": 5846 + }, + { + "epoch": 2.751529411764706, + "grad_norm": 0.01662335685855642, + "learning_rate": 0.0002147777197873538, + "loss": 0.0022, + "step": 5847 + }, + { + "epoch": 2.752, + "grad_norm": 0.07488557030129152, + "learning_rate": 0.0002147381792129274, + "loss": 0.0354, + "step": 5848 + }, + { + "epoch": 2.752470588235294, + "grad_norm": 0.06719489148427452, + "learning_rate": 0.00021469863310954296, + "loss": 0.0275, + "step": 5849 + }, + { + "epoch": 2.7529411764705882, + "grad_norm": 0.15744407088510184, + "learning_rate": 0.00021465908148057787, + "loss": 0.1199, + "step": 5850 + }, + { + "epoch": 2.7534117647058824, + "grad_norm": 0.04646780638922339, + "learning_rate": 0.00021461952432941006, + "loss": 0.0147, + "step": 5851 + }, + { + "epoch": 2.7538823529411767, + "grad_norm": 0.20517063488802104, + "learning_rate": 0.00021457996165941798, + "loss": 0.1451, + "step": 5852 + }, + { + "epoch": 2.7543529411764704, + "grad_norm": 0.09370810363596635, + "learning_rate": 0.00021454039347398033, + "loss": 0.0679, + "step": 5853 + }, + { + "epoch": 2.7548235294117647, + "grad_norm": 0.06261306130559148, + "learning_rate": 0.00021450081977647658, + "loss": 0.0294, + "step": 5854 + }, + { + "epoch": 2.755294117647059, + "grad_norm": 0.09148749197255807, + "learning_rate": 0.00021446124057028642, + "loss": 0.0662, + "step": 5855 + }, + { + "epoch": 2.7557647058823527, + "grad_norm": 0.07711639190686316, + "learning_rate": 0.00021442165585879014, + "loss": 0.0292, + "step": 5856 + }, + { + "epoch": 2.756235294117647, + "grad_norm": 0.10671357156082285, + "learning_rate": 0.00021438206564536845, + "loss": 0.061, + "step": 5857 + }, + { + "epoch": 2.756705882352941, + "grad_norm": 0.08647675465147317, + "learning_rate": 0.00021434246993340256, + "loss": 0.0457, + "step": 5858 + }, + { + "epoch": 2.7571764705882353, + "grad_norm": 0.016345402893381773, + "learning_rate": 0.00021430286872627417, + "loss": 0.0019, + "step": 5859 + }, + { + "epoch": 2.7576470588235296, + "grad_norm": 0.010725956925455594, + "learning_rate": 0.00021426326202736532, + "loss": 0.0005, + "step": 5860 + }, + { + "epoch": 2.758117647058824, + "grad_norm": 0.06263780719951814, + "learning_rate": 0.00021422364984005876, + "loss": 0.0301, + "step": 5861 + }, + { + "epoch": 2.7585882352941176, + "grad_norm": 0.05974638636973943, + "learning_rate": 0.00021418403216773746, + "loss": 0.027, + "step": 5862 + }, + { + "epoch": 2.759058823529412, + "grad_norm": 0.0720083739437832, + "learning_rate": 0.000214144409013785, + "loss": 0.032, + "step": 5863 + }, + { + "epoch": 2.759529411764706, + "grad_norm": 0.06844100628698624, + "learning_rate": 0.00021410478038158537, + "loss": 0.0373, + "step": 5864 + }, + { + "epoch": 2.76, + "grad_norm": 0.042689351186156084, + "learning_rate": 0.00021406514627452306, + "loss": 0.0126, + "step": 5865 + }, + { + "epoch": 2.760470588235294, + "grad_norm": 0.03847195302608364, + "learning_rate": 0.00021402550669598303, + "loss": 0.0101, + "step": 5866 + }, + { + "epoch": 2.7609411764705882, + "grad_norm": 0.11240224777786804, + "learning_rate": 0.00021398586164935067, + "loss": 0.063, + "step": 5867 + }, + { + "epoch": 2.7614117647058825, + "grad_norm": 0.06322067458456705, + "learning_rate": 0.0002139462111380119, + "loss": 0.0292, + "step": 5868 + }, + { + "epoch": 2.7618823529411767, + "grad_norm": 0.09374362725357668, + "learning_rate": 0.00021390655516535304, + "loss": 0.0364, + "step": 5869 + }, + { + "epoch": 2.7623529411764705, + "grad_norm": 0.04178789850997011, + "learning_rate": 0.00021386689373476087, + "loss": 0.0131, + "step": 5870 + }, + { + "epoch": 2.7628235294117647, + "grad_norm": 0.0774344618905882, + "learning_rate": 0.00021382722684962276, + "loss": 0.0305, + "step": 5871 + }, + { + "epoch": 2.763294117647059, + "grad_norm": 2.1326211077229604e-05, + "learning_rate": 0.00021378755451332637, + "loss": 0.0, + "step": 5872 + }, + { + "epoch": 2.7637647058823527, + "grad_norm": 0.1657498385082517, + "learning_rate": 0.00021374787672926, + "loss": 0.123, + "step": 5873 + }, + { + "epoch": 2.764235294117647, + "grad_norm": 0.06918229227490966, + "learning_rate": 0.00021370819350081222, + "loss": 0.0293, + "step": 5874 + }, + { + "epoch": 2.764705882352941, + "grad_norm": 0.13646257987815094, + "learning_rate": 0.00021366850483137226, + "loss": 0.0675, + "step": 5875 + }, + { + "epoch": 2.7651764705882353, + "grad_norm": 0.10680806333249675, + "learning_rate": 0.00021362881072432965, + "loss": 0.0665, + "step": 5876 + }, + { + "epoch": 2.7656470588235296, + "grad_norm": 0.1287179636171381, + "learning_rate": 0.00021358911118307455, + "loss": 0.0824, + "step": 5877 + }, + { + "epoch": 2.766117647058824, + "grad_norm": 0.04669704835920545, + "learning_rate": 0.00021354940621099745, + "loss": 0.0184, + "step": 5878 + }, + { + "epoch": 2.7665882352941176, + "grad_norm": 0.08763167431828628, + "learning_rate": 0.00021350969581148934, + "loss": 0.0437, + "step": 5879 + }, + { + "epoch": 2.767058823529412, + "grad_norm": 0.14615840905765934, + "learning_rate": 0.00021346997998794172, + "loss": 0.0801, + "step": 5880 + }, + { + "epoch": 2.767529411764706, + "grad_norm": 0.08333082925542991, + "learning_rate": 0.00021343025874374645, + "loss": 0.0276, + "step": 5881 + }, + { + "epoch": 2.768, + "grad_norm": 0.055334145613709716, + "learning_rate": 0.000213390532082296, + "loss": 0.0265, + "step": 5882 + }, + { + "epoch": 2.768470588235294, + "grad_norm": 0.10762270033512328, + "learning_rate": 0.00021335080000698317, + "loss": 0.0563, + "step": 5883 + }, + { + "epoch": 2.7689411764705882, + "grad_norm": 0.07342280275512822, + "learning_rate": 0.00021331106252120126, + "loss": 0.0317, + "step": 5884 + }, + { + "epoch": 2.7694117647058825, + "grad_norm": 0.11360432722272353, + "learning_rate": 0.00021327131962834407, + "loss": 0.0562, + "step": 5885 + }, + { + "epoch": 2.7698823529411767, + "grad_norm": 0.10975440931027003, + "learning_rate": 0.00021323157133180587, + "loss": 0.06, + "step": 5886 + }, + { + "epoch": 2.7703529411764705, + "grad_norm": 0.04235797211075377, + "learning_rate": 0.0002131918176349813, + "loss": 0.0155, + "step": 5887 + }, + { + "epoch": 2.7708235294117647, + "grad_norm": 3.98272020405836e-05, + "learning_rate": 0.0002131520585412656, + "loss": 0.0, + "step": 5888 + }, + { + "epoch": 2.771294117647059, + "grad_norm": 0.039810652949461546, + "learning_rate": 0.0002131122940540543, + "loss": 0.0077, + "step": 5889 + }, + { + "epoch": 2.7717647058823527, + "grad_norm": 0.037256493411776365, + "learning_rate": 0.00021307252417674356, + "loss": 0.0112, + "step": 5890 + }, + { + "epoch": 2.772235294117647, + "grad_norm": 0.14141592081736146, + "learning_rate": 0.0002130327489127298, + "loss": 0.0829, + "step": 5891 + }, + { + "epoch": 2.772705882352941, + "grad_norm": 0.10481073612034078, + "learning_rate": 0.0002129929682654102, + "loss": 0.0882, + "step": 5892 + }, + { + "epoch": 2.7731764705882354, + "grad_norm": 7.576706947281846e-05, + "learning_rate": 0.0002129531822381821, + "loss": 0.0, + "step": 5893 + }, + { + "epoch": 2.7736470588235296, + "grad_norm": 0.0001543356448801508, + "learning_rate": 0.00021291339083444344, + "loss": 0.0, + "step": 5894 + }, + { + "epoch": 2.774117647058824, + "grad_norm": 0.1143332342581827, + "learning_rate": 0.00021287359405759265, + "loss": 0.063, + "step": 5895 + }, + { + "epoch": 2.7745882352941176, + "grad_norm": 4.7047543932080196e-05, + "learning_rate": 0.0002128337919110285, + "loss": 0.0, + "step": 5896 + }, + { + "epoch": 2.775058823529412, + "grad_norm": 0.13110858389421046, + "learning_rate": 0.00021279398439815034, + "loss": 0.0785, + "step": 5897 + }, + { + "epoch": 2.775529411764706, + "grad_norm": 0.11071854613361617, + "learning_rate": 0.00021275417152235787, + "loss": 0.0538, + "step": 5898 + }, + { + "epoch": 2.776, + "grad_norm": 0.11360710706480871, + "learning_rate": 0.00021271435328705138, + "loss": 0.0646, + "step": 5899 + }, + { + "epoch": 2.776470588235294, + "grad_norm": 0.11914277965480967, + "learning_rate": 0.00021267452969563153, + "loss": 0.1019, + "step": 5900 + }, + { + "epoch": 2.7769411764705882, + "grad_norm": 0.08304596338413299, + "learning_rate": 0.00021263470075149937, + "loss": 0.0305, + "step": 5901 + }, + { + "epoch": 2.7774117647058825, + "grad_norm": 0.10273037205006777, + "learning_rate": 0.00021259486645805662, + "loss": 0.0644, + "step": 5902 + }, + { + "epoch": 2.7778823529411767, + "grad_norm": 0.05419807796434455, + "learning_rate": 0.00021255502681870518, + "loss": 0.0288, + "step": 5903 + }, + { + "epoch": 2.7783529411764705, + "grad_norm": 0.00023790277713484445, + "learning_rate": 0.00021251518183684767, + "loss": 0.0, + "step": 5904 + }, + { + "epoch": 2.7788235294117647, + "grad_norm": 0.17625383985855705, + "learning_rate": 0.00021247533151588693, + "loss": 0.1009, + "step": 5905 + }, + { + "epoch": 2.779294117647059, + "grad_norm": 0.0867718103536694, + "learning_rate": 0.0002124354758592265, + "loss": 0.0386, + "step": 5906 + }, + { + "epoch": 2.7797647058823527, + "grad_norm": 0.04582040141297526, + "learning_rate": 0.00021239561487027017, + "loss": 0.0148, + "step": 5907 + }, + { + "epoch": 2.780235294117647, + "grad_norm": 0.09219546718982864, + "learning_rate": 0.00021235574855242226, + "loss": 0.0536, + "step": 5908 + }, + { + "epoch": 2.780705882352941, + "grad_norm": 0.06063985018987937, + "learning_rate": 0.00021231587690908758, + "loss": 0.0265, + "step": 5909 + }, + { + "epoch": 2.7811764705882354, + "grad_norm": 0.0762525524636551, + "learning_rate": 0.00021227599994367137, + "loss": 0.0461, + "step": 5910 + }, + { + "epoch": 2.7816470588235296, + "grad_norm": 0.03171007032780225, + "learning_rate": 0.0002122361176595793, + "loss": 0.0059, + "step": 5911 + }, + { + "epoch": 2.782117647058824, + "grad_norm": 0.08203133021529203, + "learning_rate": 0.0002121962300602175, + "loss": 0.0626, + "step": 5912 + }, + { + "epoch": 2.7825882352941176, + "grad_norm": 0.2039477469758449, + "learning_rate": 0.00021215633714899264, + "loss": 0.1356, + "step": 5913 + }, + { + "epoch": 2.783058823529412, + "grad_norm": 0.04062106982562249, + "learning_rate": 0.00021211643892931167, + "loss": 0.0116, + "step": 5914 + }, + { + "epoch": 2.783529411764706, + "grad_norm": 0.07763238781584081, + "learning_rate": 0.00021207653540458209, + "loss": 0.0449, + "step": 5915 + }, + { + "epoch": 2.784, + "grad_norm": 0.09718877715875285, + "learning_rate": 0.00021203662657821198, + "loss": 0.0306, + "step": 5916 + }, + { + "epoch": 2.784470588235294, + "grad_norm": 0.11698598244061979, + "learning_rate": 0.00021199671245360964, + "loss": 0.0534, + "step": 5917 + }, + { + "epoch": 2.7849411764705883, + "grad_norm": 0.11431987247572675, + "learning_rate": 0.0002119567930341839, + "loss": 0.0481, + "step": 5918 + }, + { + "epoch": 2.7854117647058825, + "grad_norm": 9.171086160131784e-05, + "learning_rate": 0.00021191686832334423, + "loss": 0.0, + "step": 5919 + }, + { + "epoch": 2.7858823529411767, + "grad_norm": 0.07317292336667057, + "learning_rate": 0.0002118769383245003, + "loss": 0.0214, + "step": 5920 + }, + { + "epoch": 2.7863529411764705, + "grad_norm": 0.09531710104748839, + "learning_rate": 0.0002118370030410623, + "loss": 0.033, + "step": 5921 + }, + { + "epoch": 2.7868235294117647, + "grad_norm": 0.10453429687558796, + "learning_rate": 0.0002117970624764409, + "loss": 0.0657, + "step": 5922 + }, + { + "epoch": 2.787294117647059, + "grad_norm": 0.010337359009266002, + "learning_rate": 0.00021175711663404728, + "loss": 0.0008, + "step": 5923 + }, + { + "epoch": 2.7877647058823527, + "grad_norm": 0.06711597702385666, + "learning_rate": 0.000211717165517293, + "loss": 0.0351, + "step": 5924 + }, + { + "epoch": 2.788235294117647, + "grad_norm": 0.09842931238478864, + "learning_rate": 0.00021167720912959004, + "loss": 0.0581, + "step": 5925 + }, + { + "epoch": 2.788705882352941, + "grad_norm": 0.07588407380957639, + "learning_rate": 0.00021163724747435091, + "loss": 0.0272, + "step": 5926 + }, + { + "epoch": 2.7891764705882354, + "grad_norm": 0.027265669781095545, + "learning_rate": 0.00021159728055498856, + "loss": 0.0067, + "step": 5927 + }, + { + "epoch": 2.7896470588235296, + "grad_norm": 0.09533550194322818, + "learning_rate": 0.00021155730837491625, + "loss": 0.0222, + "step": 5928 + }, + { + "epoch": 2.7901176470588234, + "grad_norm": 9.583241588034345e-05, + "learning_rate": 0.0002115173309375479, + "loss": 0.0, + "step": 5929 + }, + { + "epoch": 2.7905882352941176, + "grad_norm": 0.1289107613461725, + "learning_rate": 0.00021147734824629772, + "loss": 0.0942, + "step": 5930 + }, + { + "epoch": 2.791058823529412, + "grad_norm": 0.07359047240286273, + "learning_rate": 0.0002114373603045805, + "loss": 0.0328, + "step": 5931 + }, + { + "epoch": 2.791529411764706, + "grad_norm": 0.283432854391608, + "learning_rate": 0.00021139736711581132, + "loss": 0.1417, + "step": 5932 + }, + { + "epoch": 2.792, + "grad_norm": 0.04155897907173937, + "learning_rate": 0.00021135736868340587, + "loss": 0.0124, + "step": 5933 + }, + { + "epoch": 2.792470588235294, + "grad_norm": 0.13226381355189323, + "learning_rate": 0.00021131736501078018, + "loss": 0.0862, + "step": 5934 + }, + { + "epoch": 2.7929411764705883, + "grad_norm": 0.07525722230880286, + "learning_rate": 0.0002112773561013507, + "loss": 0.0329, + "step": 5935 + }, + { + "epoch": 2.7934117647058825, + "grad_norm": 0.06195724744497593, + "learning_rate": 0.0002112373419585345, + "loss": 0.0213, + "step": 5936 + }, + { + "epoch": 2.7938823529411767, + "grad_norm": 0.10019178539188786, + "learning_rate": 0.00021119732258574893, + "loss": 0.0401, + "step": 5937 + }, + { + "epoch": 2.7943529411764705, + "grad_norm": 0.1486466118298257, + "learning_rate": 0.0002111572979864118, + "loss": 0.0883, + "step": 5938 + }, + { + "epoch": 2.7948235294117647, + "grad_norm": 0.03742327680817496, + "learning_rate": 0.00021111726816394147, + "loss": 0.012, + "step": 5939 + }, + { + "epoch": 2.795294117647059, + "grad_norm": 0.035026538046935674, + "learning_rate": 0.00021107723312175665, + "loss": 0.0045, + "step": 5940 + }, + { + "epoch": 2.7957647058823527, + "grad_norm": 8.135397152708899e-06, + "learning_rate": 0.00021103719286327658, + "loss": 0.0, + "step": 5941 + }, + { + "epoch": 2.796235294117647, + "grad_norm": 0.06293082587994538, + "learning_rate": 0.0002109971473919208, + "loss": 0.0132, + "step": 5942 + }, + { + "epoch": 2.796705882352941, + "grad_norm": 1.090130801361972e-05, + "learning_rate": 0.00021095709671110942, + "loss": 0.0, + "step": 5943 + }, + { + "epoch": 2.7971764705882354, + "grad_norm": 0.10848251301462479, + "learning_rate": 0.00021091704082426308, + "loss": 0.0395, + "step": 5944 + }, + { + "epoch": 2.7976470588235296, + "grad_norm": 0.13457910020823122, + "learning_rate": 0.0002108769797348026, + "loss": 0.0691, + "step": 5945 + }, + { + "epoch": 2.7981176470588234, + "grad_norm": 0.11353822718528438, + "learning_rate": 0.00021083691344614947, + "loss": 0.0463, + "step": 5946 + }, + { + "epoch": 2.7985882352941176, + "grad_norm": 0.07211663472375275, + "learning_rate": 0.00021079684196172554, + "loss": 0.0233, + "step": 5947 + }, + { + "epoch": 2.799058823529412, + "grad_norm": 0.018209155965689762, + "learning_rate": 0.00021075676528495307, + "loss": 0.0031, + "step": 5948 + }, + { + "epoch": 2.799529411764706, + "grad_norm": 0.04376350327765208, + "learning_rate": 0.00021071668341925482, + "loss": 0.0167, + "step": 5949 + }, + { + "epoch": 2.8, + "grad_norm": 7.839682109702846e-05, + "learning_rate": 0.00021067659636805403, + "loss": 0.0, + "step": 5950 + }, + { + "epoch": 2.800470588235294, + "grad_norm": 0.001374009847332931, + "learning_rate": 0.00021063650413477428, + "loss": 0.0, + "step": 5951 + }, + { + "epoch": 2.8009411764705883, + "grad_norm": 0.04095649708986493, + "learning_rate": 0.00021059640672283964, + "loss": 0.0106, + "step": 5952 + }, + { + "epoch": 2.8014117647058825, + "grad_norm": 0.09897127713120468, + "learning_rate": 0.0002105563041356747, + "loss": 0.0418, + "step": 5953 + }, + { + "epoch": 2.8018823529411767, + "grad_norm": 0.25527149446853553, + "learning_rate": 0.00021051619637670431, + "loss": 0.139, + "step": 5954 + }, + { + "epoch": 2.8023529411764705, + "grad_norm": 1.5917502028934896e-05, + "learning_rate": 0.00021047608344935395, + "loss": 0.0, + "step": 5955 + }, + { + "epoch": 2.8028235294117647, + "grad_norm": 9.976434614399178e-06, + "learning_rate": 0.0002104359653570494, + "loss": 0.0, + "step": 5956 + }, + { + "epoch": 2.803294117647059, + "grad_norm": 0.08981086641806509, + "learning_rate": 0.00021039584210321699, + "loss": 0.0423, + "step": 5957 + }, + { + "epoch": 2.8037647058823527, + "grad_norm": 0.10541968045646087, + "learning_rate": 0.00021035571369128341, + "loss": 0.0426, + "step": 5958 + }, + { + "epoch": 2.804235294117647, + "grad_norm": 0.10854895444502756, + "learning_rate": 0.00021031558012467582, + "loss": 0.061, + "step": 5959 + }, + { + "epoch": 2.804705882352941, + "grad_norm": 0.06712116617904601, + "learning_rate": 0.00021027544140682188, + "loss": 0.0281, + "step": 5960 + }, + { + "epoch": 2.8051764705882354, + "grad_norm": 0.03789227675367984, + "learning_rate": 0.00021023529754114956, + "loss": 0.0117, + "step": 5961 + }, + { + "epoch": 2.8056470588235296, + "grad_norm": 0.00010978654001930486, + "learning_rate": 0.0002101951485310874, + "loss": 0.0, + "step": 5962 + }, + { + "epoch": 2.8061176470588234, + "grad_norm": 0.07652369883012314, + "learning_rate": 0.0002101549943800643, + "loss": 0.0386, + "step": 5963 + }, + { + "epoch": 2.8065882352941176, + "grad_norm": 0.11610550135462704, + "learning_rate": 0.00021011483509150963, + "loss": 0.0891, + "step": 5964 + }, + { + "epoch": 2.807058823529412, + "grad_norm": 0.23685981989138755, + "learning_rate": 0.00021007467066885314, + "loss": 0.1368, + "step": 5965 + }, + { + "epoch": 2.8075294117647056, + "grad_norm": 0.04276926570872577, + "learning_rate": 0.00021003450111552514, + "loss": 0.0162, + "step": 5966 + }, + { + "epoch": 2.808, + "grad_norm": 0.0724453699360937, + "learning_rate": 0.00020999432643495626, + "loss": 0.0413, + "step": 5967 + }, + { + "epoch": 2.808470588235294, + "grad_norm": 0.05898185424630002, + "learning_rate": 0.00020995414663057768, + "loss": 0.0223, + "step": 5968 + }, + { + "epoch": 2.8089411764705883, + "grad_norm": 0.10018063473381429, + "learning_rate": 0.00020991396170582088, + "loss": 0.0805, + "step": 5969 + }, + { + "epoch": 2.8094117647058825, + "grad_norm": 0.24245470478980977, + "learning_rate": 0.00020987377166411785, + "loss": 0.1403, + "step": 5970 + }, + { + "epoch": 2.8098823529411767, + "grad_norm": 0.07747117776624701, + "learning_rate": 0.00020983357650890106, + "loss": 0.0292, + "step": 5971 + }, + { + "epoch": 2.8103529411764705, + "grad_norm": 0.07956958407918893, + "learning_rate": 0.00020979337624360336, + "loss": 0.054, + "step": 5972 + }, + { + "epoch": 2.8108235294117647, + "grad_norm": 0.07905284352876456, + "learning_rate": 0.000209753170871658, + "loss": 0.0355, + "step": 5973 + }, + { + "epoch": 2.811294117647059, + "grad_norm": 0.09773510383971981, + "learning_rate": 0.00020971296039649882, + "loss": 0.0287, + "step": 5974 + }, + { + "epoch": 2.8117647058823527, + "grad_norm": 4.939403963338431e-06, + "learning_rate": 0.00020967274482155992, + "loss": 0.0, + "step": 5975 + }, + { + "epoch": 2.812235294117647, + "grad_norm": 0.0872336543368063, + "learning_rate": 0.00020963252415027588, + "loss": 0.0384, + "step": 5976 + }, + { + "epoch": 2.812705882352941, + "grad_norm": 0.058488927130588675, + "learning_rate": 0.0002095922983860818, + "loss": 0.0232, + "step": 5977 + }, + { + "epoch": 2.8131764705882354, + "grad_norm": 0.11771224578579165, + "learning_rate": 0.00020955206753241315, + "loss": 0.0721, + "step": 5978 + }, + { + "epoch": 2.8136470588235296, + "grad_norm": 3.1203170505895292e-06, + "learning_rate": 0.00020951183159270586, + "loss": 0.0, + "step": 5979 + }, + { + "epoch": 2.8141176470588234, + "grad_norm": 0.08183711753717689, + "learning_rate": 0.0002094715905703962, + "loss": 0.0426, + "step": 5980 + }, + { + "epoch": 2.8145882352941176, + "grad_norm": 0.15156705356249148, + "learning_rate": 0.00020943134446892102, + "loss": 0.082, + "step": 5981 + }, + { + "epoch": 2.815058823529412, + "grad_norm": 0.0736022212579346, + "learning_rate": 0.00020939109329171752, + "loss": 0.0422, + "step": 5982 + }, + { + "epoch": 2.8155294117647056, + "grad_norm": 0.08426272696579833, + "learning_rate": 0.00020935083704222331, + "loss": 0.0346, + "step": 5983 + }, + { + "epoch": 2.816, + "grad_norm": 0.04688865142934053, + "learning_rate": 0.00020931057572387655, + "loss": 0.018, + "step": 5984 + }, + { + "epoch": 2.816470588235294, + "grad_norm": 2.121347316278592e-05, + "learning_rate": 0.00020927030934011568, + "loss": 0.0, + "step": 5985 + }, + { + "epoch": 2.8169411764705883, + "grad_norm": 0.07768244325862148, + "learning_rate": 0.00020923003789437967, + "loss": 0.0337, + "step": 5986 + }, + { + "epoch": 2.8174117647058825, + "grad_norm": 2.87569336449806e-06, + "learning_rate": 0.00020918976139010784, + "loss": 0.0, + "step": 5987 + }, + { + "epoch": 2.8178823529411767, + "grad_norm": 0.16213417621346965, + "learning_rate": 0.00020914947983074016, + "loss": 0.0779, + "step": 5988 + }, + { + "epoch": 2.8183529411764705, + "grad_norm": 0.09126848176562642, + "learning_rate": 0.00020910919321971675, + "loss": 0.0523, + "step": 5989 + }, + { + "epoch": 2.8188235294117647, + "grad_norm": 0.17019023272223263, + "learning_rate": 0.00020906890156047826, + "loss": 0.098, + "step": 5990 + }, + { + "epoch": 2.819294117647059, + "grad_norm": 0.06540995278449817, + "learning_rate": 0.0002090286048564659, + "loss": 0.0247, + "step": 5991 + }, + { + "epoch": 2.8197647058823527, + "grad_norm": 0.07665928788872953, + "learning_rate": 0.00020898830311112118, + "loss": 0.0387, + "step": 5992 + }, + { + "epoch": 2.820235294117647, + "grad_norm": 0.07853317667980668, + "learning_rate": 0.00020894799632788596, + "loss": 0.0508, + "step": 5993 + }, + { + "epoch": 2.820705882352941, + "grad_norm": 0.08097794843439039, + "learning_rate": 0.0002089076845102028, + "loss": 0.0451, + "step": 5994 + }, + { + "epoch": 2.8211764705882354, + "grad_norm": 0.10583555402870864, + "learning_rate": 0.00020886736766151438, + "loss": 0.0497, + "step": 5995 + }, + { + "epoch": 2.8216470588235296, + "grad_norm": 0.06934799401947472, + "learning_rate": 0.00020882704578526406, + "loss": 0.0364, + "step": 5996 + }, + { + "epoch": 2.8221176470588234, + "grad_norm": 0.14638001722208646, + "learning_rate": 0.0002087867188848954, + "loss": 0.0699, + "step": 5997 + }, + { + "epoch": 2.8225882352941176, + "grad_norm": 0.12850271684000625, + "learning_rate": 0.0002087463869638527, + "loss": 0.0865, + "step": 5998 + }, + { + "epoch": 2.823058823529412, + "grad_norm": 7.388354617001932e-07, + "learning_rate": 0.00020870605002558038, + "loss": 0.0, + "step": 5999 + }, + { + "epoch": 2.8235294117647056, + "grad_norm": 0.05109045369693018, + "learning_rate": 0.00020866570807352337, + "loss": 0.0214, + "step": 6000 + }, + { + "epoch": 2.824, + "grad_norm": 0.17812494834758988, + "learning_rate": 0.0002086253611111272, + "loss": 0.0743, + "step": 6001 + }, + { + "epoch": 2.824470588235294, + "grad_norm": 3.455246711182517e-05, + "learning_rate": 0.00020858500914183763, + "loss": 0.0, + "step": 6002 + }, + { + "epoch": 2.8249411764705883, + "grad_norm": 0.04957144194978191, + "learning_rate": 0.0002085446521691009, + "loss": 0.0234, + "step": 6003 + }, + { + "epoch": 2.8254117647058825, + "grad_norm": 5.0698909566182566e-05, + "learning_rate": 0.00020850429019636372, + "loss": 0.0, + "step": 6004 + }, + { + "epoch": 2.8258823529411767, + "grad_norm": 0.06859886851974234, + "learning_rate": 0.00020846392322707323, + "loss": 0.0288, + "step": 6005 + }, + { + "epoch": 2.8263529411764705, + "grad_norm": 0.047351939187157174, + "learning_rate": 0.00020842355126467693, + "loss": 0.0134, + "step": 6006 + }, + { + "epoch": 2.8268235294117647, + "grad_norm": 0.11525081522134753, + "learning_rate": 0.00020838317431262274, + "loss": 0.0859, + "step": 6007 + }, + { + "epoch": 2.827294117647059, + "grad_norm": 0.163576565062311, + "learning_rate": 0.0002083427923743591, + "loss": 0.1444, + "step": 6008 + }, + { + "epoch": 2.8277647058823527, + "grad_norm": 0.11055074427173946, + "learning_rate": 0.00020830240545333489, + "loss": 0.0589, + "step": 6009 + }, + { + "epoch": 2.828235294117647, + "grad_norm": 0.056778264519644336, + "learning_rate": 0.00020826201355299918, + "loss": 0.0257, + "step": 6010 + }, + { + "epoch": 2.828705882352941, + "grad_norm": 0.15124073792253517, + "learning_rate": 0.0002082216166768018, + "loss": 0.0843, + "step": 6011 + }, + { + "epoch": 2.8291764705882354, + "grad_norm": 0.04524397929980303, + "learning_rate": 0.00020818121482819277, + "loss": 0.0164, + "step": 6012 + }, + { + "epoch": 2.8296470588235296, + "grad_norm": 0.10362079958195705, + "learning_rate": 0.00020814080801062263, + "loss": 0.0629, + "step": 6013 + }, + { + "epoch": 2.8301176470588234, + "grad_norm": 0.03339809773817875, + "learning_rate": 0.00020810039622754228, + "loss": 0.0069, + "step": 6014 + }, + { + "epoch": 2.8305882352941176, + "grad_norm": 0.03940683164377656, + "learning_rate": 0.00020805997948240314, + "loss": 0.0126, + "step": 6015 + }, + { + "epoch": 2.831058823529412, + "grad_norm": 1.331976448024255e-05, + "learning_rate": 0.00020801955777865694, + "loss": 0.0, + "step": 6016 + }, + { + "epoch": 2.8315294117647056, + "grad_norm": 0.16816365655598142, + "learning_rate": 0.00020797913111975593, + "loss": 0.11, + "step": 6017 + }, + { + "epoch": 2.832, + "grad_norm": 0.10827649989416507, + "learning_rate": 0.00020793869950915276, + "loss": 0.0514, + "step": 6018 + }, + { + "epoch": 2.832470588235294, + "grad_norm": 3.777794082614477e-05, + "learning_rate": 0.00020789826295030046, + "loss": 0.0, + "step": 6019 + }, + { + "epoch": 2.8329411764705883, + "grad_norm": 0.09179326651381348, + "learning_rate": 0.0002078578214466525, + "loss": 0.049, + "step": 6020 + }, + { + "epoch": 2.8334117647058825, + "grad_norm": 0.09117370799019874, + "learning_rate": 0.00020781737500166279, + "loss": 0.0423, + "step": 6021 + }, + { + "epoch": 2.8338823529411767, + "grad_norm": 0.04275565577467268, + "learning_rate": 0.00020777692361878569, + "loss": 0.0146, + "step": 6022 + }, + { + "epoch": 2.8343529411764705, + "grad_norm": 0.041543873699498385, + "learning_rate": 0.0002077364673014759, + "loss": 0.0109, + "step": 6023 + }, + { + "epoch": 2.8348235294117647, + "grad_norm": 0.14724791401579423, + "learning_rate": 0.00020769600605318857, + "loss": 0.0962, + "step": 6024 + }, + { + "epoch": 2.835294117647059, + "grad_norm": 0.08671434747307907, + "learning_rate": 0.0002076555398773794, + "loss": 0.0419, + "step": 6025 + }, + { + "epoch": 2.8357647058823527, + "grad_norm": 0.08342392042349854, + "learning_rate": 0.00020761506877750428, + "loss": 0.0407, + "step": 6026 + }, + { + "epoch": 2.836235294117647, + "grad_norm": 0.0006028353462007786, + "learning_rate": 0.00020757459275701966, + "loss": 0.0, + "step": 6027 + }, + { + "epoch": 2.836705882352941, + "grad_norm": 0.036027354370148895, + "learning_rate": 0.0002075341118193825, + "loss": 0.0066, + "step": 6028 + }, + { + "epoch": 2.8371764705882354, + "grad_norm": 0.10035155931302801, + "learning_rate": 0.00020749362596804994, + "loss": 0.0571, + "step": 6029 + }, + { + "epoch": 2.8376470588235296, + "grad_norm": 0.08642583886388239, + "learning_rate": 0.00020745313520647971, + "loss": 0.0309, + "step": 6030 + }, + { + "epoch": 2.8381176470588234, + "grad_norm": 3.8777299948424585e-06, + "learning_rate": 0.00020741263953812992, + "loss": 0.0, + "step": 6031 + }, + { + "epoch": 2.8385882352941176, + "grad_norm": 1.1839923567915955e-05, + "learning_rate": 0.0002073721389664591, + "loss": 0.0, + "step": 6032 + }, + { + "epoch": 2.839058823529412, + "grad_norm": 0.08741735105772845, + "learning_rate": 0.00020733163349492625, + "loss": 0.0397, + "step": 6033 + }, + { + "epoch": 2.8395294117647056, + "grad_norm": 0.08951911404156342, + "learning_rate": 0.00020729112312699066, + "loss": 0.0335, + "step": 6034 + }, + { + "epoch": 2.84, + "grad_norm": 0.15638528640360438, + "learning_rate": 0.00020725060786611212, + "loss": 0.088, + "step": 6035 + }, + { + "epoch": 2.840470588235294, + "grad_norm": 0.10874032150793773, + "learning_rate": 0.0002072100877157509, + "loss": 0.0643, + "step": 6036 + }, + { + "epoch": 2.8409411764705883, + "grad_norm": 2.5704254457172113e-05, + "learning_rate": 0.00020716956267936759, + "loss": 0.0, + "step": 6037 + }, + { + "epoch": 2.8414117647058825, + "grad_norm": 6.8901897104783465e-06, + "learning_rate": 0.0002071290327604231, + "loss": 0.0, + "step": 6038 + }, + { + "epoch": 2.8418823529411767, + "grad_norm": 0.13639761502534567, + "learning_rate": 0.00020708849796237914, + "loss": 0.0577, + "step": 6039 + }, + { + "epoch": 2.8423529411764705, + "grad_norm": 0.06993030779343866, + "learning_rate": 0.00020704795828869736, + "loss": 0.0283, + "step": 6040 + }, + { + "epoch": 2.8428235294117647, + "grad_norm": 5.7505473066077535e-05, + "learning_rate": 0.00020700741374284014, + "loss": 0.0, + "step": 6041 + }, + { + "epoch": 2.843294117647059, + "grad_norm": 0.1516993370144439, + "learning_rate": 0.00020696686432827017, + "loss": 0.0956, + "step": 6042 + }, + { + "epoch": 2.8437647058823528, + "grad_norm": 2.0273413353788237e-05, + "learning_rate": 0.00020692631004845055, + "loss": 0.0, + "step": 6043 + }, + { + "epoch": 2.844235294117647, + "grad_norm": 0.059320837786101886, + "learning_rate": 0.00020688575090684484, + "loss": 0.0191, + "step": 6044 + }, + { + "epoch": 2.844705882352941, + "grad_norm": 0.17838431725507248, + "learning_rate": 0.00020684518690691696, + "loss": 0.07, + "step": 6045 + }, + { + "epoch": 2.8451764705882354, + "grad_norm": 0.09842848055011674, + "learning_rate": 0.00020680461805213133, + "loss": 0.0407, + "step": 6046 + }, + { + "epoch": 2.8456470588235296, + "grad_norm": 0.08660367036332153, + "learning_rate": 0.00020676404434595265, + "loss": 0.0405, + "step": 6047 + }, + { + "epoch": 2.8461176470588234, + "grad_norm": 0.054645806163418666, + "learning_rate": 0.00020672346579184617, + "loss": 0.0149, + "step": 6048 + }, + { + "epoch": 2.8465882352941176, + "grad_norm": 1.007010136711242e-06, + "learning_rate": 0.00020668288239327753, + "loss": 0.0, + "step": 6049 + }, + { + "epoch": 2.847058823529412, + "grad_norm": 0.1212071597702537, + "learning_rate": 0.00020664229415371266, + "loss": 0.1104, + "step": 6050 + }, + { + "epoch": 2.8475294117647056, + "grad_norm": 0.05256983420411666, + "learning_rate": 0.00020660170107661803, + "loss": 0.0075, + "step": 6051 + }, + { + "epoch": 2.848, + "grad_norm": 0.0672238270212575, + "learning_rate": 0.00020656110316546054, + "loss": 0.0327, + "step": 6052 + }, + { + "epoch": 2.848470588235294, + "grad_norm": 0.016467796084443747, + "learning_rate": 0.00020652050042370743, + "loss": 0.0024, + "step": 6053 + }, + { + "epoch": 2.8489411764705883, + "grad_norm": 0.05449116391547746, + "learning_rate": 0.0002064798928548263, + "loss": 0.0158, + "step": 6054 + }, + { + "epoch": 2.8494117647058825, + "grad_norm": 0.07588302033954081, + "learning_rate": 0.00020643928046228525, + "loss": 0.0324, + "step": 6055 + }, + { + "epoch": 2.8498823529411763, + "grad_norm": 0.06381968268796946, + "learning_rate": 0.00020639866324955285, + "loss": 0.0255, + "step": 6056 + }, + { + "epoch": 2.8503529411764705, + "grad_norm": 0.02379005262125137, + "learning_rate": 0.00020635804122009802, + "loss": 0.0058, + "step": 6057 + }, + { + "epoch": 2.8508235294117648, + "grad_norm": 0.05718370207059807, + "learning_rate": 0.00020631741437738998, + "loss": 0.0203, + "step": 6058 + }, + { + "epoch": 2.851294117647059, + "grad_norm": 0.13601721739172098, + "learning_rate": 0.00020627678272489854, + "loss": 0.0958, + "step": 6059 + }, + { + "epoch": 2.8517647058823528, + "grad_norm": 0.11432227101877086, + "learning_rate": 0.00020623614626609385, + "loss": 0.0561, + "step": 6060 + }, + { + "epoch": 2.852235294117647, + "grad_norm": 3.4855880059162375e-07, + "learning_rate": 0.0002061955050044464, + "loss": 0.0, + "step": 6061 + }, + { + "epoch": 2.852705882352941, + "grad_norm": 0.08710605690642192, + "learning_rate": 0.00020615485894342716, + "loss": 0.0361, + "step": 6062 + }, + { + "epoch": 2.8531764705882354, + "grad_norm": 3.271176078055314e-06, + "learning_rate": 0.00020611420808650758, + "loss": 0.0, + "step": 6063 + }, + { + "epoch": 2.8536470588235296, + "grad_norm": 2.829337943257451e-05, + "learning_rate": 0.00020607355243715936, + "loss": 0.0, + "step": 6064 + }, + { + "epoch": 2.8541176470588234, + "grad_norm": 0.08043347891505474, + "learning_rate": 0.00020603289199885473, + "loss": 0.0354, + "step": 6065 + }, + { + "epoch": 2.8545882352941176, + "grad_norm": 0.09290847113071556, + "learning_rate": 0.00020599222677506628, + "loss": 0.061, + "step": 6066 + }, + { + "epoch": 2.855058823529412, + "grad_norm": 9.542767168404459e-05, + "learning_rate": 0.00020595155676926704, + "loss": 0.0, + "step": 6067 + }, + { + "epoch": 2.8555294117647056, + "grad_norm": 0.06028545549020818, + "learning_rate": 0.0002059108819849304, + "loss": 0.0203, + "step": 6068 + }, + { + "epoch": 2.856, + "grad_norm": 0.10480321133698743, + "learning_rate": 0.0002058702024255302, + "loss": 0.0568, + "step": 6069 + }, + { + "epoch": 2.856470588235294, + "grad_norm": 9.171222945602541e-07, + "learning_rate": 0.00020582951809454067, + "loss": 0.0, + "step": 6070 + }, + { + "epoch": 2.8569411764705883, + "grad_norm": 5.834746547775646e-06, + "learning_rate": 0.00020578882899543644, + "loss": 0.0, + "step": 6071 + }, + { + "epoch": 2.8574117647058825, + "grad_norm": 0.10851435460935534, + "learning_rate": 0.0002057481351316926, + "loss": 0.0695, + "step": 6072 + }, + { + "epoch": 2.8578823529411763, + "grad_norm": 0.05288409150370726, + "learning_rate": 0.00020570743650678454, + "loss": 0.0216, + "step": 6073 + }, + { + "epoch": 2.8583529411764705, + "grad_norm": 0.12011040531070641, + "learning_rate": 0.00020566673312418817, + "loss": 0.0731, + "step": 6074 + }, + { + "epoch": 2.8588235294117648, + "grad_norm": 0.06137355821742023, + "learning_rate": 0.00020562602498737975, + "loss": 0.027, + "step": 6075 + }, + { + "epoch": 2.859294117647059, + "grad_norm": 0.07534089227835575, + "learning_rate": 0.00020558531209983593, + "loss": 0.0291, + "step": 6076 + }, + { + "epoch": 2.8597647058823528, + "grad_norm": 0.08089667011438352, + "learning_rate": 0.00020554459446503386, + "loss": 0.0503, + "step": 6077 + }, + { + "epoch": 2.860235294117647, + "grad_norm": 0.05752182027787807, + "learning_rate": 0.00020550387208645093, + "loss": 0.022, + "step": 6078 + }, + { + "epoch": 2.860705882352941, + "grad_norm": 0.12599692810302754, + "learning_rate": 0.0002054631449675651, + "loss": 0.0482, + "step": 6079 + }, + { + "epoch": 2.8611764705882354, + "grad_norm": 0.21101005067801268, + "learning_rate": 0.00020542241311185467, + "loss": 0.1166, + "step": 6080 + }, + { + "epoch": 2.8616470588235297, + "grad_norm": 0.060599899118610905, + "learning_rate": 0.0002053816765227983, + "loss": 0.0187, + "step": 6081 + }, + { + "epoch": 2.8621176470588234, + "grad_norm": 0.07635058198900897, + "learning_rate": 0.0002053409352038751, + "loss": 0.04, + "step": 6082 + }, + { + "epoch": 2.8625882352941177, + "grad_norm": 0.08566145820422595, + "learning_rate": 0.00020530018915856457, + "loss": 0.0492, + "step": 6083 + }, + { + "epoch": 2.863058823529412, + "grad_norm": 0.06168008379884171, + "learning_rate": 0.00020525943839034668, + "loss": 0.0248, + "step": 6084 + }, + { + "epoch": 2.8635294117647057, + "grad_norm": 0.07467609601833454, + "learning_rate": 0.0002052186829027017, + "loss": 0.0335, + "step": 6085 + }, + { + "epoch": 2.864, + "grad_norm": 0.10825061789852906, + "learning_rate": 0.00020517792269911033, + "loss": 0.0636, + "step": 6086 + }, + { + "epoch": 2.864470588235294, + "grad_norm": 0.12012465589525158, + "learning_rate": 0.0002051371577830537, + "loss": 0.0864, + "step": 6087 + }, + { + "epoch": 2.8649411764705883, + "grad_norm": 1.0295833016269998e-05, + "learning_rate": 0.00020509638815801342, + "loss": 0.0, + "step": 6088 + }, + { + "epoch": 2.8654117647058825, + "grad_norm": 0.13813928652653668, + "learning_rate": 0.00020505561382747127, + "loss": 0.0806, + "step": 6089 + }, + { + "epoch": 2.8658823529411763, + "grad_norm": 0.047198628661735756, + "learning_rate": 0.00020501483479490968, + "loss": 0.0134, + "step": 6090 + }, + { + "epoch": 2.8663529411764705, + "grad_norm": 0.11863251475263896, + "learning_rate": 0.00020497405106381138, + "loss": 0.0751, + "step": 6091 + }, + { + "epoch": 2.8668235294117648, + "grad_norm": 0.08878673931347596, + "learning_rate": 0.0002049332626376594, + "loss": 0.0544, + "step": 6092 + }, + { + "epoch": 2.8672941176470585, + "grad_norm": 0.09753243154576029, + "learning_rate": 0.00020489246951993742, + "loss": 0.0401, + "step": 6093 + }, + { + "epoch": 2.8677647058823528, + "grad_norm": 0.08218577122898553, + "learning_rate": 0.00020485167171412926, + "loss": 0.0257, + "step": 6094 + }, + { + "epoch": 2.868235294117647, + "grad_norm": 0.05026802908741613, + "learning_rate": 0.00020481086922371935, + "loss": 0.0078, + "step": 6095 + }, + { + "epoch": 2.868705882352941, + "grad_norm": 0.10579053134889982, + "learning_rate": 0.00020477006205219228, + "loss": 0.0499, + "step": 6096 + }, + { + "epoch": 2.8691764705882354, + "grad_norm": 2.446743041351058e-06, + "learning_rate": 0.0002047292502030333, + "loss": 0.0, + "step": 6097 + }, + { + "epoch": 2.8696470588235297, + "grad_norm": 3.923976826244882e-06, + "learning_rate": 0.00020468843367972786, + "loss": 0.0, + "step": 6098 + }, + { + "epoch": 2.8701176470588234, + "grad_norm": 0.13448454053261757, + "learning_rate": 0.00020464761248576203, + "loss": 0.0784, + "step": 6099 + }, + { + "epoch": 2.8705882352941177, + "grad_norm": 0.10807334139246433, + "learning_rate": 0.00020460678662462194, + "loss": 0.0825, + "step": 6100 + }, + { + "epoch": 2.871058823529412, + "grad_norm": 0.1183192232887876, + "learning_rate": 0.0002045659560997945, + "loss": 0.0752, + "step": 6101 + }, + { + "epoch": 2.8715294117647057, + "grad_norm": 0.08454572638053819, + "learning_rate": 0.00020452512091476678, + "loss": 0.0495, + "step": 6102 + }, + { + "epoch": 2.872, + "grad_norm": 0.058897209067057435, + "learning_rate": 0.0002044842810730262, + "loss": 0.0194, + "step": 6103 + }, + { + "epoch": 2.872470588235294, + "grad_norm": 0.2962225817752639, + "learning_rate": 0.00020444343657806087, + "loss": 0.1254, + "step": 6104 + }, + { + "epoch": 2.8729411764705883, + "grad_norm": 1.9945995360865205e-06, + "learning_rate": 0.000204402587433359, + "loss": 0.0, + "step": 6105 + }, + { + "epoch": 2.8734117647058826, + "grad_norm": 0.039894894525136754, + "learning_rate": 0.0002043617336424093, + "loss": 0.0106, + "step": 6106 + }, + { + "epoch": 2.8738823529411763, + "grad_norm": 0.07090332597541472, + "learning_rate": 0.00020432087520870086, + "loss": 0.0205, + "step": 6107 + }, + { + "epoch": 2.8743529411764706, + "grad_norm": 8.282194118258117e-06, + "learning_rate": 0.0002042800121357233, + "loss": 0.0, + "step": 6108 + }, + { + "epoch": 2.8748235294117648, + "grad_norm": 0.11734875775684712, + "learning_rate": 0.00020423914442696642, + "loss": 0.044, + "step": 6109 + }, + { + "epoch": 2.8752941176470586, + "grad_norm": 0.06553133525229415, + "learning_rate": 0.00020419827208592058, + "loss": 0.0353, + "step": 6110 + }, + { + "epoch": 2.875764705882353, + "grad_norm": 0.08340616491082063, + "learning_rate": 0.00020415739511607653, + "loss": 0.0462, + "step": 6111 + }, + { + "epoch": 2.876235294117647, + "grad_norm": 0.050940268313194495, + "learning_rate": 0.0002041165135209252, + "loss": 0.0119, + "step": 6112 + }, + { + "epoch": 2.8767058823529412, + "grad_norm": 0.08092943444850455, + "learning_rate": 0.0002040756273039582, + "loss": 0.0602, + "step": 6113 + }, + { + "epoch": 2.8771764705882354, + "grad_norm": 0.08358424220506941, + "learning_rate": 0.0002040347364686674, + "loss": 0.034, + "step": 6114 + }, + { + "epoch": 2.8776470588235297, + "grad_norm": 0.04381778257707931, + "learning_rate": 0.00020399384101854506, + "loss": 0.0115, + "step": 6115 + }, + { + "epoch": 2.8781176470588234, + "grad_norm": 0.09782721099867818, + "learning_rate": 0.00020395294095708387, + "loss": 0.0578, + "step": 6116 + }, + { + "epoch": 2.8785882352941177, + "grad_norm": 0.09343407544936463, + "learning_rate": 0.00020391203628777687, + "loss": 0.0487, + "step": 6117 + }, + { + "epoch": 2.879058823529412, + "grad_norm": 0.07817819651253981, + "learning_rate": 0.00020387112701411754, + "loss": 0.0295, + "step": 6118 + }, + { + "epoch": 2.8795294117647057, + "grad_norm": 0.08202053841298598, + "learning_rate": 0.00020383021313959972, + "loss": 0.0434, + "step": 6119 + }, + { + "epoch": 2.88, + "grad_norm": 0.06754883846496657, + "learning_rate": 0.00020378929466771767, + "loss": 0.0224, + "step": 6120 + }, + { + "epoch": 2.880470588235294, + "grad_norm": 0.1194814738270146, + "learning_rate": 0.000203748371601966, + "loss": 0.064, + "step": 6121 + }, + { + "epoch": 2.8809411764705883, + "grad_norm": 0.09786886190644509, + "learning_rate": 0.0002037074439458398, + "loss": 0.0558, + "step": 6122 + }, + { + "epoch": 2.8814117647058826, + "grad_norm": 0.08774579696349628, + "learning_rate": 0.0002036665117028344, + "loss": 0.0433, + "step": 6123 + }, + { + "epoch": 2.8818823529411763, + "grad_norm": 0.038792416932111946, + "learning_rate": 0.0002036255748764457, + "loss": 0.0118, + "step": 6124 + }, + { + "epoch": 2.8823529411764706, + "grad_norm": 0.09815050024838515, + "learning_rate": 0.00020358463347016988, + "loss": 0.0437, + "step": 6125 + }, + { + "epoch": 2.882823529411765, + "grad_norm": 0.12411102413302377, + "learning_rate": 0.00020354368748750357, + "loss": 0.0546, + "step": 6126 + }, + { + "epoch": 2.8832941176470586, + "grad_norm": 0.07601990984682663, + "learning_rate": 0.00020350273693194363, + "loss": 0.0286, + "step": 6127 + }, + { + "epoch": 2.883764705882353, + "grad_norm": 0.08084141493078272, + "learning_rate": 0.00020346178180698758, + "loss": 0.0297, + "step": 6128 + }, + { + "epoch": 2.884235294117647, + "grad_norm": 0.07774984192320529, + "learning_rate": 0.00020342082211613312, + "loss": 0.0377, + "step": 6129 + }, + { + "epoch": 2.8847058823529412, + "grad_norm": 0.04918351175532867, + "learning_rate": 0.00020337985786287842, + "loss": 0.0175, + "step": 6130 + }, + { + "epoch": 2.8851764705882355, + "grad_norm": 0.039559305349090065, + "learning_rate": 0.00020333888905072208, + "loss": 0.0106, + "step": 6131 + }, + { + "epoch": 2.8856470588235297, + "grad_norm": 0.24009395564000233, + "learning_rate": 0.00020329791568316294, + "loss": 0.0541, + "step": 6132 + }, + { + "epoch": 2.8861176470588235, + "grad_norm": 0.10074254795120453, + "learning_rate": 0.00020325693776370037, + "loss": 0.0471, + "step": 6133 + }, + { + "epoch": 2.8865882352941177, + "grad_norm": 0.14090424500909854, + "learning_rate": 0.0002032159552958341, + "loss": 0.0661, + "step": 6134 + }, + { + "epoch": 2.887058823529412, + "grad_norm": 0.10465757997007828, + "learning_rate": 0.00020317496828306427, + "loss": 0.0484, + "step": 6135 + }, + { + "epoch": 2.8875294117647057, + "grad_norm": 0.1396526602831536, + "learning_rate": 0.00020313397672889132, + "loss": 0.0605, + "step": 6136 + }, + { + "epoch": 2.888, + "grad_norm": 0.036297282892843175, + "learning_rate": 0.00020309298063681607, + "loss": 0.0092, + "step": 6137 + }, + { + "epoch": 2.888470588235294, + "grad_norm": 0.09637456708169101, + "learning_rate": 0.00020305198001033988, + "loss": 0.038, + "step": 6138 + }, + { + "epoch": 2.8889411764705883, + "grad_norm": 0.25823612858149125, + "learning_rate": 0.00020301097485296443, + "loss": 0.0997, + "step": 6139 + }, + { + "epoch": 2.8894117647058826, + "grad_norm": 0.059635778005357576, + "learning_rate": 0.00020296996516819163, + "loss": 0.0282, + "step": 6140 + }, + { + "epoch": 2.8898823529411763, + "grad_norm": 8.508409116645164e-06, + "learning_rate": 0.00020292895095952402, + "loss": 0.0, + "step": 6141 + }, + { + "epoch": 2.8903529411764706, + "grad_norm": 0.00010732018061523111, + "learning_rate": 0.0002028879322304644, + "loss": 0.0, + "step": 6142 + }, + { + "epoch": 2.890823529411765, + "grad_norm": 0.07904160106656179, + "learning_rate": 0.00020284690898451593, + "loss": 0.0478, + "step": 6143 + }, + { + "epoch": 2.8912941176470586, + "grad_norm": 0.09567115677468853, + "learning_rate": 0.00020280588122518223, + "loss": 0.0674, + "step": 6144 + }, + { + "epoch": 2.891764705882353, + "grad_norm": 0.06828115067267214, + "learning_rate": 0.00020276484895596726, + "loss": 0.0288, + "step": 6145 + }, + { + "epoch": 2.892235294117647, + "grad_norm": 0.07404366353077263, + "learning_rate": 0.00020272381218037542, + "loss": 0.026, + "step": 6146 + }, + { + "epoch": 2.8927058823529412, + "grad_norm": 0.062480209272902446, + "learning_rate": 0.00020268277090191133, + "loss": 0.0203, + "step": 6147 + }, + { + "epoch": 2.8931764705882355, + "grad_norm": 0.11564276278389188, + "learning_rate": 0.00020264172512408026, + "loss": 0.0615, + "step": 6148 + }, + { + "epoch": 2.8936470588235297, + "grad_norm": 0.0004820774694641735, + "learning_rate": 0.00020260067485038767, + "loss": 0.0, + "step": 6149 + }, + { + "epoch": 2.8941176470588235, + "grad_norm": 0.09732706098716222, + "learning_rate": 0.0002025596200843394, + "loss": 0.0391, + "step": 6150 + }, + { + "epoch": 2.8945882352941177, + "grad_norm": 2.2276676980828524e-05, + "learning_rate": 0.00020251856082944178, + "loss": 0.0, + "step": 6151 + }, + { + "epoch": 2.895058823529412, + "grad_norm": 0.0722215880317763, + "learning_rate": 0.00020247749708920146, + "loss": 0.0187, + "step": 6152 + }, + { + "epoch": 2.8955294117647057, + "grad_norm": 0.021424369111168712, + "learning_rate": 0.00020243642886712553, + "loss": 0.0041, + "step": 6153 + }, + { + "epoch": 2.896, + "grad_norm": 0.062489956270434896, + "learning_rate": 0.0002023953561667213, + "loss": 0.0203, + "step": 6154 + }, + { + "epoch": 2.896470588235294, + "grad_norm": 0.0008662739085941643, + "learning_rate": 0.0002023542789914967, + "loss": 0.0, + "step": 6155 + }, + { + "epoch": 2.8969411764705884, + "grad_norm": 3.399217665889974e-05, + "learning_rate": 0.0002023131973449599, + "loss": 0.0, + "step": 6156 + }, + { + "epoch": 2.8974117647058826, + "grad_norm": 0.054997348779533634, + "learning_rate": 0.00020227211123061942, + "loss": 0.0191, + "step": 6157 + }, + { + "epoch": 2.8978823529411764, + "grad_norm": 0.06913765258823783, + "learning_rate": 0.00020223102065198426, + "loss": 0.0317, + "step": 6158 + }, + { + "epoch": 2.8983529411764706, + "grad_norm": 0.11470518623431547, + "learning_rate": 0.00020218992561256375, + "loss": 0.0565, + "step": 6159 + }, + { + "epoch": 2.898823529411765, + "grad_norm": 0.06576917583159533, + "learning_rate": 0.0002021488261158676, + "loss": 0.033, + "step": 6160 + }, + { + "epoch": 2.8992941176470586, + "grad_norm": 0.1237069514475168, + "learning_rate": 0.0002021077221654059, + "loss": 0.0782, + "step": 6161 + }, + { + "epoch": 2.899764705882353, + "grad_norm": 0.07861069458283855, + "learning_rate": 0.00020206661376468912, + "loss": 0.0328, + "step": 6162 + }, + { + "epoch": 2.900235294117647, + "grad_norm": 0.20240407169352403, + "learning_rate": 0.00020202550091722818, + "loss": 0.0762, + "step": 6163 + }, + { + "epoch": 2.9007058823529412, + "grad_norm": 0.08301527482851292, + "learning_rate": 0.00020198438362653426, + "loss": 0.0354, + "step": 6164 + }, + { + "epoch": 2.9011764705882355, + "grad_norm": 0.11115528436963759, + "learning_rate": 0.00020194326189611897, + "loss": 0.0774, + "step": 6165 + }, + { + "epoch": 2.9016470588235292, + "grad_norm": 0.04846109101076087, + "learning_rate": 0.00020190213572949435, + "loss": 0.0188, + "step": 6166 + }, + { + "epoch": 2.9021176470588235, + "grad_norm": 6.221288478651775e-06, + "learning_rate": 0.0002018610051301728, + "loss": 0.0, + "step": 6167 + }, + { + "epoch": 2.9025882352941177, + "grad_norm": 0.13129226129014493, + "learning_rate": 0.00020181987010166699, + "loss": 0.0713, + "step": 6168 + }, + { + "epoch": 2.903058823529412, + "grad_norm": 2.404070701532256e-05, + "learning_rate": 0.0002017787306474901, + "loss": 0.0, + "step": 6169 + }, + { + "epoch": 2.9035294117647057, + "grad_norm": 0.04670317329684244, + "learning_rate": 0.00020173758677115568, + "loss": 0.0144, + "step": 6170 + }, + { + "epoch": 2.904, + "grad_norm": 0.2306075365666816, + "learning_rate": 0.00020169643847617753, + "loss": 0.1251, + "step": 6171 + }, + { + "epoch": 2.904470588235294, + "grad_norm": 0.05064580365209066, + "learning_rate": 0.00020165528576607, + "loss": 0.019, + "step": 6172 + }, + { + "epoch": 2.9049411764705884, + "grad_norm": 9.428419769942271e-06, + "learning_rate": 0.0002016141286443477, + "loss": 0.0, + "step": 6173 + }, + { + "epoch": 2.9054117647058826, + "grad_norm": 0.14184886390379275, + "learning_rate": 0.0002015729671145256, + "loss": 0.0919, + "step": 6174 + }, + { + "epoch": 2.9058823529411764, + "grad_norm": 0.1450349492977399, + "learning_rate": 0.0002015318011801192, + "loss": 0.0577, + "step": 6175 + }, + { + "epoch": 2.9063529411764706, + "grad_norm": 0.06583666526850188, + "learning_rate": 0.0002014906308446442, + "loss": 0.0283, + "step": 6176 + }, + { + "epoch": 2.906823529411765, + "grad_norm": 0.15152479829327978, + "learning_rate": 0.0002014494561116168, + "loss": 0.0703, + "step": 6177 + }, + { + "epoch": 2.9072941176470586, + "grad_norm": 0.0915832819431817, + "learning_rate": 0.00020140827698455338, + "loss": 0.0443, + "step": 6178 + }, + { + "epoch": 2.907764705882353, + "grad_norm": 0.06715939635420226, + "learning_rate": 0.00020136709346697107, + "loss": 0.0224, + "step": 6179 + }, + { + "epoch": 2.908235294117647, + "grad_norm": 0.04677734313273479, + "learning_rate": 0.00020132590556238696, + "loss": 0.0173, + "step": 6180 + }, + { + "epoch": 2.9087058823529413, + "grad_norm": 0.1377608055334763, + "learning_rate": 0.00020128471327431877, + "loss": 0.0676, + "step": 6181 + }, + { + "epoch": 2.9091764705882355, + "grad_norm": 0.08318942274766584, + "learning_rate": 0.0002012435166062845, + "loss": 0.0533, + "step": 6182 + }, + { + "epoch": 2.9096470588235293, + "grad_norm": 0.07862572727410254, + "learning_rate": 0.00020120231556180262, + "loss": 0.0272, + "step": 6183 + }, + { + "epoch": 2.9101176470588235, + "grad_norm": 1.1641725237657305e-05, + "learning_rate": 0.00020116111014439178, + "loss": 0.0, + "step": 6184 + }, + { + "epoch": 2.9105882352941177, + "grad_norm": 1.8775707219088632e-05, + "learning_rate": 0.00020111990035757125, + "loss": 0.0, + "step": 6185 + }, + { + "epoch": 2.911058823529412, + "grad_norm": 0.0001458105604385636, + "learning_rate": 0.00020107868620486047, + "loss": 0.0, + "step": 6186 + }, + { + "epoch": 2.9115294117647057, + "grad_norm": 0.07202683484604484, + "learning_rate": 0.0002010374676897793, + "loss": 0.0328, + "step": 6187 + }, + { + "epoch": 2.912, + "grad_norm": 0.07171336260569283, + "learning_rate": 0.00020099624481584807, + "loss": 0.0373, + "step": 6188 + }, + { + "epoch": 2.912470588235294, + "grad_norm": 0.0541188820375572, + "learning_rate": 0.0002009550175865874, + "loss": 0.018, + "step": 6189 + }, + { + "epoch": 2.9129411764705884, + "grad_norm": 0.07714774064579638, + "learning_rate": 0.00020091378600551834, + "loss": 0.0447, + "step": 6190 + }, + { + "epoch": 2.9134117647058826, + "grad_norm": 0.13200765716888657, + "learning_rate": 0.00020087255007616222, + "loss": 0.0824, + "step": 6191 + }, + { + "epoch": 2.9138823529411764, + "grad_norm": 0.04476064227545316, + "learning_rate": 0.00020083130980204074, + "loss": 0.0164, + "step": 6192 + }, + { + "epoch": 2.9143529411764706, + "grad_norm": 0.09601064114860293, + "learning_rate": 0.0002007900651866761, + "loss": 0.05, + "step": 6193 + }, + { + "epoch": 2.914823529411765, + "grad_norm": 0.1616013065556874, + "learning_rate": 0.00020074881623359083, + "loss": 0.0967, + "step": 6194 + }, + { + "epoch": 2.9152941176470586, + "grad_norm": 0.08418674975443334, + "learning_rate": 0.00020070756294630765, + "loss": 0.036, + "step": 6195 + }, + { + "epoch": 2.915764705882353, + "grad_norm": 0.08292591950088357, + "learning_rate": 0.0002006663053283499, + "loss": 0.0352, + "step": 6196 + }, + { + "epoch": 2.916235294117647, + "grad_norm": 0.039622547989876, + "learning_rate": 0.00020062504338324116, + "loss": 0.01, + "step": 6197 + }, + { + "epoch": 2.9167058823529413, + "grad_norm": 0.08536122893842081, + "learning_rate": 0.0002005837771145054, + "loss": 0.0409, + "step": 6198 + }, + { + "epoch": 2.9171764705882355, + "grad_norm": 0.06032755507184028, + "learning_rate": 0.00020054250652566696, + "loss": 0.0284, + "step": 6199 + }, + { + "epoch": 2.9176470588235293, + "grad_norm": 0.06480978850695249, + "learning_rate": 0.0002005012316202506, + "loss": 0.0222, + "step": 6200 + }, + { + "epoch": 2.9181176470588235, + "grad_norm": 0.059228586085032495, + "learning_rate": 0.00020045995240178133, + "loss": 0.0202, + "step": 6201 + }, + { + "epoch": 2.9185882352941177, + "grad_norm": 8.07498832338413e-06, + "learning_rate": 0.00020041866887378457, + "loss": 0.0, + "step": 6202 + }, + { + "epoch": 2.919058823529412, + "grad_norm": 0.23557715424369705, + "learning_rate": 0.0002003773810397862, + "loss": 0.1676, + "step": 6203 + }, + { + "epoch": 2.9195294117647057, + "grad_norm": 0.034538472436265545, + "learning_rate": 0.00020033608890331246, + "loss": 0.0095, + "step": 6204 + }, + { + "epoch": 2.92, + "grad_norm": 0.10909317620521072, + "learning_rate": 0.00020029479246788973, + "loss": 0.0513, + "step": 6205 + }, + { + "epoch": 2.920470588235294, + "grad_norm": 0.10934659820541517, + "learning_rate": 0.00020025349173704508, + "loss": 0.0672, + "step": 6206 + }, + { + "epoch": 2.9209411764705884, + "grad_norm": 0.08447511263909291, + "learning_rate": 0.00020021218671430573, + "loss": 0.0367, + "step": 6207 + }, + { + "epoch": 2.9214117647058826, + "grad_norm": 7.503233388715135e-06, + "learning_rate": 0.00020017087740319936, + "loss": 0.0, + "step": 6208 + }, + { + "epoch": 2.9218823529411764, + "grad_norm": 2.1966426108945047e-05, + "learning_rate": 0.00020012956380725393, + "loss": 0.0, + "step": 6209 + }, + { + "epoch": 2.9223529411764706, + "grad_norm": 0.07303836448904905, + "learning_rate": 0.00020008824592999795, + "loss": 0.0324, + "step": 6210 + }, + { + "epoch": 2.922823529411765, + "grad_norm": 0.03377458583772694, + "learning_rate": 0.00020004692377496004, + "loss": 0.0095, + "step": 6211 + }, + { + "epoch": 2.9232941176470586, + "grad_norm": 0.10045750203769849, + "learning_rate": 0.00020000559734566933, + "loss": 0.0511, + "step": 6212 + }, + { + "epoch": 2.923764705882353, + "grad_norm": 0.12828911243168148, + "learning_rate": 0.00019996426664565538, + "loss": 0.0714, + "step": 6213 + }, + { + "epoch": 2.924235294117647, + "grad_norm": 0.20194177942101793, + "learning_rate": 0.000199922931678448, + "loss": 0.0923, + "step": 6214 + }, + { + "epoch": 2.9247058823529413, + "grad_norm": 0.10422867513811698, + "learning_rate": 0.00019988159244757736, + "loss": 0.0447, + "step": 6215 + }, + { + "epoch": 2.9251764705882355, + "grad_norm": 0.14788819241271844, + "learning_rate": 0.000199840248956574, + "loss": 0.0912, + "step": 6216 + }, + { + "epoch": 2.9256470588235293, + "grad_norm": 0.15678407311667605, + "learning_rate": 0.000199798901208969, + "loss": 0.0582, + "step": 6217 + }, + { + "epoch": 2.9261176470588235, + "grad_norm": 0.11114831285072685, + "learning_rate": 0.0001997575492082935, + "loss": 0.0673, + "step": 6218 + }, + { + "epoch": 2.9265882352941177, + "grad_norm": 0.00013356930862335741, + "learning_rate": 0.0001997161929580793, + "loss": 0.0, + "step": 6219 + }, + { + "epoch": 2.9270588235294115, + "grad_norm": 0.0575337582898756, + "learning_rate": 0.00019967483246185834, + "loss": 0.0216, + "step": 6220 + }, + { + "epoch": 2.9275294117647057, + "grad_norm": 3.571560242089238e-06, + "learning_rate": 0.00019963346772316307, + "loss": 0.0, + "step": 6221 + }, + { + "epoch": 2.928, + "grad_norm": 0.07238160396751424, + "learning_rate": 0.00019959209874552612, + "loss": 0.0209, + "step": 6222 + }, + { + "epoch": 2.928470588235294, + "grad_norm": 0.09317545958878858, + "learning_rate": 0.00019955072553248075, + "loss": 0.0437, + "step": 6223 + }, + { + "epoch": 2.9289411764705884, + "grad_norm": 0.0673933310345762, + "learning_rate": 0.0001995093480875604, + "loss": 0.0308, + "step": 6224 + }, + { + "epoch": 2.9294117647058826, + "grad_norm": 0.10569300315976443, + "learning_rate": 0.00019946796641429884, + "loss": 0.0402, + "step": 6225 + }, + { + "epoch": 2.9298823529411764, + "grad_norm": 1.3630484816452834e-05, + "learning_rate": 0.0001994265805162303, + "loss": 0.0, + "step": 6226 + }, + { + "epoch": 2.9303529411764706, + "grad_norm": 4.932944325391246e-06, + "learning_rate": 0.00019938519039688937, + "loss": 0.0, + "step": 6227 + }, + { + "epoch": 2.930823529411765, + "grad_norm": 0.10680854092260791, + "learning_rate": 0.00019934379605981097, + "loss": 0.0481, + "step": 6228 + }, + { + "epoch": 2.9312941176470586, + "grad_norm": 0.11098425871619444, + "learning_rate": 0.0001993023975085303, + "loss": 0.0577, + "step": 6229 + }, + { + "epoch": 2.931764705882353, + "grad_norm": 0.15467640578183786, + "learning_rate": 0.00019926099474658305, + "loss": 0.0635, + "step": 6230 + }, + { + "epoch": 2.932235294117647, + "grad_norm": 0.0538683170343878, + "learning_rate": 0.0001992195877775053, + "loss": 0.0226, + "step": 6231 + }, + { + "epoch": 2.9327058823529413, + "grad_norm": 0.1142787049354796, + "learning_rate": 0.00019917817660483327, + "loss": 0.047, + "step": 6232 + }, + { + "epoch": 2.9331764705882355, + "grad_norm": 0.09088829807145939, + "learning_rate": 0.00019913676123210373, + "loss": 0.0401, + "step": 6233 + }, + { + "epoch": 2.9336470588235293, + "grad_norm": 0.10892906274615091, + "learning_rate": 0.00019909534166285382, + "loss": 0.0591, + "step": 6234 + }, + { + "epoch": 2.9341176470588235, + "grad_norm": 0.1090669265157549, + "learning_rate": 0.00019905391790062092, + "loss": 0.0594, + "step": 6235 + }, + { + "epoch": 2.9345882352941177, + "grad_norm": 0.15555953351874036, + "learning_rate": 0.00019901248994894276, + "loss": 0.0959, + "step": 6236 + }, + { + "epoch": 2.9350588235294115, + "grad_norm": 0.08336307957738823, + "learning_rate": 0.0001989710578113576, + "loss": 0.0446, + "step": 6237 + }, + { + "epoch": 2.9355294117647057, + "grad_norm": 0.19425742357613884, + "learning_rate": 0.00019892962149140392, + "loss": 0.0725, + "step": 6238 + }, + { + "epoch": 2.936, + "grad_norm": 2.3350938187369537e-05, + "learning_rate": 0.0001988881809926205, + "loss": 0.0, + "step": 6239 + }, + { + "epoch": 2.936470588235294, + "grad_norm": 0.07247101963067581, + "learning_rate": 0.00019884673631854672, + "loss": 0.0264, + "step": 6240 + }, + { + "epoch": 2.9369411764705884, + "grad_norm": 0.11938333233997503, + "learning_rate": 0.000198805287472722, + "loss": 0.0543, + "step": 6241 + }, + { + "epoch": 2.9374117647058826, + "grad_norm": 0.04788426623169324, + "learning_rate": 0.0001987638344586864, + "loss": 0.0104, + "step": 6242 + }, + { + "epoch": 2.9378823529411764, + "grad_norm": 0.06274821304478535, + "learning_rate": 0.00019872237727998004, + "loss": 0.0349, + "step": 6243 + }, + { + "epoch": 2.9383529411764706, + "grad_norm": 1.0279625555407646e-05, + "learning_rate": 0.0001986809159401438, + "loss": 0.0, + "step": 6244 + }, + { + "epoch": 2.938823529411765, + "grad_norm": 0.11299299289121326, + "learning_rate": 0.00019863945044271846, + "loss": 0.0301, + "step": 6245 + }, + { + "epoch": 2.9392941176470586, + "grad_norm": 0.08990031221491394, + "learning_rate": 0.00019859798079124554, + "loss": 0.0515, + "step": 6246 + }, + { + "epoch": 2.939764705882353, + "grad_norm": 0.30897353201335576, + "learning_rate": 0.00019855650698926666, + "loss": 0.086, + "step": 6247 + }, + { + "epoch": 2.940235294117647, + "grad_norm": 0.3202705920793857, + "learning_rate": 0.00019851502904032396, + "loss": 0.0528, + "step": 6248 + }, + { + "epoch": 2.9407058823529413, + "grad_norm": 0.11418821768740044, + "learning_rate": 0.00019847354694795978, + "loss": 0.0216, + "step": 6249 + }, + { + "epoch": 2.9411764705882355, + "grad_norm": 0.15711352167982873, + "learning_rate": 0.00019843206071571692, + "loss": 0.0363, + "step": 6250 + }, + { + "epoch": 2.9416470588235293, + "grad_norm": 0.020116152919738384, + "learning_rate": 0.00019839057034713858, + "loss": 0.0018, + "step": 6251 + }, + { + "epoch": 2.9421176470588235, + "grad_norm": 0.11633395345556712, + "learning_rate": 0.00019834907584576812, + "loss": 0.0705, + "step": 6252 + }, + { + "epoch": 2.9425882352941177, + "grad_norm": 0.07151945313834414, + "learning_rate": 0.00019830757721514943, + "loss": 0.0207, + "step": 6253 + }, + { + "epoch": 2.9430588235294115, + "grad_norm": 0.09308258442830318, + "learning_rate": 0.00019826607445882667, + "loss": 0.0573, + "step": 6254 + }, + { + "epoch": 2.9435294117647057, + "grad_norm": 0.12556558305799728, + "learning_rate": 0.00019822456758034445, + "loss": 0.1006, + "step": 6255 + }, + { + "epoch": 2.944, + "grad_norm": 0.10199388041437658, + "learning_rate": 0.0001981830565832476, + "loss": 0.0439, + "step": 6256 + }, + { + "epoch": 2.944470588235294, + "grad_norm": 0.05960253933529358, + "learning_rate": 0.0001981415414710814, + "loss": 0.0244, + "step": 6257 + }, + { + "epoch": 2.9449411764705884, + "grad_norm": 0.0384107153370468, + "learning_rate": 0.00019810002224739142, + "loss": 0.0134, + "step": 6258 + }, + { + "epoch": 2.9454117647058826, + "grad_norm": 0.052680187105649576, + "learning_rate": 0.00019805849891572363, + "loss": 0.0188, + "step": 6259 + }, + { + "epoch": 2.9458823529411764, + "grad_norm": 0.00022027686838550566, + "learning_rate": 0.00019801697147962428, + "loss": 0.0, + "step": 6260 + }, + { + "epoch": 2.9463529411764706, + "grad_norm": 0.11805756044940637, + "learning_rate": 0.00019797543994264004, + "loss": 0.0638, + "step": 6261 + }, + { + "epoch": 2.946823529411765, + "grad_norm": 0.057509935418159394, + "learning_rate": 0.00019793390430831794, + "loss": 0.0189, + "step": 6262 + }, + { + "epoch": 2.9472941176470586, + "grad_norm": 0.10853571730844071, + "learning_rate": 0.00019789236458020527, + "loss": 0.0546, + "step": 6263 + }, + { + "epoch": 2.947764705882353, + "grad_norm": 0.10505127203180659, + "learning_rate": 0.00019785082076184977, + "loss": 0.0459, + "step": 6264 + }, + { + "epoch": 2.948235294117647, + "grad_norm": 0.0007340655350022583, + "learning_rate": 0.00019780927285679945, + "loss": 0.0, + "step": 6265 + }, + { + "epoch": 2.9487058823529413, + "grad_norm": 0.07023432591110573, + "learning_rate": 0.00019776772086860277, + "loss": 0.0265, + "step": 6266 + }, + { + "epoch": 2.9491764705882355, + "grad_norm": 0.13302123380015465, + "learning_rate": 0.0001977261648008084, + "loss": 0.0894, + "step": 6267 + }, + { + "epoch": 2.9496470588235293, + "grad_norm": 0.10552891401797455, + "learning_rate": 0.00019768460465696545, + "loss": 0.0282, + "step": 6268 + }, + { + "epoch": 2.9501176470588235, + "grad_norm": 0.08741189139646743, + "learning_rate": 0.0001976430404406234, + "loss": 0.0437, + "step": 6269 + }, + { + "epoch": 2.9505882352941177, + "grad_norm": 0.113764927314054, + "learning_rate": 0.00019760147215533197, + "loss": 0.0497, + "step": 6270 + }, + { + "epoch": 2.9510588235294115, + "grad_norm": 0.10187420593923913, + "learning_rate": 0.00019755989980464137, + "loss": 0.0498, + "step": 6271 + }, + { + "epoch": 2.9515294117647057, + "grad_norm": 0.06593254073685546, + "learning_rate": 0.00019751832339210204, + "loss": 0.0007, + "step": 6272 + }, + { + "epoch": 2.952, + "grad_norm": 0.00012153783159996416, + "learning_rate": 0.00019747674292126477, + "loss": 0.0, + "step": 6273 + }, + { + "epoch": 2.952470588235294, + "grad_norm": 7.600961635654894e-05, + "learning_rate": 0.00019743515839568083, + "loss": 0.0, + "step": 6274 + }, + { + "epoch": 2.9529411764705884, + "grad_norm": 0.07203025940878091, + "learning_rate": 0.00019739356981890167, + "loss": 0.0337, + "step": 6275 + }, + { + "epoch": 2.9534117647058826, + "grad_norm": 0.042400673226341064, + "learning_rate": 0.00019735197719447916, + "loss": 0.0101, + "step": 6276 + }, + { + "epoch": 2.9538823529411764, + "grad_norm": 0.00015653789967478925, + "learning_rate": 0.00019731038052596558, + "loss": 0.0, + "step": 6277 + }, + { + "epoch": 2.9543529411764706, + "grad_norm": 0.04111282702044191, + "learning_rate": 0.00019726877981691347, + "loss": 0.0049, + "step": 6278 + }, + { + "epoch": 2.954823529411765, + "grad_norm": 0.1580196765745241, + "learning_rate": 0.0001972271750708757, + "loss": 0.095, + "step": 6279 + }, + { + "epoch": 2.9552941176470586, + "grad_norm": 0.0003923370802611753, + "learning_rate": 0.0001971855662914055, + "loss": 0.0, + "step": 6280 + }, + { + "epoch": 2.955764705882353, + "grad_norm": 0.06955639213050847, + "learning_rate": 0.00019714395348205649, + "loss": 0.0323, + "step": 6281 + }, + { + "epoch": 2.956235294117647, + "grad_norm": 0.05664641639491416, + "learning_rate": 0.00019710233664638264, + "loss": 0.0202, + "step": 6282 + }, + { + "epoch": 2.9567058823529413, + "grad_norm": 0.1288555405268743, + "learning_rate": 0.00019706071578793819, + "loss": 0.0802, + "step": 6283 + }, + { + "epoch": 2.9571764705882355, + "grad_norm": 0.08137426858756151, + "learning_rate": 0.00019701909091027778, + "loss": 0.048, + "step": 6284 + }, + { + "epoch": 2.9576470588235293, + "grad_norm": 0.09169651682362517, + "learning_rate": 0.0001969774620169564, + "loss": 0.0477, + "step": 6285 + }, + { + "epoch": 2.9581176470588235, + "grad_norm": 0.22277362827067226, + "learning_rate": 0.00019693582911152935, + "loss": 0.1223, + "step": 6286 + }, + { + "epoch": 2.9585882352941177, + "grad_norm": 0.0524189557636217, + "learning_rate": 0.0001968941921975522, + "loss": 0.0062, + "step": 6287 + }, + { + "epoch": 2.9590588235294115, + "grad_norm": 0.06526526785682077, + "learning_rate": 0.00019685255127858107, + "loss": 0.0284, + "step": 6288 + }, + { + "epoch": 2.9595294117647057, + "grad_norm": 0.14257942397692552, + "learning_rate": 0.00019681090635817228, + "loss": 0.0805, + "step": 6289 + }, + { + "epoch": 2.96, + "grad_norm": 0.14740181457385684, + "learning_rate": 0.00019676925743988244, + "loss": 0.062, + "step": 6290 + }, + { + "epoch": 2.960470588235294, + "grad_norm": 5.4606496649546154e-05, + "learning_rate": 0.00019672760452726863, + "loss": 0.0, + "step": 6291 + }, + { + "epoch": 2.9609411764705884, + "grad_norm": 4.300396279926305e-06, + "learning_rate": 0.0001966859476238882, + "loss": 0.0, + "step": 6292 + }, + { + "epoch": 2.961411764705882, + "grad_norm": 0.07386389122788992, + "learning_rate": 0.00019664428673329883, + "loss": 0.0268, + "step": 6293 + }, + { + "epoch": 2.9618823529411764, + "grad_norm": 0.1286372935849134, + "learning_rate": 0.00019660262185905856, + "loss": 0.0521, + "step": 6294 + }, + { + "epoch": 2.9623529411764706, + "grad_norm": 0.12015294738866672, + "learning_rate": 0.0001965609530047259, + "loss": 0.0929, + "step": 6295 + }, + { + "epoch": 2.962823529411765, + "grad_norm": 0.09974749361094339, + "learning_rate": 0.0001965192801738594, + "loss": 0.0572, + "step": 6296 + }, + { + "epoch": 2.9632941176470586, + "grad_norm": 0.10080766489008512, + "learning_rate": 0.0001964776033700182, + "loss": 0.0345, + "step": 6297 + }, + { + "epoch": 2.963764705882353, + "grad_norm": 0.06646143173821571, + "learning_rate": 0.00019643592259676164, + "loss": 0.0246, + "step": 6298 + }, + { + "epoch": 2.964235294117647, + "grad_norm": 0.033579549781084225, + "learning_rate": 0.0001963942378576496, + "loss": 0.0093, + "step": 6299 + }, + { + "epoch": 2.9647058823529413, + "grad_norm": 0.13018586677981137, + "learning_rate": 0.0001963525491562421, + "loss": 0.0935, + "step": 6300 + }, + { + "epoch": 2.9651764705882355, + "grad_norm": 5.023365964980909e-06, + "learning_rate": 0.0001963108564960995, + "loss": 0.0, + "step": 6301 + }, + { + "epoch": 2.9656470588235293, + "grad_norm": 1.4050583792340932e-05, + "learning_rate": 0.00019626915988078263, + "loss": 0.0, + "step": 6302 + }, + { + "epoch": 2.9661176470588235, + "grad_norm": 5.865026311030246e-06, + "learning_rate": 0.00019622745931385254, + "loss": 0.0, + "step": 6303 + }, + { + "epoch": 2.9665882352941177, + "grad_norm": 0.06839047296660415, + "learning_rate": 0.00019618575479887066, + "loss": 0.0259, + "step": 6304 + }, + { + "epoch": 2.9670588235294115, + "grad_norm": 0.06836728484388085, + "learning_rate": 0.00019614404633939886, + "loss": 0.0286, + "step": 6305 + }, + { + "epoch": 2.9675294117647057, + "grad_norm": 0.12487981993255842, + "learning_rate": 0.0001961023339389991, + "loss": 0.0517, + "step": 6306 + }, + { + "epoch": 2.968, + "grad_norm": 0.03437309920181541, + "learning_rate": 0.00019606061760123392, + "loss": 0.0065, + "step": 6307 + }, + { + "epoch": 2.968470588235294, + "grad_norm": 0.0527124642090519, + "learning_rate": 0.00019601889732966604, + "loss": 0.0202, + "step": 6308 + }, + { + "epoch": 2.9689411764705884, + "grad_norm": 0.05941544636883683, + "learning_rate": 0.0001959771731278586, + "loss": 0.0243, + "step": 6309 + }, + { + "epoch": 2.969411764705882, + "grad_norm": 0.05673682802016504, + "learning_rate": 0.0001959354449993751, + "loss": 0.0194, + "step": 6310 + }, + { + "epoch": 2.9698823529411764, + "grad_norm": 0.03254953278606602, + "learning_rate": 0.00019589371294777925, + "loss": 0.006, + "step": 6311 + }, + { + "epoch": 2.9703529411764706, + "grad_norm": 0.060968141739591426, + "learning_rate": 0.00019585197697663526, + "loss": 0.021, + "step": 6312 + }, + { + "epoch": 2.970823529411765, + "grad_norm": 0.061350223052602425, + "learning_rate": 0.0001958102370895075, + "loss": 0.0273, + "step": 6313 + }, + { + "epoch": 2.9712941176470586, + "grad_norm": 5.780101412073651e-06, + "learning_rate": 0.00019576849328996075, + "loss": 0.0, + "step": 6314 + }, + { + "epoch": 2.971764705882353, + "grad_norm": 0.0624670011712671, + "learning_rate": 0.00019572674558156024, + "loss": 0.0202, + "step": 6315 + }, + { + "epoch": 2.972235294117647, + "grad_norm": 0.08260163007634812, + "learning_rate": 0.00019568499396787134, + "loss": 0.0405, + "step": 6316 + }, + { + "epoch": 2.9727058823529413, + "grad_norm": 0.15136365328359955, + "learning_rate": 0.0001956432384524599, + "loss": 0.1029, + "step": 6317 + }, + { + "epoch": 2.9731764705882355, + "grad_norm": 0.07134926974010816, + "learning_rate": 0.000195601479038892, + "loss": 0.0312, + "step": 6318 + }, + { + "epoch": 2.9736470588235293, + "grad_norm": 0.04958499907716499, + "learning_rate": 0.0001955597157307341, + "loss": 0.0184, + "step": 6319 + }, + { + "epoch": 2.9741176470588235, + "grad_norm": 0.07490966335010564, + "learning_rate": 0.000195517948531553, + "loss": 0.0281, + "step": 6320 + }, + { + "epoch": 2.9745882352941178, + "grad_norm": 0.15186444693504156, + "learning_rate": 0.00019547617744491587, + "loss": 0.0736, + "step": 6321 + }, + { + "epoch": 2.9750588235294115, + "grad_norm": 0.04516139882402757, + "learning_rate": 0.00019543440247439008, + "loss": 0.0097, + "step": 6322 + }, + { + "epoch": 2.9755294117647058, + "grad_norm": 3.103483951315342e-05, + "learning_rate": 0.00019539262362354353, + "loss": 0.0, + "step": 6323 + }, + { + "epoch": 2.976, + "grad_norm": 0.03567622938326731, + "learning_rate": 0.00019535084089594422, + "loss": 0.0105, + "step": 6324 + }, + { + "epoch": 2.976470588235294, + "grad_norm": 0.13162980460186666, + "learning_rate": 0.00019530905429516066, + "loss": 0.0694, + "step": 6325 + }, + { + "epoch": 2.9769411764705884, + "grad_norm": 0.08363866650903644, + "learning_rate": 0.00019526726382476165, + "loss": 0.0441, + "step": 6326 + }, + { + "epoch": 2.977411764705882, + "grad_norm": 0.1404689673277005, + "learning_rate": 0.00019522546948831627, + "loss": 0.0663, + "step": 6327 + }, + { + "epoch": 2.9778823529411764, + "grad_norm": 0.14763146539901348, + "learning_rate": 0.00019518367128939392, + "loss": 0.0799, + "step": 6328 + }, + { + "epoch": 2.9783529411764706, + "grad_norm": 0.050533662232663666, + "learning_rate": 0.0001951418692315645, + "loss": 0.0148, + "step": 6329 + }, + { + "epoch": 2.978823529411765, + "grad_norm": 0.14283977629471536, + "learning_rate": 0.00019510006331839794, + "loss": 0.0945, + "step": 6330 + }, + { + "epoch": 2.9792941176470586, + "grad_norm": 0.08048136317957418, + "learning_rate": 0.00019505825355346483, + "loss": 0.0318, + "step": 6331 + }, + { + "epoch": 2.979764705882353, + "grad_norm": 0.1318148909179708, + "learning_rate": 0.0001950164399403358, + "loss": 0.0578, + "step": 6332 + }, + { + "epoch": 2.980235294117647, + "grad_norm": 0.05681095530158919, + "learning_rate": 0.00019497462248258202, + "loss": 0.0199, + "step": 6333 + }, + { + "epoch": 2.9807058823529413, + "grad_norm": 5.2623921837265716e-05, + "learning_rate": 0.0001949328011837749, + "loss": 0.0, + "step": 6334 + }, + { + "epoch": 2.9811764705882355, + "grad_norm": 0.09514716958358777, + "learning_rate": 0.00019489097604748613, + "loss": 0.0422, + "step": 6335 + }, + { + "epoch": 2.9816470588235293, + "grad_norm": 0.10675846392151157, + "learning_rate": 0.00019484914707728785, + "loss": 0.0529, + "step": 6336 + }, + { + "epoch": 2.9821176470588235, + "grad_norm": 0.014042931699529916, + "learning_rate": 0.00019480731427675238, + "loss": 0.0015, + "step": 6337 + }, + { + "epoch": 2.9825882352941178, + "grad_norm": 0.0011038075849082874, + "learning_rate": 0.0001947654776494525, + "loss": 0.0, + "step": 6338 + }, + { + "epoch": 2.9830588235294115, + "grad_norm": 3.03098196832337e-05, + "learning_rate": 0.0001947236371989613, + "loss": 0.0, + "step": 6339 + }, + { + "epoch": 2.9835294117647058, + "grad_norm": 0.08379642595342464, + "learning_rate": 0.0001946817929288521, + "loss": 0.0371, + "step": 6340 + }, + { + "epoch": 2.984, + "grad_norm": 0.1193188579618506, + "learning_rate": 0.0001946399448426986, + "loss": 0.0634, + "step": 6341 + }, + { + "epoch": 2.984470588235294, + "grad_norm": 0.12432736168629048, + "learning_rate": 0.00019459809294407486, + "loss": 0.064, + "step": 6342 + }, + { + "epoch": 2.9849411764705884, + "grad_norm": 0.09917994270911117, + "learning_rate": 0.00019455623723655521, + "loss": 0.0549, + "step": 6343 + }, + { + "epoch": 2.985411764705882, + "grad_norm": 0.32689613100694964, + "learning_rate": 0.00019451437772371438, + "loss": 0.208, + "step": 6344 + }, + { + "epoch": 2.9858823529411764, + "grad_norm": 0.09212302400411654, + "learning_rate": 0.00019447251440912736, + "loss": 0.0508, + "step": 6345 + }, + { + "epoch": 2.9863529411764707, + "grad_norm": 0.10668187189677736, + "learning_rate": 0.00019443064729636947, + "loss": 0.0647, + "step": 6346 + }, + { + "epoch": 2.9868235294117644, + "grad_norm": 0.048896618052828, + "learning_rate": 0.00019438877638901635, + "loss": 0.0126, + "step": 6347 + }, + { + "epoch": 2.9872941176470587, + "grad_norm": 0.06504900615786961, + "learning_rate": 0.00019434690169064402, + "loss": 0.0256, + "step": 6348 + }, + { + "epoch": 2.987764705882353, + "grad_norm": 0.04993674631686231, + "learning_rate": 0.00019430502320482871, + "loss": 0.0168, + "step": 6349 + }, + { + "epoch": 2.988235294117647, + "grad_norm": 0.10129406943133264, + "learning_rate": 0.00019426314093514717, + "loss": 0.0494, + "step": 6350 + }, + { + "epoch": 2.9887058823529413, + "grad_norm": 0.1444824782897118, + "learning_rate": 0.0001942212548851763, + "loss": 0.0734, + "step": 6351 + }, + { + "epoch": 2.9891764705882355, + "grad_norm": 0.28468730251162855, + "learning_rate": 0.00019417936505849328, + "loss": 0.1225, + "step": 6352 + }, + { + "epoch": 2.9896470588235293, + "grad_norm": 0.05100150504359021, + "learning_rate": 0.00019413747145867583, + "loss": 0.0175, + "step": 6353 + }, + { + "epoch": 2.9901176470588235, + "grad_norm": 0.17512506440614123, + "learning_rate": 0.00019409557408930184, + "loss": 0.1189, + "step": 6354 + }, + { + "epoch": 2.9905882352941178, + "grad_norm": 0.00011264228721620452, + "learning_rate": 0.0001940536729539495, + "loss": 0.0, + "step": 6355 + }, + { + "epoch": 2.9910588235294115, + "grad_norm": 0.07489289077429817, + "learning_rate": 0.00019401176805619746, + "loss": 0.0286, + "step": 6356 + }, + { + "epoch": 2.9915294117647058, + "grad_norm": 0.09490505123261443, + "learning_rate": 0.0001939698593996245, + "loss": 0.0339, + "step": 6357 + }, + { + "epoch": 2.992, + "grad_norm": 0.07139044427022305, + "learning_rate": 0.00019392794698780988, + "loss": 0.0344, + "step": 6358 + }, + { + "epoch": 2.992470588235294, + "grad_norm": 0.2319555094517475, + "learning_rate": 0.00019388603082433315, + "loss": 0.1264, + "step": 6359 + }, + { + "epoch": 2.9929411764705884, + "grad_norm": 0.0895954433228931, + "learning_rate": 0.00019384411091277412, + "loss": 0.0231, + "step": 6360 + }, + { + "epoch": 2.993411764705882, + "grad_norm": 0.09959573121230046, + "learning_rate": 0.00019380218725671298, + "loss": 0.0515, + "step": 6361 + }, + { + "epoch": 2.9938823529411764, + "grad_norm": 0.05856597917699238, + "learning_rate": 0.00019376025985973013, + "loss": 0.0193, + "step": 6362 + }, + { + "epoch": 2.9943529411764707, + "grad_norm": 0.09205355282109884, + "learning_rate": 0.00019371832872540652, + "loss": 0.0332, + "step": 6363 + }, + { + "epoch": 2.9948235294117644, + "grad_norm": 0.05623409757435118, + "learning_rate": 0.00019367639385732316, + "loss": 0.0154, + "step": 6364 + }, + { + "epoch": 2.9952941176470587, + "grad_norm": 0.05816232387097028, + "learning_rate": 0.00019363445525906152, + "loss": 0.0158, + "step": 6365 + }, + { + "epoch": 2.995764705882353, + "grad_norm": 0.0934306535734845, + "learning_rate": 0.00019359251293420336, + "loss": 0.0379, + "step": 6366 + }, + { + "epoch": 2.996235294117647, + "grad_norm": 0.08769186716999368, + "learning_rate": 0.00019355056688633086, + "loss": 0.0433, + "step": 6367 + }, + { + "epoch": 2.9967058823529413, + "grad_norm": 0.07092620330567934, + "learning_rate": 0.0001935086171190263, + "loss": 0.0151, + "step": 6368 + }, + { + "epoch": 2.9971764705882356, + "grad_norm": 0.06402386790031339, + "learning_rate": 0.00019346666363587238, + "loss": 0.0217, + "step": 6369 + }, + { + "epoch": 2.9976470588235293, + "grad_norm": 0.06405519304677705, + "learning_rate": 0.00019342470644045223, + "loss": 0.0315, + "step": 6370 + }, + { + "epoch": 2.9981176470588236, + "grad_norm": 0.014736687354910299, + "learning_rate": 0.00019338274553634913, + "loss": 0.0021, + "step": 6371 + }, + { + "epoch": 2.9985882352941178, + "grad_norm": 0.07541702263553823, + "learning_rate": 0.00019334078092714683, + "loss": 0.0429, + "step": 6372 + }, + { + "epoch": 2.9990588235294116, + "grad_norm": 0.05217967043872214, + "learning_rate": 0.00019329881261642918, + "loss": 0.0164, + "step": 6373 + }, + { + "epoch": 2.9995294117647058, + "grad_norm": 6.642894149272895e-05, + "learning_rate": 0.00019325684060778056, + "loss": 0.0, + "step": 6374 + }, + { + "epoch": 3.0, + "grad_norm": 0.09695494529666272, + "learning_rate": 0.00019321486490478563, + "loss": 0.0354, + "step": 6375 + }, + { + "epoch": 3.0, + "eval_loss": 0.04407524690032005, + "eval_runtime": 59.1545, + "eval_samples_per_second": 12.679, + "eval_steps_per_second": 6.339, + "step": 6375 + }, + { + "epoch": 3.0004705882352942, + "grad_norm": 3.4849059402036185e-06, + "learning_rate": 0.0001931728855110292, + "loss": 0.0, + "step": 6376 + }, + { + "epoch": 3.0009411764705884, + "grad_norm": 0.06224179931123662, + "learning_rate": 0.00019313090243009665, + "loss": 0.0188, + "step": 6377 + }, + { + "epoch": 3.0014117647058822, + "grad_norm": 0.04622299510412963, + "learning_rate": 0.00019308891566557345, + "loss": 0.0124, + "step": 6378 + }, + { + "epoch": 3.0018823529411764, + "grad_norm": 0.05158538271847483, + "learning_rate": 0.0001930469252210455, + "loss": 0.0126, + "step": 6379 + }, + { + "epoch": 3.0023529411764707, + "grad_norm": 0.07179153269604413, + "learning_rate": 0.000193004931100099, + "loss": 0.0259, + "step": 6380 + }, + { + "epoch": 3.002823529411765, + "grad_norm": 0.11416821594375169, + "learning_rate": 0.00019296293330632047, + "loss": 0.051, + "step": 6381 + }, + { + "epoch": 3.0032941176470587, + "grad_norm": 0.10123347943949078, + "learning_rate": 0.0001929209318432967, + "loss": 0.0667, + "step": 6382 + }, + { + "epoch": 3.003764705882353, + "grad_norm": 0.15183803823637315, + "learning_rate": 0.00019287892671461475, + "loss": 0.0859, + "step": 6383 + }, + { + "epoch": 3.004235294117647, + "grad_norm": 0.10330229831553164, + "learning_rate": 0.0001928369179238622, + "loss": 0.0403, + "step": 6384 + }, + { + "epoch": 3.0047058823529413, + "grad_norm": 0.1125542117342554, + "learning_rate": 0.00019279490547462674, + "loss": 0.0468, + "step": 6385 + }, + { + "epoch": 3.005176470588235, + "grad_norm": 1.0789802772158321e-06, + "learning_rate": 0.0001927528893704964, + "loss": 0.0, + "step": 6386 + }, + { + "epoch": 3.0056470588235293, + "grad_norm": 0.04890564431989754, + "learning_rate": 0.00019271086961505964, + "loss": 0.0091, + "step": 6387 + }, + { + "epoch": 3.0061176470588236, + "grad_norm": 0.10584121493568864, + "learning_rate": 0.0001926688462119051, + "loss": 0.0556, + "step": 6388 + }, + { + "epoch": 3.006588235294118, + "grad_norm": 0.15075920719610844, + "learning_rate": 0.00019262681916462187, + "loss": 0.0828, + "step": 6389 + }, + { + "epoch": 3.0070588235294116, + "grad_norm": 0.03655928139885795, + "learning_rate": 0.00019258478847679914, + "loss": 0.0046, + "step": 6390 + }, + { + "epoch": 3.007529411764706, + "grad_norm": 0.08016686573156737, + "learning_rate": 0.0001925427541520266, + "loss": 0.0201, + "step": 6391 + }, + { + "epoch": 3.008, + "grad_norm": 0.09070963183458305, + "learning_rate": 0.00019250071619389422, + "loss": 0.0555, + "step": 6392 + }, + { + "epoch": 3.0084705882352942, + "grad_norm": 0.04994501306963986, + "learning_rate": 0.00019245867460599217, + "loss": 0.0034, + "step": 6393 + }, + { + "epoch": 3.0089411764705885, + "grad_norm": 0.07452113753957333, + "learning_rate": 0.00019241662939191106, + "loss": 0.0379, + "step": 6394 + }, + { + "epoch": 3.0094117647058822, + "grad_norm": 0.09971655847117669, + "learning_rate": 0.00019237458055524178, + "loss": 0.0474, + "step": 6395 + }, + { + "epoch": 3.0098823529411765, + "grad_norm": 0.12961541422595257, + "learning_rate": 0.00019233252809957546, + "loss": 0.0506, + "step": 6396 + }, + { + "epoch": 3.0103529411764707, + "grad_norm": 0.051140476146404164, + "learning_rate": 0.00019229047202850357, + "loss": 0.0153, + "step": 6397 + }, + { + "epoch": 3.010823529411765, + "grad_norm": 0.07260327701228768, + "learning_rate": 0.00019224841234561797, + "loss": 0.0261, + "step": 6398 + }, + { + "epoch": 3.0112941176470587, + "grad_norm": 0.07959991974114589, + "learning_rate": 0.00019220634905451074, + "loss": 0.033, + "step": 6399 + }, + { + "epoch": 3.011764705882353, + "grad_norm": 0.1676830045813977, + "learning_rate": 0.00019216428215877425, + "loss": 0.0752, + "step": 6400 + }, + { + "epoch": 3.012235294117647, + "grad_norm": 0.07969188666353114, + "learning_rate": 0.00019212221166200124, + "loss": 0.0288, + "step": 6401 + }, + { + "epoch": 3.0127058823529413, + "grad_norm": 0.09967813203087873, + "learning_rate": 0.00019208013756778476, + "loss": 0.0334, + "step": 6402 + }, + { + "epoch": 3.013176470588235, + "grad_norm": 0.09745699442360255, + "learning_rate": 0.0001920380598797181, + "loss": 0.0433, + "step": 6403 + }, + { + "epoch": 3.0136470588235293, + "grad_norm": 0.042987024012314023, + "learning_rate": 0.000191995978601395, + "loss": 0.008, + "step": 6404 + }, + { + "epoch": 3.0141176470588236, + "grad_norm": 0.00015967079351093602, + "learning_rate": 0.0001919538937364093, + "loss": 0.0, + "step": 6405 + }, + { + "epoch": 3.014588235294118, + "grad_norm": 0.12465929319465868, + "learning_rate": 0.00019191180528835527, + "loss": 0.0586, + "step": 6406 + }, + { + "epoch": 3.0150588235294116, + "grad_norm": 0.06972852844480935, + "learning_rate": 0.00019186971326082752, + "loss": 0.0287, + "step": 6407 + }, + { + "epoch": 3.015529411764706, + "grad_norm": 0.10384927028231639, + "learning_rate": 0.0001918276176574209, + "loss": 0.0582, + "step": 6408 + }, + { + "epoch": 3.016, + "grad_norm": 0.4636337104125926, + "learning_rate": 0.00019178551848173054, + "loss": 0.0876, + "step": 6409 + }, + { + "epoch": 3.0164705882352942, + "grad_norm": 0.10048600029509953, + "learning_rate": 0.00019174341573735196, + "loss": 0.0435, + "step": 6410 + }, + { + "epoch": 3.016941176470588, + "grad_norm": 0.27275078455128027, + "learning_rate": 0.0001917013094278809, + "loss": 0.1286, + "step": 6411 + }, + { + "epoch": 3.0174117647058822, + "grad_norm": 0.06060061751626181, + "learning_rate": 0.0001916591995569135, + "loss": 0.0209, + "step": 6412 + }, + { + "epoch": 3.0178823529411765, + "grad_norm": 0.00010364177770113725, + "learning_rate": 0.00019161708612804615, + "loss": 0.0, + "step": 6413 + }, + { + "epoch": 3.0183529411764707, + "grad_norm": 1.3168461990745044e-05, + "learning_rate": 0.00019157496914487545, + "loss": 0.0, + "step": 6414 + }, + { + "epoch": 3.018823529411765, + "grad_norm": 0.0678833194092811, + "learning_rate": 0.00019153284861099848, + "loss": 0.0223, + "step": 6415 + }, + { + "epoch": 3.0192941176470587, + "grad_norm": 0.09700029468147034, + "learning_rate": 0.00019149072453001254, + "loss": 0.0438, + "step": 6416 + }, + { + "epoch": 3.019764705882353, + "grad_norm": 0.11435230081174255, + "learning_rate": 0.00019144859690551518, + "loss": 0.043, + "step": 6417 + }, + { + "epoch": 3.020235294117647, + "grad_norm": 0.16843829350504794, + "learning_rate": 0.00019140646574110438, + "loss": 0.0099, + "step": 6418 + }, + { + "epoch": 3.0207058823529414, + "grad_norm": 0.1288496423117344, + "learning_rate": 0.0001913643310403783, + "loss": 0.0439, + "step": 6419 + }, + { + "epoch": 3.021176470588235, + "grad_norm": 0.04303581714626996, + "learning_rate": 0.00019132219280693544, + "loss": 0.0099, + "step": 6420 + }, + { + "epoch": 3.0216470588235294, + "grad_norm": 0.08569925778216428, + "learning_rate": 0.00019128005104437462, + "loss": 0.0375, + "step": 6421 + }, + { + "epoch": 3.0221176470588236, + "grad_norm": 0.2891400737041192, + "learning_rate": 0.00019123790575629503, + "loss": 0.1188, + "step": 6422 + }, + { + "epoch": 3.022588235294118, + "grad_norm": 0.04218609907270072, + "learning_rate": 0.00019119575694629594, + "loss": 0.0089, + "step": 6423 + }, + { + "epoch": 3.0230588235294116, + "grad_norm": 0.16495979390198304, + "learning_rate": 0.00019115360461797716, + "loss": 0.0551, + "step": 6424 + }, + { + "epoch": 3.023529411764706, + "grad_norm": 0.200863498732392, + "learning_rate": 0.00019111144877493873, + "loss": 0.106, + "step": 6425 + }, + { + "epoch": 3.024, + "grad_norm": 0.055573272033451124, + "learning_rate": 0.0001910692894207809, + "loss": 0.0143, + "step": 6426 + }, + { + "epoch": 3.0244705882352942, + "grad_norm": 0.07068878171480503, + "learning_rate": 0.00019102712655910426, + "loss": 0.0283, + "step": 6427 + }, + { + "epoch": 3.024941176470588, + "grad_norm": 0.0018310967100170093, + "learning_rate": 0.00019098496019350985, + "loss": 0.0, + "step": 6428 + }, + { + "epoch": 3.0254117647058822, + "grad_norm": 0.03498977306737925, + "learning_rate": 0.0001909427903275988, + "loss": 0.0072, + "step": 6429 + }, + { + "epoch": 3.0258823529411765, + "grad_norm": 0.08532042848168417, + "learning_rate": 0.00019090061696497258, + "loss": 0.0316, + "step": 6430 + }, + { + "epoch": 3.0263529411764707, + "grad_norm": 0.1177592706180759, + "learning_rate": 0.0001908584401092331, + "loss": 0.0532, + "step": 6431 + }, + { + "epoch": 3.026823529411765, + "grad_norm": 0.15985084749172032, + "learning_rate": 0.0001908162597639824, + "loss": 0.043, + "step": 6432 + }, + { + "epoch": 3.0272941176470587, + "grad_norm": 0.0740902595730654, + "learning_rate": 0.00019077407593282294, + "loss": 0.0353, + "step": 6433 + }, + { + "epoch": 3.027764705882353, + "grad_norm": 0.14515582207825284, + "learning_rate": 0.00019073188861935735, + "loss": 0.0426, + "step": 6434 + }, + { + "epoch": 3.028235294117647, + "grad_norm": 0.07513335374632198, + "learning_rate": 0.00019068969782718874, + "loss": 0.0346, + "step": 6435 + }, + { + "epoch": 3.0287058823529414, + "grad_norm": 0.08270381614028378, + "learning_rate": 0.00019064750355992036, + "loss": 0.0361, + "step": 6436 + }, + { + "epoch": 3.029176470588235, + "grad_norm": 3.467540998698576, + "learning_rate": 0.00019060530582115578, + "loss": 0.1315, + "step": 6437 + }, + { + "epoch": 3.0296470588235294, + "grad_norm": 5.452232271315121e-06, + "learning_rate": 0.00019056310461449885, + "loss": 0.0, + "step": 6438 + }, + { + "epoch": 3.0301176470588236, + "grad_norm": 0.3043021043101579, + "learning_rate": 0.00019052089994355394, + "loss": 0.1192, + "step": 6439 + }, + { + "epoch": 3.030588235294118, + "grad_norm": 0.10883996755720415, + "learning_rate": 0.00019047869181192535, + "loss": 0.0464, + "step": 6440 + }, + { + "epoch": 3.0310588235294116, + "grad_norm": 0.06910746015083631, + "learning_rate": 0.00019043648022321793, + "loss": 0.0171, + "step": 6441 + }, + { + "epoch": 3.031529411764706, + "grad_norm": 9.459024479317151e-08, + "learning_rate": 0.00019039426518103678, + "loss": 0.0, + "step": 6442 + }, + { + "epoch": 3.032, + "grad_norm": 1.3181209125966435, + "learning_rate": 0.00019035204668898725, + "loss": 0.0514, + "step": 6443 + }, + { + "epoch": 3.0324705882352943, + "grad_norm": 0.10254186370868534, + "learning_rate": 0.00019030982475067498, + "loss": 0.0524, + "step": 6444 + }, + { + "epoch": 3.032941176470588, + "grad_norm": 2.1816821518158855e-08, + "learning_rate": 0.00019026759936970598, + "loss": 0.0, + "step": 6445 + }, + { + "epoch": 3.0334117647058823, + "grad_norm": 0.12054623316331976, + "learning_rate": 0.0001902253705496865, + "loss": 0.0454, + "step": 6446 + }, + { + "epoch": 3.0338823529411765, + "grad_norm": 0.05118283567096681, + "learning_rate": 0.00019018313829422307, + "loss": 0.0255, + "step": 6447 + }, + { + "epoch": 3.0343529411764707, + "grad_norm": 6.713532265734295e-07, + "learning_rate": 0.0001901409026069225, + "loss": 0.0, + "step": 6448 + }, + { + "epoch": 3.034823529411765, + "grad_norm": 0.0966059610519804, + "learning_rate": 0.000190098663491392, + "loss": 0.0419, + "step": 6449 + }, + { + "epoch": 3.0352941176470587, + "grad_norm": 0.11337609117780353, + "learning_rate": 0.00019005642095123895, + "loss": 0.0496, + "step": 6450 + }, + { + "epoch": 3.035764705882353, + "grad_norm": 0.0711688232718246, + "learning_rate": 0.00019001417499007103, + "loss": 0.0233, + "step": 6451 + }, + { + "epoch": 3.036235294117647, + "grad_norm": 0.049729205546931554, + "learning_rate": 0.0001899719256114963, + "loss": 0.0152, + "step": 6452 + }, + { + "epoch": 3.0367058823529414, + "grad_norm": 0.07463563683923521, + "learning_rate": 0.00018992967281912312, + "loss": 0.0282, + "step": 6453 + }, + { + "epoch": 3.037176470588235, + "grad_norm": 0.11884716125967323, + "learning_rate": 0.00018988741661655996, + "loss": 0.067, + "step": 6454 + }, + { + "epoch": 3.0376470588235294, + "grad_norm": 0.13057755732406187, + "learning_rate": 0.00018984515700741579, + "loss": 0.0479, + "step": 6455 + }, + { + "epoch": 3.0381176470588236, + "grad_norm": 0.07863489052549058, + "learning_rate": 0.0001898028939952998, + "loss": 0.0447, + "step": 6456 + }, + { + "epoch": 3.038588235294118, + "grad_norm": 0.12367583829416057, + "learning_rate": 0.00018976062758382144, + "loss": 0.0487, + "step": 6457 + }, + { + "epoch": 3.0390588235294116, + "grad_norm": 0.00014326637565764003, + "learning_rate": 0.0001897183577765904, + "loss": 0.0, + "step": 6458 + }, + { + "epoch": 3.039529411764706, + "grad_norm": 0.0004973047769676199, + "learning_rate": 0.00018967608457721684, + "loss": 0.0, + "step": 6459 + }, + { + "epoch": 3.04, + "grad_norm": 0.00026502741018792186, + "learning_rate": 0.00018963380798931102, + "loss": 0.0, + "step": 6460 + }, + { + "epoch": 3.0404705882352943, + "grad_norm": 0.1198267341314491, + "learning_rate": 0.00018959152801648364, + "loss": 0.056, + "step": 6461 + }, + { + "epoch": 3.040941176470588, + "grad_norm": 0.03898546128040708, + "learning_rate": 0.00018954924466234554, + "loss": 0.0077, + "step": 6462 + }, + { + "epoch": 3.0414117647058823, + "grad_norm": 0.05467729525225681, + "learning_rate": 0.00018950695793050796, + "loss": 0.0095, + "step": 6463 + }, + { + "epoch": 3.0418823529411765, + "grad_norm": 0.013558745845905412, + "learning_rate": 0.0001894646678245824, + "loss": 0.0012, + "step": 6464 + }, + { + "epoch": 3.0423529411764707, + "grad_norm": 0.09729755042540988, + "learning_rate": 0.00018942237434818062, + "loss": 0.0471, + "step": 6465 + }, + { + "epoch": 3.0428235294117645, + "grad_norm": 0.11400349236393484, + "learning_rate": 0.00018938007750491468, + "loss": 0.0627, + "step": 6466 + }, + { + "epoch": 3.0432941176470587, + "grad_norm": 0.0010066696261292587, + "learning_rate": 0.00018933777729839701, + "loss": 0.0, + "step": 6467 + }, + { + "epoch": 3.043764705882353, + "grad_norm": 0.10518869596036172, + "learning_rate": 0.0001892954737322402, + "loss": 0.0479, + "step": 6468 + }, + { + "epoch": 3.044235294117647, + "grad_norm": 0.12784143511769683, + "learning_rate": 0.00018925316681005718, + "loss": 0.0672, + "step": 6469 + }, + { + "epoch": 3.0447058823529414, + "grad_norm": 0.0014969493885016681, + "learning_rate": 0.0001892108565354612, + "loss": 0.0, + "step": 6470 + }, + { + "epoch": 3.045176470588235, + "grad_norm": 0.07726087364165057, + "learning_rate": 0.0001891685429120657, + "loss": 0.0299, + "step": 6471 + }, + { + "epoch": 3.0456470588235294, + "grad_norm": 0.046237949670871616, + "learning_rate": 0.00018912622594348454, + "loss": 0.0151, + "step": 6472 + }, + { + "epoch": 3.0461176470588236, + "grad_norm": 0.14309434174981303, + "learning_rate": 0.00018908390563333175, + "loss": 0.0985, + "step": 6473 + }, + { + "epoch": 3.046588235294118, + "grad_norm": 0.03692481663990113, + "learning_rate": 0.00018904158198522177, + "loss": 0.0083, + "step": 6474 + }, + { + "epoch": 3.0470588235294116, + "grad_norm": 0.1104396216935413, + "learning_rate": 0.00018899925500276912, + "loss": 0.0682, + "step": 6475 + }, + { + "epoch": 3.047529411764706, + "grad_norm": 0.0020078489009414478, + "learning_rate": 0.00018895692468958884, + "loss": 0.0, + "step": 6476 + }, + { + "epoch": 3.048, + "grad_norm": 0.10052425020327534, + "learning_rate": 0.00018891459104929608, + "loss": 0.0396, + "step": 6477 + }, + { + "epoch": 3.0484705882352943, + "grad_norm": 0.06582503735297135, + "learning_rate": 0.0001888722540855064, + "loss": 0.0076, + "step": 6478 + }, + { + "epoch": 3.048941176470588, + "grad_norm": 0.1829726094882884, + "learning_rate": 0.00018882991380183556, + "loss": 0.0665, + "step": 6479 + }, + { + "epoch": 3.0494117647058823, + "grad_norm": 0.15264478958291736, + "learning_rate": 0.0001887875702018996, + "loss": 0.062, + "step": 6480 + }, + { + "epoch": 3.0498823529411765, + "grad_norm": 0.3160564714554348, + "learning_rate": 0.00018874522328931492, + "loss": 0.1589, + "step": 6481 + }, + { + "epoch": 3.0503529411764707, + "grad_norm": 0.030695785893954112, + "learning_rate": 0.0001887028730676981, + "loss": 0.005, + "step": 6482 + }, + { + "epoch": 3.0508235294117645, + "grad_norm": 0.17743464652365012, + "learning_rate": 0.00018866051954066608, + "loss": 0.071, + "step": 6483 + }, + { + "epoch": 3.0512941176470587, + "grad_norm": 0.08420244569235234, + "learning_rate": 0.00018861816271183613, + "loss": 0.0231, + "step": 6484 + }, + { + "epoch": 3.051764705882353, + "grad_norm": 0.001393404982236597, + "learning_rate": 0.00018857580258482557, + "loss": 0.0, + "step": 6485 + }, + { + "epoch": 3.052235294117647, + "grad_norm": 0.0644594737696834, + "learning_rate": 0.0001885334391632523, + "loss": 0.0193, + "step": 6486 + }, + { + "epoch": 3.0527058823529414, + "grad_norm": 0.2070636200108786, + "learning_rate": 0.00018849107245073432, + "loss": 0.0607, + "step": 6487 + }, + { + "epoch": 3.053176470588235, + "grad_norm": 0.08820361083463464, + "learning_rate": 0.00018844870245088996, + "loss": 0.0384, + "step": 6488 + }, + { + "epoch": 3.0536470588235294, + "grad_norm": 0.07296140927396072, + "learning_rate": 0.00018840632916733777, + "loss": 0.0263, + "step": 6489 + }, + { + "epoch": 3.0541176470588236, + "grad_norm": 0.09615066322321093, + "learning_rate": 0.00018836395260369676, + "loss": 0.0233, + "step": 6490 + }, + { + "epoch": 3.054588235294118, + "grad_norm": 0.0005241971279226016, + "learning_rate": 0.000188321572763586, + "loss": 0.0, + "step": 6491 + }, + { + "epoch": 3.0550588235294116, + "grad_norm": 0.09442535619892306, + "learning_rate": 0.00018827918965062492, + "loss": 0.0321, + "step": 6492 + }, + { + "epoch": 3.055529411764706, + "grad_norm": 0.0996681409090307, + "learning_rate": 0.00018823680326843334, + "loss": 0.0395, + "step": 6493 + }, + { + "epoch": 3.056, + "grad_norm": 0.13327933464000957, + "learning_rate": 0.00018819441362063122, + "loss": 0.0127, + "step": 6494 + }, + { + "epoch": 3.0564705882352943, + "grad_norm": 0.10082343688981606, + "learning_rate": 0.00018815202071083878, + "loss": 0.0475, + "step": 6495 + }, + { + "epoch": 3.056941176470588, + "grad_norm": 0.06915891223312844, + "learning_rate": 0.00018810962454267666, + "loss": 0.0278, + "step": 6496 + }, + { + "epoch": 3.0574117647058823, + "grad_norm": 0.1280351105449218, + "learning_rate": 0.00018806722511976567, + "loss": 0.0405, + "step": 6497 + }, + { + "epoch": 3.0578823529411765, + "grad_norm": 0.11704307222975714, + "learning_rate": 0.00018802482244572693, + "loss": 0.0418, + "step": 6498 + }, + { + "epoch": 3.0583529411764707, + "grad_norm": 0.00027000653060090495, + "learning_rate": 0.00018798241652418182, + "loss": 0.0, + "step": 6499 + }, + { + "epoch": 3.0588235294117645, + "grad_norm": 0.00026042187095329553, + "learning_rate": 0.00018794000735875208, + "loss": 0.0, + "step": 6500 + }, + { + "epoch": 3.0592941176470587, + "grad_norm": 0.25242096160353983, + "learning_rate": 0.00018789759495305963, + "loss": 0.1036, + "step": 6501 + }, + { + "epoch": 3.059764705882353, + "grad_norm": 0.10870115218987154, + "learning_rate": 0.00018785517931072668, + "loss": 0.05, + "step": 6502 + }, + { + "epoch": 3.060235294117647, + "grad_norm": 0.06826077042046463, + "learning_rate": 0.0001878127604353757, + "loss": 0.0272, + "step": 6503 + }, + { + "epoch": 3.0607058823529414, + "grad_norm": 0.11907385674338677, + "learning_rate": 0.00018777033833062954, + "loss": 0.0422, + "step": 6504 + }, + { + "epoch": 3.061176470588235, + "grad_norm": 0.056894024410151435, + "learning_rate": 0.00018772791300011122, + "loss": 0.0194, + "step": 6505 + }, + { + "epoch": 3.0616470588235294, + "grad_norm": 0.050680500128523764, + "learning_rate": 0.00018768548444744404, + "loss": 0.012, + "step": 6506 + }, + { + "epoch": 3.0621176470588236, + "grad_norm": 0.20385259605320483, + "learning_rate": 0.00018764305267625169, + "loss": 0.0625, + "step": 6507 + }, + { + "epoch": 3.062588235294118, + "grad_norm": 0.0002910918297059485, + "learning_rate": 0.000187600617690158, + "loss": 0.0, + "step": 6508 + }, + { + "epoch": 3.0630588235294116, + "grad_norm": 0.10595510874785624, + "learning_rate": 0.00018755817949278706, + "loss": 0.0518, + "step": 6509 + }, + { + "epoch": 3.063529411764706, + "grad_norm": 0.11851272965385813, + "learning_rate": 0.00018751573808776346, + "loss": 0.0296, + "step": 6510 + }, + { + "epoch": 3.064, + "grad_norm": 0.1336592769386832, + "learning_rate": 0.00018747329347871178, + "loss": 0.0544, + "step": 6511 + }, + { + "epoch": 3.0644705882352943, + "grad_norm": 0.00012615349109240515, + "learning_rate": 0.00018743084566925702, + "loss": 0.0, + "step": 6512 + }, + { + "epoch": 3.064941176470588, + "grad_norm": 0.322748635532397, + "learning_rate": 0.00018738839466302443, + "loss": 0.1011, + "step": 6513 + }, + { + "epoch": 3.0654117647058823, + "grad_norm": 0.1161007994281505, + "learning_rate": 0.00018734594046363956, + "loss": 0.0481, + "step": 6514 + }, + { + "epoch": 3.0658823529411765, + "grad_norm": 0.0819903512764226, + "learning_rate": 0.00018730348307472824, + "loss": 0.0299, + "step": 6515 + }, + { + "epoch": 3.0663529411764707, + "grad_norm": 0.08050283922817068, + "learning_rate": 0.0001872610224999164, + "loss": 0.0085, + "step": 6516 + }, + { + "epoch": 3.0668235294117645, + "grad_norm": 0.1454083379056489, + "learning_rate": 0.00018721855874283058, + "loss": 0.0659, + "step": 6517 + }, + { + "epoch": 3.0672941176470587, + "grad_norm": 0.11106311291507531, + "learning_rate": 0.00018717609180709721, + "loss": 0.0513, + "step": 6518 + }, + { + "epoch": 3.067764705882353, + "grad_norm": 0.1125440167903829, + "learning_rate": 0.00018713362169634325, + "loss": 0.0318, + "step": 6519 + }, + { + "epoch": 3.068235294117647, + "grad_norm": 0.04573611829644511, + "learning_rate": 0.00018709114841419588, + "loss": 0.0131, + "step": 6520 + }, + { + "epoch": 3.0687058823529414, + "grad_norm": 0.013564487319271283, + "learning_rate": 0.0001870486719642825, + "loss": 0.0014, + "step": 6521 + }, + { + "epoch": 3.069176470588235, + "grad_norm": 0.11278146749353546, + "learning_rate": 0.00018700619235023085, + "loss": 0.0576, + "step": 6522 + }, + { + "epoch": 3.0696470588235294, + "grad_norm": 0.143074000368473, + "learning_rate": 0.00018696370957566882, + "loss": 0.0747, + "step": 6523 + }, + { + "epoch": 3.0701176470588236, + "grad_norm": 1.7376658122102474e-05, + "learning_rate": 0.0001869212236442247, + "loss": 0.0, + "step": 6524 + }, + { + "epoch": 3.070588235294118, + "grad_norm": 0.08526415215948058, + "learning_rate": 0.00018687873455952697, + "loss": 0.0382, + "step": 6525 + }, + { + "epoch": 3.0710588235294116, + "grad_norm": 0.30597319786249666, + "learning_rate": 0.00018683624232520446, + "loss": 0.0656, + "step": 6526 + }, + { + "epoch": 3.071529411764706, + "grad_norm": 0.09812113482371482, + "learning_rate": 0.00018679374694488608, + "loss": 0.0374, + "step": 6527 + }, + { + "epoch": 3.072, + "grad_norm": 0.09341875313756241, + "learning_rate": 0.00018675124842220135, + "loss": 0.0426, + "step": 6528 + }, + { + "epoch": 3.0724705882352943, + "grad_norm": 0.09722942171048471, + "learning_rate": 0.0001867087467607797, + "loss": 0.0242, + "step": 6529 + }, + { + "epoch": 3.072941176470588, + "grad_norm": 0.09036192500816607, + "learning_rate": 0.00018666624196425098, + "loss": 0.0311, + "step": 6530 + }, + { + "epoch": 3.0734117647058823, + "grad_norm": 0.11898076586602974, + "learning_rate": 0.00018662373403624538, + "loss": 0.0397, + "step": 6531 + }, + { + "epoch": 3.0738823529411765, + "grad_norm": 0.03867544914240011, + "learning_rate": 0.00018658122298039327, + "loss": 0.0072, + "step": 6532 + }, + { + "epoch": 3.0743529411764707, + "grad_norm": 0.0626791448803437, + "learning_rate": 0.00018653870880032527, + "loss": 0.025, + "step": 6533 + }, + { + "epoch": 3.0748235294117645, + "grad_norm": 0.1310303727028999, + "learning_rate": 0.00018649619149967232, + "loss": 0.0658, + "step": 6534 + }, + { + "epoch": 3.0752941176470587, + "grad_norm": 0.1387272271219273, + "learning_rate": 0.00018645367108206565, + "loss": 0.0528, + "step": 6535 + }, + { + "epoch": 3.075764705882353, + "grad_norm": 0.0960014101832223, + "learning_rate": 0.00018641114755113658, + "loss": 0.0616, + "step": 6536 + }, + { + "epoch": 3.076235294117647, + "grad_norm": 0.09346815116024726, + "learning_rate": 0.00018636862091051697, + "loss": 0.0404, + "step": 6537 + }, + { + "epoch": 3.0767058823529414, + "grad_norm": 0.04818487565318129, + "learning_rate": 0.00018632609116383874, + "loss": 0.0151, + "step": 6538 + }, + { + "epoch": 3.077176470588235, + "grad_norm": 8.375216566758682e-05, + "learning_rate": 0.00018628355831473416, + "loss": 0.0, + "step": 6539 + }, + { + "epoch": 3.0776470588235294, + "grad_norm": 0.0975096854410171, + "learning_rate": 0.00018624102236683567, + "loss": 0.0564, + "step": 6540 + }, + { + "epoch": 3.0781176470588236, + "grad_norm": 0.0938502710996837, + "learning_rate": 0.00018619848332377615, + "loss": 0.0392, + "step": 6541 + }, + { + "epoch": 3.078588235294118, + "grad_norm": 0.4502756096474406, + "learning_rate": 0.00018615594118918862, + "loss": 0.1296, + "step": 6542 + }, + { + "epoch": 3.0790588235294116, + "grad_norm": 0.10675050184488966, + "learning_rate": 0.00018611339596670632, + "loss": 0.0333, + "step": 6543 + }, + { + "epoch": 3.079529411764706, + "grad_norm": 0.00012704565351971515, + "learning_rate": 0.00018607084765996285, + "loss": 0.0, + "step": 6544 + }, + { + "epoch": 3.08, + "grad_norm": 0.05387215052101459, + "learning_rate": 0.00018602829627259213, + "loss": 0.0159, + "step": 6545 + }, + { + "epoch": 3.0804705882352943, + "grad_norm": 0.06686607646591229, + "learning_rate": 0.00018598574180822818, + "loss": 0.0227, + "step": 6546 + }, + { + "epoch": 3.080941176470588, + "grad_norm": 0.0751308676269276, + "learning_rate": 0.00018594318427050537, + "loss": 0.0273, + "step": 6547 + }, + { + "epoch": 3.0814117647058823, + "grad_norm": 0.08439406203933078, + "learning_rate": 0.00018590062366305828, + "loss": 0.0412, + "step": 6548 + }, + { + "epoch": 3.0818823529411765, + "grad_norm": 0.0505479653728472, + "learning_rate": 0.0001858580599895219, + "loss": 0.0141, + "step": 6549 + }, + { + "epoch": 3.0823529411764707, + "grad_norm": 0.07673412666672833, + "learning_rate": 0.00018581549325353126, + "loss": 0.0284, + "step": 6550 + }, + { + "epoch": 3.0828235294117645, + "grad_norm": 0.000332851742458591, + "learning_rate": 0.00018577292345872188, + "loss": 0.0, + "step": 6551 + }, + { + "epoch": 3.0832941176470587, + "grad_norm": 0.09567077010887823, + "learning_rate": 0.00018573035060872933, + "loss": 0.0239, + "step": 6552 + }, + { + "epoch": 3.083764705882353, + "grad_norm": 0.07023663053542079, + "learning_rate": 0.0001856877747071896, + "loss": 0.0234, + "step": 6553 + }, + { + "epoch": 3.084235294117647, + "grad_norm": 0.049534184336999444, + "learning_rate": 0.00018564519575773884, + "loss": 0.0147, + "step": 6554 + }, + { + "epoch": 3.084705882352941, + "grad_norm": 0.00032236440529516327, + "learning_rate": 0.00018560261376401356, + "loss": 0.0, + "step": 6555 + }, + { + "epoch": 3.085176470588235, + "grad_norm": 0.14934953610064522, + "learning_rate": 0.0001855600287296504, + "loss": 0.0766, + "step": 6556 + }, + { + "epoch": 3.0856470588235294, + "grad_norm": 0.1269902617124185, + "learning_rate": 0.00018551744065828634, + "loss": 0.0498, + "step": 6557 + }, + { + "epoch": 3.0861176470588236, + "grad_norm": 0.05872408050895639, + "learning_rate": 0.0001854748495535587, + "loss": 0.0157, + "step": 6558 + }, + { + "epoch": 3.086588235294118, + "grad_norm": 0.0005983472162716746, + "learning_rate": 0.0001854322554191049, + "loss": 0.0, + "step": 6559 + }, + { + "epoch": 3.0870588235294116, + "grad_norm": 0.13624549237493422, + "learning_rate": 0.00018538965825856264, + "loss": 0.0568, + "step": 6560 + }, + { + "epoch": 3.087529411764706, + "grad_norm": 0.04120036828508725, + "learning_rate": 0.00018534705807557003, + "loss": 0.0078, + "step": 6561 + }, + { + "epoch": 3.088, + "grad_norm": 0.00019595135181581417, + "learning_rate": 0.00018530445487376524, + "loss": 0.0, + "step": 6562 + }, + { + "epoch": 3.0884705882352943, + "grad_norm": 0.13646684825437796, + "learning_rate": 0.0001852618486567869, + "loss": 0.0911, + "step": 6563 + }, + { + "epoch": 3.088941176470588, + "grad_norm": 0.1137597466826309, + "learning_rate": 0.00018521923942827362, + "loss": 0.0536, + "step": 6564 + }, + { + "epoch": 3.0894117647058823, + "grad_norm": 0.242971483399231, + "learning_rate": 0.00018517662719186464, + "loss": 0.0509, + "step": 6565 + }, + { + "epoch": 3.0898823529411765, + "grad_norm": 0.012721858166111667, + "learning_rate": 0.0001851340119511991, + "loss": 0.0009, + "step": 6566 + }, + { + "epoch": 3.0903529411764707, + "grad_norm": 0.12200024448013501, + "learning_rate": 0.00018509139370991666, + "loss": 0.0434, + "step": 6567 + }, + { + "epoch": 3.0908235294117645, + "grad_norm": 0.00021697446991744254, + "learning_rate": 0.000185048772471657, + "loss": 0.0, + "step": 6568 + }, + { + "epoch": 3.0912941176470587, + "grad_norm": 0.07096971247112417, + "learning_rate": 0.00018500614824006035, + "loss": 0.0206, + "step": 6569 + }, + { + "epoch": 3.091764705882353, + "grad_norm": 0.09311757429985175, + "learning_rate": 0.0001849635210187669, + "loss": 0.03, + "step": 6570 + }, + { + "epoch": 3.092235294117647, + "grad_norm": 0.0001680175440647506, + "learning_rate": 0.00018492089081141724, + "loss": 0.0, + "step": 6571 + }, + { + "epoch": 3.092705882352941, + "grad_norm": 0.00018029354178975262, + "learning_rate": 0.00018487825762165223, + "loss": 0.0, + "step": 6572 + }, + { + "epoch": 3.093176470588235, + "grad_norm": 0.1665948873120008, + "learning_rate": 0.00018483562145311296, + "loss": 0.0785, + "step": 6573 + }, + { + "epoch": 3.0936470588235294, + "grad_norm": 0.15083407299363089, + "learning_rate": 0.00018479298230944071, + "loss": 0.072, + "step": 6574 + }, + { + "epoch": 3.0941176470588236, + "grad_norm": 0.0001748810552247997, + "learning_rate": 0.00018475034019427717, + "loss": 0.0, + "step": 6575 + }, + { + "epoch": 3.094588235294118, + "grad_norm": 0.09535833388890831, + "learning_rate": 0.00018470769511126409, + "loss": 0.0155, + "step": 6576 + }, + { + "epoch": 3.0950588235294116, + "grad_norm": 0.12892590934846967, + "learning_rate": 0.00018466504706404362, + "loss": 0.0767, + "step": 6577 + }, + { + "epoch": 3.095529411764706, + "grad_norm": 0.02482315178654729, + "learning_rate": 0.00018462239605625808, + "loss": 0.0034, + "step": 6578 + }, + { + "epoch": 3.096, + "grad_norm": 0.11402443149999511, + "learning_rate": 0.0001845797420915501, + "loss": 0.0481, + "step": 6579 + }, + { + "epoch": 3.0964705882352943, + "grad_norm": 0.1602828364484573, + "learning_rate": 0.00018453708517356256, + "loss": 0.0707, + "step": 6580 + }, + { + "epoch": 3.096941176470588, + "grad_norm": 0.11937066219918407, + "learning_rate": 0.0001844944253059385, + "loss": 0.0268, + "step": 6581 + }, + { + "epoch": 3.0974117647058823, + "grad_norm": 0.033142348241371265, + "learning_rate": 0.00018445176249232134, + "loss": 0.0036, + "step": 6582 + }, + { + "epoch": 3.0978823529411765, + "grad_norm": 0.00024326512897222176, + "learning_rate": 0.0001844090967363547, + "loss": 0.0, + "step": 6583 + }, + { + "epoch": 3.0983529411764708, + "grad_norm": 0.11796066792686566, + "learning_rate": 0.0001843664280416824, + "loss": 0.0486, + "step": 6584 + }, + { + "epoch": 3.0988235294117645, + "grad_norm": 0.020473012613437386, + "learning_rate": 0.0001843237564119486, + "loss": 0.0037, + "step": 6585 + }, + { + "epoch": 3.0992941176470588, + "grad_norm": 0.09174424521164817, + "learning_rate": 0.00018428108185079763, + "loss": 0.039, + "step": 6586 + }, + { + "epoch": 3.099764705882353, + "grad_norm": 0.09725777293812021, + "learning_rate": 0.00018423840436187412, + "loss": 0.0498, + "step": 6587 + }, + { + "epoch": 3.100235294117647, + "grad_norm": 0.00014464795030458416, + "learning_rate": 0.0001841957239488229, + "loss": 0.0, + "step": 6588 + }, + { + "epoch": 3.100705882352941, + "grad_norm": 0.000168062512211874, + "learning_rate": 0.00018415304061528918, + "loss": 0.0, + "step": 6589 + }, + { + "epoch": 3.101176470588235, + "grad_norm": 0.10705589095435451, + "learning_rate": 0.00018411035436491827, + "loss": 0.0225, + "step": 6590 + }, + { + "epoch": 3.1016470588235294, + "grad_norm": 0.12036355560739465, + "learning_rate": 0.0001840676652013558, + "loss": 0.0369, + "step": 6591 + }, + { + "epoch": 3.1021176470588236, + "grad_norm": 0.23065523896627457, + "learning_rate": 0.00018402497312824758, + "loss": 0.0745, + "step": 6592 + }, + { + "epoch": 3.1025882352941174, + "grad_norm": 0.07889962365915508, + "learning_rate": 0.0001839822781492398, + "loss": 0.0255, + "step": 6593 + }, + { + "epoch": 3.1030588235294116, + "grad_norm": 0.07469895552273188, + "learning_rate": 0.00018393958026797878, + "loss": 0.0308, + "step": 6594 + }, + { + "epoch": 3.103529411764706, + "grad_norm": 0.027072515792672596, + "learning_rate": 0.0001838968794881111, + "loss": 0.0044, + "step": 6595 + }, + { + "epoch": 3.104, + "grad_norm": 0.1494991940047215, + "learning_rate": 0.00018385417581328364, + "loss": 0.0703, + "step": 6596 + }, + { + "epoch": 3.1044705882352943, + "grad_norm": 0.0880728506587723, + "learning_rate": 0.00018381146924714356, + "loss": 0.0328, + "step": 6597 + }, + { + "epoch": 3.104941176470588, + "grad_norm": 0.00016216998913775026, + "learning_rate": 0.00018376875979333812, + "loss": 0.0, + "step": 6598 + }, + { + "epoch": 3.1054117647058823, + "grad_norm": 0.0596230049602456, + "learning_rate": 0.000183726047455515, + "loss": 0.017, + "step": 6599 + }, + { + "epoch": 3.1058823529411765, + "grad_norm": 0.00019790189334819607, + "learning_rate": 0.000183683332237322, + "loss": 0.0, + "step": 6600 + }, + { + "epoch": 3.1063529411764708, + "grad_norm": 0.1002010363186891, + "learning_rate": 0.00018364061414240717, + "loss": 0.0345, + "step": 6601 + }, + { + "epoch": 3.1068235294117645, + "grad_norm": 0.1457868619557146, + "learning_rate": 0.00018359789317441892, + "loss": 0.0406, + "step": 6602 + }, + { + "epoch": 3.1072941176470588, + "grad_norm": 0.08058041540166401, + "learning_rate": 0.00018355516933700578, + "loss": 0.0345, + "step": 6603 + }, + { + "epoch": 3.107764705882353, + "grad_norm": 0.09231979032780716, + "learning_rate": 0.0001835124426338166, + "loss": 0.0284, + "step": 6604 + }, + { + "epoch": 3.108235294117647, + "grad_norm": 0.04419441511107895, + "learning_rate": 0.00018346971306850045, + "loss": 0.012, + "step": 6605 + }, + { + "epoch": 3.108705882352941, + "grad_norm": 0.06157531679997028, + "learning_rate": 0.00018342698064470662, + "loss": 0.0189, + "step": 6606 + }, + { + "epoch": 3.109176470588235, + "grad_norm": 9.445414146322196e-05, + "learning_rate": 0.0001833842453660847, + "loss": 0.0, + "step": 6607 + }, + { + "epoch": 3.1096470588235294, + "grad_norm": 0.10179439896736243, + "learning_rate": 0.00018334150723628442, + "loss": 0.0335, + "step": 6608 + }, + { + "epoch": 3.1101176470588237, + "grad_norm": 0.05863786288664391, + "learning_rate": 0.00018329876625895592, + "loss": 0.0174, + "step": 6609 + }, + { + "epoch": 3.1105882352941174, + "grad_norm": 0.05842678075889254, + "learning_rate": 0.0001832560224377495, + "loss": 0.0201, + "step": 6610 + }, + { + "epoch": 3.1110588235294117, + "grad_norm": 0.16189103177774014, + "learning_rate": 0.0001832132757763156, + "loss": 0.0729, + "step": 6611 + }, + { + "epoch": 3.111529411764706, + "grad_norm": 0.09639252596717456, + "learning_rate": 0.000183170526278305, + "loss": 0.0483, + "step": 6612 + }, + { + "epoch": 3.112, + "grad_norm": 0.10958241889664747, + "learning_rate": 0.0001831277739473688, + "loss": 0.0576, + "step": 6613 + }, + { + "epoch": 3.1124705882352943, + "grad_norm": 8.362387355692563e-05, + "learning_rate": 0.00018308501878715823, + "loss": 0.0, + "step": 6614 + }, + { + "epoch": 3.112941176470588, + "grad_norm": 0.19286787327542929, + "learning_rate": 0.00018304226080132472, + "loss": 0.1123, + "step": 6615 + }, + { + "epoch": 3.1134117647058823, + "grad_norm": 0.05197189049831449, + "learning_rate": 0.00018299949999352013, + "loss": 0.022, + "step": 6616 + }, + { + "epoch": 3.1138823529411765, + "grad_norm": 0.08173204047331178, + "learning_rate": 0.00018295673636739635, + "loss": 0.0442, + "step": 6617 + }, + { + "epoch": 3.1143529411764708, + "grad_norm": 0.09492908537503947, + "learning_rate": 0.00018291396992660565, + "loss": 0.0482, + "step": 6618 + }, + { + "epoch": 3.1148235294117645, + "grad_norm": 0.06342298458384282, + "learning_rate": 0.00018287120067480046, + "loss": 0.0268, + "step": 6619 + }, + { + "epoch": 3.1152941176470588, + "grad_norm": 0.15741673944972026, + "learning_rate": 0.0001828284286156335, + "loss": 0.0681, + "step": 6620 + }, + { + "epoch": 3.115764705882353, + "grad_norm": 0.0942125612602481, + "learning_rate": 0.0001827856537527577, + "loss": 0.0245, + "step": 6621 + }, + { + "epoch": 3.116235294117647, + "grad_norm": 0.14696037418296345, + "learning_rate": 0.00018274287608982626, + "loss": 0.0478, + "step": 6622 + }, + { + "epoch": 3.116705882352941, + "grad_norm": 0.040321102932639484, + "learning_rate": 0.00018270009563049264, + "loss": 0.0059, + "step": 6623 + }, + { + "epoch": 3.117176470588235, + "grad_norm": 0.15430780027257313, + "learning_rate": 0.00018265731237841043, + "loss": 0.0951, + "step": 6624 + }, + { + "epoch": 3.1176470588235294, + "grad_norm": 5.2112830998217096e-05, + "learning_rate": 0.00018261452633723356, + "loss": 0.0, + "step": 6625 + }, + { + "epoch": 3.1181176470588237, + "grad_norm": 0.14493887662222713, + "learning_rate": 0.00018257173751061613, + "loss": 0.0633, + "step": 6626 + }, + { + "epoch": 3.1185882352941174, + "grad_norm": 0.11803168007821034, + "learning_rate": 0.00018252894590221262, + "loss": 0.0417, + "step": 6627 + }, + { + "epoch": 3.1190588235294117, + "grad_norm": 0.07570804275684286, + "learning_rate": 0.00018248615151567754, + "loss": 0.0145, + "step": 6628 + }, + { + "epoch": 3.119529411764706, + "grad_norm": 0.18375210758949992, + "learning_rate": 0.00018244335435466577, + "loss": 0.0936, + "step": 6629 + }, + { + "epoch": 3.12, + "grad_norm": 5.3299453254552684e-05, + "learning_rate": 0.00018240055442283242, + "loss": 0.0, + "step": 6630 + }, + { + "epoch": 3.1204705882352943, + "grad_norm": 0.00017302247360996637, + "learning_rate": 0.00018235775172383277, + "loss": 0.0, + "step": 6631 + }, + { + "epoch": 3.120941176470588, + "grad_norm": 0.13652301205015901, + "learning_rate": 0.00018231494626132243, + "loss": 0.0515, + "step": 6632 + }, + { + "epoch": 3.1214117647058823, + "grad_norm": 0.03744013656046588, + "learning_rate": 0.00018227213803895717, + "loss": 0.0053, + "step": 6633 + }, + { + "epoch": 3.1218823529411766, + "grad_norm": 0.16791611664840575, + "learning_rate": 0.00018222932706039307, + "loss": 0.0569, + "step": 6634 + }, + { + "epoch": 3.1223529411764708, + "grad_norm": 0.01947964149984695, + "learning_rate": 0.0001821865133292863, + "loss": 0.0027, + "step": 6635 + }, + { + "epoch": 3.1228235294117646, + "grad_norm": 0.12187394656558685, + "learning_rate": 0.00018214369684929347, + "loss": 0.0496, + "step": 6636 + }, + { + "epoch": 3.1232941176470588, + "grad_norm": 0.07542071565640769, + "learning_rate": 0.00018210087762407122, + "loss": 0.0237, + "step": 6637 + }, + { + "epoch": 3.123764705882353, + "grad_norm": 0.00010914009504197305, + "learning_rate": 0.00018205805565727658, + "loss": 0.0, + "step": 6638 + }, + { + "epoch": 3.124235294117647, + "grad_norm": 0.12281815918687636, + "learning_rate": 0.00018201523095256672, + "loss": 0.0571, + "step": 6639 + }, + { + "epoch": 3.124705882352941, + "grad_norm": 0.08413748558324544, + "learning_rate": 0.00018197240351359913, + "loss": 0.0362, + "step": 6640 + }, + { + "epoch": 3.1251764705882352, + "grad_norm": 0.09067885864235274, + "learning_rate": 0.0001819295733440315, + "loss": 0.0299, + "step": 6641 + }, + { + "epoch": 3.1256470588235294, + "grad_norm": 0.06697809254099192, + "learning_rate": 0.00018188674044752163, + "loss": 0.0275, + "step": 6642 + }, + { + "epoch": 3.1261176470588237, + "grad_norm": 0.1041395710829451, + "learning_rate": 0.00018184390482772774, + "loss": 0.0341, + "step": 6643 + }, + { + "epoch": 3.1265882352941174, + "grad_norm": 0.1551221602094729, + "learning_rate": 0.0001818010664883082, + "loss": 0.0606, + "step": 6644 + }, + { + "epoch": 3.1270588235294117, + "grad_norm": 0.07351586008811636, + "learning_rate": 0.0001817582254329216, + "loss": 0.0229, + "step": 6645 + }, + { + "epoch": 3.127529411764706, + "grad_norm": 4.100668321823578e-05, + "learning_rate": 0.0001817153816652267, + "loss": 0.0, + "step": 6646 + }, + { + "epoch": 3.128, + "grad_norm": 0.057123444996308316, + "learning_rate": 0.0001816725351888827, + "loss": 0.0217, + "step": 6647 + }, + { + "epoch": 3.1284705882352943, + "grad_norm": 0.16696775393097837, + "learning_rate": 0.00018162968600754888, + "loss": 0.0811, + "step": 6648 + }, + { + "epoch": 3.128941176470588, + "grad_norm": 0.1391619411962386, + "learning_rate": 0.00018158683412488466, + "loss": 0.0844, + "step": 6649 + }, + { + "epoch": 3.1294117647058823, + "grad_norm": 0.05980705871823291, + "learning_rate": 0.00018154397954454993, + "loss": 0.0209, + "step": 6650 + }, + { + "epoch": 3.1298823529411766, + "grad_norm": 0.13599850687273682, + "learning_rate": 0.0001815011222702046, + "loss": 0.0595, + "step": 6651 + }, + { + "epoch": 3.130352941176471, + "grad_norm": 0.04916574299349652, + "learning_rate": 0.00018145826230550892, + "loss": 0.01, + "step": 6652 + }, + { + "epoch": 3.1308235294117646, + "grad_norm": 0.0002448275504824928, + "learning_rate": 0.00018141539965412334, + "loss": 0.0, + "step": 6653 + }, + { + "epoch": 3.131294117647059, + "grad_norm": 0.014260131792616574, + "learning_rate": 0.00018137253431970853, + "loss": 0.0017, + "step": 6654 + }, + { + "epoch": 3.131764705882353, + "grad_norm": 3.7271794344499875e-05, + "learning_rate": 0.0001813296663059254, + "loss": 0.0, + "step": 6655 + }, + { + "epoch": 3.1322352941176472, + "grad_norm": 0.05118117679747207, + "learning_rate": 0.00018128679561643508, + "loss": 0.0159, + "step": 6656 + }, + { + "epoch": 3.132705882352941, + "grad_norm": 0.07711425011463888, + "learning_rate": 0.000181243922254899, + "loss": 0.0229, + "step": 6657 + }, + { + "epoch": 3.1331764705882352, + "grad_norm": 4.0079332441805563e-05, + "learning_rate": 0.0001812010462249787, + "loss": 0.0, + "step": 6658 + }, + { + "epoch": 3.1336470588235295, + "grad_norm": 0.024239760548481672, + "learning_rate": 0.00018115816753033601, + "loss": 0.0029, + "step": 6659 + }, + { + "epoch": 3.1341176470588237, + "grad_norm": 0.16403470999259318, + "learning_rate": 0.00018111528617463296, + "loss": 0.0765, + "step": 6660 + }, + { + "epoch": 3.1345882352941175, + "grad_norm": 0.1906575431730672, + "learning_rate": 0.00018107240216153184, + "loss": 0.0612, + "step": 6661 + }, + { + "epoch": 3.1350588235294117, + "grad_norm": 0.21258983037767423, + "learning_rate": 0.00018102951549469522, + "loss": 0.0919, + "step": 6662 + }, + { + "epoch": 3.135529411764706, + "grad_norm": 0.04062271751474671, + "learning_rate": 0.00018098662617778573, + "loss": 0.0138, + "step": 6663 + }, + { + "epoch": 3.136, + "grad_norm": 0.15320445314406492, + "learning_rate": 0.0001809437342144664, + "loss": 0.0759, + "step": 6664 + }, + { + "epoch": 3.1364705882352943, + "grad_norm": 0.0998370382185952, + "learning_rate": 0.00018090083960840042, + "loss": 0.0618, + "step": 6665 + }, + { + "epoch": 3.136941176470588, + "grad_norm": 0.07800142996995975, + "learning_rate": 0.0001808579423632511, + "loss": 0.0428, + "step": 6666 + }, + { + "epoch": 3.1374117647058823, + "grad_norm": 0.14129853916449864, + "learning_rate": 0.0001808150424826822, + "loss": 0.0603, + "step": 6667 + }, + { + "epoch": 3.1378823529411766, + "grad_norm": 0.07611182015078018, + "learning_rate": 0.0001807721399703575, + "loss": 0.0379, + "step": 6668 + }, + { + "epoch": 3.138352941176471, + "grad_norm": 0.00011068740836115489, + "learning_rate": 0.0001807292348299411, + "loss": 0.0, + "step": 6669 + }, + { + "epoch": 3.1388235294117646, + "grad_norm": 0.33254808329341246, + "learning_rate": 0.00018068632706509733, + "loss": 0.1533, + "step": 6670 + }, + { + "epoch": 3.139294117647059, + "grad_norm": 0.10152902212320113, + "learning_rate": 0.00018064341667949074, + "loss": 0.0432, + "step": 6671 + }, + { + "epoch": 3.139764705882353, + "grad_norm": 0.12064494977980779, + "learning_rate": 0.00018060050367678607, + "loss": 0.0228, + "step": 6672 + }, + { + "epoch": 3.1402352941176472, + "grad_norm": 0.08131281245090398, + "learning_rate": 0.00018055758806064822, + "loss": 0.0323, + "step": 6673 + }, + { + "epoch": 3.140705882352941, + "grad_norm": 3.1148990455278436e-05, + "learning_rate": 0.00018051466983474252, + "loss": 0.0, + "step": 6674 + }, + { + "epoch": 3.1411764705882352, + "grad_norm": 0.04932157563482051, + "learning_rate": 0.00018047174900273435, + "loss": 0.0107, + "step": 6675 + }, + { + "epoch": 3.1416470588235295, + "grad_norm": 0.1320793636058709, + "learning_rate": 0.00018042882556828927, + "loss": 0.0651, + "step": 6676 + }, + { + "epoch": 3.1421176470588237, + "grad_norm": 0.03290541591246981, + "learning_rate": 0.0001803858995350733, + "loss": 0.0036, + "step": 6677 + }, + { + "epoch": 3.1425882352941175, + "grad_norm": 2.7089680679721644e-05, + "learning_rate": 0.00018034297090675246, + "loss": 0.0, + "step": 6678 + }, + { + "epoch": 3.1430588235294117, + "grad_norm": 0.11293951590250835, + "learning_rate": 0.00018030003968699308, + "loss": 0.0225, + "step": 6679 + }, + { + "epoch": 3.143529411764706, + "grad_norm": 0.26056707786945443, + "learning_rate": 0.00018025710587946166, + "loss": 0.0329, + "step": 6680 + }, + { + "epoch": 3.144, + "grad_norm": 0.13542061344584236, + "learning_rate": 0.00018021416948782503, + "loss": 0.0294, + "step": 6681 + }, + { + "epoch": 3.1444705882352944, + "grad_norm": 0.09019716389212547, + "learning_rate": 0.00018017123051575012, + "loss": 0.0253, + "step": 6682 + }, + { + "epoch": 3.144941176470588, + "grad_norm": 0.10818291742870847, + "learning_rate": 0.0001801282889669041, + "loss": 0.0341, + "step": 6683 + }, + { + "epoch": 3.1454117647058824, + "grad_norm": 0.07090751848538253, + "learning_rate": 0.00018008534484495445, + "loss": 0.0259, + "step": 6684 + }, + { + "epoch": 3.1458823529411766, + "grad_norm": 0.18357799240773426, + "learning_rate": 0.0001800423981535688, + "loss": 0.0606, + "step": 6685 + }, + { + "epoch": 3.1463529411764704, + "grad_norm": 0.12495796456075753, + "learning_rate": 0.000179999448896415, + "loss": 0.0697, + "step": 6686 + }, + { + "epoch": 3.1468235294117646, + "grad_norm": 0.14105882838395203, + "learning_rate": 0.00017995649707716103, + "loss": 0.0817, + "step": 6687 + }, + { + "epoch": 3.147294117647059, + "grad_norm": 0.0767634687697281, + "learning_rate": 0.00017991354269947533, + "loss": 0.0247, + "step": 6688 + }, + { + "epoch": 3.147764705882353, + "grad_norm": 0.10014125485113151, + "learning_rate": 0.00017987058576702638, + "loss": 0.0512, + "step": 6689 + }, + { + "epoch": 3.1482352941176472, + "grad_norm": 0.1225126844030975, + "learning_rate": 0.00017982762628348286, + "loss": 0.0893, + "step": 6690 + }, + { + "epoch": 3.148705882352941, + "grad_norm": 0.13730417383184487, + "learning_rate": 0.00017978466425251375, + "loss": 0.0166, + "step": 6691 + }, + { + "epoch": 3.1491764705882352, + "grad_norm": 0.3326878080935835, + "learning_rate": 0.00017974169967778825, + "loss": 0.0767, + "step": 6692 + }, + { + "epoch": 3.1496470588235295, + "grad_norm": 0.02049116728292814, + "learning_rate": 0.00017969873256297572, + "loss": 0.0024, + "step": 6693 + }, + { + "epoch": 3.1501176470588237, + "grad_norm": 0.12886619381779288, + "learning_rate": 0.0001796557629117457, + "loss": 0.0547, + "step": 6694 + }, + { + "epoch": 3.1505882352941175, + "grad_norm": 1.2598779704810899e-05, + "learning_rate": 0.00017961279072776812, + "loss": 0.0, + "step": 6695 + }, + { + "epoch": 3.1510588235294117, + "grad_norm": 0.32501028333222054, + "learning_rate": 0.00017956981601471294, + "loss": 0.1028, + "step": 6696 + }, + { + "epoch": 3.151529411764706, + "grad_norm": 0.11653227433112952, + "learning_rate": 0.00017952683877625042, + "loss": 0.0506, + "step": 6697 + }, + { + "epoch": 3.152, + "grad_norm": 0.069821830368568, + "learning_rate": 0.00017948385901605106, + "loss": 0.0041, + "step": 6698 + }, + { + "epoch": 3.152470588235294, + "grad_norm": 2.7848179164563023e-05, + "learning_rate": 0.0001794408767377855, + "loss": 0.0, + "step": 6699 + }, + { + "epoch": 3.152941176470588, + "grad_norm": 0.11224679415926558, + "learning_rate": 0.00017939789194512472, + "loss": 0.0578, + "step": 6700 + }, + { + "epoch": 3.1534117647058824, + "grad_norm": 0.09227120375451324, + "learning_rate": 0.0001793549046417397, + "loss": 0.0402, + "step": 6701 + }, + { + "epoch": 3.1538823529411766, + "grad_norm": 0.050863125483654, + "learning_rate": 0.00017931191483130186, + "loss": 0.0171, + "step": 6702 + }, + { + "epoch": 3.1543529411764704, + "grad_norm": 0.12294499152760177, + "learning_rate": 0.00017926892251748274, + "loss": 0.0598, + "step": 6703 + }, + { + "epoch": 3.1548235294117646, + "grad_norm": 0.10342720768809964, + "learning_rate": 0.00017922592770395405, + "loss": 0.0532, + "step": 6704 + }, + { + "epoch": 3.155294117647059, + "grad_norm": 0.09043688127117291, + "learning_rate": 0.00017918293039438781, + "loss": 0.0373, + "step": 6705 + }, + { + "epoch": 3.155764705882353, + "grad_norm": 0.06674087778841652, + "learning_rate": 0.00017913993059245616, + "loss": 0.0286, + "step": 6706 + }, + { + "epoch": 3.1562352941176472, + "grad_norm": 5.720825618829838e-05, + "learning_rate": 0.00017909692830183152, + "loss": 0.0, + "step": 6707 + }, + { + "epoch": 3.156705882352941, + "grad_norm": 0.07186461390783834, + "learning_rate": 