{ "best_metric": 0.19756343960762024, "best_model_checkpoint": "final_models/lait_bur_llama/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": 1053.3900819381472, "learning_rate": 1.5e-07, "loss": 16.1288, "step": 1 }, { "epoch": 0.0009411764705882353, "grad_norm": 1003.999746052958, "learning_rate": 3e-07, "loss": 17.8195, "step": 2 }, { "epoch": 0.001411764705882353, "grad_norm": 1302.9432126513736, "learning_rate": 4.5e-07, "loss": 15.3047, "step": 3 }, { "epoch": 0.0018823529411764706, "grad_norm": 806.5289363478547, "learning_rate": 6e-07, "loss": 16.4631, "step": 4 }, { "epoch": 0.002352941176470588, "grad_norm": 817.2119672989627, "learning_rate": 7.499999999999999e-07, "loss": 17.752, "step": 5 }, { "epoch": 0.002823529411764706, "grad_norm": 323.1102095060333, "learning_rate": 9e-07, "loss": 18.8803, "step": 6 }, { "epoch": 0.0032941176470588237, "grad_norm": 1330.795245700747, "learning_rate": 1.05e-06, "loss": 13.7679, "step": 7 }, { "epoch": 0.0037647058823529413, "grad_norm": 202.8744809439263, "learning_rate": 1.2e-06, "loss": 19.2957, "step": 8 }, { "epoch": 0.0042352941176470585, "grad_norm": 225.65961612124192, "learning_rate": 1.3499999999999998e-06, "loss": 13.5266, "step": 9 }, { "epoch": 0.004705882352941176, "grad_norm": 194.175949821567, "learning_rate": 1.4999999999999998e-06, "loss": 18.6098, "step": 10 }, { "epoch": 0.0051764705882352945, "grad_norm": 191.9163072370054, "learning_rate": 1.6499999999999997e-06, "loss": 13.5329, "step": 11 }, { "epoch": 0.005647058823529412, "grad_norm": 174.03697385962977, "learning_rate": 1.8e-06, "loss": 15.0908, "step": 12 }, { "epoch": 0.00611764705882353, "grad_norm": 118.40811281566457, "learning_rate": 1.9499999999999995e-06, "loss": 10.0461, "step": 13 }, { "epoch": 0.006588235294117647, "grad_norm": 371.29265482438115, "learning_rate": 2.1e-06, "loss": 9.5583, "step": 14 }, { "epoch": 0.007058823529411765, "grad_norm": 129.1709195988487, "learning_rate": 2.2499999999999996e-06, "loss": 9.9364, "step": 15 }, { "epoch": 0.0075294117647058826, "grad_norm": 257.5089141171512, "learning_rate": 2.4e-06, "loss": 7.226, "step": 16 }, { "epoch": 0.008, "grad_norm": 134.66713128832131, "learning_rate": 2.55e-06, "loss": 3.7174, "step": 17 }, { "epoch": 0.008470588235294117, "grad_norm": 166.8986302868171, "learning_rate": 2.6999999999999996e-06, "loss": 3.4386, "step": 18 }, { "epoch": 0.008941176470588235, "grad_norm": 199.3125940423567, "learning_rate": 2.85e-06, "loss": 3.1737, "step": 19 }, { "epoch": 0.009411764705882352, "grad_norm": 98.81307593798378, "learning_rate": 2.9999999999999997e-06, "loss": 1.7396, "step": 20 }, { "epoch": 0.00988235294117647, "grad_norm": 44.8330482390545, "learning_rate": 3.15e-06, "loss": 1.2638, "step": 21 }, { "epoch": 0.010352941176470589, "grad_norm": 18.77352799486752, "learning_rate": 3.2999999999999993e-06, "loss": 0.592, "step": 22 }, { "epoch": 0.010823529411764706, "grad_norm": 72.96098694662109, "learning_rate": 3.4499999999999996e-06, "loss": 3.9644, "step": 23 }, { "epoch": 0.011294117647058824, "grad_norm": 113.79672854357362, "learning_rate": 3.6e-06, "loss": 8.2021, "step": 24 }, { "epoch": 0.011764705882352941, "grad_norm": 85.61264620500445, "learning_rate": 3.7499999999999997e-06, "loss": 5.5016, "step": 25 }, { "epoch": 0.01223529411764706, "grad_norm": 52.248936133304575, "learning_rate": 3.899999999999999e-06, "loss": 2.6352, "step": 26 }, { "epoch": 0.012705882352941176, "grad_norm": 24.062933047891626, "learning_rate": 4.049999999999999e-06, "loss": 1.2081, "step": 27 }, { "epoch": 0.013176470588235295, "grad_norm": 132.76997158859814, "learning_rate": 4.2e-06, "loss": 6.7401, "step": 28 }, { "epoch": 0.013647058823529411, "grad_norm": 20.975620332236804, "learning_rate": 4.35e-06, "loss": 0.9364, "step": 29 }, { "epoch": 0.01411764705882353, "grad_norm": 33.59017491364359, "learning_rate": 4.499999999999999e-06, "loss": 1.6937, "step": 30 }, { "epoch": 0.014588235294117647, "grad_norm": 37.06005604722755, "learning_rate": 4.6499999999999995e-06, "loss": 1.2307, "step": 31 }, { "epoch": 0.015058823529411765, "grad_norm": 46.71072185455814, "learning_rate": 4.8e-06, "loss": 2.4607, "step": 32 }, { "epoch": 0.015529411764705882, "grad_norm": 65.14173059107449, "learning_rate": 4.95e-06, "loss": 2.3423, "step": 33 }, { "epoch": 0.016, "grad_norm": 61.949936409726874, "learning_rate": 5.1e-06, "loss": 2.2889, "step": 34 }, { "epoch": 0.01647058823529412, "grad_norm": 83.9886223691891, "learning_rate": 5.25e-06, "loss": 1.4629, "step": 35 }, { "epoch": 0.016941176470588234, "grad_norm": 21.799554405449186, "learning_rate": 5.399999999999999e-06, "loss": 1.4722, "step": 36 }, { "epoch": 0.017411764705882352, "grad_norm": 43.19199795824257, "learning_rate": 5.549999999999999e-06, "loss": 1.3979, "step": 37 }, { "epoch": 0.01788235294117647, "grad_norm": 27.685119289650174, "learning_rate": 5.7e-06, "loss": 0.9254, "step": 38 }, { "epoch": 0.01835294117647059, "grad_norm": 12.324598025853454, "learning_rate": 5.85e-06, "loss": 0.2565, "step": 39 }, { "epoch": 0.018823529411764704, "grad_norm": 29.184554394421262, "learning_rate": 5.999999999999999e-06, "loss": 1.111, "step": 40 }, { "epoch": 0.019294117647058823, "grad_norm": 31.648354632020066, "learning_rate": 6.1499999999999996e-06, "loss": 1.3861, "step": 41 }, { "epoch": 0.01976470588235294, "grad_norm": 40.8403703097617, "learning_rate": 6.3e-06, "loss": 0.8583, "step": 42 }, { "epoch": 0.02023529411764706, "grad_norm": 41.94255273791031, "learning_rate": 6.449999999999999e-06, "loss": 1.4781, "step": 43 }, { "epoch": 0.020705882352941178, "grad_norm": 32.948075381051524, "learning_rate": 6.599999999999999e-06, "loss": 0.6085, "step": 44 }, { "epoch": 0.021176470588235293, "grad_norm": 29.974851980390746, "learning_rate": 6.749999999999999e-06, "loss": 0.6835, "step": 45 }, { "epoch": 0.02164705882352941, "grad_norm": 67.42808424491332, "learning_rate": 6.899999999999999e-06, "loss": 1.3755, "step": 46 }, { "epoch": 0.02211764705882353, "grad_norm": 43.547349691284644, "learning_rate": 7.049999999999999e-06, "loss": 0.9514, "step": 47 }, { "epoch": 0.02258823529411765, "grad_norm": 30.98253508019103, "learning_rate": 7.2e-06, "loss": 0.489, "step": 48 }, { "epoch": 0.023058823529411764, "grad_norm": 37.439473455748505, "learning_rate": 7.35e-06, "loss": 0.3977, "step": 49 }, { "epoch": 0.023529411764705882, "grad_norm": 36.49355592498661, "learning_rate": 7.499999999999999e-06, "loss": 0.6606, "step": 50 }, { "epoch": 0.024, "grad_norm": 8.807910147107396, "learning_rate": 7.65e-06, "loss": 0.4608, "step": 51 }, { "epoch": 0.02447058823529412, "grad_norm": 3.574145179840181, "learning_rate": 7.799999999999998e-06, "loss": 0.1526, "step": 52 }, { "epoch": 0.024941176470588234, "grad_norm": 8.326246269105294, "learning_rate": 7.949999999999998e-06, "loss": 0.7282, "step": 53 }, { "epoch": 0.025411764705882352, "grad_norm": 1.6774687168977787, "learning_rate": 8.099999999999999e-06, "loss": 0.07, "step": 54 }, { "epoch": 0.02588235294117647, "grad_norm": 1.022980850700904, "learning_rate": 8.249999999999999e-06, "loss": 0.0196, "step": 55 }, { "epoch": 0.02635294117647059, "grad_norm": 7.394354010585887, "learning_rate": 8.4e-06, "loss": 0.164, "step": 56 }, { "epoch": 0.026823529411764704, "grad_norm": 50.93547136734546, "learning_rate": 8.55e-06, "loss": 0.9636, "step": 57 }, { "epoch": 0.027294117647058823, "grad_norm": 2.251672617522615, "learning_rate": 8.7e-06, "loss": 0.0945, "step": 58 }, { "epoch": 0.02776470588235294, "grad_norm": 1.963773868335638, "learning_rate": 8.849999999999998e-06, "loss": 0.0492, "step": 59 }, { "epoch": 0.02823529411764706, "grad_norm": 9.32580593441241, "learning_rate": 8.999999999999999e-06, "loss": 0.3724, "step": 60 }, { "epoch": 0.028705882352941175, "grad_norm": 10.287714360614089, "learning_rate": 9.149999999999999e-06, "loss": 0.4388, "step": 61 }, { "epoch": 0.029176470588235293, "grad_norm": 4.537692740748704, "learning_rate": 9.299999999999999e-06, "loss": 0.4039, "step": 62 }, { "epoch": 0.029647058823529412, "grad_norm": 2.7045465024847, "learning_rate": 9.45e-06, "loss": 0.2588, "step": 63 }, { "epoch": 0.03011764705882353, "grad_norm": 3.0111275993414486, "learning_rate": 9.6e-06, "loss": 0.0778, "step": 64 }, { "epoch": 0.03058823529411765, "grad_norm": 5.87075152160422, "learning_rate": 9.75e-06, "loss": 0.5651, "step": 65 }, { "epoch": 0.031058823529411764, "grad_norm": 1.5893945110800258, "learning_rate": 9.9e-06, "loss": 0.0793, "step": 66 }, { "epoch": 0.03152941176470588, "grad_norm": 3.102922388104239, "learning_rate": 1.005e-05, "loss": 0.3956, "step": 67 }, { "epoch": 0.032, "grad_norm": 1.3642493020844426, "learning_rate": 1.02e-05, "loss": 0.0467, "step": 68 }, { "epoch": 0.03247058823529412, "grad_norm": 3.169868693663029, "learning_rate": 1.035e-05, "loss": 0.2531, "step": 69 }, { "epoch": 0.03294117647058824, "grad_norm": 1.4479045431244182, "learning_rate": 1.05e-05, "loss": 0.051, "step": 70 }, { "epoch": 0.033411764705882356, "grad_norm": 2.26858295980039, "learning_rate": 1.0649999999999998e-05, "loss": 0.1253, "step": 71 }, { "epoch": 0.03388235294117647, "grad_norm": 1.8894013431580807, "learning_rate": 1.0799999999999998e-05, "loss": 0.1179, "step": 72 }, { "epoch": 0.034352941176470586, "grad_norm": 1.7529480893202434, "learning_rate": 1.0949999999999998e-05, "loss": 0.2042, "step": 73 }, { "epoch": 0.034823529411764705, "grad_norm": 2.5265106079503417, "learning_rate": 1.1099999999999999e-05, "loss": 0.3444, "step": 74 }, { "epoch": 0.03529411764705882, "grad_norm": 3.211599055714563, "learning_rate": 1.1249999999999999e-05, "loss": 0.5266, "step": 75 }, { "epoch": 0.03576470588235294, "grad_norm": 2.429542607531107, "learning_rate": 1.14e-05, "loss": 0.3559, "step": 76 }, { "epoch": 0.03623529411764706, "grad_norm": 2.634029214968653, "learning_rate": 1.155e-05, "loss": 0.3542, "step": 77 }, { "epoch": 0.03670588235294118, "grad_norm": 2.245820931807417, "learning_rate": 1.17e-05, "loss": 0.4037, "step": 78 }, { "epoch": 0.0371764705882353, "grad_norm": 1.6172104232081386, "learning_rate": 1.1849999999999998e-05, "loss": 0.0601, "step": 79 }, { "epoch": 0.03764705882352941, "grad_norm": 2.198860467283352, "learning_rate": 1.1999999999999999e-05, "loss": 0.043, "step": 80 }, { "epoch": 0.03811764705882353, "grad_norm": 9.689918899100414, "learning_rate": 1.2149999999999999e-05, "loss": 0.2579, "step": 81 }, { "epoch": 0.038588235294117645, "grad_norm": 2.724848966400364, "learning_rate": 1.2299999999999999e-05, "loss": 0.6148, "step": 82 }, { "epoch": 0.039058823529411764, "grad_norm": 1.8704706381934046, "learning_rate": 1.245e-05, "loss": 0.2162, "step": 83 }, { "epoch": 0.03952941176470588, "grad_norm": 2.246726528751377, "learning_rate": 1.26e-05, "loss": 0.3992, "step": 84 }, { "epoch": 0.04, "grad_norm": 1.8845663460645368, "learning_rate": 1.275e-05, "loss": 0.2925, "step": 85 }, { "epoch": 0.04047058823529412, "grad_norm": 4.942180941101507, "learning_rate": 1.2899999999999998e-05, "loss": 0.6683, "step": 86 }, { "epoch": 0.04094117647058824, "grad_norm": 1.0275751597846952, "learning_rate": 1.3049999999999999e-05, "loss": 0.0478, "step": 87 }, { "epoch": 0.041411764705882356, "grad_norm": 2.364158974995223, "learning_rate": 1.3199999999999997e-05, "loss": 0.4269, "step": 88 }, { "epoch": 0.04188235294117647, "grad_norm": 0.5044429650886701, "learning_rate": 1.3349999999999998e-05, "loss": 0.0094, "step": 89 }, { "epoch": 0.042352941176470586, "grad_norm": 3.305708555129838, "learning_rate": 1.3499999999999998e-05, "loss": 0.3632, "step": 90 }, { "epoch": 0.042823529411764705, "grad_norm": 1.3048093151528424, "learning_rate": 1.3649999999999998e-05, "loss": 0.0126, "step": 91 }, { "epoch": 0.04329411764705882, "grad_norm": 0.6531175326904165, "learning_rate": 1.3799999999999998e-05, "loss": 0.0117, "step": 92 }, { "epoch": 0.04376470588235294, "grad_norm": 3.5154911022300754, "learning_rate": 1.3949999999999999e-05, "loss": 0.4349, "step": 93 }, { "epoch": 0.04423529411764706, "grad_norm": 2.1997293262298845, "learning_rate": 1.4099999999999999e-05, "loss": 0.2262, "step": 94 }, { "epoch": 0.04470588235294118, "grad_norm": 5.464299814029718, "learning_rate": 1.4249999999999999e-05, "loss": 0.6119, "step": 95 }, { "epoch": 0.0451764705882353, "grad_norm": 1.3108512830916828, "learning_rate": 1.44e-05, "loss": 0.1663, "step": 96 }, { "epoch": 0.04564705882352941, "grad_norm": 2.5290489908907907, "learning_rate": 1.455e-05, "loss": 0.2921, "step": 97 }, { "epoch": 0.04611764705882353, "grad_norm": 2.4425330713898603, "learning_rate": 1.47e-05, "loss": 0.2641, "step": 98 }, { "epoch": 0.046588235294117646, "grad_norm": 0.3319379490182039, "learning_rate": 1.485e-05, "loss": 0.0093, "step": 99 }, { "epoch": 0.047058823529411764, "grad_norm": 3.22621507121045, "learning_rate": 1.4999999999999999e-05, "loss": 0.3392, "step": 100 }, { "epoch": 0.04752941176470588, "grad_norm": 1.3137611781333656, "learning_rate": 1.5149999999999999e-05, "loss": 0.161, "step": 101 }, { "epoch": 0.048, "grad_norm": 0.5891545368605571, "learning_rate": 1.53e-05, "loss": 0.0115, "step": 102 }, { "epoch": 0.04847058823529412, "grad_norm": 1.001577018524294, "learning_rate": 1.545e-05, "loss": 0.0332, "step": 103 }, { "epoch": 0.04894117647058824, "grad_norm": 5.176244931500083, "learning_rate": 1.5599999999999996e-05, "loss": 0.4425, "step": 104 }, { "epoch": 0.04941176470588235, "grad_norm": 1.220031632486272, "learning_rate": 1.5749999999999997e-05, "loss": 0.077, "step": 105 }, { "epoch": 0.04988235294117647, "grad_norm": 1.0353396456567876, "learning_rate": 1.5899999999999997e-05, "loss": 0.0312, "step": 106 }, { "epoch": 0.050352941176470586, "grad_norm": 14.38190526441612, "learning_rate": 1.6049999999999997e-05, "loss": 0.5575, "step": 107 }, { "epoch": 0.050823529411764705, "grad_norm": 3.182223076959544, "learning_rate": 1.6199999999999997e-05, "loss": 0.3704, "step": 108 }, { "epoch": 0.05129411764705882, "grad_norm": 4.630894916673046, "learning_rate": 1.6349999999999998e-05, "loss": 0.3177, "step": 109 }, { "epoch": 0.05176470588235294, "grad_norm": 2.556798555109172, "learning_rate": 1.6499999999999998e-05, "loss": 0.1421, "step": 110 }, { "epoch": 0.05223529411764706, "grad_norm": 2.657054843538193, "learning_rate": 1.6649999999999998e-05, "loss": 0.1316, "step": 111 }, { "epoch": 0.05270588235294118, "grad_norm": 0.9665620049527295, "learning_rate": 1.68e-05, "loss": 0.0545, "step": 112 }, { "epoch": 0.0531764705882353, "grad_norm": 2.7271797969808986, "learning_rate": 1.695e-05, "loss": 0.2855, "step": 113 }, { "epoch": 0.05364705882352941, "grad_norm": 2.1295007443686806, "learning_rate": 1.71e-05, "loss": 0.1426, "step": 114 }, { "epoch": 0.05411764705882353, "grad_norm": 1.427258203742596, "learning_rate": 1.725e-05, "loss": 0.1193, "step": 115 }, { "epoch": 0.054588235294117646, "grad_norm": 5.168983296670674, "learning_rate": 1.74e-05, "loss": 0.3496, "step": 116 }, { "epoch": 0.055058823529411764, "grad_norm": 2.4221062192808045, "learning_rate": 1.755e-05, "loss": 0.4722, "step": 117 }, { "epoch": 0.05552941176470588, "grad_norm": 3.3792991494166986, "learning_rate": 1.7699999999999997e-05, "loss": 0.3605, "step": 118 }, { "epoch": 0.056, "grad_norm": 0.9268210242766988, "learning_rate": 1.7849999999999997e-05, "loss": 0.0413, "step": 119 }, { "epoch": 0.05647058823529412, "grad_norm": 3.356504740210788, "learning_rate": 1.7999999999999997e-05, "loss": 0.2778, "step": 120 }, { "epoch": 0.05694117647058824, "grad_norm": 1.4915842759082953, "learning_rate": 1.8149999999999997e-05, "loss": 0.1197, "step": 121 }, { "epoch": 0.05741176470588235, "grad_norm": 1.8852085879244627, "learning_rate": 1.8299999999999998e-05, "loss": 0.2163, "step": 122 }, { "epoch": 0.05788235294117647, "grad_norm": 1.27735638922533, "learning_rate": 1.8449999999999998e-05, "loss": 0.1386, "step": 123 }, { "epoch": 0.05835294117647059, "grad_norm": 1.9326682266037647, "learning_rate": 1.8599999999999998e-05, "loss": 0.1375, "step": 124 }, { "epoch": 0.058823529411764705, "grad_norm": 1.816816395314965, "learning_rate": 1.875e-05, "loss": 0.1245, "step": 125 }, { "epoch": 0.059294117647058824, "grad_norm": 2.9089074748851895, "learning_rate": 1.89e-05, "loss": 0.3784, "step": 126 }, { "epoch": 0.05976470588235294, "grad_norm": 0.42358771111495336, "learning_rate": 1.905e-05, "loss": 0.0229, "step": 127 }, { "epoch": 0.06023529411764706, "grad_norm": 2.6065169501379937, "learning_rate": 1.92e-05, "loss": 0.2167, "step": 128 }, { "epoch": 0.06070588235294118, "grad_norm": 1.9231010776790831, "learning_rate": 1.935e-05, "loss": 0.2229, "step": 129 }, { "epoch": 0.0611764705882353, "grad_norm": 3.160762554789088, "learning_rate": 1.95e-05, "loss": 0.3715, "step": 130 }, { "epoch": 0.06164705882352941, "grad_norm": 2.19436469451767, "learning_rate": 1.965e-05, "loss": 0.3214, "step": 131 }, { "epoch": 0.06211764705882353, "grad_norm": 2.264006666489439, "learning_rate": 1.98e-05, "loss": 0.2756, "step": 132 }, { "epoch": 0.06258823529411765, "grad_norm": 4.054363241585821, "learning_rate": 1.995e-05, "loss": 0.6223, "step": 133 }, { "epoch": 0.06305882352941176, "grad_norm": 2.079475684274301, "learning_rate": 2.01e-05, "loss": 0.2122, "step": 134 }, { "epoch": 0.06352941176470588, "grad_norm": 3.421910452642776, "learning_rate": 2.025e-05, "loss": 0.4225, "step": 135 }, { "epoch": 0.064, "grad_norm": 2.342844803856131, "learning_rate": 2.04e-05, "loss": 0.4283, "step": 136 }, { "epoch": 0.06447058823529411, "grad_norm": 1.5451085984580313, "learning_rate": 2.055e-05, "loss": 0.106, "step": 137 }, { "epoch": 0.06494117647058824, "grad_norm": 2.319549103193929, "learning_rate": 2.07e-05, "loss": 0.328, "step": 138 }, { "epoch": 0.06541176470588235, "grad_norm": 2.0270586319914394, "learning_rate": 2.085e-05, "loss": 0.2006, "step": 139 }, { "epoch": 0.06588235294117648, "grad_norm": 1.2711285130825922, "learning_rate": 2.1e-05, "loss": 0.0723, "step": 140 }, { "epoch": 0.06635294117647059, "grad_norm": 2.318927266063024, "learning_rate": 2.1149999999999996e-05, "loss": 0.3841, "step": 141 }, { "epoch": 0.06682352941176471, "grad_norm": 4.417401807111931, "learning_rate": 2.1299999999999996e-05, "loss": 0.0214, "step": 142 }, { "epoch": 0.06729411764705882, "grad_norm": 1.9211869589977606, "learning_rate": 2.1449999999999996e-05, "loss": 0.2463, "step": 143 }, { "epoch": 0.06776470588235294, "grad_norm": 1.6254680924012048, "learning_rate": 2.1599999999999996e-05, "loss": 0.241, "step": 144 }, { "epoch": 0.06823529411764706, "grad_norm": 0.7148213107587708, "learning_rate": 2.1749999999999997e-05, "loss": 0.0388, "step": 145 }, { "epoch": 0.06870588235294117, "grad_norm": 0.7998088194785156, "learning_rate": 2.1899999999999997e-05, "loss": 0.0332, "step": 146 }, { "epoch": 0.0691764705882353, "grad_norm": 2.3588754170634, "learning_rate": 2.2049999999999997e-05, "loss": 0.2562, "step": 147 }, { "epoch": 0.06964705882352941, "grad_norm": 1.446954147329684, "learning_rate": 2.2199999999999998e-05, "loss": 0.205, "step": 148 }, { "epoch": 0.07011764705882353, "grad_norm": 1.009038460997951, "learning_rate": 2.2349999999999998e-05, "loss": 0.0499, "step": 149 }, { "epoch": 0.07058823529411765, "grad_norm": 2.2148131436655807, "learning_rate": 2.2499999999999998e-05, "loss": 0.1716, "step": 150 }, { "epoch": 0.07105882352941177, "grad_norm": 1.8408629744089027, "learning_rate": 2.2649999999999998e-05, "loss": 0.3146, "step": 151 }, { "epoch": 0.07152941176470588, "grad_norm": 1.1033120686517024, "learning_rate": 2.28e-05, "loss": 0.0665, "step": 152 }, { "epoch": 0.072, "grad_norm": 4.654312189037422, "learning_rate": 2.295e-05, "loss": 0.5209, "step": 153 }, { "epoch": 0.07247058823529412, "grad_norm": 1.4360139530993687, "learning_rate": 2.31e-05, "loss": 0.1963, "step": 154 }, { "epoch": 0.07294117647058823, "grad_norm": 3.1199025681740746, "learning_rate": 2.325e-05, "loss": 0.2623, "step": 155 }, { "epoch": 0.07341176470588236, "grad_norm": 2.1919323156174646, "learning_rate": 2.34e-05, "loss": 0.1514, "step": 156 }, { "epoch": 0.07388235294117647, "grad_norm": 1.9963539762818134, "learning_rate": 2.3549999999999996e-05, "loss": 0.2943, "step": 157 }, { "epoch": 0.0743529411764706, "grad_norm": 3.2228836402456222, "learning_rate": 2.3699999999999997e-05, "loss": 0.3207, "step": 158 }, { "epoch": 0.0748235294117647, "grad_norm": 2.090458311540065, "learning_rate": 2.3849999999999997e-05, "loss": 0.383, "step": 159 }, { "epoch": 0.07529411764705882, "grad_norm": 1.1210619965885271, "learning_rate": 2.3999999999999997e-05, "loss": 0.0842, "step": 160 }, { "epoch": 0.07576470588235294, "grad_norm": 2.4227242726397447, "learning_rate": 2.4149999999999997e-05, "loss": 0.3124, "step": 161 }, { "epoch": 0.07623529411764705, "grad_norm": 1.6435347020006421, "learning_rate": 2.4299999999999998e-05, "loss": 0.1716, "step": 162 }, { "epoch": 0.07670588235294118, "grad_norm": 1.618284518612266, "learning_rate": 2.4449999999999998e-05, "loss": 0.2308, "step": 163 }, { "epoch": 0.07717647058823529, "grad_norm": 2.297330570576708, "learning_rate": 2.4599999999999998e-05, "loss": 0.2429, "step": 164 }, { "epoch": 0.07764705882352942, "grad_norm": 1.7633783359610107, "learning_rate": 2.475e-05, "loss": 0.2318, "step": 165 }, { "epoch": 0.07811764705882353, "grad_norm": 1.0322609256701536, "learning_rate": 2.49e-05, "loss": 0.0603, "step": 166 }, { "epoch": 0.07858823529411765, "grad_norm": 2.3687715328802907, "learning_rate": 2.505e-05, "loss": 0.269, "step": 167 }, { "epoch": 0.07905882352941176, "grad_norm": 2.890031143059409, "learning_rate": 2.52e-05, "loss": 0.4239, "step": 168 }, { "epoch": 0.07952941176470588, "grad_norm": 0.3945162398333535, "learning_rate": 2.535e-05, "loss": 0.0151, "step": 169 }, { "epoch": 0.08, "grad_norm": 3.3257646638234184, "learning_rate": 2.55e-05, "loss": 0.2612, "step": 170 }, { "epoch": 0.08047058823529411, "grad_norm": 0.7663406710088967, "learning_rate": 2.565e-05, "loss": 0.037, "step": 171 }, { "epoch": 0.08094117647058824, "grad_norm": 1.7343750602552834, "learning_rate": 2.5799999999999997e-05, "loss": 0.0987, "step": 172 }, { "epoch": 0.08141176470588235, "grad_norm": 1.2389520701815924, "learning_rate": 2.5949999999999997e-05, "loss": 0.1038, "step": 173 }, { "epoch": 0.08188235294117648, "grad_norm": 2.6454177356000352, "learning_rate": 2.6099999999999997e-05, "loss": 0.189, "step": 174 }, { "epoch": 0.08235294117647059, "grad_norm": 1.4965932833228082, "learning_rate": 2.6249999999999998e-05, "loss": 0.1503, "step": 175 }, { "epoch": 0.08282352941176471, "grad_norm": 1.0797752974512458, "learning_rate": 2.6399999999999995e-05, "loss": 0.1152, "step": 176 }, { "epoch": 0.08329411764705882, "grad_norm": 2.2921701399477414, "learning_rate": 2.6549999999999995e-05, "loss": 0.2221, "step": 177 }, { "epoch": 0.08376470588235294, "grad_norm": 1.6759862060666093, "learning_rate": 2.6699999999999995e-05, "loss": 0.1909, "step": 178 }, { "epoch": 0.08423529411764706, "grad_norm": 0.4589803639154801, "learning_rate": 2.6849999999999995e-05, "loss": 0.0208, "step": 179 }, { "epoch": 0.08470588235294117, "grad_norm": 4.526325758800001, "learning_rate": 2.6999999999999996e-05, "loss": 0.4966, "step": 180 }, { "epoch": 0.0851764705882353, "grad_norm": 1.6918728078926533, "learning_rate": 2.7149999999999996e-05, "loss": 0.1918, "step": 181 }, { "epoch": 0.08564705882352941, "grad_norm": 1.607262031179581, "learning_rate": 2.7299999999999996e-05, "loss": 0.132, "step": 182 }, { "epoch": 0.08611764705882353, "grad_norm": 5.10363585452211, "learning_rate": 2.7449999999999996e-05, "loss": 0.2817, "step": 183 }, { "epoch": 0.08658823529411765, "grad_norm": 2.1554402885339705, "learning_rate": 2.7599999999999997e-05, "loss": 0.2787, "step": 184 }, { "epoch": 0.08705882352941176, "grad_norm": 0.8087611384984963, "learning_rate": 2.7749999999999997e-05, "loss": 0.0129, "step": 185 }, { "epoch": 0.08752941176470588, "grad_norm": 1.4589101712619301, "learning_rate": 2.7899999999999997e-05, "loss": 0.0787, "step": 186 }, { "epoch": 0.088, "grad_norm": 2.663176684949081, "learning_rate": 2.8049999999999997e-05, "loss": 0.2188, "step": 187 }, { "epoch": 0.08847058823529412, "grad_norm": 3.484559738187892, "learning_rate": 2.8199999999999998e-05, "loss": 0.3383, "step": 188 }, { "epoch": 0.08894117647058823, "grad_norm": 1.696595591914173, "learning_rate": 2.8349999999999998e-05, "loss": 0.2524, "step": 189 }, { "epoch": 0.08941176470588236, "grad_norm": 4.694886745543893, "learning_rate": 2.8499999999999998e-05, "loss": 0.3579, "step": 190 }, { "epoch": 0.08988235294117647, "grad_norm": 1.2051106780544845, "learning_rate": 2.865e-05, "loss": 0.1595, "step": 191 }, { "epoch": 0.0903529411764706, "grad_norm": 1.614586302497747, "learning_rate": 2.88e-05, "loss": 0.2072, "step": 192 }, { "epoch": 0.0908235294117647, "grad_norm": 1.905065842856865, "learning_rate": 2.895e-05, "loss": 0.1729, "step": 193 }, { "epoch": 0.09129411764705882, "grad_norm": 2.034374741677463, "learning_rate": 2.91e-05, "loss": 0.1902, "step": 194 }, { "epoch": 0.09176470588235294, "grad_norm": 1.760564942858609, "learning_rate": 2.925e-05, "loss": 0.3053, "step": 195 }, { "epoch": 0.09223529411764705, "grad_norm": 1.7470279780483013, "learning_rate": 2.94e-05, "loss": 0.1168, "step": 196 }, { "epoch": 0.09270588235294118, "grad_norm": 2.9765053525576275, "learning_rate": 2.955e-05, "loss": 0.2024, "step": 197 }, { "epoch": 0.09317647058823529, "grad_norm": 1.1524414328490646, "learning_rate": 2.97e-05, "loss": 0.1179, "step": 198 }, { "epoch": 0.09364705882352942, "grad_norm": 2.936751832099627, "learning_rate": 2.985e-05, "loss": 0.3563, "step": 199 }, { "epoch": 0.09411764705882353, "grad_norm": 0.31315308659346697, "learning_rate": 2.9999999999999997e-05, "loss": 0.0097, "step": 200 }, { "epoch": 0.09458823529411765, "grad_norm": 0.7787420417805113, "learning_rate": 3.0149999999999998e-05, "loss": 0.0249, "step": 201 }, { "epoch": 0.09505882352941177, "grad_norm": 4.483917727277514, "learning_rate": 3.0299999999999998e-05, "loss": 0.4612, "step": 202 }, { "epoch": 0.09552941176470588, "grad_norm": 1.5534902272762332, "learning_rate": 3.0449999999999998e-05, "loss": 0.1525, "step": 203 }, { "epoch": 0.096, "grad_norm": 0.6558708748851018, "learning_rate": 3.06e-05, "loss": 0.022, "step": 204 }, { "epoch": 0.09647058823529411, "grad_norm": 1.3297945465889556, "learning_rate": 3.0749999999999995e-05, "loss": 0.0809, "step": 205 }, { "epoch": 0.09694117647058824, "grad_norm": 3.635381169891572, "learning_rate": 3.09e-05, "loss": 0.17, "step": 206 }, { "epoch": 0.09741176470588235, "grad_norm": 6.647260272164321, "learning_rate": 3.1049999999999996e-05, "loss": 0.3832, "step": 207 }, { "epoch": 0.09788235294117648, "grad_norm": 0.6860217431713749, "learning_rate": 3.119999999999999e-05, "loss": 0.0425, "step": 208 }, { "epoch": 0.09835294117647059, "grad_norm": 2.818521889434579, "learning_rate": 3.1349999999999996e-05, "loss": 0.2381, "step": 209 }, { "epoch": 0.0988235294117647, "grad_norm": 0.6588501799682182, "learning_rate": 3.149999999999999e-05, "loss": 0.0399, "step": 210 }, { "epoch": 0.09929411764705882, "grad_norm": 2.6020766669052904, "learning_rate": 3.165e-05, "loss": 0.1976, "step": 211 }, { "epoch": 0.09976470588235294, "grad_norm": 3.124925171384034, "learning_rate": 3.1799999999999994e-05, "loss": 0.2184, "step": 212 }, { "epoch": 0.10023529411764706, "grad_norm": 2.2002642667758714, "learning_rate": 3.195e-05, "loss": 0.2399, "step": 213 }, { "epoch": 0.10070588235294117, "grad_norm": 1.0378975501701615, "learning_rate": 3.2099999999999994e-05, "loss": 0.1599, "step": 214 }, { "epoch": 0.1011764705882353, "grad_norm": 0.8630313218023407, "learning_rate": 3.225e-05, "loss": 0.0511, "step": 215 }, { "epoch": 0.10164705882352941, "grad_norm": 2.837346053739339, "learning_rate": 3.2399999999999995e-05, "loss": 0.2102, "step": 216 }, { "epoch": 0.10211764705882354, "grad_norm": 3.4672459261985162, "learning_rate": 3.255e-05, "loss": 0.4688, "step": 217 }, { "epoch": 0.10258823529411765, "grad_norm": 0.9468115008785694, "learning_rate": 3.2699999999999995e-05, "loss": 0.0625, "step": 218 }, { "epoch": 0.10305882352941176, "grad_norm": 0.4348344505407089, "learning_rate": 3.285e-05, "loss": 0.0138, "step": 219 }, { "epoch": 0.10352941176470588, "grad_norm": 3.2059244467226176, "learning_rate": 3.2999999999999996e-05, "loss": 0.1934, "step": 220 }, { "epoch": 0.104, "grad_norm": 2.132488170599688, "learning_rate": 3.315e-05, "loss": 0.1203, "step": 221 }, { "epoch": 0.10447058823529412, "grad_norm": 3.4106781365687238, "learning_rate": 3.3299999999999996e-05, "loss": 0.3041, "step": 222 }, { "epoch": 0.10494117647058823, "grad_norm": 2.436631974794772, "learning_rate": 3.345e-05, "loss": 0.3953, "step": 223 }, { "epoch": 0.10541176470588236, "grad_norm": 3.5510704757473293, "learning_rate": 3.36e-05, "loss": 0.3534, "step": 224 }, { "epoch": 0.10588235294117647, "grad_norm": 2.7265313741375468, "learning_rate": 3.375e-05, "loss": 0.2713, "step": 225 }, { "epoch": 0.1063529411764706, "grad_norm": 1.2069319806418988, "learning_rate": 3.39e-05, "loss": 0.0842, "step": 226 }, { "epoch": 0.1068235294117647, "grad_norm": 7.313294321464473, "learning_rate": 3.405e-05, "loss": 0.3073, "step": 227 }, { "epoch": 0.10729411764705882, "grad_norm": 2.4640627185604904, "learning_rate": 3.42e-05, "loss": 0.3272, "step": 228 }, { "epoch": 0.10776470588235294, "grad_norm": 1.5469621584630788, "learning_rate": 3.435e-05, "loss": 0.2249, "step": 229 }, { "epoch": 0.10823529411764705, "grad_norm": 1.2923252966855605, "learning_rate": 3.45e-05, "loss": 0.1561, "step": 230 }, { "epoch": 0.10870588235294118, "grad_norm": 149.00193396638468, "learning_rate": 3.465e-05, "loss": 0.927, "step": 231 }, { "epoch": 0.10917647058823529, "grad_norm": 53.46832359697489, "learning_rate": 3.48e-05, "loss": 0.3617, "step": 232 }, { "epoch": 0.10964705882352942, "grad_norm": 2.3608529553096034, "learning_rate": 3.4949999999999996e-05, "loss": 0.4324, "step": 233 }, { "epoch": 0.11011764705882353, "grad_norm": 2.0678195471880176, "learning_rate": 3.51e-05, "loss": 0.4507, "step": 234 }, { "epoch": 0.11058823529411765, "grad_norm": 3.389447010846317, "learning_rate": 3.5249999999999996e-05, "loss": 0.3923, "step": 235 }, { "epoch": 0.11105882352941177, "grad_norm": 2.047767045929269, "learning_rate": 3.539999999999999e-05, "loss": 0.3781, "step": 236 }, { "epoch": 0.11152941176470588, "grad_norm": 24.52127936012464, "learning_rate": 3.555e-05, "loss": 0.1139, "step": 237 }, { "epoch": 0.112, "grad_norm": 24.799558003474974, "learning_rate": 3.5699999999999994e-05, "loss": 0.2011, "step": 238 }, { "epoch": 0.11247058823529411, "grad_norm": 140.18120745237582, "learning_rate": 3.585e-05, "loss": 0.1755, "step": 239 }, { "epoch": 0.11294117647058824, "grad_norm": 29.373648326139765, "learning_rate": 3.5999999999999994e-05, "loss": 0.2281, "step": 240 }, { "epoch": 0.11341176470588235, "grad_norm": 3.4227252693256305, "learning_rate": 3.615e-05, "loss": 0.2448, "step": 241 }, { "epoch": 0.11388235294117648, "grad_norm": 2.1467222420434706, "learning_rate": 3.6299999999999995e-05, "loss": 0.2973, "step": 242 }, { "epoch": 0.11435294117647059, "grad_norm": 1.3341705625663565, "learning_rate": 3.645e-05, "loss": 0.1713, "step": 243 }, { "epoch": 0.1148235294117647, "grad_norm": 2.1005861118993105, "learning_rate": 3.6599999999999995e-05, "loss": 0.2653, "step": 244 }, { "epoch": 0.11529411764705882, "grad_norm": 3.0126805526719016, "learning_rate": 3.675e-05, "loss": 0.3488, "step": 245 }, { "epoch": 0.11576470588235294, "grad_norm": 1.4659736608415526, "learning_rate": 3.6899999999999996e-05, "loss": 0.167, "step": 246 }, { "epoch": 0.11623529411764706, "grad_norm": 2.597228835339375, "learning_rate": 3.705e-05, "loss": 0.2862, "step": 247 }, { "epoch": 0.11670588235294117, "grad_norm": 3.619269545609415, "learning_rate": 3.7199999999999996e-05, "loss": 0.5514, "step": 248 }, { "epoch": 0.1171764705882353, "grad_norm": 1.9119545684318975, "learning_rate": 3.735e-05, "loss": 0.225, "step": 249 }, { "epoch": 0.11764705882352941, "grad_norm": 6.224769193307903, "learning_rate": 3.75e-05, "loss": 0.4506, "step": 250 }, { "epoch": 0.11811764705882354, "grad_norm": 2.222322778401734, "learning_rate": 3.7649999999999994e-05, "loss": 0.5386, "step": 251 }, { "epoch": 0.11858823529411765, "grad_norm": 1.6331329996242754, "learning_rate": 3.78e-05, "loss": 0.2616, "step": 252 }, { "epoch": 0.11905882352941176, "grad_norm": 1.1946639261377832, "learning_rate": 3.7949999999999994e-05, "loss": 0.0757, "step": 253 }, { "epoch": 0.11952941176470588, "grad_norm": 0.858453014459814, "learning_rate": 3.81e-05, "loss": 0.0511, "step": 254 }, { "epoch": 0.12, "grad_norm": 2.634186492703819, "learning_rate": 3.8249999999999995e-05, "loss": 0.3203, "step": 255 }, { "epoch": 0.12047058823529412, "grad_norm": 0.5023169031619533, "learning_rate": 3.84e-05, "loss": 0.0031, "step": 256 }, { "epoch": 0.12094117647058823, "grad_norm": 0.9034851416353232, "learning_rate": 3.8549999999999995e-05, "loss": 0.1126, "step": 257 }, { "epoch": 0.12141176470588236, "grad_norm": 3.608735481914626, "learning_rate": 3.87e-05, "loss": 0.3763, "step": 258 }, { "epoch": 0.12188235294117647, "grad_norm": 1.2027734707421172, "learning_rate": 3.8849999999999996e-05, "loss": 0.1634, "step": 259 }, { "epoch": 0.1223529411764706, "grad_norm": 0.13338829745457276, "learning_rate": 3.9e-05, "loss": 0.0044, "step": 260 }, { "epoch": 0.1228235294117647, "grad_norm": 2.795464044728325, "learning_rate": 3.9149999999999996e-05, "loss": 0.3391, "step": 261 }, { "epoch": 0.12329411764705882, "grad_norm": 2.5467089058373054, "learning_rate": 3.93e-05, "loss": 0.4238, "step": 262 }, { "epoch": 0.12376470588235294, "grad_norm": 1.0624276005489395, "learning_rate": 3.945e-05, "loss": 0.1103, "step": 263 }, { "epoch": 0.12423529411764705, "grad_norm": 1.5504311917740696, "learning_rate": 3.96e-05, "loss": 0.2205, "step": 264 }, { "epoch": 0.12470588235294118, "grad_norm": 1.572081369705461, "learning_rate": 3.975e-05, "loss": 0.1363, "step": 265 }, { "epoch": 0.1251764705882353, "grad_norm": 1.7594253487639815, "learning_rate": 3.99e-05, "loss": 0.2079, "step": 266 }, { "epoch": 0.12564705882352942, "grad_norm": 1.3138751779944386, "learning_rate": 4.005e-05, "loss": 0.2255, "step": 267 }, { "epoch": 0.12611764705882353, "grad_norm": 1.1438911566748737, "learning_rate": 4.02e-05, "loss": 0.1748, "step": 268 }, { "epoch": 0.12658823529411764, "grad_norm": 2.077263082436487, "learning_rate": 4.035e-05, "loss": 0.2796, "step": 269 }, { "epoch": 0.12705882352941175, "grad_norm": 2.2635397708042646, "learning_rate": 4.05e-05, "loss": 0.2092, "step": 270 }, { "epoch": 0.1275294117647059, "grad_norm": 