0.00017905392352618647, + "loss": 0.023, + "step": 6708 + }, + { + "epoch": 3.1571764705882353, + "grad_norm": 6.203859418952106e-05, + "learning_rate": 0.00017901091626919383, + "loss": 0.0, + "step": 6709 + }, + { + "epoch": 3.1576470588235295, + "grad_norm": 0.09901657178663235, + "learning_rate": 0.00017896790653452668, + "loss": 0.0366, + "step": 6710 + }, + { + "epoch": 3.1581176470588237, + "grad_norm": 0.08072737602613973, + "learning_rate": 0.00017892489432585817, + "loss": 0.0231, + "step": 6711 + }, + { + "epoch": 3.1585882352941175, + "grad_norm": 0.09923386934093319, + "learning_rate": 0.00017888187964686182, + "loss": 0.0287, + "step": 6712 + }, + { + "epoch": 3.1590588235294117, + "grad_norm": 0.06038181574431785, + "learning_rate": 0.00017883886250121133, + "loss": 0.014, + "step": 6713 + }, + { + "epoch": 3.159529411764706, + "grad_norm": 0.08928679044519106, + "learning_rate": 0.00017879584289258048, + "loss": 0.0349, + "step": 6714 + }, + { + "epoch": 3.16, + "grad_norm": 0.3301249305000096, + "learning_rate": 0.00017875282082464343, + "loss": 0.0925, + "step": 6715 + }, + { + "epoch": 3.160470588235294, + "grad_norm": 0.06727731754878351, + "learning_rate": 0.00017870979630107446, + "loss": 0.0195, + "step": 6716 + }, + { + "epoch": 3.160941176470588, + "grad_norm": 9.505003496592743e-05, + "learning_rate": 0.00017866676932554802, + "loss": 0.0, + "step": 6717 + }, + { + "epoch": 3.1614117647058824, + "grad_norm": 8.562881391360023e-05, + "learning_rate": 0.0001786237399017389, + "loss": 0.0, + "step": 6718 + }, + { + "epoch": 3.1618823529411766, + "grad_norm": 0.10152440963064474, + "learning_rate": 0.000178580708033322, + "loss": 0.0297, + "step": 6719 + }, + { + "epoch": 3.1623529411764704, + "grad_norm": 0.07114986702213129, + "learning_rate": 0.00017853767372397243, + "loss": 0.0222, + "step": 6720 + }, + { + "epoch": 3.1628235294117646, + "grad_norm": 0.07412576483855993, + "learning_rate": 0.00017849463697736554, + "loss": 0.025, + "step": 6721 + }, + { + "epoch": 3.163294117647059, + "grad_norm": 0.0001129595650668251, + "learning_rate": 0.0001784515977971769, + "loss": 0.0, + "step": 6722 + }, + { + "epoch": 3.163764705882353, + "grad_norm": 0.00011716253645142372, + "learning_rate": 0.00017840855618708227, + "loss": 0.0, + "step": 6723 + }, + { + "epoch": 3.1642352941176473, + "grad_norm": 0.0015753007852516074, + "learning_rate": 0.00017836551215075758, + "loss": 0.0, + "step": 6724 + }, + { + "epoch": 3.164705882352941, + "grad_norm": 0.11131471053745899, + "learning_rate": 0.000178322465691879, + "loss": 0.0654, + "step": 6725 + }, + { + "epoch": 3.1651764705882353, + "grad_norm": 0.11783781185672575, + "learning_rate": 0.00017827941681412298, + "loss": 0.0511, + "step": 6726 + }, + { + "epoch": 3.1656470588235295, + "grad_norm": 0.04952847088724171, + "learning_rate": 0.00017823636552116603, + "loss": 0.0179, + "step": 6727 + }, + { + "epoch": 3.1661176470588237, + "grad_norm": 0.0421640008266885, + "learning_rate": 0.00017819331181668496, + "loss": 0.0075, + "step": 6728 + }, + { + "epoch": 3.1665882352941175, + "grad_norm": 0.10800064214384143, + "learning_rate": 0.00017815025570435677, + "loss": 0.0482, + "step": 6729 + }, + { + "epoch": 3.1670588235294117, + "grad_norm": 0.21559620861204706, + "learning_rate": 0.0001781071971878587, + "loss": 0.0791, + "step": 6730 + }, + { + "epoch": 3.167529411764706, + "grad_norm": 0.029105939739985553, + "learning_rate": 0.00017806413627086807, + "loss": 0.0027, + "step": 6731 + }, + { + "epoch": 3.168, + "grad_norm": 0.09544048516968343, + "learning_rate": 0.0001780210729570626, + "loss": 0.0489, + "step": 6732 + }, + { + "epoch": 3.168470588235294, + "grad_norm": 0.12222824765975446, + "learning_rate": 0.00017797800725012002, + "loss": 0.0465, + "step": 6733 + }, + { + "epoch": 3.168941176470588, + "grad_norm": 0.07818191547425803, + "learning_rate": 0.00017793493915371842, + "loss": 0.0367, + "step": 6734 + }, + { + "epoch": 3.1694117647058824, + "grad_norm": 0.02032922671671647, + "learning_rate": 0.00017789186867153594, + "loss": 0.003, + "step": 6735 + }, + { + "epoch": 3.1698823529411766, + "grad_norm": 0.10724915707620099, + "learning_rate": 0.00017784879580725113, + "loss": 0.0535, + "step": 6736 + }, + { + "epoch": 3.1703529411764704, + "grad_norm": 0.09422692524187903, + "learning_rate": 0.00017780572056454256, + "loss": 0.0357, + "step": 6737 + }, + { + "epoch": 3.1708235294117646, + "grad_norm": 0.0735940580323194, + "learning_rate": 0.00017776264294708905, + "loss": 0.0237, + "step": 6738 + }, + { + "epoch": 3.171294117647059, + "grad_norm": 0.06753987494748093, + "learning_rate": 0.00017771956295856964, + "loss": 0.0164, + "step": 6739 + }, + { + "epoch": 3.171764705882353, + "grad_norm": 0.16594681049401858, + "learning_rate": 0.0001776764806026636, + "loss": 0.0312, + "step": 6740 + }, + { + "epoch": 3.1722352941176473, + "grad_norm": 9.664181960674567e-05, + "learning_rate": 0.00017763339588305037, + "loss": 0.0, + "step": 6741 + }, + { + "epoch": 3.172705882352941, + "grad_norm": 0.07102723054387691, + "learning_rate": 0.00017759030880340956, + "loss": 0.0242, + "step": 6742 + }, + { + "epoch": 3.1731764705882353, + "grad_norm": 0.00016571227447316787, + "learning_rate": 0.00017754721936742108, + "loss": 0.0, + "step": 6743 + }, + { + "epoch": 3.1736470588235295, + "grad_norm": 0.00011586120166105496, + "learning_rate": 0.00017750412757876492, + "loss": 0.0, + "step": 6744 + }, + { + "epoch": 3.1741176470588237, + "grad_norm": 0.05465167504862354, + "learning_rate": 0.00017746103344112136, + "loss": 0.0211, + "step": 6745 + }, + { + "epoch": 3.1745882352941175, + "grad_norm": 0.00011825761620262784, + "learning_rate": 0.00017741793695817093, + "loss": 0.0, + "step": 6746 + }, + { + "epoch": 3.1750588235294117, + "grad_norm": 7.86674910849485e-05, + "learning_rate": 0.00017737483813359413, + "loss": 0.0, + "step": 6747 + }, + { + "epoch": 3.175529411764706, + "grad_norm": 0.0859087068765284, + "learning_rate": 0.00017733173697107187, + "loss": 0.0273, + "step": 6748 + }, + { + "epoch": 3.176, + "grad_norm": 0.00013785233197320806, + "learning_rate": 0.00017728863347428524, + "loss": 0.0, + "step": 6749 + }, + { + "epoch": 3.176470588235294, + "grad_norm": 0.057957873032279124, + "learning_rate": 0.00017724552764691545, + "loss": 0.0148, + "step": 6750 + }, + { + "epoch": 3.176941176470588, + "grad_norm": 0.1722968842566579, + "learning_rate": 0.000177202419492644, + "loss": 0.0908, + "step": 6751 + }, + { + "epoch": 3.1774117647058824, + "grad_norm": 0.17745098615687102, + "learning_rate": 0.00017715930901515242, + "loss": 0.0744, + "step": 6752 + }, + { + "epoch": 3.1778823529411766, + "grad_norm": 0.12170052452693716, + "learning_rate": 0.0001771161962181227, + "loss": 0.0509, + "step": 6753 + }, + { + "epoch": 3.1783529411764704, + "grad_norm": 0.10213503595724038, + "learning_rate": 0.0001770730811052368, + "loss": 0.0318, + "step": 6754 + }, + { + "epoch": 3.1788235294117646, + "grad_norm": 0.11531150016308224, + "learning_rate": 0.00017702996368017693, + "loss": 0.0501, + "step": 6755 + }, + { + "epoch": 3.179294117647059, + "grad_norm": 0.016071972221507157, + "learning_rate": 0.00017698684394662563, + "loss": 0.0026, + "step": 6756 + }, + { + "epoch": 3.179764705882353, + "grad_norm": 0.08850134525737384, + "learning_rate": 0.0001769437219082655, + "loss": 0.0463, + "step": 6757 + }, + { + "epoch": 3.1802352941176473, + "grad_norm": 0.3356947778179355, + "learning_rate": 0.00017690059756877933, + "loss": 0.0908, + "step": 6758 + }, + { + "epoch": 3.180705882352941, + "grad_norm": 0.0783640734212206, + "learning_rate": 0.00017685747093185017, + "loss": 0.0334, + "step": 6759 + }, + { + "epoch": 3.1811764705882353, + "grad_norm": 0.09529207819518744, + "learning_rate": 0.00017681434200116132, + "loss": 0.019, + "step": 6760 + }, + { + "epoch": 3.1816470588235295, + "grad_norm": 0.09046619281065081, + "learning_rate": 0.0001767712107803961, + "loss": 0.0489, + "step": 6761 + }, + { + "epoch": 3.1821176470588237, + "grad_norm": 0.08208287863399542, + "learning_rate": 0.00017672807727323818, + "loss": 0.0265, + "step": 6762 + }, + { + "epoch": 3.1825882352941175, + "grad_norm": 0.044026836333811985, + "learning_rate": 0.00017668494148337133, + "loss": 0.007, + "step": 6763 + }, + { + "epoch": 3.1830588235294117, + "grad_norm": 0.23457140223729145, + "learning_rate": 0.00017664180341447964, + "loss": 0.0878, + "step": 6764 + }, + { + "epoch": 3.183529411764706, + "grad_norm": 0.08521603742741156, + "learning_rate": 0.0001765986630702472, + "loss": 0.0374, + "step": 6765 + }, + { + "epoch": 3.184, + "grad_norm": 0.00011225598401595839, + "learning_rate": 0.00017655552045435846, + "loss": 0.0, + "step": 6766 + }, + { + "epoch": 3.184470588235294, + "grad_norm": 0.02184634486693696, + "learning_rate": 0.00017651237557049808, + "loss": 0.0006, + "step": 6767 + }, + { + "epoch": 3.184941176470588, + "grad_norm": 0.07004369606635562, + "learning_rate": 0.00017646922842235076, + "loss": 0.0222, + "step": 6768 + }, + { + "epoch": 3.1854117647058824, + "grad_norm": 0.12949071478540314, + "learning_rate": 0.00017642607901360147, + "loss": 0.0761, + "step": 6769 + }, + { + "epoch": 3.1858823529411766, + "grad_norm": 0.3060368803750785, + "learning_rate": 0.00017638292734793547, + "loss": 0.1518, + "step": 6770 + }, + { + "epoch": 3.1863529411764704, + "grad_norm": 0.11077665983450594, + "learning_rate": 0.00017633977342903807, + "loss": 0.0621, + "step": 6771 + }, + { + "epoch": 3.1868235294117646, + "grad_norm": 0.08832896723701232, + "learning_rate": 0.0001762966172605948, + "loss": 0.0314, + "step": 6772 + }, + { + "epoch": 3.187294117647059, + "grad_norm": 1.9818202084644407e-05, + "learning_rate": 0.00017625345884629141, + "loss": 0.0, + "step": 6773 + }, + { + "epoch": 3.187764705882353, + "grad_norm": 0.09573754543476587, + "learning_rate": 0.00017621029818981387, + "loss": 0.0389, + "step": 6774 + }, + { + "epoch": 3.1882352941176473, + "grad_norm": 0.08732394510366007, + "learning_rate": 0.00017616713529484835, + "loss": 0.0311, + "step": 6775 + }, + { + "epoch": 3.188705882352941, + "grad_norm": 3.942716135007217e-05, + "learning_rate": 0.00017612397016508104, + "loss": 0.0, + "step": 6776 + }, + { + "epoch": 3.1891764705882353, + "grad_norm": 0.16986460178084173, + "learning_rate": 0.0001760808028041986, + "loss": 0.0715, + "step": 6777 + }, + { + "epoch": 3.1896470588235295, + "grad_norm": 0.033319125271821334, + "learning_rate": 0.0001760376332158877, + "loss": 0.0068, + "step": 6778 + }, + { + "epoch": 3.1901176470588237, + "grad_norm": 0.07689971673213135, + "learning_rate": 0.00017599446140383512, + "loss": 0.0204, + "step": 6779 + }, + { + "epoch": 3.1905882352941175, + "grad_norm": 0.13779470486340187, + "learning_rate": 0.00017595128737172812, + "loss": 0.0965, + "step": 6780 + }, + { + "epoch": 3.1910588235294117, + "grad_norm": 1.2708932071138013e-05, + "learning_rate": 0.00017590811112325384, + "loss": 0.0, + "step": 6781 + }, + { + "epoch": 3.191529411764706, + "grad_norm": 0.1296149623372176, + "learning_rate": 0.00017586493266209978, + "loss": 0.0706, + "step": 6782 + }, + { + "epoch": 3.192, + "grad_norm": 1.7830968963420487e-05, + "learning_rate": 0.00017582175199195365, + "loss": 0.0, + "step": 6783 + }, + { + "epoch": 3.192470588235294, + "grad_norm": 0.13801978199258932, + "learning_rate": 0.00017577856911650324, + "loss": 0.0502, + "step": 6784 + }, + { + "epoch": 3.192941176470588, + "grad_norm": 0.08080679161673036, + "learning_rate": 0.00017573538403943656, + "loss": 0.0356, + "step": 6785 + }, + { + "epoch": 3.1934117647058824, + "grad_norm": 0.0722451563154447, + "learning_rate": 0.00017569219676444187, + "loss": 0.0306, + "step": 6786 + }, + { + "epoch": 3.1938823529411766, + "grad_norm": 0.10029076193380464, + "learning_rate": 0.00017564900729520751, + "loss": 0.0499, + "step": 6787 + }, + { + "epoch": 3.1943529411764704, + "grad_norm": 0.1026142406265031, + "learning_rate": 0.0001756058156354222, + "loss": 0.0486, + "step": 6788 + }, + { + "epoch": 3.1948235294117646, + "grad_norm": 0.17803571244190616, + "learning_rate": 0.00017556262178877459, + "loss": 0.1092, + "step": 6789 + }, + { + "epoch": 3.195294117647059, + "grad_norm": 0.16725539203956497, + "learning_rate": 0.00017551942575895368, + "loss": 0.0911, + "step": 6790 + }, + { + "epoch": 3.195764705882353, + "grad_norm": 755.9763981822995, + "learning_rate": 0.00017547622754964868, + "loss": 0.3785, + "step": 6791 + }, + { + "epoch": 3.1962352941176473, + "grad_norm": 2.136045020845825e-06, + "learning_rate": 0.0001754330271645489, + "loss": 0.0, + "step": 6792 + }, + { + "epoch": 3.196705882352941, + "grad_norm": 0.0735130540573552, + "learning_rate": 0.00017538982460734375, + "loss": 0.0253, + "step": 6793 + }, + { + "epoch": 3.1971764705882353, + "grad_norm": 0.08395676917837978, + "learning_rate": 0.00017534661988172315, + "loss": 0.025, + "step": 6794 + }, + { + "epoch": 3.1976470588235295, + "grad_norm": 0.07830103081092049, + "learning_rate": 0.00017530341299137685, + "loss": 0.019, + "step": 6795 + }, + { + "epoch": 3.1981176470588237, + "grad_norm": 0.7360597021338242, + "learning_rate": 0.00017526020393999496, + "loss": 0.1478, + "step": 6796 + }, + { + "epoch": 3.1985882352941175, + "grad_norm": 4.484341498105129, + "learning_rate": 0.00017521699273126778, + "loss": 0.0543, + "step": 6797 + }, + { + "epoch": 3.1990588235294117, + "grad_norm": 4.985738858439985, + "learning_rate": 0.00017517377936888568, + "loss": 0.1393, + "step": 6798 + }, + { + "epoch": 3.199529411764706, + "grad_norm": 0.15444266290200048, + "learning_rate": 0.0001751305638565394, + "loss": 0.0448, + "step": 6799 + }, + { + "epoch": 3.2, + "grad_norm": 0.09334153113346366, + "learning_rate": 0.00017508734619791966, + "loss": 0.0486, + "step": 6800 + }, + { + "epoch": 3.200470588235294, + "grad_norm": 0.07306591211712066, + "learning_rate": 0.00017504412639671754, + "loss": 0.022, + "step": 6801 + }, + { + "epoch": 3.200941176470588, + "grad_norm": 1.1902364314311926e-10, + "learning_rate": 0.00017500090445662415, + "loss": 0.0, + "step": 6802 + }, + { + "epoch": 3.2014117647058824, + "grad_norm": 0.012269900633470408, + "learning_rate": 0.0001749576803813309, + "loss": 0.0003, + "step": 6803 + }, + { + "epoch": 3.2018823529411766, + "grad_norm": 0.021099057009861463, + "learning_rate": 0.00017491445417452934, + "loss": 0.0031, + "step": 6804 + }, + { + "epoch": 3.2023529411764704, + "grad_norm": 0.15652575788964718, + "learning_rate": 0.00017487122583991123, + "loss": 0.1164, + "step": 6805 + }, + { + "epoch": 3.2028235294117646, + "grad_norm": 0.11634077612145068, + "learning_rate": 0.00017482799538116834, + "loss": 0.0723, + "step": 6806 + }, + { + "epoch": 3.203294117647059, + "grad_norm": 0.2375554820971858, + "learning_rate": 0.00017478476280199294, + "loss": 0.1013, + "step": 6807 + }, + { + "epoch": 3.203764705882353, + "grad_norm": 0.10681817430540225, + "learning_rate": 0.0001747415281060772, + "loss": 0.043, + "step": 6808 + }, + { + "epoch": 3.2042352941176473, + "grad_norm": 0.07301312771496811, + "learning_rate": 0.00017469829129711361, + "loss": 0.0413, + "step": 6809 + }, + { + "epoch": 3.204705882352941, + "grad_norm": 0.07592777706342946, + "learning_rate": 0.00017465505237879476, + "loss": 0.0101, + "step": 6810 + }, + { + "epoch": 3.2051764705882353, + "grad_norm": 0.056321432580629834, + "learning_rate": 0.00017461181135481355, + "loss": 0.0265, + "step": 6811 + }, + { + "epoch": 3.2056470588235295, + "grad_norm": 9.481905424887444e-10, + "learning_rate": 0.00017456856822886297, + "loss": 0.0, + "step": 6812 + }, + { + "epoch": 3.2061176470588233, + "grad_norm": 0.08307225536148558, + "learning_rate": 0.00017452532300463609, + "loss": 0.0506, + "step": 6813 + }, + { + "epoch": 3.2065882352941175, + "grad_norm": 0.079318467812207, + "learning_rate": 0.00017448207568582634, + "loss": 0.0362, + "step": 6814 + }, + { + "epoch": 3.2070588235294117, + "grad_norm": 0.13202214643500193, + "learning_rate": 0.00017443882627612726, + "loss": 0.0669, + "step": 6815 + }, + { + "epoch": 3.207529411764706, + "grad_norm": 0.12052979792654082, + "learning_rate": 0.00017439557477923255, + "loss": 0.0574, + "step": 6816 + }, + { + "epoch": 3.208, + "grad_norm": 0.05530833420762806, + "learning_rate": 0.00017435232119883604, + "loss": 0.0217, + "step": 6817 + }, + { + "epoch": 3.208470588235294, + "grad_norm": 0.1133594952207288, + "learning_rate": 0.0001743090655386319, + "loss": 0.0537, + "step": 6818 + }, + { + "epoch": 3.208941176470588, + "grad_norm": 0.09307324983702697, + "learning_rate": 0.00017426580780231435, + "loss": 0.0589, + "step": 6819 + }, + { + "epoch": 3.2094117647058824, + "grad_norm": 0.08031304071891228, + "learning_rate": 0.0001742225479935777, + "loss": 0.0395, + "step": 6820 + }, + { + "epoch": 3.2098823529411766, + "grad_norm": 0.06565463133543112, + "learning_rate": 0.00017417928611611675, + "loss": 0.0337, + "step": 6821 + }, + { + "epoch": 3.2103529411764704, + "grad_norm": 0.09669683349813783, + "learning_rate": 0.0001741360221736261, + "loss": 0.0452, + "step": 6822 + }, + { + "epoch": 3.2108235294117646, + "grad_norm": 1.15686815888634e-09, + "learning_rate": 0.0001740927561698008, + "loss": 0.0, + "step": 6823 + }, + { + "epoch": 3.211294117647059, + "grad_norm": 0.1576448949909591, + "learning_rate": 0.00017404948810833594, + "loss": 0.0394, + "step": 6824 + }, + { + "epoch": 3.211764705882353, + "grad_norm": 0.05088921694779953, + "learning_rate": 0.0001740062179929268, + "loss": 0.0166, + "step": 6825 + }, + { + "epoch": 3.212235294117647, + "grad_norm": 8.674014657936149e-09, + "learning_rate": 0.00017396294582726896, + "loss": 0.0, + "step": 6826 + }, + { + "epoch": 3.212705882352941, + "grad_norm": 0.14489466341682833, + "learning_rate": 0.00017391967161505795, + "loss": 0.0478, + "step": 6827 + }, + { + "epoch": 3.2131764705882353, + "grad_norm": 0.013720853493561632, + "learning_rate": 0.00017387639535998967, + "loss": 0.0012, + "step": 6828 + }, + { + "epoch": 3.2136470588235295, + "grad_norm": 0.04965893042433472, + "learning_rate": 0.00017383311706576016, + "loss": 0.0166, + "step": 6829 + }, + { + "epoch": 3.2141176470588233, + "grad_norm": 0.15336286083927023, + "learning_rate": 0.00017378983673606548, + "loss": 0.0827, + "step": 6830 + }, + { + "epoch": 3.2145882352941175, + "grad_norm": 0.08258175850583056, + "learning_rate": 0.00017374655437460206, + "loss": 0.03, + "step": 6831 + }, + { + "epoch": 3.2150588235294117, + "grad_norm": 0.03968002361350125, + "learning_rate": 0.00017370326998506646, + "loss": 0.0072, + "step": 6832 + }, + { + "epoch": 3.215529411764706, + "grad_norm": 0.0800182213470713, + "learning_rate": 0.00017365998357115535, + "loss": 0.0248, + "step": 6833 + }, + { + "epoch": 3.216, + "grad_norm": 0.33278713738535115, + "learning_rate": 0.00017361669513656558, + "loss": 0.0557, + "step": 6834 + }, + { + "epoch": 3.216470588235294, + "grad_norm": 9.475191428172764e-09, + "learning_rate": 0.0001735734046849942, + "loss": 0.0, + "step": 6835 + }, + { + "epoch": 3.216941176470588, + "grad_norm": 0.10460472659128563, + "learning_rate": 0.0001735301122201384, + "loss": 0.0518, + "step": 6836 + }, + { + "epoch": 3.2174117647058824, + "grad_norm": 0.08507947610001268, + "learning_rate": 0.00017348681774569563, + "loss": 0.035, + "step": 6837 + }, + { + "epoch": 3.2178823529411766, + "grad_norm": 0.13821970135582906, + "learning_rate": 0.00017344352126536343, + "loss": 0.0727, + "step": 6838 + }, + { + "epoch": 3.2183529411764704, + "grad_norm": 0.11268986890697648, + "learning_rate": 0.00017340022278283953, + "loss": 0.0486, + "step": 6839 + }, + { + "epoch": 3.2188235294117646, + "grad_norm": 0.041953055104420475, + "learning_rate": 0.00017335692230182183, + "loss": 0.0059, + "step": 6840 + }, + { + "epoch": 3.219294117647059, + "grad_norm": 0.11051876149004326, + "learning_rate": 0.00017331361982600836, + "loss": 0.0614, + "step": 6841 + }, + { + "epoch": 3.219764705882353, + "grad_norm": 0.03415248738041794, + "learning_rate": 0.00017327031535909745, + "loss": 0.0073, + "step": 6842 + }, + { + "epoch": 3.220235294117647, + "grad_norm": 0.12664583291000958, + "learning_rate": 0.00017322700890478744, + "loss": 0.0524, + "step": 6843 + }, + { + "epoch": 3.220705882352941, + "grad_norm": 0.16676205535166178, + "learning_rate": 0.00017318370046677693, + "loss": 0.1039, + "step": 6844 + }, + { + "epoch": 3.2211764705882353, + "grad_norm": 0.04306646437670053, + "learning_rate": 0.00017314039004876473, + "loss": 0.0108, + "step": 6845 + }, + { + "epoch": 3.2216470588235295, + "grad_norm": 0.10333995365698563, + "learning_rate": 0.0001730970776544497, + "loss": 0.0618, + "step": 6846 + }, + { + "epoch": 3.2221176470588233, + "grad_norm": 0.1299296166332647, + "learning_rate": 0.00017305376328753092, + "loss": 0.0557, + "step": 6847 + }, + { + "epoch": 3.2225882352941175, + "grad_norm": 2.9076900089223865e-09, + "learning_rate": 0.00017301044695170769, + "loss": 0.0, + "step": 6848 + }, + { + "epoch": 3.2230588235294118, + "grad_norm": 0.07559350157785573, + "learning_rate": 0.00017296712865067944, + "loss": 0.0066, + "step": 6849 + }, + { + "epoch": 3.223529411764706, + "grad_norm": 0.212005362606209, + "learning_rate": 0.00017292380838814577, + "loss": 0.0887, + "step": 6850 + }, + { + "epoch": 3.224, + "grad_norm": 0.10772179102233446, + "learning_rate": 0.00017288048616780636, + "loss": 0.0604, + "step": 6851 + }, + { + "epoch": 3.224470588235294, + "grad_norm": 0.12212275954130672, + "learning_rate": 0.00017283716199336123, + "loss": 0.058, + "step": 6852 + }, + { + "epoch": 3.224941176470588, + "grad_norm": 0.10088152114849397, + "learning_rate": 0.0001727938358685105, + "loss": 0.0432, + "step": 6853 + }, + { + "epoch": 3.2254117647058824, + "grad_norm": 0.128878361063846, + "learning_rate": 0.0001727505077969543, + "loss": 0.0612, + "step": 6854 + }, + { + "epoch": 3.2258823529411766, + "grad_norm": 0.09318917097929266, + "learning_rate": 0.00017270717778239316, + "loss": 0.0408, + "step": 6855 + }, + { + "epoch": 3.2263529411764704, + "grad_norm": 0.09612697683767081, + "learning_rate": 0.0001726638458285277, + "loss": 0.037, + "step": 6856 + }, + { + "epoch": 3.2268235294117646, + "grad_norm": 1.2427460136550676e-09, + "learning_rate": 0.00017262051193905865, + "loss": 0.0, + "step": 6857 + }, + { + "epoch": 3.227294117647059, + "grad_norm": 1.1556577950861828e-09, + "learning_rate": 0.00017257717611768685, + "loss": 0.0, + "step": 6858 + }, + { + "epoch": 3.227764705882353, + "grad_norm": 1.1120223907331324e-09, + "learning_rate": 0.00017253383836811356, + "loss": 0.0, + "step": 6859 + }, + { + "epoch": 3.228235294117647, + "grad_norm": 1.4439236959392133e-09, + "learning_rate": 0.0001724904986940399, + "loss": 0.0, + "step": 6860 + }, + { + "epoch": 3.228705882352941, + "grad_norm": 1.4490124467557696e-09, + "learning_rate": 0.00017244715709916733, + "loss": 0.0, + "step": 6861 + }, + { + "epoch": 3.2291764705882353, + "grad_norm": 0.09952171646889668, + "learning_rate": 0.00017240381358719746, + "loss": 0.0616, + "step": 6862 + }, + { + "epoch": 3.2296470588235295, + "grad_norm": 0.04224510877715732, + "learning_rate": 0.000172360468161832, + "loss": 0.0083, + "step": 6863 + }, + { + "epoch": 3.2301176470588233, + "grad_norm": 0.1533967253149652, + "learning_rate": 0.00017231712082677291, + "loss": 0.065, + "step": 6864 + }, + { + "epoch": 3.2305882352941175, + "grad_norm": 0.1505849022637686, + "learning_rate": 0.0001722737715857222, + "loss": 0.0857, + "step": 6865 + }, + { + "epoch": 3.2310588235294118, + "grad_norm": 0.08368839471030771, + "learning_rate": 0.00017223042044238219, + "loss": 0.0363, + "step": 6866 + }, + { + "epoch": 3.231529411764706, + "grad_norm": 0.10021800334364442, + "learning_rate": 0.0001721870674004552, + "loss": 0.0243, + "step": 6867 + }, + { + "epoch": 3.232, + "grad_norm": 0.07186236851934441, + "learning_rate": 0.00017214371246364384, + "loss": 0.0321, + "step": 6868 + }, + { + "epoch": 3.232470588235294, + "grad_norm": 0.019043464456092486, + "learning_rate": 0.00017210035563565082, + "loss": 0.0009, + "step": 6869 + }, + { + "epoch": 3.232941176470588, + "grad_norm": 0.1509667130746055, + "learning_rate": 0.00017205699692017902, + "loss": 0.0601, + "step": 6870 + }, + { + "epoch": 3.2334117647058824, + "grad_norm": 0.10153604293335772, + "learning_rate": 0.0001720136363209315, + "loss": 0.054, + "step": 6871 + }, + { + "epoch": 3.2338823529411767, + "grad_norm": 0.1550213659269885, + "learning_rate": 0.0001719702738416115, + "loss": 0.0855, + "step": 6872 + }, + { + "epoch": 3.2343529411764704, + "grad_norm": 0.425281534774063, + "learning_rate": 0.0001719269094859223, + "loss": 0.158, + "step": 6873 + }, + { + "epoch": 3.2348235294117647, + "grad_norm": 0.054539553061239754, + "learning_rate": 0.0001718835432575675, + "loss": 0.0139, + "step": 6874 + }, + { + "epoch": 3.235294117647059, + "grad_norm": 1.1095229024196547e-09, + "learning_rate": 0.00017184017516025076, + "loss": 0.0, + "step": 6875 + }, + { + "epoch": 3.235764705882353, + "grad_norm": 0.06882397275876867, + "learning_rate": 0.0001717968051976759, + "loss": 0.0255, + "step": 6876 + }, + { + "epoch": 3.236235294117647, + "grad_norm": 1.0072947147824841e-09, + "learning_rate": 0.00017175343337354698, + "loss": 0.0, + "step": 6877 + }, + { + "epoch": 3.236705882352941, + "grad_norm": 0.04637941892625508, + "learning_rate": 0.00017171005969156817, + "loss": 0.0133, + "step": 6878 + }, + { + "epoch": 3.2371764705882353, + "grad_norm": 0.17250780290994416, + "learning_rate": 0.00017166668415544376, + "loss": 0.0335, + "step": 6879 + }, + { + "epoch": 3.2376470588235295, + "grad_norm": 0.06744513631119298, + "learning_rate": 0.0001716233067688783, + "loss": 0.0232, + "step": 6880 + }, + { + "epoch": 3.2381176470588233, + "grad_norm": 0.10056087555179793, + "learning_rate": 0.00017157992753557633, + "loss": 0.0331, + "step": 6881 + }, + { + "epoch": 3.2385882352941175, + "grad_norm": 5.7632300439902e-09, + "learning_rate": 0.0001715365464592427, + "loss": 0.0, + "step": 6882 + }, + { + "epoch": 3.2390588235294118, + "grad_norm": 0.07671891531581834, + "learning_rate": 0.00017149316354358237, + "loss": 0.0294, + "step": 6883 + }, + { + "epoch": 3.239529411764706, + "grad_norm": 0.06383177081273582, + "learning_rate": 0.00017144977879230046, + "loss": 0.0167, + "step": 6884 + }, + { + "epoch": 3.24, + "grad_norm": 0.15848488251270587, + "learning_rate": 0.00017140639220910222, + "loss": 0.0715, + "step": 6885 + }, + { + "epoch": 3.240470588235294, + "grad_norm": 0.09092195909759819, + "learning_rate": 0.00017136300379769312, + "loss": 0.0388, + "step": 6886 + }, + { + "epoch": 3.240941176470588, + "grad_norm": 9.21722804959515e-10, + "learning_rate": 0.00017131961356177869, + "loss": 0.0, + "step": 6887 + }, + { + "epoch": 3.2414117647058824, + "grad_norm": 0.038797364928145256, + "learning_rate": 0.00017127622150506472, + "loss": 0.0098, + "step": 6888 + }, + { + "epoch": 3.2418823529411767, + "grad_norm": 0.21786886884301018, + "learning_rate": 0.000171232827631257, + "loss": 0.0788, + "step": 6889 + }, + { + "epoch": 3.2423529411764704, + "grad_norm": 0.15056125648460222, + "learning_rate": 0.00017118943194406169, + "loss": 0.0804, + "step": 6890 + }, + { + "epoch": 3.2428235294117647, + "grad_norm": 0.1659794672225596, + "learning_rate": 0.00017114603444718496, + "loss": 0.0927, + "step": 6891 + }, + { + "epoch": 3.243294117647059, + "grad_norm": 0.1397915666101485, + "learning_rate": 0.00017110263514433317, + "loss": 0.075, + "step": 6892 + }, + { + "epoch": 3.243764705882353, + "grad_norm": 0.09841986763427278, + "learning_rate": 0.00017105923403921283, + "loss": 0.0339, + "step": 6893 + }, + { + "epoch": 3.244235294117647, + "grad_norm": 0.12988915255542136, + "learning_rate": 0.00017101583113553063, + "loss": 0.0794, + "step": 6894 + }, + { + "epoch": 3.244705882352941, + "grad_norm": 0.10708654942448635, + "learning_rate": 0.00017097242643699333, + "loss": 0.0421, + "step": 6895 + }, + { + "epoch": 3.2451764705882353, + "grad_norm": 0.0625017340438864, + "learning_rate": 0.00017092901994730797, + "loss": 0.0191, + "step": 6896 + }, + { + "epoch": 3.2456470588235296, + "grad_norm": 0.04994112748142611, + "learning_rate": 0.00017088561167018164, + "loss": 0.012, + "step": 6897 + }, + { + "epoch": 3.2461176470588233, + "grad_norm": 0.14352063675110258, + "learning_rate": 0.0001708422016093216, + "loss": 0.0465, + "step": 6898 + }, + { + "epoch": 3.2465882352941176, + "grad_norm": 0.253344365711787, + "learning_rate": 0.00017079878976843533, + "loss": 0.1325, + "step": 6899 + }, + { + "epoch": 3.2470588235294118, + "grad_norm": 0.11778821522205475, + "learning_rate": 0.00017075537615123042, + "loss": 0.0904, + "step": 6900 + }, + { + "epoch": 3.247529411764706, + "grad_norm": 0.06735417886290038, + "learning_rate": 0.0001707119607614146, + "loss": 0.0219, + "step": 6901 + }, + { + "epoch": 3.248, + "grad_norm": 0.07054817720388029, + "learning_rate": 0.0001706685436026957, + "loss": 0.0199, + "step": 6902 + }, + { + "epoch": 3.248470588235294, + "grad_norm": 0.08231960887803519, + "learning_rate": 0.0001706251246787818, + "loss": 0.039, + "step": 6903 + }, + { + "epoch": 3.248941176470588, + "grad_norm": 0.11873134534021951, + "learning_rate": 0.0001705817039933811, + "loss": 0.0336, + "step": 6904 + }, + { + "epoch": 3.2494117647058824, + "grad_norm": 0.18384716521276626, + "learning_rate": 0.00017053828155020193, + "loss": 0.0845, + "step": 6905 + }, + { + "epoch": 3.2498823529411767, + "grad_norm": 0.07993174175031073, + "learning_rate": 0.00017049485735295275, + "loss": 0.0281, + "step": 6906 + }, + { + "epoch": 3.2503529411764704, + "grad_norm": 0.02841396329580308, + "learning_rate": 0.00017045143140534226, + "loss": 0.0049, + "step": 6907 + }, + { + "epoch": 3.2508235294117647, + "grad_norm": 0.04950531644919475, + "learning_rate": 0.0001704080037110792, + "loss": 0.0136, + "step": 6908 + }, + { + "epoch": 3.251294117647059, + "grad_norm": 0.09306742765584591, + "learning_rate": 0.00017036457427387248, + "loss": 0.0354, + "step": 6909 + }, + { + "epoch": 3.251764705882353, + "grad_norm": 0.12667782939661273, + "learning_rate": 0.00017032114309743128, + "loss": 0.0713, + "step": 6910 + }, + { + "epoch": 3.252235294117647, + "grad_norm": 5.748186251586532e-09, + "learning_rate": 0.0001702777101854648, + "loss": 0.0, + "step": 6911 + }, + { + "epoch": 3.252705882352941, + "grad_norm": 0.1785369325739867, + "learning_rate": 0.0001702342755416824, + "loss": 0.0476, + "step": 6912 + }, + { + "epoch": 3.2531764705882353, + "grad_norm": 7.014282194572968e-10, + "learning_rate": 0.00017019083916979357, + "loss": 0.0, + "step": 6913 + }, + { + "epoch": 3.2536470588235296, + "grad_norm": 0.0645663537251673, + "learning_rate": 0.0001701474010735081, + "loss": 0.0198, + "step": 6914 + }, + { + "epoch": 3.2541176470588233, + "grad_norm": 0.1253566466993922, + "learning_rate": 0.00017010396125653572, + "loss": 0.0621, + "step": 6915 + }, + { + "epoch": 3.2545882352941176, + "grad_norm": 0.12338406463501077, + "learning_rate": 0.00017006051972258645, + "loss": 0.0662, + "step": 6916 + }, + { + "epoch": 3.255058823529412, + "grad_norm": 0.08965803410989145, + "learning_rate": 0.0001700170764753704, + "loss": 0.0582, + "step": 6917 + }, + { + "epoch": 3.255529411764706, + "grad_norm": 0.24617834069083835, + "learning_rate": 0.00016997363151859785, + "loss": 0.1505, + "step": 6918 + }, + { + "epoch": 3.2560000000000002, + "grad_norm": 0.0834454418123424, + "learning_rate": 0.00016993018485597915, + "loss": 0.0432, + "step": 6919 + }, + { + "epoch": 3.256470588235294, + "grad_norm": 0.037337323423927635, + "learning_rate": 0.00016988673649122492, + "loss": 0.0063, + "step": 6920 + }, + { + "epoch": 3.2569411764705882, + "grad_norm": 0.07027658545272485, + "learning_rate": 0.00016984328642804587, + "loss": 0.0275, + "step": 6921 + }, + { + "epoch": 3.2574117647058825, + "grad_norm": 0.08921402388010514, + "learning_rate": 0.00016979983467015284, + "loss": 0.0384, + "step": 6922 + }, + { + "epoch": 3.2578823529411762, + "grad_norm": 0.03947961398167123, + "learning_rate": 0.00016975638122125678, + "loss": 0.0096, + "step": 6923 + }, + { + "epoch": 3.2583529411764705, + "grad_norm": 0.08104695133494212, + "learning_rate": 0.00016971292608506888, + "loss": 0.0294, + "step": 6924 + }, + { + "epoch": 3.2588235294117647, + "grad_norm": 0.015238590448665417, + "learning_rate": 0.0001696694692653004, + "loss": 0.0019, + "step": 6925 + }, + { + "epoch": 3.259294117647059, + "grad_norm": 0.17195282116968397, + "learning_rate": 0.00016962601076566275, + "loss": 0.0751, + "step": 6926 + }, + { + "epoch": 3.259764705882353, + "grad_norm": 0.14608503567241896, + "learning_rate": 0.00016958255058986753, + "loss": 0.0775, + "step": 6927 + }, + { + "epoch": 3.260235294117647, + "grad_norm": 0.12083620002348479, + "learning_rate": 0.00016953908874162645, + "loss": 0.0445, + "step": 6928 + }, + { + "epoch": 3.260705882352941, + "grad_norm": 0.08110405264063224, + "learning_rate": 0.00016949562522465135, + "loss": 0.0395, + "step": 6929 + }, + { + "epoch": 3.2611764705882353, + "grad_norm": 0.38518544575866165, + "learning_rate": 0.00016945216004265418, + "loss": 0.0742, + "step": 6930 + }, + { + "epoch": 3.2616470588235296, + "grad_norm": 0.05566433005046702, + "learning_rate": 0.0001694086931993472, + "loss": 0.0169, + "step": 6931 + }, + { + "epoch": 3.2621176470588233, + "grad_norm": 5.678010344582703e-09, + "learning_rate": 0.0001693652246984426, + "loss": 0.0, + "step": 6932 + }, + { + "epoch": 3.2625882352941176, + "grad_norm": 0.08261079859839263, + "learning_rate": 0.0001693217545436528, + "loss": 0.0275, + "step": 6933 + }, + { + "epoch": 3.263058823529412, + "grad_norm": 0.12477273699256423, + "learning_rate": 0.0001692782827386904, + "loss": 0.0631, + "step": 6934 + }, + { + "epoch": 3.263529411764706, + "grad_norm": 1.1262036065548111e-07, + "learning_rate": 0.00016923480928726813, + "loss": 0.0, + "step": 6935 + }, + { + "epoch": 3.2640000000000002, + "grad_norm": 0.08495221442490757, + "learning_rate": 0.00016919133419309875, + "loss": 0.0275, + "step": 6936 + }, + { + "epoch": 3.264470588235294, + "grad_norm": 0.05392914661278435, + "learning_rate": 0.00016914785745989532, + "loss": 0.015, + "step": 6937 + }, + { + "epoch": 3.2649411764705882, + "grad_norm": 0.12097302150869858, + "learning_rate": 0.00016910437909137098, + "loss": 0.0346, + "step": 6938 + }, + { + "epoch": 3.2654117647058825, + "grad_norm": 0.09405763441353138, + "learning_rate": 0.00016906089909123892, + "loss": 0.0235, + "step": 6939 + }, + { + "epoch": 3.2658823529411762, + "grad_norm": 0.17846891489967945, + "learning_rate": 0.00016901741746321253, + "loss": 0.0779, + "step": 6940 + }, + { + "epoch": 3.2663529411764705, + "grad_norm": 1.7019418004006989e-07, + "learning_rate": 0.0001689739342110055, + "loss": 0.0, + "step": 6941 + }, + { + "epoch": 3.2668235294117647, + "grad_norm": 0.031873551864794264, + "learning_rate": 0.00016893044933833135, + "loss": 0.0054, + "step": 6942 + }, + { + "epoch": 3.267294117647059, + "grad_norm": 5.031080542319375e-08, + "learning_rate": 0.00016888696284890402, + "loss": 0.0, + "step": 6943 + }, + { + "epoch": 3.267764705882353, + "grad_norm": 0.23677416024913445, + "learning_rate": 0.0001688434747464374, + "loss": 0.1075, + "step": 6944 + }, + { + "epoch": 3.268235294117647, + "grad_norm": 0.05958500107507972, + "learning_rate": 0.00016879998503464561, + "loss": 0.0217, + "step": 6945 + }, + { + "epoch": 3.268705882352941, + "grad_norm": 0.1726094028372753, + "learning_rate": 0.00016875649371724292, + "loss": 0.0925, + "step": 6946 + }, + { + "epoch": 3.2691764705882354, + "grad_norm": 0.034978859696917934, + "learning_rate": 0.00016871300079794363, + "loss": 0.0036, + "step": 6947 + }, + { + "epoch": 3.2696470588235296, + "grad_norm": 0.12753383013917907, + "learning_rate": 0.0001686695062804623, + "loss": 0.059, + "step": 6948 + }, + { + "epoch": 3.2701176470588234, + "grad_norm": 1.435191996673417e-05, + "learning_rate": 0.00016862601016851358, + "loss": 0.0, + "step": 6949 + }, + { + "epoch": 3.2705882352941176, + "grad_norm": 0.07366515843955418, + "learning_rate": 0.00016858251246581216, + "loss": 0.0238, + "step": 6950 + }, + { + "epoch": 3.271058823529412, + "grad_norm": 0.09052927726173567, + "learning_rate": 0.0001685390131760731, + "loss": 0.0335, + "step": 6951 + }, + { + "epoch": 3.271529411764706, + "grad_norm": 0.12599585889309622, + "learning_rate": 0.00016849551230301135, + "loss": 0.0473, + "step": 6952 + }, + { + "epoch": 3.2720000000000002, + "grad_norm": 0.06270306584731569, + "learning_rate": 0.00016845200985034215, + "loss": 0.0201, + "step": 6953 + }, + { + "epoch": 3.272470588235294, + "grad_norm": 0.024611405148163328, + "learning_rate": 0.0001684085058217808, + "loss": 0.004, + "step": 6954 + }, + { + "epoch": 3.2729411764705882, + "grad_norm": 0.07994339560980908, + "learning_rate": 0.00016836500022104274, + "loss": 0.0347, + "step": 6955 + }, + { + "epoch": 3.2734117647058825, + "grad_norm": 1.3568748406473896e-07, + "learning_rate": 0.00016832149305184363, + "loss": 0.0, + "step": 6956 + }, + { + "epoch": 3.2738823529411762, + "grad_norm": 0.25802085615993453, + "learning_rate": 0.00016827798431789907, + "loss": 0.0998, + "step": 6957 + }, + { + "epoch": 3.2743529411764705, + "grad_norm": 0.05056433049105562, + "learning_rate": 0.00016823447402292504, + "loss": 0.0134, + "step": 6958 + }, + { + "epoch": 3.2748235294117647, + "grad_norm": 0.07502416310979287, + "learning_rate": 0.00016819096217063752, + "loss": 0.0356, + "step": 6959 + }, + { + "epoch": 3.275294117647059, + "grad_norm": 0.061480006820258934, + "learning_rate": 0.00016814744876475252, + "loss": 0.0262, + "step": 6960 + }, + { + "epoch": 3.275764705882353, + "grad_norm": 0.09190581737018828, + "learning_rate": 0.00016810393380898645, + "loss": 0.043, + "step": 6961 + }, + { + "epoch": 3.276235294117647, + "grad_norm": 0.05049139270435558, + "learning_rate": 0.0001680604173070556, + "loss": 0.0127, + "step": 6962 + }, + { + "epoch": 3.276705882352941, + "grad_norm": 0.03453663323261565, + "learning_rate": 0.00016801689926267653, + "loss": 0.0041, + "step": 6963 + }, + { + "epoch": 3.2771764705882354, + "grad_norm": 0.1419363315207877, + "learning_rate": 0.00016797337967956592, + "loss": 0.069, + "step": 6964 + }, + { + "epoch": 3.2776470588235296, + "grad_norm": 0.07570418217044586, + "learning_rate": 0.00016792985856144049, + "loss": 0.0313, + "step": 6965 + }, + { + "epoch": 3.2781176470588234, + "grad_norm": 0.03042507398626807, + "learning_rate": 0.00016788633591201723, + "loss": 0.0063, + "step": 6966 + }, + { + "epoch": 3.2785882352941176, + "grad_norm": 0.07462186928184732, + "learning_rate": 0.0001678428117350131, + "loss": 0.0296, + "step": 6967 + }, + { + "epoch": 3.279058823529412, + "grad_norm": 0.17397446602247613, + "learning_rate": 0.0001677992860341454, + "loss": 0.0798, + "step": 6968 + }, + { + "epoch": 3.279529411764706, + "grad_norm": 0.18145064952476025, + "learning_rate": 0.00016775575881313132, + "loss": 0.0997, + "step": 6969 + }, + { + "epoch": 3.2800000000000002, + "grad_norm": 0.1428614845749972, + "learning_rate": 0.00016771223007568836, + "loss": 0.0711, + "step": 6970 + }, + { + "epoch": 3.280470588235294, + "grad_norm": 0.08299989134180483, + "learning_rate": 0.00016766869982553405, + "loss": 0.0297, + "step": 6971 + }, + { + "epoch": 3.2809411764705882, + "grad_norm": 0.07524015227450166, + "learning_rate": 0.0001676251680663861, + "loss": 0.0252, + "step": 6972 + }, + { + "epoch": 3.2814117647058825, + "grad_norm": 0.03174047178564235, + "learning_rate": 0.0001675816348019624, + "loss": 0.0034, + "step": 6973 + }, + { + "epoch": 3.2818823529411763, + "grad_norm": 0.03878624659834554, + "learning_rate": 0.0001675381000359808, + "loss": 0.0061, + "step": 6974 + }, + { + "epoch": 3.2823529411764705, + "grad_norm": 0.08819914127327952, + "learning_rate": 0.0001674945637721595, + "loss": 0.027, + "step": 6975 + }, + { + "epoch": 3.2828235294117647, + "grad_norm": 7.567928081322002e-07, + "learning_rate": 0.00016745102601421663, + "loss": 0.0, + "step": 6976 + }, + { + "epoch": 3.283294117647059, + "grad_norm": 0.06216249444080256, + "learning_rate": 0.00016740748676587045, + "loss": 0.0166, + "step": 6977 + }, + { + "epoch": 3.283764705882353, + "grad_norm": 0.18062256896676082, + "learning_rate": 0.0001673639460308396, + "loss": 0.073, + "step": 6978 + }, + { + "epoch": 3.284235294117647, + "grad_norm": 0.09586238933439171, + "learning_rate": 0.00016732040381284258, + "loss": 0.0439, + "step": 6979 + }, + { + "epoch": 3.284705882352941, + "grad_norm": 0.1856910799841414, + "learning_rate": 0.00016727686011559813, + "loss": 0.102, + "step": 6980 + }, + { + "epoch": 3.2851764705882354, + "grad_norm": 0.06635963169838571, + "learning_rate": 0.00016723331494282502, + "loss": 0.0124, + "step": 6981 + }, + { + "epoch": 3.2856470588235296, + "grad_norm": 0.159570796241895, + "learning_rate": 0.00016718976829824234, + "loss": 0.087, + "step": 6982 + }, + { + "epoch": 3.2861176470588234, + "grad_norm": 4.7582304384046115e-08, + "learning_rate": 0.00016714622018556915, + "loss": 0.0, + "step": 6983 + }, + { + "epoch": 3.2865882352941176, + "grad_norm": 0.0665894324999142, + "learning_rate": 0.0001671026706085246, + "loss": 0.0274, + "step": 6984 + }, + { + "epoch": 3.287058823529412, + "grad_norm": 1.3822685150068478e-05, + "learning_rate": 0.0001670591195708281, + "loss": 0.0, + "step": 6985 + }, + { + "epoch": 3.287529411764706, + "grad_norm": 0.14595173030776493, + "learning_rate": 0.00016701556707619914, + "loss": 0.0479, + "step": 6986 + }, + { + "epoch": 3.288, + "grad_norm": 9.220798558453932e-07, + "learning_rate": 0.00016697201312835727, + "loss": 0.0, + "step": 6987 + }, + { + "epoch": 3.288470588235294, + "grad_norm": 1.2260642034115752e-07, + "learning_rate": 0.00016692845773102222, + "loss": 0.0, + "step": 6988 + }, + { + "epoch": 3.2889411764705883, + "grad_norm": 0.1427843844446008, + "learning_rate": 0.0001668849008879139, + "loss": 0.0892, + "step": 6989 + }, + { + "epoch": 3.2894117647058825, + "grad_norm": 0.03603892739377161, + "learning_rate": 0.00016684134260275217, + "loss": 0.0054, + "step": 6990 + }, + { + "epoch": 3.2898823529411763, + "grad_norm": 0.20748518479792835, + "learning_rate": 0.0001667977828792572, + "loss": 0.102, + "step": 6991 + }, + { + "epoch": 3.2903529411764705, + "grad_norm": 0.171849542603816, + "learning_rate": 0.00016675422172114917, + "loss": 0.1038, + "step": 6992 + }, + { + "epoch": 3.2908235294117647, + "grad_norm": 0.1299421040997471, + "learning_rate": 0.00016671065913214845, + "loss": 0.0513, + "step": 6993 + }, + { + "epoch": 3.291294117647059, + "grad_norm": 0.1046884019375583, + "learning_rate": 0.00016666709511597546, + "loss": 0.0493, + "step": 6994 + }, + { + "epoch": 3.291764705882353, + "grad_norm": 0.06681066707308979, + "learning_rate": 0.0001666235296763508, + "loss": 0.0247, + "step": 6995 + }, + { + "epoch": 3.292235294117647, + "grad_norm": 0.14619252017435852, + "learning_rate": 0.00016657996281699514, + "loss": 0.078, + "step": 6996 + }, + { + "epoch": 3.292705882352941, + "grad_norm": 0.13909283735886152, + "learning_rate": 0.00016653639454162938, + "loss": 0.0334, + "step": 6997 + }, + { + "epoch": 3.2931764705882354, + "grad_norm": 0.08822805228020716, + "learning_rate": 0.00016649282485397437, + "loss": 0.0332, + "step": 6998 + }, + { + "epoch": 3.2936470588235296, + "grad_norm": 0.025992819273268852, + "learning_rate": 0.00016644925375775124, + "loss": 0.0032, + "step": 6999 + }, + { + "epoch": 3.2941176470588234, + "grad_norm": 0.08756950207813911, + "learning_rate": 0.00016640568125668117, + "loss": 0.0432, + "step": 7000 + }, + { + "epoch": 3.2945882352941176, + "grad_norm": 0.06912947442649886, + "learning_rate": 0.00016636210735448539, + "loss": 0.0231, + "step": 7001 + }, + { + "epoch": 3.295058823529412, + "grad_norm": 0.10241032017820126, + "learning_rate": 0.00016631853205488546, + "loss": 0.0402, + "step": 7002 + }, + { + "epoch": 3.295529411764706, + "grad_norm": 0.14912221281116633, + "learning_rate": 0.00016627495536160284, + "loss": 0.0272, + "step": 7003 + }, + { + "epoch": 3.296, + "grad_norm": 0.08580946180399311, + "learning_rate": 0.0001662313772783592, + "loss": 0.0487, + "step": 7004 + }, + { + "epoch": 3.296470588235294, + "grad_norm": 0.0640675528694911, + "learning_rate": 0.00016618779780887627, + "loss": 0.0092, + "step": 7005 + }, + { + "epoch": 3.2969411764705883, + "grad_norm": 2.4535044254845657e-06, + "learning_rate": 0.00016614421695687606, + "loss": 0.0, + "step": 7006 + }, + { + "epoch": 3.2974117647058825, + "grad_norm": 0.026663741555928815, + "learning_rate": 0.00016610063472608053, + "loss": 0.0033, + "step": 7007 + }, + { + "epoch": 3.2978823529411763, + "grad_norm": 2.8595114514133267e-06, + "learning_rate": 0.0001660570511202118, + "loss": 0.0, + "step": 7008 + }, + { + "epoch": 3.2983529411764705, + "grad_norm": 0.05921437728918079, + "learning_rate": 0.00016601346614299214, + "loss": 0.016, + "step": 7009 + }, + { + "epoch": 3.2988235294117647, + "grad_norm": 0.06980449772365731, + "learning_rate": 0.00016596987979814395, + "loss": 0.02, + "step": 7010 + }, + { + "epoch": 3.299294117647059, + "grad_norm": 2.2615981237397578e-05, + "learning_rate": 0.0001659262920893897, + "loss": 0.0, + "step": 7011 + }, + { + "epoch": 3.299764705882353, + "grad_norm": 0.2305807611012709, + "learning_rate": 0.00016588270302045197, + "loss": 0.0945, + "step": 7012 + }, + { + "epoch": 3.300235294117647, + "grad_norm": 0.04615311344372355, + "learning_rate": 0.00016583911259505352, + "loss": 0.0118, + "step": 7013 + }, + { + "epoch": 3.300705882352941, + "grad_norm": 0.10724384662662763, + "learning_rate": 0.00016579552081691715, + "loss": 0.0271, + "step": 7014 + }, + { + "epoch": 3.3011764705882354, + "grad_norm": 0.035011860150589644, + "learning_rate": 0.0001657519276897658, + "loss": 0.008, + "step": 7015 + }, + { + "epoch": 3.3016470588235296, + "grad_norm": 0.17010576011027984, + "learning_rate": 0.0001657083332173226, + "loss": 0.089, + "step": 7016 + }, + { + "epoch": 3.3021176470588234, + "grad_norm": 0.11030743271526644, + "learning_rate": 0.0001656647374033107, + "loss": 0.0503, + "step": 7017 + }, + { + "epoch": 3.3025882352941176, + "grad_norm": 0.08319691958554856, + "learning_rate": 0.00016562114025145332, + "loss": 0.0301, + "step": 7018 + }, + { + "epoch": 3.303058823529412, + "grad_norm": 0.07618271393318896, + "learning_rate": 0.00016557754176547404, + "loss": 0.0307, + "step": 7019 + }, + { + "epoch": 3.303529411764706, + "grad_norm": 0.11307615905290164, + "learning_rate": 0.0001655339419490962, + "loss": 0.0227, + "step": 7020 + }, + { + "epoch": 3.304, + "grad_norm": 0.1167205088164627, + "learning_rate": 0.00016549034080604359, + "loss": 0.049, + "step": 7021 + }, + { + "epoch": 3.304470588235294, + "grad_norm": 0.09661068572000596, + "learning_rate": 0.00016544673834003985, + "loss": 0.0408, + "step": 7022 + }, + { + "epoch": 3.3049411764705883, + "grad_norm": 9.747447229954745e-08, + "learning_rate": 0.00016540313455480895, + "loss": 0.0, + "step": 7023 + }, + { + "epoch": 3.3054117647058825, + "grad_norm": 5.396873712233545e-07, + "learning_rate": 0.00016535952945407473, + "loss": 0.0, + "step": 7024 + }, + { + "epoch": 3.3058823529411763, + "grad_norm": 0.05711106958543939, + "learning_rate": 0.00016531592304156137, + "loss": 0.0158, + "step": 7025 + }, + { + "epoch": 3.3063529411764705, + "grad_norm": 0.07300489712856269, + "learning_rate": 0.0001652723153209931, + "loss": 0.0189, + "step": 7026 + }, + { + "epoch": 3.3068235294117647, + "grad_norm": 0.0749138814648124, + "learning_rate": 0.0001652287062960942, + "loss": 0.0272, + "step": 7027 + }, + { + "epoch": 3.307294117647059, + "grad_norm": 0.06927438255711738, + "learning_rate": 0.00016518509597058902, + "loss": 0.0292, + "step": 7028 + }, + { + "epoch": 3.307764705882353, + "grad_norm": 0.1173629167908859, + "learning_rate": 0.0001651414843482022, + "loss": 0.0509, + "step": 7029 + }, + { + "epoch": 3.308235294117647, + "grad_norm": 0.0954684028279206, + "learning_rate": 0.00016509787143265836, + "loss": 0.0331, + "step": 7030 + }, + { + "epoch": 3.308705882352941, + "grad_norm": 8.561403420059991e-08, + "learning_rate": 0.0001650542572276822, + "loss": 0.0, + "step": 7031 + }, + { + "epoch": 3.3091764705882354, + "grad_norm": 0.0331851186925965, + "learning_rate": 0.00016501064173699866, + "loss": 0.0034, + "step": 7032 + }, + { + "epoch": 3.309647058823529, + "grad_norm": 0.05113497680611098, + "learning_rate": 0.0001649670249643327, + "loss": 0.0184, + "step": 7033 + }, + { + "epoch": 3.3101176470588234, + "grad_norm": 0.05359684251305171, + "learning_rate": 0.0001649234069134094, + "loss": 0.0153, + "step": 7034 + }, + { + "epoch": 3.3105882352941176, + "grad_norm": 0.05978750775966057, + "learning_rate": 0.00016487978758795397, + "loss": 0.0161, + "step": 7035 + }, + { + "epoch": 3.311058823529412, + "grad_norm": 0.0848094375824953, + "learning_rate": 0.00016483616699169163, + "loss": 0.0375, + "step": 7036 + }, + { + "epoch": 3.311529411764706, + "grad_norm": 1.3963786948282241e-09, + "learning_rate": 0.0001647925451283479, + "loss": 0.0, + "step": 7037 + }, + { + "epoch": 3.312, + "grad_norm": 0.11588698524153533, + "learning_rate": 0.00016474892200164825, + "loss": 0.0503, + "step": 7038 + }, + { + "epoch": 3.312470588235294, + "grad_norm": 0.1317301806450278, + "learning_rate": 0.0001647052976153183, + "loss": 0.0459, + "step": 7039 + }, + { + "epoch": 3.3129411764705883, + "grad_norm": 0.2168309424085221, + "learning_rate": 0.00016466167197308384, + "loss": 0.0792, + "step": 7040 + }, + { + "epoch": 3.3134117647058825, + "grad_norm": 0.07528603023067852, + "learning_rate": 0.00016461804507867067, + "loss": 0.0261, + "step": 7041 + }, + { + "epoch": 3.3138823529411763, + "grad_norm": 7.919732379131833e-08, + "learning_rate": 0.00016457441693580472, + "loss": 0.0, + "step": 7042 + }, + { + "epoch": 3.3143529411764705, + "grad_norm": 0.04547520094128282, + "learning_rate": 0.0001645307875482121, + "loss": 0.0059, + "step": 7043 + }, + { + "epoch": 3.3148235294117647, + "grad_norm": 0.06048538238114035, + "learning_rate": 0.00016448715691961898, + "loss": 0.02, + "step": 7044 + }, + { + "epoch": 3.315294117647059, + "grad_norm": 0.03979941759890507, + "learning_rate": 0.00016444352505375155, + "loss": 0.0109, + "step": 7045 + }, + { + "epoch": 3.315764705882353, + "grad_norm": 0.08471068796657685, + "learning_rate": 0.00016439989195433626, + "loss": 0.0359, + "step": 7046 + }, + { + "epoch": 3.316235294117647, + "grad_norm": 0.033247774187087895, + "learning_rate": 0.00016435625762509955, + "loss": 0.0064, + "step": 7047 + }, + { + "epoch": 3.316705882352941, + "grad_norm": 0.08402282861739403, + "learning_rate": 0.00016431262206976805, + "loss": 0.0255, + "step": 7048 + }, + { + "epoch": 3.3171764705882354, + "grad_norm": 9.890168173659584e-07, + "learning_rate": 0.00016426898529206836, + "loss": 0.0, + "step": 7049 + }, + { + "epoch": 3.317647058823529, + "grad_norm": 0.11484752616879264, + "learning_rate": 0.00016422534729572738, + "loss": 0.0827, + "step": 7050 + }, + { + "epoch": 3.3181176470588234, + "grad_norm": 0.10421791821971489, + "learning_rate": 0.00016418170808447198, + "loss": 0.0496, + "step": 7051 + }, + { + "epoch": 3.3185882352941176, + "grad_norm": 0.09390841041205966, + "learning_rate": 0.00016413806766202912, + "loss": 0.023, + "step": 7052 + }, + { + "epoch": 3.319058823529412, + "grad_norm": 0.061122076374981925, + "learning_rate": 0.00016409442603212594, + "loss": 0.0295, + "step": 7053 + }, + { + "epoch": 3.319529411764706, + "grad_norm": 0.07475626941201276, + "learning_rate": 0.00016405078319848965, + "loss": 0.0221, + "step": 7054 + }, + { + "epoch": 3.32, + "grad_norm": 0.13381294067128896, + "learning_rate": 0.0001640071391648476, + "loss": 0.0475, + "step": 7055 + }, + { + "epoch": 3.320470588235294, + "grad_norm": 0.04004652504761698, + "learning_rate": 0.0001639634939349271, + "loss": 0.0076, + "step": 7056 + }, + { + "epoch": 3.3209411764705883, + "grad_norm": 0.09460379584822702, + "learning_rate": 0.00016391984751245572, + "loss": 0.0587, + "step": 7057 + }, + { + "epoch": 3.3214117647058825, + "grad_norm": 0.05431824945699419, + "learning_rate": 0.00016387619990116116, + "loss": 0.0151, + "step": 7058 + }, + { + "epoch": 3.3218823529411763, + "grad_norm": 1.797998781518985e-05, + "learning_rate": 0.00016383255110477102, + "loss": 0.0, + "step": 7059 + }, + { + "epoch": 3.3223529411764705, + "grad_norm": 0.060681783056909216, + "learning_rate": 0.00016378890112701312, + "loss": 0.0161, + "step": 7060 + }, + { + "epoch": 3.3228235294117647, + "grad_norm": 1.7069815011319212e-06, + "learning_rate": 0.0001637452499716155, + "loss": 0.0, + "step": 7061 + }, + { + "epoch": 3.323294117647059, + "grad_norm": 0.05911407346282067, + "learning_rate": 0.0001637015976423061, + "loss": 0.0166, + "step": 7062 + }, + { + "epoch": 3.323764705882353, + "grad_norm": 0.11545178817248919, + "learning_rate": 0.00016365794414281303, + "loss": 0.0456, + "step": 7063 + }, + { + "epoch": 3.324235294117647, + "grad_norm": 4.289915092733944e-06, + "learning_rate": 0.00016361428947686457, + "loss": 0.0, + "step": 7064 + }, + { + "epoch": 3.324705882352941, + "grad_norm": 0.07509375884593937, + "learning_rate": 0.00016357063364818902, + "loss": 0.0238, + "step": 7065 + }, + { + "epoch": 3.3251764705882354, + "grad_norm": 2.2661454678635514e-06, + "learning_rate": 0.00016352697666051474, + "loss": 0.0, + "step": 7066 + }, + { + "epoch": 3.325647058823529, + "grad_norm": 0.07629830647721457, + "learning_rate": 0.0001634833185175703, + "loss": 0.0246, + "step": 7067 + }, + { + "epoch": 3.3261176470588234, + "grad_norm": 0.16787178537493544, + "learning_rate": 0.00016343965922308438, + "loss": 0.0975, + "step": 7068 + }, + { + "epoch": 3.3265882352941176, + "grad_norm": 0.08866341227401094, + "learning_rate": 0.00016339599878078563, + "loss": 0.0335, + "step": 7069 + }, + { + "epoch": 3.327058823529412, + "grad_norm": 0.05357224942204034, + "learning_rate": 0.00016335233719440282, + "loss": 0.0175, + "step": 7070 + }, + { + "epoch": 3.327529411764706, + "grad_norm": 0.07312368869787048, + "learning_rate": 0.000163308674467665, + "loss": 0.0269, + "step": 7071 + }, + { + "epoch": 3.328, + "grad_norm": 0.17150887264101825, + "learning_rate": 0.00016326501060430105, + "loss": 0.0776, + "step": 7072 + }, + { + "epoch": 3.328470588235294, + "grad_norm": 2.247748125767313e-06, + "learning_rate": 0.00016322134560804017, + "loss": 0.0, + "step": 7073 + }, + { + "epoch": 3.3289411764705883, + "grad_norm": 1.4540902795411719e-06, + "learning_rate": 0.00016317767948261148, + "loss": 0.0, + "step": 7074 + }, + { + "epoch": 3.3294117647058825, + "grad_norm": 0.04480851505099973, + "learning_rate": 0.00016313401223174434, + "loss": 0.0132, + "step": 7075 + }, + { + "epoch": 3.3298823529411763, + "grad_norm": 0.16674713770438343, + "learning_rate": 0.00016309034385916815, + "loss": 0.0785, + "step": 7076 + }, + { + "epoch": 3.3303529411764705, + "grad_norm": 0.09100360519592486, + "learning_rate": 0.00016304667436861233, + "loss": 0.0342, + "step": 7077 + }, + { + "epoch": 3.3308235294117647, + "grad_norm": 5.4767033501739997e-08, + "learning_rate": 0.00016300300376380657, + "loss": 0.0, + "step": 7078 + }, + { + "epoch": 3.331294117647059, + "grad_norm": 0.06601789648444562, + "learning_rate": 0.00016295933204848052, + "loss": 0.0239, + "step": 7079 + }, + { + "epoch": 3.331764705882353, + "grad_norm": 0.09983142264020888, + "learning_rate": 0.00016291565922636392, + "loss": 0.0453, + "step": 7080 + }, + { + "epoch": 3.332235294117647, + "grad_norm": 0.07586107386661226, + "learning_rate": 0.0001628719853011867, + "loss": 0.0392, + "step": 7081 + }, + { + "epoch": 3.332705882352941, + "grad_norm": 0.11674398344776324, + "learning_rate": 0.00016282831027667881, + "loss": 0.039, + "step": 7082 + }, + { + "epoch": 3.3331764705882354, + "grad_norm": 0.06599510990243249, + "learning_rate": 0.00016278463415657027, + "loss": 0.0263, + "step": 7083 + }, + { + "epoch": 3.333647058823529, + "grad_norm": 4.438745933172704e-07, + "learning_rate": 0.00016274095694459128, + "loss": 0.0, + "step": 7084 + }, + { + "epoch": 3.3341176470588234, + "grad_norm": 0.07884638065474232, + "learning_rate": 0.00016269727864447206, + "loss": 0.0119, + "step": 7085 + }, + { + "epoch": 3.3345882352941176, + "grad_norm": 0.07216608086468887, + "learning_rate": 0.00016265359925994298, + "loss": 0.022, + "step": 7086 + }, + { + "epoch": 3.335058823529412, + "grad_norm": 0.04390730813758553, + "learning_rate": 0.00016260991879473441, + "loss": 0.0055, + "step": 7087 + }, + { + "epoch": 3.335529411764706, + "grad_norm": 0.0923097793677657, + "learning_rate": 0.000162566237252577, + "loss": 0.0404, + "step": 7088 + }, + { + "epoch": 3.336, + "grad_norm": 0.1035355472628327, + "learning_rate": 0.00016252255463720125, + "loss": 0.029, + "step": 7089 + }, + { + "epoch": 3.336470588235294, + "grad_norm": 0.08967221813810074, + "learning_rate": 0.0001624788709523379, + "loss": 0.0456, + "step": 7090 + }, + { + "epoch": 3.3369411764705883, + "grad_norm": 0.054003550751450066, + "learning_rate": 0.00016243518620171778, + "loss": 0.0144, + "step": 7091 + }, + { + "epoch": 3.3374117647058825, + "grad_norm": 8.220527190070775e-05, + "learning_rate": 0.00016239150038907177, + "loss": 0.0, + "step": 7092 + }, + { + "epoch": 3.3378823529411763, + "grad_norm": 0.07695063250759236, + "learning_rate": 0.00016234781351813079, + "loss": 0.0311, + "step": 7093 + }, + { + "epoch": 3.3383529411764705, + "grad_norm": 0.13986933616460512, + "learning_rate": 0.00016230412559262606, + "loss": 0.0785, + "step": 7094 + }, + { + "epoch": 3.3388235294117647, + "grad_norm": 0.0007748742715592499, + "learning_rate": 0.00016226043661628862, + "loss": 0.0, + "step": 7095 + }, + { + "epoch": 3.339294117647059, + "grad_norm": 0.10814965453206306, + "learning_rate": 0.00016221674659284976, + "loss": 0.0549, + "step": 7096 + }, + { + "epoch": 3.339764705882353, + "grad_norm": 0.11967754829739384, + "learning_rate": 0.00016217305552604076, + "loss": 0.0401, + "step": 7097 + }, + { + "epoch": 3.340235294117647, + "grad_norm": 0.00014708283860105984, + "learning_rate": 0.00016212936341959312, + "loss": 0.0, + "step": 7098 + }, + { + "epoch": 3.340705882352941, + "grad_norm": 8.876426175149178e-07, + "learning_rate": 0.0001620856702772384, + "loss": 0.0, + "step": 7099 + }, + { + "epoch": 3.3411764705882354, + "grad_norm": 0.09845769655274421, + "learning_rate": 0.00016204197610270816, + "loss": 0.0395, + "step": 7100 + }, + { + "epoch": 3.341647058823529, + "grad_norm": 0.2279564462363324, + "learning_rate": 0.00016199828089973406, + "loss": 0.1128, + "step": 7101 + }, + { + "epoch": 3.3421176470588234, + "grad_norm": 0.08070232623908809, + "learning_rate": 0.00016195458467204793, + "loss": 0.0373, + "step": 7102 + }, + { + "epoch": 3.3425882352941176, + "grad_norm": 0.10799500456510473, + "learning_rate": 0.00016191088742338166, + "loss": 0.044, + "step": 7103 + }, + { + "epoch": 3.343058823529412, + "grad_norm": 9.787428452058039e-06, + "learning_rate": 0.00016186718915746714, + "loss": 0.0, + "step": 7104 + }, + { + "epoch": 3.343529411764706, + "grad_norm": 0.11337340053799357, + "learning_rate": 0.00016182348987803648, + "loss": 0.0726, + "step": 7105 + }, + { + "epoch": 3.344, + "grad_norm": 0.15204575270262116, + "learning_rate": 0.0001617797895888218, + "loss": 0.0934, + "step": 7106 + }, + { + "epoch": 3.344470588235294, + "grad_norm": 0.07379485910514205, + "learning_rate": 0.00016173608829355527, + "loss": 0.0296, + "step": 7107 + }, + { + "epoch": 3.3449411764705883, + "grad_norm": 0.181601367382434, + "learning_rate": 0.00016169238599596927, + "loss": 0.0902, + "step": 7108 + }, + { + "epoch": 3.3454117647058825, + "grad_norm": 0.08471006449944377, + "learning_rate": 0.00016164868269979619, + "loss": 0.0296, + "step": 7109 + }, + { + "epoch": 3.3458823529411763, + "grad_norm": 1.0148041934595517e-06, + "learning_rate": 0.0001616049784087684, + "loss": 0.0, + "step": 7110 + }, + { + "epoch": 3.3463529411764705, + "grad_norm": 8.09831205297249e-06, + "learning_rate": 0.00016156127312661855, + "loss": 0.0, + "step": 7111 + }, + { + "epoch": 3.3468235294117648, + "grad_norm": 1.0492369794781799e-07, + "learning_rate": 0.0001615175668570793, + "loss": 0.0, + "step": 7112 + }, + { + "epoch": 3.347294117647059, + "grad_norm": 0.09274256560615902, + "learning_rate": 0.0001614738596038833, + "loss": 0.0353, + "step": 7113 + }, + { + "epoch": 3.3477647058823528, + "grad_norm": 0.07152425770525345, + "learning_rate": 0.00016143015137076345, + "loss": 0.0288, + "step": 7114 + }, + { + "epoch": 3.348235294117647, + "grad_norm": 0.07952096819923284, + "learning_rate": 0.0001613864421614526, + "loss": 0.0432, + "step": 7115 + }, + { + "epoch": 3.348705882352941, + "grad_norm": 0.04081818910970303, + "learning_rate": 0.00016134273197968374, + "loss": 0.0069, + "step": 7116 + }, + { + "epoch": 3.3491764705882354, + "grad_norm": 0.07639280079857405, + "learning_rate": 0.0001612990208291899, + "loss": 0.0248, + "step": 7117 + }, + { + "epoch": 3.349647058823529, + "grad_norm": 8.573259814559943e-07, + "learning_rate": 0.0001612553087137043, + "loss": 0.0, + "step": 7118 + }, + { + "epoch": 3.3501176470588234, + "grad_norm": 5.740775795477062e-07, + "learning_rate": 0.0001612115956369601, + "loss": 0.0, + "step": 7119 + }, + { + "epoch": 3.3505882352941176, + "grad_norm": 0.33422798957002814, + "learning_rate": 0.00016116788160269065, + "loss": 0.1179, + "step": 7120 + }, + { + "epoch": 3.351058823529412, + "grad_norm": 0.08828517544507765, + "learning_rate": 0.0001611241666146293, + "loss": 0.0372, + "step": 7121 + }, + { + "epoch": 3.351529411764706, + "grad_norm": 0.04441282601862234, + "learning_rate": 0.00016108045067650962, + "loss": 0.0118, + "step": 7122 + }, + { + "epoch": 3.352, + "grad_norm": 0.12875918604713282, + "learning_rate": 0.0001610367337920651, + "loss": 0.0649, + "step": 7123 + }, + { + "epoch": 3.352470588235294, + "grad_norm": 0.06865504861537512, + "learning_rate": 0.00016099301596502936, + "loss": 0.0301, + "step": 7124 + }, + { + "epoch": 3.3529411764705883, + "grad_norm": 1.0645139832769706e-06, + "learning_rate": 0.00016094929719913612, + "loss": 0.0, + "step": 7125 + }, + { + "epoch": 3.3534117647058825, + "grad_norm": 0.2419902590762652, + "learning_rate": 0.0001609055774981192, + "loss": 0.1209, + "step": 7126 + }, + { + "epoch": 3.3538823529411763, + "grad_norm": 0.06800347186397585, + "learning_rate": 0.0001608618568657125, + "loss": 0.0246, + "step": 7127 + }, + { + "epoch": 3.3543529411764705, + "grad_norm": 0.08262130985586372, + "learning_rate": 0.0001608181353056499, + "loss": 0.0246, + "step": 7128 + }, + { + "epoch": 3.3548235294117648, + "grad_norm": 0.10535440599958701, + "learning_rate": 0.0001607744128216655, + "loss": 0.0583, + "step": 7129 + }, + { + "epoch": 3.355294117647059, + "grad_norm": 0.051080859858084346, + "learning_rate": 0.00016073068941749343, + "loss": 0.0113, + "step": 7130 + }, + { + "epoch": 3.3557647058823528, + "grad_norm": 0.1493351802488477, + "learning_rate": 0.0001606869650968678, + "loss": 0.0626, + "step": 7131 + }, + { + "epoch": 3.356235294117647, + "grad_norm": 0.11024957358499438, + "learning_rate": 0.00016064323986352295, + "loss": 0.0485, + "step": 7132 + }, + { + "epoch": 3.356705882352941, + "grad_norm": 6.887995654897122e-07, + "learning_rate": 0.00016059951372119323, + "loss": 0.0, + "step": 7133 + }, + { + "epoch": 3.3571764705882354, + "grad_norm": 0.05044757505654474, + "learning_rate": 0.000160555786673613, + "loss": 0.0133, + "step": 7134 + }, + { + "epoch": 3.357647058823529, + "grad_norm": 0.11584891083860373, + "learning_rate": 0.00016051205872451686, + "loss": 0.0209, + "step": 7135 + }, + { + "epoch": 3.3581176470588234, + "grad_norm": 0.10294851474899235, + "learning_rate": 0.00016046832987763935, + "loss": 0.046, + "step": 7136 + }, + { + "epoch": 3.3585882352941177, + "grad_norm": 5.294057999494779e-07, + "learning_rate": 0.0001604246001367151, + "loss": 0.0, + "step": 7137 + }, + { + "epoch": 3.359058823529412, + "grad_norm": 0.1706354451079628, + "learning_rate": 0.00016038086950547885, + "loss": 0.0975, + "step": 7138 + }, + { + "epoch": 3.359529411764706, + "grad_norm": 0.05265171240737138, + "learning_rate": 0.00016033713798766546, + "loss": 0.0219, + "step": 7139 + }, + { + "epoch": 3.36, + "grad_norm": 0.039719418273746504, + "learning_rate": 0.00016029340558700977, + "loss": 0.0085, + "step": 7140 + }, + { + "epoch": 3.360470588235294, + "grad_norm": 0.12086744193538737, + "learning_rate": 0.00016024967230724672, + "loss": 0.06, + "step": 7141 + }, + { + "epoch": 3.3609411764705883, + "grad_norm": 0.10718210121065556, + "learning_rate": 0.00016020593815211138, + "loss": 0.0542, + "step": 7142 + }, + { + "epoch": 3.3614117647058825, + "grad_norm": 0.05967469353747254, + "learning_rate": 0.00016016220312533888, + "loss": 0.0245, + "step": 7143 + }, + { + "epoch": 3.3618823529411763, + "grad_norm": 0.12029653975561452, + "learning_rate": 0.00016011846723066445, + "loss": 0.0377, + "step": 7144 + }, + { + "epoch": 3.3623529411764705, + "grad_norm": 0.11487674891575023, + "learning_rate": 0.0001600747304718232, + "loss": 0.0296, + "step": 7145 + }, + { + "epoch": 3.3628235294117648, + "grad_norm": 0.06644545336895424, + "learning_rate": 0.00016003099285255058, + "loss": 0.0251, + "step": 7146 + }, + { + "epoch": 3.363294117647059, + "grad_norm": 0.0699421031213923, + "learning_rate": 0.000159987254376582, + "loss": 0.0276, + "step": 7147 + }, + { + "epoch": 3.3637647058823528, + "grad_norm": 0.10624040681519527, + "learning_rate": 0.00015994351504765287, + "loss": 0.0286, + "step": 7148 + }, + { + "epoch": 3.364235294117647, + "grad_norm": 0.14534815548500402, + "learning_rate": 0.00015989977486949883, + "loss": 0.0661, + "step": 7149 + }, + { + "epoch": 3.364705882352941, + "grad_norm": 0.16044314898579096, + "learning_rate": 0.00015985603384585542, + "loss": 0.115, + "step": 7150 + }, + { + "epoch": 3.3651764705882354, + "grad_norm": 1.7892821603349416e-06, + "learning_rate": 0.0001598122919804584, + "loss": 0.0, + "step": 7151 + }, + { + "epoch": 3.365647058823529, + "grad_norm": 8.628159615912296e-07, + "learning_rate": 0.0001597685492770435, + "loss": 0.0, + "step": 7152 + }, + { + "epoch": 3.3661176470588234, + "grad_norm": 0.07656512745559692, + "learning_rate": 0.00015972480573934656, + "loss": 0.022, + "step": 7153 + }, + { + "epoch": 3.3665882352941177, + "grad_norm": 0.322716147323598, + "learning_rate": 0.00015968106137110352, + "loss": 0.0741, + "step": 7154 + }, + { + "epoch": 3.367058823529412, + "grad_norm": 0.05061619288390367, + "learning_rate": 0.00015963731617605034, + "loss": 0.0138, + "step": 7155 + }, + { + "epoch": 3.367529411764706, + "grad_norm": 0.047696350717895644, + "learning_rate": 0.00015959357015792313, + "loss": 0.0085, + "step": 7156 + }, + { + "epoch": 3.368, + "grad_norm": 0.0762172812206946, + "learning_rate": 0.00015954982332045794, + "loss": 0.0286, + "step": 7157 + }, + { + "epoch": 3.368470588235294, + "grad_norm": 0.05966841728486856, + "learning_rate": 0.000159506075667391, + "loss": 0.0148, + "step": 7158 + }, + { + "epoch": 3.3689411764705883, + "grad_norm": 0.03707193017429515, + "learning_rate": 0.0001594623272024586, + "loss": 0.0056, + "step": 7159 + }, + { + "epoch": 3.369411764705882, + "grad_norm": 0.15801627257974327, + "learning_rate": 0.000159418577929397, + "loss": 0.0716, + "step": 7160 + }, + { + "epoch": 3.3698823529411763, + "grad_norm": 0.0576186756550595, + "learning_rate": 0.0001593748278519427, + "loss": 0.0203, + "step": 7161 + }, + { + "epoch": 3.3703529411764706, + "grad_norm": 0.1158247701462827, + "learning_rate": 0.00015933107697383206, + "loss": 0.0356, + "step": 7162 + }, + { + "epoch": 3.3708235294117648, + "grad_norm": 0.08234794879655666, + "learning_rate": 0.0001592873252988017, + "loss": 0.0251, + "step": 7163 + }, + { + "epoch": 3.371294117647059, + "grad_norm": 0.07936048939528077, + "learning_rate": 0.0001592435728305882, + "loss": 0.0256, + "step": 7164 + }, + { + "epoch": 3.3717647058823528, + "grad_norm": 0.09740639724585527, + "learning_rate": 0.00015919981957292825, + "loss": 0.0505, + "step": 7165 + }, + { + "epoch": 3.372235294117647, + "grad_norm": 0.15714588168768814, + "learning_rate": 0.00015915606552955856, + "loss": 0.0774, + "step": 7166 + }, + { + "epoch": 3.372705882352941, + "grad_norm": 0.042643682284309446, + "learning_rate": 0.000159112310704216, + "loss": 0.011, + "step": 7167 + }, + { + "epoch": 3.3731764705882354, + "grad_norm": 0.13856522796292853, + "learning_rate": 0.00015906855510063744, + "loss": 0.0536, + "step": 7168 + }, + { + "epoch": 3.373647058823529, + "grad_norm": 0.10967469203122217, + "learning_rate": 0.0001590247987225597, + "loss": 0.0399, + "step": 7169 + }, + { + "epoch": 3.3741176470588234, + "grad_norm": 0.15550029553249006, + "learning_rate": 0.00015898104157371993, + "loss": 0.058, + "step": 7170 + }, + { + "epoch": 3.3745882352941177, + "grad_norm": 4.2435627958127715e-06, + "learning_rate": 0.00015893728365785518, + "loss": 0.0, + "step": 7171 + }, + { + "epoch": 3.375058823529412, + "grad_norm": 0.04416995141040224, + "learning_rate": 0.00015889352497870253, + "loss": 0.0118, + "step": 7172 + }, + { + "epoch": 3.375529411764706, + "grad_norm": 0.07972480870550201, + "learning_rate": 0.00015884976553999925, + "loss": 0.027, + "step": 7173 + }, + { + "epoch": 3.376, + "grad_norm": 6.233944527583454e-06, + "learning_rate": 0.0001588060053454826, + "loss": 0.0, + "step": 7174 + }, + { + "epoch": 3.376470588235294, + "grad_norm": 0.14183818185686697, + "learning_rate": 0.0001587622443988899, + "loss": 0.0564, + "step": 7175 + }, + { + "epoch": 3.3769411764705883, + "grad_norm": 0.05847350857914236, + "learning_rate": 0.0001587184827039585, + "loss": 0.0132, + "step": 7176 + }, + { + "epoch": 3.377411764705882, + "grad_norm": 0.048681389403156874, + "learning_rate": 0.00015867472026442596, + "loss": 0.0129, + "step": 7177 + }, + { + "epoch": 3.3778823529411763, + "grad_norm": 0.32227581273665584, + "learning_rate": 0.00015863095708402972, + "loss": 0.1046, + "step": 7178 + }, + { + "epoch": 3.3783529411764706, + "grad_norm": 0.00015531815143362537, + "learning_rate": 0.00015858719316650744, + "loss": 0.0, + "step": 7179 + }, + { + "epoch": 3.378823529411765, + "grad_norm": 0.13383652688329264, + "learning_rate": 0.00015854342851559673, + "loss": 0.0447, + "step": 7180 + }, + { + "epoch": 3.379294117647059, + "grad_norm": 1.2043933238162156e-06, + "learning_rate": 0.00015849966313503537, + "loss": 0.0, + "step": 7181 + }, + { + "epoch": 3.379764705882353, + "grad_norm": 0.08303291222898858, + "learning_rate": 0.00015845589702856102, + "loss": 0.0251, + "step": 7182 + }, + { + "epoch": 3.380235294117647, + "grad_norm": 0.10454358869178945, + "learning_rate": 0.0001584121301999116, + "loss": 0.0442, + "step": 7183 + }, + { + "epoch": 3.3807058823529412, + "grad_norm": 0.10271031605061649, + "learning_rate": 0.00015836836265282502, + "loss": 0.064, + "step": 7184 + }, + { + "epoch": 3.3811764705882354, + "grad_norm": 0.06802791544771535, + "learning_rate": 0.0001583245943910392, + "loss": 0.0229, + "step": 7185 + }, + { + "epoch": 3.3816470588235292, + "grad_norm": 4.6566403019035944e-05, + "learning_rate": 0.0001582808254182922, + "loss": 0.0, + "step": 7186 + }, + { + "epoch": 3.3821176470588235, + "grad_norm": 0.10885292984483218, + "learning_rate": 0.0001582370557383221, + "loss": 0.0515, + "step": 7187 + }, + { + "epoch": 3.3825882352941177, + "grad_norm": 0.13208461589489742, + "learning_rate": 0.00015819328535486704, + "loss": 0.0891, + "step": 7188 + }, + { + "epoch": 3.383058823529412, + "grad_norm": 5.394837870963484e-07, + "learning_rate": 0.0001581495142716652, + "loss": 0.0, + "step": 7189 + }, + { + "epoch": 3.383529411764706, + "grad_norm": 1.0891189320543991e-05, + "learning_rate": 0.00015810574249245488, + "loss": 0.0, + "step": 7190 + }, + { + "epoch": 3.384, + "grad_norm": 0.096286405172042, + "learning_rate": 0.00015806197002097438, + "loss": 0.0514, + "step": 7191 + }, + { + "epoch": 3.384470588235294, + "grad_norm": 0.06989885974523122, + "learning_rate": 0.00015801819686096213, + "loss": 0.0317, + "step": 7192 + }, + { + "epoch": 3.3849411764705883, + "grad_norm": 0.10646052826678186, + "learning_rate": 0.00015797442301615652, + "loss": 0.0229, + "step": 7193 + }, + { + "epoch": 3.385411764705882, + "grad_norm": 0.059382859645908344, + "learning_rate": 0.00015793064849029608, + "loss": 0.0221, + "step": 7194 + }, + { + "epoch": 3.3858823529411763, + "grad_norm": 1.852048141917573e-05, + "learning_rate": 0.00015788687328711933, + "loss": 0.0, + "step": 7195 + }, + { + "epoch": 3.3863529411764706, + "grad_norm": 0.07380267197857575, + "learning_rate": 0.00015784309741036492, + "loss": 0.0245, + "step": 7196 + }, + { + "epoch": 3.386823529411765, + "grad_norm": 0.08511813328094497, + "learning_rate": 0.00015779932086377153, + "loss": 0.0278, + "step": 7197 + }, + { + "epoch": 3.387294117647059, + "grad_norm": 0.09089851945445464, + "learning_rate": 0.00015775554365107791, + "loss": 0.0299, + "step": 7198 + }, + { + "epoch": 3.387764705882353, + "grad_norm": 0.07546833374619485, + "learning_rate": 0.0001577117657760228, + "loss": 0.0174, + "step": 7199 + }, + { + "epoch": 3.388235294117647, + "grad_norm": 5.408830848354673e-06, + "learning_rate": 0.00015766798724234506, + "loss": 0.0, + "step": 7200 + }, + { + "epoch": 3.3887058823529412, + "grad_norm": 0.16034662353038523, + "learning_rate": 0.00015762420805378363, + "loss": 0.0725, + "step": 7201 + }, + { + "epoch": 3.3891764705882355, + "grad_norm": 0.10160411593802253, + "learning_rate": 0.00015758042821407742, + "loss": 0.0513, + "step": 7202 + }, + { + "epoch": 3.3896470588235292, + "grad_norm": 0.06550157796827542, + "learning_rate": 0.00015753664772696544, + "loss": 0.0286, + "step": 7203 + }, + { + "epoch": 3.3901176470588235, + "grad_norm": 0.15156997164264238, + "learning_rate": 0.00015749286659618678, + "loss": 0.065, + "step": 7204 + }, + { + "epoch": 3.3905882352941177, + "grad_norm": 0.04412549678277063, + "learning_rate": 0.00015744908482548059, + "loss": 0.0105, + "step": 7205 + }, + { + "epoch": 3.391058823529412, + "grad_norm": 0.12162113568496662, + "learning_rate": 0.00015740530241858597, + "loss": 0.0527, + "step": 7206 + }, + { + "epoch": 3.391529411764706, + "grad_norm": 0.06646815051953416, + "learning_rate": 0.00015736151937924224, + "loss": 0.019, + "step": 7207 + }, + { + "epoch": 3.392, + "grad_norm": 0.16365142291743637, + "learning_rate": 0.00015731773571118864, + "loss": 0.0925, + "step": 7208 + }, + { + "epoch": 3.392470588235294, + "grad_norm": 0.12590657052491652, + "learning_rate": 0.00015727395141816448, + "loss": 0.0773, + "step": 7209 + }, + { + "epoch": 3.3929411764705883, + "grad_norm": 0.035177827150895784, + "learning_rate": 0.00015723016650390922, + "loss": 0.0055, + "step": 7210 + }, + { + "epoch": 3.393411764705882, + "grad_norm": 0.030756159788346338, + "learning_rate": 0.0001571863809721623, + "loss": 0.006, + "step": 7211 + }, + { + "epoch": 3.3938823529411764, + "grad_norm": 0.1100718996621509, + "learning_rate": 0.00015714259482666317, + "loss": 0.0705, + "step": 7212 + }, + { + "epoch": 3.3943529411764706, + "grad_norm": 0.3092958570972227, + "learning_rate": 0.00015709880807115138, + "loss": 0.141, + "step": 7213 + }, + { + "epoch": 3.394823529411765, + "grad_norm": 0.09011081383683772, + "learning_rate": 0.0001570550207093666, + "loss": 0.0448, + "step": 7214 + }, + { + "epoch": 3.395294117647059, + "grad_norm": 0.07416269390402083, + "learning_rate": 0.00015701123274504845, + "loss": 0.0167, + "step": 7215 + }, + { + "epoch": 3.395764705882353, + "grad_norm": 1.0171044311254746e-06, + "learning_rate": 0.00015696744418193664, + "loss": 0.0, + "step": 7216 + }, + { + "epoch": 3.396235294117647, + "grad_norm": 0.10883974495519133, + "learning_rate": 0.00015692365502377088, + "loss": 0.0444, + "step": 7217 + }, + { + "epoch": 3.3967058823529412, + "grad_norm": 1.3072433284729458e-05, + "learning_rate": 0.00015687986527429103, + "loss": 0.0, + "step": 7218 + }, + { + "epoch": 3.3971764705882355, + "grad_norm": 0.023154813046018408, + "learning_rate": 0.00015683607493723699, + "loss": 0.0028, + "step": 7219 + }, + { + "epoch": 3.3976470588235292, + "grad_norm": 0.11483866358848341, + "learning_rate": 0.00015679228401634855, + "loss": 0.0467, + "step": 7220 + }, + { + "epoch": 3.3981176470588235, + "grad_norm": 2.557254677756989e-06, + "learning_rate": 0.0001567484925153658, + "loss": 0.0, + "step": 7221 + }, + { + "epoch": 3.3985882352941177, + "grad_norm": 8.199737120668979e-07, + "learning_rate": 0.00015670470043802867, + "loss": 0.0, + "step": 7222 + }, + { + "epoch": 3.399058823529412, + "grad_norm": 4.807237376706288e-07, + "learning_rate": 0.0001566609077880772, + "loss": 0.0, + "step": 7223 + }, + { + "epoch": 3.3995294117647057, + "grad_norm": 0.06015702662477049, + "learning_rate": 0.00015661711456925158, + "loss": 0.0203, + "step": 7224 + }, + { + "epoch": 3.4, + "grad_norm": 0.08626065113840888, + "learning_rate": 0.0001565733207852919, + "loss": 0.0355, + "step": 7225 + }, + { + "epoch": 3.400470588235294, + "grad_norm": 0.061918690673877466, + "learning_rate": 0.00015652952643993841, + "loss": 0.0228, + "step": 7226 + }, + { + "epoch": 3.4009411764705884, + "grad_norm": 0.05607464799874406, + "learning_rate": 0.00015648573153693128, + "loss": 0.0081, + "step": 7227 + }, + { + "epoch": 3.401411764705882, + "grad_norm": 1.1458592515500852e-06, + "learning_rate": 0.0001564419360800109, + "loss": 0.0, + "step": 7228 + }, + { + "epoch": 3.4018823529411764, + "grad_norm": 3.1550695688274965e-07, + "learning_rate": 0.0001563981400729176, + "loss": 0.0, + "step": 7229 + }, + { + "epoch": 3.4023529411764706, + "grad_norm": 0.12639212717235687, + "learning_rate": 0.0001563543435193917, + "loss": 0.0538, + "step": 7230 + }, + { + "epoch": 3.402823529411765, + "grad_norm": 0.08489632106021021, + "learning_rate": 0.0001563105464231737, + "loss": 0.0349, + "step": 7231 + }, + { + "epoch": 3.403294117647059, + "grad_norm": 0.026140970530257924, + "learning_rate": 0.0001562667487880041, + "loss": 0.0031, + "step": 7232 + }, + { + "epoch": 3.403764705882353, + "grad_norm": 0.03201264095330337, + "learning_rate": 0.00015622295061762346, + "loss": 0.0055, + "step": 7233 + }, + { + "epoch": 3.404235294117647, + "grad_norm": 0.1382190089623257, + "learning_rate": 0.00015617915191577222, + "loss": 0.0472, + "step": 7234 + }, + { + "epoch": 3.4047058823529412, + "grad_norm": 0.07515960977454555, + "learning_rate": 0.0001561353526861912, + "loss": 0.0289, + "step": 7235 + }, + { + "epoch": 3.4051764705882355, + "grad_norm": 0.12487916143181182, + "learning_rate": 0.0001560915529326209, + "loss": 0.0796, + "step": 7236 + }, + { + "epoch": 3.4056470588235292, + "grad_norm": 0.08994144985807317, + "learning_rate": 0.00015604775265880212, + "loss": 0.0305, + "step": 7237 + }, + { + "epoch": 3.4061176470588235, + "grad_norm": 2.680030692238753e-06, + "learning_rate": 0.0001560039518684756, + "loss": 0.0, + "step": 7238 + }, + { + "epoch": 3.4065882352941177, + "grad_norm": 0.06261961006376328, + "learning_rate": 0.00015596015056538215, + "loss": 0.0154, + "step": 7239 + }, + { + "epoch": 3.407058823529412, + "grad_norm": 0.15668013097240327, + "learning_rate": 0.0001559163487532626, + "loss": 0.0757, + "step": 7240 + }, + { + "epoch": 3.4075294117647057, + "grad_norm": 0.04277647518171639, + "learning_rate": 0.00015587254643585784, + "loss": 0.0151, + "step": 7241 + }, + { + "epoch": 3.408, + "grad_norm": 0.09039126467758499, + "learning_rate": 0.00015582874361690885, + "loss": 0.0314, + "step": 7242 + }, + { + "epoch": 3.408470588235294, + "grad_norm": 9.037131826306805e-08, + "learning_rate": 0.00015578494030015652, + "loss": 0.0, + "step": 7243 + }, + { + "epoch": 3.4089411764705884, + "grad_norm": 0.16755076587591564, + "learning_rate": 0.00015574113648934192, + "loss": 0.0696, + "step": 7244 + }, + { + "epoch": 3.409411764705882, + "grad_norm": 0.021850683159431016, + "learning_rate": 0.0001556973321882061, + "loss": 0.0024, + "step": 7245 + }, + { + "epoch": 3.4098823529411764, + "grad_norm": 0.05193017226642431, + "learning_rate": 0.0001556535274004902, + "loss": 0.0213, + "step": 7246 + }, + { + "epoch": 3.4103529411764706, + "grad_norm": 0.0815402362501232, + "learning_rate": 0.00015560972212993527, + "loss": 0.0357, + "step": 7247 + }, + { + "epoch": 3.410823529411765, + "grad_norm": 0.0008435837333257744, + "learning_rate": 0.0001555659163802826, + "loss": 0.0, + "step": 7248 + }, + { + "epoch": 3.411294117647059, + "grad_norm": 0.10978659245043253, + "learning_rate": 0.00015552211015527336, + "loss": 0.0497, + "step": 7249 + }, + { + "epoch": 3.411764705882353, + "grad_norm": 8.543901959495972e-08, + "learning_rate": 0.00015547830345864885, + "loss": 0.0, + "step": 7250 + }, + { + "epoch": 3.412235294117647, + "grad_norm": 0.06794664614291591, + "learning_rate": 0.00015543449629415028, + "loss": 0.0246, + "step": 7251 + }, + { + "epoch": 3.4127058823529413, + "grad_norm": 0.053903531230740494, + "learning_rate": 0.00015539068866551916, + "loss": 0.0222, + "step": 7252 + }, + { + "epoch": 3.4131764705882355, + "grad_norm": 0.07239116166441349, + "learning_rate": 0.00015534688057649673, + "loss": 0.0112, + "step": 7253 + }, + { + "epoch": 3.4136470588235293, + "grad_norm": 0.12545526958553943, + "learning_rate": 0.0001553030720308245, + "loss": 0.0304, + "step": 7254 + }, + { + "epoch": 3.4141176470588235, + "grad_norm": 0.11195625293619585, + "learning_rate": 0.00015525926303224387, + "loss": 0.0571, + "step": 7255 + }, + { + "epoch": 3.4145882352941177, + "grad_norm": 0.1437192533905299, + "learning_rate": 0.0001552154535844964, + "loss": 0.0728, + "step": 7256 + }, + { + "epoch": 3.415058823529412, + "grad_norm": 0.10695771127893024, + "learning_rate": 0.00015517164369132364, + "loss": 0.0762, + "step": 7257 + }, + { + "epoch": 3.4155294117647057, + "grad_norm": 0.14311405557469303, + "learning_rate": 0.0001551278333564671, + "loss": 0.0794, + "step": 7258 + }, + { + "epoch": 3.416, + "grad_norm": 0.25622843892113695, + "learning_rate": 0.00015508402258366846, + "loss": 0.0529, + "step": 7259 + }, + { + "epoch": 3.416470588235294, + "grad_norm": 0.1883445934406114, + "learning_rate": 0.00015504021137666938, + "loss": 0.079, + "step": 7260 + }, + { + "epoch": 3.4169411764705884, + "grad_norm": 0.16745343150149253, + "learning_rate": 0.00015499639973921148, + "loss": 0.0821, + "step": 7261 + }, + { + "epoch": 3.417411764705882, + "grad_norm": 1.0256345881851148e-05, + "learning_rate": 0.00015495258767503657, + "loss": 0.0, + "step": 7262 + }, + { + "epoch": 3.4178823529411764, + "grad_norm": 6.64613015390628e-05, + "learning_rate": 0.0001549087751878864, + "loss": 0.0, + "step": 7263 + }, + { + "epoch": 3.4183529411764706, + "grad_norm": 0.11128118777751306, + "learning_rate": 0.00015486496228150266, + "loss": 0.0464, + "step": 7264 + }, + { + "epoch": 3.418823529411765, + "grad_norm": 2.5841112845843848e-06, + "learning_rate": 0.00015482114895962735, + "loss": 0.0, + "step": 7265 + }, + { + "epoch": 3.419294117647059, + "grad_norm": 0.15130163214696146, + "learning_rate": 0.00015477733522600227, + "loss": 0.0705, + "step": 7266 + }, + { + "epoch": 3.419764705882353, + "grad_norm": 1.9843435582138405e-05, + "learning_rate": 0.00015473352108436934, + "loss": 0.0, + "step": 7267 + }, + { + "epoch": 3.420235294117647, + "grad_norm": 0.10297637113052326, + "learning_rate": 0.00015468970653847043, + "loss": 0.0533, + "step": 7268 + }, + { + "epoch": 3.4207058823529413, + "grad_norm": 0.1654626495439898, + "learning_rate": 0.00015464589159204764, + "loss": 0.0631, + "step": 7269 + }, + { + "epoch": 3.4211764705882355, + "grad_norm": 0.03566682325968756, + "learning_rate": 0.0001546020762488429, + "loss": 0.0061, + "step": 7270 + }, + { + "epoch": 3.4216470588235293, + "grad_norm": 3.05316478697053e-05, + "learning_rate": 0.00015455826051259826, + "loss": 0.0, + "step": 7271 + }, + { + "epoch": 3.4221176470588235, + "grad_norm": 0.019771727897960587, + "learning_rate": 0.00015451444438705584, + "loss": 0.0023, + "step": 7272 + }, + { + "epoch": 3.4225882352941177, + "grad_norm": 0.14945471623023207, + "learning_rate": 0.00015447062787595772, + "loss": 0.039, + "step": 7273 + }, + { + "epoch": 3.423058823529412, + "grad_norm": 3.915456217252457e-06, + "learning_rate": 0.00015442681098304603, + "loss": 0.0, + "step": 7274 + }, + { + "epoch": 3.4235294117647057, + "grad_norm": 0.03761394278196328, + "learning_rate": 0.000154382993712063, + "loss": 0.0048, + "step": 7275 + }, + { + "epoch": 3.424, + "grad_norm": 0.07402290796023568, + "learning_rate": 0.00015433917606675075, + "loss": 0.0134, + "step": 7276 + }, + { + "epoch": 3.424470588235294, + "grad_norm": 0.12921754485625844, + "learning_rate": 0.00015429535805085162, + "loss": 0.04, + "step": 7277 + }, + { + "epoch": 3.4249411764705884, + "grad_norm": 2.9783173116633155e-06, + "learning_rate": 0.00015425153966810783, + "loss": 0.0, + "step": 7278 + }, + { + "epoch": 3.425411764705882, + "grad_norm": 0.10042526126253795, + "learning_rate": 0.00015420772092226173, + "loss": 0.0228, + "step": 7279 + }, + { + "epoch": 3.4258823529411764, + "grad_norm": 0.10103615312990377, + "learning_rate": 0.00015416390181705562, + "loss": 0.0505, + "step": 7280 + }, + { + "epoch": 3.4263529411764706, + "grad_norm": 5.302522445375483e-05, + "learning_rate": 0.00015412008235623186, + "loss": 0.0, + "step": 7281 + }, + { + "epoch": 3.426823529411765, + "grad_norm": 0.16074269250483994, + "learning_rate": 0.00015407626254353282, + "loss": 0.0583, + "step": 7282 + }, + { + "epoch": 3.427294117647059, + "grad_norm": 0.08626359158586137, + "learning_rate": 0.00015403244238270102, + "loss": 0.0198, + "step": 7283 + }, + { + "epoch": 3.427764705882353, + "grad_norm": 0.12307653294927912, + "learning_rate": 0.00015398862187747888, + "loss": 0.0741, + "step": 7284 + }, + { + "epoch": 3.428235294117647, + "grad_norm": 0.0816095714307904, + "learning_rate": 0.0001539448010316088, + "loss": 0.0417, + "step": 7285 + }, + { + "epoch": 3.4287058823529413, + "grad_norm": 0.13986505243498645, + "learning_rate": 0.0001539009798488334, + "loss": 0.0625, + "step": 7286 + }, + { + "epoch": 3.429176470588235, + "grad_norm": 0.0002476161738108913, + "learning_rate": 0.0001538571583328952, + "loss": 0.0, + "step": 7287 + }, + { + "epoch": 3.4296470588235293, + "grad_norm": 0.07855090608051875, + "learning_rate": 0.0001538133364875367, + "loss": 0.0213, + "step": 7288 + }, + { + "epoch": 3.4301176470588235, + "grad_norm": 1.960141699241133e-08, + "learning_rate": 0.00015376951431650063, + "loss": 0.0, + "step": 7289 + }, + { + "epoch": 3.4305882352941177, + "grad_norm": 0.09522014828505258, + "learning_rate": 0.00015372569182352952, + "loss": 0.0228, + "step": 7290 + }, + { + "epoch": 3.431058823529412, + "grad_norm": 0.10865141403509801, + "learning_rate": 0.00015368186901236605, + "loss": 0.0354, + "step": 7291 + }, + { + "epoch": 3.4315294117647057, + "grad_norm": 0.14099149468070366, + "learning_rate": 0.00015363804588675292, + "loss": 0.0303, + "step": 7292 + }, + { + "epoch": 3.432, + "grad_norm": 2.3059430663947658e-07, + "learning_rate": 0.00015359422245043278, + "loss": 0.0, + "step": 7293 + }, + { + "epoch": 3.432470588235294, + "grad_norm": 0.05926613214907859, + "learning_rate": 0.00015355039870714848, + "loss": 0.0206, + "step": 7294 + }, + { + "epoch": 3.4329411764705884, + "grad_norm": 0.08691770685332637, + "learning_rate": 0.00015350657466064264, + "loss": 0.0331, + "step": 7295 + }, + { + "epoch": 3.433411764705882, + "grad_norm": 0.1140357120865439, + "learning_rate": 0.00015346275031465818, + "loss": 0.0491, + "step": 7296 + }, + { + "epoch": 3.4338823529411764, + "grad_norm": 4.194337813068567e-05, + "learning_rate": 0.00015341892567293785, + "loss": 0.0, + "step": 7297 + }, + { + "epoch": 3.4343529411764706, + "grad_norm": 0.04569561967132539, + "learning_rate": 0.0001533751007392245, + "loss": 0.0064, + "step": 7298 + }, + { + "epoch": 3.434823529411765, + "grad_norm": 9.271675065517723e-05, + "learning_rate": 0.00015333127551726096, + "loss": 0.0, + "step": 7299 + }, + { + "epoch": 3.435294117647059, + "grad_norm": 0.20866481591780983, + "learning_rate": 0.0001532874500107902, + "loss": 0.1458, + "step": 7300 + }, + { + "epoch": 3.435764705882353, + "grad_norm": 2.7385322226868173e-07, + "learning_rate": 0.00015324362422355507, + "loss": 0.0, + "step": 7301 + }, + { + "epoch": 3.436235294117647, + "grad_norm": 9.013350062578373e-06, + "learning_rate": 0.00015319979815929847, + "loss": 0.0, + "step": 7302 + }, + { + "epoch": 3.4367058823529413, + "grad_norm": 0.07223188281842964, + "learning_rate": 0.00015315597182176348, + "loss": 0.0253, + "step": 7303 + }, + { + "epoch": 3.437176470588235, + "grad_norm": 0.05851458085814704, + "learning_rate": 0.000153112145214693, + "loss": 0.0207, + "step": 7304 + }, + { + "epoch": 3.4376470588235293, + "grad_norm": 0.2046741300990326, + "learning_rate": 0.00015306831834183008, + "loss": 0.1344, + "step": 7305 + }, + { + "epoch": 3.4381176470588235, + "grad_norm": 0.023468680059600167, + "learning_rate": 0.00015302449120691765, + "loss": 0.0039, + "step": 7306 + }, + { + "epoch": 3.4385882352941177, + "grad_norm": 0.0001852721195173871, + "learning_rate": 0.00015298066381369888, + "loss": 0.0, + "step": 7307 + }, + { + "epoch": 3.439058823529412, + "grad_norm": 0.0820894307264369, + "learning_rate": 0.0001529368361659168, + "loss": 0.0311, + "step": 7308 + }, + { + "epoch": 3.4395294117647057, + "grad_norm": 0.07758261850302223, + "learning_rate": 0.0001528930082673145, + "loss": 0.04, + "step": 7309 + }, + { + "epoch": 3.44, + "grad_norm": 0.07253207492915523, + "learning_rate": 0.0001528491801216351, + "loss": 0.0305, + "step": 7310 + }, + { + "epoch": 3.440470588235294, + "grad_norm": 0.05916355713978767, + "learning_rate": 0.00015280535173262176, + "loss": 0.017, + "step": 7311 + }, + { + "epoch": 3.4409411764705884, + "grad_norm": 0.07417718131629114, + "learning_rate": 0.0001527615231040176, + "loss": 0.0131, + "step": 7312 + }, + { + "epoch": 3.441411764705882, + "grad_norm": 1.656233689282321e-07, + "learning_rate": 0.00015271769423956584, + "loss": 0.0, + "step": 7313 + }, + { + "epoch": 3.4418823529411764, + "grad_norm": 0.10827826398059194, + "learning_rate": 0.00015267386514300965, + "loss": 0.0447, + "step": 7314 + }, + { + "epoch": 3.4423529411764706, + "grad_norm": 0.06378587387694852, + "learning_rate": 0.00015263003581809226, + "loss": 0.0203, + "step": 7315 + }, + { + "epoch": 3.442823529411765, + "grad_norm": 0.10745582319969911, + "learning_rate": 0.00015258620626855691, + "loss": 0.0387, + "step": 7316 + }, + { + "epoch": 3.443294117647059, + "grad_norm": 0.042959666713636345, + "learning_rate": 0.00015254237649814687, + "loss": 0.01, + "step": 7317 + }, + { + "epoch": 3.443764705882353, + "grad_norm": 0.10258413414765553, + "learning_rate": 0.00015249854651060542, + "loss": 0.0503, + "step": 7318 + }, + { + "epoch": 3.444235294117647, + "grad_norm": 0.0378513011070538, + "learning_rate": 0.00015245471630967583, + "loss": 0.0055, + "step": 7319 + }, + { + "epoch": 3.4447058823529413, + "grad_norm": 0.11411940800605853, + "learning_rate": 0.00015241088589910138, + "loss": 0.0575, + "step": 7320 + }, + { + "epoch": 3.445176470588235, + "grad_norm": 0.04763194681521065, + "learning_rate": 0.0001523670552826255, + "loss": 0.0131, + "step": 7321 + }, + { + "epoch": 3.4456470588235293, + "grad_norm": 0.032699398938255894, + "learning_rate": 0.00015232322446399147, + "loss": 0.0046, + "step": 7322 + }, + { + "epoch": 3.4461176470588235, + "grad_norm": 0.12269610166178173, + "learning_rate": 0.00015227939344694268, + "loss": 0.0738, + "step": 7323 + }, + { + "epoch": 3.4465882352941177, + "grad_norm": 0.22237447394290752, + "learning_rate": 0.0001522355622352225, + "loss": 0.1124, + "step": 7324 + }, + { + "epoch": 3.447058823529412, + "grad_norm": 0.09608294791963985, + "learning_rate": 0.00015219173083257437, + "loss": 0.0381, + "step": 7325 + }, + { + "epoch": 3.4475294117647057, + "grad_norm": 0.0683231184975389, + "learning_rate": 0.00015214789924274168, + "loss": 0.0284, + "step": 7326 + }, + { + "epoch": 3.448, + "grad_norm": 0.11238841418996061, + "learning_rate": 0.00015210406746946783, + "loss": 0.0198, + "step": 7327 + }, + { + "epoch": 3.448470588235294, + "grad_norm": 0.03948433491485898, + "learning_rate": 0.0001520602355164963, + "loss": 0.0105, + "step": 7328 + }, + { + "epoch": 3.4489411764705884, + "grad_norm": 0.12621407998944595, + "learning_rate": 0.00015201640338757054, + "loss": 0.0471, + "step": 7329 + }, + { + "epoch": 3.449411764705882, + "grad_norm": 0.1356996307265802, + "learning_rate": 0.00015197257108643406, + "loss": 0.0792, + "step": 7330 + }, + { + "epoch": 3.4498823529411764, + "grad_norm": 0.07329229657257404, + "learning_rate": 0.00015192873861683036, + "loss": 0.0214, + "step": 7331 + }, + { + "epoch": 3.4503529411764706, + "grad_norm": 0.10258454537228709, + "learning_rate": 0.00015188490598250288, + "loss": 0.0391, + "step": 7332 + }, + { + "epoch": 3.450823529411765, + "grad_norm": 0.16169263007417867, + "learning_rate": 0.0001518410731871952, + "loss": 0.0929, + "step": 7333 + }, + { + "epoch": 3.451294117647059, + "grad_norm": 0.08538434297375545, + "learning_rate": 0.00015179724023465084, + "loss": 0.0366, + "step": 7334 + }, + { + "epoch": 3.451764705882353, + "grad_norm": 7.156303991742159e-05, + "learning_rate": 0.00015175340712861337, + "loss": 0.0, + "step": 7335 + }, + { + "epoch": 3.452235294117647, + "grad_norm": 0.0002614609907606436, + "learning_rate": 0.00015170957387282627, + "loss": 0.0, + "step": 7336 + }, + { + "epoch": 3.4527058823529413, + "grad_norm": 1.0830891390064634e-06, + "learning_rate": 0.00015166574047103322, + "loss": 0.0, + "step": 7337 + }, + { + "epoch": 3.453176470588235, + "grad_norm": 0.06546357068074062, + "learning_rate": 0.00015162190692697778, + "loss": 0.0337, + "step": 7338 + }, + { + "epoch": 3.4536470588235293, + "grad_norm": 0.04040215027157333, + "learning_rate": 0.00015157807324440347, + "loss": 0.0079, + "step": 7339 + }, + { + "epoch": 3.4541176470588235, + "grad_norm": 0.1391768192281272, + "learning_rate": 0.00015153423942705402, + "loss": 0.0711, + "step": 7340 + }, + { + "epoch": 3.4545882352941177, + "grad_norm": 0.062106699334899006, + "learning_rate": 0.000151490405478673, + "loss": 0.0165, + "step": 7341 + }, + { + "epoch": 3.455058823529412, + "grad_norm": 0.15183359311130165, + "learning_rate": 0.000151446571403004, + "loss": 0.0709, + "step": 7342 + }, + { + "epoch": 3.4555294117647057, + "grad_norm": 0.12471502435010025, + "learning_rate": 0.0001514027372037907, + "loss": 0.0595, + "step": 7343 + }, + { + "epoch": 3.456, + "grad_norm": 2.590460635415581e-06, + "learning_rate": 0.0001513589028847768, + "loss": 0.0, + "step": 7344 + }, + { + "epoch": 3.456470588235294, + "grad_norm": 0.12471838905911664, + "learning_rate": 0.00015131506844970596, + "loss": 0.0694, + "step": 7345 + }, + { + "epoch": 3.4569411764705884, + "grad_norm": 0.052577901723178, + "learning_rate": 0.00015127123390232177, + "loss": 0.0107, + "step": 7346 + }, + { + "epoch": 3.457411764705882, + "grad_norm": 0.08532160191636487, + "learning_rate": 0.00015122739924636796, + "loss": 0.0241, + "step": 7347 + }, + { + "epoch": 3.4578823529411764, + "grad_norm": 0.05039204492965688, + "learning_rate": 0.00015118356448558828, + "loss": 0.014, + "step": 7348 + }, + { + "epoch": 3.4583529411764706, + "grad_norm": 0.13780019620492773, + "learning_rate": 0.00015113972962372635, + "loss": 0.0606, + "step": 7349 + }, + { + "epoch": 3.458823529411765, + "grad_norm": 0.0575393691228786, + "learning_rate": 0.00015109589466452594, + "loss": 0.0144, + "step": 7350 + }, + { + "epoch": 3.4592941176470586, + "grad_norm": 0.08504467112128244, + "learning_rate": 0.00015105205961173074, + "loss": 0.0299, + "step": 7351 + }, + { + "epoch": 3.459764705882353, + "grad_norm": 0.041718518176412124, + "learning_rate": 0.00015100822446908452, + "loss": 0.0138, + "step": 7352 + }, + { + "epoch": 3.460235294117647, + "grad_norm": 0.09068695562772215, + "learning_rate": 0.00015096438924033095, + "loss": 0.0307, + "step": 7353 + }, + { + "epoch": 3.4607058823529413, + "grad_norm": 0.18503975973643344, + "learning_rate": 0.00015092055392921383, + "loss": 0.1076, + "step": 7354 + }, + { + "epoch": 3.461176470588235, + "grad_norm": 4.536777001589839e-05, + "learning_rate": 0.00015087671853947684, + "loss": 0.0, + "step": 7355 + }, + { + "epoch": 3.4616470588235293, + "grad_norm": 0.15800226339758872, + "learning_rate": 0.00015083288307486388, + "loss": 0.0658, + "step": 7356 + }, + { + "epoch": 3.4621176470588235, + "grad_norm": 0.0227755929813633, + "learning_rate": 0.00015078904753911853, + "loss": 0.0034, + "step": 7357 + }, + { + "epoch": 3.4625882352941177, + "grad_norm": 0.06758145530229173, + "learning_rate": 0.0001507452119359847, + "loss": 0.0209, + "step": 7358 + }, + { + "epoch": 3.463058823529412, + "grad_norm": 0.030049342881299974, + "learning_rate": 0.0001507013762692061, + "loss": 0.0065, + "step": 7359 + }, + { + "epoch": 3.4635294117647057, + "grad_norm": 0.015265673834594614, + "learning_rate": 0.0001506575405425265, + "loss": 0.0024, + "step": 7360 + }, + { + "epoch": 3.464, + "grad_norm": 0.03585413032635706, + "learning_rate": 0.00015061370475968975, + "loss": 0.0072, + "step": 7361 + }, + { + "epoch": 3.464470588235294, + "grad_norm": 0.07516041103820603, + "learning_rate": 0.0001505698689244396, + "loss": 0.027, + "step": 7362 + }, + { + "epoch": 3.4649411764705884, + "grad_norm": 0.00010249568076641974, + "learning_rate": 0.0001505260330405198, + "loss": 0.0, + "step": 7363 + }, + { + "epoch": 3.465411764705882, + "grad_norm": 0.2491746702219166, + "learning_rate": 0.00015048219711167422, + "loss": 0.0818, + "step": 7364 + }, + { + "epoch": 3.4658823529411764, + "grad_norm": 0.07483125240967664, + "learning_rate": 0.00015043836114164667, + "loss": 0.0112, + "step": 7365 + }, + { + "epoch": 3.4663529411764706, + "grad_norm": 0.04342990247862331, + "learning_rate": 0.00015039452513418094, + "loss": 0.0104, + "step": 7366 + }, + { + "epoch": 3.466823529411765, + "grad_norm": 3.5314727034543835e-08, + "learning_rate": 0.00015035068909302079, + "loss": 0.0, + "step": 7367 + }, + { + "epoch": 3.4672941176470586, + "grad_norm": 0.14678411403501365, + "learning_rate": 0.00015030685302191011, + "loss": 0.0608, + "step": 7368 + }, + { + "epoch": 3.467764705882353, + "grad_norm": 0.10464000616848868, + "learning_rate": 0.00015026301692459265, + "loss": 0.0225, + "step": 7369 + }, + { + "epoch": 3.468235294117647, + "grad_norm": 0.062289207776289365, + "learning_rate": 0.00015021918080481225, + "loss": 0.0193, + "step": 7370 + }, + { + "epoch": 3.4687058823529413, + "grad_norm": 0.07977162080936795, + "learning_rate": 0.00015017534466631272, + "loss": 0.0435, + "step": 7371 + }, + { + "epoch": 3.469176470588235, + "grad_norm": 0.09934657268597213, + "learning_rate": 0.00015013150851283795, + "loss": 0.0483, + "step": 7372 + }, + { + "epoch": 3.4696470588235293, + "grad_norm": 0.1316547731768977, + "learning_rate": 0.0001500876723481317, + "loss": 0.0562, + "step": 7373 + }, + { + "epoch": 3.4701176470588235, + "grad_norm": 0.13801125385521648, + "learning_rate": 0.00015004383617593772, + "loss": 0.0765, + "step": 7374 + }, + { + "epoch": 3.4705882352941178, + "grad_norm": 2.773690607696709e-06, + "learning_rate": 0.00015, + "loss": 0.0, + "step": 7375 + }, + { + "epoch": 3.471058823529412, + "grad_norm": 0.11891817154474006, + "learning_rate": 0.00014995616382406225, + "loss": 0.0281, + "step": 7376 + }, + { + "epoch": 3.4715294117647058, + "grad_norm": 0.06495705592668378, + "learning_rate": 0.00014991232765186832, + "loss": 0.0187, + "step": 7377 + }, + { + "epoch": 3.472, + "grad_norm": 0.09886158627220065, + "learning_rate": 0.00014986849148716205, + "loss": 0.0498, + "step": 7378 + }, + { + "epoch": 3.472470588235294, + "grad_norm": 0.07543794856804668, + "learning_rate": 0.00014982465533368725, + "loss": 0.0215, + "step": 7379 + }, + { + "epoch": 3.4729411764705884, + "grad_norm": 0.04376700388478498, + "learning_rate": 0.00014978081919518772, + "loss": 0.0122, + "step": 7380 + }, + { + "epoch": 3.473411764705882, + "grad_norm": 0.06308173283792563, + "learning_rate": 0.00014973698307540735, + "loss": 0.019, + "step": 7381 + }, + { + "epoch": 3.4738823529411764, + "grad_norm": 0.05677736721667051, + "learning_rate": 0.00014969314697808991, + "loss": 0.0205, + "step": 7382 + }, + { + "epoch": 3.4743529411764706, + "grad_norm": 0.10115113430822753, + "learning_rate": 0.00014964931090697922, + "loss": 0.0568, + "step": 7383 + }, + { + "epoch": 3.474823529411765, + "grad_norm": 0.11375847107410943, + "learning_rate": 0.00014960547486581906, + "loss": 0.0637, + "step": 7384 + }, + { + "epoch": 3.4752941176470586, + "grad_norm": 0.11654295411175822, + "learning_rate": 0.0001495616388583533, + "loss": 0.036, + "step": 7385 + }, + { + "epoch": 3.475764705882353, + "grad_norm": 0.131702712774709, + "learning_rate": 0.00014951780288832575, + "loss": 0.0701, + "step": 7386 + }, + { + "epoch": 3.476235294117647, + "grad_norm": 0.05903869918612907, + "learning_rate": 0.0001494739669594802, + "loss": 0.0109, + "step": 7387 + }, + { + "epoch": 3.4767058823529413, + "grad_norm": 0.1763263684492959, + "learning_rate": 0.00014943013107556044, + "loss": 0.0796, + "step": 7388 + }, + { + "epoch": 3.477176470588235, + "grad_norm": 0.03754162067540534, + "learning_rate": 0.00014938629524031028, + "loss": 0.0078, + "step": 7389 + }, + { + "epoch": 3.4776470588235293, + "grad_norm": 0.10326660989379331, + "learning_rate": 0.00014934245945747349, + "loss": 0.0536, + "step": 7390 + }, + { + "epoch": 3.4781176470588235, + "grad_norm": 1.427643431698549e-05, + "learning_rate": 0.00014929862373079392, + "loss": 0.0, + "step": 7391 + }, + { + "epoch": 3.4785882352941178, + "grad_norm": 0.00035322868333460827, + "learning_rate": 0.00014925478806401533, + "loss": 0.0, + "step": 7392 + }, + { + "epoch": 3.479058823529412, + "grad_norm": 0.03610029983212002, + "learning_rate": 0.0001492109524608815, + "loss": 0.0066, + "step": 7393 + }, + { + "epoch": 3.4795294117647058, + "grad_norm": 0.0644169502546018, + "learning_rate": 0.00014916711692513618, + "loss": 0.0167, + "step": 7394 + }, + { + "epoch": 3.48, + "grad_norm": 0.07059051329092672, + "learning_rate": 0.00014912328146052316, + "loss": 0.0234, + "step": 7395 + }, + { + "epoch": 3.480470588235294, + "grad_norm": 0.1562374766640375, + "learning_rate": 0.00014907944607078623, + "loss": 0.0609, + "step": 7396 + }, + { + "epoch": 3.4809411764705884, + "grad_norm": 0.13626856702762552, + "learning_rate": 0.00014903561075966908, + "loss": 0.0786, + "step": 7397 + }, + { + "epoch": 3.481411764705882, + "grad_norm": 0.03406826435733998, + "learning_rate": 0.00014899177553091543, + "loss": 0.0079, + "step": 7398 + }, + { + "epoch": 3.4818823529411764, + "grad_norm": 0.11330911469162504, + "learning_rate": 0.0001489479403882692, + "loss": 0.0399, + "step": 7399 + }, + { + "epoch": 3.4823529411764707, + "grad_norm": 0.04716242164688744, + "learning_rate": 0.00014890410533547404, + "loss": 0.0116, + "step": 7400 + }, + { + "epoch": 3.482823529411765, + "grad_norm": 0.13183076905149796, + "learning_rate": 0.00014886027037627362, + "loss": 0.052, + "step": 7401 + }, + { + "epoch": 3.4832941176470587, + "grad_norm": 0.06340723703191728, + "learning_rate": 0.0001488164355144117, + "loss": 0.028, + "step": 7402 + }, + { + "epoch": 3.483764705882353, + "grad_norm": 0.13113162237449125, + "learning_rate": 0.00014877260075363199, + "loss": 0.0762, + "step": 7403 + }, + { + "epoch": 3.484235294117647, + "grad_norm": 0.20362817885199277, + "learning_rate": 0.00014872876609767823, + "loss": 0.0593, + "step": 7404 + }, + { + "epoch": 3.4847058823529413, + "grad_norm": 6.9719097294687575e-06, + "learning_rate": 0.00014868493155029402, + "loss": 0.0, + "step": 7405 + }, + { + "epoch": 3.485176470588235, + "grad_norm": 0.0005832740497379046, + "learning_rate": 0.00014864109711522315, + "loss": 0.0, + "step": 7406 + }, + { + "epoch": 3.4856470588235293, + "grad_norm": 0.00012203587928782175, + "learning_rate": 0.00014859726279620926, + "loss": 0.0, + "step": 7407 + }, + { + "epoch": 3.4861176470588235, + "grad_norm": 0.12358397642083271, + "learning_rate": 0.00014855342859699597, + "loss": 0.046, + "step": 7408 + }, + { + "epoch": 3.4865882352941178, + "grad_norm": 1.5469445105775677e-06, + "learning_rate": 0.000148509594521327, + "loss": 0.0, + "step": 7409 + }, + { + "epoch": 3.487058823529412, + "grad_norm": 0.08559481481682339, + "learning_rate": 0.00014846576057294596, + "loss": 0.0339, + "step": 7410 + }, + { + "epoch": 3.4875294117647058, + "grad_norm": 0.035417914160917496, + "learning_rate": 0.0001484219267555965, + "loss": 0.0086, + "step": 7411 + }, + { + "epoch": 3.488, + "grad_norm": 0.07620463153113033, + "learning_rate": 0.0001483780930730222, + "loss": 0.0231, + "step": 7412 + }, + { + "epoch": 3.488470588235294, + "grad_norm": 0.028370935271378347, + "learning_rate": 0.00014833425952896675, + "loss": 0.006, + "step": 7413 + }, + { + "epoch": 3.488941176470588, + "grad_norm": 0.20178028829409375, + "learning_rate": 0.00014829042612717373, + "loss": 0.0909, + "step": 7414 + }, + { + "epoch": 3.489411764705882, + "grad_norm": 0.5036482668463073, + "learning_rate": 0.00014824659287138663, + "loss": 0.1517, + "step": 7415 + }, + { + "epoch": 3.4898823529411764, + "grad_norm": 0.08355118972304831, + "learning_rate": 0.00014820275976534916, + "loss": 0.0425, + "step": 7416 + }, + { + "epoch": 3.4903529411764707, + "grad_norm": 0.09344210946211663, + "learning_rate": 0.0001481589268128048, + "loss": 0.0309, + "step": 7417 + }, + { + "epoch": 3.490823529411765, + "grad_norm": 1.2303821674445168e-05, + "learning_rate": 0.0001481150940174971, + "loss": 0.0, + "step": 7418 + }, + { + "epoch": 3.4912941176470587, + "grad_norm": 0.04355559353505412, + "learning_rate": 0.00014807126138316964, + "loss": 0.0082, + "step": 7419 + }, + { + "epoch": 3.491764705882353, + "grad_norm": 0.07284012785183294, + "learning_rate": 0.00014802742891356592, + "loss": 0.025, + "step": 7420 + }, + { + "epoch": 3.492235294117647, + "grad_norm": 0.12001493735958418, + "learning_rate": 0.00014798359661242943, + "loss": 0.0659, + "step": 7421 + }, + { + "epoch": 3.4927058823529413, + "grad_norm": 0.15224838379806774, + "learning_rate": 0.00014793976448350366, + "loss": 0.0337, + "step": 7422 + }, + { + "epoch": 3.493176470588235, + "grad_norm": 0.06669029089420073, + "learning_rate": 0.00014789593253053217, + "loss": 0.0226, + "step": 7423 + }, + { + "epoch": 3.4936470588235293, + "grad_norm": 0.1403592499789062, + "learning_rate": 0.00014785210075725835, + "loss": 0.0383, + "step": 7424 + }, + { + "epoch": 3.4941176470588236, + "grad_norm": 0.09429424665805392, + "learning_rate": 0.0001478082691674256, + "loss": 0.0304, + "step": 7425 + }, + { + "epoch": 3.4945882352941178, + "grad_norm": 0.09595921097850209, + "learning_rate": 0.00014776443776477746, + "loss": 0.0479, + "step": 7426 + }, + { + "epoch": 3.495058823529412, + "grad_norm": 0.09780036594228775, + "learning_rate": 0.00014772060655305732, + "loss": 0.0457, + "step": 7427 + }, + { + "epoch": 3.4955294117647058, + "grad_norm": 0.05108715798622859, + "learning_rate": 0.0001476767755360085, + "loss": 0.0123, + "step": 7428 + }, + { + "epoch": 3.496, + "grad_norm": 0.07803048173603465, + "learning_rate": 0.00014763294471737448, + "loss": 0.0175, + "step": 7429 + }, + { + "epoch": 3.496470588235294, + "grad_norm": 0.1494816883215476, + "learning_rate": 0.00014758911410089862, + "loss": 0.0765, + "step": 7430 + }, + { + "epoch": 3.496941176470588, + "grad_norm": 0.16009867305461306, + "learning_rate": 0.0001475452836903242, + "loss": 0.065, + "step": 7431 + }, + { + "epoch": 3.497411764705882, + "grad_norm": 0.07338390176539263, + "learning_rate": 0.0001475014534893946, + "loss": 0.0334, + "step": 7432 + }, + { + "epoch": 3.4978823529411764, + "grad_norm": 0.05371061989690071, + "learning_rate": 0.00014745762350185316, + "loss": 0.0225, + "step": 7433 + }, + { + "epoch": 3.4983529411764707, + "grad_norm": 1.665861852407919e-06, + "learning_rate": 0.0001474137937314431, + "loss": 0.0, + "step": 7434 + }, + { + "epoch": 3.498823529411765, + "grad_norm": 1.0231361281712158e-06, + "learning_rate": 0.00014736996418190774, + "loss": 0.0, + "step": 7435 + }, + { + "epoch": 3.4992941176470587, + "grad_norm": 0.07027306013016682, + "learning_rate": 0.00014732613485699035, + "loss": 0.0222, + "step": 7436 + }, + { + "epoch": 3.499764705882353, + "grad_norm": 2.5619973184078524e-07, + "learning_rate": 0.0001472823057604342, + "loss": 0.0, + "step": 7437 + }, + { + "epoch": 3.500235294117647, + "grad_norm": 1.9921012566774195e-06, + "learning_rate": 0.0001472384768959824, + "loss": 0.0, + "step": 7438 + }, + { + "epoch": 3.5007058823529413, + "grad_norm": 0.08202899707556115, + "learning_rate": 0.00014719464826737827, + "loss": 0.0259, + "step": 7439 + }, + { + "epoch": 3.501176470588235, + "grad_norm": 0.16900164686055857, + "learning_rate": 0.0001471508198783649, + "loss": 0.0782, + "step": 7440 + }, + { + "epoch": 3.5016470588235293, + "grad_norm": 1.440577631488473e-06, + "learning_rate": 0.0001471069917326855, + "loss": 0.0, + "step": 7441 + }, + { + "epoch": 3.5021176470588236, + "grad_norm": 7.292055522735313e-06, + "learning_rate": 0.0001470631638340832, + "loss": 0.0, + "step": 7442 + }, + { + "epoch": 3.502588235294118, + "grad_norm": 0.04469094317011611, + "learning_rate": 0.00014701933618630107, + "loss": 0.0125, + "step": 7443 + }, + { + "epoch": 3.503058823529412, + "grad_norm": 0.1102444776512277, + "learning_rate": 0.00014697550879308233, + "loss": 0.0381, + "step": 7444 + }, + { + "epoch": 3.503529411764706, + "grad_norm": 0.08371615708331441, + "learning_rate": 0.00014693168165816995, + "loss": 0.0339, + "step": 7445 + }, + { + "epoch": 3.504, + "grad_norm": 0.06386678680595757, + "learning_rate": 0.00014688785478530697, + "loss": 0.0178, + "step": 7446 + }, + { + "epoch": 3.5044705882352942, + "grad_norm": 1.0090826762547583e-05, + "learning_rate": 0.00014684402817823647, + "loss": 0.0, + "step": 7447 + }, + { + "epoch": 3.504941176470588, + "grad_norm": 0.10162379170741363, + "learning_rate": 0.00014680020184070148, + "loss": 0.0399, + "step": 7448 + }, + { + "epoch": 3.5054117647058822, + "grad_norm": 0.044966132895082646, + "learning_rate": 0.0001467563757764449, + "loss": 0.0096, + "step": 7449 + }, + { + "epoch": 3.5058823529411764, + "grad_norm": 0.042276318392269446, + "learning_rate": 0.00014671254998920976, + "loss": 0.0078, + "step": 7450 + }, + { + "epoch": 3.5063529411764707, + "grad_norm": 0.12947718601818284, + "learning_rate": 0.000146668724482739, + "loss": 0.0667, + "step": 7451 + }, + { + "epoch": 3.506823529411765, + "grad_norm": 0.1803411886968798, + "learning_rate": 0.0001466248992607755, + "loss": 0.1007, + "step": 7452 + }, + { + "epoch": 3.5072941176470587, + "grad_norm": 0.08909173035922971, + "learning_rate": 0.00014658107432706212, + "loss": 0.0296, + "step": 7453 + }, + { + "epoch": 3.507764705882353, + "grad_norm": 0.09351096888391959, + "learning_rate": 0.0001465372496853418, + "loss": 0.0498, + "step": 7454 + }, + { + "epoch": 3.508235294117647, + "grad_norm": 0.06470058446960857, + "learning_rate": 0.00014649342533935733, + "loss": 0.017, + "step": 7455 + }, + { + "epoch": 3.5087058823529413, + "grad_norm": 0.1368954626192703, + "learning_rate": 0.00014644960129285152, + "loss": 0.0262, + "step": 7456 + }, + { + "epoch": 3.509176470588235, + "grad_norm": 0.04117455043673468, + "learning_rate": 0.0001464057775495672, + "loss": 0.0032, + "step": 7457 + }, + { + "epoch": 3.5096470588235293, + "grad_norm": 0.11997064895892287, + "learning_rate": 0.0001463619541132471, + "loss": 0.0545, + "step": 7458 + }, + { + "epoch": 3.5101176470588236, + "grad_norm": 1.3302263670926125e-06, + "learning_rate": 0.00014631813098763392, + "loss": 0.0, + "step": 7459 + }, + { + "epoch": 3.510588235294118, + "grad_norm": 0.15689768504364954, + "learning_rate": 0.00014627430817647048, + "loss": 0.0739, + "step": 7460 + }, + { + "epoch": 3.511058823529412, + "grad_norm": 0.0786940113890863, + "learning_rate": 0.00014623048568349937, + "loss": 0.0225, + "step": 7461 + }, + { + "epoch": 3.511529411764706, + "grad_norm": 0.04977538931555339, + "learning_rate": 0.0001461866635124633, + "loss": 0.0143, + "step": 7462 + }, + { + "epoch": 3.512, + "grad_norm": 0.0939036100837336, + "learning_rate": 0.0001461428416671048, + "loss": 0.0305, + "step": 7463 + }, + { + "epoch": 3.5124705882352942, + "grad_norm": 0.1706359358952785, + "learning_rate": 0.0001460990201511666, + "loss": 0.0828, + "step": 7464 + }, + { + "epoch": 3.512941176470588, + "grad_norm": 0.08303042743899004, + "learning_rate": 0.0001460551989683912, + "loss": 0.0364, + "step": 7465 + }, + { + "epoch": 3.5134117647058822, + "grad_norm": 0.055485551496112076, + "learning_rate": 0.00014601137812252112, + "loss": 0.0198, + "step": 7466 + }, + { + "epoch": 3.5138823529411765, + "grad_norm": 0.14793038825221197, + "learning_rate": 0.00014596755761729895, + "loss": 0.0817, + "step": 7467 + }, + { + "epoch": 3.5143529411764707, + "grad_norm": 0.15021801379516528, + "learning_rate": 0.00014592373745646715, + "loss": 0.0908, + "step": 7468 + }, + { + "epoch": 3.514823529411765, + "grad_norm": 2.0011607945778286e-05, + "learning_rate": 0.00014587991764376812, + "loss": 0.0, + "step": 7469 + }, + { + "epoch": 3.5152941176470587, + "grad_norm": 0.11050239568544772, + "learning_rate": 0.00014583609818294438, + "loss": 0.034, + "step": 7470 + }, + { + "epoch": 3.515764705882353, + "grad_norm": 8.21281971869387e-06, + "learning_rate": 0.00014579227907773827, + "loss": 0.0, + "step": 7471 + }, + { + "epoch": 3.516235294117647, + "grad_norm": 0.13688045790502998, + "learning_rate": 0.00014574846033189217, + "loss": 0.0412, + "step": 7472 + }, + { + "epoch": 3.516705882352941, + "grad_norm": 0.15416524551684196, + "learning_rate": 0.00014570464194914835, + "loss": 0.0878, + "step": 7473 + }, + { + "epoch": 3.517176470588235, + "grad_norm": 0.029319425128833433, + "learning_rate": 0.00014566082393324925, + "loss": 0.0035, + "step": 7474 + }, + { + "epoch": 3.5176470588235293, + "grad_norm": 0.061076802147451856, + "learning_rate": 0.00014561700628793706, + "loss": 0.021, + "step": 7475 + }, + { + "epoch": 3.5181176470588236, + "grad_norm": 0.06375824984321063, + "learning_rate": 0.000145573189016954, + "loss": 0.0189, + "step": 7476 + }, + { + "epoch": 3.518588235294118, + "grad_norm": 0.038373108713281236, + "learning_rate": 0.0001455293721240423, + "loss": 0.0105, + "step": 7477 + }, + { + "epoch": 3.519058823529412, + "grad_norm": 2.1706268330984615e-06, + "learning_rate": 0.0001454855556129442, + "loss": 0.0, + "step": 7478 + }, + { + "epoch": 3.519529411764706, + "grad_norm": 0.0530117869083454, + "learning_rate": 0.00014544173948740174, + "loss": 0.0169, + "step": 7479 + }, + { + "epoch": 3.52, + "grad_norm": 0.17334986991946041, + "learning_rate": 0.00014539792375115713, + "loss": 0.0356, + "step": 7480 + }, + { + "epoch": 3.5204705882352942, + "grad_norm": 0.06559479718304677, + "learning_rate": 0.0001453541084079524, + "loss": 0.012, + "step": 7481 + }, + { + "epoch": 3.520941176470588, + "grad_norm": 2.6907261431127555e-05, + "learning_rate": 0.0001453102934615296, + "loss": 0.0, + "step": 7482 + }, + { + "epoch": 3.5214117647058822, + "grad_norm": 0.16875987277326945, + "learning_rate": 0.0001452664789156307, + "loss": 0.0563, + "step": 7483 + }, + { + "epoch": 3.5218823529411765, + "grad_norm": 0.04374695434998468, + "learning_rate": 0.00014522266477399776, + "loss": 0.0081, + "step": 7484 + }, + { + "epoch": 3.5223529411764707, + "grad_norm": 0.04311116736763097, + "learning_rate": 0.00014517885104037268, + "loss": 0.0082, + "step": 7485 + }, + { + "epoch": 3.522823529411765, + "grad_norm": 0.19442588300417973, + "learning_rate": 0.00014513503771849734, + "loss": 0.0833, + "step": 7486 + }, + { + "epoch": 3.5232941176470587, + "grad_norm": 0.14249321059312875, + "learning_rate": 0.00014509122481211358, + "loss": 0.0655, + "step": 7487 + }, + { + "epoch": 3.523764705882353, + "grad_norm": 0.0680443933484471, + "learning_rate": 0.0001450474123249634, + "loss": 0.0235, + "step": 7488 + }, + { + "epoch": 3.524235294117647, + "grad_norm": 1.1702992197972407e-07, + "learning_rate": 0.00014500360026078847, + "loss": 0.0, + "step": 7489 + }, + { + "epoch": 3.524705882352941, + "grad_norm": 1.0337897352666784e-05, + "learning_rate": 0.00014495978862333063, + "loss": 0.0, + "step": 7490 + }, + { + "epoch": 3.525176470588235, + "grad_norm": 0.038609156203316466, + "learning_rate": 0.00014491597741633148, + "loss": 0.0087, + "step": 7491 + }, + { + "epoch": 3.5256470588235294, + "grad_norm": 0.08485084505761331, + "learning_rate": 0.00014487216664353287, + "loss": 0.0271, + "step": 7492 + }, + { + "epoch": 3.5261176470588236, + "grad_norm": 0.09059081458728668, + "learning_rate": 0.00014482835630867636, + "loss": 0.0275, + "step": 7493 + }, + { + "epoch": 3.526588235294118, + "grad_norm": 0.19302444780427164, + "learning_rate": 0.00014478454641550354, + "loss": 0.0393, + "step": 7494 + }, + { + "epoch": 3.527058823529412, + "grad_norm": 0.12308017399507235, + "learning_rate": 0.0001447407369677561, + "loss": 0.0224, + "step": 7495 + }, + { + "epoch": 3.527529411764706, + "grad_norm": 0.1786509806459583, + "learning_rate": 0.0001446969279691755, + "loss": 0.0305, + "step": 7496 + }, + { + "epoch": 3.528, + "grad_norm": 0.13442282331746128, + "learning_rate": 0.00014465311942350324, + "loss": 0.048, + "step": 7497 + }, + { + "epoch": 3.5284705882352942, + "grad_norm": 0.17373912617373344, + "learning_rate": 0.00014460931133448084, + "loss": 0.1106, + "step": 7498 + }, + { + "epoch": 3.528941176470588, + "grad_norm": 0.1265504259780788, + "learning_rate": 0.0001445655037058497, + "loss": 0.0809, + "step": 7499 + }, + { + "epoch": 3.5294117647058822, + "grad_norm": 0.1011817148067988, + "learning_rate": 0.00014452169654135115, + "loss": 0.0379, + "step": 7500 + }, + { + "epoch": 3.5298823529411765, + "grad_norm": 0.11154065710410895, + "learning_rate": 0.0001444778898447266, + "loss": 0.0602, + "step": 7501 + }, + { + "epoch": 3.5303529411764707, + "grad_norm": 0.12546183204252656, + "learning_rate": 0.00014443408361971739, + "loss": 0.0418, + "step": 7502 + }, + { + "epoch": 3.530823529411765, + "grad_norm": 1.783048027023863e-08, + "learning_rate": 0.0001443902778700647, + "loss": 0.0, + "step": 7503 + }, + { + "epoch": 3.5312941176470587, + "grad_norm": 0.10511580120814334, + "learning_rate": 0.0001443464725995098, + "loss": 0.0515, + "step": 7504 + }, + { + "epoch": 3.531764705882353, + "grad_norm": 0.177151745392796, + "learning_rate": 0.0001443026678117939, + "loss": 0.0848, + "step": 7505 + }, + { + "epoch": 3.532235294117647, + "grad_norm": 2.931016676690296e-06, + "learning_rate": 0.00014425886351065808, + "loss": 0.0, + "step": 7506 + }, + { + "epoch": 3.532705882352941, + "grad_norm": 0.11961494994943918, + "learning_rate": 0.00014421505969984346, + "loss": 0.05, + "step": 7507 + }, + { + "epoch": 3.533176470588235, + "grad_norm": 0.07767120210618082, + "learning_rate": 0.00014417125638309115, + "loss": 0.0313, + "step": 7508 + }, + { + "epoch": 3.5336470588235294, + "grad_norm": 0.0820872844701177, + "learning_rate": 0.00014412745356414213, + "loss": 0.0405, + "step": 7509 + }, + { + "epoch": 3.5341176470588236, + "grad_norm": 0.12051432390854754, + "learning_rate": 0.00014408365124673737, + "loss": 0.0622, + "step": 7510 + }, + { + "epoch": 3.534588235294118, + "grad_norm": 0.036722043839812805, + "learning_rate": 0.00014403984943461783, + "loss": 0.0057, + "step": 7511 + }, + { + "epoch": 3.535058823529412, + "grad_norm": 6.905022385035404e-07, + "learning_rate": 0.0001439960481315244, + "loss": 0.0, + "step": 7512 + }, + { + "epoch": 3.535529411764706, + "grad_norm": 0.25837528262348947, + "learning_rate": 0.00014395224734119788, + "loss": 0.0669, + "step": 7513 + }, + { + "epoch": 3.536, + "grad_norm": 0.08123223038303276, + "learning_rate": 0.00014390844706737907, + "loss": 0.0203, + "step": 7514 + }, + { + "epoch": 3.5364705882352943, + "grad_norm": 0.06392765013318212, + "learning_rate": 0.00014386464731380881, + "loss": 0.0282, + "step": 7515 + }, + { + "epoch": 3.536941176470588, + "grad_norm": 0.06572843559736812, + "learning_rate": 0.00014382084808422775, + "loss": 0.0154, + "step": 7516 + }, + { + "epoch": 3.5374117647058823, + "grad_norm": 0.12952220151530458, + "learning_rate": 0.00014377704938237654, + "loss": 0.0695, + "step": 7517 + }, + { + "epoch": 3.5378823529411765, + "grad_norm": 8.336298626645741e-07, + "learning_rate": 0.00014373325121199587, + "loss": 0.0, + "step": 7518 + }, + { + "epoch": 3.5383529411764707, + "grad_norm": 0.22028405084818264, + "learning_rate": 0.00014368945357682628, + "loss": 0.1143, + "step": 7519 + }, + { + "epoch": 3.538823529411765, + "grad_norm": 0.16423996988081754, + "learning_rate": 0.0001436456564806083, + "loss": 0.0435, + "step": 7520 + }, + { + "epoch": 3.5392941176470587, + "grad_norm": 1.2858134625067277e-07, + "learning_rate": 0.00014360185992708243, + "loss": 0.0, + "step": 7521 + }, + { + "epoch": 3.539764705882353, + "grad_norm": 9.911039288328462e-07, + "learning_rate": 0.00014355806391998913, + "loss": 0.0, + "step": 7522 + }, + { + "epoch": 3.540235294117647, + "grad_norm": 0.0880275255107331, + "learning_rate": 0.00014351426846306875, + "loss": 0.0424, + "step": 7523 + }, + { + "epoch": 3.540705882352941, + "grad_norm": 5.044840956955242e-07, + "learning_rate": 0.00014347047356006161, + "loss": 0.0, + "step": 7524 + }, + { + "epoch": 3.541176470588235, + "grad_norm": 3.3293005122123027e-07, + "learning_rate": 0.00014342667921470812, + "loss": 0.0, + "step": 7525 + }, + { + "epoch": 3.5416470588235294, + "grad_norm": 0.1286149887318933, + "learning_rate": 0.00014338288543074845, + "loss": 0.0649, + "step": 7526 + }, + { + "epoch": 3.5421176470588236, + "grad_norm": 0.10430135584450442, + "learning_rate": 0.00014333909221192278, + "loss": 0.0454, + "step": 7527 + }, + { + "epoch": 3.542588235294118, + "grad_norm": 0.1839949112878486, + "learning_rate": 0.00014329529956197135, + "loss": 0.063, + "step": 7528 + }, + { + "epoch": 3.543058823529412, + "grad_norm": 0.0918402339370639, + "learning_rate": 0.00014325150748463423, + "loss": 0.0344, + "step": 7529 + }, + { + "epoch": 3.543529411764706, + "grad_norm": 0.041807858761646215, + "learning_rate": 0.00014320771598365142, + "loss": 0.011, + "step": 7530 + }, + { + "epoch": 3.544, + "grad_norm": 0.09324750064991699, + "learning_rate": 0.00014316392506276302, + "loss": 0.0407, + "step": 7531 + }, + { + "epoch": 3.5444705882352943, + "grad_norm": 3.083401880806823e-07, + "learning_rate": 0.0001431201347257089, + "loss": 0.0, + "step": 7532 + }, + { + "epoch": 3.544941176470588, + "grad_norm": 4.467470180771295e-06, + "learning_rate": 0.00014307634497622907, + "loss": 0.0, + "step": 7533 + }, + { + "epoch": 3.5454117647058823, + "grad_norm": 0.09192901282189135, + "learning_rate": 0.00014303255581806336, + "loss": 0.0414, + "step": 7534 + }, + { + "epoch": 3.5458823529411765, + "grad_norm": 0.11540583472914295, + "learning_rate": 0.0001429887672549515, + "loss": 0.0496, + "step": 7535 + }, + { + "epoch": 3.5463529411764707, + "grad_norm": 0.06208223273111559, + "learning_rate": 0.00014294497929063336, + "loss": 0.0153, + "step": 7536 + }, + { + "epoch": 3.546823529411765, + "grad_norm": 0.028223966906245512, + "learning_rate": 0.0001429011919288486, + "loss": 0.005, + "step": 7537 + }, + { + "epoch": 3.5472941176470587, + "grad_norm": 0.14665741082218528, + "learning_rate": 0.0001428574051733368, + "loss": 0.0708, + "step": 7538 + }, + { + "epoch": 3.547764705882353, + "grad_norm": 0.061265344802827804, + "learning_rate": 0.0001428136190278377, + "loss": 0.0178, + "step": 7539 + }, + { + "epoch": 3.548235294117647, + "grad_norm": 0.1033712015804245, + "learning_rate": 0.00014276983349609075, + "loss": 0.047, + "step": 7540 + }, + { + "epoch": 3.548705882352941, + "grad_norm": 0.10043588698670264, + "learning_rate": 0.0001427260485818355, + "loss": 0.0286, + "step": 7541 + }, + { + "epoch": 3.549176470588235, + "grad_norm": 0.03893623762990525, + "learning_rate": 0.00014268226428881136, + "loss": 0.009, + "step": 7542 + }, + { + "epoch": 3.5496470588235294, + "grad_norm": 0.11265928904949063, + "learning_rate": 0.00014263848062075776, + "loss": 0.0543, + "step": 7543 + }, + { + "epoch": 3.5501176470588236, + "grad_norm": 0.0598181585512588, + "learning_rate": 0.00014259469758141403, + "loss": 0.0209, + "step": 7544 + }, + { + "epoch": 3.550588235294118, + "grad_norm": 6.484084744151994e-06, + "learning_rate": 0.00014255091517451941, + "loss": 0.0, + "step": 7545 + }, + { + "epoch": 3.5510588235294116, + "grad_norm": 0.10146818796923444, + "learning_rate": 0.0001425071334038132, + "loss": 0.0412, + "step": 7546 + }, + { + "epoch": 3.551529411764706, + "grad_norm": 0.08454990998593291, + "learning_rate": 0.00014246335227303456, + "loss": 0.0341, + "step": 7547 + }, + { + "epoch": 3.552, + "grad_norm": 0.11567956036851067, + "learning_rate": 0.00014241957178592258, + "loss": 0.0621, + "step": 7548 + }, + { + "epoch": 3.5524705882352943, + "grad_norm": 0.030339989877420682, + "learning_rate": 0.00014237579194621637, + "loss": 0.0049, + "step": 7549 + }, + { + "epoch": 3.552941176470588, + "grad_norm": 2.7990487640674742e-05, + "learning_rate": 0.00014233201275765494, + "loss": 0.0, + "step": 7550 + }, + { + "epoch": 3.5534117647058823, + "grad_norm": 0.037236370588046706, + "learning_rate": 0.0001422882342239772, + "loss": 0.0078, + "step": 7551 + }, + { + "epoch": 3.5538823529411765, + "grad_norm": 1.1923834267295092e-07, + "learning_rate": 0.00014224445634892206, + "loss": 0.0, + "step": 7552 + }, + { + "epoch": 3.5543529411764707, + "grad_norm": 0.03599780708290995, + "learning_rate": 0.00014220067913622844, + "loss": 0.004, + "step": 7553 + }, + { + "epoch": 3.554823529411765, + "grad_norm": 0.056353295346979575, + "learning_rate": 0.00014215690258963508, + "loss": 0.0199, + "step": 7554 + }, + { + "epoch": 3.5552941176470587, + "grad_norm": 0.07359090622413217, + "learning_rate": 0.00014211312671288064, + "loss": 0.0245, + "step": 7555 + }, + { + "epoch": 3.555764705882353, + "grad_norm": 0.06106511721570771, + "learning_rate": 0.00014206935150970392, + "loss": 0.02, + "step": 7556 + }, + { + "epoch": 3.556235294117647, + "grad_norm": 0.23755261639412523, + "learning_rate": 0.00014202557698384348, + "loss": 0.0735, + "step": 7557 + }, + { + "epoch": 3.556705882352941, + "grad_norm": 0.08654625550480433, + "learning_rate": 0.00014198180313903784, + "loss": 0.0269, + "step": 7558 + }, + { + "epoch": 3.557176470588235, + "grad_norm": 0.07854063112569781, + "learning_rate": 0.0001419380299790256, + "loss": 0.012, + "step": 7559 + }, + { + "epoch": 3.5576470588235294, + "grad_norm": 3.6193320768850115e-07, + "learning_rate": 0.00014189425750754512, + "loss": 0.0, + "step": 7560 + }, + { + "epoch": 3.5581176470588236, + "grad_norm": 0.07688499672398802, + "learning_rate": 0.00014185048572833478, + "loss": 0.0319, + "step": 7561 + }, + { + "epoch": 3.558588235294118, + "grad_norm": 0.1270130443138084, + "learning_rate": 0.00014180671464513296, + "loss": 0.0599, + "step": 7562 + }, + { + "epoch": 3.5590588235294116, + "grad_norm": 2.3356234673175965e-07, + "learning_rate": 0.0001417629442616779, + "loss": 0.0, + "step": 7563 + }, + { + "epoch": 3.559529411764706, + "grad_norm": 0.12144955537589965, + "learning_rate": 0.00014171917458170782, + "loss": 0.0595, + "step": 7564 + }, + { + "epoch": 3.56, + "grad_norm": 0.10557822404236684, + "learning_rate": 0.0001416754056089608, + "loss": 0.0464, + "step": 7565 + }, + { + "epoch": 3.5604705882352943, + "grad_norm": 0.05114485079689261, + "learning_rate": 0.000141631637347175, + "loss": 0.0104, + "step": 7566 + }, + { + "epoch": 3.560941176470588, + "grad_norm": 0.1278833150450966, + "learning_rate": 0.0001415878698000884, + "loss": 0.0652, + "step": 7567 + }, + { + "epoch": 3.5614117647058823, + "grad_norm": 0.13660912662803895, + "learning_rate": 0.00014154410297143898, + "loss": 0.0753, + "step": 7568 + }, + { + "epoch": 3.5618823529411765, + "grad_norm": 0.0525951495394156, + "learning_rate": 0.00014150033686496469, + "loss": 0.0216, + "step": 7569 + }, + { + "epoch": 3.5623529411764707, + "grad_norm": 0.06537495118194486, + "learning_rate": 0.0001414565714844033, + "loss": 0.0299, + "step": 7570 + }, + { + "epoch": 3.562823529411765, + "grad_norm": 0.1908450363637293, + "learning_rate": 0.00014141280683349253, + "loss": 0.0683, + "step": 7571 + }, + { + "epoch": 