2.4415828912970703, "learning_rate": 4.065e-05, "loss": 0.4013, "step": 271 }, { "epoch": 0.128, "grad_norm": 1.1971980827812547, "learning_rate": 4.08e-05, "loss": 0.1968, "step": 272 }, { "epoch": 0.12847058823529411, "grad_norm": 4.177949875339217, "learning_rate": 4.095e-05, "loss": 0.397, "step": 273 }, { "epoch": 0.12894117647058823, "grad_norm": 1.6432823762141175, "learning_rate": 4.11e-05, "loss": 0.1944, "step": 274 }, { "epoch": 0.12941176470588237, "grad_norm": 0.5860992823529728, "learning_rate": 4.125e-05, "loss": 0.0444, "step": 275 }, { "epoch": 0.12988235294117648, "grad_norm": 1.5793640414910255, "learning_rate": 4.14e-05, "loss": 0.1898, "step": 276 }, { "epoch": 0.1303529411764706, "grad_norm": 1.225845880725593, "learning_rate": 4.155e-05, "loss": 0.1064, "step": 277 }, { "epoch": 0.1308235294117647, "grad_norm": 2.1914380352485363, "learning_rate": 4.17e-05, "loss": 0.2589, "step": 278 }, { "epoch": 0.1312941176470588, "grad_norm": 0.5168226702552925, "learning_rate": 4.185e-05, "loss": 0.0438, "step": 279 }, { "epoch": 0.13176470588235295, "grad_norm": 1.0320396834300776, "learning_rate": 4.2e-05, "loss": 0.1525, "step": 280 }, { "epoch": 0.13223529411764706, "grad_norm": 0.9265946983516421, "learning_rate": 4.215e-05, "loss": 0.1306, "step": 281 }, { "epoch": 0.13270588235294117, "grad_norm": 1.348911151104915, "learning_rate": 4.229999999999999e-05, "loss": 0.2496, "step": 282 }, { "epoch": 0.13317647058823529, "grad_norm": 2.3062686908554437, "learning_rate": 4.2449999999999995e-05, "loss": 0.3533, "step": 283 }, { "epoch": 0.13364705882352942, "grad_norm": 1.5091097894141674, "learning_rate": 4.259999999999999e-05, "loss": 0.3019, "step": 284 }, { "epoch": 0.13411764705882354, "grad_norm": 1.1653920079282394, "learning_rate": 4.2749999999999996e-05, "loss": 0.2331, "step": 285 }, { "epoch": 0.13458823529411765, "grad_norm": 1.4876708949666153, "learning_rate": 4.289999999999999e-05, "loss": 0.4222, "step": 286 }, { "epoch": 0.13505882352941176, "grad_norm": 1.5186870525161913, "learning_rate": 4.3049999999999996e-05, "loss": 0.2481, "step": 287 }, { "epoch": 0.13552941176470587, "grad_norm": 1.8903031866078581, "learning_rate": 4.319999999999999e-05, "loss": 0.4484, "step": 288 }, { "epoch": 0.136, "grad_norm": 1.1716779496889422, "learning_rate": 4.334999999999999e-05, "loss": 0.2071, "step": 289 }, { "epoch": 0.13647058823529412, "grad_norm": 1.602707591576776, "learning_rate": 4.3499999999999993e-05, "loss": 0.2907, "step": 290 }, { "epoch": 0.13694117647058823, "grad_norm": 16.159639426505503, "learning_rate": 4.364999999999999e-05, "loss": 0.2934, "step": 291 }, { "epoch": 0.13741176470588234, "grad_norm": 1.227460627863373, "learning_rate": 4.3799999999999994e-05, "loss": 0.1813, "step": 292 }, { "epoch": 0.13788235294117648, "grad_norm": 1.708777925889935, "learning_rate": 4.394999999999999e-05, "loss": 0.1498, "step": 293 }, { "epoch": 0.1383529411764706, "grad_norm": 3.404207123892332, "learning_rate": 4.4099999999999995e-05, "loss": 0.4207, "step": 294 }, { "epoch": 0.1388235294117647, "grad_norm": 1.7882772710303372, "learning_rate": 4.424999999999999e-05, "loss": 0.1001, "step": 295 }, { "epoch": 0.13929411764705882, "grad_norm": 1.1795690636613836, "learning_rate": 4.4399999999999995e-05, "loss": 0.0674, "step": 296 }, { "epoch": 0.13976470588235293, "grad_norm": 1.1363681682085809, "learning_rate": 4.454999999999999e-05, "loss": 0.0881, "step": 297 }, { "epoch": 0.14023529411764707, "grad_norm": 1.4071460336097827, "learning_rate": 4.4699999999999996e-05, "loss": 0.0649, "step": 298 }, { "epoch": 0.14070588235294118, "grad_norm": 2.999416988870933, "learning_rate": 4.484999999999999e-05, "loss": 0.2748, "step": 299 }, { "epoch": 0.1411764705882353, "grad_norm": 2.158854171998822, "learning_rate": 4.4999999999999996e-05, "loss": 0.2764, "step": 300 }, { "epoch": 0.1416470588235294, "grad_norm": 2.2426556646514206, "learning_rate": 4.514999999999999e-05, "loss": 0.2603, "step": 301 }, { "epoch": 0.14211764705882354, "grad_norm": 4.228044484895601, "learning_rate": 4.5299999999999997e-05, "loss": 0.6165, "step": 302 }, { "epoch": 0.14258823529411765, "grad_norm": 1.7955828187627096, "learning_rate": 4.5449999999999993e-05, "loss": 0.0618, "step": 303 }, { "epoch": 0.14305882352941177, "grad_norm": 7.270216819315916, "learning_rate": 4.56e-05, "loss": 0.5771, "step": 304 }, { "epoch": 0.14352941176470588, "grad_norm": 1.8415450007444616, "learning_rate": 4.5749999999999994e-05, "loss": 0.178, "step": 305 }, { "epoch": 0.144, "grad_norm": 3.116502524126434, "learning_rate": 4.59e-05, "loss": 0.3302, "step": 306 }, { "epoch": 0.14447058823529413, "grad_norm": 3.1812040469339036, "learning_rate": 4.6049999999999994e-05, "loss": 0.354, "step": 307 }, { "epoch": 0.14494117647058824, "grad_norm": 1.9452846615123869, "learning_rate": 4.62e-05, "loss": 0.3721, "step": 308 }, { "epoch": 0.14541176470588235, "grad_norm": 1.18575390997834, "learning_rate": 4.6349999999999995e-05, "loss": 0.1615, "step": 309 }, { "epoch": 0.14588235294117646, "grad_norm": 0.4379829732667185, "learning_rate": 4.65e-05, "loss": 0.028, "step": 310 }, { "epoch": 0.14635294117647057, "grad_norm": 0.9583068165122192, "learning_rate": 4.6649999999999996e-05, "loss": 0.0457, "step": 311 }, { "epoch": 0.1468235294117647, "grad_norm": 4.125778779065409, "learning_rate": 4.68e-05, "loss": 0.3139, "step": 312 }, { "epoch": 0.14729411764705883, "grad_norm": 2.9439885905230603, "learning_rate": 4.6949999999999996e-05, "loss": 0.2698, "step": 313 }, { "epoch": 0.14776470588235294, "grad_norm": 1.0798921095204776, "learning_rate": 4.709999999999999e-05, "loss": 0.1959, "step": 314 }, { "epoch": 0.14823529411764705, "grad_norm": 0.7396382080418201, "learning_rate": 4.7249999999999997e-05, "loss": 0.0451, "step": 315 }, { "epoch": 0.1487058823529412, "grad_norm": 1.428618945322037, "learning_rate": 4.7399999999999993e-05, "loss": 0.1124, "step": 316 }, { "epoch": 0.1491764705882353, "grad_norm": 2.1911871550713236, "learning_rate": 4.755e-05, "loss": 0.1836, "step": 317 }, { "epoch": 0.1496470588235294, "grad_norm": 1.6587427928873837, "learning_rate": 4.7699999999999994e-05, "loss": 0.0764, "step": 318 }, { "epoch": 0.15011764705882352, "grad_norm": 5.174486377274196, "learning_rate": 4.785e-05, "loss": 0.4427, "step": 319 }, { "epoch": 0.15058823529411763, "grad_norm": 1.9495167332522827, "learning_rate": 4.7999999999999994e-05, "loss": 0.4216, "step": 320 }, { "epoch": 0.15105882352941177, "grad_norm": 2.0512957519189308, "learning_rate": 4.815e-05, "loss": 0.2264, "step": 321 }, { "epoch": 0.15152941176470588, "grad_norm": 3.63601540340834, "learning_rate": 4.8299999999999995e-05, "loss": 0.3176, "step": 322 }, { "epoch": 0.152, "grad_norm": 1.793285252678394, "learning_rate": 4.845e-05, "loss": 0.1421, "step": 323 }, { "epoch": 0.1524705882352941, "grad_norm": 1.8549148004459264, "learning_rate": 4.8599999999999995e-05, "loss": 0.2238, "step": 324 }, { "epoch": 0.15294117647058825, "grad_norm": 0.7651246815960717, "learning_rate": 4.875e-05, "loss": 0.072, "step": 325 }, { "epoch": 0.15341176470588236, "grad_norm": 1.100705239056733, "learning_rate": 4.8899999999999996e-05, "loss": 0.0591, "step": 326 }, { "epoch": 0.15388235294117647, "grad_norm": 1.4560185514294732, "learning_rate": 4.905e-05, "loss": 0.1548, "step": 327 }, { "epoch": 0.15435294117647058, "grad_norm": 2.0985161646028914, "learning_rate": 4.9199999999999997e-05, "loss": 0.2925, "step": 328 }, { "epoch": 0.1548235294117647, "grad_norm": 2.3304916040213146, "learning_rate": 4.935e-05, "loss": 0.2146, "step": 329 }, { "epoch": 0.15529411764705883, "grad_norm": 2.9590500216085904, "learning_rate": 4.95e-05, "loss": 0.4112, "step": 330 }, { "epoch": 0.15576470588235294, "grad_norm": 1.0400866852598623, "learning_rate": 4.965e-05, "loss": 0.0911, "step": 331 }, { "epoch": 0.15623529411764706, "grad_norm": 3.669838931491877, "learning_rate": 4.98e-05, "loss": 0.3743, "step": 332 }, { "epoch": 0.15670588235294117, "grad_norm": 1.7691561246297702, "learning_rate": 4.995e-05, "loss": 0.1774, "step": 333 }, { "epoch": 0.1571764705882353, "grad_norm": 1.1140094216922345, "learning_rate": 5.01e-05, "loss": 0.1139, "step": 334 }, { "epoch": 0.15764705882352942, "grad_norm": 0.4314900688586958, "learning_rate": 5.025e-05, "loss": 0.0228, "step": 335 }, { "epoch": 0.15811764705882353, "grad_norm": 1.6254796338159057, "learning_rate": 5.04e-05, "loss": 0.322, "step": 336 }, { "epoch": 0.15858823529411764, "grad_norm": 2.280406813817557, "learning_rate": 5.055e-05, "loss": 0.2684, "step": 337 }, { "epoch": 0.15905882352941175, "grad_norm": 2.701310242074141, "learning_rate": 5.07e-05, "loss": 0.3186, "step": 338 }, { "epoch": 0.1595294117647059, "grad_norm": 1.2143418259551937, "learning_rate": 5.0849999999999996e-05, "loss": 0.184, "step": 339 }, { "epoch": 0.16, "grad_norm": 2.4362584484047667, "learning_rate": 5.1e-05, "loss": 0.3836, "step": 340 }, { "epoch": 0.16047058823529411, "grad_norm": 2.2633974021151473, "learning_rate": 5.1149999999999996e-05, "loss": 0.262, "step": 341 }, { "epoch": 0.16094117647058823, "grad_norm": 1.7714531998501442, "learning_rate": 5.13e-05, "loss": 0.2529, "step": 342 }, { "epoch": 0.16141176470588237, "grad_norm": 2.1275639190924456, "learning_rate": 5.145e-05, "loss": 0.4102, "step": 343 }, { "epoch": 0.16188235294117648, "grad_norm": 2.1633891259989166, "learning_rate": 5.1599999999999994e-05, "loss": 0.3381, "step": 344 }, { "epoch": 0.1623529411764706, "grad_norm": 1.9012672731955844, "learning_rate": 5.174999999999999e-05, "loss": 0.103, "step": 345 }, { "epoch": 0.1628235294117647, "grad_norm": 2.460933449352158, "learning_rate": 5.1899999999999994e-05, "loss": 0.2548, "step": 346 }, { "epoch": 0.1632941176470588, "grad_norm": 1.1743784803797372, "learning_rate": 5.204999999999999e-05, "loss": 0.1536, "step": 347 }, { "epoch": 0.16376470588235295, "grad_norm": 0.15609734858159321, "learning_rate": 5.2199999999999995e-05, "loss": 0.0057, "step": 348 }, { "epoch": 0.16423529411764706, "grad_norm": 1.5065090672080397, "learning_rate": 5.234999999999999e-05, "loss": 0.2064, "step": 349 }, { "epoch": 0.16470588235294117, "grad_norm": 1.4547509944729717, "learning_rate": 5.2499999999999995e-05, "loss": 0.1967, "step": 350 }, { "epoch": 0.16517647058823529, "grad_norm": 2.2352871059203747, "learning_rate": 5.264999999999999e-05, "loss": 0.2383, "step": 351 }, { "epoch": 0.16564705882352943, "grad_norm": 2.2918458660457395, "learning_rate": 5.279999999999999e-05, "loss": 0.3166, "step": 352 }, { "epoch": 0.16611764705882354, "grad_norm": 1.0355881577639126, "learning_rate": 5.294999999999999e-05, "loss": 0.1395, "step": 353 }, { "epoch": 0.16658823529411765, "grad_norm": 1.0956702189067726, "learning_rate": 5.309999999999999e-05, "loss": 0.0726, "step": 354 }, { "epoch": 0.16705882352941176, "grad_norm": 1.161808343321967, "learning_rate": 5.324999999999999e-05, "loss": 0.0595, "step": 355 }, { "epoch": 0.16752941176470587, "grad_norm": 1.4929329101453377, "learning_rate": 5.339999999999999e-05, "loss": 0.0963, "step": 356 }, { "epoch": 0.168, "grad_norm": 2.5361254601832006, "learning_rate": 5.3549999999999994e-05, "loss": 0.1991, "step": 357 }, { "epoch": 0.16847058823529412, "grad_norm": 0.9318616844915913, "learning_rate": 5.369999999999999e-05, "loss": 0.1219, "step": 358 }, { "epoch": 0.16894117647058823, "grad_norm": 1.4500036704914077, "learning_rate": 5.3849999999999994e-05, "loss": 0.0919, "step": 359 }, { "epoch": 0.16941176470588235, "grad_norm": 0.612228100454981, "learning_rate": 5.399999999999999e-05, "loss": 0.0506, "step": 360 }, { "epoch": 0.16988235294117648, "grad_norm": 2.7488943335766454, "learning_rate": 5.4149999999999995e-05, "loss": 0.2553, "step": 361 }, { "epoch": 0.1703529411764706, "grad_norm": 1.7242854320271144, "learning_rate": 5.429999999999999e-05, "loss": 0.1248, "step": 362 }, { "epoch": 0.1708235294117647, "grad_norm": 1.0893534401619753, "learning_rate": 5.4449999999999995e-05, "loss": 0.1479, "step": 363 }, { "epoch": 0.17129411764705882, "grad_norm": 1.420178748641406, "learning_rate": 5.459999999999999e-05, "loss": 0.2493, "step": 364 }, { "epoch": 0.17176470588235293, "grad_norm": 0.7644348366833493, "learning_rate": 5.4749999999999996e-05, "loss": 0.0775, "step": 365 }, { "epoch": 0.17223529411764707, "grad_norm": 0.6339681819039006, "learning_rate": 5.489999999999999e-05, "loss": 0.0358, "step": 366 }, { "epoch": 0.17270588235294118, "grad_norm": 2.913009701566078, "learning_rate": 5.5049999999999996e-05, "loss": 0.2523, "step": 367 }, { "epoch": 0.1731764705882353, "grad_norm": 1.6129545704622683, "learning_rate": 5.519999999999999e-05, "loss": 0.1781, "step": 368 }, { "epoch": 0.1736470588235294, "grad_norm": 1.5502632022139473, "learning_rate": 5.535e-05, "loss": 0.3184, "step": 369 }, { "epoch": 0.17411764705882352, "grad_norm": 1.3196044742243065, "learning_rate": 5.5499999999999994e-05, "loss": 0.2718, "step": 370 }, { "epoch": 0.17458823529411766, "grad_norm": 1.7087033624728156, "learning_rate": 5.565e-05, "loss": 0.2053, "step": 371 }, { "epoch": 0.17505882352941177, "grad_norm": 0.9059008253971796, "learning_rate": 5.5799999999999994e-05, "loss": 0.0818, "step": 372 }, { "epoch": 0.17552941176470588, "grad_norm": 0.5177849111670224, "learning_rate": 5.595e-05, "loss": 0.0261, "step": 373 }, { "epoch": 0.176, "grad_norm": 1.866949419350413, "learning_rate": 5.6099999999999995e-05, "loss": 0.2197, "step": 374 }, { "epoch": 0.17647058823529413, "grad_norm": 2.963658765341423, "learning_rate": 5.625e-05, "loss": 0.4103, "step": 375 }, { "epoch": 0.17694117647058824, "grad_norm": 2.595132488199016, "learning_rate": 5.6399999999999995e-05, "loss": 0.5955, "step": 376 }, { "epoch": 0.17741176470588235, "grad_norm": 1.6752143165941762, "learning_rate": 5.654999999999999e-05, "loss": 0.2577, "step": 377 }, { "epoch": 0.17788235294117646, "grad_norm": 0.9047606881199116, "learning_rate": 5.6699999999999996e-05, "loss": 0.0684, "step": 378 }, { "epoch": 0.17835294117647058, "grad_norm": 0.6747015433235357, "learning_rate": 5.684999999999999e-05, "loss": 0.0779, "step": 379 }, { "epoch": 0.17882352941176471, "grad_norm": 2.115517456765072, "learning_rate": 5.6999999999999996e-05, "loss": 0.2946, "step": 380 }, { "epoch": 0.17929411764705883, "grad_norm": 1.7591162323214773, "learning_rate": 5.714999999999999e-05, "loss": 0.1764, "step": 381 }, { "epoch": 0.17976470588235294, "grad_norm": 1.7917316822878186, "learning_rate": 5.73e-05, "loss": 0.223, "step": 382 }, { "epoch": 0.18023529411764705, "grad_norm": 0.22443396380833391, "learning_rate": 5.7449999999999994e-05, "loss": 0.0175, "step": 383 }, { "epoch": 0.1807058823529412, "grad_norm": 3.091524103278536, "learning_rate": 5.76e-05, "loss": 0.3792, "step": 384 }, { "epoch": 0.1811764705882353, "grad_norm": 2.236449695019339, "learning_rate": 5.7749999999999994e-05, "loss": 0.2933, "step": 385 }, { "epoch": 0.1816470588235294, "grad_norm": 1.8927467690583584, "learning_rate": 5.79e-05, "loss": 0.2551, "step": 386 }, { "epoch": 0.18211764705882352, "grad_norm": 0.44285341522400956, "learning_rate": 5.8049999999999995e-05, "loss": 0.0372, "step": 387 }, { "epoch": 0.18258823529411763, "grad_norm": 3.8176580493766368, "learning_rate": 5.82e-05, "loss": 0.3771, "step": 388 }, { "epoch": 0.18305882352941177, "grad_norm": 0.8989408818801315, "learning_rate": 5.8349999999999995e-05, "loss": 0.1557, "step": 389 }, { "epoch": 0.18352941176470589, "grad_norm": 1.6438440220106285, "learning_rate": 5.85e-05, "loss": 0.2629, "step": 390 }, { "epoch": 0.184, "grad_norm": 1.4569295895849907, "learning_rate": 5.8649999999999996e-05, "loss": 0.3168, "step": 391 }, { "epoch": 0.1844705882352941, "grad_norm": 0.6423644157988085, "learning_rate": 5.88e-05, "loss": 0.0816, "step": 392 }, { "epoch": 0.18494117647058825, "grad_norm": 1.9879907859319077, "learning_rate": 5.8949999999999996e-05, "loss": 0.4175, "step": 393 }, { "epoch": 0.18541176470588236, "grad_norm": 0.8439555429741288, "learning_rate": 5.91e-05, "loss": 0.0923, "step": 394 }, { "epoch": 0.18588235294117647, "grad_norm": 0.9051610749975613, "learning_rate": 5.925e-05, "loss": 0.1215, "step": 395 }, { "epoch": 0.18635294117647058, "grad_norm": 2.266312043704595, "learning_rate": 5.94e-05, "loss": 0.4363, "step": 396 }, { "epoch": 0.1868235294117647, "grad_norm": 0.9036236230226791, "learning_rate": 5.955e-05, "loss": 0.1175, "step": 397 }, { "epoch": 0.18729411764705883, "grad_norm": 0.43020924039187153, "learning_rate": 5.97e-05, "loss": 0.0412, "step": 398 }, { "epoch": 0.18776470588235294, "grad_norm": 1.3805974429657877, "learning_rate": 5.985e-05, "loss": 0.2887, "step": 399 }, { "epoch": 0.18823529411764706, "grad_norm": 2.1901164027937945, "learning_rate": 5.9999999999999995e-05, "loss": 0.0782, "step": 400 }, { "epoch": 0.18870588235294117, "grad_norm": 0.7906654318599001, "learning_rate": 6.015e-05, "loss": 0.0807, "step": 401 }, { "epoch": 0.1891764705882353, "grad_norm": 1.127850032118891, "learning_rate": 6.0299999999999995e-05, "loss": 0.0907, "step": 402 }, { "epoch": 0.18964705882352942, "grad_norm": 1.750286784630517, "learning_rate": 6.045e-05, "loss": 0.3195, "step": 403 }, { "epoch": 0.19011764705882353, "grad_norm": 3.117800556878295, "learning_rate": 6.0599999999999996e-05, "loss": 0.5474, "step": 404 }, { "epoch": 0.19058823529411764, "grad_norm": 0.3656106758465673, "learning_rate": 6.075e-05, "loss": 0.0262, "step": 405 }, { "epoch": 0.19105882352941175, "grad_norm": 2.0242477268539987, "learning_rate": 6.0899999999999996e-05, "loss": 0.5465, "step": 406 }, { "epoch": 0.1915294117647059, "grad_norm": 1.3592511457675214, "learning_rate": 6.104999999999999e-05, "loss": 0.2236, "step": 407 }, { "epoch": 0.192, "grad_norm": 0.5608665387520767, "learning_rate": 6.12e-05, "loss": 0.0418, "step": 408 }, { "epoch": 0.19247058823529412, "grad_norm": 1.0558667694603017, "learning_rate": 6.134999999999999e-05, "loss": 0.2175, "step": 409 }, { "epoch": 0.19294117647058823, "grad_norm": 2.0052953217847804, "learning_rate": 6.149999999999999e-05, "loss": 0.3544, "step": 410 }, { "epoch": 0.19341176470588237, "grad_norm": 1.9396081658738138, "learning_rate": 6.165e-05, "loss": 0.299, "step": 411 }, { "epoch": 0.19388235294117648, "grad_norm": 0.7169434426792053, "learning_rate": 6.18e-05, "loss": 0.0777, "step": 412 }, { "epoch": 0.1943529411764706, "grad_norm": 0.8041938229376934, "learning_rate": 6.194999999999999e-05, "loss": 0.077, "step": 413 }, { "epoch": 0.1948235294117647, "grad_norm": 0.9101244499180058, "learning_rate": 6.209999999999999e-05, "loss": 0.0422, "step": 414 }, { "epoch": 0.1952941176470588, "grad_norm": 0.6887541262756884, "learning_rate": 6.225e-05, "loss": 0.0341, "step": 415 }, { "epoch": 0.19576470588235295, "grad_norm": 4.307004021138938, "learning_rate": 6.239999999999999e-05, "loss": 0.5819, "step": 416 }, { "epoch": 0.19623529411764706, "grad_norm": 1.62158121709159, "learning_rate": 6.254999999999999e-05, "loss": 0.3043, "step": 417 }, { "epoch": 0.19670588235294117, "grad_norm": 1.1695221269487688, "learning_rate": 6.269999999999999e-05, "loss": 0.1147, "step": 418 }, { "epoch": 0.1971764705882353, "grad_norm": 2.4603506759859934, "learning_rate": 6.285e-05, "loss": 0.2885, "step": 419 }, { "epoch": 0.1976470588235294, "grad_norm": 0.8816894538806833, "learning_rate": 6.299999999999999e-05, "loss": 0.0711, "step": 420 }, { "epoch": 0.19811764705882354, "grad_norm": 0.9069340707310544, "learning_rate": 6.314999999999999e-05, "loss": 0.1334, "step": 421 }, { "epoch": 0.19858823529411765, "grad_norm": 2.6355242341238205, "learning_rate": 6.33e-05, "loss": 0.3465, "step": 422 }, { "epoch": 0.19905882352941176, "grad_norm": 11.571739566466071, "learning_rate": 6.345e-05, "loss": 0.0441, "step": 423 }, { "epoch": 0.19952941176470587, "grad_norm": 1.315458227298009, "learning_rate": 6.359999999999999e-05, "loss": 0.309, "step": 424 }, { "epoch": 0.2, "grad_norm": 0.6203416677792172, "learning_rate": 6.374999999999999e-05, "loss": 0.0727, "step": 425 }, { "epoch": 0.20047058823529412, "grad_norm": 10.161415978981537, "learning_rate": 6.39e-05, "loss": 0.3907, "step": 426 }, { "epoch": 0.20094117647058823, "grad_norm": 1.6442442786670384, "learning_rate": 6.405e-05, "loss": 0.2806, "step": 427 }, { "epoch": 0.20141176470588235, "grad_norm": 1.5017563220228973, "learning_rate": 6.419999999999999e-05, "loss": 0.276, "step": 428 }, { "epoch": 0.20188235294117646, "grad_norm": 1.1000715280080484, "learning_rate": 6.434999999999999e-05, "loss": 0.1711, "step": 429 }, { "epoch": 0.2023529411764706, "grad_norm": 1.6679246525852351, "learning_rate": 6.45e-05, "loss": 0.2406, "step": 430 }, { "epoch": 0.2028235294117647, "grad_norm": 0.7429803344599861, "learning_rate": 6.465e-05, "loss": 0.0616, "step": 431 }, { "epoch": 0.20329411764705882, "grad_norm": 2.272720163845265, "learning_rate": 6.479999999999999e-05, "loss": 0.2246, "step": 432 }, { "epoch": 0.20376470588235293, "grad_norm": 0.7806605078536624, "learning_rate": 6.494999999999999e-05, "loss": 0.0262, "step": 433 }, { "epoch": 0.20423529411764707, "grad_norm": 2.754408406741543, "learning_rate": 6.51e-05, "loss": 0.1627, "step": 434 }, { "epoch": 0.20470588235294118, "grad_norm": 2.1851551416200183, "learning_rate": 6.525e-05, "loss": 0.2892, "step": 435 }, { "epoch": 0.2051764705882353, "grad_norm": 0.8584418637957998, "learning_rate": 6.539999999999999e-05, "loss": 0.074, "step": 436 }, { "epoch": 0.2056470588235294, "grad_norm": 2.6729868123554463, "learning_rate": 6.555e-05, "loss": 0.3139, "step": 437 }, { "epoch": 0.20611764705882352, "grad_norm": 1.3546036919063886, "learning_rate": 6.57e-05, "loss": 0.1945, "step": 438 }, { "epoch": 0.20658823529411766, "grad_norm": 1.534457328140222, "learning_rate": 6.584999999999999e-05, "loss": 0.2078, "step": 439 }, { "epoch": 0.20705882352941177, "grad_norm": 1.5241338470746613, "learning_rate": 6.599999999999999e-05, "loss": 0.2227, "step": 440 }, { "epoch": 0.20752941176470588, "grad_norm": 1.2862937102242658, "learning_rate": 6.615e-05, "loss": 0.1627, "step": 441 }, { "epoch": 0.208, "grad_norm": 3.5170410977312505, "learning_rate": 6.63e-05, "loss": 0.2268, "step": 442 }, { "epoch": 0.20847058823529413, "grad_norm": 14.394858331944432, "learning_rate": 6.644999999999999e-05, "loss": 0.2317, "step": 443 }, { "epoch": 0.20894117647058824, "grad_norm": 1.3508408439032409, "learning_rate": 6.659999999999999e-05, "loss": 0.0905, "step": 444 }, { "epoch": 0.20941176470588235, "grad_norm": 1.5239670797972658, "learning_rate": 6.675e-05, "loss": 0.3549, "step": 445 }, { "epoch": 0.20988235294117646, "grad_norm": 16.550205702300737, "learning_rate": 6.69e-05, "loss": 0.1002, "step": 446 }, { "epoch": 0.21035294117647058, "grad_norm": 1.2075607659012921, "learning_rate": 6.704999999999999e-05, "loss": 0.1461, "step": 447 }, { "epoch": 0.21082352941176472, "grad_norm": 84.03476606431452, "learning_rate": 6.72e-05, "loss": 0.4381, "step": 448 }, { "epoch": 0.21129411764705883, "grad_norm": 2.9599340615405176, "learning_rate": 6.735e-05, "loss": 0.3175, "step": 449 }, { "epoch": 0.21176470588235294, "grad_norm": 8.437226127461965, "learning_rate": 6.75e-05, "loss": 0.4525, "step": 450 }, { "epoch": 0.21223529411764705, "grad_norm": 1.2029459498432868, "learning_rate": 6.764999999999999e-05, "loss": 0.2085, "step": 451 }, { "epoch": 0.2127058823529412, "grad_norm": 2.6541449735210074, "learning_rate": 6.78e-05, "loss": 0.3506, "step": 452 }, { "epoch": 0.2131764705882353, "grad_norm": 0.7522278344802578, "learning_rate": 6.795e-05, "loss": 0.0495, "step": 453 }, { "epoch": 0.2136470588235294, "grad_norm": 2.3981761901724834, "learning_rate": 6.81e-05, "loss": 0.3138, "step": 454 }, { "epoch": 0.21411764705882352, "grad_norm": 1430.2969122751938, "learning_rate": 6.824999999999999e-05, "loss": 27.1428, "step": 455 }, { "epoch": 0.21458823529411764, "grad_norm": 365.60718394723335, "learning_rate": 6.84e-05, "loss": 7.262, "step": 456 }, { "epoch": 0.21505882352941177, "grad_norm": 3.554333621186873, "learning_rate": 6.855e-05, "loss": 0.2762, "step": 457 }, { "epoch": 0.21552941176470589, "grad_norm": 3.863309473648501, "learning_rate": 6.87e-05, "loss": 0.2826, "step": 458 }, { "epoch": 0.216, "grad_norm": 5.571009864686323, "learning_rate": 6.884999999999999e-05, "loss": 0.6177, "step": 459 }, { "epoch": 0.2164705882352941, "grad_norm": 7.66595325863811, "learning_rate": 6.9e-05, "loss": 0.7301, "step": 460 }, { "epoch": 0.21694117647058825, "grad_norm": 5.475841387725096, "learning_rate": 6.915e-05, "loss": 0.6748, "step": 461 }, { "epoch": 0.21741176470588236, "grad_norm": 4.6243445172485345, "learning_rate": 6.93e-05, "loss": 0.425, "step": 462 }, { "epoch": 0.21788235294117647, "grad_norm": 1.8385038432111478, "learning_rate": 6.945e-05, "loss": 0.1826, "step": 463 }, { "epoch": 0.21835294117647058, "grad_norm": 4.975768381265682, "learning_rate": 6.96e-05, "loss": 0.6052, "step": 464 }, { "epoch": 0.2188235294117647, "grad_norm": 1.243963478089005, "learning_rate": 6.975e-05, "loss": 0.0752, "step": 465 }, { "epoch": 0.21929411764705883, "grad_norm": 2.1816322390366096, "learning_rate": 6.989999999999999e-05, "loss": 0.2036, "step": 466 }, { "epoch": 0.21976470588235295, "grad_norm": 4.063873906578696, "learning_rate": 7.005e-05, "loss": 0.7568, "step": 467 }, { "epoch": 0.22023529411764706, "grad_norm": 2.39494052791162, "learning_rate": 7.02e-05, "loss": 0.2776, "step": 468 }, { "epoch": 0.22070588235294117, "grad_norm": 2.381595308727214, "learning_rate": 7.034999999999999e-05, "loss": 0.2904, "step": 469 }, { "epoch": 0.2211764705882353, "grad_norm": 1.7121193538610036, "learning_rate": 7.049999999999999e-05, "loss": 0.2648, "step": 470 }, { "epoch": 0.22164705882352942, "grad_norm": 1.9931518753900448, "learning_rate": 7.065e-05, "loss": 0.4402, "step": 471 }, { "epoch": 0.22211764705882353, "grad_norm": 1.8986279135657755, "learning_rate": 7.079999999999999e-05, "loss": 0.4153, "step": 472 }, { "epoch": 0.22258823529411764, "grad_norm": 2.087490577311031, "learning_rate": 7.094999999999999e-05, "loss": 0.3072, "step": 473 }, { "epoch": 0.22305882352941175, "grad_norm": 0.9244068315070885, "learning_rate": 7.11e-05, "loss": 0.0478, "step": 474 }, { "epoch": 0.2235294117647059, "grad_norm": 1.7495235843796388, "learning_rate": 7.125e-05, "loss": 0.3981, "step": 475 }, { "epoch": 0.224, "grad_norm": 2.7431275829168738, "learning_rate": 7.139999999999999e-05, "loss": 0.4285, "step": 476 }, { "epoch": 0.22447058823529412, "grad_norm": 2.6786005571431426, "learning_rate": 7.154999999999999e-05, "loss": 0.5675, "step": 477 }, { "epoch": 0.22494117647058823, "grad_norm": 1.503165954875519, "learning_rate": 7.17e-05, "loss": 0.2404, "step": 478 }, { "epoch": 0.22541176470588234, "grad_norm": 1.3736162260803517, "learning_rate": 7.184999999999998e-05, "loss": 0.1078, "step": 479 }, { "epoch": 0.22588235294117648, "grad_norm": 1.038575764127082, "learning_rate": 7.199999999999999e-05, "loss": 0.1425, "step": 480 }, { "epoch": 0.2263529411764706, "grad_norm": 4.08545879260648, "learning_rate": 7.214999999999999e-05, "loss": 0.6124, "step": 481 }, { "epoch": 0.2268235294117647, "grad_norm": 1.8244130839595094, "learning_rate": 7.23e-05, "loss": 0.3034, "step": 482 }, { "epoch": 0.2272941176470588, "grad_norm": 1.3621123127147718, "learning_rate": 7.244999999999999e-05, "loss": 0.2855, "step": 483 }, { "epoch": 0.22776470588235295, "grad_norm": 1.597207411287223, "learning_rate": 7.259999999999999e-05, "loss": 0.3247, "step": 484 }, { "epoch": 0.22823529411764706, "grad_norm": 1.45198796240823, "learning_rate": 7.274999999999999e-05, "loss": 0.1341, "step": 485 }, { "epoch": 0.22870588235294118, "grad_norm": 1.710063523713283, "learning_rate": 7.29e-05, "loss": 0.182, "step": 486 }, { "epoch": 0.2291764705882353, "grad_norm": 433.42126930914424, "learning_rate": 7.304999999999999e-05, "loss": 0.9364, "step": 487 }, { "epoch": 0.2296470588235294, "grad_norm": 270.1935865047002, "learning_rate": 7.319999999999999e-05, "loss": 0.6243, "step": 488 }, { "epoch": 0.23011764705882354, "grad_norm": 1.8398653937994704, "learning_rate": 7.335e-05, "loss": 0.1617, "step": 489 }, { "epoch": 0.23058823529411765, "grad_norm": 1.8660883721757693, "learning_rate": 7.35e-05, "loss": 0.2446, "step": 490 }, { "epoch": 0.23105882352941176, "grad_norm": 0.9616258770053021, "learning_rate": 7.364999999999999e-05, "loss": 0.0762, "step": 491 }, { "epoch": 0.23152941176470587, "grad_norm": 4.018107573073293, "learning_rate": 7.379999999999999e-05, "loss": 0.5089, "step": 492 }, { "epoch": 0.232, "grad_norm": 3.6821603600052044, "learning_rate": 7.395e-05, "loss": 0.5618, "step": 493 }, { "epoch": 0.23247058823529412, "grad_norm": 1.867372217331949, "learning_rate": 7.41e-05, "loss": 0.203, "step": 494 }, { "epoch": 0.23294117647058823, "grad_norm": 3.5066303595351345, "learning_rate": 7.424999999999999e-05, "loss": 0.5821, "step": 495 }, { "epoch": 0.23341176470588235, "grad_norm": 1.8462227191048284, "learning_rate": 7.439999999999999e-05, "loss": 0.2983, "step": 496 }, { "epoch": 0.23388235294117646, "grad_norm": 2.739377256988044, "learning_rate": 7.455e-05, "loss": 0.4341, "step": 497 }, { "epoch": 0.2343529411764706, "grad_norm": 1.1158498988216072, "learning_rate": 7.47e-05, "loss": 0.2136, "step": 498 }, { "epoch": 0.2348235294117647, "grad_norm": 2.8182512405107345, "learning_rate": 7.484999999999999e-05, "loss": 0.3548, "step": 499 }, { "epoch": 0.23529411764705882, "grad_norm": 1.600694193279675, "learning_rate": 7.5e-05, "loss": 0.1778, "step": 500 }, { "epoch": 0.23576470588235293, "grad_norm": 1.0932128910999603, "learning_rate": 7.515e-05, "loss": 0.206, "step": 501 }, { "epoch": 0.23623529411764707, "grad_norm": 2.1045597489238492, "learning_rate": 7.529999999999999e-05, "loss": 0.2052, "step": 502 }, { "epoch": 0.23670588235294118, "grad_norm": 1.6807021996807676, "learning_rate": 7.544999999999999e-05, "loss": 0.2098, "step": 503 }, { "epoch": 0.2371764705882353, "grad_norm": 1.591517053940264, "learning_rate": 7.56e-05, "loss": 0.1642, "step": 504 }, { "epoch": 0.2376470588235294, "grad_norm": 3.09688470456185, "learning_rate": 7.575e-05, "loss": 0.6541, "step": 505 }, { "epoch": 0.23811764705882352, "grad_norm": 1.1565560703177622, "learning_rate": 7.589999999999999e-05, "loss": 0.1364, "step": 506 }, { "epoch": 0.23858823529411766, "grad_norm": 2.7009322477342264, "learning_rate": 7.604999999999999e-05, "loss": 0.2838, "step": 507 }, { "epoch": 0.23905882352941177, "grad_norm": 0.6200113639352055, "learning_rate": 7.62e-05, "loss": 0.0812, "step": 508 }, { "epoch": 0.23952941176470588, "grad_norm": 3.8941468714012726, "learning_rate": 7.635e-05, "loss": 0.1876, "step": 509 }, { "epoch": 0.24, "grad_norm": 42.07801326650286, "learning_rate": 7.649999999999999e-05, "loss": 0.4836, "step": 510 }, { "epoch": 0.24047058823529413, "grad_norm": 2.1051629121499165, "learning_rate": 7.664999999999999e-05, "loss": 0.4155, "step": 511 }, { "epoch": 0.24094117647058824, "grad_norm": 2.2255244203806877, "learning_rate": 7.68e-05, "loss": 0.1127, "step": 512 }, { "epoch": 0.24141176470588235, "grad_norm": 4.9974932811377535, "learning_rate": 7.695e-05, "loss": 0.2556, "step": 513 }, { "epoch": 0.24188235294117647, "grad_norm": 1.4611226090522698, "learning_rate": 7.709999999999999e-05, "loss": 0.1464, "step": 514 }, { "epoch": 0.24235294117647058, "grad_norm": 1.7713612068764655, "learning_rate": 7.725e-05, "loss": 0.2991, "step": 515 }, { "epoch": 0.24282352941176472, "grad_norm": 1.3826257800195463, "learning_rate": 7.74e-05, "loss": 0.2569, "step": 516 }, { "epoch": 0.24329411764705883, "grad_norm": 1.939563161666686, "learning_rate": 7.755e-05, "loss": 0.2849, "step": 517 }, { "epoch": 0.24376470588235294, "grad_norm": 1.5743354670167318, "learning_rate": 7.769999999999999e-05, "loss": 0.1499, "step": 518 }, { "epoch": 0.24423529411764705, "grad_norm": 1.5182570611738717, "learning_rate": 7.785e-05, "loss": 0.2839, "step": 519 }, { "epoch": 0.2447058823529412, "grad_norm": 1.1451147917146725, "learning_rate": 7.8e-05, "loss": 0.1391, "step": 520 }, { "epoch": 0.2451764705882353, "grad_norm": 1.9210534777900268, "learning_rate": 7.815e-05, "loss": 0.2732, "step": 521 }, { "epoch": 0.2456470588235294, "grad_norm": 1.0079812247725533, "learning_rate": 7.829999999999999e-05, "loss": 0.1247, "step": 522 }, { "epoch": 0.24611764705882352, "grad_norm": 1.8959318549328266, "learning_rate": 7.845e-05, "loss": 0.2853, "step": 523 }, { "epoch": 0.24658823529411764, "grad_norm": 2.1797388064251657, "learning_rate": 7.86e-05, "loss": 0.3207, "step": 524 }, { "epoch": 0.24705882352941178, "grad_norm": 0.7663755758270961, "learning_rate": 7.874999999999999e-05, "loss": 0.0511, "step": 525 }, { "epoch": 0.2475294117647059, "grad_norm": 1.3035445304180904, "learning_rate": 7.89e-05, "loss": 0.3487, "step": 526 }, { "epoch": 0.248, "grad_norm": 1.0423095563689488, "learning_rate": 7.905e-05, "loss": 0.0918, "step": 527 }, { "epoch": 0.2484705882352941, "grad_norm": 2.0690767736994404, "learning_rate": 7.92e-05, "loss": 0.499, "step": 528 }, { "epoch": 0.24894117647058825, "grad_norm": 0.7617344019219481, "learning_rate": 7.934999999999999e-05, "loss": 0.1323, "step": 529 }, { "epoch": 0.24941176470588236, "grad_norm": 2.5861861887702524, "learning_rate": 7.95e-05, "loss": 0.7709, "step": 530 }, { "epoch": 0.24988235294117647, "grad_norm": 0.7458073526542021, "learning_rate": 7.965e-05, "loss": 0.1867, "step": 531 }, { "epoch": 0.2503529411764706, "grad_norm": 0.2704392799286716, "learning_rate": 7.98e-05, "loss": 0.0126, "step": 532 }, { "epoch": 0.2508235294117647, "grad_norm": 1.3192230594405352, "learning_rate": 7.994999999999999e-05, "loss": 0.3641, "step": 533 }, { "epoch": 0.25129411764705883, "grad_norm": 0.762665507721129, "learning_rate": 8.01e-05, "loss": 0.1543, "step": 534 }, { "epoch": 0.25176470588235295, "grad_norm": 1.146712966528985, "learning_rate": 8.025e-05, "loss": 0.335, "step": 535 }, { "epoch": 0.25223529411764706, "grad_norm": 1.1894379392377112, "learning_rate": 8.04e-05, "loss": 0.1604, "step": 536 }, { "epoch": 0.25270588235294117, "grad_norm": 0.9591519788024899, "learning_rate": 8.054999999999999e-05, "loss": 0.224, "step": 537 }, { "epoch": 0.2531764705882353, "grad_norm": 0.9572138620081773, "learning_rate": 8.07e-05, "loss": 0.2237, "step": 538 }, { "epoch": 0.2536470588235294, "grad_norm": 0.1770830901596665, "learning_rate": 8.085e-05, "loss": 0.0093, "step": 539 }, { "epoch": 0.2541176470588235, "grad_norm": 1.79236530304837, "learning_rate": 8.1e-05, "loss": 0.3378, "step": 540 }, { "epoch": 0.25458823529411767, "grad_norm": 0.9849088733864391, "learning_rate": 8.115e-05, "loss": 0.2432, "step": 541 }, { "epoch": 0.2550588235294118, "grad_norm": 