3.5632941176470587, + "grad_norm": 0.11179443937391798, + "learning_rate": 0.00014136904291597028, + "loss": 0.0518, + "step": 7572 + }, + { + "epoch": 3.563764705882353, + "grad_norm": 0.30167709677214677, + "learning_rate": 0.00014132527973557407, + "loss": 0.0912, + "step": 7573 + }, + { + "epoch": 3.564235294117647, + "grad_norm": 0.10072235756836088, + "learning_rate": 0.00014128151729604152, + "loss": 0.0275, + "step": 7574 + }, + { + "epoch": 3.564705882352941, + "grad_norm": 0.08139188219137998, + "learning_rate": 0.00014123775560111009, + "loss": 0.0289, + "step": 7575 + }, + { + "epoch": 3.565176470588235, + "grad_norm": 0.061574931824077105, + "learning_rate": 0.00014119399465451738, + "loss": 0.0192, + "step": 7576 + }, + { + "epoch": 3.5656470588235294, + "grad_norm": 0.11192044277997634, + "learning_rate": 0.0001411502344600007, + "loss": 0.0434, + "step": 7577 + }, + { + "epoch": 3.5661176470588236, + "grad_norm": 0.17256112605451937, + "learning_rate": 0.00014110647502129742, + "loss": 0.0843, + "step": 7578 + }, + { + "epoch": 3.566588235294118, + "grad_norm": 0.11841518503862523, + "learning_rate": 0.00014106271634214477, + "loss": 0.0704, + "step": 7579 + }, + { + "epoch": 3.5670588235294116, + "grad_norm": 0.09847015681896638, + "learning_rate": 0.00014101895842628001, + "loss": 0.0485, + "step": 7580 + }, + { + "epoch": 3.567529411764706, + "grad_norm": 3.874425200913387e-05, + "learning_rate": 0.00014097520127744024, + "loss": 0.0, + "step": 7581 + }, + { + "epoch": 3.568, + "grad_norm": 0.11989680746875232, + "learning_rate": 0.00014093144489936259, + "loss": 0.0393, + "step": 7582 + }, + { + "epoch": 3.5684705882352943, + "grad_norm": 0.05769715890213197, + "learning_rate": 0.00014088768929578397, + "loss": 0.0112, + "step": 7583 + }, + { + "epoch": 3.568941176470588, + "grad_norm": 0.0947360891919032, + "learning_rate": 0.00014084393447044138, + "loss": 0.0348, + "step": 7584 + }, + { + "epoch": 3.5694117647058823, + "grad_norm": 0.18697995472426693, + "learning_rate": 0.00014080018042707173, + "loss": 0.039, + "step": 7585 + }, + { + "epoch": 3.5698823529411765, + "grad_norm": 0.10414556075559218, + "learning_rate": 0.00014075642716941176, + "loss": 0.0203, + "step": 7586 + }, + { + "epoch": 3.5703529411764707, + "grad_norm": 0.08285194610820033, + "learning_rate": 0.0001407126747011983, + "loss": 0.025, + "step": 7587 + }, + { + "epoch": 3.570823529411765, + "grad_norm": 0.09692661128293273, + "learning_rate": 0.00014066892302616794, + "loss": 0.0398, + "step": 7588 + }, + { + "epoch": 3.5712941176470587, + "grad_norm": 1.388938676377408e-06, + "learning_rate": 0.0001406251721480573, + "loss": 0.0, + "step": 7589 + }, + { + "epoch": 3.571764705882353, + "grad_norm": 0.09146112429452624, + "learning_rate": 0.000140581422070603, + "loss": 0.0377, + "step": 7590 + }, + { + "epoch": 3.572235294117647, + "grad_norm": 0.10709794986026577, + "learning_rate": 0.0001405376727975414, + "loss": 0.0429, + "step": 7591 + }, + { + "epoch": 3.572705882352941, + "grad_norm": 0.08096825457165997, + "learning_rate": 0.000140493924332609, + "loss": 0.0302, + "step": 7592 + }, + { + "epoch": 3.573176470588235, + "grad_norm": 0.17657943192429154, + "learning_rate": 0.00014045017667954203, + "loss": 0.0979, + "step": 7593 + }, + { + "epoch": 3.5736470588235294, + "grad_norm": 1.009641919077098e-05, + "learning_rate": 0.00014040642984207687, + "loss": 0.0, + "step": 7594 + }, + { + "epoch": 3.5741176470588236, + "grad_norm": 0.033638752886986256, + "learning_rate": 0.00014036268382394964, + "loss": 0.0077, + "step": 7595 + }, + { + "epoch": 3.574588235294118, + "grad_norm": 0.09384951308385833, + "learning_rate": 0.00014031893862889646, + "loss": 0.0336, + "step": 7596 + }, + { + "epoch": 3.5750588235294116, + "grad_norm": 1.433412335075257e-06, + "learning_rate": 0.00014027519426065344, + "loss": 0.0, + "step": 7597 + }, + { + "epoch": 3.575529411764706, + "grad_norm": 0.10526820238831591, + "learning_rate": 0.00014023145072295654, + "loss": 0.0323, + "step": 7598 + }, + { + "epoch": 3.576, + "grad_norm": 6.38679625609216e-05, + "learning_rate": 0.0001401877080195416, + "loss": 0.0, + "step": 7599 + }, + { + "epoch": 3.576470588235294, + "grad_norm": 0.17309628401779084, + "learning_rate": 0.00014014396615414458, + "loss": 0.0886, + "step": 7600 + }, + { + "epoch": 3.576941176470588, + "grad_norm": 0.09784310862565146, + "learning_rate": 0.0001401002251305012, + "loss": 0.0497, + "step": 7601 + }, + { + "epoch": 3.5774117647058823, + "grad_norm": 0.026513452336143915, + "learning_rate": 0.0001400564849523471, + "loss": 0.0029, + "step": 7602 + }, + { + "epoch": 3.5778823529411765, + "grad_norm": 1.1457802212143451e-06, + "learning_rate": 0.00014001274562341798, + "loss": 0.0, + "step": 7603 + }, + { + "epoch": 3.5783529411764707, + "grad_norm": 0.06775599484873832, + "learning_rate": 0.0001399690071474494, + "loss": 0.0379, + "step": 7604 + }, + { + "epoch": 3.578823529411765, + "grad_norm": 0.3438402388442661, + "learning_rate": 0.0001399252695281768, + "loss": 0.0913, + "step": 7605 + }, + { + "epoch": 3.5792941176470587, + "grad_norm": 8.46286913955956e-07, + "learning_rate": 0.00013988153276933555, + "loss": 0.0, + "step": 7606 + }, + { + "epoch": 3.579764705882353, + "grad_norm": 3.836798419504578e-05, + "learning_rate": 0.0001398377968746611, + "loss": 0.0, + "step": 7607 + }, + { + "epoch": 3.580235294117647, + "grad_norm": 0.035610982441586184, + "learning_rate": 0.0001397940618478886, + "loss": 0.0061, + "step": 7608 + }, + { + "epoch": 3.580705882352941, + "grad_norm": 0.050537473521812044, + "learning_rate": 0.00013975032769275326, + "loss": 0.0131, + "step": 7609 + }, + { + "epoch": 3.581176470588235, + "grad_norm": 0.10029159425115187, + "learning_rate": 0.00013970659441299026, + "loss": 0.0404, + "step": 7610 + }, + { + "epoch": 3.5816470588235294, + "grad_norm": 0.04752043400432198, + "learning_rate": 0.00013966286201233457, + "loss": 0.0134, + "step": 7611 + }, + { + "epoch": 3.5821176470588236, + "grad_norm": 0.18884837387896866, + "learning_rate": 0.00013961913049452115, + "loss": 0.0778, + "step": 7612 + }, + { + "epoch": 3.582588235294118, + "grad_norm": 5.742294611367407e-05, + "learning_rate": 0.00013957539986328492, + "loss": 0.0, + "step": 7613 + }, + { + "epoch": 3.5830588235294116, + "grad_norm": 0.09185240279804083, + "learning_rate": 0.00013953167012236068, + "loss": 0.0362, + "step": 7614 + }, + { + "epoch": 3.583529411764706, + "grad_norm": 0.0999840697319665, + "learning_rate": 0.00013948794127548316, + "loss": 0.04, + "step": 7615 + }, + { + "epoch": 3.584, + "grad_norm": 0.18537596131827908, + "learning_rate": 0.00013944421332638698, + "loss": 0.0842, + "step": 7616 + }, + { + "epoch": 3.584470588235294, + "grad_norm": 0.07293208635276907, + "learning_rate": 0.0001394004862788068, + "loss": 0.0254, + "step": 7617 + }, + { + "epoch": 3.584941176470588, + "grad_norm": 0.09383514711610838, + "learning_rate": 0.00013935676013647708, + "loss": 0.0206, + "step": 7618 + }, + { + "epoch": 3.5854117647058823, + "grad_norm": 7.859230196185539e-07, + "learning_rate": 0.0001393130349031322, + "loss": 0.0, + "step": 7619 + }, + { + "epoch": 3.5858823529411765, + "grad_norm": 0.060505406946467055, + "learning_rate": 0.00013926931058250657, + "loss": 0.0109, + "step": 7620 + }, + { + "epoch": 3.5863529411764707, + "grad_norm": 8.341931402241555e-07, + "learning_rate": 0.00013922558717833445, + "loss": 0.0, + "step": 7621 + }, + { + "epoch": 3.586823529411765, + "grad_norm": 0.13523783587594793, + "learning_rate": 0.00013918186469435005, + "loss": 0.0679, + "step": 7622 + }, + { + "epoch": 3.5872941176470587, + "grad_norm": 0.01443145804902979, + "learning_rate": 0.00013913814313428748, + "loss": 0.001, + "step": 7623 + }, + { + "epoch": 3.587764705882353, + "grad_norm": 0.0516239849376093, + "learning_rate": 0.00013909442250188074, + "loss": 0.0107, + "step": 7624 + }, + { + "epoch": 3.588235294117647, + "grad_norm": 0.08062701471840673, + "learning_rate": 0.00013905070280086386, + "loss": 0.0358, + "step": 7625 + }, + { + "epoch": 3.588705882352941, + "grad_norm": 0.08274555885770148, + "learning_rate": 0.00013900698403497062, + "loss": 0.0389, + "step": 7626 + }, + { + "epoch": 3.589176470588235, + "grad_norm": 0.020617278249324255, + "learning_rate": 0.00013896326620793488, + "loss": 0.0028, + "step": 7627 + }, + { + "epoch": 3.5896470588235294, + "grad_norm": 0.158537109545575, + "learning_rate": 0.00013891954932349033, + "loss": 0.0837, + "step": 7628 + }, + { + "epoch": 3.5901176470588236, + "grad_norm": 0.09927986484006038, + "learning_rate": 0.00013887583338537064, + "loss": 0.0548, + "step": 7629 + }, + { + "epoch": 3.590588235294118, + "grad_norm": 1.800650272524961e-06, + "learning_rate": 0.0001388321183973093, + "loss": 0.0, + "step": 7630 + }, + { + "epoch": 3.5910588235294116, + "grad_norm": 0.09420361905488454, + "learning_rate": 0.0001387884043630399, + "loss": 0.0356, + "step": 7631 + }, + { + "epoch": 3.591529411764706, + "grad_norm": 0.1540014592318495, + "learning_rate": 0.0001387446912862957, + "loss": 0.035, + "step": 7632 + }, + { + "epoch": 3.592, + "grad_norm": 0.10044607551073734, + "learning_rate": 0.0001387009791708101, + "loss": 0.0545, + "step": 7633 + }, + { + "epoch": 3.592470588235294, + "grad_norm": 0.06587097711632839, + "learning_rate": 0.00013865726802031627, + "loss": 0.0256, + "step": 7634 + }, + { + "epoch": 3.592941176470588, + "grad_norm": 0.12943415047886012, + "learning_rate": 0.0001386135578385474, + "loss": 0.069, + "step": 7635 + }, + { + "epoch": 3.5934117647058823, + "grad_norm": 0.0002780172553670581, + "learning_rate": 0.00013856984862923655, + "loss": 0.0, + "step": 7636 + }, + { + "epoch": 3.5938823529411765, + "grad_norm": 0.06435895267177394, + "learning_rate": 0.00013852614039611667, + "loss": 0.0213, + "step": 7637 + }, + { + "epoch": 3.5943529411764708, + "grad_norm": 0.05455718174484936, + "learning_rate": 0.0001384824331429207, + "loss": 0.0161, + "step": 7638 + }, + { + "epoch": 3.594823529411765, + "grad_norm": 0.09146992068284257, + "learning_rate": 0.00013843872687338142, + "loss": 0.0496, + "step": 7639 + }, + { + "epoch": 3.5952941176470588, + "grad_norm": 1.690256160028423e-05, + "learning_rate": 0.00013839502159123158, + "loss": 0.0, + "step": 7640 + }, + { + "epoch": 3.595764705882353, + "grad_norm": 0.09721889242545828, + "learning_rate": 0.00013835131730020382, + "loss": 0.0452, + "step": 7641 + }, + { + "epoch": 3.596235294117647, + "grad_norm": 0.1429827622109635, + "learning_rate": 0.0001383076140040307, + "loss": 0.0683, + "step": 7642 + }, + { + "epoch": 3.596705882352941, + "grad_norm": 2.602330862607128e-06, + "learning_rate": 0.0001382639117064447, + "loss": 0.0, + "step": 7643 + }, + { + "epoch": 3.597176470588235, + "grad_norm": 0.07970396425825958, + "learning_rate": 0.0001382202104111782, + "loss": 0.0268, + "step": 7644 + }, + { + "epoch": 3.5976470588235294, + "grad_norm": 0.08961656574789925, + "learning_rate": 0.0001381765101219635, + "loss": 0.0131, + "step": 7645 + }, + { + "epoch": 3.5981176470588236, + "grad_norm": 0.12849559839317223, + "learning_rate": 0.00013813281084253286, + "loss": 0.04, + "step": 7646 + }, + { + "epoch": 3.598588235294118, + "grad_norm": 1.7524703354007544e-05, + "learning_rate": 0.00013808911257661834, + "loss": 0.0, + "step": 7647 + }, + { + "epoch": 3.5990588235294116, + "grad_norm": 0.05474266305254985, + "learning_rate": 0.00013804541532795204, + "loss": 0.0158, + "step": 7648 + }, + { + "epoch": 3.599529411764706, + "grad_norm": 1.6823936614798825e-06, + "learning_rate": 0.00013800171910026594, + "loss": 0.0, + "step": 7649 + }, + { + "epoch": 3.6, + "grad_norm": 0.043948698893734914, + "learning_rate": 0.00013795802389729184, + "loss": 0.0074, + "step": 7650 + }, + { + "epoch": 3.600470588235294, + "grad_norm": 0.0939912978804121, + "learning_rate": 0.0001379143297227616, + "loss": 0.0575, + "step": 7651 + }, + { + "epoch": 3.600941176470588, + "grad_norm": 4.061397739688039e-06, + "learning_rate": 0.00013787063658040686, + "loss": 0.0, + "step": 7652 + }, + { + "epoch": 3.6014117647058823, + "grad_norm": 0.0919231513170181, + "learning_rate": 0.00013782694447395926, + "loss": 0.0213, + "step": 7653 + }, + { + "epoch": 3.6018823529411765, + "grad_norm": 0.0983620974696722, + "learning_rate": 0.0001377832534071503, + "loss": 0.0325, + "step": 7654 + }, + { + "epoch": 3.6023529411764708, + "grad_norm": 2.5669916275992842e-06, + "learning_rate": 0.0001377395633837114, + "loss": 0.0, + "step": 7655 + }, + { + "epoch": 3.602823529411765, + "grad_norm": 0.10514155425601186, + "learning_rate": 0.000137695874407374, + "loss": 0.0283, + "step": 7656 + }, + { + "epoch": 3.6032941176470588, + "grad_norm": 3.821984556647021e-06, + "learning_rate": 0.0001376521864818692, + "loss": 0.0, + "step": 7657 + }, + { + "epoch": 3.603764705882353, + "grad_norm": 0.07921924892452396, + "learning_rate": 0.00013760849961092826, + "loss": 0.009, + "step": 7658 + }, + { + "epoch": 3.604235294117647, + "grad_norm": 0.1020627615147994, + "learning_rate": 0.00013756481379828225, + "loss": 0.034, + "step": 7659 + }, + { + "epoch": 3.604705882352941, + "grad_norm": 1.5615171087745993e-05, + "learning_rate": 0.00013752112904766208, + "loss": 0.0, + "step": 7660 + }, + { + "epoch": 3.605176470588235, + "grad_norm": 3.6170717484814663e-07, + "learning_rate": 0.00013747744536279878, + "loss": 0.0, + "step": 7661 + }, + { + "epoch": 3.6056470588235294, + "grad_norm": 0.05290702277631353, + "learning_rate": 0.00013743376274742304, + "loss": 0.0163, + "step": 7662 + }, + { + "epoch": 3.6061176470588236, + "grad_norm": 0.1255196292102114, + "learning_rate": 0.00013739008120526559, + "loss": 0.029, + "step": 7663 + }, + { + "epoch": 3.606588235294118, + "grad_norm": 4.698915448164272e-06, + "learning_rate": 0.00013734640074005702, + "loss": 0.0, + "step": 7664 + }, + { + "epoch": 3.6070588235294117, + "grad_norm": 2.8455970611974893e-06, + "learning_rate": 0.0001373027213555279, + "loss": 0.0, + "step": 7665 + }, + { + "epoch": 3.607529411764706, + "grad_norm": 0.019216909294975902, + "learning_rate": 0.0001372590430554087, + "loss": 0.002, + "step": 7666 + }, + { + "epoch": 3.608, + "grad_norm": 4.3178411771176564e-06, + "learning_rate": 0.0001372153658434297, + "loss": 0.0, + "step": 7667 + }, + { + "epoch": 3.608470588235294, + "grad_norm": 0.1389505783509478, + "learning_rate": 0.00013717168972332116, + "loss": 0.0446, + "step": 7668 + }, + { + "epoch": 3.608941176470588, + "grad_norm": 0.04917159040056299, + "learning_rate": 0.00013712801469881324, + "loss": 0.0198, + "step": 7669 + }, + { + "epoch": 3.6094117647058823, + "grad_norm": 0.038109183075771706, + "learning_rate": 0.00013708434077363603, + "loss": 0.0029, + "step": 7670 + }, + { + "epoch": 3.6098823529411765, + "grad_norm": 0.06836749358110492, + "learning_rate": 0.00013704066795151946, + "loss": 0.02, + "step": 7671 + }, + { + "epoch": 3.6103529411764708, + "grad_norm": 0.12462341748214538, + "learning_rate": 0.00013699699623619338, + "loss": 0.0516, + "step": 7672 + }, + { + "epoch": 3.6108235294117645, + "grad_norm": 0.1435954040528868, + "learning_rate": 0.00013695332563138762, + "loss": 0.0614, + "step": 7673 + }, + { + "epoch": 3.6112941176470588, + "grad_norm": 0.13976800152667648, + "learning_rate": 0.00013690965614083185, + "loss": 0.057, + "step": 7674 + }, + { + "epoch": 3.611764705882353, + "grad_norm": 2.6466815365536452e-06, + "learning_rate": 0.00013686598776825563, + "loss": 0.0, + "step": 7675 + }, + { + "epoch": 3.612235294117647, + "grad_norm": 0.09434315813426003, + "learning_rate": 0.00013682232051738852, + "loss": 0.0451, + "step": 7676 + }, + { + "epoch": 3.612705882352941, + "grad_norm": 0.057129586223308465, + "learning_rate": 0.00013677865439195986, + "loss": 0.01, + "step": 7677 + }, + { + "epoch": 3.613176470588235, + "grad_norm": 0.006109870383170272, + "learning_rate": 0.00013673498939569892, + "loss": 0.0001, + "step": 7678 + }, + { + "epoch": 3.6136470588235294, + "grad_norm": 0.10758945005009692, + "learning_rate": 0.000136691325532335, + "loss": 0.0359, + "step": 7679 + }, + { + "epoch": 3.6141176470588237, + "grad_norm": 0.05334280292881827, + "learning_rate": 0.00013664766280559716, + "loss": 0.0173, + "step": 7680 + }, + { + "epoch": 3.614588235294118, + "grad_norm": 0.1245878273146271, + "learning_rate": 0.00013660400121921434, + "loss": 0.0665, + "step": 7681 + }, + { + "epoch": 3.6150588235294117, + "grad_norm": 0.08454462151711055, + "learning_rate": 0.00013656034077691562, + "loss": 0.0357, + "step": 7682 + }, + { + "epoch": 3.615529411764706, + "grad_norm": 0.10028572691872176, + "learning_rate": 0.00013651668148242967, + "loss": 0.033, + "step": 7683 + }, + { + "epoch": 3.616, + "grad_norm": 0.046891811153308406, + "learning_rate": 0.00013647302333948528, + "loss": 0.0127, + "step": 7684 + }, + { + "epoch": 3.616470588235294, + "grad_norm": 0.15444669501321795, + "learning_rate": 0.000136429366351811, + "loss": 0.0658, + "step": 7685 + }, + { + "epoch": 3.616941176470588, + "grad_norm": 4.260071271278364e-05, + "learning_rate": 0.00013638571052313543, + "loss": 0.0, + "step": 7686 + }, + { + "epoch": 3.6174117647058823, + "grad_norm": 0.04248338723589229, + "learning_rate": 0.00013634205585718698, + "loss": 0.0085, + "step": 7687 + }, + { + "epoch": 3.6178823529411765, + "grad_norm": 0.06419067684928967, + "learning_rate": 0.00013629840235769388, + "loss": 0.0115, + "step": 7688 + }, + { + "epoch": 3.6183529411764708, + "grad_norm": 0.014344465904138218, + "learning_rate": 0.0001362547500283845, + "loss": 0.0018, + "step": 7689 + }, + { + "epoch": 3.6188235294117646, + "grad_norm": 0.12146529044710593, + "learning_rate": 0.00013621109887298685, + "loss": 0.0727, + "step": 7690 + }, + { + "epoch": 3.6192941176470588, + "grad_norm": 0.13301207543152122, + "learning_rate": 0.00013616744889522898, + "loss": 0.0764, + "step": 7691 + }, + { + "epoch": 3.619764705882353, + "grad_norm": 0.11467717857114818, + "learning_rate": 0.00013612380009883886, + "loss": 0.0534, + "step": 7692 + }, + { + "epoch": 3.620235294117647, + "grad_norm": 0.13549778522959957, + "learning_rate": 0.00013608015248754425, + "loss": 0.0308, + "step": 7693 + }, + { + "epoch": 3.620705882352941, + "grad_norm": 1.484111155760182e-05, + "learning_rate": 0.00013603650606507293, + "loss": 0.0, + "step": 7694 + }, + { + "epoch": 3.621176470588235, + "grad_norm": 0.11705064250383398, + "learning_rate": 0.0001359928608351524, + "loss": 0.0573, + "step": 7695 + }, + { + "epoch": 3.6216470588235294, + "grad_norm": 0.10873122559249247, + "learning_rate": 0.00013594921680151032, + "loss": 0.0551, + "step": 7696 + }, + { + "epoch": 3.6221176470588237, + "grad_norm": 0.0002931653353350995, + "learning_rate": 0.00013590557396787406, + "loss": 0.0, + "step": 7697 + }, + { + "epoch": 3.622588235294118, + "grad_norm": 0.04119900102663766, + "learning_rate": 0.00013586193233797086, + "loss": 0.0079, + "step": 7698 + }, + { + "epoch": 3.6230588235294117, + "grad_norm": 0.09264832192757216, + "learning_rate": 0.00013581829191552802, + "loss": 0.0335, + "step": 7699 + }, + { + "epoch": 3.623529411764706, + "grad_norm": 0.09018392294112802, + "learning_rate": 0.00013577465270427262, + "loss": 0.0318, + "step": 7700 + }, + { + "epoch": 3.624, + "grad_norm": 0.04199465635228173, + "learning_rate": 0.00013573101470793161, + "loss": 0.009, + "step": 7701 + }, + { + "epoch": 3.624470588235294, + "grad_norm": 0.10511064895629478, + "learning_rate": 0.00013568737793023198, + "loss": 0.0441, + "step": 7702 + }, + { + "epoch": 3.624941176470588, + "grad_norm": 0.10744779760564935, + "learning_rate": 0.00013564374237490047, + "loss": 0.0468, + "step": 7703 + }, + { + "epoch": 3.6254117647058823, + "grad_norm": 0.05562757004477547, + "learning_rate": 0.0001356001080456638, + "loss": 0.0155, + "step": 7704 + }, + { + "epoch": 3.6258823529411766, + "grad_norm": 0.127712941905918, + "learning_rate": 0.00013555647494624846, + "loss": 0.04, + "step": 7705 + }, + { + "epoch": 3.626352941176471, + "grad_norm": 2.4018138560614775e-06, + "learning_rate": 0.00013551284308038105, + "loss": 0.0, + "step": 7706 + }, + { + "epoch": 3.6268235294117646, + "grad_norm": 0.12069475171993731, + "learning_rate": 0.00013546921245178792, + "loss": 0.049, + "step": 7707 + }, + { + "epoch": 3.627294117647059, + "grad_norm": 0.04130377768577765, + "learning_rate": 0.00013542558306419526, + "loss": 0.0087, + "step": 7708 + }, + { + "epoch": 3.627764705882353, + "grad_norm": 0.07519333043297682, + "learning_rate": 0.00013538195492132928, + "loss": 0.0244, + "step": 7709 + }, + { + "epoch": 3.6282352941176472, + "grad_norm": 0.10827595561402245, + "learning_rate": 0.00013533832802691613, + "loss": 0.0484, + "step": 7710 + }, + { + "epoch": 3.628705882352941, + "grad_norm": 9.514435129146325e-07, + "learning_rate": 0.00013529470238468166, + "loss": 0.0, + "step": 7711 + }, + { + "epoch": 3.6291764705882352, + "grad_norm": 7.454153242301736e-07, + "learning_rate": 0.00013525107799835172, + "loss": 0.0, + "step": 7712 + }, + { + "epoch": 3.6296470588235294, + "grad_norm": 0.027357017080258485, + "learning_rate": 0.00013520745487165207, + "loss": 0.0055, + "step": 7713 + }, + { + "epoch": 3.6301176470588237, + "grad_norm": 7.653921897186503e-07, + "learning_rate": 0.00013516383300830835, + "loss": 0.0, + "step": 7714 + }, + { + "epoch": 3.630588235294118, + "grad_norm": 0.12189647337488378, + "learning_rate": 0.00013512021241204604, + "loss": 0.0423, + "step": 7715 + }, + { + "epoch": 3.6310588235294117, + "grad_norm": 1.024714195785652e-06, + "learning_rate": 0.00013507659308659053, + "loss": 0.0, + "step": 7716 + }, + { + "epoch": 3.631529411764706, + "grad_norm": 1.4968174771001909e-06, + "learning_rate": 0.00013503297503566727, + "loss": 0.0, + "step": 7717 + }, + { + "epoch": 3.632, + "grad_norm": 0.10490472137588708, + "learning_rate": 0.00013498935826300132, + "loss": 0.0503, + "step": 7718 + }, + { + "epoch": 3.632470588235294, + "grad_norm": 0.10613475003207486, + "learning_rate": 0.00013494574277231772, + "loss": 0.0137, + "step": 7719 + }, + { + "epoch": 3.632941176470588, + "grad_norm": 0.05462441895441145, + "learning_rate": 0.00013490212856734162, + "loss": 0.0177, + "step": 7720 + }, + { + "epoch": 3.6334117647058823, + "grad_norm": 0.03547915422515945, + "learning_rate": 0.00013485851565179778, + "loss": 0.0046, + "step": 7721 + }, + { + "epoch": 3.6338823529411766, + "grad_norm": 0.013916278707501088, + "learning_rate": 0.00013481490402941098, + "loss": 0.0013, + "step": 7722 + }, + { + "epoch": 3.634352941176471, + "grad_norm": 0.11479026415903745, + "learning_rate": 0.00013477129370390582, + "loss": 0.0361, + "step": 7723 + }, + { + "epoch": 3.6348235294117646, + "grad_norm": 0.057640864355536836, + "learning_rate": 0.00013472768467900686, + "loss": 0.0136, + "step": 7724 + }, + { + "epoch": 3.635294117647059, + "grad_norm": 0.13239261887217918, + "learning_rate": 0.0001346840769584386, + "loss": 0.0565, + "step": 7725 + }, + { + "epoch": 3.635764705882353, + "grad_norm": 0.07237674629512417, + "learning_rate": 0.00013464047054592525, + "loss": 0.0291, + "step": 7726 + }, + { + "epoch": 3.636235294117647, + "grad_norm": 0.04218836485423579, + "learning_rate": 0.00013459686544519108, + "loss": 0.0052, + "step": 7727 + }, + { + "epoch": 3.636705882352941, + "grad_norm": 0.06840741840423675, + "learning_rate": 0.00013455326165996015, + "loss": 0.0165, + "step": 7728 + }, + { + "epoch": 3.6371764705882352, + "grad_norm": 0.03728330154558678, + "learning_rate": 0.0001345096591939564, + "loss": 0.0078, + "step": 7729 + }, + { + "epoch": 3.6376470588235295, + "grad_norm": 0.11822114398088719, + "learning_rate": 0.00013446605805090376, + "loss": 0.0683, + "step": 7730 + }, + { + "epoch": 3.6381176470588237, + "grad_norm": 0.10255146536869626, + "learning_rate": 0.000134422458234526, + "loss": 0.025, + "step": 7731 + }, + { + "epoch": 3.638588235294118, + "grad_norm": 0.07945679824951318, + "learning_rate": 0.00013437885974854665, + "loss": 0.0314, + "step": 7732 + }, + { + "epoch": 3.6390588235294117, + "grad_norm": 0.10672366016483058, + "learning_rate": 0.00013433526259668931, + "loss": 0.0454, + "step": 7733 + }, + { + "epoch": 3.639529411764706, + "grad_norm": 2.263141442509931e-06, + "learning_rate": 0.0001342916667826774, + "loss": 0.0, + "step": 7734 + }, + { + "epoch": 3.64, + "grad_norm": 0.11295947535809328, + "learning_rate": 0.0001342480723102342, + "loss": 0.0437, + "step": 7735 + }, + { + "epoch": 3.640470588235294, + "grad_norm": 0.14879946518472806, + "learning_rate": 0.00013420447918308285, + "loss": 0.0782, + "step": 7736 + }, + { + "epoch": 3.640941176470588, + "grad_norm": 0.0675430867633606, + "learning_rate": 0.00013416088740494649, + "loss": 0.0302, + "step": 7737 + }, + { + "epoch": 3.6414117647058823, + "grad_norm": 4.683935272431005e-06, + "learning_rate": 0.00013411729697954803, + "loss": 0.0, + "step": 7738 + }, + { + "epoch": 3.6418823529411766, + "grad_norm": 0.10289544597502027, + "learning_rate": 0.00013407370791061027, + "loss": 0.0509, + "step": 7739 + }, + { + "epoch": 3.642352941176471, + "grad_norm": 9.296826495391615e-06, + "learning_rate": 0.00013403012020185603, + "loss": 0.0, + "step": 7740 + }, + { + "epoch": 3.6428235294117646, + "grad_norm": 0.0006630887810151186, + "learning_rate": 0.00013398653385700784, + "loss": 0.0, + "step": 7741 + }, + { + "epoch": 3.643294117647059, + "grad_norm": 0.07040969631918195, + "learning_rate": 0.00013394294887978818, + "loss": 0.0196, + "step": 7742 + }, + { + "epoch": 3.643764705882353, + "grad_norm": 0.060141231102059114, + "learning_rate": 0.0001338993652739195, + "loss": 0.0172, + "step": 7743 + }, + { + "epoch": 3.644235294117647, + "grad_norm": 0.07858059611999516, + "learning_rate": 0.00013385578304312397, + "loss": 0.0269, + "step": 7744 + }, + { + "epoch": 3.644705882352941, + "grad_norm": 0.04776611280243866, + "learning_rate": 0.00013381220219112373, + "loss": 0.0115, + "step": 7745 + }, + { + "epoch": 3.6451764705882352, + "grad_norm": 0.014894522011745752, + "learning_rate": 0.00013376862272164084, + "loss": 0.0015, + "step": 7746 + }, + { + "epoch": 3.6456470588235295, + "grad_norm": 0.051505017436941275, + "learning_rate": 0.00013372504463839718, + "loss": 0.0133, + "step": 7747 + }, + { + "epoch": 3.6461176470588237, + "grad_norm": 0.04950992752160439, + "learning_rate": 0.00013368146794511457, + "loss": 0.0089, + "step": 7748 + }, + { + "epoch": 3.646588235294118, + "grad_norm": 0.08211003901053841, + "learning_rate": 0.0001336378926455146, + "loss": 0.0466, + "step": 7749 + }, + { + "epoch": 3.6470588235294117, + "grad_norm": 0.10511042676299236, + "learning_rate": 0.00013359431874331886, + "loss": 0.024, + "step": 7750 + }, + { + "epoch": 3.647529411764706, + "grad_norm": 0.10413144574724434, + "learning_rate": 0.0001335507462422488, + "loss": 0.0409, + "step": 7751 + }, + { + "epoch": 3.648, + "grad_norm": 6.971518223328109e-06, + "learning_rate": 0.00013350717514602563, + "loss": 0.0, + "step": 7752 + }, + { + "epoch": 3.648470588235294, + "grad_norm": 4.616137384787655e-06, + "learning_rate": 0.00013346360545837062, + "loss": 0.0, + "step": 7753 + }, + { + "epoch": 3.648941176470588, + "grad_norm": 0.09565736788647554, + "learning_rate": 0.0001334200371830048, + "loss": 0.0506, + "step": 7754 + }, + { + "epoch": 3.6494117647058824, + "grad_norm": 0.07535235013738097, + "learning_rate": 0.00013337647032364919, + "loss": 0.0062, + "step": 7755 + }, + { + "epoch": 3.6498823529411766, + "grad_norm": 0.17528823636722077, + "learning_rate": 0.0001333329048840245, + "loss": 0.0699, + "step": 7756 + }, + { + "epoch": 3.650352941176471, + "grad_norm": 3.880832129185192e-06, + "learning_rate": 0.0001332893408678515, + "loss": 0.0, + "step": 7757 + }, + { + "epoch": 3.6508235294117646, + "grad_norm": 0.1000907633128471, + "learning_rate": 0.00013324577827885078, + "loss": 0.0333, + "step": 7758 + }, + { + "epoch": 3.651294117647059, + "grad_norm": 0.0977177008863613, + "learning_rate": 0.00013320221712074276, + "loss": 0.0433, + "step": 7759 + }, + { + "epoch": 3.651764705882353, + "grad_norm": 0.521570627224484, + "learning_rate": 0.00013315865739724778, + "loss": 0.1555, + "step": 7760 + }, + { + "epoch": 3.652235294117647, + "grad_norm": 0.062211997738069194, + "learning_rate": 0.00013311509911208608, + "loss": 0.0269, + "step": 7761 + }, + { + "epoch": 3.652705882352941, + "grad_norm": 7.210151013053907e-05, + "learning_rate": 0.00013307154226897773, + "loss": 0.0, + "step": 7762 + }, + { + "epoch": 3.6531764705882352, + "grad_norm": 0.42491004897100015, + "learning_rate": 0.0001330279868716427, + "loss": 0.0719, + "step": 7763 + }, + { + "epoch": 3.6536470588235295, + "grad_norm": 0.044812440485497804, + "learning_rate": 0.00013298443292380084, + "loss": 0.0079, + "step": 7764 + }, + { + "epoch": 3.6541176470588237, + "grad_norm": 0.037473469047114696, + "learning_rate": 0.00013294088042917188, + "loss": 0.0071, + "step": 7765 + }, + { + "epoch": 3.654588235294118, + "grad_norm": 0.09123009727655712, + "learning_rate": 0.0001328973293914754, + "loss": 0.0454, + "step": 7766 + }, + { + "epoch": 3.6550588235294117, + "grad_norm": 0.09861383663514109, + "learning_rate": 0.00013285377981443085, + "loss": 0.0448, + "step": 7767 + }, + { + "epoch": 3.655529411764706, + "grad_norm": 0.11771259637352846, + "learning_rate": 0.00013281023170175766, + "loss": 0.0359, + "step": 7768 + }, + { + "epoch": 3.656, + "grad_norm": 0.11345649099274556, + "learning_rate": 0.00013276668505717498, + "loss": 0.0693, + "step": 7769 + }, + { + "epoch": 3.656470588235294, + "grad_norm": 0.1127332578627169, + "learning_rate": 0.00013272313988440187, + "loss": 0.0422, + "step": 7770 + }, + { + "epoch": 3.656941176470588, + "grad_norm": 0.08453802985743007, + "learning_rate": 0.0001326795961871574, + "loss": 0.0226, + "step": 7771 + }, + { + "epoch": 3.6574117647058824, + "grad_norm": 0.1248708753356044, + "learning_rate": 0.00013263605396916038, + "loss": 0.0402, + "step": 7772 + }, + { + "epoch": 3.6578823529411766, + "grad_norm": 1.008579679517835e-05, + "learning_rate": 0.00013259251323412952, + "loss": 0.0, + "step": 7773 + }, + { + "epoch": 3.658352941176471, + "grad_norm": 0.13656547550061057, + "learning_rate": 0.0001325489739857834, + "loss": 0.0459, + "step": 7774 + }, + { + "epoch": 3.6588235294117646, + "grad_norm": 0.09012580834630049, + "learning_rate": 0.0001325054362278405, + "loss": 0.051, + "step": 7775 + }, + { + "epoch": 3.659294117647059, + "grad_norm": 0.1185761979756613, + "learning_rate": 0.00013246189996401916, + "loss": 0.0436, + "step": 7776 + }, + { + "epoch": 3.659764705882353, + "grad_norm": 0.2377159743102824, + "learning_rate": 0.00013241836519803756, + "loss": 0.0675, + "step": 7777 + }, + { + "epoch": 3.660235294117647, + "grad_norm": 0.045800400643633145, + "learning_rate": 0.00013237483193361387, + "loss": 0.0016, + "step": 7778 + }, + { + "epoch": 3.660705882352941, + "grad_norm": 0.03365463683652252, + "learning_rate": 0.00013233130017446595, + "loss": 0.0067, + "step": 7779 + }, + { + "epoch": 3.6611764705882353, + "grad_norm": 0.15594289014361176, + "learning_rate": 0.00013228776992431165, + "loss": 0.0577, + "step": 7780 + }, + { + "epoch": 3.6616470588235295, + "grad_norm": 0.09366398639427856, + "learning_rate": 0.00013224424118686868, + "loss": 0.0441, + "step": 7781 + }, + { + "epoch": 3.6621176470588237, + "grad_norm": 0.10733739609346027, + "learning_rate": 0.00013220071396585463, + "loss": 0.034, + "step": 7782 + }, + { + "epoch": 3.662588235294118, + "grad_norm": 0.09402359533653244, + "learning_rate": 0.0001321571882649869, + "loss": 0.0293, + "step": 7783 + }, + { + "epoch": 3.6630588235294117, + "grad_norm": 0.049425677498814186, + "learning_rate": 0.00013211366408798277, + "loss": 0.0148, + "step": 7784 + }, + { + "epoch": 3.663529411764706, + "grad_norm": 0.03171001443821161, + "learning_rate": 0.00013207014143855951, + "loss": 0.0067, + "step": 7785 + }, + { + "epoch": 3.664, + "grad_norm": 0.12567806655226227, + "learning_rate": 0.00013202662032043408, + "loss": 0.018, + "step": 7786 + }, + { + "epoch": 3.664470588235294, + "grad_norm": 0.07202911410238784, + "learning_rate": 0.00013198310073732344, + "loss": 0.0166, + "step": 7787 + }, + { + "epoch": 3.664941176470588, + "grad_norm": 0.07433904812971376, + "learning_rate": 0.0001319395826929444, + "loss": 0.0307, + "step": 7788 + }, + { + "epoch": 3.6654117647058824, + "grad_norm": 0.05850722089260292, + "learning_rate": 0.00013189606619101358, + "loss": 0.0145, + "step": 7789 + }, + { + "epoch": 3.6658823529411766, + "grad_norm": 0.07381398950165478, + "learning_rate": 0.00013185255123524745, + "loss": 0.0207, + "step": 7790 + }, + { + "epoch": 3.666352941176471, + "grad_norm": 0.11654383672617426, + "learning_rate": 0.00013180903782936254, + "loss": 0.0468, + "step": 7791 + }, + { + "epoch": 3.6668235294117646, + "grad_norm": 0.14481934495613658, + "learning_rate": 0.000131765525977075, + "loss": 0.0529, + "step": 7792 + }, + { + "epoch": 3.667294117647059, + "grad_norm": 5.445039039928158e-06, + "learning_rate": 0.0001317220156821009, + "loss": 0.0, + "step": 7793 + }, + { + "epoch": 3.667764705882353, + "grad_norm": 0.1090681576928032, + "learning_rate": 0.00013167850694815642, + "loss": 0.0514, + "step": 7794 + }, + { + "epoch": 3.668235294117647, + "grad_norm": 0.05218672246837338, + "learning_rate": 0.0001316349997789573, + "loss": 0.0149, + "step": 7795 + }, + { + "epoch": 3.668705882352941, + "grad_norm": 0.09420072801232982, + "learning_rate": 0.00013159149417821923, + "loss": 0.0353, + "step": 7796 + }, + { + "epoch": 3.6691764705882353, + "grad_norm": 0.565585230271141, + "learning_rate": 0.00013154799014965785, + "loss": 0.2923, + "step": 7797 + }, + { + "epoch": 3.6696470588235295, + "grad_norm": 7.563845999108482e-05, + "learning_rate": 0.00013150448769698857, + "loss": 0.0, + "step": 7798 + }, + { + "epoch": 3.6701176470588237, + "grad_norm": 0.06428331960149934, + "learning_rate": 0.00013146098682392685, + "loss": 0.0285, + "step": 7799 + }, + { + "epoch": 3.6705882352941175, + "grad_norm": 0.03357173786867922, + "learning_rate": 0.0001314174875341878, + "loss": 0.0057, + "step": 7800 + }, + { + "epoch": 3.6710588235294117, + "grad_norm": 0.0002344183193303611, + "learning_rate": 0.00013137398983148642, + "loss": 0.0, + "step": 7801 + }, + { + "epoch": 3.671529411764706, + "grad_norm": 0.11704568130633444, + "learning_rate": 0.00013133049371953767, + "loss": 0.0296, + "step": 7802 + }, + { + "epoch": 3.672, + "grad_norm": 0.08748497455289676, + "learning_rate": 0.00013128699920205634, + "loss": 0.0467, + "step": 7803 + }, + { + "epoch": 3.672470588235294, + "grad_norm": 0.07080282372948639, + "learning_rate": 0.00013124350628275708, + "loss": 0.0278, + "step": 7804 + }, + { + "epoch": 3.672941176470588, + "grad_norm": 0.1443741674307086, + "learning_rate": 0.00013120001496535433, + "loss": 0.0919, + "step": 7805 + }, + { + "epoch": 3.6734117647058824, + "grad_norm": 0.11524947329890788, + "learning_rate": 0.00013115652525356257, + "loss": 0.0467, + "step": 7806 + }, + { + "epoch": 3.6738823529411766, + "grad_norm": 0.07855762661460745, + "learning_rate": 0.00013111303715109598, + "loss": 0.0333, + "step": 7807 + }, + { + "epoch": 3.674352941176471, + "grad_norm": 0.0002914740642348558, + "learning_rate": 0.0001310695506616686, + "loss": 0.0, + "step": 7808 + }, + { + "epoch": 3.6748235294117646, + "grad_norm": 0.04376464124468642, + "learning_rate": 0.0001310260657889945, + "loss": 0.0088, + "step": 7809 + }, + { + "epoch": 3.675294117647059, + "grad_norm": 0.009410869192261473, + "learning_rate": 0.00013098258253678744, + "loss": 0.0001, + "step": 7810 + }, + { + "epoch": 3.675764705882353, + "grad_norm": 0.08834686662365986, + "learning_rate": 0.00013093910090876108, + "loss": 0.0333, + "step": 7811 + }, + { + "epoch": 3.676235294117647, + "grad_norm": 0.08434327799767097, + "learning_rate": 0.00013089562090862905, + "loss": 0.0334, + "step": 7812 + }, + { + "epoch": 3.676705882352941, + "grad_norm": 0.02527700625495939, + "learning_rate": 0.00013085214254010465, + "loss": 0.0024, + "step": 7813 + }, + { + "epoch": 3.6771764705882353, + "grad_norm": 0.17356741484697907, + "learning_rate": 0.00013080866580690126, + "loss": 0.0819, + "step": 7814 + }, + { + "epoch": 3.6776470588235295, + "grad_norm": 0.11709619754190777, + "learning_rate": 0.00013076519071273187, + "loss": 0.0518, + "step": 7815 + }, + { + "epoch": 3.6781176470588237, + "grad_norm": 0.06269132850534936, + "learning_rate": 0.00013072171726130957, + "loss": 0.0151, + "step": 7816 + }, + { + "epoch": 3.6785882352941175, + "grad_norm": 0.08641008625344249, + "learning_rate": 0.00013067824545634718, + "loss": 0.0462, + "step": 7817 + }, + { + "epoch": 3.6790588235294117, + "grad_norm": 0.07201888071771963, + "learning_rate": 0.00013063477530155738, + "loss": 0.0287, + "step": 7818 + }, + { + "epoch": 3.679529411764706, + "grad_norm": 0.04726925628643417, + "learning_rate": 0.00013059130680065278, + "loss": 0.0086, + "step": 7819 + }, + { + "epoch": 3.68, + "grad_norm": 0.06437438892717362, + "learning_rate": 0.0001305478399573458, + "loss": 0.0149, + "step": 7820 + }, + { + "epoch": 3.680470588235294, + "grad_norm": 0.0509449272339225, + "learning_rate": 0.00013050437477534862, + "loss": 0.0173, + "step": 7821 + }, + { + "epoch": 3.680941176470588, + "grad_norm": 0.12121588884876251, + "learning_rate": 0.00013046091125837352, + "loss": 0.051, + "step": 7822 + }, + { + "epoch": 3.6814117647058824, + "grad_norm": 0.00027608422117927103, + "learning_rate": 0.00013041744941013245, + "loss": 0.0, + "step": 7823 + }, + { + "epoch": 3.6818823529411766, + "grad_norm": 0.06545627642333159, + "learning_rate": 0.00013037398923433725, + "loss": 0.0248, + "step": 7824 + }, + { + "epoch": 3.682352941176471, + "grad_norm": 0.036790112740803724, + "learning_rate": 0.00013033053073469958, + "loss": 0.0106, + "step": 7825 + }, + { + "epoch": 3.6828235294117646, + "grad_norm": 0.06853736645035712, + "learning_rate": 0.00013028707391493112, + "loss": 0.0213, + "step": 7826 + }, + { + "epoch": 3.683294117647059, + "grad_norm": 0.06804434447592377, + "learning_rate": 0.00013024361877874324, + "loss": 0.0218, + "step": 7827 + }, + { + "epoch": 3.683764705882353, + "grad_norm": 0.00018053940568010362, + "learning_rate": 0.00013020016532984714, + "loss": 0.0, + "step": 7828 + }, + { + "epoch": 3.684235294117647, + "grad_norm": 0.16194033930912916, + "learning_rate": 0.00013015671357195413, + "loss": 0.0681, + "step": 7829 + }, + { + "epoch": 3.684705882352941, + "grad_norm": 0.06350430667448712, + "learning_rate": 0.00013011326350877508, + "loss": 0.0231, + "step": 7830 + }, + { + "epoch": 3.6851764705882353, + "grad_norm": 0.09215804122016329, + "learning_rate": 0.00013006981514402085, + "loss": 0.0397, + "step": 7831 + }, + { + "epoch": 3.6856470588235295, + "grad_norm": 0.11055613232508435, + "learning_rate": 0.00013002636848140217, + "loss": 0.0668, + "step": 7832 + }, + { + "epoch": 3.6861176470588237, + "grad_norm": 0.15740083327957494, + "learning_rate": 0.00012998292352462963, + "loss": 0.071, + "step": 7833 + }, + { + "epoch": 3.6865882352941175, + "grad_norm": 0.12013211871090013, + "learning_rate": 0.00012993948027741357, + "loss": 0.0477, + "step": 7834 + }, + { + "epoch": 3.6870588235294117, + "grad_norm": 0.0003215009495214263, + "learning_rate": 0.0001298960387434643, + "loss": 0.0, + "step": 7835 + }, + { + "epoch": 3.687529411764706, + "grad_norm": 0.06285033776629065, + "learning_rate": 0.00012985259892649191, + "loss": 0.0255, + "step": 7836 + }, + { + "epoch": 3.6879999999999997, + "grad_norm": 0.03748817179163108, + "learning_rate": 0.00012980916083020646, + "loss": 0.0061, + "step": 7837 + }, + { + "epoch": 3.688470588235294, + "grad_norm": 0.037526676508541364, + "learning_rate": 0.0001297657244583176, + "loss": 0.0085, + "step": 7838 + }, + { + "epoch": 3.688941176470588, + "grad_norm": 0.09274036360507813, + "learning_rate": 0.00012972228981453522, + "loss": 0.009, + "step": 7839 + }, + { + "epoch": 3.6894117647058824, + "grad_norm": 0.05165612162176589, + "learning_rate": 0.00012967885690256872, + "loss": 0.0141, + "step": 7840 + }, + { + "epoch": 3.6898823529411766, + "grad_norm": 0.05082741754534274, + "learning_rate": 0.0001296354257261275, + "loss": 0.0131, + "step": 7841 + }, + { + "epoch": 3.690352941176471, + "grad_norm": 0.12063995007233903, + "learning_rate": 0.0001295919962889208, + "loss": 0.0654, + "step": 7842 + }, + { + "epoch": 3.6908235294117646, + "grad_norm": 0.12871482101835446, + "learning_rate": 0.0001295485685946577, + "loss": 0.0408, + "step": 7843 + }, + { + "epoch": 3.691294117647059, + "grad_norm": 0.1731286284039703, + "learning_rate": 0.0001295051426470472, + "loss": 0.108, + "step": 7844 + }, + { + "epoch": 3.691764705882353, + "grad_norm": 0.09002105767739177, + "learning_rate": 0.00012946171844979804, + "loss": 0.0438, + "step": 7845 + }, + { + "epoch": 3.692235294117647, + "grad_norm": 0.07249799802951482, + "learning_rate": 0.00012941829600661886, + "loss": 0.0128, + "step": 7846 + }, + { + "epoch": 3.692705882352941, + "grad_norm": 0.08679505201670652, + "learning_rate": 0.00012937487532121818, + "loss": 0.0395, + "step": 7847 + }, + { + "epoch": 3.6931764705882353, + "grad_norm": 0.08834251999974778, + "learning_rate": 0.00012933145639730428, + "loss": 0.0452, + "step": 7848 + }, + { + "epoch": 3.6936470588235295, + "grad_norm": 0.06963234442370746, + "learning_rate": 0.00012928803923858536, + "loss": 0.0284, + "step": 7849 + }, + { + "epoch": 3.6941176470588237, + "grad_norm": 0.07913177307123982, + "learning_rate": 0.00012924462384876953, + "loss": 0.032, + "step": 7850 + }, + { + "epoch": 3.6945882352941175, + "grad_norm": 0.05795030362784744, + "learning_rate": 0.00012920121023156462, + "loss": 0.0114, + "step": 7851 + }, + { + "epoch": 3.6950588235294117, + "grad_norm": 0.1425352911348797, + "learning_rate": 0.00012915779839067838, + "loss": 0.08, + "step": 7852 + }, + { + "epoch": 3.695529411764706, + "grad_norm": 0.0003400471628285551, + "learning_rate": 0.00012911438832981836, + "loss": 0.0, + "step": 7853 + }, + { + "epoch": 3.6959999999999997, + "grad_norm": 0.09758188916673244, + "learning_rate": 0.00012907098005269203, + "loss": 0.024, + "step": 7854 + }, + { + "epoch": 3.696470588235294, + "grad_norm": 0.16940950525148707, + "learning_rate": 0.00012902757356300667, + "loss": 0.0978, + "step": 7855 + }, + { + "epoch": 3.696941176470588, + "grad_norm": 1.592301918071243e-05, + "learning_rate": 0.00012898416886446937, + "loss": 0.0, + "step": 7856 + }, + { + "epoch": 3.6974117647058824, + "grad_norm": 0.08336770971747785, + "learning_rate": 0.00012894076596078717, + "loss": 0.0296, + "step": 7857 + }, + { + "epoch": 3.6978823529411766, + "grad_norm": 0.035105222275439774, + "learning_rate": 0.00012889736485566683, + "loss": 0.0064, + "step": 7858 + }, + { + "epoch": 3.698352941176471, + "grad_norm": 0.22018018113715965, + "learning_rate": 0.000128853965552815, + "loss": 0.1048, + "step": 7859 + }, + { + "epoch": 3.6988235294117646, + "grad_norm": 0.1465468198483221, + "learning_rate": 0.0001288105680559383, + "loss": 0.0774, + "step": 7860 + }, + { + "epoch": 3.699294117647059, + "grad_norm": 0.10809672652119283, + "learning_rate": 0.000128767172368743, + "loss": 0.0394, + "step": 7861 + }, + { + "epoch": 3.699764705882353, + "grad_norm": 0.03374855039113221, + "learning_rate": 0.0001287237784949353, + "loss": 0.0065, + "step": 7862 + }, + { + "epoch": 3.700235294117647, + "grad_norm": 0.1876742954510282, + "learning_rate": 0.0001286803864382213, + "loss": 0.059, + "step": 7863 + }, + { + "epoch": 3.700705882352941, + "grad_norm": 0.05893680445307963, + "learning_rate": 0.00012863699620230688, + "loss": 0.0148, + "step": 7864 + }, + { + "epoch": 3.7011764705882353, + "grad_norm": 0.06398255090993406, + "learning_rate": 0.00012859360779089775, + "loss": 0.0226, + "step": 7865 + }, + { + "epoch": 3.7016470588235295, + "grad_norm": 2.2097677245284073e-06, + "learning_rate": 0.00012855022120769952, + "loss": 0.0, + "step": 7866 + }, + { + "epoch": 3.7021176470588237, + "grad_norm": 0.09858622892144767, + "learning_rate": 0.0001285068364564176, + "loss": 0.0534, + "step": 7867 + }, + { + "epoch": 3.7025882352941175, + "grad_norm": 0.05285423384038474, + "learning_rate": 0.0001284634535407573, + "loss": 0.0193, + "step": 7868 + }, + { + "epoch": 3.7030588235294117, + "grad_norm": 0.05365451488259158, + "learning_rate": 0.00012842007246442367, + "loss": 0.0172, + "step": 7869 + }, + { + "epoch": 3.703529411764706, + "grad_norm": 0.07539244972722492, + "learning_rate": 0.0001283766932311217, + "loss": 0.0138, + "step": 7870 + }, + { + "epoch": 3.7039999999999997, + "grad_norm": 0.03605418239315377, + "learning_rate": 0.00012833331584455622, + "loss": 0.0091, + "step": 7871 + }, + { + "epoch": 3.704470588235294, + "grad_norm": 0.242436208710035, + "learning_rate": 0.0001282899403084318, + "loss": 0.058, + "step": 7872 + }, + { + "epoch": 3.704941176470588, + "grad_norm": 0.16753782862246056, + "learning_rate": 0.000128246566626453, + "loss": 0.022, + "step": 7873 + }, + { + "epoch": 3.7054117647058824, + "grad_norm": 0.08081242416458397, + "learning_rate": 0.0001282031948023241, + "loss": 0.0346, + "step": 7874 + }, + { + "epoch": 3.7058823529411766, + "grad_norm": 1.6650749150306845e-06, + "learning_rate": 0.0001281598248397493, + "loss": 0.0, + "step": 7875 + }, + { + "epoch": 3.706352941176471, + "grad_norm": 1.0838529105371406e-05, + "learning_rate": 0.00012811645674243254, + "loss": 0.0, + "step": 7876 + }, + { + "epoch": 3.7068235294117646, + "grad_norm": 0.15294225768849068, + "learning_rate": 0.00012807309051407774, + "loss": 0.0644, + "step": 7877 + }, + { + "epoch": 3.707294117647059, + "grad_norm": 0.1293801301418611, + "learning_rate": 0.00012802972615838856, + "loss": 0.0536, + "step": 7878 + }, + { + "epoch": 3.707764705882353, + "grad_norm": 0.065644046143035, + "learning_rate": 0.00012798636367906849, + "loss": 0.0178, + "step": 7879 + }, + { + "epoch": 3.708235294117647, + "grad_norm": 0.13046994235688616, + "learning_rate": 0.00012794300307982098, + "loss": 0.0633, + "step": 7880 + }, + { + "epoch": 3.708705882352941, + "grad_norm": 0.08045072172715947, + "learning_rate": 0.0001278996443643492, + "loss": 0.0317, + "step": 7881 + }, + { + "epoch": 3.7091764705882353, + "grad_norm": 0.0749188354202972, + "learning_rate": 0.00012785628753635617, + "loss": 0.0341, + "step": 7882 + }, + { + "epoch": 3.7096470588235295, + "grad_norm": 0.12309089070315761, + "learning_rate": 0.0001278129325995448, + "loss": 0.0506, + "step": 7883 + }, + { + "epoch": 3.7101176470588237, + "grad_norm": 0.14181843914726394, + "learning_rate": 0.00012776957955761784, + "loss": 0.0465, + "step": 7884 + }, + { + "epoch": 3.7105882352941175, + "grad_norm": 9.609323126867225e-06, + "learning_rate": 0.0001277262284142778, + "loss": 0.0, + "step": 7885 + }, + { + "epoch": 3.7110588235294117, + "grad_norm": 0.10062994456265985, + "learning_rate": 0.00012768287917322706, + "loss": 0.0497, + "step": 7886 + }, + { + "epoch": 3.711529411764706, + "grad_norm": 9.871236347965776e-06, + "learning_rate": 0.00012763953183816794, + "loss": 0.0, + "step": 7887 + }, + { + "epoch": 3.7119999999999997, + "grad_norm": 0.29883495083469047, + "learning_rate": 0.00012759618641280251, + "loss": 0.0888, + "step": 7888 + }, + { + "epoch": 3.712470588235294, + "grad_norm": 0.14974226827421752, + "learning_rate": 0.00012755284290083264, + "loss": 0.0434, + "step": 7889 + }, + { + "epoch": 3.712941176470588, + "grad_norm": 0.12024779005254596, + "learning_rate": 0.00012750950130596006, + "loss": 0.0327, + "step": 7890 + }, + { + "epoch": 3.7134117647058824, + "grad_norm": 4.965625611238885e-06, + "learning_rate": 0.0001274661616318864, + "loss": 0.0, + "step": 7891 + }, + { + "epoch": 3.7138823529411766, + "grad_norm": 0.0678857686398857, + "learning_rate": 0.0001274228238823131, + "loss": 0.0183, + "step": 7892 + }, + { + "epoch": 3.714352941176471, + "grad_norm": 0.11234993586586375, + "learning_rate": 0.00012737948806094135, + "loss": 0.0328, + "step": 7893 + }, + { + "epoch": 3.7148235294117646, + "grad_norm": 0.19674440897564363, + "learning_rate": 0.00012733615417147228, + "loss": 0.0589, + "step": 7894 + }, + { + "epoch": 3.715294117647059, + "grad_norm": 0.1392623765324563, + "learning_rate": 0.0001272928222176068, + "loss": 0.0695, + "step": 7895 + }, + { + "epoch": 3.715764705882353, + "grad_norm": 0.15919440585489275, + "learning_rate": 0.0001272494922030457, + "loss": 0.0677, + "step": 7896 + }, + { + "epoch": 3.716235294117647, + "grad_norm": 0.086613382241058, + "learning_rate": 0.0001272061641314895, + "loss": 0.0325, + "step": 7897 + }, + { + "epoch": 3.716705882352941, + "grad_norm": 7.16674559151401e-05, + "learning_rate": 0.00012716283800663874, + "loss": 0.0, + "step": 7898 + }, + { + "epoch": 3.7171764705882353, + "grad_norm": 0.046439204842758874, + "learning_rate": 0.00012711951383219362, + "loss": 0.0092, + "step": 7899 + }, + { + "epoch": 3.7176470588235295, + "grad_norm": 0.1303877397065014, + "learning_rate": 0.00012707619161185423, + "loss": 0.0441, + "step": 7900 + }, + { + "epoch": 3.7181176470588237, + "grad_norm": 0.08036939736738014, + "learning_rate": 0.00012703287134932054, + "loss": 0.0303, + "step": 7901 + }, + { + "epoch": 3.7185882352941175, + "grad_norm": 0.11653975971864784, + "learning_rate": 0.0001269895530482923, + "loss": 0.0177, + "step": 7902 + }, + { + "epoch": 3.7190588235294118, + "grad_norm": 0.09444582871647421, + "learning_rate": 0.00012694623671246908, + "loss": 0.0372, + "step": 7903 + }, + { + "epoch": 3.719529411764706, + "grad_norm": 0.04237321739366382, + "learning_rate": 0.0001269029223455503, + "loss": 0.0094, + "step": 7904 + }, + { + "epoch": 3.7199999999999998, + "grad_norm": 0.09140676815546134, + "learning_rate": 0.00012685960995123527, + "loss": 0.0355, + "step": 7905 + }, + { + "epoch": 3.720470588235294, + "grad_norm": 0.18489407177234887, + "learning_rate": 0.00012681629953322307, + "loss": 0.0759, + "step": 7906 + }, + { + "epoch": 3.720941176470588, + "grad_norm": 0.11390221887685018, + "learning_rate": 0.00012677299109521256, + "loss": 0.0602, + "step": 7907 + }, + { + "epoch": 3.7214117647058824, + "grad_norm": 0.0799146553289189, + "learning_rate": 0.00012672968464090255, + "loss": 0.0255, + "step": 7908 + }, + { + "epoch": 3.7218823529411766, + "grad_norm": 0.08264734142091006, + "learning_rate": 0.00012668638017399164, + "loss": 0.01, + "step": 7909 + }, + { + "epoch": 3.722352941176471, + "grad_norm": 0.08043074352443257, + "learning_rate": 0.00012664307769817815, + "loss": 0.025, + "step": 7910 + }, + { + "epoch": 3.7228235294117646, + "grad_norm": 0.532300788180054, + "learning_rate": 0.00012659977721716045, + "loss": 0.1627, + "step": 7911 + }, + { + "epoch": 3.723294117647059, + "grad_norm": 0.1342734475742349, + "learning_rate": 0.00012655647873463657, + "loss": 0.0817, + "step": 7912 + }, + { + "epoch": 3.723764705882353, + "grad_norm": 0.08236915379872115, + "learning_rate": 0.00012651318225430437, + "loss": 0.0242, + "step": 7913 + }, + { + "epoch": 3.724235294117647, + "grad_norm": 0.04818544706472791, + "learning_rate": 0.00012646988777986157, + "loss": 0.0116, + "step": 7914 + }, + { + "epoch": 3.724705882352941, + "grad_norm": 3.474365732957765e-05, + "learning_rate": 0.00012642659531500583, + "loss": 0.0, + "step": 7915 + }, + { + "epoch": 3.7251764705882353, + "grad_norm": 0.12834931123799068, + "learning_rate": 0.00012638330486343445, + "loss": 0.0586, + "step": 7916 + }, + { + "epoch": 3.7256470588235295, + "grad_norm": 0.00032415685531000764, + "learning_rate": 0.00012634001642884465, + "loss": 0.0, + "step": 7917 + }, + { + "epoch": 3.7261176470588238, + "grad_norm": 0.0569789552002435, + "learning_rate": 0.0001262967300149335, + "loss": 0.0171, + "step": 7918 + }, + { + "epoch": 3.7265882352941175, + "grad_norm": 0.02035182832530013, + "learning_rate": 0.0001262534456253979, + "loss": 0.0028, + "step": 7919 + }, + { + "epoch": 3.7270588235294118, + "grad_norm": 0.08749575882257708, + "learning_rate": 0.0001262101632639345, + "loss": 0.0413, + "step": 7920 + }, + { + "epoch": 3.727529411764706, + "grad_norm": 0.07786021380868473, + "learning_rate": 0.00012616688293423987, + "loss": 0.0273, + "step": 7921 + }, + { + "epoch": 3.7279999999999998, + "grad_norm": 0.04813325887728521, + "learning_rate": 0.00012612360464001033, + "loss": 0.0096, + "step": 7922 + }, + { + "epoch": 3.728470588235294, + "grad_norm": 0.260539792971769, + "learning_rate": 0.00012608032838494205, + "loss": 0.1356, + "step": 7923 + }, + { + "epoch": 3.728941176470588, + "grad_norm": 0.06712897619318003, + "learning_rate": 0.00012603705417273107, + "loss": 0.0216, + "step": 7924 + }, + { + "epoch": 3.7294117647058824, + "grad_norm": 0.0673487838458464, + "learning_rate": 0.0001259937820070732, + "loss": 0.0246, + "step": 7925 + }, + { + "epoch": 3.7298823529411766, + "grad_norm": 0.03861170044246201, + "learning_rate": 0.0001259505118916641, + "loss": 0.0053, + "step": 7926 + }, + { + "epoch": 3.7303529411764704, + "grad_norm": 0.024850415829028215, + "learning_rate": 0.0001259072438301992, + "loss": 0.0031, + "step": 7927 + }, + { + "epoch": 3.7308235294117646, + "grad_norm": 0.09640815768695872, + "learning_rate": 0.00012586397782637392, + "loss": 0.042, + "step": 7928 + }, + { + "epoch": 3.731294117647059, + "grad_norm": 0.021230950669091468, + "learning_rate": 0.0001258207138838833, + "loss": 0.0021, + "step": 7929 + }, + { + "epoch": 3.731764705882353, + "grad_norm": 0.1556545485631072, + "learning_rate": 0.00012577745200642227, + "loss": 0.0751, + "step": 7930 + }, + { + "epoch": 3.732235294117647, + "grad_norm": 0.07092464040824391, + "learning_rate": 0.00012573419219768562, + "loss": 0.0253, + "step": 7931 + }, + { + "epoch": 3.732705882352941, + "grad_norm": 0.0006817656495050914, + "learning_rate": 0.00012569093446136806, + "loss": 0.0, + "step": 7932 + }, + { + "epoch": 3.7331764705882353, + "grad_norm": 0.02937014706797096, + "learning_rate": 0.0001256476788011639, + "loss": 0.0042, + "step": 7933 + }, + { + "epoch": 3.7336470588235295, + "grad_norm": 3.5450102541462735e-05, + "learning_rate": 0.00012560442522076745, + "loss": 0.0, + "step": 7934 + }, + { + "epoch": 3.7341176470588238, + "grad_norm": 0.00012642488881434077, + "learning_rate": 0.00012556117372387272, + "loss": 0.0, + "step": 7935 + }, + { + "epoch": 3.7345882352941175, + "grad_norm": 0.1344120141964087, + "learning_rate": 0.00012551792431417363, + "loss": 0.0563, + "step": 7936 + }, + { + "epoch": 3.7350588235294118, + "grad_norm": 0.07709789588540737, + "learning_rate": 0.0001254746769953639, + "loss": 0.0257, + "step": 7937 + }, + { + "epoch": 3.735529411764706, + "grad_norm": 0.10571222817515846, + "learning_rate": 0.000125431431771137, + "loss": 0.0436, + "step": 7938 + }, + { + "epoch": 3.7359999999999998, + "grad_norm": 0.0004880834348283337, + "learning_rate": 0.0001253881886451864, + "loss": 0.0, + "step": 7939 + }, + { + "epoch": 3.736470588235294, + "grad_norm": 0.0004912276041821315, + "learning_rate": 0.0001253449476212052, + "loss": 0.0, + "step": 7940 + }, + { + "epoch": 3.736941176470588, + "grad_norm": 0.06104313211270069, + "learning_rate": 0.00012530170870288636, + "loss": 0.0144, + "step": 7941 + }, + { + "epoch": 3.7374117647058824, + "grad_norm": 0.13033921418688424, + "learning_rate": 0.00012525847189392278, + "loss": 0.0581, + "step": 7942 + }, + { + "epoch": 3.7378823529411767, + "grad_norm": 0.07027374284233276, + "learning_rate": 0.00012521523719800706, + "loss": 0.0104, + "step": 7943 + }, + { + "epoch": 3.7383529411764704, + "grad_norm": 1.437091240505419e-05, + "learning_rate": 0.00012517200461883163, + "loss": 0.0, + "step": 7944 + }, + { + "epoch": 3.7388235294117647, + "grad_norm": 7.05171602009938e-05, + "learning_rate": 0.00012512877416008878, + "loss": 0.0, + "step": 7945 + }, + { + "epoch": 3.739294117647059, + "grad_norm": 0.15300235357976683, + "learning_rate": 0.00012508554582547064, + "loss": 0.0752, + "step": 7946 + }, + { + "epoch": 3.739764705882353, + "grad_norm": 0.14207456750075378, + "learning_rate": 0.00012504231961866908, + "loss": 0.0565, + "step": 7947 + }, + { + "epoch": 3.740235294117647, + "grad_norm": 0.04367176455277767, + "learning_rate": 0.00012499909554337582, + "loss": 0.0109, + "step": 7948 + }, + { + "epoch": 3.740705882352941, + "grad_norm": 9.20262475522379e-06, + "learning_rate": 0.00012495587360328246, + "loss": 0.0, + "step": 7949 + }, + { + "epoch": 3.7411764705882353, + "grad_norm": 0.13120621470688623, + "learning_rate": 0.00012491265380208032, + "loss": 0.0386, + "step": 7950 + }, + { + "epoch": 3.7416470588235295, + "grad_norm": 0.11208606231735663, + "learning_rate": 0.0001248694361434606, + "loss": 0.0446, + "step": 7951 + }, + { + "epoch": 3.7421176470588238, + "grad_norm": 0.06151162948192391, + "learning_rate": 0.0001248262206311143, + "loss": 0.0221, + "step": 7952 + }, + { + "epoch": 3.7425882352941175, + "grad_norm": 0.2886688990928001, + "learning_rate": 0.00012478300726873225, + "loss": 0.0462, + "step": 7953 + }, + { + "epoch": 3.7430588235294118, + "grad_norm": 0.05449957253727095, + "learning_rate": 0.00012473979606000504, + "loss": 0.0142, + "step": 7954 + }, + { + "epoch": 3.743529411764706, + "grad_norm": 0.04088586570180057, + "learning_rate": 0.00012469658700862315, + "loss": 0.0063, + "step": 7955 + }, + { + "epoch": 3.7439999999999998, + "grad_norm": 0.11552839487627237, + "learning_rate": 0.00012465338011827685, + "loss": 0.0385, + "step": 7956 + }, + { + "epoch": 3.744470588235294, + "grad_norm": 0.00011324395809904188, + "learning_rate": 0.00012461017539265622, + "loss": 0.0, + "step": 7957 + }, + { + "epoch": 3.744941176470588, + "grad_norm": 0.12252495665756617, + "learning_rate": 0.00012456697283545114, + "loss": 0.0539, + "step": 7958 + }, + { + "epoch": 3.7454117647058824, + "grad_norm": 0.00024309430297680766, + "learning_rate": 0.00012452377245035132, + "loss": 0.0, + "step": 7959 + }, + { + "epoch": 3.7458823529411767, + "grad_norm": 0.03142036078682889, + "learning_rate": 0.00012448057424104632, + "loss": 0.0067, + "step": 7960 + }, + { + "epoch": 3.7463529411764704, + "grad_norm": 0.08000273763810604, + "learning_rate": 0.00012443737821122541, + "loss": 0.0119, + "step": 7961 + }, + { + "epoch": 3.7468235294117647, + "grad_norm": 0.10084212188399076, + "learning_rate": 0.0001243941843645778, + "loss": 0.0171, + "step": 7962 + }, + { + "epoch": 3.747294117647059, + "grad_norm": 0.05844556251267946, + "learning_rate": 0.00012435099270479246, + "loss": 0.0029, + "step": 7963 + }, + { + "epoch": 3.7477647058823527, + "grad_norm": 0.07879700600613868, + "learning_rate": 0.00012430780323555816, + "loss": 0.0286, + "step": 7964 + }, + { + "epoch": 3.748235294117647, + "grad_norm": 0.08811188837182424, + "learning_rate": 0.00012426461596056344, + "loss": 0.0136, + "step": 7965 + }, + { + "epoch": 3.748705882352941, + "grad_norm": 0.09265269479513469, + "learning_rate": 0.00012422143088349679, + "loss": 0.0388, + "step": 7966 + }, + { + "epoch": 3.7491764705882353, + "grad_norm": 0.25280204737274764, + "learning_rate": 0.00012417824800804638, + "loss": 0.0667, + "step": 7967 + }, + { + "epoch": 3.7496470588235296, + "grad_norm": 0.08598632553376774, + "learning_rate": 0.0001241350673379002, + "loss": 0.0235, + "step": 7968 + }, + { + "epoch": 3.7501176470588238, + "grad_norm": 0.10011384589806825, + "learning_rate": 0.0001240918888767462, + "loss": 0.0374, + "step": 7969 + }, + { + "epoch": 3.7505882352941176, + "grad_norm": 0.10337397811686297, + "learning_rate": 0.00012404871262827193, + "loss": 0.0524, + "step": 7970 + }, + { + "epoch": 3.751058823529412, + "grad_norm": 0.14660979625518267, + "learning_rate": 0.00012400553859616488, + "loss": 0.0646, + "step": 7971 + }, + { + "epoch": 3.751529411764706, + "grad_norm": 0.0920143300736906, + "learning_rate": 0.00012396236678411236, + "loss": 0.0418, + "step": 7972 + }, + { + "epoch": 3.752, + "grad_norm": 1.1914372201180858e-05, + "learning_rate": 0.00012391919719580144, + "loss": 0.0, + "step": 7973 + }, + { + "epoch": 3.752470588235294, + "grad_norm": 0.06070549032127403, + "learning_rate": 0.00012387602983491894, + "loss": 0.0154, + "step": 7974 + }, + { + "epoch": 3.7529411764705882, + "grad_norm": 0.14252653505364904, + "learning_rate": 0.00012383286470515166, + "loss": 0.0681, + "step": 7975 + }, + { + "epoch": 3.7534117647058824, + "grad_norm": 0.050675894004973555, + "learning_rate": 0.0001237897018101861, + "loss": 0.0097, + "step": 7976 + }, + { + "epoch": 3.7538823529411767, + "grad_norm": 0.1285304846288674, + "learning_rate": 0.00012374654115370856, + "loss": 0.048, + "step": 7977 + }, + { + "epoch": 3.7543529411764704, + "grad_norm": 0.03581042590303132, + "learning_rate": 0.0001237033827394052, + "loss": 0.0039, + "step": 7978 + }, + { + "epoch": 3.7548235294117647, + "grad_norm": 0.14356386554547074, + "learning_rate": 0.0001236602265709619, + "loss": 0.0588, + "step": 7979 + }, + { + "epoch": 3.755294117647059, + "grad_norm": 0.1301587430674203, + "learning_rate": 0.00012361707265206448, + "loss": 0.0588, + "step": 7980 + }, + { + "epoch": 3.7557647058823527, + "grad_norm": 0.1350058459407067, + "learning_rate": 0.00012357392098639848, + "loss": 0.05, + "step": 7981 + }, + { + "epoch": 3.756235294117647, + "grad_norm": 0.10039637247847462, + "learning_rate": 0.00012353077157764922, + "loss": 0.0102, + "step": 7982 + }, + { + "epoch": 3.756705882352941, + "grad_norm": 0.0431114088574293, + "learning_rate": 0.00012348762442950187, + "loss": 0.0092, + "step": 7983 + }, + { + "epoch": 3.7571764705882353, + "grad_norm": 0.03718783830689841, + "learning_rate": 0.00012344447954564148, + "loss": 0.0059, + "step": 7984 + }, + { + "epoch": 3.7576470588235296, + "grad_norm": 0.07937009561701343, + "learning_rate": 0.0001234013369297528, + "loss": 0.0207, + "step": 7985 + }, + { + "epoch": 3.758117647058824, + "grad_norm": 0.1516518794262152, + "learning_rate": 0.00012335819658552036, + "loss": 0.0882, + "step": 7986 + }, + { + "epoch": 3.7585882352941176, + "grad_norm": 0.08196738460169702, + "learning_rate": 0.00012331505851662864, + "loss": 0.0268, + "step": 7987 + }, + { + "epoch": 3.759058823529412, + "grad_norm": 0.07930370675469464, + "learning_rate": 0.00012327192272676183, + "loss": 0.0321, + "step": 7988 + }, + { + "epoch": 3.759529411764706, + "grad_norm": 0.06242557537045177, + "learning_rate": 0.0001232287892196039, + "loss": 0.0186, + "step": 7989 + }, + { + "epoch": 3.76, + "grad_norm": 0.0735443767736928, + "learning_rate": 0.00012318565799883865, + "loss": 0.0077, + "step": 7990 + }, + { + "epoch": 3.760470588235294, + "grad_norm": 0.1150860265978786, + "learning_rate": 0.0001231425290681498, + "loss": 0.04, + "step": 7991 + }, + { + "epoch": 3.7609411764705882, + "grad_norm": 0.08966153617001042, + "learning_rate": 0.00012309940243122064, + "loss": 0.025, + "step": 7992 + }, + { + "epoch": 3.7614117647058825, + "grad_norm": 0.07124815159636587, + "learning_rate": 0.00012305627809173448, + "loss": 0.0191, + "step": 7993 + }, + { + "epoch": 3.7618823529411767, + "grad_norm": 1.6504018292841268e-05, + "learning_rate": 0.00012301315605337435, + "loss": 0.0, + "step": 7994 + }, + { + "epoch": 3.7623529411764705, + "grad_norm": 0.05538852280521633, + "learning_rate": 0.00012297003631982304, + "loss": 0.0179, + "step": 7995 + }, + { + "epoch": 3.7628235294117647, + "grad_norm": 0.05315280605051041, + "learning_rate": 0.0001229269188947632, + "loss": 0.0134, + "step": 7996 + }, + { + "epoch": 3.763294117647059, + "grad_norm": 0.13851507387932407, + "learning_rate": 0.0001228838037818773, + "loss": 0.0491, + "step": 7997 + }, + { + "epoch": 3.7637647058823527, + "grad_norm": 0.10960945618456358, + "learning_rate": 0.00012284069098484758, + "loss": 0.0541, + "step": 7998 + }, + { + "epoch": 3.764235294117647, + "grad_norm": 0.08024415744555695, + "learning_rate": 0.000122797580507356, + "loss": 0.025, + "step": 7999 + }, + { + "epoch": 3.764705882352941, + "grad_norm": 0.12689254701125438, + "learning_rate": 0.00012275447235308453, + "loss": 0.0506, + "step": 8000 + }, + { + "epoch": 3.7651764705882353, + "grad_norm": 0.12003093627322788, + "learning_rate": 0.00012271136652571474, + "loss": 0.0626, + "step": 8001 + }, + { + "epoch": 3.7656470588235296, + "grad_norm": 0.05536182354548992, + "learning_rate": 0.0001226682630289281, + "loss": 0.0178, + "step": 8002 + }, + { + "epoch": 3.766117647058824, + "grad_norm": 0.04437597892552871, + "learning_rate": 0.00012262516186640587, + "loss": 0.0084, + "step": 8003 + }, + { + "epoch": 3.7665882352941176, + "grad_norm": 0.07795046526725122, + "learning_rate": 0.0001225820630418291, + "loss": 0.0202, + "step": 8004 + }, + { + "epoch": 3.767058823529412, + "grad_norm": 0.061007784583582, + "learning_rate": 0.00012253896655887861, + "loss": 0.0172, + "step": 8005 + }, + { + "epoch": 3.767529411764706, + "grad_norm": 0.11950678609425719, + "learning_rate": 0.00012249587242123505, + "loss": 0.0449, + "step": 8006 + }, + { + "epoch": 3.768, + "grad_norm": 0.11587897609294558, + "learning_rate": 0.00012245278063257892, + "loss": 0.0597, + "step": 8007 + }, + { + "epoch": 3.768470588235294, + "grad_norm": 0.06918681840680756, + "learning_rate": 0.00012240969119659044, + "loss": 0.0209, + "step": 8008 + }, + { + "epoch": 3.7689411764705882, + "grad_norm": 0.046112396264325686, + "learning_rate": 0.00012236660411694964, + "loss": 0.0103, + "step": 8009 + }, + { + "epoch": 3.7694117647058825, + "grad_norm": 0.12440760626859441, + "learning_rate": 0.00012232351939733642, + "loss": 0.0526, + "step": 8010 + }, + { + "epoch": 3.7698823529411767, + "grad_norm": 0.14317236705318304, + "learning_rate": 0.0001222804370414304, + "loss": 0.0541, + "step": 8011 + }, + { + "epoch": 3.7703529411764705, + "grad_norm": 0.140505766947021, + "learning_rate": 0.00012223735705291098, + "loss": 0.0638, + "step": 8012 + }, + { + "epoch": 3.7708235294117647, + "grad_norm": 3.0600352897970074e-05, + "learning_rate": 0.00012219427943545744, + "loss": 0.0, + "step": 8013 + }, + { + "epoch": 3.771294117647059, + "grad_norm": 0.10946541651774022, + "learning_rate": 0.00012215120419274887, + "loss": 0.0572, + "step": 8014 + }, + { + "epoch": 3.7717647058823527, + "grad_norm": 0.13175668481865108, + "learning_rate": 0.00012210813132846406, + "loss": 0.0515, + "step": 8015 + }, + { + "epoch": 3.772235294117647, + "grad_norm": 0.09639604111811358, + "learning_rate": 0.0001220650608462816, + "loss": 0.0284, + "step": 8016 + }, + { + "epoch": 3.772705882352941, + "grad_norm": 0.057171754307580286, + "learning_rate": 0.00012202199274987999, + "loss": 0.0166, + "step": 8017 + }, + { + "epoch": 3.7731764705882354, + "grad_norm": 0.06993467905485122, + "learning_rate": 0.00012197892704293743, + "loss": 0.0218, + "step": 8018 + }, + { + "epoch": 3.7736470588235296, + "grad_norm": 0.05249527774524538, + "learning_rate": 0.00012193586372913192, + "loss": 0.0163, + "step": 8019 + }, + { + "epoch": 3.774117647058824, + "grad_norm": 0.08743584078757333, + "learning_rate": 0.00012189280281214126, + "loss": 0.025, + "step": 8020 + }, + { + "epoch": 3.7745882352941176, + "grad_norm": 0.00015531031812970352, + "learning_rate": 0.00012184974429564318, + "loss": 0.0, + "step": 8021 + }, + { + "epoch": 3.775058823529412, + "grad_norm": 0.07445478590937561, + "learning_rate": 0.000121806688183315, + "loss": 0.0286, + "step": 8022 + }, + { + "epoch": 3.775529411764706, + "grad_norm": 0.1344976458017688, + "learning_rate": 0.00012176363447883397, + "loss": 0.0791, + "step": 8023 + }, + { + "epoch": 3.776, + "grad_norm": 0.15603355473939734, + "learning_rate": 0.00012172058318587698, + "loss": 0.0994, + "step": 8024 + }, + { + "epoch": 3.776470588235294, + "grad_norm": 0.2123573022805047, + "learning_rate": 0.00012167753430812093, + "loss": 0.0631, + "step": 8025 + }, + { + "epoch": 3.7769411764705882, + "grad_norm": 0.09196045622793729, + "learning_rate": 0.0001216344878492424, + "loss": 0.0299, + "step": 8026 + }, + { + "epoch": 3.7774117647058825, + "grad_norm": 0.1779967322431395, + "learning_rate": 0.0001215914438129177, + "loss": 0.0734, + "step": 8027 + }, + { + "epoch": 3.7778823529411767, + "grad_norm": 0.02141715209997693, + "learning_rate": 0.00012154840220282306, + "loss": 0.0032, + "step": 8028 + }, + { + "epoch": 3.7783529411764705, + "grad_norm": 0.06420182603602796, + "learning_rate": 0.00012150536302263444, + "loss": 0.0257, + "step": 8029 + }, + { + "epoch": 3.7788235294117647, + "grad_norm": 1.5193971342928253e-05, + "learning_rate": 0.00012146232627602753, + "loss": 0.0, + "step": 8030 + }, + { + "epoch": 3.779294117647059, + "grad_norm": 0.11005978738363771, + "learning_rate": 0.00012141929196667798, + "loss": 0.0202, + "step": 8031 + }, + { + "epoch": 3.7797647058823527, + "grad_norm": 0.07212736516749906, + "learning_rate": 0.00012137626009826108, + "loss": 0.0219, + "step": 8032 + }, + { + "epoch": 3.780235294117647, + "grad_norm": 0.04947467557310754, + "learning_rate": 0.00012133323067445198, + "loss": 0.0126, + "step": 8033 + }, + { + "epoch": 3.780705882352941, + "grad_norm": 0.06685849523903724, + "learning_rate": 0.00012129020369892555, + "loss": 0.0241, + "step": 8034 + }, + { + "epoch": 3.7811764705882354, + "grad_norm": 6.698941547110073e-06, + "learning_rate": 0.00012124717917535656, + "loss": 0.0, + "step": 8035 + }, + { + "epoch": 3.7816470588235296, + "grad_norm": 0.04314780586740552, + "learning_rate": 0.00012120415710741951, + "loss": 0.0104, + "step": 8036 + }, + { + "epoch": 3.782117647058824, + "grad_norm": 0.12293173380633465, + "learning_rate": 0.00012116113749878867, + "loss": 0.0509, + "step": 8037 + }, + { + "epoch": 3.7825882352941176, + "grad_norm": 0.14762501430680758, + "learning_rate": 0.00012111812035313814, + "loss": 0.0713, + "step": 8038 + }, + { + "epoch": 3.783058823529412, + "grad_norm": 6.037256380850382e-06, + "learning_rate": 0.00012107510567414183, + "loss": 0.0, + "step": 8039 + }, + { + "epoch": 3.783529411764706, + "grad_norm": 0.03519777445346762, + "learning_rate": 0.00012103209346547332, + "loss": 0.0042, + "step": 8040 + }, + { + "epoch": 3.784, + "grad_norm": 0.09067007767993059, + "learning_rate": 0.00012098908373080617, + "loss": 0.0309, + "step": 8041 + }, + { + "epoch": 3.784470588235294, + "grad_norm": 5.305357315263809e-05, + "learning_rate": 0.00012094607647381355, + "loss": 0.0, + "step": 8042 + }, + { + "epoch": 3.7849411764705883, + "grad_norm": 2.7442747429710984e-05, + "learning_rate": 0.00012090307169816847, + "loss": 0.0, + "step": 8043 + }, + { + "epoch": 3.7854117647058825, + "grad_norm": 6.41930782716958e-05, + "learning_rate": 0.00012086006940754382, + "loss": 0.0, + "step": 8044 + }, + { + "epoch": 3.7858823529411767, + "grad_norm": 0.08902927718441468, + "learning_rate": 0.00012081706960561217, + "loss": 0.0363, + "step": 8045 + }, + { + "epoch": 3.7863529411764705, + "grad_norm": 0.09998137322687314, + "learning_rate": 0.00012077407229604593, + "loss": 0.0395, + "step": 8046 + }, + { + "epoch": 3.7868235294117647, + "grad_norm": 0.014552124450152384, + "learning_rate": 0.00012073107748251724, + "loss": 0.0013, + "step": 8047 + }, + { + "epoch": 3.787294117647059, + "grad_norm": 0.04605625615332728, + "learning_rate": 0.00012068808516869812, + "loss": 0.0108, + "step": 8048 + }, + { + "epoch": 3.7877647058823527, + "grad_norm": 0.1003533122012289, + "learning_rate": 0.00012064509535826029, + "loss": 0.0435, + "step": 8049 + }, + { + "epoch": 3.788235294117647, + "grad_norm": 0.09380707605050367, + "learning_rate": 0.00012060210805487529, + "loss": 0.029, + "step": 8050 + }, + { + "epoch": 3.788705882352941, + "grad_norm": 0.0006811757903002939, + "learning_rate": 0.00012055912326221448, + "loss": 0.0, + "step": 8051 + }, + { + "epoch": 3.7891764705882354, + "grad_norm": 0.0655271890435591, + "learning_rate": 0.00012051614098394894, + "loss": 0.0112, + "step": 8052 + }, + { + "epoch": 3.7896470588235296, + "grad_norm": 0.030849041574438917, + "learning_rate": 0.00012047316122374957, + "loss": 0.0056, + "step": 8053 + }, + { + "epoch": 3.7901176470588234, + "grad_norm": 0.09500303840740935, + "learning_rate": 0.00012043018398528707, + "loss": 0.0355, + "step": 8054 + }, + { + "epoch": 3.7905882352941176, + "grad_norm": 0.11529759381070843, + "learning_rate": 0.0001203872092722319, + "loss": 0.0407, + "step": 8055 + }, + { + "epoch": 3.791058823529412, + "grad_norm": 0.4816178844362894, + "learning_rate": 0.00012034423708825431, + "loss": 0.1758, + "step": 8056 + }, + { + "epoch": 3.791529411764706, + "grad_norm": 0.12773436941580732, + "learning_rate": 0.00012030126743702431, + "loss": 0.053, + "step": 8057 + }, + { + "epoch": 3.792, + "grad_norm": 0.10098195807947535, + "learning_rate": 0.00012025830032221177, + "loss": 0.0349, + "step": 8058 + }, + { + "epoch": 3.792470588235294, + "grad_norm": 0.029818392080803283, + "learning_rate": 0.00012021533574748625, + "loss": 0.0038, + "step": 8059 + }, + { + "epoch": 3.7929411764705883, + "grad_norm": 0.07384869858273464, + "learning_rate": 0.00012017237371651714, + "loss": 0.0238, + "step": 8060 + }, + { + "epoch": 3.7934117647058825, + "grad_norm": 0.25571056972195333, + "learning_rate": 0.00012012941423297363, + "loss": 0.1041, + "step": 8061 + }, + { + "epoch": 3.7938823529411767, + "grad_norm": 0.08973273512080729, + "learning_rate": 0.00012008645730052468, + "loss": 0.0335, + "step": 8062 + }, + { + "epoch": 3.7943529411764705, + "grad_norm": 0.07998284522822745, + "learning_rate": 0.00012004350292283896, + "loss": 0.0286, + "step": 8063 + }, + { + "epoch": 3.7948235294117647, + "grad_norm": 0.19753100064282972, + "learning_rate": 0.00012000055110358502, + "loss": 0.0832, + "step": 8064 + }, + { + "epoch": 3.795294117647059, + "grad_norm": 0.17393382790015097, + "learning_rate": 0.00011995760184643118, + "loss": 0.1036, + "step": 8065 + }, + { + "epoch": 3.7957647058823527, + "grad_norm": 0.113595849185006, + "learning_rate": 0.00011991465515504551, + "loss": 0.0535, + "step": 8066 + }, + { + "epoch": 3.796235294117647, + "grad_norm": 0.0006247456723241141, + "learning_rate": 0.00011987171103309586, + "loss": 0.0, + "step": 8067 + }, + { + "epoch": 3.796705882352941, + "grad_norm": 0.06725105908967222, + "learning_rate": 0.00011982876948424986, + "loss": 0.0188, + "step": 8068 + }, + { + "epoch": 3.7971764705882354, + "grad_norm": 0.11261958161816593, + "learning_rate": 0.00011978583051217493, + "loss": 0.0341, + "step": 8069 + }, + { + "epoch": 3.7976470588235296, + "grad_norm": 0.06930639668219649, + "learning_rate": 0.0001197428941205383, + "loss": 0.025, + "step": 8070 + }, + { + "epoch": 3.7981176470588234, + "grad_norm": 0.16530836180474676, + "learning_rate": 0.00011969996031300687, + "loss": 0.0686, + "step": 8071 + }, + { + "epoch": 3.7985882352941176, + "grad_norm": 0.07038802186877573, + "learning_rate": 0.00011965702909324751, + "loss": 0.0237, + "step": 8072 + }, + { + "epoch": 3.799058823529412, + "grad_norm": 0.0885225466785112, + "learning_rate": 0.00011961410046492667, + "loss": 0.0294, + "step": 8073 + }, + { + "epoch": 3.799529411764706, + "grad_norm": 0.00016496350657207667, + "learning_rate": 0.0001195711744317107, + "loss": 0.0, + "step": 8074 + }, + { + "epoch": 3.8, + "grad_norm": 0.05228469685253621, + "learning_rate": 0.00011952825099726567, + "loss": 0.0148, + "step": 8075 + }, + { + "epoch": 3.800470588235294, + "grad_norm": 9.012891450512473e-06, + "learning_rate": 0.00011948533016525749, + "loss": 0.0, + "step": 8076 + }, + { + "epoch": 3.8009411764705883, + "grad_norm": 0.07196078064405875, + "learning_rate": 0.00011944241193935178, + "loss": 0.0188, + "step": 8077 + }, + { + "epoch": 3.8014117647058825, + "grad_norm": 0.18491425789411858, + "learning_rate": 0.00011939949632321395, + "loss": 0.083, + "step": 8078 + }, + { + "epoch": 3.8018823529411767, + "grad_norm": 0.07252755604839249, + "learning_rate": 0.00011935658332050925, + "loss": 0.0182, + "step": 8079 + }, + { + "epoch": 3.8023529411764705, + "grad_norm": 0.12736640834733998, + "learning_rate": 0.00011931367293490265, + "loss": 0.0325, + "step": 8080 + }, + { + "epoch": 3.8028235294117647, + "grad_norm": 0.07779686529800016, + "learning_rate": 0.00011927076517005885, + "loss": 0.0223, + "step": 8081 + }, + { + "epoch": 3.803294117647059, + "grad_norm": 0.0005542392174789026, + "learning_rate": 0.0001192278600296425, + "loss": 0.0, + "step": 8082 + }, + { + "epoch": 3.8037647058823527, + "grad_norm": 3.795256614650674e-05, + "learning_rate": 0.00011918495751731781, + "loss": 0.0, + "step": 8083 + }, + { + "epoch": 3.804235294117647, + "grad_norm": 0.08879339334848643, + "learning_rate": 0.0001191420576367489, + "loss": 0.0357, + "step": 8084 + }, + { + "epoch": 3.804705882352941, + "grad_norm": 0.003804310658529706, + "learning_rate": 0.00011909916039159958, + "loss": 0.0001, + "step": 8085 + }, + { + "epoch": 3.8051764705882354, + "grad_norm": 0.13909081053508301, + "learning_rate": 0.00011905626578553359, + "loss": 0.0629, + "step": 8086 + }, + { + "epoch": 3.8056470588235296, + "grad_norm": 0.041067806168480304, + "learning_rate": 0.00011901337382221425, + "loss": 0.0094, + "step": 8087 + }, + { + "epoch": 3.8061176470588234, + "grad_norm": 0.12620620990462617, + "learning_rate": 0.00011897048450530477, + "loss": 0.0487, + "step": 8088 + }, + { + "epoch": 3.8065882352941176, + "grad_norm": 0.09008842024823238, + "learning_rate": 0.00011892759783846812, + "loss": 0.0289, + "step": 8089 + }, + { + "epoch": 3.807058823529412, + "grad_norm": 0.1959467711541802, + "learning_rate": 0.00011888471382536704, + "loss": 0.1047, + "step": 8090 + }, + { + "epoch": 3.8075294117647056, + "grad_norm": 0.10432583596306493, + "learning_rate": 0.00011884183246966399, + "loss": 0.0447, + "step": 8091 + }, + { + "epoch": 3.808, + "grad_norm": 0.11407462316205545, + "learning_rate": 0.00011879895377502129, + "loss": 0.046, + "step": 8092 + }, + { + "epoch": 3.808470588235294, + "grad_norm": 0.08123031201166087, + "learning_rate": 0.00011875607774510099, + "loss": 0.0322, + "step": 8093 + }, + { + "epoch": 3.8089411764705883, + "grad_norm": 0.05229376122440641, + "learning_rate": 0.00011871320438356491, + "loss": 0.0109, + "step": 8094 + }, + { + "epoch": 3.8094117647058825, + "grad_norm": 0.06516196248916091, + "learning_rate": 0.0001186703336940746, + "loss": 0.0186, + "step": 8095 + }, + { + "epoch": 3.8098823529411767, + "grad_norm": 6.292051755274176e-06, + "learning_rate": 0.00011862746568029149, + "loss": 0.0, + "step": 8096 + }, + { + "epoch": 3.8103529411764705, + "grad_norm": 0.1373640801784113, + "learning_rate": 0.00011858460034587669, + "loss": 0.0752, + "step": 8097 + }, + { + "epoch": 3.8108235294117647, + "grad_norm": 0.09256336835069677, + "learning_rate": 0.0001185417376944911, + "loss": 0.0505, + "step": 8098 + }, + { + "epoch": 3.811294117647059, + "grad_norm": 0.09115765643539633, + "learning_rate": 0.00011849887772979541, + "loss": 0.0414, + "step": 8099 + }, + { + "epoch": 3.8117647058823527, + "grad_norm": 0.04650157413148482, + "learning_rate": 0.00011845602045545008, + "loss": 0.0114, + "step": 8100 + }, + { + "epoch": 3.812235294117647, + "grad_norm": 0.0749953325583172, + "learning_rate": 0.00011841316587511531, + "loss": 0.0216, + "step": 8101 + }, + { + "epoch": 3.812705882352941, + "grad_norm": 0.047536070045529406, + "learning_rate": 0.00011837031399245115, + "loss": 0.0076, + "step": 8102 + }, + { + "epoch": 3.8131764705882354, + "grad_norm": 0.10420113039581907, + "learning_rate": 0.00011832746481111728, + "loss": 0.0217, + "step": 8103 + }, + { + "epoch": 3.8136470588235296, + "grad_norm": 0.0974020967792852, + "learning_rate": 0.00011828461833477326, + "loss": 0.0394, + "step": 8104 + }, + { + "epoch": 3.8141176470588234, + "grad_norm": 8.911061449650535e-05, + "learning_rate": 0.00011824177456707844, + "loss": 0.0, + "step": 8105 + }, + { + "epoch": 3.8145882352941176, + "grad_norm": 0.1170610301712427, + "learning_rate": 0.00011819893351169183, + "loss": 0.0307, + "step": 8106 + }, + { + "epoch": 3.815058823529412, + "grad_norm": 0.1710463133509819, + "learning_rate": 0.00011815609517227228, + "loss": 0.0719, + "step": 8107 + }, + { + "epoch": 3.8155294117647056, + "grad_norm": 0.12867279211821508, + "learning_rate": 0.00011811325955247833, + "loss": 0.0731, + "step": 8108 + }, + { + "epoch": 3.816, + "grad_norm": 0.14828931316473404, + "learning_rate": 0.00011807042665596846, + "loss": 0.075, + "step": 8109 + }, + { + "epoch": 3.816470588235294, + "grad_norm": 0.1869966022974447, + "learning_rate": 0.0001180275964864008, + "loss": 0.0811, + "step": 8110 + }, + { + "epoch": 3.8169411764705883, + "grad_norm": 0.10698460056626467, + "learning_rate": 0.00011798476904743321, + "loss": 0.0518, + "step": 8111 + }, + { + "epoch": 3.8174117647058825, + "grad_norm": 0.014423243227823812, + "learning_rate": 0.00011794194434272336, + "loss": 0.0011, + "step": 8112 + }, + { + "epoch": 3.8178823529411767, + "grad_norm": 0.07635238060592396, + "learning_rate": 0.00011789912237592874, + "loss": 0.0277, + "step": 8113 + }, + { + "epoch": 3.8183529411764705, + "grad_norm": 1.5546680330834555e-05, + "learning_rate": 0.00011785630315070651, + "loss": 0.0, + "step": 8114 + }, + { + "epoch": 3.8188235294117647, + "grad_norm": 0.10839073111874752, + "learning_rate": 0.00011781348667071365, + "loss": 0.0209, + "step": 8115 + }, + { + "epoch": 3.819294117647059, + "grad_norm": 0.07306809779067448, + "learning_rate": 0.0001177706729396069, + "loss": 0.0201, + "step": 8116 + }, + { + "epoch": 3.8197647058823527, + "grad_norm": 0.18998581929460098, + "learning_rate": 0.00011772786196104276, + "loss": 0.1014, + "step": 8117 + }, + { + "epoch": 3.820235294117647, + "grad_norm": 0.1590579979207402, + "learning_rate": 0.00011768505373867753, + "loss": 0.0689, + "step": 8118 + }, + { + "epoch": 3.820705882352941, + "grad_norm": 0.16965702201364916, + "learning_rate": 0.00011764224827616717, + "loss": 0.0552, + "step": 8119 + }, + { + "epoch": 3.8211764705882354, + "grad_norm": 0.11850865307760723, + "learning_rate": 0.00011759944557716755, + "loss": 0.0562, + "step": 8120 + }, + { + "epoch": 3.8216470588235296, + "grad_norm": 0.1705268980773089, + "learning_rate": 0.00011755664564533422, + "loss": 0.0541, + "step": 8121 + }, + { + "epoch": 3.8221176470588234, + "grad_norm": 4.0017553067458774e-05, + "learning_rate": 0.00011751384848432245, + "loss": 0.0, + "step": 8122 + }, + { + "epoch": 3.8225882352941176, + "grad_norm": 0.019450173091533452, + "learning_rate": 0.00011747105409778738, + "loss": 0.0026, + "step": 8123 + }, + { + "epoch": 3.823058823529412, + "grad_norm": 0.040356474335878044, + "learning_rate": 0.00011742826248938384, + "loss": 0.0094, + "step": 8124 + }, + { + "epoch": 3.8235294117647056, + "grad_norm": 0.1217362698592575, + "learning_rate": 0.00011738547366276645, + "loss": 0.0574, + "step": 8125 + }, + { + "epoch": 3.824, + "grad_norm": 0.08354153043664231, + "learning_rate": 0.00011734268762158957, + "loss": 0.0374, + "step": 8126 + }, + { + "epoch": 3.824470588235294, + "grad_norm": 0.09817770042497612, + "learning_rate": 0.00011729990436950737, + "loss": 0.0397, + "step": 8127 + }, + { + "epoch": 3.8249411764705883, + "grad_norm": 0.009987029199843508, + "learning_rate": 0.00011725712391017374, + "loss": 0.0005, + "step": 8128 + }, + { + "epoch": 3.8254117647058825, + "grad_norm": 0.13053177510223943, + "learning_rate": 0.00011721434624724227, + "loss": 0.0706, + "step": 8129 + }, + { + "epoch": 3.8258823529411767, + "grad_norm": 0.00010364677580095877, + "learning_rate": 0.0001171715713843665, + "loss": 0.0, + "step": 8130 + }, + { + "epoch": 3.8263529411764705, + "grad_norm": 6.313873777113297e-05, + "learning_rate": 0.00011712879932519954, + "loss": 0.0, + "step": 8131 + }, + { + "epoch": 3.8268235294117647, + "grad_norm": 0.10143764006591884, + "learning_rate": 0.00011708603007339433, + "loss": 0.0336, + "step": 8132 + }, + { + "epoch": 3.827294117647059, + "grad_norm": 0.0990293909279818, + "learning_rate": 0.00011704326363260365, + "loss": 0.0345, + "step": 8133 + }, + { + "epoch": 3.8277647058823527, + "grad_norm": 0.07504139782890044, + "learning_rate": 0.00011700050000647986, + "loss": 0.0228, + "step": 8134 + }, + { + "epoch": 3.828235294117647, + "grad_norm": 0.10654380687340985, + "learning_rate": 0.00011695773919867525, + "loss": 0.0513, + "step": 8135 + }, + { + "epoch": 3.828705882352941, + "grad_norm": 0.0987804540423448, + "learning_rate": 0.00011691498121284176, + "loss": 0.041, + "step": 8136 + }, + { + "epoch": 3.8291764705882354, + "grad_norm": 6.07121250061387e-06, + "learning_rate": 0.00011687222605263118, + "loss": 0.0, + "step": 8137 + }, + { + "epoch": 3.8296470588235296, + "grad_norm": 0.09698952936507238, + "learning_rate": 0.000116829473721695, + "loss": 0.031, + "step": 8138 + }, + { + "epoch": 3.8301176470588234, + "grad_norm": 0.05287564892226539, + "learning_rate": 0.00011678672422368442, + "loss": 0.0154, + "step": 8139 + }, + { + "epoch": 3.8305882352941176, + "grad_norm": 0.07886002926468182, + "learning_rate": 0.00011674397756225053, + "loss": 0.0276, + "step": 8140 + }, + { + "epoch": 3.831058823529412, + "grad_norm": 0.05701827385230981, + "learning_rate": 0.00011670123374104409, + "loss": 0.0163, + "step": 8141 + }, + { + "epoch": 3.8315294117647056, + "grad_norm": 0.08407739163046292, + "learning_rate": 0.00011665849276371557, + "loss": 0.0399, + "step": 8142 + }, + { + "epoch": 3.832, + "grad_norm": 0.1488264243079231, + "learning_rate": 0.00011661575463391534, + "loss": 0.0757, + "step": 8143 + }, + { + "epoch": 3.832470588235294, + "grad_norm": 0.07944513503712651, + "learning_rate": 0.0001165730193552934, + "loss": 0.0255, + "step": 8144 + }, + { + "epoch": 3.8329411764705883, + "grad_norm": 2.3994053793468435e-06, + "learning_rate": 0.00011653028693149961, + "loss": 0.0, + "step": 8145 + }, + { + "epoch": 3.8334117647058825, + "grad_norm": 0.14670876951506556, + "learning_rate": 0.00011648755736618342, + "loss": 0.0561, + "step": 8146 + }, + { + "epoch": 3.8338823529411767, + "grad_norm": 0.0720838284919784, + "learning_rate": 0.00011644483066299424, + "loss": 0.0183, + "step": 8147 + }, + { + "epoch": 3.8343529411764705, + "grad_norm": 8.886676887324295e-05, + "learning_rate": 0.00011640210682558112, + "loss": 0.0, + "step": 8148 + }, + { + "epoch": 3.8348235294117647, + "grad_norm": 0.12464027543347785, + "learning_rate": 0.00011635938585759284, + "loss": 0.0512, + "step": 8149 + }, + { + "epoch": 3.835294117647059, + "grad_norm": 0.15051806369966575, + "learning_rate": 0.00011631666776267803, + "loss": 0.0771, + "step": 8150 + }, + { + "epoch": 3.8357647058823527, + "grad_norm": 0.07683726971312654, + "learning_rate": 0.00011627395254448504, + "loss": 0.0173, + "step": 8151 + }, + { + "epoch": 3.836235294117647, + "grad_norm": 0.0414401432976053, + "learning_rate": 0.00011623124020666186, + "loss": 0.0094, + "step": 8152 + }, + { + "epoch": 3.836705882352941, + "grad_norm": 0.3918962026695084, + "learning_rate": 0.00011618853075285644, + "loss": 0.0862, + "step": 8153 + }, + { + "epoch": 3.8371764705882354, + "grad_norm": 0.08594562286782356, + "learning_rate": 0.00011614582418671629, + "loss": 0.023, + "step": 8154 + }, + { + "epoch": 3.8376470588235296, + "grad_norm": 0.09490631740440988, + "learning_rate": 0.00011610312051188888, + "loss": 0.0403, + "step": 8155 + }, + { + "epoch": 3.8381176470588234, + "grad_norm": 0.06653645497705554, + "learning_rate": 0.00011606041973202123, + "loss": 0.0239, + "step": 8156 + }, + { + "epoch": 3.8385882352941176, + "grad_norm": 0.0002948737309094353, + "learning_rate": 0.00011601772185076018, + "loss": 0.0, + "step": 8157 + }, + { + "epoch": 3.839058823529412, + "grad_norm": 0.07557716522557216, + "learning_rate": 0.00011597502687175238, + "loss": 0.0242, + "step": 8158 + }, + { + "epoch": 3.8395294117647056, + "grad_norm": 0.07154464782290168, + "learning_rate": 0.0001159323347986442, + "loss": 0.0264, + "step": 8159 + }, + { + "epoch": 3.84, + "grad_norm": 0.09088528381732357, + "learning_rate": 0.00011588964563508167, + "loss": 0.0288, + "step": 8160 + }, + { + "epoch": 3.840470588235294, + "grad_norm": 0.09099257190217314, + "learning_rate": 0.00011584695938471077, + "loss": 0.0255, + "step": 8161 + }, + { + "epoch": 3.8409411764705883, + "grad_norm": 0.07903330019467648, + "learning_rate": 0.00011580427605117705, + "loss": 0.0214, + "step": 8162 + }, + { + "epoch": 3.8414117647058825, + "grad_norm": 0.11678133298143692, + "learning_rate": 0.00011576159563812589, + "loss": 0.0456, + "step": 8163 + }, + { + "epoch": 3.8418823529411767, + "grad_norm": 0.12469215820586159, + "learning_rate": 0.00011571891814920236, + "loss": 0.0197, + "step": 8164 + }, + { + "epoch": 3.8423529411764705, + "grad_norm": 0.0848757304936781, + "learning_rate": 0.00011567624358805139, + "loss": 0.0296, + "step": 8165 + }, + { + "epoch": 3.8428235294117647, + "grad_norm": 8.952881644138163e-06, + "learning_rate": 0.0001156335719583176, + "loss": 0.0, + "step": 8166 + }, + { + "epoch": 3.843294117647059, + "grad_norm": 0.11070124425878519, + "learning_rate": 0.00011559090326364527, + "loss": 0.0287, + "step": 8167 + }, + { + "epoch": 3.8437647058823528, + "grad_norm": 0.09427408732003285, + "learning_rate": 0.00011554823750767863, + "loss": 0.0433, + "step": 8168 + }, + { + "epoch": 3.844235294117647, + "grad_norm": 0.13419865172094794, + "learning_rate": 0.0001155055746940615, + "loss": 0.0513, + "step": 8169 + }, + { + "epoch": 3.844705882352941, + "grad_norm": 0.08176761294910537, + "learning_rate": 0.00011546291482643744, + "loss": 0.0292, + "step": 8170 + }, + { + "epoch": 3.8451764705882354, + "grad_norm": 0.15813530448421373, + "learning_rate": 0.00011542025790844988, + "loss": 0.0668, + "step": 8171 + }, + { + "epoch": 3.8456470588235296, + "grad_norm": 0.15584541048991501, + "learning_rate": 0.00011537760394374192, + "loss": 0.0596, + "step": 8172 + }, + { + "epoch": 3.8461176470588234, + "grad_norm": 0.10280265271158383, + "learning_rate": 0.00011533495293595637, + "loss": 0.0442, + "step": 8173 + }, + { + "epoch": 3.8465882352941176, + "grad_norm": 0.13218769514277165, + "learning_rate": 0.0001152923048887359, + "loss": 0.0414, + "step": 8174 + }, + { + "epoch": 3.847058823529412, + "grad_norm": 0.08572382104639135, + "learning_rate": 0.00011524965980572284, + "loss": 0.0246, + "step": 8175 + }, + { + "epoch": 3.8475294117647056, + "grad_norm": 0.1355264626527529, + "learning_rate": 0.00011520701769055927, + "loss": 0.0725, + "step": 8176 + }, + { + "epoch": 3.848, + "grad_norm": 0.060311020509981106, + "learning_rate": 0.00011516437854688703, + "loss": 0.0181, + "step": 8177 + }, + { + "epoch": 3.848470588235294, + "grad_norm": 0.05925205561128273, + "learning_rate": 0.00011512174237834776, + "loss": 0.0054, + "step": 8178 + }, + { + "epoch": 3.8489411764705883, + "grad_norm": 7.349843428831901e-06, + "learning_rate": 0.00011507910918858276, + "loss": 0.0, + "step": 8179 + }, + { + "epoch": 3.8494117647058825, + "grad_norm": 2.325849492967816e-05, + "learning_rate": 0.0001150364789812331, + "loss": 0.0, + "step": 8180 + }, + { + "epoch": 3.8498823529411763, + "grad_norm": 0.08662489583933253, + "learning_rate": 0.00011499385175993965, + "loss": 0.0348, + "step": 8181 + }, + { + "epoch": 3.8503529411764705, + "grad_norm": 0.14492083393935126, + "learning_rate": 0.00011495122752834296, + "loss": 0.0515, + "step": 8182 + }, + { + "epoch": 3.8508235294117648, + "grad_norm": 0.07081084488101506, + "learning_rate": 0.00011490860629008333, + "loss": 0.02, + "step": 8183 + }, + { + "epoch": 3.851294117647059, + "grad_norm": 3.416954700277361e-05, + "learning_rate": 0.00011486598804880088, + "loss": 0.0, + "step": 8184 + }, + { + "epoch": 3.8517647058823528, + "grad_norm": 0.2333139427790965, + "learning_rate": 0.00011482337280813536, + "loss": 0.0518, + "step": 8185 + }, + { + "epoch": 3.852235294117647, + "grad_norm": 0.07680892928584881, + "learning_rate": 0.00011478076057172638, + "loss": 0.0225, + "step": 8186 + }, + { + "epoch": 3.852705882352941, + "grad_norm": 0.06791194215087219, + "learning_rate": 0.00011473815134321315, + "loss": 0.0245, + "step": 8187 + }, + { + "epoch": 3.8531764705882354, + "grad_norm": 0.10098644064430048, + "learning_rate": 0.00011469554512623477, + "loss": 0.0402, + "step": 8188 + }, + { + "epoch": 3.8536470588235296, + "grad_norm": 1.9205262490054794e-05, + "learning_rate": 0.00011465294192443001, + "loss": 0.0, + "step": 8189 + }, + { + "epoch": 3.8541176470588234, + "grad_norm": 0.2402950810857929, + "learning_rate": 0.00011461034174143736, + "loss": 0.1166, + "step": 8190 + }, + { + "epoch": 3.8545882352941176, + "grad_norm": 0.06282828012566873, + "learning_rate": 0.00011456774458089515, + "loss": 0.0147, + "step": 8191 + }, + { + "epoch": 3.855058823529412, + "grad_norm": 0.04649426818616502, + "learning_rate": 0.00011452515044644132, + "loss": 0.0099, + "step": 8192 + }, + { + "epoch": 3.8555294117647056, + "grad_norm": 0.16903229254957744, + "learning_rate": 0.00011448255934171363, + "loss": 0.0696, + "step": 8193 + }, + { + "epoch": 3.856, + "grad_norm": 0.06574231287136181, + "learning_rate": 0.00011443997127034962, + "loss": 0.0159, + "step": 8194 + }, + { + "epoch": 3.856470588235294, + "grad_norm": 0.07928249733148612, + "learning_rate": 0.0001143973862359865, + "loss": 0.0273, + "step": 8195 + }, + { + "epoch": 3.8569411764705883, + "grad_norm": 0.15600408950195033, + "learning_rate": 0.0001143548042422612, + "loss": 0.0777, + "step": 8196 + }, + { + "epoch": 3.8574117647058825, + "grad_norm": 0.07223435500139837, + "learning_rate": 0.0001143122252928104, + "loss": 0.0261, + "step": 8197 + }, + { + "epoch": 3.8578823529411763, + "grad_norm": 0.1024544931767108, + "learning_rate": 0.00011426964939127063, + "loss": 0.0537, + "step": 8198 + }, + { + "epoch": 3.8583529411764705, + "grad_norm": 0.11575183291673163, + "learning_rate": 0.00011422707654127809, + "loss": 0.0494, + "step": 8199 + }, + { + "epoch": 3.8588235294117648, + "grad_norm": 0.11471302707978508, + "learning_rate": 0.00011418450674646868, + "loss": 0.0523, + "step": 8200 + }, + { + "epoch": 3.859294117647059, + "grad_norm": 0.06261139301647813, + "learning_rate": 0.00011414194001047806, + "loss": 0.0208, + "step": 8201 + }, + { + "epoch": 3.8597647058823528, + "grad_norm": 0.10313346522751295, + "learning_rate": 0.00011409937633694165, + "loss": 0.0407, + "step": 8202 + }, + { + "epoch": 3.860235294117647, + "grad_norm": 0.04418609338863883, + "learning_rate": 0.0001140568157294946, + "loss": 0.0086, + "step": 8203 + }, + { + "epoch": 3.860705882352941, + "grad_norm": 0.062276878141390754, + "learning_rate": 0.00011401425819177179, + "loss": 0.0178, + "step": 8204 + }, + { + "epoch": 3.8611764705882354, + "grad_norm": 0.0739377199182968, + "learning_rate": 0.00011397170372740782, + "loss": 0.0178, + "step": 8205 + }, + { + "epoch": 3.8616470588235297, + "grad_norm": 0.16038827214353035, + "learning_rate": 0.0001139291523400371, + "loss": 0.0226, + "step": 8206 + }, + { + "epoch": 3.8621176470588234, + "grad_norm": 0.05493566302531725, + "learning_rate": 0.00011388660403329368, + "loss": 0.0113, + "step": 8207 + }, + { + "epoch": 3.8625882352941177, + "grad_norm": 0.02026754118461213, + "learning_rate": 0.0001138440588108114, + "loss": 0.0002, + "step": 8208 + }, + { + "epoch": 3.863058823529412, + "grad_norm": 0.20002696360465708, + "learning_rate": 0.00011380151667622384, + "loss": 0.0674, + "step": 8209 + }, + { + "epoch": 3.8635294117647057, + "grad_norm": 0.11200538483079073, + "learning_rate": 0.00011375897763316433, + "loss": 0.0356, + "step": 8210 + }, + { + "epoch": 3.864, + "grad_norm": 0.12113175203933428, + "learning_rate": 0.00011371644168526586, + "loss": 0.0398, + "step": 8211 + }, + { + "epoch": 3.864470588235294, + "grad_norm": 0.10484374587595965, + "learning_rate": 0.00011367390883616126, + "loss": 0.0255, + "step": 8212 + }, + { + "epoch": 3.8649411764705883, + "grad_norm": 0.19789906591949313, + "learning_rate": 0.00011363137908948303, + "loss": 0.0827, + "step": 8213 + }, + { + "epoch": 3.8654117647058825, + "grad_norm": 0.0005741439783736157, + "learning_rate": 0.00011358885244886342, + "loss": 0.0, + "step": 8214 + }, + { + "epoch": 3.8658823529411763, + "grad_norm": 0.003955094434381428, + "learning_rate": 0.00011354632891793437, + "loss": 0.0, + "step": 8215 + }, + { + "epoch": 3.8663529411764705, + "grad_norm": 0.04122621935254908, + "learning_rate": 0.00011350380850032766, + "loss": 0.0068, + "step": 8216 + }, + { + "epoch": 3.8668235294117648, + "grad_norm": 0.031814941735531274, + "learning_rate": 0.00011346129119967473, + "loss": 0.0061, + "step": 8217 + }, + { + "epoch": 3.8672941176470585, + "grad_norm": 0.0229552666070933, + "learning_rate": 0.0001134187770196067, + "loss": 0.0032, + "step": 8218 + }, + { + "epoch": 3.8677647058823528, + "grad_norm": 0.10728076023064653, + "learning_rate": 0.00011337626596375458, + "loss": 0.0413, + "step": 8219 + }, + { + "epoch": 3.868235294117647, + "grad_norm": 0.03350944371694588, + "learning_rate": 0.000113333758035749, + "loss": 0.0004, + "step": 8220 + }, + { + "epoch": 3.868705882352941, + "grad_norm": 0.00035744245036152983, + "learning_rate": 0.00011329125323922031, + "loss": 0.0, + "step": 8221 + }, + { + "epoch": 3.8691764705882354, + "grad_norm": 0.15667439735387945, + "learning_rate": 0.00011324875157779865, + "loss": 0.0348, + "step": 8222 + }, + { + "epoch": 3.8696470588235297, + "grad_norm": 0.11640681905688276, + "learning_rate": 0.00011320625305511388, + "loss": 0.0491, + "step": 8223 + }, + { + "epoch": 3.8701176470588234, + "grad_norm": 0.00010315126091595077, + "learning_rate": 0.00011316375767479554, + "loss": 0.0, + "step": 8224 + }, + { + "epoch": 3.8705882352941177, + "grad_norm": 2.6407555444527876e-05, + "learning_rate": 0.00011312126544047302, + "loss": 0.0, + "step": 8225 + }, + { + "epoch": 3.871058823529412, + "grad_norm": 0.12363327872221061, + "learning_rate": 0.0001130787763557753, + "loss": 0.052, + "step": 8226 + }, + { + "epoch": 3.8715294117647057, + "grad_norm": 3.601585544629254e-05, + "learning_rate": 0.00011303629042433118, + "loss": 0.0, + "step": 8227 + }, + { + "epoch": 3.872, + "grad_norm": 0.1443634278816907, + "learning_rate": 0.00011299380764976915, + "loss": 0.0768, + "step": 8228 + }, + { + "epoch": 3.872470588235294, + "grad_norm": 0.144223854377411, + "learning_rate": 0.00011295132803571747, + "loss": 0.0554, + "step": 8229 + }, + { + "epoch": 3.8729411764705883, + "grad_norm": 0.061021536144299844, + "learning_rate": 0.00011290885158580411, + "loss": 0.0218, + "step": 8230 + }, + { + "epoch": 3.8734117647058826, + "grad_norm": 0.06130067919287852, + "learning_rate": 0.00011286637830365672, + "loss": 0.0122, + "step": 8231 + }, + { + "epoch": 3.8738823529411763, + "grad_norm": 0.06884848899488995, + "learning_rate": 0.0001128239081929028, + "loss": 0.0157, + "step": 8232 + }, + { + "epoch": 3.8743529411764706, + "grad_norm": 0.07877757591498881, + "learning_rate": 0.00011278144125716946, + "loss": 0.0233, + "step": 8233 + }, + { + "epoch": 3.8748235294117648, + "grad_norm": 8.574105573454134e-05, + "learning_rate": 0.00011273897750008358, + "loss": 0.0, + "step": 8234 + }, + { + "epoch": 3.8752941176470586, + "grad_norm": 0.06618753064114988, + "learning_rate": 0.0001126965169252718, + "loss": 0.0233, + "step": 8235 + }, + { + "epoch": 3.875764705882353, + "grad_norm": 0.1591361798576309, + "learning_rate": 0.00011265405953636044, + "loss": 0.086, + "step": 8236 + }, + { + "epoch": 3.876235294117647, + "grad_norm": 0.055612710920680195, + "learning_rate": 0.0001126116053369756, + "loss": 0.0102, + "step": 8237 + }, + { + "epoch": 3.8767058823529412, + "grad_norm": 0.1480633890299921, + "learning_rate": 0.000112569154330743, + "loss": 0.0419, + "step": 8238 + }, + { + "epoch": 3.8771764705882354, + "grad_norm": 0.15256115146493315, + "learning_rate": 0.00011252670652128825, + "loss": 0.0782, + "step": 8239 + }, + { + "epoch": 3.8776470588235297, + "grad_norm": 0.2013208375073922, + "learning_rate": 0.00011248426191223658, + "loss": 0.1083, + "step": 8240 + }, + { + "epoch": 3.8781176470588234, + "grad_norm": 0.08176768849274312, + "learning_rate": 0.00011244182050721293, + "loss": 0.028, + "step": 8241 + }, + { + "epoch": 3.8785882352941177, + "grad_norm": 0.09572241290123698, + "learning_rate": 0.00011239938230984198, + "loss": 0.0361, + "step": 8242 + }, + { + "epoch": 3.879058823529412, + "grad_norm": 0.0006618447085418808, + "learning_rate": 0.00011235694732374829, + "loss": 0.0, + "step": 8243 + }, + { + "epoch": 3.8795294117647057, + "grad_norm": 4.5895602163098376e-05, + "learning_rate": 0.00011231451555255592, + "loss": 0.0, + "step": 8244 + }, + { + "epoch": 3.88, + "grad_norm": 0.02477097477169735, + "learning_rate": 0.00011227208699988879, + "loss": 0.0035, + "step": 8245 + }, + { + "epoch": 3.880470588235294, + "grad_norm": 0.06468830864507517, + "learning_rate": 0.00011222966166937044, + "loss": 0.0181, + "step": 8246 + }, + { + "epoch": 3.8809411764705883, + "grad_norm": 0.17453145386104485, + "learning_rate": 0.00011218723956462427, + "loss": 0.0484, + "step": 8247 + }, + { + "epoch": 3.8814117647058826, + "grad_norm": 0.16008227945028342, + "learning_rate": 0.00011214482068927332, + "loss": 0.0296, + "step": 8248 + }, + { + "epoch": 3.8818823529411763, + "grad_norm": 0.06828349569016642, + "learning_rate": 0.00011210240504694033, + "loss": 0.014, + "step": 8249 + }, + { + "epoch": 3.8823529411764706, + "grad_norm": 0.11621676209329854, + "learning_rate": 0.00011205999264124786, + "loss": 0.0289, + "step": 8250 + }, + { + "epoch": 3.882823529411765, + "grad_norm": 0.09523568401352825, + "learning_rate": 0.00011201758347581814, + "loss": 0.0248, + "step": 8251 + }, + { + "epoch": 3.8832941176470586, + "grad_norm": 0.185694319098424, + "learning_rate": 0.00011197517755427303, + "loss": 0.0827, + "step": 8252 + }, + { + "epoch": 3.883764705882353, + "grad_norm": 0.00017463323070868997, + "learning_rate": 0.00011193277488023432, + "loss": 0.0, + "step": 8253 + }, + { + "epoch": 3.884235294117647, + "grad_norm": 0.04968254967956194, + "learning_rate": 0.00011189037545732335, + "loss": 0.0119, + "step": 8254 + }, + { + "epoch": 3.8847058823529412, + "grad_norm": 0.13586894291252638, + "learning_rate": 0.00011184797928916124, + "loss": 0.045, + "step": 8255 + }, + { + "epoch": 3.8851764705882355, + "grad_norm": 0.05215572454335708, + "learning_rate": 0.0001118055863793688, + "loss": 0.0186, + "step": 8256 + }, + { + "epoch": 3.8856470588235297, + "grad_norm": 0.11744936224932431, + "learning_rate": 0.00011176319673156666, + "loss": 0.0471, + "step": 8257 + }, + { + "epoch": 3.8861176470588235, + "grad_norm": 0.1579802119742822, + "learning_rate": 0.00011172081034937507, + "loss": 0.0422, + "step": 8258 + }, + { + "epoch": 3.8865882352941177, + "grad_norm": 0.00018821631876552404, + "learning_rate": 0.00011167842723641398, + "loss": 0.0, + "step": 8259 + }, + { + "epoch": 3.887058823529412, + "grad_norm": 0.0006021337972338723, + "learning_rate": 0.00011163604739630324, + "loss": 0.0, + "step": 8260 + }, + { + "epoch": 3.8875294117647057, + "grad_norm": 0.11845513851732674, + "learning_rate": 0.00011159367083266219, + "loss": 0.0424, + "step": 8261 + }, + { + "epoch": 3.888, + "grad_norm": 0.03833866825668094, + "learning_rate": 0.00011155129754911002, + "loss": 0.0024, + "step": 8262 + }, + { + "epoch": 3.888470588235294, + "grad_norm": 0.040690776893506045, + "learning_rate": 0.00011150892754926568, + "loss": 0.0052, + "step": 8263 + }, + { + "epoch": 3.8889411764705883, + "grad_norm": 0.1625512983533014, + "learning_rate": 0.00011146656083674768, + "loss": 0.0757, + "step": 8264 + }, + { + "epoch": 3.8894117647058826, + "grad_norm": 0.06518580153943762, + "learning_rate": 0.00011142419741517443, + "loss": 0.021, + "step": 8265 + }, + { + "epoch": 3.8898823529411763, + "grad_norm": 0.09024003640925767, + "learning_rate": 0.00011138183728816389, + "loss": 0.0235, + "step": 8266 + }, + { + "epoch": 3.8903529411764706, + "grad_norm": 0.002794237608087876, + "learning_rate": 0.0001113394804593339, + "loss": 0.0, + "step": 8267 + }, + { + "epoch": 3.890823529411765, + "grad_norm": 0.11944113510475297, + "learning_rate": 0.00011129712693230191, + "loss": 0.0413, + "step": 8268 + }, + { + "epoch": 3.8912941176470586, + "grad_norm": 0.09090395467844219, + "learning_rate": 0.00011125477671068506, + "loss": 0.0208, + "step": 8269 + }, + { + "epoch": 3.891764705882353, + "grad_norm": 0.11044167712557082, + "learning_rate": 0.0001112124297981004, + "loss": 0.0379, + "step": 8270 + }, + { + "epoch": 3.892235294117647, + "grad_norm": 0.10087102155747983, + "learning_rate": 0.00011117008619816445, + "loss": 0.0382, + "step": 8271 + }, + { + "epoch": 3.8927058823529412, + "grad_norm": 0.09090876375110694, + "learning_rate": 0.00011112774591449356, + "loss": 0.0297, + "step": 8272 + }, + { + "epoch": 3.8931764705882355, + "grad_norm": 0.11273173135729896, + "learning_rate": 0.00011108540895070388, + "loss": 0.0068, + "step": 8273 + }, + { + "epoch": 3.8936470588235297, + "grad_norm": 0.0870950859063967, + "learning_rate": 0.00011104307531041117, + "loss": 0.0321, + "step": 8274 + }, + { + "epoch": 3.8941176470588235, + "grad_norm": 0.16757987176498437, + "learning_rate": 0.0001110007449972309, + "loss": 0.0653, + "step": 8275 + }, + { + "epoch": 3.8945882352941177, + "grad_norm": 0.14789372359003472, + "learning_rate": 0.00011095841801477826, + "loss": 0.0727, + "step": 8276 + }, + { + "epoch": 3.895058823529412, + "grad_norm": 0.0009450873198237404, + "learning_rate": 0.00011091609436666825, + "loss": 0.0, + "step": 8277 + }, + { + "epoch": 3.8955294117647057, + "grad_norm": 0.13338244643842379, + "learning_rate": 0.00011087377405651549, + "loss": 0.0519, + "step": 8278 + }, + { + "epoch": 3.896, + "grad_norm": 0.1902147479780184, + "learning_rate": 0.00011083145708793431, + "loss": 0.091, + "step": 8279 + }, + { + "epoch": 3.896470588235294, + "grad_norm": 0.1122802798138599, + "learning_rate": 0.00011078914346453883, + "loss": 0.0444, + "step": 8280 + }, + { + "epoch": 3.8969411764705884, + "grad_norm": 0.0002738626908303041, + "learning_rate": 0.00011074683318994286, + "loss": 0.0, + "step": 8281 + }, + { + "epoch": 3.8974117647058826, + "grad_norm": 0.3212497963109424, + "learning_rate": 0.00011070452626775981, + "loss": 0.1043, + "step": 8282 + }, + { + "epoch": 3.8978823529411764, + "grad_norm": 0.0660265145992734, + "learning_rate": 0.000110662222701603, + "loss": 0.0199, + "step": 8283 + }, + { + "epoch": 3.8983529411764706, + "grad_norm": 0.08452988169075308, + "learning_rate": 0.00011061992249508533, + "loss": 0.0245, + "step": 8284 + }, + { + "epoch": 3.898823529411765, + "grad_norm": 0.1400652977046068, + "learning_rate": 0.0001105776256518194, + "loss": 0.0813, + "step": 8285 + }, + { + "epoch": 3.8992941176470586, + "grad_norm": 0.12792015846949464, + "learning_rate": 0.0001105353321754176, + "loss": 0.0584, + "step": 8286 + }, + { + "epoch": 3.899764705882353, + "grad_norm": 0.0873797985907659, + "learning_rate": 0.000110493042069492, + "loss": 0.0305, + "step": 8287 + }, + { + "epoch": 3.900235294117647, + "grad_norm": 0.12528828586995622, + "learning_rate": 0.00011045075533765444, + "loss": 0.048, + "step": 8288 + }, + { + "epoch": 3.9007058823529412, + "grad_norm": 0.10048779842368862, + "learning_rate": 0.00011040847198351635, + "loss": 0.0138, + "step": 8289 + }, + { + "epoch": 3.9011764705882355, + "grad_norm": 0.18122999373252668, + "learning_rate": 0.00011036619201068892, + "loss": 0.0799, + "step": 8290 + }, + { + "epoch": 3.9016470588235292, + "grad_norm": 0.0935318989699674, + "learning_rate": 0.00011032391542278313, + "loss": 0.0298, + "step": 8291 + }, + { + "epoch": 3.9021176470588235, + "grad_norm": 0.21296532165079476, + "learning_rate": 0.00011028164222340957, + "loss": 0.0942, + "step": 8292 + }, + { + "epoch": 3.9025882352941177, + "grad_norm": 0.0020243478605332954, + "learning_rate": 0.00011023937241617853, + "loss": 0.0, + "step": 8293 + }, + { + "epoch": 3.903058823529412, + "grad_norm": 0.0576600028785034, + "learning_rate": 0.00011019710600470016, + "loss": 0.0141, + "step": 8294 + }, + { + "epoch": 3.9035294117647057, + "grad_norm": 0.015900150562718703, + "learning_rate": 0.00011015484299258417, + "loss": 0.0013, + "step": 8295 + }, + { + "epoch": 3.904, + "grad_norm": 0.12613005750078363, + "learning_rate": 0.00011011258338344002, + "loss": 0.0432, + "step": 8296 + }, + { + "epoch": 3.904470588235294, + "grad_norm": 0.10431923843129247, + "learning_rate": 0.00011007032718087688, + "loss": 0.0328, + "step": 8297 + }, + { + "epoch": 3.9049411764705884, + "grad_norm": 0.053544142881373755, + "learning_rate": 0.00011002807438850367, + "loss": 0.0043, + "step": 8298 + }, + { + "epoch": 3.9054117647058826, + "grad_norm": 0.09861380351256029, + "learning_rate": 0.00010998582500992897, + "loss": 0.0314, + "step": 8299 + }, + { + "epoch": 3.9058823529411764, + "grad_norm": 0.16244637538647796, + "learning_rate": 0.00010994357904876106, + "loss": 0.0711, + "step": 8300 + }, + { + "epoch": 3.9063529411764706, + "grad_norm": 0.14246406703418263, + "learning_rate": 0.000109901336508608, + "loss": 0.0239, + "step": 8301 + }, + { + "epoch": 3.906823529411765, + "grad_norm": 3.4363764621114426e-05, + "learning_rate": 0.00010985909739307748, + "loss": 0.0, + "step": 8302 + }, + { + "epoch": 3.9072941176470586, + "grad_norm": 0.09222171078326, + "learning_rate": 0.0001098168617057769, + "loss": 0.0319, + "step": 8303 + }, + { + "epoch": 3.907764705882353, + "grad_norm": 0.04173130912688382, + "learning_rate": 0.00010977462945031347, + "loss": 0.0102, + "step": 8304 + }, + { + "epoch": 3.908235294117647, + "grad_norm": 0.3282813961782495, + "learning_rate": 0.00010973240063029398, + "loss": 0.1105, + "step": 8305 + }, + { + "epoch": 3.9087058823529413, + "grad_norm": 0.037869802285929476, + "learning_rate": 0.000109690175249325, + "loss": 0.0065, + "step": 8306 + }, + { + "epoch": 3.9091764705882355, + "grad_norm": 0.00010893408573768771, + "learning_rate": 0.00010964795331101273, + "loss": 0.0, + "step": 8307 + }, + { + "epoch": 3.9096470588235293, + "grad_norm": 0.15252313958953778, + "learning_rate": 0.00010960573481896322, + "loss": 0.0392, + "step": 8308 + }, + { + "epoch": 3.9101176470588235, + "grad_norm": 0.10367577454637338, + "learning_rate": 0.00010956351977678206, + "loss": 0.0307, + "step": 8309 + }, + { + "epoch": 3.9105882352941177, + "grad_norm": 1.5475388555042325e-05, + "learning_rate": 0.00010952130818807465, + "loss": 0.0, + "step": 8310 + }, + { + "epoch": 3.911058823529412, + "grad_norm": 0.13519291437627134, + "learning_rate": 0.00010947910005644608, + "loss": 0.0526, + "step": 8311 + }, + { + "epoch": 3.9115294117647057, + "grad_norm": 0.0003527881431196002, + "learning_rate": 0.00010943689538550111, + "loss": 0.0, + "step": 8312 + }, + { + "epoch": 3.912, + "grad_norm": 0.09696636633224363, + "learning_rate": 0.00010939469417884424, + "loss": 0.034, + "step": 8313 + }, + { + "epoch": 3.912470588235294, + "grad_norm": 0.10350475101473533, + "learning_rate": 0.00010935249644007966, + "loss": 0.0453, + "step": 8314 + }, + { + "epoch": 3.9129411764705884, + "grad_norm": 0.06353797821877902, + "learning_rate": 0.00010931030217281125, + "loss": 0.0124, + "step": 8315 + }, + { + "epoch": 3.9134117647058826, + "grad_norm": 0.08261068527180733, + "learning_rate": 0.00010926811138064264, + "loss": 0.0269, + "step": 8316 + }, + { + "epoch": 3.9138823529411764, + "grad_norm": 0.04689218193688836, + "learning_rate": 0.00010922592406717706, + "loss": 0.0113, + "step": 8317 + }, + { + "epoch": 3.9143529411764706, + "grad_norm": 0.09638218623916121, + "learning_rate": 0.0001091837402360176, + "loss": 0.0516, + "step": 8318 + }, + { + "epoch": 3.914823529411765, + "grad_norm": 0.11020676801271774, + "learning_rate": 0.00010914155989076692, + "loss": 0.0492, + "step": 8319 + }, + { + "epoch": 3.9152941176470586, + "grad_norm": 0.0006921787944493422, + "learning_rate": 0.00010909938303502741, + "loss": 0.0, + "step": 8320 + }, + { + "epoch": 3.915764705882353, + "grad_norm": 0.07413199959861305, + "learning_rate": 0.00010905720967240122, + "loss": 0.0179, + "step": 8321 + }, + { + "epoch": 3.916235294117647, + "grad_norm": 0.13138212275727726, + "learning_rate": 0.00010901503980649017, + "loss": 0.0585, + "step": 8322 + }, + { + "epoch": 3.9167058823529413, + "grad_norm": 0.16924386430259122, + "learning_rate": 0.0001089728734408957, + "loss": 0.0653, + "step": 8323 + }, + { + "epoch": 3.9171764705882355, + "grad_norm": 0.19312290207133054, + "learning_rate": 0.00010893071057921911, + "loss": 0.028, + "step": 8324 + }, + { + "epoch": 3.9176470588235293, + "grad_norm": 0.09278365244772083, + "learning_rate": 0.0001088885512250613, + "loss": 0.0325, + "step": 8325 + }, + { + "epoch": 3.9181176470588235, + "grad_norm": 0.05073843509805297, + "learning_rate": 0.00010884639538202285, + "loss": 0.0101, + "step": 8326 + }, + { + "epoch": 3.9185882352941177, + "grad_norm": 0.08301116864808482, + "learning_rate": 0.00010880424305370406, + "loss": 0.0233, + "step": 8327 + }, + { + "epoch": 3.919058823529412, + "grad_norm": 0.057686921591114854, + "learning_rate": 0.00010876209424370502, + "loss": 0.0104, + "step": 8328 + }, + { + "epoch": 3.9195294117647057, + "grad_norm": 0.17227769303571988, + "learning_rate": 0.00010871994895562538, + "loss": 0.0482, + "step": 8329 + }, + { + "epoch": 3.92, + "grad_norm": 0.07502670188602417, + "learning_rate": 0.00010867780719306458, + "loss": 0.028, + "step": 8330 + }, + { + "epoch": 3.920470588235294, + "grad_norm": 0.12906588913916234, + "learning_rate": 0.00010863566895962166, + "loss": 0.0588, + "step": 8331 + }, + { + "epoch": 3.9209411764705884, + "grad_norm": 0.011373949641799426, + "learning_rate": 0.0001085935342588956, + "loss": 0.0001, + "step": 8332 + }, + { + "epoch": 3.9214117647058826, + "grad_norm": 0.12124207738282958, + "learning_rate": 0.00010855140309448479, + "loss": 0.023, + "step": 8333 + }, + { + "epoch": 3.9218823529411764, + "grad_norm": 0.009201741834440603, + "learning_rate": 0.00010850927546998744, + "loss": 0.0001, + "step": 8334 + }, + { + "epoch": 3.9223529411764706, + "grad_norm": 0.10938241955614962, + "learning_rate": 0.00010846715138900148, + "loss": 0.036, + "step": 8335 + }, + { + "epoch": 3.922823529411765, + "grad_norm": 0.009149828710735008, + "learning_rate": 0.00010842503085512453, + "loss": 0.0001, + "step": 8336 + }, + { + "epoch": 3.9232941176470586, + "grad_norm": 0.001763283628512743, + "learning_rate": 0.00010838291387195387, + "loss": 0.0, + "step": 8337 + }, + { + "epoch": 3.923764705882353, + "grad_norm": 0.0006653938077695796, + "learning_rate": 0.00010834080044308646, + "loss": 0.0, + "step": 8338 + }, + { + "epoch": 3.924235294117647, + "grad_norm": 0.06137521309213455, + "learning_rate": 0.00010829869057211906, + "loss": 0.009, + "step": 8339 + }, + { + "epoch": 3.9247058823529413, + "grad_norm": 0.11647554238452458, + "learning_rate": 0.00010825658426264803, + "loss": 0.0471, + "step": 8340 + }, + { + "epoch": 3.9251764705882355, + "grad_norm": 0.05858925998971238, + "learning_rate": 0.00010821448151826941, + "loss": 0.0139, + "step": 8341 + }, + { + "epoch": 3.9256470588235293, + "grad_norm": 0.1882115727561178, + "learning_rate": 0.00010817238234257907, + "loss": 0.061, + "step": 8342 + }, + { + "epoch": 3.9261176470588235, + "grad_norm": 0.06618952312868069, + "learning_rate": 0.00010813028673917245, + "loss": 0.0119, + "step": 8343 + }, + { + "epoch": 3.9265882352941177, + "grad_norm": 0.1631644578747136, + "learning_rate": 0.0001080881947116447, + "loss": 0.0885, + "step": 8344 + }, + { + "epoch": 3.9270588235294115, + "grad_norm": 1.4907246521946941e-05, + "learning_rate": 0.00010804610626359068, + "loss": 0.0, + "step": 8345 + }, + { + "epoch": 3.9275294117647057, + "grad_norm": 0.033050669294983515, + "learning_rate": 0.000108004021398605, + "loss": 0.004, + "step": 8346 + }, + { + "epoch": 3.928, + "grad_norm": 0.026409196461302107, + "learning_rate": 0.00010796194012028185, + "loss": 0.0039, + "step": 8347 + }, + { + "epoch": 3.928470588235294, + "grad_norm": 0.17852802420678637, + "learning_rate": 0.00010791986243221521, + "loss": 0.0555, + "step": 8348 + }, + { + "epoch": 3.9289411764705884, + "grad_norm": 0.07363835903484496, + "learning_rate": 0.00010787778833799875, + "loss": 0.0203, + "step": 8349 + }, + { + "epoch": 3.9294117647058826, + "grad_norm": 0.08135010155331707, + "learning_rate": 0.00010783571784122577, + "loss": 0.0189, + "step": 8350 + }, + { + "epoch": 3.9298823529411764, + "grad_norm": 0.14097928200330997, + "learning_rate": 0.00010779365094548926, + "loss": 0.0443, + "step": 8351 + }, + { + "epoch": 3.9303529411764706, + "grad_norm": 0.039517599718045957, + "learning_rate": 0.00010775158765438202, + "loss": 0.0073, + "step": 8352 + }, + { + "epoch": 3.930823529411765, + "grad_norm": 6.106259987607555e-05, + "learning_rate": 0.0001077095279714964, + "loss": 0.0, + "step": 8353 + }, + { + "epoch": 3.9312941176470586, + "grad_norm": 0.07109990123507927, + "learning_rate": 0.00010766747190042452, + "loss": 0.0059, + "step": 8354 + }, + { + "epoch": 3.931764705882353, + "grad_norm": 0.08641955909663891, + "learning_rate": 0.00010762541944475821, + "loss": 0.0161, + "step": 8355 + }, + { + "epoch": 3.932235294117647, + "grad_norm": 0.11325545622103912, + "learning_rate": 0.0001075833706080889, + "loss": 0.0405, + "step": 8356 + }, + { + "epoch": 3.9327058823529413, + "grad_norm": 0.11504051076159982, + "learning_rate": 0.00010754132539400782, + "loss": 0.0501, + "step": 8357 + }, + { + "epoch": 3.9331764705882355, + "grad_norm": 0.16831191651864255, + "learning_rate": 0.00010749928380610577, + "loss": 0.0809, + "step": 8358 + }, + { + "epoch": 3.9336470588235293, + "grad_norm": 0.2439425576322929, + "learning_rate": 0.00010745724584797338, + "loss": 0.0861, + "step": 8359 + }, + { + "epoch": 3.9341176470588235, + "grad_norm": 0.10559479910229316, + "learning_rate": 0.00010741521152320088, + "loss": 0.0213, + "step": 8360 + }, + { + "epoch": 3.9345882352941177, + "grad_norm": 0.1062933646643645, + "learning_rate": 0.00010737318083537813, + "loss": 0.0195, + "step": 8361 + }, + { + "epoch": 3.9350588235294115, + "grad_norm": 0.0926134376592837, + "learning_rate": 0.00010733115378809485, + "loss": 0.0164, + "step": 8362 + }, + { + "epoch": 3.9355294117647057, + "grad_norm": 0.03601407667084887, + "learning_rate": 0.00010728913038494038, + "loss": 0.0056, + "step": 8363 + }, + { + "epoch": 3.936, + "grad_norm": 0.15543718414798416, + "learning_rate": 0.00010724711062950358, + "loss": 0.0561, + "step": 8364 + }, + { + "epoch": 3.936470588235294, + "grad_norm": 0.06937882890002411, + "learning_rate": 0.00010720509452537329, + "loss": 0.015, + "step": 8365 + }, + { + "epoch": 3.9369411764705884, + "grad_norm": 0.12347825601808621, + "learning_rate": 0.00010716308207613784, + "loss": 0.0329, + "step": 8366 + }, + { + "epoch": 3.9374117647058826, + "grad_norm": 0.0001491890819018638, + "learning_rate": 0.00010712107328538528, + "loss": 0.0, + "step": 8367 + }, + { + "epoch": 3.9378823529411764, + "grad_norm": 0.1088378223483605, + "learning_rate": 0.00010707906815670337, + "loss": 0.036, + "step": 8368 + }, + { + "epoch": 3.9383529411764706, + "grad_norm": 0.05037280082418643, + "learning_rate": 0.00010703706669367956, + "loss": 0.0094, + "step": 8369 + }, + { + "epoch": 3.938823529411765, + "grad_norm": 0.0983833993911197, + "learning_rate": 0.00010699506889990102, + "loss": 0.0121, + "step": 8370 + }, + { + "epoch": 3.9392941176470586, + "grad_norm": 0.002150901697310025, + "learning_rate": 0.00010695307477895449, + "loss": 0.0, + "step": 8371 + }, + { + "epoch": 3.939764705882353, + "grad_norm": 0.03200885397466619, + "learning_rate": 0.00010691108433442656, + "loss": 0.004, + "step": 8372 + }, + { + "epoch": 3.940235294117647, + "grad_norm": 0.13490076881929258, + "learning_rate": 0.00010686909756990337, + "loss": 0.0518, + "step": 8373 + }, + { + "epoch": 3.9407058823529413, + "grad_norm": 0.11418622627508934, + "learning_rate": 0.00010682711448897077, + "loss": 0.0259, + "step": 8374 + }, + { + "epoch": 3.9411764705882355, + "grad_norm": 0.2018178521908675, + "learning_rate": 0.00010678513509521435, + "loss": 0.0868, + "step": 8375 + }, + { + "epoch": 3.9416470588235293, + "grad_norm": 0.10949348424462335, + "learning_rate": 0.00010674315939221937, + "loss": 0.0319, + "step": 8376 + }, + { + "epoch": 3.9421176470588235, + "grad_norm": 0.11384442597219628, + "learning_rate": 0.00010670118738357078, + "loss": 0.0389, + "step": 8377 + }, + { + "epoch": 3.9425882352941177, + "grad_norm": 0.16465474045636672, + "learning_rate": 0.00010665921907285317, + "loss": 0.0226, + "step": 8378 + }, + { + "epoch": 3.9430588235294115, + "grad_norm": 0.04917243891991978, + "learning_rate": 0.0001066172544636508, + "loss": 0.0082, + "step": 8379 + }, + { + "epoch": 3.9435294117647057, + "grad_norm": 0.12245763131606495, + "learning_rate": 0.00010657529355954771, + "loss": 0.0362, + "step": 8380 + }, + { + "epoch": 3.944, + "grad_norm": 0.08072775795886324, + "learning_rate": 0.00010653333636412758, + "loss": 0.0164, + "step": 8381 + }, + { + "epoch": 3.944470588235294, + "grad_norm": 0.07095560675764738, + "learning_rate": 0.00010649138288097366, + "loss": 0.0146, + "step": 8382 + }, + { + "epoch": 3.9449411764705884, + "grad_norm": 0.026522647528985754, + "learning_rate": 0.00010644943311366911, + "loss": 0.0024, + "step": 8383 + }, + { + "epoch": 3.9454117647058826, + "grad_norm": 0.07085575631522432, + "learning_rate": 0.00010640748706579658, + "loss": 0.0118, + "step": 8384 + }, + { + "epoch": 3.9458823529411764, + "grad_norm": 0.05997255340078366, + "learning_rate": 0.00010636554474093846, + "loss": 0.0103, + "step": 8385 + }, + { + "epoch": 3.9463529411764706, + "grad_norm": 0.11832678801639841, + "learning_rate": 0.00010632360614267683, + "loss": 0.0403, + "step": 8386 + }, + { + "epoch": 3.946823529411765, + "grad_norm": 0.11058382051525548, + "learning_rate": 0.0001062816712745935, + "loss": 0.0307, + "step": 8387 + }, + { + "epoch": 3.9472941176470586, + "grad_norm": 0.08020732354993226, + "learning_rate": 0.00010623974014026986, + "loss": 0.0211, + "step": 8388 + }, + { + "epoch": 3.947764705882353, + "grad_norm": 0.07384866644905858, + "learning_rate": 0.00010619781274328703, + "loss": 0.0183, + "step": 8389 + }, + { + "epoch": 3.948235294117647, + "grad_norm": 0.05253505069853845, + "learning_rate": 0.00010615588908722588, + "loss": 0.0126, + "step": 8390 + }, + { + "epoch": 3.9487058823529413, + "grad_norm": 0.07934755618792656, + "learning_rate": 0.00010611396917566685, + "loss": 0.0279, + "step": 8391 + }, + { + "epoch": 3.9491764705882355, + "grad_norm": 0.03923825023256322, + "learning_rate": 0.00010607205301219008, + "loss": 0.0032, + "step": 8392 + }, + { + "epoch": 3.9496470588235293, + "grad_norm": 0.07906956236489127, + "learning_rate": 0.00010603014060037549, + "loss": 0.0143, + "step": 8393 + }, + { + "epoch": 3.9501176470588235, + "grad_norm": 0.18529127731643466, + "learning_rate": 0.00010598823194380256, + "loss": 0.0717, + "step": 8394 + }, + { + "epoch": 3.9505882352941177, + "grad_norm": 0.178271257091716, + "learning_rate": 0.0001059463270460505, + "loss": 0.0728, + "step": 8395 + }, + { + "epoch": 3.9510588235294115, + "grad_norm": 0.12663416721064927, + "learning_rate": 0.00010590442591069816, + "loss": 0.0414, + "step": 8396 + }, + { + "epoch": 3.9515294117647057, + "grad_norm": 0.05363505023117792, + "learning_rate": 0.00010586252854132415, + "loss": 0.0129, + "step": 8397 + }, + { + "epoch": 3.952, + "grad_norm": 0.09911321864043098, + "learning_rate": 0.0001058206349415067, + "loss": 0.0376, + "step": 8398 + }, + { + "epoch": 3.952470588235294, + "grad_norm": 0.055517799362307205, + "learning_rate": 0.00010577874511482371, + "loss": 0.0063, + "step": 8399 + }, + { + "epoch": 3.9529411764705884, + "grad_norm": 0.0003473247412896221, + "learning_rate": 0.00010573685906485282, + "loss": 0.0, + "step": 8400 + }, + { + "epoch": 3.9534117647058826, + "grad_norm": 0.07280676908320595, + "learning_rate": 0.00010569497679517126, + "loss": 0.0175, + "step": 8401 + }, + { + "epoch": 3.9538823529411764, + "grad_norm": 0.03440231470942449, + "learning_rate": 0.00010565309830935599, + "loss": 0.0035, + "step": 8402 + }, + { + "epoch": 3.9543529411764706, + "grad_norm": 0.09061543918547554, + "learning_rate": 0.00010561122361098365, + "loss": 0.0336, + "step": 8403 + }, + { + "epoch": 3.954823529411765, + "grad_norm": 0.07700667144126384, + "learning_rate": 0.00010556935270363055, + "loss": 0.0203, + "step": 8404 + }, + { + "epoch": 3.9552941176470586, + "grad_norm": 0.15027915661530428, + "learning_rate": 0.00010552748559087261, + "loss": 0.0671, + "step": 8405 + }, + { + "epoch": 3.955764705882353, + "grad_norm": 0.000667814332837261, + "learning_rate": 0.00010548562227628559, + "loss": 0.0, + "step": 8406 + }, + { + "epoch": 3.956235294117647, + "grad_norm": 0.005093202867432266, + "learning_rate": 0.00010544376276344477, + "loss": 0.0001, + "step": 8407 + }, + { + "epoch": 3.9567058823529413, + "grad_norm": 0.1272994982225479, + "learning_rate": 0.00010540190705592516, + "loss": 0.0532, + "step": 8408 + }, + { + "epoch": 3.9571764705882355, + "grad_norm": 0.02485560119083181, + "learning_rate": 0.00010536005515730141, + "loss": 0.0022, + "step": 8409 + }, + { + "epoch": 3.9576470588235293, + "grad_norm": 0.11261240242772474, + "learning_rate": 0.00010531820707114793, + "loss": 0.0258, + "step": 8410 + }, + { + "epoch": 3.9581176470588235, + "grad_norm": 0.0040665125946583395, + "learning_rate": 0.00010527636280103872, + "loss": 0.0, + "step": 8411 + }, + { + "epoch": 3.9585882352941177, + "grad_norm": 0.055706164391103714, + "learning_rate": 0.00010523452235054746, + "loss": 0.0139, + "step": 8412 + }, + { + "epoch": 3.9590588235294115, + "grad_norm": 0.07362040243929065, + "learning_rate": 0.00010519268572324763, + "loss": 0.0205, + "step": 8413 + }, + { + "epoch": 3.9595294117647057, + "grad_norm": 0.16752758703114512, + "learning_rate": 0.00010515085292271219, + "loss": 0.058, + "step": 8414 + }, + { + "epoch": 3.96, + "grad_norm": 0.023325457228150402, + "learning_rate": 0.00010510902395251387, + "loss": 0.003, + "step": 8415 + }, + { + "epoch": 3.960470588235294, + "grad_norm": 0.12589658526254505, + "learning_rate": 0.00010506719881622514, + "loss": 0.0593, + "step": 8416 + }, + { + "epoch": 3.9609411764705884, + "grad_norm": 0.11238732386857485, + "learning_rate": 0.00010502537751741801, + "loss": 0.0397, + "step": 8417 + }, + { + "epoch": 3.961411764705882, + "grad_norm": 0.03157095496335117, + "learning_rate": 0.00010498356005966422, + "loss": 0.0032, + "step": 8418 + }, + { + "epoch": 3.9618823529411764, + "grad_norm": 0.0025894770336173773, + "learning_rate": 0.00010494174644653518, + "loss": 0.0, + "step": 8419 + }, + { + "epoch": 3.9623529411764706, + "grad_norm": 0.17908482061113346, + "learning_rate": 0.000104899936681602, + "loss": 0.0711, + "step": 8420 + }, + { + "epoch": 3.962823529411765, + "grad_norm": 0.1678943170370805, + "learning_rate": 0.0001048581307684355, + "loss": 0.0623, + "step": 8421 + }, + { + "epoch": 3.9632941176470586, + "grad_norm": 0.1742354825772918, + "learning_rate": 0.00010481632871060605, + "loss": 0.0485, + "step": 8422 + }, + { + "epoch": 3.963764705882353, + "grad_norm": 0.10771984512103198, + "learning_rate": 0.0001047745305116837, + "loss": 0.0266, + "step": 8423 + }, + { + "epoch": 3.964235294117647, + "grad_norm": 0.1329765349007004, + "learning_rate": 0.00010473273617523831, + "loss": 0.0633, + "step": 8424 + }, + { + "epoch": 3.9647058823529413, + "grad_norm": 0.09986642777311733, + "learning_rate": 0.00010469094570483928, + "loss": 0.02, + "step": 8425 + }, + { + "epoch": 3.9651764705882355, + "grad_norm": 0.04155475689337542, + "learning_rate": 0.00010464915910405576, + "loss": 0.0081, + "step": 8426 + }, + { + "epoch": 3.9656470588235293, + "grad_norm": 4.051018014851306e-05, + "learning_rate": 0.00010460737637645645, + "loss": 0.0, + "step": 8427 + }, + { + "epoch": 3.9661176470588235, + "grad_norm": 0.10256043425670633, + "learning_rate": 0.00010456559752560987, + "loss": 0.0445, + "step": 8428 + }, + { + "epoch": 3.9665882352941177, + "grad_norm": 5.6839334002230484e-05, + "learning_rate": 0.00010452382255508412, + "loss": 0.0, + "step": 8429 + }, + { + "epoch": 3.9670588235294115, + "grad_norm": 0.0774688756676679, + "learning_rate": 0.00010448205146844696, + "loss": 0.0142, + "step": 8430 + }, + { + "epoch": 3.9675294117647057, + "grad_norm": 0.09340218169026294, + "learning_rate": 0.00010444028426926588, + "loss": 0.0332, + "step": 8431 + }, + { + "epoch": 3.968, + "grad_norm": 0.09287276294200336, + "learning_rate": 0.000104398520961108, + "loss": 0.0318, + "step": 8432 + }, + { + "epoch": 3.968470588235294, + "grad_norm": 0.10453529014713989, + "learning_rate": 0.00010435676154754009, + "loss": 0.0247, + "step": 8433 + }, + { + "epoch": 3.9689411764705884, + "grad_norm": 0.06324584712439386, + "learning_rate": 