0.665263447879446, "learning_rate": 8.13e-05, "loss": 0.1103, "step": 542 }, { "epoch": 0.2555294117647059, "grad_norm": 0.9272160375494655, "learning_rate": 8.145e-05, "loss": 0.1656, "step": 543 }, { "epoch": 0.256, "grad_norm": 1.6725261634822863, "learning_rate": 8.16e-05, "loss": 0.2288, "step": 544 }, { "epoch": 0.2564705882352941, "grad_norm": 2.204959543243234, "learning_rate": 8.175e-05, "loss": 0.4351, "step": 545 }, { "epoch": 0.25694117647058823, "grad_norm": 1.0187070127503273, "learning_rate": 8.19e-05, "loss": 0.2781, "step": 546 }, { "epoch": 0.25741176470588234, "grad_norm": 3.288097898875031, "learning_rate": 8.205e-05, "loss": 0.3593, "step": 547 }, { "epoch": 0.25788235294117645, "grad_norm": 2.6001015806905348, "learning_rate": 8.22e-05, "loss": 0.4381, "step": 548 }, { "epoch": 0.25835294117647056, "grad_norm": 0.31763115136901193, "learning_rate": 8.235e-05, "loss": 0.0334, "step": 549 }, { "epoch": 0.25882352941176473, "grad_norm": 2.3289304328848006, "learning_rate": 8.25e-05, "loss": 0.216, "step": 550 }, { "epoch": 0.25929411764705884, "grad_norm": 2.376608184820889, "learning_rate": 8.265e-05, "loss": 0.2522, "step": 551 }, { "epoch": 0.25976470588235295, "grad_norm": 0.9234929372161903, "learning_rate": 8.28e-05, "loss": 0.0948, "step": 552 }, { "epoch": 0.26023529411764706, "grad_norm": 1.3955872906336608, "learning_rate": 8.295e-05, "loss": 0.2206, "step": 553 }, { "epoch": 0.2607058823529412, "grad_norm": 1.3484604015391495, "learning_rate": 8.31e-05, "loss": 0.4202, "step": 554 }, { "epoch": 0.2611764705882353, "grad_norm": 2.8125333029838973, "learning_rate": 8.325e-05, "loss": 0.5376, "step": 555 }, { "epoch": 0.2616470588235294, "grad_norm": 1.6600483676448674, "learning_rate": 8.34e-05, "loss": 0.3014, "step": 556 }, { "epoch": 0.2621176470588235, "grad_norm": 0.9980884986139974, "learning_rate": 8.355e-05, "loss": 0.3117, "step": 557 }, { "epoch": 0.2625882352941176, "grad_norm": 0.9214406404308718, "learning_rate": 8.37e-05, "loss": 0.2083, "step": 558 }, { "epoch": 0.2630588235294118, "grad_norm": 2.073883027445651, "learning_rate": 8.385e-05, "loss": 0.2946, "step": 559 }, { "epoch": 0.2635294117647059, "grad_norm": 1.4667478471841633, "learning_rate": 8.4e-05, "loss": 0.2188, "step": 560 }, { "epoch": 0.264, "grad_norm": 0.7923194484627955, "learning_rate": 8.415e-05, "loss": 0.1574, "step": 561 }, { "epoch": 0.2644705882352941, "grad_norm": 0.5052502962676018, "learning_rate": 8.43e-05, "loss": 0.0501, "step": 562 }, { "epoch": 0.26494117647058824, "grad_norm": 0.4107046158564295, "learning_rate": 8.444999999999998e-05, "loss": 0.0574, "step": 563 }, { "epoch": 0.26541176470588235, "grad_norm": 2.4818562839198353, "learning_rate": 8.459999999999998e-05, "loss": 0.202, "step": 564 }, { "epoch": 0.26588235294117646, "grad_norm": 1.7411424401900744, "learning_rate": 8.474999999999999e-05, "loss": 0.2926, "step": 565 }, { "epoch": 0.26635294117647057, "grad_norm": 1.7472224704363724, "learning_rate": 8.489999999999999e-05, "loss": 0.4241, "step": 566 }, { "epoch": 0.2668235294117647, "grad_norm": 2.3853943844183076, "learning_rate": 8.504999999999998e-05, "loss": 0.3935, "step": 567 }, { "epoch": 0.26729411764705885, "grad_norm": 0.8357086590854694, "learning_rate": 8.519999999999998e-05, "loss": 0.0726, "step": 568 }, { "epoch": 0.26776470588235296, "grad_norm": 0.774557103811741, "learning_rate": 8.534999999999999e-05, "loss": 0.0743, "step": 569 }, { "epoch": 0.26823529411764707, "grad_norm": 1.4829612888436454, "learning_rate": 8.549999999999999e-05, "loss": 0.308, "step": 570 }, { "epoch": 0.2687058823529412, "grad_norm": 1.0760039758676951, "learning_rate": 8.564999999999998e-05, "loss": 0.1484, "step": 571 }, { "epoch": 0.2691764705882353, "grad_norm": 3.388715300706427, "learning_rate": 8.579999999999998e-05, "loss": 0.3736, "step": 572 }, { "epoch": 0.2696470588235294, "grad_norm": 0.9412042714564468, "learning_rate": 8.594999999999999e-05, "loss": 0.1114, "step": 573 }, { "epoch": 0.2701176470588235, "grad_norm": 1.6253170123618976, "learning_rate": 8.609999999999999e-05, "loss": 0.2016, "step": 574 }, { "epoch": 0.27058823529411763, "grad_norm": 0.6970324837159821, "learning_rate": 8.624999999999998e-05, "loss": 0.1377, "step": 575 }, { "epoch": 0.27105882352941174, "grad_norm": 2.0695356217086047, "learning_rate": 8.639999999999999e-05, "loss": 0.2853, "step": 576 }, { "epoch": 0.2715294117647059, "grad_norm": 2.178291452083412, "learning_rate": 8.654999999999999e-05, "loss": 0.323, "step": 577 }, { "epoch": 0.272, "grad_norm": 1.0102662647428966, "learning_rate": 8.669999999999998e-05, "loss": 0.2139, "step": 578 }, { "epoch": 0.27247058823529413, "grad_norm": 0.8186253039736947, "learning_rate": 8.684999999999998e-05, "loss": 0.1929, "step": 579 }, { "epoch": 0.27294117647058824, "grad_norm": 1.570196617450508, "learning_rate": 8.699999999999999e-05, "loss": 0.2547, "step": 580 }, { "epoch": 0.27341176470588235, "grad_norm": 0.5102383093076348, "learning_rate": 8.714999999999999e-05, "loss": 0.0684, "step": 581 }, { "epoch": 0.27388235294117647, "grad_norm": 0.8452659646559534, "learning_rate": 8.729999999999998e-05, "loss": 0.1019, "step": 582 }, { "epoch": 0.2743529411764706, "grad_norm": 1.4422511719216848, "learning_rate": 8.744999999999998e-05, "loss": 0.2689, "step": 583 }, { "epoch": 0.2748235294117647, "grad_norm": 1.560559170224838, "learning_rate": 8.759999999999999e-05, "loss": 0.2949, "step": 584 }, { "epoch": 0.2752941176470588, "grad_norm": 1.5784488927595597, "learning_rate": 8.774999999999999e-05, "loss": 0.1884, "step": 585 }, { "epoch": 0.27576470588235297, "grad_norm": 0.3698736904648655, "learning_rate": 8.789999999999998e-05, "loss": 0.0355, "step": 586 }, { "epoch": 0.2762352941176471, "grad_norm": 3.2434041517225287, "learning_rate": 8.804999999999999e-05, "loss": 0.4785, "step": 587 }, { "epoch": 0.2767058823529412, "grad_norm": 1.0538380438234, "learning_rate": 8.819999999999999e-05, "loss": 0.049, "step": 588 }, { "epoch": 0.2771764705882353, "grad_norm": 1.1978614466601876, "learning_rate": 8.834999999999999e-05, "loss": 0.1931, "step": 589 }, { "epoch": 0.2776470588235294, "grad_norm": 3.129613191503247, "learning_rate": 8.849999999999998e-05, "loss": 0.3721, "step": 590 }, { "epoch": 0.2781176470588235, "grad_norm": 0.9460230606427275, "learning_rate": 8.864999999999999e-05, "loss": 0.2059, "step": 591 }, { "epoch": 0.27858823529411764, "grad_norm": 0.2779686141061049, "learning_rate": 8.879999999999999e-05, "loss": 0.0249, "step": 592 }, { "epoch": 0.27905882352941175, "grad_norm": 0.26887874620690627, "learning_rate": 8.895e-05, "loss": 0.0107, "step": 593 }, { "epoch": 0.27952941176470586, "grad_norm": 1.8399336513604725, "learning_rate": 8.909999999999998e-05, "loss": 0.2107, "step": 594 }, { "epoch": 0.28, "grad_norm": 2.6597301501294566, "learning_rate": 8.924999999999999e-05, "loss": 0.371, "step": 595 }, { "epoch": 0.28047058823529414, "grad_norm": 0.775190321478634, "learning_rate": 8.939999999999999e-05, "loss": 0.1754, "step": 596 }, { "epoch": 0.28094117647058825, "grad_norm": 1.1175128938900407, "learning_rate": 8.955e-05, "loss": 0.3625, "step": 597 }, { "epoch": 0.28141176470588236, "grad_norm": 1.4488992920609627, "learning_rate": 8.969999999999998e-05, "loss": 0.1733, "step": 598 }, { "epoch": 0.2818823529411765, "grad_norm": 2.3330515474809728, "learning_rate": 8.984999999999999e-05, "loss": 0.3648, "step": 599 }, { "epoch": 0.2823529411764706, "grad_norm": 0.8154776280124868, "learning_rate": 8.999999999999999e-05, "loss": 0.0943, "step": 600 }, { "epoch": 0.2828235294117647, "grad_norm": 0.777597786384569, "learning_rate": 9.014999999999998e-05, "loss": 0.1274, "step": 601 }, { "epoch": 0.2832941176470588, "grad_norm": 0.36746155031260175, "learning_rate": 9.029999999999999e-05, "loss": 0.0394, "step": 602 }, { "epoch": 0.2837647058823529, "grad_norm": 1.0668342671204427, "learning_rate": 9.044999999999999e-05, "loss": 0.1689, "step": 603 }, { "epoch": 0.2842352941176471, "grad_norm": 0.9033870967765061, "learning_rate": 9.059999999999999e-05, "loss": 0.158, "step": 604 }, { "epoch": 0.2847058823529412, "grad_norm": 1.4470819017527718, "learning_rate": 9.074999999999998e-05, "loss": 0.2161, "step": 605 }, { "epoch": 0.2851764705882353, "grad_norm": 0.46207905281215556, "learning_rate": 9.089999999999999e-05, "loss": 0.0792, "step": 606 }, { "epoch": 0.2856470588235294, "grad_norm": 1.382595649391288, "learning_rate": 9.104999999999999e-05, "loss": 0.4433, "step": 607 }, { "epoch": 0.28611764705882353, "grad_norm": 0.6924145468812933, "learning_rate": 9.12e-05, "loss": 0.0372, "step": 608 }, { "epoch": 0.28658823529411764, "grad_norm": 2.171351238075373, "learning_rate": 9.134999999999998e-05, "loss": 0.2355, "step": 609 }, { "epoch": 0.28705882352941176, "grad_norm": 0.5554465288245505, "learning_rate": 9.149999999999999e-05, "loss": 0.0524, "step": 610 }, { "epoch": 0.28752941176470587, "grad_norm": 1.4255935898735277, "learning_rate": 9.164999999999999e-05, "loss": 0.3369, "step": 611 }, { "epoch": 0.288, "grad_norm": 1.1458791174703618, "learning_rate": 9.18e-05, "loss": 0.1149, "step": 612 }, { "epoch": 0.28847058823529415, "grad_norm": 0.7976418861819703, "learning_rate": 9.194999999999999e-05, "loss": 0.1198, "step": 613 }, { "epoch": 0.28894117647058826, "grad_norm": 0.4062352509379705, "learning_rate": 9.209999999999999e-05, "loss": 0.0496, "step": 614 }, { "epoch": 0.28941176470588237, "grad_norm": 0.8375219375503137, "learning_rate": 9.224999999999999e-05, "loss": 0.1965, "step": 615 }, { "epoch": 0.2898823529411765, "grad_norm": 1.4703974335280094, "learning_rate": 9.24e-05, "loss": 0.236, "step": 616 }, { "epoch": 0.2903529411764706, "grad_norm": 0.7235874551308148, "learning_rate": 9.254999999999999e-05, "loss": 0.0637, "step": 617 }, { "epoch": 0.2908235294117647, "grad_norm": 0.8976771435056353, "learning_rate": 9.269999999999999e-05, "loss": 0.1333, "step": 618 }, { "epoch": 0.2912941176470588, "grad_norm": 1.0338410382600292, "learning_rate": 9.285e-05, "loss": 0.206, "step": 619 }, { "epoch": 0.2917647058823529, "grad_norm": 1.2674215853172988, "learning_rate": 9.3e-05, "loss": 0.243, "step": 620 }, { "epoch": 0.29223529411764704, "grad_norm": 0.728067018393702, "learning_rate": 9.314999999999999e-05, "loss": 0.1213, "step": 621 }, { "epoch": 0.29270588235294115, "grad_norm": 1.0241755899757201, "learning_rate": 9.329999999999999e-05, "loss": 0.281, "step": 622 }, { "epoch": 0.2931764705882353, "grad_norm": 0.6749983621182956, "learning_rate": 9.345e-05, "loss": 0.0809, "step": 623 }, { "epoch": 0.2936470588235294, "grad_norm": 2.1757844121447696, "learning_rate": 9.36e-05, "loss": 0.4807, "step": 624 }, { "epoch": 0.29411764705882354, "grad_norm": 0.9248507857806791, "learning_rate": 9.374999999999999e-05, "loss": 0.1373, "step": 625 }, { "epoch": 0.29458823529411765, "grad_norm": 1.0244674091816106, "learning_rate": 9.389999999999999e-05, "loss": 0.2432, "step": 626 }, { "epoch": 0.29505882352941176, "grad_norm": 0.9185152384093118, "learning_rate": 9.405e-05, "loss": 0.091, "step": 627 }, { "epoch": 0.2955294117647059, "grad_norm": 1.7059612922522378, "learning_rate": 9.419999999999999e-05, "loss": 0.2373, "step": 628 }, { "epoch": 0.296, "grad_norm": 1.0883885105505384, "learning_rate": 9.434999999999999e-05, "loss": 0.2071, "step": 629 }, { "epoch": 0.2964705882352941, "grad_norm": 0.9127497196760247, "learning_rate": 9.449999999999999e-05, "loss": 0.1976, "step": 630 }, { "epoch": 0.2969411764705882, "grad_norm": 1.4627521531781305, "learning_rate": 9.465e-05, "loss": 0.268, "step": 631 }, { "epoch": 0.2974117647058824, "grad_norm": 2.079191840207052, "learning_rate": 9.479999999999999e-05, "loss": 0.2014, "step": 632 }, { "epoch": 0.2978823529411765, "grad_norm": 0.604971197894348, "learning_rate": 9.494999999999999e-05, "loss": 0.0335, "step": 633 }, { "epoch": 0.2983529411764706, "grad_norm": 0.9391549981246418, "learning_rate": 9.51e-05, "loss": 0.1096, "step": 634 }, { "epoch": 0.2988235294117647, "grad_norm": 0.9476094441816123, "learning_rate": 9.525e-05, "loss": 0.1288, "step": 635 }, { "epoch": 0.2992941176470588, "grad_norm": 1.0802031815374034, "learning_rate": 9.539999999999999e-05, "loss": 0.1871, "step": 636 }, { "epoch": 0.29976470588235293, "grad_norm": 2.397250998797779, "learning_rate": 9.554999999999999e-05, "loss": 0.5991, "step": 637 }, { "epoch": 0.30023529411764704, "grad_norm": 0.7158191961071144, "learning_rate": 9.57e-05, "loss": 0.1502, "step": 638 }, { "epoch": 0.30070588235294116, "grad_norm": 1.499746225820254, "learning_rate": 9.585e-05, "loss": 0.449, "step": 639 }, { "epoch": 0.30117647058823527, "grad_norm": 0.8475583414289695, "learning_rate": 9.599999999999999e-05, "loss": 0.1337, "step": 640 }, { "epoch": 0.30164705882352943, "grad_norm": 1.9139437737101583, "learning_rate": 9.614999999999999e-05, "loss": 0.3807, "step": 641 }, { "epoch": 0.30211764705882355, "grad_norm": 1.0193811859916733, "learning_rate": 9.63e-05, "loss": 0.2976, "step": 642 }, { "epoch": 0.30258823529411766, "grad_norm": 0.7291027817886574, "learning_rate": 9.645e-05, "loss": 0.1344, "step": 643 }, { "epoch": 0.30305882352941177, "grad_norm": 0.6880889724256226, "learning_rate": 9.659999999999999e-05, "loss": 0.142, "step": 644 }, { "epoch": 0.3035294117647059, "grad_norm": 2.0978530987206825, "learning_rate": 9.675e-05, "loss": 0.5671, "step": 645 }, { "epoch": 0.304, "grad_norm": 0.4610284768799396, "learning_rate": 9.69e-05, "loss": 0.1003, "step": 646 }, { "epoch": 0.3044705882352941, "grad_norm": 0.5745339216405309, "learning_rate": 9.705e-05, "loss": 0.0945, "step": 647 }, { "epoch": 0.3049411764705882, "grad_norm": 2.2866735896618873, "learning_rate": 9.719999999999999e-05, "loss": 0.4608, "step": 648 }, { "epoch": 0.3054117647058823, "grad_norm": 0.9707208225631089, "learning_rate": 9.735e-05, "loss": 0.2377, "step": 649 }, { "epoch": 0.3058823529411765, "grad_norm": 0.9828276762178595, "learning_rate": 9.75e-05, "loss": 0.2257, "step": 650 }, { "epoch": 0.3063529411764706, "grad_norm": 0.6686133525617737, "learning_rate": 9.764999999999999e-05, "loss": 0.1217, "step": 651 }, { "epoch": 0.3068235294117647, "grad_norm": 1.3349126569296503, "learning_rate": 9.779999999999999e-05, "loss": 0.1795, "step": 652 }, { "epoch": 0.30729411764705883, "grad_norm": 0.6421293230138665, "learning_rate": 9.795e-05, "loss": 0.1026, "step": 653 }, { "epoch": 0.30776470588235294, "grad_norm": 0.7061203809933085, "learning_rate": 9.81e-05, "loss": 0.1142, "step": 654 }, { "epoch": 0.30823529411764705, "grad_norm": 1.6279835461113803, "learning_rate": 9.824999999999999e-05, "loss": 0.3431, "step": 655 }, { "epoch": 0.30870588235294116, "grad_norm": 0.7538296119753832, "learning_rate": 9.839999999999999e-05, "loss": 0.12, "step": 656 }, { "epoch": 0.3091764705882353, "grad_norm": 1.1377345807191959, "learning_rate": 9.855e-05, "loss": 0.2801, "step": 657 }, { "epoch": 0.3096470588235294, "grad_norm": 0.9813583752298995, "learning_rate": 9.87e-05, "loss": 0.1488, "step": 658 }, { "epoch": 0.31011764705882355, "grad_norm": 0.8546929685877113, "learning_rate": 9.884999999999999e-05, "loss": 0.1371, "step": 659 }, { "epoch": 0.31058823529411766, "grad_norm": 0.4279864865322814, "learning_rate": 9.9e-05, "loss": 0.0274, "step": 660 }, { "epoch": 0.3110588235294118, "grad_norm": 1.8134186944994983, "learning_rate": 9.915e-05, "loss": 0.3982, "step": 661 }, { "epoch": 0.3115294117647059, "grad_norm": 0.6766125401170027, "learning_rate": 9.93e-05, "loss": 0.1145, "step": 662 }, { "epoch": 0.312, "grad_norm": 0.38364401731770803, "learning_rate": 9.944999999999999e-05, "loss": 0.0475, "step": 663 }, { "epoch": 0.3124705882352941, "grad_norm": 0.8186446126386959, "learning_rate": 9.96e-05, "loss": 0.1793, "step": 664 }, { "epoch": 0.3129411764705882, "grad_norm": 0.5871403418740873, "learning_rate": 9.975e-05, "loss": 0.0631, "step": 665 }, { "epoch": 0.31341176470588233, "grad_norm": 1.084890338226896, "learning_rate": 9.99e-05, "loss": 0.3021, "step": 666 }, { "epoch": 0.31388235294117645, "grad_norm": 1.0599681084595416, "learning_rate": 0.00010004999999999999, "loss": 0.1762, "step": 667 }, { "epoch": 0.3143529411764706, "grad_norm": 3.423059094889314, "learning_rate": 0.0001002, "loss": 0.3573, "step": 668 }, { "epoch": 0.3148235294117647, "grad_norm": 0.7353172069269749, "learning_rate": 0.00010035, "loss": 0.1237, "step": 669 }, { "epoch": 0.31529411764705884, "grad_norm": 1.1346315774818263, "learning_rate": 0.0001005, "loss": 0.1046, "step": 670 }, { "epoch": 0.31576470588235295, "grad_norm": 1.300915143730808, "learning_rate": 0.00010065, "loss": 0.2386, "step": 671 }, { "epoch": 0.31623529411764706, "grad_norm": 2.1191956922777817, "learning_rate": 0.0001008, "loss": 0.3598, "step": 672 }, { "epoch": 0.31670588235294117, "grad_norm": 1.2198255355393364, "learning_rate": 0.00010095, "loss": 0.2888, "step": 673 }, { "epoch": 0.3171764705882353, "grad_norm": 3.526846554195552, "learning_rate": 0.0001011, "loss": 0.4595, "step": 674 }, { "epoch": 0.3176470588235294, "grad_norm": 0.903144746760479, "learning_rate": 0.00010125, "loss": 0.25, "step": 675 }, { "epoch": 0.3181176470588235, "grad_norm": 0.4864767843853096, "learning_rate": 0.0001014, "loss": 0.0485, "step": 676 }, { "epoch": 0.31858823529411767, "grad_norm": 2.401137739971787, "learning_rate": 0.00010155, "loss": 0.2996, "step": 677 }, { "epoch": 0.3190588235294118, "grad_norm": 3.252747114690808, "learning_rate": 0.00010169999999999999, "loss": 0.5597, "step": 678 }, { "epoch": 0.3195294117647059, "grad_norm": 1.193519128340458, "learning_rate": 0.00010185, "loss": 0.2405, "step": 679 }, { "epoch": 0.32, "grad_norm": 2.4753242330115834, "learning_rate": 0.000102, "loss": 0.3496, "step": 680 }, { "epoch": 0.3204705882352941, "grad_norm": 1.624630248335363, "learning_rate": 0.00010215, "loss": 0.2528, "step": 681 }, { "epoch": 0.32094117647058823, "grad_norm": 1.039257649417399, "learning_rate": 0.00010229999999999999, "loss": 0.2699, "step": 682 }, { "epoch": 0.32141176470588234, "grad_norm": 2.5145443574946684, "learning_rate": 0.00010245, "loss": 0.2146, "step": 683 }, { "epoch": 0.32188235294117645, "grad_norm": 2.030345694981069, "learning_rate": 0.0001026, "loss": 0.2409, "step": 684 }, { "epoch": 0.32235294117647056, "grad_norm": 0.8207931188406135, "learning_rate": 0.00010275, "loss": 0.0519, "step": 685 }, { "epoch": 0.32282352941176473, "grad_norm": 0.39974624944758835, "learning_rate": 0.0001029, "loss": 0.0525, "step": 686 }, { "epoch": 0.32329411764705884, "grad_norm": 0.8942591146412717, "learning_rate": 0.00010305, "loss": 0.2187, "step": 687 }, { "epoch": 0.32376470588235295, "grad_norm": 1.2050169539255242, "learning_rate": 0.00010319999999999999, "loss": 0.1641, "step": 688 }, { "epoch": 0.32423529411764707, "grad_norm": 1.6118951745629386, "learning_rate": 0.00010334999999999998, "loss": 0.2959, "step": 689 }, { "epoch": 0.3247058823529412, "grad_norm": 1.377532737656485, "learning_rate": 0.00010349999999999998, "loss": 0.3591, "step": 690 }, { "epoch": 0.3251764705882353, "grad_norm": 2.3469803151829756, "learning_rate": 0.00010364999999999999, "loss": 0.2254, "step": 691 }, { "epoch": 0.3256470588235294, "grad_norm": 1.9798325456508932, "learning_rate": 0.00010379999999999999, "loss": 0.3015, "step": 692 }, { "epoch": 0.3261176470588235, "grad_norm": 2.278171164464319, "learning_rate": 0.00010394999999999998, "loss": 0.1497, "step": 693 }, { "epoch": 0.3265882352941176, "grad_norm": 1.1134540574898346, "learning_rate": 0.00010409999999999998, "loss": 0.2127, "step": 694 }, { "epoch": 0.3270588235294118, "grad_norm": 0.7302157998153611, "learning_rate": 0.00010424999999999999, "loss": 0.1477, "step": 695 }, { "epoch": 0.3275294117647059, "grad_norm": 0.46523687157271626, "learning_rate": 0.00010439999999999999, "loss": 0.0779, "step": 696 }, { "epoch": 0.328, "grad_norm": 2.2750637041683532, "learning_rate": 0.00010454999999999998, "loss": 0.4826, "step": 697 }, { "epoch": 0.3284705882352941, "grad_norm": 0.6875695583962245, "learning_rate": 0.00010469999999999998, "loss": 0.1442, "step": 698 }, { "epoch": 0.32894117647058824, "grad_norm": 1615.7080251507737, "learning_rate": 0.00010484999999999999, "loss": 0.6773, "step": 699 }, { "epoch": 0.32941176470588235, "grad_norm": 106.1593107016583, "learning_rate": 0.00010499999999999999, "loss": 6.7413, "step": 700 }, { "epoch": 0.32988235294117646, "grad_norm": 2.3723308804978935, "learning_rate": 0.00010514999999999998, "loss": 0.3858, "step": 701 }, { "epoch": 0.33035294117647057, "grad_norm": 24.976637038946883, "learning_rate": 0.00010529999999999998, "loss": 0.9641, "step": 702 }, { "epoch": 0.3308235294117647, "grad_norm": 16.722039863818996, "learning_rate": 0.00010544999999999999, "loss": 0.6964, "step": 703 }, { "epoch": 0.33129411764705885, "grad_norm": 1.9109688834764114, "learning_rate": 0.00010559999999999998, "loss": 0.1178, "step": 704 }, { "epoch": 0.33176470588235296, "grad_norm": 6.710857112879257, "learning_rate": 0.00010574999999999998, "loss": 0.6812, "step": 705 }, { "epoch": 0.3322352941176471, "grad_norm": 5.7489194810228055, "learning_rate": 0.00010589999999999999, "loss": 0.6153, "step": 706 }, { "epoch": 0.3327058823529412, "grad_norm": 4.075876460555557, "learning_rate": 0.00010604999999999999, "loss": 0.4134, "step": 707 }, { "epoch": 0.3331764705882353, "grad_norm": 8.901751247060888, "learning_rate": 0.00010619999999999998, "loss": 1.4468, "step": 708 }, { "epoch": 0.3336470588235294, "grad_norm": 14.988479706584375, "learning_rate": 0.00010634999999999998, "loss": 1.441, "step": 709 }, { "epoch": 0.3341176470588235, "grad_norm": 10.300217138364607, "learning_rate": 0.00010649999999999999, "loss": 1.2334, "step": 710 }, { "epoch": 0.33458823529411763, "grad_norm": 17.61878024212167, "learning_rate": 0.00010664999999999999, "loss": 0.633, "step": 711 }, { "epoch": 0.33505882352941174, "grad_norm": 0.4773587407362364, "learning_rate": 0.00010679999999999998, "loss": 0.1987, "step": 712 }, { "epoch": 0.3355294117647059, "grad_norm": 11.380889457886854, "learning_rate": 0.00010694999999999998, "loss": 1.0334, "step": 713 }, { "epoch": 0.336, "grad_norm": 1.0914597746800492, "learning_rate": 0.00010709999999999999, "loss": 0.1356, "step": 714 }, { "epoch": 0.33647058823529413, "grad_norm": 21.82768322524575, "learning_rate": 0.00010724999999999999, "loss": 1.5513, "step": 715 }, { "epoch": 0.33694117647058824, "grad_norm": 8.755436422212329, "learning_rate": 0.00010739999999999998, "loss": 2.589, "step": 716 }, { "epoch": 0.33741176470588236, "grad_norm": 16.244103706467556, "learning_rate": 0.00010754999999999999, "loss": 1.078, "step": 717 }, { "epoch": 0.33788235294117647, "grad_norm": 10.006278137436698, "learning_rate": 0.00010769999999999999, "loss": 1.4069, "step": 718 }, { "epoch": 0.3383529411764706, "grad_norm": 10.06080454636576, "learning_rate": 0.00010784999999999999, "loss": 0.7692, "step": 719 }, { "epoch": 0.3388235294117647, "grad_norm": 4.464449625641073, "learning_rate": 0.00010799999999999998, "loss": 0.4619, "step": 720 }, { "epoch": 0.3392941176470588, "grad_norm": 1.424205762444851, "learning_rate": 0.00010814999999999999, "loss": 0.2028, "step": 721 }, { "epoch": 0.33976470588235297, "grad_norm": 10.477668481610129, "learning_rate": 0.00010829999999999999, "loss": 0.8966, "step": 722 }, { "epoch": 0.3402352941176471, "grad_norm": 6.536665760053284, "learning_rate": 0.00010845, "loss": 0.5687, "step": 723 }, { "epoch": 0.3407058823529412, "grad_norm": 8.912886058786617, "learning_rate": 0.00010859999999999998, "loss": 0.9229, "step": 724 }, { "epoch": 0.3411764705882353, "grad_norm": 10.965258034170539, "learning_rate": 0.00010874999999999999, "loss": 0.5887, "step": 725 }, { "epoch": 0.3416470588235294, "grad_norm": 10.409606486258852, "learning_rate": 0.00010889999999999999, "loss": 1.1054, "step": 726 }, { "epoch": 0.3421176470588235, "grad_norm": 10.869592346454663, "learning_rate": 0.00010904999999999998, "loss": 0.694, "step": 727 }, { "epoch": 0.34258823529411764, "grad_norm": 18.824843077089174, "learning_rate": 0.00010919999999999998, "loss": 1.5357, "step": 728 }, { "epoch": 0.34305882352941175, "grad_norm": 3.975208137499112, "learning_rate": 0.00010934999999999999, "loss": 0.9231, "step": 729 }, { "epoch": 0.34352941176470586, "grad_norm": 4.958589475979727, "learning_rate": 0.00010949999999999999, "loss": 0.7115, "step": 730 }, { "epoch": 0.344, "grad_norm": 7.869185895652223, "learning_rate": 0.00010964999999999998, "loss": 1.1244, "step": 731 }, { "epoch": 0.34447058823529414, "grad_norm": 2.7185581223769018, "learning_rate": 0.00010979999999999999, "loss": 0.9986, "step": 732 }, { "epoch": 0.34494117647058825, "grad_norm": 4.787579551567938, "learning_rate": 0.00010994999999999999, "loss": 0.4706, "step": 733 }, { "epoch": 0.34541176470588236, "grad_norm": 15.688510932740714, "learning_rate": 0.00011009999999999999, "loss": 0.8121, "step": 734 }, { "epoch": 0.3458823529411765, "grad_norm": 35.418711840086644, "learning_rate": 0.00011024999999999998, "loss": 3.9232, "step": 735 }, { "epoch": 0.3463529411764706, "grad_norm": 57.85460771633729, "learning_rate": 0.00011039999999999999, "loss": 0.9, "step": 736 }, { "epoch": 0.3468235294117647, "grad_norm": 4.447412152585237, "learning_rate": 0.00011054999999999999, "loss": 0.3192, "step": 737 }, { "epoch": 0.3472941176470588, "grad_norm": 3.943930149749368, "learning_rate": 0.0001107, "loss": 0.6563, "step": 738 }, { "epoch": 0.3477647058823529, "grad_norm": 7.859993631779855, "learning_rate": 0.00011084999999999998, "loss": 1.1008, "step": 739 }, { "epoch": 0.34823529411764703, "grad_norm": 3.2085411503509977, "learning_rate": 0.00011099999999999999, "loss": 0.6125, "step": 740 }, { "epoch": 0.3487058823529412, "grad_norm": 25.27930121818923, "learning_rate": 0.00011114999999999999, "loss": 1.623, "step": 741 }, { "epoch": 0.3491764705882353, "grad_norm": 9.045202294521298, "learning_rate": 0.0001113, "loss": 0.5854, "step": 742 }, { "epoch": 0.3496470588235294, "grad_norm": 3.276081618248379, "learning_rate": 0.00011144999999999998, "loss": 0.5288, "step": 743 }, { "epoch": 0.35011764705882353, "grad_norm": 14.024913687866201, "learning_rate": 0.00011159999999999999, "loss": 1.0799, "step": 744 }, { "epoch": 0.35058823529411764, "grad_norm": 8.423106090387785, "learning_rate": 0.00011174999999999999, "loss": 2.3701, "step": 745 }, { "epoch": 0.35105882352941176, "grad_norm": 1.7941940408881256, "learning_rate": 0.0001119, "loss": 0.5984, "step": 746 }, { "epoch": 0.35152941176470587, "grad_norm": 14.229377758300522, "learning_rate": 0.00011204999999999999, "loss": 0.8211, "step": 747 }, { "epoch": 0.352, "grad_norm": 13.85986892952042, "learning_rate": 0.00011219999999999999, "loss": 0.4568, "step": 748 }, { "epoch": 0.3524705882352941, "grad_norm": 15.898318802471199, "learning_rate": 0.00011235, "loss": 1.5615, "step": 749 }, { "epoch": 0.35294117647058826, "grad_norm": 4.4969932031146325, "learning_rate": 0.0001125, "loss": 0.672, "step": 750 }, { "epoch": 0.35341176470588237, "grad_norm": 4.5691754337948165, "learning_rate": 0.00011264999999999999, "loss": 0.8199, "step": 751 }, { "epoch": 0.3538823529411765, "grad_norm": 2.6621225675800875, "learning_rate": 0.00011279999999999999, "loss": 0.2983, "step": 752 }, { "epoch": 0.3543529411764706, "grad_norm": 11.469584134811427, "learning_rate": 0.00011295, "loss": 1.2005, "step": 753 }, { "epoch": 0.3548235294117647, "grad_norm": 1.2633031966985893, "learning_rate": 0.00011309999999999998, "loss": 0.2637, "step": 754 }, { "epoch": 0.3552941176470588, "grad_norm": 2.7349002409608683, "learning_rate": 0.00011324999999999999, "loss": 0.5775, "step": 755 }, { "epoch": 0.3557647058823529, "grad_norm": 1.417155455511152, "learning_rate": 0.00011339999999999999, "loss": 0.5377, "step": 756 }, { "epoch": 0.35623529411764704, "grad_norm": 8.688643320392174, "learning_rate": 0.00011355, "loss": 0.7102, "step": 757 }, { "epoch": 0.35670588235294115, "grad_norm": 2.8010813894738353, "learning_rate": 0.00011369999999999999, "loss": 0.4839, "step": 758 }, { "epoch": 0.3571764705882353, "grad_norm": 10.398703866219766, "learning_rate": 0.00011384999999999999, "loss": 0.4122, "step": 759 }, { "epoch": 0.35764705882352943, "grad_norm": 6.192681135282548, "learning_rate": 0.00011399999999999999, "loss": 0.3217, "step": 760 }, { "epoch": 0.35811764705882354, "grad_norm": 3.0465710948037033, "learning_rate": 0.00011415, "loss": 0.5822, "step": 761 }, { "epoch": 0.35858823529411765, "grad_norm": 2.0511547695566352, "learning_rate": 0.00011429999999999999, "loss": 0.448, "step": 762 }, { "epoch": 0.35905882352941176, "grad_norm": 6.610663461203885, "learning_rate": 0.00011444999999999999, "loss": 0.5494, "step": 763 }, { "epoch": 0.3595294117647059, "grad_norm": 1.4321768426555999, "learning_rate": 0.0001146, "loss": 0.1387, "step": 764 }, { "epoch": 0.36, "grad_norm": 1.7513298432855209, "learning_rate": 0.00011475, "loss": 0.3024, "step": 765 }, { "epoch": 0.3604705882352941, "grad_norm": 1.6411804298044108, "learning_rate": 0.00011489999999999999, "loss": 0.4612, "step": 766 }, { "epoch": 0.3609411764705882, "grad_norm": 2.537949873089706, "learning_rate": 0.00011504999999999999, "loss": 0.1794, "step": 767 }, { "epoch": 0.3614117647058824, "grad_norm": 3.219661297232049, "learning_rate": 0.0001152, "loss": 0.2069, "step": 768 }, { "epoch": 0.3618823529411765, "grad_norm": 2.912349641540771, "learning_rate": 0.00011535, "loss": 0.6793, "step": 769 }, { "epoch": 0.3623529411764706, "grad_norm": 1.6916449304350538, "learning_rate": 0.00011549999999999999, "loss": 0.5672, "step": 770 }, { "epoch": 0.3628235294117647, "grad_norm": 1.8543592611496045, "learning_rate": 0.00011564999999999999, "loss": 0.2702, "step": 771 }, { "epoch": 0.3632941176470588, "grad_norm": 1.3454089915907073, "learning_rate": 0.0001158, "loss": 0.34, "step": 772 }, { "epoch": 0.36376470588235293, "grad_norm": 2.0063383500547496, "learning_rate": 0.00011595, "loss": 0.4569, "step": 773 }, { "epoch": 0.36423529411764705, "grad_norm": 2.298152953760203, "learning_rate": 0.00011609999999999999, "loss": 0.7657, "step": 774 }, { "epoch": 0.36470588235294116, "grad_norm": 1.9893655181674923, "learning_rate": 0.00011624999999999999, "loss": 0.6346, "step": 775 }, { "epoch": 0.36517647058823527, "grad_norm": 0.9625999230995761, "learning_rate": 0.0001164, "loss": 0.1999, "step": 776 }, { "epoch": 0.36564705882352944, "grad_norm": 1.1573747366796563, "learning_rate": 0.00011654999999999999, "loss": 0.3538, "step": 777 }, { "epoch": 0.36611764705882355, "grad_norm": 0.8698362861066057, "learning_rate": 0.00011669999999999999, "loss": 0.3107, "step": 778 }, { "epoch": 0.36658823529411766, "grad_norm": 1.0940694914304117, "learning_rate": 0.00011685, "loss": 0.2158, "step": 779 }, { "epoch": 0.36705882352941177, "grad_norm": 1.8050300440912113, "learning_rate": 0.000117, "loss": 0.4609, "step": 780 }, { "epoch": 0.3675294117647059, "grad_norm": 0.815558976375099, "learning_rate": 0.00011714999999999999, "loss": 0.1933, "step": 781 }, { "epoch": 0.368, "grad_norm": 0.3403395604639789, "learning_rate": 0.00011729999999999999, "loss": 0.063, "step": 782 }, { "epoch": 0.3684705882352941, "grad_norm": 1.6336807819824122, "learning_rate": 0.00011745, "loss": 0.2299, "step": 783 }, { "epoch": 0.3689411764705882, "grad_norm": 0.709743018891166, "learning_rate": 0.0001176, "loss": 0.2128, "step": 784 }, { "epoch": 0.36941176470588233, "grad_norm": 0.39660977412177945, "learning_rate": 0.00011774999999999999, "loss": 0.0766, "step": 785 }, { "epoch": 0.3698823529411765, "grad_norm": 0.5176434031173397, "learning_rate": 0.00011789999999999999, "loss": 0.0448, "step": 786 }, { "epoch": 0.3703529411764706, "grad_norm": 0.9827041444930918, "learning_rate": 0.00011805, "loss": 0.2509, "step": 787 }, { "epoch": 0.3708235294117647, "grad_norm": 2.628296908062467, "learning_rate": 0.0001182, "loss": 0.7189, "step": 788 }, { "epoch": 0.37129411764705883, "grad_norm": 2.9045719796337366, "learning_rate": 0.00011834999999999999, "loss": 0.535, "step": 789 }, { "epoch": 0.37176470588235294, "grad_norm": 0.5652037649297759, "learning_rate": 0.0001185, "loss": 0.0606, "step": 790 }, { "epoch": 0.37223529411764705, "grad_norm": 3.017846954222186, "learning_rate": 0.00011865, "loss": 0.4139, "step": 791 }, { "epoch": 0.37270588235294116, "grad_norm": 1.3686444900245893, "learning_rate": 0.0001188, "loss": 0.3024, "step": 792 }, { "epoch": 0.3731764705882353, "grad_norm": 1.9015821661401608, "learning_rate": 0.00011894999999999999, "loss": 0.3455, "step": 793 }, { "epoch": 0.3736470588235294, "grad_norm": 1.2070724400391433, "learning_rate": 0.0001191, "loss": 0.2689, "step": 794 }, { "epoch": 0.37411764705882355, "grad_norm": 1.1216432158613496, "learning_rate": 0.00011925, "loss": 0.3287, "step": 795 }, { "epoch": 0.37458823529411767, "grad_norm": 1.2877153627502784, "learning_rate": 0.0001194, "loss": 0.3755, "step": 796 }, { "epoch": 0.3750588235294118, "grad_norm": 0.6409413548883194, "learning_rate": 0.00011954999999999999, "loss": 0.1115, "step": 797 }, { "epoch": 0.3755294117647059, "grad_norm": 3.721848638495115, "learning_rate": 0.0001197, "loss": 0.038, "step": 798 }, { "epoch": 0.376, "grad_norm": 1.2518439463631463, "learning_rate": 0.00011985, "loss": 0.3616, "step": 799 }, { "epoch": 0.3764705882352941, "grad_norm": 1.260444500186898, "learning_rate": 0.00011999999999999999, "loss": 0.5555, "step": 800 }, { "epoch": 0.3769411764705882, "grad_norm": 4.537247064452471, "learning_rate": 0.00012014999999999999, "loss": 0.0592, "step": 801 }, { "epoch": 0.37741176470588234, "grad_norm": 0.5809686556894118, "learning_rate": 0.0001203, "loss": 0.0837, "step": 802 }, { "epoch": 0.37788235294117645, "grad_norm": 0.5905695320570278, "learning_rate": 0.00012045, "loss": 0.08, "step": 803 }, { "epoch": 0.3783529411764706, "grad_norm": 0.8008838535147307, "learning_rate": 0.00012059999999999999, "loss": 0.1555, "step": 804 }, { "epoch": 0.3788235294117647, "grad_norm": 0.7279382740306704, "learning_rate": 0.00012075, "loss": 0.1711, "step": 805 }, { "epoch": 0.37929411764705884, "grad_norm": 1.6608167668008755, "learning_rate": 0.0001209, "loss": 0.4551, "step": 806 }, { "epoch": 0.37976470588235295, "grad_norm": 0.4589289305760174, "learning_rate": 0.00012105, "loss": 0.0755, "step": 807 }, { "epoch": 0.38023529411764706, "grad_norm": 2.0295076588176952, "learning_rate": 0.00012119999999999999, "loss": 0.4187, "step": 808 }, { "epoch": 0.38070588235294117, "grad_norm": 1.8641851512497667, "learning_rate": 0.00012135, "loss": 0.2232, "step": 809 }, { "epoch": 0.3811764705882353, "grad_norm": 0.7574013759104068, "learning_rate": 0.0001215, "loss": 0.0763, "step": 810 }, { "epoch": 0.3816470588235294, "grad_norm": 0.9775093955492022, "learning_rate": 0.00012165, "loss": 0.2458, "step": 811 }, { "epoch": 0.3821176470588235, "grad_norm": 0.590425000816024, "learning_rate": 0.00012179999999999999, "loss": 0.0848, "step": 812 }, { "epoch": 0.3825882352941177, "grad_norm": 0.5110576278936305, "learning_rate": 0.00012194999999999998, "loss": 0.0612, "step": 813 }, { "epoch": 0.3830588235294118, "grad_norm": 1.1800104580779, "learning_rate": 0.00012209999999999999, "loss": 0.1334, "step": 814 }, { "epoch": 0.3835294117647059, "grad_norm": 1.57967709461727, "learning_rate": 0.00012225, "loss": 0.1474, "step": 815 }, { "epoch": 0.384, "grad_norm": 0.3986699381528179, "learning_rate": 0.0001224, "loss": 0.0321, "step": 816 }, { "epoch": 0.3844705882352941, "grad_norm": 0.8306680794118789, "learning_rate": 0.00012254999999999997, "loss": 0.2148, "step": 817 }, { "epoch": 0.38494117647058823, "grad_norm": 0.6577573533879014, "learning_rate": 0.00012269999999999997, "loss": 0.1341, "step": 818 }, { "epoch": 0.38541176470588234, "grad_norm": 0.9124290415587041, "learning_rate": 0.00012284999999999998, "loss": 0.1171, "step": 819 }, { "epoch": 0.38588235294117645, "grad_norm": 2.979795764410915, "learning_rate": 0.00012299999999999998, "loss": 0.2588, "step": 820 }, { "epoch": 0.38635294117647057, "grad_norm": 1.4616501556596337, "learning_rate": 0.00012314999999999998, "loss": 0.379, "step": 821 }, { "epoch": 0.38682352941176473, "grad_norm": 0.6371137972687665, "learning_rate": 0.0001233, "loss": 0.2723, "step": 822 }, { "epoch": 0.38729411764705884, "grad_norm": 0.11764067852516616, "learning_rate": 0.00012345, "loss": 0.0109, "step": 823 }, { "epoch": 0.38776470588235296, "grad_norm": 1.969793032905129, "learning_rate": 0.0001236, "loss": 0.3043, "step": 824 }, { "epoch": 0.38823529411764707, "grad_norm": 1.7977067910206388, "learning_rate": 0.00012374999999999997, "loss": 0.2584, "step": 825 }, { "epoch": 0.3887058823529412, "grad_norm": 0.541751099650033, "learning_rate": 0.00012389999999999998, "loss": 0.128, "step": 826 }, { "epoch": 0.3891764705882353, "grad_norm": 7.163342906297623, "learning_rate": 0.00012404999999999998, "loss": 0.2554, "step": 827 }, { "epoch": 0.3896470588235294, "grad_norm": 2.1192674104665583, "learning_rate": 0.00012419999999999998, "loss": 0.3358, "step": 828 }, { "epoch": 0.3901176470588235, "grad_norm": 1.655245229655318, "learning_rate": 0.00012435, "loss": 0.2609, "step": 829 }, { "epoch": 0.3905882352941176, "grad_norm": 1.9175572909565266, "learning_rate": 0.0001245, "loss": 0.378, "step": 830 }, { "epoch": 0.3910588235294118, "grad_norm": 1.3658700480966817, "learning_rate": 0.00012465, "loss": 0.1468, "step": 831 }, { "epoch": 0.3915294117647059, "grad_norm": 1.1781184339568171, "learning_rate": 0.00012479999999999997, "loss": 0.2822, "step": 832 }, { "epoch": 0.392, "grad_norm": 1.6502298973979028, "learning_rate": 0.00012494999999999997, "loss": 0.2416, "step": 833 }, { "epoch": 0.3924705882352941, "grad_norm": 0.8946253585349778, "learning_rate": 0.00012509999999999998, "loss": 0.3158, "step": 834 }, { "epoch": 0.39294117647058824, "grad_norm": 2.6050150517097066, "learning_rate": 0.00012524999999999998, "loss": 0.1711, "step": 835 }, { "epoch": 0.39341176470588235, "grad_norm": 0.7566859746142262, "learning_rate": 0.00012539999999999999, "loss": 0.0841, "step": 836 }, { "epoch": 0.39388235294117646, "grad_norm": 0.40174185692324876, "learning_rate": 0.00012555, "loss": 0.0648, "step": 837 }, { "epoch": 0.3943529411764706, "grad_norm": 1.2613459473793542, "learning_rate": 0.0001257, "loss": 0.3633, "step": 838 }, { "epoch": 0.3948235294117647, "grad_norm": 2.0426161233289255, "learning_rate": 0.00012585, "loss": 0.5909, "step": 839 }, { "epoch": 0.3952941176470588, "grad_norm": 3.4110806893343093, "learning_rate": 0.00012599999999999997, "loss": 0.5514, "step": 840 }, { "epoch": 0.39576470588235296, "grad_norm": 0.7478847325997703, "learning_rate": 0.00012614999999999998, "loss": 0.0912, "step": 841 }, { "epoch": 0.3962352941176471, "grad_norm": 3.3708323380986553, "learning_rate": 0.00012629999999999998, "loss": 0.4551, "step": 842 }, { "epoch": 0.3967058823529412, "grad_norm": 0.9173760951805641, "learning_rate": 0.00012644999999999998, "loss": 0.1863, "step": 843 }, { "epoch": 0.3971764705882353, "grad_norm": 0.6370488590821378, "learning_rate": 0.0001266, "loss": 0.136, "step": 844 }, { "epoch": 0.3976470588235294, "grad_norm": 2.108543100840635, "learning_rate": 0.00012675, "loss": 0.471, "step": 845 }, { "epoch": 0.3981176470588235, "grad_norm": 0.4333004292010908, "learning_rate": 0.0001269, "loss": 0.0422, "step": 846 }, { "epoch": 0.39858823529411763, "grad_norm": 1.20811601711382, "learning_rate": 0.00012705, "loss": 0.14, "step": 847 }, { "epoch": 0.39905882352941174, "grad_norm": 0.5466814261883969, "learning_rate": 0.00012719999999999997, "loss": 0.0241, "step": 848 }, { "epoch": 0.39952941176470586, "grad_norm": 0.61801763883678, "learning_rate": 0.00012734999999999998, "loss": 0.0596, "step": 849 }, { "epoch": 0.4, "grad_norm": 1.5855153626100729, "learning_rate": 0.00012749999999999998, "loss": 0.4123, "step": 850 }, { "epoch": 0.40047058823529413, "grad_norm": 0.6518061572772145, "learning_rate": 0.00012764999999999999, "loss": 0.134, "step": 851 }, { "epoch": 0.40094117647058825, "grad_norm": 1.0297705593905875, "learning_rate": 0.0001278, "loss": 0.1924, "step": 852 }, { "epoch": 0.40141176470588236, "grad_norm": 0.12580757060778044, "learning_rate": 0.00012795, "loss": 0.0099, "step": 853 }, { "epoch": 0.40188235294117647, "grad_norm": 1.6189005634180946, "learning_rate": 0.0001281, "loss": 0.2573, "step": 854 }, { "epoch": 0.4023529411764706, "grad_norm": 0.8490767545879948, "learning_rate": 0.00012824999999999997, "loss": 0.1021, "step": 855 }, { "epoch": 0.4028235294117647, "grad_norm": 2.0305771954683696, "learning_rate": 0.00012839999999999998, "loss": 0.2263, "step": 856 }, { "epoch": 0.4032941176470588, "grad_norm": 1.6530117600262715, "learning_rate": 0.00012854999999999998, "loss": 0.4044, "step": 857 }, { "epoch": 0.4037647058823529, "grad_norm": 8.24475094551377, "learning_rate": 0.00012869999999999998, "loss": 0.4409, "step": 858 }, { "epoch": 0.4042352941176471, "grad_norm": 0.8517953889556293, "learning_rate": 0.00012885, "loss": 0.1017, "step": 859 }, { "epoch": 0.4047058823529412, "grad_norm": 2.3691251928945136, "learning_rate": 0.000129, "loss": 0.4206, "step": 860 }, { "epoch": 0.4051764705882353, "grad_norm": 1.484639842788778, "learning_rate": 0.00012915, "loss": 0.3324, "step": 861 }, { "epoch": 0.4056470588235294, "grad_norm": 1.3909181434516542, "learning_rate": 0.0001293, "loss": 0.3361, "step": 862 }, { "epoch": 0.4061176470588235, "grad_norm": 1.7016424023593189, "learning_rate": 0.00012944999999999998, "loss": 0.2843, "step": 863 }, { "epoch": 0.40658823529411764, "grad_norm": 2.2398416037450644, "learning_rate": 0.00012959999999999998, "loss": 0.332, "step": 864 }, { "epoch": 0.40705882352941175, "grad_norm": 0.5195262118449631, "learning_rate": 0.00012974999999999998, "loss": 0.0838, "step": 865 }, { "epoch": 0.40752941176470586, "grad_norm": 40.32231784141888, "learning_rate": 0.00012989999999999999, "loss": 0.3964, "step": 866 }, { "epoch": 0.408, "grad_norm": 0.6908504047412448, "learning_rate": 0.00013005, "loss": 0.0937, "step": 867 }, { "epoch": 0.40847058823529414, "grad_norm": 31.371538951554058, "learning_rate": 0.0001302, "loss": 0.8947, "step": 868 }, { "epoch": 0.40894117647058825, "grad_norm": 27.900998946170635, "learning_rate": 0.00013035, "loss": 1.5241, "step": 869 }, { "epoch": 0.40941176470588236, "grad_norm": 10.12420755113767, "learning_rate": 0.0001305, "loss": 0.9614, "step": 870 }, { "epoch": 0.4098823529411765, "grad_norm": 7.248385357471507, "learning_rate": 0.00013064999999999998, "loss": 0.8431, "step": 871 }, { "epoch": 0.4103529411764706, "grad_norm": 15.373930484794814, "learning_rate": 0.00013079999999999998, "loss": 0.6715, "step": 872 }, { "epoch": 0.4108235294117647, "grad_norm": 4.591384138815599, "learning_rate": 0.00013094999999999998, "loss": 0.6276, "step": 873 }, { "epoch": 0.4112941176470588, "grad_norm": 17.24532944903512, "learning_rate": 0.0001311, "loss": 1.178, "step": 874 }, { "epoch": 0.4117647058823529, "grad_norm": 1.0212847444924584, "learning_rate": 0.00013125, "loss": 0.3316, "step": 875 }, { "epoch": 0.41223529411764703, "grad_norm": 1.8216254532713105, "learning_rate": 0.0001314, "loss": 0.1759, "step": 876 }, { "epoch": 0.4127058823529412, "grad_norm": 2.2317610483237393, "learning_rate": 0.00013155, "loss": 0.1634, "step": 877 }, { "epoch": 0.4131764705882353, "grad_norm": 3.3205964459522908, "learning_rate": 0.00013169999999999998, "loss": 0.3237, "step": 878 }, { "epoch": 0.4136470588235294, "grad_norm": 3.25670561003987, "learning_rate": 0.00013184999999999998, "loss": 0.3216, "step": 879 }, { "epoch": 0.41411764705882353, "grad_norm": 0.7142006247450977, "learning_rate": 0.00013199999999999998, "loss": 0.1525, "step": 880 }, { "epoch": 0.41458823529411765, "grad_norm": 5.879932988866183, "learning_rate": 0.00013215, "loss": 0.3347, "step": 881 }, { "epoch": 0.41505882352941176, "grad_norm": 10.805205987994984, "learning_rate": 0.0001323, "loss": 1.6675, "step": 882 }, { "epoch": 0.41552941176470587, "grad_norm": 19.190588398077942, "learning_rate": 0.00013245, "loss": 1.3968, "step": 883 }, { "epoch": 0.416, "grad_norm": 5.294961360395567, "learning_rate": 0.0001326, "loss": 0.2196, "step": 884 }, { "epoch": 0.4164705882352941, "grad_norm": 11.203466350779806, "learning_rate": 0.00013275, "loss": 0.9252, "step": 885 }, { "epoch": 0.41694117647058826, "grad_norm": 14.209005039108927, "learning_rate": 0.00013289999999999998, "loss": 1.2765, "step": 886 }, { "epoch": 0.41741176470588237, "grad_norm": 20.424121869413604, "learning_rate": 0.00013304999999999998, "loss": 2.1345, "step": 887 }, { "epoch": 0.4178823529411765, "grad_norm": 6.367803476623538, "learning_rate": 0.00013319999999999999, "loss": 0.8886, "step": 888 }, { "epoch": 0.4183529411764706, "grad_norm": 10.75926965560021, "learning_rate": 0.00013335, "loss": 0.8647, "step": 889 }, { "epoch": 0.4188235294117647, "grad_norm": 2.78024074578634, "learning_rate": 0.0001335, "loss": 0.2656, "step": 890 }, { "epoch": 0.4192941176470588, "grad_norm": 10.74149929960007, "learning_rate": 0.00013365, "loss": 0.8689, "step": 891 }, { "epoch": 0.41976470588235293, "grad_norm": 1.2860499493866777, "learning_rate": 0.0001338, "loss": 0.1989, "step": 892 }, { "epoch": 0.42023529411764704, "grad_norm": 4.702275249394482, "learning_rate": 0.00013395, "loss": 0.5017, "step": 893 }, { "epoch": 0.42070588235294115, "grad_norm": 2.3226371481763786, "learning_rate": 0.00013409999999999998, "loss": 0.2767, "step": 894 }, { "epoch": 0.4211764705882353, "grad_norm": 5.184380492329177, "learning_rate": 0.00013424999999999998, "loss": 0.5765, "step": 895 }, { "epoch": 0.42164705882352943, "grad_norm": 1.9424884072668702, "learning_rate": 0.0001344, "loss": 0.5745, "step": 896 }, { "epoch": 0.42211764705882354, "grad_norm": 1.5011794029021543, "learning_rate": 0.00013455, "loss": 0.2528, "step": 897 }, { "epoch": 0.42258823529411765, "grad_norm": 18.910321517461153, "learning_rate": 0.0001347, "loss": 1.7327, "step": 898 }, { "epoch": 0.42305882352941176, "grad_norm": 4.020680004815721, "learning_rate": 0.00013485, "loss": 0.7598, "step": 899 }, { "epoch": 0.4235294117647059, "grad_norm": 1.6050928679380392, "learning_rate": 0.000135, "loss": 0.2973, "step": 900 }, { "epoch": 0.424, "grad_norm": 1.2696742449792164, "learning_rate": 0.00013514999999999998, "loss": 0.2642, "step": 901 }, { "epoch": 0.4244705882352941, "grad_norm": 3.8222554949176817, "learning_rate": 0.00013529999999999998, "loss": 0.4538, "step": 902 }, { "epoch": 0.4249411764705882, "grad_norm": 5.961705240111002, "learning_rate": 0.00013544999999999999, "loss": 0.5383, "step": 903 }, { "epoch": 0.4254117647058824, "grad_norm": 0.6733913043796804, "learning_rate": 0.0001356, "loss": 0.1628, "step": 904 }, { "epoch": 0.4258823529411765, "grad_norm": 12.993171599560366, "learning_rate": 0.00013575, "loss": 0.626, "step": 905 }, { "epoch": 0.4263529411764706, "grad_norm": 6.421268138736397, "learning_rate": 0.0001359, "loss": 0.7777, "step": 906 }, { "epoch": 0.4268235294117647, "grad_norm": 3.1793786809647226, "learning_rate": 0.00013605, "loss": 0.4054, "step": 907 }, { "epoch": 0.4272941176470588, "grad_norm": 6.709527237274319, "learning_rate": 0.0001362, "loss": 0.7061, "step": 908 }, { "epoch": 0.42776470588235294, "grad_norm": 1.162395745794459, "learning_rate": 0.00013634999999999998, "loss": 0.2201, "step": 909 }, { "epoch": 0.42823529411764705, "grad_norm": 8.805556313644093, "learning_rate": 0.00013649999999999998, "loss": 0.7755, "step": 910 }, { "epoch": 0.42870588235294116, "grad_norm": 6.548656901416772, "learning_rate": 0.00013665, "loss": 0.5331, "step": 911 }, { "epoch": 0.42917647058823527, "grad_norm": 4.839869732250564, "learning_rate": 0.0001368, "loss": 0.6794, "step": 912 }, { "epoch": 0.42964705882352944, "grad_norm": 3.1947605670843076, "learning_rate": 0.00013695, "loss": 0.1686, "step": 913 }, { "epoch": 0.43011764705882355, "grad_norm": 6.089981402675963, "learning_rate": 0.0001371, "loss": 0.5213, "step": 914 }, { "epoch": 0.43058823529411766, "grad_norm": 2.532791707678698, "learning_rate": 0.00013725, "loss": 0.3388, "step": 915 }, { "epoch": 0.43105882352941177, "grad_norm": 6.274671368041178, "learning_rate": 0.0001374, "loss": 0.5356, "step": 916 }, { "epoch": 0.4315294117647059, "grad_norm": 5.263775458839623, "learning_rate": 0.00013754999999999998, "loss": 0.4215, "step": 917 }, { "epoch": 0.432, "grad_norm": 4.544213631622543, "learning_rate": 0.00013769999999999999, "loss": 0.8469, "step": 918 }, { "epoch": 0.4324705882352941, "grad_norm": 4.264330064118444, "learning_rate": 0.00013785, "loss": 0.5559, "step": 919 }, { "epoch": 0.4329411764705882, "grad_norm": 9.013846354913229, "learning_rate": 0.000138, "loss": 0.9102, "step": 920 }, { "epoch": 0.43341176470588233, "grad_norm": 2.6318767369321447, "learning_rate": 0.00013815, "loss": 0.1157, "step": 921 }, { "epoch": 0.4338823529411765, "grad_norm": 7.302430229545679, "learning_rate": 0.0001383, "loss": 0.4276, "step": 922 }, { "epoch": 0.4343529411764706, "grad_norm": 3.5083021409210686, "learning_rate": 0.00013845, "loss": 0.2005, "step": 923 }, { "epoch": 0.4348235294117647, "grad_norm": 5.442876963091229, "learning_rate": 0.0001386, "loss": 0.3162, "step": 924 }, { "epoch": 0.43529411764705883, "grad_norm": 12.853547153970618, "learning_rate": 0.00013874999999999998, "loss": 0.789, "step": 925 }, { "epoch": 0.43576470588235294, "grad_norm": 6.375508165376869, "learning_rate": 0.0001389, "loss": 0.4383, "step": 926 }, { "epoch": 0.43623529411764705, "grad_norm": 1.4208072522013784, "learning_rate": 0.00013905, "loss": 0.4767, "step": 927 }, { "epoch": 0.43670588235294117, "grad_norm": 7.591756301929449, "learning_rate": 0.0001392, "loss": 0.4624, "step": 928 }, { "epoch": 0.4371764705882353, "grad_norm": 12.748578378594651, "learning_rate": 0.00013935, "loss": 0.7799, "step": 929 }, { "epoch": 0.4376470588235294, "grad_norm": 5.141513981903017, "learning_rate": 0.0001395, "loss": 0.1799, "step": 930 }, { "epoch": 0.43811764705882356, "grad_norm": 6.338307570865685, "learning_rate": 0.00013965, "loss": 0.5124, "step": 931 }, { "epoch": 0.43858823529411767, "grad_norm": 3.0121387229076504, "learning_rate": 0.00013979999999999998, "loss": 0.5421, "step": 932 }, { "epoch": 0.4390588235294118, "grad_norm": 3.3659289323572907, "learning_rate": 0.00013995, "loss": 0.559, "step": 933 }, { "epoch": 0.4395294117647059, "grad_norm": 2.386252523454274, "learning_rate": 0.0001401, "loss": 0.3236, "step": 934 }, { "epoch": 0.44, "grad_norm": 1.672077342891995, "learning_rate": 0.00014025, "loss": 0.3189, "step": 935 }, { "epoch": 0.4404705882352941, "grad_norm": 1.9095440892944802, "learning_rate": 0.0001404, "loss": 0.5196, "step": 936 }, { "epoch": 0.4409411764705882, "grad_norm": 1.580261372440491, "learning_rate": 0.00014055, "loss": 0.3467, "step": 937 }, { "epoch": 0.44141176470588234, "grad_norm": 2.777159161376466, "learning_rate": 0.00014069999999999998, "loss": 0.2164, "step": 938 }, { "epoch": 0.44188235294117645, "grad_norm": 3.8423489412913714, "learning_rate": 0.00014084999999999998, "loss": 0.1104, "step": 939 }, { "epoch": 0.4423529411764706, "grad_norm": 2.565270456836593, "learning_rate": 0.00014099999999999998, "loss": 0.0634, "step": 940 }, { "epoch": 0.4428235294117647, "grad_norm": 5.691570639308471, "learning_rate": 0.00014115, "loss": 0.2142, "step": 941 }, { "epoch": 0.44329411764705884, "grad_norm": 0.810984296569636, "learning_rate": 0.0001413, "loss": 0.141, "step": 942 }, { "epoch": 0.44376470588235295, "grad_norm": 0.8140959498119239, "learning_rate": 0.00014144999999999997, "loss": 0.0727, "step": 943 }, { "epoch": 0.44423529411764706, "grad_norm": 21.047546589525282, "learning_rate": 0.00014159999999999997, "loss": 0.41, "step": 944 }, { "epoch": 0.4447058823529412, "grad_norm": 0.8188355666005358, "learning_rate": 0.00014174999999999998, "loss": 0.2592, "step": 945 }, { "epoch": 0.4451764705882353, "grad_norm": 0.9039369926495444, "learning_rate": 0.00014189999999999998, "loss": 0.3633, "step": 946 }, { "epoch": 0.4456470588235294, "grad_norm": 7.2912774867340575, "learning_rate": 0.00014204999999999998, "loss": 0.5428, "step": 947 }, { "epoch": 0.4461176470588235, "grad_norm": 0.7869749299706491, "learning_rate": 0.0001422, "loss": 0.1809, "step": 948 }, { "epoch": 0.4465882352941177, "grad_norm": 2.4297429095805785, "learning_rate": 0.00014235, "loss": 0.1832, "step": 949 }, { "epoch": 0.4470588235294118, "grad_norm": 7.707944099814535, "learning_rate": 0.0001425, "loss": 0.3041, "step": 950 }, { "epoch": 0.4475294117647059, "grad_norm": 1.0843093483134787, "learning_rate": 0.00014264999999999997, "loss": 0.4351, "step": 951 }, { "epoch": 0.448, "grad_norm": 1.2233579829187584, "learning_rate": 0.00014279999999999997, "loss": 0.6034, "step": 952 }, { "epoch": 0.4484705882352941, "grad_norm": 1.3086587668959855, "learning_rate": 0.00014294999999999998, "loss": 0.2618, "step": 953 }, { "epoch": 0.44894117647058823, "grad_norm": 7.450452132571332, "learning_rate": 0.00014309999999999998, "loss": 0.4105, "step": 954 }, { "epoch": 0.44941176470588234, "grad_norm": 0.7191420080487996, "learning_rate": 0.00014324999999999999, "loss": 0.2121, "step": 955 }, { "epoch": 0.44988235294117646, "grad_norm": 0.7102595234006356, "learning_rate": 0.0001434, "loss": 0.1318, "step": 956 }, { "epoch": 0.45035294117647057, "grad_norm": 0.5917642999289192, "learning_rate": 0.00014355, "loss": 0.1889, "step": 957 }, { "epoch": 0.4508235294117647, "grad_norm": 0.6871308741057746, "learning_rate": 0.00014369999999999997, "loss": 0.2051, "step": 958 }, { "epoch": 0.45129411764705885, "grad_norm": 0.6706078359931759, "learning_rate": 0.00014384999999999997, "loss": 0.0712, "step": 959 }, { "epoch": 0.45176470588235296, "grad_norm": 1.1177369483554696, "learning_rate": 0.00014399999999999998, "loss": 0.3177, "step": 960 }, { "epoch": 0.45223529411764707, "grad_norm": 0.6983025646724733, "learning_rate": 0.00014414999999999998, "loss": 0.1821, "step": 961 }, { "epoch": 0.4527058823529412, "grad_norm": 1.6677027655201446, "learning_rate": 0.00014429999999999998, "loss": 0.1732, "step": 962 }, { "epoch": 0.4531764705882353, "grad_norm": 1.5554527643629936, "learning_rate": 0.00014445, "loss": 0.2008, "step": 963 }, { "epoch": 0.4536470588235294, "grad_norm": 0.6244435097999916, "learning_rate": 0.0001446, "loss": 0.1159, "step": 964 }, { "epoch": 0.4541176470588235, "grad_norm": 1.1901139856115301, "learning_rate": 0.00014475, "loss": 0.4961, "step": 965 }, { "epoch": 0.4545882352941176, "grad_norm": 1.140167074147596, "learning_rate": 0.00014489999999999997, "loss": 0.2721, "step": 966 }, { "epoch": 0.45505882352941174, "grad_norm": 0.6002723034147419, "learning_rate": 0.00014504999999999997, "loss": 0.1259, "step": 967 }, { "epoch": 0.4555294117647059, "grad_norm": 1.0985947218528371, "learning_rate": 0.00014519999999999998, "loss": 0.3446, "step": 968 }, { "epoch": 0.456, "grad_norm": 0.8721432314197498, "learning_rate": 0.00014534999999999998, "loss": 0.3744, "step": 969 }, { "epoch": 0.45647058823529413, "grad_norm": 7.232018281328646, "learning_rate": 0.00014549999999999999, "loss": 0.6358, "step": 970 }, { "epoch": 0.45694117647058824, "grad_norm": 38.80681160720234, "learning_rate": 0.00014565, "loss": 1.046, "step": 971 }, { "epoch": 0.45741176470588235, "grad_norm": 11.086706323486135, "learning_rate": 0.0001458, "loss": 0.4921, "step": 972 }, { "epoch": 0.45788235294117646, "grad_norm": 29.025229540509617, "learning_rate": 0.00014595, "loss": 0.5408, "step": 973 }, { "epoch": 0.4583529411764706, "grad_norm": 15.064286346003952, "learning_rate": 0.00014609999999999997, "loss": 0.7316, "step": 974 }, { "epoch": 0.4588235294117647, "grad_norm": 0.9228732026801424, "learning_rate": 0.00014624999999999998, "loss": 0.0749, "step": 975 }, { "epoch": 0.4592941176470588, "grad_norm": 2.991573758943565, "learning_rate": 0.00014639999999999998, "loss": 0.5325, "step": 976 }, { "epoch": 0.45976470588235296, "grad_norm": 0.9305078476179732, "learning_rate": 0.00014654999999999998, "loss": 0.2597, "step": 977 }, { "epoch": 0.4602352941176471, "grad_norm": 1.037901387879196, "learning_rate": 0.0001467, "loss": 0.3152, "step": 978 }, { "epoch": 0.4607058823529412, "grad_norm": 0.8613966983635162, "learning_rate": 0.00014685, "loss": 0.2334, "step": 979 }, { "epoch": 0.4611764705882353, "grad_norm": 0.5733663516928473, "learning_rate": 0.000147, "loss": 0.0878, "step": 980 }, { "epoch": 0.4616470588235294, "grad_norm": 7.262072300703794, "learning_rate": 0.00014714999999999997, "loss": 0.3094, "step": 981 }, { "epoch": 0.4621176470588235, "grad_norm": 0.5854359800721964, "learning_rate": 0.00014729999999999998, "loss": 0.1772, "step": 982 }, { "epoch": 0.46258823529411763, "grad_norm": 0.8681210419225494, "learning_rate": 0.00014744999999999998, "loss": 0.3014, "step": 983 }, { "epoch": 0.46305882352941174, "grad_norm": 1.7115993722265477, "learning_rate": 0.00014759999999999998, "loss": 0.2431, "step": 984 }, { "epoch": 0.46352941176470586, "grad_norm": 1.442421140228279, "learning_rate": 0.00014774999999999999, "loss": 0.4263, "step": 985 }, { "epoch": 0.464, "grad_norm": 0.9554536846247302, "learning_rate": 0.0001479, "loss": 0.3015, "step": 986 }, { "epoch": 0.46447058823529414, "grad_norm": 0.7203866721113437, "learning_rate": 0.00014805, "loss": 0.2598, "step": 987 }, { "epoch": 0.46494117647058825, "grad_norm": 0.7752140597402443, "learning_rate": 0.0001482, "loss": 0.0747, "step": 988 }, { "epoch": 0.46541176470588236, "grad_norm": 0.7062895950254907, "learning_rate": 0.00014834999999999997, "loss": 0.1327, "step": 989 }, { "epoch": 0.46588235294117647, "grad_norm": 1.617380355947511, "learning_rate": 0.00014849999999999998, "loss": 0.4537, "step": 990 }, { "epoch": 0.4663529411764706, "grad_norm": 0.7146491843138265, "learning_rate": 0.00014864999999999998, "loss": 0.3465, "step": 991 }, { "epoch": 0.4668235294117647, "grad_norm": 0.7575458212668422, "learning_rate": 0.00014879999999999998, "loss": 0.1024, "step": 992 }, { "epoch": 0.4672941176470588, "grad_norm": 0.9521947012780693, "learning_rate": 0.00014895, "loss": 0.1826, "step": 993 }, { "epoch": 0.4677647058823529, "grad_norm": 2.590052488630991, "learning_rate": 0.0001491, "loss": 0.1728, "step": 994 }, { "epoch": 0.4682352941176471, "grad_norm": 0.5894474925176637, "learning_rate": 0.00014925, "loss": 0.1269, "step": 995 }, { "epoch": 0.4687058823529412, "grad_norm": 0.5645507089254999, "learning_rate": 0.0001494, "loss": 0.1346, "step": 996 }, { "epoch": 0.4691764705882353, "grad_norm": 1.571834598468916, "learning_rate": 0.00014954999999999998, "loss": 0.2517, "step": 997 }, { "epoch": 0.4696470588235294, "grad_norm": 2.341256815096256, "learning_rate": 0.00014969999999999998, "loss": 0.6647, "step": 998 }, { "epoch": 0.47011764705882353, "grad_norm": 0.8654724788781635, "learning_rate": 0.00014984999999999998, "loss": 0.3362, "step": 999 }, { "epoch": 0.47058823529411764, "grad_norm": 1.0961875068825564, "learning_rate": 0.00015, "loss": 0.4022, "step": 1000 }, { "epoch": 0.47105882352941175, "grad_norm": 0.4424967468119188, "learning_rate": 0.00015014999999999996, "loss": 0.0507, "step": 1001 }, { "epoch": 0.47152941176470586, "grad_norm": 0.9652596092356498, "learning_rate": 0.0001503, "loss": 0.2844, "step": 1002 }, { "epoch": 0.472, "grad_norm": 0.5763264038235892, "learning_rate": 0.00015044999999999997, "loss": 0.1214, "step": 1003 }, { "epoch": 0.47247058823529414, "grad_norm": 1.1507002196469562, "learning_rate": 0.00015059999999999997, "loss": 0.2996, "step": 1004 }, { "epoch": 0.47294117647058825, "grad_norm": 0.5200618034793709, "learning_rate": 0.00015074999999999998, "loss": 0.1437, "step": 1005 }, { "epoch": 0.47341176470588237, "grad_norm": 0.6085987721935882, "learning_rate": 0.00015089999999999998, "loss": 0.0958, "step": 1006 }, { "epoch": 0.4738823529411765, "grad_norm": 1.4186926249688996, "learning_rate": 0.00015104999999999996, "loss": 0.419, "step": 1007 }, { "epoch": 0.4743529411764706, "grad_norm": 0.38642710525181256, "learning_rate": 0.0001512, "loss": 0.0554, "step": 1008 }, { "epoch": 0.4748235294117647, "grad_norm": 1.151334942199187, "learning_rate": 0.00015134999999999997, "loss": 0.36, "step": 1009 }, { "epoch": 0.4752941176470588, "grad_norm": 0.8369683129814086, "learning_rate": 0.0001515, "loss": 0.3176, "step": 1010 }, { "epoch": 0.4757647058823529, "grad_norm": 1.307175571885699, "learning_rate": 0.00015164999999999997, "loss": 0.3545, "step": 1011 }, { "epoch": 0.47623529411764703, "grad_norm": 1.0145930420968439, "learning_rate": 0.00015179999999999998, "loss": 0.3219, "step": 1012 }, { "epoch": 0.4767058823529412, "grad_norm": 1.3925369352947918, "learning_rate": 0.00015194999999999998, "loss": 0.3297, "step": 1013 }, { "epoch": 0.4771764705882353, "grad_norm": 0.32118664134616637, "learning_rate": 0.00015209999999999998, "loss": 0.054, "step": 1014 }, { "epoch": 0.4776470588235294, "grad_norm": 0.6590928370482014, "learning_rate": 0.00015224999999999996, "loss": 0.1662, "step": 1015 }, { "epoch": 0.47811764705882354, "grad_norm": 0.9863886472392943, "learning_rate": 0.0001524, "loss": 0.2721, "step": 1016 }, { "epoch": 0.47858823529411765, "grad_norm": 0.4607060347639457, "learning_rate": 0.00015254999999999997, "loss": 0.1671, "step": 1017 }, { "epoch": 0.47905882352941176, "grad_norm": 1.2680851571055867, "learning_rate": 0.0001527, "loss": 0.3991, "step": 1018 }, { "epoch": 0.47952941176470587, "grad_norm": 0.905155012640033, "learning_rate": 0.00015284999999999997, "loss": 0.1595, "step": 1019 }, { "epoch": 0.48, "grad_norm": 0.359224523209878, "learning_rate": 0.00015299999999999998, "loss": 0.1123, "step": 1020 }, { "epoch": 0.4804705882352941, "grad_norm": 1.4602507598302135, "learning_rate": 0.00015314999999999998, "loss": 0.4814, "step": 1021 }, { "epoch": 0.48094117647058826, "grad_norm": 0.6768395756247969, "learning_rate": 0.00015329999999999999, "loss": 0.2958, "step": 1022 }, { "epoch": 0.4814117647058824, "grad_norm": 0.7430265481127665, "learning_rate": 0.00015344999999999996, "loss": 0.19, "step": 1023 }, { "epoch": 0.4818823529411765, "grad_norm": 0.8822880338513929, "learning_rate": 0.0001536, "loss": 0.1554, "step": 1024 }, { "epoch": 0.4823529411764706, "grad_norm": 0.635178601483674, "learning_rate": 0.00015374999999999997, "loss": 0.0992, "step": 1025 }, { "epoch": 0.4828235294117647, "grad_norm": 0.8726166859685486, "learning_rate": 0.0001539, "loss": 0.3487, "step": 1026 }, { "epoch": 0.4832941176470588, "grad_norm": 0.420651675430575, "learning_rate": 0.00015404999999999998, "loss": 0.1034, "step": 1027 }, { "epoch": 0.48376470588235293, "grad_norm": 0.5235242638748837, "learning_rate": 0.00015419999999999998, "loss": 0.0958, "step": 1028 }, { "epoch": 0.48423529411764704, "grad_norm": 0.07013813662443037, "learning_rate": 0.00015434999999999998, "loss": 0.0106, "step": 1029 }, { "epoch": 0.48470588235294115, "grad_norm": 4.626051489243828, "learning_rate": 0.0001545, "loss": 0.3214, "step": 1030 }, { "epoch": 0.4851764705882353, "grad_norm": 0.9798253380195472, "learning_rate": 0.00015464999999999996, "loss": 0.3158, "step": 1031 }, { "epoch": 0.48564705882352943, "grad_norm": 0.43716876142628935, "learning_rate": 0.0001548, "loss": 0.0812, "step": 1032 }, { "epoch": 0.48611764705882354, "grad_norm": 0.8912199596742821, "learning_rate": 0.00015494999999999997, "loss": 0.318, "step": 1033 }, { "epoch": 0.48658823529411765, "grad_norm": 1.9213372917497435, "learning_rate": 0.0001551, "loss": 0.5262, "step": 1034 }, { "epoch": 0.48705882352941177, "grad_norm": 0.306563600452669, "learning_rate": 0.00015524999999999998, "loss": 0.0986, "step": 1035 }, { "epoch": 0.4875294117647059, "grad_norm": 1.1686929322020168, "learning_rate": 0.00015539999999999998, "loss": 0.2839, "step": 1036 }, { "epoch": 0.488, "grad_norm": 0.29952140658448323, "learning_rate": 0.00015554999999999999, "loss": 0.0662, "step": 1037 }, { "epoch": 0.4884705882352941, "grad_norm": 1.4569128713235902, "learning_rate": 0.0001557, "loss": 0.4877, "step": 1038 }, { "epoch": 0.4889411764705882, "grad_norm": 0.5714966856821587, "learning_rate": 0.00015584999999999997, "loss": 0.084, "step": 1039 }, { "epoch": 0.4894117647058824, "grad_norm": 0.7021419825908152, "learning_rate": 0.000156, "loss": 0.2795, "step": 1040 }, { "epoch": 0.4898823529411765, "grad_norm": 0.5476481003682916, "learning_rate": 0.00015614999999999997, "loss": 0.2177, "step": 1041 }, { "epoch": 0.4903529411764706, "grad_norm": 0.6827349515561578, "learning_rate": 0.0001563, "loss": 0.1705, "step": 1042 }, { "epoch": 0.4908235294117647, "grad_norm": 0.5836572883900867, "learning_rate": 0.00015644999999999998, "loss": 0.2218, "step": 1043 }, { "epoch": 0.4912941176470588, "grad_norm": 5.473561012981751, "learning_rate": 0.00015659999999999998, "loss": 0.5186, "step": 1044 }, { "epoch": 0.49176470588235294, "grad_norm": 3.6107966039873545, "learning_rate": 0.00015675, "loss": 0.5252, "step": 1045 }, { "epoch": 0.49223529411764705, "grad_norm": 1.506763263144273, "learning_rate": 0.0001569, "loss": 0.2541, "step": 1046 }, { "epoch": 0.49270588235294116, "grad_norm": 0.5109201183150187, "learning_rate": 0.00015704999999999997, "loss": 0.1249, "step": 1047 }, { "epoch": 0.49317647058823527, "grad_norm": 1.6495497507230155, "learning_rate": 0.0001572, "loss": 0.5413, "step": 1048 }, { "epoch": 0.49364705882352944, "grad_norm": 1.3761567409706597, "learning_rate": 0.00015734999999999998, "loss": 0.2794, "step": 1049 }, { "epoch": 0.49411764705882355, "grad_norm": 0.2445769723570262, "learning_rate": 0.00015749999999999998, "loss": 0.0227, "step": 1050 }, { "epoch": 0.49458823529411766, "grad_norm": 0.8930689408964441, "learning_rate": 0.00015764999999999998, "loss": 0.2497, "step": 1051 }, { "epoch": 0.4950588235294118, "grad_norm": 1.6714742328832988, "learning_rate": 0.0001578, "loss": 0.3161, "step": 1052 }, { "epoch": 0.4955294117647059, "grad_norm": 0.8757544274548296, "learning_rate": 0.00015794999999999996, "loss": 0.1739, "step": 1053 }, { "epoch": 0.496, "grad_norm": 1.2518773847089628, "learning_rate": 0.0001581, "loss": 0.3737, "step": 1054 }, { "epoch": 0.4964705882352941, "grad_norm": 0.44254882017361347, "learning_rate": 0.00015824999999999997, "loss": 0.1704, "step": 1055 }, { "epoch": 0.4969411764705882, "grad_norm": 1.2688079499099034, "learning_rate": 0.0001584, "loss": 0.5068, "step": 1056 }, { "epoch": 0.49741176470588233, "grad_norm": 1.1260833589746972, "learning_rate": 0.00015854999999999998, "loss": 0.3979, "step": 1057 }, { "epoch": 0.4978823529411765, "grad_norm": 1.6248912711710035, "learning_rate": 0.00015869999999999998, "loss": 0.4712, "step": 1058 }, { "epoch": 0.4983529411764706, "grad_norm": 0.4586677616604469, "learning_rate": 0.00015884999999999999, "loss": 0.0934, "step": 1059 }, { "epoch": 0.4988235294117647, "grad_norm": 0.18839232602805606, "learning_rate": 0.000159, "loss": 0.016, "step": 1060 }, { "epoch": 0.49929411764705883, "grad_norm": 0.4941092877910045, "learning_rate": 0.00015914999999999997, "loss": 0.1077, "step": 1061 }, { "epoch": 0.49976470588235294, "grad_norm": 0.5649325768700973, "learning_rate": 0.0001593, "loss": 0.1242, "step": 1062 }, { "epoch": 0.5002352941176471, "grad_norm": 0.48711393875972914, "learning_rate": 0.00015944999999999997, "loss": 0.0486, "step": 1063 }, { "epoch": 0.5007058823529412, "grad_norm": 0.6044276860090443, "learning_rate": 0.0001596, "loss": 0.1721, "step": 1064 }, { "epoch": 0.5011764705882353, "grad_norm": 0.2923527696903157, "learning_rate": 0.00015974999999999998, "loss": 0.0411, "step": 1065 }, { "epoch": 0.5016470588235294, "grad_norm": 1.2031986723036616, "learning_rate": 0.00015989999999999998, "loss": 0.2592, "step": 1066 }, { "epoch": 0.5021176470588236, "grad_norm": 0.8382344267522083, "learning_rate": 0.00016005, "loss": 0.3638, "step": 1067 }, { "epoch": 0.5025882352941177, "grad_norm": 1.2452608683114745, "learning_rate": 0.0001602, "loss": 0.3999, "step": 1068 }, { "epoch": 0.5030588235294118, "grad_norm": 0.521484235969348, "learning_rate": 0.00016034999999999997, "loss": 0.0766, "step": 1069 }, { "epoch": 0.5035294117647059, "grad_norm": 0.3606989226724685, "learning_rate": 0.0001605, "loss": 0.0677, "step": 1070 }, { "epoch": 0.504, "grad_norm": 0.8422304720221719, "learning_rate": 0.00016064999999999997, "loss": 0.2392, "step": 1071 }, { "epoch": 0.5044705882352941, "grad_norm": 0.4584641740218889, "learning_rate": 0.0001608, "loss": 0.0716, "step": 1072 }, { "epoch": 0.5049411764705882, "grad_norm": 1.0147294104828415, "learning_rate": 0.00016094999999999998, "loss": 0.2459, "step": 1073 }, { "epoch": 0.5054117647058823, "grad_norm": 1.1971144192850374, "learning_rate": 0.00016109999999999999, "loss": 0.2204, "step": 1074 }, { "epoch": 0.5058823529411764, "grad_norm": 1.0007520098087324, "learning_rate": 0.00016125, "loss": 0.2413, "step": 1075 }, { "epoch": 0.5063529411764706, "grad_norm": 0.8499159009882845, "learning_rate": 0.0001614, "loss": 0.2272, "step": 1076 }, { "epoch": 0.5068235294117647, "grad_norm": 0.2028411661274071, "learning_rate": 0.00016154999999999997, "loss": 0.0259, "step": 1077 }, { "epoch": 0.5072941176470588, "grad_norm": 0.39843566926229546, "learning_rate": 0.0001617, "loss": 0.055, "step": 1078 }, { "epoch": 0.5077647058823529, "grad_norm": 6.124762218397902, "learning_rate": 0.00016184999999999998, "loss": 0.204, "step": 1079 }, { "epoch": 0.508235294117647, "grad_norm": 0.6248466310103998, "learning_rate": 0.000162, "loss": 0.083, "step": 1080 }, { "epoch": 0.5087058823529412, "grad_norm": 1.4616966264532902, "learning_rate": 0.00016214999999999998, "loss": 0.3615, "step": 1081 }, { "epoch": 0.5091764705882353, "grad_norm": 0.9656382174780354, "learning_rate": 0.0001623, "loss": 0.1325, "step": 1082 }, { "epoch": 0.5096470588235295, "grad_norm": 2.2229045514429875, "learning_rate": 0.00016245, "loss": 0.2077, "step": 1083 }, { "epoch": 0.5101176470588236, "grad_norm": 1.9984113499561504, "learning_rate": 0.0001626, "loss": 0.4697, "step": 1084 }, { "epoch": 0.5105882352941177, "grad_norm": 0.6467920913339389, "learning_rate": 0.00016274999999999997, "loss": 0.1637, "step": 1085 }, { "epoch": 0.5110588235294118, "grad_norm": 1.5069173389693145, "learning_rate": 0.0001629, "loss": 0.1684, "step": 1086 }, { "epoch": 0.5115294117647059, "grad_norm": 0.7205048698597094, "learning_rate": 0.00016304999999999998, "loss": 0.166, "step": 1087 }, { "epoch": 0.512, "grad_norm": 0.5046264298552752, "learning_rate": 0.0001632, "loss": 0.0797, "step": 1088 }, { "epoch": 0.5124705882352941, "grad_norm": 1.466926082976974, "learning_rate": 0.00016334999999999999, "loss": 0.566, "step": 1089 }, { "epoch": 0.5129411764705882, "grad_norm": 0.622998680003944, "learning_rate": 0.0001635, "loss": 0.2364, "step": 1090 }, { "epoch": 0.5134117647058823, "grad_norm": 0.8448349521708642, "learning_rate": 0.00016365, "loss": 0.2681, "step": 1091 }, { "epoch": 0.5138823529411765, "grad_norm": 0.4653923670140467, "learning_rate": 0.0001638, "loss": 0.0647, "step": 1092 }, { "epoch": 0.5143529411764706, "grad_norm": 1.1382386720484856, "learning_rate": 0.00016394999999999997, "loss": 0.2819, "step": 1093 }, { "epoch": 0.5148235294117647, "grad_norm": 0.6634320595774346, "learning_rate": 0.0001641, "loss": 0.2597, "step": 1094 }, { "epoch": 0.5152941176470588, "grad_norm": 1.1130536225219156, "learning_rate": 0.00016424999999999998, "loss": 0.3162, "step": 1095 }, { "epoch": 0.5157647058823529, "grad_norm": 0.722146267077936, "learning_rate": 0.0001644, "loss": 0.1014, "step": 1096 }, { "epoch": 0.516235294117647, "grad_norm": 0.5676973407696262, "learning_rate": 0.00016455, "loss": 0.3024, "step": 1097 }, { "epoch": 0.5167058823529411, "grad_norm": 0.3542296257956678, "learning_rate": 0.0001647, "loss": 0.0523, "step": 1098 }, { "epoch": 0.5171764705882353, "grad_norm": 1.3980365852433927, "learning_rate": 0.00016485, "loss": 0.3383, "step": 1099 }, { "epoch": 0.5176470588235295, "grad_norm": 0.6835490674962503, "learning_rate": 0.000165, "loss": 0.0765, "step": 1100 }, { "epoch": 0.5181176470588236, "grad_norm": 0.8433532354869749, "learning_rate": 0.00016514999999999998, "loss": 0.2064, "step": 1101 }, { "epoch": 0.5185882352941177, "grad_norm": 0.890415407631117, "learning_rate": 0.0001653, "loss": 0.3185, "step": 1102 }, { "epoch": 0.5190588235294118, "grad_norm": 0.7516781977746991, "learning_rate": 0.00016544999999999998, "loss": 0.3514, "step": 1103 }, { "epoch": 0.5195294117647059, "grad_norm": 0.9117093430442811, "learning_rate": 0.0001656, "loss": 0.38, "step": 1104 }, { "epoch": 0.52, "grad_norm": 0.5250442004653912, "learning_rate": 0.00016575, "loss": 0.2359, "step": 1105 }, { "epoch": 0.5204705882352941, "grad_norm": 0.21657962438439432, "learning_rate": 0.0001659, "loss": 0.0352, "step": 1106 }, { "epoch": 0.5209411764705882, "grad_norm": 0.3503976497754981, "learning_rate": 0.00016604999999999997, "loss": 0.0568, "step": 1107 }, { "epoch": 0.5214117647058824, "grad_norm": 0.4036100840409614, "learning_rate": 0.0001662, "loss": 0.1039, "step": 1108 }, { "epoch": 0.5218823529411765, "grad_norm": 1.090556117822395, "learning_rate": 0.00016634999999999998, "loss": 0.4509, "step": 1109 }, { "epoch": 0.5223529411764706, "grad_norm": 0.7649046249370128, "learning_rate": 0.0001665, "loss": 0.3047, "step": 1110 }, { "epoch": 0.5228235294117647, "grad_norm": 0.4715823194426645, "learning_rate": 0.00016664999999999998, "loss": 0.1446, "step": 1111 }, { "epoch": 0.5232941176470588, "grad_norm": 0.7921294708962765, "learning_rate": 0.0001668, "loss": 0.2819, "step": 1112 }, { "epoch": 0.5237647058823529, "grad_norm": 0.683866845054071, "learning_rate": 0.00016695, "loss": 0.2465, "step": 1113 }, { "epoch": 0.524235294117647, "grad_norm": 1.5887265913898314, "learning_rate": 0.0001671, "loss": 0.2531, "step": 1114 }, { "epoch": 0.5247058823529411, "grad_norm": 0.6513838745726716, "learning_rate": 0.00016724999999999997, "loss": 0.2472, "step": 1115 }, { "epoch": 0.5251764705882352, "grad_norm": 0.6170771709528896, "learning_rate": 0.0001674, "loss": 0.1501, "step": 1116 }, { "epoch": 0.5256470588235295, "grad_norm": 0.5475810777663344, "learning_rate": 0.00016754999999999998, "loss": 0.1803, "step": 1117 }, { "epoch": 0.5261176470588236, "grad_norm": 1.173555715775318, "learning_rate": 0.0001677, "loss": 0.4608, "step": 1118 }, { "epoch": 0.5265882352941177, "grad_norm": 0.6202660426588124, "learning_rate": 0.00016785, "loss": 0.1802, "step": 1119 }, { "epoch": 0.5270588235294118, "grad_norm": 0.6596773200775811, "learning_rate": 0.000168, "loss": 0.3992, "step": 1120 }, { "epoch": 0.5275294117647059, "grad_norm": 0.6038645976289674, "learning_rate": 0.00016815, "loss": 0.2527, "step": 1121 }, { "epoch": 0.528, "grad_norm": 0.3667367230393716, "learning_rate": 0.0001683, "loss": 0.0929, "step": 1122 }, { "epoch": 0.5284705882352941, "grad_norm": 0.5129111991202076, "learning_rate": 0.00016844999999999997, "loss": 0.1053, "step": 1123 }, { "epoch": 0.5289411764705882, "grad_norm": 0.8242075787958765, "learning_rate": 0.0001686, "loss": 0.3498, "step": 1124 }, { "epoch": 0.5294117647058824, "grad_norm": 1.300639861426465, "learning_rate": 0.00016874999999999998, "loss": 0.3477, "step": 1125 }, { "epoch": 0.5298823529411765, "grad_norm": 0.600188725104145, "learning_rate": 0.00016889999999999996, "loss": 0.0791, "step": 1126 }, { "epoch": 0.5303529411764706, "grad_norm": 0.481552609700077, "learning_rate": 0.00016905, "loss": 0.1226, "step": 1127 }, { "epoch": 0.5308235294117647, "grad_norm": 0.8387049056137528, "learning_rate": 0.00016919999999999997, "loss": 0.2745, "step": 1128 }, { "epoch": 0.5312941176470588, "grad_norm": 1.2518357133400504, "learning_rate": 0.00016935, "loss": 0.1569, "step": 1129 }, { "epoch": 0.5317647058823529, "grad_norm": 0.9835193300895564, "learning_rate": 0.00016949999999999997, "loss": 0.2465, "step": 1130 }, { "epoch": 0.532235294117647, "grad_norm": 0.6601024510077071, "learning_rate": 0.00016964999999999998, "loss": 0.2357, "step": 1131 }, { "epoch": 0.5327058823529411, "grad_norm": 0.556039505926003, "learning_rate": 0.00016979999999999998, "loss": 0.1803, "step": 1132 }, { "epoch": 0.5331764705882353, "grad_norm": 0.8924443728796566, "learning_rate": 0.00016994999999999998, "loss": 0.3141, "step": 1133 }, { "epoch": 0.5336470588235294, "grad_norm": 2.1081855824391558, "learning_rate": 0.00017009999999999996, "loss": 0.0604, "step": 1134 }, { "epoch": 0.5341176470588235, "grad_norm": 0.8642314028989494, "learning_rate": 0.00017025, "loss": 0.1914, "step": 1135 }, { "epoch": 0.5345882352941177, "grad_norm": 0.7258417454221743, "learning_rate": 0.00017039999999999997, "loss": 0.2545, "step": 1136 }, { "epoch": 0.5350588235294118, "grad_norm": 0.6225152866260734, "learning_rate": 0.00017055, "loss": 0.203, "step": 1137 }, { "epoch": 0.5355294117647059, "grad_norm": 0.5395548897870507, "learning_rate": 0.00017069999999999998, "loss": 0.1296, "step": 1138 }, { "epoch": 0.536, "grad_norm": 0.44480956141090366, "learning_rate": 0.00017084999999999998, "loss": 0.0954, "step": 1139 }, { "epoch": 0.5364705882352941, "grad_norm": 0.3066740088539107, "learning_rate": 0.00017099999999999998, "loss": 0.0236, "step": 1140 }, { "epoch": 0.5369411764705883, "grad_norm": 1.0232061677149364, "learning_rate": 0.00017114999999999999, "loss": 0.3981, "step": 1141 }, { "epoch": 0.5374117647058824, "grad_norm": 0.9205191987019409, "learning_rate": 0.00017129999999999996, "loss": 0.4122, "step": 1142 }, { "epoch": 0.5378823529411765, "grad_norm": 1.3515600779741224, "learning_rate": 0.00017145, "loss": 0.5001, "step": 1143 }, { "epoch": 0.5383529411764706, "grad_norm": 0.5075514128496493, "learning_rate": 0.00017159999999999997, "loss": 0.1201, "step": 1144 }, { "epoch": 0.5388235294117647, "grad_norm": 1.2367398960121638, "learning_rate": 0.00017175, "loss": 0.2446, "step": 1145 }, { "epoch": 0.5392941176470588, "grad_norm": 2.2537161304374402, "learning_rate": 0.00017189999999999998, "loss": 0.459, "step": 1146 }, { "epoch": 0.5397647058823529, "grad_norm": 0.17225631566964264, "learning_rate": 0.00017204999999999998, "loss": 0.0377, "step": 1147 }, { "epoch": 0.540235294117647, "grad_norm": 0.8415325590326309, "learning_rate": 0.00017219999999999998, "loss": 0.1531, "step": 1148 }, { "epoch": 0.5407058823529411, "grad_norm": 1.675759367677962, "learning_rate": 0.00017235, "loss": 0.294, "step": 1149 }, { "epoch": 0.5411764705882353, "grad_norm": 0.8726109655408427, "learning_rate": 0.00017249999999999996, "loss": 0.1571, "step": 1150 }, { "epoch": 0.5416470588235294, "grad_norm": 0.3044920894761076, "learning_rate": 0.00017265, "loss": 0.0326, "step": 1151 }, { "epoch": 0.5421176470588235, "grad_norm": 0.47851498586522195, "learning_rate": 0.00017279999999999997, "loss": 0.0931, "step": 1152 }, { "epoch": 0.5425882352941176, "grad_norm": 0.6038070200541388, "learning_rate": 0.00017294999999999998, "loss": 0.1399, "step": 1153 }, { "epoch": 0.5430588235294118, "grad_norm": 1.378593897097814, "learning_rate": 0.00017309999999999998, "loss": 0.2559, "step": 1154 }, { "epoch": 0.5435294117647059, "grad_norm": 0.345496865326896, "learning_rate": 0.00017324999999999998, "loss": 0.0196, "step": 1155 }, { "epoch": 0.544, "grad_norm": 0.4258252553187074, "learning_rate": 0.00017339999999999996, "loss": 0.1315, "step": 1156 }, { "epoch": 0.5444705882352942, "grad_norm": 0.33034981725147833, "learning_rate": 0.00017355, "loss": 0.0663, "step": 1157 }, { "epoch": 0.5449411764705883, "grad_norm": 0.7478835598624191, "learning_rate": 0.00017369999999999997, "loss": 0.2773, "step": 1158 }, { "epoch": 0.5454117647058824, "grad_norm": 0.96638000409688, "learning_rate": 0.00017385, "loss": 0.3026, "step": 1159 }, { "epoch": 0.5458823529411765, "grad_norm": 0.5398397221506103, "learning_rate": 0.00017399999999999997, "loss": 0.262, "step": 1160 }, { "epoch": 0.5463529411764706, "grad_norm": 0.6590179400800917, "learning_rate": 0.00017414999999999998, "loss": 0.1845, "step": 1161 }, { "epoch": 0.5468235294117647, "grad_norm": 0.5626786984573875, "learning_rate": 0.00017429999999999998, "loss": 0.1642, "step": 1162 }, { "epoch": 0.5472941176470588, "grad_norm": 0.9083885817910294, "learning_rate": 0.00017444999999999998, "loss": 0.2183, "step": 1163 }, { "epoch": 0.5477647058823529, "grad_norm": 0.3418960742488844, "learning_rate": 0.00017459999999999996, "loss": 0.0702, "step": 1164 }, { "epoch": 0.548235294117647, "grad_norm": 0.4102394389252403, "learning_rate": 0.00017475, "loss": 0.0902, "step": 1165 }, { "epoch": 0.5487058823529412, "grad_norm": 0.571921666865139, "learning_rate": 0.00017489999999999997, "loss": 0.2169, "step": 1166 }, { "epoch": 0.5491764705882353, "grad_norm": 0.8341117455519208, "learning_rate": 0.00017505, "loss": 0.2073, "step": 1167 }, { "epoch": 0.5496470588235294, "grad_norm": 0.33076407610769265, "learning_rate": 0.00017519999999999998, "loss": 0.0852, "step": 1168 }, { "epoch": 0.5501176470588235, "grad_norm": 0.5499266408869229, "learning_rate": 0.00017534999999999998, "loss": 0.1595, "step": 1169 }, { "epoch": 0.5505882352941176, "grad_norm": 0.11800646376398853, "learning_rate": 0.00017549999999999998, "loss": 0.0103, "step": 1170 }, { "epoch": 0.5510588235294117, "grad_norm": 0.3954778393653873, "learning_rate": 0.00017565, "loss": 0.08, "step": 1171 }, { "epoch": 0.5515294117647059, "grad_norm": 0.37027428287854336, "learning_rate": 0.00017579999999999996, "loss": 0.1158, "step": 1172 }, { "epoch": 0.552, "grad_norm": 0.6835984500596188, "learning_rate": 0.00017595, "loss": 0.2342, "step": 1173 }, { "epoch": 0.5524705882352942, "grad_norm": 0.36706912787339663, "learning_rate": 0.00017609999999999997, "loss": 0.0499, "step": 1174 }, { "epoch": 0.5529411764705883, "grad_norm": 0.5858823471589791, "learning_rate": 0.00017625, "loss": 0.2118, "step": 1175 }, { "epoch": 0.5534117647058824, "grad_norm": 0.5585870984458298, "learning_rate": 0.00017639999999999998, "loss": 0.0568, "step": 1176 }, { "epoch": 0.5538823529411765, "grad_norm": 0.5948697762233845, "learning_rate": 0.00017654999999999998, "loss": 0.1724, "step": 1177 }, { "epoch": 0.5543529411764706, "grad_norm": 0.5620359685072426, "learning_rate": 0.00017669999999999999, "loss": 0.1769, "step": 1178 }, { "epoch": 0.5548235294117647, "grad_norm": 0.5371750006875435, "learning_rate": 0.00017685, "loss": 0.0775, "step": 1179 }, { "epoch": 0.5552941176470588, "grad_norm": 0.617167313555291, "learning_rate": 0.00017699999999999997, "loss": 0.0969, "step": 1180 }, { "epoch": 0.5557647058823529, "grad_norm": 0.8642660171237468, "learning_rate": 0.00017715, "loss": 0.2784, "step": 1181 }, { "epoch": 0.556235294117647, "grad_norm": 0.5293128740107668, "learning_rate": 0.00017729999999999997, "loss": 0.2048, "step": 1182 }, { "epoch": 0.5567058823529412, "grad_norm": 0.4452219038838237, "learning_rate": 0.00017745, "loss": 0.1279, "step": 1183 }, { "epoch": 0.5571764705882353, "grad_norm": 0.43722574783766865, "learning_rate": 0.00017759999999999998, "loss": 0.1665, "step": 1184 }, { "epoch": 0.5576470588235294, "grad_norm": 0.8100020608293627, "learning_rate": 0.00017774999999999998, "loss": 0.2694, "step": 1185 }, { "epoch": 0.5581176470588235, "grad_norm": 0.45984481171695, "learning_rate": 0.0001779, "loss": 0.0419, "step": 1186 }, { "epoch": 0.5585882352941176, "grad_norm": 0.696980264508867, "learning_rate": 0.00017805, "loss": 0.307, "step": 1187 }, { "epoch": 0.5590588235294117, "grad_norm": 0.9195121667570466, "learning_rate": 0.00017819999999999997, "loss": 0.228, "step": 1188 }, { "epoch": 0.5595294117647058, "grad_norm": 0.4665433816434737, "learning_rate": 0.00017835, "loss": 0.126, "step": 1189 }, { "epoch": 0.56, "grad_norm": 0.4587294401812934, "learning_rate": 0.00017849999999999997, "loss": 0.1655, "step": 1190 }, { "epoch": 0.5604705882352942, "grad_norm": 0.5404280556396007, "learning_rate": 0.00017865, "loss": 0.1615, "step": 1191 }, { "epoch": 0.5609411764705883, "grad_norm": 1.3368992877508854, "learning_rate": 0.00017879999999999998, "loss": 0.2614, "step": 1192 }, { "epoch": 0.5614117647058824, "grad_norm": 0.35161392533290314, "learning_rate": 0.00017894999999999999, "loss": 0.0684, "step": 1193 }, { "epoch": 0.5618823529411765, "grad_norm": 0.2629940244389818, "learning_rate": 0.0001791, "loss": 0.0436, "step": 1194 }, { "epoch": 0.5623529411764706, "grad_norm": 0.37419694951875465, "learning_rate": 0.00017925, "loss": 0.0476, "step": 1195 }, { "epoch": 0.5628235294117647, "grad_norm": 1.1796242099370517, "learning_rate": 0.00017939999999999997, "loss": 0.251, "step": 1196 }, { "epoch": 0.5632941176470588, "grad_norm": 0.5555967937878394, "learning_rate": 0.00017955, "loss": 0.1125, "step": 1197 }, { "epoch": 0.563764705882353, "grad_norm": 1.7661930269307213, "learning_rate": 0.00017969999999999998, "loss": 0.3358, "step": 1198 }, { "epoch": 0.5642352941176471, "grad_norm": 1.1916847695228348, "learning_rate": 0.00017984999999999998, "loss": 0.3518, "step": 1199 }, { "epoch": 0.5647058823529412, "grad_norm": 0.7796599331981615, "learning_rate": 0.00017999999999999998, "loss": 0.2631, "step": 1200 }, { "epoch": 0.5651764705882353, "grad_norm": 1.015136587325478, "learning_rate": 0.00018015, "loss": 0.2888, "step": 1201 }, { "epoch": 0.5656470588235294, "grad_norm": 0.3408756388818853, "learning_rate": 0.00018029999999999996, "loss": 0.1047, "step": 1202 }, { "epoch": 0.5661176470588235, "grad_norm": 0.5645812138306623, "learning_rate": 0.00018045, "loss": 0.0938, "step": 1203 }, { "epoch": 0.5665882352941176, "grad_norm": 0.5242780163650917, "learning_rate": 0.00018059999999999997, "loss": 0.0905, "step": 1204 }, { "epoch": 0.5670588235294117, "grad_norm": 2.4956617551906946, "learning_rate": 0.00018075, "loss": 0.6031, "step": 1205 }, { "epoch": 0.5675294117647058, "grad_norm": 1.3755413143145436, "learning_rate": 0.00018089999999999998, "loss": 0.2927, "step": 1206 }, { "epoch": 0.568, "grad_norm": 0.7360682591037941, "learning_rate": 0.00018104999999999998, "loss": 0.108, "step": 1207 }, { "epoch": 0.5684705882352942, "grad_norm": 0.46221572797046473, "learning_rate": 0.00018119999999999999, "loss": 0.0878, "step": 1208 }, { "epoch": 0.5689411764705883, "grad_norm": 1.81437924197642, "learning_rate": 0.00018135, "loss": 0.3007, "step": 1209 }, { "epoch": 0.5694117647058824, "grad_norm": 0.9510785159161858, "learning_rate": 0.00018149999999999997, "loss": 0.2789, "step": 1210 }, { "epoch": 0.5698823529411765, "grad_norm": 0.6927960900624988, "learning_rate": 0.00018165, "loss": 0.2672, "step": 1211 }, { "epoch": 0.5703529411764706, "grad_norm": 1.5395142200934233, "learning_rate": 0.00018179999999999997, "loss": 0.3067, "step": 1212 }, { "epoch": 0.5708235294117647, "grad_norm": 1.5925140349035114, "learning_rate": 0.00018195, "loss": 0.3324, "step": 1213 }, { "epoch": 0.5712941176470588, "grad_norm": 1.548577461524083, "learning_rate": 0.00018209999999999998, "loss": 0.4125, "step": 1214 }, { "epoch": 0.571764705882353, "grad_norm": 0.7218851909115418, "learning_rate": 0.00018224999999999998, "loss": 0.2432, "step": 1215 }, { "epoch": 0.5722352941176471, "grad_norm": 0.7990913556479552, "learning_rate": 0.0001824, "loss": 0.1722, "step": 1216 }, { "epoch": 0.5727058823529412, "grad_norm": 1.8623861070983756, "learning_rate": 0.00018255, "loss": 0.3577, "step": 1217 }, { "epoch": 0.5731764705882353, "grad_norm": 0.6769094488283385, "learning_rate": 0.00018269999999999997, "loss": 0.1759, "step": 1218 }, { "epoch": 0.5736470588235294, "grad_norm": 0.8068893733858393, "learning_rate": 0.00018285, "loss": 0.3851, "step": 1219 }, { "epoch": 0.5741176470588235, "grad_norm": 0.8184059889301434, "learning_rate": 0.00018299999999999998, "loss": 0.23, "step": 1220 }, { "epoch": 0.5745882352941176, "grad_norm": 0.9448853829867973, "learning_rate": 0.00018315, "loss": 0.3341, "step": 1221 }, { "epoch": 0.5750588235294117, "grad_norm": 0.8508798975645289, "learning_rate": 0.00018329999999999998, "loss": 0.2915, "step": 1222 }, { "epoch": 0.5755294117647058, "grad_norm": 0.4904719142405973, "learning_rate": 0.00018345, "loss": 0.2062, "step": 1223 }, { "epoch": 0.576, "grad_norm": 1.2177671694511232, "learning_rate": 0.0001836, "loss": 0.3756, "step": 1224 }, { "epoch": 0.5764705882352941, "grad_norm": 0.3452649443275163, "learning_rate": 0.00018375, "loss": 0.1034, "step": 1225 }, { "epoch": 0.5769411764705883, "grad_norm": 0.761736053561204, "learning_rate": 0.00018389999999999997, "loss": 0.2724, "step": 1226 }, { "epoch": 0.5774117647058824, "grad_norm": 0.7671721620374026, "learning_rate": 0.00018405, "loss": 0.2722, "step": 1227 }, { "epoch": 0.5778823529411765, "grad_norm": 0.4232330445696333, "learning_rate": 0.00018419999999999998, "loss": 0.156, "step": 1228 }, { "epoch": 0.5783529411764706, "grad_norm": 0.6151983920288158, "learning_rate": 0.00018435, "loss": 0.2455, "step": 1229 }, { "epoch": 0.5788235294117647, "grad_norm": 0.47256317490945865, "learning_rate": 0.00018449999999999999, "loss": 0.118, "step": 1230 }, { "epoch": 0.5792941176470588, "grad_norm": 0.2684027545332771, "learning_rate": 0.00018465, "loss": 0.0466, "step": 1231 }, { "epoch": 0.579764705882353, "grad_norm": 0.20369837002904306, "learning_rate": 0.0001848, "loss": 0.0199, "step": 1232 }, { "epoch": 0.5802352941176471, "grad_norm": 0.5218128310633723, "learning_rate": 0.00018495, "loss": 0.1318, "step": 1233 }, { "epoch": 0.5807058823529412, "grad_norm": 0.6616069580386046, "learning_rate": 0.00018509999999999997, "loss": 0.249, "step": 1234 }, { "epoch": 0.5811764705882353, "grad_norm": 0.2765198036046553, "learning_rate": 0.00018525, "loss": 0.0443, "step": 1235 }, { "epoch": 0.5816470588235294, "grad_norm": 0.5308220839226144, "learning_rate": 0.00018539999999999998, "loss": 0.1725, "step": 1236 }, { "epoch": 0.5821176470588235, "grad_norm": 0.48856816593264346, "learning_rate": 0.00018555, "loss": 0.1205, "step": 1237 }, { "epoch": 0.5825882352941176, "grad_norm": 0.644293067374527, "learning_rate": 0.0001857, "loss": 0.1916, "step": 1238 }, { "epoch": 0.5830588235294117, "grad_norm": 1.392501479217379, "learning_rate": 0.00018585, "loss": 0.493, "step": 1239 }, { "epoch": 0.5835294117647059, "grad_norm": 0.23546594890990868, "learning_rate": 0.000186, "loss": 0.0366, "step": 1240 }, { "epoch": 0.584, "grad_norm": 0.8654977444365201, "learning_rate": 0.00018615, "loss": 0.2554, "step": 1241 }, { "epoch": 0.5844705882352941, "grad_norm": 0.11895265677954937, "learning_rate": 0.00018629999999999997, "loss": 0.0203, "step": 1242 }, { "epoch": 0.5849411764705882, "grad_norm": 0.5530013290677592, "learning_rate": 0.00018645, "loss": 0.1734, "step": 1243 }, { "epoch": 0.5854117647058823, "grad_norm": 0.6704684359909427, "learning_rate": 0.00018659999999999998, "loss": 0.2874, "step": 1244 }, { "epoch": 0.5858823529411765, "grad_norm": 0.6878832059432105, "learning_rate": 0.00018675, "loss": 0.3136, "step": 1245 }, { "epoch": 0.5863529411764706, "grad_norm": 0.09140760323394637, "learning_rate": 0.0001869, "loss": 0.0036, "step": 1246 }, { "epoch": 0.5868235294117647, "grad_norm": 0.5087243384559359, "learning_rate": 0.00018705, "loss": 0.1327, "step": 1247 }, { "epoch": 0.5872941176470589, "grad_norm": 0.6476932084301972, "learning_rate": 0.0001872, "loss": 0.2782, "step": 1248 }, { "epoch": 0.587764705882353, "grad_norm": 0.6783813071820032, "learning_rate": 0.00018735, "loss": 0.274, "step": 1249 }, { "epoch": 0.5882352941176471, "grad_norm": 0.5015316668206583, "learning_rate": 0.00018749999999999998, "loss": 0.26, "step": 1250 }, { "epoch": 0.5887058823529412, "grad_norm": 0.33084720326049344, "learning_rate": 0.00018764999999999998, "loss": 0.0932, "step": 1251 }, { "epoch": 0.5891764705882353, "grad_norm": 0.754109966435222, "learning_rate": 0.00018779999999999998, "loss": 0.2396, "step": 1252 }, { "epoch": 0.5896470588235294, "grad_norm": 0.6112584094213203, "learning_rate": 0.00018794999999999996, "loss": 0.2272, "step": 1253 }, { "epoch": 0.5901176470588235, "grad_norm": 0.19595477754764173, "learning_rate": 0.0001881, "loss": 0.023, "step": 1254 }, { "epoch": 0.5905882352941176, "grad_norm": 0.36912305100406456, "learning_rate": 0.00018824999999999997, "loss": 0.0463, "step": 1255 }, { "epoch": 0.5910588235294117, "grad_norm": 0.5558105333315, "learning_rate": 0.00018839999999999997, "loss": 0.1516, "step": 1256 }, { "epoch": 0.5915294117647059, "grad_norm": 0.41765003016650976, "learning_rate": 0.00018854999999999998, "loss": 0.1149, "step": 1257 }, { "epoch": 0.592, "grad_norm": 0.7454711129417538, "learning_rate": 0.00018869999999999998, "loss": 0.2788, "step": 1258 }, { "epoch": 0.5924705882352941, "grad_norm": 0.9416354596479205, "learning_rate": 0.00018884999999999996, "loss": 0.4145, "step": 1259 }, { "epoch": 0.5929411764705882, "grad_norm": 0.48770627731210303, "learning_rate": 0.00018899999999999999, "loss": 0.1472, "step": 1260 }, { "epoch": 0.5934117647058823, "grad_norm": 0.6121168187052872, "learning_rate": 0.00018914999999999996, "loss": 0.1541, "step": 1261 }, { "epoch": 0.5938823529411764, "grad_norm": 0.5427676927675665, "learning_rate": 0.0001893, "loss": 0.2039, "step": 1262 }, { "epoch": 0.5943529411764706, "grad_norm": 1.394346719590549, "learning_rate": 0.00018944999999999997, "loss": 0.2686, "step": 1263 }, { "epoch": 0.5948235294117648, "grad_norm": 0.3831616706165368, "learning_rate": 0.00018959999999999997, "loss": 0.131, "step": 1264 }, { "epoch": 0.5952941176470589, "grad_norm": 0.48759447242809223, "learning_rate": 0.00018974999999999998, "loss": 0.1378, "step": 1265 }, { "epoch": 0.595764705882353, "grad_norm": 0.2511500173009873, "learning_rate": 0.00018989999999999998, "loss": 0.0685, "step": 1266 }, { "epoch": 0.5962352941176471, "grad_norm": 0.312743673010797, "learning_rate": 0.00019004999999999996, "loss": 0.0561, "step": 1267 }, { "epoch": 0.5967058823529412, "grad_norm": 0.7272083206487074, "learning_rate": 0.0001902, "loss": 0.2732, "step": 1268 }, { "epoch": 0.5971764705882353, "grad_norm": 0.4886563800053658, "learning_rate": 0.00019034999999999996, "loss": 0.1727, "step": 1269 }, { "epoch": 0.5976470588235294, "grad_norm": 0.5487707741122276, "learning_rate": 0.0001905, "loss": 0.1911, "step": 1270 }, { "epoch": 0.5981176470588235, "grad_norm": 1.048350614034549, "learning_rate": 0.00019064999999999997, "loss": 0.303, "step": 1271 }, { "epoch": 0.5985882352941176, "grad_norm": 0.7096870092343963, "learning_rate": 0.00019079999999999998, "loss": 0.2894, "step": 1272 }, { "epoch": 0.5990588235294118, "grad_norm": 0.45997028334098794, "learning_rate": 0.00019094999999999998, "loss": 0.107, "step": 1273 }, { "epoch": 0.5995294117647059, "grad_norm": 0.43670295932001285, "learning_rate": 0.00019109999999999998, "loss": 0.1135, "step": 1274 }, { "epoch": 0.6, "grad_norm": 0.11257921232649097, "learning_rate": 0.00019124999999999996, "loss": 0.0184, "step": 1275 }, { "epoch": 0.6004705882352941, "grad_norm": 1.3759748003927745, "learning_rate": 0.0001914, "loss": 0.2636, "step": 1276 }, { "epoch": 0.6009411764705882, "grad_norm": 40.421358634530236, "learning_rate": 0.00019154999999999997, "loss": 0.649, "step": 1277 }, { "epoch": 0.6014117647058823, "grad_norm": 64.87946361699363, "learning_rate": 0.0001917, "loss": 0.562, "step": 1278 }, { "epoch": 0.6018823529411764, "grad_norm": 253.13622558227925, "learning_rate": 0.00019184999999999997, "loss": 7.6107, "step": 1279 }, { "epoch": 0.6023529411764705, "grad_norm": 832.6178197201467, "learning_rate": 0.00019199999999999998, "loss": 9.0775, "step": 1280 }, { "epoch": 0.6028235294117648, "grad_norm": 11.04874544534692, "learning_rate": 0.00019214999999999998, "loss": 0.6959, "step": 1281 }, { "epoch": 0.6032941176470589, "grad_norm": 3.913248237491206, "learning_rate": 0.00019229999999999999, "loss": 0.6518, "step": 1282 }, { "epoch": 0.603764705882353, "grad_norm": 9.513935388732529, "learning_rate": 0.00019244999999999996, "loss": 0.6078, "step": 1283 }, { "epoch": 0.6042352941176471, "grad_norm": 41.93209608395266, "learning_rate": 0.0001926, "loss": 0.8641, "step": 1284 }, { "epoch": 0.6047058823529412, "grad_norm": 18.716077796605962, "learning_rate": 0.00019274999999999997, "loss": 0.4991, "step": 1285 }, { "epoch": 0.6051764705882353, "grad_norm": 11.296572262095609, "learning_rate": 0.0001929, "loss": 0.11, "step": 1286 }, { "epoch": 0.6056470588235294, "grad_norm": 7.668959447057747, "learning_rate": 0.00019304999999999998, "loss": 0.6947, "step": 1287 }, { "epoch": 0.6061176470588235, "grad_norm": 4.027779977264231, "learning_rate": 0.00019319999999999998, "loss": 0.5592, "step": 1288 }, { "epoch": 0.6065882352941177, "grad_norm": 5.022618681951433, "learning_rate": 0.00019334999999999998, "loss": 0.4685, "step": 1289 }, { "epoch": 0.6070588235294118, "grad_norm": 8.327979092209834, "learning_rate": 0.0001935, "loss": 0.8857, "step": 1290 }, { "epoch": 0.6075294117647059, "grad_norm": 0.22169205062567474, "learning_rate": 0.00019364999999999996, "loss": 0.0212, "step": 1291 }, { "epoch": 0.608, "grad_norm": 0.9199417188663693, "learning_rate": 0.0001938, "loss": 0.2726, "step": 1292 }, { "epoch": 0.6084705882352941, "grad_norm": 1.001067728878187, "learning_rate": 0.00019394999999999997, "loss": 0.2391, "step": 1293 }, { "epoch": 0.6089411764705882, "grad_norm": 0.7141641353716768, "learning_rate": 0.0001941, "loss": 0.1021, "step": 1294 }, { "epoch": 0.6094117647058823, "grad_norm": 1.723633515511285, "learning_rate": 0.00019424999999999998, "loss": 0.3009, "step": 1295 }, { "epoch": 0.6098823529411764, "grad_norm": 0.49599164251815214, "learning_rate": 0.00019439999999999998, "loss": 0.1118, "step": 1296 }, { "epoch": 0.6103529411764705, "grad_norm": 0.9776267354171025, "learning_rate": 0.00019454999999999999, "loss": 0.1473, "step": 1297 }, { "epoch": 0.6108235294117647, "grad_norm": 2.0432759994850342, "learning_rate": 0.0001947, "loss": 0.2181, "step": 1298 }, { "epoch": 0.6112941176470589, "grad_norm": 1.4401757551194243, "learning_rate": 0.00019484999999999997, "loss": 0.3259, "step": 1299 }, { "epoch": 0.611764705882353, "grad_norm": 1.6341481943592888, "learning_rate": 0.000195, "loss": 0.1851, "step": 1300 }, { "epoch": 0.6122352941176471, "grad_norm": 1.5462398858929618, "learning_rate": 0.00019514999999999997, "loss": 0.389, "step": 1301 }, { "epoch": 0.6127058823529412, "grad_norm": 1.79398515066942, "learning_rate": 0.00019529999999999998, "loss": 0.0814, "step": 1302 }, { "epoch": 0.6131764705882353, "grad_norm": 0.714748722126376, "learning_rate": 0.00019544999999999998, "loss": 0.3829, "step": 1303 }, { "epoch": 0.6136470588235294, "grad_norm": 0.9247813537125331, "learning_rate": 0.00019559999999999998, "loss": 0.1951, "step": 1304 }, { "epoch": 0.6141176470588235, "grad_norm": 0.7216246361712417, "learning_rate": 0.00019574999999999996, "loss": 0.278, "step": 1305 }, { "epoch": 0.6145882352941177, "grad_norm": 1.737176039045819, "learning_rate": 0.0001959, "loss": 0.3322, "step": 1306 }, { "epoch": 0.6150588235294118, "grad_norm": 0.7105211680470791, "learning_rate": 0.00019604999999999997, "loss": 0.0846, "step": 1307 }, { "epoch": 0.6155294117647059, "grad_norm": 1.9283136478461216, "learning_rate": 0.0001962, "loss": 0.6774, "step": 1308 }, { "epoch": 0.616, "grad_norm": 1.48110969564765, "learning_rate": 0.00019634999999999998, "loss": 0.0171, "step": 1309 }, { "epoch": 0.6164705882352941, "grad_norm": 1.0275770055291609, "learning_rate": 0.00019649999999999998, "loss": 0.3432, "step": 1310 }, { "epoch": 0.6169411764705882, "grad_norm": 1.2092217432233556, "learning_rate": 0.00019664999999999998, "loss": 0.378, "step": 1311 }, { "epoch": 0.6174117647058823, "grad_norm": 0.4929720105203299, "learning_rate": 0.00019679999999999999, "loss": 0.0623, "step": 1312 }, { "epoch": 0.6178823529411764, "grad_norm": 0.5918401843731186, "learning_rate": 0.00019694999999999996, "loss": 0.1286, "step": 1313 }, { "epoch": 0.6183529411764705, "grad_norm": 1.5594343502682202, "learning_rate": 0.0001971, "loss": 0.6003, "step": 1314 }, { "epoch": 0.6188235294117647, "grad_norm": 0.6658012684266439, "learning_rate": 0.00019724999999999997, "loss": 0.3586, "step": 1315 }, { "epoch": 0.6192941176470588, "grad_norm": 0.6324449051365418, "learning_rate": 0.0001974, "loss": 0.2744, "step": 1316 }, { "epoch": 0.619764705882353, "grad_norm": 0.5457595426607229, "learning_rate": 0.00019754999999999998, "loss": 0.2541, "step": 1317 }, { "epoch": 0.6202352941176471, "grad_norm": 0.5099831498521809, "learning_rate": 0.00019769999999999998, "loss": 0.3009, "step": 1318 }, { "epoch": 0.6207058823529412, "grad_norm": 0.755495882084854, "learning_rate": 0.00019784999999999998, "loss": 0.2565, "step": 1319 }, { "epoch": 0.6211764705882353, "grad_norm": 0.5580167459994507, "learning_rate": 0.000198, "loss": 0.1493, "step": 1320 }, { "epoch": 0.6216470588235294, "grad_norm": 0.3634962018574332, "learning_rate": 0.00019814999999999996, "loss": 0.0778, "step": 1321 }, { "epoch": 0.6221176470588236, "grad_norm": 0.46601988713338854, "learning_rate": 0.0001983, "loss": 0.0825, "step": 1322 }, { "epoch": 0.6225882352941177, "grad_norm": 0.7584602351734748, "learning_rate": 0.00019844999999999997, "loss": 0.1265, "step": 1323 }, { "epoch": 0.6230588235294118, "grad_norm": 0.310784748351069, "learning_rate": 0.0001986, "loss": 0.0884, "step": 1324 }, { "epoch": 0.6235294117647059, "grad_norm": 5.475736899884403, "learning_rate": 0.00019874999999999998, "loss": 0.143, "step": 1325 }, { "epoch": 0.624, "grad_norm": 0.7880618060001792, "learning_rate": 0.00019889999999999998, "loss": 0.2519, "step": 1326 }, { "epoch": 0.6244705882352941, "grad_norm": 4.329357392424001, "learning_rate": 0.00019905, "loss": 1.0371, "step": 1327 }, { "epoch": 0.6249411764705882, "grad_norm": 0.20932468702279652, "learning_rate": 0.0001992, "loss": 0.0236, "step": 1328 }, { "epoch": 0.6254117647058823, "grad_norm": 0.910449955674078, "learning_rate": 0.00019934999999999997, "loss": 0.1375, "step": 1329 }, { "epoch": 0.6258823529411764, "grad_norm": 0.5773590626409246, "learning_rate": 0.0001995, "loss": 0.1052, "step": 1330 }, { "epoch": 0.6263529411764706, "grad_norm": 1.0356254830634284, "learning_rate": 0.00019964999999999997, "loss": 0.1526, "step": 1331 }, { "epoch": 0.6268235294117647, "grad_norm": 3.0131663510746316, "learning_rate": 0.0001998, "loss": 0.0855, "step": 1332 }, { "epoch": 0.6272941176470588, "grad_norm": 0.7847735317093094, "learning_rate": 0.00019994999999999998, "loss": 0.1947, "step": 1333 }, { "epoch": 0.6277647058823529, "grad_norm": 1.1942422149831637, "learning_rate": 0.00020009999999999998, "loss": 0.4082, "step": 1334 }, { "epoch": 0.6282352941176471, "grad_norm": 0.4155673835108706, "learning_rate": 0.00020025, "loss": 0.1103, "step": 1335 }, { "epoch": 0.6287058823529412, "grad_norm": 0.6934997349617189, "learning_rate": 0.0002004, "loss": 0.2132, "step": 1336 }, { "epoch": 0.6291764705882353, "grad_norm": 0.5213791198423691, "learning_rate": 0.00020054999999999997, "loss": 0.2035, "step": 1337 }, { "epoch": 0.6296470588235294, "grad_norm": 0.5124694997510283, "learning_rate": 0.0002007, "loss": 0.1466, "step": 1338 }, { "epoch": 0.6301176470588236, "grad_norm": 0.7454657075483208, "learning_rate": 0.00020084999999999998, "loss": 0.1562, "step": 1339 }, { "epoch": 0.6305882352941177, "grad_norm": 0.41465973246207255, "learning_rate": 0.000201, "loss": 0.1179, "step": 1340 }, { "epoch": 0.6310588235294118, "grad_norm": 0.51732587224294, "learning_rate": 0.00020114999999999998, "loss": 0.1791, "step": 1341 }, { "epoch": 0.6315294117647059, "grad_norm": 0.43164112758032713, "learning_rate": 0.0002013, "loss": 0.1401, "step": 1342 }, { "epoch": 0.632, "grad_norm": 0.6589637954506469, "learning_rate": 0.00020145, "loss": 0.2396, "step": 1343 }, { "epoch": 0.6324705882352941, "grad_norm": 0.42320282440845314, "learning_rate": 0.0002016, "loss": 0.0501, "step": 1344 }, { "epoch": 0.6329411764705882, "grad_norm": 0.3870349229875479, "learning_rate": 0.00020174999999999997, "loss": 0.0913, "step": 1345 }, { "epoch": 0.6334117647058823, "grad_norm": 0.2460673515328249, "learning_rate": 0.0002019, "loss": 0.042, "step": 1346 }, { "epoch": 0.6338823529411765, "grad_norm": 