0.00010431500603212863, + "loss": 0.0139, + "step": 8434 + }, + { + "epoch": 3.969411764705882, + "grad_norm": 0.00900906338125592, + "learning_rate": 0.00010427325441843976, + "loss": 0.0003, + "step": 8435 + }, + { + "epoch": 3.9698823529411764, + "grad_norm": 0.0001572592274200756, + "learning_rate": 0.00010423150671003924, + "loss": 0.0, + "step": 8436 + }, + { + "epoch": 3.9703529411764706, + "grad_norm": 0.10530795775843817, + "learning_rate": 0.00010418976291049251, + "loss": 0.0367, + "step": 8437 + }, + { + "epoch": 3.970823529411765, + "grad_norm": 0.04781232517532583, + "learning_rate": 0.00010414802302336474, + "loss": 0.0096, + "step": 8438 + }, + { + "epoch": 3.9712941176470586, + "grad_norm": 0.13655363864844083, + "learning_rate": 0.00010410628705222073, + "loss": 0.0468, + "step": 8439 + }, + { + "epoch": 3.971764705882353, + "grad_norm": 0.1436016954725754, + "learning_rate": 0.00010406455500062487, + "loss": 0.05, + "step": 8440 + }, + { + "epoch": 3.972235294117647, + "grad_norm": 0.07501607195628485, + "learning_rate": 0.00010402282687214137, + "loss": 0.0153, + "step": 8441 + }, + { + "epoch": 3.9727058823529413, + "grad_norm": 0.10213387645892506, + "learning_rate": 0.00010398110267033396, + "loss": 0.0304, + "step": 8442 + }, + { + "epoch": 3.9731764705882355, + "grad_norm": 0.14636043723495293, + "learning_rate": 0.00010393938239876608, + "loss": 0.0634, + "step": 8443 + }, + { + "epoch": 3.9736470588235293, + "grad_norm": 0.07534267147781527, + "learning_rate": 0.0001038976660610009, + "loss": 0.0259, + "step": 8444 + }, + { + "epoch": 3.9741176470588235, + "grad_norm": 0.22411888541846547, + "learning_rate": 0.00010385595366060116, + "loss": 0.0797, + "step": 8445 + }, + { + "epoch": 3.9745882352941178, + "grad_norm": 0.152629540094638, + "learning_rate": 0.0001038142452011293, + "loss": 0.0533, + "step": 8446 + }, + { + "epoch": 3.9750588235294115, + "grad_norm": 0.11447550331468087, + "learning_rate": 0.00010377254068614743, + "loss": 0.046, + "step": 8447 + }, + { + "epoch": 3.9755294117647058, + "grad_norm": 0.27877899395946415, + "learning_rate": 0.00010373084011921736, + "loss": 0.0906, + "step": 8448 + }, + { + "epoch": 3.976, + "grad_norm": 0.086371073126526, + "learning_rate": 0.00010368914350390048, + "loss": 0.0063, + "step": 8449 + }, + { + "epoch": 3.976470588235294, + "grad_norm": 0.07001846018142771, + "learning_rate": 0.0001036474508437579, + "loss": 0.0101, + "step": 8450 + }, + { + "epoch": 3.9769411764705884, + "grad_norm": 0.023419359934249862, + "learning_rate": 0.00010360576214235036, + "loss": 0.0016, + "step": 8451 + }, + { + "epoch": 3.977411764705882, + "grad_norm": 0.18264373718234816, + "learning_rate": 0.00010356407740323832, + "loss": 0.0486, + "step": 8452 + }, + { + "epoch": 3.9778823529411764, + "grad_norm": 0.1394935243718984, + "learning_rate": 0.00010352239662998182, + "loss": 0.046, + "step": 8453 + }, + { + "epoch": 3.9783529411764706, + "grad_norm": 0.09424855495234087, + "learning_rate": 0.00010348071982614063, + "loss": 0.0182, + "step": 8454 + }, + { + "epoch": 3.978823529411765, + "grad_norm": 0.12293929484494563, + "learning_rate": 0.00010343904699527415, + "loss": 0.0377, + "step": 8455 + }, + { + "epoch": 3.9792941176470586, + "grad_norm": 0.00011679469260043313, + "learning_rate": 0.00010339737814094144, + "loss": 0.0, + "step": 8456 + }, + { + "epoch": 3.979764705882353, + "grad_norm": 0.1028560382215545, + "learning_rate": 0.00010335571326670117, + "loss": 0.0258, + "step": 8457 + }, + { + "epoch": 3.980235294117647, + "grad_norm": 0.09289069394888756, + "learning_rate": 0.00010331405237611182, + "loss": 0.0249, + "step": 8458 + }, + { + "epoch": 3.9807058823529413, + "grad_norm": 0.07192413168189322, + "learning_rate": 0.00010327239547273138, + "loss": 0.0183, + "step": 8459 + }, + { + "epoch": 3.9811764705882355, + "grad_norm": 0.151903148320546, + "learning_rate": 0.00010323074256011756, + "loss": 0.065, + "step": 8460 + }, + { + "epoch": 3.9816470588235293, + "grad_norm": 0.04409697398362677, + "learning_rate": 0.00010318909364182774, + "loss": 0.0063, + "step": 8461 + }, + { + "epoch": 3.9821176470588235, + "grad_norm": 0.10590278053622405, + "learning_rate": 0.00010314744872141893, + "loss": 0.021, + "step": 8462 + }, + { + "epoch": 3.9825882352941178, + "grad_norm": 0.1267132009651941, + "learning_rate": 0.00010310580780244779, + "loss": 0.0429, + "step": 8463 + }, + { + "epoch": 3.9830588235294115, + "grad_norm": 0.14600819156209244, + "learning_rate": 0.00010306417088847067, + "loss": 0.0371, + "step": 8464 + }, + { + "epoch": 3.9835294117647058, + "grad_norm": 0.21137808436902183, + "learning_rate": 0.00010302253798304356, + "loss": 0.0871, + "step": 8465 + }, + { + "epoch": 3.984, + "grad_norm": 0.0564920782847441, + "learning_rate": 0.00010298090908972217, + "loss": 0.0116, + "step": 8466 + }, + { + "epoch": 3.984470588235294, + "grad_norm": 0.00035622997223291, + "learning_rate": 0.00010293928421206179, + "loss": 0.0, + "step": 8467 + }, + { + "epoch": 3.9849411764705884, + "grad_norm": 0.16160632841454325, + "learning_rate": 0.00010289766335361733, + "loss": 0.0518, + "step": 8468 + }, + { + "epoch": 3.985411764705882, + "grad_norm": 0.0006378665177613702, + "learning_rate": 0.00010285604651794347, + "loss": 0.0, + "step": 8469 + }, + { + "epoch": 3.9858823529411764, + "grad_norm": 0.00037360791419933296, + "learning_rate": 0.00010281443370859449, + "loss": 0.0, + "step": 8470 + }, + { + "epoch": 3.9863529411764707, + "grad_norm": 0.00038029526697294424, + "learning_rate": 0.00010277282492912429, + "loss": 0.0, + "step": 8471 + }, + { + "epoch": 3.9868235294117644, + "grad_norm": 0.2045679894039005, + "learning_rate": 0.0001027312201830865, + "loss": 0.0775, + "step": 8472 + }, + { + "epoch": 3.9872941176470587, + "grad_norm": 0.10566028365666637, + "learning_rate": 0.00010268961947403437, + "loss": 0.0326, + "step": 8473 + }, + { + "epoch": 3.987764705882353, + "grad_norm": 0.1487251850034798, + "learning_rate": 0.00010264802280552077, + "loss": 0.0445, + "step": 8474 + }, + { + "epoch": 3.988235294117647, + "grad_norm": 0.13148172182976878, + "learning_rate": 0.0001026064301810983, + "loss": 0.0327, + "step": 8475 + }, + { + "epoch": 3.9887058823529413, + "grad_norm": 0.12011989652698943, + "learning_rate": 0.00010256484160431917, + "loss": 0.0259, + "step": 8476 + }, + { + "epoch": 3.9891764705882355, + "grad_norm": 3.634309027415781e-05, + "learning_rate": 0.00010252325707873523, + "loss": 0.0, + "step": 8477 + }, + { + "epoch": 3.9896470588235293, + "grad_norm": 0.09926605808290577, + "learning_rate": 0.00010248167660789797, + "loss": 0.0387, + "step": 8478 + }, + { + "epoch": 3.9901176470588235, + "grad_norm": 0.0007391041075208336, + "learning_rate": 0.00010244010019535864, + "loss": 0.0, + "step": 8479 + }, + { + "epoch": 3.9905882352941178, + "grad_norm": 0.07762565560118437, + "learning_rate": 0.00010239852784466803, + "loss": 0.0207, + "step": 8480 + }, + { + "epoch": 3.9910588235294115, + "grad_norm": 0.06703641779874744, + "learning_rate": 0.00010235695955937661, + "loss": 0.0136, + "step": 8481 + }, + { + "epoch": 3.9915294117647058, + "grad_norm": 0.054789406374239465, + "learning_rate": 0.00010231539534303454, + "loss": 0.0144, + "step": 8482 + }, + { + "epoch": 3.992, + "grad_norm": 0.07507939826468783, + "learning_rate": 0.00010227383519919161, + "loss": 0.0093, + "step": 8483 + }, + { + "epoch": 3.992470588235294, + "grad_norm": 0.16783890424151998, + "learning_rate": 0.00010223227913139722, + "loss": 0.045, + "step": 8484 + }, + { + "epoch": 3.9929411764705884, + "grad_norm": 0.015983038460497957, + "learning_rate": 0.0001021907271432005, + "loss": 0.0013, + "step": 8485 + }, + { + "epoch": 3.993411764705882, + "grad_norm": 0.06643649956471345, + "learning_rate": 0.00010214917923815022, + "loss": 0.0148, + "step": 8486 + }, + { + "epoch": 3.9938823529411764, + "grad_norm": 0.0878812804093988, + "learning_rate": 0.00010210763541979473, + "loss": 0.0283, + "step": 8487 + }, + { + "epoch": 3.9943529411764707, + "grad_norm": 1.9730596237251773e-05, + "learning_rate": 0.00010206609569168204, + "loss": 0.0, + "step": 8488 + }, + { + "epoch": 3.9948235294117644, + "grad_norm": 0.1230258809424398, + "learning_rate": 0.00010202456005735993, + "loss": 0.0356, + "step": 8489 + }, + { + "epoch": 3.9952941176470587, + "grad_norm": 0.06973996991587265, + "learning_rate": 0.00010198302852037573, + "loss": 0.0087, + "step": 8490 + }, + { + "epoch": 3.995764705882353, + "grad_norm": 0.10477962826453796, + "learning_rate": 0.00010194150108427637, + "loss": 0.0468, + "step": 8491 + }, + { + "epoch": 3.996235294117647, + "grad_norm": 0.06896750223365182, + "learning_rate": 0.00010189997775260857, + "loss": 0.0035, + "step": 8492 + }, + { + "epoch": 3.9967058823529413, + "grad_norm": 8.557085476390085e-06, + "learning_rate": 0.00010185845852891858, + "loss": 0.0, + "step": 8493 + }, + { + "epoch": 3.9971764705882356, + "grad_norm": 0.143672075934897, + "learning_rate": 0.00010181694341675235, + "loss": 0.052, + "step": 8494 + }, + { + "epoch": 3.9976470588235293, + "grad_norm": 0.10204780061248078, + "learning_rate": 0.00010177543241965555, + "loss": 0.0158, + "step": 8495 + }, + { + "epoch": 3.9981176470588236, + "grad_norm": 0.07416099701853279, + "learning_rate": 0.00010173392554117333, + "loss": 0.0276, + "step": 8496 + }, + { + "epoch": 3.9985882352941178, + "grad_norm": 0.18201966008733653, + "learning_rate": 0.00010169242278485063, + "loss": 0.0311, + "step": 8497 + }, + { + "epoch": 3.9990588235294116, + "grad_norm": 0.06767952679084234, + "learning_rate": 0.00010165092415423193, + "loss": 0.0128, + "step": 8498 + }, + { + "epoch": 3.9995294117647058, + "grad_norm": 0.05573339812520215, + "learning_rate": 0.00010160942965286148, + "loss": 0.0083, + "step": 8499 + }, + { + "epoch": 4.0, + "grad_norm": 0.1007066824809765, + "learning_rate": 0.0001015679392842831, + "loss": 0.039, + "step": 8500 + }, + { + "epoch": 4.0, + "eval_loss": 0.03484344482421875, + "eval_runtime": 59.3385, + "eval_samples_per_second": 12.639, + "eval_steps_per_second": 6.32, + "step": 8500 + }, + { + "epoch": 4.000470588235294, + "grad_norm": 2.8417607952065056e-05, + "learning_rate": 0.00010152645305204022, + "loss": 0.0, + "step": 8501 + }, + { + "epoch": 4.0009411764705884, + "grad_norm": 0.10756027014413584, + "learning_rate": 0.00010148497095967606, + "loss": 0.0252, + "step": 8502 + }, + { + "epoch": 4.001411764705883, + "grad_norm": 0.12363949581421824, + "learning_rate": 0.00010144349301073334, + "loss": 0.0409, + "step": 8503 + }, + { + "epoch": 4.001882352941177, + "grad_norm": 0.11151222700747472, + "learning_rate": 0.00010140201920875445, + "loss": 0.0247, + "step": 8504 + }, + { + "epoch": 4.00235294117647, + "grad_norm": 0.06789652527312612, + "learning_rate": 0.00010136054955728154, + "loss": 0.0107, + "step": 8505 + }, + { + "epoch": 4.0028235294117644, + "grad_norm": 0.11878911714404135, + "learning_rate": 0.00010131908405985625, + "loss": 0.0174, + "step": 8506 + }, + { + "epoch": 4.003294117647059, + "grad_norm": 0.04441757436513361, + "learning_rate": 0.00010127762272001996, + "loss": 0.0058, + "step": 8507 + }, + { + "epoch": 4.003764705882353, + "grad_norm": 8.035616144595426e-05, + "learning_rate": 0.00010123616554131363, + "loss": 0.0, + "step": 8508 + }, + { + "epoch": 4.004235294117647, + "grad_norm": 0.00019988362405700668, + "learning_rate": 0.00010119471252727796, + "loss": 0.0, + "step": 8509 + }, + { + "epoch": 4.004705882352941, + "grad_norm": 0.0007803299758189957, + "learning_rate": 0.00010115326368145327, + "loss": 0.0, + "step": 8510 + }, + { + "epoch": 4.005176470588236, + "grad_norm": 0.0954443753945327, + "learning_rate": 0.00010111181900737945, + "loss": 0.0326, + "step": 8511 + }, + { + "epoch": 4.00564705882353, + "grad_norm": 0.04083278193538026, + "learning_rate": 0.00010107037850859605, + "loss": 0.0035, + "step": 8512 + }, + { + "epoch": 4.006117647058823, + "grad_norm": 0.02717770608008402, + "learning_rate": 0.00010102894218864235, + "loss": 0.0026, + "step": 8513 + }, + { + "epoch": 4.006588235294117, + "grad_norm": 0.1078792686071838, + "learning_rate": 0.00010098751005105719, + "loss": 0.0436, + "step": 8514 + }, + { + "epoch": 4.007058823529412, + "grad_norm": 0.1365696496096252, + "learning_rate": 0.0001009460820993791, + "loss": 0.0316, + "step": 8515 + }, + { + "epoch": 4.007529411764706, + "grad_norm": 0.0778536865229515, + "learning_rate": 0.00010090465833714616, + "loss": 0.0118, + "step": 8516 + }, + { + "epoch": 4.008, + "grad_norm": 7.525051340908036e-05, + "learning_rate": 0.00010086323876789621, + "loss": 0.0, + "step": 8517 + }, + { + "epoch": 4.008470588235294, + "grad_norm": 7.30180251441519e-05, + "learning_rate": 0.00010082182339516674, + "loss": 0.0, + "step": 8518 + }, + { + "epoch": 4.0089411764705885, + "grad_norm": 0.11144782874130212, + "learning_rate": 0.00010078041222249469, + "loss": 0.033, + "step": 8519 + }, + { + "epoch": 4.009411764705883, + "grad_norm": 0.11562833018063164, + "learning_rate": 0.00010073900525341692, + "loss": 0.0312, + "step": 8520 + }, + { + "epoch": 4.009882352941177, + "grad_norm": 0.13969457741684513, + "learning_rate": 0.00010069760249146972, + "loss": 0.0213, + "step": 8521 + }, + { + "epoch": 4.01035294117647, + "grad_norm": 0.11919228661387932, + "learning_rate": 0.00010065620394018904, + "loss": 0.0337, + "step": 8522 + }, + { + "epoch": 4.0108235294117645, + "grad_norm": 0.19603221215274777, + "learning_rate": 0.00010061480960311061, + "loss": 0.0399, + "step": 8523 + }, + { + "epoch": 4.011294117647059, + "grad_norm": 0.16690031121293605, + "learning_rate": 0.0001005734194837697, + "loss": 0.075, + "step": 8524 + }, + { + "epoch": 4.011764705882353, + "grad_norm": 0.06961288095789463, + "learning_rate": 0.00010053203358570115, + "loss": 0.0112, + "step": 8525 + }, + { + "epoch": 4.012235294117647, + "grad_norm": 0.07243402337052529, + "learning_rate": 0.00010049065191243961, + "loss": 0.0078, + "step": 8526 + }, + { + "epoch": 4.012705882352941, + "grad_norm": 0.10979762447360883, + "learning_rate": 0.00010044927446751923, + "loss": 0.0294, + "step": 8527 + }, + { + "epoch": 4.013176470588236, + "grad_norm": 0.1447836538502959, + "learning_rate": 0.00010040790125447386, + "loss": 0.0588, + "step": 8528 + }, + { + "epoch": 4.01364705882353, + "grad_norm": 0.058576865957220094, + "learning_rate": 0.00010036653227683694, + "loss": 0.0161, + "step": 8529 + }, + { + "epoch": 4.014117647058823, + "grad_norm": 1.5754889353061805e-05, + "learning_rate": 0.00010032516753814166, + "loss": 0.0, + "step": 8530 + }, + { + "epoch": 4.014588235294117, + "grad_norm": 0.023023803935528376, + "learning_rate": 0.0001002838070419207, + "loss": 0.0013, + "step": 8531 + }, + { + "epoch": 4.015058823529412, + "grad_norm": 2.8560734067075312e-06, + "learning_rate": 0.00010024245079170645, + "loss": 0.0, + "step": 8532 + }, + { + "epoch": 4.015529411764706, + "grad_norm": 0.14542952582038304, + "learning_rate": 0.000100201098791031, + "loss": 0.0586, + "step": 8533 + }, + { + "epoch": 4.016, + "grad_norm": 0.12482602091179074, + "learning_rate": 0.00010015975104342598, + "loss": 0.0373, + "step": 8534 + }, + { + "epoch": 4.016470588235294, + "grad_norm": 1.2950765916479351e-05, + "learning_rate": 0.00010011840755242266, + "loss": 0.0, + "step": 8535 + }, + { + "epoch": 4.0169411764705885, + "grad_norm": 0.10257037744266084, + "learning_rate": 0.00010007706832155201, + "loss": 0.0332, + "step": 8536 + }, + { + "epoch": 4.017411764705883, + "grad_norm": 9.688133328016736e-06, + "learning_rate": 0.00010003573335434462, + "loss": 0.0, + "step": 8537 + }, + { + "epoch": 4.017882352941177, + "grad_norm": 2.0350994902873888e-05, + "learning_rate": 9.999440265433065e-05, + "loss": 0.0, + "step": 8538 + }, + { + "epoch": 4.01835294117647, + "grad_norm": 0.01683336558314614, + "learning_rate": 9.995307622503993e-05, + "loss": 0.0013, + "step": 8539 + }, + { + "epoch": 4.0188235294117645, + "grad_norm": 0.085874180187912, + "learning_rate": 9.991175407000206e-05, + "loss": 0.0249, + "step": 8540 + }, + { + "epoch": 4.019294117647059, + "grad_norm": 0.09817493306033861, + "learning_rate": 9.987043619274604e-05, + "loss": 0.0331, + "step": 8541 + }, + { + "epoch": 4.019764705882353, + "grad_norm": 9.801462988902657e-06, + "learning_rate": 9.982912259680063e-05, + "loss": 0.0, + "step": 8542 + }, + { + "epoch": 4.020235294117647, + "grad_norm": 0.06553876809875839, + "learning_rate": 9.978781328569427e-05, + "loss": 0.0117, + "step": 8543 + }, + { + "epoch": 4.020705882352941, + "grad_norm": 0.06134048893325843, + "learning_rate": 9.974650826295493e-05, + "loss": 0.0149, + "step": 8544 + }, + { + "epoch": 4.021176470588236, + "grad_norm": 0.05200431099738357, + "learning_rate": 9.970520753211027e-05, + "loss": 0.0113, + "step": 8545 + }, + { + "epoch": 4.02164705882353, + "grad_norm": 0.14802224428115093, + "learning_rate": 9.966391109668758e-05, + "loss": 0.063, + "step": 8546 + }, + { + "epoch": 4.022117647058823, + "grad_norm": 0.1321867415056869, + "learning_rate": 9.96226189602138e-05, + "loss": 0.0462, + "step": 8547 + }, + { + "epoch": 4.022588235294117, + "grad_norm": 0.08316012851447743, + "learning_rate": 9.958133112621547e-05, + "loss": 0.0223, + "step": 8548 + }, + { + "epoch": 4.023058823529412, + "grad_norm": 0.037145250688466684, + "learning_rate": 9.954004759821871e-05, + "loss": 0.004, + "step": 8549 + }, + { + "epoch": 4.023529411764706, + "grad_norm": 0.13046647871855105, + "learning_rate": 9.949876837974944e-05, + "loss": 0.0601, + "step": 8550 + }, + { + "epoch": 4.024, + "grad_norm": 0.04370745846510334, + "learning_rate": 9.945749347433304e-05, + "loss": 0.0069, + "step": 8551 + }, + { + "epoch": 4.024470588235294, + "grad_norm": 1.9589978685798576e-05, + "learning_rate": 9.941622288549458e-05, + "loss": 0.0, + "step": 8552 + }, + { + "epoch": 4.0249411764705885, + "grad_norm": 4.772659475320799e-07, + "learning_rate": 9.937495661675878e-05, + "loss": 0.0, + "step": 8553 + }, + { + "epoch": 4.025411764705883, + "grad_norm": 0.15943238487731526, + "learning_rate": 9.933369467165004e-05, + "loss": 0.0279, + "step": 8554 + }, + { + "epoch": 4.025882352941177, + "grad_norm": 0.12296543998130863, + "learning_rate": 9.929243705369232e-05, + "loss": 0.0405, + "step": 8555 + }, + { + "epoch": 4.02635294117647, + "grad_norm": 4.014526863113109e-05, + "learning_rate": 9.925118376640918e-05, + "loss": 0.0, + "step": 8556 + }, + { + "epoch": 4.0268235294117645, + "grad_norm": 0.11188236742835661, + "learning_rate": 9.920993481332383e-05, + "loss": 0.0459, + "step": 8557 + }, + { + "epoch": 4.027294117647059, + "grad_norm": 0.03795048007134726, + "learning_rate": 9.916869019795922e-05, + "loss": 0.0048, + "step": 8558 + }, + { + "epoch": 4.027764705882353, + "grad_norm": 0.14333865694739356, + "learning_rate": 9.912744992383777e-05, + "loss": 0.0753, + "step": 8559 + }, + { + "epoch": 4.028235294117647, + "grad_norm": 3.5766638535024284e-07, + "learning_rate": 9.90862139944816e-05, + "loss": 0.0, + "step": 8560 + }, + { + "epoch": 4.028705882352941, + "grad_norm": 1.346380285864279e-06, + "learning_rate": 9.904498241341252e-05, + "loss": 0.0, + "step": 8561 + }, + { + "epoch": 4.029176470588236, + "grad_norm": 0.06397901947626092, + "learning_rate": 9.900375518415188e-05, + "loss": 0.0133, + "step": 8562 + }, + { + "epoch": 4.02964705882353, + "grad_norm": 0.08541878676683622, + "learning_rate": 9.896253231022066e-05, + "loss": 0.0086, + "step": 8563 + }, + { + "epoch": 4.030117647058823, + "grad_norm": 0.13021949916413247, + "learning_rate": 9.892131379513953e-05, + "loss": 0.0498, + "step": 8564 + }, + { + "epoch": 4.030588235294117, + "grad_norm": 0.11114802753344202, + "learning_rate": 9.888009964242876e-05, + "loss": 0.0572, + "step": 8565 + }, + { + "epoch": 4.031058823529412, + "grad_norm": 0.059545589936595714, + "learning_rate": 9.883888985560822e-05, + "loss": 0.0137, + "step": 8566 + }, + { + "epoch": 4.031529411764706, + "grad_norm": 0.09671146821087805, + "learning_rate": 9.879768443819737e-05, + "loss": 0.0191, + "step": 8567 + }, + { + "epoch": 4.032, + "grad_norm": 0.11547001696632489, + "learning_rate": 9.875648339371546e-05, + "loss": 0.0405, + "step": 8568 + }, + { + "epoch": 4.032470588235294, + "grad_norm": 0.10414665389610445, + "learning_rate": 9.871528672568124e-05, + "loss": 0.0363, + "step": 8569 + }, + { + "epoch": 4.0329411764705885, + "grad_norm": 0.09361744330009793, + "learning_rate": 9.867409443761304e-05, + "loss": 0.0195, + "step": 8570 + }, + { + "epoch": 4.033411764705883, + "grad_norm": 0.09585911271928645, + "learning_rate": 9.863290653302894e-05, + "loss": 0.0269, + "step": 8571 + }, + { + "epoch": 4.033882352941176, + "grad_norm": 1.2505295146135288e-05, + "learning_rate": 9.859172301544659e-05, + "loss": 0.0, + "step": 8572 + }, + { + "epoch": 4.03435294117647, + "grad_norm": 0.2634925449044399, + "learning_rate": 9.855054388838322e-05, + "loss": 0.0691, + "step": 8573 + }, + { + "epoch": 4.0348235294117645, + "grad_norm": 0.024523101061137465, + "learning_rate": 9.85093691553558e-05, + "loss": 0.0018, + "step": 8574 + }, + { + "epoch": 4.035294117647059, + "grad_norm": 0.131479821880853, + "learning_rate": 9.84681988198808e-05, + "loss": 0.0549, + "step": 8575 + }, + { + "epoch": 4.035764705882353, + "grad_norm": 0.05222723945746317, + "learning_rate": 9.842703288547438e-05, + "loss": 0.0108, + "step": 8576 + }, + { + "epoch": 4.036235294117647, + "grad_norm": 2.2941930677415657e-05, + "learning_rate": 9.838587135565229e-05, + "loss": 0.0, + "step": 8577 + }, + { + "epoch": 4.036705882352941, + "grad_norm": 0.22203020250279634, + "learning_rate": 9.834471423393e-05, + "loss": 0.0845, + "step": 8578 + }, + { + "epoch": 4.037176470588236, + "grad_norm": 0.08953091586404155, + "learning_rate": 9.830356152382245e-05, + "loss": 0.0154, + "step": 8579 + }, + { + "epoch": 4.03764705882353, + "grad_norm": 0.1180559483879748, + "learning_rate": 9.826241322884432e-05, + "loss": 0.024, + "step": 8580 + }, + { + "epoch": 4.038117647058823, + "grad_norm": 0.11866354288840317, + "learning_rate": 9.822126935250987e-05, + "loss": 0.0441, + "step": 8581 + }, + { + "epoch": 4.038588235294117, + "grad_norm": 0.11633555450855768, + "learning_rate": 9.818012989833301e-05, + "loss": 0.036, + "step": 8582 + }, + { + "epoch": 4.039058823529412, + "grad_norm": 0.1627096925365646, + "learning_rate": 9.81389948698272e-05, + "loss": 0.0417, + "step": 8583 + }, + { + "epoch": 4.039529411764706, + "grad_norm": 0.2250297951214914, + "learning_rate": 9.809786427050562e-05, + "loss": 0.0852, + "step": 8584 + }, + { + "epoch": 4.04, + "grad_norm": 0.06566185236712446, + "learning_rate": 9.805673810388103e-05, + "loss": 0.0132, + "step": 8585 + }, + { + "epoch": 4.040470588235294, + "grad_norm": 0.12808255803637444, + "learning_rate": 9.801561637346575e-05, + "loss": 0.0475, + "step": 8586 + }, + { + "epoch": 4.0409411764705885, + "grad_norm": 0.05879130612457567, + "learning_rate": 9.797449908277184e-05, + "loss": 0.0161, + "step": 8587 + }, + { + "epoch": 4.041411764705883, + "grad_norm": 0.12517784086293762, + "learning_rate": 9.793338623531089e-05, + "loss": 0.0363, + "step": 8588 + }, + { + "epoch": 4.041882352941176, + "grad_norm": 0.0012303234800508025, + "learning_rate": 9.789227783459415e-05, + "loss": 0.0, + "step": 8589 + }, + { + "epoch": 4.04235294117647, + "grad_norm": 0.08684892294793761, + "learning_rate": 9.785117388413244e-05, + "loss": 0.0176, + "step": 8590 + }, + { + "epoch": 4.0428235294117645, + "grad_norm": 0.06234983599521278, + "learning_rate": 9.781007438743629e-05, + "loss": 0.015, + "step": 8591 + }, + { + "epoch": 4.043294117647059, + "grad_norm": 0.1023595812018188, + "learning_rate": 9.776897934801577e-05, + "loss": 0.0262, + "step": 8592 + }, + { + "epoch": 4.043764705882353, + "grad_norm": 0.1583801804624424, + "learning_rate": 9.77278887693806e-05, + "loss": 0.0461, + "step": 8593 + }, + { + "epoch": 4.044235294117647, + "grad_norm": 0.07658519064850886, + "learning_rate": 9.768680265504013e-05, + "loss": 0.0137, + "step": 8594 + }, + { + "epoch": 4.044705882352941, + "grad_norm": 0.06720058632415217, + "learning_rate": 9.764572100850331e-05, + "loss": 0.0125, + "step": 8595 + }, + { + "epoch": 4.045176470588236, + "grad_norm": 0.053917001412533916, + "learning_rate": 9.760464383327868e-05, + "loss": 0.007, + "step": 8596 + }, + { + "epoch": 4.04564705882353, + "grad_norm": 0.12967860347600385, + "learning_rate": 9.756357113287449e-05, + "loss": 0.0391, + "step": 8597 + }, + { + "epoch": 4.046117647058823, + "grad_norm": 0.1312189410380971, + "learning_rate": 9.75225029107985e-05, + "loss": 0.0401, + "step": 8598 + }, + { + "epoch": 4.046588235294117, + "grad_norm": 0.10099334857155405, + "learning_rate": 9.74814391705582e-05, + "loss": 0.0151, + "step": 8599 + }, + { + "epoch": 4.047058823529412, + "grad_norm": 0.12524693739468848, + "learning_rate": 9.744037991566058e-05, + "loss": 0.037, + "step": 8600 + }, + { + "epoch": 4.047529411764706, + "grad_norm": 0.21411111435821517, + "learning_rate": 9.739932514961232e-05, + "loss": 0.0816, + "step": 8601 + }, + { + "epoch": 4.048, + "grad_norm": 3.018968450592048e-05, + "learning_rate": 9.73582748759197e-05, + "loss": 0.0, + "step": 8602 + }, + { + "epoch": 4.048470588235294, + "grad_norm": 0.11201099779534439, + "learning_rate": 9.731722909808863e-05, + "loss": 0.0182, + "step": 8603 + }, + { + "epoch": 4.0489411764705885, + "grad_norm": 0.11717980929345871, + "learning_rate": 9.727618781962457e-05, + "loss": 0.0445, + "step": 8604 + }, + { + "epoch": 4.049411764705883, + "grad_norm": 0.10606226039503026, + "learning_rate": 9.723515104403268e-05, + "loss": 0.0333, + "step": 8605 + }, + { + "epoch": 4.049882352941176, + "grad_norm": 1.9661141845074217e-05, + "learning_rate": 9.719411877481776e-05, + "loss": 0.0, + "step": 8606 + }, + { + "epoch": 4.05035294117647, + "grad_norm": 0.08498143462053222, + "learning_rate": 9.715309101548406e-05, + "loss": 0.0225, + "step": 8607 + }, + { + "epoch": 4.0508235294117645, + "grad_norm": 1.777034754554485e-05, + "learning_rate": 9.711206776953559e-05, + "loss": 0.0, + "step": 8608 + }, + { + "epoch": 4.051294117647059, + "grad_norm": 0.11511674171150946, + "learning_rate": 9.707104904047596e-05, + "loss": 0.0354, + "step": 8609 + }, + { + "epoch": 4.051764705882353, + "grad_norm": 0.15088992200299636, + "learning_rate": 9.703003483180836e-05, + "loss": 0.0611, + "step": 8610 + }, + { + "epoch": 4.052235294117647, + "grad_norm": 0.04130084073526795, + "learning_rate": 9.698902514703558e-05, + "loss": 0.0018, + "step": 8611 + }, + { + "epoch": 4.052705882352941, + "grad_norm": 0.049488144934057295, + "learning_rate": 9.694801998966011e-05, + "loss": 0.0072, + "step": 8612 + }, + { + "epoch": 4.053176470588236, + "grad_norm": 0.0886397027903891, + "learning_rate": 9.690701936318395e-05, + "loss": 0.0225, + "step": 8613 + }, + { + "epoch": 4.05364705882353, + "grad_norm": 0.16559733410509173, + "learning_rate": 9.686602327110869e-05, + "loss": 0.0306, + "step": 8614 + }, + { + "epoch": 4.054117647058823, + "grad_norm": 3.115061576028838e-05, + "learning_rate": 9.682503171693573e-05, + "loss": 0.0, + "step": 8615 + }, + { + "epoch": 4.054588235294117, + "grad_norm": 0.11125908008570397, + "learning_rate": 9.678404470416587e-05, + "loss": 0.0226, + "step": 8616 + }, + { + "epoch": 4.055058823529412, + "grad_norm": 4.400590676839937e-06, + "learning_rate": 9.674306223629963e-05, + "loss": 0.0, + "step": 8617 + }, + { + "epoch": 4.055529411764706, + "grad_norm": 0.05623087058833887, + "learning_rate": 9.670208431683704e-05, + "loss": 0.0141, + "step": 8618 + }, + { + "epoch": 4.056, + "grad_norm": 0.06013848337241392, + "learning_rate": 9.666111094927792e-05, + "loss": 0.0056, + "step": 8619 + }, + { + "epoch": 4.056470588235294, + "grad_norm": 0.09441404391601325, + "learning_rate": 9.662014213712155e-05, + "loss": 0.0253, + "step": 8620 + }, + { + "epoch": 4.0569411764705885, + "grad_norm": 0.09885724585248051, + "learning_rate": 9.657917788386684e-05, + "loss": 0.0295, + "step": 8621 + }, + { + "epoch": 4.057411764705883, + "grad_norm": 1.1358864088811349e-05, + "learning_rate": 9.65382181930124e-05, + "loss": 0.0, + "step": 8622 + }, + { + "epoch": 4.057882352941176, + "grad_norm": 0.0655381975329103, + "learning_rate": 9.649726306805635e-05, + "loss": 0.0141, + "step": 8623 + }, + { + "epoch": 4.05835294117647, + "grad_norm": 0.01882561497919694, + "learning_rate": 9.645631251249643e-05, + "loss": 0.0004, + "step": 8624 + }, + { + "epoch": 4.0588235294117645, + "grad_norm": 0.0623643599384916, + "learning_rate": 9.641536652983008e-05, + "loss": 0.0118, + "step": 8625 + }, + { + "epoch": 4.059294117647059, + "grad_norm": 0.054398331851092566, + "learning_rate": 9.637442512355428e-05, + "loss": 0.009, + "step": 8626 + }, + { + "epoch": 4.059764705882353, + "grad_norm": 0.20245243356041942, + "learning_rate": 9.63334882971656e-05, + "loss": 0.0517, + "step": 8627 + }, + { + "epoch": 4.060235294117647, + "grad_norm": 3.133157393400679e-05, + "learning_rate": 9.629255605416021e-05, + "loss": 0.0, + "step": 8628 + }, + { + "epoch": 4.060705882352941, + "grad_norm": 0.12429208996528447, + "learning_rate": 9.6251628398034e-05, + "loss": 0.0282, + "step": 8629 + }, + { + "epoch": 4.061176470588236, + "grad_norm": 0.08714457689735645, + "learning_rate": 9.621070533228236e-05, + "loss": 0.0256, + "step": 8630 + }, + { + "epoch": 4.06164705882353, + "grad_norm": 0.03650003479587631, + "learning_rate": 9.616978686040028e-05, + "loss": 0.0057, + "step": 8631 + }, + { + "epoch": 4.062117647058823, + "grad_norm": 6.813042925062326e-05, + "learning_rate": 9.612887298588246e-05, + "loss": 0.0, + "step": 8632 + }, + { + "epoch": 4.062588235294117, + "grad_norm": 0.06126791680426345, + "learning_rate": 9.608796371222316e-05, + "loss": 0.0137, + "step": 8633 + }, + { + "epoch": 4.063058823529412, + "grad_norm": 0.13402375438809713, + "learning_rate": 9.604705904291614e-05, + "loss": 0.0589, + "step": 8634 + }, + { + "epoch": 4.063529411764706, + "grad_norm": 0.07534421067547407, + "learning_rate": 9.600615898145496e-05, + "loss": 0.0108, + "step": 8635 + }, + { + "epoch": 4.064, + "grad_norm": 0.06172225494855758, + "learning_rate": 9.596526353133261e-05, + "loss": 0.01, + "step": 8636 + }, + { + "epoch": 4.064470588235294, + "grad_norm": 0.10407339965018204, + "learning_rate": 9.59243726960418e-05, + "loss": 0.0255, + "step": 8637 + }, + { + "epoch": 4.0649411764705885, + "grad_norm": 0.12084492862257556, + "learning_rate": 9.588348647907481e-05, + "loss": 0.035, + "step": 8638 + }, + { + "epoch": 4.065411764705883, + "grad_norm": 0.0026706613958033666, + "learning_rate": 9.584260488392353e-05, + "loss": 0.0, + "step": 8639 + }, + { + "epoch": 4.065882352941176, + "grad_norm": 0.001906791082302359, + "learning_rate": 9.580172791407942e-05, + "loss": 0.0, + "step": 8640 + }, + { + "epoch": 4.06635294117647, + "grad_norm": 0.10775500763470934, + "learning_rate": 9.576085557303355e-05, + "loss": 0.0305, + "step": 8641 + }, + { + "epoch": 4.0668235294117645, + "grad_norm": 0.09406889137754361, + "learning_rate": 9.571998786427665e-05, + "loss": 0.0153, + "step": 8642 + }, + { + "epoch": 4.067294117647059, + "grad_norm": 0.09321029049774154, + "learning_rate": 9.567912479129907e-05, + "loss": 0.0285, + "step": 8643 + }, + { + "epoch": 4.067764705882353, + "grad_norm": 0.07857165927892451, + "learning_rate": 9.563826635759068e-05, + "loss": 0.0133, + "step": 8644 + }, + { + "epoch": 4.068235294117647, + "grad_norm": 0.04259715173943248, + "learning_rate": 9.559741256664098e-05, + "loss": 0.0051, + "step": 8645 + }, + { + "epoch": 4.068705882352941, + "grad_norm": 0.06335672389323067, + "learning_rate": 9.555656342193907e-05, + "loss": 0.012, + "step": 8646 + }, + { + "epoch": 4.069176470588236, + "grad_norm": 0.08227079502576064, + "learning_rate": 9.551571892697372e-05, + "loss": 0.0181, + "step": 8647 + }, + { + "epoch": 4.06964705882353, + "grad_norm": 8.808219239745497e-05, + "learning_rate": 9.547487908523322e-05, + "loss": 0.0, + "step": 8648 + }, + { + "epoch": 4.070117647058823, + "grad_norm": 0.12449086431372261, + "learning_rate": 9.543404390020543e-05, + "loss": 0.0237, + "step": 8649 + }, + { + "epoch": 4.070588235294117, + "grad_norm": 0.061821153346593154, + "learning_rate": 9.5393213375378e-05, + "loss": 0.0117, + "step": 8650 + }, + { + "epoch": 4.071058823529412, + "grad_norm": 0.15593026086163883, + "learning_rate": 9.535238751423799e-05, + "loss": 0.0621, + "step": 8651 + }, + { + "epoch": 4.071529411764706, + "grad_norm": 0.09000165438735384, + "learning_rate": 9.531156632027209e-05, + "loss": 0.0194, + "step": 8652 + }, + { + "epoch": 4.072, + "grad_norm": 0.036343981101421605, + "learning_rate": 9.52707497969667e-05, + "loss": 0.0052, + "step": 8653 + }, + { + "epoch": 4.072470588235294, + "grad_norm": 0.13129197862887884, + "learning_rate": 9.522993794780773e-05, + "loss": 0.0285, + "step": 8654 + }, + { + "epoch": 4.0729411764705885, + "grad_norm": 0.08864786676714712, + "learning_rate": 9.518913077628066e-05, + "loss": 0.0185, + "step": 8655 + }, + { + "epoch": 4.073411764705883, + "grad_norm": 0.3841060365616038, + "learning_rate": 9.514832828587071e-05, + "loss": 0.0879, + "step": 8656 + }, + { + "epoch": 4.073882352941176, + "grad_norm": 0.10136088277900687, + "learning_rate": 9.510753048006257e-05, + "loss": 0.0162, + "step": 8657 + }, + { + "epoch": 4.07435294117647, + "grad_norm": 0.16211099298245657, + "learning_rate": 9.506673736234057e-05, + "loss": 0.0475, + "step": 8658 + }, + { + "epoch": 4.0748235294117645, + "grad_norm": 0.17980368934958504, + "learning_rate": 9.502594893618861e-05, + "loss": 0.0509, + "step": 8659 + }, + { + "epoch": 4.075294117647059, + "grad_norm": 0.1263979617627512, + "learning_rate": 9.49851652050903e-05, + "loss": 0.0552, + "step": 8660 + }, + { + "epoch": 4.075764705882353, + "grad_norm": 0.00024529852583871645, + "learning_rate": 9.494438617252875e-05, + "loss": 0.0, + "step": 8661 + }, + { + "epoch": 4.076235294117647, + "grad_norm": 0.08018299696675789, + "learning_rate": 9.490361184198659e-05, + "loss": 0.0251, + "step": 8662 + }, + { + "epoch": 4.076705882352941, + "grad_norm": 0.15832859696503665, + "learning_rate": 9.486284221694626e-05, + "loss": 0.0376, + "step": 8663 + }, + { + "epoch": 4.077176470588236, + "grad_norm": 0.09030160275937012, + "learning_rate": 9.482207730088968e-05, + "loss": 0.0288, + "step": 8664 + }, + { + "epoch": 4.07764705882353, + "grad_norm": 0.1047417549020077, + "learning_rate": 9.47813170972983e-05, + "loss": 0.0323, + "step": 8665 + }, + { + "epoch": 4.078117647058823, + "grad_norm": 0.2331910959293354, + "learning_rate": 9.474056160965332e-05, + "loss": 0.0631, + "step": 8666 + }, + { + "epoch": 4.078588235294117, + "grad_norm": 0.09137989445820659, + "learning_rate": 9.469981084143541e-05, + "loss": 0.0277, + "step": 8667 + }, + { + "epoch": 4.079058823529412, + "grad_norm": 0.04573007092735135, + "learning_rate": 9.465906479612492e-05, + "loss": 0.0055, + "step": 8668 + }, + { + "epoch": 4.079529411764706, + "grad_norm": 0.19650277835726393, + "learning_rate": 9.461832347720168e-05, + "loss": 0.0686, + "step": 8669 + }, + { + "epoch": 4.08, + "grad_norm": 0.08490369785879336, + "learning_rate": 9.457758688814533e-05, + "loss": 0.0033, + "step": 8670 + }, + { + "epoch": 4.080470588235294, + "grad_norm": 3.110213719300062e-05, + "learning_rate": 9.453685503243489e-05, + "loss": 0.0, + "step": 8671 + }, + { + "epoch": 4.0809411764705885, + "grad_norm": 0.06619161615173208, + "learning_rate": 9.449612791354904e-05, + "loss": 0.0104, + "step": 8672 + }, + { + "epoch": 4.081411764705883, + "grad_norm": 0.04159107934474993, + "learning_rate": 9.445540553496614e-05, + "loss": 0.0036, + "step": 8673 + }, + { + "epoch": 4.081882352941176, + "grad_norm": 0.05999957065082279, + "learning_rate": 9.441468790016406e-05, + "loss": 0.0124, + "step": 8674 + }, + { + "epoch": 4.08235294117647, + "grad_norm": 0.1517398728835476, + "learning_rate": 9.437397501262026e-05, + "loss": 0.0296, + "step": 8675 + }, + { + "epoch": 4.0828235294117645, + "grad_norm": 0.0667006656425981, + "learning_rate": 9.433326687581183e-05, + "loss": 0.0057, + "step": 8676 + }, + { + "epoch": 4.083294117647059, + "grad_norm": 0.0951411337681855, + "learning_rate": 9.429256349321548e-05, + "loss": 0.018, + "step": 8677 + }, + { + "epoch": 4.083764705882353, + "grad_norm": 0.11588323052313322, + "learning_rate": 9.425186486830745e-05, + "loss": 0.05, + "step": 8678 + }, + { + "epoch": 4.084235294117647, + "grad_norm": 0.150875674772977, + "learning_rate": 9.421117100456356e-05, + "loss": 0.0139, + "step": 8679 + }, + { + "epoch": 4.084705882352941, + "grad_norm": 9.030605645725139e-06, + "learning_rate": 9.417048190545935e-05, + "loss": 0.0, + "step": 8680 + }, + { + "epoch": 4.085176470588236, + "grad_norm": 0.12454556752028745, + "learning_rate": 9.412979757446983e-05, + "loss": 0.0389, + "step": 8681 + }, + { + "epoch": 4.085647058823529, + "grad_norm": 6.204457967624041e-07, + "learning_rate": 9.40891180150696e-05, + "loss": 0.0, + "step": 8682 + }, + { + "epoch": 4.086117647058823, + "grad_norm": 5.341851791562881e-06, + "learning_rate": 9.404844323073296e-05, + "loss": 0.0, + "step": 8683 + }, + { + "epoch": 4.086588235294117, + "grad_norm": 0.12337418850093201, + "learning_rate": 9.400777322493373e-05, + "loss": 0.0363, + "step": 8684 + }, + { + "epoch": 4.087058823529412, + "grad_norm": 0.8043465613830371, + "learning_rate": 9.396710800114526e-05, + "loss": 0.1008, + "step": 8685 + }, + { + "epoch": 4.087529411764706, + "grad_norm": 0.14517689104805237, + "learning_rate": 9.39264475628406e-05, + "loss": 0.0518, + "step": 8686 + }, + { + "epoch": 4.088, + "grad_norm": 0.24609962980444522, + "learning_rate": 9.388579191349237e-05, + "loss": 0.0821, + "step": 8687 + }, + { + "epoch": 4.088470588235294, + "grad_norm": 0.10804984262241318, + "learning_rate": 9.384514105657277e-05, + "loss": 0.0482, + "step": 8688 + }, + { + "epoch": 4.0889411764705885, + "grad_norm": 0.11604046385515071, + "learning_rate": 9.380449499555358e-05, + "loss": 0.0239, + "step": 8689 + }, + { + "epoch": 4.089411764705883, + "grad_norm": 0.0328102862211353, + "learning_rate": 9.376385373390611e-05, + "loss": 0.0055, + "step": 8690 + }, + { + "epoch": 4.089882352941176, + "grad_norm": 0.05733468967369114, + "learning_rate": 9.372321727510139e-05, + "loss": 0.0089, + "step": 8691 + }, + { + "epoch": 4.09035294117647, + "grad_norm": 0.05325374742807881, + "learning_rate": 9.368258562260999e-05, + "loss": 0.0019, + "step": 8692 + }, + { + "epoch": 4.0908235294117645, + "grad_norm": 2.0153621932637654e-06, + "learning_rate": 9.364195877990195e-05, + "loss": 0.0, + "step": 8693 + }, + { + "epoch": 4.091294117647059, + "grad_norm": 4.502496992502036e-06, + "learning_rate": 9.360133675044708e-05, + "loss": 0.0, + "step": 8694 + }, + { + "epoch": 4.091764705882353, + "grad_norm": 8.303069760214034e-06, + "learning_rate": 9.356071953771472e-05, + "loss": 0.0, + "step": 8695 + }, + { + "epoch": 4.092235294117647, + "grad_norm": 0.1363286130002827, + "learning_rate": 9.352010714517371e-05, + "loss": 0.037, + "step": 8696 + }, + { + "epoch": 4.092705882352941, + "grad_norm": 0.11756144487876145, + "learning_rate": 9.347949957629258e-05, + "loss": 0.0466, + "step": 8697 + }, + { + "epoch": 4.093176470588236, + "grad_norm": 0.0991971360909595, + "learning_rate": 9.343889683453945e-05, + "loss": 0.0166, + "step": 8698 + }, + { + "epoch": 4.09364705882353, + "grad_norm": 0.18383961419161435, + "learning_rate": 9.339829892338193e-05, + "loss": 0.048, + "step": 8699 + }, + { + "epoch": 4.094117647058823, + "grad_norm": 0.18088205273351085, + "learning_rate": 9.33577058462873e-05, + "loss": 0.0804, + "step": 8700 + }, + { + "epoch": 4.094588235294117, + "grad_norm": 0.0653508149816456, + "learning_rate": 9.331711760672246e-05, + "loss": 0.0107, + "step": 8701 + }, + { + "epoch": 4.095058823529412, + "grad_norm": 0.1026483315910598, + "learning_rate": 9.32765342081538e-05, + "loss": 0.0315, + "step": 8702 + }, + { + "epoch": 4.095529411764706, + "grad_norm": 3.0032636471855048e-05, + "learning_rate": 9.323595565404731e-05, + "loss": 0.0, + "step": 8703 + }, + { + "epoch": 4.096, + "grad_norm": 0.10026390620981201, + "learning_rate": 9.319538194786867e-05, + "loss": 0.0283, + "step": 8704 + }, + { + "epoch": 4.096470588235294, + "grad_norm": 0.08253563093707217, + "learning_rate": 9.315481309308303e-05, + "loss": 0.0176, + "step": 8705 + }, + { + "epoch": 4.0969411764705885, + "grad_norm": 5.00074002793648e-06, + "learning_rate": 9.311424909315514e-05, + "loss": 0.0, + "step": 8706 + }, + { + "epoch": 4.097411764705883, + "grad_norm": 0.08850640465739963, + "learning_rate": 9.307368995154945e-05, + "loss": 0.0275, + "step": 8707 + }, + { + "epoch": 4.097882352941176, + "grad_norm": 0.13270473519872264, + "learning_rate": 9.303313567172985e-05, + "loss": 0.0313, + "step": 8708 + }, + { + "epoch": 4.09835294117647, + "grad_norm": 0.20511313273714252, + "learning_rate": 9.299258625715988e-05, + "loss": 0.082, + "step": 8709 + }, + { + "epoch": 4.0988235294117645, + "grad_norm": 0.00015384941441168692, + "learning_rate": 9.295204171130264e-05, + "loss": 0.0, + "step": 8710 + }, + { + "epoch": 4.099294117647059, + "grad_norm": 0.1070675119663478, + "learning_rate": 9.291150203762087e-05, + "loss": 0.0306, + "step": 8711 + }, + { + "epoch": 4.099764705882353, + "grad_norm": 0.14974475289911712, + "learning_rate": 9.287096723957685e-05, + "loss": 0.0694, + "step": 8712 + }, + { + "epoch": 4.100235294117647, + "grad_norm": 0.06652499996036138, + "learning_rate": 9.283043732063243e-05, + "loss": 0.0109, + "step": 8713 + }, + { + "epoch": 4.100705882352941, + "grad_norm": 0.17373709765037126, + "learning_rate": 9.27899122842491e-05, + "loss": 0.0706, + "step": 8714 + }, + { + "epoch": 4.101176470588236, + "grad_norm": 0.1652398314742085, + "learning_rate": 9.274939213388785e-05, + "loss": 0.0336, + "step": 8715 + }, + { + "epoch": 4.101647058823529, + "grad_norm": 0.14738394090154608, + "learning_rate": 9.270887687300932e-05, + "loss": 0.0577, + "step": 8716 + }, + { + "epoch": 4.102117647058823, + "grad_norm": 0.11669307220776526, + "learning_rate": 9.266836650507375e-05, + "loss": 0.0356, + "step": 8717 + }, + { + "epoch": 4.102588235294117, + "grad_norm": 0.06535449457899767, + "learning_rate": 9.26278610335409e-05, + "loss": 0.0093, + "step": 8718 + }, + { + "epoch": 4.103058823529412, + "grad_norm": 0.13558464552659252, + "learning_rate": 9.25873604618701e-05, + "loss": 0.0411, + "step": 8719 + }, + { + "epoch": 4.103529411764706, + "grad_norm": 0.039432904679400284, + "learning_rate": 9.254686479352029e-05, + "loss": 0.0041, + "step": 8720 + }, + { + "epoch": 4.104, + "grad_norm": 0.053202472367163516, + "learning_rate": 9.250637403195008e-05, + "loss": 0.0062, + "step": 8721 + }, + { + "epoch": 4.104470588235294, + "grad_norm": 0.08985507056544911, + "learning_rate": 9.246588818061753e-05, + "loss": 0.023, + "step": 8722 + }, + { + "epoch": 4.1049411764705885, + "grad_norm": 0.10669747209967104, + "learning_rate": 9.242540724298032e-05, + "loss": 0.029, + "step": 8723 + }, + { + "epoch": 4.105411764705883, + "grad_norm": 0.059887207932250344, + "learning_rate": 9.238493122249574e-05, + "loss": 0.0113, + "step": 8724 + }, + { + "epoch": 4.105882352941176, + "grad_norm": 0.2259248332793086, + "learning_rate": 9.234446012262065e-05, + "loss": 0.0488, + "step": 8725 + }, + { + "epoch": 4.10635294117647, + "grad_norm": 0.06189593858058604, + "learning_rate": 9.230399394681141e-05, + "loss": 0.0108, + "step": 8726 + }, + { + "epoch": 4.1068235294117645, + "grad_norm": 0.13952424261649293, + "learning_rate": 9.226353269852413e-05, + "loss": 0.0295, + "step": 8727 + }, + { + "epoch": 4.107294117647059, + "grad_norm": 0.09304391381890316, + "learning_rate": 9.222307638121436e-05, + "loss": 0.0157, + "step": 8728 + }, + { + "epoch": 4.107764705882353, + "grad_norm": 0.06433407076440646, + "learning_rate": 9.218262499833723e-05, + "loss": 0.007, + "step": 8729 + }, + { + "epoch": 4.108235294117647, + "grad_norm": 0.1190830269838495, + "learning_rate": 9.21421785533475e-05, + "loss": 0.0336, + "step": 8730 + }, + { + "epoch": 4.108705882352941, + "grad_norm": 0.08303308484139266, + "learning_rate": 9.210173704969952e-05, + "loss": 0.0169, + "step": 8731 + }, + { + "epoch": 4.109176470588236, + "grad_norm": 0.0003749865415572049, + "learning_rate": 9.206130049084721e-05, + "loss": 0.0, + "step": 8732 + }, + { + "epoch": 4.109647058823529, + "grad_norm": 0.13993283549822302, + "learning_rate": 9.202086888024404e-05, + "loss": 0.0511, + "step": 8733 + }, + { + "epoch": 4.110117647058823, + "grad_norm": 0.08412019979277117, + "learning_rate": 9.1980442221343e-05, + "loss": 0.0162, + "step": 8734 + }, + { + "epoch": 4.110588235294117, + "grad_norm": 0.1018834583625694, + "learning_rate": 9.194002051759685e-05, + "loss": 0.0226, + "step": 8735 + }, + { + "epoch": 4.111058823529412, + "grad_norm": 0.12347781998383162, + "learning_rate": 9.189960377245768e-05, + "loss": 0.0408, + "step": 8736 + }, + { + "epoch": 4.111529411764706, + "grad_norm": 0.027533560360987357, + "learning_rate": 9.185919198937736e-05, + "loss": 0.0027, + "step": 8737 + }, + { + "epoch": 4.112, + "grad_norm": 0.10559641217042467, + "learning_rate": 9.181878517180719e-05, + "loss": 0.0238, + "step": 8738 + }, + { + "epoch": 4.112470588235294, + "grad_norm": 0.03274873021003732, + "learning_rate": 9.177838332319818e-05, + "loss": 0.0018, + "step": 8739 + }, + { + "epoch": 4.1129411764705885, + "grad_norm": 0.09707935940015626, + "learning_rate": 9.173798644700079e-05, + "loss": 0.027, + "step": 8740 + }, + { + "epoch": 4.113411764705883, + "grad_norm": 0.2037726011817132, + "learning_rate": 9.169759454666511e-05, + "loss": 0.0782, + "step": 8741 + }, + { + "epoch": 4.113882352941176, + "grad_norm": 0.08361701236541119, + "learning_rate": 9.165720762564086e-05, + "loss": 0.0198, + "step": 8742 + }, + { + "epoch": 4.11435294117647, + "grad_norm": 0.0580089041199207, + "learning_rate": 9.161682568737726e-05, + "loss": 0.0081, + "step": 8743 + }, + { + "epoch": 4.1148235294117645, + "grad_norm": 0.0649930931738553, + "learning_rate": 9.157644873532308e-05, + "loss": 0.0104, + "step": 8744 + }, + { + "epoch": 4.115294117647059, + "grad_norm": 0.042357903318228585, + "learning_rate": 9.153607677292676e-05, + "loss": 0.006, + "step": 8745 + }, + { + "epoch": 4.115764705882353, + "grad_norm": 0.02552870640881248, + "learning_rate": 9.149570980363624e-05, + "loss": 0.0022, + "step": 8746 + }, + { + "epoch": 4.116235294117647, + "grad_norm": 0.0015285819920587177, + "learning_rate": 9.145534783089909e-05, + "loss": 0.0, + "step": 8747 + }, + { + "epoch": 4.116705882352941, + "grad_norm": 0.11210515057811829, + "learning_rate": 9.141499085816237e-05, + "loss": 0.0257, + "step": 8748 + }, + { + "epoch": 4.117176470588236, + "grad_norm": 0.07872755565545402, + "learning_rate": 9.137463888887279e-05, + "loss": 0.0136, + "step": 8749 + }, + { + "epoch": 4.117647058823529, + "grad_norm": 0.0002185549857520524, + "learning_rate": 9.133429192647661e-05, + "loss": 0.0, + "step": 8750 + }, + { + "epoch": 4.118117647058823, + "grad_norm": 0.23232274049474072, + "learning_rate": 9.129394997441963e-05, + "loss": 0.0692, + "step": 8751 + }, + { + "epoch": 4.118588235294117, + "grad_norm": 0.045058497000726866, + "learning_rate": 9.12536130361473e-05, + "loss": 0.0041, + "step": 8752 + }, + { + "epoch": 4.119058823529412, + "grad_norm": 0.09064454267834116, + "learning_rate": 9.121328111510457e-05, + "loss": 0.017, + "step": 8753 + }, + { + "epoch": 4.119529411764706, + "grad_norm": 0.09526023202131068, + "learning_rate": 9.117295421473595e-05, + "loss": 0.0122, + "step": 8754 + }, + { + "epoch": 4.12, + "grad_norm": 0.22009739823098917, + "learning_rate": 9.113263233848562e-05, + "loss": 0.1281, + "step": 8755 + }, + { + "epoch": 4.120470588235294, + "grad_norm": 0.11370110994122891, + "learning_rate": 9.109231548979722e-05, + "loss": 0.0474, + "step": 8756 + }, + { + "epoch": 4.1209411764705886, + "grad_norm": 0.11593195693808281, + "learning_rate": 9.105200367211401e-05, + "loss": 0.0182, + "step": 8757 + }, + { + "epoch": 4.121411764705883, + "grad_norm": 0.09724479398953005, + "learning_rate": 9.101169688887882e-05, + "loss": 0.0351, + "step": 8758 + }, + { + "epoch": 4.121882352941176, + "grad_norm": 0.15168973964509658, + "learning_rate": 9.097139514353407e-05, + "loss": 0.0546, + "step": 8759 + }, + { + "epoch": 4.12235294117647, + "grad_norm": 0.1780576291003483, + "learning_rate": 9.093109843952171e-05, + "loss": 0.0782, + "step": 8760 + }, + { + "epoch": 4.1228235294117646, + "grad_norm": 0.048043994817281194, + "learning_rate": 9.089080678028325e-05, + "loss": 0.0089, + "step": 8761 + }, + { + "epoch": 4.123294117647059, + "grad_norm": 0.09801054263384304, + "learning_rate": 9.085052016925984e-05, + "loss": 0.0257, + "step": 8762 + }, + { + "epoch": 4.123764705882353, + "grad_norm": 0.09237006621009308, + "learning_rate": 9.081023860989213e-05, + "loss": 0.021, + "step": 8763 + }, + { + "epoch": 4.124235294117647, + "grad_norm": 0.1243558501331554, + "learning_rate": 9.076996210562035e-05, + "loss": 0.0472, + "step": 8764 + }, + { + "epoch": 4.124705882352941, + "grad_norm": 0.15845622518074465, + "learning_rate": 9.072969065988435e-05, + "loss": 0.0463, + "step": 8765 + }, + { + "epoch": 4.125176470588236, + "grad_norm": 0.05396860497256934, + "learning_rate": 9.06894242761235e-05, + "loss": 0.0093, + "step": 8766 + }, + { + "epoch": 4.125647058823529, + "grad_norm": 0.07922235728745064, + "learning_rate": 9.064916295777669e-05, + "loss": 0.0197, + "step": 8767 + }, + { + "epoch": 4.126117647058823, + "grad_norm": 0.08949933705924858, + "learning_rate": 9.06089067082825e-05, + "loss": 0.02, + "step": 8768 + }, + { + "epoch": 4.1265882352941174, + "grad_norm": 0.08472409866663262, + "learning_rate": 9.0568655531079e-05, + "loss": 0.0171, + "step": 8769 + }, + { + "epoch": 4.127058823529412, + "grad_norm": 0.07750285625745362, + "learning_rate": 9.052840942960383e-05, + "loss": 0.0167, + "step": 8770 + }, + { + "epoch": 4.127529411764706, + "grad_norm": 0.1439860451245051, + "learning_rate": 9.048816840729415e-05, + "loss": 0.0413, + "step": 8771 + }, + { + "epoch": 4.128, + "grad_norm": 1.1162871631161932e-05, + "learning_rate": 9.044793246758685e-05, + "loss": 0.0, + "step": 8772 + }, + { + "epoch": 4.128470588235294, + "grad_norm": 0.06023320173352627, + "learning_rate": 9.04077016139182e-05, + "loss": 0.0155, + "step": 8773 + }, + { + "epoch": 4.128941176470589, + "grad_norm": 0.00018283269143071156, + "learning_rate": 9.03674758497241e-05, + "loss": 0.0, + "step": 8774 + }, + { + "epoch": 4.129411764705883, + "grad_norm": 4.967731389455512e-06, + "learning_rate": 9.032725517844004e-05, + "loss": 0.0, + "step": 8775 + }, + { + "epoch": 4.129882352941176, + "grad_norm": 0.08616925641652265, + "learning_rate": 9.028703960350113e-05, + "loss": 0.0219, + "step": 8776 + }, + { + "epoch": 4.13035294117647, + "grad_norm": 0.1299144182911142, + "learning_rate": 9.024682912834196e-05, + "loss": 0.0454, + "step": 8777 + }, + { + "epoch": 4.130823529411765, + "grad_norm": 0.09025564156579821, + "learning_rate": 9.020662375639663e-05, + "loss": 0.0211, + "step": 8778 + }, + { + "epoch": 4.131294117647059, + "grad_norm": 9.162222580683361e-05, + "learning_rate": 9.016642349109892e-05, + "loss": 0.0, + "step": 8779 + }, + { + "epoch": 4.131764705882353, + "grad_norm": 0.03756545383369887, + "learning_rate": 9.012622833588213e-05, + "loss": 0.0011, + "step": 8780 + }, + { + "epoch": 4.132235294117647, + "grad_norm": 0.08589906304619513, + "learning_rate": 9.008603829417913e-05, + "loss": 0.0225, + "step": 8781 + }, + { + "epoch": 4.1327058823529415, + "grad_norm": 0.12328896418603734, + "learning_rate": 9.00458533694223e-05, + "loss": 0.0463, + "step": 8782 + }, + { + "epoch": 4.133176470588236, + "grad_norm": 0.20174010364952322, + "learning_rate": 9.000567356504368e-05, + "loss": 0.0917, + "step": 8783 + }, + { + "epoch": 4.133647058823529, + "grad_norm": 0.09214425294241432, + "learning_rate": 8.996549888447484e-05, + "loss": 0.034, + "step": 8784 + }, + { + "epoch": 4.134117647058823, + "grad_norm": 0.10161989154690813, + "learning_rate": 8.99253293311468e-05, + "loss": 0.0257, + "step": 8785 + }, + { + "epoch": 4.1345882352941175, + "grad_norm": 0.04688065960576178, + "learning_rate": 8.988516490849037e-05, + "loss": 0.0059, + "step": 8786 + }, + { + "epoch": 4.135058823529412, + "grad_norm": 0.152524217453188, + "learning_rate": 8.984500561993569e-05, + "loss": 0.044, + "step": 8787 + }, + { + "epoch": 4.135529411764706, + "grad_norm": 0.05800285863277231, + "learning_rate": 8.98048514689126e-05, + "loss": 0.0191, + "step": 8788 + }, + { + "epoch": 4.136, + "grad_norm": 0.00014965882070823913, + "learning_rate": 8.976470245885041e-05, + "loss": 0.0, + "step": 8789 + }, + { + "epoch": 4.136470588235294, + "grad_norm": 0.12110295571888872, + "learning_rate": 8.972455859317811e-05, + "loss": 0.0626, + "step": 8790 + }, + { + "epoch": 4.136941176470589, + "grad_norm": 0.000718499015505179, + "learning_rate": 8.968441987532417e-05, + "loss": 0.0, + "step": 8791 + }, + { + "epoch": 4.137411764705883, + "grad_norm": 0.14378802964174334, + "learning_rate": 8.964428630871657e-05, + "loss": 0.0397, + "step": 8792 + }, + { + "epoch": 4.137882352941176, + "grad_norm": 0.1143221875530277, + "learning_rate": 8.960415789678303e-05, + "loss": 0.0404, + "step": 8793 + }, + { + "epoch": 4.13835294117647, + "grad_norm": 0.009213189489124817, + "learning_rate": 8.95640346429506e-05, + "loss": 0.0004, + "step": 8794 + }, + { + "epoch": 4.138823529411765, + "grad_norm": 0.05336123383932797, + "learning_rate": 8.952391655064605e-05, + "loss": 0.0116, + "step": 8795 + }, + { + "epoch": 4.139294117647059, + "grad_norm": 0.27149382104017183, + "learning_rate": 8.948380362329567e-05, + "loss": 0.0582, + "step": 8796 + }, + { + "epoch": 4.139764705882353, + "grad_norm": 0.1456945327758322, + "learning_rate": 8.944369586432531e-05, + "loss": 0.047, + "step": 8797 + }, + { + "epoch": 4.140235294117647, + "grad_norm": 0.10668339258987453, + "learning_rate": 8.940359327716032e-05, + "loss": 0.0098, + "step": 8798 + }, + { + "epoch": 4.1407058823529415, + "grad_norm": 3.1652863793706885e-05, + "learning_rate": 8.936349586522569e-05, + "loss": 0.0, + "step": 8799 + }, + { + "epoch": 4.141176470588236, + "grad_norm": 0.12275602071489525, + "learning_rate": 8.932340363194595e-05, + "loss": 0.0441, + "step": 8800 + }, + { + "epoch": 4.141647058823529, + "grad_norm": 0.05768058297132241, + "learning_rate": 8.928331658074516e-05, + "loss": 0.0138, + "step": 8801 + }, + { + "epoch": 4.142117647058823, + "grad_norm": 0.08347860621589435, + "learning_rate": 8.92432347150469e-05, + "loss": 0.0247, + "step": 8802 + }, + { + "epoch": 4.1425882352941175, + "grad_norm": 0.025458977282225643, + "learning_rate": 8.920315803827446e-05, + "loss": 0.0022, + "step": 8803 + }, + { + "epoch": 4.143058823529412, + "grad_norm": 0.07815812096212083, + "learning_rate": 8.916308655385054e-05, + "loss": 0.0197, + "step": 8804 + }, + { + "epoch": 4.143529411764706, + "grad_norm": 0.09723132691670583, + "learning_rate": 8.912302026519738e-05, + "loss": 0.0375, + "step": 8805 + }, + { + "epoch": 4.144, + "grad_norm": 0.05775847199439111, + "learning_rate": 8.908295917573692e-05, + "loss": 0.0118, + "step": 8806 + }, + { + "epoch": 4.144470588235294, + "grad_norm": 0.06427942014838497, + "learning_rate": 8.904290328889055e-05, + "loss": 0.0135, + "step": 8807 + }, + { + "epoch": 4.144941176470589, + "grad_norm": 0.10909544233435368, + "learning_rate": 8.900285260807922e-05, + "loss": 0.0509, + "step": 8808 + }, + { + "epoch": 4.145411764705882, + "grad_norm": 0.08558217778882407, + "learning_rate": 8.896280713672344e-05, + "loss": 0.0214, + "step": 8809 + }, + { + "epoch": 4.145882352941176, + "grad_norm": 2.9059474184904122e-05, + "learning_rate": 8.892276687824335e-05, + "loss": 0.0, + "step": 8810 + }, + { + "epoch": 4.14635294117647, + "grad_norm": 0.09121809496220963, + "learning_rate": 8.888273183605856e-05, + "loss": 0.0234, + "step": 8811 + }, + { + "epoch": 4.146823529411765, + "grad_norm": 0.08626140422552309, + "learning_rate": 8.884270201358821e-05, + "loss": 0.0272, + "step": 8812 + }, + { + "epoch": 4.147294117647059, + "grad_norm": 0.04195922785785434, + "learning_rate": 8.880267741425111e-05, + "loss": 0.0039, + "step": 8813 + }, + { + "epoch": 4.147764705882353, + "grad_norm": 0.05449793649364326, + "learning_rate": 8.876265804146554e-05, + "loss": 0.0129, + "step": 8814 + }, + { + "epoch": 4.148235294117647, + "grad_norm": 0.050417602866798195, + "learning_rate": 8.872264389864929e-05, + "loss": 0.0048, + "step": 8815 + }, + { + "epoch": 4.1487058823529415, + "grad_norm": 0.060712084777529804, + "learning_rate": 8.868263498921987e-05, + "loss": 0.003, + "step": 8816 + }, + { + "epoch": 4.149176470588236, + "grad_norm": 0.22376448008617164, + "learning_rate": 8.864263131659417e-05, + "loss": 0.0508, + "step": 8817 + }, + { + "epoch": 4.149647058823529, + "grad_norm": 0.14989781340461192, + "learning_rate": 8.860263288418866e-05, + "loss": 0.0466, + "step": 8818 + }, + { + "epoch": 4.150117647058823, + "grad_norm": 0.12582796768339327, + "learning_rate": 8.856263969541948e-05, + "loss": 0.0441, + "step": 8819 + }, + { + "epoch": 4.1505882352941175, + "grad_norm": 0.07819855507216568, + "learning_rate": 8.852265175370223e-05, + "loss": 0.0182, + "step": 8820 + }, + { + "epoch": 4.151058823529412, + "grad_norm": 1.1579313891288796e-05, + "learning_rate": 8.848266906245207e-05, + "loss": 0.0, + "step": 8821 + }, + { + "epoch": 4.151529411764706, + "grad_norm": 0.172579760031571, + "learning_rate": 8.844269162508373e-05, + "loss": 0.0613, + "step": 8822 + }, + { + "epoch": 4.152, + "grad_norm": 0.18640771177901846, + "learning_rate": 8.840271944501143e-05, + "loss": 0.1021, + "step": 8823 + }, + { + "epoch": 4.152470588235294, + "grad_norm": 0.06236281570193063, + "learning_rate": 8.836275252564903e-05, + "loss": 0.0178, + "step": 8824 + }, + { + "epoch": 4.152941176470589, + "grad_norm": 0.10748945945092908, + "learning_rate": 8.832279087040992e-05, + "loss": 0.028, + "step": 8825 + }, + { + "epoch": 4.153411764705883, + "grad_norm": 0.07555243960855076, + "learning_rate": 8.828283448270698e-05, + "loss": 0.02, + "step": 8826 + }, + { + "epoch": 4.153882352941176, + "grad_norm": 0.03536346724108963, + "learning_rate": 8.824288336595267e-05, + "loss": 0.0042, + "step": 8827 + }, + { + "epoch": 4.15435294117647, + "grad_norm": 0.12008682641986353, + "learning_rate": 8.820293752355906e-05, + "loss": 0.0243, + "step": 8828 + }, + { + "epoch": 4.154823529411765, + "grad_norm": 0.08245862837860744, + "learning_rate": 8.816299695893773e-05, + "loss": 0.0183, + "step": 8829 + }, + { + "epoch": 4.155294117647059, + "grad_norm": 0.09763568052527968, + "learning_rate": 8.81230616754997e-05, + "loss": 0.0362, + "step": 8830 + }, + { + "epoch": 4.155764705882353, + "grad_norm": 0.0670552643873098, + "learning_rate": 8.808313167665575e-05, + "loss": 0.0104, + "step": 8831 + }, + { + "epoch": 4.156235294117647, + "grad_norm": 0.06766728887365513, + "learning_rate": 8.804320696581606e-05, + "loss": 0.0049, + "step": 8832 + }, + { + "epoch": 4.1567058823529415, + "grad_norm": 0.002828525464975322, + "learning_rate": 8.800328754639038e-05, + "loss": 0.0, + "step": 8833 + }, + { + "epoch": 4.157176470588236, + "grad_norm": 0.00029064255039635146, + "learning_rate": 8.796337342178804e-05, + "loss": 0.0, + "step": 8834 + }, + { + "epoch": 4.157647058823529, + "grad_norm": 0.07926915487001263, + "learning_rate": 8.792346459541789e-05, + "loss": 0.0139, + "step": 8835 + }, + { + "epoch": 4.158117647058823, + "grad_norm": 0.00013488142065161938, + "learning_rate": 8.788356107068833e-05, + "loss": 0.0, + "step": 8836 + }, + { + "epoch": 4.1585882352941175, + "grad_norm": 0.06692002580246487, + "learning_rate": 8.784366285100738e-05, + "loss": 0.0119, + "step": 8837 + }, + { + "epoch": 4.159058823529412, + "grad_norm": 0.0010169121843268184, + "learning_rate": 8.780376993978246e-05, + "loss": 0.0, + "step": 8838 + }, + { + "epoch": 4.159529411764706, + "grad_norm": 0.02536692483968324, + "learning_rate": 8.77638823404207e-05, + "loss": 0.0023, + "step": 8839 + }, + { + "epoch": 4.16, + "grad_norm": 0.1454778793546268, + "learning_rate": 8.772400005632859e-05, + "loss": 0.0351, + "step": 8840 + }, + { + "epoch": 4.160470588235294, + "grad_norm": 0.1484142405700003, + "learning_rate": 8.768412309091238e-05, + "loss": 0.032, + "step": 8841 + }, + { + "epoch": 4.160941176470589, + "grad_norm": 0.07543180292254405, + "learning_rate": 8.764425144757771e-05, + "loss": 0.014, + "step": 8842 + }, + { + "epoch": 4.161411764705882, + "grad_norm": 0.0006203616955569247, + "learning_rate": 8.760438512972982e-05, + "loss": 0.0, + "step": 8843 + }, + { + "epoch": 4.161882352941176, + "grad_norm": 0.0013198340791444343, + "learning_rate": 8.756452414077351e-05, + "loss": 0.0, + "step": 8844 + }, + { + "epoch": 4.16235294117647, + "grad_norm": 0.030255880757552308, + "learning_rate": 8.752466848411305e-05, + "loss": 0.0024, + "step": 8845 + }, + { + "epoch": 4.162823529411765, + "grad_norm": 0.25985305348537197, + "learning_rate": 8.748481816315233e-05, + "loss": 0.074, + "step": 8846 + }, + { + "epoch": 4.163294117647059, + "grad_norm": 0.00029570871089779196, + "learning_rate": 8.744497318129483e-05, + "loss": 0.0, + "step": 8847 + }, + { + "epoch": 4.163764705882353, + "grad_norm": 0.08236350543909074, + "learning_rate": 8.74051335419434e-05, + "loss": 0.0145, + "step": 8848 + }, + { + "epoch": 4.164235294117647, + "grad_norm": 0.06088940044166927, + "learning_rate": 8.736529924850057e-05, + "loss": 0.0125, + "step": 8849 + }, + { + "epoch": 4.1647058823529415, + "grad_norm": 0.07385893725052865, + "learning_rate": 8.73254703043685e-05, + "loss": 0.0154, + "step": 8850 + }, + { + "epoch": 4.165176470588236, + "grad_norm": 0.09315515915158327, + "learning_rate": 8.72856467129486e-05, + "loss": 0.0229, + "step": 8851 + }, + { + "epoch": 4.165647058823529, + "grad_norm": 0.0049927206334830715, + "learning_rate": 8.72458284776421e-05, + "loss": 0.0, + "step": 8852 + }, + { + "epoch": 4.166117647058823, + "grad_norm": 3.1966104706534606e-05, + "learning_rate": 8.72060156018497e-05, + "loss": 0.0, + "step": 8853 + }, + { + "epoch": 4.1665882352941175, + "grad_norm": 0.0483318967162558, + "learning_rate": 8.716620808897151e-05, + "loss": 0.0021, + "step": 8854 + }, + { + "epoch": 4.167058823529412, + "grad_norm": 0.1496686684236608, + "learning_rate": 8.712640594240735e-05, + "loss": 0.0561, + "step": 8855 + }, + { + "epoch": 4.167529411764706, + "grad_norm": 1.8899325673602666e-05, + "learning_rate": 8.708660916555657e-05, + "loss": 0.0, + "step": 8856 + }, + { + "epoch": 4.168, + "grad_norm": 0.17988958138764297, + "learning_rate": 8.70468177618179e-05, + "loss": 0.0771, + "step": 8857 + }, + { + "epoch": 4.168470588235294, + "grad_norm": 7.814987823887271e-05, + "learning_rate": 8.700703173458978e-05, + "loss": 0.0, + "step": 8858 + }, + { + "epoch": 4.168941176470589, + "grad_norm": 0.03355163607773589, + "learning_rate": 8.69672510872702e-05, + "loss": 0.0033, + "step": 8859 + }, + { + "epoch": 4.169411764705882, + "grad_norm": 0.08239469741884489, + "learning_rate": 8.692747582325646e-05, + "loss": 0.0133, + "step": 8860 + }, + { + "epoch": 4.169882352941176, + "grad_norm": 7.264788099907946e-06, + "learning_rate": 8.68877059459457e-05, + "loss": 0.0, + "step": 8861 + }, + { + "epoch": 4.17035294117647, + "grad_norm": 0.09748811826748553, + "learning_rate": 8.684794145873443e-05, + "loss": 0.0278, + "step": 8862 + }, + { + "epoch": 4.170823529411765, + "grad_norm": 0.08451259963065602, + "learning_rate": 8.680818236501867e-05, + "loss": 0.0163, + "step": 8863 + }, + { + "epoch": 4.171294117647059, + "grad_norm": 0.0482425698996267, + "learning_rate": 8.67684286681941e-05, + "loss": 0.0061, + "step": 8864 + }, + { + "epoch": 4.171764705882353, + "grad_norm": 0.09802932544752378, + "learning_rate": 8.672868037165587e-05, + "loss": 0.0297, + "step": 8865 + }, + { + "epoch": 4.172235294117647, + "grad_norm": 0.14427921480966896, + "learning_rate": 8.668893747879869e-05, + "loss": 0.064, + "step": 8866 + }, + { + "epoch": 4.1727058823529415, + "grad_norm": 4.8207204336822034e-05, + "learning_rate": 8.664919999301683e-05, + "loss": 0.0, + "step": 8867 + }, + { + "epoch": 4.173176470588236, + "grad_norm": 0.05658068168041468, + "learning_rate": 8.660946791770397e-05, + "loss": 0.0062, + "step": 8868 + }, + { + "epoch": 4.173647058823529, + "grad_norm": 0.1015314856001692, + "learning_rate": 8.656974125625349e-05, + "loss": 0.0221, + "step": 8869 + }, + { + "epoch": 4.174117647058823, + "grad_norm": 0.13243059989670902, + "learning_rate": 8.653002001205827e-05, + "loss": 0.0474, + "step": 8870 + }, + { + "epoch": 4.1745882352941175, + "grad_norm": 0.1637434924185687, + "learning_rate": 8.649030418851061e-05, + "loss": 0.0724, + "step": 8871 + }, + { + "epoch": 4.175058823529412, + "grad_norm": 0.06242622194293982, + "learning_rate": 8.645059378900255e-05, + "loss": 0.0139, + "step": 8872 + }, + { + "epoch": 4.175529411764706, + "grad_norm": 0.15199689829975266, + "learning_rate": 8.641088881692542e-05, + "loss": 0.0161, + "step": 8873 + }, + { + "epoch": 4.176, + "grad_norm": 0.05832562110103251, + "learning_rate": 8.637118927567028e-05, + "loss": 0.0113, + "step": 8874 + }, + { + "epoch": 4.176470588235294, + "grad_norm": 0.08213418362181656, + "learning_rate": 8.633149516862775e-05, + "loss": 0.023, + "step": 8875 + }, + { + "epoch": 4.176941176470589, + "grad_norm": 0.05584736575221259, + "learning_rate": 8.629180649918775e-05, + "loss": 0.0045, + "step": 8876 + }, + { + "epoch": 4.177411764705882, + "grad_norm": 0.10967637046207719, + "learning_rate": 8.625212327073998e-05, + "loss": 0.0404, + "step": 8877 + }, + { + "epoch": 4.177882352941176, + "grad_norm": 0.1474372749538616, + "learning_rate": 8.621244548667363e-05, + "loss": 0.0543, + "step": 8878 + }, + { + "epoch": 4.17835294117647, + "grad_norm": 0.05079817501168171, + "learning_rate": 8.617277315037723e-05, + "loss": 0.0105, + "step": 8879 + }, + { + "epoch": 4.178823529411765, + "grad_norm": 0.15115666374866474, + "learning_rate": 8.613310626523909e-05, + "loss": 0.0575, + "step": 8880 + }, + { + "epoch": 4.179294117647059, + "grad_norm": 0.09658663737744583, + "learning_rate": 8.609344483464697e-05, + "loss": 0.033, + "step": 8881 + }, + { + "epoch": 4.179764705882353, + "grad_norm": 0.14503994120127667, + "learning_rate": 8.60537888619881e-05, + "loss": 0.0544, + "step": 8882 + }, + { + "epoch": 4.180235294117647, + "grad_norm": 0.3786810953555372, + "learning_rate": 8.601413835064929e-05, + "loss": 0.1658, + "step": 8883 + }, + { + "epoch": 4.1807058823529415, + "grad_norm": 0.067187570014794, + "learning_rate": 8.597449330401698e-05, + "loss": 0.0101, + "step": 8884 + }, + { + "epoch": 4.181176470588236, + "grad_norm": 0.033701048827620726, + "learning_rate": 8.593485372547693e-05, + "loss": 0.0057, + "step": 8885 + }, + { + "epoch": 4.181647058823529, + "grad_norm": 0.06353588710926661, + "learning_rate": 8.589521961841462e-05, + "loss": 0.0082, + "step": 8886 + }, + { + "epoch": 4.182117647058823, + "grad_norm": 0.025069053203358154, + "learning_rate": 8.585559098621503e-05, + "loss": 0.0017, + "step": 8887 + }, + { + "epoch": 4.1825882352941175, + "grad_norm": 0.1185278947638375, + "learning_rate": 8.581596783226254e-05, + "loss": 0.0394, + "step": 8888 + }, + { + "epoch": 4.183058823529412, + "grad_norm": 0.16006141875269042, + "learning_rate": 8.577635015994122e-05, + "loss": 0.0694, + "step": 8889 + }, + { + "epoch": 4.183529411764706, + "grad_norm": 8.101862332253126e-06, + "learning_rate": 8.573673797263466e-05, + "loss": 0.0, + "step": 8890 + }, + { + "epoch": 4.184, + "grad_norm": 0.07970485941792887, + "learning_rate": 8.569713127372583e-05, + "loss": 0.0173, + "step": 8891 + }, + { + "epoch": 4.184470588235294, + "grad_norm": 0.1051244431829093, + "learning_rate": 8.565753006659741e-05, + "loss": 0.0236, + "step": 8892 + }, + { + "epoch": 4.184941176470589, + "grad_norm": 0.13683547641188232, + "learning_rate": 8.561793435463157e-05, + "loss": 0.0373, + "step": 8893 + }, + { + "epoch": 4.185411764705882, + "grad_norm": 0.03445150257126628, + "learning_rate": 8.557834414120986e-05, + "loss": 0.0037, + "step": 8894 + }, + { + "epoch": 4.185882352941176, + "grad_norm": 0.07649448542142574, + "learning_rate": 8.553875942971362e-05, + "loss": 0.0109, + "step": 8895 + }, + { + "epoch": 4.18635294117647, + "grad_norm": 0.15714610133050225, + "learning_rate": 8.549918022352343e-05, + "loss": 0.035, + "step": 8896 + }, + { + "epoch": 4.186823529411765, + "grad_norm": 0.11188059734045906, + "learning_rate": 8.545960652601964e-05, + "loss": 0.0252, + "step": 8897 + }, + { + "epoch": 4.187294117647059, + "grad_norm": 0.1167403527347724, + "learning_rate": 8.542003834058207e-05, + "loss": 0.0214, + "step": 8898 + }, + { + "epoch": 4.187764705882353, + "grad_norm": 0.12935464150381237, + "learning_rate": 8.538047567058994e-05, + "loss": 0.028, + "step": 8899 + }, + { + "epoch": 4.188235294117647, + "grad_norm": 0.15417985643444063, + "learning_rate": 8.534091851942214e-05, + "loss": 0.0312, + "step": 8900 + }, + { + "epoch": 4.1887058823529415, + "grad_norm": 0.13234057373058977, + "learning_rate": 8.530136689045712e-05, + "loss": 0.0332, + "step": 8901 + }, + { + "epoch": 4.189176470588236, + "grad_norm": 0.0004665570260003958, + "learning_rate": 8.526182078707261e-05, + "loss": 0.0, + "step": 8902 + }, + { + "epoch": 4.189647058823529, + "grad_norm": 0.055199491602312294, + "learning_rate": 8.522228021264618e-05, + "loss": 0.0077, + "step": 8903 + }, + { + "epoch": 4.190117647058823, + "grad_norm": 0.024829002624897383, + "learning_rate": 8.518274517055482e-05, + "loss": 0.0015, + "step": 8904 + }, + { + "epoch": 4.1905882352941175, + "grad_norm": 0.044677348160472606, + "learning_rate": 8.514321566417488e-05, + "loss": 0.0025, + "step": 8905 + }, + { + "epoch": 4.191058823529412, + "grad_norm": 0.04683241356754664, + "learning_rate": 8.510369169688243e-05, + "loss": 0.0003, + "step": 8906 + }, + { + "epoch": 4.191529411764706, + "grad_norm": 0.17226197299934742, + "learning_rate": 8.506417327205309e-05, + "loss": 0.0101, + "step": 8907 + }, + { + "epoch": 4.192, + "grad_norm": 0.002845325425528005, + "learning_rate": 8.502466039306175e-05, + "loss": 0.0, + "step": 8908 + }, + { + "epoch": 4.192470588235294, + "grad_norm": 0.08596107348058835, + "learning_rate": 8.498515306328323e-05, + "loss": 0.0167, + "step": 8909 + }, + { + "epoch": 4.192941176470589, + "grad_norm": 0.05338429777765089, + "learning_rate": 8.494565128609146e-05, + "loss": 0.0075, + "step": 8910 + }, + { + "epoch": 4.193411764705882, + "grad_norm": 0.1260301557326225, + "learning_rate": 8.490615506486018e-05, + "loss": 0.0223, + "step": 8911 + }, + { + "epoch": 4.193882352941176, + "grad_norm": 0.10275334845257118, + "learning_rate": 8.486666440296259e-05, + "loss": 0.0245, + "step": 8912 + }, + { + "epoch": 4.19435294117647, + "grad_norm": 0.1473976517632043, + "learning_rate": 8.482717930377129e-05, + "loss": 0.0291, + "step": 8913 + }, + { + "epoch": 4.194823529411765, + "grad_norm": 0.057125824901246576, + "learning_rate": 8.478769977065853e-05, + "loss": 0.0104, + "step": 8914 + }, + { + "epoch": 4.195294117647059, + "grad_norm": 0.07521277438890842, + "learning_rate": 8.474822580699614e-05, + "loss": 0.0149, + "step": 8915 + }, + { + "epoch": 4.195764705882353, + "grad_norm": 0.16776076965016776, + "learning_rate": 8.470875741615527e-05, + "loss": 0.0491, + "step": 8916 + }, + { + "epoch": 4.196235294117647, + "grad_norm": 0.12917422439247175, + "learning_rate": 8.466929460150678e-05, + "loss": 0.0363, + "step": 8917 + }, + { + "epoch": 4.1967058823529415, + "grad_norm": 0.12484919373522095, + "learning_rate": 8.462983736642101e-05, + "loss": 0.0139, + "step": 8918 + }, + { + "epoch": 4.197176470588236, + "grad_norm": 0.17983524876696333, + "learning_rate": 8.459038571426772e-05, + "loss": 0.0838, + "step": 8919 + }, + { + "epoch": 4.197647058823529, + "grad_norm": 0.07468610896006127, + "learning_rate": 8.455093964841632e-05, + "loss": 0.0176, + "step": 8920 + }, + { + "epoch": 4.198117647058823, + "grad_norm": 0.019845869958494765, + "learning_rate": 8.451149917223576e-05, + "loss": 0.0012, + "step": 8921 + }, + { + "epoch": 4.1985882352941175, + "grad_norm": 0.06454073678631096, + "learning_rate": 8.447206428909434e-05, + "loss": 0.0091, + "step": 8922 + }, + { + "epoch": 4.199058823529412, + "grad_norm": 0.1126948391472068, + "learning_rate": 8.44326350023601e-05, + "loss": 0.0209, + "step": 8923 + }, + { + "epoch": 4.199529411764706, + "grad_norm": 0.1632092230042234, + "learning_rate": 8.439321131540036e-05, + "loss": 0.0778, + "step": 8924 + }, + { + "epoch": 4.2, + "grad_norm": 0.11305635422134, + "learning_rate": 8.435379323158218e-05, + "loss": 0.0236, + "step": 8925 + }, + { + "epoch": 4.200470588235294, + "grad_norm": 0.15013193996208887, + "learning_rate": 8.431438075427213e-05, + "loss": 0.0554, + "step": 8926 + }, + { + "epoch": 4.200941176470589, + "grad_norm": 0.06785687356873368, + "learning_rate": 8.427497388683607e-05, + "loss": 0.0135, + "step": 8927 + }, + { + "epoch": 4.201411764705882, + "grad_norm": 0.038793828947473186, + "learning_rate": 8.423557263263961e-05, + "loss": 0.003, + "step": 8928 + }, + { + "epoch": 4.201882352941176, + "grad_norm": 0.10802746278033046, + "learning_rate": 8.419617699504791e-05, + "loss": 0.0338, + "step": 8929 + }, + { + "epoch": 4.20235294117647, + "grad_norm": 0.14464026260397314, + "learning_rate": 8.415678697742538e-05, + "loss": 0.0407, + "step": 8930 + }, + { + "epoch": 4.202823529411765, + "grad_norm": 0.059883591138094536, + "learning_rate": 8.41174025831362e-05, + "loss": 0.016, + "step": 8931 + }, + { + "epoch": 4.203294117647059, + "grad_norm": 0.10683264886335775, + "learning_rate": 8.407802381554406e-05, + "loss": 0.0314, + "step": 8932 + }, + { + "epoch": 4.203764705882353, + "grad_norm": 0.11569308249858277, + "learning_rate": 8.403865067801197e-05, + "loss": 0.0266, + "step": 8933 + }, + { + "epoch": 4.204235294117647, + "grad_norm": 0.08931139956567088, + "learning_rate": 8.399928317390265e-05, + "loss": 0.0261, + "step": 8934 + }, + { + "epoch": 4.2047058823529415, + "grad_norm": 0.10406573140276641, + "learning_rate": 8.395992130657834e-05, + "loss": 0.0364, + "step": 8935 + }, + { + "epoch": 4.205176470588235, + "grad_norm": 0.14796069603494794, + "learning_rate": 8.392056507940061e-05, + "loss": 0.0602, + "step": 8936 + }, + { + "epoch": 4.205647058823529, + "grad_norm": 0.13380959433684314, + "learning_rate": 8.388121449573075e-05, + "loss": 0.0613, + "step": 8937 + }, + { + "epoch": 4.206117647058823, + "grad_norm": 0.15007307421584415, + "learning_rate": 8.384186955892954e-05, + "loss": 0.0514, + "step": 8938 + }, + { + "epoch": 4.2065882352941175, + "grad_norm": 0.08519090646572859, + "learning_rate": 8.380253027235709e-05, + "loss": 0.0234, + "step": 8939 + }, + { + "epoch": 4.207058823529412, + "grad_norm": 3.2561735631662715e-06, + "learning_rate": 8.37631966393733e-05, + "loss": 0.0, + "step": 8940 + }, + { + "epoch": 4.207529411764706, + "grad_norm": 0.25488589207374934, + "learning_rate": 8.372386866333743e-05, + "loss": 0.067, + "step": 8941 + }, + { + "epoch": 4.208, + "grad_norm": 0.04907004208311268, + "learning_rate": 8.36845463476082e-05, + "loss": 0.0081, + "step": 8942 + }, + { + "epoch": 4.208470588235294, + "grad_norm": 0.04592691980137685, + "learning_rate": 8.364522969554399e-05, + "loss": 0.0064, + "step": 8943 + }, + { + "epoch": 4.208941176470589, + "grad_norm": 0.09057194682579343, + "learning_rate": 8.36059187105027e-05, + "loss": 0.0207, + "step": 8944 + }, + { + "epoch": 4.209411764705882, + "grad_norm": 0.009279144592035648, + "learning_rate": 8.356661339584156e-05, + "loss": 0.0005, + "step": 8945 + }, + { + "epoch": 4.209882352941176, + "grad_norm": 0.02520631255462584, + "learning_rate": 8.352731375491754e-05, + "loss": 0.0024, + "step": 8946 + }, + { + "epoch": 4.21035294117647, + "grad_norm": 1.966644323622921e-06, + "learning_rate": 8.348801979108691e-05, + "loss": 0.0, + "step": 8947 + }, + { + "epoch": 4.210823529411765, + "grad_norm": 0.10590173304295165, + "learning_rate": 8.344873150770562e-05, + "loss": 0.0393, + "step": 8948 + }, + { + "epoch": 4.211294117647059, + "grad_norm": 0.062309571010645645, + "learning_rate": 8.340944890812916e-05, + "loss": 0.0098, + "step": 8949 + }, + { + "epoch": 4.211764705882353, + "grad_norm": 0.14199130011519492, + "learning_rate": 8.337017199571235e-05, + "loss": 0.0505, + "step": 8950 + }, + { + "epoch": 4.212235294117647, + "grad_norm": 0.1141596295857219, + "learning_rate": 8.333090077380964e-05, + "loss": 0.0099, + "step": 8951 + }, + { + "epoch": 4.2127058823529415, + "grad_norm": 0.07080424006904627, + "learning_rate": 8.329163524577504e-05, + "loss": 0.0189, + "step": 8952 + }, + { + "epoch": 4.213176470588236, + "grad_norm": 0.14828850592850054, + "learning_rate": 8.325237541496197e-05, + "loss": 0.0676, + "step": 8953 + }, + { + "epoch": 4.213647058823529, + "grad_norm": 6.267171486623541e-07, + "learning_rate": 8.32131212847235e-05, + "loss": 0.0, + "step": 8954 + }, + { + "epoch": 4.214117647058823, + "grad_norm": 0.0322851999000979, + "learning_rate": 8.3173872858412e-05, + "loss": 0.0027, + "step": 8955 + }, + { + "epoch": 4.2145882352941175, + "grad_norm": 0.10236575420543743, + "learning_rate": 8.313463013937953e-05, + "loss": 0.0236, + "step": 8956 + }, + { + "epoch": 4.215058823529412, + "grad_norm": 0.16950237602787227, + "learning_rate": 8.309539313097769e-05, + "loss": 0.06, + "step": 8957 + }, + { + "epoch": 4.215529411764706, + "grad_norm": 0.14410385190773348, + "learning_rate": 8.305616183655738e-05, + "loss": 0.0469, + "step": 8958 + }, + { + "epoch": 4.216, + "grad_norm": 0.086282592035106, + "learning_rate": 8.30169362594692e-05, + "loss": 0.0226, + "step": 8959 + }, + { + "epoch": 4.216470588235294, + "grad_norm": 0.20549287597570964, + "learning_rate": 8.29777164030633e-05, + "loss": 0.058, + "step": 8960 + }, + { + "epoch": 4.216941176470589, + "grad_norm": 0.07609317411397482, + "learning_rate": 8.293850227068908e-05, + "loss": 0.0152, + "step": 8961 + }, + { + "epoch": 4.217411764705882, + "grad_norm": 0.08435562369563472, + "learning_rate": 8.289929386569569e-05, + "loss": 0.0232, + "step": 8962 + }, + { + "epoch": 4.217882352941176, + "grad_norm": 0.09546594755764884, + "learning_rate": 8.28600911914318e-05, + "loss": 0.0126, + "step": 8963 + }, + { + "epoch": 4.21835294117647, + "grad_norm": 0.07464849544140513, + "learning_rate": 8.282089425124538e-05, + "loss": 0.0049, + "step": 8964 + }, + { + "epoch": 4.218823529411765, + "grad_norm": 0.098426425955991, + "learning_rate": 8.278170304848413e-05, + "loss": 0.0213, + "step": 8965 + }, + { + "epoch": 4.219294117647059, + "grad_norm": 0.19303197271339623, + "learning_rate": 8.274251758649518e-05, + "loss": 0.0654, + "step": 8966 + }, + { + "epoch": 4.219764705882353, + "grad_norm": 2.216794634956462e-06, + "learning_rate": 8.270333786862506e-05, + "loss": 0.0, + "step": 8967 + }, + { + "epoch": 4.220235294117647, + "grad_norm": 0.13283303221640655, + "learning_rate": 8.266416389821997e-05, + "loss": 0.0613, + "step": 8968 + }, + { + "epoch": 4.2207058823529415, + "grad_norm": 0.05087290357737215, + "learning_rate": 8.262499567862565e-05, + "loss": 0.0102, + "step": 8969 + }, + { + "epoch": 4.221176470588235, + "grad_norm": 0.056454918119722475, + "learning_rate": 8.258583321318708e-05, + "loss": 0.0091, + "step": 8970 + }, + { + "epoch": 4.221647058823529, + "grad_norm": 2.0702524299142336e-05, + "learning_rate": 8.254667650524904e-05, + "loss": 0.0, + "step": 8971 + }, + { + "epoch": 4.222117647058823, + "grad_norm": 0.0019264575428708899, + "learning_rate": 8.250752555815574e-05, + "loss": 0.0, + "step": 8972 + }, + { + "epoch": 4.2225882352941175, + "grad_norm": 2.559048309800865e-06, + "learning_rate": 8.246838037525073e-05, + "loss": 0.0, + "step": 8973 + }, + { + "epoch": 4.223058823529412, + "grad_norm": 3.691721225577581e-06, + "learning_rate": 8.242924095987736e-05, + "loss": 0.0, + "step": 8974 + }, + { + "epoch": 4.223529411764706, + "grad_norm": 8.463999373252024e-06, + "learning_rate": 8.239010731537816e-05, + "loss": 0.0, + "step": 8975 + }, + { + "epoch": 4.224, + "grad_norm": 0.0989949638327847, + "learning_rate": 8.235097944509545e-05, + "loss": 0.0253, + "step": 8976 + }, + { + "epoch": 4.224470588235294, + "grad_norm": 0.12089954032623466, + "learning_rate": 8.231185735237096e-05, + "loss": 0.022, + "step": 8977 + }, + { + "epoch": 4.224941176470589, + "grad_norm": 1.1763372263812925e-06, + "learning_rate": 8.22727410405458e-05, + "loss": 0.0, + "step": 8978 + }, + { + "epoch": 4.225411764705882, + "grad_norm": 7.2684786025664804e-06, + "learning_rate": 8.223363051296074e-05, + "loss": 0.0, + "step": 8979 + }, + { + "epoch": 4.225882352941176, + "grad_norm": 0.056851782424082974, + "learning_rate": 8.21945257729561e-05, + "loss": 0.0111, + "step": 8980 + }, + { + "epoch": 4.22635294117647, + "grad_norm": 0.04669714867942592, + "learning_rate": 8.215542682387148e-05, + "loss": 0.0043, + "step": 8981 + }, + { + "epoch": 4.226823529411765, + "grad_norm": 9.355911572646417e-05, + "learning_rate": 8.211633366904618e-05, + "loss": 0.0, + "step": 8982 + }, + { + "epoch": 4.227294117647059, + "grad_norm": 0.1119190067814316, + "learning_rate": 8.207724631181904e-05, + "loss": 0.0412, + "step": 8983 + }, + { + "epoch": 4.227764705882353, + "grad_norm": 0.2375584398379637, + "learning_rate": 8.203816475552813e-05, + "loss": 0.0561, + "step": 8984 + }, + { + "epoch": 4.228235294117647, + "grad_norm": 0.06586292237387026, + "learning_rate": 8.19990890035113e-05, + "loss": 0.0111, + "step": 8985 + }, + { + "epoch": 4.2287058823529415, + "grad_norm": 0.00016961357393143096, + "learning_rate": 8.196001905910589e-05, + "loss": 0.0, + "step": 8986 + }, + { + "epoch": 4.229176470588235, + "grad_norm": 0.12751424891804053, + "learning_rate": 8.192095492564851e-05, + "loss": 0.0496, + "step": 8987 + }, + { + "epoch": 4.229647058823529, + "grad_norm": 6.37475474457353e-06, + "learning_rate": 8.188189660647554e-05, + "loss": 0.0, + "step": 8988 + }, + { + "epoch": 4.230117647058823, + "grad_norm": 6.83215628302085e-06, + "learning_rate": 8.184284410492274e-05, + "loss": 0.0, + "step": 8989 + }, + { + "epoch": 4.2305882352941175, + "grad_norm": 2.0598168243127022e-05, + "learning_rate": 8.180379742432533e-05, + "loss": 0.0, + "step": 8990 + }, + { + "epoch": 4.231058823529412, + "grad_norm": 0.1531072072560808, + "learning_rate": 8.176475656801811e-05, + "loss": 0.0425, + "step": 8991 + }, + { + "epoch": 4.231529411764706, + "grad_norm": 0.09245738178246209, + "learning_rate": 8.172572153933541e-05, + "loss": 0.0207, + "step": 8992 + }, + { + "epoch": 4.232, + "grad_norm": 0.10618525670024637, + "learning_rate": 8.168669234161094e-05, + "loss": 0.0208, + "step": 8993 + }, + { + "epoch": 4.232470588235294, + "grad_norm": 0.10764659723091743, + "learning_rate": 8.164766897817806e-05, + "loss": 0.033, + "step": 8994 + }, + { + "epoch": 4.232941176470589, + "grad_norm": 9.998859500132147e-06, + "learning_rate": 8.160865145236946e-05, + "loss": 0.0, + "step": 8995 + }, + { + "epoch": 4.233411764705882, + "grad_norm": 1.9631783562562344e-05, + "learning_rate": 8.15696397675175e-05, + "loss": 0.0, + "step": 8996 + }, + { + "epoch": 4.233882352941176, + "grad_norm": 1.0878311456127676e-05, + "learning_rate": 8.153063392695394e-05, + "loss": 0.0, + "step": 8997 + }, + { + "epoch": 4.23435294117647, + "grad_norm": 0.11403329260505522, + "learning_rate": 8.14916339340101e-05, + "loss": 0.0296, + "step": 8998 + }, + { + "epoch": 4.234823529411765, + "grad_norm": 0.14705796404553312, + "learning_rate": 8.145263979201672e-05, + "loss": 0.0397, + "step": 8999 + }, + { + "epoch": 4.235294117647059, + "grad_norm": 0.18253089166905354, + "learning_rate": 8.141365150430421e-05, + "loss": 0.0833, + "step": 9000 + }, + { + "epoch": 4.235764705882353, + "grad_norm": 0.0978988208723438, + "learning_rate": 8.137466907420222e-05, + "loss": 0.0217, + "step": 9001 + }, + { + "epoch": 4.236235294117647, + "grad_norm": 0.1372383746687185, + "learning_rate": 8.133569250504013e-05, + "loss": 0.0362, + "step": 9002 + }, + { + "epoch": 4.2367058823529415, + "grad_norm": 0.011301978322834203, + "learning_rate": 8.129672180014666e-05, + "loss": 0.0002, + "step": 9003 + }, + { + "epoch": 4.237176470588235, + "grad_norm": 0.11090454761164618, + "learning_rate": 8.125775696285013e-05, + "loss": 0.032, + "step": 9004 + }, + { + "epoch": 4.237647058823529, + "grad_norm": 0.06944493321706366, + "learning_rate": 8.121879799647839e-05, + "loss": 0.0093, + "step": 9005 + }, + { + "epoch": 4.238117647058823, + "grad_norm": 0.022339877046281386, + "learning_rate": 8.117984490435861e-05, + "loss": 0.0018, + "step": 9006 + }, + { + "epoch": 4.2385882352941175, + "grad_norm": 0.04541652990930953, + "learning_rate": 8.114089768981765e-05, + "loss": 0.0049, + "step": 9007 + }, + { + "epoch": 4.239058823529412, + "grad_norm": 0.00037859850681411884, + "learning_rate": 8.110195635618184e-05, + "loss": 0.0, + "step": 9008 + }, + { + "epoch": 4.239529411764706, + "grad_norm": 0.13018377182494453, + "learning_rate": 8.106302090677682e-05, + "loss": 0.0336, + "step": 9009 + }, + { + "epoch": 4.24, + "grad_norm": 0.12027662318940834, + "learning_rate": 8.102409134492795e-05, + "loss": 0.0342, + "step": 9010 + }, + { + "epoch": 4.240470588235294, + "grad_norm": 4.947474368768321e-06, + "learning_rate": 8.098516767396008e-05, + "loss": 0.0, + "step": 9011 + }, + { + "epoch": 4.240941176470589, + "grad_norm": 0.1370223501607456, + "learning_rate": 8.094624989719734e-05, + "loss": 0.0468, + "step": 9012 + }, + { + "epoch": 4.241411764705882, + "grad_norm": 0.04200476463485459, + "learning_rate": 8.090733801796356e-05, + "loss": 0.0041, + "step": 9013 + }, + { + "epoch": 4.241882352941176, + "grad_norm": 0.018124701251922785, + "learning_rate": 8.086843203958207e-05, + "loss": 0.001, + "step": 9014 + }, + { + "epoch": 4.24235294117647, + "grad_norm": 0.09075996667953701, + "learning_rate": 8.082953196537553e-05, + "loss": 0.0174, + "step": 9015 + }, + { + "epoch": 4.242823529411765, + "grad_norm": 0.10029030560992488, + "learning_rate": 8.079063779866622e-05, + "loss": 0.0272, + "step": 9016 + }, + { + "epoch": 4.243294117647059, + "grad_norm": 0.08500197194818998, + "learning_rate": 8.075174954277598e-05, + "loss": 0.0249, + "step": 9017 + }, + { + "epoch": 4.243764705882353, + "grad_norm": 0.13091542677855586, + "learning_rate": 8.071286720102594e-05, + "loss": 0.0247, + "step": 9018 + }, + { + "epoch": 4.244235294117647, + "grad_norm": 0.056909915329386704, + "learning_rate": 8.067399077673688e-05, + "loss": 0.0054, + "step": 9019 + }, + { + "epoch": 4.2447058823529416, + "grad_norm": 0.06586531596417741, + "learning_rate": 8.06351202732291e-05, + "loss": 0.0134, + "step": 9020 + }, + { + "epoch": 4.245176470588235, + "grad_norm": 0.06087198786725079, + "learning_rate": 8.059625569382223e-05, + "loss": 0.0076, + "step": 9021 + }, + { + "epoch": 4.245647058823529, + "grad_norm": 0.12548891975461812, + "learning_rate": 8.055739704183553e-05, + "loss": 0.0263, + "step": 9022 + }, + { + "epoch": 4.246117647058823, + "grad_norm": 0.1145148586865351, + "learning_rate": 8.051854432058785e-05, + "loss": 0.0334, + "step": 9023 + }, + { + "epoch": 4.2465882352941176, + "grad_norm": 0.07821593640404084, + "learning_rate": 8.047969753339717e-05, + "loss": 0.0087, + "step": 9024 + }, + { + "epoch": 4.247058823529412, + "grad_norm": 0.1301779249224235, + "learning_rate": 8.04408566835814e-05, + "loss": 0.0287, + "step": 9025 + }, + { + "epoch": 4.247529411764706, + "grad_norm": 0.13871072862711534, + "learning_rate": 8.040202177445761e-05, + "loss": 0.0433, + "step": 9026 + }, + { + "epoch": 4.248, + "grad_norm": 0.005419459387744593, + "learning_rate": 8.036319280934251e-05, + "loss": 0.0001, + "step": 9027 + }, + { + "epoch": 4.248470588235294, + "grad_norm": 0.03151477745988527, + "learning_rate": 8.032436979155237e-05, + "loss": 0.0015, + "step": 9028 + }, + { + "epoch": 4.248941176470589, + "grad_norm": 0.019025231151696183, + "learning_rate": 8.02855527244028e-05, + "loss": 0.0009, + "step": 9029 + }, + { + "epoch": 4.249411764705882, + "grad_norm": 0.08073634843860378, + "learning_rate": 8.024674161120892e-05, + "loss": 0.016, + "step": 9030 + }, + { + "epoch": 4.249882352941176, + "grad_norm": 3.553878716517744e-06, + "learning_rate": 8.020793645528552e-05, + "loss": 0.0, + "step": 9031 + }, + { + "epoch": 4.2503529411764704, + "grad_norm": 0.02779831804333469, + "learning_rate": 8.016913725994661e-05, + "loss": 0.0027, + "step": 9032 + }, + { + "epoch": 4.250823529411765, + "grad_norm": 0.04846913979339466, + "learning_rate": 8.01303440285059e-05, + "loss": 0.0037, + "step": 9033 + }, + { + "epoch": 4.251294117647059, + "grad_norm": 0.053460045722153356, + "learning_rate": 8.009155676427659e-05, + "loss": 0.0042, + "step": 9034 + }, + { + "epoch": 4.251764705882353, + "grad_norm": 0.13751754083852288, + "learning_rate": 8.005277547057114e-05, + "loss": 0.0399, + "step": 9035 + }, + { + "epoch": 4.252235294117647, + "grad_norm": 0.0735239743166236, + "learning_rate": 8.001400015070178e-05, + "loss": 0.0162, + "step": 9036 + }, + { + "epoch": 4.252705882352942, + "grad_norm": 0.1268705669099121, + "learning_rate": 7.997523080798014e-05, + "loss": 0.0309, + "step": 9037 + }, + { + "epoch": 4.253176470588235, + "grad_norm": 0.07789619791103787, + "learning_rate": 7.993646744571719e-05, + "loss": 0.0117, + "step": 9038 + }, + { + "epoch": 4.253647058823529, + "grad_norm": 0.03730490065594283, + "learning_rate": 7.98977100672236e-05, + "loss": 0.002, + "step": 9039 + }, + { + "epoch": 4.254117647058823, + "grad_norm": 0.07332768923679622, + "learning_rate": 7.985895867580948e-05, + "loss": 0.0107, + "step": 9040 + }, + { + "epoch": 4.254588235294118, + "grad_norm": 0.14054375290234233, + "learning_rate": 7.982021327478419e-05, + "loss": 0.0463, + "step": 9041 + }, + { + "epoch": 4.255058823529412, + "grad_norm": 0.08795969744279532, + "learning_rate": 7.978147386745705e-05, + "loss": 0.0161, + "step": 9042 + }, + { + "epoch": 4.255529411764706, + "grad_norm": 0.1808181441421032, + "learning_rate": 7.974274045713638e-05, + "loss": 0.0697, + "step": 9043 + }, + { + "epoch": 4.256, + "grad_norm": 0.07560534194047605, + "learning_rate": 7.97040130471303e-05, + "loss": 0.0175, + "step": 9044 + }, + { + "epoch": 4.2564705882352944, + "grad_norm": 0.12237731209908742, + "learning_rate": 7.966529164074636e-05, + "loss": 0.0345, + "step": 9045 + }, + { + "epoch": 4.256941176470589, + "grad_norm": 0.14218125385407776, + "learning_rate": 7.962657624129147e-05, + "loss": 0.0445, + "step": 9046 + }, + { + "epoch": 4.257411764705882, + "grad_norm": 0.08601909025413404, + "learning_rate": 7.958786685207212e-05, + "loss": 0.0166, + "step": 9047 + }, + { + "epoch": 4.257882352941176, + "grad_norm": 0.03829455467396489, + "learning_rate": 7.954916347639438e-05, + "loss": 0.0042, + "step": 9048 + }, + { + "epoch": 4.2583529411764705, + "grad_norm": 0.00014404944865939413, + "learning_rate": 7.951046611756358e-05, + "loss": 0.0, + "step": 9049 + }, + { + "epoch": 4.258823529411765, + "grad_norm": 0.10246496374214238, + "learning_rate": 7.947177477888472e-05, + "loss": 0.0142, + "step": 9050 + }, + { + "epoch": 4.259294117647059, + "grad_norm": 0.1882579787151485, + "learning_rate": 7.94330894636623e-05, + "loss": 0.0171, + "step": 9051 + }, + { + "epoch": 4.259764705882353, + "grad_norm": 0.14132072299304838, + "learning_rate": 7.939441017520011e-05, + "loss": 0.0162, + "step": 9052 + }, + { + "epoch": 4.260235294117647, + "grad_norm": 0.0003923486675235949, + "learning_rate": 7.935573691680166e-05, + "loss": 0.0, + "step": 9053 + }, + { + "epoch": 4.260705882352942, + "grad_norm": 0.15069760077494884, + "learning_rate": 7.931706969176972e-05, + "loss": 0.0459, + "step": 9054 + }, + { + "epoch": 4.261176470588235, + "grad_norm": 5.1870089984318536e-05, + "learning_rate": 7.927840850340673e-05, + "loss": 0.0, + "step": 9055 + }, + { + "epoch": 4.261647058823529, + "grad_norm": 0.10832806203462694, + "learning_rate": 7.92397533550146e-05, + "loss": 0.0294, + "step": 9056 + }, + { + "epoch": 4.262117647058823, + "grad_norm": 0.12857304041198717, + "learning_rate": 7.920110424989455e-05, + "loss": 0.0347, + "step": 9057 + }, + { + "epoch": 4.262588235294118, + "grad_norm": 0.18618692770253115, + "learning_rate": 7.916246119134748e-05, + "loss": 0.0288, + "step": 9058 + }, + { + "epoch": 4.263058823529412, + "grad_norm": 0.16011993567432775, + "learning_rate": 7.912382418267374e-05, + "loss": 0.0401, + "step": 9059 + }, + { + "epoch": 4.263529411764706, + "grad_norm": 0.0936050127979105, + "learning_rate": 7.908519322717297e-05, + "loss": 0.0148, + "step": 9060 + }, + { + "epoch": 4.264, + "grad_norm": 2.5848679467910583e-05, + "learning_rate": 7.904656832814457e-05, + "loss": 0.0, + "step": 9061 + }, + { + "epoch": 4.2644705882352945, + "grad_norm": 0.0003727961419752774, + "learning_rate": 7.900794948888728e-05, + "loss": 0.0, + "step": 9062 + }, + { + "epoch": 4.264941176470588, + "grad_norm": 0.13358608495195337, + "learning_rate": 7.896933671269929e-05, + "loss": 0.0444, + "step": 9063 + }, + { + "epoch": 4.265411764705882, + "grad_norm": 0.10256046656746133, + "learning_rate": 7.893073000287834e-05, + "loss": 0.0238, + "step": 9064 + }, + { + "epoch": 4.265882352941176, + "grad_norm": 0.07028300489935857, + "learning_rate": 7.88921293627217e-05, + "loss": 0.0119, + "step": 9065 + }, + { + "epoch": 4.2663529411764705, + "grad_norm": 0.15377470871493923, + "learning_rate": 7.885353479552592e-05, + "loss": 0.0516, + "step": 9066 + }, + { + "epoch": 4.266823529411765, + "grad_norm": 0.0015940637026016597, + "learning_rate": 7.881494630458726e-05, + "loss": 0.0, + "step": 9067 + }, + { + "epoch": 4.267294117647059, + "grad_norm": 0.0005896529943901621, + "learning_rate": 7.87763638932014e-05, + "loss": 0.0, + "step": 9068 + }, + { + "epoch": 4.267764705882353, + "grad_norm": 0.05612926079020917, + "learning_rate": 7.873778756466335e-05, + "loss": 0.0083, + "step": 9069 + }, + { + "epoch": 4.268235294117647, + "grad_norm": 0.1465605966275138, + "learning_rate": 7.869921732226779e-05, + "loss": 0.0618, + "step": 9070 + }, + { + "epoch": 4.268705882352942, + "grad_norm": 0.31019656672221246, + "learning_rate": 7.866065316930886e-05, + "loss": 0.0885, + "step": 9071 + }, + { + "epoch": 4.269176470588235, + "grad_norm": 0.11247211640907824, + "learning_rate": 7.862209510907999e-05, + "loss": 0.0379, + "step": 9072 + }, + { + "epoch": 4.269647058823529, + "grad_norm": 0.1487677192782596, + "learning_rate": 7.858354314487432e-05, + "loss": 0.0343, + "step": 9073 + }, + { + "epoch": 4.270117647058823, + "grad_norm": 0.08836937242248069, + "learning_rate": 7.854499727998442e-05, + "loss": 0.0133, + "step": 9074 + }, + { + "epoch": 4.270588235294118, + "grad_norm": 0.1313770921338554, + "learning_rate": 7.850645751770217e-05, + "loss": 0.0307, + "step": 9075 + }, + { + "epoch": 4.271058823529412, + "grad_norm": 0.002948519723064487, + "learning_rate": 7.846792386131917e-05, + "loss": 0.0001, + "step": 9076 + }, + { + "epoch": 4.271529411764706, + "grad_norm": 0.0828853356285608, + "learning_rate": 7.84293963141263e-05, + "loss": 0.0203, + "step": 9077 + }, + { + "epoch": 4.272, + "grad_norm": 0.0492137203373845, + "learning_rate": 7.839087487941403e-05, + "loss": 0.0082, + "step": 9078 + }, + { + "epoch": 4.2724705882352945, + "grad_norm": 0.09604739827739278, + "learning_rate": 7.835235956047234e-05, + "loss": 0.0332, + "step": 9079 + }, + { + "epoch": 4.272941176470589, + "grad_norm": 0.12437284938734662, + "learning_rate": 7.83138503605905e-05, + "loss": 0.0419, + "step": 9080 + }, + { + "epoch": 4.273411764705882, + "grad_norm": 0.14545890624295982, + "learning_rate": 7.827534728305747e-05, + "loss": 0.032, + "step": 9081 + }, + { + "epoch": 4.273882352941176, + "grad_norm": 0.04554674462425348, + "learning_rate": 7.823685033116166e-05, + "loss": 0.0059, + "step": 9082 + }, + { + "epoch": 4.2743529411764705, + "grad_norm": 0.05766653513205523, + "learning_rate": 7.819835950819078e-05, + "loss": 0.0125, + "step": 9083 + }, + { + "epoch": 4.274823529411765, + "grad_norm": 1.755688015521612e-05, + "learning_rate": 7.815987481743215e-05, + "loss": 0.0, + "step": 9084 + }, + { + "epoch": 4.275294117647059, + "grad_norm": 0.00025821038395676957, + "learning_rate": 7.812139626217266e-05, + "loss": 0.0, + "step": 9085 + }, + { + "epoch": 4.275764705882353, + "grad_norm": 2.2654428372799835e-05, + "learning_rate": 7.808292384569838e-05, + "loss": 0.0, + "step": 9086 + }, + { + "epoch": 4.276235294117647, + "grad_norm": 0.033384092745447715, + "learning_rate": 7.804445757129525e-05, + "loss": 0.0017, + "step": 9087 + }, + { + "epoch": 4.276705882352942, + "grad_norm": 0.03158161462114735, + "learning_rate": 7.800599744224834e-05, + "loss": 0.0033, + "step": 9088 + }, + { + "epoch": 4.277176470588235, + "grad_norm": 4.443216011252222e-05, + "learning_rate": 7.796754346184238e-05, + "loss": 0.0, + "step": 9089 + }, + { + "epoch": 4.277647058823529, + "grad_norm": 0.1673818292487371, + "learning_rate": 7.792909563336157e-05, + "loss": 0.0675, + "step": 9090 + }, + { + "epoch": 4.278117647058823, + "grad_norm": 0.10378882844510998, + "learning_rate": 7.789065396008945e-05, + "loss": 0.0346, + "step": 9091 + }, + { + "epoch": 4.278588235294118, + "grad_norm": 0.0002882910665862272, + "learning_rate": 7.785221844530918e-05, + "loss": 0.0, + "step": 9092 + }, + { + "epoch": 4.279058823529412, + "grad_norm": 0.08117365671499893, + "learning_rate": 7.78137890923034e-05, + "loss": 0.0182, + "step": 9093 + }, + { + "epoch": 4.279529411764706, + "grad_norm": 0.02139915843988383, + "learning_rate": 7.777536590435404e-05, + "loss": 0.0019, + "step": 9094 + }, + { + "epoch": 4.28, + "grad_norm": 0.16132183782267878, + "learning_rate": 7.773694888474267e-05, + "loss": 0.0661, + "step": 9095 + }, + { + "epoch": 4.2804705882352945, + "grad_norm": 0.10673532160993554, + "learning_rate": 7.769853803675041e-05, + "loss": 0.0352, + "step": 9096 + }, + { + "epoch": 4.280941176470588, + "grad_norm": 0.06489298383104052, + "learning_rate": 7.766013336365755e-05, + "loss": 0.007, + "step": 9097 + }, + { + "epoch": 4.281411764705882, + "grad_norm": 1.95379707326027e-06, + "learning_rate": 7.762173486874414e-05, + "loss": 0.0, + "step": 9098 + }, + { + "epoch": 4.281882352941176, + "grad_norm": 0.061577222798221566, + "learning_rate": 7.758334255528963e-05, + "loss": 0.0113, + "step": 9099 + }, + { + "epoch": 4.2823529411764705, + "grad_norm": 8.116745171946966e-06, + "learning_rate": 7.754495642657282e-05, + "loss": 0.0, + "step": 9100 + }, + { + "epoch": 4.282823529411765, + "grad_norm": 0.14869328260784556, + "learning_rate": 7.750657648587213e-05, + "loss": 0.0438, + "step": 9101 + }, + { + "epoch": 4.283294117647059, + "grad_norm": 0.08410911003146965, + "learning_rate": 7.746820273646542e-05, + "loss": 0.0247, + "step": 9102 + }, + { + "epoch": 4.283764705882353, + "grad_norm": 0.12401952434883097, + "learning_rate": 7.742983518162993e-05, + "loss": 0.0312, + "step": 9103 + }, + { + "epoch": 4.284235294117647, + "grad_norm": 0.05666903634486668, + "learning_rate": 7.73914738246425e-05, + "loss": 0.0124, + "step": 9104 + }, + { + "epoch": 4.284705882352942, + "grad_norm": 0.13240900716115536, + "learning_rate": 7.735311866877933e-05, + "loss": 0.0411, + "step": 9105 + }, + { + "epoch": 4.285176470588235, + "grad_norm": 0.14524744963060776, + "learning_rate": 7.731476971731612e-05, + "loss": 0.0304, + "step": 9106 + }, + { + "epoch": 4.285647058823529, + "grad_norm": 0.07950540014017726, + "learning_rate": 7.727642697352817e-05, + "loss": 0.0129, + "step": 9107 + }, + { + "epoch": 4.286117647058823, + "grad_norm": 0.10458495929247709, + "learning_rate": 7.723809044069e-05, + "loss": 0.0367, + "step": 9108 + }, + { + "epoch": 4.286588235294118, + "grad_norm": 0.0003762763405567025, + "learning_rate": 7.719976012207579e-05, + "loss": 0.0, + "step": 9109 + }, + { + "epoch": 4.287058823529412, + "grad_norm": 0.1107518945285297, + "learning_rate": 7.716143602095924e-05, + "loss": 0.0084, + "step": 9110 + }, + { + "epoch": 4.287529411764706, + "grad_norm": 0.09471702324167386, + "learning_rate": 7.712311814061323e-05, + "loss": 0.0196, + "step": 9111 + }, + { + "epoch": 4.288, + "grad_norm": 0.11136254706281282, + "learning_rate": 7.708480648431043e-05, + "loss": 0.0322, + "step": 9112 + }, + { + "epoch": 4.2884705882352945, + "grad_norm": 0.14860127520973296, + "learning_rate": 7.704650105532284e-05, + "loss": 0.0455, + "step": 9113 + }, + { + "epoch": 4.288941176470589, + "grad_norm": 0.20536811821878181, + "learning_rate": 7.700820185692184e-05, + "loss": 0.081, + "step": 9114 + }, + { + "epoch": 4.289411764705882, + "grad_norm": 0.16387332278949254, + "learning_rate": 7.69699088923784e-05, + "loss": 0.063, + "step": 9115 + }, + { + "epoch": 4.289882352941176, + "grad_norm": 0.054362926656327484, + "learning_rate": 7.693162216496303e-05, + "loss": 0.0063, + "step": 9116 + }, + { + "epoch": 4.2903529411764705, + "grad_norm": 0.17628369391463664, + "learning_rate": 7.689334167794547e-05, + "loss": 0.0293, + "step": 9117 + }, + { + "epoch": 4.290823529411765, + "grad_norm": 0.16243869139131517, + "learning_rate": 7.685506743459511e-05, + "loss": 0.0574, + "step": 9118 + }, + { + "epoch": 4.291294117647059, + "grad_norm": 0.11134345003568602, + "learning_rate": 7.681679943818082e-05, + "loss": 0.0323, + "step": 9119 + }, + { + "epoch": 4.291764705882353, + "grad_norm": 0.06752992197370272, + "learning_rate": 7.677853769197078e-05, + "loss": 0.01, + "step": 9120 + }, + { + "epoch": 4.292235294117647, + "grad_norm": 0.19951296060203633, + "learning_rate": 7.674028219923275e-05, + "loss": 0.0324, + "step": 9121 + }, + { + "epoch": 4.292705882352941, + "grad_norm": 0.08375783533706761, + "learning_rate": 7.670203296323403e-05, + "loss": 0.0146, + "step": 9122 + }, + { + "epoch": 4.293176470588235, + "grad_norm": 0.1255835592121245, + "learning_rate": 7.666378998724116e-05, + "loss": 0.0391, + "step": 9123 + }, + { + "epoch": 4.293647058823529, + "grad_norm": 0.03577919032396299, + "learning_rate": 7.662555327452031e-05, + "loss": 0.003, + "step": 9124 + }, + { + "epoch": 4.294117647058823, + "grad_norm": 7.859281174247551e-06, + "learning_rate": 7.65873228283372e-05, + "loss": 0.0, + "step": 9125 + }, + { + "epoch": 4.294588235294118, + "grad_norm": 0.13147594517884004, + "learning_rate": 7.654909865195675e-05, + "loss": 0.0227, + "step": 9126 + }, + { + "epoch": 4.295058823529412, + "grad_norm": 5.125554810514404e-05, + "learning_rate": 7.65108807486436e-05, + "loss": 0.0, + "step": 9127 + }, + { + "epoch": 4.295529411764706, + "grad_norm": 0.12221636766572745, + "learning_rate": 7.647266912166162e-05, + "loss": 0.0305, + "step": 9128 + }, + { + "epoch": 4.296, + "grad_norm": 0.14403926888950627, + "learning_rate": 7.643446377427438e-05, + "loss": 0.0412, + "step": 9129 + }, + { + "epoch": 4.2964705882352945, + "grad_norm": 0.2621020728954112, + "learning_rate": 7.639626470974474e-05, + "loss": 0.0455, + "step": 9130 + }, + { + "epoch": 4.296941176470588, + "grad_norm": 6.841615544678296e-06, + "learning_rate": 7.635807193133512e-05, + "loss": 0.0, + "step": 9131 + }, + { + "epoch": 4.297411764705882, + "grad_norm": 0.10851796670081973, + "learning_rate": 7.631988544230745e-05, + "loss": 0.0231, + "step": 9132 + }, + { + "epoch": 4.297882352941176, + "grad_norm": 1.3673060854873239e-05, + "learning_rate": 7.628170524592288e-05, + "loss": 0.0, + "step": 9133 + }, + { + "epoch": 4.2983529411764705, + "grad_norm": 0.06505436418823836, + "learning_rate": 7.624353134544229e-05, + "loss": 0.0087, + "step": 9134 + }, + { + "epoch": 4.298823529411765, + "grad_norm": 0.06341718892696072, + "learning_rate": 7.620536374412594e-05, + "loss": 0.0116, + "step": 9135 + }, + { + "epoch": 4.299294117647059, + "grad_norm": 0.08912553000232067, + "learning_rate": 7.616720244523342e-05, + "loss": 0.0224, + "step": 9136 + }, + { + "epoch": 4.299764705882353, + "grad_norm": 0.08334437733938888, + "learning_rate": 7.612904745202399e-05, + "loss": 0.0136, + "step": 9137 + }, + { + "epoch": 4.300235294117647, + "grad_norm": 0.053678571251303396, + "learning_rate": 7.609089876775628e-05, + "loss": 0.007, + "step": 9138 + }, + { + "epoch": 4.300705882352942, + "grad_norm": 0.10935865580339316, + "learning_rate": 7.605275639568828e-05, + "loss": 0.0198, + "step": 9139 + }, + { + "epoch": 4.301176470588235, + "grad_norm": 0.09771808338881159, + "learning_rate": 7.60146203390776e-05, + "loss": 0.0357, + "step": 9140 + }, + { + "epoch": 4.301647058823529, + "grad_norm": 0.05952921687393543, + "learning_rate": 7.59764906011813e-05, + "loss": 0.0096, + "step": 9141 + }, + { + "epoch": 4.302117647058823, + "grad_norm": 0.11802613980844545, + "learning_rate": 7.593836718525572e-05, + "loss": 0.0324, + "step": 9142 + }, + { + "epoch": 4.302588235294118, + "grad_norm": 0.032960339825009684, + "learning_rate": 7.590025009455684e-05, + "loss": 0.0021, + "step": 9143 + }, + { + "epoch": 4.303058823529412, + "grad_norm": 0.11359468327388328, + "learning_rate": 7.586213933234013e-05, + "loss": 0.0347, + "step": 9144 + }, + { + "epoch": 4.303529411764706, + "grad_norm": 0.16471985662121275, + "learning_rate": 7.58240349018603e-05, + "loss": 0.0366, + "step": 9145 + }, + { + "epoch": 4.304, + "grad_norm": 0.043507588801921765, + "learning_rate": 7.578593680637173e-05, + "loss": 0.0042, + "step": 9146 + }, + { + "epoch": 4.3044705882352945, + "grad_norm": 0.1226894392359756, + "learning_rate": 7.574784504912823e-05, + "loss": 0.0427, + "step": 9147 + }, + { + "epoch": 4.304941176470588, + "grad_norm": 0.05740524587520855, + "learning_rate": 7.57097596333829e-05, + "loss": 0.0062, + "step": 9148 + }, + { + "epoch": 4.305411764705882, + "grad_norm": 0.058542904462396086, + "learning_rate": 7.567168056238851e-05, + "loss": 0.0076, + "step": 9149 + }, + { + "epoch": 4.305882352941176, + "grad_norm": 0.03343536139498916, + "learning_rate": 7.563360783939722e-05, + "loss": 0.0036, + "step": 9150 + }, + { + "epoch": 4.3063529411764705, + "grad_norm": 0.08172109313060066, + "learning_rate": 7.559554146766054e-05, + "loss": 0.0205, + "step": 9151 + }, + { + "epoch": 4.306823529411765, + "grad_norm": 0.1556583699416785, + "learning_rate": 7.555748145042957e-05, + "loss": 0.0585, + "step": 9152 + }, + { + "epoch": 4.307294117647059, + "grad_norm": 0.05018305486618013, + "learning_rate": 7.551942779095486e-05, + "loss": 0.0079, + "step": 9153 + }, + { + "epoch": 4.307764705882353, + "grad_norm": 0.16617290039794058, + "learning_rate": 7.54813804924863e-05, + "loss": 0.0551, + "step": 9154 + }, + { + "epoch": 4.308235294117647, + "grad_norm": 0.1803972147716636, + "learning_rate": 7.544333955827342e-05, + "loss": 0.0794, + "step": 9155 + }, + { + "epoch": 4.308705882352941, + "grad_norm": 0.15685852610073348, + "learning_rate": 7.540530499156496e-05, + "loss": 0.0319, + "step": 9156 + }, + { + "epoch": 4.309176470588235, + "grad_norm": 0.15394071191606928, + "learning_rate": 7.536727679560937e-05, + "loss": 0.0628, + "step": 9157 + }, + { + "epoch": 4.309647058823529, + "grad_norm": 0.12402437955767975, + "learning_rate": 7.532925497365444e-05, + "loss": 0.0497, + "step": 9158 + }, + { + "epoch": 4.310117647058823, + "grad_norm": 0.058646145418223786, + "learning_rate": 7.529123952894734e-05, + "loss": 0.0086, + "step": 9159 + }, + { + "epoch": 4.310588235294118, + "grad_norm": 0.00013292491307659546, + "learning_rate": 7.525323046473485e-05, + "loss": 0.0, + "step": 9160 + }, + { + "epoch": 4.311058823529412, + "grad_norm": 0.045668116056230515, + "learning_rate": 7.521522778426315e-05, + "loss": 0.0036, + "step": 9161 + }, + { + "epoch": 4.311529411764706, + "grad_norm": 0.1763279282819271, + "learning_rate": 7.517723149077776e-05, + "loss": 0.0569, + "step": 9162 + }, + { + "epoch": 4.312, + "grad_norm": 0.08639999340058108, + "learning_rate": 7.513924158752383e-05, + "loss": 0.0206, + "step": 9163 + }, + { + "epoch": 4.3124705882352945, + "grad_norm": 0.05138584146186482, + "learning_rate": 7.510125807774589e-05, + "loss": 0.0093, + "step": 9164 + }, + { + "epoch": 4.312941176470588, + "grad_norm": 0.11536106138033671, + "learning_rate": 7.506328096468782e-05, + "loss": 0.0369, + "step": 9165 + }, + { + "epoch": 4.313411764705882, + "grad_norm": 0.07652912968279932, + "learning_rate": 7.502531025159313e-05, + "loss": 0.0099, + "step": 9166 + }, + { + "epoch": 4.313882352941176, + "grad_norm": 0.08227768787613936, + "learning_rate": 7.498734594170474e-05, + "loss": 0.0189, + "step": 9167 + }, + { + "epoch": 4.3143529411764705, + "grad_norm": 0.04332538596153698, + "learning_rate": 7.49493880382649e-05, + "loss": 0.0056, + "step": 9168 + }, + { + "epoch": 4.314823529411765, + "grad_norm": 0.09502947474000036, + "learning_rate": 7.491143654451544e-05, + "loss": 0.0255, + "step": 9169 + }, + { + "epoch": 4.315294117647059, + "grad_norm": 2.929966124034244e-05, + "learning_rate": 7.487349146369766e-05, + "loss": 0.0, + "step": 9170 + }, + { + "epoch": 4.315764705882353, + "grad_norm": 0.027759978371502188, + "learning_rate": 7.483555279905215e-05, + "loss": 0.0027, + "step": 9171 + }, + { + "epoch": 4.316235294117647, + "grad_norm": 1.4555318934977263e-05, + "learning_rate": 7.479762055381909e-05, + "loss": 0.0, + "step": 9172 + }, + { + "epoch": 4.316705882352942, + "grad_norm": 0.058474345386294785, + "learning_rate": 7.475969473123819e-05, + "loss": 0.0061, + "step": 9173 + }, + { + "epoch": 4.317176470588235, + "grad_norm": 3.5509282865666025e-05, + "learning_rate": 7.472177533454833e-05, + "loss": 0.0, + "step": 9174 + }, + { + "epoch": 4.317647058823529, + "grad_norm": 0.00021662123688312568, + "learning_rate": 7.46838623669881e-05, + "loss": 0.0, + "step": 9175 + }, + { + "epoch": 4.318117647058823, + "grad_norm": 0.11037214330856349, + "learning_rate": 7.464595583179544e-05, + "loss": 0.0272, + "step": 9176 + }, + { + "epoch": 4.318588235294118, + "grad_norm": 0.1177569877309998, + "learning_rate": 7.460805573220779e-05, + "loss": 0.0399, + "step": 9177 + }, + { + "epoch": 4.319058823529412, + "grad_norm": 0.23134133250345537, + "learning_rate": 7.4570162071462e-05, + "loss": 0.065, + "step": 9178 + }, + { + "epoch": 4.319529411764706, + "grad_norm": 0.26739382488440433, + "learning_rate": 7.453227485279432e-05, + "loss": 0.0327, + "step": 9179 + }, + { + "epoch": 4.32, + "grad_norm": 0.0386270872107832, + "learning_rate": 7.449439407944052e-05, + "loss": 0.0051, + "step": 9180 + }, + { + "epoch": 4.3204705882352945, + "grad_norm": 0.18328064559398496, + "learning_rate": 7.445651975463588e-05, + "loss": 0.0694, + "step": 9181 + }, + { + "epoch": 4.320941176470588, + "grad_norm": 0.0003204164563362761, + "learning_rate": 7.441865188161494e-05, + "loss": 0.0, + "step": 9182 + }, + { + "epoch": 4.321411764705882, + "grad_norm": 0.15201791551539742, + "learning_rate": 7.43807904636119e-05, + "loss": 0.045, + "step": 9183 + }, + { + "epoch": 4.321882352941176, + "grad_norm": 0.13908418301735287, + "learning_rate": 7.434293550386021e-05, + "loss": 0.0461, + "step": 9184 + }, + { + "epoch": 4.3223529411764705, + "grad_norm": 0.00014516178629817616, + "learning_rate": 7.430508700559292e-05, + "loss": 0.0, + "step": 9185 + }, + { + "epoch": 4.322823529411765, + "grad_norm": 0.17155447612755503, + "learning_rate": 7.426724497204254e-05, + "loss": 0.0524, + "step": 9186 + }, + { + "epoch": 4.323294117647059, + "grad_norm": 0.15464290358242824, + "learning_rate": 7.422940940644085e-05, + "loss": 0.0322, + "step": 9187 + }, + { + "epoch": 4.323764705882353, + "grad_norm": 0.10004918443138644, + "learning_rate": 7.419158031201924e-05, + "loss": 0.0141, + "step": 9188 + }, + { + "epoch": 4.324235294117647, + "grad_norm": 0.10155244298218843, + "learning_rate": 7.415375769200859e-05, + "loss": 0.0215, + "step": 9189 + }, + { + "epoch": 4.324705882352941, + "grad_norm": 0.10302580676667868, + "learning_rate": 7.411594154963897e-05, + "loss": 0.0346, + "step": 9190 + }, + { + "epoch": 4.325176470588235, + "grad_norm": 0.0013897079223064114, + "learning_rate": 7.407813188814017e-05, + "loss": 0.0, + "step": 9191 + }, + { + "epoch": 4.325647058823529, + "grad_norm": 0.15470772687682202, + "learning_rate": 7.404032871074135e-05, + "loss": 0.0346, + "step": 9192 + }, + { + "epoch": 4.326117647058823, + "grad_norm": 0.1452334962891178, + "learning_rate": 7.400253202067101e-05, + "loss": 0.0662, + "step": 9193 + }, + { + "epoch": 4.326588235294118, + "grad_norm": 4.8787093005527446e-05, + "learning_rate": 7.396474182115717e-05, + "loss": 0.0, + "step": 9194 + }, + { + "epoch": 4.327058823529412, + "grad_norm": 0.016381561900159373, + "learning_rate": 7.392695811542739e-05, + "loss": 0.0006, + "step": 9195 + }, + { + "epoch": 4.327529411764706, + "grad_norm": 0.08887118818133936, + "learning_rate": 7.388918090670847e-05, + "loss": 0.0224, + "step": 9196 + }, + { + "epoch": 4.328, + "grad_norm": 0.12744910454205938, + "learning_rate": 7.385141019822683e-05, + "loss": 0.0239, + "step": 9197 + }, + { + "epoch": 4.3284705882352945, + "grad_norm": 0.0002670195820849015, + "learning_rate": 7.381364599320831e-05, + "loss": 0.0, + "step": 9198 + }, + { + "epoch": 4.328941176470588, + "grad_norm": 0.0015098939017732, + "learning_rate": 7.377588829487806e-05, + "loss": 0.0, + "step": 9199 + }, + { + "epoch": 4.329411764705882, + "grad_norm": 0.13462790043519712, + "learning_rate": 7.373813710646083e-05, + "loss": 0.0461, + "step": 9200 + }, + { + "epoch": 4.329882352941176, + "grad_norm": 0.05569217337460067, + "learning_rate": 7.370039243118082e-05, + "loss": 0.0076, + "step": 9201 + }, + { + "epoch": 4.3303529411764705, + "grad_norm": 0.08278477624957532, + "learning_rate": 7.366265427226148e-05, + "loss": 0.0124, + "step": 9202 + }, + { + "epoch": 4.330823529411765, + "grad_norm": 0.08491420417017959, + "learning_rate": 7.36249226329259e-05, + "loss": 0.024, + "step": 9203 + }, + { + "epoch": 4.331294117647059, + "grad_norm": 0.12391255109697223, + "learning_rate": 7.358719751639659e-05, + "loss": 0.0403, + "step": 9204 + }, + { + "epoch": 4.331764705882353, + "grad_norm": 0.0072780791480799405, + "learning_rate": 7.354947892589536e-05, + "loss": 0.0001, + "step": 9205 + }, + { + "epoch": 4.332235294117647, + "grad_norm": 0.054078599149942196, + "learning_rate": 7.351176686464368e-05, + "loss": 0.0082, + "step": 9206 + }, + { + "epoch": 4.332705882352942, + "grad_norm": 4.696958509497212e-05, + "learning_rate": 7.347406133586223e-05, + "loss": 0.0, + "step": 9207 + }, + { + "epoch": 4.333176470588235, + "grad_norm": 0.06608410898726634, + "learning_rate": 7.34363623427713e-05, + "loss": 0.0119, + "step": 9208 + }, + { + "epoch": 4.333647058823529, + "grad_norm": 0.008104213175312014, + "learning_rate": 7.339866988859061e-05, + "loss": 0.0003, + "step": 9209 + }, + { + "epoch": 4.334117647058823, + "grad_norm": 0.07675766046483391, + "learning_rate": 7.33609839765392e-05, + "loss": 0.0149, + "step": 9210 + }, + { + "epoch": 4.334588235294118, + "grad_norm": 0.04868979556741057, + "learning_rate": 7.332330460983565e-05, + "loss": 0.0086, + "step": 9211 + }, + { + "epoch": 4.335058823529412, + "grad_norm": 0.11813935608097877, + "learning_rate": 7.328563179169806e-05, + "loss": 0.0387, + "step": 9212 + }, + { + "epoch": 4.335529411764706, + "grad_norm": 0.1256003965849359, + "learning_rate": 7.324796552534373e-05, + "loss": 0.0472, + "step": 9213 + }, + { + "epoch": 4.336, + "grad_norm": 0.10038241888983976, + "learning_rate": 7.321030581398961e-05, + "loss": 0.0236, + "step": 9214 + }, + { + "epoch": 4.3364705882352945, + "grad_norm": 0.056748798689457806, + "learning_rate": 7.317265266085209e-05, + "loss": 0.005, + "step": 9215 + }, + { + "epoch": 4.336941176470588, + "grad_norm": 0.027123989644303074, + "learning_rate": 7.31350060691468e-05, + "loss": 0.0026, + "step": 9216 + }, + { + "epoch": 4.337411764705882, + "grad_norm": 0.21276982418622628, + "learning_rate": 7.309736604208902e-05, + "loss": 0.0656, + "step": 9217 + }, + { + "epoch": 4.337882352941176, + "grad_norm": 0.0957039910220642, + "learning_rate": 7.305973258289343e-05, + "loss": 0.0269, + "step": 9218 + }, + { + "epoch": 4.3383529411764705, + "grad_norm": 0.09220889054034233, + "learning_rate": 7.302210569477397e-05, + "loss": 0.0143, + "step": 9219 + }, + { + "epoch": 4.338823529411765, + "grad_norm": 0.21768903203859913, + "learning_rate": 7.298448538094436e-05, + "loss": 0.092, + "step": 9220 + }, + { + "epoch": 4.339294117647059, + "grad_norm": 0.21963657525439734, + "learning_rate": 7.294687164461741e-05, + "loss": 0.0409, + "step": 9221 + }, + { + "epoch": 4.339764705882353, + "grad_norm": 0.11304900075478412, + "learning_rate": 7.290926448900557e-05, + "loss": 0.0319, + "step": 9222 + }, + { + "epoch": 4.340235294117647, + "grad_norm": 0.1312969542100761, + "learning_rate": 7.287166391732071e-05, + "loss": 0.0377, + "step": 9223 + }, + { + "epoch": 4.340705882352941, + "grad_norm": 2.527503807373301e-05, + "learning_rate": 7.283406993277401e-05, + "loss": 0.0, + "step": 9224 + }, + { + "epoch": 4.341176470588235, + "grad_norm": 0.12844077100393458, + "learning_rate": 7.279648253857626e-05, + "loss": 0.0455, + "step": 9225 + }, + { + "epoch": 4.341647058823529, + "grad_norm": 0.03839418868749131, + "learning_rate": 7.275890173793761e-05, + "loss": 0.0056, + "step": 9226 + }, + { + "epoch": 4.342117647058823, + "grad_norm": 0.10019803618345739, + "learning_rate": 7.272132753406757e-05, + "loss": 0.0366, + "step": 9227 + }, + { + "epoch": 4.342588235294118, + "grad_norm": 5.6859833381563295e-05, + "learning_rate": 7.26837599301752e-05, + "loss": 0.0, + "step": 9228 + }, + { + "epoch": 4.343058823529412, + "grad_norm": 0.11804002952819495, + "learning_rate": 7.264619892946904e-05, + "loss": 0.0284, + "step": 9229 + }, + { + "epoch": 4.343529411764706, + "grad_norm": 0.1694021157827478, + "learning_rate": 7.260864453515683e-05, + "loss": 0.0665, + "step": 9230 + }, + { + "epoch": 4.344, + "grad_norm": 1.147259320808829e-05, + "learning_rate": 7.2571096750446e-05, + "loss": 0.0, + "step": 9231 + }, + { + "epoch": 4.3444705882352945, + "grad_norm": 0.09667668697493298, + "learning_rate": 7.253355557854334e-05, + "loss": 0.0095, + "step": 9232 + }, + { + "epoch": 4.344941176470588, + "grad_norm": 0.10435032371221017, + "learning_rate": 7.249602102265496e-05, + "loss": 0.0176, + "step": 9233 + }, + { + "epoch": 4.345411764705882, + "grad_norm": 0.14170310104262668, + "learning_rate": 7.245849308598658e-05, + "loss": 0.0502, + "step": 9234 + }, + { + "epoch": 4.345882352941176, + "grad_norm": 0.08830405912324717, + "learning_rate": 7.242097177174318e-05, + "loss": 0.0237, + "step": 9235 + }, + { + "epoch": 4.3463529411764705, + "grad_norm": 0.02527995865919958, + "learning_rate": 7.238345708312931e-05, + "loss": 0.0027, + "step": 9236 + }, + { + "epoch": 4.346823529411765, + "grad_norm": 0.019069217682334653, + "learning_rate": 7.234594902334897e-05, + "loss": 0.0016, + "step": 9237 + }, + { + "epoch": 4.347294117647059, + "grad_norm": 0.0391357187737878, + "learning_rate": 7.230844759560544e-05, + "loss": 0.0051, + "step": 9238 + }, + { + "epoch": 4.347764705882353, + "grad_norm": 0.06207380432876529, + "learning_rate": 7.227095280310152e-05, + "loss": 0.0136, + "step": 9239 + }, + { + "epoch": 4.348235294117647, + "grad_norm": 0.13890476085205536, + "learning_rate": 7.223346464903958e-05, + "loss": 0.0449, + "step": 9240 + }, + { + "epoch": 4.348705882352942, + "grad_norm": 0.17237222220644605, + "learning_rate": 7.219598313662113e-05, + "loss": 0.0799, + "step": 9241 + }, + { + "epoch": 4.349176470588235, + "grad_norm": 0.052411432152976514, + "learning_rate": 7.215850826904734e-05, + "loss": 0.0047, + "step": 9242 + }, + { + "epoch": 4.349647058823529, + "grad_norm": 0.03541487272875028, + "learning_rate": 7.212104004951883e-05, + "loss": 0.005, + "step": 9243 + }, + { + "epoch": 4.350117647058823, + "grad_norm": 0.06923325412808501, + "learning_rate": 7.208357848123543e-05, + "loss": 0.0139, + "step": 9244 + }, + { + "epoch": 4.350588235294118, + "grad_norm": 0.0714399477516654, + "learning_rate": 7.204612356739658e-05, + "loss": 0.0152, + "step": 9245 + }, + { + "epoch": 4.351058823529412, + "grad_norm": 0.14359909221538528, + "learning_rate": 7.200867531120124e-05, + "loss": 0.0346, + "step": 9246 + }, + { + "epoch": 4.351529411764706, + "grad_norm": 0.033707823835732766, + "learning_rate": 7.197123371584749e-05, + "loss": 0.0021, + "step": 9247 + }, + { + "epoch": 4.352, + "grad_norm": 0.13739580857625183, + "learning_rate": 7.193379878453312e-05, + "loss": 0.0218, + "step": 9248 + }, + { + "epoch": 4.352470588235294, + "grad_norm": 0.06789539254646057, + "learning_rate": 7.189637052045529e-05, + "loss": 0.0109, + "step": 9249 + }, + { + "epoch": 4.352941176470588, + "grad_norm": 0.03860637115015528, + "learning_rate": 7.185894892681048e-05, + "loss": 0.0023, + "step": 9250 + }, + { + "epoch": 4.353411764705882, + "grad_norm": 0.06139524131039863, + "learning_rate": 7.182153400679468e-05, + "loss": 0.0071, + "step": 9251 + }, + { + "epoch": 4.353882352941176, + "grad_norm": 0.02415777367505969, + "learning_rate": 7.17841257636034e-05, + "loss": 0.0008, + "step": 9252 + }, + { + "epoch": 4.3543529411764705, + "grad_norm": 0.11713483326040054, + "learning_rate": 7.17467242004314e-05, + "loss": 0.0379, + "step": 9253 + }, + { + "epoch": 4.354823529411765, + "grad_norm": 0.05120106719383209, + "learning_rate": 7.170932932047296e-05, + "loss": 0.0019, + "step": 9254 + }, + { + "epoch": 4.355294117647059, + "grad_norm": 1.499588423478975e-05, + "learning_rate": 7.167194112692186e-05, + "loss": 0.0, + "step": 9255 + }, + { + "epoch": 4.355764705882353, + "grad_norm": 0.1734511903042767, + "learning_rate": 7.163455962297116e-05, + "loss": 0.0652, + "step": 9256 + }, + { + "epoch": 4.356235294117647, + "grad_norm": 0.018412980327131834, + "learning_rate": 7.159718481181346e-05, + "loss": 0.0009, + "step": 9257 + }, + { + "epoch": 4.356705882352941, + "grad_norm": 0.00030902725205490827, + "learning_rate": 7.155981669664072e-05, + "loss": 0.0, + "step": 9258 + }, + { + "epoch": 4.357176470588235, + "grad_norm": 5.037313613830219e-06, + "learning_rate": 7.152245528064437e-05, + "loss": 0.0, + "step": 9259 + }, + { + "epoch": 4.357647058823529, + "grad_norm": 0.10411777628792208, + "learning_rate": 7.148510056701532e-05, + "loss": 0.029, + "step": 9260 + }, + { + "epoch": 4.358117647058823, + "grad_norm": 0.02886121016626531, + "learning_rate": 7.144775255894375e-05, + "loss": 0.0023, + "step": 9261 + }, + { + "epoch": 4.358588235294118, + "grad_norm": 0.07843624645852543, + "learning_rate": 7.14104112596194e-05, + "loss": 0.0158, + "step": 9262 + }, + { + "epoch": 4.359058823529412, + "grad_norm": 0.11172755134424722, + "learning_rate": 7.137307667223141e-05, + "loss": 0.032, + "step": 9263 + }, + { + "epoch": 4.359529411764706, + "grad_norm": 0.08699238991596527, + "learning_rate": 7.133574879996833e-05, + "loss": 0.0228, + "step": 9264 + }, + { + "epoch": 4.36, + "grad_norm": 0.02785306211727677, + "learning_rate": 7.129842764601821e-05, + "loss": 0.0027, + "step": 9265 + }, + { + "epoch": 4.3604705882352945, + "grad_norm": 0.1338863303517063, + "learning_rate": 7.126111321356834e-05, + "loss": 0.0512, + "step": 9266 + }, + { + "epoch": 4.360941176470588, + "grad_norm": 0.2210979038539248, + "learning_rate": 7.122380550580562e-05, + "loss": 0.0612, + "step": 9267 + }, + { + "epoch": 4.361411764705882, + "grad_norm": 0.08964649220851034, + "learning_rate": 7.118650452591634e-05, + "loss": 0.017, + "step": 9268 + }, + { + "epoch": 4.361882352941176, + "grad_norm": 0.0948209526318025, + "learning_rate": 7.114921027708613e-05, + "loss": 0.0235, + "step": 9269 + }, + { + "epoch": 4.3623529411764705, + "grad_norm": 0.039241660178174785, + "learning_rate": 7.111192276250011e-05, + "loss": 0.0041, + "step": 9270 + }, + { + "epoch": 4.362823529411765, + "grad_norm": 1.4304635906128643e-05, + "learning_rate": 7.107464198534287e-05, + "loss": 0.0, + "step": 9271 + }, + { + "epoch": 4.363294117647059, + "grad_norm": 0.11229736264397966, + "learning_rate": 7.103736794879829e-05, + "loss": 0.0126, + "step": 9272 + }, + { + "epoch": 4.363764705882353, + "grad_norm": 0.22900110622542877, + "learning_rate": 7.100010065604978e-05, + "loss": 0.05, + "step": 9273 + }, + { + "epoch": 4.364235294117647, + "grad_norm": 0.1351443049974075, + "learning_rate": 7.096284011028024e-05, + "loss": 0.0537, + "step": 9274 + }, + { + "epoch": 4.364705882352941, + "grad_norm": 0.1937138951871671, + "learning_rate": 7.092558631467177e-05, + "loss": 0.0855, + "step": 9275 + }, + { + "epoch": 4.365176470588235, + "grad_norm": 0.19775105887995004, + "learning_rate": 7.088833927240608e-05, + "loss": 0.0458, + "step": 9276 + }, + { + "epoch": 4.365647058823529, + "grad_norm": 0.053066163482678626, + "learning_rate": 7.085109898666433e-05, + "loss": 0.008, + "step": 9277 + }, + { + "epoch": 4.366117647058823, + "grad_norm": 0.10210439313777786, + "learning_rate": 7.081386546062687e-05, + "loss": 0.0358, + "step": 9278 + }, + { + "epoch": 4.366588235294118, + "grad_norm": 0.05933298369163001, + "learning_rate": 7.077663869747372e-05, + "loss": 0.0073, + "step": 9279 + }, + { + "epoch": 4.367058823529412, + "grad_norm": 0.051608068388554515, + "learning_rate": 7.073941870038427e-05, + "loss": 0.0112, + "step": 9280 + }, + { + "epoch": 4.367529411764706, + "grad_norm": 0.17565202328535967, + "learning_rate": 7.070220547253717e-05, + "loss": 0.0808, + "step": 9281 + }, + { + "epoch": 4.368, + "grad_norm": 0.005237849963832775, + "learning_rate": 7.066499901711068e-05, + "loss": 0.0001, + "step": 9282 + }, + { + "epoch": 4.368470588235294, + "grad_norm": 0.10628943090529182, + "learning_rate": 7.062779933728247e-05, + "loss": 0.0122, + "step": 9283 + }, + { + "epoch": 4.368941176470588, + "grad_norm": 0.1459164318204064, + "learning_rate": 7.059060643622944e-05, + "loss": 0.0659, + "step": 9284 + }, + { + "epoch": 4.369411764705882, + "grad_norm": 0.1000452854286699, + "learning_rate": 7.05534203171282e-05, + "loss": 0.0223, + "step": 9285 + }, + { + "epoch": 4.369882352941176, + "grad_norm": 0.07632777340160571, + "learning_rate": 7.051624098315449e-05, + "loss": 0.0165, + "step": 9286 + }, + { + "epoch": 4.3703529411764706, + "grad_norm": 0.14638545967999236, + "learning_rate": 7.047906843748365e-05, + "loss": 0.0699, + "step": 9287 + }, + { + "epoch": 4.370823529411765, + "grad_norm": 0.11057044337216675, + "learning_rate": 7.044190268329049e-05, + "loss": 0.0334, + "step": 9288 + }, + { + "epoch": 4.371294117647059, + "grad_norm": 0.07777560101418317, + "learning_rate": 7.0404743723749e-05, + "loss": 0.0159, + "step": 9289 + }, + { + "epoch": 4.371764705882353, + "grad_norm": 0.10281906513197624, + "learning_rate": 7.036759156203281e-05, + "loss": 0.0231, + "step": 9290 + }, + { + "epoch": 4.372235294117647, + "grad_norm": 0.10628319765691348, + "learning_rate": 7.033044620131495e-05, + "loss": 0.032, + "step": 9291 + }, + { + "epoch": 4.372705882352941, + "grad_norm": 0.14161499977404857, + "learning_rate": 7.02933076447677e-05, + "loss": 0.0244, + "step": 9292 + }, + { + "epoch": 4.373176470588235, + "grad_norm": 0.05866132185165888, + "learning_rate": 7.025617589556297e-05, + "loss": 0.0087, + "step": 9293 + }, + { + "epoch": 4.373647058823529, + "grad_norm": 0.11476513900177684, + "learning_rate": 7.021905095687198e-05, + "loss": 0.0156, + "step": 9294 + }, + { + "epoch": 4.374117647058823, + "grad_norm": 0.049912114433077956, + "learning_rate": 7.018193283186531e-05, + "loss": 0.0034, + "step": 9295 + }, + { + "epoch": 4.374588235294118, + "grad_norm": 0.10897422399615604, + "learning_rate": 7.01448215237131e-05, + "loss": 0.0413, + "step": 9296 + }, + { + "epoch": 4.375058823529412, + "grad_norm": 0.11903161253500534, + "learning_rate": 7.010771703558488e-05, + "loss": 0.0171, + "step": 9297 + }, + { + "epoch": 4.375529411764706, + "grad_norm": 0.07858796750452834, + "learning_rate": 7.007061937064941e-05, + "loss": 0.0054, + "step": 9298 + }, + { + "epoch": 4.376, + "grad_norm": 0.061086110239596945, + "learning_rate": 7.00335285320751e-05, + "loss": 0.0045, + "step": 9299 + }, + { + "epoch": 4.376470588235295, + "grad_norm": 0.08088666219005224, + "learning_rate": 6.999644452302975e-05, + "loss": 0.0143, + "step": 9300 + }, + { + "epoch": 4.376941176470588, + "grad_norm": 0.09054533329656855, + "learning_rate": 6.995936734668042e-05, + "loss": 0.0254, + "step": 9301 + }, + { + "epoch": 4.377411764705882, + "grad_norm": 0.05297179228134606, + "learning_rate": 6.992229700619369e-05, + "loss": 0.0073, + "step": 9302 + }, + { + "epoch": 4.377882352941176, + "grad_norm": 0.08915358675912853, + "learning_rate": 