0.4310642995265607, "learning_rate": 0.00020204999999999998, "loss": 0.1091, "step": 1347 }, { "epoch": 0.6343529411764706, "grad_norm": 0.3107581067571762, "learning_rate": 0.0002022, "loss": 0.0616, "step": 1348 }, { "epoch": 0.6348235294117647, "grad_norm": 0.5559510642837658, "learning_rate": 0.00020234999999999999, "loss": 0.1767, "step": 1349 }, { "epoch": 0.6352941176470588, "grad_norm": 0.6593409245514864, "learning_rate": 0.0002025, "loss": 0.261, "step": 1350 }, { "epoch": 0.6357647058823529, "grad_norm": 0.7030450674372832, "learning_rate": 0.00020264999999999997, "loss": 0.3024, "step": 1351 }, { "epoch": 0.636235294117647, "grad_norm": 1.8784465621050543, "learning_rate": 0.0002028, "loss": 0.0664, "step": 1352 }, { "epoch": 0.6367058823529411, "grad_norm": 0.646875439552133, "learning_rate": 0.00020294999999999997, "loss": 0.3626, "step": 1353 }, { "epoch": 0.6371764705882353, "grad_norm": 0.45002444050849594, "learning_rate": 0.0002031, "loss": 0.156, "step": 1354 }, { "epoch": 0.6376470588235295, "grad_norm": 0.5276794056497647, "learning_rate": 0.00020324999999999998, "loss": 0.2362, "step": 1355 }, { "epoch": 0.6381176470588236, "grad_norm": 0.6424653554971173, "learning_rate": 0.00020339999999999998, "loss": 0.1051, "step": 1356 }, { "epoch": 0.6385882352941177, "grad_norm": 0.6957311414837812, "learning_rate": 0.00020355, "loss": 0.2231, "step": 1357 }, { "epoch": 0.6390588235294118, "grad_norm": 0.6013203951158893, "learning_rate": 0.0002037, "loss": 0.1993, "step": 1358 }, { "epoch": 0.6395294117647059, "grad_norm": 0.45851025620709057, "learning_rate": 0.00020384999999999997, "loss": 0.1392, "step": 1359 }, { "epoch": 0.64, "grad_norm": 0.2477172987643188, "learning_rate": 0.000204, "loss": 0.0582, "step": 1360 }, { "epoch": 0.6404705882352941, "grad_norm": 1.067552580033563, "learning_rate": 0.00020414999999999997, "loss": 0.5445, "step": 1361 }, { "epoch": 0.6409411764705882, "grad_norm": 0.4202242443786375, "learning_rate": 0.0002043, "loss": 0.1411, "step": 1362 }, { "epoch": 0.6414117647058823, "grad_norm": 0.7745621922500252, "learning_rate": 0.00020444999999999998, "loss": 0.2788, "step": 1363 }, { "epoch": 0.6418823529411765, "grad_norm": 0.7020110912702237, "learning_rate": 0.00020459999999999999, "loss": 0.3275, "step": 1364 }, { "epoch": 0.6423529411764706, "grad_norm": 0.7716563378524686, "learning_rate": 0.00020475, "loss": 0.3906, "step": 1365 }, { "epoch": 0.6428235294117647, "grad_norm": 0.6192088360000066, "learning_rate": 0.0002049, "loss": 0.2307, "step": 1366 }, { "epoch": 0.6432941176470588, "grad_norm": 0.3532972811475707, "learning_rate": 0.00020504999999999997, "loss": 0.0795, "step": 1367 }, { "epoch": 0.6437647058823529, "grad_norm": 0.4962637511289129, "learning_rate": 0.0002052, "loss": 0.259, "step": 1368 }, { "epoch": 0.644235294117647, "grad_norm": 0.9907254059875578, "learning_rate": 0.00020534999999999998, "loss": 0.4119, "step": 1369 }, { "epoch": 0.6447058823529411, "grad_norm": 0.5485822444077534, "learning_rate": 0.0002055, "loss": 0.2535, "step": 1370 }, { "epoch": 0.6451764705882352, "grad_norm": 0.3340148944457657, "learning_rate": 0.00020564999999999998, "loss": 0.1044, "step": 1371 }, { "epoch": 0.6456470588235295, "grad_norm": 0.4826338349069654, "learning_rate": 0.0002058, "loss": 0.1222, "step": 1372 }, { "epoch": 0.6461176470588236, "grad_norm": 0.6485449172029237, "learning_rate": 0.00020595, "loss": 0.2144, "step": 1373 }, { "epoch": 0.6465882352941177, "grad_norm": 1.0417631054400571, "learning_rate": 0.0002061, "loss": 0.177, "step": 1374 }, { "epoch": 0.6470588235294118, "grad_norm": 0.38610610585138283, "learning_rate": 0.00020624999999999997, "loss": 0.0917, "step": 1375 }, { "epoch": 0.6475294117647059, "grad_norm": 1.4267625394281767, "learning_rate": 0.00020639999999999998, "loss": 0.3293, "step": 1376 }, { "epoch": 0.648, "grad_norm": 0.7765704413071743, "learning_rate": 0.00020654999999999998, "loss": 0.3815, "step": 1377 }, { "epoch": 0.6484705882352941, "grad_norm": 0.28555463080463817, "learning_rate": 0.00020669999999999996, "loss": 0.0871, "step": 1378 }, { "epoch": 0.6489411764705882, "grad_norm": 0.33234555972454444, "learning_rate": 0.00020684999999999999, "loss": 0.0816, "step": 1379 }, { "epoch": 0.6494117647058824, "grad_norm": 0.976698335839472, "learning_rate": 0.00020699999999999996, "loss": 0.2531, "step": 1380 }, { "epoch": 0.6498823529411765, "grad_norm": 0.5321875694458712, "learning_rate": 0.00020715, "loss": 0.1839, "step": 1381 }, { "epoch": 0.6503529411764706, "grad_norm": 0.3645094102805714, "learning_rate": 0.00020729999999999997, "loss": 0.0671, "step": 1382 }, { "epoch": 0.6508235294117647, "grad_norm": 0.4171744685079755, "learning_rate": 0.00020744999999999997, "loss": 0.0899, "step": 1383 }, { "epoch": 0.6512941176470588, "grad_norm": 0.5148390402062887, "learning_rate": 0.00020759999999999998, "loss": 0.1759, "step": 1384 }, { "epoch": 0.6517647058823529, "grad_norm": 0.37715645738709247, "learning_rate": 0.00020774999999999998, "loss": 0.048, "step": 1385 }, { "epoch": 0.652235294117647, "grad_norm": 1.3515382522620167, "learning_rate": 0.00020789999999999996, "loss": 0.2743, "step": 1386 }, { "epoch": 0.6527058823529411, "grad_norm": 0.344266152006046, "learning_rate": 0.00020805, "loss": 0.0386, "step": 1387 }, { "epoch": 0.6531764705882352, "grad_norm": 1.1336750033076275, "learning_rate": 0.00020819999999999996, "loss": 0.3906, "step": 1388 }, { "epoch": 0.6536470588235294, "grad_norm": 0.7122512052429899, "learning_rate": 0.00020835, "loss": 0.2856, "step": 1389 }, { "epoch": 0.6541176470588236, "grad_norm": 0.361236283773461, "learning_rate": 0.00020849999999999997, "loss": 0.1145, "step": 1390 }, { "epoch": 0.6545882352941177, "grad_norm": 1.2097376725621283, "learning_rate": 0.00020864999999999998, "loss": 0.3079, "step": 1391 }, { "epoch": 0.6550588235294118, "grad_norm": 0.48613963469562643, "learning_rate": 0.00020879999999999998, "loss": 0.197, "step": 1392 }, { "epoch": 0.6555294117647059, "grad_norm": 0.15554520178932765, "learning_rate": 0.00020894999999999998, "loss": 0.0201, "step": 1393 }, { "epoch": 0.656, "grad_norm": 0.6776444323930052, "learning_rate": 0.00020909999999999996, "loss": 0.1937, "step": 1394 }, { "epoch": 0.6564705882352941, "grad_norm": 0.7339880761107306, "learning_rate": 0.00020925, "loss": 0.2617, "step": 1395 }, { "epoch": 0.6569411764705883, "grad_norm": 1.1519581100478733, "learning_rate": 0.00020939999999999997, "loss": 0.4827, "step": 1396 }, { "epoch": 0.6574117647058824, "grad_norm": 0.6618584422631851, "learning_rate": 0.00020955, "loss": 0.2866, "step": 1397 }, { "epoch": 0.6578823529411765, "grad_norm": 0.757857046432025, "learning_rate": 0.00020969999999999997, "loss": 0.2869, "step": 1398 }, { "epoch": 0.6583529411764706, "grad_norm": 0.34239427582750337, "learning_rate": 0.00020984999999999998, "loss": 0.0174, "step": 1399 }, { "epoch": 0.6588235294117647, "grad_norm": 0.5354047148975941, "learning_rate": 0.00020999999999999998, "loss": 0.1755, "step": 1400 }, { "epoch": 0.6592941176470588, "grad_norm": 0.5073760687568103, "learning_rate": 0.00021014999999999999, "loss": 0.2578, "step": 1401 }, { "epoch": 0.6597647058823529, "grad_norm": 0.7671612812150604, "learning_rate": 0.00021029999999999996, "loss": 0.306, "step": 1402 }, { "epoch": 0.660235294117647, "grad_norm": 0.26218746641839064, "learning_rate": 0.00021045, "loss": 0.0391, "step": 1403 }, { "epoch": 0.6607058823529411, "grad_norm": 0.9602204451504907, "learning_rate": 0.00021059999999999997, "loss": 0.2275, "step": 1404 }, { "epoch": 0.6611764705882353, "grad_norm": 0.8683039614777716, "learning_rate": 0.00021074999999999997, "loss": 0.3259, "step": 1405 }, { "epoch": 0.6616470588235294, "grad_norm": 0.7699642803205914, "learning_rate": 0.00021089999999999998, "loss": 0.3098, "step": 1406 }, { "epoch": 0.6621176470588235, "grad_norm": 0.38072560245736625, "learning_rate": 0.00021104999999999998, "loss": 0.1465, "step": 1407 }, { "epoch": 0.6625882352941177, "grad_norm": 0.7867176496311262, "learning_rate": 0.00021119999999999996, "loss": 0.2392, "step": 1408 }, { "epoch": 0.6630588235294118, "grad_norm": 0.15279232582466853, "learning_rate": 0.00021135, "loss": 0.022, "step": 1409 }, { "epoch": 0.6635294117647059, "grad_norm": 0.7867913473453798, "learning_rate": 0.00021149999999999996, "loss": 0.3976, "step": 1410 }, { "epoch": 0.664, "grad_norm": 0.45343008270373464, "learning_rate": 0.00021165, "loss": 0.1629, "step": 1411 }, { "epoch": 0.6644705882352941, "grad_norm": 0.32830365777182247, "learning_rate": 0.00021179999999999997, "loss": 0.0815, "step": 1412 }, { "epoch": 0.6649411764705883, "grad_norm": 0.5633372058845785, "learning_rate": 0.00021194999999999997, "loss": 0.2407, "step": 1413 }, { "epoch": 0.6654117647058824, "grad_norm": 0.7136721393416304, "learning_rate": 0.00021209999999999998, "loss": 0.3614, "step": 1414 }, { "epoch": 0.6658823529411765, "grad_norm": 0.6426361526754136, "learning_rate": 0.00021224999999999998, "loss": 0.2508, "step": 1415 }, { "epoch": 0.6663529411764706, "grad_norm": 0.20559881531530166, "learning_rate": 0.00021239999999999996, "loss": 0.0179, "step": 1416 }, { "epoch": 0.6668235294117647, "grad_norm": 0.5518525950043764, "learning_rate": 0.00021255, "loss": 0.1437, "step": 1417 }, { "epoch": 0.6672941176470588, "grad_norm": 0.7394452249956519, "learning_rate": 0.00021269999999999997, "loss": 0.2986, "step": 1418 }, { "epoch": 0.6677647058823529, "grad_norm": 0.4601174544917248, "learning_rate": 0.00021285, "loss": 0.0928, "step": 1419 }, { "epoch": 0.668235294117647, "grad_norm": 0.3682806261133285, "learning_rate": 0.00021299999999999997, "loss": 0.138, "step": 1420 }, { "epoch": 0.6687058823529411, "grad_norm": 0.7336952589257433, "learning_rate": 0.00021314999999999998, "loss": 0.2313, "step": 1421 }, { "epoch": 0.6691764705882353, "grad_norm": 0.8516968073759018, "learning_rate": 0.00021329999999999998, "loss": 0.4437, "step": 1422 }, { "epoch": 0.6696470588235294, "grad_norm": 0.44691374743105117, "learning_rate": 0.00021344999999999998, "loss": 0.161, "step": 1423 }, { "epoch": 0.6701176470588235, "grad_norm": 0.7340470242371181, "learning_rate": 0.00021359999999999996, "loss": 0.3961, "step": 1424 }, { "epoch": 0.6705882352941176, "grad_norm": 0.5286321594154684, "learning_rate": 0.00021375, "loss": 0.2855, "step": 1425 }, { "epoch": 0.6710588235294118, "grad_norm": 0.6761222178202746, "learning_rate": 0.00021389999999999997, "loss": 0.2841, "step": 1426 }, { "epoch": 0.6715294117647059, "grad_norm": 0.9449800093834893, "learning_rate": 0.00021405, "loss": 0.3457, "step": 1427 }, { "epoch": 0.672, "grad_norm": 1.2679426227218962, "learning_rate": 0.00021419999999999998, "loss": 0.4922, "step": 1428 }, { "epoch": 0.6724705882352942, "grad_norm": 0.8610276684881861, "learning_rate": 0.00021434999999999998, "loss": 0.357, "step": 1429 }, { "epoch": 0.6729411764705883, "grad_norm": 0.7019595642144557, "learning_rate": 0.00021449999999999998, "loss": 0.2332, "step": 1430 }, { "epoch": 0.6734117647058824, "grad_norm": 0.5567886287876513, "learning_rate": 0.00021464999999999999, "loss": 0.0969, "step": 1431 }, { "epoch": 0.6738823529411765, "grad_norm": 0.8648546303135177, "learning_rate": 0.00021479999999999996, "loss": 0.3631, "step": 1432 }, { "epoch": 0.6743529411764706, "grad_norm": 0.4999057364176092, "learning_rate": 0.00021495, "loss": 0.2446, "step": 1433 }, { "epoch": 0.6748235294117647, "grad_norm": 0.47288419819545235, "learning_rate": 0.00021509999999999997, "loss": 0.13, "step": 1434 }, { "epoch": 0.6752941176470588, "grad_norm": 0.653632920455304, "learning_rate": 0.00021525, "loss": 0.2276, "step": 1435 }, { "epoch": 0.6757647058823529, "grad_norm": 1.1535050792906907, "learning_rate": 0.00021539999999999998, "loss": 0.4251, "step": 1436 }, { "epoch": 0.676235294117647, "grad_norm": 1.057240097771952, "learning_rate": 0.00021554999999999998, "loss": 0.4281, "step": 1437 }, { "epoch": 0.6767058823529412, "grad_norm": 0.1274392826971882, "learning_rate": 0.00021569999999999998, "loss": 0.0143, "step": 1438 }, { "epoch": 0.6771764705882353, "grad_norm": 1.1883823809894536, "learning_rate": 0.00021585, "loss": 0.2631, "step": 1439 }, { "epoch": 0.6776470588235294, "grad_norm": 1.0029739827076818, "learning_rate": 0.00021599999999999996, "loss": 0.2254, "step": 1440 }, { "epoch": 0.6781176470588235, "grad_norm": 1.1918522042885122, "learning_rate": 0.00021615, "loss": 0.2997, "step": 1441 }, { "epoch": 0.6785882352941176, "grad_norm": 0.9124761528163419, "learning_rate": 0.00021629999999999997, "loss": 0.3051, "step": 1442 }, { "epoch": 0.6790588235294117, "grad_norm": 1.2025465942750277, "learning_rate": 0.00021645, "loss": 0.4299, "step": 1443 }, { "epoch": 0.6795294117647059, "grad_norm": 0.39915223834074404, "learning_rate": 0.00021659999999999998, "loss": 0.1077, "step": 1444 }, { "epoch": 0.68, "grad_norm": 48.76537588329299, "learning_rate": 0.00021674999999999998, "loss": 0.8279, "step": 1445 }, { "epoch": 0.6804705882352942, "grad_norm": 0.7571498138087254, "learning_rate": 0.0002169, "loss": 0.2028, "step": 1446 }, { "epoch": 0.6809411764705883, "grad_norm": 0.8272523337280038, "learning_rate": 0.00021705, "loss": 0.0937, "step": 1447 }, { "epoch": 0.6814117647058824, "grad_norm": 2.3306830358778146, "learning_rate": 0.00021719999999999997, "loss": 0.2672, "step": 1448 }, { "epoch": 0.6818823529411765, "grad_norm": 0.9052688645493937, "learning_rate": 0.00021735, "loss": 0.2356, "step": 1449 }, { "epoch": 0.6823529411764706, "grad_norm": 0.7871861025288657, "learning_rate": 0.00021749999999999997, "loss": 0.1631, "step": 1450 }, { "epoch": 0.6828235294117647, "grad_norm": 1.4631466803623738, "learning_rate": 0.00021764999999999998, "loss": 0.286, "step": 1451 }, { "epoch": 0.6832941176470588, "grad_norm": 2.2467372023451597, "learning_rate": 0.00021779999999999998, "loss": 0.1791, "step": 1452 }, { "epoch": 0.6837647058823529, "grad_norm": 0.4561378297929561, "learning_rate": 0.00021794999999999999, "loss": 0.0993, "step": 1453 }, { "epoch": 0.684235294117647, "grad_norm": 1.186419808813212, "learning_rate": 0.00021809999999999996, "loss": 0.342, "step": 1454 }, { "epoch": 0.6847058823529412, "grad_norm": 0.9005239398794598, "learning_rate": 0.00021825, "loss": 0.3796, "step": 1455 }, { "epoch": 0.6851764705882353, "grad_norm": 1.134496731987878, "learning_rate": 0.00021839999999999997, "loss": 0.4881, "step": 1456 }, { "epoch": 0.6856470588235294, "grad_norm": 1.1674061249473981, "learning_rate": 0.00021855, "loss": 0.3399, "step": 1457 }, { "epoch": 0.6861176470588235, "grad_norm": 0.23025141035939806, "learning_rate": 0.00021869999999999998, "loss": 0.0429, "step": 1458 }, { "epoch": 0.6865882352941176, "grad_norm": 0.3102421771863425, "learning_rate": 0.00021884999999999998, "loss": 0.0643, "step": 1459 }, { "epoch": 0.6870588235294117, "grad_norm": 17.130306226131047, "learning_rate": 0.00021899999999999998, "loss": 0.0949, "step": 1460 }, { "epoch": 0.6875294117647058, "grad_norm": 0.559568979965395, "learning_rate": 0.00021915, "loss": 0.1341, "step": 1461 }, { "epoch": 0.688, "grad_norm": 0.9607707565808109, "learning_rate": 0.00021929999999999996, "loss": 0.1125, "step": 1462 }, { "epoch": 0.6884705882352942, "grad_norm": 1.882280996647234, "learning_rate": 0.00021945, "loss": 0.5721, "step": 1463 }, { "epoch": 0.6889411764705883, "grad_norm": 1.1265986631140754, "learning_rate": 0.00021959999999999997, "loss": 0.3057, "step": 1464 }, { "epoch": 0.6894117647058824, "grad_norm": 2.0705939307358423, "learning_rate": 0.00021975, "loss": 0.6466, "step": 1465 }, { "epoch": 0.6898823529411765, "grad_norm": 6.850114606597415, "learning_rate": 0.00021989999999999998, "loss": 1.0383, "step": 1466 }, { "epoch": 0.6903529411764706, "grad_norm": 7.156738185712797, "learning_rate": 0.00022004999999999998, "loss": 0.2915, "step": 1467 }, { "epoch": 0.6908235294117647, "grad_norm": 6.924770548127318, "learning_rate": 0.00022019999999999999, "loss": 1.0696, "step": 1468 }, { "epoch": 0.6912941176470588, "grad_norm": 16.985688236448453, "learning_rate": 0.00022035, "loss": 1.7588, "step": 1469 }, { "epoch": 0.691764705882353, "grad_norm": 15.486361468308328, "learning_rate": 0.00022049999999999997, "loss": 1.7348, "step": 1470 }, { "epoch": 0.6922352941176471, "grad_norm": 13.951244004326188, "learning_rate": 0.00022065, "loss": 1.8897, "step": 1471 }, { "epoch": 0.6927058823529412, "grad_norm": 16.711234057196243, "learning_rate": 0.00022079999999999997, "loss": 1.4215, "step": 1472 }, { "epoch": 0.6931764705882353, "grad_norm": 16.4390267893176, "learning_rate": 0.00022095, "loss": 1.6118, "step": 1473 }, { "epoch": 0.6936470588235294, "grad_norm": 13.386393601495755, "learning_rate": 0.00022109999999999998, "loss": 1.6181, "step": 1474 }, { "epoch": 0.6941176470588235, "grad_norm": 7.041985135288613, "learning_rate": 0.00022124999999999998, "loss": 2.0307, "step": 1475 }, { "epoch": 0.6945882352941176, "grad_norm": 5.063764644849268, "learning_rate": 0.0002214, "loss": 1.5259, "step": 1476 }, { "epoch": 0.6950588235294117, "grad_norm": 8.360891253690262, "learning_rate": 0.00022155, "loss": 1.3296, "step": 1477 }, { "epoch": 0.6955294117647058, "grad_norm": 20.531962939858378, "learning_rate": 0.00022169999999999997, "loss": 1.1509, "step": 1478 }, { "epoch": 0.696, "grad_norm": 3.5583680948560494, "learning_rate": 0.00022185, "loss": 1.2002, "step": 1479 }, { "epoch": 0.6964705882352941, "grad_norm": 8.570454047296295, "learning_rate": 0.00022199999999999998, "loss": 1.3008, "step": 1480 }, { "epoch": 0.6969411764705883, "grad_norm": 12.693873396439438, "learning_rate": 0.00022215, "loss": 1.3391, "step": 1481 }, { "epoch": 0.6974117647058824, "grad_norm": 19.665014957324093, "learning_rate": 0.00022229999999999998, "loss": 0.3976, "step": 1482 }, { "epoch": 0.6978823529411765, "grad_norm": 11.50611266323349, "learning_rate": 0.00022244999999999999, "loss": 0.5275, "step": 1483 }, { "epoch": 0.6983529411764706, "grad_norm": 9.569532808332918, "learning_rate": 0.0002226, "loss": 1.1019, "step": 1484 }, { "epoch": 0.6988235294117647, "grad_norm": 36.03406750628513, "learning_rate": 0.00022275, "loss": 2.1909, "step": 1485 }, { "epoch": 0.6992941176470588, "grad_norm": 4.542806709699389, "learning_rate": 0.00022289999999999997, "loss": 0.9344, "step": 1486 }, { "epoch": 0.699764705882353, "grad_norm": 10.591744064353364, "learning_rate": 0.00022305, "loss": 0.5176, "step": 1487 }, { "epoch": 0.7002352941176471, "grad_norm": 18.220337567584092, "learning_rate": 0.00022319999999999998, "loss": 0.5365, "step": 1488 }, { "epoch": 0.7007058823529412, "grad_norm": 4.820788591059744, "learning_rate": 0.00022335, "loss": 1.0853, "step": 1489 }, { "epoch": 0.7011764705882353, "grad_norm": 10.865114063948258, "learning_rate": 0.00022349999999999998, "loss": 1.1192, "step": 1490 }, { "epoch": 0.7016470588235294, "grad_norm": 1.290886555139775, "learning_rate": 0.00022365, "loss": 0.7752, "step": 1491 }, { "epoch": 0.7021176470588235, "grad_norm": 6.168038555574171, "learning_rate": 0.0002238, "loss": 0.6535, "step": 1492 }, { "epoch": 0.7025882352941176, "grad_norm": 4.18811440468323, "learning_rate": 0.00022395, "loss": 0.5781, "step": 1493 }, { "epoch": 0.7030588235294117, "grad_norm": 1.81868285880162, "learning_rate": 0.00022409999999999997, "loss": 0.8761, "step": 1494 }, { "epoch": 0.7035294117647058, "grad_norm": 6.123359581256626, "learning_rate": 0.00022425, "loss": 0.4918, "step": 1495 }, { "epoch": 0.704, "grad_norm": 26.985675695455033, "learning_rate": 0.00022439999999999998, "loss": 1.7819, "step": 1496 }, { "epoch": 0.7044705882352941, "grad_norm": 5.962031667836617, "learning_rate": 0.00022455, "loss": 0.6389, "step": 1497 }, { "epoch": 0.7049411764705882, "grad_norm": 7.774331836860654, "learning_rate": 0.0002247, "loss": 0.6636, "step": 1498 }, { "epoch": 0.7054117647058824, "grad_norm": 19.891407535369922, "learning_rate": 0.00022485, "loss": 1.5813, "step": 1499 }, { "epoch": 0.7058823529411765, "grad_norm": 23.178073438104494, "learning_rate": 0.000225, "loss": 1.9363, "step": 1500 }, { "epoch": 0.7063529411764706, "grad_norm": 2.695597247001396, "learning_rate": 0.00022514999999999997, "loss": 1.4395, "step": 1501 }, { "epoch": 0.7068235294117647, "grad_norm": 22.02146206038378, "learning_rate": 0.00022529999999999997, "loss": 1.1523, "step": 1502 }, { "epoch": 0.7072941176470589, "grad_norm": 36.45432198903929, "learning_rate": 0.00022544999999999995, "loss": 1.2183, "step": 1503 }, { "epoch": 0.707764705882353, "grad_norm": 3.628525262103395, "learning_rate": 0.00022559999999999998, "loss": 1.6129, "step": 1504 }, { "epoch": 0.7082352941176471, "grad_norm": 23.738796887329496, "learning_rate": 0.00022574999999999996, "loss": 0.7789, "step": 1505 }, { "epoch": 0.7087058823529412, "grad_norm": 1.2895753793784677, "learning_rate": 0.0002259, "loss": 0.9777, "step": 1506 }, { "epoch": 0.7091764705882353, "grad_norm": 11.055067082516022, "learning_rate": 0.00022604999999999997, "loss": 0.2565, "step": 1507 }, { "epoch": 0.7096470588235294, "grad_norm": 2.574603613795424, "learning_rate": 0.00022619999999999997, "loss": 0.6142, "step": 1508 }, { "epoch": 0.7101176470588235, "grad_norm": 2.03623619320144, "learning_rate": 0.00022634999999999997, "loss": 0.2703, "step": 1509 }, { "epoch": 0.7105882352941176, "grad_norm": 13.980061715214616, "learning_rate": 0.00022649999999999998, "loss": 1.0517, "step": 1510 }, { "epoch": 0.7110588235294117, "grad_norm": 40.66162472181378, "learning_rate": 0.00022664999999999995, "loss": 2.3621, "step": 1511 }, { "epoch": 0.7115294117647059, "grad_norm": 13.693361909866674, "learning_rate": 0.00022679999999999998, "loss": 0.9805, "step": 1512 }, { "epoch": 0.712, "grad_norm": 11.348247631098126, "learning_rate": 0.00022694999999999996, "loss": 1.2058, "step": 1513 }, { "epoch": 0.7124705882352941, "grad_norm": 25.93832408525648, "learning_rate": 0.0002271, "loss": 1.5966, "step": 1514 }, { "epoch": 0.7129411764705882, "grad_norm": 3.566647674561617, "learning_rate": 0.00022724999999999997, "loss": 0.6417, "step": 1515 }, { "epoch": 0.7134117647058823, "grad_norm": 22.29191997617386, "learning_rate": 0.00022739999999999997, "loss": 0.3933, "step": 1516 }, { "epoch": 0.7138823529411765, "grad_norm": 4.799334249702979, "learning_rate": 0.00022754999999999997, "loss": 1.102, "step": 1517 }, { "epoch": 0.7143529411764706, "grad_norm": 21.32878624215212, "learning_rate": 0.00022769999999999998, "loss": 1.1359, "step": 1518 }, { "epoch": 0.7148235294117647, "grad_norm": 11.095466750220051, "learning_rate": 0.00022784999999999995, "loss": 1.3079, "step": 1519 }, { "epoch": 0.7152941176470589, "grad_norm": 16.382366826422228, "learning_rate": 0.00022799999999999999, "loss": 0.7141, "step": 1520 }, { "epoch": 0.715764705882353, "grad_norm": 10.718285179090353, "learning_rate": 0.00022814999999999996, "loss": 0.7745, "step": 1521 }, { "epoch": 0.7162352941176471, "grad_norm": 19.070789162600498, "learning_rate": 0.0002283, "loss": 1.5775, "step": 1522 }, { "epoch": 0.7167058823529412, "grad_norm": 20.345451560306934, "learning_rate": 0.00022844999999999997, "loss": 1.4579, "step": 1523 }, { "epoch": 0.7171764705882353, "grad_norm": 16.083034481300917, "learning_rate": 0.00022859999999999997, "loss": 1.4357, "step": 1524 }, { "epoch": 0.7176470588235294, "grad_norm": 15.3269955992204, "learning_rate": 0.00022874999999999998, "loss": 1.3978, "step": 1525 }, { "epoch": 0.7181176470588235, "grad_norm": 2.2145794535830348, "learning_rate": 0.00022889999999999998, "loss": 1.2747, "step": 1526 }, { "epoch": 0.7185882352941176, "grad_norm": 13.883952441636147, "learning_rate": 0.00022904999999999996, "loss": 0.8486, "step": 1527 }, { "epoch": 0.7190588235294118, "grad_norm": 25.30261993302244, "learning_rate": 0.0002292, "loss": 0.7435, "step": 1528 }, { "epoch": 0.7195294117647059, "grad_norm": 27.78183122230098, "learning_rate": 0.00022934999999999996, "loss": 0.6077, "step": 1529 }, { "epoch": 0.72, "grad_norm": 13.676161367803006, "learning_rate": 0.0002295, "loss": 0.7055, "step": 1530 }, { "epoch": 0.7204705882352941, "grad_norm": 13.677675263885707, "learning_rate": 0.00022964999999999997, "loss": 1.2773, "step": 1531 }, { "epoch": 0.7209411764705882, "grad_norm": 1.3446629805334753, "learning_rate": 0.00022979999999999997, "loss": 0.726, "step": 1532 }, { "epoch": 0.7214117647058823, "grad_norm": 21.813695851372405, "learning_rate": 0.00022994999999999998, "loss": 1.4794, "step": 1533 }, { "epoch": 0.7218823529411764, "grad_norm": 5.976259947979106, "learning_rate": 0.00023009999999999998, "loss": 0.8205, "step": 1534 }, { "epoch": 0.7223529411764706, "grad_norm": 15.015837357017498, "learning_rate": 0.00023024999999999996, "loss": 1.1931, "step": 1535 }, { "epoch": 0.7228235294117648, "grad_norm": 21.822190312131713, "learning_rate": 0.0002304, "loss": 1.5062, "step": 1536 }, { "epoch": 0.7232941176470589, "grad_norm": 20.476365307315277, "learning_rate": 0.00023054999999999997, "loss": 1.5266, "step": 1537 }, { "epoch": 0.723764705882353, "grad_norm": 5.823726949836751, "learning_rate": 0.0002307, "loss": 1.2527, "step": 1538 }, { "epoch": 0.7242352941176471, "grad_norm": 34.26640447353048, "learning_rate": 0.00023084999999999997, "loss": 0.505, "step": 1539 }, { "epoch": 0.7247058823529412, "grad_norm": 14.399124806012445, "learning_rate": 0.00023099999999999998, "loss": 1.2115, "step": 1540 }, { "epoch": 0.7251764705882353, "grad_norm": 44.45701460529234, "learning_rate": 0.00023114999999999998, "loss": 0.536, "step": 1541 }, { "epoch": 0.7256470588235294, "grad_norm": 27.671454513875613, "learning_rate": 0.00023129999999999998, "loss": 0.8129, "step": 1542 }, { "epoch": 0.7261176470588235, "grad_norm": 2.902907813734804, "learning_rate": 0.00023144999999999996, "loss": 1.0291, "step": 1543 }, { "epoch": 0.7265882352941176, "grad_norm": 15.650414239362206, "learning_rate": 0.0002316, "loss": 0.247, "step": 1544 }, { "epoch": 0.7270588235294118, "grad_norm": 26.57027760685872, "learning_rate": 0.00023174999999999997, "loss": 1.6191, "step": 1545 }, { "epoch": 0.7275294117647059, "grad_norm": 2.1693016804925724, "learning_rate": 0.0002319, "loss": 0.6715, "step": 1546 }, { "epoch": 0.728, "grad_norm": 6.503458572445864, "learning_rate": 0.00023204999999999998, "loss": 0.7129, "step": 1547 }, { "epoch": 0.7284705882352941, "grad_norm": 1.7087425793799058, "learning_rate": 0.00023219999999999998, "loss": 0.2181, "step": 1548 }, { "epoch": 0.7289411764705882, "grad_norm": 5.458021422902418, "learning_rate": 0.00023234999999999998, "loss": 0.5388, "step": 1549 }, { "epoch": 0.7294117647058823, "grad_norm": 27.859715144066122, "learning_rate": 0.00023249999999999999, "loss": 1.6616, "step": 1550 }, { "epoch": 0.7298823529411764, "grad_norm": 44.32051653609039, "learning_rate": 0.00023264999999999996, "loss": 2.5661, "step": 1551 }, { "epoch": 0.7303529411764705, "grad_norm": 47.676309429079595, "learning_rate": 0.0002328, "loss": 2.718, "step": 1552 }, { "epoch": 0.7308235294117648, "grad_norm": 15.714952115293464, "learning_rate": 0.00023294999999999997, "loss": 1.2245, "step": 1553 }, { "epoch": 0.7312941176470589, "grad_norm": 5.29208778770597, "learning_rate": 0.00023309999999999997, "loss": 0.9588, "step": 1554 }, { "epoch": 0.731764705882353, "grad_norm": 21.858150111579373, "learning_rate": 0.00023324999999999998, "loss": 1.3638, "step": 1555 }, { "epoch": 0.7322352941176471, "grad_norm": 43.295876349522956, "learning_rate": 0.00023339999999999998, "loss": 1.0867, "step": 1556 }, { "epoch": 0.7327058823529412, "grad_norm": 57.53998810943559, "learning_rate": 0.00023354999999999996, "loss": 1.1884, "step": 1557 }, { "epoch": 0.7331764705882353, "grad_norm": 17.03220216540181, "learning_rate": 0.0002337, "loss": 1.3005, "step": 1558 }, { "epoch": 0.7336470588235294, "grad_norm": 30.130923441320668, "learning_rate": 0.00023384999999999997, "loss": 0.8128, "step": 1559 }, { "epoch": 0.7341176470588235, "grad_norm": 26.3002200035577, "learning_rate": 0.000234, "loss": 0.3898, "step": 1560 }, { "epoch": 0.7345882352941177, "grad_norm": 11.892177180840442, "learning_rate": 0.00023414999999999997, "loss": 0.2993, "step": 1561 }, { "epoch": 0.7350588235294118, "grad_norm": 1.981683658664869, "learning_rate": 0.00023429999999999998, "loss": 0.4874, "step": 1562 }, { "epoch": 0.7355294117647059, "grad_norm": 10.311268370690417, "learning_rate": 0.00023444999999999998, "loss": 0.8333, "step": 1563 }, { "epoch": 0.736, "grad_norm": 31.444432434500268, "learning_rate": 0.00023459999999999998, "loss": 1.8724, "step": 1564 }, { "epoch": 0.7364705882352941, "grad_norm": 7.650705831599901, "learning_rate": 0.00023474999999999996, "loss": 0.5877, "step": 1565 }, { "epoch": 0.7369411764705882, "grad_norm": 25.30668941541228, "learning_rate": 0.0002349, "loss": 1.5734, "step": 1566 }, { "epoch": 0.7374117647058823, "grad_norm": 21.464468640854697, "learning_rate": 0.00023504999999999997, "loss": 1.319, "step": 1567 }, { "epoch": 0.7378823529411764, "grad_norm": 17.2982172058739, "learning_rate": 0.0002352, "loss": 1.1695, "step": 1568 }, { "epoch": 0.7383529411764705, "grad_norm": 3.2315753671267644, "learning_rate": 0.00023534999999999997, "loss": 0.4708, "step": 1569 }, { "epoch": 0.7388235294117647, "grad_norm": 30.22575825823417, "learning_rate": 0.00023549999999999998, "loss": 1.7922, "step": 1570 }, { "epoch": 0.7392941176470588, "grad_norm": 19.444949082564676, "learning_rate": 0.00023564999999999998, "loss": 1.4037, "step": 1571 }, { "epoch": 0.739764705882353, "grad_norm": 14.57379626216806, "learning_rate": 0.00023579999999999999, "loss": 0.5796, "step": 1572 }, { "epoch": 0.7402352941176471, "grad_norm": 22.641206822628888, "learning_rate": 0.00023594999999999996, "loss": 2.193, "step": 1573 }, { "epoch": 0.7407058823529412, "grad_norm": 16.639240042829083, "learning_rate": 0.0002361, "loss": 1.2673, "step": 1574 }, { "epoch": 0.7411764705882353, "grad_norm": 18.862633975971416, "learning_rate": 0.00023624999999999997, "loss": 1.6682, "step": 1575 }, { "epoch": 0.7416470588235294, "grad_norm": 21.213835016580195, "learning_rate": 0.0002364, "loss": 1.4556, "step": 1576 }, { "epoch": 0.7421176470588235, "grad_norm": 63.51897453061666, "learning_rate": 0.00023654999999999998, "loss": 0.9089, "step": 1577 }, { "epoch": 0.7425882352941177, "grad_norm": 29.971265576394156, "learning_rate": 0.00023669999999999998, "loss": 1.1469, "step": 1578 }, { "epoch": 0.7430588235294118, "grad_norm": 25.83467947712038, "learning_rate": 0.00023684999999999998, "loss": 2.6984, "step": 1579 }, { "epoch": 0.7435294117647059, "grad_norm": 11.509561464109936, "learning_rate": 0.000237, "loss": 0.5692, "step": 1580 }, { "epoch": 0.744, "grad_norm": 14.02050917857935, "learning_rate": 0.00023714999999999996, "loss": 0.1881, "step": 1581 }, { "epoch": 0.7444705882352941, "grad_norm": 15.06896725203845, "learning_rate": 0.0002373, "loss": 1.1934, "step": 1582 }, { "epoch": 0.7449411764705882, "grad_norm": 38.28526194301361, "learning_rate": 0.00023744999999999997, "loss": 2.2053, "step": 1583 }, { "epoch": 0.7454117647058823, "grad_norm": 6.476263234829788, "learning_rate": 0.0002376, "loss": 0.7863, "step": 1584 }, { "epoch": 0.7458823529411764, "grad_norm": 45.28764136843151, "learning_rate": 0.00023774999999999998, "loss": 2.5861, "step": 1585 }, { "epoch": 0.7463529411764706, "grad_norm": 20.041028192943312, "learning_rate": 0.00023789999999999998, "loss": 1.4289, "step": 1586 }, { "epoch": 0.7468235294117647, "grad_norm": 16.11077662574239, "learning_rate": 0.00023804999999999999, "loss": 1.324, "step": 1587 }, { "epoch": 0.7472941176470588, "grad_norm": 6.367958403936993, "learning_rate": 0.0002382, "loss": 0.7896, "step": 1588 }, { "epoch": 0.7477647058823529, "grad_norm": 25.81711553702177, "learning_rate": 0.00023834999999999997, "loss": 0.3772, "step": 1589 }, { "epoch": 0.7482352941176471, "grad_norm": 24.78933995540634, "learning_rate": 0.0002385, "loss": 0.9476, "step": 1590 }, { "epoch": 0.7487058823529412, "grad_norm": 4.095947850604833, "learning_rate": 0.00023864999999999997, "loss": 1.3833, "step": 1591 }, { "epoch": 0.7491764705882353, "grad_norm": 1.7142684459468638, "learning_rate": 0.0002388, "loss": 1.6719, "step": 1592 }, { "epoch": 0.7496470588235294, "grad_norm": 8.501531925602523, "learning_rate": 0.00023894999999999998, "loss": 1.0397, "step": 1593 }, { "epoch": 0.7501176470588236, "grad_norm": 21.923928342590543, "learning_rate": 0.00023909999999999998, "loss": 0.5421, "step": 1594 }, { "epoch": 0.7505882352941177, "grad_norm": 2.713285819876659, "learning_rate": 0.00023925, "loss": 1.3399, "step": 1595 }, { "epoch": 0.7510588235294118, "grad_norm": 11.021859598388358, "learning_rate": 0.0002394, "loss": 0.9231, "step": 1596 }, { "epoch": 0.7515294117647059, "grad_norm": 3.7560812429604793, "learning_rate": 0.00023954999999999997, "loss": 0.8684, "step": 1597 }, { "epoch": 0.752, "grad_norm": 16.06241171103871, "learning_rate": 0.0002397, "loss": 1.5029, "step": 1598 }, { "epoch": 0.7524705882352941, "grad_norm": 2.3236012139063367, "learning_rate": 0.00023984999999999998, "loss": 0.9671, "step": 1599 }, { "epoch": 0.7529411764705882, "grad_norm": 2.786355782664073, "learning_rate": 0.00023999999999999998, "loss": 0.9056, "step": 1600 }, { "epoch": 0.7534117647058823, "grad_norm": 31.398500639996833, "learning_rate": 0.00024014999999999998, "loss": 2.0365, "step": 1601 }, { "epoch": 0.7538823529411764, "grad_norm": 22.45951561607525, "learning_rate": 0.00024029999999999999, "loss": 1.7336, "step": 1602 }, { "epoch": 0.7543529411764706, "grad_norm": 11.764251993605132, "learning_rate": 0.00024044999999999996, "loss": 0.5202, "step": 1603 }, { "epoch": 0.7548235294117647, "grad_norm": 9.812445327021155, "learning_rate": 0.0002406, "loss": 1.569, "step": 1604 }, { "epoch": 0.7552941176470588, "grad_norm": 2.553325218229198, "learning_rate": 0.00024074999999999997, "loss": 1.246, "step": 1605 }, { "epoch": 0.7557647058823529, "grad_norm": 7.812084041057643, "learning_rate": 0.0002409, "loss": 1.4534, "step": 1606 }, { "epoch": 0.756235294117647, "grad_norm": 26.566763704453933, "learning_rate": 0.00024104999999999998, "loss": 0.8411, "step": 1607 }, { "epoch": 0.7567058823529412, "grad_norm": 28.681313148895214, "learning_rate": 0.00024119999999999998, "loss": 0.8173, "step": 1608 }, { "epoch": 0.7571764705882353, "grad_norm": 16.823678461662702, "learning_rate": 0.00024134999999999998, "loss": 2.0817, "step": 1609 }, { "epoch": 0.7576470588235295, "grad_norm": 21.39519706648097, "learning_rate": 0.0002415, "loss": 2.1878, "step": 1610 }, { "epoch": 0.7581176470588236, "grad_norm": 13.152447111354137, "learning_rate": 0.00024164999999999996, "loss": 1.1881, "step": 1611 }, { "epoch": 0.7585882352941177, "grad_norm": 4.69643041219363, "learning_rate": 0.0002418, "loss": 1.3039, "step": 1612 }, { "epoch": 0.7590588235294118, "grad_norm": 28.263104939220558, "learning_rate": 0.00024194999999999997, "loss": 0.6041, "step": 1613 }, { "epoch": 0.7595294117647059, "grad_norm": 13.11412598349419, "learning_rate": 0.0002421, "loss": 1.7701, "step": 1614 }, { "epoch": 0.76, "grad_norm": 5.834811647851324, "learning_rate": 0.00024224999999999998, "loss": 1.4065, "step": 1615 }, { "epoch": 0.7604705882352941, "grad_norm": 2.2094957073333172, "learning_rate": 0.00024239999999999998, "loss": 1.018, "step": 1616 }, { "epoch": 0.7609411764705882, "grad_norm": 6.601249496721915, "learning_rate": 0.00024255, "loss": 1.0269, "step": 1617 }, { "epoch": 0.7614117647058823, "grad_norm": 21.56054702044041, "learning_rate": 0.0002427, "loss": 0.3686, "step": 1618 }, { "epoch": 0.7618823529411765, "grad_norm": 1.1254795455330782, "learning_rate": 0.00024284999999999997, "loss": 0.8669, "step": 1619 }, { "epoch": 0.7623529411764706, "grad_norm": 14.524117691725895, "learning_rate": 0.000243, "loss": 1.2224, "step": 1620 }, { "epoch": 0.7628235294117647, "grad_norm": 18.386105540770686, "learning_rate": 0.00024314999999999997, "loss": 1.3619, "step": 1621 }, { "epoch": 0.7632941176470588, "grad_norm": 4.890500634831446, "learning_rate": 0.0002433, "loss": 0.5807, "step": 1622 }, { "epoch": 0.7637647058823529, "grad_norm": 39.20309867203434, "learning_rate": 0.00024344999999999998, "loss": 2.5771, "step": 1623 }, { "epoch": 0.764235294117647, "grad_norm": 16.402953104389997, "learning_rate": 0.00024359999999999999, "loss": 1.4095, "step": 1624 }, { "epoch": 0.7647058823529411, "grad_norm": 6.8315569969807655, "learning_rate": 0.00024375, "loss": 0.8992, "step": 1625 }, { "epoch": 0.7651764705882353, "grad_norm": 13.002381284011092, "learning_rate": 0.00024389999999999997, "loss": 0.8036, "step": 1626 }, { "epoch": 0.7656470588235295, "grad_norm": 36.15715960173398, "learning_rate": 0.00024404999999999997, "loss": 0.4844, "step": 1627 }, { "epoch": 0.7661176470588236, "grad_norm": 11.959865619135984, "learning_rate": 0.00024419999999999997, "loss": 1.1776, "step": 1628 }, { "epoch": 0.7665882352941177, "grad_norm": 9.764449406415988, "learning_rate": 0.00024435, "loss": 0.9795, "step": 1629 }, { "epoch": 0.7670588235294118, "grad_norm": 6.222344502685213, "learning_rate": 0.0002445, "loss": 1.505, "step": 1630 }, { "epoch": 0.7675294117647059, "grad_norm": 32.27677421273894, "learning_rate": 0.00024464999999999996, "loss": 2.3086, "step": 1631 }, { "epoch": 0.768, "grad_norm": 12.542273297048267, "learning_rate": 0.0002448, "loss": 1.3904, "step": 1632 }, { "epoch": 0.7684705882352941, "grad_norm": 9.29588096352356, "learning_rate": 0.00024494999999999996, "loss": 1.3562, "step": 1633 }, { "epoch": 0.7689411764705882, "grad_norm": 24.23004414728491, "learning_rate": 0.00024509999999999994, "loss": 0.6723, "step": 1634 }, { "epoch": 0.7694117647058824, "grad_norm": 31.501442867407864, "learning_rate": 0.00024524999999999997, "loss": 0.521, "step": 1635 }, { "epoch": 0.7698823529411765, "grad_norm": 19.114528418350098, "learning_rate": 0.00024539999999999995, "loss": 0.6781, "step": 1636 }, { "epoch": 0.7703529411764706, "grad_norm": 23.727504108530734, "learning_rate": 0.00024555, "loss": 1.9574, "step": 1637 }, { "epoch": 0.7708235294117647, "grad_norm": 1.11016603334719, "learning_rate": 0.00024569999999999995, "loss": 0.9521, "step": 1638 }, { "epoch": 0.7712941176470588, "grad_norm": 33.00639222950281, "learning_rate": 0.00024585, "loss": 2.1677, "step": 1639 }, { "epoch": 0.7717647058823529, "grad_norm": 3.792145567121205, "learning_rate": 0.00024599999999999996, "loss": 0.9608, "step": 1640 }, { "epoch": 0.772235294117647, "grad_norm": 7.677118416506635, "learning_rate": 0.00024615, "loss": 0.799, "step": 1641 }, { "epoch": 0.7727058823529411, "grad_norm": 13.089339246985418, "learning_rate": 0.00024629999999999997, "loss": 0.6222, "step": 1642 }, { "epoch": 0.7731764705882352, "grad_norm": 9.419512016531833, "learning_rate": 0.00024645, "loss": 1.1855, "step": 1643 }, { "epoch": 0.7736470588235295, "grad_norm": 9.585062450817215, "learning_rate": 0.0002466, "loss": 0.6382, "step": 1644 }, { "epoch": 0.7741176470588236, "grad_norm": 25.73857483342142, "learning_rate": 0.00024675, "loss": 1.9697, "step": 1645 }, { "epoch": 0.7745882352941177, "grad_norm": 12.903525599175968, "learning_rate": 0.0002469, "loss": 0.4422, "step": 1646 }, { "epoch": 0.7750588235294118, "grad_norm": 9.096709746609315, "learning_rate": 0.00024704999999999996, "loss": 1.4352, "step": 1647 }, { "epoch": 0.7755294117647059, "grad_norm": 6.933714886607528, "learning_rate": 0.0002472, "loss": 0.6692, "step": 1648 }, { "epoch": 0.776, "grad_norm": 30.685976340196962, "learning_rate": 0.00024734999999999997, "loss": 2.1902, "step": 1649 }, { "epoch": 0.7764705882352941, "grad_norm": 18.498725065644795, "learning_rate": 0.00024749999999999994, "loss": 0.4577, "step": 1650 }, { "epoch": 0.7769411764705882, "grad_norm": 8.790350192049281, "learning_rate": 0.00024765, "loss": 0.7529, "step": 1651 }, { "epoch": 0.7774117647058824, "grad_norm": 18.48358423467156, "learning_rate": 0.00024779999999999995, "loss": 1.9081, "step": 1652 }, { "epoch": 0.7778823529411765, "grad_norm": 13.297378940098744, "learning_rate": 0.00024795, "loss": 0.7394, "step": 1653 }, { "epoch": 0.7783529411764706, "grad_norm": 3.805127017223193, "learning_rate": 0.00024809999999999996, "loss": 1.2525, "step": 1654 }, { "epoch": 0.7788235294117647, "grad_norm": 17.578515268338695, "learning_rate": 0.00024825, "loss": 2.1094, "step": 1655 }, { "epoch": 0.7792941176470588, "grad_norm": 2.551568208006472, "learning_rate": 0.00024839999999999997, "loss": 1.2429, "step": 1656 }, { "epoch": 0.7797647058823529, "grad_norm": 21.7612065472126, "learning_rate": 0.00024855, "loss": 1.2198, "step": 1657 }, { "epoch": 0.780235294117647, "grad_norm": 24.594151694305598, "learning_rate": 0.0002487, "loss": 2.448, "step": 1658 }, { "epoch": 0.7807058823529411, "grad_norm": 21.61653392270677, "learning_rate": 0.00024885, "loss": 1.1117, "step": 1659 }, { "epoch": 0.7811764705882352, "grad_norm": 2.5712147953701687, "learning_rate": 0.000249, "loss": 1.5417, "step": 1660 }, { "epoch": 0.7816470588235294, "grad_norm": 6.857610795998875, "learning_rate": 0.00024914999999999996, "loss": 1.3348, "step": 1661 }, { "epoch": 0.7821176470588236, "grad_norm": 2.5159028774630814, "learning_rate": 0.0002493, "loss": 1.2387, "step": 1662 }, { "epoch": 0.7825882352941177, "grad_norm": 23.871859907446932, "learning_rate": 0.00024944999999999996, "loss": 0.6502, "step": 1663 }, { "epoch": 0.7830588235294118, "grad_norm": 16.545847990981564, "learning_rate": 0.00024959999999999994, "loss": 0.5171, "step": 1664 }, { "epoch": 0.7835294117647059, "grad_norm": 15.36863354716723, "learning_rate": 0.00024974999999999997, "loss": 1.4414, "step": 1665 }, { "epoch": 0.784, "grad_norm": 9.053430223319447, "learning_rate": 0.00024989999999999995, "loss": 0.4537, "step": 1666 }, { "epoch": 0.7844705882352941, "grad_norm": 10.223827420677083, "learning_rate": 0.00025005, "loss": 0.1037, "step": 1667 }, { "epoch": 0.7849411764705883, "grad_norm": 1.2774314263859388, "learning_rate": 0.00025019999999999996, "loss": 0.4956, "step": 1668 }, { "epoch": 0.7854117647058824, "grad_norm": 1.5189629006617877, "learning_rate": 0.00025035, "loss": 0.4934, "step": 1669 }, { "epoch": 0.7858823529411765, "grad_norm": 2.1188329745278, "learning_rate": 0.00025049999999999996, "loss": 0.4979, "step": 1670 }, { "epoch": 0.7863529411764706, "grad_norm": 8.216203786845965, "learning_rate": 0.00025065, "loss": 0.7041, "step": 1671 }, { "epoch": 0.7868235294117647, "grad_norm": 16.829006548957747, "learning_rate": 0.00025079999999999997, "loss": 1.0777, "step": 1672 }, { "epoch": 0.7872941176470588, "grad_norm": 9.513981308503999, "learning_rate": 0.00025095, "loss": 0.7145, "step": 1673 }, { "epoch": 0.7877647058823529, "grad_norm": 25.94969969719837, "learning_rate": 0.0002511, "loss": 1.5957, "step": 1674 }, { "epoch": 0.788235294117647, "grad_norm": 6.225183251688678, "learning_rate": 0.00025125, "loss": 0.0514, "step": 1675 }, { "epoch": 0.7887058823529411, "grad_norm": 46.09598088911999, "learning_rate": 0.0002514, "loss": 2.6497, "step": 1676 }, { "epoch": 0.7891764705882353, "grad_norm": 27.094890731216257, "learning_rate": 0.00025154999999999996, "loss": 1.7191, "step": 1677 }, { "epoch": 0.7896470588235294, "grad_norm": 6.413132754341096, "learning_rate": 0.0002517, "loss": 1.0815, "step": 1678 }, { "epoch": 0.7901176470588235, "grad_norm": 5.674145594800056, "learning_rate": 0.00025184999999999997, "loss": 1.3432, "step": 1679 }, { "epoch": 0.7905882352941176, "grad_norm": 44.547268537793066, "learning_rate": 0.00025199999999999995, "loss": 0.5891, "step": 1680 }, { "epoch": 0.7910588235294118, "grad_norm": 13.527725640658325, "learning_rate": 0.00025215, "loss": 1.4641, "step": 1681 }, { "epoch": 0.7915294117647059, "grad_norm": 3.8222635086561847, "learning_rate": 0.00025229999999999995, "loss": 2.2255, "step": 1682 }, { "epoch": 0.792, "grad_norm": 51.94533183217214, "learning_rate": 0.00025245, "loss": 0.8973, "step": 1683 }, { "epoch": 0.7924705882352941, "grad_norm": 15.435766253918926, "learning_rate": 0.00025259999999999996, "loss": 1.2529, "step": 1684 }, { "epoch": 0.7929411764705883, "grad_norm": 5.391330135578681, "learning_rate": 0.00025275, "loss": 1.3388, "step": 1685 }, { "epoch": 0.7934117647058824, "grad_norm": 13.1398032654144, "learning_rate": 0.00025289999999999997, "loss": 1.4735, "step": 1686 }, { "epoch": 0.7938823529411765, "grad_norm": 15.746889199728647, "learning_rate": 0.00025305, "loss": 1.4679, "step": 1687 }, { "epoch": 0.7943529411764706, "grad_norm": 5.630651549144933, "learning_rate": 0.0002532, "loss": 1.0228, "step": 1688 }, { "epoch": 0.7948235294117647, "grad_norm": 11.510565469884481, "learning_rate": 0.00025335, "loss": 0.4968, "step": 1689 }, { "epoch": 0.7952941176470588, "grad_norm": 10.268327852167507, "learning_rate": 0.0002535, "loss": 1.0763, "step": 1690 }, { "epoch": 0.7957647058823529, "grad_norm": 23.491627286638405, "learning_rate": 0.00025365, "loss": 1.7805, "step": 1691 }, { "epoch": 0.796235294117647, "grad_norm": 24.52435558432357, "learning_rate": 0.0002538, "loss": 2.0094, "step": 1692 }, { "epoch": 0.7967058823529412, "grad_norm": 18.42911676225681, "learning_rate": 0.00025394999999999997, "loss": 0.443, "step": 1693 }, { "epoch": 0.7971764705882353, "grad_norm": 1.8442008262337581, "learning_rate": 0.0002541, "loss": 0.9469, "step": 1694 }, { "epoch": 0.7976470588235294, "grad_norm": 6.114670544321379, "learning_rate": 0.00025425, "loss": 1.0677, "step": 1695 }, { "epoch": 0.7981176470588235, "grad_norm": 3.1049349732912983, "learning_rate": 0.00025439999999999995, "loss": 1.4226, "step": 1696 }, { "epoch": 0.7985882352941176, "grad_norm": 19.41365135051918, "learning_rate": 0.00025455, "loss": 0.6567, "step": 1697 }, { "epoch": 0.7990588235294117, "grad_norm": 3.286102195434951, "learning_rate": 0.00025469999999999996, "loss": 1.0573, "step": 1698 }, { "epoch": 0.7995294117647059, "grad_norm": 29.255644875574976, "learning_rate": 0.00025485, "loss": 2.2145, "step": 1699 }, { "epoch": 0.8, "grad_norm": 7.790769379483668, "learning_rate": 0.00025499999999999996, "loss": 0.8548, "step": 1700 }, { "epoch": 0.8004705882352942, "grad_norm": 13.640646706698579, "learning_rate": 0.00025515, "loss": 0.6274, "step": 1701 }, { "epoch": 0.8009411764705883, "grad_norm": 23.5827220449429, "learning_rate": 0.00025529999999999997, "loss": 2.1102, "step": 1702 }, { "epoch": 0.8014117647058824, "grad_norm": 11.606597723364567, "learning_rate": 0.00025545, "loss": 1.7348, "step": 1703 }, { "epoch": 0.8018823529411765, "grad_norm": 3.700040746558668, "learning_rate": 0.0002556, "loss": 1.1257, "step": 1704 }, { "epoch": 0.8023529411764706, "grad_norm": 14.965010504834144, "learning_rate": 0.00025575, "loss": 2.5448, "step": 1705 }, { "epoch": 0.8028235294117647, "grad_norm": 4.426399627893217, "learning_rate": 0.0002559, "loss": 0.5643, "step": 1706 }, { "epoch": 0.8032941176470588, "grad_norm": 7.5393505883321215, "learning_rate": 0.00025604999999999996, "loss": 1.0578, "step": 1707 }, { "epoch": 0.8037647058823529, "grad_norm": 22.084897077690048, "learning_rate": 0.0002562, "loss": 2.3485, "step": 1708 }, { "epoch": 0.804235294117647, "grad_norm": 27.09404770209842, "learning_rate": 0.00025634999999999997, "loss": 3.3642, "step": 1709 }, { "epoch": 0.8047058823529412, "grad_norm": 7.643462251988084, "learning_rate": 0.00025649999999999995, "loss": 1.3244, "step": 1710 }, { "epoch": 0.8051764705882353, "grad_norm": 5.216143017060381, "learning_rate": 0.00025665, "loss": 1.3705, "step": 1711 }, { "epoch": 0.8056470588235294, "grad_norm": 4.591504569524252, "learning_rate": 0.00025679999999999995, "loss": 1.7446, "step": 1712 }, { "epoch": 0.8061176470588235, "grad_norm": 15.416071884114325, "learning_rate": 0.00025695, "loss": 5.0552, "step": 1713 }, { "epoch": 0.8065882352941176, "grad_norm": 7.545822520241012, "learning_rate": 0.00025709999999999996, "loss": 1.9805, "step": 1714 }, { "epoch": 0.8070588235294117, "grad_norm": 5.320964258557006, "learning_rate": 0.00025725, "loss": 1.8566, "step": 1715 }, { "epoch": 0.8075294117647058, "grad_norm": 2.6105582680054926, "learning_rate": 0.00025739999999999997, "loss": 0.9135, "step": 1716 }, { "epoch": 0.808, "grad_norm": 11.230247825580651, "learning_rate": 0.00025755, "loss": 2.2953, "step": 1717 }, { "epoch": 0.8084705882352942, "grad_norm": 2.1238119749316944, "learning_rate": 0.0002577, "loss": 0.568, "step": 1718 }, { "epoch": 0.8089411764705883, "grad_norm": 11.94848223792255, "learning_rate": 0.00025785, "loss": 1.0986, "step": 1719 }, { "epoch": 0.8094117647058824, "grad_norm": 1.3648266023957674, "learning_rate": 0.000258, "loss": 0.0688, "step": 1720 }, { "epoch": 0.8098823529411765, "grad_norm": 17.696017343146757, "learning_rate": 0.00025815, "loss": 2.2737, "step": 1721 }, { "epoch": 0.8103529411764706, "grad_norm": 3.5159217446715125, "learning_rate": 0.0002583, "loss": 0.7431, "step": 1722 }, { "epoch": 0.8108235294117647, "grad_norm": 1.5106328730970509, "learning_rate": 0.00025844999999999997, "loss": 0.3014, "step": 1723 }, { "epoch": 0.8112941176470588, "grad_norm": 19.713195010023828, "learning_rate": 0.0002586, "loss": 2.3014, "step": 1724 }, { "epoch": 0.8117647058823529, "grad_norm": 1.4428545756798066, "learning_rate": 0.00025875, "loss": 0.1326, "step": 1725 }, { "epoch": 0.812235294117647, "grad_norm": 7.453590694096505, "learning_rate": 0.00025889999999999995, "loss": 0.9237, "step": 1726 }, { "epoch": 0.8127058823529412, "grad_norm": 2.508798441937304, "learning_rate": 0.00025905, "loss": 0.5764, "step": 1727 }, { "epoch": 0.8131764705882353, "grad_norm": 6.221638366455181, "learning_rate": 0.00025919999999999996, "loss": 1.5368, "step": 1728 }, { "epoch": 0.8136470588235294, "grad_norm": 2.5574301285622716, "learning_rate": 0.00025935, "loss": 0.6845, "step": 1729 }, { "epoch": 0.8141176470588235, "grad_norm": 0.7910580459249579, "learning_rate": 0.00025949999999999997, "loss": 0.1773, "step": 1730 }, { "epoch": 0.8145882352941176, "grad_norm": 5.446573172862555, "learning_rate": 0.00025965, "loss": 1.4136, "step": 1731 }, { "epoch": 0.8150588235294117, "grad_norm": 2.2278643492951393, "learning_rate": 0.00025979999999999997, "loss": 0.518, "step": 1732 }, { "epoch": 0.8155294117647058, "grad_norm": 8.825480256081114, "learning_rate": 0.00025995, "loss": 1.237, "step": 1733 }, { "epoch": 0.816, "grad_norm": 1.4796173868036393, "learning_rate": 0.0002601, "loss": 0.2633, "step": 1734 }, { "epoch": 0.8164705882352942, "grad_norm": 4.28901244803398, "learning_rate": 0.00026025, "loss": 1.2068, "step": 1735 }, { "epoch": 0.8169411764705883, "grad_norm": 1.8185460362908699, "learning_rate": 0.0002604, "loss": 0.6394, "step": 1736 }, { "epoch": 0.8174117647058824, "grad_norm": 7.562671148192199, "learning_rate": 0.00026055, "loss": 0.773, "step": 1737 }, { "epoch": 0.8178823529411765, "grad_norm": 6.177518271112749, "learning_rate": 0.0002607, "loss": 0.6769, "step": 1738 }, { "epoch": 0.8183529411764706, "grad_norm": 0.8911097600204985, "learning_rate": 0.00026084999999999997, "loss": 0.1995, "step": 1739 }, { "epoch": 0.8188235294117647, "grad_norm": 5.495490951323176, "learning_rate": 0.000261, "loss": 1.0982, "step": 1740 }, { "epoch": 0.8192941176470588, "grad_norm": 0.6312480174391479, "learning_rate": 0.00026115, "loss": 0.2047, "step": 1741 }, { "epoch": 0.819764705882353, "grad_norm": 0.7353251140332505, "learning_rate": 0.00026129999999999995, "loss": 0.255, "step": 1742 }, { "epoch": 0.8202352941176471, "grad_norm": 0.4978751538159515, "learning_rate": 0.00026145, "loss": 0.1436, "step": 1743 }, { "epoch": 0.8207058823529412, "grad_norm": 4.409262932113217, "learning_rate": 0.00026159999999999996, "loss": 1.0101, "step": 1744 }, { "epoch": 0.8211764705882353, "grad_norm": 8.929017879850983, "learning_rate": 0.00026175, "loss": 1.7948, "step": 1745 }, { "epoch": 0.8216470588235294, "grad_norm": 2.0629490651707054, "learning_rate": 0.00026189999999999997, "loss": 1.5383, "step": 1746 }, { "epoch": 0.8221176470588235, "grad_norm": 10.829202008359031, "learning_rate": 0.00026205, "loss": 1.8856, "step": 1747 }, { "epoch": 0.8225882352941176, "grad_norm": 2.4512709466065035, "learning_rate": 0.0002622, "loss": 0.7216, "step": 1748 }, { "epoch": 0.8230588235294117, "grad_norm": 2.9565344257461574, "learning_rate": 0.00026235, "loss": 0.885, "step": 1749 }, { "epoch": 0.8235294117647058, "grad_norm": 5.3652004992279725, "learning_rate": 0.0002625, "loss": 1.5027, "step": 1750 }, { "epoch": 0.824, "grad_norm": 2.7306510673537825, "learning_rate": 0.00026264999999999996, "loss": 1.3542, "step": 1751 }, { "epoch": 0.8244705882352941, "grad_norm": 4.251835324115278, "learning_rate": 0.0002628, "loss": 1.4299, "step": 1752 }, { "epoch": 0.8249411764705883, "grad_norm": 4.10502285914191, "learning_rate": 0.00026294999999999997, "loss": 1.1028, "step": 1753 }, { "epoch": 0.8254117647058824, "grad_norm": 2.537346098123724, "learning_rate": 0.0002631, "loss": 0.884, "step": 1754 }, { "epoch": 0.8258823529411765, "grad_norm": 1.5941023256452564, "learning_rate": 0.00026325, "loss": 0.5853, "step": 1755 }, { "epoch": 0.8263529411764706, "grad_norm": 1.967595448298134, "learning_rate": 0.00026339999999999995, "loss": 0.8216, "step": 1756 }, { "epoch": 0.8268235294117647, "grad_norm": 4.090469146214864, "learning_rate": 0.00026355, "loss": 1.6061, "step": 1757 }, { "epoch": 0.8272941176470588, "grad_norm": 1.6049921447635982, "learning_rate": 0.00026369999999999996, "loss": 0.8386, "step": 1758 }, { "epoch": 0.827764705882353, "grad_norm": 1.1249600871972099, "learning_rate": 0.00026384999999999994, "loss": 0.681, "step": 1759 }, { "epoch": 0.8282352941176471, "grad_norm": 1.1543899087905103, "learning_rate": 0.00026399999999999997, "loss": 0.8458, "step": 1760 }, { "epoch": 0.8287058823529412, "grad_norm": 196.90414562580844, "learning_rate": 0.00026414999999999994, "loss": 0.685, "step": 1761 }, { "epoch": 0.8291764705882353, "grad_norm": 1.1510045892198135, "learning_rate": 0.0002643, "loss": 0.5531, "step": 1762 }, { "epoch": 0.8296470588235294, "grad_norm": 2.698631348888115, "learning_rate": 0.00026444999999999995, "loss": 0.832, "step": 1763 }, { "epoch": 0.8301176470588235, "grad_norm": 6.378647910038557, "learning_rate": 0.0002646, "loss": 0.736, "step": 1764 }, { "epoch": 0.8305882352941176, "grad_norm": 7.251539060834629, "learning_rate": 0.00026474999999999996, "loss": 1.1692, "step": 1765 }, { "epoch": 0.8310588235294117, "grad_norm": 1.1358124387638462, "learning_rate": 0.0002649, "loss": 0.7523, "step": 1766 }, { "epoch": 0.8315294117647059, "grad_norm": 0.5753387289359211, "learning_rate": 0.00026504999999999996, "loss": 0.3712, "step": 1767 }, { "epoch": 0.832, "grad_norm": 0.3832502730772982, "learning_rate": 0.0002652, "loss": 0.1237, "step": 1768 }, { "epoch": 0.8324705882352941, "grad_norm": 3.9441570991350208, "learning_rate": 0.00026534999999999997, "loss": 1.0014, "step": 1769 }, { "epoch": 0.8329411764705882, "grad_norm": 48.528463519572014, "learning_rate": 0.0002655, "loss": 0.4015, "step": 1770 }, { "epoch": 0.8334117647058824, "grad_norm": 21.96301025296617, "learning_rate": 0.00026565, "loss": 2.1651, "step": 1771 }, { "epoch": 0.8338823529411765, "grad_norm": 0.817898610156038, "learning_rate": 0.00026579999999999996, "loss": 0.4084, "step": 1772 }, { "epoch": 0.8343529411764706, "grad_norm": 3.2773416182569366, "learning_rate": 0.00026595, "loss": 1.4001, "step": 1773 }, { "epoch": 0.8348235294117647, "grad_norm": 13.666968242349725, "learning_rate": 0.00026609999999999996, "loss": 0.6426, "step": 1774 }, { "epoch": 0.8352941176470589, "grad_norm": 2.83771388359688, "learning_rate": 0.00026624999999999994, "loss": 1.4918, "step": 1775 }, { "epoch": 0.835764705882353, "grad_norm": 3.4919859960996646, "learning_rate": 0.00026639999999999997, "loss": 0.6373, "step": 1776 }, { "epoch": 0.8362352941176471, "grad_norm": 2.2852941264076976, "learning_rate": 0.00026654999999999995, "loss": 0.5115, "step": 1777 }, { "epoch": 0.8367058823529412, "grad_norm": 0.8296192271127951, "learning_rate": 0.0002667, "loss": 0.3973, "step": 1778 }, { "epoch": 0.8371764705882353, "grad_norm": 0.2789907462589035, "learning_rate": 0.00026684999999999995, "loss": 0.1326, "step": 1779 }, { "epoch": 0.8376470588235294, "grad_norm": 6.666787291051869, "learning_rate": 0.000267, "loss": 1.1154, "step": 1780 }, { "epoch": 0.8381176470588235, "grad_norm": 0.34076825088914053, "learning_rate": 0.00026714999999999996, "loss": 0.075, "step": 1781 }, { "epoch": 0.8385882352941176, "grad_norm": 2.805750027428565, "learning_rate": 0.0002673, "loss": 1.1138, "step": 1782 }, { "epoch": 0.8390588235294117, "grad_norm": 1.310111977272033, "learning_rate": 0.00026744999999999997, "loss": 0.6966, "step": 1783 }, { "epoch": 0.8395294117647059, "grad_norm": 0.9834787106159917, "learning_rate": 0.0002676, "loss": 0.561, "step": 1784 }, { "epoch": 0.84, "grad_norm": 1.0790338015599386, "learning_rate": 0.00026775, "loss": 0.6052, "step": 1785 }, { "epoch": 0.8404705882352941, "grad_norm": 1.662545021672833, "learning_rate": 0.0002679, "loss": 0.9756, "step": 1786 }, { "epoch": 0.8409411764705882, "grad_norm": 0.3300524591614462, "learning_rate": 0.00026805, "loss": 0.1843, "step": 1787 }, { "epoch": 0.8414117647058823, "grad_norm": 2.2077878716762656, "learning_rate": 0.00026819999999999996, "loss": 1.0534, "step": 1788 }, { "epoch": 0.8418823529411765, "grad_norm": 0.8228158401358862, "learning_rate": 0.00026835, "loss": 1.0737, "step": 1789 }, { "epoch": 0.8423529411764706, "grad_norm": 0.8008894667437101, "learning_rate": 0.00026849999999999997, "loss": 0.3089, "step": 1790 }, { "epoch": 0.8428235294117647, "grad_norm": 1.7329978444696488, "learning_rate": 0.00026864999999999994, "loss": 0.5397, "step": 1791 }, { "epoch": 0.8432941176470589, "grad_norm": 2.281455187964231, "learning_rate": 0.0002688, "loss": 0.897, "step": 1792 }, { "epoch": 0.843764705882353, "grad_norm": 1.3815267000270073, "learning_rate": 0.00026894999999999995, "loss": 1.1412, "step": 1793 }, { "epoch": 0.8442352941176471, "grad_norm": 3.236624782297551, "learning_rate": 0.0002691, "loss": 1.2755, "step": 1794 }, { "epoch": 0.8447058823529412, "grad_norm": 1.3073318118540405, "learning_rate": 0.00026924999999999996, "loss": 1.0556, "step": 1795 }, { "epoch": 0.8451764705882353, "grad_norm": 1.365758011120906, "learning_rate": 0.0002694, "loss": 0.7296, "step": 1796 }, { "epoch": 0.8456470588235294, "grad_norm": 2.5006907951909483, "learning_rate": 0.00026954999999999997, "loss": 1.3482, "step": 1797 }, { "epoch": 0.8461176470588235, "grad_norm": 0.6628383414733569, "learning_rate": 0.0002697, "loss": 0.3796, "step": 1798 }, { "epoch": 0.8465882352941176, "grad_norm": 0.9554650576442598, "learning_rate": 0.00026984999999999997, "loss": 0.3902, "step": 1799 }, { "epoch": 0.8470588235294118, "grad_norm": 1.4317507853441265, "learning_rate": 0.00027, "loss": 0.3749, "step": 1800 }, { "epoch": 0.8475294117647059, "grad_norm": 1.560856126483493, "learning_rate": 0.00027015, "loss": 1.0766, "step": 1801 }, { "epoch": 0.848, "grad_norm": 0.3358836825576194, "learning_rate": 0.00027029999999999996, "loss": 0.1361, "step": 1802 }, { "epoch": 0.8484705882352941, "grad_norm": 3.3954144595978892, "learning_rate": 0.00027045, "loss": 1.1083, "step": 1803 }, { "epoch": 0.8489411764705882, "grad_norm": 0.2010428154930345, "learning_rate": 0.00027059999999999996, "loss": 0.0818, "step": 1804 }, { "epoch": 0.8494117647058823, "grad_norm": 0.7948014577796855, "learning_rate": 0.00027074999999999994, "loss": 0.3894, "step": 1805 }, { "epoch": 0.8498823529411764, "grad_norm": 1.051051912710073, "learning_rate": 0.00027089999999999997, "loss": 0.7544, "step": 1806 }, { "epoch": 0.8503529411764705, "grad_norm": 1.5735314868677328, "learning_rate": 0.00027104999999999995, "loss": 0.7365, "step": 1807 }, { "epoch": 0.8508235294117648, "grad_norm": 1.2592106555274853, "learning_rate": 0.0002712, "loss": 0.6886, "step": 1808 }, { "epoch": 0.8512941176470589, "grad_norm": 0.3462554597425785, "learning_rate": 0.00027134999999999995, "loss": 0.361, "step": 1809 }, { "epoch": 0.851764705882353, "grad_norm": 0.9282660098692865, "learning_rate": 0.0002715, "loss": 0.5418, "step": 1810 }, { "epoch": 0.8522352941176471, "grad_norm": 1.8399336572003446, "learning_rate": 0.00027164999999999996, "loss": 0.8455, "step": 1811 }, { "epoch": 0.8527058823529412, "grad_norm": 0.49166473200484173, "learning_rate": 0.0002718, "loss": 0.3635, "step": 1812 }, { "epoch": 0.8531764705882353, "grad_norm": 1.3198502273319026, "learning_rate": 0.00027194999999999997, "loss": 1.4106, "step": 1813 }, { "epoch": 0.8536470588235294, "grad_norm": 1.0928445372631308, "learning_rate": 0.0002721, "loss": 0.9003, "step": 1814 }, { "epoch": 0.8541176470588235, "grad_norm": 0.5737182700769438, "learning_rate": 0.00027225, "loss": 0.3212, "step": 1815 }, { "epoch": 0.8545882352941176, "grad_norm": 0.8267791716369893, "learning_rate": 0.0002724, "loss": 0.6275, "step": 1816 }, { "epoch": 0.8550588235294118, "grad_norm": 0.8587550606495005, "learning_rate": 0.00027255, "loss": 0.5187, "step": 1817 }, { "epoch": 0.8555294117647059, "grad_norm": 0.4947052220774127, "learning_rate": 0.00027269999999999996, "loss": 0.3993, "step": 1818 }, { "epoch": 0.856, "grad_norm": 12.00697536613261, "learning_rate": 0.00027285, "loss": 1.2788, "step": 1819 }, { "epoch": 0.8564705882352941, "grad_norm": 0.35548587184112707, "learning_rate": 0.00027299999999999997, "loss": 0.2608, "step": 1820 }, { "epoch": 0.8569411764705882, "grad_norm": 1.580234105265976, "learning_rate": 0.00027314999999999994, "loss": 0.8822, "step": 1821 }, { "epoch": 0.8574117647058823, "grad_norm": 9.712694901953858, "learning_rate": 0.0002733, "loss": 0.6797, "step": 1822 }, { "epoch": 0.8578823529411764, "grad_norm": 0.7609121045218805, "learning_rate": 0.00027344999999999995, "loss": 0.4531, "step": 1823 }, { "epoch": 0.8583529411764705, "grad_norm": 0.12450736239767317, "learning_rate": 0.0002736, "loss": 0.0343, "step": 1824 }, { "epoch": 0.8588235294117647, "grad_norm": 1.269090197124888, "learning_rate": 0.00027374999999999996, "loss": 0.4658, "step": 1825 }, { "epoch": 0.8592941176470589, "grad_norm": 0.7942473189728645, "learning_rate": 0.0002739, "loss": 0.3536, "step": 1826 }, { "epoch": 0.859764705882353, "grad_norm": 0.4836241733576445, "learning_rate": 0.00027404999999999997, "loss": 0.3687, "step": 1827 }, { "epoch": 0.8602352941176471, "grad_norm": 1.5392282686390177, "learning_rate": 0.0002742, "loss": 1.2077, "step": 1828 }, { "epoch": 0.8607058823529412, "grad_norm": 1.0640809794123511, "learning_rate": 0.00027435, "loss": 0.8022, "step": 1829 }, { "epoch": 0.8611764705882353, "grad_norm": 0.8191930986765936, "learning_rate": 0.0002745, "loss": 0.614, "step": 1830 }, { "epoch": 0.8616470588235294, "grad_norm": 0.762668641849838, "learning_rate": 0.00027465, "loss": 0.8677, "step": 1831 }, { "epoch": 0.8621176470588235, "grad_norm": 1.4006040267517763, "learning_rate": 0.0002748, "loss": 0.8798, "step": 1832 }, { "epoch": 0.8625882352941177, "grad_norm": 0.5117486556807868, "learning_rate": 0.00027495, "loss": 0.3489, "step": 1833 }, { "epoch": 0.8630588235294118, "grad_norm": 0.5582208253405668, "learning_rate": 0.00027509999999999996, "loss": 0.2406, "step": 1834 }, { "epoch": 0.8635294117647059, "grad_norm": 2.076137618121632, "learning_rate": 0.00027525, "loss": 1.3559, "step": 1835 }, { "epoch": 0.864, "grad_norm": 0.8119012183504468, "learning_rate": 0.00027539999999999997, "loss": 0.6149, "step": 1836 }, { "epoch": 0.8644705882352941, "grad_norm": 0.7379131461040187, "learning_rate": 0.00027554999999999995, "loss": 0.4955, "step": 1837 }, { "epoch": 0.8649411764705882, "grad_norm": 0.8351195430914822, "learning_rate": 0.0002757, "loss": 0.5209, "step": 1838 }, { "epoch": 0.8654117647058823, "grad_norm": 1.1536496855053413, "learning_rate": 0.00027584999999999996, "loss": 0.8544, "step": 1839 }, { "epoch": 0.8658823529411764, "grad_norm": 0.48615169703452243, "learning_rate": 0.000276, "loss": 0.3013, "step": 1840 }, { "epoch": 0.8663529411764705, "grad_norm": 0.5313419211684409, "learning_rate": 0.00027614999999999996, "loss": 0.4254, "step": 1841 }, { "epoch": 0.8668235294117647, "grad_norm": 2.6563422581524745, "learning_rate": 0.0002763, "loss": 1.5563, "step": 1842 }, { "epoch": 0.8672941176470588, "grad_norm": 0.5935076229406695, "learning_rate": 0.00027644999999999997, "loss": 0.4742, "step": 1843 }, { "epoch": 0.867764705882353, "grad_norm": 0.2555816395590088, "learning_rate": 0.0002766, "loss": 0.098, "step": 1844 }, { "epoch": 0.8682352941176471, "grad_norm": 1.3633091876758858, "learning_rate": 0.00027675, "loss": 0.6655, "step": 1845 }, { "epoch": 0.8687058823529412, "grad_norm": 1.110265681031719, "learning_rate": 0.0002769, "loss": 0.6638, "step": 1846 }, { "epoch": 0.8691764705882353, "grad_norm": 0.6846586288089598, "learning_rate": 0.00027705, "loss": 0.1984, "step": 1847 }, { "epoch": 0.8696470588235294, "grad_norm": 0.5715409808730861, "learning_rate": 0.0002772, "loss": 0.34, "step": 1848 }, { "epoch": 0.8701176470588236, "grad_norm": 1.380296033517468, "learning_rate": 0.00027735, "loss": 0.8761, "step": 1849 }, { "epoch": 0.8705882352941177, "grad_norm": 2.5593144174416613, "learning_rate": 0.00027749999999999997, "loss": 0.8535, "step": 1850 }, { "epoch": 0.8710588235294118, "grad_norm": 0.8045321421608818, "learning_rate": 0.00027764999999999995, "loss": 0.8637, "step": 1851 }, { "epoch": 0.8715294117647059, "grad_norm": 0.23285766712515976, "learning_rate": 0.0002778, "loss": 0.1358, "step": 1852 }, { "epoch": 0.872, "grad_norm": 0.9545030538924315, "learning_rate": 0.00027794999999999995, "loss": 0.374, "step": 1853 }, { "epoch": 0.8724705882352941, "grad_norm": 2.812482020081087, "learning_rate": 0.0002781, "loss": 1.5258, "step": 1854 }, { "epoch": 0.8729411764705882, "grad_norm": 1.2278848065348473, "learning_rate": 0.00027824999999999996, "loss": 1.0524, "step": 1855 }, { "epoch": 0.8734117647058823, "grad_norm": 0.4420074061840446, "learning_rate": 0.0002784, "loss": 0.2185, "step": 1856 }, { "epoch": 0.8738823529411764, "grad_norm": 1.231125314766152, "learning_rate": 0.00027854999999999997, "loss": 0.8422, "step": 1857 }, { "epoch": 0.8743529411764706, "grad_norm": 0.7301806347593586, "learning_rate": 0.0002787, "loss": 0.4212, "step": 1858 }, { "epoch": 0.8748235294117647, "grad_norm": 1.231329395550635, "learning_rate": 0.00027885, "loss": 0.8059, "step": 1859 }, { "epoch": 0.8752941176470588, "grad_norm": 1.0762674440469995, "learning_rate": 0.000279, "loss": 0.734, "step": 1860 }, { "epoch": 0.8757647058823529, "grad_norm": 1.1052152317383828, "learning_rate": 0.00027915, "loss": 0.7168, "step": 1861 }, { "epoch": 0.8762352941176471, "grad_norm": 1.9879368391142493, "learning_rate": 0.0002793, "loss": 1.1304, "step": 1862 }, { "epoch": 0.8767058823529412, "grad_norm": 0.5731101719495212, "learning_rate": 0.00027945, "loss": 0.3964, "step": 1863 }, { "epoch": 0.8771764705882353, "grad_norm": 2.3445481577694776, "learning_rate": 0.00027959999999999997, "loss": 0.6733, "step": 1864 }, { "epoch": 0.8776470588235294, "grad_norm": 0.7255452675836992, "learning_rate": 0.00027975, "loss": 0.2992, "step": 1865 }, { "epoch": 0.8781176470588236, "grad_norm": 1.2437327400368126, "learning_rate": 0.0002799, "loss": 0.4002, "step": 1866 }, { "epoch": 0.8785882352941177, "grad_norm": 1.087357897806919, "learning_rate": 0.00028004999999999995, "loss": 0.648, "step": 1867 }, { "epoch": 0.8790588235294118, "grad_norm": 0.7572869591136421, "learning_rate": 0.0002802, "loss": 0.5521, "step": 1868 }, { "epoch": 0.8795294117647059, "grad_norm": 0.6566694198216998, "learning_rate": 0.00028034999999999996, "loss": 0.5615, "step": 1869 }, { "epoch": 0.88, "grad_norm": 1.1827217157021068, "learning_rate": 0.0002805, "loss": 0.5535, "step": 1870 }, { "epoch": 0.8804705882352941, "grad_norm": 0.9596553146991411, "learning_rate": 0.00028064999999999996, "loss": 0.603, "step": 1871 }, { "epoch": 0.8809411764705882, "grad_norm": 0.7112616741431553, "learning_rate": 0.0002808, "loss": 0.4459, "step": 1872 }, { "epoch": 0.8814117647058823, "grad_norm": 1.15732120236837, "learning_rate": 0.00028094999999999997, "loss": 0.5892, "step": 1873 }, { "epoch": 0.8818823529411765, "grad_norm": 0.485402009097053, "learning_rate": 0.0002811, "loss": 0.3374, "step": 1874 }, { "epoch": 0.8823529411764706, "grad_norm": 0.68915938459674, "learning_rate": 0.00028125, "loss": 0.6608, "step": 1875 }, { "epoch": 0.8828235294117647, "grad_norm": 1.3039999858312605, "learning_rate": 0.00028139999999999996, "loss": 0.7923, "step": 1876 }, { "epoch": 0.8832941176470588, "grad_norm": 0.559640668062854, "learning_rate": 0.00028155, "loss": 0.3868, "step": 1877 }, { "epoch": 0.8837647058823529, "grad_norm": 0.9744361469619051, "learning_rate": 0.00028169999999999996, "loss": 0.8597, "step": 1878 }, { "epoch": 0.884235294117647, "grad_norm": 1.0175710218716063, "learning_rate": 0.00028185, "loss": 0.6232, "step": 1879 }, { "epoch": 0.8847058823529412, "grad_norm": 0.475144649041125, "learning_rate": 0.00028199999999999997, "loss": 0.3927, "step": 1880 }, { "epoch": 0.8851764705882353, "grad_norm": 0.7105033143337668, "learning_rate": 0.00028215, "loss": 0.4846, "step": 1881 }, { "epoch": 0.8856470588235295, "grad_norm": 0.8625133563186781, "learning_rate": 0.0002823, "loss": 0.6333, "step": 1882 }, { "epoch": 0.8861176470588236, "grad_norm": 