6.988523350473563e-05, + "loss": 0.0252, + "step": 9303 + }, + { + "epoch": 4.378352941176471, + "grad_norm": 0.10256251861773977, + "learning_rate": 6.984817684547155e-05, + "loss": 0.0302, + "step": 9304 + }, + { + "epoch": 4.378823529411765, + "grad_norm": 6.987951352762358e-05, + "learning_rate": 6.981112703156628e-05, + "loss": 0.0, + "step": 9305 + }, + { + "epoch": 4.379294117647059, + "grad_norm": 0.15606835631265753, + "learning_rate": 6.977408406618418e-05, + "loss": 0.048, + "step": 9306 + }, + { + "epoch": 4.379764705882353, + "grad_norm": 0.13830374469288034, + "learning_rate": 6.973704795248871e-05, + "loss": 0.0124, + "step": 9307 + }, + { + "epoch": 4.3802352941176474, + "grad_norm": 0.1063071729944994, + "learning_rate": 6.970001869364306e-05, + "loss": 0.0179, + "step": 9308 + }, + { + "epoch": 4.380705882352941, + "grad_norm": 0.06583296855088869, + "learning_rate": 6.966299629280968e-05, + "loss": 0.0127, + "step": 9309 + }, + { + "epoch": 4.381176470588235, + "grad_norm": 0.09332405974581504, + "learning_rate": 6.962598075315046e-05, + "loss": 0.0238, + "step": 9310 + }, + { + "epoch": 4.381647058823529, + "grad_norm": 0.10882947130961662, + "learning_rate": 6.958897207782677e-05, + "loss": 0.0217, + "step": 9311 + }, + { + "epoch": 4.3821176470588235, + "grad_norm": 0.2096976158358767, + "learning_rate": 6.955197026999925e-05, + "loss": 0.0487, + "step": 9312 + }, + { + "epoch": 4.382588235294118, + "grad_norm": 0.15804087142430356, + "learning_rate": 6.951497533282809e-05, + "loss": 0.0451, + "step": 9313 + }, + { + "epoch": 4.383058823529412, + "grad_norm": 0.26835482832004287, + "learning_rate": 6.947798726947277e-05, + "loss": 0.0698, + "step": 9314 + }, + { + "epoch": 4.383529411764706, + "grad_norm": 0.11920303488038377, + "learning_rate": 6.94410060830923e-05, + "loss": 0.0455, + "step": 9315 + }, + { + "epoch": 4.384, + "grad_norm": 0.08156546705779422, + "learning_rate": 6.940403177684509e-05, + "loss": 0.023, + "step": 9316 + }, + { + "epoch": 4.384470588235294, + "grad_norm": 0.09111602452721976, + "learning_rate": 6.936706435388882e-05, + "loss": 0.02, + "step": 9317 + }, + { + "epoch": 4.384941176470588, + "grad_norm": 9.823107311904973e-05, + "learning_rate": 6.93301038173808e-05, + "loss": 0.0, + "step": 9318 + }, + { + "epoch": 4.385411764705882, + "grad_norm": 0.09594486380088801, + "learning_rate": 6.929315017047762e-05, + "loss": 0.02, + "step": 9319 + }, + { + "epoch": 4.385882352941176, + "grad_norm": 0.0026614121422235855, + "learning_rate": 6.925620341633522e-05, + "loss": 0.0, + "step": 9320 + }, + { + "epoch": 4.386352941176471, + "grad_norm": 0.16581770786996372, + "learning_rate": 6.921926355810911e-05, + "loss": 0.0402, + "step": 9321 + }, + { + "epoch": 4.386823529411765, + "grad_norm": 0.09514895943263194, + "learning_rate": 6.918233059895418e-05, + "loss": 0.0247, + "step": 9322 + }, + { + "epoch": 4.387294117647059, + "grad_norm": 0.06024877639506182, + "learning_rate": 6.914540454202457e-05, + "loss": 0.0057, + "step": 9323 + }, + { + "epoch": 4.387764705882353, + "grad_norm": 0.0005236312837737221, + "learning_rate": 6.910848539047399e-05, + "loss": 0.0, + "step": 9324 + }, + { + "epoch": 4.3882352941176475, + "grad_norm": 0.057379004314715334, + "learning_rate": 6.907157314745558e-05, + "loss": 0.0049, + "step": 9325 + }, + { + "epoch": 4.388705882352941, + "grad_norm": 0.03400967279243298, + "learning_rate": 6.903466781612173e-05, + "loss": 0.003, + "step": 9326 + }, + { + "epoch": 4.389176470588235, + "grad_norm": 0.09008335978358747, + "learning_rate": 6.899776939962438e-05, + "loss": 0.0202, + "step": 9327 + }, + { + "epoch": 4.389647058823529, + "grad_norm": 0.11656730461783499, + "learning_rate": 6.89608779011149e-05, + "loss": 0.0309, + "step": 9328 + }, + { + "epoch": 4.3901176470588235, + "grad_norm": 0.10745855396202995, + "learning_rate": 6.892399332374387e-05, + "loss": 0.0194, + "step": 9329 + }, + { + "epoch": 4.390588235294118, + "grad_norm": 8.348385109454392e-05, + "learning_rate": 6.888711567066149e-05, + "loss": 0.0, + "step": 9330 + }, + { + "epoch": 4.391058823529412, + "grad_norm": 0.12598611241327926, + "learning_rate": 6.885024494501735e-05, + "loss": 0.0289, + "step": 9331 + }, + { + "epoch": 4.391529411764706, + "grad_norm": 0.03888033703612695, + "learning_rate": 6.881338114996027e-05, + "loss": 0.0033, + "step": 9332 + }, + { + "epoch": 4.392, + "grad_norm": 0.04235872134726044, + "learning_rate": 6.877652428863866e-05, + "loss": 0.0044, + "step": 9333 + }, + { + "epoch": 4.392470588235295, + "grad_norm": 0.11419687472700904, + "learning_rate": 6.873967436420033e-05, + "loss": 0.0223, + "step": 9334 + }, + { + "epoch": 4.392941176470588, + "grad_norm": 0.09836105721650962, + "learning_rate": 6.870283137979233e-05, + "loss": 0.0261, + "step": 9335 + }, + { + "epoch": 4.393411764705882, + "grad_norm": 0.10020269820786663, + "learning_rate": 6.866599533856135e-05, + "loss": 0.0307, + "step": 9336 + }, + { + "epoch": 4.393882352941176, + "grad_norm": 0.10474897251032313, + "learning_rate": 6.862916624365325e-05, + "loss": 0.0186, + "step": 9337 + }, + { + "epoch": 4.394352941176471, + "grad_norm": 0.09399572275577478, + "learning_rate": 6.859234409821347e-05, + "loss": 0.0209, + "step": 9338 + }, + { + "epoch": 4.394823529411765, + "grad_norm": 6.137801090227845e-05, + "learning_rate": 6.855552890538688e-05, + "loss": 0.0, + "step": 9339 + }, + { + "epoch": 4.395294117647059, + "grad_norm": 0.11389063089992947, + "learning_rate": 6.851872066831751e-05, + "loss": 0.0312, + "step": 9340 + }, + { + "epoch": 4.395764705882353, + "grad_norm": 0.16603323426975763, + "learning_rate": 6.848191939014909e-05, + "loss": 0.0737, + "step": 9341 + }, + { + "epoch": 4.3962352941176475, + "grad_norm": 0.1505473094423678, + "learning_rate": 6.844512507402465e-05, + "loss": 0.065, + "step": 9342 + }, + { + "epoch": 4.396705882352941, + "grad_norm": 0.09585168501448754, + "learning_rate": 6.84083377230865e-05, + "loss": 0.0102, + "step": 9343 + }, + { + "epoch": 4.397176470588235, + "grad_norm": 0.05787660993307096, + "learning_rate": 6.837155734047652e-05, + "loss": 0.0093, + "step": 9344 + }, + { + "epoch": 4.397647058823529, + "grad_norm": 0.05829212090474988, + "learning_rate": 6.833478392933597e-05, + "loss": 0.0071, + "step": 9345 + }, + { + "epoch": 4.3981176470588235, + "grad_norm": 0.06349843330908993, + "learning_rate": 6.829801749280537e-05, + "loss": 0.0069, + "step": 9346 + }, + { + "epoch": 4.398588235294118, + "grad_norm": 0.1324271117215764, + "learning_rate": 6.826125803402484e-05, + "loss": 0.0455, + "step": 9347 + }, + { + "epoch": 4.399058823529412, + "grad_norm": 0.07605589323463358, + "learning_rate": 6.822450555613385e-05, + "loss": 0.0183, + "step": 9348 + }, + { + "epoch": 4.399529411764706, + "grad_norm": 0.09698387461620041, + "learning_rate": 6.818776006227114e-05, + "loss": 0.017, + "step": 9349 + }, + { + "epoch": 4.4, + "grad_norm": 0.04587777264191798, + "learning_rate": 6.815102155557501e-05, + "loss": 0.0041, + "step": 9350 + }, + { + "epoch": 4.400470588235294, + "grad_norm": 0.09945126957082152, + "learning_rate": 6.811429003918317e-05, + "loss": 0.0245, + "step": 9351 + }, + { + "epoch": 4.400941176470588, + "grad_norm": 0.08462527404951496, + "learning_rate": 6.807756551623248e-05, + "loss": 0.0128, + "step": 9352 + }, + { + "epoch": 4.401411764705882, + "grad_norm": 0.07078683423558049, + "learning_rate": 6.804084798985964e-05, + "loss": 0.0124, + "step": 9353 + }, + { + "epoch": 4.401882352941176, + "grad_norm": 1.3878120998338858e-05, + "learning_rate": 6.800413746320034e-05, + "loss": 0.0, + "step": 9354 + }, + { + "epoch": 4.402352941176471, + "grad_norm": 0.05722233946887292, + "learning_rate": 6.796743393938987e-05, + "loss": 0.01, + "step": 9355 + }, + { + "epoch": 4.402823529411765, + "grad_norm": 1.2351026889668281e-05, + "learning_rate": 6.793073742156298e-05, + "loss": 0.0, + "step": 9356 + }, + { + "epoch": 4.403294117647059, + "grad_norm": 0.11137465651671027, + "learning_rate": 6.78940479128536e-05, + "loss": 0.0176, + "step": 9357 + }, + { + "epoch": 4.403764705882353, + "grad_norm": 0.11012415646211947, + "learning_rate": 6.785736541639526e-05, + "loss": 0.0157, + "step": 9358 + }, + { + "epoch": 4.4042352941176475, + "grad_norm": 0.0772544746021971, + "learning_rate": 6.782068993532085e-05, + "loss": 0.0044, + "step": 9359 + }, + { + "epoch": 4.404705882352941, + "grad_norm": 0.09940969867965566, + "learning_rate": 6.778402147276259e-05, + "loss": 0.0212, + "step": 9360 + }, + { + "epoch": 4.405176470588235, + "grad_norm": 0.07575242053454205, + "learning_rate": 6.774736003185215e-05, + "loss": 0.0194, + "step": 9361 + }, + { + "epoch": 4.405647058823529, + "grad_norm": 0.07447225832458151, + "learning_rate": 6.771070561572067e-05, + "loss": 0.0155, + "step": 9362 + }, + { + "epoch": 4.4061176470588235, + "grad_norm": 2.9616790623998767e-05, + "learning_rate": 6.767405822749848e-05, + "loss": 0.0, + "step": 9363 + }, + { + "epoch": 4.406588235294118, + "grad_norm": 0.020928893331556335, + "learning_rate": 6.76374178703156e-05, + "loss": 0.0018, + "step": 9364 + }, + { + "epoch": 4.407058823529412, + "grad_norm": 0.06859320842247514, + "learning_rate": 6.760078454730115e-05, + "loss": 0.0083, + "step": 9365 + }, + { + "epoch": 4.407529411764706, + "grad_norm": 0.12780295607223305, + "learning_rate": 6.756415826158388e-05, + "loss": 0.04, + "step": 9366 + }, + { + "epoch": 4.408, + "grad_norm": 1.7266394370280014e-05, + "learning_rate": 6.752753901629189e-05, + "loss": 0.0, + "step": 9367 + }, + { + "epoch": 4.408470588235295, + "grad_norm": 0.11841717361319001, + "learning_rate": 6.749092681455255e-05, + "loss": 0.0283, + "step": 9368 + }, + { + "epoch": 4.408941176470588, + "grad_norm": 1.6858761371466975e-05, + "learning_rate": 6.745432165949274e-05, + "loss": 0.0, + "step": 9369 + }, + { + "epoch": 4.409411764705882, + "grad_norm": 0.026702389725653437, + "learning_rate": 6.741772355423882e-05, + "loss": 0.0013, + "step": 9370 + }, + { + "epoch": 4.409882352941176, + "grad_norm": 0.08970344646041234, + "learning_rate": 6.73811325019163e-05, + "loss": 0.0257, + "step": 9371 + }, + { + "epoch": 4.410352941176471, + "grad_norm": 0.11118691166039099, + "learning_rate": 6.734454850565031e-05, + "loss": 0.0206, + "step": 9372 + }, + { + "epoch": 4.410823529411765, + "grad_norm": 0.04032819144424041, + "learning_rate": 6.730797156856537e-05, + "loss": 0.0067, + "step": 9373 + }, + { + "epoch": 4.411294117647059, + "grad_norm": 0.08725713953991882, + "learning_rate": 6.727140169378521e-05, + "loss": 0.0201, + "step": 9374 + }, + { + "epoch": 4.411764705882353, + "grad_norm": 6.423876585226202e-06, + "learning_rate": 6.72348388844331e-05, + "loss": 0.0, + "step": 9375 + }, + { + "epoch": 4.412235294117647, + "grad_norm": 0.11654098353402526, + "learning_rate": 6.719828314363178e-05, + "loss": 0.0373, + "step": 9376 + }, + { + "epoch": 4.412705882352941, + "grad_norm": 0.07201001550832563, + "learning_rate": 6.716173447450317e-05, + "loss": 0.015, + "step": 9377 + }, + { + "epoch": 4.413176470588235, + "grad_norm": 0.07659864780183662, + "learning_rate": 6.712519288016875e-05, + "loss": 0.0183, + "step": 9378 + }, + { + "epoch": 4.413647058823529, + "grad_norm": 0.016578767616810042, + "learning_rate": 6.70886583637494e-05, + "loss": 0.0005, + "step": 9379 + }, + { + "epoch": 4.4141176470588235, + "grad_norm": 0.1091048051139824, + "learning_rate": 6.705213092836527e-05, + "loss": 0.0124, + "step": 9380 + }, + { + "epoch": 4.414588235294118, + "grad_norm": 0.06625828782845371, + "learning_rate": 6.701561057713599e-05, + "loss": 0.0113, + "step": 9381 + }, + { + "epoch": 4.415058823529412, + "grad_norm": 0.09497477660255774, + "learning_rate": 6.697909731318069e-05, + "loss": 0.0191, + "step": 9382 + }, + { + "epoch": 4.415529411764706, + "grad_norm": 0.09276174368905472, + "learning_rate": 6.694259113961761e-05, + "loss": 0.0112, + "step": 9383 + }, + { + "epoch": 4.416, + "grad_norm": 0.07993135554052823, + "learning_rate": 6.690609205956464e-05, + "loss": 0.013, + "step": 9384 + }, + { + "epoch": 4.416470588235294, + "grad_norm": 0.10174990992396613, + "learning_rate": 6.686960007613906e-05, + "loss": 0.0312, + "step": 9385 + }, + { + "epoch": 4.416941176470588, + "grad_norm": 0.1349966611869578, + "learning_rate": 6.683311519245733e-05, + "loss": 0.0351, + "step": 9386 + }, + { + "epoch": 4.417411764705882, + "grad_norm": 0.11073462329095941, + "learning_rate": 6.679663741163553e-05, + "loss": 0.0262, + "step": 9387 + }, + { + "epoch": 4.417882352941176, + "grad_norm": 0.13411012654162296, + "learning_rate": 6.676016673678896e-05, + "loss": 0.0373, + "step": 9388 + }, + { + "epoch": 4.418352941176471, + "grad_norm": 0.13236794916757325, + "learning_rate": 6.672370317103245e-05, + "loss": 0.0376, + "step": 9389 + }, + { + "epoch": 4.418823529411765, + "grad_norm": 0.04445756861751524, + "learning_rate": 6.668724671748018e-05, + "loss": 0.0032, + "step": 9390 + }, + { + "epoch": 4.419294117647059, + "grad_norm": 0.15216830679172777, + "learning_rate": 6.665079737924566e-05, + "loss": 0.0476, + "step": 9391 + }, + { + "epoch": 4.419764705882353, + "grad_norm": 2.749838973034549e-06, + "learning_rate": 6.661435515944186e-05, + "loss": 0.0, + "step": 9392 + }, + { + "epoch": 4.4202352941176475, + "grad_norm": 0.1016508044266716, + "learning_rate": 6.65779200611812e-05, + "loss": 0.0316, + "step": 9393 + }, + { + "epoch": 4.420705882352941, + "grad_norm": 0.08828391687362795, + "learning_rate": 6.654149208757529e-05, + "loss": 0.0229, + "step": 9394 + }, + { + "epoch": 4.421176470588235, + "grad_norm": 4.708745979059548e-06, + "learning_rate": 6.65050712417353e-05, + "loss": 0.0, + "step": 9395 + }, + { + "epoch": 4.421647058823529, + "grad_norm": 0.10497484220197703, + "learning_rate": 6.646865752677186e-05, + "loss": 0.0278, + "step": 9396 + }, + { + "epoch": 4.4221176470588235, + "grad_norm": 0.05728838698931865, + "learning_rate": 6.643225094579462e-05, + "loss": 0.0042, + "step": 9397 + }, + { + "epoch": 4.422588235294118, + "grad_norm": 0.1525032062619632, + "learning_rate": 6.63958515019132e-05, + "loss": 0.034, + "step": 9398 + }, + { + "epoch": 4.423058823529412, + "grad_norm": 0.12048096365604219, + "learning_rate": 6.635945919823607e-05, + "loss": 0.0174, + "step": 9399 + }, + { + "epoch": 4.423529411764706, + "grad_norm": 0.056282078351299794, + "learning_rate": 6.632307403787138e-05, + "loss": 0.0074, + "step": 9400 + }, + { + "epoch": 4.424, + "grad_norm": 0.11611347435476564, + "learning_rate": 6.628669602392665e-05, + "loss": 0.0207, + "step": 9401 + }, + { + "epoch": 4.424470588235294, + "grad_norm": 0.12270628292550698, + "learning_rate": 6.625032515950865e-05, + "loss": 0.0449, + "step": 9402 + }, + { + "epoch": 4.424941176470588, + "grad_norm": 0.08841332774294425, + "learning_rate": 6.621396144772365e-05, + "loss": 0.0078, + "step": 9403 + }, + { + "epoch": 4.425411764705882, + "grad_norm": 0.1427579561631702, + "learning_rate": 6.617760489167738e-05, + "loss": 0.0251, + "step": 9404 + }, + { + "epoch": 4.425882352941176, + "grad_norm": 0.1781667852368541, + "learning_rate": 6.614125549447475e-05, + "loss": 0.06, + "step": 9405 + }, + { + "epoch": 4.426352941176471, + "grad_norm": 0.037122516193346826, + "learning_rate": 6.610491325922022e-05, + "loss": 0.005, + "step": 9406 + }, + { + "epoch": 4.426823529411765, + "grad_norm": 2.4094387139055175e-06, + "learning_rate": 6.606857818901765e-05, + "loss": 0.0, + "step": 9407 + }, + { + "epoch": 4.427294117647059, + "grad_norm": 3.267729721590301e-06, + "learning_rate": 6.603225028697013e-05, + "loss": 0.0, + "step": 9408 + }, + { + "epoch": 4.427764705882353, + "grad_norm": 5.168367727346821e-06, + "learning_rate": 6.59959295561803e-05, + "loss": 0.0, + "step": 9409 + }, + { + "epoch": 4.428235294117647, + "grad_norm": 0.11625731809331942, + "learning_rate": 6.595961599975016e-05, + "loss": 0.0436, + "step": 9410 + }, + { + "epoch": 4.428705882352941, + "grad_norm": 0.08770515336704905, + "learning_rate": 6.5923309620781e-05, + "loss": 0.0138, + "step": 9411 + }, + { + "epoch": 4.429176470588235, + "grad_norm": 0.05137987190082104, + "learning_rate": 6.588701042237358e-05, + "loss": 0.004, + "step": 9412 + }, + { + "epoch": 4.429647058823529, + "grad_norm": 0.0482259107307213, + "learning_rate": 6.585071840762809e-05, + "loss": 0.006, + "step": 9413 + }, + { + "epoch": 4.4301176470588235, + "grad_norm": 0.10644804647242473, + "learning_rate": 6.581443357964395e-05, + "loss": 0.0049, + "step": 9414 + }, + { + "epoch": 4.430588235294118, + "grad_norm": 0.06346132169537629, + "learning_rate": 6.577815594152017e-05, + "loss": 0.013, + "step": 9415 + }, + { + "epoch": 4.431058823529412, + "grad_norm": 0.10738943736434985, + "learning_rate": 6.574188549635493e-05, + "loss": 0.0257, + "step": 9416 + }, + { + "epoch": 4.431529411764706, + "grad_norm": 0.06751103955396708, + "learning_rate": 6.570562224724593e-05, + "loss": 0.0131, + "step": 9417 + }, + { + "epoch": 4.432, + "grad_norm": 2.19314074120265e-05, + "learning_rate": 6.566936619729033e-05, + "loss": 0.0, + "step": 9418 + }, + { + "epoch": 4.432470588235294, + "grad_norm": 0.09756127874790368, + "learning_rate": 6.563311734958444e-05, + "loss": 0.0114, + "step": 9419 + }, + { + "epoch": 4.432941176470588, + "grad_norm": 0.05122277363854297, + "learning_rate": 6.559687570722413e-05, + "loss": 0.0064, + "step": 9420 + }, + { + "epoch": 4.433411764705882, + "grad_norm": 0.08840216475400871, + "learning_rate": 6.556064127330467e-05, + "loss": 0.0251, + "step": 9421 + }, + { + "epoch": 4.433882352941176, + "grad_norm": 0.15425490876531808, + "learning_rate": 6.552441405092058e-05, + "loss": 0.0537, + "step": 9422 + }, + { + "epoch": 4.434352941176471, + "grad_norm": 4.74484494935179e-06, + "learning_rate": 6.54881940431659e-05, + "loss": 0.0, + "step": 9423 + }, + { + "epoch": 4.434823529411765, + "grad_norm": 0.14138032734358502, + "learning_rate": 6.5451981253134e-05, + "loss": 0.0284, + "step": 9424 + }, + { + "epoch": 4.435294117647059, + "grad_norm": 0.14071314493668, + "learning_rate": 6.541577568391758e-05, + "loss": 0.0191, + "step": 9425 + }, + { + "epoch": 4.435764705882353, + "grad_norm": 0.13885759409792003, + "learning_rate": 6.537957733860879e-05, + "loss": 0.017, + "step": 9426 + }, + { + "epoch": 4.4362352941176475, + "grad_norm": 0.13854927426391217, + "learning_rate": 6.53433862202992e-05, + "loss": 0.0486, + "step": 9427 + }, + { + "epoch": 4.436705882352941, + "grad_norm": 0.06009847619534588, + "learning_rate": 6.530720233207962e-05, + "loss": 0.0103, + "step": 9428 + }, + { + "epoch": 4.437176470588235, + "grad_norm": 0.06192566764684202, + "learning_rate": 6.527102567704036e-05, + "loss": 0.0096, + "step": 9429 + }, + { + "epoch": 4.437647058823529, + "grad_norm": 0.07782752374877719, + "learning_rate": 6.523485625827119e-05, + "loss": 0.0128, + "step": 9430 + }, + { + "epoch": 4.4381176470588235, + "grad_norm": 5.630798101008824e-06, + "learning_rate": 6.519869407886098e-05, + "loss": 0.0, + "step": 9431 + }, + { + "epoch": 4.438588235294118, + "grad_norm": 0.0530463023669995, + "learning_rate": 6.516253914189825e-05, + "loss": 0.0094, + "step": 9432 + }, + { + "epoch": 4.439058823529412, + "grad_norm": 0.0013533122849770944, + "learning_rate": 6.512639145047086e-05, + "loss": 0.0, + "step": 9433 + }, + { + "epoch": 4.439529411764706, + "grad_norm": 0.13615586134391572, + "learning_rate": 6.509025100766591e-05, + "loss": 0.0472, + "step": 9434 + }, + { + "epoch": 4.44, + "grad_norm": 0.15230249138603738, + "learning_rate": 6.505411781656997e-05, + "loss": 0.0476, + "step": 9435 + }, + { + "epoch": 4.440470588235294, + "grad_norm": 0.019190681999978934, + "learning_rate": 6.50179918802691e-05, + "loss": 0.0007, + "step": 9436 + }, + { + "epoch": 4.440941176470588, + "grad_norm": 0.10119812376497613, + "learning_rate": 6.49818732018485e-05, + "loss": 0.0198, + "step": 9437 + }, + { + "epoch": 4.441411764705882, + "grad_norm": 0.1039857029449342, + "learning_rate": 6.494576178439298e-05, + "loss": 0.0167, + "step": 9438 + }, + { + "epoch": 4.441882352941176, + "grad_norm": 0.08637372911806662, + "learning_rate": 6.490965763098654e-05, + "loss": 0.0235, + "step": 9439 + }, + { + "epoch": 4.442352941176471, + "grad_norm": 3.032035112346808e-05, + "learning_rate": 6.487356074471272e-05, + "loss": 0.0, + "step": 9440 + }, + { + "epoch": 4.442823529411765, + "grad_norm": 0.03504081642949196, + "learning_rate": 6.483747112865436e-05, + "loss": 0.0051, + "step": 9441 + }, + { + "epoch": 4.443294117647059, + "grad_norm": 0.0001309232978339181, + "learning_rate": 6.480138878589368e-05, + "loss": 0.0, + "step": 9442 + }, + { + "epoch": 4.443764705882353, + "grad_norm": 1.6342644203559563e-05, + "learning_rate": 6.476531371951229e-05, + "loss": 0.0, + "step": 9443 + }, + { + "epoch": 4.444235294117647, + "grad_norm": 0.00024014084236856153, + "learning_rate": 6.472924593259122e-05, + "loss": 0.0, + "step": 9444 + }, + { + "epoch": 4.444705882352941, + "grad_norm": 0.10571332519409893, + "learning_rate": 6.469318542821078e-05, + "loss": 0.0153, + "step": 9445 + }, + { + "epoch": 4.445176470588235, + "grad_norm": 0.09001282994995914, + "learning_rate": 6.465713220945075e-05, + "loss": 0.0253, + "step": 9446 + }, + { + "epoch": 4.445647058823529, + "grad_norm": 3.046826144099116e-06, + "learning_rate": 6.46210862793902e-05, + "loss": 0.0, + "step": 9447 + }, + { + "epoch": 4.4461176470588235, + "grad_norm": 0.1052706246806325, + "learning_rate": 6.458504764110764e-05, + "loss": 0.0325, + "step": 9448 + }, + { + "epoch": 4.446588235294118, + "grad_norm": 0.1430822944411002, + "learning_rate": 6.454901629768105e-05, + "loss": 0.0378, + "step": 9449 + }, + { + "epoch": 4.447058823529412, + "grad_norm": 0.08493210219617345, + "learning_rate": 6.451299225218754e-05, + "loss": 0.0226, + "step": 9450 + }, + { + "epoch": 4.447529411764706, + "grad_norm": 0.03937484065616031, + "learning_rate": 6.447697550770381e-05, + "loss": 0.0068, + "step": 9451 + }, + { + "epoch": 4.448, + "grad_norm": 0.11391800826396442, + "learning_rate": 6.44409660673059e-05, + "loss": 0.0298, + "step": 9452 + }, + { + "epoch": 4.448470588235294, + "grad_norm": 4.769828496596303e-05, + "learning_rate": 6.440496393406912e-05, + "loss": 0.0, + "step": 9453 + }, + { + "epoch": 4.448941176470588, + "grad_norm": 0.09818612389991822, + "learning_rate": 6.436896911106824e-05, + "loss": 0.0322, + "step": 9454 + }, + { + "epoch": 4.449411764705882, + "grad_norm": 0.13216413361044557, + "learning_rate": 6.433298160137749e-05, + "loss": 0.049, + "step": 9455 + }, + { + "epoch": 4.449882352941176, + "grad_norm": 5.028111371218778e-05, + "learning_rate": 6.429700140807023e-05, + "loss": 0.0, + "step": 9456 + }, + { + "epoch": 4.450352941176471, + "grad_norm": 0.1455478585136845, + "learning_rate": 6.426102853421942e-05, + "loss": 0.0407, + "step": 9457 + }, + { + "epoch": 4.450823529411765, + "grad_norm": 0.15438418373843854, + "learning_rate": 6.42250629828974e-05, + "loss": 0.0404, + "step": 9458 + }, + { + "epoch": 4.451294117647059, + "grad_norm": 0.10493712802351526, + "learning_rate": 6.418910475717565e-05, + "loss": 0.0247, + "step": 9459 + }, + { + "epoch": 4.451764705882353, + "grad_norm": 0.11002054044739427, + "learning_rate": 6.415315386012525e-05, + "loss": 0.0347, + "step": 9460 + }, + { + "epoch": 4.4522352941176475, + "grad_norm": 0.0010598367165655946, + "learning_rate": 6.411721029481665e-05, + "loss": 0.0, + "step": 9461 + }, + { + "epoch": 4.452705882352941, + "grad_norm": 0.00013504593491573717, + "learning_rate": 6.40812740643195e-05, + "loss": 0.0, + "step": 9462 + }, + { + "epoch": 4.453176470588235, + "grad_norm": 0.1534567967616275, + "learning_rate": 6.404534517170296e-05, + "loss": 0.0504, + "step": 9463 + }, + { + "epoch": 4.453647058823529, + "grad_norm": 0.04602090058023425, + "learning_rate": 6.40094236200356e-05, + "loss": 0.0061, + "step": 9464 + }, + { + "epoch": 4.4541176470588235, + "grad_norm": 0.20075389999137047, + "learning_rate": 6.397350941238519e-05, + "loss": 0.0638, + "step": 9465 + }, + { + "epoch": 4.454588235294118, + "grad_norm": 0.10527184163986249, + "learning_rate": 6.39376025518191e-05, + "loss": 0.0256, + "step": 9466 + }, + { + "epoch": 4.455058823529412, + "grad_norm": 0.07872292940740283, + "learning_rate": 6.390170304140382e-05, + "loss": 0.0157, + "step": 9467 + }, + { + "epoch": 4.455529411764706, + "grad_norm": 0.0789341982415684, + "learning_rate": 6.386581088420542e-05, + "loss": 0.013, + "step": 9468 + }, + { + "epoch": 4.456, + "grad_norm": 0.0663710633987957, + "learning_rate": 6.382992608328929e-05, + "loss": 0.0122, + "step": 9469 + }, + { + "epoch": 4.456470588235294, + "grad_norm": 0.16088213431275752, + "learning_rate": 6.37940486417201e-05, + "loss": 0.0642, + "step": 9470 + }, + { + "epoch": 4.456941176470588, + "grad_norm": 0.13357448703059238, + "learning_rate": 6.375817856256199e-05, + "loss": 0.0457, + "step": 9471 + }, + { + "epoch": 4.457411764705882, + "grad_norm": 6.047747877541728e-06, + "learning_rate": 6.37223158488785e-05, + "loss": 0.0, + "step": 9472 + }, + { + "epoch": 4.457882352941176, + "grad_norm": 0.027153130400538322, + "learning_rate": 6.368646050373236e-05, + "loss": 0.0038, + "step": 9473 + }, + { + "epoch": 4.458352941176471, + "grad_norm": 0.11746083189601055, + "learning_rate": 6.365061253018586e-05, + "loss": 0.0294, + "step": 9474 + }, + { + "epoch": 4.458823529411765, + "grad_norm": 0.021002176276992784, + "learning_rate": 6.361477193130065e-05, + "loss": 0.0025, + "step": 9475 + }, + { + "epoch": 4.459294117647059, + "grad_norm": 0.11321937979292944, + "learning_rate": 6.35789387101376e-05, + "loss": 0.0299, + "step": 9476 + }, + { + "epoch": 4.459764705882353, + "grad_norm": 0.021857231278123074, + "learning_rate": 6.354311286975705e-05, + "loss": 0.0023, + "step": 9477 + }, + { + "epoch": 4.460235294117647, + "grad_norm": 0.04301487680968027, + "learning_rate": 6.35072944132188e-05, + "loss": 0.005, + "step": 9478 + }, + { + "epoch": 4.460705882352941, + "grad_norm": 0.05167387051023001, + "learning_rate": 6.34714833435818e-05, + "loss": 0.0092, + "step": 9479 + }, + { + "epoch": 4.461176470588235, + "grad_norm": 0.13415140962364758, + "learning_rate": 6.343567966390452e-05, + "loss": 0.0239, + "step": 9480 + }, + { + "epoch": 4.461647058823529, + "grad_norm": 9.890716018751324e-06, + "learning_rate": 6.339988337724487e-05, + "loss": 0.0, + "step": 9481 + }, + { + "epoch": 4.4621176470588235, + "grad_norm": 0.07970341858871151, + "learning_rate": 6.336409448665988e-05, + "loss": 0.0198, + "step": 9482 + }, + { + "epoch": 4.462588235294118, + "grad_norm": 0.13903065988340077, + "learning_rate": 6.332831299520617e-05, + "loss": 0.0636, + "step": 9483 + }, + { + "epoch": 4.463058823529412, + "grad_norm": 0.15818964035521893, + "learning_rate": 6.32925389059397e-05, + "loss": 0.0377, + "step": 9484 + }, + { + "epoch": 4.463529411764706, + "grad_norm": 0.09181091817299881, + "learning_rate": 6.325677222191568e-05, + "loss": 0.0178, + "step": 9485 + }, + { + "epoch": 4.464, + "grad_norm": 0.10257817885301597, + "learning_rate": 6.322101294618875e-05, + "loss": 0.0298, + "step": 9486 + }, + { + "epoch": 4.464470588235294, + "grad_norm": 0.07537479345143741, + "learning_rate": 6.318526108181299e-05, + "loss": 0.0095, + "step": 9487 + }, + { + "epoch": 4.464941176470588, + "grad_norm": 0.10255728252317863, + "learning_rate": 6.314951663184175e-05, + "loss": 0.0346, + "step": 9488 + }, + { + "epoch": 4.465411764705882, + "grad_norm": 0.013325134574343062, + "learning_rate": 6.311377959932783e-05, + "loss": 0.0006, + "step": 9489 + }, + { + "epoch": 4.465882352941176, + "grad_norm": 0.11946185110337601, + "learning_rate": 6.307804998732328e-05, + "loss": 0.0252, + "step": 9490 + }, + { + "epoch": 4.466352941176471, + "grad_norm": 0.10245660177727423, + "learning_rate": 6.304232779887958e-05, + "loss": 0.0291, + "step": 9491 + }, + { + "epoch": 4.466823529411765, + "grad_norm": 0.017747292243986874, + "learning_rate": 6.300661303704768e-05, + "loss": 0.0013, + "step": 9492 + }, + { + "epoch": 4.467294117647059, + "grad_norm": 0.03479006845024289, + "learning_rate": 6.297090570487766e-05, + "loss": 0.0017, + "step": 9493 + }, + { + "epoch": 4.467764705882353, + "grad_norm": 0.11719250144345751, + "learning_rate": 6.293520580541925e-05, + "loss": 0.0388, + "step": 9494 + }, + { + "epoch": 4.4682352941176475, + "grad_norm": 0.18973121157696196, + "learning_rate": 6.289951334172123e-05, + "loss": 0.0464, + "step": 9495 + }, + { + "epoch": 4.468705882352941, + "grad_norm": 0.06406043060303496, + "learning_rate": 6.2863828316832e-05, + "loss": 0.0113, + "step": 9496 + }, + { + "epoch": 4.469176470588235, + "grad_norm": 0.06268881768601184, + "learning_rate": 6.282815073379932e-05, + "loss": 0.0052, + "step": 9497 + }, + { + "epoch": 4.469647058823529, + "grad_norm": 0.1662354725718986, + "learning_rate": 6.279248059567006e-05, + "loss": 0.0451, + "step": 9498 + }, + { + "epoch": 4.4701176470588235, + "grad_norm": 0.11169321264945777, + "learning_rate": 6.275681790549072e-05, + "loss": 0.0216, + "step": 9499 + }, + { + "epoch": 4.470588235294118, + "grad_norm": 0.11133405974846755, + "learning_rate": 6.27211626663071e-05, + "loss": 0.0156, + "step": 9500 + }, + { + "epoch": 4.471058823529412, + "grad_norm": 0.1275086762073142, + "learning_rate": 6.268551488116423e-05, + "loss": 0.0351, + "step": 9501 + }, + { + "epoch": 4.471529411764706, + "grad_norm": 0.0360099776021776, + "learning_rate": 6.264987455310667e-05, + "loss": 0.0031, + "step": 9502 + }, + { + "epoch": 4.4719999999999995, + "grad_norm": 0.06115125587287235, + "learning_rate": 6.261424168517832e-05, + "loss": 0.0087, + "step": 9503 + }, + { + "epoch": 4.472470588235294, + "grad_norm": 0.09289591781715477, + "learning_rate": 6.257861628042233e-05, + "loss": 0.0155, + "step": 9504 + }, + { + "epoch": 4.472941176470588, + "grad_norm": 0.14602525597996185, + "learning_rate": 6.254299834188128e-05, + "loss": 0.0438, + "step": 9505 + }, + { + "epoch": 4.473411764705882, + "grad_norm": 0.054805008593916175, + "learning_rate": 6.25073878725972e-05, + "loss": 0.0061, + "step": 9506 + }, + { + "epoch": 4.473882352941176, + "grad_norm": 4.8520081729251984e-05, + "learning_rate": 6.24717848756113e-05, + "loss": 0.0, + "step": 9507 + }, + { + "epoch": 4.474352941176471, + "grad_norm": 0.11054185323822038, + "learning_rate": 6.24361893539643e-05, + "loss": 0.0191, + "step": 9508 + }, + { + "epoch": 4.474823529411765, + "grad_norm": 0.16680846314142841, + "learning_rate": 6.240060131069628e-05, + "loss": 0.0694, + "step": 9509 + }, + { + "epoch": 4.475294117647059, + "grad_norm": 0.1997465785587632, + "learning_rate": 6.23650207488465e-05, + "loss": 0.0895, + "step": 9510 + }, + { + "epoch": 4.475764705882353, + "grad_norm": 1.9820585544438904e-05, + "learning_rate": 6.232944767145384e-05, + "loss": 0.0, + "step": 9511 + }, + { + "epoch": 4.476235294117647, + "grad_norm": 0.04825260080878019, + "learning_rate": 6.229388208155639e-05, + "loss": 0.0054, + "step": 9512 + }, + { + "epoch": 4.476705882352941, + "grad_norm": 0.1459769706070962, + "learning_rate": 6.225832398219158e-05, + "loss": 0.0696, + "step": 9513 + }, + { + "epoch": 4.477176470588235, + "grad_norm": 0.08070049057625428, + "learning_rate": 6.222277337639624e-05, + "loss": 0.0183, + "step": 9514 + }, + { + "epoch": 4.477647058823529, + "grad_norm": 0.130862530849279, + "learning_rate": 6.218723026720668e-05, + "loss": 0.0114, + "step": 9515 + }, + { + "epoch": 4.4781176470588235, + "grad_norm": 0.14276908457705903, + "learning_rate": 6.21516946576583e-05, + "loss": 0.0494, + "step": 9516 + }, + { + "epoch": 4.478588235294118, + "grad_norm": 0.13735858340848545, + "learning_rate": 6.211616655078616e-05, + "loss": 0.0441, + "step": 9517 + }, + { + "epoch": 4.479058823529412, + "grad_norm": 0.05574083322737599, + "learning_rate": 6.208064594962441e-05, + "loss": 0.0074, + "step": 9518 + }, + { + "epoch": 4.479529411764706, + "grad_norm": 7.75196342119819e-07, + "learning_rate": 6.204513285720673e-05, + "loss": 0.0, + "step": 9519 + }, + { + "epoch": 4.48, + "grad_norm": 9.232449916147562e-06, + "learning_rate": 6.200962727656616e-05, + "loss": 0.0, + "step": 9520 + }, + { + "epoch": 4.480470588235294, + "grad_norm": 0.11691561458618625, + "learning_rate": 6.197412921073498e-05, + "loss": 0.0382, + "step": 9521 + }, + { + "epoch": 4.480941176470588, + "grad_norm": 0.11903941953680104, + "learning_rate": 6.193863866274492e-05, + "loss": 0.0361, + "step": 9522 + }, + { + "epoch": 4.481411764705882, + "grad_norm": 0.16102255980075195, + "learning_rate": 6.190315563562711e-05, + "loss": 0.0357, + "step": 9523 + }, + { + "epoch": 4.481882352941176, + "grad_norm": 0.14936378040818304, + "learning_rate": 6.186768013241188e-05, + "loss": 0.0415, + "step": 9524 + }, + { + "epoch": 4.482352941176471, + "grad_norm": 3.241136379895256e-05, + "learning_rate": 6.183221215612904e-05, + "loss": 0.0, + "step": 9525 + }, + { + "epoch": 4.482823529411765, + "grad_norm": 0.05800720635853753, + "learning_rate": 6.179675170980779e-05, + "loss": 0.0124, + "step": 9526 + }, + { + "epoch": 4.483294117647059, + "grad_norm": 5.534289432986759e-05, + "learning_rate": 6.176129879647654e-05, + "loss": 0.0, + "step": 9527 + }, + { + "epoch": 4.483764705882353, + "grad_norm": 0.11324193369274159, + "learning_rate": 6.172585341916317e-05, + "loss": 0.0257, + "step": 9528 + }, + { + "epoch": 4.484235294117647, + "grad_norm": 9.607341237251009e-06, + "learning_rate": 6.169041558089496e-05, + "loss": 0.0, + "step": 9529 + }, + { + "epoch": 4.484705882352941, + "grad_norm": 0.1345427667072349, + "learning_rate": 6.16549852846983e-05, + "loss": 0.0439, + "step": 9530 + }, + { + "epoch": 4.485176470588235, + "grad_norm": 0.09355201322372352, + "learning_rate": 6.161956253359935e-05, + "loss": 0.0276, + "step": 9531 + }, + { + "epoch": 4.485647058823529, + "grad_norm": 7.22556110194029e-05, + "learning_rate": 6.15841473306232e-05, + "loss": 0.0, + "step": 9532 + }, + { + "epoch": 4.4861176470588235, + "grad_norm": 0.16573218776582, + "learning_rate": 6.154873967879455e-05, + "loss": 0.059, + "step": 9533 + }, + { + "epoch": 4.486588235294118, + "grad_norm": 0.08889216429399263, + "learning_rate": 6.151333958113743e-05, + "loss": 0.0211, + "step": 9534 + }, + { + "epoch": 4.487058823529412, + "grad_norm": 0.07038802267319208, + "learning_rate": 6.147794704067508e-05, + "loss": 0.0177, + "step": 9535 + }, + { + "epoch": 4.487529411764706, + "grad_norm": 0.00015658079321142042, + "learning_rate": 6.144256206043025e-05, + "loss": 0.0, + "step": 9536 + }, + { + "epoch": 4.4879999999999995, + "grad_norm": 7.2539502139677e-05, + "learning_rate": 6.140718464342505e-05, + "loss": 0.0, + "step": 9537 + }, + { + "epoch": 4.488470588235294, + "grad_norm": 0.06721100827960666, + "learning_rate": 6.137181479268077e-05, + "loss": 0.005, + "step": 9538 + }, + { + "epoch": 4.488941176470588, + "grad_norm": 0.0986674531882506, + "learning_rate": 6.133645251121823e-05, + "loss": 0.0072, + "step": 9539 + }, + { + "epoch": 4.489411764705882, + "grad_norm": 0.14339977371003568, + "learning_rate": 6.130109780205759e-05, + "loss": 0.0235, + "step": 9540 + }, + { + "epoch": 4.489882352941176, + "grad_norm": 5.72955300252386e-05, + "learning_rate": 6.126575066821821e-05, + "loss": 0.0, + "step": 9541 + }, + { + "epoch": 4.490352941176471, + "grad_norm": 0.07769112494758564, + "learning_rate": 6.123041111271896e-05, + "loss": 0.0196, + "step": 9542 + }, + { + "epoch": 4.490823529411765, + "grad_norm": 0.024218762696550503, + "learning_rate": 6.119507913857807e-05, + "loss": 0.0015, + "step": 9543 + }, + { + "epoch": 4.491294117647059, + "grad_norm": 0.15228016818073276, + "learning_rate": 6.115975474881296e-05, + "loss": 0.0217, + "step": 9544 + }, + { + "epoch": 4.491764705882353, + "grad_norm": 0.11967269254164435, + "learning_rate": 6.11244379464406e-05, + "loss": 0.0506, + "step": 9545 + }, + { + "epoch": 4.492235294117647, + "grad_norm": 0.06601309108930392, + "learning_rate": 6.108912873447712e-05, + "loss": 0.0182, + "step": 9546 + }, + { + "epoch": 4.492705882352941, + "grad_norm": 4.760513669431617e-05, + "learning_rate": 6.105382711593814e-05, + "loss": 0.0, + "step": 9547 + }, + { + "epoch": 4.493176470588235, + "grad_norm": 0.14768563805082738, + "learning_rate": 6.101853309383866e-05, + "loss": 0.0503, + "step": 9548 + }, + { + "epoch": 4.493647058823529, + "grad_norm": 0.14737287559637577, + "learning_rate": 6.0983246671192866e-05, + "loss": 0.0201, + "step": 9549 + }, + { + "epoch": 4.4941176470588236, + "grad_norm": 0.051182984240090844, + "learning_rate": 6.0947967851014405e-05, + "loss": 0.0035, + "step": 9550 + }, + { + "epoch": 4.494588235294118, + "grad_norm": 4.921453347956073e-05, + "learning_rate": 6.091269663631636e-05, + "loss": 0.0, + "step": 9551 + }, + { + "epoch": 4.495058823529412, + "grad_norm": 0.1588222642437883, + "learning_rate": 6.0877433030110935e-05, + "loss": 0.0551, + "step": 9552 + }, + { + "epoch": 4.495529411764706, + "grad_norm": 6.472905104154432e-05, + "learning_rate": 6.084217703540988e-05, + "loss": 0.0, + "step": 9553 + }, + { + "epoch": 4.496, + "grad_norm": 6.559099929692535e-05, + "learning_rate": 6.0806928655224266e-05, + "loss": 0.0, + "step": 9554 + }, + { + "epoch": 4.496470588235294, + "grad_norm": 0.07119565358969876, + "learning_rate": 6.0771687892564394e-05, + "loss": 0.015, + "step": 9555 + }, + { + "epoch": 4.496941176470588, + "grad_norm": 0.07588306697938423, + "learning_rate": 6.0736454750440025e-05, + "loss": 0.0284, + "step": 9556 + }, + { + "epoch": 4.497411764705882, + "grad_norm": 0.09941883043223583, + "learning_rate": 6.070122923186032e-05, + "loss": 0.0207, + "step": 9557 + }, + { + "epoch": 4.497882352941176, + "grad_norm": 0.1320919224533403, + "learning_rate": 6.0666011339833584e-05, + "loss": 0.0454, + "step": 9558 + }, + { + "epoch": 4.498352941176471, + "grad_norm": 0.13048458282189676, + "learning_rate": 6.063080107736767e-05, + "loss": 0.0323, + "step": 9559 + }, + { + "epoch": 4.498823529411765, + "grad_norm": 0.0002877000503882659, + "learning_rate": 6.0595598447469726e-05, + "loss": 0.0, + "step": 9560 + }, + { + "epoch": 4.499294117647059, + "grad_norm": 0.24124968710603265, + "learning_rate": 6.056040345314618e-05, + "loss": 0.0435, + "step": 9561 + }, + { + "epoch": 4.499764705882353, + "grad_norm": 0.00021276039721746545, + "learning_rate": 6.0525216097402865e-05, + "loss": 0.0, + "step": 9562 + }, + { + "epoch": 4.5002352941176476, + "grad_norm": 0.17358123529228017, + "learning_rate": 6.049003638324501e-05, + "loss": 0.0276, + "step": 9563 + }, + { + "epoch": 4.500705882352941, + "grad_norm": 0.15573606248547783, + "learning_rate": 6.0454864313677054e-05, + "loss": 0.0573, + "step": 9564 + }, + { + "epoch": 4.501176470588235, + "grad_norm": 0.1537749263313831, + "learning_rate": 6.041969989170289e-05, + "loss": 0.0511, + "step": 9565 + }, + { + "epoch": 4.501647058823529, + "grad_norm": 0.00013172512555258435, + "learning_rate": 6.0384543120325796e-05, + "loss": 0.0, + "step": 9566 + }, + { + "epoch": 4.502117647058824, + "grad_norm": 0.06945707126579816, + "learning_rate": 6.034939400254824e-05, + "loss": 0.013, + "step": 9567 + }, + { + "epoch": 4.502588235294118, + "grad_norm": 0.0898601595921015, + "learning_rate": 6.031425254137222e-05, + "loss": 0.0159, + "step": 9568 + }, + { + "epoch": 4.503058823529412, + "grad_norm": 0.018924312909038347, + "learning_rate": 6.0279118739798874e-05, + "loss": 0.0013, + "step": 9569 + }, + { + "epoch": 4.503529411764706, + "grad_norm": 0.0005524389646594921, + "learning_rate": 6.0243992600828875e-05, + "loss": 0.0, + "step": 9570 + }, + { + "epoch": 4.504, + "grad_norm": 0.1168289741144243, + "learning_rate": 6.020887412746221e-05, + "loss": 0.024, + "step": 9571 + }, + { + "epoch": 4.504470588235294, + "grad_norm": 0.001238113999061704, + "learning_rate": 6.017376332269807e-05, + "loss": 0.0, + "step": 9572 + }, + { + "epoch": 4.504941176470588, + "grad_norm": 0.10977277553680553, + "learning_rate": 6.0138660189535126e-05, + "loss": 0.0297, + "step": 9573 + }, + { + "epoch": 4.505411764705882, + "grad_norm": 4.530625040926578e-05, + "learning_rate": 6.0103564730971375e-05, + "loss": 0.0, + "step": 9574 + }, + { + "epoch": 4.5058823529411764, + "grad_norm": 1.906364060418076e-05, + "learning_rate": 6.006847695000415e-05, + "loss": 0.0, + "step": 9575 + }, + { + "epoch": 4.506352941176471, + "grad_norm": 0.09734978293856543, + "learning_rate": 6.0033396849630136e-05, + "loss": 0.0137, + "step": 9576 + }, + { + "epoch": 4.506823529411765, + "grad_norm": 0.00018712577719446583, + "learning_rate": 5.999832443284527e-05, + "loss": 0.0, + "step": 9577 + }, + { + "epoch": 4.507294117647059, + "grad_norm": 0.13168957279699864, + "learning_rate": 5.9963259702644963e-05, + "loss": 0.0434, + "step": 9578 + }, + { + "epoch": 4.507764705882353, + "grad_norm": 0.12654094419792536, + "learning_rate": 5.992820266202394e-05, + "loss": 0.0515, + "step": 9579 + }, + { + "epoch": 4.508235294117647, + "grad_norm": 0.09150611557392363, + "learning_rate": 5.989315331397616e-05, + "loss": 0.016, + "step": 9580 + }, + { + "epoch": 4.508705882352941, + "grad_norm": 0.10138422234320624, + "learning_rate": 5.985811166149507e-05, + "loss": 0.0304, + "step": 9581 + }, + { + "epoch": 4.509176470588235, + "grad_norm": 0.06244688920976745, + "learning_rate": 5.982307770757343e-05, + "loss": 0.0136, + "step": 9582 + }, + { + "epoch": 4.509647058823529, + "grad_norm": 0.139876945224299, + "learning_rate": 5.978805145520322e-05, + "loss": 0.0307, + "step": 9583 + }, + { + "epoch": 4.510117647058824, + "grad_norm": 0.06386524090200063, + "learning_rate": 5.97530329073759e-05, + "loss": 0.0061, + "step": 9584 + }, + { + "epoch": 4.510588235294118, + "grad_norm": 0.10904990248519801, + "learning_rate": 5.971802206708229e-05, + "loss": 0.0224, + "step": 9585 + }, + { + "epoch": 4.511058823529412, + "grad_norm": 0.23933592399412065, + "learning_rate": 5.968301893731237e-05, + "loss": 0.1066, + "step": 9586 + }, + { + "epoch": 4.511529411764706, + "grad_norm": 0.15345297480971262, + "learning_rate": 5.964802352105564e-05, + "loss": 0.0421, + "step": 9587 + }, + { + "epoch": 4.5120000000000005, + "grad_norm": 0.14861523120085451, + "learning_rate": 5.961303582130092e-05, + "loss": 0.046, + "step": 9588 + }, + { + "epoch": 4.512470588235294, + "grad_norm": 3.7508368675338444e-06, + "learning_rate": 5.957805584103623e-05, + "loss": 0.0, + "step": 9589 + }, + { + "epoch": 4.512941176470588, + "grad_norm": 0.04524708551058661, + "learning_rate": 5.954308358324909e-05, + "loss": 0.0056, + "step": 9590 + }, + { + "epoch": 4.513411764705882, + "grad_norm": 0.05873635715109745, + "learning_rate": 5.950811905092636e-05, + "loss": 0.0121, + "step": 9591 + }, + { + "epoch": 4.5138823529411765, + "grad_norm": 0.17696163083090277, + "learning_rate": 5.947316224705407e-05, + "loss": 0.0575, + "step": 9592 + }, + { + "epoch": 4.514352941176471, + "grad_norm": 0.14672228847955437, + "learning_rate": 5.943821317461775e-05, + "loss": 0.0316, + "step": 9593 + }, + { + "epoch": 4.514823529411765, + "grad_norm": 0.16596718129878993, + "learning_rate": 5.940327183660228e-05, + "loss": 0.0527, + "step": 9594 + }, + { + "epoch": 4.515294117647059, + "grad_norm": 0.12320615024473228, + "learning_rate": 5.936833823599172e-05, + "loss": 0.0124, + "step": 9595 + }, + { + "epoch": 4.5157647058823525, + "grad_norm": 0.10165754527851827, + "learning_rate": 5.933341237576968e-05, + "loss": 0.0201, + "step": 9596 + }, + { + "epoch": 4.516235294117647, + "grad_norm": 0.05550003439964147, + "learning_rate": 5.9298494258918875e-05, + "loss": 0.0061, + "step": 9597 + }, + { + "epoch": 4.516705882352941, + "grad_norm": 6.477567210254469e-06, + "learning_rate": 5.926358388842156e-05, + "loss": 0.0, + "step": 9598 + }, + { + "epoch": 4.517176470588235, + "grad_norm": 0.08801222814824591, + "learning_rate": 5.922868126725929e-05, + "loss": 0.0056, + "step": 9599 + }, + { + "epoch": 4.517647058823529, + "grad_norm": 7.889428628563676e-06, + "learning_rate": 5.919378639841281e-05, + "loss": 0.0, + "step": 9600 + }, + { + "epoch": 4.518117647058824, + "grad_norm": 0.12019177238244638, + "learning_rate": 5.915889928486239e-05, + "loss": 0.0318, + "step": 9601 + }, + { + "epoch": 4.518588235294118, + "grad_norm": 0.10826006748019745, + "learning_rate": 5.912401992958757e-05, + "loss": 0.0334, + "step": 9602 + }, + { + "epoch": 4.519058823529412, + "grad_norm": 0.007890314170507144, + "learning_rate": 5.908914833556715e-05, + "loss": 0.0003, + "step": 9603 + }, + { + "epoch": 4.519529411764706, + "grad_norm": 0.05016701355862504, + "learning_rate": 5.9054284505779396e-05, + "loss": 0.008, + "step": 9604 + }, + { + "epoch": 4.52, + "grad_norm": 0.04800647228540285, + "learning_rate": 5.901942844320187e-05, + "loss": 0.0043, + "step": 9605 + }, + { + "epoch": 4.520470588235294, + "grad_norm": 1.4375633907874325e-06, + "learning_rate": 5.898458015081136e-05, + "loss": 0.0, + "step": 9606 + }, + { + "epoch": 4.520941176470588, + "grad_norm": 0.1478885527374291, + "learning_rate": 5.8949739631584136e-05, + "loss": 0.0265, + "step": 9607 + }, + { + "epoch": 4.521411764705882, + "grad_norm": 0.17794477228734132, + "learning_rate": 5.8914906888495797e-05, + "loss": 0.0761, + "step": 9608 + }, + { + "epoch": 4.5218823529411765, + "grad_norm": 0.0928617532268873, + "learning_rate": 5.888008192452112e-05, + "loss": 0.0145, + "step": 9609 + }, + { + "epoch": 4.522352941176471, + "grad_norm": 1.9802415799394708e-06, + "learning_rate": 5.8845264742634394e-05, + "loss": 0.0, + "step": 9610 + }, + { + "epoch": 4.522823529411765, + "grad_norm": 0.2276510559201213, + "learning_rate": 5.881045534580922e-05, + "loss": 0.0759, + "step": 9611 + }, + { + "epoch": 4.523294117647059, + "grad_norm": 0.1216451434334607, + "learning_rate": 5.87756537370184e-05, + "loss": 0.0463, + "step": 9612 + }, + { + "epoch": 4.523764705882353, + "grad_norm": 0.06195824994389862, + "learning_rate": 5.874085991923419e-05, + "loss": 0.0055, + "step": 9613 + }, + { + "epoch": 4.524235294117647, + "grad_norm": 0.06006164662273669, + "learning_rate": 5.8706073895428225e-05, + "loss": 0.0102, + "step": 9614 + }, + { + "epoch": 4.524705882352941, + "grad_norm": 0.06860046347395281, + "learning_rate": 5.867129566857129e-05, + "loss": 0.0056, + "step": 9615 + }, + { + "epoch": 4.525176470588235, + "grad_norm": 0.09383754617733504, + "learning_rate": 5.8636525241633673e-05, + "loss": 0.0176, + "step": 9616 + }, + { + "epoch": 4.525647058823529, + "grad_norm": 0.08273876447375911, + "learning_rate": 5.860176261758499e-05, + "loss": 0.0195, + "step": 9617 + }, + { + "epoch": 4.526117647058824, + "grad_norm": 0.11381671572476866, + "learning_rate": 5.8567007799394015e-05, + "loss": 0.0358, + "step": 9618 + }, + { + "epoch": 4.526588235294118, + "grad_norm": 0.15078923375482808, + "learning_rate": 5.853226079002906e-05, + "loss": 0.0595, + "step": 9619 + }, + { + "epoch": 4.527058823529412, + "grad_norm": 0.13213712560788196, + "learning_rate": 5.849752159245769e-05, + "loss": 0.0267, + "step": 9620 + }, + { + "epoch": 4.527529411764706, + "grad_norm": 0.12185096460428696, + "learning_rate": 5.84627902096468e-05, + "loss": 0.0521, + "step": 9621 + }, + { + "epoch": 4.5280000000000005, + "grad_norm": 0.16298958750982986, + "learning_rate": 5.842806664456264e-05, + "loss": 0.047, + "step": 9622 + }, + { + "epoch": 4.528470588235294, + "grad_norm": 0.07478961863162614, + "learning_rate": 5.839335090017072e-05, + "loss": 0.0163, + "step": 9623 + }, + { + "epoch": 4.528941176470588, + "grad_norm": 0.15600971199160624, + "learning_rate": 5.835864297943594e-05, + "loss": 0.0483, + "step": 9624 + }, + { + "epoch": 4.529411764705882, + "grad_norm": 0.07625728318337856, + "learning_rate": 5.8323942885322605e-05, + "loss": 0.0163, + "step": 9625 + }, + { + "epoch": 4.5298823529411765, + "grad_norm": 0.12905234670528745, + "learning_rate": 5.828925062079416e-05, + "loss": 0.0175, + "step": 9626 + }, + { + "epoch": 4.530352941176471, + "grad_norm": 0.22772180392968758, + "learning_rate": 5.8254566188813615e-05, + "loss": 0.0274, + "step": 9627 + }, + { + "epoch": 4.530823529411765, + "grad_norm": 0.08696298195764941, + "learning_rate": 5.821988959234308e-05, + "loss": 0.0234, + "step": 9628 + }, + { + "epoch": 4.531294117647059, + "grad_norm": 0.10146738545907537, + "learning_rate": 5.8185220834344124e-05, + "loss": 0.0332, + "step": 9629 + }, + { + "epoch": 4.5317647058823525, + "grad_norm": 0.14362075117474335, + "learning_rate": 5.815055991777773e-05, + "loss": 0.0513, + "step": 9630 + }, + { + "epoch": 4.532235294117647, + "grad_norm": 9.04443618577092e-05, + "learning_rate": 5.811590684560398e-05, + "loss": 0.0, + "step": 9631 + }, + { + "epoch": 4.532705882352941, + "grad_norm": 0.11360509847178087, + "learning_rate": 5.808126162078249e-05, + "loss": 0.0345, + "step": 9632 + }, + { + "epoch": 4.533176470588235, + "grad_norm": 1.4237977102004278e-05, + "learning_rate": 5.8046624246272146e-05, + "loss": 0.0, + "step": 9633 + }, + { + "epoch": 4.533647058823529, + "grad_norm": 0.03675308002062025, + "learning_rate": 5.801199472503109e-05, + "loss": 0.0008, + "step": 9634 + }, + { + "epoch": 4.534117647058824, + "grad_norm": 0.04954288022520024, + "learning_rate": 5.797737306001686e-05, + "loss": 0.0079, + "step": 9635 + }, + { + "epoch": 4.534588235294118, + "grad_norm": 0.15737792433917508, + "learning_rate": 5.794275925418641e-05, + "loss": 0.0438, + "step": 9636 + }, + { + "epoch": 4.535058823529412, + "grad_norm": 0.09519968351823134, + "learning_rate": 5.79081533104958e-05, + "loss": 0.0321, + "step": 9637 + }, + { + "epoch": 4.535529411764706, + "grad_norm": 0.07269146369941695, + "learning_rate": 5.787355523190059e-05, + "loss": 0.0131, + "step": 9638 + }, + { + "epoch": 4.536, + "grad_norm": 0.04552545143220528, + "learning_rate": 5.783896502135572e-05, + "loss": 0.0056, + "step": 9639 + }, + { + "epoch": 4.536470588235294, + "grad_norm": 0.04820631942514957, + "learning_rate": 5.780438268181522e-05, + "loss": 0.0072, + "step": 9640 + }, + { + "epoch": 4.536941176470588, + "grad_norm": 0.07270861296185867, + "learning_rate": 5.776980821623264e-05, + "loss": 0.0118, + "step": 9641 + }, + { + "epoch": 4.537411764705882, + "grad_norm": 0.0778814350777456, + "learning_rate": 5.7735241627560905e-05, + "loss": 0.0184, + "step": 9642 + }, + { + "epoch": 4.5378823529411765, + "grad_norm": 0.11377596969786936, + "learning_rate": 5.770068291875202e-05, + "loss": 0.0319, + "step": 9643 + }, + { + "epoch": 4.538352941176471, + "grad_norm": 0.0794712852545769, + "learning_rate": 5.766613209275754e-05, + "loss": 0.012, + "step": 9644 + }, + { + "epoch": 4.538823529411765, + "grad_norm": 0.0675199603150438, + "learning_rate": 5.763158915252833e-05, + "loss": 0.0141, + "step": 9645 + }, + { + "epoch": 4.539294117647059, + "grad_norm": 4.588899013212172e-06, + "learning_rate": 5.759705410101443e-05, + "loss": 0.0, + "step": 9646 + }, + { + "epoch": 4.539764705882353, + "grad_norm": 5.46634467932741e-06, + "learning_rate": 5.756252694116538e-05, + "loss": 0.0, + "step": 9647 + }, + { + "epoch": 4.540235294117647, + "grad_norm": 0.15657664412050196, + "learning_rate": 5.752800767592988e-05, + "loss": 0.0313, + "step": 9648 + }, + { + "epoch": 4.540705882352941, + "grad_norm": 0.10496716721553605, + "learning_rate": 5.74934963082561e-05, + "loss": 0.0363, + "step": 9649 + }, + { + "epoch": 4.541176470588235, + "grad_norm": 0.09256745036849595, + "learning_rate": 5.745899284109154e-05, + "loss": 0.0212, + "step": 9650 + }, + { + "epoch": 4.541647058823529, + "grad_norm": 5.463514186058092e-06, + "learning_rate": 5.742449727738285e-05, + "loss": 0.0, + "step": 9651 + }, + { + "epoch": 4.542117647058824, + "grad_norm": 0.05831630855994978, + "learning_rate": 5.739000962007616e-05, + "loss": 0.0099, + "step": 9652 + }, + { + "epoch": 4.542588235294118, + "grad_norm": 0.08157075222604487, + "learning_rate": 5.7355529872116957e-05, + "loss": 0.0187, + "step": 9653 + }, + { + "epoch": 4.543058823529412, + "grad_norm": 0.09181335491146178, + "learning_rate": 5.732105803644986e-05, + "loss": 0.0289, + "step": 9654 + }, + { + "epoch": 4.543529411764706, + "grad_norm": 0.08354942341669448, + "learning_rate": 5.7286594116019e-05, + "loss": 0.0157, + "step": 9655 + }, + { + "epoch": 4.5440000000000005, + "grad_norm": 0.16703661664066669, + "learning_rate": 5.725213811376782e-05, + "loss": 0.0362, + "step": 9656 + }, + { + "epoch": 4.544470588235294, + "grad_norm": 0.17117120478706951, + "learning_rate": 5.7217690032638916e-05, + "loss": 0.0678, + "step": 9657 + }, + { + "epoch": 4.544941176470588, + "grad_norm": 5.300303385224314e-06, + "learning_rate": 5.718324987557438e-05, + "loss": 0.0, + "step": 9658 + }, + { + "epoch": 4.545411764705882, + "grad_norm": 0.11719554337147735, + "learning_rate": 5.714881764551561e-05, + "loss": 0.0222, + "step": 9659 + }, + { + "epoch": 4.5458823529411765, + "grad_norm": 0.14622377987034854, + "learning_rate": 5.7114393345403194e-05, + "loss": 0.0121, + "step": 9660 + }, + { + "epoch": 4.546352941176471, + "grad_norm": 0.10614089254861339, + "learning_rate": 5.707997697817719e-05, + "loss": 0.0251, + "step": 9661 + }, + { + "epoch": 4.546823529411765, + "grad_norm": 0.12899878129673822, + "learning_rate": 5.704556854677698e-05, + "loss": 0.0412, + "step": 9662 + }, + { + "epoch": 4.547294117647059, + "grad_norm": 0.20476188308658355, + "learning_rate": 5.701116805414104e-05, + "loss": 0.086, + "step": 9663 + }, + { + "epoch": 4.5477647058823525, + "grad_norm": 0.13834897813750538, + "learning_rate": 5.6976775503207574e-05, + "loss": 0.0381, + "step": 9664 + }, + { + "epoch": 4.548235294117647, + "grad_norm": 0.09161848924097785, + "learning_rate": 5.6942390896913686e-05, + "loss": 0.0161, + "step": 9665 + }, + { + "epoch": 4.548705882352941, + "grad_norm": 0.019873804135829985, + "learning_rate": 5.690801423819608e-05, + "loss": 0.0015, + "step": 9666 + }, + { + "epoch": 4.549176470588235, + "grad_norm": 1.672628700663299e-05, + "learning_rate": 5.687364552999071e-05, + "loss": 0.0, + "step": 9667 + }, + { + "epoch": 4.549647058823529, + "grad_norm": 0.06385697742601998, + "learning_rate": 5.683928477523275e-05, + "loss": 0.0115, + "step": 9668 + }, + { + "epoch": 4.550117647058824, + "grad_norm": 0.03755409452931381, + "learning_rate": 5.680493197685681e-05, + "loss": 0.0036, + "step": 9669 + }, + { + "epoch": 4.550588235294118, + "grad_norm": 0.11728476198597382, + "learning_rate": 5.677058713779687e-05, + "loss": 0.0391, + "step": 9670 + }, + { + "epoch": 4.551058823529412, + "grad_norm": 0.06687421875081782, + "learning_rate": 5.6736250260986e-05, + "loss": 0.0084, + "step": 9671 + }, + { + "epoch": 4.551529411764706, + "grad_norm": 0.05431147544234562, + "learning_rate": 5.6701921349356826e-05, + "loss": 0.0059, + "step": 9672 + }, + { + "epoch": 4.552, + "grad_norm": 0.23980852266776126, + "learning_rate": 5.666760040584126e-05, + "loss": 0.0567, + "step": 9673 + }, + { + "epoch": 4.552470588235294, + "grad_norm": 0.07168096540663471, + "learning_rate": 5.663328743337034e-05, + "loss": 0.014, + "step": 9674 + }, + { + "epoch": 4.552941176470588, + "grad_norm": 0.13658143457833724, + "learning_rate": 5.659898243487463e-05, + "loss": 0.0338, + "step": 9675 + }, + { + "epoch": 4.553411764705882, + "grad_norm": 0.06159467148893831, + "learning_rate": 5.6564685413284014e-05, + "loss": 0.0145, + "step": 9676 + }, + { + "epoch": 4.5538823529411765, + "grad_norm": 0.12607105000896238, + "learning_rate": 5.6530396371527515e-05, + "loss": 0.0307, + "step": 9677 + }, + { + "epoch": 4.554352941176471, + "grad_norm": 0.09253814590311858, + "learning_rate": 5.649611531253366e-05, + "loss": 0.0055, + "step": 9678 + }, + { + "epoch": 4.554823529411765, + "grad_norm": 0.11494957864571047, + "learning_rate": 5.646184223923017e-05, + "loss": 0.0393, + "step": 9679 + }, + { + "epoch": 4.555294117647059, + "grad_norm": 0.0816231555563016, + "learning_rate": 5.642757715454413e-05, + "loss": 0.0132, + "step": 9680 + }, + { + "epoch": 4.555764705882353, + "grad_norm": 0.07204396812648127, + "learning_rate": 5.639332006140203e-05, + "loss": 0.0098, + "step": 9681 + }, + { + "epoch": 4.556235294117647, + "grad_norm": 0.00025353235929667395, + "learning_rate": 5.635907096272949e-05, + "loss": 0.0, + "step": 9682 + }, + { + "epoch": 4.556705882352941, + "grad_norm": 0.11300633201849099, + "learning_rate": 5.632482986145159e-05, + "loss": 0.0415, + "step": 9683 + }, + { + "epoch": 4.557176470588235, + "grad_norm": 3.7762748010961926e-06, + "learning_rate": 5.6290596760492745e-05, + "loss": 0.0, + "step": 9684 + }, + { + "epoch": 4.557647058823529, + "grad_norm": 5.880678848053691e-05, + "learning_rate": 5.625637166277654e-05, + "loss": 0.0, + "step": 9685 + }, + { + "epoch": 4.558117647058824, + "grad_norm": 0.08303090414009197, + "learning_rate": 5.622215457122598e-05, + "loss": 0.0175, + "step": 9686 + }, + { + "epoch": 4.558588235294118, + "grad_norm": 0.03902071309141595, + "learning_rate": 5.618794548876346e-05, + "loss": 0.0032, + "step": 9687 + }, + { + "epoch": 4.559058823529412, + "grad_norm": 0.12795893655061105, + "learning_rate": 5.6153744418310505e-05, + "loss": 0.0288, + "step": 9688 + }, + { + "epoch": 4.559529411764705, + "grad_norm": 9.796979592762058e-06, + "learning_rate": 5.611955136278808e-05, + "loss": 0.0, + "step": 9689 + }, + { + "epoch": 4.5600000000000005, + "grad_norm": 4.939212460401496e-06, + "learning_rate": 5.608536632511649e-05, + "loss": 0.0, + "step": 9690 + }, + { + "epoch": 4.560470588235294, + "grad_norm": 0.15034466008097508, + "learning_rate": 5.6051189308215225e-05, + "loss": 0.0541, + "step": 9691 + }, + { + "epoch": 4.560941176470588, + "grad_norm": 0.07342205854899676, + "learning_rate": 5.6017020315003195e-05, + "loss": 0.0136, + "step": 9692 + }, + { + "epoch": 4.561411764705882, + "grad_norm": 0.12463484074928312, + "learning_rate": 5.598285934839868e-05, + "loss": 0.0437, + "step": 9693 + }, + { + "epoch": 4.5618823529411765, + "grad_norm": 0.022904105851640114, + "learning_rate": 5.594870641131907e-05, + "loss": 0.0015, + "step": 9694 + }, + { + "epoch": 4.562352941176471, + "grad_norm": 0.057069388532333544, + "learning_rate": 5.591456150668126e-05, + "loss": 0.0066, + "step": 9695 + }, + { + "epoch": 4.562823529411765, + "grad_norm": 0.1487410723878026, + "learning_rate": 5.588042463740143e-05, + "loss": 0.0491, + "step": 9696 + }, + { + "epoch": 4.563294117647059, + "grad_norm": 0.1022647772841539, + "learning_rate": 5.584629580639494e-05, + "loss": 0.0138, + "step": 9697 + }, + { + "epoch": 4.5637647058823525, + "grad_norm": 0.12731848916745547, + "learning_rate": 5.581217501657666e-05, + "loss": 0.0098, + "step": 9698 + }, + { + "epoch": 4.564235294117647, + "grad_norm": 0.06901507604198295, + "learning_rate": 5.577806227086057e-05, + "loss": 0.0156, + "step": 9699 + }, + { + "epoch": 4.564705882352941, + "grad_norm": 0.08550009901447585, + "learning_rate": 5.57439575721601e-05, + "loss": 0.0233, + "step": 9700 + }, + { + "epoch": 4.565176470588235, + "grad_norm": 0.036657156992292925, + "learning_rate": 5.570986092338806e-05, + "loss": 0.0039, + "step": 9701 + }, + { + "epoch": 4.565647058823529, + "grad_norm": 0.08316402882375945, + "learning_rate": 5.5675772327456313e-05, + "loss": 0.0209, + "step": 9702 + }, + { + "epoch": 4.566117647058824, + "grad_norm": 0.09564426465383205, + "learning_rate": 5.5641691787276286e-05, + "loss": 0.0244, + "step": 9703 + }, + { + "epoch": 4.566588235294118, + "grad_norm": 0.032308889118487234, + "learning_rate": 5.560761930575864e-05, + "loss": 0.0027, + "step": 9704 + }, + { + "epoch": 4.567058823529412, + "grad_norm": 0.23017536646293582, + "learning_rate": 5.557355488581325e-05, + "loss": 0.0324, + "step": 9705 + }, + { + "epoch": 4.567529411764706, + "grad_norm": 0.08307686970134352, + "learning_rate": 5.553949853034943e-05, + "loss": 0.0206, + "step": 9706 + }, + { + "epoch": 4.568, + "grad_norm": 0.09714573588532585, + "learning_rate": 5.550545024227582e-05, + "loss": 0.026, + "step": 9707 + }, + { + "epoch": 4.568470588235294, + "grad_norm": 0.07626711994985445, + "learning_rate": 5.5471410024500156e-05, + "loss": 0.0087, + "step": 9708 + }, + { + "epoch": 4.568941176470588, + "grad_norm": 0.029178713896439345, + "learning_rate": 5.5437377879929826e-05, + "loss": 0.0034, + "step": 9709 + }, + { + "epoch": 4.569411764705882, + "grad_norm": 0.17647981893941245, + "learning_rate": 5.5403353811471194e-05, + "loss": 0.0807, + "step": 9710 + }, + { + "epoch": 4.5698823529411765, + "grad_norm": 0.09470587184344524, + "learning_rate": 5.5369337822030154e-05, + "loss": 0.0268, + "step": 9711 + }, + { + "epoch": 4.570352941176471, + "grad_norm": 0.0564783222390222, + "learning_rate": 5.533532991451187e-05, + "loss": 0.0127, + "step": 9712 + }, + { + "epoch": 4.570823529411765, + "grad_norm": 0.0547845663532426, + "learning_rate": 5.53013300918207e-05, + "loss": 0.0051, + "step": 9713 + }, + { + "epoch": 4.571294117647059, + "grad_norm": 0.05833443822424075, + "learning_rate": 5.526733835686041e-05, + "loss": 0.0083, + "step": 9714 + }, + { + "epoch": 4.571764705882353, + "grad_norm": 0.1615050057033388, + "learning_rate": 5.523335471253416e-05, + "loss": 0.0463, + "step": 9715 + }, + { + "epoch": 4.572235294117647, + "grad_norm": 0.11332603583559778, + "learning_rate": 5.519937916174419e-05, + "loss": 0.0319, + "step": 9716 + }, + { + "epoch": 4.572705882352941, + "grad_norm": 0.03404530794337465, + "learning_rate": 5.516541170739223e-05, + "loss": 0.0023, + "step": 9717 + }, + { + "epoch": 4.573176470588235, + "grad_norm": 0.14388375242469956, + "learning_rate": 5.5131452352379314e-05, + "loss": 0.0426, + "step": 9718 + }, + { + "epoch": 4.573647058823529, + "grad_norm": 0.02515106381195701, + "learning_rate": 5.509750109960566e-05, + "loss": 0.0029, + "step": 9719 + }, + { + "epoch": 4.574117647058824, + "grad_norm": 0.05347272434165539, + "learning_rate": 5.5063557951970885e-05, + "loss": 0.0066, + "step": 9720 + }, + { + "epoch": 4.574588235294118, + "grad_norm": 0.07114618658592047, + "learning_rate": 5.502962291237399e-05, + "loss": 0.0116, + "step": 9721 + }, + { + "epoch": 4.575058823529412, + "grad_norm": 0.04166530165405972, + "learning_rate": 5.4995695983713076e-05, + "loss": 0.003, + "step": 9722 + }, + { + "epoch": 4.575529411764705, + "grad_norm": 0.0721810190422581, + "learning_rate": 5.496177716888571e-05, + "loss": 0.01, + "step": 9723 + }, + { + "epoch": 4.576, + "grad_norm": 0.15858132095700533, + "learning_rate": 5.4927866470788786e-05, + "loss": 0.0611, + "step": 9724 + }, + { + "epoch": 4.576470588235294, + "grad_norm": 0.05879039939903153, + "learning_rate": 5.489396389231834e-05, + "loss": 0.0053, + "step": 9725 + }, + { + "epoch": 4.576941176470588, + "grad_norm": 0.051506453180203894, + "learning_rate": 5.4860069436369916e-05, + "loss": 0.0041, + "step": 9726 + }, + { + "epoch": 4.577411764705882, + "grad_norm": 0.02486891314003292, + "learning_rate": 5.4826183105838184e-05, + "loss": 0.0028, + "step": 9727 + }, + { + "epoch": 4.5778823529411765, + "grad_norm": 3.4321212012343866e-06, + "learning_rate": 5.4792304903617226e-05, + "loss": 0.0, + "step": 9728 + }, + { + "epoch": 4.578352941176471, + "grad_norm": 0.1278575645337952, + "learning_rate": 5.4758434832600475e-05, + "loss": 0.0302, + "step": 9729 + }, + { + "epoch": 4.578823529411765, + "grad_norm": 0.06034673076153889, + "learning_rate": 5.47245728956805e-05, + "loss": 0.0083, + "step": 9730 + }, + { + "epoch": 4.579294117647059, + "grad_norm": 0.00015516719275417013, + "learning_rate": 5.469071909574933e-05, + "loss": 0.0, + "step": 9731 + }, + { + "epoch": 4.5797647058823525, + "grad_norm": 2.751877966639074e-06, + "learning_rate": 5.465687343569828e-05, + "loss": 0.0, + "step": 9732 + }, + { + "epoch": 4.580235294117647, + "grad_norm": 0.09701273710065464, + "learning_rate": 5.462303591841783e-05, + "loss": 0.0043, + "step": 9733 + }, + { + "epoch": 4.580705882352941, + "grad_norm": 0.10997401755040807, + "learning_rate": 5.458920654679795e-05, + "loss": 0.0399, + "step": 9734 + }, + { + "epoch": 4.581176470588235, + "grad_norm": 0.11227632019108638, + "learning_rate": 5.4555385323727876e-05, + "loss": 0.0264, + "step": 9735 + }, + { + "epoch": 4.581647058823529, + "grad_norm": 0.01611085032695886, + "learning_rate": 5.4521572252095985e-05, + "loss": 0.0009, + "step": 9736 + }, + { + "epoch": 4.582117647058824, + "grad_norm": 1.694164282100649e-05, + "learning_rate": 5.4487767334790145e-05, + "loss": 0.0, + "step": 9737 + }, + { + "epoch": 4.582588235294118, + "grad_norm": 0.1209153282858297, + "learning_rate": 5.445397057469751e-05, + "loss": 0.0321, + "step": 9738 + }, + { + "epoch": 4.583058823529412, + "grad_norm": 0.15471554937147877, + "learning_rate": 5.4420181974704386e-05, + "loss": 0.0402, + "step": 9739 + }, + { + "epoch": 4.583529411764706, + "grad_norm": 1.5265348623909275e-06, + "learning_rate": 5.4386401537696536e-05, + "loss": 0.0, + "step": 9740 + }, + { + "epoch": 4.584, + "grad_norm": 0.14437408975044902, + "learning_rate": 5.435262926655903e-05, + "loss": 0.0535, + "step": 9741 + }, + { + "epoch": 4.584470588235294, + "grad_norm": 0.09212300528289004, + "learning_rate": 5.43188651641761e-05, + "loss": 0.0215, + "step": 9742 + }, + { + "epoch": 4.584941176470588, + "grad_norm": 0.20781446647193855, + "learning_rate": 5.428510923343136e-05, + "loss": 0.0904, + "step": 9743 + }, + { + "epoch": 4.585411764705882, + "grad_norm": 0.16696456264792625, + "learning_rate": 5.425136147720785e-05, + "loss": 0.0564, + "step": 9744 + }, + { + "epoch": 4.5858823529411765, + "grad_norm": 0.0988478561874596, + "learning_rate": 5.421762189838765e-05, + "loss": 0.029, + "step": 9745 + }, + { + "epoch": 4.586352941176471, + "grad_norm": 0.11946976883809271, + "learning_rate": 5.4183890499852364e-05, + "loss": 0.0363, + "step": 9746 + }, + { + "epoch": 4.586823529411765, + "grad_norm": 0.08683242824916294, + "learning_rate": 5.4150167284482844e-05, + "loss": 0.0187, + "step": 9747 + }, + { + "epoch": 4.587294117647059, + "grad_norm": 4.074962523517212e-06, + "learning_rate": 5.4116452255159126e-05, + "loss": 0.0, + "step": 9748 + }, + { + "epoch": 4.587764705882353, + "grad_norm": 0.16159872013992044, + "learning_rate": 5.408274541476075e-05, + "loss": 0.0576, + "step": 9749 + }, + { + "epoch": 4.588235294117647, + "grad_norm": 0.041402649288330304, + "learning_rate": 5.4049046766166335e-05, + "loss": 0.0044, + "step": 9750 + }, + { + "epoch": 4.588705882352941, + "grad_norm": 0.06618898796307224, + "learning_rate": 5.401535631225396e-05, + "loss": 0.0168, + "step": 9751 + }, + { + "epoch": 4.589176470588235, + "grad_norm": 0.15270192003780306, + "learning_rate": 5.398167405590096e-05, + "loss": 0.0489, + "step": 9752 + }, + { + "epoch": 4.589647058823529, + "grad_norm": 9.97089255339545e-06, + "learning_rate": 5.3947999999983967e-05, + "loss": 0.0, + "step": 9753 + }, + { + "epoch": 4.590117647058824, + "grad_norm": 0.21505429855346714, + "learning_rate": 5.3914334147378894e-05, + "loss": 0.0568, + "step": 9754 + }, + { + "epoch": 4.590588235294118, + "grad_norm": 0.13390910666392433, + "learning_rate": 5.388067650096106e-05, + "loss": 0.0193, + "step": 9755 + }, + { + "epoch": 4.591058823529412, + "grad_norm": 0.06684679691896246, + "learning_rate": 5.384702706360483e-05, + "loss": 0.0122, + "step": 9756 + }, + { + "epoch": 4.591529411764705, + "grad_norm": 0.06760903983101495, + "learning_rate": 5.3813385838184185e-05, + "loss": 0.0085, + "step": 9757 + }, + { + "epoch": 4.592, + "grad_norm": 0.05084346752114461, + "learning_rate": 5.3779752827572126e-05, + "loss": 0.0013, + "step": 9758 + }, + { + "epoch": 4.592470588235294, + "grad_norm": 0.12984779831497908, + "learning_rate": 5.374612803464115e-05, + "loss": 0.0184, + "step": 9759 + }, + { + "epoch": 4.592941176470588, + "grad_norm": 0.07623589035206227, + "learning_rate": 5.3712511462262994e-05, + "loss": 0.007, + "step": 9760 + }, + { + "epoch": 4.593411764705882, + "grad_norm": 0.09693655675140322, + "learning_rate": 5.36789031133086e-05, + "loss": 0.022, + "step": 9761 + }, + { + "epoch": 4.5938823529411765, + "grad_norm": 0.08843169351508107, + "learning_rate": 5.364530299064834e-05, + "loss": 0.0246, + "step": 9762 + }, + { + "epoch": 4.594352941176471, + "grad_norm": 0.023982934294246448, + "learning_rate": 5.361171109715186e-05, + "loss": 0.0021, + "step": 9763 + }, + { + "epoch": 4.594823529411765, + "grad_norm": 0.1165400176329958, + "learning_rate": 5.357812743568799e-05, + "loss": 0.0412, + "step": 9764 + }, + { + "epoch": 4.595294117647059, + "grad_norm": 0.0958204002198537, + "learning_rate": 5.354455200912498e-05, + "loss": 0.0238, + "step": 9765 + }, + { + "epoch": 4.5957647058823525, + "grad_norm": 8.101413615338146e-06, + "learning_rate": 5.35109848203304e-05, + "loss": 0.0, + "step": 9766 + }, + { + "epoch": 4.596235294117647, + "grad_norm": 0.1852886341382096, + "learning_rate": 5.3477425872170926e-05, + "loss": 0.0816, + "step": 9767 + }, + { + "epoch": 4.596705882352941, + "grad_norm": 0.10496867328239844, + "learning_rate": 5.3443875167512736e-05, + "loss": 0.0111, + "step": 9768 + }, + { + "epoch": 4.597176470588235, + "grad_norm": 0.014851884548261538, + "learning_rate": 5.341033270922126e-05, + "loss": 0.0006, + "step": 9769 + }, + { + "epoch": 4.597647058823529, + "grad_norm": 0.07749803250382235, + "learning_rate": 5.337679850016109e-05, + "loss": 0.0106, + "step": 9770 + }, + { + "epoch": 4.598117647058824, + "grad_norm": 8.708048827402911e-06, + "learning_rate": 5.3343272543196264e-05, + "loss": 0.0, + "step": 9771 + }, + { + "epoch": 4.598588235294118, + "grad_norm": 3.9689556704895994e-05, + "learning_rate": 5.33097548411901e-05, + "loss": 0.0, + "step": 9772 + }, + { + "epoch": 4.599058823529412, + "grad_norm": 0.10537593066393651, + "learning_rate": 5.3276245397005094e-05, + "loss": 0.0243, + "step": 9773 + }, + { + "epoch": 4.599529411764706, + "grad_norm": 9.318920881841226e-05, + "learning_rate": 5.324274421350316e-05, + "loss": 0.0, + "step": 9774 + }, + { + "epoch": 4.6, + "grad_norm": 0.03187870445080469, + "learning_rate": 5.3209251293545514e-05, + "loss": 0.0007, + "step": 9775 + }, + { + "epoch": 4.600470588235294, + "grad_norm": 1.551884229182888e-05, + "learning_rate": 5.317576663999251e-05, + "loss": 0.0, + "step": 9776 + }, + { + "epoch": 4.600941176470588, + "grad_norm": 0.10490178859977031, + "learning_rate": 5.314229025570401e-05, + "loss": 0.0263, + "step": 9777 + }, + { + "epoch": 4.601411764705882, + "grad_norm": 0.03017038377348614, + "learning_rate": 5.3108822143538944e-05, + "loss": 0.0034, + "step": 9778 + }, + { + "epoch": 4.6018823529411765, + "grad_norm": 0.14743056601972274, + "learning_rate": 5.307536230635571e-05, + "loss": 0.0161, + "step": 9779 + }, + { + "epoch": 4.602352941176471, + "grad_norm": 0.15137921205293914, + "learning_rate": 5.304191074701201e-05, + "loss": 0.0453, + "step": 9780 + }, + { + "epoch": 4.602823529411765, + "grad_norm": 0.08296593156374454, + "learning_rate": 5.3008467468364675e-05, + "loss": 0.0147, + "step": 9781 + }, + { + "epoch": 4.603294117647059, + "grad_norm": 0.11474991894405413, + "learning_rate": 5.297503247326993e-05, + "loss": 0.0274, + "step": 9782 + }, + { + "epoch": 4.603764705882353, + "grad_norm": 0.14662397404112262, + "learning_rate": 5.294160576458339e-05, + "loss": 0.0389, + "step": 9783 + }, + { + "epoch": 4.604235294117647, + "grad_norm": 0.030481340291091695, + "learning_rate": 5.290818734515972e-05, + "loss": 0.0024, + "step": 9784 + }, + { + "epoch": 4.604705882352941, + "grad_norm": 1.7066608738336587e-06, + "learning_rate": 5.287477721785309e-05, + "loss": 0.0, + "step": 9785 + }, + { + "epoch": 4.605176470588235, + "grad_norm": 0.1331310323661661, + "learning_rate": 5.284137538551693e-05, + "loss": 0.0438, + "step": 9786 + }, + { + "epoch": 4.605647058823529, + "grad_norm": 0.08806973816794865, + "learning_rate": 5.2807981851003825e-05, + "loss": 0.0064, + "step": 9787 + }, + { + "epoch": 4.606117647058824, + "grad_norm": 0.13341022331496338, + "learning_rate": 5.2774596617165793e-05, + "loss": 0.0371, + "step": 9788 + }, + { + "epoch": 4.606588235294118, + "grad_norm": 0.03845121397994253, + "learning_rate": 5.2741219686854146e-05, + "loss": 0.0054, + "step": 9789 + }, + { + "epoch": 4.607058823529412, + "grad_norm": 0.12328723775477851, + "learning_rate": 5.270785106291934e-05, + "loss": 0.046, + "step": 9790 + }, + { + "epoch": 4.607529411764705, + "grad_norm": 0.00034628140760293387, + "learning_rate": 5.2674490748211265e-05, + "loss": 0.0, + "step": 9791 + }, + { + "epoch": 4.608, + "grad_norm": 0.16308979135585705, + "learning_rate": 5.2641138745579126e-05, + "loss": 0.0603, + "step": 9792 + }, + { + "epoch": 4.608470588235294, + "grad_norm": 6.8022441014814e-05, + "learning_rate": 5.2607795057871216e-05, + "loss": 0.0, + "step": 9793 + }, + { + "epoch": 4.608941176470588, + "grad_norm": 0.05142693131115921, + "learning_rate": 5.257445968793531e-05, + "loss": 0.0074, + "step": 9794 + }, + { + "epoch": 4.609411764705882, + "grad_norm": 0.06436486441865576, + "learning_rate": 5.254113263861848e-05, + "loss": 0.0123, + "step": 9795 + }, + { + "epoch": 4.6098823529411765, + "grad_norm": 0.09909879514295118, + "learning_rate": 5.250781391276684e-05, + "loss": 0.0189, + "step": 9796 + }, + { + "epoch": 4.610352941176471, + "grad_norm": 0.13453245169644001, + "learning_rate": 5.24745035132262e-05, + "loss": 0.0296, + "step": 9797 + }, + { + "epoch": 4.610823529411765, + "grad_norm": 0.06806931003750871, + "learning_rate": 5.2441201442841264e-05, + "loss": 0.0106, + "step": 9798 + }, + { + "epoch": 4.611294117647059, + "grad_norm": 0.12044138361771685, + "learning_rate": 5.240790770445625e-05, + "loss": 0.0215, + "step": 9799 + }, + { + "epoch": 4.6117647058823525, + "grad_norm": 0.12004236922852199, + "learning_rate": 5.237462230091467e-05, + "loss": 0.0347, + "step": 9800 + }, + { + "epoch": 4.612235294117647, + "grad_norm": 0.0001234303222874255, + "learning_rate": 5.234134523505912e-05, + "loss": 0.0, + "step": 9801 + }, + { + "epoch": 4.612705882352941, + "grad_norm": 0.12181692135961641, + "learning_rate": 5.230807650973172e-05, + "loss": 0.0265, + "step": 9802 + }, + { + "epoch": 4.613176470588235, + "grad_norm": 0.16205505892709943, + "learning_rate": 5.227481612777379e-05, + "loss": 0.0195, + "step": 9803 + }, + { + "epoch": 4.613647058823529, + "grad_norm": 0.11398699215985818, + "learning_rate": 5.224156409202587e-05, + "loss": 0.026, + "step": 9804 + }, + { + "epoch": 4.614117647058824, + "grad_norm": 7.2915278128747784e-06, + "learning_rate": 5.2208320405327865e-05, + "loss": 0.0, + "step": 9805 + }, + { + "epoch": 4.614588235294118, + "grad_norm": 0.1308671377064916, + "learning_rate": 5.2175085070519024e-05, + "loss": 0.0301, + "step": 9806 + }, + { + "epoch": 4.615058823529412, + "grad_norm": 0.12810828786637207, + "learning_rate": 5.2141858090437696e-05, + "loss": 0.0486, + "step": 9807 + }, + { + "epoch": 4.615529411764706, + "grad_norm": 1.5791697561267258e-05, + "learning_rate": 5.210863946792172e-05, + "loss": 0.0, + "step": 9808 + }, + { + "epoch": 4.616, + "grad_norm": 0.03008247294017937, + "learning_rate": 5.207542920580805e-05, + "loss": 0.0024, + "step": 9809 + }, + { + "epoch": 4.616470588235294, + "grad_norm": 0.12008931324557276, + "learning_rate": 5.204222730693303e-05, + "loss": 0.0473, + "step": 9810 + }, + { + "epoch": 4.616941176470588, + "grad_norm": 3.191200696071812e-05, + "learning_rate": 5.2009033774132345e-05, + "loss": 0.0, + "step": 9811 + }, + { + "epoch": 4.617411764705882, + "grad_norm": 0.10832720894962206, + "learning_rate": 5.197584861024076e-05, + "loss": 0.0349, + "step": 9812 + }, + { + "epoch": 4.6178823529411765, + "grad_norm": 0.3551537910528631, + "learning_rate": 5.194267181809251e-05, + "loss": 0.1137, + "step": 9813 + }, + { + "epoch": 4.618352941176471, + "grad_norm": 0.03677575749514487, + "learning_rate": 5.1909503400521104e-05, + "loss": 0.0022, + "step": 9814 + }, + { + "epoch": 4.618823529411765, + "grad_norm": 1.3392501506846057e-06, + "learning_rate": 5.187634336035918e-05, + "loss": 0.0, + "step": 9815 + }, + { + "epoch": 4.619294117647058, + "grad_norm": 0.09196231863948208, + "learning_rate": 5.184319170043884e-05, + "loss": 0.0207, + "step": 9816 + }, + { + "epoch": 4.619764705882353, + "grad_norm": 0.09186440829552336, + "learning_rate": 5.1810048423591424e-05, + "loss": 0.0149, + "step": 9817 + }, + { + "epoch": 4.620235294117647, + "grad_norm": 7.864471756527968e-06, + "learning_rate": 5.177691353264746e-05, + "loss": 0.0, + "step": 9818 + }, + { + "epoch": 4.620705882352941, + "grad_norm": 0.04787142648177645, + "learning_rate": 5.174378703043684e-05, + "loss": 0.0062, + "step": 9819 + }, + { + "epoch": 4.621176470588235, + "grad_norm": 0.08210361055044557, + "learning_rate": 5.171066891978881e-05, + "loss": 0.0151, + "step": 9820 + }, + { + "epoch": 4.621647058823529, + "grad_norm": 0.06443051205981484, + "learning_rate": 5.167755920353171e-05, + "loss": 0.0017, + "step": 9821 + }, + { + "epoch": 4.622117647058824, + "grad_norm": 0.1568526613298764, + "learning_rate": 5.164445788449334e-05, + "loss": 0.0236, + "step": 9822 + }, + { + "epoch": 4.622588235294118, + "grad_norm": 3.0113132527179885e-05, + "learning_rate": 5.161136496550072e-05, + "loss": 0.0, + "step": 9823 + }, + { + "epoch": 4.623058823529412, + "grad_norm": 0.051617738020450034, + "learning_rate": 5.15782804493801e-05, + "loss": 0.0076, + "step": 9824 + }, + { + "epoch": 4.623529411764705, + "grad_norm": 0.04312595138679415, + "learning_rate": 5.1545204338957066e-05, + "loss": 0.0038, + "step": 9825 + }, + { + "epoch": 4.624, + "grad_norm": 0.09007523897295944, + "learning_rate": 5.151213663705655e-05, + "loss": 0.0219, + "step": 9826 + }, + { + "epoch": 4.624470588235294, + "grad_norm": 0.04238979131619688, + "learning_rate": 5.14790773465026e-05, + "loss": 0.0058, + "step": 9827 + }, + { + "epoch": 4.624941176470588, + "grad_norm": 0.1777255428487514, + "learning_rate": 5.144602647011874e-05, + "loss": 0.022, + "step": 9828 + }, + { + "epoch": 4.625411764705882, + "grad_norm": 0.12518830801451789, + "learning_rate": 5.141298401072758e-05, + "loss": 0.0311, + "step": 9829 + }, + { + "epoch": 4.625882352941177, + "grad_norm": 0.16772337553099653, + "learning_rate": 5.137994997115115e-05, + "loss": 0.0438, + "step": 9830 + }, + { + "epoch": 4.626352941176471, + "grad_norm": 0.14762715968990783, + "learning_rate": 5.134692435421077e-05, + "loss": 0.0454, + "step": 9831 + }, + { + "epoch": 4.626823529411765, + "grad_norm": 0.08265158716960977, + "learning_rate": 5.1313907162726883e-05, + "loss": 0.0132, + "step": 9832 + }, + { + "epoch": 4.627294117647059, + "grad_norm": 0.17835228096890418, + "learning_rate": 5.128089839951937e-05, + "loss": 0.0629, + "step": 9833 + }, + { + "epoch": 4.627764705882353, + "grad_norm": 0.08704892734282163, + "learning_rate": 5.124789806740742e-05, + "loss": 0.0064, + "step": 9834 + }, + { + "epoch": 4.628235294117647, + "grad_norm": 0.04607573374438899, + "learning_rate": 5.12149061692093e-05, + "loss": 0.0023, + "step": 9835 + }, + { + "epoch": 4.628705882352941, + "grad_norm": 0.0001316408138069647, + "learning_rate": 5.118192270774271e-05, + "loss": 0.0, + "step": 9836 + }, + { + "epoch": 4.629176470588235, + "grad_norm": 0.20030786660503486, + "learning_rate": 5.114894768582468e-05, + "loss": 0.0625, + "step": 9837 + }, + { + "epoch": 4.6296470588235294, + "grad_norm": 0.058111427339389184, + "learning_rate": 5.111598110627133e-05, + "loss": 0.011, + "step": 9838 + }, + { + "epoch": 4.630117647058824, + "grad_norm": 0.08877743422273514, + "learning_rate": 5.108302297189822e-05, + "loss": 0.0092, + "step": 9839 + }, + { + "epoch": 4.630588235294118, + "grad_norm": 0.019954033798645396, + "learning_rate": 5.105007328552019e-05, + "loss": 0.0018, + "step": 9840 + }, + { + "epoch": 4.631058823529412, + "grad_norm": 0.1773907900965248, + "learning_rate": 5.101713204995114e-05, + "loss": 0.063, + "step": 9841 + }, + { + "epoch": 4.631529411764706, + "grad_norm": 0.10820722937100888, + "learning_rate": 5.0984199268004615e-05, + "loss": 0.0321, + "step": 9842 + }, + { + "epoch": 4.632, + "grad_norm": 0.10334381727853612, + "learning_rate": 5.095127494249309e-05, + "loss": 0.0263, + "step": 9843 + }, + { + "epoch": 4.632470588235294, + "grad_norm": 0.03710566869674459, + "learning_rate": 5.091835907622854e-05, + "loss": 0.0028, + "step": 9844 + }, + { + "epoch": 4.632941176470588, + "grad_norm": 0.08083236652869041, + "learning_rate": 5.0885451672022156e-05, + "loss": 0.0087, + "step": 9845 + }, + { + "epoch": 4.633411764705882, + "grad_norm": 0.09470352498197468, + "learning_rate": 5.0852552732684296e-05, + "loss": 0.0226, + "step": 9846 + }, + { + "epoch": 4.633882352941177, + "grad_norm": 0.08872566662807845, + "learning_rate": 5.081966226102474e-05, + "loss": 0.0208, + "step": 9847 + }, + { + "epoch": 4.634352941176471, + "grad_norm": 6.0829672367962425e-06, + "learning_rate": 5.078678025985257e-05, + "loss": 0.0, + "step": 9848 + }, + { + "epoch": 4.634823529411765, + "grad_norm": 0.016012615822739495, + "learning_rate": 5.0753906731975954e-05, + "loss": 0.0011, + "step": 9849 + }, + { + "epoch": 4.635294117647058, + "grad_norm": 0.09886887143170658, + "learning_rate": 5.07210416802025e-05, + "loss": 0.0137, + "step": 9850 + }, + { + "epoch": 4.635764705882353, + "grad_norm": 0.02995491621003326, + "learning_rate": 5.0688185107339084e-05, + "loss": 0.0029, + "step": 9851 + }, + { + "epoch": 4.636235294117647, + "grad_norm": 6.315161710475152e-07, + "learning_rate": 5.065533701619175e-05, + "loss": 0.0, + "step": 9852 + }, + { + "epoch": 4.636705882352941, + "grad_norm": 0.08696540872633361, + "learning_rate": 5.06224974095659e-05, + "loss": 0.012, + "step": 9853 + }, + { + "epoch": 4.637176470588235, + "grad_norm": 0.1293148999047516, + "learning_rate": 5.058966629026629e-05, + "loss": 0.0429, + "step": 9854 + }, + { + "epoch": 4.6376470588235295, + "grad_norm": 0.05837428028683645, + "learning_rate": 5.0556843661096714e-05, + "loss": 0.0087, + "step": 9855 + }, + { + "epoch": 4.638117647058824, + "grad_norm": 0.09667950443114202, + "learning_rate": 5.052402952486046e-05, + "loss": 0.0121, + "step": 9856 + }, + { + "epoch": 4.638588235294118, + "grad_norm": 0.047982810267086895, + "learning_rate": 5.0491223884360054e-05, + "loss": 0.0066, + "step": 9857 + }, + { + "epoch": 4.639058823529412, + "grad_norm": 1.372499255948439e-06, + "learning_rate": 5.0458426742397174e-05, + "loss": 0.0, + "step": 9858 + }, + { + "epoch": 4.6395294117647055, + "grad_norm": 0.18400322649341466, + "learning_rate": 5.042563810177296e-05, + "loss": 0.0753, + "step": 9859 + }, + { + "epoch": 4.64, + "grad_norm": 0.1123061151589766, + "learning_rate": 5.039285796528759e-05, + "loss": 0.0342, + "step": 9860 + }, + { + "epoch": 4.640470588235294, + "grad_norm": 0.10493968716412566, + "learning_rate": 5.0360086335740734e-05, + "loss": 0.0236, + "step": 9861 + }, + { + "epoch": 4.640941176470588, + "grad_norm": 0.13032972371234258, + "learning_rate": 5.032732321593128e-05, + "loss": 0.0381, + "step": 9862 + }, + { + "epoch": 4.641411764705882, + "grad_norm": 0.08593598613668374, + "learning_rate": 5.029456860865727e-05, + "loss": 0.0165, + "step": 9863 + }, + { + "epoch": 4.641882352941177, + "grad_norm": 0.019328571573791174, + "learning_rate": 5.0261822516716146e-05, + "loss": 0.0016, + "step": 9864 + }, + { + "epoch": 4.642352941176471, + "grad_norm": 4.792589161633424e-06, + "learning_rate": 5.022908494290465e-05, + "loss": 0.0, + "step": 9865 + }, + { + "epoch": 4.642823529411765, + "grad_norm": 0.14843045303360375, + "learning_rate": 5.0196355890018605e-05, + "loss": 0.0351, + "step": 9866 + }, + { + "epoch": 4.643294117647059, + "grad_norm": 0.1272251680475537, + "learning_rate": 5.016363536085333e-05, + "loss": 0.0321, + "step": 9867 + }, + { + "epoch": 4.643764705882353, + "grad_norm": 6.273541716142514e-07, + "learning_rate": 5.0130923358203316e-05, + "loss": 0.0, + "step": 9868 + }, + { + "epoch": 4.644235294117647, + "grad_norm": 0.1535246055725065, + "learning_rate": 5.0098219884862265e-05, + "loss": 0.0286, + "step": 9869 + }, + { + "epoch": 4.644705882352941, + "grad_norm": 0.053031005337130024, + "learning_rate": 5.006552494362325e-05, + "loss": 0.0071, + "step": 9870 + }, + { + "epoch": 4.645176470588235, + "grad_norm": 0.14449553019355407, + "learning_rate": 5.0032838537278627e-05, + "loss": 0.0619, + "step": 9871 + }, + { + "epoch": 4.6456470588235295, + "grad_norm": 1.5837453727904033e-06, + "learning_rate": 5.000016066861988e-05, + "loss": 0.0, + "step": 9872 + }, + { + "epoch": 4.646117647058824, + "grad_norm": 1.2741536676813291e-06, + "learning_rate": 4.9967491340437904e-05, + "loss": 0.0, + "step": 9873 + }, + { + "epoch": 4.646588235294118, + "grad_norm": 1.131880440202529e-06, + "learning_rate": 4.993483055552289e-05, + "loss": 0.0, + "step": 9874 + }, + { + "epoch": 4.647058823529412, + "grad_norm": 1.3234738153407407e-06, + "learning_rate": 4.990217831666409e-05, + "loss": 0.0, + "step": 9875 + }, + { + "epoch": 4.647529411764706, + "grad_norm": 0.09058092500629414, + "learning_rate": 4.986953462665026e-05, + "loss": 0.0157, + "step": 9876 + }, + { + "epoch": 4.648, + "grad_norm": 0.1269035177379307, + "learning_rate": 4.9836899488269345e-05, + "loss": 0.0258, + "step": 9877 + }, + { + "epoch": 4.648470588235294, + "grad_norm": 0.04873219997671769, + "learning_rate": 4.9804272904308465e-05, + "loss": 0.0064, + "step": 9878 + }, + { + "epoch": 4.648941176470588, + "grad_norm": 0.08157886829335065, + "learning_rate": 4.97716548775542e-05, + "loss": 0.0125, + "step": 9879 + }, + { + "epoch": 4.649411764705882, + "grad_norm": 0.01904911223033895, + "learning_rate": 4.9739045410792164e-05, + "loss": 0.001, + "step": 9880 + }, + { + "epoch": 4.649882352941177, + "grad_norm": 0.12515213910921233, + "learning_rate": 4.970644450680742e-05, + "loss": 0.0322, + "step": 9881 + }, + { + "epoch": 4.650352941176471, + "grad_norm": 0.1615936302483777, + "learning_rate": 4.9673852168384316e-05, + "loss": 0.0526, + "step": 9882 + }, + { + "epoch": 4.650823529411765, + "grad_norm": 8.346486952541276e-06, + "learning_rate": 4.964126839830628e-05, + "loss": 0.0, + "step": 9883 + }, + { + "epoch": 4.651294117647058, + "grad_norm": 0.1749742758774735, + "learning_rate": 4.960869319935616e-05, + "loss": 0.0663, + "step": 9884 + }, + { + "epoch": 4.651764705882353, + "grad_norm": 0.14092903550309213, + "learning_rate": 4.9576126574316086e-05, + "loss": 0.0378, + "step": 9885 + }, + { + "epoch": 4.652235294117647, + "grad_norm": 0.048648996346475924, + "learning_rate": 4.9543568525967376e-05, + "loss": 0.0056, + "step": 9886 + }, + { + "epoch": 4.652705882352941, + "grad_norm": 0.13972165261042918, + "learning_rate": 4.9511019057090684e-05, + "loss": 0.0543, + "step": 9887 + }, + { + "epoch": 4.653176470588235, + "grad_norm": 0.0946181171206674, + "learning_rate": 4.947847817046582e-05, + "loss": 0.0195, + "step": 9888 + }, + { + "epoch": 4.6536470588235295, + "grad_norm": 0.11549777520526278, + "learning_rate": 4.9445945868871974e-05, + "loss": 0.0417, + "step": 9889 + }, + { + "epoch": 4.654117647058824, + "grad_norm": 0.09141099013214266, + "learning_rate": 4.941342215508761e-05, + "loss": 0.0187, + "step": 9890 + }, + { + "epoch": 4.654588235294118, + "grad_norm": 0.09930042415694892, + "learning_rate": 4.9380907031890315e-05, + "loss": 0.0088, + "step": 9891 + }, + { + "epoch": 4.655058823529412, + "grad_norm": 0.07330471894383706, + "learning_rate": 4.934840050205711e-05, + "loss": 0.0162, + "step": 9892 + }, + { + "epoch": 4.6555294117647055, + "grad_norm": 0.1218707284744377, + "learning_rate": 4.931590256836424e-05, + "loss": 0.0636, + "step": 9893 + }, + { + "epoch": 4.656, + "grad_norm": 3.3200698137363057e-06, + "learning_rate": 4.928341323358708e-05, + "loss": 0.0, + "step": 9894 + }, + { + "epoch": 4.656470588235294, + "grad_norm": 0.06868193291771586, + "learning_rate": 4.925093250050045e-05, + "loss": 0.0097, + "step": 9895 + }, + { + "epoch": 4.656941176470588, + "grad_norm": 0.08935226773755678, + "learning_rate": 4.921846037187841e-05, + "loss": 0.0172, + "step": 9896 + }, + { + "epoch": 4.657411764705882, + "grad_norm": 0.07987509964944874, + "learning_rate": 4.918599685049414e-05, + "loss": 0.0173, + "step": 9897 + }, + { + "epoch": 4.657882352941177, + "grad_norm": 0.13531795518285952, + "learning_rate": 4.915354193912021e-05, + "loss": 0.046, + "step": 9898 + }, + { + "epoch": 4.658352941176471, + "grad_norm": 0.2923445866734368, + "learning_rate": 4.9121095640528516e-05, + "loss": 0.0611, + "step": 9899 + }, + { + "epoch": 4.658823529411765, + "grad_norm": 0.10234418612072768, + "learning_rate": 4.908865795748999e-05, + "loss": 0.0208, + "step": 9900 + }, + { + "epoch": 4.659294117647059, + "grad_norm": 2.577118533562696e-06, + "learning_rate": 4.905622889277506e-05, + "loss": 0.0, + "step": 9901 + }, + { + "epoch": 4.659764705882353, + "grad_norm": 0.00016229947399534636, + "learning_rate": 4.902380844915337e-05, + "loss": 0.0, + "step": 9902 + }, + { + "epoch": 4.660235294117647, + "grad_norm": 0.1696301858620285, + "learning_rate": 4.899139662939366e-05, + "loss": 0.0549, + "step": 9903 + }, + { + "epoch": 4.660705882352941, + "grad_norm": 1.3921399931415776e-05, + "learning_rate": 4.895899343626412e-05, + "loss": 0.0, + "step": 9904 + }, + { + "epoch": 4.661176470588235, + "grad_norm": 0.1694223317240272, + "learning_rate": 4.892659887253219e-05, + "loss": 0.0575, + "step": 9905 + }, + { + "epoch": 4.6616470588235295, + "grad_norm": 0.09563487663089495, + "learning_rate": 4.8894212940964445e-05, + "loss": 0.026, + "step": 9906 + }, + { + "epoch": 4.662117647058824, + "grad_norm": 3.3990576619861956e-06, + "learning_rate": 4.8861835644326866e-05, + "loss": 0.0, + "step": 9907 + }, + { + "epoch": 4.662588235294118, + "grad_norm": 0.06948083529379868, + "learning_rate": 4.882946698538457e-05, + "loss": 0.0134, + "step": 9908 + }, + { + "epoch": 4.663058823529412, + "grad_norm": 0.07742327391424808, + "learning_rate": 4.879710696690202e-05, + "loss": 0.0162, + "step": 9909 + }, + { + "epoch": 4.663529411764706, + "grad_norm": 3.716860669033278e-05, + "learning_rate": 4.876475559164299e-05, + "loss": 0.0, + "step": 9910 + }, + { + "epoch": 4.664, + "grad_norm": 0.1631146378897924, + "learning_rate": 4.873241286237033e-05, + "loss": 0.0404, + "step": 9911 + }, + { + "epoch": 4.664470588235294, + "grad_norm": 0.09782018367929379, + "learning_rate": 4.8700078781846326e-05, + "loss": 0.0173, + "step": 9912 + }, + { + "epoch": 4.664941176470588, + "grad_norm": 0.07203217012006004, + "learning_rate": 4.866775335283253e-05, + "loss": 0.0211, + "step": 9913 + }, + { + "epoch": 4.665411764705882, + "grad_norm": 0.04665083448860009, + "learning_rate": 4.863543657808956e-05, + "loss": 0.0028, + "step": 9914 + }, + { + "epoch": 4.665882352941177, + "grad_norm": 0.11088345899049805, + "learning_rate": 4.860312846037753e-05, + "loss": 0.021, + "step": 9915 + }, + { + "epoch": 4.666352941176471, + "grad_norm": 0.13345590952077577, + "learning_rate": 4.857082900245569e-05, + "loss": 0.0335, + "step": 9916 + }, + { + "epoch": 4.666823529411765, + "grad_norm": 0.10341092371277093, + "learning_rate": 4.8538538207082534e-05, + "loss": 0.0324, + "step": 9917 + }, + { + "epoch": 4.667294117647058, + "grad_norm": 6.425965496142988e-06, + "learning_rate": 4.850625607701588e-05, + "loss": 0.0, + "step": 9918 + }, + { + "epoch": 4.667764705882353, + "grad_norm": 0.1335969012790192, + "learning_rate": 4.8473982615012816e-05, + "loss": 0.0473, + "step": 9919 + }, + { + "epoch": 4.668235294117647, + "grad_norm": 0.1036121273417736, + "learning_rate": 4.844171782382959e-05, + "loss": 0.0276, + "step": 9920 + }, + { + "epoch": 4.668705882352941, + "grad_norm": 6.076165759889563e-06, + "learning_rate": 4.840946170622179e-05, + "loss": 0.0, + "step": 9921 + }, + { + "epoch": 4.669176470588235, + "grad_norm": 0.16369746628674123, + "learning_rate": 4.8377214264944296e-05, + "loss": 0.046, + "step": 9922 + }, + { + "epoch": 4.6696470588235295, + "grad_norm": 0.08935614957053781, + "learning_rate": 4.834497550275114e-05, + "loss": 0.0116, + "step": 9923 + }, + { + "epoch": 4.670117647058824, + "grad_norm": 0.09907792257385437, + "learning_rate": 4.831274542239566e-05, + "loss": 0.0287, + "step": 9924 + }, + { + "epoch": 4.670588235294118, + "grad_norm": 0.10694787551263514, + "learning_rate": 4.8280524026630565e-05, + "loss": 0.0088, + "step": 9925 + }, + { + "epoch": 4.671058823529412, + "grad_norm": 0.20490496343151607, + "learning_rate": 4.824831131820759e-05, + "loss": 0.0779, + "step": 9926 + }, + { + "epoch": 4.6715294117647055, + "grad_norm": 0.18017301776844943, + "learning_rate": 4.821610729987792e-05, + "loss": 0.0681, + "step": 9927 + }, + { + "epoch": 4.672, + "grad_norm": 1.9819450832726404e-05, + "learning_rate": 4.818391197439199e-05, + "loss": 0.0, + "step": 9928 + }, + { + "epoch": 4.672470588235294, + "grad_norm": 0.09165405063064147, + "learning_rate": 4.815172534449933e-05, + "loss": 0.0142, + "step": 9929 + }, + { + "epoch": 4.672941176470588, + "grad_norm": 2.1614796953680702e-06, + "learning_rate": 4.811954741294889e-05, + "loss": 0.0, + "step": 9930 + }, + { + "epoch": 4.673411764705882, + "grad_norm": 0.04948001449910844, + "learning_rate": 4.808737818248882e-05, + "loss": 0.005, + "step": 9931 + }, + { + "epoch": 4.673882352941177, + "grad_norm": 0.08891358348957043, + "learning_rate": 4.805521765586653e-05, + "loss": 0.0107, + "step": 9932 + }, + { + "epoch": 4.674352941176471, + "grad_norm": 0.13565688491143396, + "learning_rate": 4.802306583582873e-05, + "loss": 0.0481, + "step": 9933 + }, + { + "epoch": 4.674823529411765, + "grad_norm": 0.0986000829190239, + "learning_rate": 4.7990922725121266e-05, + "loss": 0.0327, + "step": 9934 + }, + { + "epoch": 4.675294117647059, + "grad_norm": 3.7769891668451003e-06, + "learning_rate": 4.795878832648935e-05, + "loss": 0.0, + "step": 9935 + }, + { + "epoch": 4.675764705882353, + "grad_norm": 0.17454691448183238, + "learning_rate": 4.7926662642677466e-05, + "loss": 0.0472, + "step": 9936 + }, + { + "epoch": 4.676235294117647, + "grad_norm": 0.13626183358354235, + "learning_rate": 4.7894545676429206e-05, + "loss": 0.0228, + "step": 9937 + }, + { + "epoch": 4.676705882352941, + "grad_norm": 0.0002558724391955357, + "learning_rate": 4.7862437430487635e-05, + "loss": 0.0, + "step": 9938 + }, + { + "epoch": 4.677176470588235, + "grad_norm": 0.034978006093781044, + "learning_rate": 4.783033790759481e-05, + "loss": 0.0019, + "step": 9939 + }, + { + "epoch": 4.6776470588235295, + "grad_norm": 0.0625097972181991, + "learning_rate": 4.77982471104923e-05, + "loss": 0.0111, + "step": 9940 + }, + { + "epoch": 4.678117647058824, + "grad_norm": 0.14562708778422545, + "learning_rate": 4.776616504192082e-05, + "loss": 0.0297, + "step": 9941 + }, + { + "epoch": 4.678588235294118, + "grad_norm": 0.13756675515755282, + "learning_rate": 4.7734091704620244e-05, + "loss": 0.0467, + "step": 9942 + }, + { + "epoch": 4.679058823529411, + "grad_norm": 0.025683697345930043, + "learning_rate": 4.7702027101329846e-05, + "loss": 0.0016, + "step": 9943 + }, + { + "epoch": 4.679529411764706, + "grad_norm": 0.10624084810089525, + "learning_rate": 4.7669971234788165e-05, + "loss": 0.0226, + "step": 9944 + }, + { + "epoch": 4.68, + "grad_norm": 0.1171123087213518, + "learning_rate": 4.763792410773281e-05, + "loss": 0.0172, + "step": 9945 + }, + { + "epoch": 4.680470588235294, + "grad_norm": 0.16953291489879102, + "learning_rate": 4.760588572290082e-05, + "loss": 0.0311, + "step": 9946 + }, + { + "epoch": 4.680941176470588, + "grad_norm": 0.021197838632752472, + "learning_rate": 4.7573856083028475e-05, + "loss": 0.0014, + "step": 9947 + }, + { + "epoch": 4.681411764705882, + "grad_norm": 0.09981261421363548, + "learning_rate": 4.7541835190851175e-05, + "loss": 0.0156, + "step": 9948 + }, + { + "epoch": 4.681882352941177, + "grad_norm": 0.021021723337502386, + "learning_rate": 4.7509823049103706e-05, + "loss": 0.0008, + "step": 9949 + }, + { + "epoch": 4.682352941176471, + "grad_norm": 0.029571987589860467, + "learning_rate": 4.74778196605201e-05, + "loss": 0.0024, + "step": 9950 + }, + { + "epoch": 4.682823529411765, + "grad_norm": 0.025808782409261148, + "learning_rate": 4.744582502783352e-05, + "loss": 0.0017, + "step": 9951 + }, + { + "epoch": 4.683294117647058, + "grad_norm": 0.23086928510617466, + "learning_rate": 4.7413839153776504e-05, + "loss": 0.0585, + "step": 9952 + }, + { + "epoch": 4.683764705882353, + "grad_norm": 0.06465988426465386, + "learning_rate": 4.7381862041080865e-05, + "loss": 0.0062, + "step": 9953 + }, + { + "epoch": 4.684235294117647, + "grad_norm": 0.06316103465787067, + "learning_rate": 4.7349893692477495e-05, + "loss": 0.0122, + "step": 9954 + }, + { + "epoch": 4.684705882352941, + "grad_norm": 0.14674796462108863, + "learning_rate": 4.7317934110696685e-05, + "loss": 0.0301, + "step": 9955 + }, + { + "epoch": 4.685176470588235, + "grad_norm": 9.075675925391688e-06, + "learning_rate": 4.728598329846802e-05, + "loss": 0.0, + "step": 9956 + }, + { + "epoch": 4.6856470588235295, + "grad_norm": 4.740772258741246e-06, + "learning_rate": 4.725404125852013e-05, + "loss": 0.0, + "step": 9957 + }, + { + "epoch": 4.686117647058824, + "grad_norm": 0.08686850107124074, + "learning_rate": 4.7222107993581135e-05, + "loss": 0.0186, + "step": 9958 + }, + { + "epoch": 4.686588235294118, + "grad_norm": 0.16299473207613874, + "learning_rate": 4.7190183506378205e-05, + "loss": 0.0645, + "step": 9959 + }, + { + "epoch": 4.687058823529412, + "grad_norm": 0.23523110137458164, + "learning_rate": 4.7158267799637864e-05, + "loss": 0.065, + "step": 9960 + }, + { + "epoch": 4.6875294117647055, + "grad_norm": 0.14348768635567835, + "learning_rate": 4.712636087608594e-05, + "loss": 0.026, + "step": 9961 + }, + { + "epoch": 4.688, + "grad_norm": 0.07190520032420604, + "learning_rate": 4.709446273844736e-05, + "loss": 0.0052, + "step": 9962 + }, + { + "epoch": 4.688470588235294, + "grad_norm": 0.025263665087531147, + "learning_rate": 4.7062573389446395e-05, + "loss": 0.002, + "step": 9963 + }, + { + "epoch": 4.688941176470588, + "grad_norm": 0.036786942282050296, + "learning_rate": 4.70306928318066e-05, + "loss": 0.0044, + "step": 9964 + }, + { + "epoch": 4.689411764705882, + "grad_norm": 0.03375584545840104, + "learning_rate": 4.6998821068250675e-05, + "loss": 0.0028, + "step": 9965 + }, + { + "epoch": 4.689882352941177, + "grad_norm": 0.09885786317507343, + "learning_rate": 4.696695810150065e-05, + "loss": 0.0183, + "step": 9966 + }, + { + "epoch": 4.690352941176471, + "grad_norm": 0.09596101295546779, + "learning_rate": 4.693510393427781e-05, + "loss": 0.0135, + "step": 9967 + }, + { + "epoch": 4.690823529411765, + "grad_norm": 0.0678746323425108, + "learning_rate": 4.690325856930258e-05, + "loss": 0.014, + "step": 9968 + }, + { + "epoch": 4.691294117647059, + "grad_norm": 0.04163230018742749, + "learning_rate": 4.6871422009294764e-05, + "loss": 0.0056, + "step": 9969 + }, + { + "epoch": 4.691764705882353, + "grad_norm": 0.12514302798357113, + "learning_rate": 4.6839594256973406e-05, + "loss": 0.0171, + "step": 9970 + }, + { + "epoch": 4.692235294117647, + "grad_norm": 0.2818295005950447, + "learning_rate": 4.680777531505662e-05, + "loss": 0.0881, + "step": 9971 + }, + { + "epoch": 4.692705882352941, + "grad_norm": 6.446284515218135e-05, + "learning_rate": 4.6775965186262e-05, + "loss": 0.0, + "step": 9972 + }, + { + "epoch": 4.693176470588235, + "grad_norm": 0.1446040628037964, + "learning_rate": 4.6744163873306314e-05, + "loss": 0.0399, + "step": 9973 + }, + { + "epoch": 4.6936470588235295, + "grad_norm": 0.11633548810698956, + "learning_rate": 4.671237137890539e-05, + "loss": 0.0212, + "step": 9974 + }, + { + "epoch": 4.694117647058824, + "grad_norm": 0.08610314730574835, + "learning_rate": 4.668058770577468e-05, + "loss": 0.0254, + "step": 9975 + }, + { + "epoch": 4.694588235294118, + "grad_norm": 0.05819916580624722, + "learning_rate": 4.6648812856628515e-05, + "loss": 0.0043, + "step": 9976 + }, + { + "epoch": 4.695058823529411, + "grad_norm": 0.16694903420282542, + "learning_rate": 4.6617046834180656e-05, + "loss": 0.066, + "step": 9977 + }, + { + "epoch": 4.6955294117647055, + "grad_norm": 0.2231180695609724, + "learning_rate": 4.658528964114413e-05, + "loss": 0.0669, + "step": 9978 + }, + { + "epoch": 4.696, + "grad_norm": 0.116424989228766, + "learning_rate": 4.655354128023107e-05, + "loss": 0.0382, + "step": 9979 + }, + { + "epoch": 4.696470588235294, + "grad_norm": 0.11541581243161966, + "learning_rate": 4.6521801754152995e-05, + "loss": 0.0157, + "step": 9980 + }, + { + "epoch": 4.696941176470588, + "grad_norm": 5.349071918579716e-06, + "learning_rate": 4.649007106562066e-05, + "loss": 0.0, + "step": 9981 + }, + { + "epoch": 4.697411764705882, + "grad_norm": 0.17648722116389517, + "learning_rate": 4.645834921734392e-05, + "loss": 0.0487, + "step": 9982 + }, + { + "epoch": 4.697882352941177, + "grad_norm": 0.07964782156226444, + "learning_rate": 4.642663621203201e-05, + "loss": 0.018, + "step": 9983 + }, + { + "epoch": 4.698352941176471, + "grad_norm": 0.15099132961950468, + "learning_rate": 4.639493205239346e-05, + "loss": 0.0389, + "step": 9984 + }, + { + "epoch": 4.698823529411765, + "grad_norm": 1.3781805934847663e-05, + "learning_rate": 4.636323674113583e-05, + "loss": 0.0, + "step": 9985 + }, + { + "epoch": 4.699294117647058, + "grad_norm": 0.14204584896723316, + "learning_rate": 4.6331550280966126e-05, + "loss": 0.0312, + "step": 9986 + }, + { + "epoch": 4.699764705882353, + "grad_norm": 0.19907568920074215, + "learning_rate": 4.629987267459056e-05, + "loss": 0.0469, + "step": 9987 + }, + { + "epoch": 4.700235294117647, + "grad_norm": 0.05220954514036557, + "learning_rate": 4.626820392471446e-05, + "loss": 0.0027, + "step": 9988 + }, + { + "epoch": 4.700705882352941, + "grad_norm": 0.10882481541133905, + "learning_rate": 4.6236544034042604e-05, + "loss": 0.0271, + "step": 9989 + }, + { + "epoch": 4.701176470588235, + "grad_norm": 0.09818964044294201, + "learning_rate": 4.62048930052788e-05, + "loss": 0.0169, + "step": 9990 + }, + { + "epoch": 4.7016470588235295, + "grad_norm": 0.11086613425192128, + "learning_rate": 4.6173250841126245e-05, + "loss": 0.0325, + "step": 9991 + }, + { + "epoch": 4.702117647058824, + "grad_norm": 0.056281999942841575, + "learning_rate": 4.614161754428737e-05, + "loss": 0.0082, + "step": 9992 + }, + { + "epoch": 4.702588235294118, + "grad_norm": 0.10268151745182193, + "learning_rate": 4.6109993117463756e-05, + "loss": 0.033, + "step": 9993 + }, + { + "epoch": 4.703058823529412, + "grad_norm": 1.0402254852160722e-06, + "learning_rate": 4.607837756335629e-05, + "loss": 0.0, + "step": 9994 + }, + { + "epoch": 4.7035294117647055, + "grad_norm": 8.963669555178854e-06, + "learning_rate": 4.604677088466516e-05, + "loss": 0.0, + "step": 9995 + }, + { + "epoch": 4.704, + "grad_norm": 0.24517406964410643, + "learning_rate": 4.6015173084089635e-05, + "loss": 0.091, + "step": 9996 + }, + { + "epoch": 4.704470588235294, + "grad_norm": 0.11938709609003671, + "learning_rate": 4.598358416432838e-05, + "loss": 0.0259, + "step": 9997 + }, + { + "epoch": 4.704941176470588, + "grad_norm": 0.08194173988781986, + "learning_rate": 4.595200412807927e-05, + "loss": 0.0003, + "step": 9998 + }, + { + "epoch": 4.705411764705882, + "grad_norm": 0.11501131647061111, + "learning_rate": 4.592043297803932e-05, + "loss": 0.0433, + "step": 9999 + }, + { + "epoch": 4.705882352941177, + "grad_norm": 7.966805026509399e-06, + "learning_rate": 4.58888707169049e-05, + "loss": 0.0, + "step": 10000 + }, + { + "epoch": 4.706352941176471, + "grad_norm": 0.3098629946910887, + "learning_rate": 4.585731734737164e-05, + "loss": 0.1114, + "step": 10001 + }, + { + "epoch": 4.706823529411765, + "grad_norm": 0.03307295098874109, + "learning_rate": 4.5825772872134234e-05, + "loss": 0.0024, + "step": 10002 + }, + { + "epoch": 4.707294117647059, + "grad_norm": 0.12789955085516874, + "learning_rate": 4.579423729388681e-05, + "loss": 0.0462, + "step": 10003 + }, + { + "epoch": 4.707764705882353, + "grad_norm": 0.18962467389509766, + "learning_rate": 4.576271061532268e-05, + "loss": 0.0764, + "step": 10004 + }, + { + "epoch": 4.708235294117647, + "grad_norm": 0.036920944736257406, + "learning_rate": 4.573119283913431e-05, + "loss": 0.004, + "step": 10005 + }, + { + "epoch": 4.708705882352941, + "grad_norm": 0.04930979045095041, + "learning_rate": 4.569968396801351e-05, + "loss": 0.0045, + "step": 10006 + }, + { + "epoch": 4.709176470588235, + "grad_norm": 0.12664082867389465, + "learning_rate": 4.566818400465133e-05, + "loss": 0.0372, + "step": 10007 + }, + { + "epoch": 4.7096470588235295, + "grad_norm": 0.1654478886267782, + "learning_rate": 4.563669295173793e-05, + "loss": 0.0602, + "step": 10008 + }, + { + "epoch": 4.710117647058824, + "grad_norm": 0.10239188131134214, + "learning_rate": 4.560521081196289e-05, + "loss": 0.0226, + "step": 10009 + }, + { + "epoch": 4.710588235294118, + "grad_norm": 0.08123322611688666, + "learning_rate": 4.5573737588014875e-05, + "loss": 0.0205, + "step": 10010 + }, + { + "epoch": 4.711058823529411, + "grad_norm": 0.14252033408174997, + "learning_rate": 4.554227328258185e-05, + "loss": 0.0383, + "step": 10011 + }, + { + "epoch": 4.7115294117647055, + "grad_norm": 0.0528296319689994, + "learning_rate": 4.551081789835113e-05, + "loss": 0.006, + "step": 10012 + }, + { + "epoch": 4.712, + "grad_norm": 0.010978366016035678, + "learning_rate": 4.547937143800901e-05, + "loss": 0.0004, + "step": 10013 + }, + { + "epoch": 4.712470588235294, + "grad_norm": 0.08798641705927757, + "learning_rate": 4.544793390424124e-05, + "loss": 0.0156, + "step": 10014 + }, + { + "epoch": 4.712941176470588, + "grad_norm": 4.584432935568031e-06, + "learning_rate": 4.541650529973279e-05, + "loss": 0.0, + "step": 10015 + }, + { + "epoch": 4.713411764705882, + "grad_norm": 0.16124591676063155, + "learning_rate": 4.5385085627167717e-05, + "loss": 0.0407, + "step": 10016 + }, + { + "epoch": 4.713882352941177, + "grad_norm": 0.08893370204480146, + "learning_rate": 4.535367488922947e-05, + "loss": 0.0168, + "step": 10017 + }, + { + "epoch": 4.714352941176471, + "grad_norm": 0.08167589231218322, + "learning_rate": 4.5322273088600714e-05, + "loss": 0.0131, + "step": 10018 + }, + { + "epoch": 4.714823529411765, + "grad_norm": 0.10255691171883821, + "learning_rate": 4.5290880227963186e-05, + "loss": 0.0209, + "step": 10019 + }, + { + "epoch": 4.715294117647058, + "grad_norm": 1.1949344323534991e-05, + "learning_rate": 4.525949630999818e-05, + "loss": 0.0, + "step": 10020 + }, + { + "epoch": 4.715764705882353, + "grad_norm": 0.0889434591893724, + "learning_rate": 4.522812133738589e-05, + "loss": 0.0046, + "step": 10021 + }, + { + "epoch": 4.716235294117647, + "grad_norm": 0.06105844693694135, + "learning_rate": 4.519675531280594e-05, + "loss": 0.0099, + "step": 10022 + }, + { + "epoch": 4.716705882352941, + "grad_norm": 0.03747866044697236, + "learning_rate": 4.5165398238937194e-05, + "loss": 0.0024, + "step": 10023 + }, + { + "epoch": 4.717176470588235, + "grad_norm": 0.04974960875145044, + "learning_rate": 4.51340501184576e-05, + "loss": 0.0035, + "step": 10024 + }, + { + "epoch": 4.7176470588235295, + "grad_norm": 0.09948837138836747, + "learning_rate": 4.510271095404449e-05, + "loss": 0.0215, + "step": 10025 + }, + { + "epoch": 4.718117647058824, + "grad_norm": 0.028459875522673306, + "learning_rate": 4.507138074837443e-05, + "loss": 0.0027, + "step": 10026 + }, + { + "epoch": 4.718588235294118, + "grad_norm": 0.07529371317077789, + "learning_rate": 4.504005950412308e-05, + "loss": 0.009, + "step": 10027 + }, + { + "epoch": 4.719058823529412, + "grad_norm": 0.15099697936564482, + "learning_rate": 4.5008747223965456e-05, + "loss": 0.0393, + "step": 10028 + }, + { + "epoch": 4.7195294117647055, + "grad_norm": 0.0001593241521290637, + "learning_rate": 4.4977443910575866e-05, + "loss": 0.0, + "step": 10029 + }, + { + "epoch": 4.72, + "grad_norm": 0.09511583190837336, + "learning_rate": 4.494614956662765e-05, + "loss": 0.0245, + "step": 10030 + }, + { + "epoch": 4.720470588235294, + "grad_norm": 0.14718640955041049, + "learning_rate": 4.491486419479355e-05, + "loss": 0.0394, + "step": 10031 + }, + { + "epoch": 4.720941176470588, + "grad_norm": 0.09122829167067686, + "learning_rate": 4.488358779774551e-05, + "loss": 0.0144, + "step": 10032 + }, + { + "epoch": 4.721411764705882, + "grad_norm": 0.15442073606651954, + "learning_rate": 4.485232037815465e-05, + "loss": 0.0412, + "step": 10033 + }, + { + "epoch": 4.721882352941177, + "grad_norm": 0.07412962250706465, + "learning_rate": 4.4821061938691355e-05, + "loss": 0.0151, + "step": 10034 + }, + { + "epoch": 4.722352941176471, + "grad_norm": 0.1705933515467233, + "learning_rate": 4.478981248202533e-05, + "loss": 0.0321, + "step": 10035 + }, + { + "epoch": 4.722823529411765, + "grad_norm": 0.24064299003991688, + "learning_rate": 4.4758572010825324e-05, + "loss": 0.0754, + "step": 10036 + }, + { + "epoch": 4.723294117647059, + "grad_norm": 0.08272919566769438, + "learning_rate": 4.472734052775947e-05, + "loss": 0.0188, + "step": 10037 + }, + { + "epoch": 4.723764705882353, + "grad_norm": 2.251139746129542e-05, + "learning_rate": 4.4696118035495136e-05, + "loss": 0.0, + "step": 10038 + }, + { + "epoch": 4.724235294117647, + "grad_norm": 3.2138842892885552e-06, + "learning_rate": 4.4664904536698796e-05, + "loss": 0.0, + "step": 10039 + }, + { + "epoch": 4.724705882352941, + "grad_norm": 0.08839506075037139, + "learning_rate": 4.463370003403632e-05, + "loss": 0.0141, + "step": 10040 + }, + { + "epoch": 4.725176470588235, + "grad_norm": 4.613059221015111e-06, + "learning_rate": 4.4602504530172636e-05, + "loss": 0.0, + "step": 10041 + }, + { + "epoch": 4.7256470588235295, + "grad_norm": 1.686738579972217e-06, + "learning_rate": 4.457131802777202e-05, + "loss": 0.0, + "step": 10042 + }, + { + "epoch": 4.726117647058824, + "grad_norm": 0.21407070917814483, + "learning_rate": 4.454014052949804e-05, + "loss": 0.0446, + "step": 10043 + }, + { + "epoch": 4.726588235294118, + "grad_norm": 0.07606060567242591, + "learning_rate": 4.4508972038013274e-05, + "loss": 0.0076, + "step": 10044 + }, + { + "epoch": 4.727058823529411, + "grad_norm": 0.2046076114151625, + "learning_rate": 4.4477812555979724e-05, + "loss": 0.0654, + "step": 10045 + }, + { + "epoch": 4.7275294117647055, + "grad_norm": 0.00030489977949307986, + "learning_rate": 4.4446662086058634e-05, + "loss": 0.0, + "step": 10046 + }, + { + "epoch": 4.728, + "grad_norm": 0.2475412454993642, + "learning_rate": 4.4415520630910265e-05, + "loss": 0.043, + "step": 10047 + }, + { + "epoch": 4.728470588235294, + "grad_norm": 1.053694345018148e-06, + "learning_rate": 4.438438819319434e-05, + "loss": 0.0, + "step": 10048 + }, + { + "epoch": 4.728941176470588, + "grad_norm": 0.03828519918440996, + "learning_rate": 4.435326477556975e-05, + "loss": 0.0026, + "step": 10049 + }, + { + "epoch": 4.729411764705882, + "grad_norm": 0.0953148093548943, + "learning_rate": 4.432215038069449e-05, + "loss": 0.0228, + "step": 10050 + }, + { + "epoch": 4.729882352941177, + "grad_norm": 0.10306643276170263, + "learning_rate": 4.4291045011225934e-05, + "loss": 0.0379, + "step": 10051 + }, + { + "epoch": 4.730352941176471, + "grad_norm": 0.09840981709310313, + "learning_rate": 4.4259948669820676e-05, + "loss": 0.0151, + "step": 10052 + }, + { + "epoch": 4.730823529411765, + "grad_norm": 9.084644882201401e-06, + "learning_rate": 4.422886135913441e-05, + "loss": 0.0, + "step": 10053 + }, + { + "epoch": 4.731294117647058, + "grad_norm": 0.15630422136853347, + "learning_rate": 4.419778308182219e-05, + "loss": 0.0365, + "step": 10054 + }, + { + "epoch": 4.731764705882353, + "grad_norm": 0.15607951357444316, + "learning_rate": 4.41667138405383e-05, + "loss": 0.0658, + "step": 10055 + }, + { + "epoch": 4.732235294117647, + "grad_norm": 0.1828642129179832, + "learning_rate": 4.413565363793611e-05, + "loss": 0.0705, + "step": 10056 + }, + { + "epoch": 4.732705882352941, + "grad_norm": 0.11563933191931425, + "learning_rate": 4.4104602476668364e-05, + "loss": 0.0241, + "step": 10057 + }, + { + "epoch": 4.733176470588235, + "grad_norm": 0.00023480739788175845, + "learning_rate": 4.4073560359387035e-05, + "loss": 0.0, + "step": 10058 + }, + { + "epoch": 4.7336470588235295, + "grad_norm": 0.06723969372430104, + "learning_rate": 4.4042527288743155e-05, + "loss": 0.0037, + "step": 10059 + }, + { + "epoch": 4.734117647058824, + "grad_norm": 0.05329138772800472, + "learning_rate": 4.401150326738722e-05, + "loss": 0.0061, + "step": 10060 + }, + { + "epoch": 4.734588235294118, + "grad_norm": 0.1376100445264864, + "learning_rate": 4.3980488297968726e-05, + "loss": 0.0398, + "step": 10061 + }, + { + "epoch": 4.735058823529412, + "grad_norm": 0.16056733510952667, + "learning_rate": 4.394948238313655e-05, + "loss": 0.0217, + "step": 10062 + }, + { + "epoch": 4.7355294117647055, + "grad_norm": 0.09616610615919231, + "learning_rate": 4.391848552553875e-05, + "loss": 0.0274, + "step": 10063 + }, + { + "epoch": 4.736, + "grad_norm": 0.05333375712702846, + "learning_rate": 4.3887497727822616e-05, + "loss": 0.0072, + "step": 10064 + }, + { + "epoch": 4.736470588235294, + "grad_norm": 0.07843054158174186, + "learning_rate": 4.385651899263467e-05, + "loss": 0.0194, + "step": 10065 + }, + { + "epoch": 4.736941176470588, + "grad_norm": 0.12497857606655587, + "learning_rate": 4.3825549322620657e-05, + "loss": 0.0367, + "step": 10066 + }, + { + "epoch": 4.737411764705882, + "grad_norm": 0.021041123009541994, + "learning_rate": 4.3794588720425486e-05, + "loss": 0.0009, + "step": 10067 + }, + { + "epoch": 4.737882352941177, + "grad_norm": 0.11196338552111362, + "learning_rate": 4.3763637188693404e-05, + "loss": 0.0395, + "step": 10068 + }, + { + "epoch": 4.738352941176471, + "grad_norm": 0.09485769390451142, + "learning_rate": 4.3732694730067766e-05, + "loss": 0.0138, + "step": 10069 + }, + { + "epoch": 4.738823529411764, + "grad_norm": 0.09358028474860194, + "learning_rate": 4.370176134719121e-05, + "loss": 0.0238, + "step": 10070 + }, + { + "epoch": 4.739294117647058, + "grad_norm": 1.0886353231751628e-06, + "learning_rate": 4.36708370427057e-05, + "loss": 0.0, + "step": 10071 + }, + { + "epoch": 4.739764705882353, + "grad_norm": 7.082241360820362e-07, + "learning_rate": 4.363992181925218e-05, + "loss": 0.0, + "step": 10072 + }, + { + "epoch": 4.740235294117647, + "grad_norm": 0.06261730172914678, + "learning_rate": 4.3609015679471036e-05, + "loss": 0.0092, + "step": 10073 + }, + { + "epoch": 4.740705882352941, + "grad_norm": 0.06919223950368135, + "learning_rate": 4.357811862600185e-05, + "loss": 0.0091, + "step": 10074 + }, + { + "epoch": 4.741176470588235, + "grad_norm": 0.18934222073790521, + "learning_rate": 4.3547230661483276e-05, + "loss": 0.0523, + "step": 10075 + }, + { + "epoch": 4.7416470588235295, + "grad_norm": 0.03460817057243014, + "learning_rate": 4.351635178855336e-05, + "loss": 0.0023, + "step": 10076 + }, + { + "epoch": 4.742117647058824, + "grad_norm": 1.1001005437817126e-05, + "learning_rate": 4.348548200984934e-05, + "loss": 0.0, + "step": 10077 + }, + { + "epoch": 4.742588235294118, + "grad_norm": 0.09598338824507559, + "learning_rate": 4.345462132800756e-05, + "loss": 0.0262, + "step": 10078 + }, + { + "epoch": 4.743058823529411, + "grad_norm": 0.09742164755662, + "learning_rate": 4.342376974566372e-05, + "loss": 0.0185, + "step": 10079 + }, + { + "epoch": 4.7435294117647056, + "grad_norm": 0.07303920005402645, + "learning_rate": 4.339292726545275e-05, + "loss": 0.0165, + "step": 10080 + }, + { + "epoch": 4.744, + "grad_norm": 0.1674494501212873, + "learning_rate": 4.3362093890008645e-05, + "loss": 0.0216, + "step": 10081 + }, + { + "epoch": 4.744470588235294, + "grad_norm": 0.11039100647825555, + "learning_rate": 4.333126962196478e-05, + "loss": 0.0466, + "step": 10082 + }, + { + "epoch": 4.744941176470588, + "grad_norm": 3.406347907555256e-06, + "learning_rate": 4.330045446395374e-05, + "loss": 0.0, + "step": 10083 + }, + { + "epoch": 4.745411764705882, + "grad_norm": 0.08228426420803482, + "learning_rate": 4.3269648418607194e-05, + "loss": 0.0161, + "step": 10084 + }, + { + "epoch": 4.745882352941177, + "grad_norm": 0.12128563232888222, + "learning_rate": 4.32388514885562e-05, + "loss": 0.047, + "step": 10085 + }, + { + "epoch": 4.746352941176471, + "grad_norm": 0.0949935466730887, + "learning_rate": 4.320806367643098e-05, + "loss": 0.0242, + "step": 10086 + }, + { + "epoch": 4.746823529411765, + "grad_norm": 0.06398318276234567, + "learning_rate": 4.3177284984860895e-05, + "loss": 0.0132, + "step": 10087 + }, + { + "epoch": 4.747294117647058, + "grad_norm": 0.02888110299954374, + "learning_rate": 4.314651541647468e-05, + "loss": 0.0023, + "step": 10088 + }, + { + "epoch": 4.747764705882353, + "grad_norm": 0.051028422401454866, + "learning_rate": 4.3115754973900104e-05, + "loss": 0.0018, + "step": 10089 + }, + { + "epoch": 4.748235294117647, + "grad_norm": 3.9442977244181025e-06, + "learning_rate": 4.308500365976434e-05, + "loss": 0.0, + "step": 10090 + }, + { + "epoch": 4.748705882352941, + "grad_norm": 0.04258609070525581, + "learning_rate": 4.305426147669371e-05, + "loss": 0.0052, + "step": 10091 + }, + { + "epoch": 4.749176470588235, + "grad_norm": 0.06666589435037329, + "learning_rate": 4.302352842731368e-05, + "loss": 0.0089, + "step": 10092 + }, + { + "epoch": 4.7496470588235296, + "grad_norm": 0.1684555444892639, + "learning_rate": 4.299280451424903e-05, + "loss": 0.0434, + "step": 10093 + }, + { + "epoch": 4.750117647058824, + "grad_norm": 0.15381776816243073, + "learning_rate": 4.296208974012379e-05, + "loss": 0.0458, + "step": 10094 + }, + { + "epoch": 4.750588235294118, + "grad_norm": 8.707928182745973e-06, + "learning_rate": 4.2931384107561065e-05, + "loss": 0.0, + "step": 10095 + }, + { + "epoch": 4.751058823529412, + "grad_norm": 0.06853642663037879, + "learning_rate": 4.290068761918329e-05, + "loss": 0.0146, + "step": 10096 + }, + { + "epoch": 4.751529411764706, + "grad_norm": 0.053874030227427726, + "learning_rate": 4.287000027761217e-05, + "loss": 0.0075, + "step": 10097 + }, + { + "epoch": 4.752, + "grad_norm": 0.10619596995552417, + "learning_rate": 4.2839322085468445e-05, + "loss": 0.0224, + "step": 10098 + }, + { + "epoch": 4.752470588235294, + "grad_norm": 0.06888740106495335, + "learning_rate": 4.280865304537223e-05, + "loss": 0.0054, + "step": 10099 + }, + { + "epoch": 4.752941176470588, + "grad_norm": 0.1506480064634225, + "learning_rate": 4.277799315994286e-05, + "loss": 0.0436, + "step": 10100 + }, + { + "epoch": 4.7534117647058824, + "grad_norm": 0.14943123657936158, + "learning_rate": 4.2747342431798767e-05, + "loss": 0.0341, + "step": 10101 + }, + { + "epoch": 4.753882352941177, + "grad_norm": 0.10134050845572312, + "learning_rate": 4.271670086355768e-05, + "loss": 0.0236, + "step": 10102 + }, + { + "epoch": 4.754352941176471, + "grad_norm": 9.35218291682037e-06, + "learning_rate": 4.268606845783663e-05, + "loss": 0.0, + "step": 10103 + }, + { + "epoch": 4.754823529411764, + "grad_norm": 0.18644795505786937, + "learning_rate": 4.265544521725165e-05, + "loss": 0.0918, + "step": 10104 + }, + { + "epoch": 4.7552941176470584, + "grad_norm": 0.08866888874150079, + "learning_rate": 4.2624831144418166e-05, + "loss": 0.0329, + "step": 10105 + }, + { + "epoch": 4.755764705882353, + "grad_norm": 0.07073584670644936, + "learning_rate": 4.259422624195082e-05, + "loss": 0.0124, + "step": 10106 + }, + { + "epoch": 4.756235294117647, + "grad_norm": 0.12764137122701283, + "learning_rate": 4.2563630512463284e-05, + "loss": 0.0437, + "step": 10107 + }, + { + "epoch": 4.756705882352941, + "grad_norm": 0.11056944146219012, + "learning_rate": 4.253304395856876e-05, + "loss": 0.0284, + "step": 10108 + }, + { + "epoch": 4.757176470588235, + "grad_norm": 3.479235708705421e-05, + "learning_rate": 4.2502466582879375e-05, + "loss": 0.0, + "step": 10109 + }, + { + "epoch": 4.75764705882353, + "grad_norm": 0.09824467186655998, + "learning_rate": 4.2471898388006606e-05, + "loss": 0.0283, + "step": 10110 + }, + { + "epoch": 4.758117647058824, + "grad_norm": 6.191280936061282e-05, + "learning_rate": 4.244133937656118e-05, + "loss": 0.0, + "step": 10111 + }, + { + "epoch": 4.758588235294118, + "grad_norm": 0.16348692656506464, + "learning_rate": 4.2410789551152894e-05, + "loss": 0.0614, + "step": 10112 + }, + { + "epoch": 4.759058823529411, + "grad_norm": 0.06309430882925841, + "learning_rate": 4.238024891439092e-05, + "loss": 0.0077, + "step": 10113 + }, + { + "epoch": 4.759529411764706, + "grad_norm": 0.19531409062775235, + "learning_rate": 4.234971746888359e-05, + "loss": 0.0538, + "step": 10114 + }, + { + "epoch": 4.76, + "grad_norm": 0.1179664325305082, + "learning_rate": 4.231919521723836e-05, + "loss": 0.0315, + "step": 10115 + }, + { + "epoch": 4.760470588235294, + "grad_norm": 0.09884690114831403, + "learning_rate": 4.228868216206206e-05, + "loss": 0.0087, + "step": 10116 + }, + { + "epoch": 4.760941176470588, + "grad_norm": 0.12300185010854411, + "learning_rate": 4.225817830596064e-05, + "loss": 0.0227, + "step": 10117 + }, + { + "epoch": 4.7614117647058825, + "grad_norm": 0.14443821589247574, + "learning_rate": 4.2227683651539236e-05, + "loss": 0.0514, + "step": 10118 + }, + { + "epoch": 4.761882352941177, + "grad_norm": 0.1272939956179257, + "learning_rate": 4.21971982014023e-05, + "loss": 0.0147, + "step": 10119 + }, + { + "epoch": 4.762352941176471, + "grad_norm": 0.02454487716194121, + "learning_rate": 4.2166721958153365e-05, + "loss": 0.0017, + "step": 10120 + }, + { + "epoch": 4.762823529411765, + "grad_norm": 0.2108118662542334, + "learning_rate": 4.2136254924395305e-05, + "loss": 0.0581, + "step": 10121 + }, + { + "epoch": 4.7632941176470585, + "grad_norm": 0.018958192815335492, + "learning_rate": 4.210579710273018e-05, + "loss": 0.0014, + "step": 10122 + }, + { + "epoch": 4.763764705882353, + "grad_norm": 9.338753527975592e-07, + "learning_rate": 4.207534849575916e-05, + "loss": 0.0, + "step": 10123 + }, + { + "epoch": 4.764235294117647, + "grad_norm": 1.502158541383548e-05, + "learning_rate": 4.204490910608274e-05, + "loss": 0.0, + "step": 10124 + }, + { + "epoch": 4.764705882352941, + "grad_norm": 0.08930659107157594, + "learning_rate": 4.201447893630065e-05, + "loss": 0.021, + "step": 10125 + }, + { + "epoch": 4.765176470588235, + "grad_norm": 0.14586636072235815, + "learning_rate": 4.1984057989011674e-05, + "loss": 0.0599, + "step": 10126 + }, + { + "epoch": 4.76564705882353, + "grad_norm": 5.6774098080404425e-06, + "learning_rate": 4.195364626681396e-05, + "loss": 0.0, + "step": 10127 + }, + { + "epoch": 4.766117647058824, + "grad_norm": 0.0892989070185484, + "learning_rate": 4.192324377230487e-05, + "loss": 0.0185, + "step": 10128 + }, + { + "epoch": 4.766588235294118, + "grad_norm": 0.10533714024385506, + "learning_rate": 4.1892850508080824e-05, + "loss": 0.0361, + "step": 10129 + }, + { + "epoch": 4.767058823529412, + "grad_norm": 1.0452507375817591e-05, + "learning_rate": 4.18624664767376e-05, + "loss": 0.0, + "step": 10130 + }, + { + "epoch": 4.767529411764706, + "grad_norm": 0.07221444309698753, + "learning_rate": 4.1832091680870214e-05, + "loss": 0.0075, + "step": 10131 + }, + { + "epoch": 4.768, + "grad_norm": 0.08137078950457544, + "learning_rate": 4.1801726123072696e-05, + "loss": 0.0142, + "step": 10132 + }, + { + "epoch": 4.768470588235294, + "grad_norm": 0.039529532875908144, + "learning_rate": 4.1771369805938486e-05, + "loss": 0.0048, + "step": 10133 + }, + { + "epoch": 4.768941176470588, + "grad_norm": 0.10738818572077759, + "learning_rate": 4.1741022732060204e-05, + "loss": 0.0261, + "step": 10134 + }, + { + "epoch": 4.7694117647058825, + "grad_norm": 0.06278590056406556, + "learning_rate": 4.171068490402954e-05, + "loss": 0.0089, + "step": 10135 + }, + { + "epoch": 4.769882352941177, + "grad_norm": 0.04551089529297629, + "learning_rate": 4.168035632443754e-05, + "loss": 0.0037, + "step": 10136 + }, + { + "epoch": 4.770352941176471, + "grad_norm": 0.03249421109328039, + "learning_rate": 4.165003699587445e-05, + "loss": 0.003, + "step": 10137 + }, + { + "epoch": 4.770823529411764, + "grad_norm": 0.08249632560853735, + "learning_rate": 4.161972692092962e-05, + "loss": 0.0185, + "step": 10138 + }, + { + "epoch": 4.7712941176470585, + "grad_norm": 0.13044059518658363, + "learning_rate": 4.158942610219175e-05, + "loss": 0.0419, + "step": 10139 + }, + { + "epoch": 4.771764705882353, + "grad_norm": 0.12541775163214744, + "learning_rate": 4.155913454224862e-05, + "loss": 0.0229, + "step": 10140 + }, + { + "epoch": 4.772235294117647, + "grad_norm": 0.14546239851374257, + "learning_rate": 4.1528852243687276e-05, + "loss": 0.0266, + "step": 10141 + }, + { + "epoch": 4.772705882352941, + "grad_norm": 0.05917957327958128, + "learning_rate": 4.1498579209094046e-05, + "loss": 0.0084, + "step": 10142 + }, + { + "epoch": 4.773176470588235, + "grad_norm": 0.20262822879178222, + "learning_rate": 4.146831544105431e-05, + "loss": 0.0241, + "step": 10143 + }, + { + "epoch": 4.77364705882353, + "grad_norm": 0.04828432072124388, + "learning_rate": 4.143806094215277e-05, + "loss": 0.0061, + "step": 10144 + }, + { + "epoch": 4.774117647058824, + "grad_norm": 0.11444000827408972, + "learning_rate": 4.140781571497336e-05, + "loss": 0.0288, + "step": 10145 + }, + { + "epoch": 4.774588235294118, + "grad_norm": 0.12084010695782796, + "learning_rate": 4.137757976209911e-05, + "loss": 0.0341, + "step": 10146 + }, + { + "epoch": 4.775058823529411, + "grad_norm": 0.04602498925104279, + "learning_rate": 4.134735308611232e-05, + "loss": 0.0052, + "step": 10147 + }, + { + "epoch": 4.775529411764706, + "grad_norm": 0.09263184992819914, + "learning_rate": 4.131713568959457e-05, + "loss": 0.0133, + "step": 10148 + }, + { + "epoch": 4.776, + "grad_norm": 0.09016633303159638, + "learning_rate": 4.128692757512646e-05, + "loss": 0.0214, + "step": 10149 + }, + { + "epoch": 4.776470588235294, + "grad_norm": 0.3296556313095173, + "learning_rate": 4.125672874528797e-05, + "loss": 0.0561, + "step": 10150 + }, + { + "epoch": 4.776941176470588, + "grad_norm": 1.0943190546781697e-05, + "learning_rate": 4.122653920265828e-05, + "loss": 0.0, + "step": 10151 + }, + { + "epoch": 4.7774117647058825, + "grad_norm": 0.0925213914980529, + "learning_rate": 4.119635894981558e-05, + "loss": 0.0087, + "step": 10152 + }, + { + "epoch": 4.777882352941177, + "grad_norm": 0.15313723656822384, + "learning_rate": 4.1166187989337596e-05, + "loss": 0.0379, + "step": 10153 + }, + { + "epoch": 4.778352941176471, + "grad_norm": 0.09906267277939214, + "learning_rate": 4.113602632380094e-05, + "loss": 0.0149, + "step": 10154 + }, + { + "epoch": 4.778823529411765, + "grad_norm": 0.06978020109219432, + "learning_rate": 4.110587395578161e-05, + "loss": 0.0094, + "step": 10155 + }, + { + "epoch": 4.7792941176470585, + "grad_norm": 2.5479955621178644e-06, + "learning_rate": 4.1075730887854805e-05, + "loss": 0.0, + "step": 10156 + }, + { + "epoch": 4.779764705882353, + "grad_norm": 0.1096210485335013, + "learning_rate": 4.104559712259482e-05, + "loss": 0.0154, + "step": 10157 + }, + { + "epoch": 4.780235294117647, + "grad_norm": 0.1400426619124492, + "learning_rate": 4.101547266257526e-05, + "loss": 0.0511, + "step": 10158 + }, + { + "epoch": 4.780705882352941, + "grad_norm": 1.925296650139417e-05, + "learning_rate": 4.098535751036894e-05, + "loss": 0.0, + "step": 10159 + }, + { + "epoch": 4.781176470588235, + "grad_norm": 0.0619864118028257, + "learning_rate": 4.0955251668547754e-05, + "loss": 0.0057, + "step": 10160 + }, + { + "epoch": 4.78164705882353, + "grad_norm": 0.15001348277436802, + "learning_rate": 4.092515513968294e-05, + "loss": 0.0674, + "step": 10161 + }, + { + "epoch": 4.782117647058824, + "grad_norm": 0.5238668647604621, + "learning_rate": 4.0895067926344935e-05, + "loss": 0.1486, + "step": 10162 + }, + { + "epoch": 4.782588235294118, + "grad_norm": 0.10631229914077461, + "learning_rate": 4.0864990031103235e-05, + "loss": 0.0183, + "step": 10163 + }, + { + "epoch": 4.783058823529412, + "grad_norm": 0.07617131758218161, + "learning_rate": 4.0834921456526684e-05, + "loss": 0.003, + "step": 10164 + }, + { + "epoch": 4.783529411764706, + "grad_norm": 0.20318104931329514, + "learning_rate": 4.080486220518336e-05, + "loss": 0.079, + "step": 10165 + }, + { + "epoch": 4.784, + "grad_norm": 0.05638754638673951, + "learning_rate": 4.0774812279640325e-05, + "loss": 0.0052, + "step": 10166 + }, + { + "epoch": 4.784470588235294, + "grad_norm": 0.08623484412823561, + "learning_rate": 4.074477168246409e-05, + "loss": 0.0116, + "step": 10167 + }, + { + "epoch": 4.784941176470588, + "grad_norm": 0.10078435517841108, + "learning_rate": 4.071474041622029e-05, + "loss": 0.0208, + "step": 10168 + }, + { + "epoch": 4.7854117647058825, + "grad_norm": 0.08183973432012567, + "learning_rate": 4.0684718483473656e-05, + "loss": 0.0082, + "step": 10169 + }, + { + "epoch": 4.785882352941177, + "grad_norm": 0.10625784627513907, + "learning_rate": 4.06547058867883e-05, + "loss": 0.0251, + "step": 10170 + }, + { + "epoch": 4.786352941176471, + "grad_norm": 0.06263561333795183, + "learning_rate": 4.0624702628727345e-05, + "loss": 0.0087, + "step": 10171 + }, + { + "epoch": 4.786823529411764, + "grad_norm": 0.3022301157699738, + "learning_rate": 4.0594708711853267e-05, + "loss": 0.0893, + "step": 10172 + }, + { + "epoch": 4.7872941176470585, + "grad_norm": 0.21634156412894623, + "learning_rate": 4.0564724138727734e-05, + "loss": 0.0256, + "step": 10173 + }, + { + "epoch": 4.787764705882353, + "grad_norm": 0.15534338502236622, + "learning_rate": 4.0534748911911505e-05, + "loss": 0.0489, + "step": 10174 + }, + { + "epoch": 4.788235294117647, + "grad_norm": 2.439170067636663e-05, + "learning_rate": 4.0504783033964645e-05, + "loss": 0.0, + "step": 10175 + }, + { + "epoch": 4.788705882352941, + "grad_norm": 5.818867217261314e-05, + "learning_rate": 4.047482650744641e-05, + "loss": 0.0, + "step": 10176 + }, + { + "epoch": 4.789176470588235, + "grad_norm": 0.00042056025118152664, + "learning_rate": 4.044487933491518e-05, + "loss": 0.0, + "step": 10177 + }, + { + "epoch": 4.78964705882353, + "grad_norm": 0.1555357588234508, + "learning_rate": 4.041494151892862e-05, + "loss": 0.0427, + "step": 10178 + }, + { + "epoch": 4.790117647058824, + "grad_norm": 0.0024299610845198937, + "learning_rate": 4.038501306204361e-05, + "loss": 0.0, + "step": 10179 + }, + { + "epoch": 4.790588235294118, + "grad_norm": 1.1940814055838385e-05, + "learning_rate": 4.0355093966816095e-05, + "loss": 0.0, + "step": 10180 + }, + { + "epoch": 4.791058823529411, + "grad_norm": 0.09456563910053468, + "learning_rate": 4.032518423580135e-05, + "loss": 0.0236, + "step": 10181 + }, + { + "epoch": 4.791529411764706, + "grad_norm": 0.11849602752480477, + "learning_rate": 4.029528387155389e-05, + "loss": 0.0331, + "step": 10182 + }, + { + "epoch": 4.792, + "grad_norm": 3.4792516630707624e-05, + "learning_rate": 4.026539287662722e-05, + "loss": 0.0, + "step": 10183 + }, + { + "epoch": 4.792470588235294, + "grad_norm": 0.055252314599190396, + "learning_rate": 4.0235511253574256e-05, + "loss": 0.0106, + "step": 10184 + }, + { + "epoch": 4.792941176470588, + "grad_norm": 0.16950556092380364, + "learning_rate": 4.020563900494707e-05, + "loss": 0.0706, + "step": 10185 + }, + { + "epoch": 4.7934117647058825, + "grad_norm": 0.1162471686993331, + "learning_rate": 4.01757761332968e-05, + "loss": 0.0288, + "step": 10186 + }, + { + "epoch": 4.793882352941177, + "grad_norm": 0.1426132483852827, + "learning_rate": 4.0145922641173934e-05, + "loss": 0.0367, + "step": 10187 + }, + { + "epoch": 4.794352941176471, + "grad_norm": 0.08217785159995004, + "learning_rate": 4.011607853112815e-05, + "loss": 0.0193, + "step": 10188 + }, + { + "epoch": 4.794823529411765, + "grad_norm": 0.038353042536662114, + "learning_rate": 4.008624380570819e-05, + "loss": 0.0036, + "step": 10189 + }, + { + "epoch": 4.7952941176470585, + "grad_norm": 0.12264823839591099, + "learning_rate": 4.0056418467462184e-05, + "loss": 0.0339, + "step": 10190 + }, + { + "epoch": 4.795764705882353, + "grad_norm": 0.0008200655334735699, + "learning_rate": 4.002660251893727e-05, + "loss": 0.0, + "step": 10191 + }, + { + "epoch": 4.796235294117647, + "grad_norm": 0.13785111841430064, + "learning_rate": 3.9996795962679896e-05, + "loss": 0.0438, + "step": 10192 + }, + { + "epoch": 4.796705882352941, + "grad_norm": 0.05003546078707346, + "learning_rate": 3.9966998801235766e-05, + "loss": 0.0043, + "step": 10193 + }, + { + "epoch": 4.797176470588235, + "grad_norm": 6.128094978314959e-05, + "learning_rate": 3.9937211037149616e-05, + "loss": 0.0, + "step": 10194 + }, + { + "epoch": 4.79764705882353, + "grad_norm": 0.11026537683555618, + "learning_rate": 3.9907432672965485e-05, + "loss": 0.0091, + "step": 10195 + }, + { + "epoch": 4.798117647058824, + "grad_norm": 0.07792141271008213, + "learning_rate": 3.987766371122661e-05, + "loss": 0.01, + "step": 10196 + }, + { + "epoch": 4.798588235294117, + "grad_norm": 1.3419279323935346e-05, + "learning_rate": 3.984790415447539e-05, + "loss": 0.0, + "step": 10197 + }, + { + "epoch": 4.799058823529411, + "grad_norm": 0.20058234307046294, + "learning_rate": 3.9818154005253514e-05, + "loss": 0.0442, + "step": 10198 + }, + { + "epoch": 4.799529411764706, + "grad_norm": 0.0006214800082499538, + "learning_rate": 3.978841326610166e-05, + "loss": 0.0, + "step": 10199 + }, + { + "epoch": 4.8, + "grad_norm": 0.1837274315334236, + "learning_rate": 3.97586819395599e-05, + "loss": 0.0418, + "step": 10200 + }, + { + "epoch": 4.800470588235294, + "grad_norm": 0.09140615936515263, + "learning_rate": 3.972896002816749e-05, + "loss": 0.0247, + "step": 10201 + }, + { + "epoch": 4.800941176470588, + "grad_norm": 0.1452904231472526, + "learning_rate": 3.969924753446271e-05, + "loss": 0.0145, + "step": 10202 + }, + { + "epoch": 4.8014117647058825, + "grad_norm": 0.1257922056762048, + "learning_rate": 3.966954446098322e-05, + "loss": 0.0398, + "step": 10203 + }, + { + "epoch": 4.801882352941177, + "grad_norm": 0.18974166459597339, + "learning_rate": 3.9639850810265845e-05, + "loss": 0.0017, + "step": 10204 + }, + { + "epoch": 4.802352941176471, + "grad_norm": 0.09326872315890067, + "learning_rate": 3.961016658484648e-05, + "loss": 0.0203, + "step": 10205 + }, + { + "epoch": 4.802823529411764, + "grad_norm": 3.203508839923403e-05, + "learning_rate": 3.9580491787260346e-05, + "loss": 0.0, + "step": 10206 + }, + { + "epoch": 4.8032941176470585, + "grad_norm": 0.12215552484484415, + "learning_rate": 3.955082642004187e-05, + "loss": 0.0337, + "step": 10207 + }, + { + "epoch": 4.803764705882353, + "grad_norm": 0.07147098918266982, + "learning_rate": 3.952117048572452e-05, + "loss": 0.0057, + "step": 10208 + }, + { + "epoch": 4.804235294117647, + "grad_norm": 0.03782440220883456, + "learning_rate": 3.94915239868411e-05, + "loss": 0.0019, + "step": 10209 + }, + { + "epoch": 4.804705882352941, + "grad_norm": 2.8341325658460646e-05, + "learning_rate": 3.946188692592362e-05, + "loss": 0.0, + "step": 10210 + }, + { + "epoch": 4.805176470588235, + "grad_norm": 0.13437117032185972, + "learning_rate": 3.943225930550315e-05, + "loss": 0.0318, + "step": 10211 + }, + { + "epoch": 4.80564705882353, + "grad_norm": 0.16090725064949285, + "learning_rate": 3.9402641128110054e-05, + "loss": 0.0234, + "step": 10212 + }, + { + "epoch": 4.806117647058824, + "grad_norm": 0.06660693296435544, + "learning_rate": 3.937303239627392e-05, + "loss": 0.0046, + "step": 10213 + }, + { + "epoch": 4.806588235294118, + "grad_norm": 0.07924018949606662, + "learning_rate": 3.934343311252341e-05, + "loss": 0.0143, + "step": 10214 + }, + { + "epoch": 4.807058823529411, + "grad_norm": 0.19092954267456816, + "learning_rate": 3.9313843279386486e-05, + "loss": 0.0467, + "step": 10215 + }, + { + "epoch": 4.807529411764706, + "grad_norm": 0.06858284192675046, + "learning_rate": 3.9284262899390285e-05, + "loss": 0.01, + "step": 10216 + }, + { + "epoch": 4.808, + "grad_norm": 0.15119089195523358, + "learning_rate": 3.925469197506107e-05, + "loss": 0.0354, + "step": 10217 + }, + { + "epoch": 4.808470588235294, + "grad_norm": 0.2422120197586943, + "learning_rate": 3.922513050892434e-05, + "loss": 0.0595, + "step": 10218 + }, + { + "epoch": 4.808941176470588, + "grad_norm": 0.06626983589518019, + "learning_rate": 3.919557850350486e-05, + "loss": 0.0093, + "step": 10219 + }, + { + "epoch": 4.8094117647058825, + "grad_norm": 9.875174489752668e-05, + "learning_rate": 3.916603596132642e-05, + "loss": 0.0, + "step": 10220 + }, + { + "epoch": 4.809882352941177, + "grad_norm": 0.19523238255757658, + "learning_rate": 3.91365028849122e-05, + "loss": 0.0119, + "step": 10221 + }, + { + "epoch": 4.810352941176471, + "grad_norm": 0.14459299590796076, + "learning_rate": 3.910697927678437e-05, + "loss": 0.0304, + "step": 10222 + }, + { + "epoch": 4.810823529411765, + "grad_norm": 0.04385978800043969, + "learning_rate": 3.9077465139464446e-05, + "loss": 0.0031, + "step": 10223 + }, + { + "epoch": 4.8112941176470585, + "grad_norm": 0.1299049977025089, + "learning_rate": 3.9047960475473095e-05, + "loss": 0.0317, + "step": 10224 + }, + { + "epoch": 4.811764705882353, + "grad_norm": 2.7312837385952546e-05, + "learning_rate": 3.901846528733011e-05, + "loss": 0.0, + "step": 10225 + }, + { + "epoch": 4.812235294117647, + "grad_norm": 3.0284432323391996e-06, + "learning_rate": 3.898897957755454e-05, + "loss": 0.0, + "step": 10226 + }, + { + "epoch": 4.812705882352941, + "grad_norm": 1.0607970427793172e-05, + "learning_rate": 3.895950334866468e-05, + "loss": 0.0, + "step": 10227 + }, + { + "epoch": 4.813176470588235, + "grad_norm": 0.08864621415209936, + "learning_rate": 3.893003660317783e-05, + "loss": 0.0244, + "step": 10228 + }, + { + "epoch": 4.81364705882353, + "grad_norm": 0.06050407654088297, + "learning_rate": 3.8900579343610646e-05, + "loss": 0.0044, + "step": 10229 + }, + { + "epoch": 4.814117647058824, + "grad_norm": 0.15725800643647173, + "learning_rate": 3.887113157247897e-05, + "loss": 0.0244, + "step": 10230 + }, + { + "epoch": 4.814588235294117, + "grad_norm": 0.17845971517518408, + "learning_rate": 3.884169329229769e-05, + "loss": 0.0274, + "step": 10231 + }, + { + "epoch": 4.815058823529411, + "grad_norm": 0.11755862748872299, + "learning_rate": 3.881226450558104e-05, + "loss": 0.0309, + "step": 10232 + }, + { + "epoch": 4.815529411764706, + "grad_norm": 7.651520658264773e-06, + "learning_rate": 3.878284521484242e-05, + "loss": 0.0, + "step": 10233 + }, + { + "epoch": 4.816, + "grad_norm": 0.11126491577877218, + "learning_rate": 3.8753435422594295e-05, + "loss": 0.02, + "step": 10234 + }, + { + "epoch": 4.816470588235294, + "grad_norm": 0.09969749866307315, + "learning_rate": 3.8724035131348437e-05, + "loss": 0.0132, + "step": 10235 + }, + { + "epoch": 4.816941176470588, + "grad_norm": 0.08118771280524509, + "learning_rate": 3.8694644343615836e-05, + "loss": 0.014, + "step": 10236 + }, + { + "epoch": 4.8174117647058825, + "grad_norm": 0.08371123215555737, + "learning_rate": 3.866526306190651e-05, + "loss": 0.0111, + "step": 10237 + }, + { + "epoch": 4.817882352941177, + "grad_norm": 0.02409735185817624, + "learning_rate": 3.863589128872981e-05, + "loss": 0.0021, + "step": 10238 + }, + { + "epoch": 4.818352941176471, + "grad_norm": 0.18367527562535604, + "learning_rate": 3.8606529026594276e-05, + "loss": 0.0892, + "step": 10239 + }, + { + "epoch": 4.818823529411764, + "grad_norm": 0.13190892619645841, + "learning_rate": 3.85771762780075e-05, + "loss": 0.0298, + "step": 10240 + }, + { + "epoch": 4.8192941176470585, + "grad_norm": 0.05490149727776203, + "learning_rate": 3.854783304547639e-05, + "loss": 0.0031, + "step": 10241 + }, + { + "epoch": 4.819764705882353, + "grad_norm": 0.098714013149239, + "learning_rate": 3.851849933150703e-05, + "loss": 0.026, + "step": 10242 + }, + { + "epoch": 4.820235294117647, + "grad_norm": 0.06393205522830624, + "learning_rate": 3.848917513860463e-05, + "loss": 0.0057, + "step": 10243 + }, + { + "epoch": 4.820705882352941, + "grad_norm": 0.12650824177237413, + "learning_rate": 3.8459860469273666e-05, + "loss": 0.0394, + "step": 10244 + }, + { + "epoch": 4.821176470588235, + "grad_norm": 0.13895838556001394, + "learning_rate": 3.8430555326017684e-05, + "loss": 0.0274, + "step": 10245 + }, + { + "epoch": 4.82164705882353, + "grad_norm": 0.13550915967854463, + "learning_rate": 3.840125971133952e-05, + "loss": 0.0363, + "step": 10246 + }, + { + "epoch": 4.822117647058824, + "grad_norm": 0.029805375997707363, + "learning_rate": 3.8371973627741204e-05, + "loss": 0.003, + "step": 10247 + }, + { + "epoch": 4.822588235294118, + "grad_norm": 0.09613899819403444, + "learning_rate": 3.8342697077723845e-05, + "loss": 0.0115, + "step": 10248 + }, + { + "epoch": 4.823058823529411, + "grad_norm": 0.001180798561662751, + "learning_rate": 3.8313430063787856e-05, + "loss": 0.0, + "step": 10249 + }, + { + "epoch": 4.823529411764706, + "grad_norm": 6.197750609879865e-05, + "learning_rate": 3.8284172588432716e-05, + "loss": 0.0, + "step": 10250 + }, + { + "epoch": 4.824, + "grad_norm": 1.665577532245181e-05, + "learning_rate": 3.8254924654157214e-05, + "loss": 0.0, + "step": 10251 + }, + { + "epoch": 4.824470588235294, + "grad_norm": 0.16459370575960156, + "learning_rate": 3.8225686263459296e-05, + "loss": 0.045, + "step": 10252 + }, + { + "epoch": 4.824941176470588, + "grad_norm": 0.12276849612401404, + "learning_rate": 3.819645741883597e-05, + "loss": 0.0293, + "step": 10253 + }, + { + "epoch": 4.8254117647058825, + "grad_norm": 4.7981682187062135e-06, + "learning_rate": 3.816723812278358e-05, + "loss": 0.0, + "step": 10254 + }, + { + "epoch": 4.825882352941177, + "grad_norm": 0.09949041525349482, + "learning_rate": 3.813802837779762e-05, + "loss": 0.0216, + "step": 10255 + }, + { + "epoch": 4.826352941176471, + "grad_norm": 0.14252197102286462, + "learning_rate": 3.810882818637268e-05, + "loss": 0.0524, + "step": 10256 + }, + { + "epoch": 4.826823529411765, + "grad_norm": 0.08489354996972326, + "learning_rate": 3.807963755100263e-05, + "loss": 0.0155, + "step": 10257 + }, + { + "epoch": 4.8272941176470585, + "grad_norm": 0.10897939987926974, + "learning_rate": 3.8050456474180554e-05, + "loss": 0.0407, + "step": 10258 + }, + { + "epoch": 4.827764705882353, + "grad_norm": 0.1576799243290461, + "learning_rate": 3.802128495839856e-05, + "loss": 0.0296, + "step": 10259 + }, + { + "epoch": 4.828235294117647, + "grad_norm": 0.1918882246764564, + "learning_rate": 3.799212300614809e-05, + "loss": 0.0654, + "step": 10260 + }, + { + "epoch": 4.828705882352941, + "grad_norm": 0.11155668976996394, + "learning_rate": 3.7962970619919745e-05, + "loss": 0.0342, + "step": 10261 + }, + { + "epoch": 4.829176470588235, + "grad_norm": 0.032405198148932154, + "learning_rate": 3.7933827802203207e-05, + "loss": 0.0019, + "step": 10262 + }, + { + "epoch": 4.82964705882353, + "grad_norm": 0.1709385712661745, + "learning_rate": 3.790469455548745e-05, + "loss": 0.0297, + "step": 10263 + }, + { + "epoch": 4.830117647058824, + "grad_norm": 0.12797610527772246, + "learning_rate": 3.787557088226065e-05, + "loss": 0.0478, + "step": 10264 + }, + { + "epoch": 4.830588235294117, + "grad_norm": 0.1310295097929403, + "learning_rate": 3.7846456785010014e-05, + "loss": 0.0183, + "step": 10265 + }, + { + "epoch": 4.831058823529411, + "grad_norm": 0.08494145124090131, + "learning_rate": 3.7817352266222086e-05, + "loss": 0.0173, + "step": 10266 + }, + { + "epoch": 4.831529411764706, + "grad_norm": 0.12189146557230275, + "learning_rate": 3.778825732838257e-05, + "loss": 0.0344, + "step": 10267 + }, + { + "epoch": 4.832, + "grad_norm": 7.473082794237978e-06, + "learning_rate": 3.775917197397622e-05, + "loss": 0.0, + "step": 10268 + }, + { + "epoch": 4.832470588235294, + "grad_norm": 0.13724386808245773, + "learning_rate": 3.773009620548718e-05, + "loss": 0.0376, + "step": 10269 + }, + { + "epoch": 4.832941176470588, + "grad_norm": 0.10060775285063396, + "learning_rate": 3.770103002539857e-05, + "loss": 0.0214, + "step": 10270 + }, + { + "epoch": 4.8334117647058825, + "grad_norm": 0.13343786929458354, + "learning_rate": 3.767197343619279e-05, + "loss": 0.0279, + "step": 10271 + }, + { + "epoch": 4.833882352941177, + "grad_norm": 0.07091777293804812, + "learning_rate": 3.7642926440351514e-05, + "loss": 0.0142, + "step": 10272 + }, + { + "epoch": 4.834352941176471, + "grad_norm": 0.08458576947918496, + "learning_rate": 3.761388904035537e-05, + "loss": 0.0047, + "step": 10273 + }, + { + "epoch": 4.834823529411764, + "grad_norm": 4.0778103839345917e-05, + "learning_rate": 3.758486123868435e-05, + "loss": 0.0, + "step": 10274 + }, + { + "epoch": 4.8352941176470585, + "grad_norm": 0.06762344759876397, + "learning_rate": 3.755584303781763e-05, + "loss": 0.0088, + "step": 10275 + }, + { + "epoch": 4.835764705882353, + "grad_norm": 0.12490154951482897, + "learning_rate": 3.7526834440233395e-05, + "loss": 0.0303, + "step": 10276 + }, + { + "epoch": 4.836235294117647, + "grad_norm": 0.07602751910580945, + "learning_rate": 3.749783544840918e-05, + "loss": 0.0145, + "step": 10277 + }, + { + "epoch": 4.836705882352941, + "grad_norm": 0.16251223822750394, + "learning_rate": 3.746884606482169e-05, + "loss": 0.0452, + "step": 10278 + }, + { + "epoch": 4.837176470588235, + "grad_norm": 0.035754874173517845, + "learning_rate": 3.743986629194666e-05, + "loss": 0.0034, + "step": 10279 + }, + { + "epoch": 4.83764705882353, + "grad_norm": 0.014369868734909048, + "learning_rate": 3.741089613225914e-05, + "loss": 0.0006, + "step": 10280 + }, + { + "epoch": 4.838117647058824, + "grad_norm": 0.0175879373507749, + "learning_rate": 3.7381935588233385e-05, + "loss": 0.0008, + "step": 10281 + }, + { + "epoch": 4.838588235294118, + "grad_norm": 2.565078204384331e-05, + "learning_rate": 3.735298466234269e-05, + "loss": 0.0, + "step": 10282 + }, + { + "epoch": 4.839058823529411, + "grad_norm": 0.07906300017823066, + "learning_rate": 3.732404335705962e-05, + "loss": 0.0083, + "step": 10283 + }, + { + "epoch": 4.839529411764706, + "grad_norm": 0.15528885896861966, + "learning_rate": 3.729511167485597e-05, + "loss": 0.0354, + "step": 10284 + }, + { + "epoch": 4.84, + "grad_norm": 0.10582483169145965, + "learning_rate": 3.72661896182025e-05, + "loss": 0.0139, + "step": 10285 + }, + { + "epoch": 4.840470588235294, + "grad_norm": 0.00023082832886370513, + "learning_rate": 3.7237277189569505e-05, + "loss": 0.0, + "step": 10286 + }, + { + "epoch": 4.840941176470588, + "grad_norm": 0.0997175810437193, + "learning_rate": 3.720837439142608e-05, + "loss": 0.0174, + "step": 10287 + }, + { + "epoch": 4.8414117647058825, + "grad_norm": 0.152368748275123, + "learning_rate": 3.717948122624072e-05, + "loss": 0.0465, + "step": 10288 + }, + { + "epoch": 4.841882352941177, + "grad_norm": 6.060901055644139e-05, + "learning_rate": 3.7150597696481076e-05, + "loss": 0.0, + "step": 10289 + }, + { + "epoch": 4.842352941176471, + "grad_norm": 0.11740605478184965, + "learning_rate": 3.712172380461387e-05, + "loss": 0.0144, + "step": 10290 + }, + { + "epoch": 4.842823529411765, + "grad_norm": 4.033736279884763e-05, + "learning_rate": 3.709285955310513e-05, + "loss": 0.0, + "step": 10291 + }, + { + "epoch": 4.8432941176470585, + "grad_norm": 0.11156309733678092, + "learning_rate": 3.7064004944420026e-05, + "loss": 0.0311, + "step": 10292 + }, + { + "epoch": 4.843764705882353, + "grad_norm": 0.07730560197556201, + "learning_rate": 3.703515998102281e-05, + "loss": 0.0174, + "step": 10293 + }, + { + "epoch": 4.844235294117647, + "grad_norm": 0.09745486721103164, + "learning_rate": 3.7006324665377025e-05, + "loss": 0.0163, + "step": 10294 + }, + { + "epoch": 4.844705882352941, + "grad_norm": 8.404028159695578e-06, + "learning_rate": 3.6977498999945384e-05, + "loss": 0.0, + "step": 10295 + }, + { + "epoch": 4.845176470588235, + "grad_norm": 0.09906480338787768, + "learning_rate": 3.694868298718965e-05, + "loss": 0.0156, + "step": 10296 + }, + { + "epoch": 4.84564705882353, + "grad_norm": 0.1218064042281058, + "learning_rate": 3.6919876629570905e-05, + "loss": 0.0175, + "step": 10297 + }, + { + "epoch": 4.846117647058824, + "grad_norm": 0.07212452935129147, + "learning_rate": 3.68910799295494e-05, + "loss": 0.0099, + "step": 10298 + }, + { + "epoch": 4.846588235294117, + "grad_norm": 0.15035003099939928, + "learning_rate": 3.686229288958442e-05, + "loss": 0.0296, + "step": 10299 + }, + { + "epoch": 4.847058823529411, + "grad_norm": 4.648586331333891e-05, + "learning_rate": 3.6833515512134606e-05, + "loss": 0.0, + "step": 10300 + }, + { + "epoch": 4.847529411764706, + "grad_norm": 0.14468066063972043, + "learning_rate": 3.6804747799657594e-05, + "loss": 0.0408, + "step": 10301 + }, + { + "epoch": 4.848, + "grad_norm": 1.5821261613321057e-05, + "learning_rate": 3.677598975461036e-05, + "loss": 0.0, + "step": 10302 + }, + { + "epoch": 4.848470588235294, + "grad_norm": 0.031633073972811766, + "learning_rate": 3.6747241379448986e-05, + "loss": 0.002, + "step": 10303 + }, + { + "epoch": 4.848941176470588, + "grad_norm": 0.1659888495130976, + "learning_rate": 3.671850267662866e-05, + "loss": 0.0511, + "step": 10304 + }, + { + "epoch": 4.8494117647058825, + "grad_norm": 1.9135680334626034e-06, + "learning_rate": 3.668977364860386e-05, + "loss": 0.0, + "step": 10305 + }, + { + "epoch": 4.849882352941177, + "grad_norm": 0.10381437251927642, + "learning_rate": 3.666105429782821e-05, + "loss": 0.0246, + "step": 10306 + }, + { + "epoch": 4.850352941176471, + "grad_norm": 2.252922918773994e-05, + "learning_rate": 3.6632344626754405e-05, + "loss": 0.0, + "step": 10307 + }, + { + "epoch": 4.850823529411764, + "grad_norm": 0.13330499647816194, + "learning_rate": 3.660364463783443e-05, + "loss": 0.0194, + "step": 10308 + }, + { + "epoch": 4.8512941176470585, + "grad_norm": 0.2772038838177635, + "learning_rate": 3.657495433351945e-05, + "loss": 0.0531, + "step": 10309 + }, + { + "epoch": 4.851764705882353, + "grad_norm": 9.892217156614427e-06, + "learning_rate": 3.6546273716259685e-05, + "loss": 0.0, + "step": 10310 + }, + { + "epoch": 4.852235294117647, + "grad_norm": 0.11563921543710133, + "learning_rate": 3.6517602788504613e-05, + "loss": 0.0303, + "step": 10311 + }, + { + "epoch": 4.852705882352941, + "grad_norm": 0.05519240070349334, + "learning_rate": 3.648894155270295e-05, + "loss": 0.0063, + "step": 10312 + }, + { + "epoch": 4.853176470588235, + "grad_norm": 0.10712832095445796, + "learning_rate": 3.64602900113024e-05, + "loss": 0.0073, + "step": 10313 + }, + { + "epoch": 4.85364705882353, + "grad_norm": 0.2822340884606282, + "learning_rate": 3.643164816674999e-05, + "loss": 0.0486, + "step": 10314 + }, + { + "epoch": 4.854117647058824, + "grad_norm": 0.12706328268039324, + "learning_rate": 3.6403016021491924e-05, + "loss": 0.0298, + "step": 10315 + }, + { + "epoch": 4.854588235294118, + "grad_norm": 0.007569546526435212, + "learning_rate": 3.637439357797343e-05, + "loss": 0.0003, + "step": 10316 + }, + { + "epoch": 4.855058823529411, + "grad_norm": 0.0696095975780951, + "learning_rate": 3.634578083863905e-05, + "loss": 0.01, + "step": 10317 + }, + { + "epoch": 4.855529411764706, + "grad_norm": 0.10097400191034481, + "learning_rate": 3.631717780593252e-05, + "loss": 0.0193, + "step": 10318 + }, + { + "epoch": 4.856, + "grad_norm": 2.1779055090149483e-06, + "learning_rate": 3.628858448229655e-05, + "loss": 0.0, + "step": 10319 + }, + { + "epoch": 4.856470588235294, + "grad_norm": 0.17775811902402158, + "learning_rate": 3.6260000870173266e-05, + "loss": 0.069, + "step": 10320 + }, + { + "epoch": 4.856941176470588, + "grad_norm": 4.264081725800815e-05, + "learning_rate": 3.6231426972003766e-05, + "loss": 0.0, + "step": 10321 + }, + { + "epoch": 4.8574117647058825, + "grad_norm": 4.5787262055508365e-06, + "learning_rate": 3.620286279022843e-05, + "loss": 0.0, + "step": 10322 + }, + { + "epoch": 4.857882352941177, + "grad_norm": 0.12008469020882614, + "learning_rate": 3.617430832728682e-05, + "loss": 0.0243, + "step": 10323 + }, + { + "epoch": 4.85835294117647, + "grad_norm": 0.13905658970057305, + "learning_rate": 3.614576358561757e-05, + "loss": 0.0532, + "step": 10324 + }, + { + "epoch": 4.858823529411764, + "grad_norm": 0.0002800482373042885, + "learning_rate": 3.611722856765855e-05, + "loss": 0.0, + "step": 10325 + }, + { + "epoch": 4.8592941176470585, + "grad_norm": 0.18083381718618013, + "learning_rate": 3.608870327584683e-05, + "loss": 0.0623, + "step": 10326 + }, + { + "epoch": 4.859764705882353, + "grad_norm": 0.19816352217892086, + "learning_rate": 3.606018771261856e-05, + "loss": 0.0658, + "step": 10327 + }, + { + "epoch": 4.860235294117647, + "grad_norm": 0.11631623224684096, + "learning_rate": 3.6031681880409106e-05, + "loss": 0.0292, + "step": 10328 + }, + { + "epoch": 4.860705882352941, + "grad_norm": 0.13076207140933946, + "learning_rate": 3.6003185781653097e-05, + "loss": 0.0425, + "step": 10329 + }, + { + "epoch": 4.861176470588235, + "grad_norm": 2.7553849576008545e-06, + "learning_rate": 3.5974699418784084e-05, + "loss": 0.0, + "step": 10330 + }, + { + "epoch": 4.86164705882353, + "grad_norm": 0.0475964042531947, + "learning_rate": 3.5946222794235105e-05, + "loss": 0.0082, + "step": 10331 + }, + { + "epoch": 4.862117647058824, + "grad_norm": 0.18781249102660527, + "learning_rate": 3.5917755910438094e-05, + "loss": 0.0607, + "step": 10332 + }, + { + "epoch": 4.862588235294117, + "grad_norm": 0.16507003560987932, + "learning_rate": 3.58892987698243e-05, + "loss": 0.0405, + "step": 10333 + }, + { + "epoch": 4.863058823529411, + "grad_norm": 0.10000465398390536, + "learning_rate": 3.5860851374824134e-05, + "loss": 0.0228, + "step": 10334 + }, + { + "epoch": 4.863529411764706, + "grad_norm": 0.12262728694970712, + "learning_rate": 3.5832413727867075e-05, + "loss": 0.0269, + "step": 10335 + }, + { + "epoch": 4.864, + "grad_norm": 0.04727464156168676, + "learning_rate": 3.580398583138188e-05, + "loss": 0.0068, + "step": 10336 + }, + { + "epoch": 4.864470588235294, + "grad_norm": 7.26988420108291e-06, + "learning_rate": 3.577556768779646e-05, + "loss": 0.0, + "step": 10337 + }, + { + "epoch": 4.864941176470588, + "grad_norm": 0.08262699698292982, + "learning_rate": 3.574715929953778e-05, + "loss": 0.0122, + "step": 10338 + }, + { + "epoch": 4.8654117647058825, + "grad_norm": 0.00042669081047075043, + "learning_rate": 3.571876066903211e-05, + "loss": 0.0, + "step": 10339 + }, + { + "epoch": 4.865882352941177, + "grad_norm": 1.3437378287236289e-06, + "learning_rate": 3.569037179870487e-05, + "loss": 0.0, + "step": 10340 + }, + { + "epoch": 4.866352941176471, + "grad_norm": 0.20867229577688243, + "learning_rate": 3.5661992690980506e-05, + "loss": 0.0464, + "step": 10341 + }, + { + "epoch": 4.866823529411764, + "grad_norm": 0.13828463806678354, + "learning_rate": 3.56336233482828e-05, + "loss": 0.0252, + "step": 10342 + }, + { + "epoch": 4.8672941176470585, + "grad_norm": 0.03694878586834459, + "learning_rate": 3.560526377303467e-05, + "loss": 0.0032, + "step": 10343 + }, + { + "epoch": 4.867764705882353, + "grad_norm": 0.16124208405274432, + "learning_rate": 3.557691396765807e-05, + "loss": 0.0166, + "step": 10344 + }, + { + "epoch": 4.868235294117647, + "grad_norm": 0.12132529168117104, + "learning_rate": 3.554857393457426e-05, + "loss": 0.0201, + "step": 10345 + }, + { + "epoch": 4.868705882352941, + "grad_norm": 0.0007978264571603653, + "learning_rate": 3.5520243676203664e-05, + "loss": 0.0, + "step": 10346 + }, + { + "epoch": 4.869176470588235, + "grad_norm": 5.824112835470439e-05, + "learning_rate": 3.5491923194965735e-05, + "loss": 0.0, + "step": 10347 + }, + { + "epoch": 4.86964705882353, + "grad_norm": 0.000123827713067681, + "learning_rate": 3.546361249327922e-05, + "loss": 0.0, + "step": 10348 + }, + { + "epoch": 4.870117647058824, + "grad_norm": 0.15362467280275166, + "learning_rate": 3.5435311573562024e-05, + "loss": 0.0278, + "step": 10349 + }, + { + "epoch": 4.870588235294118, + "grad_norm": 1.2998953205679534e-06, + "learning_rate": 3.540702043823113e-05, + "loss": 0.0, + "step": 10350 + }, + { + "epoch": 4.871058823529411, + "grad_norm": 0.27128036442934994, + "learning_rate": 3.537873908970282e-05, + "loss": 0.0719, + "step": 10351 + }, + { + "epoch": 4.871529411764706, + "grad_norm": 0.09044302456817248, + "learning_rate": 3.535046753039235e-05, + "loss": 0.019, + "step": 10352 + }, + { + "epoch": 4.872, + "grad_norm": 0.12358503309526972, + "learning_rate": 3.53222057627143e-05, + "loss": 0.0204, + "step": 10353 + }, + { + "epoch": 4.872470588235294, + "grad_norm": 0.09710453579864192, + "learning_rate": 3.5293953789082425e-05, + "loss": 0.018, + "step": 10354 + }, + { + "epoch": 4.872941176470588, + "grad_norm": 0.13369258909639384, + "learning_rate": 3.5265711611909476e-05, + "loss": 0.0463, + "step": 10355 + }, + { + "epoch": 4.8734117647058826, + "grad_norm": 7.645722541652724e-05, + "learning_rate": 3.5237479233607525e-05, + "loss": 0.0, + "step": 10356 + }, + { + "epoch": 4.873882352941177, + "grad_norm": 0.12430440154655048, + "learning_rate": 3.520925665658779e-05, + "loss": 0.0394, + "step": 10357 + }, + { + "epoch": 4.87435294117647, + "grad_norm": 0.12178631108203931, + "learning_rate": 3.518104388326055e-05, + "loss": 0.0261, + "step": 10358 + }, + { + "epoch": 4.874823529411764, + "grad_norm": 0.0869731480766004, + "learning_rate": 3.5152840916035344e-05, + "loss": 0.0126, + "step": 10359 + }, + { + "epoch": 4.875294117647059, + "grad_norm": 0.16392984124190635, + "learning_rate": 3.5124647757320883e-05, + "loss": 0.0274, + "step": 10360 + }, + { + "epoch": 4.875764705882353, + "grad_norm": 2.5236761753029855e-06, + "learning_rate": 3.509646440952492e-05, + "loss": 0.0, + "step": 10361 + }, + { + "epoch": 4.876235294117647, + "grad_norm": 0.08895226441910709, + "learning_rate": 3.506829087505451e-05, + "loss": 0.0194, + "step": 10362 + }, + { + "epoch": 4.876705882352941, + "grad_norm": 0.10607478736461708, + "learning_rate": 3.504012715631583e-05, + "loss": 0.0154, + "step": 10363 + }, + { + "epoch": 4.8771764705882354, + "grad_norm": 0.1159911907665075, + "learning_rate": 3.501197325571411e-05, + "loss": 0.0244, + "step": 10364 + }, + { + "epoch": 4.87764705882353, + "grad_norm": 0.10125105151269301, + "learning_rate": 3.49838291756539e-05, + "loss": 0.021, + "step": 10365 + }, + { + "epoch": 4.878117647058824, + "grad_norm": 0.2431569676686514, + "learning_rate": 3.495569491853887e-05, + "loss": 0.0758, + "step": 10366 + }, + { + "epoch": 4.878588235294117, + "grad_norm": 6.7830874093196676e-06, + "learning_rate": 3.492757048677175e-05, + "loss": 0.0, + "step": 10367 + }, + { + "epoch": 4.8790588235294114, + "grad_norm": 0.06454861739964095, + "learning_rate": 3.4899455882754526e-05, + "loss": 0.0091, + "step": 10368 + }, + { + "epoch": 4.879529411764706, + "grad_norm": 0.1818914056335886, + "learning_rate": 3.4871351108888375e-05, + "loss": 0.0373, + "step": 10369 + }, + { + "epoch": 4.88, + "grad_norm": 5.047829946019108e-06, + "learning_rate": 3.4843256167573514e-05, + "loss": 0.0, + "step": 10370 + }, + { + "epoch": 4.880470588235294, + "grad_norm": 0.16816555256199558, + "learning_rate": 3.481517106120944e-05, + "loss": 0.0498, + "step": 10371 + }, + { + "epoch": 4.880941176470588, + "grad_norm": 0.04281470758459172, + "learning_rate": 3.478709579219472e-05, + "loss": 0.0027, + "step": 10372 + }, + { + "epoch": 4.881411764705883, + "grad_norm": 0.11850394066868976, + "learning_rate": 3.4759030362927115e-05, + "loss": 0.0376, + "step": 10373 + }, + { + "epoch": 4.881882352941177, + "grad_norm": 0.09974748010730008, + "learning_rate": 3.4730974775803574e-05, + "loss": 0.0197, + "step": 10374 + }, + { + "epoch": 4.882352941176471, + "grad_norm": 0.05570911499420847, + "learning_rate": 3.4702929033220174e-05, + "loss": 0.0022, + "step": 10375 + }, + { + "epoch": 4.882823529411764, + "grad_norm": 0.09107197468252615, + "learning_rate": 3.467489313757217e-05, + "loss": 0.0134, + "step": 10376 + }, + { + "epoch": 4.883294117647059, + "grad_norm": 0.1346073314984889, + "learning_rate": 3.4646867091254e-05, + "loss": 0.0264, + "step": 10377 + }, + { + "epoch": 4.883764705882353, + "grad_norm": 0.09755638292821558, + "learning_rate": 3.4618850896659135e-05, + "loss": 0.016, + "step": 10378 + }, + { + "epoch": 4.884235294117647, + "grad_norm": 0.16531655148784394, + "learning_rate": 3.4590844556180395e-05, + "loss": 0.0518, + "step": 10379 + }, + { + "epoch": 4.884705882352941, + "grad_norm": 0.04406392737547115, + "learning_rate": 3.456284807220958e-05, + "loss": 0.0027, + "step": 10380 + }, + { + "epoch": 4.8851764705882355, + "grad_norm": 0.14244802680228386, + "learning_rate": 3.453486144713772e-05, + "loss": 0.0344, + "step": 10381 + }, + { + "epoch": 4.88564705882353, + "grad_norm": 0.023573678094529517, + "learning_rate": 3.450688468335512e-05, + "loss": 0.0035, + "step": 10382 + }, + { + "epoch": 4.886117647058824, + "grad_norm": 0.10976695833602468, + "learning_rate": 3.4478917783251e-05, + "loss": 0.0305, + "step": 10383 + }, + { + "epoch": 4.886588235294118, + "grad_norm": 0.10745468427018545, + "learning_rate": 3.445096074921393e-05, + "loss": 0.0149, + "step": 10384 + }, + { + "epoch": 4.8870588235294115, + "grad_norm": 0.11108962570002195, + "learning_rate": 3.442301358363163e-05, + "loss": 0.0341, + "step": 10385 + }, + { + "epoch": 4.887529411764706, + "grad_norm": 7.153789439657679e-06, + "learning_rate": 3.4395076288890825e-05, + "loss": 0.0, + "step": 10386 + }, + { + "epoch": 4.888, + "grad_norm": 0.11590244645638344, + "learning_rate": 3.436714886737753e-05, + "loss": 0.0259, + "step": 10387 + }, + { + "epoch": 4.888470588235294, + "grad_norm": 0.2010868142382106, + "learning_rate": 3.4339231321476965e-05, + "loss": 0.0454, + "step": 10388 + }, + { + "epoch": 4.888941176470588, + "grad_norm": 0.00013150275890427607, + "learning_rate": 3.43113236535733e-05, + "loss": 0.0, + "step": 10389 + }, + { + "epoch": 4.889411764705883, + "grad_norm": 0.11466379715898938, + "learning_rate": 3.428342586605005e-05, + "loss": 0.0152, + "step": 10390 + }, + { + "epoch": 4.889882352941177, + "grad_norm": 8.409556900800826e-06, + "learning_rate": 3.425553796128985e-05, + "loss": 0.0, + "step": 10391 + }, + { + "epoch": 4.89035294117647, + "grad_norm": 1.60369513640021e-05, + "learning_rate": 3.42276599416744e-05, + "loss": 0.0, + "step": 10392 + }, + { + "epoch": 4.890823529411764, + "grad_norm": 0.02951714518409482, + "learning_rate": 3.419979180958464e-05, + "loss": 0.0026, + "step": 10393 + }, + { + "epoch": 4.891294117647059, + "grad_norm": 0.12509257683376615, + "learning_rate": 3.417193356740071e-05, + "loss": 0.0172, + "step": 10394 + }, + { + "epoch": 4.891764705882353, + "grad_norm": 0.08387309485168265, + "learning_rate": 3.414408521750173e-05, + "loss": 0.0077, + "step": 10395 + }, + { + "epoch": 4.892235294117647, + "grad_norm": 0.17301829171754757, + "learning_rate": 3.411624676226616e-05, + "loss": 0.031, + "step": 10396 + }, + { + "epoch": 4.892705882352941, + "grad_norm": 1.2851008105479395e-05, + "learning_rate": 3.408841820407157e-05, + "loss": 0.0, + "step": 10397 + }, + { + "epoch": 4.8931764705882355, + "grad_norm": 0.00012823932722985656, + "learning_rate": 3.406059954529454e-05, + "loss": 0.0, + "step": 10398 + }, + { + "epoch": 4.89364705882353, + "grad_norm": 3.916855359443606e-06, + "learning_rate": 3.403279078831101e-05, + "loss": 0.0, + "step": 10399 + }, + { + "epoch": 4.894117647058824, + "grad_norm": 0.17503620704145595, + "learning_rate": 3.4004991935496004e-05, + "loss": 0.0582, + "step": 10400 + }, + { + "epoch": 4.894588235294117, + "grad_norm": 0.0410722745335093, + "learning_rate": 3.397720298922358e-05, + "loss": 0.0022, + "step": 10401 + }, + { + "epoch": 4.8950588235294115, + "grad_norm": 0.16226360674986096, + "learning_rate": 3.394942395186716e-05, + "loss": 0.0615, + "step": 10402 + }, + { + "epoch": 4.895529411764706, + "grad_norm": 0.0823363798717234, + "learning_rate": 3.392165482579912e-05, + "loss": 0.0106, + "step": 10403 + }, + { + "epoch": 4.896, + "grad_norm": 0.18267137740090575, + "learning_rate": 3.389389561339112e-05, + "loss": 0.0526, + "step": 10404 + }, + { + "epoch": 4.896470588235294, + "grad_norm": 0.09653736573697175, + "learning_rate": 3.3866146317013976e-05, + "loss": 0.0209, + "step": 10405 + }, + { + "epoch": 4.896941176470588, + "grad_norm": 0.14613351388276125, + "learning_rate": 3.3838406939037525e-05, + "loss": 0.0504, + "step": 10406 + }, + { + "epoch": 4.897411764705883, + "grad_norm": 0.1028451954224322, + "learning_rate": 3.381067748183087e-05, + "loss": 0.0337, + "step": 10407 + }, + { + "epoch": 4.897882352941177, + "grad_norm": 0.22865794461529898, + "learning_rate": 3.3782957947762336e-05, + "loss": 0.0814, + "step": 10408 + }, + { + "epoch": 4.898352941176471, + "grad_norm": 0.028984604248949688, + "learning_rate": 3.375524833919918e-05, + "loss": 0.0019, + "step": 10409 + }, + { + "epoch": 4.898823529411764, + "grad_norm": 0.07043243825927598, + "learning_rate": 3.372754865850799e-05, + "loss": 0.009, + "step": 10410 + }, + { + "epoch": 4.899294117647059, + "grad_norm": 1.5376731862475173e-06, + "learning_rate": 3.3699858908054506e-05, + "loss": 0.0, + "step": 10411 + }, + { + "epoch": 4.899764705882353, + "grad_norm": 0.1270294299749701, + "learning_rate": 3.367217909020348e-05, + "loss": 0.0358, + "step": 10412 + }, + { + "epoch": 4.900235294117647, + "grad_norm": 0.09446011349327631, + "learning_rate": 3.364450920731894e-05, + "loss": 0.0113, + "step": 10413 + }, + { + "epoch": 4.900705882352941, + "grad_norm": 0.1879087379005086, + "learning_rate": 3.361684926176408e-05, + "loss": 0.0414, + "step": 10414 + }, + { + "epoch": 4.9011764705882355, + "grad_norm": 0.22568779983381085, + "learning_rate": 3.35891992559011e-05, + "loss": 0.051, + "step": 10415 + }, + { + "epoch": 4.90164705882353, + "grad_norm": 2.327397188310788e-06, + "learning_rate": 3.35615591920915e-05, + "loss": 0.0, + "step": 10416 + }, + { + "epoch": 4.902117647058824, + "grad_norm": 0.09050760642762168, + "learning_rate": 3.353392907269591e-05, + "loss": 0.0219, + "step": 10417 + }, + { + "epoch": 4.902588235294118, + "grad_norm": 0.14694071202978487, + "learning_rate": 3.3506308900073964e-05, + "loss": 0.0315, + "step": 10418 + }, + { + "epoch": 4.9030588235294115, + "grad_norm": 0.10327233656258256, + "learning_rate": 3.3478698676584714e-05, + "loss": 0.021, + "step": 10419 + }, + { + "epoch": 4.903529411764706, + "grad_norm": 0.0548677861477486, + "learning_rate": 3.34510984045861e-05, + "loss": 0.0074, + "step": 10420 + }, + { + "epoch": 4.904, + "grad_norm": 0.08344093153658677, + "learning_rate": 3.3423508086435326e-05, + "loss": 0.014, + "step": 10421 + }, + { + "epoch": 4.904470588235294, + "grad_norm": 0.08107585551135926, + "learning_rate": 3.339592772448883e-05, + "loss": 0.0163, + "step": 10422 + }, + { + "epoch": 4.904941176470588, + "grad_norm": 2.7283258737966883e-05, + "learning_rate": 3.3368357321102e-05, + "loss": 0.0, + "step": 10423 + }, + { + "epoch": 4.905411764705883, + "grad_norm": 1.0032362324268588e-05, + "learning_rate": 3.334079687862952e-05, + "loss": 0.0, + "step": 10424 + }, + { + "epoch": 4.905882352941177, + "grad_norm": 0.12632692253277925, + "learning_rate": 3.331324639942526e-05, + "loss": 0.0275, + "step": 10425 + }, + { + "epoch": 4.90635294117647, + "grad_norm": 0.08315960142312574, + "learning_rate": 3.328570588584204e-05, + "loss": 0.013, + "step": 10426 + }, + { + "epoch": 4.906823529411764, + "grad_norm": 0.04104523871645813, + "learning_rate": 3.325817534023202e-05, + "loss": 0.0022, + "step": 10427 + }, + { + "epoch": 4.907294117647059, + "grad_norm": 3.317124855855208e-05, + "learning_rate": 3.32306547649465e-05, + "loss": 0.0, + "step": 10428 + }, + { + "epoch": 4.907764705882353, + "grad_norm": 0.16987302276221908, + "learning_rate": 3.320314416233574e-05, + "loss": 0.0276, + "step": 10429 + }, + { + "epoch": 4.908235294117647, + "grad_norm": 0.13047583250462333, + "learning_rate": 3.317564353474943e-05, + "loss": 0.0173, + "step": 10430 + }, + { + "epoch": 4.908705882352941, + "grad_norm": 0.11662966419890723, + "learning_rate": 3.314815288453613e-05, + "loss": 0.0255, + "step": 10431 + }, + { + "epoch": 4.9091764705882355, + "grad_norm": 0.11208092344402829, + "learning_rate": 3.312067221404371e-05, + "loss": 0.0314, + "step": 10432 + }, + { + "epoch": 4.90964705882353, + "grad_norm": 0.08840272330901543, + "learning_rate": 3.309320152561923e-05, + "loss": 0.0243, + "step": 10433 + }, + { + "epoch": 4.910117647058824, + "grad_norm": 0.11116827556448405, + "learning_rate": 3.3065740821608725e-05, + "loss": 0.017, + "step": 10434 + }, + { + "epoch": 4.910588235294117, + "grad_norm": 0.16009617818022798, + "learning_rate": 3.303829010435751e-05, + "loss": 0.0408, + "step": 10435 + }, + { + "epoch": 4.9110588235294115, + "grad_norm": 0.12409173803489607, + "learning_rate": 3.3010849376210056e-05, + "loss": 0.0378, + "step": 10436 + }, + { + "epoch": 4.911529411764706, + "grad_norm": 0.08985971194589255, + "learning_rate": 3.298341863950986e-05, + "loss": 0.0186, + "step": 10437 + }, + { + "epoch": 4.912, + "grad_norm": 0.15289390630958896, + "learning_rate": 3.295599789659967e-05, + "loss": 0.0419, + "step": 10438 + }, + { + "epoch": 4.912470588235294, + "grad_norm": 0.11290438255215623, + "learning_rate": 3.2928587149821385e-05, + "loss": 0.0314, + "step": 10439 + }, + { + "epoch": 4.912941176470588, + "grad_norm": 0.11334690618059183, + "learning_rate": 3.290118640151597e-05, + "loss": 0.0244, + "step": 10440 + }, + { + "epoch": 4.913411764705883, + "grad_norm": 0.14298267582754126, + "learning_rate": 3.287379565402358e-05, + "loss": 0.0227, + "step": 10441 + }, + { + "epoch": 4.913882352941177, + "grad_norm": 0.10801428701480376, + "learning_rate": 3.284641490968359e-05, + "loss": 0.0213, + "step": 10442 + }, + { + "epoch": 4.914352941176471, + "grad_norm": 0.16468887567589396, + "learning_rate": 3.281904417083436e-05, + "loss": 0.0209, + "step": 10443 + }, + { + "epoch": 4.914823529411764, + "grad_norm": 0.05469719106739956, + "learning_rate": 3.279168343981355e-05, + "loss": 0.0082, + "step": 10444 + }, + { + "epoch": 4.915294117647059, + "grad_norm": 0.09512779331548402, + "learning_rate": 3.2764332718957894e-05, + "loss": 0.0164, + "step": 10445 + }, + { + "epoch": 4.915764705882353, + "grad_norm": 0.053267444623416184, + "learning_rate": 3.2736992010603226e-05, + "loss": 0.0039, + "step": 10446 + }, + { + "epoch": 4.916235294117647, + "grad_norm": 0.14385421368918908, + "learning_rate": 3.270966131708461e-05, + "loss": 0.0465, + "step": 10447 + }, + { + "epoch": 4.916705882352941, + "grad_norm": 0.06791471401000453, + "learning_rate": 3.268234064073628e-05, + "loss": 0.01, + "step": 10448 + }, + { + "epoch": 4.9171764705882355, + "grad_norm": 0.0764572677102827, + "learning_rate": 3.265502998389143e-05, + "loss": 0.0067, + "step": 10449 + }, + { + "epoch": 4.91764705882353, + "grad_norm": 0.11436017817786724, + "learning_rate": 3.262772934888265e-05, + "loss": 0.0278, + "step": 10450 + }, + { + "epoch": 4.918117647058823, + "grad_norm": 0.11644920221240466, + "learning_rate": 3.260043873804147e-05, + "loss": 0.0074, + "step": 10451 + }, + { + "epoch": 4.918588235294117, + "grad_norm": 0.011609015097374097, + "learning_rate": 3.257315815369863e-05, + "loss": 0.0003, + "step": 10452 + }, + { + "epoch": 4.9190588235294115, + "grad_norm": 0.10969935549492763, + "learning_rate": 3.2545887598184114e-05, + "loss": 0.0212, + "step": 10453 + }, + { + "epoch": 4.919529411764706, + "grad_norm": 0.16192058859842415, + "learning_rate": 3.251862707382686e-05, + "loss": 0.0238, + "step": 10454 + }, + { + "epoch": 4.92, + "grad_norm": 0.1538158583952233, + "learning_rate": 3.249137658295511e-05, + "loss": 0.0368, + "step": 10455 + }, + { + "epoch": 4.920470588235294, + "grad_norm": 0.0959211807696299, + "learning_rate": 3.246413612789622e-05, + "loss": 0.0245, + "step": 10456 + }, + { + "epoch": 4.920941176470588, + "grad_norm": 0.059855982742581734, + "learning_rate": 3.243690571097656e-05, + "loss": 0.0125, + "step": 10457 + }, + { + "epoch": 4.921411764705883, + "grad_norm": 0.029396887071707196, + "learning_rate": 3.24096853345218e-05, + "loss": 0.003, + "step": 10458 + }, + { + "epoch": 4.921882352941177, + "grad_norm": 0.00010014012644889114, + "learning_rate": 3.2382475000856725e-05, + "loss": 0.0, + "step": 10459 + }, + { + "epoch": 4.92235294117647, + "grad_norm": 0.08427675158493271, + "learning_rate": 3.2355274712305154e-05, + "loss": 0.0152, + "step": 10460 + }, + { + "epoch": 4.922823529411764, + "grad_norm": 0.06348631742199354, + "learning_rate": 3.232808447119018e-05, + "loss": 0.0102, + "step": 10461 + }, + { + "epoch": 4.923294117647059, + "grad_norm": 0.10608670131412516, + "learning_rate": 3.2300904279834e-05, + "loss": 0.0265, + "step": 10462 + }, + { + "epoch": 4.923764705882353, + "grad_norm": 0.1000180884764378, + "learning_rate": 3.227373414055782e-05, + "loss": 0.0066, + "step": 10463 + }, + { + "epoch": 4.924235294117647, + "grad_norm": 0.10958184148788812, + "learning_rate": 3.224657405568229e-05, + "loss": 0.0137, + "step": 10464 + }, + { + "epoch": 4.924705882352941, + "grad_norm": 0.04368344300313554, + "learning_rate": 3.221942402752685e-05, + "loss": 0.0027, + "step": 10465 + }, + { + "epoch": 4.9251764705882355, + "grad_norm": 0.03128838928275174, + "learning_rate": 3.2192284058410324e-05, + "loss": 0.0022, + "step": 10466 + }, + { + "epoch": 4.92564705882353, + "grad_norm": 0.10851473885891548, + "learning_rate": 3.216515415065061e-05, + "loss": 0.0153, + "step": 10467 + }, + { + "epoch": 4.926117647058824, + "grad_norm": 0.13195587029703018, + "learning_rate": 3.213803430656468e-05, + "loss": 0.0272, + "step": 10468 + }, + { + "epoch": 4.926588235294117, + "grad_norm": 0.09328555967944144, + "learning_rate": 3.2110924528468725e-05, + "loss": 0.0128, + "step": 10469 + }, + { + "epoch": 4.9270588235294115, + "grad_norm": 0.09753765712894293, + "learning_rate": 3.208382481867809e-05, + "loss": 0.0137, + "step": 10470 + }, + { + "epoch": 4.927529411764706, + "grad_norm": 0.0840619312157088, + "learning_rate": 3.205673517950716e-05, + "loss": 0.0091, + "step": 10471 + }, + { + "epoch": 4.928, + "grad_norm": 6.545830107739793e-05, + "learning_rate": 3.202965561326955e-05, + "loss": 0.0, + "step": 10472 + }, + { + "epoch": 4.928470588235294, + "grad_norm": 0.06532609879598224, + "learning_rate": 3.200258612227803e-05, + "loss": 0.0048, + "step": 10473 + }, + { + "epoch": 4.928941176470588, + "grad_norm": 0.061927495848923306, + "learning_rate": 3.19755267088444e-05, + "loss": 0.0044, + "step": 10474 + }, + { + "epoch": 4.929411764705883, + "grad_norm": 0.060290326260846135, + "learning_rate": 3.194847737527969e-05, + "loss": 0.0073, + "step": 10475 + }, + { + "epoch": 4.929882352941177, + "grad_norm": 0.10457462003526038, + "learning_rate": 3.1921438123894086e-05, + "loss": 0.0284, + "step": 10476 + }, + { + "epoch": 4.930352941176471, + "grad_norm": 0.1791360143599475, + "learning_rate": 3.189440895699681e-05, + "loss": 0.0575, + "step": 10477 + }, + { + "epoch": 4.930823529411764, + "grad_norm": 0.12737961250211496, + "learning_rate": 3.1867389876896296e-05, + "loss": 0.0171, + "step": 10478 + }, + { + "epoch": 4.931294117647059, + "grad_norm": 0.09991690979999014, + "learning_rate": 3.184038088590019e-05, + "loss": 0.0071, + "step": 10479 + }, + { + "epoch": 4.931764705882353, + "grad_norm": 0.38016777475863905, + "learning_rate": 3.181338198631506e-05, + "loss": 0.0892, + "step": 10480 + }, + { + "epoch": 4.932235294117647, + "grad_norm": 0.10712366474953268, + "learning_rate": 3.178639318044687e-05, + "loss": 0.0278, + "step": 10481 + }, + { + "epoch": 4.932705882352941, + "grad_norm": 0.036353666073811054, + "learning_rate": 3.175941447060048e-05, + "loss": 0.0032, + "step": 10482 + }, + { + "epoch": 4.9331764705882355, + "grad_norm": 0.23669238808270754, + "learning_rate": 3.173244585908007e-05, + "loss": 0.0668, + "step": 10483 + }, + { + "epoch": 4.93364705882353, + "grad_norm": 0.12113237604790146, + "learning_rate": 3.170548734818893e-05, + "loss": 0.0351, + "step": 10484 + }, + { + "epoch": 4.934117647058823, + "grad_norm": 0.0001273448666570605, + "learning_rate": 3.1678538940229356e-05, + "loss": 0.0, + "step": 10485 + }, + { + "epoch": 4.934588235294117, + "grad_norm": 0.17814999004067178, + "learning_rate": 3.16516006375029e-05, + "loss": 0.0616, + "step": 10486 + }, + { + "epoch": 4.9350588235294115, + "grad_norm": 0.14647545823638697, + "learning_rate": 3.1624672442310304e-05, + "loss": 0.0482, + "step": 10487 + }, + { + "epoch": 4.935529411764706, + "grad_norm": 5.069210164631016e-06, + "learning_rate": 3.159775435695126e-05, + "loss": 0.0, + "step": 10488 + }, + { + "epoch": 4.936, + "grad_norm": 0.09078959614853437, + "learning_rate": 3.1570846383724775e-05, + "loss": 0.0284, + "step": 10489 + }, + { + "epoch": 4.936470588235294, + "grad_norm": 0.1309688144034285, + "learning_rate": 3.154394852492892e-05, + "loss": 0.0369, + "step": 10490 + }, + { + "epoch": 4.936941176470588, + "grad_norm": 1.9286695106432918e-05, + "learning_rate": 3.1517060782860865e-05, + "loss": 0.0, + "step": 10491 + }, + { + "epoch": 4.937411764705883, + "grad_norm": 0.13611747955073195, + "learning_rate": 3.1490183159816956e-05, + "loss": 0.0414, + "step": 10492 + }, + { + "epoch": 4.937882352941177, + "grad_norm": 0.07638465082009938, + "learning_rate": 3.146331565809273e-05, + "loss": 0.0199, + "step": 10493 + }, + { + "epoch": 4.93835294117647, + "grad_norm": 0.1105751499366224, + "learning_rate": 3.143645827998274e-05, + "loss": 0.0379, + "step": 10494 + }, + { + "epoch": 4.938823529411764, + "grad_norm": 0.162812001131581, + "learning_rate": 3.140961102778075e-05, + "loss": 0.0508, + "step": 10495 + }, + { + "epoch": 4.939294117647059, + "grad_norm": 0.1001903819229936, + "learning_rate": 3.138277390377971e-05, + "loss": 0.0226, + "step": 10496 + }, + { + "epoch": 4.939764705882353, + "grad_norm": 0.2254217799594357, + "learning_rate": 3.135594691027155e-05, + "loss": 0.0529, + "step": 10497 + }, + { + "epoch": 4.940235294117647, + "grad_norm": 0.14275173045827974, + "learning_rate": 3.132913004954747e-05, + "loss": 0.06, + "step": 10498 + }, + { + "epoch": 4.940705882352941, + "grad_norm": 0.2358781399170877, + "learning_rate": 3.130232332389781e-05, + "loss": 0.0315, + "step": 10499 + }, + { + "epoch": 4.9411764705882355, + "grad_norm": 1.6641955092152537e-05, + "learning_rate": 3.1275526735611896e-05, + "loss": 0.0, + "step": 10500 + }, + { + "epoch": 4.94164705882353, + "grad_norm": 0.15743963032092628, + "learning_rate": 3.1248740286978375e-05, + "loss": 0.0607, + "step": 10501 + }, + { + "epoch": 4.942117647058824, + "grad_norm": 0.130379249690677, + "learning_rate": 3.1221963980284876e-05, + "loss": 0.0453, + "step": 10502 + }, + { + "epoch": 4.942588235294117, + "grad_norm": 0.10468712502564698, + "learning_rate": 3.119519781781825e-05, + "loss": 0.0132, + "step": 10503 + }, + { + "epoch": 4.9430588235294115, + "grad_norm": 0.2721923642890777, + "learning_rate": 3.1168441801864505e-05, + "loss": 0.0904, + "step": 10504 + }, + { + "epoch": 4.943529411764706, + "grad_norm": 0.10511472700207856, + "learning_rate": 3.1141695934708664e-05, + "loss": 0.0195, + "step": 10505 + }, + { + "epoch": 4.944, + "grad_norm": 0.1071921568082802, + "learning_rate": 3.1114960218634986e-05, + "loss": 0.0116, + "step": 10506 + }, + { + "epoch": 4.944470588235294, + "grad_norm": 0.1327204869387028, + "learning_rate": 3.1088234655926815e-05, + "loss": 0.0349, + "step": 10507 + }, + { + "epoch": 4.944941176470588, + "grad_norm": 0.046161335449533804, + "learning_rate": 3.106151924886668e-05, + "loss": 0.0034, + "step": 10508 + }, + { + "epoch": 4.945411764705883, + "grad_norm": 0.10309215946608806, + "learning_rate": 3.103481399973622e-05, + "loss": 0.0155, + "step": 10509 + }, + { + "epoch": 4.945882352941177, + "grad_norm": 0.08263239521535584, + "learning_rate": 3.100811891081614e-05, + "loss": 0.0138, + "step": 10510 + }, + { + "epoch": 4.946352941176471, + "grad_norm": 0.12113274520532655, + "learning_rate": 3.098143398438635e-05, + "loss": 0.0255, + "step": 10511 + }, + { + "epoch": 4.946823529411764, + "grad_norm": 0.08619415103267693, + "learning_rate": 3.0954759222725916e-05, + "loss": 0.0195, + "step": 10512 + }, + { + "epoch": 4.947294117647059, + "grad_norm": 0.1590054566526807, + "learning_rate": 3.092809462811292e-05, + "loss": 0.0478, + "step": 10513 + }, + { + "epoch": 4.947764705882353, + "grad_norm": 0.09604944145490849, + "learning_rate": 3.090144020282469e-05, + "loss": 0.021, + "step": 10514 + }, + { + "epoch": 4.948235294117647, + "grad_norm": 0.12670447401288354, + "learning_rate": 3.087479594913767e-05, + "loss": 0.0266, + "step": 10515 + }, + { + "epoch": 4.948705882352941, + "grad_norm": 0.050910272696040246, + "learning_rate": 3.0848161869327356e-05, + "loss": 0.0069, + "step": 10516 + }, + { + "epoch": 4.9491764705882355, + "grad_norm": 0.09700173945000963, + "learning_rate": 3.0821537965668454e-05, + "loss": 0.0137, + "step": 10517 + }, + { + "epoch": 4.94964705882353, + "grad_norm": 0.1884804362982825, + "learning_rate": 3.079492424043481e-05, + "loss": 0.028, + "step": 10518 + }, + { + "epoch": 4.950117647058823, + "grad_norm": 0.11551190345197808, + "learning_rate": 3.0768320695899294e-05, + "loss": 0.0374, + "step": 10519 + }, + { + "epoch": 4.950588235294117, + "grad_norm": 0.15522108943968735, + "learning_rate": 3.0741727334334026e-05, + "loss": 0.0225, + "step": 10520 + }, + { + "epoch": 4.9510588235294115, + "grad_norm": 0.01582118161130421, + "learning_rate": 3.071514415801024e-05, + "loss": 0.0006, + "step": 10521 + }, + { + "epoch": 4.951529411764706, + "grad_norm": 0.1021515802828207, + "learning_rate": 3.068857116919819e-05, + "loss": 0.0192, + "step": 10522 + }, + { + "epoch": 4.952, + "grad_norm": 0.014245000462474736, + "learning_rate": 3.066200837016737e-05, + "loss": 0.0005, + "step": 10523 + }, + { + "epoch": 4.952470588235294, + "grad_norm": 0.12998411395228962, + "learning_rate": 3.063545576318644e-05, + "loss": 0.0346, + "step": 10524 + }, + { + "epoch": 4.952941176470588, + "grad_norm": 0.1719147548026593, + "learning_rate": 3.060891335052302e-05, + "loss": 0.068, + "step": 10525 + }, + { + "epoch": 4.953411764705883, + "grad_norm": 0.171745692579676, + "learning_rate": 3.058238113444401e-05, + "loss": 0.057, + "step": 10526 + }, + { + "epoch": 4.953882352941177, + "grad_norm": 0.043999250391212795, + "learning_rate": 3.0555859117215446e-05, + "loss": 0.0048, + "step": 10527 + }, + { + "epoch": 4.95435294117647, + "grad_norm": 2.9494469109091562e-06, + "learning_rate": 3.052934730110232e-05, + "loss": 0.0, + "step": 10528 + }, + { + "epoch": 4.954823529411764, + "grad_norm": 0.07744024113727521, + "learning_rate": 3.050284568836895e-05, + "loss": 0.0053, + "step": 10529 + }, + { + "epoch": 4.955294117647059, + "grad_norm": 0.030930282070611916, + "learning_rate": 3.0476354281278726e-05, + "loss": 0.0029, + "step": 10530 + }, + { + "epoch": 4.955764705882353, + "grad_norm": 0.22698562208922782, + "learning_rate": 3.044987308209407e-05, + "loss": 0.0968, + "step": 10531 + }, + { + "epoch": 4.956235294117647, + "grad_norm": 0.06228370562853515, + "learning_rate": 3.0423402093076683e-05, + "loss": 0.009, + "step": 10532 + }, + { + "epoch": 4.956705882352941, + "grad_norm": 0.1839286175864068, + "learning_rate": 3.039694131648723e-05, + "loss": 0.0594, + "step": 10533 + }, + { + "epoch": 4.9571764705882355, + "grad_norm": 0.0740201993965377, + "learning_rate": 3.0370490754585636e-05, + "loss": 0.0126, + "step": 10534 + }, + { + "epoch": 4.95764705882353, + "grad_norm": 0.1463639064007453, + "learning_rate": 3.034405040963095e-05, + "loss": 0.0463, + "step": 10535 + }, + { + "epoch": 4.958117647058824, + "grad_norm": 0.11731904601892318, + "learning_rate": 3.0317620283881243e-05, + "loss": 0.0317, + "step": 10536 + }, + { + "epoch": 4.958588235294117, + "grad_norm": 0.12427146947743373, + "learning_rate": 3.0291200379593782e-05, + "loss": 0.0474, + "step": 10537 + }, + { + "epoch": 4.9590588235294115, + "grad_norm": 0.15792380445237314, + "learning_rate": 3.0264790699025027e-05, + "loss": 0.0434, + "step": 10538 + }, + { + "epoch": 4.959529411764706, + "grad_norm": 0.1816737839731549, + "learning_rate": 3.0238391244430405e-05, + "loss": 0.0395, + "step": 10539 + }, + { + "epoch": 4.96, + "grad_norm": 0.07868636988999896, + "learning_rate": 3.0212002018064614e-05, + "loss": 0.02, + "step": 10540 + }, + { + "epoch": 4.960470588235294, + "grad_norm": 0.19643062914290696, + "learning_rate": 3.0185623022181415e-05, + "loss": 0.0648, + "step": 10541 + }, + { + "epoch": 4.960941176470588, + "grad_norm": 0.09163048096274319, + "learning_rate": 3.015925425903368e-05, + "loss": 0.0043, + "step": 10542 + }, + { + "epoch": 4.961411764705883, + "grad_norm": 0.07792418720651134, + "learning_rate": 3.0132895730873448e-05, + "loss": 0.0159, + "step": 10543 + }, + { + "epoch": 4.961882352941177, + "grad_norm": 0.06828197253927912, + "learning_rate": 3.010654743995189e-05, + "loss": 0.011, + "step": 10544 + }, + { + "epoch": 4.962352941176471, + "grad_norm": 0.10859224941291448, + "learning_rate": 3.0080209388519222e-05, + "loss": 0.0297, + "step": 10545 + }, + { + "epoch": 4.962823529411764, + "grad_norm": 0.023275563666405145, + "learning_rate": 3.005388157882488e-05, + "loss": 0.0014, + "step": 10546 + }, + { + "epoch": 4.963294117647059, + "grad_norm": 0.1135970284239842, + "learning_rate": 3.002756401311741e-05, + "loss": 0.0418, + "step": 10547 + }, + { + "epoch": 4.963764705882353, + "grad_norm": 0.11930942289981784, + "learning_rate": 3.0001256693644405e-05, + "loss": 0.0344, + "step": 10548 + }, + { + "epoch": 4.964235294117647, + "grad_norm": 0.2110844462761398, + "learning_rate": 2.997495962265266e-05, + "loss": 0.087, + "step": 10549 + }, + { + "epoch": 4.964705882352941, + "grad_norm": 4.9526880320055594e-05, + "learning_rate": 2.9948672802388135e-05, + "loss": 0.0, + "step": 10550 + }, + { + "epoch": 4.9651764705882355, + "grad_norm": 0.1970985727664661, + "learning_rate": 2.9922396235095762e-05, + "loss": 0.0598, + "step": 10551 + }, + { + "epoch": 4.96564705882353, + "grad_norm": 9.846732559496053e-05, + "learning_rate": 2.9896129923019717e-05, + "loss": 0.0, + "step": 10552 + }, + { + "epoch": 4.966117647058823, + "grad_norm": 0.14539922036277697, + "learning_rate": 2.9869873868403287e-05, + "loss": 0.054, + "step": 10553 + }, + { + "epoch": 4.966588235294117, + "grad_norm": 0.1454280565136928, + "learning_rate": 2.9843628073488863e-05, + "loss": 0.0178, + "step": 10554 + }, + { + "epoch": 4.9670588235294115, + "grad_norm": 0.07963671342393469, + "learning_rate": 2.9817392540518014e-05, + "loss": 0.0103, + "step": 10555 + }, + { + "epoch": 4.967529411764706, + "grad_norm": 0.21278865905519947, + "learning_rate": 2.9791167271731287e-05, + "loss": 0.0703, + "step": 10556 + }, + { + "epoch": 4.968, + "grad_norm": 0.10739404527805982, + "learning_rate": 2.9764952269368487e-05, + "loss": 0.0226, + "step": 10557 + }, + { + "epoch": 4.968470588235294, + "grad_norm": 0.10393176284907836, + "learning_rate": 2.973874753566857e-05, + "loss": 0.0287, + "step": 10558 + }, + { + "epoch": 4.968941176470588, + "grad_norm": 1.4672880808735596e-06, + "learning_rate": 2.9712553072869437e-05, + "loss": 0.0, + "step": 10559 + }, + { + "epoch": 4.969411764705883, + "grad_norm": 0.04160179727688323, + "learning_rate": 2.968636888320833e-05, + "loss": 0.0064, + "step": 10560 + }, + { + "epoch": 4.969882352941177, + "grad_norm": 0.11132589355243656, + "learning_rate": 2.9660194968921403e-05, + "loss": 0.0123, + "step": 10561 + }, + { + "epoch": 4.97035294117647, + "grad_norm": 2.4836853295755797e-06, + "learning_rate": 2.963403133224409e-05, + "loss": 0.0, + "step": 10562 + }, + { + "epoch": 4.970823529411764, + "grad_norm": 0.0677397331152983, + "learning_rate": 2.9607877975410936e-05, + "loss": 0.0078, + "step": 10563 + }, + { + "epoch": 4.971294117647059, + "grad_norm": 0.07161976767485187, + "learning_rate": 2.958173490065549e-05, + "loss": 0.0086, + "step": 10564 + }, + { + "epoch": 4.971764705882353, + "grad_norm": 0.08224388606332803, + "learning_rate": 2.955560211021053e-05, + "loss": 0.0148, + "step": 10565 + }, + { + "epoch": 4.972235294117647, + "grad_norm": 0.17996544857834712, + "learning_rate": 2.952947960630797e-05, + "loss": 0.0541, + "step": 10566 + }, + { + "epoch": 4.972705882352941, + "grad_norm": 0.014959277318115593, + "learning_rate": 2.9503367391178713e-05, + "loss": 0.0005, + "step": 10567 + }, + { + "epoch": 4.9731764705882355, + "grad_norm": 0.13859595370106984, + "learning_rate": 2.94772654670529e-05, + "loss": 0.0562, + "step": 10568 + }, + { + "epoch": 4.97364705882353, + "grad_norm": 8.769431465852821e-06, + "learning_rate": 2.9451173836159826e-05, + "loss": 0.0, + "step": 10569 + }, + { + "epoch": 4.974117647058824, + "grad_norm": 0.14521191213806967, + "learning_rate": 2.9425092500727764e-05, + "loss": 0.0534, + "step": 10570 + }, + { + "epoch": 4.974588235294117, + "grad_norm": 0.032888903369305215, + "learning_rate": 2.9399021462984206e-05, + "loss": 0.0039, + "step": 10571 + }, + { + "epoch": 4.9750588235294115, + "grad_norm": 0.14042947450930998, + "learning_rate": 2.9372960725155804e-05, + "loss": 0.0372, + "step": 10572 + }, + { + "epoch": 4.975529411764706, + "grad_norm": 0.14641002057541022, + "learning_rate": 2.934691028946819e-05, + "loss": 0.0114, + "step": 10573 + }, + { + "epoch": 4.976, + "grad_norm": 0.10246056451499516, + "learning_rate": 2.9320870158146235e-05, + "loss": 0.0308, + "step": 10574 + }, + { + "epoch": 4.976470588235294, + "grad_norm": 0.05701710581141495, + "learning_rate": 2.9294840333413932e-05, + "loss": 0.0035, + "step": 10575 + }, + { + "epoch": 4.976941176470588, + "grad_norm": 0.02906573248824732, + "learning_rate": 2.926882081749428e-05, + "loss": 0.002, + "step": 10576 + }, + { + "epoch": 4.977411764705883, + "grad_norm": 0.04144738562568653, + "learning_rate": 2.9242811612609522e-05, + "loss": 0.0037, + "step": 10577 + }, + { + "epoch": 4.977882352941176, + "grad_norm": 0.08349280133719843, + "learning_rate": 2.9216812720980998e-05, + "loss": 0.0123, + "step": 10578 + }, + { + "epoch": 4.97835294117647, + "grad_norm": 0.07203435540156847, + "learning_rate": 2.919082414482908e-05, + "loss": 0.0057, + "step": 10579 + }, + { + "epoch": 4.978823529411764, + "grad_norm": 0.11967491160506354, + "learning_rate": 2.9164845886373323e-05, + "loss": 0.0265, + "step": 10580 + }, + { + "epoch": 4.979294117647059, + "grad_norm": 0.04313525214544001, + "learning_rate": 2.9138877947832474e-05, + "loss": 0.0048, + "step": 10581 + }, + { + "epoch": 4.979764705882353, + "grad_norm": 0.08847892680058861, + "learning_rate": 2.911292033142422e-05, + "loss": 0.0213, + "step": 10582 + }, + { + "epoch": 4.980235294117647, + "grad_norm": 0.1492235859377117, + "learning_rate": 2.9086973039365562e-05, + "loss": 0.0101, + "step": 10583 + }, + { + "epoch": 4.980705882352941, + "grad_norm": 0.08966540007653706, + "learning_rate": 2.906103607387244e-05, + "loss": 0.0184, + "step": 10584 + }, + { + "epoch": 4.9811764705882355, + "grad_norm": 3.040880216513466e-06, + "learning_rate": 2.9035109437160047e-05, + "loss": 0.0, + "step": 10585 + }, + { + "epoch": 4.98164705882353, + "grad_norm": 0.1007791479471551, + "learning_rate": 2.9009193131442665e-05, + "loss": 0.0192, + "step": 10586 + }, + { + "epoch": 4.982117647058823, + "grad_norm": 0.0613264706291032, + "learning_rate": 2.8983287158933622e-05, + "loss": 0.0053, + "step": 10587 + }, + { + "epoch": 4.982588235294117, + "grad_norm": 0.03694500847957993, + "learning_rate": 2.895739152184543e-05, + "loss": 0.005, + "step": 10588 + }, + { + "epoch": 4.9830588235294115, + "grad_norm": 0.31624571440380045, + "learning_rate": 2.8931506222389756e-05, + "loss": 0.0743, + "step": 10589 + }, + { + "epoch": 4.983529411764706, + "grad_norm": 0.07875530639383896, + "learning_rate": 2.8905631262777245e-05, + "loss": 0.0142, + "step": 10590 + }, + { + "epoch": 4.984, + "grad_norm": 1.5011082112781748e-05, + "learning_rate": 2.8879766645217784e-05, + "loss": 0.0, + "step": 10591 + }, + { + "epoch": 4.984470588235294, + "grad_norm": 4.113837380436122e-05, + "learning_rate": 2.8853912371920384e-05, + "loss": 0.0, + "step": 10592 + }, + { + "epoch": 4.984941176470588, + "grad_norm": 0.15380732362310973, + "learning_rate": 2.882806844509305e-05, + "loss": 0.065, + "step": 10593 + }, + { + "epoch": 4.985411764705883, + "grad_norm": 0.0967109316099082, + "learning_rate": 2.8802234866943012e-05, + "loss": 0.0146, + "step": 10594 + }, + { + "epoch": 4.985882352941177, + "grad_norm": 0.16184697439587112, + "learning_rate": 2.8776411639676644e-05, + "loss": 0.049, + "step": 10595 + }, + { + "epoch": 4.98635294117647, + "grad_norm": 0.14111320059385205, + "learning_rate": 2.8750598765499226e-05, + "loss": 0.0348, + "step": 10596 + }, + { + "epoch": 4.986823529411764, + "grad_norm": 0.1226796630122715, + "learning_rate": 2.8724796246615485e-05, + "loss": 0.031, + "step": 10597 + }, + { + "epoch": 4.987294117647059, + "grad_norm": 0.06144498327326442, + "learning_rate": 2.8699004085228948e-05, + "loss": 0.0049, + "step": 10598 + }, + { + "epoch": 4.987764705882353, + "grad_norm": 0.060858003676317184, + "learning_rate": 2.867322228354243e-05, + "loss": 0.0051, + "step": 10599 + }, + { + "epoch": 4.988235294117647, + "grad_norm": 0.1516958957753182, + "learning_rate": 2.8647450843757897e-05, + "loss": 0.0331, + "step": 10600 + }, + { + "epoch": 4.988705882352941, + "grad_norm": 0.06362188058802483, + "learning_rate": 2.862168976807624e-05, + "loss": 0.0135, + "step": 10601 + }, + { + "epoch": 4.9891764705882355, + "grad_norm": 0.10968623694936092, + "learning_rate": 2.859593905869763e-05, + "loss": 0.018, + "step": 10602 + }, + { + "epoch": 4.98964705882353, + "grad_norm": 0.149275649170429, + "learning_rate": 2.8570198717821362e-05, + "loss": 0.0283, + "step": 10603 + }, + { + "epoch": 4.990117647058824, + "grad_norm": 5.5459027130072066e-05, + "learning_rate": 2.8544468747645676e-05, + "loss": 0.0, + "step": 10604 + }, + { + "epoch": 4.990588235294117, + "grad_norm": 0.11672789238016085, + "learning_rate": 2.8518749150368114e-05, + "loss": 0.0138, + "step": 10605 + }, + { + "epoch": 4.9910588235294115, + "grad_norm": 0.09154298991429131, + "learning_rate": 2.849303992818527e-05, + "loss": 0.0131, + "step": 10606 + }, + { + "epoch": 4.991529411764706, + "grad_norm": 1.1566643742561748e-06, + "learning_rate": 2.846734108329275e-05, + "loss": 0.0, + "step": 10607 + }, + { + "epoch": 4.992, + "grad_norm": 0.14700281972936224, + "learning_rate": 2.8441652617885437e-05, + "loss": 0.0286, + "step": 10608 + }, + { + "epoch": 4.992470588235294, + "grad_norm": 0.049300789835258055, + "learning_rate": 2.8415974534157255e-05, + "loss": 0.0062, + "step": 10609 + }, + { + "epoch": 4.992941176470588, + "grad_norm": 0.14203753259568994, + "learning_rate": 2.8390306834301186e-05, + "loss": 0.0426, + "step": 10610 + }, + { + "epoch": 4.993411764705883, + "grad_norm": 0.1297647993922102, + "learning_rate": 2.836464952050944e-05, + "loss": 0.0184, + "step": 10611 + }, + { + "epoch": 4.993882352941176, + "grad_norm": 0.12470779890818742, + "learning_rate": 2.8339002594973204e-05, + "loss": 0.0277, + "step": 10612 + }, + { + "epoch": 4.99435294117647, + "grad_norm": 0.12007338095606668, + "learning_rate": 2.8313366059882887e-05, + "loss": 0.0286, + "step": 10613 + }, + { + "epoch": 4.994823529411764, + "grad_norm": 0.1153024194644917, + "learning_rate": 2.828773991742802e-05, + "loss": 0.041, + "step": 10614 + }, + { + "epoch": 4.995294117647059, + "grad_norm": 0.14028812864471335, + "learning_rate": 2.8262124169797106e-05, + "loss": 0.0215, + "step": 10615 + }, + { + "epoch": 4.995764705882353, + "grad_norm": 0.217965074648704, + "learning_rate": 2.823651881917793e-05, + "loss": 0.0586, + "step": 10616 + }, + { + "epoch": 4.996235294117647, + "grad_norm": 0.02468113953675946, + "learning_rate": 2.8210923867757317e-05, + "loss": 0.0008, + "step": 10617 + }, + { + "epoch": 4.996705882352941, + "grad_norm": 0.08894419547919026, + "learning_rate": 2.8185339317721146e-05, + "loss": 0.0195, + "step": 10618 + }, + { + "epoch": 4.9971764705882356, + "grad_norm": 1.328923065126066e-05, + "learning_rate": 2.8159765171254495e-05, + "loss": 0.0, + "step": 10619 + }, + { + "epoch": 4.99764705882353, + "grad_norm": 0.12944523681699216, + "learning_rate": 2.813420143054157e-05, + "loss": 0.0245, + "step": 10620 + }, + { + "epoch": 4.998117647058823, + "grad_norm": 0.12173338505445051, + "learning_rate": 2.810864809776554e-05, + "loss": 0.0381, + "step": 10621 + }, + { + "epoch": 4.998588235294117, + "grad_norm": 0.13990963983268706, + "learning_rate": 2.8083105175108845e-05, + "loss": 0.0568, + "step": 10622 + }, + { + "epoch": 4.9990588235294116, + "grad_norm": 0.09262964183379674, + "learning_rate": 2.8057572664752986e-05, + "loss": 0.0172, + "step": 10623 + }, + { + "epoch": 4.999529411764706, + "grad_norm": 0.15038476342377996, + "learning_rate": 2.8032050568878524e-05, + "loss": 0.0666, + "step": 10624 + }, + { + "epoch": 5.0, + "grad_norm": 0.07360460996250309, + "learning_rate": 2.800653888966519e-05, + "loss": 0.0161, + "step": 10625 + }, + { + "epoch": 5.0, + "eval_loss": 0.03002466820180416, + "eval_runtime": 59.206, + "eval_samples_per_second": 12.668, + "eval_steps_per_second": 6.334, + "step": 10625 + }, + { + "epoch": 5.000470588235294, + "grad_norm": 0.17422785968221863, + "learning_rate": 2.7981037629291852e-05, + "loss": 0.0395, + "step": 10626 + }, + { + "epoch": 5.0009411764705884, + "grad_norm": 0.10456507820779962, + "learning_rate": 2.7955546789936363e-05, + "loss": 0.0112, + "step": 10627 + }, + { + "epoch": 5.001411764705883, + "grad_norm": 0.182874376421964, + "learning_rate": 2.7930066373775794e-05, + "loss": 0.0885, + "step": 10628 + }, + { + "epoch": 5.001882352941177, + "grad_norm": 0.025947556051507847, + "learning_rate": 2.7904596382986343e-05, + "loss": 0.0013, + "step": 10629 + }, + { + "epoch": 5.00235294117647, + "grad_norm": 0.08744162683789974, + "learning_rate": 2.7879136819743203e-05, + "loss": 0.0178, + "step": 10630 + }, + { + "epoch": 5.0028235294117644, + "grad_norm": 2.6909782469341134e-05, + "learning_rate": 2.7853687686220766e-05, + "loss": 0.0, + "step": 10631 + }, + { + "epoch": 5.003294117647059, + "grad_norm": 1.045476466836052e-05, + "learning_rate": 2.782824898459258e-05, + "loss": 0.0, + "step": 10632 + }, + { + "epoch": 5.003764705882353, + "grad_norm": 0.03525009060802935, + "learning_rate": 2.7802820717031132e-05, + "loss": 0.0005, + "step": 10633 + }, + { + "epoch": 5.004235294117647, + "grad_norm": 0.09940666697058083, + "learning_rate": 2.7777402885708212e-05, + "loss": 0.0236, + "step": 10634 + }, + { + "epoch": 5.004705882352941, + "grad_norm": 0.08289963410740137, + "learning_rate": 2.7751995492794544e-05, + "loss": 0.0132, + "step": 10635 + }, + { + "epoch": 5.005176470588236, + "grad_norm": 6.661265552423695e-06, + "learning_rate": 2.7726598540460076e-05, + "loss": 0.0, + "step": 10636 + }, + { + "epoch": 5.00564705882353, + "grad_norm": 1.7003274056652703e-05, + "learning_rate": 2.770121203087389e-05, + "loss": 0.0, + "step": 10637 + }, + { + "epoch": 5.006117647058823, + "grad_norm": 0.0972405445218899, + "learning_rate": 2.7675835966204024e-05, + "loss": 0.0135, + "step": 10638 + }, + { + "epoch": 5.006588235294117, + "grad_norm": 0.00010788374080887111, + "learning_rate": 2.7650470348617772e-05, + "loss": 0.0, + "step": 10639 + }, + { + "epoch": 5.007058823529412, + "grad_norm": 0.05191739840902971, + "learning_rate": 2.7625115180281464e-05, + "loss": 0.0043, + "step": 10640 + }, + { + "epoch": 5.007529411764706, + "grad_norm": 0.09263381422140853, + "learning_rate": 2.7599770463360576e-05, + "loss": 0.0259, + "step": 10641 + }, + { + "epoch": 5.008, + "grad_norm": 0.13972658092862106, + "learning_rate": 2.7574436200019688e-05, + "loss": 0.0298, + "step": 10642 + }, + { + "epoch": 5.008470588235294, + "grad_norm": 0.05250157484007339, + "learning_rate": 2.7549112392422405e-05, + "loss": 0.0043, + "step": 10643 + }, + { + "epoch": 5.0089411764705885, + "grad_norm": 0.10928456360117579, + "learning_rate": 2.7523799042731548e-05, + "loss": 0.0084, + "step": 10644 + }, + { + "epoch": 5.009411764705883, + "grad_norm": 0.07334020839015694, + "learning_rate": 2.749849615310902e-05, + "loss": 0.0109, + "step": 10645 + }, + { + "epoch": 5.009882352941177, + "grad_norm": 0.1219382335077675, + "learning_rate": 2.7473203725715754e-05, + "loss": 0.0243, + "step": 10646 + }, + { + "epoch": 5.01035294117647, + "grad_norm": 3.0384346299831156e-05, + "learning_rate": 2.7447921762711894e-05, + "loss": 0.0, + "step": 10647 + }, + { + "epoch": 5.0108235294117645, + "grad_norm": 0.09289881271640517, + "learning_rate": 2.7422650266256663e-05, + "loss": 0.018, + "step": 10648 + }, + { + "epoch": 5.011294117647059, + "grad_norm": 0.09341367943997375, + "learning_rate": 2.7397389238508282e-05, + "loss": 0.0133, + "step": 10649 + }, + { + "epoch": 5.011764705882353, + "grad_norm": 0.07302167119153859, + "learning_rate": 2.7372138681624244e-05, + "loss": 0.0063, + "step": 10650 + }, + { + "epoch": 5.012235294117647, + "grad_norm": 0.03609474103824484, + "learning_rate": 2.7346898597761073e-05, + "loss": 0.0044, + "step": 10651 + }, + { + "epoch": 5.012705882352941, + "grad_norm": 0.10219965100294347, + "learning_rate": 2.7321668989074347e-05, + "loss": 0.0277, + "step": 10652 + }, + { + "epoch": 5.013176470588236, + "grad_norm": 0.13738374324151414, + "learning_rate": 2.7296449857718806e-05, + "loss": 0.028, + "step": 10653 + }, + { + "epoch": 5.01364705882353, + "grad_norm": 0.09720563670264007, + "learning_rate": 2.727124120584836e-05, + "loss": 0.0218, + "step": 10654 + }, + { + "epoch": 5.014117647058823, + "grad_norm": 0.1001253128653449, + "learning_rate": 2.724604303561584e-05, + "loss": 0.0134, + "step": 10655 + }, + { + "epoch": 5.014588235294117, + "grad_norm": 0.04672103181322193, + "learning_rate": 2.7220855349173353e-05, + "loss": 0.0023, + "step": 10656 + }, + { + "epoch": 5.015058823529412, + "grad_norm": 0.04061049133292854, + "learning_rate": 2.71956781486721e-05, + "loss": 0.0041, + "step": 10657 + }, + { + "epoch": 5.015529411764706, + "grad_norm": 0.08558281906278517, + "learning_rate": 2.7170511436262243e-05, + "loss": 0.0143, + "step": 10658 + }, + { + "epoch": 5.016, + "grad_norm": 3.594892038888735e-05, + "learning_rate": 2.7145355214093178e-05, + "loss": 0.0, + "step": 10659 + }, + { + "epoch": 5.016470588235294, + "grad_norm": 0.19203497579758222, + "learning_rate": 2.7120209484313422e-05, + "loss": 0.0514, + "step": 10660 + }, + { + "epoch": 5.0169411764705885, + "grad_norm": 0.07531130650544184, + "learning_rate": 2.7095074249070455e-05, + "loss": 0.0066, + "step": 10661 + }, + { + "epoch": 5.017411764705883, + "grad_norm": 0.06390560769633363, + "learning_rate": 2.7069949510511023e-05, + "loss": 0.0085, + "step": 10662 + }, + { + "epoch": 5.017882352941177, + "grad_norm": 0.12415907162018483, + "learning_rate": 2.7044835270780836e-05, + "loss": 0.0373, + "step": 10663 + }, + { + "epoch": 5.01835294117647, + "grad_norm": 1.2716370620640218e-05, + "learning_rate": 2.701973153202482e-05, + "loss": 0.0, + "step": 10664 + }, + { + "epoch": 5.0188235294117645, + "grad_norm": 0.03244539198119347, + "learning_rate": 2.6994638296386977e-05, + "loss": 0.002, + "step": 10665 + }, + { + "epoch": 5.019294117647059, + "grad_norm": 0.1634982740678334, + "learning_rate": 2.6969555566010316e-05, + "loss": 0.0261, + "step": 10666 + }, + { + "epoch": 5.019764705882353, + "grad_norm": 2.150379581746472e-05, + "learning_rate": 2.694448334303708e-05, + "loss": 0.0, + "step": 10667 + }, + { + "epoch": 5.020235294117647, + "grad_norm": 0.20443121445510812, + "learning_rate": 2.6919421629608584e-05, + "loss": 0.0505, + "step": 10668 + }, + { + "epoch": 5.020705882352941, + "grad_norm": 0.12333905870915592, + "learning_rate": 2.6894370427865157e-05, + "loss": 0.0253, + "step": 10669 + }, + { + "epoch": 5.021176470588236, + "grad_norm": 0.16441984986195224, + "learning_rate": 2.6869329739946312e-05, + "loss": 0.0486, + "step": 10670 + }, + { + "epoch": 5.02164705882353, + "grad_norm": 0.05559585549436808, + "learning_rate": 2.6844299567990713e-05, + "loss": 0.0065, + "step": 10671 + }, + { + "epoch": 5.022117647058823, + "grad_norm": 1.8174470503308056e-05, + "learning_rate": 2.6819279914135956e-05, + "loss": 0.0, + "step": 10672 + }, + { + "epoch": 5.022588235294117, + "grad_norm": 0.1661516230629207, + "learning_rate": 2.6794270780518895e-05, + "loss": 0.0183, + "step": 10673 + }, + { + "epoch": 5.023058823529412, + "grad_norm": 0.04422416872652042, + "learning_rate": 2.6769272169275474e-05, + "loss": 0.0019, + "step": 10674 + }, + { + "epoch": 5.023529411764706, + "grad_norm": 0.04846291545586831, + "learning_rate": 2.67442840825406e-05, + "loss": 0.0067, + "step": 10675 + }, + { + "epoch": 5.024, + "grad_norm": 1.8766108398736784e-05, + "learning_rate": 2.671930652244845e-05, + "loss": 0.0, + "step": 10676 + }, + { + "epoch": 5.024470588235294, + "grad_norm": 1.2206187638345015e-05, + "learning_rate": 2.6694339491132237e-05, + "loss": 0.0, + "step": 10677 + }, + { + "epoch": 5.0249411764705885, + "grad_norm": 0.1146179559823975, + "learning_rate": 2.666938299072422e-05, + "loss": 0.0206, + "step": 10678 + }, + { + "epoch": 5.025411764705883, + "grad_norm": 0.19441527397871203, + "learning_rate": 2.664443702335581e-05, + "loss": 0.0333, + "step": 10679 + }, + { + "epoch": 5.025882352941177, + "grad_norm": 0.11897377402293523, + "learning_rate": 2.6619501591157583e-05, + "loss": 0.0381, + "step": 10680 + }, + { + "epoch": 5.02635294117647, + "grad_norm": 0.1099623282162366, + "learning_rate": 2.6594576696259055e-05, + "loss": 0.0204, + "step": 10681 + }, + { + "epoch": 5.0268235294117645, + "grad_norm": 0.00013962079274224565, + "learning_rate": 2.6569662340789012e-05, + "loss": 0.0, + "step": 10682 + }, + { + "epoch": 5.027294117647059, + "grad_norm": 0.07482463394845006, + "learning_rate": 2.6544758526875193e-05, + "loss": 0.0113, + "step": 10683 + }, + { + "epoch": 5.027764705882353, + "grad_norm": 0.07773986941978271, + "learning_rate": 2.651986525664455e-05, + "loss": 0.0121, + "step": 10684 + }, + { + "epoch": 5.028235294117647, + "grad_norm": 0.06667440510742825, + "learning_rate": 2.649498253222307e-05, + "loss": 0.0089, + "step": 10685 + }, + { + "epoch": 5.028705882352941, + "grad_norm": 2.6390537155128435e-05, + "learning_rate": 2.647011035573588e-05, + "loss": 0.0, + "step": 10686 + }, + { + "epoch": 5.029176470588236, + "grad_norm": 0.10362172516341106, + "learning_rate": 2.6445248729307165e-05, + "loss": 0.0278, + "step": 10687 + }, + { + "epoch": 5.02964705882353, + "grad_norm": 0.009247868288784784, + "learning_rate": 2.6420397655060282e-05, + "loss": 0.0002, + "step": 10688 + }, + { + "epoch": 5.030117647058823, + "grad_norm": 0.1451347933974505, + "learning_rate": 2.639555713511756e-05, + "loss": 0.0413, + "step": 10689 + }, + { + "epoch": 5.030588235294117, + "grad_norm": 0.10611779415476767, + "learning_rate": 2.637072717160056e-05, + "loss": 0.0241, + "step": 10690 + }, + { + "epoch": 5.031058823529412, + "grad_norm": 0.11864614589019401, + "learning_rate": 2.6345907766629843e-05, + "loss": 0.0116, + "step": 10691 + }, + { + "epoch": 5.031529411764706, + "grad_norm": 0.1276232021816311, + "learning_rate": 2.6321098922325102e-05, + "loss": 0.0382, + "step": 10692 + }, + { + "epoch": 5.032, + "grad_norm": 0.05212970438666624, + "learning_rate": 2.6296300640805206e-05, + "loss": 0.0029, + "step": 10693 + }, + { + "epoch": 5.032470588235294, + "grad_norm": 0.1367459535604145, + "learning_rate": 2.627151292418795e-05, + "loss": 0.0396, + "step": 10694 + }, + { + "epoch": 5.0329411764705885, + "grad_norm": 0.19594013121012682, + "learning_rate": 2.6246735774590383e-05, + "loss": 0.035, + "step": 10695 + }, + { + "epoch": 5.033411764705883, + "grad_norm": 0.1082619175845701, + "learning_rate": 2.6221969194128617e-05, + "loss": 0.017, + "step": 10696 + }, + { + "epoch": 5.033882352941176, + "grad_norm": 0.08693548237589764, + "learning_rate": 2.6197213184917775e-05, + "loss": 0.0203, + "step": 10697 + }, + { + "epoch": 5.03435294117647, + "grad_norm": 0.08987910526833238, + "learning_rate": 2.6172467749072185e-05, + "loss": 0.0086, + "step": 10698 + }, + { + "epoch": 5.0348235294117645, + "grad_norm": 7.807861517590128e-06, + "learning_rate": 2.6147732888705235e-05, + "loss": 0.0, + "step": 10699 + }, + { + "epoch": 5.035294117647059, + "grad_norm": 0.11340925480296507, + "learning_rate": 2.6123008605929375e-05, + "loss": 0.0216, + "step": 10700 + }, + { + "epoch": 5.035764705882353, + "grad_norm": 0.16145562993941398, + "learning_rate": 2.6098294902856177e-05, + "loss": 0.0394, + "step": 10701 + }, + { + "epoch": 5.036235294117647, + "grad_norm": 0.1048793712865994, + "learning_rate": 2.607359178159637e-05, + "loss": 0.0261, + "step": 10702 + }, + { + "epoch": 5.036705882352941, + "grad_norm": 0.13577084559657804, + "learning_rate": 2.604889924425965e-05, + "loss": 0.0349, + "step": 10703 + }, + { + "epoch": 5.037176470588236, + "grad_norm": 0.07321508961862501, + "learning_rate": 2.6024217292954917e-05, + "loss": 0.0097, + "step": 10704 + }, + { + "epoch": 5.03764705882353, + "grad_norm": 0.04436124758403717, + "learning_rate": 2.599954592979015e-05, + "loss": 0.0039, + "step": 10705 + }, + { + "epoch": 5.038117647058823, + "grad_norm": 1.5661820784088396e-06, + "learning_rate": 2.5974885156872345e-05, + "loss": 0.0, + "step": 10706 + }, + { + "epoch": 5.038588235294117, + "grad_norm": 0.010723343572840133, + "learning_rate": 2.5950234976307682e-05, + "loss": 0.0002, + "step": 10707 + }, + { + "epoch": 5.039058823529412, + "grad_norm": 1.4023668290017918e-05, + "learning_rate": 2.5925595390201466e-05, + "loss": 0.0, + "step": 10708 + }, + { + "epoch": 5.039529411764706, + "grad_norm": 0.052283181892717315, + "learning_rate": 2.5900966400657923e-05, + "loss": 0.002, + "step": 10709 + }, + { + "epoch": 5.04, + "grad_norm": 0.02806044022039621, + "learning_rate": 2.5876348009780564e-05, + "loss": 0.0012, + "step": 10710 + }, + { + "epoch": 5.040470588235294, + "grad_norm": 0.09834426999001232, + "learning_rate": 2.5851740219671945e-05, + "loss": 0.0183, + "step": 10711 + }, + { + "epoch": 5.0409411764705885, + "grad_norm": 0.0805731733438974, + "learning_rate": 2.5827143032433627e-05, + "loss": 0.0092, + "step": 10712 + }, + { + "epoch": 5.041411764705883, + "grad_norm": 0.2651812445997511, + "learning_rate": 2.5802556450166396e-05, + "loss": 0.038, + "step": 10713 + }, + { + "epoch": 5.041882352941176, + "grad_norm": 0.19610313439202204, + "learning_rate": 2.5777980474969983e-05, + "loss": 0.0513, + "step": 10714 + }, + { + "epoch": 5.04235294117647, + "grad_norm": 0.10932037902719242, + "learning_rate": 2.5753415108943332e-05, + "loss": 0.0275, + "step": 10715 + }, + { + "epoch": 5.0428235294117645, + "grad_norm": 0.0762057098902362, + "learning_rate": 2.572886035418451e-05, + "loss": 0.0076, + "step": 10716 + }, + { + "epoch": 5.043294117647059, + "grad_norm": 0.061253777244891464, + "learning_rate": 2.5704316212790527e-05, + "loss": 0.006, + "step": 10717 + }, + { + "epoch": 5.043764705882353, + "grad_norm": 0.1082440716761263, + "learning_rate": 2.5679782686857586e-05, + "loss": 0.0149, + "step": 10718 + }, + { + "epoch": 5.044235294117647, + "grad_norm": 0.0711172086274759, + "learning_rate": 2.5655259778481045e-05, + "loss": 0.0105, + "step": 10719 + }, + { + "epoch": 5.044705882352941, + "grad_norm": 0.148832281621529, + "learning_rate": 2.5630747489755182e-05, + "loss": 0.0497, + "step": 10720 + }, + { + "epoch": 5.045176470588236, + "grad_norm": 0.05656515788304277, + "learning_rate": 2.560624582277352e-05, + "loss": 0.0118, + "step": 10721 + }, + { + "epoch": 5.04564705882353, + "grad_norm": 0.05867396180464138, + "learning_rate": 2.558175477962865e-05, + "loss": 0.0059, + "step": 10722 + }, + { + "epoch": 5.046117647058823, + "grad_norm": 0.39753834151108225, + "learning_rate": 2.5557274362412132e-05, + "loss": 0.0831, + "step": 10723 + }, + { + "epoch": 5.046588235294117, + "grad_norm": 0.04514858512905867, + "learning_rate": 2.55328045732148e-05, + "loss": 0.0074, + "step": 10724 + }, + { + "epoch": 5.047058823529412, + "grad_norm": 0.10402472581324906, + "learning_rate": 2.5508345414126474e-05, + "loss": 0.0167, + "step": 10725 + }, + { + "epoch": 5.047529411764706, + "grad_norm": 0.13391855849581102, + "learning_rate": 2.5483896887236056e-05, + "loss": 0.0247, + "step": 10726 + }, + { + "epoch": 5.048, + "grad_norm": 0.05777204782651226, + "learning_rate": 2.5459458994631604e-05, + "loss": 0.0083, + "step": 10727 + }, + { + "epoch": 5.048470588235294, + "grad_norm": 0.11300266356608367, + "learning_rate": 2.5435031738400234e-05, + "loss": 0.0382, + "step": 10728 + }, + { + "epoch": 5.0489411764705885, + "grad_norm": 0.0001460743227106122, + "learning_rate": 2.541061512062808e-05, + "loss": 0.0, + "step": 10729 + }, + { + "epoch": 5.049411764705883, + "grad_norm": 0.0496229178416605, + "learning_rate": 2.538620914340057e-05, + "loss": 0.0052, + "step": 10730 + }, + { + "epoch": 5.049882352941176, + "grad_norm": 0.13991355634098562, + "learning_rate": 2.5361813808802007e-05, + "loss": 0.0316, + "step": 10731 + }, + { + "epoch": 5.05035294117647, + "grad_norm": 2.20285029846882e-05, + "learning_rate": 2.533742911891587e-05, + "loss": 0.0, + "step": 10732 + }, + { + "epoch": 5.0508235294117645, + "grad_norm": 0.05695602927657431, + "learning_rate": 2.531305507582481e-05, + "loss": 0.0066, + "step": 10733 + }, + { + "epoch": 5.051294117647059, + "grad_norm": 2.7984560488337432e-05, + "learning_rate": 2.5288691681610383e-05, + "loss": 0.0, + "step": 10734 + }, + { + "epoch": 5.051764705882353, + "grad_norm": 0.0628503868392903, + "learning_rate": 2.526433893835339e-05, + "loss": 0.0083, + "step": 10735 + }, + { + "epoch": 5.052235294117647, + "grad_norm": 0.10482044789958445, + "learning_rate": 2.5239996848133715e-05, + "loss": 0.0215, + "step": 10736 + }, + { + "epoch": 5.052705882352941, + "grad_norm": 0.1338787420531987, + "learning_rate": 2.5215665413030213e-05, + "loss": 0.0178, + "step": 10737 + }, + { + "epoch": 5.053176470588236, + "grad_norm": 0.07860239773408065, + "learning_rate": 2.519134463512095e-05, + "loss": 0.0105, + "step": 10738 + }, + { + "epoch": 5.05364705882353, + "grad_norm": 0.15574361514363202, + "learning_rate": 2.516703451648309e-05, + "loss": 0.0457, + "step": 10739 + }, + { + "epoch": 5.054117647058823, + "grad_norm": 0.14146683354907397, + "learning_rate": 2.5142735059192737e-05, + "loss": 0.0402, + "step": 10740 + }, + { + "epoch": 5.054588235294117, + "grad_norm": 0.07686648379241003, + "learning_rate": 2.511844626532526e-05, + "loss": 0.0165, + "step": 10741 + }, + { + "epoch": 5.055058823529412, + "grad_norm": 3.174025490278303e-05, + "learning_rate": 2.5094168136954985e-05, + "loss": 0.0, + "step": 10742 + }, + { + "epoch": 5.055529411764706, + "grad_norm": 0.14270810955810784, + "learning_rate": 2.506990067615542e-05, + "loss": 0.0422, + "step": 10743 + }, + { + "epoch": 5.056, + "grad_norm": 0.2020009151864963, + "learning_rate": 2.504564388499914e-05, + "loss": 0.0643, + "step": 10744 + }, + { + "epoch": 5.056470588235294, + "grad_norm": 0.12116391158769645, + "learning_rate": 2.502139776555772e-05, + "loss": 0.0185, + "step": 10745 + }, + { + "epoch": 5.0569411764705885, + "grad_norm": 0.09572299161022936, + "learning_rate": 2.4997162319901977e-05, + "loss": 0.0237, + "step": 10746 + }, + { + "epoch": 5.057411764705883, + "grad_norm": 0.0676262265941171, + "learning_rate": 2.497293755010173e-05, + "loss": 0.0041, + "step": 10747 + }, + { + "epoch": 5.057882352941176, + "grad_norm": 0.039970817176321415, + "learning_rate": 2.4948723458225827e-05, + "loss": 0.004, + "step": 10748 + }, + { + "epoch": 5.05835294117647, + "grad_norm": 0.05511830009966054, + "learning_rate": 2.4924520046342317e-05, + "loss": 0.0056, + "step": 10749 + }, + { + "epoch": 5.0588235294117645, + "grad_norm": 0.15603907991423038, + "learning_rate": 2.4900327316518326e-05, + "loss": 0.0458, + "step": 10750 + }, + { + "epoch": 5.059294117647059, + "grad_norm": 0.05860586105303238, + "learning_rate": 2.4876145270819954e-05, + "loss": 0.0052, + "step": 10751 + }, + { + "epoch": 5.059764705882353, + "grad_norm": 5.776469061652841e-06, + "learning_rate": 2.485197391131251e-05, + "loss": 0.0, + "step": 10752 + }, + { + "epoch": 5.060235294117647, + "grad_norm": 0.022021616428352014, + "learning_rate": 2.4827813240060367e-05, + "loss": 0.0013, + "step": 10753 + }, + { + "epoch": 5.060705882352941, + "grad_norm": 0.08759682968686337, + "learning_rate": 2.480366325912692e-05, + "loss": 0.019, + "step": 10754 + }, + { + "epoch": 5.061176470588236, + "grad_norm": 0.164184261428605, + "learning_rate": 2.4779523970574703e-05, + "loss": 0.048, + "step": 10755 + }, + { + "epoch": 5.06164705882353, + "grad_norm": 0.0846192591564175, + "learning_rate": 2.4755395376465386e-05, + "loss": 0.0166, + "step": 10756 + }, + { + "epoch": 5.062117647058823, + "grad_norm": 0.10763782939386933, + "learning_rate": 2.4731277478859586e-05, + "loss": 0.0278, + "step": 10757 + }, + { + "epoch": 5.062588235294117, + "grad_norm": 0.08527074549150146, + "learning_rate": 2.4707170279817118e-05, + "loss": 0.0186, + "step": 10758 + }, + { + "epoch": 5.063058823529412, + "grad_norm": 0.08633637510367315, + "learning_rate": 2.4683073781396924e-05, + "loss": 0.0154, + "step": 10759 + }, + { + "epoch": 5.063529411764706, + "grad_norm": 2.1668104058138994e-05, + "learning_rate": 2.465898798565685e-05, + "loss": 0.0, + "step": 10760 + }, + { + "epoch": 5.064, + "grad_norm": 5.473731349052838e-05, + "learning_rate": 2.4634912894653996e-05, + "loss": 0.0, + "step": 10761 + }, + { + "epoch": 5.064470588235294, + "grad_norm": 0.08796394965393262, + "learning_rate": 2.461084851044454e-05, + "loss": 0.0161, + "step": 10762 + }, + { + "epoch": 5.0649411764705885, + "grad_norm": 3.978076921143843e-05, + "learning_rate": 2.4586794835083613e-05, + "loss": 0.0, + "step": 10763 + }, + { + "epoch": 5.065411764705883, + "grad_norm": 0.034064810217727864, + "learning_rate": 2.4562751870625597e-05, + "loss": 0.0014, + "step": 10764 + }, + { + "epoch": 5.065882352941176, + "grad_norm": 0.1302056221810056, + "learning_rate": 2.4538719619123786e-05, + "loss": 0.0327, + "step": 10765 + }, + { + "epoch": 5.06635294117647, + "grad_norm": 0.11559295526607603, + "learning_rate": 2.4514698082630706e-05, + "loss": 0.0343, + "step": 10766 + }, + { + "epoch": 5.0668235294117645, + "grad_norm": 0.1011968690490934, + "learning_rate": 2.449068726319795e-05, + "loss": 0.0141, + "step": 10767 + }, + { + "epoch": 5.067294117647059, + "grad_norm": 8.101409250982994e-05, + "learning_rate": 2.4466687162876093e-05, + "loss": 0.0, + "step": 10768 + }, + { + "epoch": 5.067764705882353, + "grad_norm": 0.1031249688023667, + "learning_rate": 2.4442697783714894e-05, + "loss": 0.0179, + "step": 10769 + }, + { + "epoch": 5.068235294117647, + "grad_norm": 0.07290883469765574, + "learning_rate": 2.4418719127763183e-05, + "loss": 0.0103, + "step": 10770 + }, + { + "epoch": 5.068705882352941, + "grad_norm": 0.12523269265330814, + "learning_rate": 2.43947511970688e-05, + "loss": 0.0235, + "step": 10771 + }, + { + "epoch": 5.069176470588236, + "grad_norm": 0.07521700725666182, + "learning_rate": 2.437079399367875e-05, + "loss": 0.01, + "step": 10772 + }, + { + "epoch": 5.06964705882353, + "grad_norm": 0.10499060462243663, + "learning_rate": 2.434684751963914e-05, + "loss": 0.0155, + "step": 10773 + }, + { + "epoch": 5.070117647058823, + "grad_norm": 0.09175191839733396, + "learning_rate": 2.4322911776995002e-05, + "loss": 0.0135, + "step": 10774 + }, + { + "epoch": 5.070588235294117, + "grad_norm": 0.15940819348004964, + "learning_rate": 2.429898676779072e-05, + "loss": 0.0381, + "step": 10775 + }, + { + "epoch": 5.071058823529412, + "grad_norm": 0.12610510359948923, + "learning_rate": 2.4275072494069487e-05, + "loss": 0.0315, + "step": 10776 + }, + { + "epoch": 5.071529411764706, + "grad_norm": 2.024596542825884e-05, + "learning_rate": 2.425116895787374e-05, + "loss": 0.0, + "step": 10777 + }, + { + "epoch": 5.072, + "grad_norm": 0.1500862113231407, + "learning_rate": 2.4227276161244995e-05, + "loss": 0.0354, + "step": 10778 + }, + { + "epoch": 5.072470588235294, + "grad_norm": 0.1334007956101148, + "learning_rate": 2.4203394106223755e-05, + "loss": 0.0455, + "step": 10779 + }, + { + "epoch": 5.0729411764705885, + "grad_norm": 0.13449051571875797, + "learning_rate": 2.417952279484967e-05, + "loss": 0.0281, + "step": 10780 + }, + { + "epoch": 5.073411764705883, + "grad_norm": 0.13721948290571864, + "learning_rate": 2.4155662229161527e-05, + "loss": 0.0325, + "step": 10781 + }, + { + "epoch": 5.073882352941176, + "grad_norm": 0.1433453929237897, + "learning_rate": 2.4131812411197045e-05, + "loss": 0.0309, + "step": 10782 + }, + { + "epoch": 5.07435294117647, + "grad_norm": 0.1661302602833352, + "learning_rate": 2.4107973342993174e-05, + "loss": 0.0647, + "step": 10783 + }, + { + "epoch": 5.0748235294117645, + "grad_norm": 0.1516493769791845, + "learning_rate": 2.408414502658591e-05, + "loss": 0.0338, + "step": 10784 + }, + { + "epoch": 5.075294117647059, + "grad_norm": 0.1435814455179054, + "learning_rate": 2.4060327464010254e-05, + "loss": 0.0421, + "step": 10785 + }, + { + "epoch": 5.075764705882353, + "grad_norm": 0.11020205892856773, + "learning_rate": 2.4036520657300346e-05, + "loss": 0.018, + "step": 10786 + }, + { + "epoch": 5.076235294117647, + "grad_norm": 0.00016977776393631088, + "learning_rate": 2.401272460848946e-05, + "loss": 0.0, + "step": 10787 + }, + { + "epoch": 5.076705882352941, + "grad_norm": 0.059876135971561974, + "learning_rate": 2.3988939319609817e-05, + "loss": 0.0061, + "step": 10788 + }, + { + "epoch": 5.077176470588236, + "grad_norm": 0.174070414335404, + "learning_rate": 2.3965164792692833e-05, + "loss": 0.0549, + "step": 10789 + }, + { + "epoch": 5.07764705882353, + "grad_norm": 0.00011712846068637796, + "learning_rate": 2.3941401029769024e-05, + "loss": 0.0, + "step": 10790 + }, + { + "epoch": 5.078117647058823, + "grad_norm": 0.11472719410220801, + "learning_rate": 2.3917648032867844e-05, + "loss": 0.0254, + "step": 10791 + }, + { + "epoch": 5.078588235294117, + "grad_norm": 2.6840257670385536e-06, + "learning_rate": 2.3893905804017977e-05, + "loss": 0.0, + "step": 10792 + }, + { + "epoch": 5.079058823529412, + "grad_norm": 0.07958478543313446, + "learning_rate": 2.387017434524707e-05, + "loss": 0.0084, + "step": 10793 + }, + { + "epoch": 5.079529411764706, + "grad_norm": 0.14414280198000157, + "learning_rate": 2.3846453658581927e-05, + "loss": 0.0382, + "step": 10794 + }, + { + "epoch": 5.08, + "grad_norm": 0.10145067904120494, + "learning_rate": 2.382274374604847e-05, + "loss": 0.0037, + "step": 10795 + }, + { + "epoch": 5.080470588235294, + "grad_norm": 0.10902585954658625, + "learning_rate": 2.3799044609671558e-05, + "loss": 0.0186, + "step": 10796 + }, + { + "epoch": 5.0809411764705885, + "grad_norm": 0.12535968565923786, + "learning_rate": 2.3775356251475248e-05, + "loss": 0.0287, + "step": 10797 + }, + { + "epoch": 5.081411764705883, + "grad_norm": 9.635093978434492e-05, + "learning_rate": 2.375167867348268e-05, + "loss": 0.0, + "step": 10798 + }, + { + "epoch": 5.081882352941176, + "grad_norm": 0.10061714474871375, + "learning_rate": 2.3728011877715947e-05, + "loss": 0.0206, + "step": 10799 + }, + { + "epoch": 5.08235294117647, + "grad_norm": 0.15672965882510445, + "learning_rate": 2.3704355866196373e-05, + "loss": 0.0391, + "step": 10800 + }, + { + "epoch": 5.0828235294117645, + "grad_norm": 0.0002791739261756954, + "learning_rate": 2.3680710640944328e-05, + "loss": 0.0, + "step": 10801 + }, + { + "epoch": 5.083294117647059, + "grad_norm": 0.12479575437674224, + "learning_rate": 2.3657076203979132e-05, + "loss": 0.0282, + "step": 10802 + }, + { + "epoch": 5.083764705882353, + "grad_norm": 0.08485473580378687, + "learning_rate": 2.363345255731935e-05, + "loss": 0.0142, + "step": 10803 + }, + { + "epoch": 5.084235294117647, + "grad_norm": 3.7117634731715662e-06, + "learning_rate": 2.360983970298257e-05, + "loss": 0.0, + "step": 10804 + }, + { + "epoch": 5.084705882352941, + "grad_norm": 0.1047059516890447, + "learning_rate": 2.3586237642985396e-05, + "loss": 0.0186, + "step": 10805 + }, + { + "epoch": 5.085176470588236, + "grad_norm": 4.407510530655227e-06, + "learning_rate": 2.356264637934358e-05, + "loss": 0.0, + "step": 10806 + }, + { + "epoch": 5.085647058823529, + "grad_norm": 0.08147080546316607, + "learning_rate": 2.3539065914071963e-05, + "loss": 0.0218, + "step": 10807 + }, + { + "epoch": 5.086117647058823, + "grad_norm": 0.14887798721719467, + "learning_rate": 2.351549624918437e-05, + "loss": 0.0414, + "step": 10808 + }, + { + "epoch": 5.086588235294117, + "grad_norm": 0.1364972790271801, + "learning_rate": 2.34919373866938e-05, + "loss": 0.0286, + "step": 10809 + }, + { + "epoch": 5.087058823529412, + "grad_norm": 0.12934563303077348, + "learning_rate": 2.3468389328612348e-05, + "loss": 0.0311, + "step": 10810 + }, + { + "epoch": 5.087529411764706, + "grad_norm": 0.042248937309695866, + "learning_rate": 2.344485207695102e-05, + "loss": 0.0034, + "step": 10811 + }, + { + "epoch": 5.088, + "grad_norm": 0.07858125969215776, + "learning_rate": 2.3421325633720094e-05, + "loss": 0.0161, + "step": 10812 + }, + { + "epoch": 5.088470588235294, + "grad_norm": 0.07197404328901984, + "learning_rate": 2.3397810000928845e-05, + "loss": 0.0096, + "step": 10813 + }, + { + "epoch": 5.0889411764705885, + "grad_norm": 0.10499266115779539, + "learning_rate": 2.337430518058557e-05, + "loss": 0.0217, + "step": 10814 + }, + { + "epoch": 5.089411764705883, + "grad_norm": 0.07580722034247324, + "learning_rate": 2.335081117469777e-05, + "loss": 0.0079, + "step": 10815 + }, + { + "epoch": 5.089882352941176, + "grad_norm": 0.16910753605326176, + "learning_rate": 2.3327327985271865e-05, + "loss": 0.0505, + "step": 10816 + }, + { + "epoch": 5.09035294117647, + "grad_norm": 0.16020740028450978, + "learning_rate": 2.3303855614313483e-05, + "loss": 0.0352, + "step": 10817 + }, + { + "epoch": 5.0908235294117645, + "grad_norm": 0.08185107273437003, + "learning_rate": 2.328039406382727e-05, + "loss": 0.0108, + "step": 10818 + }, + { + "epoch": 5.091294117647059, + "grad_norm": 0.07366445508512258, + "learning_rate": 2.325694333581695e-05, + "loss": 0.018, + "step": 10819 + }, + { + "epoch": 5.091764705882353, + "grad_norm": 8.621488159485683e-06, + "learning_rate": 2.32335034322854e-05, + "loss": 0.0, + "step": 10820 + }, + { + "epoch": 5.092235294117647, + "grad_norm": 0.10424651554941615, + "learning_rate": 2.3210074355234392e-05, + "loss": 0.0319, + "step": 10821 + }, + { + "epoch": 5.092705882352941, + "grad_norm": 0.09941168686024449, + "learning_rate": 2.318665610666493e-05, + "loss": 0.0182, + "step": 10822 + }, + { + "epoch": 5.093176470588236, + "grad_norm": 0.07859708967849985, + "learning_rate": 2.3163248688577098e-05, + "loss": 0.0168, + "step": 10823 + }, + { + "epoch": 5.09364705882353, + "grad_norm": 0.05150505448455728, + "learning_rate": 2.3139852102969906e-05, + "loss": 0.0077, + "step": 10824 + }, + { + "epoch": 5.094117647058823, + "grad_norm": 9.865264284214879e-05, + "learning_rate": 2.31164663518416e-05, + "loss": 0.0, + "step": 10825 + }, + { + "epoch": 5.094588235294117, + "grad_norm": 0.30538526441385194, + "learning_rate": 2.3093091437189476e-05, + "loss": 0.0413, + "step": 10826 + }, + { + "epoch": 5.095058823529412, + "grad_norm": 0.13231150394646454, + "learning_rate": 2.306972736100977e-05, + "loss": 0.0398, + "step": 10827 + }, + { + "epoch": 5.095529411764706, + "grad_norm": 0.00010225342815731853, + "learning_rate": 2.304637412529794e-05, + "loss": 0.0, + "step": 10828 + }, + { + "epoch": 5.096, + "grad_norm": 0.08595821860617604, + "learning_rate": 2.3023031732048504e-05, + "loss": 0.0156, + "step": 10829 + }, + { + "epoch": 5.096470588235294, + "grad_norm": 0.0960910337915622, + "learning_rate": 2.2999700183254934e-05, + "loss": 0.0137, + "step": 10830 + }, + { + "epoch": 5.0969411764705885, + "grad_norm": 0.09069125719220399, + "learning_rate": 2.29763794809099e-05, + "loss": 0.0144, + "step": 10831 + }, + { + "epoch": 5.097411764705883, + "grad_norm": 0.1361067657581076, + "learning_rate": 2.2953069627005132e-05, + "loss": 0.0317, + "step": 10832 + }, + { + "epoch": 5.097882352941176, + "grad_norm": 1.8884881148634673e-05, + "learning_rate": 2.2929770623531368e-05, + "loss": 0.0, + "step": 10833 + }, + { + "epoch": 5.09835294117647, + "grad_norm": 0.07544852682203855, + "learning_rate": 2.290648247247845e-05, + "loss": 0.0087, + "step": 10834 + }, + { + "epoch": 5.0988235294117645, + "grad_norm": 0.09327429381763833, + "learning_rate": 2.2883205175835368e-05, + "loss": 0.0104, + "step": 10835 + }, + { + "epoch": 5.099294117647059, + "grad_norm": 0.1125798764226048, + "learning_rate": 2.2859938735590017e-05, + "loss": 0.0258, + "step": 10836 + }, + { + "epoch": 5.099764705882353, + "grad_norm": 0.15859055288209434, + "learning_rate": 2.2836683153729522e-05, + "loss": 0.0588, + "step": 10837 + }, + { + "epoch": 5.100235294117647, + "grad_norm": 0.027872311649955343, + "learning_rate": 2.2813438432240067e-05, + "loss": 0.0015, + "step": 10838 + }, + { + "epoch": 5.100705882352941, + "grad_norm": 0.140473218669162, + "learning_rate": 2.2790204573106756e-05, + "loss": 0.0304, + "step": 10839 + }, + { + "epoch": 5.101176470588236, + "grad_norm": 0.10253195858123237, + "learning_rate": 2.276698157831396e-05, + "loss": 0.0259, + "step": 10840 + }, + { + "epoch": 5.101647058823529, + "grad_norm": 0.1356189666632632, + "learning_rate": 2.2743769449845035e-05, + "loss": 0.0192, + "step": 10841 + }, + { + "epoch": 5.102117647058823, + "grad_norm": 0.16618426153918486, + "learning_rate": 2.272056818968235e-05, + "loss": 0.0517, + "step": 10842 + }, + { + "epoch": 5.102588235294117, + "grad_norm": 0.048702509318706254, + "learning_rate": 2.269737779980748e-05, + "loss": 0.0049, + "step": 10843 + }, + { + "epoch": 5.103058823529412, + "grad_norm": 0.024800467634840465, + "learning_rate": 2.2674198282200934e-05, + "loss": 0.0013, + "step": 10844 + }, + { + "epoch": 5.103529411764706, + "grad_norm": 0.11327573301218821, + "learning_rate": 2.2651029638842372e-05, + "loss": 0.0203, + "step": 10845 + }, + { + "epoch": 5.104, + "grad_norm": 0.1323343819846485, + "learning_rate": 2.262787187171058e-05, + "loss": 0.0352, + "step": 10846 + }, + { + "epoch": 5.104470588235294, + "grad_norm": 5.532258859673949e-05, + "learning_rate": 2.2604724982783224e-05, + "loss": 0.0, + "step": 10847 + }, + { + "epoch": 5.1049411764705885, + "grad_norm": 0.14816944552143202, + "learning_rate": 2.258158897403724e-05, + "loss": 0.0377, + "step": 10848 + }, + { + "epoch": 5.105411764705883, + "grad_norm": 0.15193109582070544, + "learning_rate": 2.2558463847448566e-05, + "loss": 0.0284, + "step": 10849 + }, + { + "epoch": 5.105882352941176, + "grad_norm": 0.14471426201073107, + "learning_rate": 2.2535349604992153e-05, + "loss": 0.0465, + "step": 10850 + }, + { + "epoch": 5.10635294117647, + "grad_norm": 0.2062039581740649, + "learning_rate": 2.251224624864209e-05, + "loss": 0.0766, + "step": 10851 + }, + { + "epoch": 5.1068235294117645, + "grad_norm": 0.07294004017267017, + "learning_rate": 2.2489153780371566e-05, + "loss": 0.0078, + "step": 10852 + }, + { + "epoch": 5.107294117647059, + "grad_norm": 0.11236395215697159, + "learning_rate": 2.2466072202152702e-05, + "loss": 0.0134, + "step": 10853 + }, + { + "epoch": 5.107764705882353, + "grad_norm": 1.97463396912349e-06, + "learning_rate": 2.2443001515956823e-05, + "loss": 0.0, + "step": 10854 + }, + { + "epoch": 5.108235294117647, + "grad_norm": 3.7733261104082318e-06, + "learning_rate": 2.24199417237543e-05, + "loss": 0.0, + "step": 10855 + }, + { + "epoch": 5.108705882352941, + "grad_norm": 3.159077179553223e-05, + "learning_rate": 2.2396892827514507e-05, + "loss": 0.0, + "step": 10856 + }, + { + "epoch": 5.109176470588236, + "grad_norm": 4.867506607187974e-06, + "learning_rate": 2.2373854829205957e-05, + "loss": 0.0, + "step": 10857 + }, + { + "epoch": 5.109647058823529, + "grad_norm": 1.3923382839610311e-05, + "learning_rate": 2.235082773079624e-05, + "loss": 0.0, + "step": 10858 + }, + { + "epoch": 5.110117647058823, + "grad_norm": 0.1035949960372764, + "learning_rate": 2.2327811534251933e-05, + "loss": 0.0245, + "step": 10859 + }, + { + "epoch": 5.110588235294117, + "grad_norm": 0.019873301096177548, + "learning_rate": 2.230480624153873e-05, + "loss": 0.0008, + "step": 10860 + }, + { + "epoch": 5.111058823529412, + "grad_norm": 0.08450337027886849, + "learning_rate": 2.2281811854621445e-05, + "loss": 0.0147, + "step": 10861 + }, + { + "epoch": 5.111529411764706, + "grad_norm": 0.1493129183416146, + "learning_rate": 2.2258828375463855e-05, + "loss": 0.0455, + "step": 10862 + }, + { + "epoch": 5.112, + "grad_norm": 0.15322643263766203, + "learning_rate": 2.223585580602889e-05, + "loss": 0.0456, + "step": 10863 + }, + { + "epoch": 5.112470588235294, + "grad_norm": 0.09074325577679562, + "learning_rate": 2.221289414827853e-05, + "loss": 0.0177, + "step": 10864 + }, + { + "epoch": 5.1129411764705885, + "grad_norm": 3.469825713139835e-05, + "learning_rate": 2.218994340417378e-05, + "loss": 0.0, + "step": 10865 + }, + { + "epoch": 5.113411764705883, + "grad_norm": 9.651304043725844e-06, + "learning_rate": 2.2167003575674824e-05, + "loss": 0.0, + "step": 10866 + }, + { + "epoch": 5.113882352941176, + "grad_norm": 0.059740709422471915, + "learning_rate": 2.2144074664740723e-05, + "loss": 0.0068, + "step": 10867 + }, + { + "epoch": 5.11435294117647, + "grad_norm": 0.13686359717103364, + "learning_rate": 2.2121156673329794e-05, + "loss": 0.0413, + "step": 10868 + }, + { + "epoch": 5.1148235294117645, + "grad_norm": 0.0008786452222502697, + "learning_rate": 2.2098249603399348e-05, + "loss": 0.0, + "step": 10869 + }, + { + "epoch": 5.115294117647059, + "grad_norm": 0.03353035061592402, + "learning_rate": 2.2075353456905727e-05, + "loss": 0.001, + "step": 10870 + }, + { + "epoch": 5.115764705882353, + "grad_norm": 0.12388740878433221, + "learning_rate": 2.2052468235804393e-05, + "loss": 0.0175, + "step": 10871 + }, + { + "epoch": 5.116235294117647, + "grad_norm": 0.05924705338884559, + "learning_rate": 2.2029593942049838e-05, + "loss": 0.0086, + "step": 10872 + }, + { + "epoch": 5.116705882352941, + "grad_norm": 0.17446703851701773, + "learning_rate": 2.200673057759565e-05, + "loss": 0.0498, + "step": 10873 + }, + { + "epoch": 5.117176470588236, + "grad_norm": 1.8326498583959543e-05, + "learning_rate": 2.1983878144394497e-05, + "loss": 0.0, + "step": 10874 + }, + { + "epoch": 5.117647058823529, + "grad_norm": 8.300053672662378e-05, + "learning_rate": 2.1961036644398035e-05, + "loss": 0.0, + "step": 10875 + }, + { + "epoch": 5.118117647058823, + "grad_norm": 0.019574682222459703, + "learning_rate": 2.193820607955707e-05, + "loss": 0.001, + "step": 10876 + }, + { + "epoch": 5.118588235294117, + "grad_norm": 0.08953034227566306, + "learning_rate": 2.1915386451821494e-05, + "loss": 0.0161, + "step": 10877 + }, + { + "epoch": 5.119058823529412, + "grad_norm": 0.15198094657947978, + "learning_rate": 2.1892577763140112e-05, + "loss": 0.0422, + "step": 10878 + }, + { + "epoch": 5.119529411764706, + "grad_norm": 8.091832951190561e-05, + "learning_rate": 2.186978001546097e-05, + "loss": 0.0, + "step": 10879 + }, + { + "epoch": 5.12, + "grad_norm": 0.21043098306128385, + "learning_rate": 2.1846993210731108e-05, + "loss": 0.0509, + "step": 10880 + }, + { + "epoch": 5.120470588235294, + "grad_norm": 0.0798734001354702, + "learning_rate": 2.1824217350896585e-05, + "loss": 0.0145, + "step": 10881 + }, + { + "epoch": 5.1209411764705886, + "grad_norm": 0.13592182702230216, + "learning_rate": 2.180145243790261e-05, + "loss": 0.0377, + "step": 10882 + }, + { + "epoch": 5.121411764705883, + "grad_norm": 0.1371917508151013, + "learning_rate": 2.1778698473693435e-05, + "loss": 0.0198, + "step": 10883 + }, + { + "epoch": 5.121882352941176, + "grad_norm": 0.08840975736830285, + "learning_rate": 2.1755955460212282e-05, + "loss": 0.012, + "step": 10884 + }, + { + "epoch": 5.12235294117647, + "grad_norm": 0.10048627599733037, + "learning_rate": 2.1733223399401585e-05, + "loss": 0.0199, + "step": 10885 + }, + { + "epoch": 5.1228235294117646, + "grad_norm": 0.08018640566382876, + "learning_rate": 2.1710502293202786e-05, + "loss": 0.0147, + "step": 10886 + }, + { + "epoch": 5.123294117647059, + "grad_norm": 2.5744670986107652e-05, + "learning_rate": 2.1687792143556303e-05, + "loss": 0.0, + "step": 10887 + }, + { + "epoch": 5.123764705882353, + "grad_norm": 0.1289996068240859, + "learning_rate": 2.1665092952401754e-05, + "loss": 0.0295, + "step": 10888 + }, + { + "epoch": 5.124235294117647, + "grad_norm": 0.09068235618447207, + "learning_rate": 2.1642404721677765e-05, + "loss": 0.0168, + "step": 10889 + }, + { + "epoch": 5.124705882352941, + "grad_norm": 2.160029529954212e-06, + "learning_rate": 2.1619727453321974e-05, + "loss": 0.0, + "step": 10890 + }, + { + "epoch": 5.125176470588236, + "grad_norm": 0.16767435267546207, + "learning_rate": 2.159706114927116e-05, + "loss": 0.0402, + "step": 10891 + }, + { + "epoch": 5.125647058823529, + "grad_norm": 3.381287531777593e-05, + "learning_rate": 2.157440581146118e-05, + "loss": 0.0, + "step": 10892 + }, + { + "epoch": 5.126117647058823, + "grad_norm": 0.06240648729191534, + "learning_rate": 2.155176144182682e-05, + "loss": 0.0088, + "step": 10893 + }, + { + "epoch": 5.1265882352941174, + "grad_norm": 0.09067150712423543, + "learning_rate": 2.1529128042302113e-05, + "loss": 0.0113, + "step": 10894 + }, + { + "epoch": 5.127058823529412, + "grad_norm": 0.1375910471963175, + "learning_rate": 2.150650561481998e-05, + "loss": 0.033, + "step": 10895 + }, + { + "epoch": 5.127529411764706, + "grad_norm": 0.00014278471816448705, + "learning_rate": 2.1483894161312525e-05, + "loss": 0.0, + "step": 10896 + }, + { + "epoch": 5.128, + "grad_norm": 0.02008632313804647, + "learning_rate": 2.1461293683710905e-05, + "loss": 0.0008, + "step": 10897 + }, + { + "epoch": 5.128470588235294, + "grad_norm": 0.11364435625639983, + "learning_rate": 2.143870418394526e-05, + "loss": 0.0288, + "step": 10898 + }, + { + "epoch": 5.128941176470589, + "grad_norm": 0.0682977694382856, + "learning_rate": 2.141612566394486e-05, + "loss": 0.0087, + "step": 10899 + }, + { + "epoch": 5.129411764705883, + "grad_norm": 0.00011767689678013803, + "learning_rate": 2.1393558125638066e-05, + "loss": 0.0, + "step": 10900 + }, + { + "epoch": 5.129882352941176, + "grad_norm": 0.28265730235950365, + "learning_rate": 2.1371001570952186e-05, + "loss": 0.045, + "step": 10901 + }, + { + "epoch": 5.13035294117647, + "grad_norm": 0.024639263750275357, + "learning_rate": 2.1348456001813674e-05, + "loss": 0.0009, + "step": 10902 + }, + { + "epoch": 5.130823529411765, + "grad_norm": 0.15491785928176705, + "learning_rate": 2.132592142014811e-05, + "loss": 0.0334, + "step": 10903 + }, + { + "epoch": 5.131294117647059, + "grad_norm": 0.11895759232865744, + "learning_rate": 2.130339782787996e-05, + "loss": 0.0205, + "step": 10904 + }, + { + "epoch": 5.131764705882353, + "grad_norm": 0.029446521164632627, + "learning_rate": 2.1280885226932877e-05, + "loss": 0.0015, + "step": 10905 + }, + { + "epoch": 5.132235294117647, + "grad_norm": 0.06307703445149665, + "learning_rate": 2.125838361922961e-05, + "loss": 0.0031, + "step": 10906 + }, + { + "epoch": 5.1327058823529415, + "grad_norm": 0.07258768872002622, + "learning_rate": 2.1235893006691786e-05, + "loss": 0.0118, + "step": 10907 + }, + { + "epoch": 5.133176470588236, + "grad_norm": 0.07162558568757144, + "learning_rate": 2.1213413391240335e-05, + "loss": 0.012, + "step": 10908 + }, + { + "epoch": 5.133647058823529, + "grad_norm": 0.1831643329355533, + "learning_rate": 2.119094477479506e-05, + "loss": 0.0278, + "step": 10909 + }, + { + "epoch": 5.134117647058823, + "grad_norm": 0.07949570660198482, + "learning_rate": 2.116848715927489e-05, + "loss": 0.0063, + "step": 10910 + }, + { + "epoch": 5.1345882352941175, + "grad_norm": 0.08088165841980662, + "learning_rate": 2.114604054659787e-05, + "loss": 0.007, + "step": 10911 + }, + { + "epoch": 5.135058823529412, + "grad_norm": 0.09018873128137017, + "learning_rate": 2.1123604938680987e-05, + "loss": 0.0083, + "step": 10912 + }, + { + "epoch": 5.135529411764706, + "grad_norm": 0.047197225089568536, + "learning_rate": 2.1101180337440378e-05, + "loss": 0.0031, + "step": 10913 + }, + { + "epoch": 5.136, + "grad_norm": 0.18843197287220648, + "learning_rate": 2.1078766744791238e-05, + "loss": 0.0871, + "step": 10914 + }, + { + "epoch": 5.136470588235294, + "grad_norm": 5.195379859595377e-06, + "learning_rate": 2.1056364162647736e-05, + "loss": 0.0, + "step": 10915 + }, + { + "epoch": 5.136941176470589, + "grad_norm": 0.0812620725026524, + "learning_rate": 2.1033972592923214e-05, + "loss": 0.0097, + "step": 10916 + }, + { + "epoch": 5.137411764705883, + "grad_norm": 0.10456671884828168, + "learning_rate": 2.1011592037530023e-05, + "loss": 0.017, + "step": 10917 + }, + { + "epoch": 5.137882352941176, + "grad_norm": 0.07405895137464333, + "learning_rate": 2.098922249837953e-05, + "loss": 0.0114, + "step": 10918 + }, + { + "epoch": 5.13835294117647, + "grad_norm": 0.11699473464793482, + "learning_rate": 2.096686397738223e-05, + "loss": 0.0203, + "step": 10919 + }, + { + "epoch": 5.138823529411765, + "grad_norm": 0.13336704840401295, + "learning_rate": 2.094451647644769e-05, + "loss": 0.0422, + "step": 10920 + }, + { + "epoch": 5.139294117647059, + "grad_norm": 0.10234561573276806, + "learning_rate": 2.0922179997484424e-05, + "loss": 0.0249, + "step": 10921 + }, + { + "epoch": 5.139764705882353, + "grad_norm": 0.1224481555130787, + "learning_rate": 2.0899854542400147e-05, + "loss": 0.0298, + "step": 10922 + }, + { + "epoch": 5.140235294117647, + "grad_norm": 0.08861443740128569, + "learning_rate": 2.087754011310147e-05, + "loss": 0.008, + "step": 10923 + }, + { + "epoch": 5.1407058823529415, + "grad_norm": 0.06658745164303415, + "learning_rate": 2.0855236711494223e-05, + "loss": 0.0045, + "step": 10924 + }, + { + "epoch": 5.141176470588236, + "grad_norm": 0.11019435759495534, + "learning_rate": 2.083294433948324e-05, + "loss": 0.0145, + "step": 10925 + }, + { + "epoch": 5.141647058823529, + "grad_norm": 0.1330790228089585, + "learning_rate": 2.0810662998972345e-05, + "loss": 0.0267, + "step": 10926 + }, + { + "epoch": 5.142117647058823, + "grad_norm": 0.06305553928949686, + "learning_rate": 2.0788392691864504e-05, + "loss": 0.0034, + "step": 10927 + }, + { + "epoch": 5.1425882352941175, + "grad_norm": 0.056274680747020946, + "learning_rate": 2.0766133420061732e-05, + "loss": 0.0054, + "step": 10928 + }, + { + "epoch": 5.143058823529412, + "grad_norm": 0.13552089635308492, + "learning_rate": 2.0743885185465025e-05, + "loss": 0.0347, + "step": 10929 + }, + { + "epoch": 5.143529411764706, + "grad_norm": 0.00010479333739027564, + "learning_rate": 2.0721647989974514e-05, + "loss": 0.0, + "step": 10930 + }, + { + "epoch": 5.144, + "grad_norm": 0.12193512687032851, + "learning_rate": 2.0699421835489417e-05, + "loss": 0.0321, + "step": 10931 + }, + { + "epoch": 5.144470588235294, + "grad_norm": 0.15926193238280806, + "learning_rate": 2.0677206723907857e-05, + "loss": 0.0458, + "step": 10932 + }, + { + "epoch": 5.144941176470589, + "grad_norm": 0.03917375273989006, + "learning_rate": 2.0655002657127188e-05, + "loss": 0.0051, + "step": 10933 + }, + { + "epoch": 5.145411764705882, + "grad_norm": 0.1006762124895159, + "learning_rate": 2.0632809637043747e-05, + "loss": 0.0203, + "step": 10934 + }, + { + "epoch": 5.145882352941176, + "grad_norm": 0.15567502834319533, + "learning_rate": 2.0610627665552875e-05, + "loss": 0.0384, + "step": 10935 + }, + { + "epoch": 5.14635294117647, + "grad_norm": 0.06156828663581854, + "learning_rate": 2.058845674454905e-05, + "loss": 0.0089, + "step": 10936 + }, + { + "epoch": 5.146823529411765, + "grad_norm": 0.16769647179802327, + "learning_rate": 2.0566296875925813e-05, + "loss": 0.0401, + "step": 10937 + }, + { + "epoch": 5.147294117647059, + "grad_norm": 2.8945968654773655e-06, + "learning_rate": 2.054414806157566e-05, + "loss": 0.0, + "step": 10938 + }, + { + "epoch": 5.147764705882353, + "grad_norm": 0.04893098595898152, + "learning_rate": 2.052201030339024e-05, + "loss": 0.0033, + "step": 10939 + }, + { + "epoch": 5.148235294117647, + "grad_norm": 0.11154754384308307, + "learning_rate": 2.0499883603260274e-05, + "loss": 0.0219, + "step": 10940 + }, + { + "epoch": 5.1487058823529415, + "grad_norm": 0.04975389555780204, + "learning_rate": 2.0477767963075386e-05, + "loss": 0.0046, + "step": 10941 + }, + { + "epoch": 5.149176470588236, + "grad_norm": 0.22372589039736715, + "learning_rate": 2.045566338472443e-05, + "loss": 0.1088, + "step": 10942 + }, + { + "epoch": 5.149647058823529, + "grad_norm": 0.03991974803245738, + "learning_rate": 2.043356987009526e-05, + "loss": 0.0038, + "step": 10943 + }, + { + "epoch": 5.150117647058823, + "grad_norm": 0.00037641924818116417, + "learning_rate": 2.041148742107471e-05, + "loss": 0.0, + "step": 10944 + }, + { + "epoch": 5.1505882352941175, + "grad_norm": 0.13318866740198168, + "learning_rate": 2.038941603954879e-05, + "loss": 0.0376, + "step": 10945 + }, + { + "epoch": 5.151058823529412, + "grad_norm": 4.357257377198396e-06, + "learning_rate": 2.036735572740245e-05, + "loss": 0.0, + "step": 10946 + }, + { + "epoch": 5.151529411764706, + "grad_norm": 0.10243258499849976, + "learning_rate": 2.0345306486519787e-05, + "loss": 0.0136, + "step": 10947 + }, + { + "epoch": 5.152, + "grad_norm": 0.12039107195611506, + "learning_rate": 2.032326831878392e-05, + "loss": 0.0234, + "step": 10948 + }, + { + "epoch": 5.152470588235294, + "grad_norm": 0.1848367197788069, + "learning_rate": 2.0301241226076963e-05, + "loss": 0.0472, + "step": 10949 + }, + { + "epoch": 5.152941176470589, + "grad_norm": 0.12705719518594075, + "learning_rate": 2.027922521028018e-05, + "loss": 0.0182, + "step": 10950 + }, + { + "epoch": 5.153411764705883, + "grad_norm": 0.07407293449675786, + "learning_rate": 2.025722027327385e-05, + "loss": 0.0102, + "step": 10951 + }, + { + "epoch": 5.153882352941176, + "grad_norm": 5.300801115364222e-06, + "learning_rate": 2.0235226416937284e-05, + "loss": 0.0, + "step": 10952 + }, + { + "epoch": 5.15435294117647, + "grad_norm": 0.3233904670377581, + "learning_rate": 2.0213243643148914e-05, + "loss": 0.046, + "step": 10953 + }, + { + "epoch": 5.154823529411765, + "grad_norm": 1.3435604525090078e-05, + "learning_rate": 2.0191271953786104e-05, + "loss": 0.0, + "step": 10954 + }, + { + "epoch": 5.155294117647059, + "grad_norm": 0.07340654143926728, + "learning_rate": 2.016931135072537e-05, + "loss": 0.0135, + "step": 10955 + }, + { + "epoch": 5.155764705882353, + "grad_norm": 0.13174341512739832, + "learning_rate": 2.0147361835842296e-05, + "loss": 0.0142, + "step": 10956 + }, + { + "epoch": 5.156235294117647, + "grad_norm": 0.18951601225558193, + "learning_rate": 2.0125423411011416e-05, + "loss": 0.0837, + "step": 10957 + }, + { + "epoch": 5.1567058823529415, + "grad_norm": 0.052004607035658845, + "learning_rate": 2.0103496078106396e-05, + "loss": 0.0054, + "step": 10958 + }, + { + "epoch": 5.157176470588236, + "grad_norm": 0.06000528241069923, + "learning_rate": 2.0081579839000005e-05, + "loss": 0.0055, + "step": 10959 + }, + { + "epoch": 5.157647058823529, + "grad_norm": 0.08878816659438478, + "learning_rate": 2.00596746955639e-05, + "loss": 0.0177, + "step": 10960 + }, + { + "epoch": 5.158117647058823, + "grad_norm": 0.08678914153467758, + "learning_rate": 2.0037780649668912e-05, + "loss": 0.0225, + "step": 10961 + }, + { + "epoch": 5.1585882352941175, + "grad_norm": 0.04408436501781631, + "learning_rate": 2.001589770318498e-05, + "loss": 0.0026, + "step": 10962 + }, + { + "epoch": 5.159058823529412, + "grad_norm": 4.889093391581954e-06, + "learning_rate": 1.9994025857980896e-05, + "loss": 0.0, + "step": 10963 + }, + { + "epoch": 5.159529411764706, + "grad_norm": 0.11906081399044495, + "learning_rate": 1.9972165115924695e-05, + "loss": 0.0321, + "step": 10964 + }, + { + "epoch": 5.16, + "grad_norm": 0.17520940041735172, + "learning_rate": 1.99503154788834e-05, + "loss": 0.0579, + "step": 10965 + }, + { + "epoch": 5.160470588235294, + "grad_norm": 0.07768157419912347, + "learning_rate": 1.9928476948723022e-05, + "loss": 0.0134, + "step": 10966 + }, + { + "epoch": 5.160941176470589, + "grad_norm": 0.12150828862256799, + "learning_rate": 1.99066495273087e-05, + "loss": 0.0395, + "step": 10967 + }, + { + "epoch": 5.161411764705882, + "grad_norm": 0.12875154867596783, + "learning_rate": 1.9884833216504657e-05, + "loss": 0.028, + "step": 10968 + }, + { + "epoch": 5.161882352941176, + "grad_norm": 0.0712460134147518, + "learning_rate": 1.9863028018174043e-05, + "loss": 0.012, + "step": 10969 + }, + { + "epoch": 5.16235294117647, + "grad_norm": 0.13650400069375582, + "learning_rate": 1.9841233934179145e-05, + "loss": 0.0422, + "step": 10970 + }, + { + "epoch": 5.162823529411765, + "grad_norm": 0.06371142582777001, + "learning_rate": 1.9819450966381328e-05, + "loss": 0.0071, + "step": 10971 + }, + { + "epoch": 5.163294117647059, + "grad_norm": 0.08028939531347928, + "learning_rate": 1.9797679116640903e-05, + "loss": 0.0152, + "step": 10972 + }, + { + "epoch": 5.163764705882353, + "grad_norm": 0.026463669576656593, + "learning_rate": 1.9775918386817346e-05, + "loss": 0.002, + "step": 10973 + }, + { + "epoch": 5.164235294117647, + "grad_norm": 0.13253989359026094, + "learning_rate": 1.9754168778769058e-05, + "loss": 0.0341, + "step": 10974 + }, + { + "epoch": 5.1647058823529415, + "grad_norm": 0.05675525949251725, + "learning_rate": 1.9732430294353636e-05, + "loss": 0.0118, + "step": 10975 + }, + { + "epoch": 5.165176470588236, + "grad_norm": 0.14141818955114324, + "learning_rate": 1.9710702935427642e-05, + "loss": 0.0383, + "step": 10976 + }, + { + "epoch": 5.165647058823529, + "grad_norm": 0.009552813489394298, + "learning_rate": 1.9688986703846658e-05, + "loss": 0.0003, + "step": 10977 + }, + { + "epoch": 5.166117647058823, + "grad_norm": 0.12416125687231067, + "learning_rate": 1.9667281601465372e-05, + "loss": 0.03, + "step": 10978 + }, + { + "epoch": 5.1665882352941175, + "grad_norm": 0.146088894029408, + "learning_rate": 1.964558763013755e-05, + "loss": 0.0346, + "step": 10979 + }, + { + "epoch": 5.167058823529412, + "grad_norm": 0.15017776711503592, + "learning_rate": 1.9623904791715907e-05, + "loss": 0.0412, + "step": 10980 + }, + { + "epoch": 5.167529411764706, + "grad_norm": 0.07077154445465339, + "learning_rate": 1.9602233088052277e-05, + "loss": 0.0072, + "step": 10981 + }, + { + "epoch": 5.168, + "grad_norm": 2.097547098040328e-06, + "learning_rate": 1.9580572520997566e-05, + "loss": 0.0, + "step": 10982 + }, + { + "epoch": 5.168470588235294, + "grad_norm": 0.1707171420005328, + "learning_rate": 1.9558923092401645e-05, + "loss": 0.0631, + "step": 10983 + }, + { + "epoch": 5.168941176470589, + "grad_norm": 0.08421049511561392, + "learning_rate": 1.95372848041135e-05, + "loss": 0.0113, + "step": 10984 + }, + { + "epoch": 5.169411764705882, + "grad_norm": 0.04676924626046463, + "learning_rate": 1.951565765798117e-05, + "loss": 0.0022, + "step": 10985 + }, + { + "epoch": 5.169882352941176, + "grad_norm": 0.11233714784013539, + "learning_rate": 1.9494041655851677e-05, + "loss": 0.0289, + "step": 10986 + }, + { + "epoch": 5.17035294117647, + "grad_norm": 0.05147880208152909, + "learning_rate": 1.9472436799571145e-05, + "loss": 0.0074, + "step": 10987 + }, + { + "epoch": 5.170823529411765, + "grad_norm": 0.14837217262775132, + "learning_rate": 1.9450843090984797e-05, + "loss": 0.0564, + "step": 10988 + }, + { + "epoch": 5.171294117647059, + "grad_norm": 0.06209040813707118, + "learning_rate": 1.9429260531936743e-05, + "loss": 0.007, + "step": 10989 + }, + { + "epoch": 5.171764705882353, + "grad_norm": 0.04876732077708952, + "learning_rate": 1.940768912427027e-05, + "loss": 0.0054, + "step": 10990 + }, + { + "epoch": 5.172235294117647, + "grad_norm": 0.0960388101704592, + "learning_rate": 1.9386128869827743e-05, + "loss": 0.02, + "step": 10991 + }, + { + "epoch": 5.1727058823529415, + "grad_norm": 0.13692487203356216, + "learning_rate": 1.936457977045044e-05, + "loss": 0.0304, + "step": 10992 + }, + { + "epoch": 5.173176470588236, + "grad_norm": 3.721901104523469e-05, + "learning_rate": 1.9343041827978785e-05, + "loss": 0.0, + "step": 10993 + }, + { + "epoch": 5.173647058823529, + "grad_norm": 0.1590763250861667, + "learning_rate": 1.9321515044252273e-05, + "loss": 0.0459, + "step": 10994 + }, + { + "epoch": 5.174117647058823, + "grad_norm": 0.15997223532803304, + "learning_rate": 1.9299999421109303e-05, + "loss": 0.0377, + "step": 10995 + }, + { + "epoch": 5.1745882352941175, + "grad_norm": 0.10847467708399447, + "learning_rate": 1.9278494960387452e-05, + "loss": 0.0214, + "step": 10996 + }, + { + "epoch": 5.175058823529412, + "grad_norm": 0.108713906780482, + "learning_rate": 1.9257001663923338e-05, + "loss": 0.0163, + "step": 10997 + }, + { + "epoch": 5.175529411764706, + "grad_norm": 0.06892279826350833, + "learning_rate": 1.9235519533552556e-05, + "loss": 0.0028, + "step": 10998 + }, + { + "epoch": 5.176, + "grad_norm": 0.09318299271506368, + "learning_rate": 1.921404857110984e-05, + "loss": 0.0217, + "step": 10999 + }, + { + "epoch": 5.176470588235294, + "grad_norm": 0.10202904846460291, + "learning_rate": 1.9192588778428842e-05, + "loss": 0.0208, + "step": 11000 + }, + { + "epoch": 5.176941176470589, + "grad_norm": 0.06445839937852169, + "learning_rate": 1.9171140157342407e-05, + "loss": 0.0115, + "step": 11001 + }, + { + "epoch": 5.177411764705882, + "grad_norm": 0.10398875702710543, + "learning_rate": 1.9149702709682292e-05, + "loss": 0.0221, + "step": 11002 + }, + { + "epoch": 5.177882352941176, + "grad_norm": 0.0794209532902397, + "learning_rate": 1.9128276437279362e-05, + "loss": 0.0101, + "step": 11003 + }, + { + "epoch": 5.17835294117647, + "grad_norm": 0.08682995561604986, + "learning_rate": 1.9106861341963587e-05, + "loss": 0.016, + "step": 11004 + }, + { + "epoch": 5.178823529411765, + "grad_norm": 0.00033507503180295393, + "learning_rate": 1.9085457425563855e-05, + "loss": 0.0, + "step": 11005 + }, + { + "epoch": 5.179294117647059, + "grad_norm": 0.11855094117006587, + "learning_rate": 1.9064064689908187e-05, + "loss": 0.0151, + "step": 11006 + }, + { + "epoch": 5.179764705882353, + "grad_norm": 0.08978514752343342, + "learning_rate": 1.9042683136823654e-05, + "loss": 0.0174, + "step": 11007 + }, + { + "epoch": 5.180235294117647, + "grad_norm": 5.264555216028413e-05, + "learning_rate": 1.9021312768136314e-05, + "loss": 0.0, + "step": 11008 + }, + { + "epoch": 5.1807058823529415, + "grad_norm": 0.12611668445255964, + "learning_rate": 1.899995358567129e-05, + "loss": 0.0235, + "step": 11009 + }, + { + "epoch": 5.181176470588236, + "grad_norm": 0.10917580402418545, + "learning_rate": 1.897860559125282e-05, + "loss": 0.0328, + "step": 11010 + }, + { + "epoch": 5.181647058823529, + "grad_norm": 0.08406565082592644, + "learning_rate": 1.895726878670405e-05, + "loss": 0.0186, + "step": 11011 + }, + { + "epoch": 5.182117647058823, + "grad_norm": 4.47074897287456e-05, + "learning_rate": 1.8935943173847302e-05, + "loss": 0.0, + "step": 11012 + }, + { + "epoch": 5.1825882352941175, + "grad_norm": 0.09158278286216143, + "learning_rate": 1.8914628754503885e-05, + "loss": 0.0103, + "step": 11013 + }, + { + "epoch": 5.183058823529412, + "grad_norm": 0.07866145039293851, + "learning_rate": 1.889332553049413e-05, + "loss": 0.0151, + "step": 11014 + }, + { + "epoch": 5.183529411764706, + "grad_norm": 0.24478204301545003, + "learning_rate": 1.887203350363743e-05, + "loss": 0.0514, + "step": 11015 + }, + { + "epoch": 5.184, + "grad_norm": 0.07232949356392214, + "learning_rate": 1.8850752675752295e-05, + "loss": 0.0162, + "step": 11016 + }, + { + "epoch": 5.184470588235294, + "grad_norm": 1.167664633899334e-05, + "learning_rate": 1.8829483048656124e-05, + "loss": 0.0, + "step": 11017 + }, + { + "epoch": 5.184941176470589, + "grad_norm": 0.11004636617192572, + "learning_rate": 1.8808224624165497e-05, + "loss": 0.0207, + "step": 11018 + }, + { + "epoch": 5.185411764705882, + "grad_norm": 0.08658294558026748, + "learning_rate": 1.878697740409601e-05, + "loss": 0.0141, + "step": 11019 + }, + { + "epoch": 5.185882352941176, + "grad_norm": 9.88092976562158e-06, + "learning_rate": 1.8765741390262212e-05, + "loss": 0.0, + "step": 11020 + }, + { + "epoch": 5.18635294117647, + "grad_norm": 0.04342720858241799, + "learning_rate": 1.8744516584477803e-05, + "loss": 0.002, + "step": 11021 + }, + { + "epoch": 5.186823529411765, + "grad_norm": 0.10818042175339103, + "learning_rate": 1.87233029885555e-05, + "loss": 0.0225, + "step": 11022 + }, + { + "epoch": 5.187294117647059, + "grad_norm": 0.04686047759705054, + "learning_rate": 1.8702100604307018e-05, + "loss": 0.0022, + "step": 11023 + }, + { + "epoch": 5.187764705882353, + "grad_norm": 0.14524782228278252, + "learning_rate": 1.868090943354318e-05, + "loss": 0.0275, + "step": 11024 + }, + { + "epoch": 5.188235294117647, + "grad_norm": 0.16514135204267105, + "learning_rate": 1.8659729478073764e-05, + "loss": 0.0499, + "step": 11025 + }, + { + "epoch": 5.1887058823529415, + "grad_norm": 0.15243943701031343, + "learning_rate": 1.8638560739707664e-05, + "loss": 0.046, + "step": 11026 + }, + { + "epoch": 5.189176470588236, + "grad_norm": 0.03845855886117324, + "learning_rate": 1.8617403220252846e-05, + "loss": 0.0033, + "step": 11027 + }, + { + "epoch": 5.189647058823529, + "grad_norm": 0.1100577660725482, + "learning_rate": 1.8596256921516166e-05, + "loss": 0.0323, + "step": 11028 + }, + { + "epoch": 5.190117647058823, + "grad_norm": 0.049730433430144466, + "learning_rate": 1.8575121845303698e-05, + "loss": 0.0046, + "step": 11029 + }, + { + "epoch": 5.1905882352941175, + "grad_norm": 0.06353048821839659, + "learning_rate": 1.8553997993420495e-05, + "loss": 0.0034, + "step": 11030 + }, + { + "epoch": 5.191058823529412, + "grad_norm": 0.040232600449157774, + "learning_rate": 1.8532885367670547e-05, + "loss": 0.0033, + "step": 11031 + }, + { + "epoch": 5.191529411764706, + "grad_norm": 0.10121495918191512, + "learning_rate": 1.8511783969857046e-05, + "loss": 0.0191, + "step": 11032 + }, + { + "epoch": 5.192, + "grad_norm": 2.4577163852417777e-06, + "learning_rate": 1.849069380178217e-05, + "loss": 0.0, + "step": 11033 + }, + { + "epoch": 5.192470588235294, + "grad_norm": 0.047710611301082444, + "learning_rate": 1.8469614865247075e-05, + "loss": 0.0027, + "step": 11034 + }, + { + "epoch": 5.192941176470589, + "grad_norm": 0.00037555232592187, + "learning_rate": 1.8448547162052002e-05, + "loss": 0.0, + "step": 11035 + }, + { + "epoch": 5.193411764705882, + "grad_norm": 0.11840306663990845, + "learning_rate": 1.8427490693996298e-05, + "loss": 0.0163, + "step": 11036 + }, + { + "epoch": 5.193882352941176, + "grad_norm": 0.05889772305029274, + "learning_rate": 1.840644546287821e-05, + "loss": 0.0068, + "step": 11037 + }, + { + "epoch": 5.19435294117647, + "grad_norm": 0.11535834411989852, + "learning_rate": 1.838541147049515e-05, + "loss": 0.0171, + "step": 11038 + }, + { + "epoch": 5.194823529411765, + "grad_norm": 0.08151424184931845, + "learning_rate": 1.8364388718643542e-05, + "loss": 0.0127, + "step": 11039 + }, + { + "epoch": 5.195294117647059, + "grad_norm": 0.0940028972288212, + "learning_rate": 1.8343377209118755e-05, + "loss": 0.0206, + "step": 11040 + }, + { + "epoch": 5.195764705882353, + "grad_norm": 0.02049421838187868, + "learning_rate": 1.8322376943715382e-05, + "loss": 0.0009, + "step": 11041 + }, + { + "epoch": 5.196235294117647, + "grad_norm": 0.00017378264653054182, + "learning_rate": 1.8301387924226863e-05, + "loss": 0.0, + "step": 11042 + }, + { + "epoch": 5.1967058823529415, + "grad_norm": 0.1741299334829222, + "learning_rate": 1.828041015244579e-05, + "loss": 0.05, + "step": 11043 + }, + { + "epoch": 5.197176470588236, + "grad_norm": 0.14630332802081347, + "learning_rate": 1.82594436301638e-05, + "loss": 0.006, + "step": 11044 + }, + { + "epoch": 5.197647058823529, + "grad_norm": 0.20985654049862856, + "learning_rate": 1.8238488359171476e-05, + "loss": 0.0506, + "step": 11045 + }, + { + "epoch": 5.198117647058823, + "grad_norm": 0.11760759521144543, + "learning_rate": 1.821754434125854e-05, + "loss": 0.0152, + "step": 11046 + }, + { + "epoch": 5.1985882352941175, + "grad_norm": 0.05153102100825848, + "learning_rate": 1.8196611578213738e-05, + "loss": 0.0049, + "step": 11047 + }, + { + "epoch": 5.199058823529412, + "grad_norm": 0.127374587620534, + "learning_rate": 1.817569007182476e-05, + "loss": 0.0156, + "step": 11048 + }, + { + "epoch": 5.199529411764706, + "grad_norm": 0.11937429035223962, + "learning_rate": 1.815477982387846e-05, + "loss": 0.0224, + "step": 11049 + }, + { + "epoch": 5.2, + "grad_norm": 0.1265757278015304, + "learning_rate": 1.813388083616068e-05, + "loss": 0.0348, + "step": 11050 + }, + { + "epoch": 5.200470588235294, + "grad_norm": 1.7179591973029426e-05, + "learning_rate": 1.8112993110456235e-05, + "loss": 0.0, + "step": 11051 + }, + { + "epoch": 5.200941176470589, + "grad_norm": 3.6672390484543015e-06, + "learning_rate": 1.8092116648549136e-05, + "loss": 0.0, + "step": 11052 + }, + { + "epoch": 5.201411764705882, + "grad_norm": 0.016931146194991052, + "learning_rate": 1.8071251452222224e-05, + "loss": 0.0009, + "step": 11053 + }, + { + "epoch": 5.201882352941176, + "grad_norm": 0.016867275547657986, + "learning_rate": 1.8050397523257555e-05, + "loss": 0.0004, + "step": 11054 + }, + { + "epoch": 5.20235294117647, + "grad_norm": 4.725071744200968e-05, + "learning_rate": 1.8029554863436186e-05, + "loss": 0.0, + "step": 11055 + }, + { + "epoch": 5.202823529411765, + "grad_norm": 0.08955141692704902, + "learning_rate": 1.8008723474538095e-05, + "loss": 0.0204, + "step": 11056 + }, + { + "epoch": 5.203294117647059, + "grad_norm": 0.08026385163788048, + "learning_rate": 1.798790335834244e-05, + "loss": 0.0126, + "step": 11057 + }, + { + "epoch": 5.203764705882353, + "grad_norm": 0.041690152879134725, + "learning_rate": 1.796709451662739e-05, + "loss": 0.0052, + "step": 11058 + }, + { + "epoch": 5.204235294117647, + "grad_norm": 1.984265320366553e-05, + "learning_rate": 1.7946296951170065e-05, + "loss": 0.0, + "step": 11059 + }, + { + "epoch": 5.2047058823529415, + "grad_norm": 0.0886166765382393, + "learning_rate": 1.792551066374668e-05, + "loss": 0.0094, + "step": 11060 + }, + { + "epoch": 5.205176470588235, + "grad_norm": 0.1974570805202034, + "learning_rate": 1.7904735656132545e-05, + "loss": 0.0365, + "step": 11061 + }, + { + "epoch": 5.205647058823529, + "grad_norm": 0.10455701788415142, + "learning_rate": 1.7883971930101883e-05, + "loss": 0.0095, + "step": 11062 + }, + { + "epoch": 5.206117647058823, + "grad_norm": 0.07684175240151055, + "learning_rate": 1.786321948742803e-05, + "loss": 0.0121, + "step": 11063 + }, + { + "epoch": 5.2065882352941175, + "grad_norm": 0.07932644664962167, + "learning_rate": 1.78424783298834e-05, + "loss": 0.0099, + "step": 11064 + }, + { + "epoch": 5.207058823529412, + "grad_norm": 0.07558679334824112, + "learning_rate": 1.782174845923931e-05, + "loss": 0.0061, + "step": 11065 + }, + { + "epoch": 5.207529411764706, + "grad_norm": 0.13918779255108576, + "learning_rate": 1.7801029877266237e-05, + "loss": 0.029, + "step": 11066 + }, + { + "epoch": 5.208, + "grad_norm": 2.4886967766307584e-05, + "learning_rate": 1.778032258573368e-05, + "loss": 0.0, + "step": 11067 + }, + { + "epoch": 5.208470588235294, + "grad_norm": 0.05282725945042853, + "learning_rate": 1.7759626586410074e-05, + "loss": 0.004, + "step": 11068 + }, + { + "epoch": 5.208941176470589, + "grad_norm": 0.16763571856565182, + "learning_rate": 1.7738941881063007e-05, + "loss": 0.0385, + "step": 11069 + }, + { + "epoch": 5.209411764705882, + "grad_norm": 0.11260103348758271, + "learning_rate": 1.7718268471459045e-05, + "loss": 0.0222, + "step": 11070 + }, + { + "epoch": 5.209882352941176, + "grad_norm": 0.11050607673517149, + "learning_rate": 1.769760635936378e-05, + "loss": 0.017, + "step": 11071 + }, + { + "epoch": 5.21035294117647, + "grad_norm": 0.14289394868538843, + "learning_rate": 1.7676955546541867e-05, + "loss": 0.0354, + "step": 11072 + }, + { + "epoch": 5.210823529411765, + "grad_norm": 0.00010570455117170718, + "learning_rate": 1.7656316034757024e-05, + "loss": 0.0, + "step": 11073 + }, + { + "epoch": 5.211294117647059, + "grad_norm": 0.06994977634047918, + "learning_rate": 1.7635687825771894e-05, + "loss": 0.007, + "step": 11074 + }, + { + "epoch": 5.211764705882353, + "grad_norm": 0.11510624669222397, + "learning_rate": 1.7615070921348316e-05, + "loss": 0.0174, + "step": 11075 + }, + { + "epoch": 5.212235294117647, + "grad_norm": 0.16056761712382325, + "learning_rate": 1.7594465323246987e-05, + "loss": 0.0203, + "step": 11076 + }, + { + "epoch": 5.2127058823529415, + "grad_norm": 9.777768844870936e-05, + "learning_rate": 1.757387103322776e-05, + "loss": 0.0, + "step": 11077 + }, + { + "epoch": 5.213176470588236, + "grad_norm": 0.0001290567536000516, + "learning_rate": 1.7553288053049515e-05, + "loss": 0.0, + "step": 11078 + }, + { + "epoch": 5.213647058823529, + "grad_norm": 0.11699086550975399, + "learning_rate": 1.7532716384470108e-05, + "loss": 0.018, + "step": 11079 + }, + { + "epoch": 5.214117647058823, + "grad_norm": 0.1256100212470251, + "learning_rate": 1.7512156029246453e-05, + "loss": 0.0255, + "step": 11080 + }, + { + "epoch": 5.2145882352941175, + "grad_norm": 0.1486562824940997, + "learning_rate": 1.7491606989134566e-05, + "loss": 0.0365, + "step": 11081 + }, + { + "epoch": 5.215058823529412, + "grad_norm": 2.5990507081943123e-05, + "learning_rate": 1.7471069265889353e-05, + "loss": 0.0, + "step": 11082 + }, + { + "epoch": 5.215529411764706, + "grad_norm": 0.00014559301332206872, + "learning_rate": 1.7450542861264883e-05, + "loss": 0.0, + "step": 11083 + }, + { + "epoch": 5.216, + "grad_norm": 0.1306873317143905, + "learning_rate": 1.7430027777014217e-05, + "loss": 0.0287, + "step": 11084 + }, + { + "epoch": 5.216470588235294, + "grad_norm": 0.060064400856149475, + "learning_rate": 1.7409524014889388e-05, + "loss": 0.0061, + "step": 11085 + }, + { + "epoch": 5.216941176470589, + "grad_norm": 1.0436243551556179e-05, + "learning_rate": 1.738903157664161e-05, + "loss": 0.0, + "step": 11086 + }, + { + "epoch": 5.217411764705882, + "grad_norm": 0.15712107776075745, + "learning_rate": 1.736855046402097e-05, + "loss": 0.0329, + "step": 11087 + }, + { + "epoch": 5.217882352941176, + "grad_norm": 0.10985623061623946, + "learning_rate": 1.7348080678776665e-05, + "loss": 0.0195, + "step": 11088 + }, + { + "epoch": 5.21835294117647, + "grad_norm": 0.20135024299481016, + "learning_rate": 1.732762222265695e-05, + "loss": 0.0267, + "step": 11089 + }, + { + "epoch": 5.218823529411765, + "grad_norm": 0.10340542027426614, + "learning_rate": 1.7307175097409044e-05, + "loss": 0.0222, + "step": 11090 + }, + { + "epoch": 5.219294117647059, + "grad_norm": 7.325148175046648e-06, + "learning_rate": 1.7286739304779214e-05, + "loss": 0.0, + "step": 11091 + }, + { + "epoch": 5.219764705882353, + "grad_norm": 0.09040408498713384, + "learning_rate": 1.7266314846512848e-05, + "loss": 0.0081, + "step": 11092 + }, + { + "epoch": 5.220235294117647, + "grad_norm": 0.11345831100585906, + "learning_rate": 1.7245901724354218e-05, + "loss": 0.0195, + "step": 11093 + }, + { + "epoch": 5.2207058823529415, + "grad_norm": 2.6722730187636947e-05, + "learning_rate": 1.7225499940046726e-05, + "loss": 0.0, + "step": 11094 + }, + { + "epoch": 5.221176470588235, + "grad_norm": 0.09238332672321298, + "learning_rate": 1.720510949533282e-05, + "loss": 0.0138, + "step": 11095 + }, + { + "epoch": 5.221647058823529, + "grad_norm": 0.0432623594328975, + "learning_rate": 1.71847303919539e-05, + "loss": 0.0036, + "step": 11096 + }, + { + "epoch": 5.222117647058823, + "grad_norm": 0.18884105013572158, + "learning_rate": 1.7164362631650457e-05, + "loss": 0.0356, + "step": 11097 + }, + { + "epoch": 5.2225882352941175, + "grad_norm": 0.06516321616922581, + "learning_rate": 1.7144006216162036e-05, + "loss": 0.0071, + "step": 11098 + }, + { + "epoch": 5.223058823529412, + "grad_norm": 0.15485299217468612, + "learning_rate": 1.7123661147227096e-05, + "loss": 0.0432, + "step": 11099 + }, + { + "epoch": 5.223529411764706, + "grad_norm": 0.08264740166721334, + "learning_rate": 1.7103327426583265e-05, + "loss": 0.0077, + "step": 11100 + }, + { + "epoch": 5.224, + "grad_norm": 1.0536320706974932e-05, + "learning_rate": 1.7083005055967146e-05, + "loss": 0.0, + "step": 11101 + }, + { + "epoch": 5.224470588235294, + "grad_norm": 0.1342795091286109, + "learning_rate": 1.7062694037114326e-05, + "loss": 0.0318, + "step": 11102 + }, + { + "epoch": 5.224941176470589, + "grad_norm": 0.039091107606567706, + "learning_rate": 1.7042394371759516e-05, + "loss": 0.0023, + "step": 11103 + }, + { + "epoch": 5.225411764705882, + "grad_norm": 1.927752964174622e-05, + "learning_rate": 1.7022106061636342e-05, + "loss": 0.0, + "step": 11104 + }, + { + "epoch": 5.225882352941176, + "grad_norm": 0.10366657803864314, + "learning_rate": 1.700182910847757e-05, + "loss": 0.0241, + "step": 11105 + }, + { + "epoch": 5.22635294117647, + "grad_norm": 0.15884209210632036, + "learning_rate": 1.6981563514014967e-05, + "loss": 0.0545, + "step": 11106 + }, + { + "epoch": 5.226823529411765, + "grad_norm": 0.1747723685365847, + "learning_rate": 1.6961309279979278e-05, + "loss": 0.0366, + "step": 11107 + }, + { + "epoch": 5.227294117647059, + "grad_norm": 0.07251942492889518, + "learning_rate": 1.6941066408100308e-05, + "loss": 0.0123, + "step": 11108 + }, + { + "epoch": 5.227764705882353, + "grad_norm": 0.11334024957942303, + "learning_rate": 1.6920834900106967e-05, + "loss": 0.0247, + "step": 11109 + }, + { + "epoch": 5.228235294117647, + "grad_norm": 0.021694583702454157, + "learning_rate": 1.690061475772703e-05, + "loss": 0.0011, + "step": 11110 + }, + { + "epoch": 5.2287058823529415, + "grad_norm": 0.08925845012496371, + "learning_rate": 1.6880405982687455e-05, + "loss": 0.016, + "step": 11111 + }, + { + "epoch": 5.229176470588235, + "grad_norm": 0.10291925615446837, + "learning_rate": 1.6860208576714173e-05, + "loss": 0.0231, + "step": 11112 + }, + { + "epoch": 5.229647058823529, + "grad_norm": 0.1145647823693458, + "learning_rate": 1.68400225415321e-05, + "loss": 0.0262, + "step": 11113 + }, + { + "epoch": 5.230117647058823, + "grad_norm": 0.15591889869832393, + "learning_rate": 1.6819847878865255e-05, + "loss": 0.0216, + "step": 11114 + }, + { + "epoch": 5.2305882352941175, + "grad_norm": 0.040515793317661336, + "learning_rate": 1.6799684590436663e-05, + "loss": 0.0039, + "step": 11115 + }, + { + "epoch": 5.231058823529412, + "grad_norm": 0.0031206331909720323, + "learning_rate": 1.6779532677968327e-05, + "loss": 0.0, + "step": 11116 + }, + { + "epoch": 5.231529411764706, + "grad_norm": 0.04458133202675895, + "learning_rate": 1.6759392143181354e-05, + "loss": 0.005, + "step": 11117 + }, + { + "epoch": 5.232, + "grad_norm": 0.029178712672582634, + "learning_rate": 1.6739262987795854e-05, + "loss": 0.0021, + "step": 11118 + }, + { + "epoch": 5.232470588235294, + "grad_norm": 0.07508947266381577, + "learning_rate": 1.6719145213530898e-05, + "loss": 0.0088, + "step": 11119 + }, + { + "epoch": 5.232941176470589, + "grad_norm": 0.12755824312776837, + "learning_rate": 1.669903882210468e-05, + "loss": 0.0392, + "step": 11120 + }, + { + "epoch": 5.233411764705882, + "grad_norm": 8.636341930579641e-05, + "learning_rate": 1.667894381523443e-05, + "loss": 0.0, + "step": 11121 + }, + { + "epoch": 5.233882352941176, + "grad_norm": 0.04487032552308356, + "learning_rate": 1.665886019463627e-05, + "loss": 0.002, + "step": 11122 + }, + { + "epoch": 5.23435294117647, + "grad_norm": 0.18423916332284507, + "learning_rate": 1.6638787962025495e-05, + "loss": 0.0577, + "step": 11123 + }, + { + "epoch": 5.234823529411765, + "grad_norm": 0.10988575201766511, + "learning_rate": 1.6618727119116388e-05, + "loss": 0.0317, + "step": 11124 + }, + { + "epoch": 5.235294117647059, + "grad_norm": 0.156109835749567, + "learning_rate": 1.6598677667622175e-05, + "loss": 0.0334, + "step": 11125 + }, + { + "epoch": 5.235764705882353, + "grad_norm": 0.10135766394375552, + "learning_rate": 1.657863960925525e-05, + "loss": 0.0211, + "step": 11126 + }, + { + "epoch": 5.236235294117647, + "grad_norm": 0.09478968885012844, + "learning_rate": 1.655861294572692e-05, + "loss": 0.0256, + "step": 11127 + }, + { + "epoch": 5.2367058823529415, + "grad_norm": 0.09884542237979245, + "learning_rate": 1.6538597678747555e-05, + "loss": 0.0155, + "step": 11128 + }, + { + "epoch": 5.237176470588235, + "grad_norm": 0.02943181167619731, + "learning_rate": 1.651859381002658e-05, + "loss": 0.0027, + "step": 11129 + }, + { + "epoch": 5.237647058823529, + "grad_norm": 0.0638968893533856, + "learning_rate": 1.649860134127241e-05, + "loss": 0.0066, + "step": 11130 + }, + { + "epoch": 5.238117647058823, + "grad_norm": 0.1318548232878897, + "learning_rate": 1.647862027419251e-05, + "loss": 0.0345, + "step": 11131 + }, + { + "epoch": 5.2385882352941175, + "grad_norm": 0.09181749679199828, + "learning_rate": 1.6458650610493385e-05, + "loss": 0.0175, + "step": 11132 + }, + { + "epoch": 5.239058823529412, + "grad_norm": 0.1719563413194591, + "learning_rate": 1.6438692351880482e-05, + "loss": 0.0406, + "step": 11133 + }, + { + "epoch": 5.239529411764706, + "grad_norm": 9.248181102263003e-06, + "learning_rate": 1.6418745500058395e-05, + "loss": 0.0, + "step": 11134 + }, + { + "epoch": 5.24, + "grad_norm": 0.025733890686920717, + "learning_rate": 1.639881005673064e-05, + "loss": 0.0016, + "step": 11135 + }, + { + "epoch": 5.240470588235294, + "grad_norm": 0.057242421406439245, + "learning_rate": 1.6378886023599803e-05, + "loss": 0.0029, + "step": 11136 + }, + { + "epoch": 5.240941176470589, + "grad_norm": 0.0006154750753876947, + "learning_rate": 1.6358973402367524e-05, + "loss": 0.0, + "step": 11137 + }, + { + "epoch": 5.241411764705882, + "grad_norm": 0.09088206060776488, + "learning_rate": 1.6339072194734414e-05, + "loss": 0.0138, + "step": 11138 + }, + { + "epoch": 5.241882352941176, + "grad_norm": 0.06825507888387157, + "learning_rate": 1.6319182402400144e-05, + "loss": 0.0038, + "step": 11139 + }, + { + "epoch": 5.24235294117647, + "grad_norm": 0.1406578968357627, + "learning_rate": 1.629930402706342e-05, + "loss": 0.0536, + "step": 11140 + }, + { + "epoch": 5.242823529411765, + "grad_norm": 0.1540866315967823, + "learning_rate": 1.6279437070421885e-05, + "loss": 0.0427, + "step": 11141 + }, + { + "epoch": 5.243294117647059, + "grad_norm": 0.06720119842374761, + "learning_rate": 1.625958153417235e-05, + "loss": 0.0035, + "step": 11142 + }, + { + "epoch": 5.243764705882353, + "grad_norm": 7.628812347181156e-05, + "learning_rate": 1.6239737420010558e-05, + "loss": 0.0, + "step": 11143 + }, + { + "epoch": 5.244235294117647, + "grad_norm": 0.07093687035727976, + "learning_rate": 1.621990472963125e-05, + "loss": 0.0055, + "step": 11144 + }, + { + "epoch": 5.2447058823529416, + "grad_norm": 0.14491257154128256, + "learning_rate": 1.620008346472826e-05, + "loss": 0.0292, + "step": 11145 + }, + { + "epoch": 5.245176470588235, + "grad_norm": 0.11059628297270534, + "learning_rate": 1.618027362699446e-05, + "loss": 0.0318, + "step": 11146 + }, + { + "epoch": 5.245647058823529, + "grad_norm": 0.11174946569156394, + "learning_rate": 1.6160475218121654e-05, + "loss": 0.0209, + "step": 11147 + }, + { + "epoch": 5.246117647058823, + "grad_norm": 0.14533956877245877, + "learning_rate": 1.6140688239800736e-05, + "loss": 0.0425, + "step": 11148 + }, + { + "epoch": 5.2465882352941176, + "grad_norm": 0.18707708966751424, + "learning_rate": 1.6120912693721654e-05, + "loss": 0.0346, + "step": 11149 + }, + { + "epoch": 5.247058823529412, + "grad_norm": 0.1336523669804377, + "learning_rate": 1.6101148581573274e-05, + "loss": 0.0234, + "step": 11150 + }, + { + "epoch": 5.247529411764706, + "grad_norm": 0.12301694513312253, + "learning_rate": 1.6081395905043565e-05, + "loss": 0.028, + "step": 11151 + }, + { + "epoch": 5.248, + "grad_norm": 0.022280695041556706, + "learning_rate": 1.606165466581954e-05, + "loss": 0.0018, + "step": 11152 + }, + { + "epoch": 5.248470588235294, + "grad_norm": 0.07562321068472429, + "learning_rate": 1.6041924865587152e-05, + "loss": 0.0098, + "step": 11153 + }, + { + "epoch": 5.248941176470589, + "grad_norm": 2.2759048079508397e-06, + "learning_rate": 1.602220650603147e-05, + "loss": 0.0, + "step": 11154 + }, + { + "epoch": 5.249411764705882, + "grad_norm": 0.15233948720337173, + "learning_rate": 1.600249958883648e-05, + "loss": 0.0288, + "step": 11155 + }, + { + "epoch": 5.249882352941176, + "grad_norm": 0.1252983031308779, + "learning_rate": 1.5982804115685273e-05, + "loss": 0.0309, + "step": 11156 + }, + { + "epoch": 5.2503529411764704, + "grad_norm": 0.13962741910861548, + "learning_rate": 1.5963120088259977e-05, + "loss": 0.0283, + "step": 11157 + }, + { + "epoch": 5.250823529411765, + "grad_norm": 0.07798408623840773, + "learning_rate": 1.594344750824164e-05, + "loss": 0.0102, + "step": 11158 + }, + { + "epoch": 5.251294117647059, + "grad_norm": 1.8410081471091373e-05, + "learning_rate": 1.5923786377310433e-05, + "loss": 0.0, + "step": 11159 + }, + { + "epoch": 5.251764705882353, + "grad_norm": 0.11620068392277516, + "learning_rate": 1.5904136697145542e-05, + "loss": 0.0221, + "step": 11160 + }, + { + "epoch": 5.252235294117647, + "grad_norm": 0.1577643553800168, + "learning_rate": 1.5884498469425078e-05, + "loss": 0.0371, + "step": 11161 + }, + { + "epoch": 5.252705882352942, + "grad_norm": 0.11497334920255732, + "learning_rate": 1.5864871695826282e-05, + "loss": 0.0209, + "step": 11162 + }, + { + "epoch": 5.253176470588235, + "grad_norm": 4.178737949656466e-06, + "learning_rate": 1.5845256378025395e-05, + "loss": 0.0, + "step": 11163 + }, + { + "epoch": 5.253647058823529, + "grad_norm": 0.10022787520563657, + "learning_rate": 1.582565251769761e-05, + "loss": 0.0205, + "step": 11164 + }, + { + "epoch": 5.254117647058823, + "grad_norm": 0.024329797758373704, + "learning_rate": 1.5806060116517227e-05, + "loss": 0.0015, + "step": 11165 + }, + { + "epoch": 5.254588235294118, + "grad_norm": 7.725502541430126e-05, + "learning_rate": 1.578647917615757e-05, + "loss": 0.0, + "step": 11166 + }, + { + "epoch": 5.255058823529412, + "grad_norm": 0.03258003938178377, + "learning_rate": 1.5766909698290863e-05, + "loss": 0.0019, + "step": 11167 + }, + { + "epoch": 5.255529411764706, + "grad_norm": 0.1391822495072797, + "learning_rate": 1.574735168458846e-05, + "loss": 0.0366, + "step": 11168 + }, + { + "epoch": 5.256, + "grad_norm": 0.03930247348678843, + "learning_rate": 1.572780513672079e-05, + "loss": 0.0023, + "step": 11169 + }, + { + "epoch": 5.2564705882352944, + "grad_norm": 0.20030147084668393, + "learning_rate": 1.5708270056357115e-05, + "loss": 0.0186, + "step": 11170 + }, + { + "epoch": 5.256941176470589, + "grad_norm": 0.11699020970290173, + "learning_rate": 1.5688746445165867e-05, + "loss": 0.0244, + "step": 11171 + }, + { + "epoch": 5.257411764705882, + "grad_norm": 0.1485231187125174, + "learning_rate": 1.5669234304814515e-05, + "loss": 0.0158, + "step": 11172 + }, + { + "epoch": 5.257882352941176, + "grad_norm": 4.474932716594178e-06, + "learning_rate": 1.564973363696939e-05, + "loss": 0.0, + "step": 11173 + }, + { + "epoch": 5.2583529411764705, + "grad_norm": 0.13199967697724374, + "learning_rate": 1.5630244443296013e-05, + "loss": 0.0481, + "step": 11174 + }, + { + "epoch": 5.258823529411765, + "grad_norm": 0.15542295548152876, + "learning_rate": 1.5610766725458834e-05, + "loss": 0.0381, + "step": 11175 + }, + { + "epoch": 5.259294117647059, + "grad_norm": 0.048701912258471265, + "learning_rate": 1.5591300485121354e-05, + "loss": 0.0039, + "step": 11176 + }, + { + "epoch": 5.259764705882353, + "grad_norm": 0.041418186742556075, + "learning_rate": 1.5571845723946113e-05, + "loss": 0.0042, + "step": 11177 + }, + { + "epoch": 5.260235294117647, + "grad_norm": 0.025271483220445742, + "learning_rate": 1.555240244359458e-05, + "loss": 0.001, + "step": 11178 + }, + { + "epoch": 5.260705882352942, + "grad_norm": 0.08699406762671541, + "learning_rate": 1.5532970645727344e-05, + "loss": 0.0059, + "step": 11179 + }, + { + "epoch": 5.261176470588235, + "grad_norm": 1.688600498637325e-05, + "learning_rate": 1.551355033200401e-05, + "loss": 0.0, + "step": 11180 + }, + { + "epoch": 5.261647058823529, + "grad_norm": 0.11832528613291156, + "learning_rate": 1.5494141504083087e-05, + "loss": 0.0265, + "step": 11181 + }, + { + "epoch": 5.262117647058823, + "grad_norm": 6.234921961028282e-05, + "learning_rate": 1.547474416362226e-05, + "loss": 0.0, + "step": 11182 + }, + { + "epoch": 5.262588235294118, + "grad_norm": 3.257262665672058e-05, + "learning_rate": 1.5455358312278093e-05, + "loss": 0.0, + "step": 11183 + }, + { + "epoch": 5.263058823529412, + "grad_norm": 0.050999843907945035, + "learning_rate": 1.5435983951706276e-05, + "loss": 0.0019, + "step": 11184 + }, + { + "epoch": 5.263529411764706, + "grad_norm": 0.009085107704286093, + "learning_rate": 1.5416621083561486e-05, + "loss": 0.0003, + "step": 11185 + }, + { + "epoch": 5.264, + "grad_norm": 1.1187542719810471e-05, + "learning_rate": 1.5397269709497352e-05, + "loss": 0.0, + "step": 11186 + }, + { + "epoch": 5.2644705882352945, + "grad_norm": 0.03766036720221196, + "learning_rate": 1.5377929831166597e-05, + "loss": 0.0025, + "step": 11187 + }, + { + "epoch": 5.264941176470588, + "grad_norm": 0.17121966753053935, + "learning_rate": 1.5358601450220988e-05, + "loss": 0.0394, + "step": 11188 + }, + { + "epoch": 5.265411764705882, + "grad_norm": 7.382240822654516e-05, + "learning_rate": 1.53392845683112e-05, + "loss": 0.0, + "step": 11189 + }, + { + "epoch": 5.265882352941176, + "grad_norm": 0.13882046526767905, + "learning_rate": 1.5319979187087018e-05, + "loss": 0.029, + "step": 11190 + }, + { + "epoch": 5.2663529411764705, + "grad_norm": 3.240504984853054e-06, + "learning_rate": 1.5300685308197254e-05, + "loss": 0.0, + "step": 11191 + }, + { + "epoch": 5.266823529411765, + "grad_norm": 0.14584012790232295, + "learning_rate": 1.528140293328962e-05, + "loss": 0.0382, + "step": 11192 + }, + { + "epoch": 5.267294117647059, + "grad_norm": 0.08485100181163654, + "learning_rate": 1.5262132064010956e-05, + "loss": 0.0099, + "step": 11193 + }, + { + "epoch": 5.267764705882353, + "grad_norm": 0.2282785587578391, + "learning_rate": 1.5242872702007142e-05, + "loss": 0.0515, + "step": 11194 + }, + { + "epoch": 5.268235294117647, + "grad_norm": 0.11877907851408713, + "learning_rate": 1.522362484892296e-05, + "loss": 0.0262, + "step": 11195 + }, + { + "epoch": 5.268705882352942, + "grad_norm": 0.12937505796517504, + "learning_rate": 1.5204388506402281e-05, + "loss": 0.0313, + "step": 11196 + }, + { + "epoch": 5.269176470588235, + "grad_norm": 0.025540429535010987, + "learning_rate": 1.5185163676088007e-05, + "loss": 0.001, + "step": 11197 + }, + { + "epoch": 5.269647058823529, + "grad_norm": 0.12032370618779281, + "learning_rate": 1.5165950359622009e-05, + "loss": 0.0302, + "step": 11198 + }, + { + "epoch": 5.270117647058823, + "grad_norm": 0.12301017484013604, + "learning_rate": 1.5146748558645205e-05, + "loss": 0.0322, + "step": 11199 + }, + { + "epoch": 5.270588235294118, + "grad_norm": 0.11667615753591697, + "learning_rate": 1.5127558274797535e-05, + "loss": 0.0158, + "step": 11200 + }, + { + "epoch": 5.271058823529412, + "grad_norm": 0.12801788660951402, + "learning_rate": 1.5108379509717922e-05, + "loss": 0.0256, + "step": 11201 + }, + { + "epoch": 5.271529411764706, + "grad_norm": 0.05700070947782296, + "learning_rate": 1.5089212265044336e-05, + "loss": 0.0027, + "step": 11202 + }, + { + "epoch": 5.272, + "grad_norm": 0.06920004130688562, + "learning_rate": 1.5070056542413788e-05, + "loss": 0.0121, + "step": 11203 + }, + { + "epoch": 5.2724705882352945, + "grad_norm": 0.12246991625858206, + "learning_rate": 1.5050912343462202e-05, + "loss": 0.022, + "step": 11204 + }, + { + "epoch": 5.272941176470589, + "grad_norm": 0.0561489182317621, + "learning_rate": 1.5031779669824668e-05, + "loss": 0.0039, + "step": 11205 + }, + { + "epoch": 5.273411764705882, + "grad_norm": 0.1692887993176183, + "learning_rate": 1.5012658523135113e-05, + "loss": 0.0576, + "step": 11206 + }, + { + "epoch": 5.273882352941176, + "grad_norm": 1.612852682102685e-05, + "learning_rate": 1.4993548905026648e-05, + "loss": 0.0, + "step": 11207 + }, + { + "epoch": 5.2743529411764705, + "grad_norm": 0.09124472105257517, + "learning_rate": 1.4974450817131334e-05, + "loss": 0.0084, + "step": 11208 + }, + { + "epoch": 5.274823529411765, + "grad_norm": 0.0682384760092308, + "learning_rate": 1.4955364261080183e-05, + "loss": 0.0083, + "step": 11209 + }, + { + "epoch": 5.275294117647059, + "grad_norm": 0.09831867249987687, + "learning_rate": 1.4936289238503324e-05, + "loss": 0.013, + "step": 11210 + }, + { + "epoch": 5.275764705882353, + "grad_norm": 0.06598186483770696, + "learning_rate": 1.491722575102987e-05, + "loss": 0.0063, + "step": 11211 + }, + { + "epoch": 5.276235294117647, + "grad_norm": 0.03873774771937692, + "learning_rate": 1.4898173800287888e-05, + "loss": 0.0039, + "step": 11212 + }, + { + "epoch": 5.276705882352942, + "grad_norm": 0.102658387403289, + "learning_rate": 1.4879133387904523e-05, + "loss": 0.0186, + "step": 11213 + }, + { + "epoch": 5.277176470588235, + "grad_norm": 0.13654046778947104, + "learning_rate": 1.4860104515505977e-05, + "loss": 0.0213, + "step": 11214 + }, + { + "epoch": 5.277647058823529, + "grad_norm": 0.09153939841670511, + "learning_rate": 1.4841087184717315e-05, + "loss": 0.0124, + "step": 11215 + }, + { + "epoch": 5.278117647058823, + "grad_norm": 0.029457391123694863, + "learning_rate": 1.4822081397162771e-05, + "loss": 0.0027, + "step": 11216 + }, + { + "epoch": 5.278588235294118, + "grad_norm": 0.0865955653123713, + "learning_rate": 1.4803087154465548e-05, + "loss": 0.0148, + "step": 11217 + }, + { + "epoch": 5.279058823529412, + "grad_norm": 6.702087040129537e-05, + "learning_rate": 1.4784104458247764e-05, + "loss": 0.0, + "step": 11218 + }, + { + "epoch": 5.279529411764706, + "grad_norm": 0.10956994066326474, + "learning_rate": 1.476513331013074e-05, + "loss": 0.0218, + "step": 11219 + }, + { + "epoch": 5.28, + "grad_norm": 8.348111122024166e-05, + "learning_rate": 1.4746173711734627e-05, + "loss": 0.0, + "step": 11220 + }, + { + "epoch": 5.2804705882352945, + "grad_norm": 0.12272291003345663, + "learning_rate": 1.4727225664678699e-05, + "loss": 0.0176, + "step": 11221 + }, + { + "epoch": 5.280941176470588, + "grad_norm": 1.4104508045583942e-05, + "learning_rate": 1.4708289170581227e-05, + "loss": 0.0, + "step": 11222 + }, + { + "epoch": 5.281411764705882, + "grad_norm": 0.12030632113720277, + "learning_rate": 1.4689364231059436e-05, + "loss": 0.0297, + "step": 11223 + }, + { + "epoch": 5.281882352941176, + "grad_norm": 0.14146408772855246, + "learning_rate": 1.4670450847729648e-05, + "loss": 0.0405, + "step": 11224 + }, + { + "epoch": 5.2823529411764705, + "grad_norm": 0.10654111270788291, + "learning_rate": 1.465154902220717e-05, + "loss": 0.0251, + "step": 11225 + }, + { + "epoch": 5.282823529411765, + "grad_norm": 0.01744983931624246, + "learning_rate": 1.4632658756106247e-05, + "loss": 0.0007, + "step": 11226 + }, + { + "epoch": 5.283294117647059, + "grad_norm": 0.024634943629474812, + "learning_rate": 1.4613780051040236e-05, + "loss": 0.0006, + "step": 11227 + }, + { + "epoch": 5.283764705882353, + "grad_norm": 0.07361000777800877, + "learning_rate": 1.4594912908621515e-05, + "loss": 0.0111, + "step": 11228 + }, + { + "epoch": 5.284235294117647, + "grad_norm": 0.11491450770995855, + "learning_rate": 1.4576057330461361e-05, + "loss": 0.025, + "step": 11229 + }, + { + "epoch": 5.284705882352942, + "grad_norm": 7.813858282529305e-06, + "learning_rate": 1.4557213318170152e-05, + "loss": 0.0, + "step": 11230 + }, + { + "epoch": 5.285176470588235, + "grad_norm": 1.350832863111448e-05, + "learning_rate": 1.453838087335732e-05, + "loss": 0.0, + "step": 11231 + }, + { + "epoch": 5.285647058823529, + "grad_norm": 0.030057198073284264, + "learning_rate": 1.4519559997631142e-05, + "loss": 0.0015, + "step": 11232 + }, + { + "epoch": 5.286117647058823, + "grad_norm": 0.20368552956544944, + "learning_rate": 1.45007506925991e-05, + "loss": 0.0263, + "step": 11233 + }, + { + "epoch": 5.286588235294118, + "grad_norm": 0.06148660367777262, + "learning_rate": 1.448195295986756e-05, + "loss": 0.0052, + "step": 11234 + }, + { + "epoch": 5.287058823529412, + "grad_norm": 0.10262276957291357, + "learning_rate": 1.4463166801041936e-05, + "loss": 0.026, + "step": 11235 + }, + { + "epoch": 5.287529411764706, + "grad_norm": 1.6248083429781552e-06, + "learning_rate": 1.4444392217726697e-05, + "loss": 0.0, + "step": 11236 + }, + { + "epoch": 5.288, + "grad_norm": 0.14925858160569358, + "learning_rate": 1.442562921152524e-05, + "loss": 0.0581, + "step": 11237 + }, + { + "epoch": 5.2884705882352945, + "grad_norm": 0.13040597876251614, + "learning_rate": 1.4406877784040021e-05, + "loss": 0.0411, + "step": 11238 + }, + { + "epoch": 5.288941176470589, + "grad_norm": 0.068283756419533, + "learning_rate": 1.4388137936872556e-05, + "loss": 0.006, + "step": 11239 + }, + { + "epoch": 5.289411764705882, + "grad_norm": 2.0101681827457577e-06, + "learning_rate": 1.4369409671623267e-05, + "loss": 0.0, + "step": 11240 + }, + { + "epoch": 5.289882352941176, + "grad_norm": 2.5890161295741127e-06, + "learning_rate": 1.4350692989891638e-05, + "loss": 0.0, + "step": 11241 + }, + { + "epoch": 5.2903529411764705, + "grad_norm": 0.09104936989635896, + "learning_rate": 1.4331987893276225e-05, + "loss": 0.0075, + "step": 11242 + }, + { + "epoch": 5.290823529411765, + "grad_norm": 5.314489821500177e-05, + "learning_rate": 1.4313294383374468e-05, + "loss": 0.0, + "step": 11243 + }, + { + "epoch": 5.291294117647059, + "grad_norm": 0.12907455016997135, + "learning_rate": 1.429461246178289e-05, + "loss": 0.0349, + "step": 11244 + }, + { + "epoch": 5.291764705882353, + "grad_norm": 1.6261627747011547e-06, + "learning_rate": 1.4275942130097096e-05, + "loss": 0.0, + "step": 11245 + }, + { + "epoch": 5.292235294117647, + "grad_norm": 0.07026774703154665, + "learning_rate": 1.4257283389911512e-05, + "loss": 0.004, + "step": 11246 + }, + { + "epoch": 5.292705882352941, + "grad_norm": 4.4503685977449935e-06, + "learning_rate": 1.4238636242819761e-05, + "loss": 0.0, + "step": 11247 + }, + { + "epoch": 5.293176470588235, + "grad_norm": 0.12179650062056503, + "learning_rate": 1.422000069041439e-05, + "loss": 0.0213, + "step": 11248 + }, + { + "epoch": 5.293647058823529, + "grad_norm": 0.06961812782918385, + "learning_rate": 1.420137673428694e-05, + "loss": 0.0076, + "step": 11249 + }, + { + "epoch": 5.294117647058823, + "grad_norm": 0.2167601701485837, + "learning_rate": 1.4182764376028006e-05, + "loss": 0.0771, + "step": 11250 + }, + { + "epoch": 5.294588235294118, + "grad_norm": 0.11232004550976674, + "learning_rate": 1.4164163617227197e-05, + "loss": 0.0235, + "step": 11251 + }, + { + "epoch": 5.295058823529412, + "grad_norm": 3.3583826555347257e-06, + "learning_rate": 1.4145574459473064e-05, + "loss": 0.0, + "step": 11252 + }, + { + "epoch": 5.295529411764706, + "grad_norm": 0.08059787420209015, + "learning_rate": 1.4126996904353233e-05, + "loss": 0.0085, + "step": 11253 + }, + { + "epoch": 5.296, + "grad_norm": 0.07437066906352213, + "learning_rate": 1.4108430953454352e-05, + "loss": 0.0113, + "step": 11254 + }, + { + "epoch": 5.2964705882352945, + "grad_norm": 0.0367069012855344, + "learning_rate": 1.4089876608361971e-05, + "loss": 0.0038, + "step": 11255 + }, + { + "epoch": 5.296941176470588, + "grad_norm": 0.08511590141709156, + "learning_rate": 1.4071333870660807e-05, + "loss": 0.0117, + "step": 11256 + }, + { + "epoch": 5.297411764705882, + "grad_norm": 0.08876133268142623, + "learning_rate": 1.4052802741934404e-05, + "loss": 0.0253, + "step": 11257 + }, + { + "epoch": 5.297882352941176, + "grad_norm": 0.4036520095914506, + "learning_rate": 1.4034283223765485e-05, + "loss": 0.0843, + "step": 11258 + }, + { + "epoch": 5.2983529411764705, + "grad_norm": 4.537970925145731e-06, + "learning_rate": 1.4015775317735717e-05, + "loss": 0.0, + "step": 11259 + }, + { + "epoch": 5.298823529411765, + "grad_norm": 0.10864996965722301, + "learning_rate": 1.3997279025425701e-05, + "loss": 0.0128, + "step": 11260 + }, + { + "epoch": 5.299294117647059, + "grad_norm": 4.7663611351035545e-05, + "learning_rate": 1.397879434841514e-05, + "loss": 0.0, + "step": 11261 + }, + { + "epoch": 5.299764705882353, + "grad_norm": 0.15033171027583153, + "learning_rate": 1.3960321288282722e-05, + "loss": 0.0515, + "step": 11262 + }, + { + "epoch": 5.300235294117647, + "grad_norm": 1.8630260427772542e-06, + "learning_rate": 1.3941859846606152e-05, + "loss": 0.0, + "step": 11263 + }, + { + "epoch": 5.300705882352942, + "grad_norm": 0.03362262183291869, + "learning_rate": 1.3923410024962117e-05, + "loss": 0.0022, + "step": 11264 + }, + { + "epoch": 5.301176470588235, + "grad_norm": 0.09571468280170485, + "learning_rate": 1.390497182492631e-05, + "loss": 0.0187, + "step": 11265 + }, + { + "epoch": 5.301647058823529, + "grad_norm": 0.1890665311855036, + "learning_rate": 1.388654524807345e-05, + "loss": 0.0499, + "step": 11266 + }, + { + "epoch": 5.302117647058823, + "grad_norm": 0.03142292986430104, + "learning_rate": 1.3868130295977282e-05, + "loss": 0.0022, + "step": 11267 + }, + { + "epoch": 5.302588235294118, + "grad_norm": 0.09263604890196143, + "learning_rate": 1.3849726970210479e-05, + "loss": 0.0098, + "step": 11268 + }, + { + "epoch": 5.303058823529412, + "grad_norm": 0.13685207813264896, + "learning_rate": 1.3831335272344801e-05, + "loss": 0.0246, + "step": 11269 + }, + { + "epoch": 5.303529411764706, + "grad_norm": 0.13764925539051764, + "learning_rate": 1.3812955203951026e-05, + "loss": 0.0334, + "step": 11270 + }, + { + "epoch": 5.304, + "grad_norm": 0.019616090600239434, + "learning_rate": 1.3794586766598848e-05, + "loss": 0.001, + "step": 11271 + }, + { + "epoch": 5.3044705882352945, + "grad_norm": 0.05333848014812232, + "learning_rate": 1.3776229961857043e-05, + "loss": 0.0012, + "step": 11272 + }, + { + "epoch": 5.304941176470588, + "grad_norm": 0.11502170064612284, + "learning_rate": 1.3757884791293394e-05, + "loss": 0.0199, + "step": 11273 + }, + { + "epoch": 5.305411764705882, + "grad_norm": 0.13734765058326956, + "learning_rate": 1.373955125647463e-05, + "loss": 0.0275, + "step": 11274 + }, + { + "epoch": 5.305882352941176, + "grad_norm": 0.07583178433976547, + "learning_rate": 1.3721229358966545e-05, + "loss": 0.0105, + "step": 11275 + }, + { + "epoch": 5.3063529411764705, + "grad_norm": 2.7430704117880948e-05, + "learning_rate": 1.3702919100333925e-05, + "loss": 0.0, + "step": 11276 + }, + { + "epoch": 5.306823529411765, + "grad_norm": 0.11660878873456808, + "learning_rate": 1.3684620482140534e-05, + "loss": 0.0291, + "step": 11277 + }, + { + "epoch": 5.307294117647059, + "grad_norm": 0.12320176015514112, + "learning_rate": 1.3666333505949173e-05, + "loss": 0.0231, + "step": 11278 + }, + { + "epoch": 5.307764705882353, + "grad_norm": 0.017360143181596392, + "learning_rate": 1.3648058173321675e-05, + "loss": 0.0008, + "step": 11279 + }, + { + "epoch": 5.308235294117647, + "grad_norm": 0.1096416605020681, + "learning_rate": 1.3629794485818791e-05, + "loss": 0.0241, + "step": 11280 + }, + { + "epoch": 5.308705882352941, + "grad_norm": 0.03224178619613451, + "learning_rate": 1.361154244500034e-05, + "loss": 0.0023, + "step": 11281 + }, + { + "epoch": 5.309176470588235, + "grad_norm": 0.14051886970459687, + "learning_rate": 1.3593302052425159e-05, + "loss": 0.0405, + "step": 11282 + }, + { + "epoch": 5.309647058823529, + "grad_norm": 0.13826573223016794, + "learning_rate": 1.3575073309651052e-05, + "loss": 0.0322, + "step": 11283 + }, + { + "epoch": 5.310117647058823, + "grad_norm": 0.10691280055513959, + "learning_rate": 1.3556856218234858e-05, + "loss": 0.0328, + "step": 11284 + }, + { + "epoch": 5.310588235294118, + "grad_norm": 0.024889171909316693, + "learning_rate": 1.3538650779732362e-05, + "loss": 0.0008, + "step": 11285 + }, + { + "epoch": 5.311058823529412, + "grad_norm": 2.3572057592114658e-05, + "learning_rate": 1.3520456995698443e-05, + "loss": 0.0, + "step": 11286 + }, + { + "epoch": 5.311529411764706, + "grad_norm": 0.11527846640801456, + "learning_rate": 1.3502274867686936e-05, + "loss": 0.0277, + "step": 11287 + }, + { + "epoch": 5.312, + "grad_norm": 0.06047836128629979, + "learning_rate": 1.348410439725065e-05, + "loss": 0.003, + "step": 11288 + }, + { + "epoch": 5.3124705882352945, + "grad_norm": 0.08293498137728049, + "learning_rate": 1.3465945585941462e-05, + "loss": 0.0147, + "step": 11289 + }, + { + "epoch": 5.312941176470588, + "grad_norm": 3.212842637001044e-05, + "learning_rate": 1.3447798435310247e-05, + "loss": 0.0, + "step": 11290 + }, + { + "epoch": 5.313411764705882, + "grad_norm": 0.07793234594679169, + "learning_rate": 1.3429662946906799e-05, + "loss": 0.0083, + "step": 11291 + }, + { + "epoch": 5.313882352941176, + "grad_norm": 0.13777340485217454, + "learning_rate": 1.3411539122279996e-05, + "loss": 0.0444, + "step": 11292 + }, + { + "epoch": 5.3143529411764705, + "grad_norm": 0.0568006924818266, + "learning_rate": 1.3393426962977765e-05, + "loss": 0.0039, + "step": 11293 + }, + { + "epoch": 5.314823529411765, + "grad_norm": 0.12444857785897054, + "learning_rate": 1.3375326470546872e-05, + "loss": 0.0255, + "step": 11294 + }, + { + "epoch": 5.315294117647059, + "grad_norm": 0.1597440435224322, + "learning_rate": 1.3357237646533258e-05, + "loss": 0.0262, + "step": 11295 + }, + { + "epoch": 5.315764705882353, + "grad_norm": 6.744829252356587e-05, + "learning_rate": 1.3339160492481793e-05, + "loss": 0.0, + "step": 11296 + }, + { + "epoch": 5.316235294117647, + "grad_norm": 7.1798150856687645e-06, + "learning_rate": 1.3321095009936321e-05, + "loss": 0.0, + "step": 11297 + }, + { + "epoch": 5.316705882352942, + "grad_norm": 0.08191584841951861, + "learning_rate": 1.3303041200439729e-05, + "loss": 0.013, + "step": 11298 + }, + { + "epoch": 5.317176470588235, + "grad_norm": 0.09413582846801964, + "learning_rate": 1.3284999065533946e-05, + "loss": 0.0216, + "step": 11299 + }, + { + "epoch": 5.317647058823529, + "grad_norm": 0.11824744465493893, + "learning_rate": 1.326696860675981e-05, + "loss": 0.0351, + "step": 11300 + }, + { + "epoch": 5.318117647058823, + "grad_norm": 0.053208229408166316, + "learning_rate": 1.3248949825657217e-05, + "loss": 0.0036, + "step": 11301 + }, + { + "epoch": 5.318588235294118, + "grad_norm": 0.1585801055340356, + "learning_rate": 1.3230942723765092e-05, + "loss": 0.0373, + "step": 11302 + }, + { + "epoch": 5.319058823529412, + "grad_norm": 0.11045096556945508, + "learning_rate": 1.3212947302621284e-05, + "loss": 0.0182, + "step": 11303 + }, + { + "epoch": 5.319529411764706, + "grad_norm": 0.01981734148668451, + "learning_rate": 1.3194963563762718e-05, + "loss": 0.0008, + "step": 11304 + }, + { + "epoch": 5.32, + "grad_norm": 0.09154655661076405, + "learning_rate": 1.317699150872531e-05, + "loss": 0.0126, + "step": 11305 + }, + { + "epoch": 5.3204705882352945, + "grad_norm": 0.06669041189356428, + "learning_rate": 1.3159031139043918e-05, + "loss": 0.0101, + "step": 11306 + }, + { + "epoch": 5.320941176470588, + "grad_norm": 9.597882385741355e-06, + "learning_rate": 1.3141082456252467e-05, + "loss": 0.0, + "step": 11307 + }, + { + "epoch": 5.321411764705882, + "grad_norm": 0.12968531341034634, + "learning_rate": 1.312314546188386e-05, + "loss": 0.0255, + "step": 11308 + }, + { + "epoch": 5.321882352941176, + "grad_norm": 2.323523998527493e-07, + "learning_rate": 1.3105220157470009e-05, + "loss": 0.0, + "step": 11309 + }, + { + "epoch": 5.3223529411764705, + "grad_norm": 0.09325563038392261, + "learning_rate": 1.3087306544541836e-05, + "loss": 0.0144, + "step": 11310 + }, + { + "epoch": 5.322823529411765, + "grad_norm": 0.07839154734007668, + "learning_rate": 1.306940462462922e-05, + "loss": 0.0065, + "step": 11311 + }, + { + "epoch": 5.323294117647059, + "grad_norm": 0.1436699527353076, + "learning_rate": 1.3051514399261104e-05, + "loss": 0.0337, + "step": 11312 + }, + { + "epoch": 5.323764705882353, + "grad_norm": 1.3312298895075501e-06, + "learning_rate": 1.3033635869965398e-05, + "loss": 0.0, + "step": 11313 + }, + { + "epoch": 5.324235294117647, + "grad_norm": 0.13976547451603794, + "learning_rate": 1.3015769038268981e-05, + "loss": 0.0267, + "step": 11314 + }, + { + "epoch": 5.324705882352941, + "grad_norm": 9.231985702148844e-06, + "learning_rate": 1.2997913905697816e-05, + "loss": 0.0, + "step": 11315 + }, + { + "epoch": 5.325176470588235, + "grad_norm": 0.09427918509516434, + "learning_rate": 1.2980070473776782e-05, + "loss": 0.0172, + "step": 11316 + }, + { + "epoch": 5.325647058823529, + "grad_norm": 0.06119779196868691, + "learning_rate": 1.2962238744029795e-05, + "loss": 0.0093, + "step": 11317 + }, + { + "epoch": 5.326117647058823, + "grad_norm": 0.09066481418035305, + "learning_rate": 1.2944418717979804e-05, + "loss": 0.0093, + "step": 11318 + }, + { + "epoch": 5.326588235294118, + "grad_norm": 0.12166074926820104, + "learning_rate": 1.2926610397148707e-05, + "loss": 0.0232, + "step": 11319 + }, + { + "epoch": 5.327058823529412, + "grad_norm": 0.15672727136735332, + "learning_rate": 1.2908813783057404e-05, + "loss": 0.0304, + "step": 11320 + }, + { + "epoch": 5.327529411764706, + "grad_norm": 0.01037631875956053, + "learning_rate": 1.2891028877225863e-05, + "loss": 0.0004, + "step": 11321 + }, + { + "epoch": 5.328, + "grad_norm": 1.1387995461864652e-05, + "learning_rate": 1.287325568117295e-05, + "loss": 0.0, + "step": 11322 + }, + { + "epoch": 5.3284705882352945, + "grad_norm": 0.130871443128462, + "learning_rate": 1.2855494196416604e-05, + "loss": 0.0206, + "step": 11323 + }, + { + "epoch": 5.328941176470588, + "grad_norm": 9.070108066350595e-06, + "learning_rate": 1.2837744424473761e-05, + "loss": 0.0, + "step": 11324 + }, + { + "epoch": 5.329411764705882, + "grad_norm": 0.14538127642214724, + "learning_rate": 1.2820006366860292e-05, + "loss": 0.0187, + "step": 11325 + }, + { + "epoch": 5.329882352941176, + "grad_norm": 0.06225598470494037, + "learning_rate": 1.2802280025091165e-05, + "loss": 0.0086, + "step": 11326 + }, + { + "epoch": 5.3303529411764705, + "grad_norm": 0.20350710776415631, + "learning_rate": 1.2784565400680291e-05, + "loss": 0.073, + "step": 11327 + }, + { + "epoch": 5.330823529411765, + "grad_norm": 0.12726459436948517, + "learning_rate": 1.276686249514054e-05, + "loss": 0.0358, + "step": 11328 + }, + { + "epoch": 5.331294117647059, + "grad_norm": 0.019174623837249802, + "learning_rate": 1.2749171309983852e-05, + "loss": 0.0013, + "step": 11329 + }, + { + "epoch": 5.331764705882353, + "grad_norm": 4.287532520627512e-05, + "learning_rate": 1.273149184672117e-05, + "loss": 0.0, + "step": 11330 + }, + { + "epoch": 5.332235294117647, + "grad_norm": 0.06746779778809478, + "learning_rate": 1.2713824106862369e-05, + "loss": 0.0099, + "step": 11331 + }, + { + "epoch": 5.332705882352942, + "grad_norm": 0.051357329045502075, + "learning_rate": 1.2696168091916376e-05, + "loss": 0.0031, + "step": 11332 + }, + { + "epoch": 5.333176470588235, + "grad_norm": 0.044002741699354087, + "learning_rate": 1.2678523803391116e-05, + "loss": 0.0055, + "step": 11333 + }, + { + "epoch": 5.333647058823529, + "grad_norm": 0.05397775846750715, + "learning_rate": 1.2660891242793453e-05, + "loss": 0.0053, + "step": 11334 + }, + { + "epoch": 5.334117647058823, + "grad_norm": 0.09919247106313496, + "learning_rate": 1.2643270411629346e-05, + "loss": 0.0159, + "step": 11335 + }, + { + "epoch": 5.334588235294118, + "grad_norm": 0.00018266662010340602, + "learning_rate": 1.2625661311403662e-05, + "loss": 0.0, + "step": 11336 + }, + { + "epoch": 5.335058823529412, + "grad_norm": 0.05978137451990501, + "learning_rate": 1.2608063943620312e-05, + "loss": 0.0052, + "step": 11337 + }, + { + "epoch": 5.335529411764706, + "grad_norm": 0.11021008174775047, + "learning_rate": 1.2590478309782243e-05, + "loss": 0.0194, + "step": 11338 + }, + { + "epoch": 5.336, + "grad_norm": 0.17315379167266304, + "learning_rate": 1.2572904411391272e-05, + "loss": 0.0417, + "step": 11339 + }, + { + "epoch": 5.3364705882352945, + "grad_norm": 0.09597564185648408, + "learning_rate": 1.2555342249948347e-05, + "loss": 0.0119, + "step": 11340 + }, + { + "epoch": 5.336941176470588, + "grad_norm": 3.842457139171285e-05, + "learning_rate": 1.2537791826953387e-05, + "loss": 0.0, + "step": 11341 + }, + { + "epoch": 5.337411764705882, + "grad_norm": 0.13347586548515655, + "learning_rate": 1.2520253143905223e-05, + "loss": 0.0149, + "step": 11342 + }, + { + "epoch": 5.337882352941176, + "grad_norm": 0.07732309871190898, + "learning_rate": 1.2502726202301777e-05, + "loss": 0.0088, + "step": 11343 + }, + { + "epoch": 5.3383529411764705, + "grad_norm": 0.1658308916527984, + "learning_rate": 1.2485211003639967e-05, + "loss": 0.0377, + "step": 11344 + }, + { + "epoch": 5.338823529411765, + "grad_norm": 0.14379571439106834, + "learning_rate": 1.2467707549415612e-05, + "loss": 0.0358, + "step": 11345 + }, + { + "epoch": 5.339294117647059, + "grad_norm": 0.04862348901306389, + "learning_rate": 1.2450215841123618e-05, + "loss": 0.0062, + "step": 11346 + }, + { + "epoch": 5.339764705882353, + "grad_norm": 1.7245632265149692e-05, + "learning_rate": 1.2432735880257904e-05, + "loss": 0.0, + "step": 11347 + }, + { + "epoch": 5.340235294117647, + "grad_norm": 0.10929625255919191, + "learning_rate": 1.2415267668311263e-05, + "loss": 0.021, + "step": 11348 + }, + { + "epoch": 5.340705882352941, + "grad_norm": 0.18698151121875534, + "learning_rate": 1.2397811206775632e-05, + "loss": 0.0624, + "step": 11349 + }, + { + "epoch": 5.341176470588235, + "grad_norm": 0.015160030616658554, + "learning_rate": 1.2380366497141886e-05, + "loss": 0.0005, + "step": 11350 + }, + { + "epoch": 5.341647058823529, + "grad_norm": 0.07184544970747479, + "learning_rate": 1.23629335408998e-05, + "loss": 0.0108, + "step": 11351 + }, + { + "epoch": 5.342117647058823, + "grad_norm": 0.15892709163355057, + "learning_rate": 1.234551233953835e-05, + "loss": 0.0394, + "step": 11352 + }, + { + "epoch": 5.342588235294118, + "grad_norm": 0.06488743057028368, + "learning_rate": 1.232810289454531e-05, + "loss": 0.0074, + "step": 11353 + }, + { + "epoch": 5.343058823529412, + "grad_norm": 0.10238250869828064, + "learning_rate": 1.2310705207407561e-05, + "loss": 0.0127, + "step": 11354 + }, + { + "epoch": 5.343529411764706, + "grad_norm": 0.08364471068543722, + "learning_rate": 1.229331927961098e-05, + "loss": 0.0151, + "step": 11355 + }, + { + "epoch": 5.344, + "grad_norm": 1.2409329604401406e-05, + "learning_rate": 1.2275945112640363e-05, + "loss": 0.0, + "step": 11356 + }, + { + "epoch": 5.3444705882352945, + "grad_norm": 0.14418149345556489, + "learning_rate": 1.2258582707979554e-05, + "loss": 0.0553, + "step": 11357 + }, + { + "epoch": 5.344941176470588, + "grad_norm": 0.09276922832468375, + "learning_rate": 1.2241232067111418e-05, + "loss": 0.0155, + "step": 11358 + }, + { + "epoch": 5.345411764705882, + "grad_norm": 0.06608406273992988, + "learning_rate": 1.2223893191517753e-05, + "loss": 0.0092, + "step": 11359 + }, + { + "epoch": 5.345882352941176, + "grad_norm": 0.03222062672519753, + "learning_rate": 1.2206566082679392e-05, + "loss": 0.0028, + "step": 11360 + }, + { + "epoch": 5.3463529411764705, + "grad_norm": 0.09798386077082033, + "learning_rate": 1.2189250742076184e-05, + "loss": 0.009, + "step": 11361 + }, + { + "epoch": 5.346823529411765, + "grad_norm": 0.1608051026516072, + "learning_rate": 1.2171947171186897e-05, + "loss": 0.0497, + "step": 11362 + }, + { + "epoch": 5.347294117647059, + "grad_norm": 1.8017733224935477e-06, + "learning_rate": 1.2154655371489397e-05, + "loss": 0.0, + "step": 11363 + }, + { + "epoch": 5.347764705882353, + "grad_norm": 0.07904430644998288, + "learning_rate": 1.213737534446042e-05, + "loss": 0.0108, + "step": 11364 + }, + { + "epoch": 5.348235294117647, + "grad_norm": 0.1386514998073052, + "learning_rate": 1.2120107091575803e-05, + "loss": 0.0113, + "step": 11365 + }, + { + "epoch": 5.348705882352942, + "grad_norm": 0.11923763149848739, + "learning_rate": 1.2102850614310366e-05, + "loss": 0.0085, + "step": 11366 + }, + { + "epoch": 5.349176470588235, + "grad_norm": 0.035030814736745734, + "learning_rate": 1.2085605914137847e-05, + "loss": 0.0018, + "step": 11367 + }, + { + "epoch": 5.349647058823529, + "grad_norm": 0.00010278149283111873, + "learning_rate": 1.206837299253105e-05, + "loss": 0.0, + "step": 11368 + }, + { + "epoch": 5.350117647058823, + "grad_norm": 0.08008480866495683, + "learning_rate": 1.2051151850961766e-05, + "loss": 0.0083, + "step": 11369 + }, + { + "epoch": 5.350588235294118, + "grad_norm": 0.13636182995575188, + "learning_rate": 1.2033942490900716e-05, + "loss": 0.0398, + "step": 11370 + }, + { + "epoch": 5.351058823529412, + "grad_norm": 1.945646728942214e-05, + "learning_rate": 1.201674491381771e-05, + "loss": 0.0, + "step": 11371 + }, + { + "epoch": 5.351529411764706, + "grad_norm": 0.06155158163061773, + "learning_rate": 1.1999559121181523e-05, + "loss": 0.0081, + "step": 11372 + }, + { + "epoch": 5.352, + "grad_norm": 0.1958812698958816, + "learning_rate": 1.1982385114459846e-05, + "loss": 0.0644, + "step": 11373 + }, + { + "epoch": 5.352470588235294, + "grad_norm": 0.11982118825171356, + "learning_rate": 1.1965222895119442e-05, + "loss": 0.0204, + "step": 11374 + }, + { + "epoch": 5.352941176470588, + "grad_norm": 0.06380985937216853, + "learning_rate": 1.1948072464626101e-05, + "loss": 0.0111, + "step": 11375 + }, + { + "epoch": 5.353411764705882, + "grad_norm": 0.10736719039562441, + "learning_rate": 1.1930933824444472e-05, + "loss": 0.0159, + "step": 11376 + }, + { + "epoch": 5.353882352941176, + "grad_norm": 0.11963074361414004, + "learning_rate": 1.1913806976038331e-05, + "loss": 0.0301, + "step": 11377 + }, + { + "epoch": 5.3543529411764705, + "grad_norm": 0.005484829965968598, + "learning_rate": 1.1896691920870394e-05, + "loss": 0.0001, + "step": 11378 + }, + { + "epoch": 5.354823529411765, + "grad_norm": 3.683206446001957e-06, + "learning_rate": 1.1879588660402356e-05, + "loss": 0.0, + "step": 11379 + }, + { + "epoch": 5.355294117647059, + "grad_norm": 0.049478801833108155, + "learning_rate": 1.18624971960949e-05, + "loss": 0.005, + "step": 11380 + }, + { + "epoch": 5.355764705882353, + "grad_norm": 0.1162821969622775, + "learning_rate": 1.1845417529407791e-05, + "loss": 0.0185, + "step": 11381 + }, + { + "epoch": 5.356235294117647, + "grad_norm": 0.10029705870170813, + "learning_rate": 1.1828349661799629e-05, + "loss": 0.0225, + "step": 11382 + }, + { + "epoch": 5.356705882352941, + "grad_norm": 0.0196946193173927, + "learning_rate": 1.1811293594728133e-05, + "loss": 0.0011, + "step": 11383 + }, + { + "epoch": 5.357176470588235, + "grad_norm": 0.08700035948738363, + "learning_rate": 1.179424932965002e-05, + "loss": 0.0044, + "step": 11384 + }, + { + "epoch": 5.357647058823529, + "grad_norm": 0.05415709082265531, + "learning_rate": 1.1777216868020872e-05, + "loss": 0.0026, + "step": 11385 + }, + { + "epoch": 5.358117647058823, + "grad_norm": 0.09048809030518179, + "learning_rate": 1.1760196211295398e-05, + "loss": 0.0132, + "step": 11386 + }, + { + "epoch": 5.358588235294118, + "grad_norm": 0.08138222869269392, + "learning_rate": 1.1743187360927214e-05, + "loss": 0.0143, + "step": 11387 + }, + { + "epoch": 5.359058823529412, + "grad_norm": 7.517841907870598e-06, + "learning_rate": 1.1726190318368977e-05, + "loss": 0.0, + "step": 11388 + }, + { + "epoch": 5.359529411764706, + "grad_norm": 0.030351537369913844, + "learning_rate": 1.1709205085072344e-05, + "loss": 0.0011, + "step": 11389 + }, + { + "epoch": 5.36, + "grad_norm": 0.00016553358411717996, + "learning_rate": 1.1692231662487866e-05, + "loss": 0.0, + "step": 11390 + }, + { + "epoch": 5.3604705882352945, + "grad_norm": 4.0924550206377546e-05, + "learning_rate": 1.1675270052065222e-05, + "loss": 0.0, + "step": 11391 + }, + { + "epoch": 5.360941176470588, + "grad_norm": 0.00029638613307445916, + "learning_rate": 1.1658320255253e-05, + "loss": 0.0, + "step": 11392 + }, + { + "epoch": 5.361411764705882, + "grad_norm": 0.0023790745786091316, + "learning_rate": 1.1641382273498777e-05, + "loss": 0.0, + "step": 11393 + }, + { + "epoch": 5.361882352941176, + "grad_norm": 5.59376791711932e-06, + "learning_rate": 1.1624456108249164e-05, + "loss": 0.0, + "step": 11394 + }, + { + "epoch": 5.3623529411764705, + "grad_norm": 0.05377415285544973, + "learning_rate": 1.1607541760949735e-05, + "loss": 0.0076, + "step": 11395 + }, + { + "epoch": 5.362823529411765, + "grad_norm": 0.13177307381014336, + "learning_rate": 1.1590639233045002e-05, + "loss": 0.0226, + "step": 11396 + }, + { + "epoch": 5.363294117647059, + "grad_norm": 0.053750100642597366, + "learning_rate": 1.1573748525978626e-05, + "loss": 0.0067, + "step": 11397 + }, + { + "epoch": 5.363764705882353, + "grad_norm": 0.15169786429174903, + "learning_rate": 1.155686964119309e-05, + "loss": 0.0347, + "step": 11398 + }, + { + "epoch": 5.364235294117647, + "grad_norm": 0.007931838125387445, + "learning_rate": 1.154000258012992e-05, + "loss": 0.0002, + "step": 11399 + }, + { + "epoch": 5.364705882352941, + "grad_norm": 0.06260446968663409, + "learning_rate": 1.1523147344229716e-05, + "loss": 0.0092, + "step": 11400 + }, + { + "epoch": 5.365176470588235, + "grad_norm": 0.09164306417522991, + "learning_rate": 1.1506303934931926e-05, + "loss": 0.0079, + "step": 11401 + }, + { + "epoch": 5.365647058823529, + "grad_norm": 0.10137507518386012, + "learning_rate": 1.1489472353675083e-05, + "loss": 0.0115, + "step": 11402 + }, + { + "epoch": 5.366117647058823, + "grad_norm": 0.11158361272536364, + "learning_rate": 1.147265260189672e-05, + "loss": 0.0259, + "step": 11403 + }, + { + "epoch": 5.366588235294118, + "grad_norm": 0.05892495888976122, + "learning_rate": 1.1455844681033272e-05, + "loss": 0.0059, + "step": 11404 + }, + { + "epoch": 5.367058823529412, + "grad_norm": 0.1326058298676813, + "learning_rate": 1.1439048592520256e-05, + "loss": 0.0314, + "step": 11405 + }, + { + "epoch": 5.367529411764706, + "grad_norm": 0.0859995438225053, + "learning_rate": 1.1422264337792125e-05, + "loss": 0.0149, + "step": 11406 + }, + { + "epoch": 5.368, + "grad_norm": 8.53178248207134e-05, + "learning_rate": 1.1405491918282333e-05, + "loss": 0.0, + "step": 11407 + }, + { + "epoch": 5.368470588235294, + "grad_norm": 7.344717167287457e-05, + "learning_rate": 1.1388731335423334e-05, + "loss": 0.0, + "step": 11408 + }, + { + "epoch": 5.368941176470588, + "grad_norm": 0.15196315665605384, + "learning_rate": 1.1371982590646584e-05, + "loss": 0.0238, + "step": 11409 + }, + { + "epoch": 5.369411764705882, + "grad_norm": 0.13961336727136103, + "learning_rate": 1.135524568538247e-05, + "loss": 0.0322, + "step": 11410 + }, + { + "epoch": 5.369882352941176, + "grad_norm": 9.038445993347804e-06, + "learning_rate": 1.1338520621060415e-05, + "loss": 0.0, + "step": 11411 + }, + { + "epoch": 5.3703529411764706, + "grad_norm": 0.108042021499631, + "learning_rate": 1.1321807399108845e-05, + "loss": 0.0156, + "step": 11412 + }, + { + "epoch": 5.370823529411765, + "grad_norm": 0.027855620965241194, + "learning_rate": 1.1305106020955135e-05, + "loss": 0.0027, + "step": 11413 + }, + { + "epoch": 5.371294117647059, + "grad_norm": 0.0710517825608976, + "learning_rate": 1.1288416488025675e-05, + "loss": 0.0108, + "step": 11414 + }, + { + "epoch": 5.371764705882353, + "grad_norm": 0.04818444817958116, + "learning_rate": 1.1271738801745795e-05, + "loss": 0.0038, + "step": 11415 + }, + { + "epoch": 5.372235294117647, + "grad_norm": 0.1006649727030748, + "learning_rate": 1.1255072963539885e-05, + "loss": 0.0085, + "step": 11416 + }, + { + "epoch": 5.372705882352941, + "grad_norm": 8.462202514738157e-06, + "learning_rate": 1.123841897483131e-05, + "loss": 0.0, + "step": 11417 + }, + { + "epoch": 5.373176470588235, + "grad_norm": 0.10343419439416955, + "learning_rate": 1.1221776837042345e-05, + "loss": 0.0227, + "step": 11418 + }, + { + "epoch": 5.373647058823529, + "grad_norm": 0.05581431952428873, + "learning_rate": 1.1205146551594357e-05, + "loss": 0.0027, + "step": 11419 + }, + { + "epoch": 5.374117647058823, + "grad_norm": 5.397390936494171e-05, + "learning_rate": 1.1188528119907642e-05, + "loss": 0.0, + "step": 11420 + }, + { + "epoch": 5.374588235294118, + "grad_norm": 0.07389830809304385, + "learning_rate": 1.117192154340148e-05, + "loss": 0.0102, + "step": 11421 + }, + { + "epoch": 5.375058823529412, + "grad_norm": 0.025014808803994638, + "learning_rate": 1.115532682349417e-05, + "loss": 0.0013, + "step": 11422 + }, + { + "epoch": 5.375529411764706, + "grad_norm": 0.038349381874790964, + "learning_rate": 1.1138743961602998e-05, + "loss": 0.0037, + "step": 11423 + }, + { + "epoch": 5.376, + "grad_norm": 0.06569458851216596, + "learning_rate": 1.1122172959144177e-05, + "loss": 0.011, + "step": 11424 + }, + { + "epoch": 5.376470588235295, + "grad_norm": 0.11778077633960406, + "learning_rate": 1.1105613817532976e-05, + "loss": 0.0258, + "step": 11425 + }, + { + "epoch": 5.376941176470588, + "grad_norm": 0.1059299501871643, + "learning_rate": 1.108906653818365e-05, + "loss": 0.0129, + "step": 11426 + }, + { + "epoch": 5.377411764705882, + "grad_norm": 9.695266158902517e-06, + "learning_rate": 1.1072531122509381e-05, + "loss": 0.0, + "step": 11427 + }, + { + "epoch": 5.377882352941176, + "grad_norm": 0.10023728881070895, + "learning_rate": 1.1056007571922377e-05, + "loss": 0.0183, + "step": 11428 + }, + { + "epoch": 5.378352941176471, + "grad_norm": 0.19334670274919855, + "learning_rate": 1.1039495887833872e-05, + "loss": 0.0858, + "step": 11429 + }, + { + "epoch": 5.378823529411765, + "grad_norm": 0.10025019591186819, + "learning_rate": 1.1022996071653988e-05, + "loss": 0.0212, + "step": 11430 + }, + { + "epoch": 5.379294117647059, + "grad_norm": 0.02075359970198113, + "learning_rate": 1.1006508124791935e-05, + "loss": 0.0003, + "step": 11431 + }, + { + "epoch": 5.379764705882353, + "grad_norm": 0.08076024032469069, + "learning_rate": 1.099003204865585e-05, + "loss": 0.0096, + "step": 11432 + }, + { + "epoch": 5.3802352941176474, + "grad_norm": 2.351146076747921e-05, + "learning_rate": 1.0973567844652858e-05, + "loss": 0.0, + "step": 11433 + }, + { + "epoch": 5.380705882352941, + "grad_norm": 0.1296431069114084, + "learning_rate": 1.0957115514189086e-05, + "loss": 0.0341, + "step": 11434 + }, + { + "epoch": 5.381176470588235, + "grad_norm": 0.10706169341915144, + "learning_rate": 1.0940675058669673e-05, + "loss": 0.0223, + "step": 11435 + }, + { + "epoch": 5.381647058823529, + "grad_norm": 0.08875568007199938, + "learning_rate": 1.0924246479498682e-05, + "loss": 0.0144, + "step": 11436 + }, + { + "epoch": 5.3821176470588235, + "grad_norm": 0.0588188047848312, + "learning_rate": 1.0907829778079207e-05, + "loss": 0.0059, + "step": 11437 + }, + { + "epoch": 5.382588235294118, + "grad_norm": 0.14827051359229967, + "learning_rate": 1.089142495581331e-05, + "loss": 0.0345, + "step": 11438 + }, + { + "epoch": 5.383058823529412, + "grad_norm": 0.04168557214425714, + "learning_rate": 1.0875032014102036e-05, + "loss": 0.0025, + "step": 11439 + }, + { + "epoch": 5.383529411764706, + "grad_norm": 0.13985452311151447, + "learning_rate": 1.0858650954345433e-05, + "loss": 0.0352, + "step": 11440 + }, + { + "epoch": 5.384, + "grad_norm": 0.05528258796491239, + "learning_rate": 1.0842281777942514e-05, + "loss": 0.0062, + "step": 11441 + }, + { + "epoch": 5.384470588235294, + "grad_norm": 0.1193088629405463, + "learning_rate": 1.082592448629131e-05, + "loss": 0.0149, + "step": 11442 + }, + { + "epoch": 5.384941176470588, + "grad_norm": 7.777398926502415e-05, + "learning_rate": 1.0809579080788805e-05, + "loss": 0.0, + "step": 11443 + }, + { + "epoch": 5.385411764705882, + "grad_norm": 0.13358250819131678, + "learning_rate": 1.0793245562830966e-05, + "loss": 0.0359, + "step": 11444 + }, + { + "epoch": 5.385882352941176, + "grad_norm": 0.12455696748448909, + "learning_rate": 1.0776923933812775e-05, + "loss": 0.0307, + "step": 11445 + }, + { + "epoch": 5.386352941176471, + "grad_norm": 0.11579415163138575, + "learning_rate": 1.0760614195128153e-05, + "loss": 0.0247, + "step": 11446 + }, + { + "epoch": 5.386823529411765, + "grad_norm": 0.025961362436708643, + "learning_rate": 1.0744316348170035e-05, + "loss": 0.0011, + "step": 11447 + }, + { + "epoch": 5.387294117647059, + "grad_norm": 0.09582464360522026, + "learning_rate": 1.0728030394330373e-05, + "loss": 0.0074, + "step": 11448 + }, + { + "epoch": 5.387764705882353, + "grad_norm": 0.14502216240270419, + "learning_rate": 1.0711756335000005e-05, + "loss": 0.0292, + "step": 11449 + }, + { + "epoch": 5.3882352941176475, + "grad_norm": 0.06091890167347177, + "learning_rate": 1.069549417156887e-05, + "loss": 0.0039, + "step": 11450 + }, + { + "epoch": 5.388705882352941, + "grad_norm": 0.00011451075109464695, + "learning_rate": 1.0679243905425822e-05, + "loss": 0.0, + "step": 11451 + }, + { + "epoch": 5.389176470588235, + "grad_norm": 0.05822523350695746, + "learning_rate": 1.0663005537958702e-05, + "loss": 0.0045, + "step": 11452 + }, + { + "epoch": 5.389647058823529, + "grad_norm": 0.17325905346788575, + "learning_rate": 1.0646779070554335e-05, + "loss": 0.0505, + "step": 11453 + }, + { + "epoch": 5.3901176470588235, + "grad_norm": 5.061755313018279e-06, + "learning_rate": 1.0630564504598593e-05, + "loss": 0.0, + "step": 11454 + }, + { + "epoch": 5.390588235294118, + "grad_norm": 0.14155832487090952, + "learning_rate": 1.0614361841476238e-05, + "loss": 0.0371, + "step": 11455 + }, + { + "epoch": 5.391058823529412, + "grad_norm": 7.294352430330017e-05, + "learning_rate": 1.0598171082571044e-05, + "loss": 0.0, + "step": 11456 + }, + { + "epoch": 5.391529411764706, + "grad_norm": 0.09097128284635829, + "learning_rate": 1.0581992229265823e-05, + "loss": 0.0118, + "step": 11457 + }, + { + "epoch": 5.392, + "grad_norm": 0.08810714884731723, + "learning_rate": 1.0565825282942303e-05, + "loss": 0.017, + "step": 11458 + }, + { + "epoch": 5.392470588235295, + "grad_norm": 0.1094592292398399, + "learning_rate": 1.0549670244981211e-05, + "loss": 0.0309, + "step": 11459 + }, + { + "epoch": 5.392941176470588, + "grad_norm": 0.14110862608387764, + "learning_rate": 1.0533527116762296e-05, + "loss": 0.0521, + "step": 11460 + }, + { + "epoch": 5.393411764705882, + "grad_norm": 0.12263280270789224, + "learning_rate": 1.0517395899664239e-05, + "loss": 0.0349, + "step": 11461 + }, + { + "epoch": 5.393882352941176, + "grad_norm": 0.009599085894278988, + "learning_rate": 1.0501276595064716e-05, + "loss": 0.0003, + "step": 11462 + }, + { + "epoch": 5.394352941176471, + "grad_norm": 0.1442808432588958, + "learning_rate": 1.0485169204340432e-05, + "loss": 0.0331, + "step": 11463 + }, + { + "epoch": 5.394823529411765, + "grad_norm": 0.15676828977203675, + "learning_rate": 1.0469073728867e-05, + "loss": 0.0349, + "step": 11464 + }, + { + "epoch": 5.395294117647059, + "grad_norm": 0.04171734260608962, + "learning_rate": 1.0452990170019088e-05, + "loss": 0.0035, + "step": 11465 + }, + { + "epoch": 5.395764705882353, + "grad_norm": 0.11262655286623088, + "learning_rate": 1.0436918529170264e-05, + "loss": 0.0273, + "step": 11466 + }, + { + "epoch": 5.3962352941176475, + "grad_norm": 0.13419388957564798, + "learning_rate": 1.0420858807693162e-05, + "loss": 0.0183, + "step": 11467 + }, + { + "epoch": 5.396705882352941, + "grad_norm": 0.03457924877783323, + "learning_rate": 1.0404811006959352e-05, + "loss": 0.0019, + "step": 11468 + }, + { + "epoch": 5.397176470588235, + "grad_norm": 0.007943899880395454, + "learning_rate": 1.0388775128339389e-05, + "loss": 0.0002, + "step": 11469 + }, + { + "epoch": 5.397647058823529, + "grad_norm": 0.02450688923198187, + "learning_rate": 1.037275117320281e-05, + "loss": 0.0004, + "step": 11470 + }, + { + "epoch": 5.3981176470588235, + "grad_norm": 0.08095419904733193, + "learning_rate": 1.0356739142918169e-05, + "loss": 0.0174, + "step": 11471 + }, + { + "epoch": 5.398588235294118, + "grad_norm": 0.1320695666413575, + "learning_rate": 1.0340739038852924e-05, + "loss": 0.0343, + "step": 11472 + }, + { + "epoch": 5.399058823529412, + "grad_norm": 0.1962799054845615, + "learning_rate": 1.0324750862373598e-05, + "loss": 0.0606, + "step": 11473 + }, + { + "epoch": 5.399529411764706, + "grad_norm": 0.027436765714389324, + "learning_rate": 1.0308774614845683e-05, + "loss": 0.0015, + "step": 11474 + }, + { + "epoch": 5.4, + "grad_norm": 0.06964460532409338, + "learning_rate": 1.0292810297633569e-05, + "loss": 0.0087, + "step": 11475 + }, + { + "epoch": 5.400470588235294, + "grad_norm": 0.07217268935720043, + "learning_rate": 1.02768579121007e-05, + "loss": 0.0129, + "step": 11476 + }, + { + "epoch": 5.400941176470588, + "grad_norm": 2.062115549963502e-05, + "learning_rate": 1.0260917459609535e-05, + "loss": 0.0, + "step": 11477 + }, + { + "epoch": 5.401411764705882, + "grad_norm": 0.15404434263742342, + "learning_rate": 1.024498894152142e-05, + "loss": 0.0247, + "step": 11478 + }, + { + "epoch": 5.401882352941176, + "grad_norm": 0.201955799124028, + "learning_rate": 1.0229072359196749e-05, + "loss": 0.0435, + "step": 11479 + }, + { + "epoch": 5.402352941176471, + "grad_norm": 0.14501214417135036, + "learning_rate": 1.0213167713994885e-05, + "loss": 0.0327, + "step": 11480 + }, + { + "epoch": 5.402823529411765, + "grad_norm": 0.09857951800788003, + "learning_rate": 1.0197275007274126e-05, + "loss": 0.0202, + "step": 11481 + }, + { + "epoch": 5.403294117647059, + "grad_norm": 1.1491255992616901e-05, + "learning_rate": 1.0181394240391798e-05, + "loss": 0.0, + "step": 11482 + }, + { + "epoch": 5.403764705882353, + "grad_norm": 0.08148638687679993, + "learning_rate": 1.016552541470424e-05, + "loss": 0.0113, + "step": 11483 + }, + { + "epoch": 5.4042352941176475, + "grad_norm": 0.0714993241370041, + "learning_rate": 1.0149668531566662e-05, + "loss": 0.0104, + "step": 11484 + }, + { + "epoch": 5.404705882352941, + "grad_norm": 0.0797936979200351, + "learning_rate": 1.0133823592333385e-05, + "loss": 0.0099, + "step": 11485 + }, + { + "epoch": 5.405176470588235, + "grad_norm": 0.00012635497306835556, + "learning_rate": 1.0117990598357589e-05, + "loss": 0.0, + "step": 11486 + }, + { + "epoch": 5.405647058823529, + "grad_norm": 0.015789796923701784, + "learning_rate": 1.0102169550991529e-05, + "loss": 0.0006, + "step": 11487 + }, + { + "epoch": 5.4061176470588235, + "grad_norm": 0.03022072567081932, + "learning_rate": 1.008636045158639e-05, + "loss": 0.0031, + "step": 11488 + }, + { + "epoch": 5.406588235294118, + "grad_norm": 0.1413783977343892, + "learning_rate": 1.0070563301492323e-05, + "loss": 0.0215, + "step": 11489 + }, + { + "epoch": 5.407058823529412, + "grad_norm": 0.15597671021688855, + "learning_rate": 1.0054778102058502e-05, + "loss": 0.0512, + "step": 11490 + }, + { + "epoch": 5.407529411764706, + "grad_norm": 0.1332633112647605, + "learning_rate": 1.0039004854633082e-05, + "loss": 0.0292, + "step": 11491 + }, + { + "epoch": 5.408, + "grad_norm": 0.0899018689586343, + "learning_rate": 1.0023243560563132e-05, + "loss": 0.0174, + "step": 11492 + }, + { + "epoch": 5.408470588235295, + "grad_norm": 0.04886241340324756, + "learning_rate": 1.0007494221194762e-05, + "loss": 0.0067, + "step": 11493 + }, + { + "epoch": 5.408941176470588, + "grad_norm": 0.17119845764197317, + "learning_rate": 9.991756837873061e-06, + "loss": 0.0432, + "step": 11494 + }, + { + "epoch": 5.409411764705882, + "grad_norm": 0.08119159360102318, + "learning_rate": 9.976031411942054e-06, + "loss": 0.0069, + "step": 11495 + }, + { + "epoch": 5.409882352941176, + "grad_norm": 0.05536832043629539, + "learning_rate": 9.960317944744783e-06, + "loss": 0.0023, + "step": 11496 + }, + { + "epoch": 5.410352941176471, + "grad_norm": 2.7660483233594154e-05, + "learning_rate": 9.944616437623243e-06, + "loss": 0.0, + "step": 11497 + }, + { + "epoch": 5.410823529411765, + "grad_norm": 0.11751905653563835, + "learning_rate": 9.928926891918426e-06, + "loss": 0.0302, + "step": 11498 + }, + { + "epoch": 5.411294117647059, + "grad_norm": 0.09342416399194522, + "learning_rate": 9.913249308970312e-06, + "loss": 0.0111, + "step": 11499 + }, + { + "epoch": 5.411764705882353, + "grad_norm": 0.08184128272897145, + "learning_rate": 9.89758369011781e-06, + "loss": 0.0151, + "step": 11500 + }, + { + "epoch": 5.412235294117647, + "grad_norm": 0.18270940096670585, + "learning_rate": 9.881930036698854e-06, + "loss": 0.0548, + "step": 11501 + }, + { + "epoch": 5.412705882352941, + "grad_norm": 0.0001098051623082245, + "learning_rate": 9.866288350050372e-06, + "loss": 0.0, + "step": 11502 + }, + { + "epoch": 5.413176470588235, + "grad_norm": 0.14448904593894793, + "learning_rate": 9.850658631508196e-06, + "loss": 0.0236, + "step": 11503 + }, + { + "epoch": 5.413647058823529, + "grad_norm": 0.04706145322882199, + "learning_rate": 9.835040882407207e-06, + "loss": 0.0016, + "step": 11504 + }, + { + "epoch": 5.4141176470588235, + "grad_norm": 0.13831375231051554, + "learning_rate": 9.819435104081242e-06, + "loss": 0.0139, + "step": 11505 + }, + { + "epoch": 5.414588235294118, + "grad_norm": 0.14400824303951018, + "learning_rate": 9.803841297863096e-06, + "loss": 0.0359, + "step": 11506 + }, + { + "epoch": 5.415058823529412, + "grad_norm": 0.077733836481698, + "learning_rate": 9.788259465084541e-06, + "loss": 0.0126, + "step": 11507 + }, + { + "epoch": 5.415529411764706, + "grad_norm": 0.051434227591437946, + "learning_rate": 9.772689607076394e-06, + "loss": 0.003, + "step": 11508 + }, + { + "epoch": 5.416, + "grad_norm": 0.10698990935852046, + "learning_rate": 9.757131725168343e-06, + "loss": 0.0207, + "step": 11509 + }, + { + "epoch": 5.416470588235294, + "grad_norm": 0.07233902838604568, + "learning_rate": 9.74158582068914e-06, + "loss": 0.0096, + "step": 11510 + }, + { + "epoch": 5.416941176470588, + "grad_norm": 0.10947779252210371, + "learning_rate": 9.726051894966492e-06, + "loss": 0.0208, + "step": 11511 + }, + { + "epoch": 5.417411764705882, + "grad_norm": 0.17529527766794328, + "learning_rate": 9.710529949327018e-06, + "loss": 0.042, + "step": 11512 + }, + { + "epoch": 5.417882352941176, + "grad_norm": 5.157845103522782e-06, + "learning_rate": 9.695019985096409e-06, + "loss": 0.0, + "step": 11513 + }, + { + "epoch": 5.418352941176471, + "grad_norm": 0.0636745584559615, + "learning_rate": 9.679522003599305e-06, + "loss": 0.0089, + "step": 11514 + }, + { + "epoch": 5.418823529411765, + "grad_norm": 8.240034304130533e-06, + "learning_rate": 9.664036006159266e-06, + "loss": 0.0, + "step": 11515 + }, + { + "epoch": 5.419294117647059, + "grad_norm": 0.07726071915573125, + "learning_rate": 9.648561994098913e-06, + "loss": 0.0119, + "step": 11516 + }, + { + "epoch": 5.419764705882353, + "grad_norm": 0.000685624474081005, + "learning_rate": 9.633099968739777e-06, + "loss": 0.0, + "step": 11517 + }, + { + "epoch": 5.4202352941176475, + "grad_norm": 0.1144451910595786, + "learning_rate": 9.617649931402399e-06, + "loss": 0.0152, + "step": 11518 + }, + { + "epoch": 5.420705882352941, + "grad_norm": 0.0855552697243157, + "learning_rate": 9.602211883406308e-06, + "loss": 0.0216, + "step": 11519 + }, + { + "epoch": 5.421176470588235, + "grad_norm": 0.0925617680771165, + "learning_rate": 9.586785826069932e-06, + "loss": 0.0137, + "step": 11520 + }, + { + "epoch": 5.421647058823529, + "grad_norm": 0.03920627495765503, + "learning_rate": 9.571371760710783e-06, + "loss": 0.0022, + "step": 11521 + }, + { + "epoch": 5.4221176470588235, + "grad_norm": 0.11364504888826397, + "learning_rate": 9.55596968864531e-06, + "loss": 0.0257, + "step": 11522 + }, + { + "epoch": 5.422588235294118, + "grad_norm": 1.1517262724982081e-05, + "learning_rate": 9.540579611188875e-06, + "loss": 0.0, + "step": 11523 + }, + { + "epoch": 5.423058823529412, + "grad_norm": 0.10944966727664293, + "learning_rate": 9.52520152965588e-06, + "loss": 0.0196, + "step": 11524 + }, + { + "epoch": 5.423529411764706, + "grad_norm": 0.12978030910491728, + "learning_rate": 9.509835445359736e-06, + "loss": 0.0299, + "step": 11525 + }, + { + "epoch": 5.424, + "grad_norm": 0.11922431169265088, + "learning_rate": 9.494481359612715e-06, + "loss": 0.0244, + "step": 11526 + }, + { + "epoch": 5.424470588235294, + "grad_norm": 0.10612909031811724, + "learning_rate": 9.479139273726166e-06, + "loss": 0.0151, + "step": 11527 + }, + { + "epoch": 5.424941176470588, + "grad_norm": 0.15538284605966227, + "learning_rate": 9.46380918901039e-06, + "loss": 0.0346, + "step": 11528 + }, + { + "epoch": 5.425411764705882, + "grad_norm": 0.1261284901640556, + "learning_rate": 9.448491106774591e-06, + "loss": 0.0262, + "step": 11529 + }, + { + "epoch": 5.425882352941176, + "grad_norm": 0.07928522059518048, + "learning_rate": 9.433185028327106e-06, + "loss": 0.0113, + "step": 11530 + }, + { + "epoch": 5.426352941176471, + "grad_norm": 3.882048095744349e-06, + "learning_rate": 9.417890954975071e-06, + "loss": 0.0, + "step": 11531 + }, + { + "epoch": 5.426823529411765, + "grad_norm": 0.00010395531255639101, + "learning_rate": 9.402608888024692e-06, + "loss": 0.0, + "step": 11532 + }, + { + "epoch": 5.427294117647059, + "grad_norm": 1.5809306382772808e-05, + "learning_rate": 9.38733882878116e-06, + "loss": 0.0, + "step": 11533 + }, + { + "epoch": 5.427764705882353, + "grad_norm": 0.0630720874091089, + "learning_rate": 9.37208077854858e-06, + "loss": 0.0043, + "step": 11534 + }, + { + "epoch": 5.428235294117647, + "grad_norm": 1.5442613406807288e-06, + "learning_rate": 9.356834738630075e-06, + "loss": 0.0, + "step": 11535 + }, + { + "epoch": 5.428705882352941, + "grad_norm": 0.13011154171938227, + "learning_rate": 9.341600710327773e-06, + "loss": 0.0279, + "step": 11536 + }, + { + "epoch": 5.429176470588235, + "grad_norm": 0.07316033729364389, + "learning_rate": 9.32637869494265e-06, + "loss": 0.0079, + "step": 11537 + }, + { + "epoch": 5.429647058823529, + "grad_norm": 0.08164120808017705, + "learning_rate": 9.311168693774796e-06, + "loss": 0.011, + "step": 11538 + }, + { + "epoch": 5.4301176470588235, + "grad_norm": 0.10114543232788022, + "learning_rate": 9.295970708123223e-06, + "loss": 0.0205, + "step": 11539 + }, + { + "epoch": 5.430588235294118, + "grad_norm": 0.05349298988817293, + "learning_rate": 9.280784739285896e-06, + "loss": 0.0028, + "step": 11540 + }, + { + "epoch": 5.431058823529412, + "grad_norm": 0.044111011307947944, + "learning_rate": 9.265610788559758e-06, + "loss": 0.002, + "step": 11541 + }, + { + "epoch": 5.431529411764706, + "grad_norm": 0.12443200334700985, + "learning_rate": 9.250448857240789e-06, + "loss": 0.0211, + "step": 11542 + }, + { + "epoch": 5.432, + "grad_norm": 0.11606406083490023, + "learning_rate": 9.235298946623838e-06, + "loss": 0.0302, + "step": 11543 + }, + { + "epoch": 5.432470588235294, + "grad_norm": 6.366629281482808e-06, + "learning_rate": 9.220161058002834e-06, + "loss": 0.0, + "step": 11544 + }, + { + "epoch": 5.432941176470588, + "grad_norm": 0.15908384572874912, + "learning_rate": 9.205035192670563e-06, + "loss": 0.0388, + "step": 11545 + }, + { + "epoch": 5.433411764705882, + "grad_norm": 0.08020864137215912, + "learning_rate": 9.189921351918888e-06, + "loss": 0.0125, + "step": 11546 + }, + { + "epoch": 5.433882352941176, + "grad_norm": 0.1506963609999855, + "learning_rate": 9.174819537038614e-06, + "loss": 0.033, + "step": 11547 + }, + { + "epoch": 5.434352941176471, + "grad_norm": 2.1623169070615746e-05, + "learning_rate": 9.159729749319473e-06, + "loss": 0.0, + "step": 11548 + }, + { + "epoch": 5.434823529411765, + "grad_norm": 0.06228630945662201, + "learning_rate": 9.144651990050234e-06, + "loss": 0.008, + "step": 11549 + }, + { + "epoch": 5.435294117647059, + "grad_norm": 5.693332282372742e-05, + "learning_rate": 9.129586260518634e-06, + "loss": 0.0, + "step": 11550 + }, + { + "epoch": 5.435764705882353, + "grad_norm": 7.714771021483707e-05, + "learning_rate": 9.114532562011312e-06, + "loss": 0.0, + "step": 11551 + }, + { + "epoch": 5.4362352941176475, + "grad_norm": 0.23774157556391143, + "learning_rate": 9.099490895813938e-06, + "loss": 0.0302, + "step": 11552 + }, + { + "epoch": 5.436705882352941, + "grad_norm": 0.13771921288675532, + "learning_rate": 9.084461263211168e-06, + "loss": 0.0161, + "step": 11553 + }, + { + "epoch": 5.437176470588235, + "grad_norm": 0.14800702522915093, + "learning_rate": 9.069443665486593e-06, + "loss": 0.0557, + "step": 11554 + }, + { + "epoch": 5.437647058823529, + "grad_norm": 0.10363390997620227, + "learning_rate": 9.054438103922767e-06, + "loss": 0.0174, + "step": 11555 + }, + { + "epoch": 5.4381176470588235, + "grad_norm": 0.06692877805052831, + "learning_rate": 9.039444579801302e-06, + "loss": 0.0048, + "step": 11556 + }, + { + "epoch": 5.438588235294118, + "grad_norm": 0.06867888683854914, + "learning_rate": 9.024463094402656e-06, + "loss": 0.0071, + "step": 11557 + }, + { + "epoch": 5.439058823529412, + "grad_norm": 0.07406213797517568, + "learning_rate": 9.009493649006322e-06, + "loss": 0.0102, + "step": 11558 + }, + { + "epoch": 5.439529411764706, + "grad_norm": 0.1293002130022213, + "learning_rate": 8.994536244890827e-06, + "loss": 0.0251, + "step": 11559 + }, + { + "epoch": 5.44, + "grad_norm": 0.1243807090722061, + "learning_rate": 8.979590883333532e-06, + "loss": 0.0087, + "step": 11560 + }, + { + "epoch": 5.440470588235294, + "grad_norm": 0.12651520925093038, + "learning_rate": 8.964657565610884e-06, + "loss": 0.0124, + "step": 11561 + }, + { + "epoch": 5.440941176470588, + "grad_norm": 0.08089519968471608, + "learning_rate": 8.949736292998277e-06, + "loss": 0.0086, + "step": 11562 + }, + { + "epoch": 5.441411764705882, + "grad_norm": 3.66294709180785e-06, + "learning_rate": 8.934827066770012e-06, + "loss": 0.0, + "step": 11563 + }, + { + "epoch": 5.441882352941176, + "grad_norm": 0.0002735436498112837, + "learning_rate": 8.919929888199435e-06, + "loss": 0.0, + "step": 11564 + }, + { + "epoch": 5.442352941176471, + "grad_norm": 0.12351498689803313, + "learning_rate": 8.90504475855886e-06, + "loss": 0.0191, + "step": 11565 + }, + { + "epoch": 5.442823529411765, + "grad_norm": 0.17922416726018353, + "learning_rate": 8.890171679119506e-06, + "loss": 0.0603, + "step": 11566 + }, + { + "epoch": 5.443294117647059, + "grad_norm": 0.13212588622387839, + "learning_rate": 8.875310651151657e-06, + "loss": 0.043, + "step": 11567 + }, + { + "epoch": 5.443764705882353, + "grad_norm": 0.11664963760341021, + "learning_rate": 8.86046167592448e-06, + "loss": 0.0207, + "step": 11568 + }, + { + "epoch": 5.444235294117647, + "grad_norm": 0.1592346665748749, + "learning_rate": 8.845624754706144e-06, + "loss": 0.0357, + "step": 11569 + }, + { + "epoch": 5.444705882352941, + "grad_norm": 0.00030270954860259984, + "learning_rate": 8.830799888763834e-06, + "loss": 0.0, + "step": 11570 + }, + { + "epoch": 5.445176470588235, + "grad_norm": 0.04436887976618817, + "learning_rate": 8.81598707936364e-06, + "loss": 0.0042, + "step": 11571 + }, + { + "epoch": 5.445647058823529, + "grad_norm": 0.08878444462292466, + "learning_rate": 8.801186327770632e-06, + "loss": 0.0049, + "step": 11572 + }, + { + "epoch": 5.4461176470588235, + "grad_norm": 7.430713741350549e-06, + "learning_rate": 8.7863976352489e-06, + "loss": 0.0, + "step": 11573 + }, + { + "epoch": 5.446588235294118, + "grad_norm": 0.13863423552231768, + "learning_rate": 8.771621003061463e-06, + "loss": 0.026, + "step": 11574 + }, + { + "epoch": 5.447058823529412, + "grad_norm": 0.1474329421808157, + "learning_rate": 8.756856432470315e-06, + "loss": 0.0426, + "step": 11575 + }, + { + "epoch": 5.447529411764706, + "grad_norm": 0.14650790533753247, + "learning_rate": 8.742103924736416e-06, + "loss": 0.0355, + "step": 11576 + }, + { + "epoch": 5.448, + "grad_norm": 0.15479011565877498, + "learning_rate": 8.727363481119704e-06, + "loss": 0.0375, + "step": 11577 + }, + { + "epoch": 5.448470588235294, + "grad_norm": 0.16949223094841262, + "learning_rate": 8.712635102879124e-06, + "loss": 0.044, + "step": 11578 + }, + { + "epoch": 5.448941176470588, + "grad_norm": 0.03154182240050428, + "learning_rate": 8.697918791272485e-06, + "loss": 0.0032, + "step": 11579 + }, + { + "epoch": 5.449411764705882, + "grad_norm": 0.07253287824183957, + "learning_rate": 8.68321454755665e-06, + "loss": 0.009, + "step": 11580 + }, + { + "epoch": 5.449882352941176, + "grad_norm": 0.018997997905108123, + "learning_rate": 8.668522372987479e-06, + "loss": 0.0007, + "step": 11581 + }, + { + "epoch": 5.450352941176471, + "grad_norm": 0.0902922831411641, + "learning_rate": 8.653842268819723e-06, + "loss": 0.0124, + "step": 11582 + }, + { + "epoch": 5.450823529411765, + "grad_norm": 0.08563507636188183, + "learning_rate": 8.639174236307106e-06, + "loss": 0.014, + "step": 11583 + }, + { + "epoch": 5.451294117647059, + "grad_norm": 0.08928294954311992, + "learning_rate": 8.624518276702414e-06, + "loss": 0.0109, + "step": 11584 + }, + { + "epoch": 5.451764705882353, + "grad_norm": 0.049435247647084304, + "learning_rate": 8.609874391257293e-06, + "loss": 0.0007, + "step": 11585 + }, + { + "epoch": 5.4522352941176475, + "grad_norm": 0.1291629361979709, + "learning_rate": 8.595242581222394e-06, + "loss": 0.0306, + "step": 11586 + }, + { + "epoch": 5.452705882352941, + "grad_norm": 0.15471752077755613, + "learning_rate": 8.580622847847402e-06, + "loss": 0.0273, + "step": 11587 + }, + { + "epoch": 5.453176470588235, + "grad_norm": 0.07324033316253206, + "learning_rate": 8.566015192380849e-06, + "loss": 0.0118, + "step": 11588 + }, + { + "epoch": 5.453647058823529, + "grad_norm": 0.1242521604168063, + "learning_rate": 8.551419616070322e-06, + "loss": 0.0363, + "step": 11589 + }, + { + "epoch": 5.4541176470588235, + "grad_norm": 0.13206575104557539, + "learning_rate": 8.536836120162389e-06, + "loss": 0.043, + "step": 11590 + }, + { + "epoch": 5.454588235294118, + "grad_norm": 0.05288472302012494, + "learning_rate": 8.522264705902504e-06, + "loss": 0.0085, + "step": 11591 + }, + { + "epoch": 5.455058823529412, + "grad_norm": 0.022177342739908422, + "learning_rate": 8.507705374535157e-06, + "loss": 0.001, + "step": 11592 + }, + { + "epoch": 5.455529411764706, + "grad_norm": 0.13180137581246007, + "learning_rate": 8.4931581273038e-06, + "loss": 0.0118, + "step": 11593 + }, + { + "epoch": 5.456, + "grad_norm": 0.16651434712191382, + "learning_rate": 8.47862296545081e-06, + "loss": 0.0418, + "step": 11594 + }, + { + "epoch": 5.456470588235294, + "grad_norm": 1.6613294139732517e-05, + "learning_rate": 8.464099890217607e-06, + "loss": 0.0, + "step": 11595 + }, + { + "epoch": 5.456941176470588, + "grad_norm": 0.07101323387143968, + "learning_rate": 8.449588902844484e-06, + "loss": 0.0117, + "step": 11596 + }, + { + "epoch": 5.457411764705882, + "grad_norm": 0.10768143970682524, + "learning_rate": 8.43509000457075e-06, + "loss": 0.0166, + "step": 11597 + }, + { + "epoch": 5.457882352941176, + "grad_norm": 0.14890881166169104, + "learning_rate": 8.420603196634745e-06, + "loss": 0.0285, + "step": 11598 + }, + { + "epoch": 5.458352941176471, + "grad_norm": 0.013799460256390523, + "learning_rate": 8.406128480273633e-06, + "loss": 0.0009, + "step": 11599 + }, + { + "epoch": 5.458823529411765, + "grad_norm": 0.08653553094664718, + "learning_rate": 8.391665856723655e-06, + "loss": 0.0162, + "step": 11600 + }, + { + "epoch": 5.459294117647059, + "grad_norm": 0.12355284388548968, + "learning_rate": 8.377215327220043e-06, + "loss": 0.0299, + "step": 11601 + }, + { + "epoch": 5.459764705882353, + "grad_norm": 0.20979208721843348, + "learning_rate": 8.362776892996858e-06, + "loss": 0.0616, + "step": 11602 + }, + { + "epoch": 5.460235294117647, + "grad_norm": 0.12382533500069755, + "learning_rate": 8.348350555287264e-06, + "loss": 0.0225, + "step": 11603 + }, + { + "epoch": 5.460705882352941, + "grad_norm": 1.937475136451223e-05, + "learning_rate": 8.333936315323343e-06, + "loss": 0.0, + "step": 11604 + }, + { + "epoch": 5.461176470588235, + "grad_norm": 0.08762271619963206, + "learning_rate": 8.319534174336106e-06, + "loss": 0.0124, + "step": 11605 + }, + { + "epoch": 5.461647058823529, + "grad_norm": 0.09589670121120379, + "learning_rate": 8.305144133555608e-06, + "loss": 0.0153, + "step": 11606 + }, + { + "epoch": 5.4621176470588235, + "grad_norm": 0.20336072995730908, + "learning_rate": 8.290766194210813e-06, + "loss": 0.0491, + "step": 11607 + }, + { + "epoch": 5.462588235294118, + "grad_norm": 0.12434987391093423, + "learning_rate": 8.276400357529655e-06, + "loss": 0.0387, + "step": 11608 + }, + { + "epoch": 5.463058823529412, + "grad_norm": 0.13549724680412245, + "learning_rate": 8.262046624739038e-06, + "loss": 0.0407, + "step": 11609 + }, + { + "epoch": 5.463529411764706, + "grad_norm": 0.10723120226377938, + "learning_rate": 8.247704997064896e-06, + "loss": 0.017, + "step": 11610 + }, + { + "epoch": 5.464, + "grad_norm": 2.5515292606114374e-06, + "learning_rate": 8.233375475732018e-06, + "loss": 0.0, + "step": 11611 + }, + { + "epoch": 5.464470588235294, + "grad_norm": 0.1540119435107742, + "learning_rate": 8.21905806196424e-06, + "loss": 0.0379, + "step": 11612 + }, + { + "epoch": 5.464941176470588, + "grad_norm": 0.0822141139627958, + "learning_rate": 8.204752756984351e-06, + "loss": 0.0123, + "step": 11613 + }, + { + "epoch": 5.465411764705882, + "grad_norm": 0.01463725937225799, + "learning_rate": 8.19045956201404e-06, + "loss": 0.0005, + "step": 11614 + }, + { + "epoch": 5.465882352941176, + "grad_norm": 0.07928340221988352, + "learning_rate": 8.176178478274082e-06, + "loss": 0.0139, + "step": 11615 + }, + { + "epoch": 5.466352941176471, + "grad_norm": 1.838810826800296e-05, + "learning_rate": 8.161909506984132e-06, + "loss": 0.0, + "step": 11616 + }, + { + "epoch": 5.466823529411765, + "grad_norm": 0.11516137004468967, + "learning_rate": 8.147652649362802e-06, + "loss": 0.0227, + "step": 11617 + }, + { + "epoch": 5.467294117647059, + "grad_norm": 0.09730556491576993, + "learning_rate": 8.133407906627715e-06, + "loss": 0.0144, + "step": 11618 + }, + { + "epoch": 5.467764705882353, + "grad_norm": 0.05152556805151635, + "learning_rate": 8.119175279995448e-06, + "loss": 0.0061, + "step": 11619 + }, + { + "epoch": 5.4682352941176475, + "grad_norm": 0.08935124928084746, + "learning_rate": 8.104954770681532e-06, + "loss": 0.0077, + "step": 11620 + }, + { + "epoch": 5.468705882352941, + "grad_norm": 0.12153800233672196, + "learning_rate": 8.090746379900475e-06, + "loss": 0.0341, + "step": 11621 + }, + { + "epoch": 5.469176470588235, + "grad_norm": 0.08572145613595102, + "learning_rate": 8.076550108865743e-06, + "loss": 0.0153, + "step": 11622 + }, + { + "epoch": 5.469647058823529, + "grad_norm": 0.07460124273866028, + "learning_rate": 8.062365958789746e-06, + "loss": 0.0056, + "step": 11623 + }, + { + "epoch": 5.4701176470588235, + "grad_norm": 0.14310052323594152, + "learning_rate": 8.048193930883933e-06, + "loss": 0.0391, + "step": 11624 + }, + { + "epoch": 5.470588235294118, + "grad_norm": 0.12742608424028, + "learning_rate": 8.034034026358587e-06, + "loss": 0.0348, + "step": 11625 + }, + { + "epoch": 5.471058823529412, + "grad_norm": 0.18654760977279147, + "learning_rate": 8.019886246423107e-06, + "loss": 0.0459, + "step": 11626 + }, + { + "epoch": 5.471529411764706, + "grad_norm": 0.16280007382733785, + "learning_rate": 8.005750592285725e-06, + "loss": 0.0514, + "step": 11627 + }, + { + "epoch": 5.4719999999999995, + "grad_norm": 0.06452374260328204, + "learning_rate": 7.991627065153727e-06, + "loss": 0.013, + "step": 11628 + }, + { + "epoch": 5.472470588235294, + "grad_norm": 0.07099059808163484, + "learning_rate": 7.977515666233347e-06, + "loss": 0.0074, + "step": 11629 + }, + { + "epoch": 5.472941176470588, + "grad_norm": 0.11799698561388208, + "learning_rate": 7.963416396729722e-06, + "loss": 0.0144, + "step": 11630 + }, + { + "epoch": 5.473411764705882, + "grad_norm": 0.024544025574726895, + "learning_rate": 7.94932925784702e-06, + "loss": 0.0012, + "step": 11631 + }, + { + "epoch": 5.473882352941176, + "grad_norm": 0.086351275128727, + "learning_rate": 7.935254250788364e-06, + "loss": 0.0167, + "step": 11632 + }, + { + "epoch": 5.474352941176471, + "grad_norm": 0.13904226850510287, + "learning_rate": 7.921191376755808e-06, + "loss": 0.0337, + "step": 11633 + }, + { + "epoch": 5.474823529411765, + "grad_norm": 0.0001955894841354819, + "learning_rate": 7.90714063695041e-06, + "loss": 0.0, + "step": 11634 + }, + { + "epoch": 5.475294117647059, + "grad_norm": 0.06698623718457875, + "learning_rate": 7.893102032572174e-06, + "loss": 0.0072, + "step": 11635 + }, + { + "epoch": 5.475764705882353, + "grad_norm": 1.1906087682992222e-05, + "learning_rate": 7.879075564820025e-06, + "loss": 0.0, + "step": 11636 + }, + { + "epoch": 5.476235294117647, + "grad_norm": 0.025879097998139665, + "learning_rate": 7.86506123489194e-06, + "loss": 0.001, + "step": 11637 + }, + { + "epoch": 5.476705882352941, + "grad_norm": 0.02955320946957077, + "learning_rate": 7.85105904398481e-06, + "loss": 0.0018, + "step": 11638 + }, + { + "epoch": 5.477176470588235, + "grad_norm": 0.08444145303298937, + "learning_rate": 7.837068993294443e-06, + "loss": 0.0143, + "step": 11639 + }, + { + "epoch": 5.477647058823529, + "grad_norm": 0.11269211752887345, + "learning_rate": 7.823091084015687e-06, + "loss": 0.0142, + "step": 11640 + }, + { + "epoch": 5.4781176470588235, + "grad_norm": 0.10157669340978646, + "learning_rate": 7.809125317342369e-06, + "loss": 0.0207, + "step": 11641 + }, + { + "epoch": 5.478588235294118, + "grad_norm": 0.18918355142758672, + "learning_rate": 7.79517169446715e-06, + "loss": 0.0542, + "step": 11642 + }, + { + "epoch": 5.479058823529412, + "grad_norm": 0.12509268050585026, + "learning_rate": 7.781230216581779e-06, + "loss": 0.0223, + "step": 11643 + }, + { + "epoch": 5.479529411764706, + "grad_norm": 0.09611889467012094, + "learning_rate": 7.767300884876971e-06, + "loss": 0.0181, + "step": 11644 + }, + { + "epoch": 5.48, + "grad_norm": 0.13495248454179845, + "learning_rate": 7.753383700542271e-06, + "loss": 0.0193, + "step": 11645 + }, + { + "epoch": 5.480470588235294, + "grad_norm": 0.03679295649465987, + "learning_rate": 7.739478664766347e-06, + "loss": 0.0048, + "step": 11646 + }, + { + "epoch": 5.480941176470588, + "grad_norm": 0.08077030100210131, + "learning_rate": 7.725585778736698e-06, + "loss": 0.0077, + "step": 11647 + }, + { + "epoch": 5.481411764705882, + "grad_norm": 4.281487883861129e-05, + "learning_rate": 7.711705043639893e-06, + "loss": 0.0, + "step": 11648 + }, + { + "epoch": 5.481882352941176, + "grad_norm": 0.11555791801200865, + "learning_rate": 7.697836460661399e-06, + "loss": 0.0344, + "step": 11649 + }, + { + "epoch": 5.482352941176471, + "grad_norm": 0.09591184653396054, + "learning_rate": 7.683980030985654e-06, + "loss": 0.0111, + "step": 11650 + }, + { + "epoch": 5.482823529411765, + "grad_norm": 0.13867976798488502, + "learning_rate": 7.670135755796074e-06, + "loss": 0.0245, + "step": 11651 + }, + { + "epoch": 5.483294117647059, + "grad_norm": 1.2824551699297628e-05, + "learning_rate": 7.656303636275036e-06, + "loss": 0.0, + "step": 11652 + }, + { + "epoch": 5.483764705882353, + "grad_norm": 1.0818318637236871e-05, + "learning_rate": 7.64248367360384e-06, + "loss": 0.0, + "step": 11653 + }, + { + "epoch": 5.484235294117647, + "grad_norm": 0.1174594457813624, + "learning_rate": 7.6286758689627915e-06, + "loss": 0.0317, + "step": 11654 + }, + { + "epoch": 5.484705882352941, + "grad_norm": 0.029473322718056628, + "learning_rate": 7.614880223531183e-06, + "loss": 0.0014, + "step": 11655 + }, + { + "epoch": 5.485176470588235, + "grad_norm": 0.12520778862488235, + "learning_rate": 7.6010967384871725e-06, + "loss": 0.0305, + "step": 11656 + }, + { + "epoch": 5.485647058823529, + "grad_norm": 0.045463667951083586, + "learning_rate": 7.587325415007967e-06, + "loss": 0.0044, + "step": 11657 + }, + { + "epoch": 5.4861176470588235, + "grad_norm": 0.04986531297067987, + "learning_rate": 7.573566254269708e-06, + "loss": 0.006, + "step": 11658 + }, + { + "epoch": 5.486588235294118, + "grad_norm": 0.00018188138565047358, + "learning_rate": 7.559819257447492e-06, + "loss": 0.0, + "step": 11659 + }, + { + "epoch": 5.487058823529412, + "grad_norm": 0.07950368023087936, + "learning_rate": 7.546084425715359e-06, + "loss": 0.0122, + "step": 11660 + }, + { + "epoch": 5.487529411764706, + "grad_norm": 0.0996185755960802, + "learning_rate": 7.532361760246375e-06, + "loss": 0.014, + "step": 11661 + }, + { + "epoch": 5.4879999999999995, + "grad_norm": 0.0753223277871135, + "learning_rate": 7.5186512622124495e-06, + "loss": 0.0079, + "step": 11662 + }, + { + "epoch": 5.488470588235294, + "grad_norm": 3.476480693721395e-06, + "learning_rate": 7.5049529327846135e-06, + "loss": 0.0, + "step": 11663 + }, + { + "epoch": 5.488941176470588, + "grad_norm": 0.10424206490514037, + "learning_rate": 7.491266773132715e-06, + "loss": 0.0198, + "step": 11664 + }, + { + "epoch": 5.489411764705882, + "grad_norm": 0.035307130338107034, + "learning_rate": 7.477592784425634e-06, + "loss": 0.0025, + "step": 11665 + }, + { + "epoch": 5.489882352941176, + "grad_norm": 0.14380915658582583, + "learning_rate": 7.463930967831222e-06, + "loss": 0.0382, + "step": 11666 + }, + { + "epoch": 5.490352941176471, + "grad_norm": 0.018522237275750367, + "learning_rate": 7.450281324516228e-06, + "loss": 0.0007, + "step": 11667 + }, + { + "epoch": 5.490823529411765, + "grad_norm": 0.12023404091489698, + "learning_rate": 7.436643855646402e-06, + "loss": 0.0265, + "step": 11668 + }, + { + "epoch": 5.491294117647059, + "grad_norm": 0.13686419170166078, + "learning_rate": 7.423018562386479e-06, + "loss": 0.0331, + "step": 11669 + }, + { + "epoch": 5.491764705882353, + "grad_norm": 0.07449942133874407, + "learning_rate": 7.40940544590008e-06, + "loss": 0.0093, + "step": 11670 + }, + { + "epoch": 5.492235294117647, + "grad_norm": 0.044723408616198265, + "learning_rate": 7.395804507349873e-06, + "loss": 0.0031, + "step": 11671 + }, + { + "epoch": 5.492705882352941, + "grad_norm": 1.593800451743682e-06, + "learning_rate": 7.3822157478974465e-06, + "loss": 0.0, + "step": 11672 + }, + { + "epoch": 5.493176470588235, + "grad_norm": 0.08759888387076717, + "learning_rate": 7.368639168703306e-06, + "loss": 0.0191, + "step": 11673 + }, + { + "epoch": 5.493647058823529, + "grad_norm": 0.13371688369859241, + "learning_rate": 7.3550747709269885e-06, + "loss": 0.0204, + "step": 11674 + }, + { + "epoch": 5.4941176470588236, + "grad_norm": 0.040644973905340255, + "learning_rate": 7.34152255572697e-06, + "loss": 0.0051, + "step": 11675 + }, + { + "epoch": 5.494588235294118, + "grad_norm": 0.03923807621359142, + "learning_rate": 7.327982524260639e-06, + "loss": 0.0044, + "step": 11676 + }, + { + "epoch": 5.495058823529412, + "grad_norm": 0.09237694671773274, + "learning_rate": 7.314454677684439e-06, + "loss": 0.013, + "step": 11677 + }, + { + "epoch": 5.495529411764706, + "grad_norm": 0.23818997669308967, + "learning_rate": 7.300939017153645e-06, + "loss": 0.0745, + "step": 11678 + }, + { + "epoch": 5.496, + "grad_norm": 0.11395116837454612, + "learning_rate": 7.287435543822584e-06, + "loss": 0.0169, + "step": 11679 + }, + { + "epoch": 5.496470588235294, + "grad_norm": 0.13648408646734342, + "learning_rate": 7.2739442588445666e-06, + "loss": 0.0402, + "step": 11680 + }, + { + "epoch": 5.496941176470588, + "grad_norm": 0.12534705314013897, + "learning_rate": 7.260465163371754e-06, + "loss": 0.0209, + "step": 11681 + }, + { + "epoch": 5.497411764705882, + "grad_norm": 0.13695039823010607, + "learning_rate": 7.246998258555342e-06, + "loss": 0.0227, + "step": 11682 + }, + { + "epoch": 5.497882352941176, + "grad_norm": 0.14822707746208746, + "learning_rate": 7.233543545545511e-06, + "loss": 0.0372, + "step": 11683 + }, + { + "epoch": 5.498352941176471, + "grad_norm": 0.09828731865679728, + "learning_rate": 7.220101025491293e-06, + "loss": 0.0111, + "step": 11684 + }, + { + "epoch": 5.498823529411765, + "grad_norm": 0.3575840106593083, + "learning_rate": 7.2066706995407845e-06, + "loss": 0.0458, + "step": 11685 + }, + { + "epoch": 5.499294117647059, + "grad_norm": 2.4762874141715075e-05, + "learning_rate": 7.193252568841018e-06, + "loss": 0.0, + "step": 11686 + }, + { + "epoch": 5.499764705882353, + "grad_norm": 0.14504764351105, + "learning_rate": 7.1798466345379115e-06, + "loss": 0.03, + "step": 11687 + }, + { + "epoch": 5.5002352941176476, + "grad_norm": 0.18801282243969036, + "learning_rate": 7.166452897776448e-06, + "loss": 0.0596, + "step": 11688 + }, + { + "epoch": 5.500705882352941, + "grad_norm": 0.093315155088284, + "learning_rate": 7.153071359700513e-06, + "loss": 0.013, + "step": 11689 + }, + { + "epoch": 5.501176470588235, + "grad_norm": 0.15777099504291145, + "learning_rate": 7.1397020214529255e-06, + "loss": 0.0547, + "step": 11690 + }, + { + "epoch": 5.501647058823529, + "grad_norm": 0.0874155653480602, + "learning_rate": 7.126344884175522e-06, + "loss": 0.0167, + "step": 11691 + }, + { + "epoch": 5.502117647058824, + "grad_norm": 0.10446951094646768, + "learning_rate": 7.1129999490090735e-06, + "loss": 0.0115, + "step": 11692 + }, + { + "epoch": 5.502588235294118, + "grad_norm": 0.06373309143658581, + "learning_rate": 7.099667217093252e-06, + "loss": 0.0062, + "step": 11693 + }, + { + "epoch": 5.503058823529412, + "grad_norm": 0.045037318826208476, + "learning_rate": 7.08634668956678e-06, + "loss": 0.0049, + "step": 11694 + }, + { + "epoch": 5.503529411764706, + "grad_norm": 0.13853087244906717, + "learning_rate": 7.073038367567313e-06, + "loss": 0.0372, + "step": 11695 + }, + { + "epoch": 5.504, + "grad_norm": 0.14399362404774382, + "learning_rate": 7.059742252231393e-06, + "loss": 0.0261, + "step": 11696 + }, + { + "epoch": 5.504470588235294, + "grad_norm": 0.15967955339448464, + "learning_rate": 7.046458344694628e-06, + "loss": 0.0485, + "step": 11697 + }, + { + "epoch": 5.504941176470588, + "grad_norm": 0.12520966362286065, + "learning_rate": 7.033186646091477e-06, + "loss": 0.0242, + "step": 11698 + }, + { + "epoch": 5.505411764705882, + "grad_norm": 3.471565404318381e-05, + "learning_rate": 7.019927157555449e-06, + "loss": 0.0, + "step": 11699 + }, + { + "epoch": 5.5058823529411764, + "grad_norm": 0.12624844244666725, + "learning_rate": 7.006679880218974e-06, + "loss": 0.021, + "step": 11700 + }, + { + "epoch": 5.506352941176471, + "grad_norm": 0.10217272636345663, + "learning_rate": 6.993444815213378e-06, + "loss": 0.0148, + "step": 11701 + }, + { + "epoch": 5.506823529411765, + "grad_norm": 0.1198173716680431, + "learning_rate": 6.9802219636690585e-06, + "loss": 0.0387, + "step": 11702 + }, + { + "epoch": 5.507294117647059, + "grad_norm": 0.08593710971660126, + "learning_rate": 6.967011326715294e-06, + "loss": 0.0146, + "step": 11703 + }, + { + "epoch": 5.507764705882353, + "grad_norm": 0.1493736818743365, + "learning_rate": 6.953812905480316e-06, + "loss": 0.0191, + "step": 11704 + }, + { + "epoch": 5.508235294117647, + "grad_norm": 0.11766881176611076, + "learning_rate": 6.940626701091356e-06, + "loss": 0.0304, + "step": 11705 + }, + { + "epoch": 5.508705882352941, + "grad_norm": 2.0778353675171337e-05, + "learning_rate": 6.927452714674614e-06, + "loss": 0.0, + "step": 11706 + }, + { + "epoch": 5.509176470588235, + "grad_norm": 0.06990805757699321, + "learning_rate": 6.914290947355122e-06, + "loss": 0.0067, + "step": 11707 + }, + { + "epoch": 5.509647058823529, + "grad_norm": 0.14003700187499543, + "learning_rate": 6.901141400257048e-06, + "loss": 0.0137, + "step": 11708 + }, + { + "epoch": 5.510117647058824, + "grad_norm": 0.10396032213319759, + "learning_rate": 6.888004074503395e-06, + "loss": 0.0119, + "step": 11709 + }, + { + "epoch": 5.510588235294118, + "grad_norm": 7.719401504744899e-06, + "learning_rate": 6.87487897121613e-06, + "loss": 0.0, + "step": 11710 + }, + { + "epoch": 5.511058823529412, + "grad_norm": 0.06254993044902407, + "learning_rate": 6.861766091516274e-06, + "loss": 0.002, + "step": 11711 + }, + { + "epoch": 5.511529411764706, + "grad_norm": 0.08068976265403834, + "learning_rate": 6.848665436523631e-06, + "loss": 0.0104, + "step": 11712 + }, + { + "epoch": 5.5120000000000005, + "grad_norm": 0.07053897628451812, + "learning_rate": 6.83557700735714e-06, + "loss": 0.0053, + "step": 11713 + }, + { + "epoch": 5.512470588235294, + "grad_norm": 0.10464048147641317, + "learning_rate": 6.82250080513459e-06, + "loss": 0.0232, + "step": 11714 + }, + { + "epoch": 5.512941176470588, + "grad_norm": 0.08499578086258185, + "learning_rate": 6.809436830972753e-06, + "loss": 0.0107, + "step": 11715 + }, + { + "epoch": 5.513411764705882, + "grad_norm": 0.08196070396969746, + "learning_rate": 6.796385085987338e-06, + "loss": 0.0104, + "step": 11716 + }, + { + "epoch": 5.5138823529411765, + "grad_norm": 0.09954879510614754, + "learning_rate": 6.783345571293069e-06, + "loss": 0.0238, + "step": 11717 + }, + { + "epoch": 5.514352941176471, + "grad_norm": 0.20059139176566584, + "learning_rate": 6.770318288003556e-06, + "loss": 0.0272, + "step": 11718 + }, + { + "epoch": 5.514823529411765, + "grad_norm": 0.16694446648953126, + "learning_rate": 6.757303237231393e-06, + "loss": 0.0451, + "step": 11719 + }, + { + "epoch": 5.515294117647059, + "grad_norm": 0.1270427753449674, + "learning_rate": 6.744300420088156e-06, + "loss": 0.0233, + "step": 11720 + }, + { + "epoch": 5.5157647058823525, + "grad_norm": 0.08624755679159772, + "learning_rate": 6.73130983768429e-06, + "loss": 0.0154, + "step": 11721 + }, + { + "epoch": 5.516235294117647, + "grad_norm": 0.04309175623377095, + "learning_rate": 6.718331491129309e-06, + "loss": 0.0039, + "step": 11722 + }, + { + "epoch": 5.516705882352941, + "grad_norm": 3.321913686911883e-05, + "learning_rate": 6.705365381531641e-06, + "loss": 0.0, + "step": 11723 + }, + { + "epoch": 5.517176470588235, + "grad_norm": 0.17189145218556814, + "learning_rate": 6.6924115099985856e-06, + "loss": 0.0566, + "step": 11724 + }, + { + "epoch": 5.517647058823529, + "grad_norm": 0.05449862397395614, + "learning_rate": 6.679469877636523e-06, + "loss": 0.0037, + "step": 11725 + }, + { + "epoch": 5.518117647058824, + "grad_norm": 0.0733365204824804, + "learning_rate": 6.6665404855507195e-06, + "loss": 0.009, + "step": 11726 + }, + { + "epoch": 5.518588235294118, + "grad_norm": 0.19858241807532973, + "learning_rate": 6.6536233348453925e-06, + "loss": 0.0273, + "step": 11727 + }, + { + "epoch": 5.519058823529412, + "grad_norm": 0.10721149346849439, + "learning_rate": 6.640718426623759e-06, + "loss": 0.0287, + "step": 11728 + }, + { + "epoch": 5.519529411764706, + "grad_norm": 0.11595373057198423, + "learning_rate": 6.627825761987937e-06, + "loss": 0.0222, + "step": 11729 + }, + { + "epoch": 5.52, + "grad_norm": 5.7329509619262545e-06, + "learning_rate": 6.61494534203903e-06, + "loss": 0.0, + "step": 11730 + }, + { + "epoch": 5.520470588235294, + "grad_norm": 0.07547274917702596, + "learning_rate": 6.6020771678771066e-06, + "loss": 0.0084, + "step": 11731 + }, + { + "epoch": 5.520941176470588, + "grad_norm": 0.14428287204241977, + "learning_rate": 6.589221240601123e-06, + "loss": 0.0245, + "step": 11732 + }, + { + "epoch": 5.521411764705882, + "grad_norm": 0.13095938955082123, + "learning_rate": 6.576377561309082e-06, + "loss": 0.0278, + "step": 11733 + }, + { + "epoch": 5.5218823529411765, + "grad_norm": 0.11296462634452946, + "learning_rate": 6.563546131097907e-06, + "loss": 0.0101, + "step": 11734 + }, + { + "epoch": 5.522352941176471, + "grad_norm": 0.14425465114119598, + "learning_rate": 6.550726951063423e-06, + "loss": 0.0309, + "step": 11735 + }, + { + "epoch": 5.522823529411765, + "grad_norm": 1.3797630691220732e-05, + "learning_rate": 6.5379200223004855e-06, + "loss": 0.0, + "step": 11736 + }, + { + "epoch": 5.523294117647059, + "grad_norm": 0.11456198236970798, + "learning_rate": 6.5251253459028545e-06, + "loss": 0.0241, + "step": 11737 + }, + { + "epoch": 5.523764705882353, + "grad_norm": 0.1497842363514595, + "learning_rate": 6.512342922963254e-06, + "loss": 0.0198, + "step": 11738 + }, + { + "epoch": 5.524235294117647, + "grad_norm": 0.13008561622794051, + "learning_rate": 6.499572754573362e-06, + "loss": 0.0213, + "step": 11739 + }, + { + "epoch": 5.524705882352941, + "grad_norm": 0.0808545760346126, + "learning_rate": 6.48681484182384e-06, + "loss": 0.0145, + "step": 11740 + }, + { + "epoch": 5.525176470588235, + "grad_norm": 0.17294527769611096, + "learning_rate": 6.474069185804248e-06, + "loss": 0.0327, + "step": 11741 + }, + { + "epoch": 5.525647058823529, + "grad_norm": 0.0002705243287646678, + "learning_rate": 6.461335787603134e-06, + "loss": 0.0, + "step": 11742 + }, + { + "epoch": 5.526117647058824, + "grad_norm": 0.15131973916897534, + "learning_rate": 6.448614648308009e-06, + "loss": 0.031, + "step": 11743 + }, + { + "epoch": 5.526588235294118, + "grad_norm": 0.13988165503962197, + "learning_rate": 6.435905769005306e-06, + "loss": 0.0262, + "step": 11744 + }, + { + "epoch": 5.527058823529412, + "grad_norm": 0.10258976605089953, + "learning_rate": 6.423209150780423e-06, + "loss": 0.021, + "step": 11745 + }, + { + "epoch": 5.527529411764706, + "grad_norm": 0.07088139665215448, + "learning_rate": 6.4105247947177256e-06, + "loss": 0.0114, + "step": 11746 + }, + { + "epoch": 5.5280000000000005, + "grad_norm": 0.11477796763092561, + "learning_rate": 6.3978527019004964e-06, + "loss": 0.0275, + "step": 11747 + }, + { + "epoch": 5.528470588235294, + "grad_norm": 0.07853297597491411, + "learning_rate": 6.385192873411038e-06, + "loss": 0.0109, + "step": 11748 + }, + { + "epoch": 5.528941176470588, + "grad_norm": 0.08528125607477285, + "learning_rate": 6.372545310330518e-06, + "loss": 0.0083, + "step": 11749 + }, + { + "epoch": 5.529411764705882, + "grad_norm": 4.502527537345583e-05, + "learning_rate": 6.359910013739122e-06, + "loss": 0.0, + "step": 11750 + }, + { + "epoch": 5.5298823529411765, + "grad_norm": 0.11449665372271527, + "learning_rate": 6.347286984715938e-06, + "loss": 0.0237, + "step": 11751 + }, + { + "epoch": 5.530352941176471, + "grad_norm": 0.1385925170011942, + "learning_rate": 6.334676224339086e-06, + "loss": 0.0335, + "step": 11752 + }, + { + "epoch": 5.530823529411765, + "grad_norm": 0.08944295994113517, + "learning_rate": 6.322077733685538e-06, + "loss": 0.0148, + "step": 11753 + }, + { + "epoch": 5.531294117647059, + "grad_norm": 0.014792099280764718, + "learning_rate": 6.309491513831317e-06, + "loss": 0.0005, + "step": 11754 + }, + { + "epoch": 5.5317647058823525, + "grad_norm": 0.1758210560281767, + "learning_rate": 6.296917565851295e-06, + "loss": 0.0444, + "step": 11755 + }, + { + "epoch": 5.532235294117647, + "grad_norm": 2.828723322378707e-06, + "learning_rate": 6.284355890819398e-06, + "loss": 0.0, + "step": 11756 + }, + { + "epoch": 5.532705882352941, + "grad_norm": 0.13348993794165662, + "learning_rate": 6.271806489808401e-06, + "loss": 0.0196, + "step": 11757 + }, + { + "epoch": 5.533176470588235, + "grad_norm": 0.22508781793441815, + "learning_rate": 6.2592693638901125e-06, + "loss": 0.0837, + "step": 11758 + }, + { + "epoch": 5.533647058823529, + "grad_norm": 0.05356707644371873, + "learning_rate": 6.246744514135293e-06, + "loss": 0.0032, + "step": 11759 + }, + { + "epoch": 5.534117647058824, + "grad_norm": 0.010597751219077696, + "learning_rate": 6.234231941613571e-06, + "loss": 0.0004, + "step": 11760 + }, + { + "epoch": 5.534588235294118, + "grad_norm": 0.1432063951904986, + "learning_rate": 6.221731647393607e-06, + "loss": 0.0413, + "step": 11761 + }, + { + "epoch": 5.535058823529412, + "grad_norm": 0.04594574288411448, + "learning_rate": 6.209243632543015e-06, + "loss": 0.0024, + "step": 11762 + }, + { + "epoch": 5.535529411764706, + "grad_norm": 0.2629484794742319, + "learning_rate": 6.196767898128291e-06, + "loss": 0.0707, + "step": 11763 + }, + { + "epoch": 5.536, + "grad_norm": 0.016505089185790422, + "learning_rate": 6.184304445214932e-06, + "loss": 0.0005, + "step": 11764 + }, + { + "epoch": 5.536470588235294, + "grad_norm": 0.1144950039634875, + "learning_rate": 6.171853274867406e-06, + "loss": 0.0243, + "step": 11765 + }, + { + "epoch": 5.536941176470588, + "grad_norm": 0.0003993671123656153, + "learning_rate": 6.159414388149059e-06, + "loss": 0.0, + "step": 11766 + }, + { + "epoch": 5.537411764705882, + "grad_norm": 0.18960188846124656, + "learning_rate": 6.146987786122259e-06, + "loss": 0.0604, + "step": 11767 + }, + { + "epoch": 5.5378823529411765, + "grad_norm": 0.010227253047327271, + "learning_rate": 6.134573469848325e-06, + "loss": 0.0002, + "step": 11768 + }, + { + "epoch": 5.538352941176471, + "grad_norm": 0.07255226358311016, + "learning_rate": 6.1221714403874415e-06, + "loss": 0.0052, + "step": 11769 + }, + { + "epoch": 5.538823529411765, + "grad_norm": 0.10709496116282335, + "learning_rate": 6.109781698798843e-06, + "loss": 0.0078, + "step": 11770 + }, + { + "epoch": 5.539294117647059, + "grad_norm": 0.16225887243016843, + "learning_rate": 6.097404246140669e-06, + "loss": 0.0424, + "step": 11771 + }, + { + "epoch": 5.539764705882353, + "grad_norm": 0.06341562068236693, + "learning_rate": 6.085039083470006e-06, + "loss": 0.0059, + "step": 11772 + }, + { + "epoch": 5.540235294117647, + "grad_norm": 0.04341300719712362, + "learning_rate": 6.072686211842909e-06, + "loss": 0.0017, + "step": 11773 + }, + { + "epoch": 5.540705882352941, + "grad_norm": 0.08363681373832126, + "learning_rate": 6.060345632314384e-06, + "loss": 0.0108, + "step": 11774 + }, + { + "epoch": 5.541176470588235, + "grad_norm": 0.16088668088841865, + "learning_rate": 6.048017345938339e-06, + "loss": 0.0313, + "step": 11775 + }, + { + "epoch": 5.541647058823529, + "grad_norm": 0.08643631424688365, + "learning_rate": 6.035701353767713e-06, + "loss": 0.0107, + "step": 11776 + }, + { + "epoch": 5.542117647058824, + "grad_norm": 0.11898080248497692, + "learning_rate": 6.023397656854301e-06, + "loss": 0.0266, + "step": 11777 + }, + { + "epoch": 5.542588235294118, + "grad_norm": 0.2287530851722357, + "learning_rate": 6.011106256248943e-06, + "loss": 0.0353, + "step": 11778 + }, + { + "epoch": 5.543058823529412, + "grad_norm": 3.751911632046347e-05, + "learning_rate": 5.998827153001401e-06, + "loss": 0.0, + "step": 11779 + }, + { + "epoch": 5.543529411764706, + "grad_norm": 0.16798143522274914, + "learning_rate": 5.986560348160302e-06, + "loss": 0.048, + "step": 11780 + }, + { + "epoch": 5.5440000000000005, + "grad_norm": 0.14989893821238182, + "learning_rate": 5.9743058427733414e-06, + "loss": 0.0401, + "step": 11781 + }, + { + "epoch": 5.544470588235294, + "grad_norm": 0.06030512758435143, + "learning_rate": 5.9620636378871325e-06, + "loss": 0.0044, + "step": 11782 + }, + { + "epoch": 5.544941176470588, + "grad_norm": 0.36698055624488485, + "learning_rate": 5.9498337345471544e-06, + "loss": 0.0775, + "step": 11783 + }, + { + "epoch": 5.545411764705882, + "grad_norm": 2.508647747401858e-06, + "learning_rate": 5.93761613379794e-06, + "loss": 0.0, + "step": 11784 + }, + { + "epoch": 5.5458823529411765, + "grad_norm": 4.996432642513282e-05, + "learning_rate": 5.925410836682953e-06, + "loss": 0.0, + "step": 11785 + }, + { + "epoch": 5.546352941176471, + "grad_norm": 0.08936636927356405, + "learning_rate": 5.913217844244545e-06, + "loss": 0.0099, + "step": 11786 + }, + { + "epoch": 5.546823529411765, + "grad_norm": 0.04167230450511353, + "learning_rate": 5.90103715752408e-06, + "loss": 0.004, + "step": 11787 + }, + { + "epoch": 5.547294117647059, + "grad_norm": 0.08532471244495718, + "learning_rate": 5.888868777561862e-06, + "loss": 0.0182, + "step": 11788 + }, + { + "epoch": 5.5477647058823525, + "grad_norm": 0.015373790966662271, + "learning_rate": 5.8767127053970745e-06, + "loss": 0.0009, + "step": 11789 + }, + { + "epoch": 5.548235294117647, + "grad_norm": 0.17994633531777135, + "learning_rate": 5.864568942067954e-06, + "loss": 0.0458, + "step": 11790 + }, + { + "epoch": 5.548705882352941, + "grad_norm": 0.026508041070998935, + "learning_rate": 5.852437488611656e-06, + "loss": 0.0014, + "step": 11791 + }, + { + "epoch": 5.549176470588235, + "grad_norm": 0.09381167290135582, + "learning_rate": 5.840318346064199e-06, + "loss": 0.0195, + "step": 11792 + }, + { + "epoch": 5.549647058823529, + "grad_norm": 0.11381289881626479, + "learning_rate": 5.828211515460673e-06, + "loss": 0.0261, + "step": 11793 + }, + { + "epoch": 5.550117647058824, + "grad_norm": 0.1629746151898404, + "learning_rate": 5.816116997835052e-06, + "loss": 0.0279, + "step": 11794 + }, + { + "epoch": 5.550588235294118, + "grad_norm": 0.020915658035355743, + "learning_rate": 5.804034794220208e-06, + "loss": 0.0005, + "step": 11795 + }, + { + "epoch": 5.551058823529412, + "grad_norm": 0.0594022502776353, + "learning_rate": 5.791964905648116e-06, + "loss": 0.0062, + "step": 11796 + }, + { + "epoch": 5.551529411764706, + "grad_norm": 5.806776830000057e-06, + "learning_rate": 5.779907333149553e-06, + "loss": 0.0, + "step": 11797 + }, + { + "epoch": 5.552, + "grad_norm": 0.06639815835413543, + "learning_rate": 5.767862077754276e-06, + "loss": 0.0082, + "step": 11798 + }, + { + "epoch": 5.552470588235294, + "grad_norm": 0.07799369460744976, + "learning_rate": 5.755829140491064e-06, + "loss": 0.0087, + "step": 11799 + }, + { + "epoch": 5.552941176470588, + "grad_norm": 0.0008178167012518925, + "learning_rate": 5.743808522387544e-06, + "loss": 0.0, + "step": 11800 + }, + { + "epoch": 5.553411764705882, + "grad_norm": 0.09830445995917855, + "learning_rate": 5.7318002244703285e-06, + "loss": 0.0202, + "step": 11801 + }, + { + "epoch": 5.5538823529411765, + "grad_norm": 0.00052566984422711, + "learning_rate": 5.7198042477650475e-06, + "loss": 0.0, + "step": 11802 + }, + { + "epoch": 5.554352941176471, + "grad_norm": 0.09639943367480501, + "learning_rate": 5.7078205932961476e-06, + "loss": 0.0129, + "step": 11803 + }, + { + "epoch": 5.554823529411765, + "grad_norm": 0.14376232777307457, + "learning_rate": 5.69584926208711e-06, + "loss": 0.0308, + "step": 11804 + }, + { + "epoch": 5.555294117647059, + "grad_norm": 0.14839360532481766, + "learning_rate": 5.683890255160367e-06, + "loss": 0.0418, + "step": 11805 + }, + { + "epoch": 5.555764705882353, + "grad_norm": 0.12205459836764149, + "learning_rate": 5.671943573537269e-06, + "loss": 0.0126, + "step": 11806 + }, + { + "epoch": 5.556235294117647, + "grad_norm": 0.09564017855932379, + "learning_rate": 5.6600092182381156e-06, + "loss": 0.0141, + "step": 11807 + }, + { + "epoch": 5.556705882352941, + "grad_norm": 0.04594240438669743, + "learning_rate": 5.648087190282141e-06, + "loss": 0.0024, + "step": 11808 + }, + { + "epoch": 5.557176470588235, + "grad_norm": 0.07399697145952513, + "learning_rate": 5.636177490687565e-06, + "loss": 0.0128, + "step": 11809 + }, + { + "epoch": 5.557647058823529, + "grad_norm": 0.0724049809643844, + "learning_rate": 5.624280120471542e-06, + "loss": 0.0039, + "step": 11810 + }, + { + "epoch": 5.558117647058824, + "grad_norm": 0.09917362146837407, + "learning_rate": 5.612395080650156e-06, + "loss": 0.0224, + "step": 11811 + }, + { + "epoch": 5.558588235294118, + "grad_norm": 1.9684470318053785e-06, + "learning_rate": 5.600522372238431e-06, + "loss": 0.0, + "step": 11812 + }, + { + "epoch": 5.559058823529412, + "grad_norm": 7.590700361547429e-06, + "learning_rate": 5.588661996250387e-06, + "loss": 0.0, + "step": 11813 + }, + { + "epoch": 5.559529411764705, + "grad_norm": 8.581827291818919e-06, + "learning_rate": 5.576813953698933e-06, + "loss": 0.0, + "step": 11814 + }, + { + "epoch": 5.5600000000000005, + "grad_norm": 3.190681915711402e-05, + "learning_rate": 5.564978245595958e-06, + "loss": 0.0, + "step": 11815 + }, + { + "epoch": 5.560470588235294, + "grad_norm": 0.0732446355913278, + "learning_rate": 5.553154872952303e-06, + "loss": 0.0124, + "step": 11816 + }, + { + "epoch": 5.560941176470588, + "grad_norm": 0.11971303230732985, + "learning_rate": 5.541343836777729e-06, + "loss": 0.0325, + "step": 11817 + }, + { + "epoch": 5.561411764705882, + "grad_norm": 0.13945977014044156, + "learning_rate": 5.529545138080943e-06, + "loss": 0.0453, + "step": 11818 + }, + { + "epoch": 5.5618823529411765, + "grad_norm": 3.217973176466593e-05, + "learning_rate": 5.517758777869641e-06, + "loss": 0.0, + "step": 11819 + }, + { + "epoch": 5.562352941176471, + "grad_norm": 0.1721904352955385, + "learning_rate": 5.5059847571504e-06, + "loss": 0.0479, + "step": 11820 + }, + { + "epoch": 5.562823529411765, + "grad_norm": 3.5125857871864265e-05, + "learning_rate": 5.494223076928817e-06, + "loss": 0.0, + "step": 11821 + }, + { + "epoch": 5.563294117647059, + "grad_norm": 0.07821218953322191, + "learning_rate": 5.482473738209386e-06, + "loss": 0.0052, + "step": 11822 + }, + { + "epoch": 5.5637647058823525, + "grad_norm": 2.9938577052384795e-05, + "learning_rate": 5.470736741995541e-06, + "loss": 0.0, + "step": 11823 + }, + { + "epoch": 5.564235294117647, + "grad_norm": 0.047563797995626926, + "learning_rate": 5.4590120892896775e-06, + "loss": 0.0032, + "step": 11824 + }, + { + "epoch": 5.564705882352941, + "grad_norm": 1.2698137238696712e-06, + "learning_rate": 5.447299781093178e-06, + "loss": 0.0, + "step": 11825 + }, + { + "epoch": 5.565176470588235, + "grad_norm": 0.11527850525129728, + "learning_rate": 5.435599818406278e-06, + "loss": 0.0217, + "step": 11826 + }, + { + "epoch": 5.565647058823529, + "grad_norm": 0.11073094481804997, + "learning_rate": 5.423912202228275e-06, + "loss": 0.0206, + "step": 11827 + }, + { + "epoch": 5.566117647058824, + "grad_norm": 0.06674300010586902, + "learning_rate": 5.412236933557274e-06, + "loss": 0.0102, + "step": 11828 + }, + { + "epoch": 5.566588235294118, + "grad_norm": 0.12849625915768847, + "learning_rate": 5.400574013390441e-06, + "loss": 0.0332, + "step": 11829 + }, + { + "epoch": 5.567058823529412, + "grad_norm": 0.11355315213241991, + "learning_rate": 5.388923442723864e-06, + "loss": 0.025, + "step": 11830 + }, + { + "epoch": 5.567529411764706, + "grad_norm": 0.13588864003220805, + "learning_rate": 5.37728522255253e-06, + "loss": 0.0373, + "step": 11831 + }, + { + "epoch": 5.568, + "grad_norm": 0.06885922741858466, + "learning_rate": 5.365659353870394e-06, + "loss": 0.0119, + "step": 11832 + }, + { + "epoch": 5.568470588235294, + "grad_norm": 0.032212162119704686, + "learning_rate": 5.354045837670412e-06, + "loss": 0.0019, + "step": 11833 + }, + { + "epoch": 5.568941176470588, + "grad_norm": 0.03631764994820864, + "learning_rate": 5.342444674944374e-06, + "loss": 0.0042, + "step": 11834 + }, + { + "epoch": 5.569411764705882, + "grad_norm": 2.353842404996106e-05, + "learning_rate": 5.330855866683087e-06, + "loss": 0.0, + "step": 11835 + }, + { + "epoch": 5.5698823529411765, + "grad_norm": 0.09864156421440028, + "learning_rate": 5.319279413876343e-06, + "loss": 0.0196, + "step": 11836 + }, + { + "epoch": 5.570352941176471, + "grad_norm": 0.05058528674955643, + "learning_rate": 5.307715317512767e-06, + "loss": 0.0075, + "step": 11837 + }, + { + "epoch": 5.570823529411765, + "grad_norm": 0.0749405935621593, + "learning_rate": 5.29616357858002e-06, + "loss": 0.0106, + "step": 11838 + }, + { + "epoch": 5.571294117647059, + "grad_norm": 3.868164602595654e-06, + "learning_rate": 5.2846241980646795e-06, + "loss": 0.0, + "step": 11839 + }, + { + "epoch": 5.571764705882353, + "grad_norm": 0.20174395638212547, + "learning_rate": 5.273097176952224e-06, + "loss": 0.0523, + "step": 11840 + }, + { + "epoch": 5.572235294117647, + "grad_norm": 0.12002938220313134, + "learning_rate": 5.261582516227197e-06, + "loss": 0.0323, + "step": 11841 + }, + { + "epoch": 5.572705882352941, + "grad_norm": 0.0674199229946224, + "learning_rate": 5.250080216872932e-06, + "loss": 0.0126, + "step": 11842 + }, + { + "epoch": 5.573176470588235, + "grad_norm": 1.1716697762278552e-05, + "learning_rate": 5.238590279871824e-06, + "loss": 0.0, + "step": 11843 + }, + { + "epoch": 5.573647058823529, + "grad_norm": 0.07484628833139599, + "learning_rate": 5.227112706205155e-06, + "loss": 0.0093, + "step": 11844 + }, + { + "epoch": 5.574117647058824, + "grad_norm": 1.6610381077030607e-05, + "learning_rate": 5.2156474968531595e-06, + "loss": 0.0, + "step": 11845 + }, + { + "epoch": 5.574588235294118, + "grad_norm": 5.082590128067216e-06, + "learning_rate": 5.204194652795035e-06, + "loss": 0.0, + "step": 11846 + }, + { + "epoch": 5.575058823529412, + "grad_norm": 0.1600038014229532, + "learning_rate": 5.1927541750089175e-06, + "loss": 0.0257, + "step": 11847 + }, + { + "epoch": 5.575529411764705, + "grad_norm": 1.7433132363035097e-05, + "learning_rate": 5.181326064471858e-06, + "loss": 0.0, + "step": 11848 + }, + { + "epoch": 5.576, + "grad_norm": 0.02068669127872789, + "learning_rate": 5.169910322159876e-06, + "loss": 0.0007, + "step": 11849 + }, + { + "epoch": 5.576470588235294, + "grad_norm": 0.14948947396347617, + "learning_rate": 5.158506949047975e-06, + "loss": 0.0484, + "step": 11850 + }, + { + "epoch": 5.576941176470588, + "grad_norm": 0.08508667731097155, + "learning_rate": 5.147115946109992e-06, + "loss": 0.0127, + "step": 11851 + }, + { + "epoch": 5.577411764705882, + "grad_norm": 0.09858240815516049, + "learning_rate": 5.1357373143188e-06, + "loss": 0.0224, + "step": 11852 + }, + { + "epoch": 5.5778823529411765, + "grad_norm": 0.13911212663609093, + "learning_rate": 5.124371054646237e-06, + "loss": 0.0284, + "step": 11853 + }, + { + "epoch": 5.578352941176471, + "grad_norm": 0.1261171928411726, + "learning_rate": 5.11301716806296e-06, + "loss": 0.0312, + "step": 11854 + }, + { + "epoch": 5.578823529411765, + "grad_norm": 0.04361620647365961, + "learning_rate": 5.101675655538695e-06, + "loss": 0.0019, + "step": 11855 + }, + { + "epoch": 5.579294117647059, + "grad_norm": 1.5646865111139466e-06, + "learning_rate": 5.090346518042049e-06, + "loss": 0.0, + "step": 11856 + }, + { + "epoch": 5.5797647058823525, + "grad_norm": 0.2638711516307839, + "learning_rate": 5.079029756540598e-06, + "loss": 0.0274, + "step": 11857 + }, + { + "epoch": 5.580235294117647, + "grad_norm": 0.0678850081804814, + "learning_rate": 5.067725372000836e-06, + "loss": 0.0056, + "step": 11858 + }, + { + "epoch": 5.580705882352941, + "grad_norm": 0.04865347690565554, + "learning_rate": 5.056433365388207e-06, + "loss": 0.0059, + "step": 11859 + }, + { + "epoch": 5.581176470588235, + "grad_norm": 0.08141692086387431, + "learning_rate": 5.045153737667107e-06, + "loss": 0.0062, + "step": 11860 + }, + { + "epoch": 5.581647058823529, + "grad_norm": 0.06569051467862681, + "learning_rate": 5.033886489800898e-06, + "loss": 0.0054, + "step": 11861 + }, + { + "epoch": 5.582117647058824, + "grad_norm": 0.14618754747602497, + "learning_rate": 5.022631622751827e-06, + "loss": 0.032, + "step": 11862 + }, + { + "epoch": 5.582588235294118, + "grad_norm": 0.06368720039602435, + "learning_rate": 5.011389137481109e-06, + "loss": 0.0104, + "step": 11863 + }, + { + "epoch": 5.583058823529412, + "grad_norm": 0.08484028862417912, + "learning_rate": 5.000159034948958e-06, + "loss": 0.0202, + "step": 11864 + }, + { + "epoch": 5.583529411764706, + "grad_norm": 0.06493326877735414, + "learning_rate": 4.988941316114425e-06, + "loss": 0.0089, + "step": 11865 + }, + { + "epoch": 5.584, + "grad_norm": 0.17556052378442502, + "learning_rate": 4.9777359819355765e-06, + "loss": 0.0376, + "step": 11866 + }, + { + "epoch": 5.584470588235294, + "grad_norm": 0.1350557267330222, + "learning_rate": 4.9665430333694135e-06, + "loss": 0.0376, + "step": 11867 + }, + { + "epoch": 5.584941176470588, + "grad_norm": 0.00010225226628488274, + "learning_rate": 4.955362471371854e-06, + "loss": 0.0, + "step": 11868 + }, + { + "epoch": 5.585411764705882, + "grad_norm": 1.4093741423759015e-05, + "learning_rate": 4.944194296897785e-06, + "loss": 0.0, + "step": 11869 + }, + { + "epoch": 5.5858823529411765, + "grad_norm": 0.1291575480722455, + "learning_rate": 4.933038510901027e-06, + "loss": 0.0198, + "step": 11870 + }, + { + "epoch": 5.586352941176471, + "grad_norm": 4.6567654532620745e-06, + "learning_rate": 4.921895114334301e-06, + "loss": 0.0, + "step": 11871 + }, + { + "epoch": 5.586823529411765, + "grad_norm": 0.11552697871367609, + "learning_rate": 4.910764108149346e-06, + "loss": 0.0246, + "step": 11872 + }, + { + "epoch": 5.587294117647059, + "grad_norm": 0.10164124665080089, + "learning_rate": 4.899645493296817e-06, + "loss": 0.0185, + "step": 11873 + }, + { + "epoch": 5.587764705882353, + "grad_norm": 0.23648794468304765, + "learning_rate": 4.888539270726255e-06, + "loss": 0.0383, + "step": 11874 + }, + { + "epoch": 5.588235294117647, + "grad_norm": 0.15374949845722807, + "learning_rate": 4.877445441386218e-06, + "loss": 0.0455, + "step": 11875 + }, + { + "epoch": 5.588705882352941, + "grad_norm": 0.20376760890295448, + "learning_rate": 4.866364006224166e-06, + "loss": 0.0672, + "step": 11876 + }, + { + "epoch": 5.589176470588235, + "grad_norm": 0.09612555735315118, + "learning_rate": 4.85529496618649e-06, + "loss": 0.0216, + "step": 11877 + }, + { + "epoch": 5.589647058823529, + "grad_norm": 4.005220291061137e-06, + "learning_rate": 4.844238322218568e-06, + "loss": 0.0, + "step": 11878 + }, + { + "epoch": 5.590117647058824, + "grad_norm": 0.08177059996950493, + "learning_rate": 4.833194075264663e-06, + "loss": 0.0134, + "step": 11879 + }, + { + "epoch": 5.590588235294118, + "grad_norm": 3.438677877978516e-05, + "learning_rate": 4.8221622262680336e-06, + "loss": 0.0, + "step": 11880 + }, + { + "epoch": 5.591058823529412, + "grad_norm": 0.12373216810073984, + "learning_rate": 4.811142776170845e-06, + "loss": 0.0367, + "step": 11881 + }, + { + "epoch": 5.591529411764705, + "grad_norm": 0.0723858626200888, + "learning_rate": 4.800135725914195e-06, + "loss": 0.0099, + "step": 11882 + }, + { + "epoch": 5.592, + "grad_norm": 0.08457855866580637, + "learning_rate": 4.789141076438163e-06, + "loss": 0.0046, + "step": 11883 + }, + { + "epoch": 5.592470588235294, + "grad_norm": 0.03618925409837708, + "learning_rate": 4.778158828681733e-06, + "loss": 0.0032, + "step": 11884 + }, + { + "epoch": 5.592941176470588, + "grad_norm": 0.08027955425737274, + "learning_rate": 4.767188983582854e-06, + "loss": 0.0137, + "step": 11885 + }, + { + "epoch": 5.593411764705882, + "grad_norm": 0.06702854038304099, + "learning_rate": 4.75623154207841e-06, + "loss": 0.0101, + "step": 11886 + }, + { + "epoch": 5.5938823529411765, + "grad_norm": 0.03731478233585267, + "learning_rate": 4.745286505104201e-06, + "loss": 0.0008, + "step": 11887 + }, + { + "epoch": 5.594352941176471, + "grad_norm": 0.10378773365079387, + "learning_rate": 4.734353873594998e-06, + "loss": 0.0238, + "step": 11888 + }, + { + "epoch": 5.594823529411765, + "grad_norm": 0.07290487422144803, + "learning_rate": 4.723433648484503e-06, + "loss": 0.0066, + "step": 11889 + }, + { + "epoch": 5.595294117647059, + "grad_norm": 0.018137234617746963, + "learning_rate": 4.712525830705338e-06, + "loss": 0.0008, + "step": 11890 + }, + { + "epoch": 5.5957647058823525, + "grad_norm": 0.12973320307714548, + "learning_rate": 4.70163042118909e-06, + "loss": 0.022, + "step": 11891 + }, + { + "epoch": 5.596235294117647, + "grad_norm": 0.18449628716339994, + "learning_rate": 4.690747420866314e-06, + "loss": 0.0581, + "step": 11892 + }, + { + "epoch": 5.596705882352941, + "grad_norm": 0.06370491049154133, + "learning_rate": 4.679876830666434e-06, + "loss": 0.0069, + "step": 11893 + }, + { + "epoch": 5.597176470588235, + "grad_norm": 0.06755375282775049, + "learning_rate": 4.669018651517859e-06, + "loss": 0.0121, + "step": 11894 + }, + { + "epoch": 5.597647058823529, + "grad_norm": 0.0003517530014539098, + "learning_rate": 4.658172884347943e-06, + "loss": 0.0, + "step": 11895 + }, + { + "epoch": 5.598117647058824, + "grad_norm": 0.0861842561057261, + "learning_rate": 4.6473395300829495e-06, + "loss": 0.0089, + "step": 11896 + }, + { + "epoch": 5.598588235294118, + "grad_norm": 0.1068580430199434, + "learning_rate": 4.636518589648119e-06, + "loss": 0.0264, + "step": 11897 + }, + { + "epoch": 5.599058823529412, + "grad_norm": 0.1371391975300747, + "learning_rate": 4.625710063967613e-06, + "loss": 0.0217, + "step": 11898 + }, + { + "epoch": 5.599529411764706, + "grad_norm": 2.3290631393706405e-06, + "learning_rate": 4.614913953964511e-06, + "loss": 0.0, + "step": 11899 + }, + { + "epoch": 5.6, + "grad_norm": 3.91934546200164e-05, + "learning_rate": 4.604130260560873e-06, + "loss": 0.0, + "step": 11900 + }, + { + "epoch": 5.600470588235294, + "grad_norm": 0.10328293586678879, + "learning_rate": 4.593358984677681e-06, + "loss": 0.0119, + "step": 11901 + }, + { + "epoch": 5.600941176470588, + "grad_norm": 0.09759335751542728, + "learning_rate": 4.5826001272348315e-06, + "loss": 0.0217, + "step": 11902 + }, + { + "epoch": 5.601411764705882, + "grad_norm": 0.09685939817238001, + "learning_rate": 4.571853689151206e-06, + "loss": 0.0092, + "step": 11903 + }, + { + "epoch": 5.6018823529411765, + "grad_norm": 0.07925139511023674, + "learning_rate": 4.5611196713446045e-06, + "loss": 0.0062, + "step": 11904 + }, + { + "epoch": 5.602352941176471, + "grad_norm": 0.04630847734477738, + "learning_rate": 4.5503980747317425e-06, + "loss": 0.0014, + "step": 11905 + }, + { + "epoch": 5.602823529411765, + "grad_norm": 0.06015998714273541, + "learning_rate": 4.539688900228322e-06, + "loss": 0.0082, + "step": 11906 + }, + { + "epoch": 5.603294117647059, + "grad_norm": 0.07068478379854352, + "learning_rate": 4.528992148748944e-06, + "loss": 0.0079, + "step": 11907 + }, + { + "epoch": 5.603764705882353, + "grad_norm": 0.1298785827910519, + "learning_rate": 4.518307821207162e-06, + "loss": 0.0156, + "step": 11908 + }, + { + "epoch": 5.604235294117647, + "grad_norm": 0.1097170468935704, + "learning_rate": 4.507635918515496e-06, + "loss": 0.0275, + "step": 11909 + }, + { + "epoch": 5.604705882352941, + "grad_norm": 0.13266059551438603, + "learning_rate": 4.496976441585348e-06, + "loss": 0.0239, + "step": 11910 + }, + { + "epoch": 5.605176470588235, + "grad_norm": 0.004546262287167969, + "learning_rate": 4.486329391327093e-06, + "loss": 0.0001, + "step": 11911 + }, + { + "epoch": 5.605647058823529, + "grad_norm": 4.969363379843467e-06, + "learning_rate": 4.475694768650051e-06, + "loss": 0.0, + "step": 11912 + }, + { + "epoch": 5.606117647058824, + "grad_norm": 2.3658286290207893e-05, + "learning_rate": 4.4650725744624635e-06, + "loss": 0.0, + "step": 11913 + }, + { + "epoch": 5.606588235294118, + "grad_norm": 0.00018594856278748314, + "learning_rate": 4.454462809671538e-06, + "loss": 0.0, + "step": 11914 + }, + { + "epoch": 5.607058823529412, + "grad_norm": 0.018687550996284028, + "learning_rate": 4.443865475183384e-06, + "loss": 0.0007, + "step": 11915 + }, + { + "epoch": 5.607529411764705, + "grad_norm": 0.07643298048824508, + "learning_rate": 4.4332805719030586e-06, + "loss": 0.0142, + "step": 11916 + }, + { + "epoch": 5.608, + "grad_norm": 0.04969921001527494, + "learning_rate": 4.4227081007345565e-06, + "loss": 0.0024, + "step": 11917 + }, + { + "epoch": 5.608470588235294, + "grad_norm": 0.08658201513294846, + "learning_rate": 4.412148062580873e-06, + "loss": 0.0208, + "step": 11918 + }, + { + "epoch": 5.608941176470588, + "grad_norm": 0.10711570819370599, + "learning_rate": 4.401600458343818e-06, + "loss": 0.0125, + "step": 11919 + }, + { + "epoch": 5.609411764705882, + "grad_norm": 4.634375736530531e-05, + "learning_rate": 4.39106528892424e-06, + "loss": 0.0, + "step": 11920 + }, + { + "epoch": 5.6098823529411765, + "grad_norm": 0.07630683131356673, + "learning_rate": 4.380542555221917e-06, + "loss": 0.0142, + "step": 11921 + }, + { + "epoch": 5.610352941176471, + "grad_norm": 0.15428560736534289, + "learning_rate": 4.370032258135514e-06, + "loss": 0.0373, + "step": 11922 + }, + { + "epoch": 5.610823529411765, + "grad_norm": 0.08474939109585264, + "learning_rate": 4.3595343985626476e-06, + "loss": 0.0114, + "step": 11923 + }, + { + "epoch": 5.611294117647059, + "grad_norm": 0.1443070410646398, + "learning_rate": 4.349048977399933e-06, + "loss": 0.0305, + "step": 11924 + }, + { + "epoch": 5.6117647058823525, + "grad_norm": 0.12231121353627755, + "learning_rate": 4.33857599554282e-06, + "loss": 0.0286, + "step": 11925 + }, + { + "epoch": 5.612235294117647, + "grad_norm": 2.5897894083182595e-05, + "learning_rate": 4.328115453885794e-06, + "loss": 0.0, + "step": 11926 + }, + { + "epoch": 5.612705882352941, + "grad_norm": 0.019340672586607273, + "learning_rate": 4.317667353322241e-06, + "loss": 0.0008, + "step": 11927 + }, + { + "epoch": 5.613176470588235, + "grad_norm": 0.13545305602809882, + "learning_rate": 4.307231694744445e-06, + "loss": 0.0353, + "step": 11928 + }, + { + "epoch": 5.613647058823529, + "grad_norm": 0.13139561280464526, + "learning_rate": 4.296808479043678e-06, + "loss": 0.0213, + "step": 11929 + }, + { + "epoch": 5.614117647058824, + "grad_norm": 0.0540387910231098, + "learning_rate": 4.286397707110145e-06, + "loss": 0.0079, + "step": 11930 + }, + { + "epoch": 5.614588235294118, + "grad_norm": 1.397571427870231e-06, + "learning_rate": 4.275999379832967e-06, + "loss": 0.0, + "step": 11931 + }, + { + "epoch": 5.615058823529412, + "grad_norm": 0.03471679235614778, + "learning_rate": 4.265613498100234e-06, + "loss": 0.0022, + "step": 11932 + }, + { + "epoch": 5.615529411764706, + "grad_norm": 0.2047073986630357, + "learning_rate": 4.255240062798904e-06, + "loss": 0.0795, + "step": 11933 + }, + { + "epoch": 5.616, + "grad_norm": 0.07092052273682968, + "learning_rate": 4.24487907481495e-06, + "loss": 0.0095, + "step": 11934 + }, + { + "epoch": 5.616470588235294, + "grad_norm": 0.0888765917452598, + "learning_rate": 4.234530535033265e-06, + "loss": 0.0089, + "step": 11935 + }, + { + "epoch": 5.616941176470588, + "grad_norm": 0.17777425108295397, + "learning_rate": 4.224194444337642e-06, + "loss": 0.0557, + "step": 11936 + }, + { + "epoch": 5.617411764705882, + "grad_norm": 0.19163095679069822, + "learning_rate": 4.213870803610842e-06, + "loss": 0.0348, + "step": 11937 + }, + { + "epoch": 5.6178823529411765, + "grad_norm": 0.15076858089878187, + "learning_rate": 4.203559613734542e-06, + "loss": 0.0381, + "step": 11938 + }, + { + "epoch": 5.618352941176471, + "grad_norm": 2.3757977157078945e-05, + "learning_rate": 4.193260875589371e-06, + "loss": 0.0, + "step": 11939 + }, + { + "epoch": 5.618823529411765, + "grad_norm": 0.13695020423973234, + "learning_rate": 4.1829745900549275e-06, + "loss": 0.0374, + "step": 11940 + }, + { + "epoch": 5.619294117647058, + "grad_norm": 0.014573127721883959, + "learning_rate": 4.1727007580096565e-06, + "loss": 0.0003, + "step": 11941 + }, + { + "epoch": 5.619764705882353, + "grad_norm": 0.11210544188692308, + "learning_rate": 4.162439380331023e-06, + "loss": 0.0095, + "step": 11942 + }, + { + "epoch": 5.620235294117647, + "grad_norm": 0.16470320653180043, + "learning_rate": 4.15219045789541e-06, + "loss": 0.056, + "step": 11943 + }, + { + "epoch": 5.620705882352941, + "grad_norm": 0.1647226487370528, + "learning_rate": 4.141953991578101e-06, + "loss": 0.0525, + "step": 11944 + }, + { + "epoch": 5.621176470588235, + "grad_norm": 0.0705985918500962, + "learning_rate": 4.131729982253345e-06, + "loss": 0.0062, + "step": 11945 + }, + { + "epoch": 5.621647058823529, + "grad_norm": 0.15528915481665564, + "learning_rate": 4.121518430794346e-06, + "loss": 0.0423, + "step": 11946 + }, + { + "epoch": 5.622117647058824, + "grad_norm": 0.0683722727581149, + "learning_rate": 4.111319338073188e-06, + "loss": 0.0085, + "step": 11947 + }, + { + "epoch": 5.622588235294118, + "grad_norm": 0.15199809596753355, + "learning_rate": 4.101132704960941e-06, + "loss": 0.0475, + "step": 11948 + }, + { + "epoch": 5.623058823529412, + "grad_norm": 0.19181480039959944, + "learning_rate": 4.09095853232761e-06, + "loss": 0.0991, + "step": 11949 + }, + { + "epoch": 5.623529411764705, + "grad_norm": 0.02396542734334048, + "learning_rate": 4.080796821042082e-06, + "loss": 0.001, + "step": 11950 + }, + { + "epoch": 5.624, + "grad_norm": 0.1569282605628211, + "learning_rate": 4.07064757197223e-06, + "loss": 0.0467, + "step": 11951 + }, + { + "epoch": 5.624470588235294, + "grad_norm": 0.048157663674301784, + "learning_rate": 4.0605107859848604e-06, + "loss": 0.0047, + "step": 11952 + }, + { + "epoch": 5.624941176470588, + "grad_norm": 0.1990476440074421, + "learning_rate": 4.050386463945699e-06, + "loss": 0.0687, + "step": 11953 + }, + { + "epoch": 5.625411764705882, + "grad_norm": 0.10182511620541732, + "learning_rate": 4.040274606719418e-06, + "loss": 0.0118, + "step": 11954 + }, + { + "epoch": 5.625882352941177, + "grad_norm": 0.13697540887101023, + "learning_rate": 4.030175215169612e-06, + "loss": 0.0387, + "step": 11955 + }, + { + "epoch": 5.626352941176471, + "grad_norm": 0.17289430271407374, + "learning_rate": 4.020088290158807e-06, + "loss": 0.0544, + "step": 11956 + }, + { + "epoch": 5.626823529411765, + "grad_norm": 0.0714420066645704, + "learning_rate": 4.010013832548514e-06, + "loss": 0.0054, + "step": 11957 + }, + { + "epoch": 5.627294117647059, + "grad_norm": 0.14728816517475474, + "learning_rate": 3.999951843199095e-06, + "loss": 0.0435, + "step": 11958 + }, + { + "epoch": 5.627764705882353, + "grad_norm": 0.10602966701101488, + "learning_rate": 3.989902322969929e-06, + "loss": 0.0217, + "step": 11959 + }, + { + "epoch": 5.628235294117647, + "grad_norm": 0.01843994005287432, + "learning_rate": 3.979865272719279e-06, + "loss": 0.0006, + "step": 11960 + }, + { + "epoch": 5.628705882352941, + "grad_norm": 4.1514027928777325e-06, + "learning_rate": 3.969840693304344e-06, + "loss": 0.0, + "step": 11961 + }, + { + "epoch": 5.629176470588235, + "grad_norm": 0.14071465965800947, + "learning_rate": 3.959828585581287e-06, + "loss": 0.0426, + "step": 11962 + }, + { + "epoch": 5.6296470588235294, + "grad_norm": 0.17598290346226747, + "learning_rate": 3.949828950405226e-06, + "loss": 0.0566, + "step": 11963 + }, + { + "epoch": 5.630117647058824, + "grad_norm": 0.0636176498936461, + "learning_rate": 3.939841788630111e-06, + "loss": 0.009, + "step": 11964 + }, + { + "epoch": 5.630588235294118, + "grad_norm": 0.09962347985815788, + "learning_rate": 3.9298671011089245e-06, + "loss": 0.0215, + "step": 11965 + }, + { + "epoch": 5.631058823529412, + "grad_norm": 1.0094270376494489e-05, + "learning_rate": 3.9199048886935865e-06, + "loss": 0.0, + "step": 11966 + }, + { + "epoch": 5.631529411764706, + "grad_norm": 0.10428195570282414, + "learning_rate": 3.909955152234867e-06, + "loss": 0.0252, + "step": 11967 + }, + { + "epoch": 5.632, + "grad_norm": 0.06897090927903035, + "learning_rate": 3.900017892582552e-06, + "loss": 0.0081, + "step": 11968 + }, + { + "epoch": 5.632470588235294, + "grad_norm": 0.08984404475346797, + "learning_rate": 3.8900931105853294e-06, + "loss": 0.0101, + "step": 11969 + }, + { + "epoch": 5.632941176470588, + "grad_norm": 0.12592133173736902, + "learning_rate": 3.8801808070908215e-06, + "loss": 0.0398, + "step": 11970 + }, + { + "epoch": 5.633411764705882, + "grad_norm": 0.13584691637803786, + "learning_rate": 3.870280982945567e-06, + "loss": 0.0325, + "step": 11971 + }, + { + "epoch": 5.633882352941177, + "grad_norm": 0.16232301353415704, + "learning_rate": 3.860393638995107e-06, + "loss": 0.0494, + "step": 11972 + }, + { + "epoch": 5.634352941176471, + "grad_norm": 0.1015678016337652, + "learning_rate": 3.850518776083817e-06, + "loss": 0.0167, + "step": 11973 + }, + { + "epoch": 5.634823529411765, + "grad_norm": 0.05779639554019712, + "learning_rate": 3.840656395055103e-06, + "loss": 0.0066, + "step": 11974 + }, + { + "epoch": 5.635294117647058, + "grad_norm": 5.114813237489257e-06, + "learning_rate": 3.8308064967512095e-06, + "loss": 0.0, + "step": 11975 + }, + { + "epoch": 5.635764705882353, + "grad_norm": 0.07812725442478873, + "learning_rate": 3.820969082013414e-06, + "loss": 0.0144, + "step": 11976 + }, + { + "epoch": 5.636235294117647, + "grad_norm": 0.06033712269232138, + "learning_rate": 3.8111441516818776e-06, + "loss": 0.0061, + "step": 11977 + }, + { + "epoch": 5.636705882352941, + "grad_norm": 0.08653817252700542, + "learning_rate": 3.8013317065956784e-06, + "loss": 0.0092, + "step": 11978 + }, + { + "epoch": 5.637176470588235, + "grad_norm": 2.2096063461911236e-05, + "learning_rate": 3.7915317475928296e-06, + "loss": 0.0, + "step": 11979 + }, + { + "epoch": 5.6376470588235295, + "grad_norm": 0.1657247301665719, + "learning_rate": 3.7817442755103457e-06, + "loss": 0.047, + "step": 11980 + }, + { + "epoch": 5.638117647058824, + "grad_norm": 0.19113239213955638, + "learning_rate": 3.7719692911840905e-06, + "loss": 0.0266, + "step": 11981 + }, + { + "epoch": 5.638588235294118, + "grad_norm": 0.08697400078646324, + "learning_rate": 3.76220679544888e-06, + "loss": 0.0206, + "step": 11982 + }, + { + "epoch": 5.639058823529412, + "grad_norm": 0.050956709743134265, + "learning_rate": 3.7524567891385305e-06, + "loss": 0.0061, + "step": 11983 + }, + { + "epoch": 5.6395294117647055, + "grad_norm": 0.10876365111171271, + "learning_rate": 3.7427192730856927e-06, + "loss": 0.0304, + "step": 11984 + }, + { + "epoch": 5.64, + "grad_norm": 0.1957526048484072, + "learning_rate": 3.7329942481220176e-06, + "loss": 0.0613, + "step": 11985 + }, + { + "epoch": 5.640470588235294, + "grad_norm": 0.16076363908418842, + "learning_rate": 3.723281715078091e-06, + "loss": 0.0402, + "step": 11986 + }, + { + "epoch": 5.640941176470588, + "grad_norm": 0.08021917290905459, + "learning_rate": 3.7135816747833667e-06, + "loss": 0.0116, + "step": 11987 + }, + { + "epoch": 5.641411764705882, + "grad_norm": 0.16145311223158867, + "learning_rate": 3.7038941280663147e-06, + "loss": 0.0168, + "step": 11988 + }, + { + "epoch": 5.641882352941177, + "grad_norm": 0.10095275041558725, + "learning_rate": 3.6942190757542735e-06, + "loss": 0.0202, + "step": 11989 + }, + { + "epoch": 5.642352941176471, + "grad_norm": 0.021215558976415222, + "learning_rate": 3.684556518673548e-06, + "loss": 0.001, + "step": 11990 + }, + { + "epoch": 5.642823529411765, + "grad_norm": 0.06645135982724899, + "learning_rate": 3.6749064576493793e-06, + "loss": 0.01, + "step": 11991 + }, + { + "epoch": 5.643294117647059, + "grad_norm": 0.08319092310287153, + "learning_rate": 3.6652688935058904e-06, + "loss": 0.014, + "step": 11992 + }, + { + "epoch": 5.643764705882353, + "grad_norm": 0.13260344795165577, + "learning_rate": 3.6556438270662236e-06, + "loss": 0.031, + "step": 11993 + }, + { + "epoch": 5.644235294117647, + "grad_norm": 0.00018451991470538735, + "learning_rate": 3.6460312591524044e-06, + "loss": 0.0, + "step": 11994 + }, + { + "epoch": 5.644705882352941, + "grad_norm": 0.1173434179217957, + "learning_rate": 3.6364311905853594e-06, + "loss": 0.0213, + "step": 11995 + }, + { + "epoch": 5.645176470588235, + "grad_norm": 0.08513526941269103, + "learning_rate": 3.6268436221849994e-06, + "loss": 0.0144, + "step": 11996 + }, + { + "epoch": 5.6456470588235295, + "grad_norm": 0.12443030496419773, + "learning_rate": 3.6172685547701695e-06, + "loss": 0.0451, + "step": 11997 + }, + { + "epoch": 5.646117647058824, + "grad_norm": 0.3148626642343323, + "learning_rate": 3.6077059891585826e-06, + "loss": 0.0656, + "step": 11998 + }, + { + "epoch": 5.646588235294118, + "grad_norm": 0.07818907805016279, + "learning_rate": 3.598155926166968e-06, + "loss": 0.0123, + "step": 11999 + }, + { + "epoch": 5.647058823529412, + "grad_norm": 0.14420985676910944, + "learning_rate": 3.5886183666109405e-06, + "loss": 0.0443, + "step": 12000 + }, + { + "epoch": 5.647529411764706, + "grad_norm": 0.00010065600483459772, + "learning_rate": 3.579093311305031e-06, + "loss": 0.0, + "step": 12001 + }, + { + "epoch": 5.648, + "grad_norm": 0.00010478733755233984, + "learning_rate": 3.5695807610627402e-06, + "loss": 0.0, + "step": 12002 + }, + { + "epoch": 5.648470588235294, + "grad_norm": 1.5863261299835932e-05, + "learning_rate": 3.5600807166965173e-06, + "loss": 0.0, + "step": 12003 + }, + { + "epoch": 5.648941176470588, + "grad_norm": 0.1241137016501488, + "learning_rate": 3.5505931790176635e-06, + "loss": 0.0216, + "step": 12004 + }, + { + "epoch": 5.649411764705882, + "grad_norm": 0.07332342669909565, + "learning_rate": 3.5411181488364802e-06, + "loss": 0.0125, + "step": 12005 + }, + { + "epoch": 5.649882352941177, + "grad_norm": 0.05693117339927355, + "learning_rate": 3.531655626962204e-06, + "loss": 0.0027, + "step": 12006 + }, + { + "epoch": 5.650352941176471, + "grad_norm": 0.12590030949620637, + "learning_rate": 3.5222056142029387e-06, + "loss": 0.0308, + "step": 12007 + }, + { + "epoch": 5.650823529411765, + "grad_norm": 0.03523887352656832, + "learning_rate": 3.5127681113658044e-06, + "loss": 0.0041, + "step": 12008 + }, + { + "epoch": 5.651294117647058, + "grad_norm": 0.1497963479399851, + "learning_rate": 3.503343119256774e-06, + "loss": 0.0497, + "step": 12009 + }, + { + "epoch": 5.651764705882353, + "grad_norm": 0.06389134085130138, + "learning_rate": 3.493930638680803e-06, + "loss": 0.0089, + "step": 12010 + }, + { + "epoch": 5.652235294117647, + "grad_norm": 0.13992525184413074, + "learning_rate": 3.484530670441782e-06, + "loss": 0.0299, + "step": 12011 + }, + { + "epoch": 5.652705882352941, + "grad_norm": 2.7225515185986097e-05, + "learning_rate": 3.4751432153424687e-06, + "loss": 0.0, + "step": 12012 + }, + { + "epoch": 5.653176470588235, + "grad_norm": 0.12194936895502394, + "learning_rate": 3.465768274184638e-06, + "loss": 0.0342, + "step": 12013 + }, + { + "epoch": 5.6536470588235295, + "grad_norm": 0.05280249747035291, + "learning_rate": 3.4564058477689503e-06, + "loss": 0.0053, + "step": 12014 + }, + { + "epoch": 5.654117647058824, + "grad_norm": 0.16863255894354193, + "learning_rate": 3.4470559368949825e-06, + "loss": 0.0594, + "step": 12015 + }, + { + "epoch": 5.654588235294118, + "grad_norm": 0.03581386440701323, + "learning_rate": 3.4377185423612785e-06, + "loss": 0.0039, + "step": 12016 + }, + { + "epoch": 5.655058823529412, + "grad_norm": 0.08224736705504508, + "learning_rate": 3.428393664965301e-06, + "loss": 0.0168, + "step": 12017 + }, + { + "epoch": 5.6555294117647055, + "grad_norm": 0.16961807116871114, + "learning_rate": 3.4190813055034127e-06, + "loss": 0.0253, + "step": 12018 + }, + { + "epoch": 5.656, + "grad_norm": 0.05058944792731109, + "learning_rate": 3.4097814647709775e-06, + "loss": 0.0024, + "step": 12019 + }, + { + "epoch": 5.656470588235294, + "grad_norm": 0.08920776768023161, + "learning_rate": 3.400494143562227e-06, + "loss": 0.0087, + "step": 12020 + }, + { + "epoch": 5.656941176470588, + "grad_norm": 0.11258927328896892, + "learning_rate": 3.391219342670326e-06, + "loss": 0.0115, + "step": 12021 + }, + { + "epoch": 5.657411764705882, + "grad_norm": 0.12484953891631269, + "learning_rate": 3.3819570628874414e-06, + "loss": 0.0243, + "step": 12022 + }, + { + "epoch": 5.657882352941177, + "grad_norm": 0.09540598293068452, + "learning_rate": 3.3727073050045406e-06, + "loss": 0.0119, + "step": 12023 + }, + { + "epoch": 5.658352941176471, + "grad_norm": 0.11879409032044214, + "learning_rate": 3.3634700698116577e-06, + "loss": 0.0227, + "step": 12024 + }, + { + "epoch": 5.658823529411765, + "grad_norm": 0.09830718297115106, + "learning_rate": 3.354245358097679e-06, + "loss": 0.0114, + "step": 12025 + }, + { + "epoch": 5.659294117647059, + "grad_norm": 0.03465196771987324, + "learning_rate": 3.345033170650441e-06, + "loss": 0.0021, + "step": 12026 + }, + { + "epoch": 5.659764705882353, + "grad_norm": 0.04293350806084995, + "learning_rate": 3.3358335082566966e-06, + "loss": 0.0019, + "step": 12027 + }, + { + "epoch": 5.660235294117647, + "grad_norm": 0.08578976348400105, + "learning_rate": 3.326646371702185e-06, + "loss": 0.0174, + "step": 12028 + }, + { + "epoch": 5.660705882352941, + "grad_norm": 0.07882153498191867, + "learning_rate": 3.317471761771462e-06, + "loss": 0.0187, + "step": 12029 + }, + { + "epoch": 5.661176470588235, + "grad_norm": 0.035689568541031426, + "learning_rate": 3.30830967924815e-06, + "loss": 0.0032, + "step": 12030 + }, + { + "epoch": 5.6616470588235295, + "grad_norm": 5.788485529521528e-06, + "learning_rate": 3.299160124914707e-06, + "loss": 0.0, + "step": 12031 + }, + { + "epoch": 5.662117647058824, + "grad_norm": 0.15005782793181374, + "learning_rate": 3.290023099552541e-06, + "loss": 0.0461, + "step": 12032 + }, + { + "epoch": 5.662588235294118, + "grad_norm": 0.10764701534277804, + "learning_rate": 3.2808986039420115e-06, + "loss": 0.0101, + "step": 12033 + }, + { + "epoch": 5.663058823529412, + "grad_norm": 0.1167302468318489, + "learning_rate": 3.2717866388624114e-06, + "loss": 0.0226, + "step": 12034 + }, + { + "epoch": 5.663529411764706, + "grad_norm": 0.14815986229174785, + "learning_rate": 3.2626872050919185e-06, + "loss": 0.042, + "step": 12035 + }, + { + "epoch": 5.664, + "grad_norm": 0.11969644050943322, + "learning_rate": 3.2536003034076774e-06, + "loss": 0.0193, + "step": 12036 + }, + { + "epoch": 5.664470588235294, + "grad_norm": 1.0310499135237527e-05, + "learning_rate": 3.2445259345857676e-06, + "loss": 0.0, + "step": 12037 + }, + { + "epoch": 5.664941176470588, + "grad_norm": 0.00010924616285162985, + "learning_rate": 3.2354640994011693e-06, + "loss": 0.0, + "step": 12038 + }, + { + "epoch": 5.665411764705882, + "grad_norm": 0.04147460615444237, + "learning_rate": 3.22641479862783e-06, + "loss": 0.0023, + "step": 12039 + }, + { + "epoch": 5.665882352941177, + "grad_norm": 0.02478242976799133, + "learning_rate": 3.2173780330385647e-06, + "loss": 0.0009, + "step": 12040 + }, + { + "epoch": 5.666352941176471, + "grad_norm": 0.12205662137535617, + "learning_rate": 3.2083538034051726e-06, + "loss": 0.0138, + "step": 12041 + }, + { + "epoch": 5.666823529411765, + "grad_norm": 0.06385795387975436, + "learning_rate": 3.199342110498404e-06, + "loss": 0.0105, + "step": 12042 + }, + { + "epoch": 5.667294117647058, + "grad_norm": 0.0048879766163057, + "learning_rate": 3.1903429550878435e-06, + "loss": 0.0001, + "step": 12043 + }, + { + "epoch": 5.667764705882353, + "grad_norm": 0.054528791388959524, + "learning_rate": 3.181356337942109e-06, + "loss": 0.0052, + "step": 12044 + }, + { + "epoch": 5.668235294117647, + "grad_norm": 0.09871240637715324, + "learning_rate": 3.1723822598286875e-06, + "loss": 0.0136, + "step": 12045 + }, + { + "epoch": 5.668705882352941, + "grad_norm": 0.04198186825950555, + "learning_rate": 3.163420721513982e-06, + "loss": 0.0027, + "step": 12046 + }, + { + "epoch": 5.669176470588235, + "grad_norm": 0.058324969950650375, + "learning_rate": 3.1544717237633965e-06, + "loss": 0.0089, + "step": 12047 + }, + { + "epoch": 5.6696470588235295, + "grad_norm": 0.10402517219691106, + "learning_rate": 3.145535267341204e-06, + "loss": 0.018, + "step": 12048 + }, + { + "epoch": 5.670117647058824, + "grad_norm": 0.00010013323012719521, + "learning_rate": 3.1366113530106097e-06, + "loss": 0.0, + "step": 12049 + }, + { + "epoch": 5.670588235294118, + "grad_norm": 0.14061645178893867, + "learning_rate": 3.1276999815337544e-06, + "loss": 0.0247, + "step": 12050 + }, + { + "epoch": 5.671058823529412, + "grad_norm": 0.1299328260252708, + "learning_rate": 3.118801153671746e-06, + "loss": 0.0263, + "step": 12051 + }, + { + "epoch": 5.6715294117647055, + "grad_norm": 0.013164343234173342, + "learning_rate": 3.1099148701845435e-06, + "loss": 0.0004, + "step": 12052 + }, + { + "epoch": 5.672, + "grad_norm": 1.1556142462941001e-05, + "learning_rate": 3.1010411318311226e-06, + "loss": 0.0, + "step": 12053 + }, + { + "epoch": 5.672470588235294, + "grad_norm": 0.08235310412823428, + "learning_rate": 3.092179939369327e-06, + "loss": 0.0116, + "step": 12054 + }, + { + "epoch": 5.672941176470588, + "grad_norm": 0.09035159223513876, + "learning_rate": 3.083331293555935e-06, + "loss": 0.0093, + "step": 12055 + }, + { + "epoch": 5.673411764705882, + "grad_norm": 0.08267882387694055, + "learning_rate": 3.074495195146659e-06, + "loss": 0.0125, + "step": 12056 + }, + { + "epoch": 5.673882352941177, + "grad_norm": 1.6480387054140523e-05, + "learning_rate": 3.0656716448961784e-06, + "loss": 0.0, + "step": 12057 + }, + { + "epoch": 5.674352941176471, + "grad_norm": 0.0953283836691165, + "learning_rate": 3.05686064355804e-06, + "loss": 0.0133, + "step": 12058 + }, + { + "epoch": 5.674823529411765, + "grad_norm": 0.11162517234937569, + "learning_rate": 3.048062191884759e-06, + "loss": 0.029, + "step": 12059 + }, + { + "epoch": 5.675294117647059, + "grad_norm": 0.066775207164431, + "learning_rate": 3.039276290627768e-06, + "loss": 0.0086, + "step": 12060 + }, + { + "epoch": 5.675764705882353, + "grad_norm": 0.14384153941769862, + "learning_rate": 3.0305029405373993e-06, + "loss": 0.0431, + "step": 12061 + }, + { + "epoch": 5.676235294117647, + "grad_norm": 0.14427889661262458, + "learning_rate": 3.0217421423629706e-06, + "loss": 0.0474, + "step": 12062 + }, + { + "epoch": 5.676705882352941, + "grad_norm": 0.025155316142337625, + "learning_rate": 3.0129938968526836e-06, + "loss": 0.0014, + "step": 12063 + }, + { + "epoch": 5.677176470588235, + "grad_norm": 6.422948578158027e-06, + "learning_rate": 3.00425820475369e-06, + "loss": 0.0, + "step": 12064 + }, + { + "epoch": 5.6776470588235295, + "grad_norm": 0.07192154219212657, + "learning_rate": 2.9955350668120605e-06, + "loss": 0.0105, + "step": 12065 + }, + { + "epoch": 5.678117647058824, + "grad_norm": 0.029337667473462944, + "learning_rate": 2.9868244837727815e-06, + "loss": 0.0016, + "step": 12066 + }, + { + "epoch": 5.678588235294118, + "grad_norm": 3.8893501163069955e-06, + "learning_rate": 2.978126456379809e-06, + "loss": 0.0, + "step": 12067 + }, + { + "epoch": 5.679058823529411, + "grad_norm": 0.027732944026916374, + "learning_rate": 2.969440985375948e-06, + "loss": 0.0017, + "step": 12068 + }, + { + "epoch": 5.679529411764706, + "grad_norm": 0.14308224959760057, + "learning_rate": 2.960768071503006e-06, + "loss": 0.0383, + "step": 12069 + }, + { + "epoch": 5.68, + "grad_norm": 9.312502626725866e-06, + "learning_rate": 2.9521077155017224e-06, + "loss": 0.0, + "step": 12070 + }, + { + "epoch": 5.680470588235294, + "grad_norm": 0.15893537759634796, + "learning_rate": 2.9434599181116736e-06, + "loss": 0.0371, + "step": 12071 + }, + { + "epoch": 5.680941176470588, + "grad_norm": 0.1411004837445591, + "learning_rate": 2.934824680071468e-06, + "loss": 0.0376, + "step": 12072 + }, + { + "epoch": 5.681411764705882, + "grad_norm": 0.14198827381545592, + "learning_rate": 2.9262020021185993e-06, + "loss": 0.0322, + "step": 12073 + }, + { + "epoch": 5.681882352941177, + "grad_norm": 0.0840752478299307, + "learning_rate": 2.917591884989462e-06, + "loss": 0.0082, + "step": 12074 + }, + { + "epoch": 5.682352941176471, + "grad_norm": 4.871444144554615e-05, + "learning_rate": 2.908994329419401e-06, + "loss": 0.0, + "step": 12075 + }, + { + "epoch": 5.682823529411765, + "grad_norm": 0.09511940312470897, + "learning_rate": 2.9004093361427282e-06, + "loss": 0.0139, + "step": 12076 + }, + { + "epoch": 5.683294117647058, + "grad_norm": 2.3612389558484998e-05, + "learning_rate": 2.891836905892592e-06, + "loss": 0.0, + "step": 12077 + }, + { + "epoch": 5.683764705882353, + "grad_norm": 0.06471433594656816, + "learning_rate": 2.883277039401155e-06, + "loss": 0.0066, + "step": 12078 + }, + { + "epoch": 5.684235294117647, + "grad_norm": 2.3918280797227335e-05, + "learning_rate": 2.874729737399467e-06, + "loss": 0.0, + "step": 12079 + }, + { + "epoch": 5.684705882352941, + "grad_norm": 0.06334217563003276, + "learning_rate": 2.866195000617494e-06, + "loss": 0.0067, + "step": 12080 + }, + { + "epoch": 5.685176470588235, + "grad_norm": 0.13056072456447576, + "learning_rate": 2.857672829784169e-06, + "loss": 0.0432, + "step": 12081 + }, + { + "epoch": 5.6856470588235295, + "grad_norm": 0.07002334234441109, + "learning_rate": 2.84916322562731e-06, + "loss": 0.0057, + "step": 12082 + }, + { + "epoch": 5.686117647058824, + "grad_norm": 0.10843484487137688, + "learning_rate": 2.840666188873669e-06, + "loss": 0.0249, + "step": 12083 + }, + { + "epoch": 5.686588235294118, + "grad_norm": 0.04033121227142799, + "learning_rate": 2.8321817202489493e-06, + "loss": 0.0027, + "step": 12084 + }, + { + "epoch": 5.687058823529412, + "grad_norm": 0.13290329261372805, + "learning_rate": 2.8237098204777874e-06, + "loss": 0.0195, + "step": 12085 + }, + { + "epoch": 5.6875294117647055, + "grad_norm": 0.12570098538162278, + "learning_rate": 2.815250490283688e-06, + "loss": 0.0302, + "step": 12086 + }, + { + "epoch": 5.688, + "grad_norm": 4.056326230622208e-06, + "learning_rate": 2.8068037303891233e-06, + "loss": 0.0, + "step": 12087 + }, + { + "epoch": 5.688470588235294, + "grad_norm": 0.09333851528379887, + "learning_rate": 2.7983695415155162e-06, + "loss": 0.0103, + "step": 12088 + }, + { + "epoch": 5.688941176470588, + "grad_norm": 0.018582202038870024, + "learning_rate": 2.789947924383157e-06, + "loss": 0.0012, + "step": 12089 + }, + { + "epoch": 5.689411764705882, + "grad_norm": 0.14572359877102645, + "learning_rate": 2.78153887971132e-06, + "loss": 0.0411, + "step": 12090 + }, + { + "epoch": 5.689882352941177, + "grad_norm": 6.7777275634156e-05, + "learning_rate": 2.773142408218165e-06, + "loss": 0.0, + "step": 12091 + }, + { + "epoch": 5.690352941176471, + "grad_norm": 0.0956553112816854, + "learning_rate": 2.7647585106207682e-06, + "loss": 0.0089, + "step": 12092 + }, + { + "epoch": 5.690823529411765, + "grad_norm": 0.14268247518597385, + "learning_rate": 2.7563871876352063e-06, + "loss": 0.0285, + "step": 12093 + }, + { + "epoch": 5.691294117647059, + "grad_norm": 0.1527538774095619, + "learning_rate": 2.7480284399763907e-06, + "loss": 0.0282, + "step": 12094 + }, + { + "epoch": 5.691764705882353, + "grad_norm": 0.142311663973379, + "learning_rate": 2.7396822683582177e-06, + "loss": 0.0416, + "step": 12095 + }, + { + "epoch": 5.692235294117647, + "grad_norm": 0.07868455260755322, + "learning_rate": 2.731348673493483e-06, + "loss": 0.0109, + "step": 12096 + }, + { + "epoch": 5.692705882352941, + "grad_norm": 0.06376978370262133, + "learning_rate": 2.723027656093918e-06, + "loss": 0.0076, + "step": 12097 + }, + { + "epoch": 5.693176470588235, + "grad_norm": 0.1553562886617033, + "learning_rate": 2.714719216870187e-06, + "loss": 0.0414, + "step": 12098 + }, + { + "epoch": 5.6936470588235295, + "grad_norm": 3.069105076032387e-05, + "learning_rate": 2.706423356531873e-06, + "loss": 0.0, + "step": 12099 + }, + { + "epoch": 5.694117647058824, + "grad_norm": 0.11362184950532746, + "learning_rate": 2.6981400757874584e-06, + "loss": 0.0134, + "step": 12100 + }, + { + "epoch": 5.694588235294118, + "grad_norm": 0.08767819927242475, + "learning_rate": 2.6898693753443945e-06, + "loss": 0.0169, + "step": 12101 + }, + { + "epoch": 5.695058823529411, + "grad_norm": 0.18207047734040346, + "learning_rate": 2.68161125590905e-06, + "loss": 0.0519, + "step": 12102 + }, + { + "epoch": 5.6955294117647055, + "grad_norm": 0.057151735113326434, + "learning_rate": 2.673365718186693e-06, + "loss": 0.0023, + "step": 12103 + }, + { + "epoch": 5.696, + "grad_norm": 0.05589984253478728, + "learning_rate": 2.6651327628815277e-06, + "loss": 0.0061, + "step": 12104 + }, + { + "epoch": 5.696470588235294, + "grad_norm": 0.02888514002627308, + "learning_rate": 2.656912390696708e-06, + "loss": 0.002, + "step": 12105 + }, + { + "epoch": 5.696941176470588, + "grad_norm": 0.10578885632891995, + "learning_rate": 2.6487046023342563e-06, + "loss": 0.0131, + "step": 12106 + }, + { + "epoch": 5.697411764705882, + "grad_norm": 0.11643169317588518, + "learning_rate": 2.6405093984952275e-06, + "loss": 0.0189, + "step": 12107 + }, + { + "epoch": 5.697882352941177, + "grad_norm": 0.11983367781689801, + "learning_rate": 2.6323267798794625e-06, + "loss": 0.0193, + "step": 12108 + }, + { + "epoch": 5.698352941176471, + "grad_norm": 0.21310372539682002, + "learning_rate": 2.624156747185835e-06, + "loss": 0.0563, + "step": 12109 + }, + { + "epoch": 5.698823529411765, + "grad_norm": 1.6320865062547687e-05, + "learning_rate": 2.615999301112104e-06, + "loss": 0.0, + "step": 12110 + }, + { + "epoch": 5.699294117647058, + "grad_norm": 0.0672304682139431, + "learning_rate": 2.6078544423549286e-06, + "loss": 0.0047, + "step": 12111 + }, + { + "epoch": 5.699764705882353, + "grad_norm": 4.198503395011646e-05, + "learning_rate": 2.599722171609936e-06, + "loss": 0.0, + "step": 12112 + }, + { + "epoch": 5.700235294117647, + "grad_norm": 0.11291317480077559, + "learning_rate": 2.591602489571687e-06, + "loss": 0.0197, + "step": 12113 + }, + { + "epoch": 5.700705882352941, + "grad_norm": 0.05653146892174251, + "learning_rate": 2.5834953969336105e-06, + "loss": 0.0086, + "step": 12114 + }, + { + "epoch": 5.701176470588235, + "grad_norm": 0.15131128492157508, + "learning_rate": 2.5754008943880855e-06, + "loss": 0.0504, + "step": 12115 + }, + { + "epoch": 5.7016470588235295, + "grad_norm": 0.12450318915131577, + "learning_rate": 2.5673189826264596e-06, + "loss": 0.0212, + "step": 12116 + }, + { + "epoch": 5.702117647058824, + "grad_norm": 0.00011109598367534686, + "learning_rate": 2.559249662338947e-06, + "loss": 0.0, + "step": 12117 + }, + { + "epoch": 5.702588235294118, + "grad_norm": 0.1344597810318018, + "learning_rate": 2.5511929342146964e-06, + "loss": 0.0376, + "step": 12118 + }, + { + "epoch": 5.703058823529412, + "grad_norm": 0.10369028368309736, + "learning_rate": 2.543148798941791e-06, + "loss": 0.011, + "step": 12119 + }, + { + "epoch": 5.7035294117647055, + "grad_norm": 0.12782965898395166, + "learning_rate": 2.535117257207264e-06, + "loss": 0.0273, + "step": 12120 + }, + { + "epoch": 5.704, + "grad_norm": 2.8596991856870553e-06, + "learning_rate": 2.5270983096970344e-06, + "loss": 0.0, + "step": 12121 + }, + { + "epoch": 5.704470588235294, + "grad_norm": 0.10967626061713732, + "learning_rate": 2.5190919570959533e-06, + "loss": 0.0185, + "step": 12122 + }, + { + "epoch": 5.704941176470588, + "grad_norm": 0.1376018063767929, + "learning_rate": 2.511098200087808e-06, + "loss": 0.0246, + "step": 12123 + }, + { + "epoch": 5.705411764705882, + "grad_norm": 1.4260823189222033e-05, + "learning_rate": 2.503117039355318e-06, + "loss": 0.0, + "step": 12124 + }, + { + "epoch": 5.705882352941177, + "grad_norm": 1.1477076376392654e-05, + "learning_rate": 2.4951484755800886e-06, + "loss": 0.0, + "step": 12125 + }, + { + "epoch": 5.706352941176471, + "grad_norm": 0.048132120171444595, + "learning_rate": 2.487192509442676e-06, + "loss": 0.0042, + "step": 12126 + }, + { + "epoch": 5.706823529411765, + "grad_norm": 0.05845560297330665, + "learning_rate": 2.4792491416226024e-06, + "loss": 0.0029, + "step": 12127 + }, + { + "epoch": 5.707294117647059, + "grad_norm": 0.15128988839415988, + "learning_rate": 2.471318372798209e-06, + "loss": 0.021, + "step": 12128 + }, + { + "epoch": 5.707764705882353, + "grad_norm": 0.06649776728186302, + "learning_rate": 2.4634002036468536e-06, + "loss": 0.0074, + "step": 12129 + }, + { + "epoch": 5.708235294117647, + "grad_norm": 5.318330032015454e-06, + "learning_rate": 2.4554946348447958e-06, + "loss": 0.0, + "step": 12130 + }, + { + "epoch": 5.708705882352941, + "grad_norm": 0.10192981459250808, + "learning_rate": 2.447601667067195e-06, + "loss": 0.0217, + "step": 12131 + }, + { + "epoch": 5.709176470588235, + "grad_norm": 0.055295712348036784, + "learning_rate": 2.4397213009881456e-06, + "loss": 0.0053, + "step": 12132 + }, + { + "epoch": 5.7096470588235295, + "grad_norm": 0.1575553687321482, + "learning_rate": 2.4318535372806757e-06, + "loss": 0.0401, + "step": 12133 + }, + { + "epoch": 5.710117647058824, + "grad_norm": 0.0721333216534127, + "learning_rate": 2.4239983766167315e-06, + "loss": 0.0105, + "step": 12134 + }, + { + "epoch": 5.710588235294118, + "grad_norm": 0.07081057969897876, + "learning_rate": 2.416155819667193e-06, + "loss": 0.0067, + "step": 12135 + }, + { + "epoch": 5.711058823529411, + "grad_norm": 0.13822813302847622, + "learning_rate": 2.4083258671018403e-06, + "loss": 0.0233, + "step": 12136 + }, + { + "epoch": 5.7115294117647055, + "grad_norm": 0.055002407156041044, + "learning_rate": 2.4005085195893725e-06, + "loss": 0.0073, + "step": 12137 + }, + { + "epoch": 5.712, + "grad_norm": 0.0668603967621147, + "learning_rate": 2.3927037777974722e-06, + "loss": 0.0079, + "step": 12138 + }, + { + "epoch": 5.712470588235294, + "grad_norm": 0.10641307310318604, + "learning_rate": 2.384911642392673e-06, + "loss": 0.0303, + "step": 12139 + }, + { + "epoch": 5.712941176470588, + "grad_norm": 0.08710660338824838, + "learning_rate": 2.377132114040459e-06, + "loss": 0.0146, + "step": 12140 + }, + { + "epoch": 5.713411764705882, + "grad_norm": 0.06762327725545346, + "learning_rate": 2.369365193405248e-06, + "loss": 0.006, + "step": 12141 + }, + { + "epoch": 5.713882352941177, + "grad_norm": 8.233117241060551e-06, + "learning_rate": 2.3616108811503775e-06, + "loss": 0.0, + "step": 12142 + }, + { + "epoch": 5.714352941176471, + "grad_norm": 0.15624311966270604, + "learning_rate": 2.3538691779381003e-06, + "loss": 0.0556, + "step": 12143 + }, + { + "epoch": 5.714823529411765, + "grad_norm": 0.1708278254554651, + "learning_rate": 2.346140084429604e-06, + "loss": 0.0102, + "step": 12144 + }, + { + "epoch": 5.715294117647058, + "grad_norm": 0.014928161269987621, + "learning_rate": 2.3384236012849778e-06, + "loss": 0.0007, + "step": 12145 + }, + { + "epoch": 5.715764705882353, + "grad_norm": 0.055616491362130094, + "learning_rate": 2.330719729163244e-06, + "loss": 0.0066, + "step": 12146 + }, + { + "epoch": 5.716235294117647, + "grad_norm": 0.03555609473869061, + "learning_rate": 2.32302846872236e-06, + "loss": 0.0027, + "step": 12147 + }, + { + "epoch": 5.716705882352941, + "grad_norm": 0.08384406035193565, + "learning_rate": 2.3153498206192e-06, + "loss": 0.0117, + "step": 12148 + }, + { + "epoch": 5.717176470588235, + "grad_norm": 0.12470583169696431, + "learning_rate": 2.3076837855095563e-06, + "loss": 0.0271, + "step": 12149 + }, + { + "epoch": 5.7176470588235295, + "grad_norm": 0.17310085334481073, + "learning_rate": 2.3000303640481386e-06, + "loss": 0.0377, + "step": 12150 + }, + { + "epoch": 5.718117647058824, + "grad_norm": 0.07088910524151577, + "learning_rate": 2.292389556888591e-06, + "loss": 0.0076, + "step": 12151 + }, + { + "epoch": 5.718588235294118, + "grad_norm": 0.10885069052846644, + "learning_rate": 2.2847613646834905e-06, + "loss": 0.025, + "step": 12152 + }, + { + "epoch": 5.719058823529412, + "grad_norm": 0.07974729025818651, + "learning_rate": 2.2771457880842834e-06, + "loss": 0.0084, + "step": 12153 + }, + { + "epoch": 5.7195294117647055, + "grad_norm": 0.03248292945590198, + "learning_rate": 2.2695428277414165e-06, + "loss": 0.0015, + "step": 12154 + }, + { + "epoch": 5.72, + "grad_norm": 7.063294423950523e-05, + "learning_rate": 2.261952484304219e-06, + "loss": 0.0, + "step": 12155 + }, + { + "epoch": 5.720470588235294, + "grad_norm": 7.971067325159088e-05, + "learning_rate": 2.254374758420924e-06, + "loss": 0.0, + "step": 12156 + }, + { + "epoch": 5.720941176470588, + "grad_norm": 0.12562510145049488, + "learning_rate": 2.2468096507386956e-06, + "loss": 0.0204, + "step": 12157 + }, + { + "epoch": 5.721411764705882, + "grad_norm": 0.11461899116892316, + "learning_rate": 2.239257161903668e-06, + "loss": 0.0239, + "step": 12158 + }, + { + "epoch": 5.721882352941177, + "grad_norm": 0.0021120397122538852, + "learning_rate": 2.231717292560825e-06, + "loss": 0.0, + "step": 12159 + }, + { + "epoch": 5.722352941176471, + "grad_norm": 0.026954931020227064, + "learning_rate": 2.224190043354135e-06, + "loss": 0.0018, + "step": 12160 + }, + { + "epoch": 5.722823529411765, + "grad_norm": 0.04345098676549673, + "learning_rate": 2.2166754149264667e-06, + "loss": 0.0016, + "step": 12161 + }, + { + "epoch": 5.723294117647059, + "grad_norm": 0.12490226179426202, + "learning_rate": 2.2091734079195565e-06, + "loss": 0.0162, + "step": 12162 + }, + { + "epoch": 5.723764705882353, + "grad_norm": 0.00015144102481019333, + "learning_rate": 2.2016840229741593e-06, + "loss": 0.0, + "step": 12163 + }, + { + "epoch": 5.724235294117647, + "grad_norm": 0.08701518388687013, + "learning_rate": 2.1942072607299135e-06, + "loss": 0.0172, + "step": 12164 + }, + { + "epoch": 5.724705882352941, + "grad_norm": 0.10248231004137395, + "learning_rate": 2.1867431218253406e-06, + "loss": 0.0192, + "step": 12165 + }, + { + "epoch": 5.725176470588235, + "grad_norm": 0.15309267778379196, + "learning_rate": 2.179291606897915e-06, + "loss": 0.0379, + "step": 12166 + }, + { + "epoch": 5.7256470588235295, + "grad_norm": 0.14398265496291626, + "learning_rate": 2.171852716584044e-06, + "loss": 0.033, + "step": 12167 + }, + { + "epoch": 5.726117647058824, + "grad_norm": 0.12657583897306435, + "learning_rate": 2.1644264515190524e-06, + "loss": 0.0237, + "step": 12168 + }, + { + "epoch": 5.726588235294118, + "grad_norm": 0.1182412051844058, + "learning_rate": 2.1570128123371667e-06, + "loss": 0.0232, + "step": 12169 + }, + { + "epoch": 5.727058823529411, + "grad_norm": 0.0709435859720137, + "learning_rate": 2.1496117996715636e-06, + "loss": 0.0062, + "step": 12170 + }, + { + "epoch": 5.7275294117647055, + "grad_norm": 0.10034306182031016, + "learning_rate": 2.142223414154287e-06, + "loss": 0.009, + "step": 12171 + }, + { + "epoch": 5.728, + "grad_norm": 0.1118093283947306, + "learning_rate": 2.1348476564163997e-06, + "loss": 0.0235, + "step": 12172 + }, + { + "epoch": 5.728470588235294, + "grad_norm": 0.07246117799940714, + "learning_rate": 2.1274845270877804e-06, + "loss": 0.0076, + "step": 12173 + }, + { + "epoch": 5.728941176470588, + "grad_norm": 0.07736471142221173, + "learning_rate": 2.1201340267972933e-06, + "loss": 0.0137, + "step": 12174 + }, + { + "epoch": 5.729411764705882, + "grad_norm": 0.024759689996320663, + "learning_rate": 2.1127961561727193e-06, + "loss": 0.0013, + "step": 12175 + }, + { + "epoch": 5.729882352941177, + "grad_norm": 0.06692192946380679, + "learning_rate": 2.1054709158407235e-06, + "loss": 0.0093, + "step": 12176 + }, + { + "epoch": 5.730352941176471, + "grad_norm": 0.04321081093510302, + "learning_rate": 2.0981583064269226e-06, + "loss": 0.0022, + "step": 12177 + }, + { + "epoch": 5.730823529411765, + "grad_norm": 0.10063208931699966, + "learning_rate": 2.0908583285558834e-06, + "loss": 0.0059, + "step": 12178 + }, + { + "epoch": 5.731294117647058, + "grad_norm": 4.767563367758709e-06, + "learning_rate": 2.083570982851007e-06, + "loss": 0.0, + "step": 12179 + }, + { + "epoch": 5.731764705882353, + "grad_norm": 0.04350689978159308, + "learning_rate": 2.0762962699346965e-06, + "loss": 0.006, + "step": 12180 + }, + { + "epoch": 5.732235294117647, + "grad_norm": 0.10327421574895165, + "learning_rate": 2.06903419042827e-06, + "loss": 0.0154, + "step": 12181 + }, + { + "epoch": 5.732705882352941, + "grad_norm": 0.11607820648005095, + "learning_rate": 2.0617847449519e-06, + "loss": 0.0143, + "step": 12182 + }, + { + "epoch": 5.733176470588235, + "grad_norm": 0.11390964112463337, + "learning_rate": 2.054547934124756e-06, + "loss": 0.0127, + "step": 12183 + }, + { + "epoch": 5.7336470588235295, + "grad_norm": 0.00014127694278878117, + "learning_rate": 2.047323758564878e-06, + "loss": 0.0, + "step": 12184 + }, + { + "epoch": 5.734117647058824, + "grad_norm": 0.1484197040546351, + "learning_rate": 2.040112218889256e-06, + "loss": 0.0454, + "step": 12185 + }, + { + "epoch": 5.734588235294118, + "grad_norm": 0.023664707629581973, + "learning_rate": 2.0329133157137965e-06, + "loss": 0.0024, + "step": 12186 + }, + { + "epoch": 5.735058823529412, + "grad_norm": 0.13079652747621526, + "learning_rate": 2.0257270496533086e-06, + "loss": 0.0309, + "step": 12187 + }, + { + "epoch": 5.7355294117647055, + "grad_norm": 0.1374653950728759, + "learning_rate": 2.0185534213215506e-06, + "loss": 0.0277, + "step": 12188 + }, + { + "epoch": 5.736, + "grad_norm": 8.251292613532799e-07, + "learning_rate": 2.011392431331166e-06, + "loss": 0.0, + "step": 12189 + }, + { + "epoch": 5.736470588235294, + "grad_norm": 0.11634805603129401, + "learning_rate": 2.0042440802937485e-06, + "loss": 0.0171, + "step": 12190 + }, + { + "epoch": 5.736941176470588, + "grad_norm": 0.046441281202567686, + "learning_rate": 1.9971083688197943e-06, + "loss": 0.0041, + "step": 12191 + }, + { + "epoch": 5.737411764705882, + "grad_norm": 0.07639575099118169, + "learning_rate": 1.989985297518748e-06, + "loss": 0.0041, + "step": 12192 + }, + { + "epoch": 5.737882352941177, + "grad_norm": 0.09807663350008668, + "learning_rate": 1.9828748669989405e-06, + "loss": 0.0177, + "step": 12193 + }, + { + "epoch": 5.738352941176471, + "grad_norm": 0.04431705404418493, + "learning_rate": 1.9757770778676183e-06, + "loss": 0.0032, + "step": 12194 + }, + { + "epoch": 5.738823529411764, + "grad_norm": 0.00015456634063938603, + "learning_rate": 1.9686919307310145e-06, + "loss": 0.0, + "step": 12195 + }, + { + "epoch": 5.739294117647058, + "grad_norm": 0.0899827843573531, + "learning_rate": 1.961619426194194e-06, + "loss": 0.0157, + "step": 12196 + }, + { + "epoch": 5.739764705882353, + "grad_norm": 0.05042526953845562, + "learning_rate": 1.9545595648612078e-06, + "loss": 0.0048, + "step": 12197 + }, + { + "epoch": 5.740235294117647, + "grad_norm": 0.03725879715749793, + "learning_rate": 1.9475123473349897e-06, + "loss": 0.0005, + "step": 12198 + }, + { + "epoch": 5.740705882352941, + "grad_norm": 0.18987987054195288, + "learning_rate": 1.9404777742174083e-06, + "loss": 0.0719, + "step": 12199 + }, + { + "epoch": 5.741176470588235, + "grad_norm": 0.1453870514432099, + "learning_rate": 1.9334558461092663e-06, + "loss": 0.0338, + "step": 12200 + }, + { + "epoch": 5.7416470588235295, + "grad_norm": 0.09045803264937764, + "learning_rate": 1.9264465636102342e-06, + "loss": 0.0157, + "step": 12201 + }, + { + "epoch": 5.742117647058824, + "grad_norm": 1.963056216801579e-05, + "learning_rate": 1.9194499273189834e-06, + "loss": 0.0, + "step": 12202 + }, + { + "epoch": 5.742588235294118, + "grad_norm": 0.09539158286310803, + "learning_rate": 1.912465937833035e-06, + "loss": 0.0077, + "step": 12203 + }, + { + "epoch": 5.743058823529411, + "grad_norm": 0.12767842894755665, + "learning_rate": 1.9054945957488466e-06, + "loss": 0.0299, + "step": 12204 + }, + { + "epoch": 5.7435294117647056, + "grad_norm": 0.1682654155883357, + "learning_rate": 1.898535901661824e-06, + "loss": 0.051, + "step": 12205 + }, + { + "epoch": 5.744, + "grad_norm": 0.09130616334581236, + "learning_rate": 1.8915898561662923e-06, + "loss": 0.0067, + "step": 12206 + }, + { + "epoch": 5.744470588235294, + "grad_norm": 5.116595193123826e-06, + "learning_rate": 1.884656459855427e-06, + "loss": 0.0, + "step": 12207 + }, + { + "epoch": 5.744941176470588, + "grad_norm": 0.08144149496801525, + "learning_rate": 1.8777357133214044e-06, + "loss": 0.014, + "step": 12208 + }, + { + "epoch": 5.745411764705882, + "grad_norm": 0.10333154675392062, + "learning_rate": 1.8708276171553017e-06, + "loss": 0.0241, + "step": 12209 + }, + { + "epoch": 5.745882352941177, + "grad_norm": 0.09735232148883735, + "learning_rate": 1.86393217194708e-06, + "loss": 0.0178, + "step": 12210 + }, + { + "epoch": 5.746352941176471, + "grad_norm": 0.04322381636205474, + "learning_rate": 1.857049378285652e-06, + "loss": 0.0012, + "step": 12211 + }, + { + "epoch": 5.746823529411765, + "grad_norm": 0.15624425937219497, + "learning_rate": 1.8501792367588642e-06, + "loss": 0.0493, + "step": 12212 + }, + { + "epoch": 5.747294117647058, + "grad_norm": 0.1344626080260461, + "learning_rate": 1.84332174795343e-06, + "loss": 0.0376, + "step": 12213 + }, + { + "epoch": 5.747764705882353, + "grad_norm": 0.1343833517212998, + "learning_rate": 1.8364769124550149e-06, + "loss": 0.0318, + "step": 12214 + }, + { + "epoch": 5.748235294117647, + "grad_norm": 0.059181412884087424, + "learning_rate": 1.8296447308482176e-06, + "loss": 0.0076, + "step": 12215 + }, + { + "epoch": 5.748705882352941, + "grad_norm": 0.13995440084529093, + "learning_rate": 1.8228252037165215e-06, + "loss": 0.0287, + "step": 12216 + }, + { + "epoch": 5.749176470588235, + "grad_norm": 0.1266361218562542, + "learning_rate": 1.8160183316423605e-06, + "loss": 0.0295, + "step": 12217 + }, + { + "epoch": 5.7496470588235296, + "grad_norm": 0.16674517543363424, + "learning_rate": 1.8092241152070696e-06, + "loss": 0.0421, + "step": 12218 + }, + { + "epoch": 5.750117647058824, + "grad_norm": 0.021403292044939366, + "learning_rate": 1.8024425549909183e-06, + "loss": 0.0003, + "step": 12219 + }, + { + "epoch": 5.750588235294118, + "grad_norm": 2.2064466906661596e-05, + "learning_rate": 1.795673651573093e-06, + "loss": 0.0, + "step": 12220 + }, + { + "epoch": 5.751058823529412, + "grad_norm": 0.16667747617295392, + "learning_rate": 1.788917405531648e-06, + "loss": 0.0512, + "step": 12221 + }, + { + "epoch": 5.751529411764706, + "grad_norm": 0.09335409407581303, + "learning_rate": 1.7821738174436384e-06, + "loss": 0.018, + "step": 12222 + }, + { + "epoch": 5.752, + "grad_norm": 0.13548551490291022, + "learning_rate": 1.7754428878849868e-06, + "loss": 0.0198, + "step": 12223 + }, + { + "epoch": 5.752470588235294, + "grad_norm": 0.13387901130364993, + "learning_rate": 1.7687246174305502e-06, + "loss": 0.0389, + "step": 12224 + }, + { + "epoch": 5.752941176470588, + "grad_norm": 5.05066186529143e-06, + "learning_rate": 1.762019006654103e-06, + "loss": 0.0, + "step": 12225 + }, + { + "epoch": 5.7534117647058824, + "grad_norm": 0.0254708219498747, + "learning_rate": 1.7553260561283533e-06, + "loss": 0.0017, + "step": 12226 + }, + { + "epoch": 5.753882352941177, + "grad_norm": 0.10755357570221839, + "learning_rate": 1.748645766424861e-06, + "loss": 0.0213, + "step": 12227 + }, + { + "epoch": 5.754352941176471, + "grad_norm": 0.12155757638889535, + "learning_rate": 1.7419781381142028e-06, + "loss": 0.0323, + "step": 12228 + }, + { + "epoch": 5.754823529411764, + "grad_norm": 0.1039632042041574, + "learning_rate": 1.7353231717658233e-06, + "loss": 0.0262, + "step": 12229 + }, + { + "epoch": 5.7552941176470584, + "grad_norm": 3.923677747659333e-05, + "learning_rate": 1.728680867948068e-06, + "loss": 0.0, + "step": 12230 + }, + { + "epoch": 5.755764705882353, + "grad_norm": 0.061446032451495276, + "learning_rate": 1.7220512272282328e-06, + "loss": 0.008, + "step": 12231 + }, + { + "epoch": 5.756235294117647, + "grad_norm": 0.058663242176697515, + "learning_rate": 1.7154342501725316e-06, + "loss": 0.0075, + "step": 12232 + }, + { + "epoch": 5.756705882352941, + "grad_norm": 0.14380806799331414, + "learning_rate": 1.708829937346079e-06, + "loss": 0.0333, + "step": 12233 + }, + { + "epoch": 5.757176470588235, + "grad_norm": 0.14780157528517482, + "learning_rate": 1.7022382893129072e-06, + "loss": 0.0434, + "step": 12234 + }, + { + "epoch": 5.75764705882353, + "grad_norm": 0.12301745553831675, + "learning_rate": 1.6956593066359825e-06, + "loss": 0.0217, + "step": 12235 + }, + { + "epoch": 5.758117647058824, + "grad_norm": 0.0599239227378881, + "learning_rate": 1.6890929898771722e-06, + "loss": 0.0059, + "step": 12236 + }, + { + "epoch": 5.758588235294118, + "grad_norm": 0.030088967152362363, + "learning_rate": 1.6825393395972942e-06, + "loss": 0.0021, + "step": 12237 + }, + { + "epoch": 5.759058823529411, + "grad_norm": 0.06816765471689319, + "learning_rate": 1.6759983563560508e-06, + "loss": 0.0046, + "step": 12238 + }, + { + "epoch": 5.759529411764706, + "grad_norm": 0.07712806679061515, + "learning_rate": 1.669470040712062e-06, + "loss": 0.0132, + "step": 12239 + }, + { + "epoch": 5.76, + "grad_norm": 0.08771959942310764, + "learning_rate": 1.6629543932228817e-06, + "loss": 0.0156, + "step": 12240 + }, + { + "epoch": 5.760470588235294, + "grad_norm": 0.07246629720194822, + "learning_rate": 1.6564514144449982e-06, + "loss": 0.0107, + "step": 12241 + }, + { + "epoch": 5.760941176470588, + "grad_norm": 0.17073570364178764, + "learning_rate": 1.6499611049337835e-06, + "loss": 0.0575, + "step": 12242 + }, + { + "epoch": 5.7614117647058825, + "grad_norm": 0.07237525024152243, + "learning_rate": 1.6434834652435447e-06, + "loss": 0.0065, + "step": 12243 + }, + { + "epoch": 5.761882352941177, + "grad_norm": 0.05567597499286117, + "learning_rate": 1.6370184959275058e-06, + "loss": 0.0029, + "step": 12244 + }, + { + "epoch": 5.762352941176471, + "grad_norm": 2.4428706481179118e-05, + "learning_rate": 1.6305661975377916e-06, + "loss": 0.0, + "step": 12245 + }, + { + "epoch": 5.762823529411765, + "grad_norm": 0.00018411719218584265, + "learning_rate": 1.6241265706254947e-06, + "loss": 0.0, + "step": 12246 + }, + { + "epoch": 5.7632941176470585, + "grad_norm": 0.06510559711418493, + "learning_rate": 1.617699615740542e-06, + "loss": 0.0086, + "step": 12247 + }, + { + "epoch": 5.763764705882353, + "grad_norm": 0.05452222739868156, + "learning_rate": 1.6112853334318775e-06, + "loss": 0.0057, + "step": 12248 + }, + { + "epoch": 5.764235294117647, + "grad_norm": 0.11111359458496742, + "learning_rate": 1.6048837242472634e-06, + "loss": 0.0239, + "step": 12249 + }, + { + "epoch": 5.764705882352941, + "grad_norm": 8.495636832387226e-05, + "learning_rate": 1.598494788733462e-06, + "loss": 0.0, + "step": 12250 + }, + { + "epoch": 5.765176470588235, + "grad_norm": 0.02245715930601732, + "learning_rate": 1.5921185274361203e-06, + "loss": 0.0014, + "step": 12251 + }, + { + "epoch": 5.76564705882353, + "grad_norm": 0.09840752154902976, + "learning_rate": 1.585754940899786e-06, + "loss": 0.0176, + "step": 12252 + }, + { + "epoch": 5.766117647058824, + "grad_norm": 0.10518028320024747, + "learning_rate": 1.5794040296679411e-06, + "loss": 0.0226, + "step": 12253 + }, + { + "epoch": 5.766588235294118, + "grad_norm": 0.1028379208350046, + "learning_rate": 1.5730657942830183e-06, + "loss": 0.0213, + "step": 12254 + }, + { + "epoch": 5.767058823529412, + "grad_norm": 0.08453638042644149, + "learning_rate": 1.5667402352862846e-06, + "loss": 0.0204, + "step": 12255 + }, + { + "epoch": 5.767529411764706, + "grad_norm": 0.0969490774674269, + "learning_rate": 1.5604273532179913e-06, + "loss": 0.0211, + "step": 12256 + }, + { + "epoch": 5.768, + "grad_norm": 0.052761015656095485, + "learning_rate": 1.554127148617307e-06, + "loss": 0.0077, + "step": 12257 + }, + { + "epoch": 5.768470588235294, + "grad_norm": 0.1007836043048028, + "learning_rate": 1.5478396220222678e-06, + "loss": 0.0134, + "step": 12258 + }, + { + "epoch": 5.768941176470588, + "grad_norm": 4.201733804876394e-06, + "learning_rate": 1.5415647739698944e-06, + "loss": 0.0, + "step": 12259 + }, + { + "epoch": 5.7694117647058825, + "grad_norm": 0.10410606732204462, + "learning_rate": 1.5353026049960748e-06, + "loss": 0.0155, + "step": 12260 + }, + { + "epoch": 5.769882352941177, + "grad_norm": 0.14078687338110205, + "learning_rate": 1.5290531156356146e-06, + "loss": 0.0449, + "step": 12261 + }, + { + "epoch": 5.770352941176471, + "grad_norm": 0.06750581639121583, + "learning_rate": 1.5228163064222698e-06, + "loss": 0.0116, + "step": 12262 + }, + { + "epoch": 5.770823529411764, + "grad_norm": 0.15569199261816807, + "learning_rate": 1.516592177888698e-06, + "loss": 0.0435, + "step": 12263 + }, + { + "epoch": 5.7712941176470585, + "grad_norm": 0.09053148136030877, + "learning_rate": 1.5103807305664573e-06, + "loss": 0.0176, + "step": 12264 + }, + { + "epoch": 5.771764705882353, + "grad_norm": 1.2033901638340624e-05, + "learning_rate": 1.5041819649860399e-06, + "loss": 0.0, + "step": 12265 + }, + { + "epoch": 5.772235294117647, + "grad_norm": 0.08284221747598118, + "learning_rate": 1.4979958816768556e-06, + "loss": 0.0045, + "step": 12266 + }, + { + "epoch": 5.772705882352941, + "grad_norm": 0.05090714105132926, + "learning_rate": 1.4918224811672153e-06, + "loss": 0.0034, + "step": 12267 + }, + { + "epoch": 5.773176470588235, + "grad_norm": 0.17103174649562133, + "learning_rate": 1.485661763984347e-06, + "loss": 0.0504, + "step": 12268 + }, + { + "epoch": 5.77364705882353, + "grad_norm": 0.12643094267364346, + "learning_rate": 1.479513730654447e-06, + "loss": 0.0276, + "step": 12269 + }, + { + "epoch": 5.774117647058824, + "grad_norm": 3.163362403710762e-05, + "learning_rate": 1.4733783817025613e-06, + "loss": 0.0, + "step": 12270 + }, + { + "epoch": 5.774588235294118, + "grad_norm": 0.1020079278489475, + "learning_rate": 1.4672557176526712e-06, + "loss": 0.0139, + "step": 12271 + }, + { + "epoch": 5.775058823529411, + "grad_norm": 0.0860705219320392, + "learning_rate": 1.4611457390276915e-06, + "loss": 0.0175, + "step": 12272 + }, + { + "epoch": 5.775529411764706, + "grad_norm": 1.6319112381268344e-05, + "learning_rate": 1.4550484463494384e-06, + "loss": 0.0, + "step": 12273 + }, + { + "epoch": 5.776, + "grad_norm": 0.07387857120773209, + "learning_rate": 1.4489638401386783e-06, + "loss": 0.0065, + "step": 12274 + }, + { + "epoch": 5.776470588235294, + "grad_norm": 0.06942608915193009, + "learning_rate": 1.442891920915029e-06, + "loss": 0.0082, + "step": 12275 + }, + { + "epoch": 5.776941176470588, + "grad_norm": 0.21542971456595061, + "learning_rate": 1.436832689197076e-06, + "loss": 0.0264, + "step": 12276 + }, + { + "epoch": 5.7774117647058825, + "grad_norm": 0.09916919820782427, + "learning_rate": 1.4307861455023218e-06, + "loss": 0.0078, + "step": 12277 + }, + { + "epoch": 5.777882352941177, + "grad_norm": 0.08050771210660766, + "learning_rate": 1.4247522903471531e-06, + "loss": 0.005, + "step": 12278 + }, + { + "epoch": 5.778352941176471, + "grad_norm": 0.09589940981123725, + "learning_rate": 1.4187311242468914e-06, + "loss": 0.0203, + "step": 12279 + }, + { + "epoch": 5.778823529411765, + "grad_norm": 0.14863566668392944, + "learning_rate": 1.4127226477157915e-06, + "loss": 0.0425, + "step": 12280 + }, + { + "epoch": 5.7792941176470585, + "grad_norm": 0.010130875275354282, + "learning_rate": 1.4067268612669769e-06, + "loss": 0.0002, + "step": 12281 + }, + { + "epoch": 5.779764705882353, + "grad_norm": 0.11401947090485044, + "learning_rate": 1.4007437654125375e-06, + "loss": 0.0204, + "step": 12282 + }, + { + "epoch": 5.780235294117647, + "grad_norm": 0.07129171891368737, + "learning_rate": 1.3947733606634647e-06, + "loss": 0.0087, + "step": 12283 + }, + { + "epoch": 5.780705882352941, + "grad_norm": 0.1853168469138306, + "learning_rate": 1.388815647529634e-06, + "loss": 0.0459, + "step": 12284 + }, + { + "epoch": 5.781176470588235, + "grad_norm": 0.088006124635006, + "learning_rate": 1.3828706265199051e-06, + "loss": 0.0124, + "step": 12285 + }, + { + "epoch": 5.78164705882353, + "grad_norm": 0.00020896261760092497, + "learning_rate": 1.376938298141972e-06, + "loss": 0.0, + "step": 12286 + }, + { + "epoch": 5.782117647058824, + "grad_norm": 0.06935809647461222, + "learning_rate": 1.3710186629025121e-06, + "loss": 0.004, + "step": 12287 + }, + { + "epoch": 5.782588235294118, + "grad_norm": 0.16633320634189505, + "learning_rate": 1.3651117213070716e-06, + "loss": 0.0318, + "step": 12288 + }, + { + "epoch": 5.783058823529412, + "grad_norm": 5.124045586198971e-06, + "learning_rate": 1.3592174738601302e-06, + "loss": 0.0, + "step": 12289 + }, + { + "epoch": 5.783529411764706, + "grad_norm": 0.09793324671635062, + "learning_rate": 1.3533359210651184e-06, + "loss": 0.0102, + "step": 12290 + }, + { + "epoch": 5.784, + "grad_norm": 0.12202378685545073, + "learning_rate": 1.3474670634243178e-06, + "loss": 0.0327, + "step": 12291 + }, + { + "epoch": 5.784470588235294, + "grad_norm": 0.1625713594652457, + "learning_rate": 1.341610901438961e-06, + "loss": 0.0464, + "step": 12292 + }, + { + "epoch": 5.784941176470588, + "grad_norm": 3.668069768763851e-05, + "learning_rate": 1.3357674356091974e-06, + "loss": 0.0, + "step": 12293 + }, + { + "epoch": 5.7854117647058825, + "grad_norm": 9.373282032743176e-05, + "learning_rate": 1.3299366664340948e-06, + "loss": 0.0, + "step": 12294 + }, + { + "epoch": 5.785882352941177, + "grad_norm": 0.1441633402483336, + "learning_rate": 1.3241185944116217e-06, + "loss": 0.0384, + "step": 12295 + }, + { + "epoch": 5.786352941176471, + "grad_norm": 0.04529984180160355, + "learning_rate": 1.3183132200386637e-06, + "loss": 0.0038, + "step": 12296 + }, + { + "epoch": 5.786823529411764, + "grad_norm": 0.05867615519105044, + "learning_rate": 1.3125205438110408e-06, + "loss": 0.0078, + "step": 12297 + }, + { + "epoch": 5.7872941176470585, + "grad_norm": 0.07823651946867316, + "learning_rate": 1.3067405662234743e-06, + "loss": 0.0042, + "step": 12298 + }, + { + "epoch": 5.787764705882353, + "grad_norm": 2.697882992092874e-05, + "learning_rate": 1.3009732877696022e-06, + "loss": 0.0, + "step": 12299 + }, + { + "epoch": 5.788235294117647, + "grad_norm": 0.0871309624429393, + "learning_rate": 1.2952187089419642e-06, + "loss": 0.0151, + "step": 12300 + }, + { + "epoch": 5.788705882352941, + "grad_norm": 1.6189036868268577e-05, + "learning_rate": 1.28947683023205e-06, + "loss": 0.0, + "step": 12301 + }, + { + "epoch": 5.789176470588235, + "grad_norm": 0.04425291141645657, + "learning_rate": 1.2837476521302514e-06, + "loss": 0.0009, + "step": 12302 + }, + { + "epoch": 5.78964705882353, + "grad_norm": 6.831475189712752e-05, + "learning_rate": 1.2780311751258265e-06, + "loss": 0.0, + "step": 12303 + }, + { + "epoch": 5.790117647058824, + "grad_norm": 0.11062230586033618, + "learning_rate": 1.2723273997070182e-06, + "loss": 0.0203, + "step": 12304 + }, + { + "epoch": 5.790588235294118, + "grad_norm": 0.05836384730951641, + "learning_rate": 1.2666363263609702e-06, + "loss": 0.0045, + "step": 12305 + }, + { + "epoch": 5.791058823529411, + "grad_norm": 0.06519521659940462, + "learning_rate": 1.2609579555737104e-06, + "loss": 0.0061, + "step": 12306 + }, + { + "epoch": 5.791529411764706, + "grad_norm": 0.12291068784493135, + "learning_rate": 1.2552922878301841e-06, + "loss": 0.023, + "step": 12307 + }, + { + "epoch": 5.792, + "grad_norm": 3.2837749520307075e-06, + "learning_rate": 1.2496393236143044e-06, + "loss": 0.0, + "step": 12308 + }, + { + "epoch": 5.792470588235294, + "grad_norm": 7.314691612500837e-05, + "learning_rate": 1.2439990634088348e-06, + "loss": 0.0, + "step": 12309 + }, + { + "epoch": 5.792941176470588, + "grad_norm": 0.09671560851391865, + "learning_rate": 1.2383715076954902e-06, + "loss": 0.0113, + "step": 12310 + }, + { + "epoch": 5.7934117647058825, + "grad_norm": 0.02881605471617403, + "learning_rate": 1.232756656954903e-06, + "loss": 0.0023, + "step": 12311 + }, + { + "epoch": 5.793882352941177, + "grad_norm": 0.0413361901812794, + "learning_rate": 1.2271545116665727e-06, + "loss": 0.0028, + "step": 12312 + }, + { + "epoch": 5.794352941176471, + "grad_norm": 0.07279179304143824, + "learning_rate": 1.2215650723089832e-06, + "loss": 0.0114, + "step": 12313 + }, + { + "epoch": 5.794823529411765, + "grad_norm": 0.07070623470844273, + "learning_rate": 1.2159883393595026e-06, + "loss": 0.0116, + "step": 12314 + }, + { + "epoch": 5.7952941176470585, + "grad_norm": 0.1320884109439995, + "learning_rate": 1.2104243132943836e-06, + "loss": 0.0282, + "step": 12315 + }, + { + "epoch": 5.795764705882353, + "grad_norm": 0.10347631745352094, + "learning_rate": 1.2048729945888458e-06, + "loss": 0.0221, + "step": 12316 + }, + { + "epoch": 5.796235294117647, + "grad_norm": 0.12427168812456713, + "learning_rate": 1.1993343837169933e-06, + "loss": 0.025, + "step": 12317 + }, + { + "epoch": 5.796705882352941, + "grad_norm": 0.029838295324772897, + "learning_rate": 1.1938084811518311e-06, + "loss": 0.0024, + "step": 12318 + }, + { + "epoch": 5.797176470588235, + "grad_norm": 0.00011310319394156065, + "learning_rate": 1.1882952873653317e-06, + "loss": 0.0, + "step": 12319 + }, + { + "epoch": 5.79764705882353, + "grad_norm": 0.020506259435194206, + "learning_rate": 1.1827948028283352e-06, + "loss": 0.0008, + "step": 12320 + }, + { + "epoch": 5.798117647058824, + "grad_norm": 3.876761465301254e-05, + "learning_rate": 1.1773070280105822e-06, + "loss": 0.0, + "step": 12321 + }, + { + "epoch": 5.798588235294117, + "grad_norm": 0.16897340151606832, + "learning_rate": 1.171831963380815e-06, + "loss": 0.0415, + "step": 12322 + }, + { + "epoch": 5.799058823529411, + "grad_norm": 0.15974664513529938, + "learning_rate": 1.1663696094065755e-06, + "loss": 0.0286, + "step": 12323 + }, + { + "epoch": 5.799529411764706, + "grad_norm": 0.10437484265867279, + "learning_rate": 1.160919966554391e-06, + "loss": 0.0175, + "step": 12324 + }, + { + "epoch": 5.8, + "grad_norm": 0.019535092130006183, + "learning_rate": 1.155483035289706e-06, + "loss": 0.001, + "step": 12325 + }, + { + "epoch": 5.800470588235294, + "grad_norm": 0.04356610233369766, + "learning_rate": 1.150058816076832e-06, + "loss": 0.0015, + "step": 12326 + }, + { + "epoch": 5.800941176470588, + "grad_norm": 0.21522221410776982, + "learning_rate": 1.1446473093790487e-06, + "loss": 0.0795, + "step": 12327 + }, + { + "epoch": 5.8014117647058825, + "grad_norm": 0.0708595327615368, + "learning_rate": 1.1392485156585196e-06, + "loss": 0.0089, + "step": 12328 + }, + { + "epoch": 5.801882352941177, + "grad_norm": 0.08668828072280575, + "learning_rate": 1.1338624353763093e-06, + "loss": 0.0127, + "step": 12329 + }, + { + "epoch": 5.802352941176471, + "grad_norm": 0.16232702495501347, + "learning_rate": 1.1284890689924497e-06, + "loss": 0.0297, + "step": 12330 + }, + { + "epoch": 5.802823529411764, + "grad_norm": 0.04163920081142287, + "learning_rate": 1.1231284169658239e-06, + "loss": 0.0037, + "step": 12331 + }, + { + "epoch": 5.8032941176470585, + "grad_norm": 0.09112963737112727, + "learning_rate": 1.1177804797542821e-06, + "loss": 0.0128, + "step": 12332 + }, + { + "epoch": 5.803764705882353, + "grad_norm": 0.06418846475610465, + "learning_rate": 1.1124452578145425e-06, + "loss": 0.0033, + "step": 12333 + }, + { + "epoch": 5.804235294117647, + "grad_norm": 0.1227260163752379, + "learning_rate": 1.1071227516022574e-06, + "loss": 0.0314, + "step": 12334 + }, + { + "epoch": 5.804705882352941, + "grad_norm": 0.04547080988396775, + "learning_rate": 1.1018129615720129e-06, + "loss": 0.0059, + "step": 12335 + }, + { + "epoch": 5.805176470588235, + "grad_norm": 0.10014652020088818, + "learning_rate": 1.0965158881772961e-06, + "loss": 0.0162, + "step": 12336 + }, + { + "epoch": 5.80564705882353, + "grad_norm": 1.3420822207711758e-05, + "learning_rate": 1.091231531870479e-06, + "loss": 0.0, + "step": 12337 + }, + { + "epoch": 5.806117647058824, + "grad_norm": 0.12851697713528507, + "learning_rate": 1.0859598931028668e-06, + "loss": 0.0249, + "step": 12338 + }, + { + "epoch": 5.806588235294118, + "grad_norm": 0.1476516555457313, + "learning_rate": 1.0807009723247328e-06, + "loss": 0.0499, + "step": 12339 + }, + { + "epoch": 5.807058823529411, + "grad_norm": 0.07845671147161143, + "learning_rate": 1.0754547699851513e-06, + "loss": 0.0122, + "step": 12340 + }, + { + "epoch": 5.807529411764706, + "grad_norm": 0.15817763375505242, + "learning_rate": 1.0702212865322135e-06, + "loss": 0.043, + "step": 12341 + }, + { + "epoch": 5.808, + "grad_norm": 0.20069276067132205, + "learning_rate": 1.0650005224128788e-06, + "loss": 0.0313, + "step": 12342 + }, + { + "epoch": 5.808470588235294, + "grad_norm": 0.04374771721713079, + "learning_rate": 1.059792478073007e-06, + "loss": 0.0037, + "step": 12343 + }, + { + "epoch": 5.808941176470588, + "grad_norm": 0.09646754584708596, + "learning_rate": 1.0545971539574094e-06, + "loss": 0.026, + "step": 12344 + }, + { + "epoch": 5.8094117647058825, + "grad_norm": 0.08394537393200582, + "learning_rate": 1.0494145505097973e-06, + "loss": 0.0126, + "step": 12345 + }, + { + "epoch": 5.809882352941177, + "grad_norm": 7.250368539724007e-06, + "learning_rate": 1.0442446681727834e-06, + "loss": 0.0, + "step": 12346 + }, + { + "epoch": 5.810352941176471, + "grad_norm": 0.09283164126659232, + "learning_rate": 1.0390875073878812e-06, + "loss": 0.0191, + "step": 12347 + }, + { + "epoch": 5.810823529411765, + "grad_norm": 0.047804012098681875, + "learning_rate": 1.0339430685955719e-06, + "loss": 0.0053, + "step": 12348 + }, + { + "epoch": 5.8112941176470585, + "grad_norm": 0.033791298346625184, + "learning_rate": 1.0288113522352038e-06, + "loss": 0.0009, + "step": 12349 + }, + { + "epoch": 5.811764705882353, + "grad_norm": 4.417172122207749e-05, + "learning_rate": 1.0236923587450263e-06, + "loss": 0.0, + "step": 12350 + }, + { + "epoch": 5.812235294117647, + "grad_norm": 0.04619608730598235, + "learning_rate": 1.0185860885622565e-06, + "loss": 0.0025, + "step": 12351 + }, + { + "epoch": 5.812705882352941, + "grad_norm": 0.1873035753896617, + "learning_rate": 1.013492542122979e-06, + "loss": 0.0625, + "step": 12352 + }, + { + "epoch": 5.813176470588235, + "grad_norm": 0.11613027126329116, + "learning_rate": 1.0084117198622287e-06, + "loss": 0.0239, + "step": 12353 + }, + { + "epoch": 5.81364705882353, + "grad_norm": 0.15267987379297737, + "learning_rate": 1.0033436222138924e-06, + "loss": 0.0564, + "step": 12354 + }, + { + "epoch": 5.814117647058824, + "grad_norm": 8.457144216907643e-06, + "learning_rate": 9.98288249610857e-07, + "loss": 0.0, + "step": 12355 + }, + { + "epoch": 5.814588235294117, + "grad_norm": 0.12573596196892603, + "learning_rate": 9.932456024848435e-07, + "loss": 0.0378, + "step": 12356 + }, + { + "epoch": 5.815058823529411, + "grad_norm": 0.11472128437741375, + "learning_rate": 9.882156812665244e-07, + "loss": 0.0232, + "step": 12357 + }, + { + "epoch": 5.815529411764706, + "grad_norm": 0.060702389966575544, + "learning_rate": 9.831984863854891e-07, + "loss": 0.0081, + "step": 12358 + }, + { + "epoch": 5.816, + "grad_norm": 0.13132512260172644, + "learning_rate": 9.781940182702451e-07, + "loss": 0.0257, + "step": 12359 + }, + { + "epoch": 5.816470588235294, + "grad_norm": 0.0784230175031894, + "learning_rate": 9.732022773481506e-07, + "loss": 0.0065, + "step": 12360 + }, + { + "epoch": 5.816941176470588, + "grad_norm": 0.13674398519332848, + "learning_rate": 9.682232640455644e-07, + "loss": 0.0301, + "step": 12361 + }, + { + "epoch": 5.8174117647058825, + "grad_norm": 0.16914890619841688, + "learning_rate": 9.632569787877298e-07, + "loss": 0.0487, + "step": 12362 + }, + { + "epoch": 5.817882352941177, + "grad_norm": 0.10836407037776423, + "learning_rate": 9.583034219987406e-07, + "loss": 0.0283, + "step": 12363 + }, + { + "epoch": 5.818352941176471, + "grad_norm": 0.00020601304840871514, + "learning_rate": 9.53362594101692e-07, + "loss": 0.0, + "step": 12364 + }, + { + "epoch": 5.818823529411764, + "grad_norm": 0.07599662573010675, + "learning_rate": 9.48434495518563e-07, + "loss": 0.0107, + "step": 12365 + }, + { + "epoch": 5.8192941176470585, + "grad_norm": 0.06510413797454664, + "learning_rate": 9.435191266702003e-07, + "loss": 0.0066, + "step": 12366 + }, + { + "epoch": 5.819764705882353, + "grad_norm": 0.11090230370734946, + "learning_rate": 9.386164879764346e-07, + "loss": 0.0109, + "step": 12367 + }, + { + "epoch": 5.820235294117647, + "grad_norm": 0.10549028623341045, + "learning_rate": 9.337265798559645e-07, + "loss": 0.0191, + "step": 12368 + }, + { + "epoch": 5.820705882352941, + "grad_norm": 0.20200770901436785, + "learning_rate": 9.28849402726406e-07, + "loss": 0.0289, + "step": 12369 + }, + { + "epoch": 5.821176470588235, + "grad_norm": 0.15739456104728747, + "learning_rate": 9.239849570042924e-07, + "loss": 0.0401, + "step": 12370 + }, + { + "epoch": 5.82164705882353, + "grad_norm": 0.13260893199445628, + "learning_rate": 9.191332431050747e-07, + "loss": 0.0358, + "step": 12371 + }, + { + "epoch": 5.822117647058824, + "grad_norm": 0.053046459614532165, + "learning_rate": 9.142942614431215e-07, + "loss": 0.0044, + "step": 12372 + }, + { + "epoch": 5.822588235294118, + "grad_norm": 0.04028749344112354, + "learning_rate": 9.094680124317022e-07, + "loss": 0.0022, + "step": 12373 + }, + { + "epoch": 5.823058823529411, + "grad_norm": 0.08886066466972391, + "learning_rate": 9.046544964830038e-07, + "loss": 0.0119, + "step": 12374 + }, + { + "epoch": 5.823529411764706, + "grad_norm": 0.07286468651271316, + "learning_rate": 8.998537140081141e-07, + "loss": 0.0124, + "step": 12375 + }, + { + "epoch": 5.824, + "grad_norm": 0.14524030746408947, + "learning_rate": 8.950656654170552e-07, + "loss": 0.0347, + "step": 12376 + }, + { + "epoch": 5.824470588235294, + "grad_norm": 0.10656554693851036, + "learning_rate": 8.902903511187332e-07, + "loss": 0.0157, + "step": 12377 + }, + { + "epoch": 5.824941176470588, + "grad_norm": 0.15467482745119202, + "learning_rate": 8.855277715210052e-07, + "loss": 0.0348, + "step": 12378 + }, + { + "epoch": 5.8254117647058825, + "grad_norm": 5.8882183615058945e-06, + "learning_rate": 8.807779270305959e-07, + "loss": 0.0, + "step": 12379 + }, + { + "epoch": 5.825882352941177, + "grad_norm": 0.17312980313004647, + "learning_rate": 8.760408180531809e-07, + "loss": 0.0528, + "step": 12380 + }, + { + "epoch": 5.826352941176471, + "grad_norm": 0.10864615407132366, + "learning_rate": 8.713164449933364e-07, + "loss": 0.0241, + "step": 12381 + }, + { + "epoch": 5.826823529411765, + "grad_norm": 0.1381038938950426, + "learning_rate": 8.666048082545396e-07, + "loss": 0.0173, + "step": 12382 + }, + { + "epoch": 5.8272941176470585, + "grad_norm": 8.580758874458066e-06, + "learning_rate": 8.619059082391689e-07, + "loss": 0.0, + "step": 12383 + }, + { + "epoch": 5.827764705882353, + "grad_norm": 0.15565652164180227, + "learning_rate": 8.572197453485696e-07, + "loss": 0.0422, + "step": 12384 + }, + { + "epoch": 5.828235294117647, + "grad_norm": 0.07754906439355135, + "learning_rate": 8.525463199829385e-07, + "loss": 0.0103, + "step": 12385 + }, + { + "epoch": 5.828705882352941, + "grad_norm": 0.16285171778448618, + "learning_rate": 8.47885632541423e-07, + "loss": 0.0461, + "step": 12386 + }, + { + "epoch": 5.829176470588235, + "grad_norm": 0.048090561740041046, + "learning_rate": 8.432376834220544e-07, + "loss": 0.007, + "step": 12387 + }, + { + "epoch": 5.82964705882353, + "grad_norm": 0.10892858070782634, + "learning_rate": 8.386024730217988e-07, + "loss": 0.0155, + "step": 12388 + }, + { + "epoch": 5.830117647058824, + "grad_norm": 0.10031044758733666, + "learning_rate": 8.339800017365228e-07, + "loss": 0.0198, + "step": 12389 + }, + { + "epoch": 5.830588235294117, + "grad_norm": 2.2006753369958332e-05, + "learning_rate": 8.293702699610272e-07, + "loss": 0.0, + "step": 12390 + }, + { + "epoch": 5.831058823529411, + "grad_norm": 0.16049503027230513, + "learning_rate": 8.247732780889637e-07, + "loss": 0.0432, + "step": 12391 + }, + { + "epoch": 5.831529411764706, + "grad_norm": 0.10235516906030859, + "learning_rate": 8.201890265129684e-07, + "loss": 0.0228, + "step": 12392 + }, + { + "epoch": 5.832, + "grad_norm": 0.05696942259223178, + "learning_rate": 8.156175156245614e-07, + "loss": 0.0052, + "step": 12393 + }, + { + "epoch": 5.832470588235294, + "grad_norm": 0.04298609226455238, + "learning_rate": 8.110587458141637e-07, + "loss": 0.0031, + "step": 12394 + }, + { + "epoch": 5.832941176470588, + "grad_norm": 0.15714094338900056, + "learning_rate": 8.065127174711139e-07, + "loss": 0.0358, + "step": 12395 + }, + { + "epoch": 5.8334117647058825, + "grad_norm": 0.13248974490915205, + "learning_rate": 8.01979430983668e-07, + "loss": 0.0347, + "step": 12396 + }, + { + "epoch": 5.833882352941177, + "grad_norm": 0.09009534180487799, + "learning_rate": 7.97458886738983e-07, + "loss": 0.0169, + "step": 12397 + }, + { + "epoch": 5.834352941176471, + "grad_norm": 0.09396366385364509, + "learning_rate": 7.929510851231502e-07, + "loss": 0.0166, + "step": 12398 + }, + { + "epoch": 5.834823529411764, + "grad_norm": 0.12150209125716584, + "learning_rate": 7.884560265211615e-07, + "loss": 0.0146, + "step": 12399 + }, + { + "epoch": 5.8352941176470585, + "grad_norm": 0.07509492077434976, + "learning_rate": 7.839737113168931e-07, + "loss": 0.0099, + "step": 12400 + }, + { + "epoch": 5.835764705882353, + "grad_norm": 9.061447354722466e-05, + "learning_rate": 7.795041398931723e-07, + "loss": 0.0, + "step": 12401 + }, + { + "epoch": 5.836235294117647, + "grad_norm": 0.0004959661760583693, + "learning_rate": 7.750473126317269e-07, + "loss": 0.0, + "step": 12402 + }, + { + "epoch": 5.836705882352941, + "grad_norm": 0.12228133398612354, + "learning_rate": 7.706032299131692e-07, + "loss": 0.0271, + "step": 12403 + }, + { + "epoch": 5.837176470588235, + "grad_norm": 0.06305320833217562, + "learning_rate": 7.66171892117079e-07, + "loss": 0.0039, + "step": 12404 + }, + { + "epoch": 5.83764705882353, + "grad_norm": 0.0686614658916789, + "learning_rate": 7.617532996218867e-07, + "loss": 0.0062, + "step": 12405 + }, + { + "epoch": 5.838117647058824, + "grad_norm": 0.09328092701050955, + "learning_rate": 7.573474528049739e-07, + "loss": 0.0152, + "step": 12406 + }, + { + "epoch": 5.838588235294118, + "grad_norm": 0.1051185342171726, + "learning_rate": 7.529543520426228e-07, + "loss": 0.0255, + "step": 12407 + }, + { + "epoch": 5.839058823529411, + "grad_norm": 0.0747018427372878, + "learning_rate": 7.485739977100336e-07, + "loss": 0.0076, + "step": 12408 + }, + { + "epoch": 5.839529411764706, + "grad_norm": 0.06407286044928377, + "learning_rate": 7.442063901812734e-07, + "loss": 0.0086, + "step": 12409 + }, + { + "epoch": 5.84, + "grad_norm": 0.04141692228232666, + "learning_rate": 7.398515298294106e-07, + "loss": 0.0047, + "step": 12410 + }, + { + "epoch": 5.840470588235294, + "grad_norm": 0.18595273313371566, + "learning_rate": 7.355094170263143e-07, + "loss": 0.0213, + "step": 12411 + }, + { + "epoch": 5.840941176470588, + "grad_norm": 0.07538236281444363, + "learning_rate": 7.311800521428712e-07, + "loss": 0.0084, + "step": 12412 + }, + { + "epoch": 5.8414117647058825, + "grad_norm": 0.11150971352605218, + "learning_rate": 7.26863435548819e-07, + "loss": 0.0235, + "step": 12413 + }, + { + "epoch": 5.841882352941177, + "grad_norm": 0.16735775265518815, + "learning_rate": 7.22559567612796e-07, + "loss": 0.0601, + "step": 12414 + }, + { + "epoch": 5.842352941176471, + "grad_norm": 0.06749023957831916, + "learning_rate": 7.182684487023748e-07, + "loss": 0.0068, + "step": 12415 + }, + { + "epoch": 5.842823529411765, + "grad_norm": 7.404826387363045e-06, + "learning_rate": 7.139900791840792e-07, + "loss": 0.0, + "step": 12416 + }, + { + "epoch": 5.8432941176470585, + "grad_norm": 0.10710457445094156, + "learning_rate": 7.097244594232498e-07, + "loss": 0.0208, + "step": 12417 + }, + { + "epoch": 5.843764705882353, + "grad_norm": 0.10583447450589462, + "learning_rate": 7.054715897842289e-07, + "loss": 0.0101, + "step": 12418 + }, + { + "epoch": 5.844235294117647, + "grad_norm": 0.04222606712270024, + "learning_rate": 7.012314706302257e-07, + "loss": 0.0022, + "step": 12419 + }, + { + "epoch": 5.844705882352941, + "grad_norm": 0.15859213292554328, + "learning_rate": 6.970041023233508e-07, + "loss": 0.034, + "step": 12420 + }, + { + "epoch": 5.845176470588235, + "grad_norm": 1.673971484365116e-05, + "learning_rate": 6.927894852246485e-07, + "loss": 0.0, + "step": 12421 + }, + { + "epoch": 5.84564705882353, + "grad_norm": 0.12629780572099214, + "learning_rate": 6.88587619694081e-07, + "loss": 0.0318, + "step": 12422 + }, + { + "epoch": 5.846117647058824, + "grad_norm": 0.09247673022574644, + "learning_rate": 6.843985060904944e-07, + "loss": 0.0139, + "step": 12423 + }, + { + "epoch": 5.846588235294117, + "grad_norm": 0.09548164386334124, + "learning_rate": 6.802221447716694e-07, + "loss": 0.02, + "step": 12424 + }, + { + "epoch": 5.847058823529411, + "grad_norm": 0.10375383130229915, + "learning_rate": 6.760585360942872e-07, + "loss": 0.0266, + "step": 12425 + }, + { + "epoch": 5.847529411764706, + "grad_norm": 4.619140800899931e-06, + "learning_rate": 6.719076804139134e-07, + "loss": 0.0, + "step": 12426 + }, + { + "epoch": 5.848, + "grad_norm": 4.476787870041475e-06, + "learning_rate": 6.677695780850978e-07, + "loss": 0.0, + "step": 12427 + }, + { + "epoch": 5.848470588235294, + "grad_norm": 0.09448210424252701, + "learning_rate": 6.636442294612243e-07, + "loss": 0.0196, + "step": 12428 + }, + { + "epoch": 5.848941176470588, + "grad_norm": 0.17530582267536926, + "learning_rate": 6.595316348946278e-07, + "loss": 0.0184, + "step": 12429 + }, + { + "epoch": 5.8494117647058825, + "grad_norm": 0.07371654505935597, + "learning_rate": 6.554317947365273e-07, + "loss": 0.0085, + "step": 12430 + }, + { + "epoch": 5.849882352941177, + "grad_norm": 0.22012342361947754, + "learning_rate": 6.513447093370927e-07, + "loss": 0.0795, + "step": 12431 + }, + { + "epoch": 5.850352941176471, + "grad_norm": 0.05794334217466675, + "learning_rate": 6.472703790453781e-07, + "loss": 0.0073, + "step": 12432 + }, + { + "epoch": 5.850823529411764, + "grad_norm": 0.1608290166517619, + "learning_rate": 6.43208804209322e-07, + "loss": 0.0434, + "step": 12433 + }, + { + "epoch": 5.8512941176470585, + "grad_norm": 0.13608915930535936, + "learning_rate": 6.391599851758466e-07, + "loss": 0.0373, + "step": 12434 + }, + { + "epoch": 5.851764705882353, + "grad_norm": 0.09969776783588011, + "learning_rate": 6.351239222907089e-07, + "loss": 0.0074, + "step": 12435 + }, + { + "epoch": 5.852235294117647, + "grad_norm": 0.051464565514197945, + "learning_rate": 6.311006158986165e-07, + "loss": 0.0047, + "step": 12436 + }, + { + "epoch": 5.852705882352941, + "grad_norm": 6.82569166494666e-05, + "learning_rate": 6.270900663431777e-07, + "loss": 0.0, + "step": 12437 + }, + { + "epoch": 5.853176470588235, + "grad_norm": 0.06731704799746949, + "learning_rate": 6.230922739669353e-07, + "loss": 0.0087, + "step": 12438 + }, + { + "epoch": 5.85364705882353, + "grad_norm": 3.265746795490269e-05, + "learning_rate": 6.191072391112828e-07, + "loss": 0.0, + "step": 12439 + }, + { + "epoch": 5.854117647058824, + "grad_norm": 0.10426446909804717, + "learning_rate": 6.151349621165979e-07, + "loss": 0.0305, + "step": 12440 + }, + { + "epoch": 5.854588235294118, + "grad_norm": 0.01940741545323678, + "learning_rate": 6.111754433221094e-07, + "loss": 0.0009, + "step": 12441 + }, + { + "epoch": 5.855058823529411, + "grad_norm": 2.6250331895550457e-06, + "learning_rate": 6.0722868306598e-07, + "loss": 0.0, + "step": 12442 + }, + { + "epoch": 5.855529411764706, + "grad_norm": 0.1494764124006097, + "learning_rate": 6.032946816852902e-07, + "loss": 0.0366, + "step": 12443 + }, + { + "epoch": 5.856, + "grad_norm": 0.22419495068276163, + "learning_rate": 5.99373439516021e-07, + "loss": 0.0573, + "step": 12444 + }, + { + "epoch": 5.856470588235294, + "grad_norm": 1.8697813204359504e-06, + "learning_rate": 5.954649568930714e-07, + "loss": 0.0, + "step": 12445 + }, + { + "epoch": 5.856941176470588, + "grad_norm": 0.07354635050733607, + "learning_rate": 5.91569234150241e-07, + "loss": 0.0101, + "step": 12446 + }, + { + "epoch": 5.8574117647058825, + "grad_norm": 0.12521033755209784, + "learning_rate": 5.876862716202302e-07, + "loss": 0.0381, + "step": 12447 + }, + { + "epoch": 5.857882352941177, + "grad_norm": 0.07147837570256173, + "learning_rate": 5.838160696346906e-07, + "loss": 0.0069, + "step": 12448 + }, + { + "epoch": 5.85835294117647, + "grad_norm": 0.13085126740781708, + "learning_rate": 5.799586285241242e-07, + "loss": 0.0283, + "step": 12449 + }, + { + "epoch": 5.858823529411764, + "grad_norm": 0.07471474955866911, + "learning_rate": 5.761139486180178e-07, + "loss": 0.0074, + "step": 12450 + }, + { + "epoch": 5.8592941176470585, + "grad_norm": 0.0850248911431414, + "learning_rate": 5.722820302446751e-07, + "loss": 0.0124, + "step": 12451 + }, + { + "epoch": 5.859764705882353, + "grad_norm": 0.12003772378134016, + "learning_rate": 5.684628737314012e-07, + "loss": 0.024, + "step": 12452 + }, + { + "epoch": 5.860235294117647, + "grad_norm": 8.90081196947081e-06, + "learning_rate": 5.646564794043518e-07, + "loss": 0.0, + "step": 12453 + }, + { + "epoch": 5.860705882352941, + "grad_norm": 0.008432408314801076, + "learning_rate": 5.608628475886001e-07, + "loss": 0.0002, + "step": 12454 + }, + { + "epoch": 5.861176470588235, + "grad_norm": 0.06302901694995543, + "learning_rate": 5.570819786081704e-07, + "loss": 0.009, + "step": 12455 + }, + { + "epoch": 5.86164705882353, + "grad_norm": 0.09460417902581343, + "learning_rate": 5.533138727859543e-07, + "loss": 0.016, + "step": 12456 + }, + { + "epoch": 5.862117647058824, + "grad_norm": 0.12209242723654594, + "learning_rate": 5.495585304437611e-07, + "loss": 0.0104, + "step": 12457 + }, + { + "epoch": 5.862588235294117, + "grad_norm": 0.0014973092338853179, + "learning_rate": 5.458159519023175e-07, + "loss": 0.0, + "step": 12458 + }, + { + "epoch": 5.863058823529411, + "grad_norm": 0.07038343596298502, + "learning_rate": 5.42086137481268e-07, + "loss": 0.0082, + "step": 12459 + }, + { + "epoch": 5.863529411764706, + "grad_norm": 0.0716847339102754, + "learning_rate": 5.383690874991408e-07, + "loss": 0.0055, + "step": 12460 + }, + { + "epoch": 5.864, + "grad_norm": 0.1400271149597515, + "learning_rate": 5.34664802273399e-07, + "loss": 0.0326, + "step": 12461 + }, + { + "epoch": 5.864470588235294, + "grad_norm": 0.14722453944944486, + "learning_rate": 5.30973282120406e-07, + "loss": 0.0194, + "step": 12462 + }, + { + "epoch": 5.864941176470588, + "grad_norm": 0.048790226558825686, + "learning_rate": 5.272945273554429e-07, + "loss": 0.0021, + "step": 12463 + }, + { + "epoch": 5.8654117647058825, + "grad_norm": 0.05769507709731675, + "learning_rate": 5.236285382926752e-07, + "loss": 0.0041, + "step": 12464 + }, + { + "epoch": 5.865882352941177, + "grad_norm": 3.9401235001295846e-05, + "learning_rate": 5.19975315245219e-07, + "loss": 0.0, + "step": 12465 + }, + { + "epoch": 5.866352941176471, + "grad_norm": 0.08187939609728462, + "learning_rate": 5.163348585250748e-07, + "loss": 0.0126, + "step": 12466 + }, + { + "epoch": 5.866823529411764, + "grad_norm": 0.11083912269247588, + "learning_rate": 5.127071684431272e-07, + "loss": 0.0143, + "step": 12467 + }, + { + "epoch": 5.8672941176470585, + "grad_norm": 0.05854816272355994, + "learning_rate": 5.090922453092283e-07, + "loss": 0.0077, + "step": 12468 + }, + { + "epoch": 5.867764705882353, + "grad_norm": 8.996305367590734e-05, + "learning_rate": 5.054900894320979e-07, + "loss": 0.0, + "step": 12469 + }, + { + "epoch": 5.868235294117647, + "grad_norm": 0.0797031384374495, + "learning_rate": 5.019007011193732e-07, + "loss": 0.0128, + "step": 12470 + }, + { + "epoch": 5.868705882352941, + "grad_norm": 0.26932323499861677, + "learning_rate": 4.983240806776257e-07, + "loss": 0.0409, + "step": 12471 + }, + { + "epoch": 5.869176470588235, + "grad_norm": 0.07643512652139728, + "learning_rate": 4.947602284122942e-07, + "loss": 0.0068, + "step": 12472 + }, + { + "epoch": 5.86964705882353, + "grad_norm": 0.15230877434239926, + "learning_rate": 4.912091446277688e-07, + "loss": 0.0519, + "step": 12473 + }, + { + "epoch": 5.870117647058824, + "grad_norm": 0.1542126982666121, + "learning_rate": 4.876708296273068e-07, + "loss": 0.0396, + "step": 12474 + }, + { + "epoch": 5.870588235294118, + "grad_norm": 0.1345420437344985, + "learning_rate": 4.841452837131166e-07, + "loss": 0.0058, + "step": 12475 + }, + { + "epoch": 5.871058823529411, + "grad_norm": 3.0138947011536198e-05, + "learning_rate": 4.806325071862905e-07, + "loss": 0.0, + "step": 12476 + }, + { + "epoch": 5.871529411764706, + "grad_norm": 3.317128495696372e-05, + "learning_rate": 4.771325003468385e-07, + "loss": 0.0, + "step": 12477 + }, + { + "epoch": 5.872, + "grad_norm": 0.02157132173085822, + "learning_rate": 4.736452634936716e-07, + "loss": 0.0009, + "step": 12478 + }, + { + "epoch": 5.872470588235294, + "grad_norm": 9.939588461213456e-06, + "learning_rate": 4.701707969246182e-07, + "loss": 0.0, + "step": 12479 + }, + { + "epoch": 5.872941176470588, + "grad_norm": 1.3715931279535492e-05, + "learning_rate": 4.6670910093642435e-07, + "loss": 0.0, + "step": 12480 + }, + { + "epoch": 5.8734117647058826, + "grad_norm": 0.16967522298274004, + "learning_rate": 4.6326017582473674e-07, + "loss": 0.0368, + "step": 12481 + }, + { + "epoch": 5.873882352941177, + "grad_norm": 0.08441194591734469, + "learning_rate": 4.5982402188408653e-07, + "loss": 0.0115, + "step": 12482 + }, + { + "epoch": 5.87435294117647, + "grad_norm": 0.03171903707523113, + "learning_rate": 4.5640063940797223e-07, + "loss": 0.0016, + "step": 12483 + }, + { + "epoch": 5.874823529411764, + "grad_norm": 0.08057579703395992, + "learning_rate": 4.529900286887433e-07, + "loss": 0.0122, + "step": 12484 + }, + { + "epoch": 5.875294117647059, + "grad_norm": 7.103643412644947e-05, + "learning_rate": 4.4959219001768355e-07, + "loss": 0.0, + "step": 12485 + }, + { + "epoch": 5.875764705882353, + "grad_norm": 0.13065789740817033, + "learning_rate": 4.4620712368501065e-07, + "loss": 0.0281, + "step": 12486 + }, + { + "epoch": 5.876235294117647, + "grad_norm": 8.216889328382877e-06, + "learning_rate": 4.428348299797768e-07, + "loss": 0.0, + "step": 12487 + }, + { + "epoch": 5.876705882352941, + "grad_norm": 0.0910813884626965, + "learning_rate": 4.394753091900516e-07, + "loss": 0.0178, + "step": 12488 + }, + { + "epoch": 5.8771764705882354, + "grad_norm": 0.14196367184996692, + "learning_rate": 4.361285616027055e-07, + "loss": 0.0314, + "step": 12489 + }, + { + "epoch": 5.87764705882353, + "grad_norm": 4.276585011916541e-05, + "learning_rate": 4.327945875035932e-07, + "loss": 0.0, + "step": 12490 + }, + { + "epoch": 5.878117647058824, + "grad_norm": 0.11668567861501628, + "learning_rate": 4.2947338717743694e-07, + "loss": 0.0204, + "step": 12491 + }, + { + "epoch": 5.878588235294117, + "grad_norm": 7.446438528451654e-07, + "learning_rate": 4.2616496090790983e-07, + "loss": 0.0, + "step": 12492 + }, + { + "epoch": 5.8790588235294114, + "grad_norm": 0.12215203540601788, + "learning_rate": 4.228693089775526e-07, + "loss": 0.0238, + "step": 12493 + }, + { + "epoch": 5.879529411764706, + "grad_norm": 0.1694493630879718, + "learning_rate": 4.195864316678066e-07, + "loss": 0.0509, + "step": 12494 + }, + { + "epoch": 5.88, + "grad_norm": 0.1532590892546947, + "learning_rate": 4.16316329259081e-07, + "loss": 0.0348, + "step": 12495 + }, + { + "epoch": 5.880470588235294, + "grad_norm": 0.2011084397801173, + "learning_rate": 4.130590020306523e-07, + "loss": 0.0669, + "step": 12496 + }, + { + "epoch": 5.880941176470588, + "grad_norm": 0.1192938043750281, + "learning_rate": 4.0981445026068147e-07, + "loss": 0.0369, + "step": 12497 + }, + { + "epoch": 5.881411764705883, + "grad_norm": 0.051540982396694184, + "learning_rate": 4.0658267422631343e-07, + "loss": 0.0065, + "step": 12498 + }, + { + "epoch": 5.881882352941177, + "grad_norm": 0.08121419155413893, + "learning_rate": 4.033636742035273e-07, + "loss": 0.0147, + "step": 12499 + }, + { + "epoch": 5.882352941176471, + "grad_norm": 9.093810395111837e-06, + "learning_rate": 4.0015745046725336e-07, + "loss": 0.0, + "step": 12500 + }, + { + "epoch": 5.882823529411764, + "grad_norm": 0.0723031458933078, + "learning_rate": 3.9696400329132236e-07, + "loss": 0.0057, + "step": 12501 + }, + { + "epoch": 5.883294117647059, + "grad_norm": 1.0755429420391484e-05, + "learning_rate": 3.9378333294844967e-07, + "loss": 0.0, + "step": 12502 + }, + { + "epoch": 5.883764705882353, + "grad_norm": 0.12022918505785848, + "learning_rate": 3.906154397103012e-07, + "loss": 0.0285, + "step": 12503 + }, + { + "epoch": 5.884235294117647, + "grad_norm": 0.00013640983514228848, + "learning_rate": 3.874603238474272e-07, + "loss": 0.0, + "step": 12504 + }, + { + "epoch": 5.884705882352941, + "grad_norm": 0.02996143862898737, + "learning_rate": 3.8431798562927887e-07, + "loss": 0.0027, + "step": 12505 + }, + { + "epoch": 5.8851764705882355, + "grad_norm": 0.09125621750501524, + "learning_rate": 3.8118842532424145e-07, + "loss": 0.0068, + "step": 12506 + }, + { + "epoch": 5.88564705882353, + "grad_norm": 0.14217276802830425, + "learning_rate": 3.780716431995845e-07, + "loss": 0.0338, + "step": 12507 + }, + { + "epoch": 5.886117647058824, + "grad_norm": 0.06355514815892484, + "learning_rate": 3.7496763952151177e-07, + "loss": 0.0127, + "step": 12508 + }, + { + "epoch": 5.886588235294118, + "grad_norm": 0.048780117347351416, + "learning_rate": 3.7187641455509455e-07, + "loss": 0.0027, + "step": 12509 + }, + { + "epoch": 5.8870588235294115, + "grad_norm": 1.1420818536714917e-05, + "learning_rate": 3.687979685643549e-07, + "loss": 0.0, + "step": 12510 + }, + { + "epoch": 5.887529411764706, + "grad_norm": 2.1899135885511412e-05, + "learning_rate": 3.6573230181221604e-07, + "loss": 0.0, + "step": 12511 + }, + { + "epoch": 5.888, + "grad_norm": 0.11259994956919735, + "learning_rate": 3.62679414560485e-07, + "loss": 0.017, + "step": 12512 + }, + { + "epoch": 5.888470588235294, + "grad_norm": 0.09223325546701219, + "learning_rate": 3.596393070699033e-07, + "loss": 0.0177, + "step": 12513 + }, + { + "epoch": 5.888941176470588, + "grad_norm": 0.11439950582237748, + "learning_rate": 3.566119796000966e-07, + "loss": 0.0214, + "step": 12514 + }, + { + "epoch": 5.889411764705883, + "grad_norm": 0.12822593042352873, + "learning_rate": 3.535974324096413e-07, + "loss": 0.0277, + "step": 12515 + }, + { + "epoch": 5.889882352941177, + "grad_norm": 0.07135172346459563, + "learning_rate": 3.5059566575596497e-07, + "loss": 0.0082, + "step": 12516 + }, + { + "epoch": 5.89035294117647, + "grad_norm": 0.14386773669454464, + "learning_rate": 3.4760667989544575e-07, + "loss": 0.0384, + "step": 12517 + }, + { + "epoch": 5.890823529411764, + "grad_norm": 0.11089091025248053, + "learning_rate": 3.446304750833462e-07, + "loss": 0.0219, + "step": 12518 + }, + { + "epoch": 5.891294117647059, + "grad_norm": 0.06303370356287294, + "learning_rate": 3.416670515738629e-07, + "loss": 0.0036, + "step": 12519 + }, + { + "epoch": 5.891764705882353, + "grad_norm": 0.12418762503294171, + "learning_rate": 3.387164096200934e-07, + "loss": 0.0253, + "step": 12520 + }, + { + "epoch": 5.892235294117647, + "grad_norm": 0.07111393893106152, + "learning_rate": 3.3577854947403615e-07, + "loss": 0.0057, + "step": 12521 + }, + { + "epoch": 5.892705882352941, + "grad_norm": 0.10307102448360886, + "learning_rate": 3.3285347138657383e-07, + "loss": 0.0209, + "step": 12522 + }, + { + "epoch": 5.8931764705882355, + "grad_norm": 0.037660980434601275, + "learning_rate": 3.2994117560753986e-07, + "loss": 0.0028, + "step": 12523 + }, + { + "epoch": 5.89364705882353, + "grad_norm": 3.8285946642065975e-05, + "learning_rate": 3.2704166238565197e-07, + "loss": 0.0, + "step": 12524 + }, + { + "epoch": 5.894117647058824, + "grad_norm": 0.04667395574529366, + "learning_rate": 3.241549319685621e-07, + "loss": 0.0033, + "step": 12525 + }, + { + "epoch": 5.894588235294117, + "grad_norm": 5.9287182610069446e-06, + "learning_rate": 3.212809846027897e-07, + "loss": 0.0, + "step": 12526 + }, + { + "epoch": 5.8950588235294115, + "grad_norm": 0.1212213866036892, + "learning_rate": 3.1841982053378846e-07, + "loss": 0.0138, + "step": 12527 + }, + { + "epoch": 5.895529411764706, + "grad_norm": 0.09820941929680865, + "learning_rate": 3.155714400059128e-07, + "loss": 0.016, + "step": 12528 + }, + { + "epoch": 5.896, + "grad_norm": 0.10254698424043285, + "learning_rate": 3.127358432624516e-07, + "loss": 0.0204, + "step": 12529 + }, + { + "epoch": 5.896470588235294, + "grad_norm": 0.060271109043092824, + "learning_rate": 3.099130305455444e-07, + "loss": 0.0092, + "step": 12530 + }, + { + "epoch": 5.896941176470588, + "grad_norm": 0.07843573213757417, + "learning_rate": 3.0710300209629836e-07, + "loss": 0.0085, + "step": 12531 + }, + { + "epoch": 5.897411764705883, + "grad_norm": 0.08661107689943351, + "learning_rate": 3.043057581547048e-07, + "loss": 0.0176, + "step": 12532 + }, + { + "epoch": 5.897882352941177, + "grad_norm": 0.06367383822484819, + "learning_rate": 3.0152129895963937e-07, + "loss": 0.0065, + "step": 12533 + }, + { + "epoch": 5.898352941176471, + "grad_norm": 0.13904361854290662, + "learning_rate": 2.987496247489285e-07, + "loss": 0.0375, + "step": 12534 + }, + { + "epoch": 5.898823529411764, + "grad_norm": 0.10433285662003035, + "learning_rate": 2.959907357592661e-07, + "loss": 0.0072, + "step": 12535 + }, + { + "epoch": 5.899294117647059, + "grad_norm": 0.06920101645133163, + "learning_rate": 2.9324463222629715e-07, + "loss": 0.0054, + "step": 12536 + }, + { + "epoch": 5.899764705882353, + "grad_norm": 3.2094006577457825e-05, + "learning_rate": 2.9051131438455055e-07, + "loss": 0.0, + "step": 12537 + }, + { + "epoch": 5.900235294117647, + "grad_norm": 0.14510466133312921, + "learning_rate": 2.8779078246745635e-07, + "loss": 0.0328, + "step": 12538 + }, + { + "epoch": 5.900705882352941, + "grad_norm": 0.16854110872491948, + "learning_rate": 2.85083036707362e-07, + "loss": 0.0642, + "step": 12539 + }, + { + "epoch": 5.9011764705882355, + "grad_norm": 6.620869133342877e-06, + "learning_rate": 2.8238807733551583e-07, + "loss": 0.0, + "step": 12540 + }, + { + "epoch": 5.90164705882353, + "grad_norm": 0.035830314689959016, + "learning_rate": 2.797059045821004e-07, + "loss": 0.0019, + "step": 12541 + }, + { + "epoch": 5.902117647058824, + "grad_norm": 0.09629288027061171, + "learning_rate": 2.7703651867616583e-07, + "loss": 0.0145, + "step": 12542 + }, + { + "epoch": 5.902588235294118, + "grad_norm": 0.1099329517249626, + "learning_rate": 2.743799198457131e-07, + "loss": 0.0277, + "step": 12543 + }, + { + "epoch": 5.9030588235294115, + "grad_norm": 0.1385189816662821, + "learning_rate": 2.71736108317594e-07, + "loss": 0.0159, + "step": 12544 + }, + { + "epoch": 5.903529411764706, + "grad_norm": 0.12557800483462278, + "learning_rate": 2.6910508431764457e-07, + "loss": 0.0186, + "step": 12545 + }, + { + "epoch": 5.904, + "grad_norm": 0.43457496917067784, + "learning_rate": 2.6648684807053513e-07, + "loss": 0.0863, + "step": 12546 + }, + { + "epoch": 5.904470588235294, + "grad_norm": 0.14429901871821058, + "learning_rate": 2.6388139979988677e-07, + "loss": 0.0425, + "step": 12547 + }, + { + "epoch": 5.904941176470588, + "grad_norm": 0.024924407552392322, + "learning_rate": 2.612887397282215e-07, + "loss": 0.0011, + "step": 12548 + }, + { + "epoch": 5.905411764705883, + "grad_norm": 0.09388475894961927, + "learning_rate": 2.5870886807696225e-07, + "loss": 0.0144, + "step": 12549 + }, + { + "epoch": 5.905882352941177, + "grad_norm": 0.002883775469580297, + "learning_rate": 2.5614178506644934e-07, + "loss": 0.0, + "step": 12550 + }, + { + "epoch": 5.90635294117647, + "grad_norm": 0.10543885603616641, + "learning_rate": 2.5358749091590746e-07, + "loss": 0.0133, + "step": 12551 + }, + { + "epoch": 5.906823529411764, + "grad_norm": 5.691077972305867e-06, + "learning_rate": 2.5104598584349543e-07, + "loss": 0.0, + "step": 12552 + }, + { + "epoch": 5.907294117647059, + "grad_norm": 0.06875282154623824, + "learning_rate": 2.4851727006627297e-07, + "loss": 0.0105, + "step": 12553 + }, + { + "epoch": 5.907764705882353, + "grad_norm": 2.7411715319456855e-05, + "learning_rate": 2.460013438002173e-07, + "loss": 0.0, + "step": 12554 + }, + { + "epoch": 5.908235294117647, + "grad_norm": 0.12510734300649545, + "learning_rate": 2.434982072601732e-07, + "loss": 0.0225, + "step": 12555 + }, + { + "epoch": 5.908705882352941, + "grad_norm": 0.13304911189843366, + "learning_rate": 2.4100786065991975e-07, + "loss": 0.018, + "step": 12556 + }, + { + "epoch": 5.9091764705882355, + "grad_norm": 0.09495804287222268, + "learning_rate": 2.3853030421218667e-07, + "loss": 0.0147, + "step": 12557 + }, + { + "epoch": 5.90964705882353, + "grad_norm": 0.1595212870294914, + "learning_rate": 2.3606553812852146e-07, + "loss": 0.0245, + "step": 12558 + }, + { + "epoch": 5.910117647058824, + "grad_norm": 0.15965557324581067, + "learning_rate": 2.3361356261945575e-07, + "loss": 0.0505, + "step": 12559 + }, + { + "epoch": 5.910588235294117, + "grad_norm": 0.07741007391954147, + "learning_rate": 2.3117437789438865e-07, + "loss": 0.0125, + "step": 12560 + }, + { + "epoch": 5.9110588235294115, + "grad_norm": 5.171632910923388e-06, + "learning_rate": 2.2874798416163686e-07, + "loss": 0.0, + "step": 12561 + }, + { + "epoch": 5.911529411764706, + "grad_norm": 0.12031196943815221, + "learning_rate": 2.2633438162843465e-07, + "loss": 0.0249, + "step": 12562 + }, + { + "epoch": 5.912, + "grad_norm": 0.11903150709115126, + "learning_rate": 2.239335705009171e-07, + "loss": 0.0256, + "step": 12563 + }, + { + "epoch": 5.912470588235294, + "grad_norm": 0.08893703604895402, + "learning_rate": 2.215455509841202e-07, + "loss": 0.0189, + "step": 12564 + }, + { + "epoch": 5.912941176470588, + "grad_norm": 0.15139134535517754, + "learning_rate": 2.1917032328199747e-07, + "loss": 0.0442, + "step": 12565 + }, + { + "epoch": 5.913411764705883, + "grad_norm": 0.056274584495772946, + "learning_rate": 2.1680788759738664e-07, + "loss": 0.006, + "step": 12566 + }, + { + "epoch": 5.913882352941177, + "grad_norm": 0.09079446884805162, + "learning_rate": 2.144582441320597e-07, + "loss": 0.0115, + "step": 12567 + }, + { + "epoch": 5.914352941176471, + "grad_norm": 0.11407654983164912, + "learning_rate": 2.1212139308670607e-07, + "loss": 0.0231, + "step": 12568 + }, + { + "epoch": 5.914823529411764, + "grad_norm": 0.2199168720303507, + "learning_rate": 2.0979733466088277e-07, + "loss": 0.0846, + "step": 12569 + }, + { + "epoch": 5.915294117647059, + "grad_norm": 0.08003808801173216, + "learning_rate": 2.0748606905309774e-07, + "loss": 0.0115, + "step": 12570 + }, + { + "epoch": 5.915764705882353, + "grad_norm": 0.07685112655388414, + "learning_rate": 2.051875964607097e-07, + "loss": 0.0037, + "step": 12571 + }, + { + "epoch": 5.916235294117647, + "grad_norm": 0.09906763777508777, + "learning_rate": 2.0290191708006165e-07, + "loss": 0.0157, + "step": 12572 + }, + { + "epoch": 5.916705882352941, + "grad_norm": 0.024802240484808023, + "learning_rate": 2.0062903110633077e-07, + "loss": 0.0016, + "step": 12573 + }, + { + "epoch": 5.9171764705882355, + "grad_norm": 0.10054739983452077, + "learning_rate": 1.9836893873364513e-07, + "loss": 0.0211, + "step": 12574 + }, + { + "epoch": 5.91764705882353, + "grad_norm": 0.1352118170468906, + "learning_rate": 1.9612164015501696e-07, + "loss": 0.0393, + "step": 12575 + }, + { + "epoch": 5.918117647058823, + "grad_norm": 0.07738130436718238, + "learning_rate": 1.9388713556239277e-07, + "loss": 0.0068, + "step": 12576 + }, + { + "epoch": 5.918588235294117, + "grad_norm": 0.14136493736109376, + "learning_rate": 1.9166542514658656e-07, + "loss": 0.0275, + "step": 12577 + }, + { + "epoch": 5.9190588235294115, + "grad_norm": 0.12470919555323895, + "learning_rate": 1.8945650909737985e-07, + "loss": 0.0234, + "step": 12578 + }, + { + "epoch": 5.919529411764706, + "grad_norm": 0.012533859736395614, + "learning_rate": 1.8726038760338847e-07, + "loss": 0.0005, + "step": 12579 + }, + { + "epoch": 5.92, + "grad_norm": 0.04288220519360926, + "learning_rate": 1.8507706085219566e-07, + "loss": 0.0021, + "step": 12580 + }, + { + "epoch": 5.920470588235294, + "grad_norm": 0.09172917335542755, + "learning_rate": 1.8290652903025228e-07, + "loss": 0.0166, + "step": 12581 + }, + { + "epoch": 5.920941176470588, + "grad_norm": 0.00015852877336613118, + "learning_rate": 1.8074879232296002e-07, + "loss": 0.0, + "step": 12582 + }, + { + "epoch": 5.921411764705883, + "grad_norm": 0.05282416099709311, + "learning_rate": 1.7860385091455488e-07, + "loss": 0.004, + "step": 12583 + }, + { + "epoch": 5.921882352941177, + "grad_norm": 0.08557480825974444, + "learning_rate": 1.764717049882569e-07, + "loss": 0.0067, + "step": 12584 + }, + { + "epoch": 5.92235294117647, + "grad_norm": 0.1585814717686222, + "learning_rate": 1.743523547261705e-07, + "loss": 0.0472, + "step": 12585 + }, + { + "epoch": 5.922823529411764, + "grad_norm": 0.16230895707512072, + "learning_rate": 1.722458003092675e-07, + "loss": 0.0452, + "step": 12586 + }, + { + "epoch": 5.923294117647059, + "grad_norm": 0.13799662868556653, + "learning_rate": 1.701520419174873e-07, + "loss": 0.0195, + "step": 12587 + }, + { + "epoch": 5.923764705882353, + "grad_norm": 0.043456150044451206, + "learning_rate": 1.680710797296203e-07, + "loss": 0.0036, + "step": 12588 + }, + { + "epoch": 5.924235294117647, + "grad_norm": 0.023955058510479545, + "learning_rate": 1.660029139234076e-07, + "loss": 0.0012, + "step": 12589 + }, + { + "epoch": 5.924705882352941, + "grad_norm": 0.04637050748221147, + "learning_rate": 1.639475446754912e-07, + "loss": 0.0032, + "step": 12590 + }, + { + "epoch": 5.9251764705882355, + "grad_norm": 0.14617259568652735, + "learning_rate": 1.6190497216138076e-07, + "loss": 0.0322, + "step": 12591 + }, + { + "epoch": 5.92564705882353, + "grad_norm": 0.12088561263069018, + "learning_rate": 1.598751965555367e-07, + "loss": 0.0288, + "step": 12592 + }, + { + "epoch": 5.926117647058824, + "grad_norm": 0.18459954626825437, + "learning_rate": 1.5785821803132037e-07, + "loss": 0.0469, + "step": 12593 + }, + { + "epoch": 5.926588235294117, + "grad_norm": 3.815487407506894e-06, + "learning_rate": 1.558540367609773e-07, + "loss": 0.0, + "step": 12594 + }, + { + "epoch": 5.9270588235294115, + "grad_norm": 0.11939730446969582, + "learning_rate": 1.5386265291568723e-07, + "loss": 0.0217, + "step": 12595 + }, + { + "epoch": 5.927529411764706, + "grad_norm": 0.04231916768772083, + "learning_rate": 1.5188406666551412e-07, + "loss": 0.0015, + "step": 12596 + }, + { + "epoch": 5.928, + "grad_norm": 0.10876434377228376, + "learning_rate": 1.4991827817943947e-07, + "loss": 0.0275, + "step": 12597 + }, + { + "epoch": 5.928470588235294, + "grad_norm": 0.07767467851664869, + "learning_rate": 1.479652876253623e-07, + "loss": 0.0096, + "step": 12598 + }, + { + "epoch": 5.928941176470588, + "grad_norm": 0.05904778701846131, + "learning_rate": 1.4602509517006588e-07, + "loss": 0.0024, + "step": 12599 + }, + { + "epoch": 5.929411764705883, + "grad_norm": 9.819746876368311e-05, + "learning_rate": 1.4409770097926765e-07, + "loss": 0.0, + "step": 12600 + }, + { + "epoch": 5.929882352941177, + "grad_norm": 0.12772711236189505, + "learning_rate": 1.421831052175526e-07, + "loss": 0.0289, + "step": 12601 + }, + { + "epoch": 5.930352941176471, + "grad_norm": 0.20752280996751765, + "learning_rate": 1.4028130804843995e-07, + "loss": 0.0794, + "step": 12602 + }, + { + "epoch": 5.930823529411764, + "grad_norm": 0.15236472549373517, + "learning_rate": 1.3839230963436642e-07, + "loss": 0.0529, + "step": 12603 + }, + { + "epoch": 5.931294117647059, + "grad_norm": 0.10851772709567087, + "learning_rate": 1.3651611013666964e-07, + "loss": 0.0166, + "step": 12604 + }, + { + "epoch": 5.931764705882353, + "grad_norm": 2.2784970384776125e-05, + "learning_rate": 1.346527097155714e-07, + "loss": 0.0, + "step": 12605 + }, + { + "epoch": 5.932235294117647, + "grad_norm": 0.06203547723453308, + "learning_rate": 1.3280210853019447e-07, + "loss": 0.0057, + "step": 12606 + }, + { + "epoch": 5.932705882352941, + "grad_norm": 0.05463421218779208, + "learning_rate": 1.3096430673862901e-07, + "loss": 0.0035, + "step": 12607 + }, + { + "epoch": 5.9331764705882355, + "grad_norm": 0.13353590674547924, + "learning_rate": 1.2913930449781619e-07, + "loss": 0.0277, + "step": 12608 + }, + { + "epoch": 5.93364705882353, + "grad_norm": 0.07417216470558961, + "learning_rate": 1.2732710196361463e-07, + "loss": 0.0088, + "step": 12609 + }, + { + "epoch": 5.934117647058823, + "grad_norm": 1.8807362154725437e-05, + "learning_rate": 1.2552769929081717e-07, + "loss": 0.0, + "step": 12610 + }, + { + "epoch": 5.934588235294117, + "grad_norm": 0.06820881222369207, + "learning_rate": 1.2374109663306763e-07, + "loss": 0.0094, + "step": 12611 + }, + { + "epoch": 5.9350588235294115, + "grad_norm": 1.597267962535584e-05, + "learning_rate": 1.2196729414296058e-07, + "loss": 0.0, + "step": 12612 + }, + { + "epoch": 5.935529411764706, + "grad_norm": 0.18027583065798983, + "learning_rate": 1.2020629197200814e-07, + "loss": 0.0516, + "step": 12613 + }, + { + "epoch": 5.936, + "grad_norm": 0.09535489323444, + "learning_rate": 1.1845809027059006e-07, + "loss": 0.0219, + "step": 12614 + }, + { + "epoch": 5.936470588235294, + "grad_norm": 0.10623414919906768, + "learning_rate": 1.1672268918800353e-07, + "loss": 0.0173, + "step": 12615 + }, + { + "epoch": 5.936941176470588, + "grad_norm": 0.09781237654865714, + "learning_rate": 1.1500008887249668e-07, + "loss": 0.0194, + "step": 12616 + }, + { + "epoch": 5.937411764705883, + "grad_norm": 0.04756336561097821, + "learning_rate": 1.1329028947115181e-07, + "loss": 0.0021, + "step": 12617 + }, + { + "epoch": 5.937882352941177, + "grad_norm": 4.604439695949535e-06, + "learning_rate": 1.1159329113000215e-07, + "loss": 0.0, + "step": 12618 + }, + { + "epoch": 5.93835294117647, + "grad_norm": 0.08331064217713918, + "learning_rate": 1.0990909399398173e-07, + "loss": 0.0129, + "step": 12619 + }, + { + "epoch": 5.938823529411764, + "grad_norm": 8.604630418874224e-06, + "learning_rate": 1.0823769820694218e-07, + "loss": 0.0, + "step": 12620 + }, + { + "epoch": 5.939294117647059, + "grad_norm": 0.1304621328778094, + "learning_rate": 1.0657910391161928e-07, + "loss": 0.0329, + "step": 12621 + }, + { + "epoch": 5.939764705882353, + "grad_norm": 0.18413459498775786, + "learning_rate": 1.0493331124964976e-07, + "loss": 0.0527, + "step": 12622 + }, + { + "epoch": 5.940235294117647, + "grad_norm": 0.0004830872686581653, + "learning_rate": 1.0330032036162117e-07, + "loss": 0.0, + "step": 12623 + }, + { + "epoch": 5.940705882352941, + "grad_norm": 1.021705070462798e-05, + "learning_rate": 1.0168013138697196e-07, + "loss": 0.0, + "step": 12624 + }, + { + "epoch": 5.9411764705882355, + "grad_norm": 6.195090269028357e-05, + "learning_rate": 1.0007274446409141e-07, + "loss": 0.0, + "step": 12625 + }, + { + "epoch": 5.94164705882353, + "grad_norm": 0.13435979997393377, + "learning_rate": 9.847815973025308e-08, + "loss": 0.0301, + "step": 12626 + }, + { + "epoch": 5.942117647058824, + "grad_norm": 0.09179180845274498, + "learning_rate": 9.689637732164802e-08, + "loss": 0.0166, + "step": 12627 + }, + { + "epoch": 5.942588235294117, + "grad_norm": 0.08661815206802781, + "learning_rate": 9.532739737336814e-08, + "loss": 0.0115, + "step": 12628 + }, + { + "epoch": 5.9430588235294115, + "grad_norm": 1.6063704265781244e-05, + "learning_rate": 9.377122001938963e-08, + "loss": 0.0, + "step": 12629 + }, + { + "epoch": 5.943529411764706, + "grad_norm": 0.11245455828890477, + "learning_rate": 9.222784539263949e-08, + "loss": 0.0105, + "step": 12630 + }, + { + "epoch": 5.944, + "grad_norm": 0.1118732997802094, + "learning_rate": 9.069727362492896e-08, + "loss": 0.029, + "step": 12631 + }, + { + "epoch": 5.944470588235294, + "grad_norm": 0.1275339963814382, + "learning_rate": 8.917950484697012e-08, + "loss": 0.0291, + "step": 12632 + }, + { + "epoch": 5.944941176470588, + "grad_norm": 0.06591006992422055, + "learning_rate": 8.767453918839263e-08, + "loss": 0.0069, + "step": 12633 + }, + { + "epoch": 5.945411764705883, + "grad_norm": 0.10346106034431772, + "learning_rate": 8.618237677772699e-08, + "loss": 0.0214, + "step": 12634 + }, + { + "epoch": 5.945882352941177, + "grad_norm": 0.2396670969292464, + "learning_rate": 8.470301774240463e-08, + "loss": 0.0811, + "step": 12635 + }, + { + "epoch": 5.946352941176471, + "grad_norm": 3.6922589936308693e-06, + "learning_rate": 8.323646220877444e-08, + "loss": 0.0, + "step": 12636 + }, + { + "epoch": 5.946823529411764, + "grad_norm": 0.12939675098130254, + "learning_rate": 8.178271030210292e-08, + "loss": 0.03, + "step": 12637 + }, + { + "epoch": 5.947294117647059, + "grad_norm": 0.11083257696525593, + "learning_rate": 8.034176214650745e-08, + "loss": 0.0247, + "step": 12638 + }, + { + "epoch": 5.947764705882353, + "grad_norm": 0.11120333204394191, + "learning_rate": 7.891361786508953e-08, + "loss": 0.0185, + "step": 12639 + }, + { + "epoch": 5.948235294117647, + "grad_norm": 0.16334807801940573, + "learning_rate": 7.749827757981831e-08, + "loss": 0.0426, + "step": 12640 + }, + { + "epoch": 5.948705882352941, + "grad_norm": 0.07703730284906757, + "learning_rate": 7.609574141154706e-08, + "loss": 0.0108, + "step": 12641 + }, + { + "epoch": 5.9491764705882355, + "grad_norm": 0.061577092061345956, + "learning_rate": 7.470600948006333e-08, + "loss": 0.0075, + "step": 12642 + }, + { + "epoch": 5.94964705882353, + "grad_norm": 0.2159141117857497, + "learning_rate": 7.332908190408882e-08, + "loss": 0.0523, + "step": 12643 + }, + { + "epoch": 5.950117647058823, + "grad_norm": 0.1363550353193975, + "learning_rate": 7.196495880117947e-08, + "loss": 0.0137, + "step": 12644 + }, + { + "epoch": 5.950588235294117, + "grad_norm": 0.1639299135039193, + "learning_rate": 7.061364028785876e-08, + "loss": 0.0623, + "step": 12645 + }, + { + "epoch": 5.9510588235294115, + "grad_norm": 0.15100431707162296, + "learning_rate": 6.927512647953437e-08, + "loss": 0.0384, + "step": 12646 + }, + { + "epoch": 5.951529411764706, + "grad_norm": 0.07677675226899403, + "learning_rate": 6.794941749053151e-08, + "loss": 0.0096, + "step": 12647 + }, + { + "epoch": 5.952, + "grad_norm": 0.07207470585575229, + "learning_rate": 6.663651343404297e-08, + "loss": 0.0118, + "step": 12648 + }, + { + "epoch": 5.952470588235294, + "grad_norm": 3.747375918832959e-05, + "learning_rate": 6.533641442222904e-08, + "loss": 0.0, + "step": 12649 + }, + { + "epoch": 5.952941176470588, + "grad_norm": 2.4804865013442553e-05, + "learning_rate": 6.40491205661009e-08, + "loss": 0.0, + "step": 12650 + }, + { + "epoch": 5.953411764705883, + "grad_norm": 0.15175154242268785, + "learning_rate": 6.277463197560395e-08, + "loss": 0.0305, + "step": 12651 + }, + { + "epoch": 5.953882352941177, + "grad_norm": 7.532936558759039e-06, + "learning_rate": 6.151294875960111e-08, + "loss": 0.0, + "step": 12652 + }, + { + "epoch": 5.95435294117647, + "grad_norm": 0.12246758353620416, + "learning_rate": 6.026407102583952e-08, + "loss": 0.0172, + "step": 12653 + }, + { + "epoch": 5.954823529411764, + "grad_norm": 0.012969305086087627, + "learning_rate": 5.902799888096721e-08, + "loss": 0.0005, + "step": 12654 + }, + { + "epoch": 5.955294117647059, + "grad_norm": 0.1488509556502514, + "learning_rate": 5.780473243056638e-08, + "loss": 0.0261, + "step": 12655 + }, + { + "epoch": 5.955764705882353, + "grad_norm": 0.12215425165782093, + "learning_rate": 5.659427177910347e-08, + "loss": 0.0204, + "step": 12656 + }, + { + "epoch": 5.956235294117647, + "grad_norm": 0.11245466676933918, + "learning_rate": 5.539661702996245e-08, + "loss": 0.0143, + "step": 12657 + }, + { + "epoch": 5.956705882352941, + "grad_norm": 7.303813769644167e-06, + "learning_rate": 5.421176828541152e-08, + "loss": 0.0, + "step": 12658 + }, + { + "epoch": 5.9571764705882355, + "grad_norm": 0.09172630285379821, + "learning_rate": 5.30397256466697e-08, + "loss": 0.0133, + "step": 12659 + }, + { + "epoch": 5.95764705882353, + "grad_norm": 0.15737933168755333, + "learning_rate": 5.1880489213806944e-08, + "loss": 0.0336, + "step": 12660 + }, + { + "epoch": 5.958117647058824, + "grad_norm": 0.04968130743201021, + "learning_rate": 5.0734059085860704e-08, + "loss": 0.0049, + "step": 12661 + }, + { + "epoch": 5.958588235294117, + "grad_norm": 0.15981365854049218, + "learning_rate": 4.9600435360702685e-08, + "loss": 0.0489, + "step": 12662 + }, + { + "epoch": 5.9590588235294115, + "grad_norm": 2.5010300963222174e-05, + "learning_rate": 4.847961813517209e-08, + "loss": 0.0, + "step": 12663 + }, + { + "epoch": 5.959529411764706, + "grad_norm": 0.07615861797214836, + "learning_rate": 4.737160750500901e-08, + "loss": 0.0074, + "step": 12664 + }, + { + "epoch": 5.96, + "grad_norm": 0.05736806128760697, + "learning_rate": 4.6276403564804444e-08, + "loss": 0.0061, + "step": 12665 + }, + { + "epoch": 5.960470588235294, + "grad_norm": 0.0941869904533468, + "learning_rate": 4.519400640811688e-08, + "loss": 0.009, + "step": 12666 + }, + { + "epoch": 5.960941176470588, + "grad_norm": 0.08070429999046275, + "learning_rate": 4.412441612738904e-08, + "loss": 0.0059, + "step": 12667 + }, + { + "epoch": 5.961411764705883, + "grad_norm": 0.05804349920348714, + "learning_rate": 4.3067632813964524e-08, + "loss": 0.0084, + "step": 12668 + }, + { + "epoch": 5.961882352941177, + "grad_norm": 0.02298594306990876, + "learning_rate": 4.2023656558104466e-08, + "loss": 0.0014, + "step": 12669 + }, + { + "epoch": 5.962352941176471, + "grad_norm": 0.10434538036130987, + "learning_rate": 4.099248744897088e-08, + "loss": 0.0149, + "step": 12670 + }, + { + "epoch": 5.962823529411764, + "grad_norm": 0.07815804002282259, + "learning_rate": 3.9974125574609994e-08, + "loss": 0.0088, + "step": 12671 + }, + { + "epoch": 5.963294117647059, + "grad_norm": 0.1615586997631036, + "learning_rate": 3.896857102200224e-08, + "loss": 0.047, + "step": 12672 + }, + { + "epoch": 5.963764705882353, + "grad_norm": 0.023023715779474496, + "learning_rate": 3.797582387704556e-08, + "loss": 0.001, + "step": 12673 + }, + { + "epoch": 5.964235294117647, + "grad_norm": 0.00010840634591823017, + "learning_rate": 3.69958842245055e-08, + "loss": 0.0, + "step": 12674 + }, + { + "epoch": 5.964705882352941, + "grad_norm": 0.09908250663275411, + "learning_rate": 3.6028752148081766e-08, + "loss": 0.0155, + "step": 12675 + }, + { + "epoch": 5.9651764705882355, + "grad_norm": 6.502169112261727e-07, + "learning_rate": 3.5074427730374945e-08, + "loss": 0.0, + "step": 12676 + }, + { + "epoch": 5.96564705882353, + "grad_norm": 0.08880096031694365, + "learning_rate": 3.4132911052886514e-08, + "loss": 0.0104, + "step": 12677 + }, + { + "epoch": 5.966117647058823, + "grad_norm": 5.672870688553575e-06, + "learning_rate": 3.320420219603548e-08, + "loss": 0.0, + "step": 12678 + }, + { + "epoch": 5.966588235294117, + "grad_norm": 0.1448599152046468, + "learning_rate": 3.228830123910842e-08, + "loss": 0.0386, + "step": 12679 + }, + { + "epoch": 5.9670588235294115, + "grad_norm": 0.04386143939531608, + "learning_rate": 3.13852082603594e-08, + "loss": 0.0033, + "step": 12680 + }, + { + "epoch": 5.967529411764706, + "grad_norm": 0.10122701889233264, + "learning_rate": 3.049492333691006e-08, + "loss": 0.0221, + "step": 12681 + }, + { + "epoch": 5.968, + "grad_norm": 0.14423255623136116, + "learning_rate": 2.961744654478293e-08, + "loss": 0.0366, + "step": 12682 + }, + { + "epoch": 5.968470588235294, + "grad_norm": 0.047671255500670015, + "learning_rate": 2.875277795891806e-08, + "loss": 0.0029, + "step": 12683 + }, + { + "epoch": 5.968941176470588, + "grad_norm": 0.11773828221476548, + "learning_rate": 2.7900917653189692e-08, + "loss": 0.0327, + "step": 12684 + }, + { + "epoch": 5.969411764705883, + "grad_norm": 0.07881496201243264, + "learning_rate": 2.706186570032298e-08, + "loss": 0.0085, + "step": 12685 + }, + { + "epoch": 5.969882352941177, + "grad_norm": 6.985257829645276e-06, + "learning_rate": 2.6235622171977278e-08, + "loss": 0.0, + "step": 12686 + }, + { + "epoch": 5.97035294117647, + "grad_norm": 0.00019056370041678012, + "learning_rate": 2.542218713874611e-08, + "loss": 0.0, + "step": 12687 + }, + { + "epoch": 5.970823529411764, + "grad_norm": 0.10058946459326061, + "learning_rate": 2.462156067005727e-08, + "loss": 0.027, + "step": 12688 + }, + { + "epoch": 5.971294117647059, + "grad_norm": 0.16156123632160077, + "learning_rate": 2.3833742834339365e-08, + "loss": 0.0507, + "step": 12689 + }, + { + "epoch": 5.971764705882353, + "grad_norm": 1.2359487765795829e-05, + "learning_rate": 2.305873369882194e-08, + "loss": 0.0, + "step": 12690 + }, + { + "epoch": 5.972235294117647, + "grad_norm": 0.07162662941436473, + "learning_rate": 2.2296533329735356e-08, + "loss": 0.0069, + "step": 12691 + }, + { + "epoch": 5.972705882352941, + "grad_norm": 0.042339584794961665, + "learning_rate": 2.1547141792160882e-08, + "loss": 0.0028, + "step": 12692 + }, + { + "epoch": 5.9731764705882355, + "grad_norm": 0.1000037731638267, + "learning_rate": 2.0810559150097327e-08, + "loss": 0.0109, + "step": 12693 + }, + { + "epoch": 5.97364705882353, + "grad_norm": 0.11166391996171256, + "learning_rate": 2.008678546646103e-08, + "loss": 0.0236, + "step": 12694 + }, + { + "epoch": 5.974117647058824, + "grad_norm": 0.08857959352824511, + "learning_rate": 1.9375820803052555e-08, + "loss": 0.0186, + "step": 12695 + }, + { + "epoch": 5.974588235294117, + "grad_norm": 0.2239069343083768, + "learning_rate": 1.8677665220606653e-08, + "loss": 0.0361, + "step": 12696 + }, + { + "epoch": 5.9750588235294115, + "grad_norm": 0.1450678001643061, + "learning_rate": 1.7992318778742298e-08, + "loss": 0.0455, + "step": 12697 + }, + { + "epoch": 5.975529411764706, + "grad_norm": 0.12333855340456926, + "learning_rate": 1.7319781535996002e-08, + "loss": 0.0268, + "step": 12698 + }, + { + "epoch": 5.976, + "grad_norm": 0.07238076285397001, + "learning_rate": 1.6660053549788498e-08, + "loss": 0.0131, + "step": 12699 + }, + { + "epoch": 5.976470588235294, + "grad_norm": 0.1011520668043566, + "learning_rate": 1.6013134876491362e-08, + "loss": 0.0125, + "step": 12700 + }, + { + "epoch": 5.976941176470588, + "grad_norm": 0.043155926093842695, + "learning_rate": 1.5379025571327088e-08, + "loss": 0.0012, + "step": 12701 + }, + { + "epoch": 5.977411764705883, + "grad_norm": 0.13617292633784672, + "learning_rate": 1.4757725688469002e-08, + "loss": 0.0206, + "step": 12702 + }, + { + "epoch": 5.977882352941176, + "grad_norm": 0.11793207707404009, + "learning_rate": 1.414923528097467e-08, + "loss": 0.022, + "step": 12703 + }, + { + "epoch": 5.97835294117647, + "grad_norm": 0.10627913040967261, + "learning_rate": 1.3553554400802524e-08, + "loss": 0.0218, + "step": 12704 + }, + { + "epoch": 5.978823529411764, + "grad_norm": 9.451435315118127e-06, + "learning_rate": 1.2970683098845192e-08, + "loss": 0.0, + "step": 12705 + }, + { + "epoch": 5.979294117647059, + "grad_norm": 0.004970210378335209, + "learning_rate": 1.240062142487952e-08, + "loss": 0.0001, + "step": 12706 + }, + { + "epoch": 5.979764705882353, + "grad_norm": 0.047002318346103376, + "learning_rate": 1.1843369427583238e-08, + "loss": 0.0032, + "step": 12707 + }, + { + "epoch": 5.980235294117647, + "grad_norm": 7.703019861427638e-06, + "learning_rate": 1.1298927154551608e-08, + "loss": 0.0, + "step": 12708 + }, + { + "epoch": 5.980705882352941, + "grad_norm": 0.05676414361080745, + "learning_rate": 1.0767294652280767e-08, + "loss": 0.0019, + "step": 12709 + }, + { + "epoch": 5.9811764705882355, + "grad_norm": 0.02840824722511952, + "learning_rate": 1.0248471966167737e-08, + "loss": 0.0015, + "step": 12710 + }, + { + "epoch": 5.98164705882353, + "grad_norm": 0.1494660819577726, + "learning_rate": 9.74245914052707e-09, + "loss": 0.039, + "step": 12711 + }, + { + "epoch": 5.982117647058823, + "grad_norm": 4.443840741431304e-05, + "learning_rate": 9.24925621859085e-09, + "loss": 0.0, + "step": 12712 + }, + { + "epoch": 5.982588235294117, + "grad_norm": 0.16822407515564824, + "learning_rate": 8.768863242475389e-09, + "loss": 0.049, + "step": 12713 + }, + { + "epoch": 5.9830588235294115, + "grad_norm": 0.0871753385991792, + "learning_rate": 8.30128025318122e-09, + "loss": 0.0137, + "step": 12714 + }, + { + "epoch": 5.983529411764706, + "grad_norm": 0.1356559198119888, + "learning_rate": 7.846507290676373e-09, + "loss": 0.022, + "step": 12715 + }, + { + "epoch": 5.984, + "grad_norm": 0.055011114034031444, + "learning_rate": 7.404544393779799e-09, + "loss": 0.0054, + "step": 12716 + }, + { + "epoch": 5.984470588235294, + "grad_norm": 0.06180764184128554, + "learning_rate": 6.975391600261282e-09, + "loss": 0.0029, + "step": 12717 + }, + { + "epoch": 5.984941176470588, + "grad_norm": 0.03617873651134167, + "learning_rate": 6.55904894674153e-09, + "loss": 0.0032, + "step": 12718 + }, + { + "epoch": 5.985411764705883, + "grad_norm": 0.08858946594908805, + "learning_rate": 6.155516468792088e-09, + "loss": 0.0039, + "step": 12719 + }, + { + "epoch": 5.985882352941177, + "grad_norm": 0.03901691351282177, + "learning_rate": 5.764794200885381e-09, + "loss": 0.002, + "step": 12720 + }, + { + "epoch": 5.98635294117647, + "grad_norm": 0.07659380684508436, + "learning_rate": 5.38688217637806e-09, + "loss": 0.0085, + "step": 12721 + }, + { + "epoch": 5.986823529411764, + "grad_norm": 0.11268449461074338, + "learning_rate": 5.0217804275443084e-09, + "loss": 0.0254, + "step": 12722 + }, + { + "epoch": 5.987294117647059, + "grad_norm": 0.07666766634586765, + "learning_rate": 4.669488985575842e-09, + "loss": 0.0074, + "step": 12723 + }, + { + "epoch": 5.987764705882353, + "grad_norm": 0.09366489431902052, + "learning_rate": 4.3300078805652564e-09, + "loss": 0.0183, + "step": 12724 + }, + { + "epoch": 5.988235294117647, + "grad_norm": 0.1065927833566755, + "learning_rate": 4.003337141489371e-09, + "loss": 0.0163, + "step": 12725 + }, + { + "epoch": 5.988705882352941, + "grad_norm": 0.0585339785095025, + "learning_rate": 3.689476796242541e-09, + "loss": 0.0044, + "step": 12726 + }, + { + "epoch": 5.9891764705882355, + "grad_norm": 0.1335958237569649, + "learning_rate": 3.388426871653305e-09, + "loss": 0.0285, + "step": 12727 + }, + { + "epoch": 5.98964705882353, + "grad_norm": 0.16325804425629645, + "learning_rate": 3.100187393434428e-09, + "loss": 0.033, + "step": 12728 + }, + { + "epoch": 5.990117647058824, + "grad_norm": 0.11272832159116913, + "learning_rate": 2.8247583861662484e-09, + "loss": 0.018, + "step": 12729 + }, + { + "epoch": 5.990588235294117, + "grad_norm": 0.04538058581323288, + "learning_rate": 2.562139873413249e-09, + "loss": 0.0026, + "step": 12730 + }, + { + "epoch": 5.9910588235294115, + "grad_norm": 0.14437992992661902, + "learning_rate": 2.3123318775741805e-09, + "loss": 0.0015, + "step": 12731 + }, + { + "epoch": 5.991529411764706, + "grad_norm": 1.561891350455505e-05, + "learning_rate": 2.075334419998631e-09, + "loss": 0.0, + "step": 12732 + }, + { + "epoch": 5.992, + "grad_norm": 0.11269221663821344, + "learning_rate": 1.8511475209370685e-09, + "loss": 0.0198, + "step": 12733 + }, + { + "epoch": 5.992470588235294, + "grad_norm": 0.16015005043933359, + "learning_rate": 1.6397711995241869e-09, + "loss": 0.0284, + "step": 12734 + }, + { + "epoch": 5.992941176470588, + "grad_norm": 0.013308206059482729, + "learning_rate": 1.441205473795559e-09, + "loss": 0.0002, + "step": 12735 + }, + { + "epoch": 5.993411764705883, + "grad_norm": 0.11394332311022883, + "learning_rate": 1.2554503607375976e-09, + "loss": 0.0302, + "step": 12736 + }, + { + "epoch": 5.993882352941176, + "grad_norm": 0.12660866416574823, + "learning_rate": 1.082505876204287e-09, + "loss": 0.0258, + "step": 12737 + }, + { + "epoch": 5.99435294117647, + "grad_norm": 1.580409287095315e-05, + "learning_rate": 9.223720349671449e-10, + "loss": 0.0, + "step": 12738 + }, + { + "epoch": 5.994823529411764, + "grad_norm": 0.036723512973749046, + "learning_rate": 7.750488506985674e-10, + "loss": 0.0021, + "step": 12739 + }, + { + "epoch": 5.995294117647059, + "grad_norm": 0.173690404325134, + "learning_rate": 6.405363359718308e-10, + "loss": 0.0717, + "step": 12740 + }, + { + "epoch": 5.995764705882353, + "grad_norm": 0.12256551268071715, + "learning_rate": 5.188345022943963e-10, + "loss": 0.0302, + "step": 12741 + }, + { + "epoch": 5.996235294117647, + "grad_norm": 0.0465875809557938, + "learning_rate": 4.099433600579516e-10, + "loss": 0.0023, + "step": 12742 + }, + { + "epoch": 5.996705882352941, + "grad_norm": 1.5945551584383783e-05, + "learning_rate": 3.138629185550634e-10, + "loss": 0.0, + "step": 12743 + }, + { + "epoch": 5.9971764705882356, + "grad_norm": 0.14647513651962443, + "learning_rate": 2.3059318597917763e-10, + "loss": 0.0407, + "step": 12744 + }, + { + "epoch": 5.99764705882353, + "grad_norm": 0.09546889811644263, + "learning_rate": 1.6013416947457946e-10, + "loss": 0.0089, + "step": 12745 + }, + { + "epoch": 5.998117647058823, + "grad_norm": 0.1652031907562031, + "learning_rate": 1.0248587501981986e-10, + "loss": 0.0409, + "step": 12746 + }, + { + "epoch": 5.998588235294117, + "grad_norm": 0.11507665105627636, + "learning_rate": 5.7648307577595756e-11, + "loss": 0.0218, + "step": 12747 + }, + { + "epoch": 5.9990588235294116, + "grad_norm": 0.028536941303977804, + "learning_rate": 2.5621470944869883e-11, + "loss": 0.002, + "step": 12748 + }, + { + "epoch": 5.999529411764706, + "grad_norm": 0.14346979985767105, + "learning_rate": 6.4053678694442335e-12, + "loss": 0.0371, + "step": 12749 + }, + { + "epoch": 6.0, + "grad_norm": 0.1289526373112582, + "learning_rate": 0.0, + "loss": 0.0186, + "step": 12750 + }, + { + "epoch": 6.0, + "eval_loss": 0.029851149767637253, + "eval_runtime": 59.3049, + "eval_samples_per_second": 12.647, + "eval_steps_per_second": 6.323, + "step": 12750 + }, + { + "epoch": 6.0, + "step": 12750, + "total_flos": 141728309084160.0, + "train_loss": 0.11299328358555366, + "train_runtime": 9965.4705, + "train_samples_per_second": 2.559, + "train_steps_per_second": 1.279 + } + ], + "logging_steps": 1.0, + "max_steps": 12750, + "num_input_tokens_seen": 0, + "num_train_epochs": 6, + "save_steps": 500, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 141728309084160.0, + "train_batch_size": 2, + "trial_name": null, + "trial_params": null +}