0.7281427800742382, "learning_rate": 0.00028244999999999995, "loss": 0.7328, "step": 1883 }, { "epoch": 0.8865882352941177, "grad_norm": 0.57972039678983, "learning_rate": 0.0002826, "loss": 0.4784, "step": 1884 }, { "epoch": 0.8870588235294118, "grad_norm": 0.5963486626484353, "learning_rate": 0.00028274999999999996, "loss": 0.4264, "step": 1885 }, { "epoch": 0.8875294117647059, "grad_norm": 0.5278108949033374, "learning_rate": 0.00028289999999999994, "loss": 0.3488, "step": 1886 }, { "epoch": 0.888, "grad_norm": 0.8186624042739592, "learning_rate": 0.00028304999999999997, "loss": 0.4476, "step": 1887 }, { "epoch": 0.8884705882352941, "grad_norm": 0.7929644728948042, "learning_rate": 0.00028319999999999994, "loss": 0.6064, "step": 1888 }, { "epoch": 0.8889411764705882, "grad_norm": 0.6106008527175643, "learning_rate": 0.00028335, "loss": 0.6665, "step": 1889 }, { "epoch": 0.8894117647058823, "grad_norm": 0.11192273409740974, "learning_rate": 0.00028349999999999995, "loss": 0.0748, "step": 1890 }, { "epoch": 0.8898823529411765, "grad_norm": 0.5825493394473364, "learning_rate": 0.00028365, "loss": 0.3455, "step": 1891 }, { "epoch": 0.8903529411764706, "grad_norm": 0.353580874123296, "learning_rate": 0.00028379999999999996, "loss": 0.1431, "step": 1892 }, { "epoch": 0.8908235294117647, "grad_norm": 0.5043795053713614, "learning_rate": 0.00028395, "loss": 0.4217, "step": 1893 }, { "epoch": 0.8912941176470588, "grad_norm": 0.7246396929051523, "learning_rate": 0.00028409999999999997, "loss": 0.2858, "step": 1894 }, { "epoch": 0.8917647058823529, "grad_norm": 19.451210579730546, "learning_rate": 0.00028425, "loss": 0.5111, "step": 1895 }, { "epoch": 0.892235294117647, "grad_norm": 0.9051717585285579, "learning_rate": 0.0002844, "loss": 0.5758, "step": 1896 }, { "epoch": 0.8927058823529411, "grad_norm": 3.8838858832781358, "learning_rate": 0.00028455, "loss": 0.8024, "step": 1897 }, { "epoch": 0.8931764705882353, "grad_norm": 125.82532491812121, "learning_rate": 0.0002847, "loss": 3.2905, "step": 1898 }, { "epoch": 0.8936470588235295, "grad_norm": 1.566700575076739, "learning_rate": 0.00028484999999999996, "loss": 0.3649, "step": 1899 }, { "epoch": 0.8941176470588236, "grad_norm": 2.0155819736424427, "learning_rate": 0.000285, "loss": 1.0351, "step": 1900 }, { "epoch": 0.8945882352941177, "grad_norm": 0.4675249264501612, "learning_rate": 0.00028514999999999997, "loss": 0.4117, "step": 1901 }, { "epoch": 0.8950588235294118, "grad_norm": 1.2029960952459153, "learning_rate": 0.00028529999999999994, "loss": 0.6608, "step": 1902 }, { "epoch": 0.8955294117647059, "grad_norm": 1.2214987089195808, "learning_rate": 0.00028544999999999997, "loss": 0.5994, "step": 1903 }, { "epoch": 0.896, "grad_norm": 2.5342688516211522, "learning_rate": 0.00028559999999999995, "loss": 0.113, "step": 1904 }, { "epoch": 0.8964705882352941, "grad_norm": 2.3289890533164543, "learning_rate": 0.00028575, "loss": 0.9836, "step": 1905 }, { "epoch": 0.8969411764705882, "grad_norm": 3.9315469378883425, "learning_rate": 0.00028589999999999996, "loss": 1.2419, "step": 1906 }, { "epoch": 0.8974117647058824, "grad_norm": 1.5274442528321701, "learning_rate": 0.00028605, "loss": 0.3878, "step": 1907 }, { "epoch": 0.8978823529411765, "grad_norm": 2.9504817934018885, "learning_rate": 0.00028619999999999996, "loss": 0.987, "step": 1908 }, { "epoch": 0.8983529411764706, "grad_norm": 0.4063591026738378, "learning_rate": 0.00028635, "loss": 0.1451, "step": 1909 }, { "epoch": 0.8988235294117647, "grad_norm": 0.9823714984498583, "learning_rate": 0.00028649999999999997, "loss": 0.2471, "step": 1910 }, { "epoch": 0.8992941176470588, "grad_norm": 5.7810259774061095, "learning_rate": 0.00028665, "loss": 1.4602, "step": 1911 }, { "epoch": 0.8997647058823529, "grad_norm": 0.7576295459683507, "learning_rate": 0.0002868, "loss": 0.262, "step": 1912 }, { "epoch": 0.900235294117647, "grad_norm": 1.3605988527010944, "learning_rate": 0.00028694999999999995, "loss": 0.3201, "step": 1913 }, { "epoch": 0.9007058823529411, "grad_norm": 0.2277172492687653, "learning_rate": 0.0002871, "loss": 0.0617, "step": 1914 }, { "epoch": 0.9011764705882352, "grad_norm": 0.6827015937730273, "learning_rate": 0.00028724999999999996, "loss": 0.2501, "step": 1915 }, { "epoch": 0.9016470588235294, "grad_norm": 8.282096631685715, "learning_rate": 0.00028739999999999994, "loss": 1.5253, "step": 1916 }, { "epoch": 0.9021176470588236, "grad_norm": 2.9957151442463275, "learning_rate": 0.00028754999999999997, "loss": 0.8275, "step": 1917 }, { "epoch": 0.9025882352941177, "grad_norm": 0.06903586591546478, "learning_rate": 0.00028769999999999995, "loss": 0.0166, "step": 1918 }, { "epoch": 0.9030588235294118, "grad_norm": 2.1758499096460873, "learning_rate": 0.00028785, "loss": 0.7479, "step": 1919 }, { "epoch": 0.9035294117647059, "grad_norm": 1.6666658993920997, "learning_rate": 0.00028799999999999995, "loss": 0.4998, "step": 1920 }, { "epoch": 0.904, "grad_norm": 1.642120968242702, "learning_rate": 0.00028815, "loss": 0.3242, "step": 1921 }, { "epoch": 0.9044705882352941, "grad_norm": 5.109555238147288, "learning_rate": 0.00028829999999999996, "loss": 0.9101, "step": 1922 }, { "epoch": 0.9049411764705882, "grad_norm": 1.8653172378566767, "learning_rate": 0.00028845, "loss": 0.5354, "step": 1923 }, { "epoch": 0.9054117647058824, "grad_norm": 1.7416740007306128, "learning_rate": 0.00028859999999999997, "loss": 0.3552, "step": 1924 }, { "epoch": 0.9058823529411765, "grad_norm": 2.4723738978307432, "learning_rate": 0.00028875, "loss": 0.9847, "step": 1925 }, { "epoch": 0.9063529411764706, "grad_norm": 1.7348567919295095, "learning_rate": 0.0002889, "loss": 0.4797, "step": 1926 }, { "epoch": 0.9068235294117647, "grad_norm": 2.634326483806482, "learning_rate": 0.00028905, "loss": 0.8664, "step": 1927 }, { "epoch": 0.9072941176470588, "grad_norm": 1.761607949632746, "learning_rate": 0.0002892, "loss": 0.8537, "step": 1928 }, { "epoch": 0.9077647058823529, "grad_norm": 1.0985711525876223, "learning_rate": 0.00028934999999999996, "loss": 1.0558, "step": 1929 }, { "epoch": 0.908235294117647, "grad_norm": 2.151278000592147, "learning_rate": 0.0002895, "loss": 0.8051, "step": 1930 }, { "epoch": 0.9087058823529411, "grad_norm": 1.0795103761787148, "learning_rate": 0.00028964999999999997, "loss": 0.5083, "step": 1931 }, { "epoch": 0.9091764705882353, "grad_norm": 1.4097084639334874, "learning_rate": 0.00028979999999999994, "loss": 0.8445, "step": 1932 }, { "epoch": 0.9096470588235294, "grad_norm": 0.704404252703207, "learning_rate": 0.00028995, "loss": 0.4337, "step": 1933 }, { "epoch": 0.9101176470588235, "grad_norm": 1.4092922841713282, "learning_rate": 0.00029009999999999995, "loss": 0.7181, "step": 1934 }, { "epoch": 0.9105882352941177, "grad_norm": 0.5457391077874673, "learning_rate": 0.00029025, "loss": 0.2775, "step": 1935 }, { "epoch": 0.9110588235294118, "grad_norm": 0.5944788846524878, "learning_rate": 0.00029039999999999996, "loss": 0.343, "step": 1936 }, { "epoch": 0.9115294117647059, "grad_norm": 0.31758501122278904, "learning_rate": 0.00029055, "loss": 0.2177, "step": 1937 }, { "epoch": 0.912, "grad_norm": 0.5236911124816829, "learning_rate": 0.00029069999999999996, "loss": 0.4748, "step": 1938 }, { "epoch": 0.9124705882352941, "grad_norm": 0.7640475170832921, "learning_rate": 0.00029085, "loss": 0.4393, "step": 1939 }, { "epoch": 0.9129411764705883, "grad_norm": 1.253513394330692, "learning_rate": 0.00029099999999999997, "loss": 1.0481, "step": 1940 }, { "epoch": 0.9134117647058824, "grad_norm": 1.060844886616055, "learning_rate": 0.00029115, "loss": 1.1855, "step": 1941 }, { "epoch": 0.9138823529411765, "grad_norm": 0.3030738911744998, "learning_rate": 0.0002913, "loss": 0.2027, "step": 1942 }, { "epoch": 0.9143529411764706, "grad_norm": 0.857957528566246, "learning_rate": 0.00029145, "loss": 0.5484, "step": 1943 }, { "epoch": 0.9148235294117647, "grad_norm": 0.6434054480675472, "learning_rate": 0.0002916, "loss": 0.3951, "step": 1944 }, { "epoch": 0.9152941176470588, "grad_norm": 1.5172276714489636, "learning_rate": 0.00029174999999999996, "loss": 0.798, "step": 1945 }, { "epoch": 0.9157647058823529, "grad_norm": 0.6916802708171395, "learning_rate": 0.0002919, "loss": 0.5349, "step": 1946 }, { "epoch": 0.916235294117647, "grad_norm": 1.126786066673913, "learning_rate": 0.00029204999999999997, "loss": 0.5669, "step": 1947 }, { "epoch": 0.9167058823529411, "grad_norm": 1.035111045305413, "learning_rate": 0.00029219999999999995, "loss": 0.6513, "step": 1948 }, { "epoch": 0.9171764705882353, "grad_norm": 0.5405310465989341, "learning_rate": 0.00029235, "loss": 0.3349, "step": 1949 }, { "epoch": 0.9176470588235294, "grad_norm": 0.525725396541348, "learning_rate": 0.00029249999999999995, "loss": 0.409, "step": 1950 }, { "epoch": 0.9181176470588235, "grad_norm": 0.49200147814988493, "learning_rate": 0.00029265, "loss": 0.2333, "step": 1951 }, { "epoch": 0.9185882352941176, "grad_norm": 0.556464094721536, "learning_rate": 0.00029279999999999996, "loss": 0.1916, "step": 1952 }, { "epoch": 0.9190588235294118, "grad_norm": 0.9192100851248811, "learning_rate": 0.00029295, "loss": 0.3499, "step": 1953 }, { "epoch": 0.9195294117647059, "grad_norm": 1.0027796616554165, "learning_rate": 0.00029309999999999997, "loss": 0.7417, "step": 1954 }, { "epoch": 0.92, "grad_norm": 0.7441838632073738, "learning_rate": 0.00029325, "loss": 0.371, "step": 1955 }, { "epoch": 0.9204705882352942, "grad_norm": 0.4018505025194592, "learning_rate": 0.0002934, "loss": 0.204, "step": 1956 }, { "epoch": 0.9209411764705883, "grad_norm": 0.42388957668549504, "learning_rate": 0.00029355, "loss": 0.209, "step": 1957 }, { "epoch": 0.9214117647058824, "grad_norm": 0.26462857955121827, "learning_rate": 0.0002937, "loss": 0.1905, "step": 1958 }, { "epoch": 0.9218823529411765, "grad_norm": 0.8542030362662039, "learning_rate": 0.00029384999999999996, "loss": 0.571, "step": 1959 }, { "epoch": 0.9223529411764706, "grad_norm": 0.8471279613692504, "learning_rate": 0.000294, "loss": 0.6416, "step": 1960 }, { "epoch": 0.9228235294117647, "grad_norm": 0.5580907954116568, "learning_rate": 0.00029414999999999997, "loss": 0.3012, "step": 1961 }, { "epoch": 0.9232941176470588, "grad_norm": 1.4852203785459257, "learning_rate": 0.00029429999999999994, "loss": 1.1226, "step": 1962 }, { "epoch": 0.9237647058823529, "grad_norm": 0.6212996200136491, "learning_rate": 0.00029445, "loss": 0.4549, "step": 1963 }, { "epoch": 0.924235294117647, "grad_norm": 0.40715526921688205, "learning_rate": 0.00029459999999999995, "loss": 0.2595, "step": 1964 }, { "epoch": 0.9247058823529412, "grad_norm": 1.4076790103087586, "learning_rate": 0.00029475, "loss": 0.731, "step": 1965 }, { "epoch": 0.9251764705882353, "grad_norm": 0.6514887865556417, "learning_rate": 0.00029489999999999996, "loss": 0.3153, "step": 1966 }, { "epoch": 0.9256470588235294, "grad_norm": 0.4786449428565789, "learning_rate": 0.00029505, "loss": 0.2585, "step": 1967 }, { "epoch": 0.9261176470588235, "grad_norm": 0.31707736420484206, "learning_rate": 0.00029519999999999997, "loss": 0.1618, "step": 1968 }, { "epoch": 0.9265882352941176, "grad_norm": 0.4355603143562199, "learning_rate": 0.00029535, "loss": 0.1767, "step": 1969 }, { "epoch": 0.9270588235294117, "grad_norm": 1.0576032879419166, "learning_rate": 0.00029549999999999997, "loss": 0.6122, "step": 1970 }, { "epoch": 0.9275294117647059, "grad_norm": 1.0716759489327496, "learning_rate": 0.00029565, "loss": 0.5482, "step": 1971 }, { "epoch": 0.928, "grad_norm": 0.40490502702256487, "learning_rate": 0.0002958, "loss": 0.2988, "step": 1972 }, { "epoch": 0.9284705882352942, "grad_norm": 0.6845936672863838, "learning_rate": 0.00029595, "loss": 0.6043, "step": 1973 }, { "epoch": 0.9289411764705883, "grad_norm": 0.8325093811046855, "learning_rate": 0.0002961, "loss": 0.5781, "step": 1974 }, { "epoch": 0.9294117647058824, "grad_norm": 0.3217578894492605, "learning_rate": 0.00029624999999999996, "loss": 0.2174, "step": 1975 }, { "epoch": 0.9298823529411765, "grad_norm": 0.5536109063253879, "learning_rate": 0.0002964, "loss": 0.4376, "step": 1976 }, { "epoch": 0.9303529411764706, "grad_norm": 0.5551623033014561, "learning_rate": 0.00029654999999999997, "loss": 0.3815, "step": 1977 }, { "epoch": 0.9308235294117647, "grad_norm": 0.7342139643281442, "learning_rate": 0.00029669999999999995, "loss": 0.5566, "step": 1978 }, { "epoch": 0.9312941176470588, "grad_norm": 0.21722916932256797, "learning_rate": 0.00029685, "loss": 0.1793, "step": 1979 }, { "epoch": 0.9317647058823529, "grad_norm": 0.4636583476547475, "learning_rate": 0.00029699999999999996, "loss": 0.2711, "step": 1980 }, { "epoch": 0.932235294117647, "grad_norm": 0.7116976805142942, "learning_rate": 0.00029715, "loss": 0.4966, "step": 1981 }, { "epoch": 0.9327058823529412, "grad_norm": 0.3525117700252987, "learning_rate": 0.00029729999999999996, "loss": 0.2446, "step": 1982 }, { "epoch": 0.9331764705882353, "grad_norm": 1.3667523663477936, "learning_rate": 0.00029745, "loss": 1.0022, "step": 1983 }, { "epoch": 0.9336470588235294, "grad_norm": 0.4986088408692412, "learning_rate": 0.00029759999999999997, "loss": 0.4568, "step": 1984 }, { "epoch": 0.9341176470588235, "grad_norm": 0.47127417866885307, "learning_rate": 0.00029775, "loss": 0.5213, "step": 1985 }, { "epoch": 0.9345882352941176, "grad_norm": 0.43737739039109375, "learning_rate": 0.0002979, "loss": 0.2215, "step": 1986 }, { "epoch": 0.9350588235294117, "grad_norm": 1.058278296159893, "learning_rate": 0.00029805, "loss": 0.5873, "step": 1987 }, { "epoch": 0.9355294117647058, "grad_norm": 0.4893682212789742, "learning_rate": 0.0002982, "loss": 0.3478, "step": 1988 }, { "epoch": 0.936, "grad_norm": 1.0936045744365193, "learning_rate": 0.00029835, "loss": 0.6554, "step": 1989 }, { "epoch": 0.9364705882352942, "grad_norm": 0.6974133786564026, "learning_rate": 0.0002985, "loss": 0.6037, "step": 1990 }, { "epoch": 0.9369411764705883, "grad_norm": 0.41053740452199045, "learning_rate": 0.00029864999999999997, "loss": 0.3745, "step": 1991 }, { "epoch": 0.9374117647058824, "grad_norm": 0.3141438351492975, "learning_rate": 0.0002988, "loss": 0.2464, "step": 1992 }, { "epoch": 0.9378823529411765, "grad_norm": 0.28683301815153006, "learning_rate": 0.00029895, "loss": 0.1185, "step": 1993 }, { "epoch": 0.9383529411764706, "grad_norm": 0.5384479484154534, "learning_rate": 0.00029909999999999995, "loss": 0.3644, "step": 1994 }, { "epoch": 0.9388235294117647, "grad_norm": 0.39569963816986975, "learning_rate": 0.00029925, "loss": 0.2226, "step": 1995 }, { "epoch": 0.9392941176470588, "grad_norm": 0.7429445736803828, "learning_rate": 0.00029939999999999996, "loss": 0.5639, "step": 1996 }, { "epoch": 0.939764705882353, "grad_norm": 0.8919778892501027, "learning_rate": 0.00029955, "loss": 0.6543, "step": 1997 }, { "epoch": 0.9402352941176471, "grad_norm": 0.23659118607363222, "learning_rate": 0.00029969999999999997, "loss": 0.1161, "step": 1998 }, { "epoch": 0.9407058823529412, "grad_norm": 0.7293191413143553, "learning_rate": 0.00029985, "loss": 0.6406, "step": 1999 }, { "epoch": 0.9411764705882353, "grad_norm": 0.9537196956886117, "learning_rate": 0.0003, "loss": 0.9815, "step": 2000 }, { "epoch": 0.9416470588235294, "grad_norm": 0.6843834919739176, "learning_rate": 0.0002999999935946321, "loss": 0.6633, "step": 2001 }, { "epoch": 0.9421176470588235, "grad_norm": 0.8249335907369949, "learning_rate": 0.00029999997437852905, "loss": 0.548, "step": 2002 }, { "epoch": 0.9425882352941176, "grad_norm": 0.49780367858438357, "learning_rate": 0.0002999999423516924, "loss": 0.4032, "step": 2003 }, { "epoch": 0.9430588235294117, "grad_norm": 0.32871581035952824, "learning_rate": 0.00029999989751412494, "loss": 0.1198, "step": 2004 }, { "epoch": 0.9435294117647058, "grad_norm": 0.8691926541031894, "learning_rate": 0.00029999983986583054, "loss": 0.5838, "step": 2005 }, { "epoch": 0.944, "grad_norm": 0.1275280788535422, "learning_rate": 0.000299999769406814, "loss": 0.0677, "step": 2006 }, { "epoch": 0.9444705882352942, "grad_norm": 1.0386497119188172, "learning_rate": 0.0002999996861370814, "loss": 0.6151, "step": 2007 }, { "epoch": 0.9449411764705883, "grad_norm": 0.6098674603403286, "learning_rate": 0.0002999995900566399, "loss": 0.3863, "step": 2008 }, { "epoch": 0.9454117647058824, "grad_norm": 0.5204718833337892, "learning_rate": 0.0002999994811654977, "loss": 0.2923, "step": 2009 }, { "epoch": 0.9458823529411765, "grad_norm": 0.7180566601011956, "learning_rate": 0.000299999359463664, "loss": 0.6243, "step": 2010 }, { "epoch": 0.9463529411764706, "grad_norm": 0.11135598897978455, "learning_rate": 0.0002999992249511493, "loss": 0.0451, "step": 2011 }, { "epoch": 0.9468235294117647, "grad_norm": 1.110136420715212, "learning_rate": 0.00029999907762796505, "loss": 0.74, "step": 2012 }, { "epoch": 0.9472941176470588, "grad_norm": 0.44470242353718054, "learning_rate": 0.00029999891749412375, "loss": 0.3806, "step": 2013 }, { "epoch": 0.947764705882353, "grad_norm": 0.34270662673641133, "learning_rate": 0.0002999987445496392, "loss": 0.1699, "step": 2014 }, { "epoch": 0.9482352941176471, "grad_norm": 1.201696647593327, "learning_rate": 0.00029999855879452616, "loss": 0.5928, "step": 2015 }, { "epoch": 0.9487058823529412, "grad_norm": 0.5975998999396662, "learning_rate": 0.00029999836022880045, "loss": 0.5629, "step": 2016 }, { "epoch": 0.9491764705882353, "grad_norm": 1.9152446960745906, "learning_rate": 0.00029999814885247906, "loss": 0.5667, "step": 2017 }, { "epoch": 0.9496470588235294, "grad_norm": 1.3348643089438903, "learning_rate": 0.00029999792466558, "loss": 0.8553, "step": 2018 }, { "epoch": 0.9501176470588235, "grad_norm": 1.5833797818809512, "learning_rate": 0.0002999976876681224, "loss": 0.6337, "step": 2019 }, { "epoch": 0.9505882352941176, "grad_norm": 0.6022911803748947, "learning_rate": 0.00029999743786012656, "loss": 0.1706, "step": 2020 }, { "epoch": 0.9510588235294117, "grad_norm": 3.888616121688205, "learning_rate": 0.0002999971752416138, "loss": 1.3665, "step": 2021 }, { "epoch": 0.9515294117647058, "grad_norm": 1.0992541391089674, "learning_rate": 0.00029999689981260655, "loss": 0.8011, "step": 2022 }, { "epoch": 0.952, "grad_norm": 0.5013150470658452, "learning_rate": 0.0002999966115731283, "loss": 0.1314, "step": 2023 }, { "epoch": 0.9524705882352941, "grad_norm": 2.0243926119055957, "learning_rate": 0.00029999631052320374, "loss": 0.5925, "step": 2024 }, { "epoch": 0.9529411764705882, "grad_norm": 0.7190840174529312, "learning_rate": 0.0002999959966628585, "loss": 0.208, "step": 2025 }, { "epoch": 0.9534117647058824, "grad_norm": 1.7082210299540577, "learning_rate": 0.0002999956699921194, "loss": 0.5107, "step": 2026 }, { "epoch": 0.9538823529411765, "grad_norm": 0.8389185183344465, "learning_rate": 0.0002999953305110144, "loss": 0.3763, "step": 2027 }, { "epoch": 0.9543529411764706, "grad_norm": 1.2827824502536362, "learning_rate": 0.0002999949782195724, "loss": 0.7253, "step": 2028 }, { "epoch": 0.9548235294117647, "grad_norm": 1.481292098390588, "learning_rate": 0.0002999946131178236, "loss": 0.8017, "step": 2029 }, { "epoch": 0.9552941176470588, "grad_norm": 1.5766951392936317, "learning_rate": 0.0002999942352057991, "loss": 0.9457, "step": 2030 }, { "epoch": 0.955764705882353, "grad_norm": 0.5284426313619629, "learning_rate": 0.00029999384448353114, "loss": 0.3606, "step": 2031 }, { "epoch": 0.9562352941176471, "grad_norm": 0.5484775583828929, "learning_rate": 0.00029999344095105326, "loss": 0.3168, "step": 2032 }, { "epoch": 0.9567058823529412, "grad_norm": 1.7419723510055276, "learning_rate": 0.0002999930246083997, "loss": 0.7932, "step": 2033 }, { "epoch": 0.9571764705882353, "grad_norm": 0.3269659456399855, "learning_rate": 0.0002999925954556062, "loss": 0.1587, "step": 2034 }, { "epoch": 0.9576470588235294, "grad_norm": 1.3745477321378665, "learning_rate": 0.0002999921534927093, "loss": 0.5983, "step": 2035 }, { "epoch": 0.9581176470588235, "grad_norm": 0.4849773195143058, "learning_rate": 0.00029999169871974677, "loss": 0.4163, "step": 2036 }, { "epoch": 0.9585882352941176, "grad_norm": 0.738465489298845, "learning_rate": 0.0002999912311367575, "loss": 0.3279, "step": 2037 }, { "epoch": 0.9590588235294117, "grad_norm": 1.596924474960341, "learning_rate": 0.00029999075074378136, "loss": 0.6729, "step": 2038 }, { "epoch": 0.9595294117647059, "grad_norm": 0.9683154883182822, "learning_rate": 0.00029999025754085944, "loss": 0.7435, "step": 2039 }, { "epoch": 0.96, "grad_norm": 0.6603678105027498, "learning_rate": 0.00029998975152803383, "loss": 0.3439, "step": 2040 }, { "epoch": 0.9604705882352941, "grad_norm": 0.8791933340884796, "learning_rate": 0.0002999892327053477, "loss": 0.5549, "step": 2041 }, { "epoch": 0.9609411764705882, "grad_norm": 2.5290486925622546, "learning_rate": 0.0002999887010728454, "loss": 0.9263, "step": 2042 }, { "epoch": 0.9614117647058823, "grad_norm": 0.25665793737474674, "learning_rate": 0.0002999881566305724, "loss": 0.1365, "step": 2043 }, { "epoch": 0.9618823529411765, "grad_norm": 2.394207269946121, "learning_rate": 0.0002999875993785751, "loss": 1.1413, "step": 2044 }, { "epoch": 0.9623529411764706, "grad_norm": 0.9317767531406158, "learning_rate": 0.0002999870293169011, "loss": 0.6743, "step": 2045 }, { "epoch": 0.9628235294117647, "grad_norm": 0.6619954628673101, "learning_rate": 0.00029998644644559916, "loss": 0.5346, "step": 2046 }, { "epoch": 0.9632941176470589, "grad_norm": 1.1775765806232228, "learning_rate": 0.000299985850764719, "loss": 0.6098, "step": 2047 }, { "epoch": 0.963764705882353, "grad_norm": 1.0391677476298178, "learning_rate": 0.00029998524227431146, "loss": 0.8799, "step": 2048 }, { "epoch": 0.9642352941176471, "grad_norm": 0.28184613660836577, "learning_rate": 0.00029998462097442866, "loss": 0.124, "step": 2049 }, { "epoch": 0.9647058823529412, "grad_norm": 0.551880192927129, "learning_rate": 0.0002999839868651235, "loss": 0.1761, "step": 2050 }, { "epoch": 0.9651764705882353, "grad_norm": 0.9273552879676961, "learning_rate": 0.0002999833399464502, "loss": 0.5085, "step": 2051 }, { "epoch": 0.9656470588235294, "grad_norm": 0.6695958430141518, "learning_rate": 0.000299982680218464, "loss": 0.2994, "step": 2052 }, { "epoch": 0.9661176470588235, "grad_norm": 1.0554550733626331, "learning_rate": 0.00029998200768122123, "loss": 0.6552, "step": 2053 }, { "epoch": 0.9665882352941176, "grad_norm": 0.9261309606110987, "learning_rate": 0.00029998132233477936, "loss": 0.6329, "step": 2054 }, { "epoch": 0.9670588235294117, "grad_norm": 0.9773639379549196, "learning_rate": 0.0002999806241791969, "loss": 0.6854, "step": 2055 }, { "epoch": 0.9675294117647059, "grad_norm": 0.9136442524148203, "learning_rate": 0.0002999799132145335, "loss": 0.6414, "step": 2056 }, { "epoch": 0.968, "grad_norm": 1.7420543705573253, "learning_rate": 0.00029997918944084986, "loss": 0.7121, "step": 2057 }, { "epoch": 0.9684705882352941, "grad_norm": 0.8317139357924429, "learning_rate": 0.0002999784528582078, "loss": 0.5779, "step": 2058 }, { "epoch": 0.9689411764705882, "grad_norm": 0.3918025370326948, "learning_rate": 0.0002999777034666702, "loss": 0.3141, "step": 2059 }, { "epoch": 0.9694117647058823, "grad_norm": 1.2800828534407562, "learning_rate": 0.00029997694126630115, "loss": 0.5353, "step": 2060 }, { "epoch": 0.9698823529411764, "grad_norm": 0.13444117975043868, "learning_rate": 0.00029997616625716567, "loss": 0.0431, "step": 2061 }, { "epoch": 0.9703529411764706, "grad_norm": 0.41241843314111565, "learning_rate": 0.0002999753784393299, "loss": 0.2894, "step": 2062 }, { "epoch": 0.9708235294117648, "grad_norm": 0.452269609288911, "learning_rate": 0.00029997457781286125, "loss": 0.3247, "step": 2063 }, { "epoch": 0.9712941176470589, "grad_norm": 0.6639082043860238, "learning_rate": 0.000299973764377828, "loss": 0.5199, "step": 2064 }, { "epoch": 0.971764705882353, "grad_norm": 0.22309327890834724, "learning_rate": 0.00029997293813429965, "loss": 0.1991, "step": 2065 }, { "epoch": 0.9722352941176471, "grad_norm": 0.4426413997130908, "learning_rate": 0.0002999720990823468, "loss": 0.245, "step": 2066 }, { "epoch": 0.9727058823529412, "grad_norm": 0.2644875232627595, "learning_rate": 0.0002999712472220411, "loss": 0.2122, "step": 2067 }, { "epoch": 0.9731764705882353, "grad_norm": 0.7465870277365274, "learning_rate": 0.0002999703825534552, "loss": 0.53, "step": 2068 }, { "epoch": 0.9736470588235294, "grad_norm": 0.4696836414111978, "learning_rate": 0.0002999695050766631, "loss": 0.4391, "step": 2069 }, { "epoch": 0.9741176470588235, "grad_norm": 0.47538824226541326, "learning_rate": 0.0002999686147917396, "loss": 0.5059, "step": 2070 }, { "epoch": 0.9745882352941176, "grad_norm": 0.634224901325582, "learning_rate": 0.0002999677116987609, "loss": 0.6717, "step": 2071 }, { "epoch": 0.9750588235294118, "grad_norm": 0.5324071345130168, "learning_rate": 0.00029996679579780394, "loss": 0.5742, "step": 2072 }, { "epoch": 0.9755294117647059, "grad_norm": 0.5558116579308355, "learning_rate": 0.0002999658670889471, "loss": 0.599, "step": 2073 }, { "epoch": 0.976, "grad_norm": 0.47067067944460517, "learning_rate": 0.0002999649255722696, "loss": 0.464, "step": 2074 }, { "epoch": 0.9764705882352941, "grad_norm": 0.5202462642362726, "learning_rate": 0.0002999639712478519, "loss": 0.3436, "step": 2075 }, { "epoch": 0.9769411764705882, "grad_norm": 0.4763783398810852, "learning_rate": 0.00029996300411577547, "loss": 0.5536, "step": 2076 }, { "epoch": 0.9774117647058823, "grad_norm": 0.47834561380475304, "learning_rate": 0.00029996202417612294, "loss": 0.4188, "step": 2077 }, { "epoch": 0.9778823529411764, "grad_norm": 0.2868161159261352, "learning_rate": 0.000299961031428978, "loss": 0.1346, "step": 2078 }, { "epoch": 0.9783529411764705, "grad_norm": 0.3694000638634789, "learning_rate": 0.00029996002587442537, "loss": 0.3374, "step": 2079 }, { "epoch": 0.9788235294117648, "grad_norm": 0.7718814682313127, "learning_rate": 0.000299959007512551, "loss": 0.4755, "step": 2080 }, { "epoch": 0.9792941176470589, "grad_norm": 0.28414167840940413, "learning_rate": 0.00029995797634344186, "loss": 0.1877, "step": 2081 }, { "epoch": 0.979764705882353, "grad_norm": 0.5030684810396159, "learning_rate": 0.000299956932367186, "loss": 0.5535, "step": 2082 }, { "epoch": 0.9802352941176471, "grad_norm": 0.8882956964423085, "learning_rate": 0.0002999558755838726, "loss": 0.5186, "step": 2083 }, { "epoch": 0.9807058823529412, "grad_norm": 0.40462465522767055, "learning_rate": 0.00029995480599359186, "loss": 0.154, "step": 2084 }, { "epoch": 0.9811764705882353, "grad_norm": 0.4801334419273831, "learning_rate": 0.00029995372359643515, "loss": 0.3528, "step": 2085 }, { "epoch": 0.9816470588235294, "grad_norm": 0.9145486909818276, "learning_rate": 0.00029995262839249497, "loss": 0.5478, "step": 2086 }, { "epoch": 0.9821176470588235, "grad_norm": 0.8812972681296553, "learning_rate": 0.0002999515203818648, "loss": 0.6279, "step": 2087 }, { "epoch": 0.9825882352941177, "grad_norm": 1.194666285126326, "learning_rate": 0.0002999503995646393, "loss": 0.497, "step": 2088 }, { "epoch": 0.9830588235294118, "grad_norm": 0.5141759017289256, "learning_rate": 0.00029994926594091413, "loss": 0.1478, "step": 2089 }, { "epoch": 0.9835294117647059, "grad_norm": 3.3763653072724753, "learning_rate": 0.00029994811951078616, "loss": 0.6302, "step": 2090 }, { "epoch": 0.984, "grad_norm": 0.44425094888109595, "learning_rate": 0.0002999469602743533, "loss": 0.3162, "step": 2091 }, { "epoch": 0.9844705882352941, "grad_norm": 0.5419782579849287, "learning_rate": 0.0002999457882317146, "loss": 0.2809, "step": 2092 }, { "epoch": 0.9849411764705882, "grad_norm": 1.0673720307123735, "learning_rate": 0.00029994460338297, "loss": 0.4259, "step": 2093 }, { "epoch": 0.9854117647058823, "grad_norm": 0.3315184784298938, "learning_rate": 0.0002999434057282209, "loss": 0.1349, "step": 2094 }, { "epoch": 0.9858823529411764, "grad_norm": 0.9371305675705159, "learning_rate": 0.0002999421952675694, "loss": 0.3634, "step": 2095 }, { "epoch": 0.9863529411764705, "grad_norm": 0.5171840024166621, "learning_rate": 0.00029994097200111903, "loss": 0.211, "step": 2096 }, { "epoch": 0.9868235294117647, "grad_norm": 0.37243932598580337, "learning_rate": 0.0002999397359289741, "loss": 0.1675, "step": 2097 }, { "epoch": 0.9872941176470589, "grad_norm": 2.005160103174579, "learning_rate": 0.00029993848705124035, "loss": 0.9995, "step": 2098 }, { "epoch": 0.987764705882353, "grad_norm": 0.6330031753616193, "learning_rate": 0.0002999372253680244, "loss": 0.5888, "step": 2099 }, { "epoch": 0.9882352941176471, "grad_norm": 0.4198434714583396, "learning_rate": 0.0002999359508794339, "loss": 0.181, "step": 2100 }, { "epoch": 0.9887058823529412, "grad_norm": 0.6609322011978163, "learning_rate": 0.0002999346635855778, "loss": 0.1914, "step": 2101 }, { "epoch": 0.9891764705882353, "grad_norm": 2.689450881140823, "learning_rate": 0.00029993336348656594, "loss": 0.831, "step": 2102 }, { "epoch": 0.9896470588235294, "grad_norm": 0.1376775103042258, "learning_rate": 0.00029993205058250947, "loss": 0.0533, "step": 2103 }, { "epoch": 0.9901176470588235, "grad_norm": 0.5916778395303349, "learning_rate": 0.00029993072487352043, "loss": 0.3697, "step": 2104 }, { "epoch": 0.9905882352941177, "grad_norm": 1.0242215679293534, "learning_rate": 0.0002999293863597121, "loss": 0.4865, "step": 2105 }, { "epoch": 0.9910588235294118, "grad_norm": 1.1136801254677462, "learning_rate": 0.00029992803504119876, "loss": 0.5926, "step": 2106 }, { "epoch": 0.9915294117647059, "grad_norm": 0.495771036053292, "learning_rate": 0.0002999266709180959, "loss": 0.2554, "step": 2107 }, { "epoch": 0.992, "grad_norm": 0.4085833876012488, "learning_rate": 0.0002999252939905199, "loss": 0.1437, "step": 2108 }, { "epoch": 0.9924705882352941, "grad_norm": 0.7375074548592836, "learning_rate": 0.0002999239042585884, "loss": 0.337, "step": 2109 }, { "epoch": 0.9929411764705882, "grad_norm": 1.1416119858267555, "learning_rate": 0.0002999225017224201, "loss": 0.4913, "step": 2110 }, { "epoch": 0.9934117647058823, "grad_norm": 0.876911191846862, "learning_rate": 0.0002999210863821349, "loss": 0.2129, "step": 2111 }, { "epoch": 0.9938823529411764, "grad_norm": 1.2859071139354556, "learning_rate": 0.0002999196582378535, "loss": 0.7149, "step": 2112 }, { "epoch": 0.9943529411764706, "grad_norm": 1.0728598009793706, "learning_rate": 0.0002999182172896979, "loss": 0.6133, "step": 2113 }, { "epoch": 0.9948235294117647, "grad_norm": 0.37062384092109907, "learning_rate": 0.0002999167635377912, "loss": 0.0992, "step": 2114 }, { "epoch": 0.9952941176470588, "grad_norm": 0.25166538650470455, "learning_rate": 0.00029991529698225757, "loss": 0.1476, "step": 2115 }, { "epoch": 0.995764705882353, "grad_norm": 1.2992108273699488, "learning_rate": 0.0002999138176232223, "loss": 0.5774, "step": 2116 }, { "epoch": 0.9962352941176471, "grad_norm": 0.7718196175901528, "learning_rate": 0.00029991232546081156, "loss": 0.4133, "step": 2117 }, { "epoch": 0.9967058823529412, "grad_norm": 0.644724933422021, "learning_rate": 0.000299910820495153, "loss": 0.4749, "step": 2118 }, { "epoch": 0.9971764705882353, "grad_norm": 0.39074855707520045, "learning_rate": 0.00029990930272637506, "loss": 0.3097, "step": 2119 }, { "epoch": 0.9976470588235294, "grad_norm": 1.4827929278882073, "learning_rate": 0.00029990777215460737, "loss": 1.0481, "step": 2120 }, { "epoch": 0.9981176470588236, "grad_norm": 0.7554150963261933, "learning_rate": 0.0002999062287799806, "loss": 0.6179, "step": 2121 }, { "epoch": 0.9985882352941177, "grad_norm": 0.6135814964990327, "learning_rate": 0.0002999046726026266, "loss": 0.3677, "step": 2122 }, { "epoch": 0.9990588235294118, "grad_norm": 0.7920443888958754, "learning_rate": 0.00029990310362267833, "loss": 0.5885, "step": 2123 }, { "epoch": 0.9995294117647059, "grad_norm": 0.599317882439723, "learning_rate": 0.0002999015218402697, "loss": 0.3715, "step": 2124 }, { "epoch": 1.0, "grad_norm": 0.8512774410656933, "learning_rate": 0.0002998999272555359, "loss": 0.7069, "step": 2125 }, { "epoch": 1.0, "eval_loss": 0.4317689538002014, "eval_runtime": 65.1415, "eval_samples_per_second": 11.513, "eval_steps_per_second": 5.757, "step": 2125 }, { "epoch": 1.0004705882352942, "grad_norm": 0.2790585496086798, "learning_rate": 0.000299898319868613, "loss": 0.1628, "step": 2126 }, { "epoch": 1.0009411764705882, "grad_norm": 0.5351668919514218, "learning_rate": 0.00029989669967963835, "loss": 0.4898, "step": 2127 }, { "epoch": 1.0014117647058824, "grad_norm": 0.46501323476527795, "learning_rate": 0.00029989506668875033, "loss": 0.2861, "step": 2128 }, { "epoch": 1.0018823529411764, "grad_norm": 0.48450737088558365, "learning_rate": 0.00029989342089608835, "loss": 0.2906, "step": 2129 }, { "epoch": 1.0023529411764707, "grad_norm": 0.36707652951235037, "learning_rate": 0.000299891762301793, "loss": 0.1857, "step": 2130 }, { "epoch": 1.0028235294117647, "grad_norm": 0.9030962615585265, "learning_rate": 0.00029989009090600595, "loss": 0.5593, "step": 2131 }, { "epoch": 1.003294117647059, "grad_norm": 0.8333014450134161, "learning_rate": 0.00029988840670887, "loss": 0.4639, "step": 2132 }, { "epoch": 1.003764705882353, "grad_norm": 0.5879204625344215, "learning_rate": 0.0002998867097105288, "loss": 0.6548, "step": 2133 }, { "epoch": 1.0042352941176471, "grad_norm": 0.6371682675412547, "learning_rate": 0.00029988499991112747, "loss": 0.4664, "step": 2134 }, { "epoch": 1.0047058823529411, "grad_norm": 1.4761998442434616, "learning_rate": 0.000299883277310812, "loss": 0.5886, "step": 2135 }, { "epoch": 1.0051764705882353, "grad_norm": 0.3360445025088113, "learning_rate": 0.0002998815419097294, "loss": 0.0806, "step": 2136 }, { "epoch": 1.0056470588235293, "grad_norm": 0.8666723715869502, "learning_rate": 0.0002998797937080279, "loss": 0.537, "step": 2137 }, { "epoch": 1.0061176470588236, "grad_norm": 0.49652691043087743, "learning_rate": 0.000299878032705857, "loss": 0.3943, "step": 2138 }, { "epoch": 1.0065882352941176, "grad_norm": 0.34992674622109843, "learning_rate": 0.0002998762589033669, "loss": 0.2962, "step": 2139 }, { "epoch": 1.0070588235294118, "grad_norm": 0.9795360984346474, "learning_rate": 0.00029987447230070917, "loss": 0.6061, "step": 2140 }, { "epoch": 1.0075294117647058, "grad_norm": 1.1515617200968176, "learning_rate": 0.00029987267289803635, "loss": 0.8437, "step": 2141 }, { "epoch": 1.008, "grad_norm": 0.36016240617146406, "learning_rate": 0.00029987086069550216, "loss": 0.3451, "step": 2142 }, { "epoch": 1.008470588235294, "grad_norm": 0.4899308503813044, "learning_rate": 0.0002998690356932613, "loss": 0.2692, "step": 2143 }, { "epoch": 1.0089411764705882, "grad_norm": 1.0545514791325652, "learning_rate": 0.0002998671978914698, "loss": 0.5154, "step": 2144 }, { "epoch": 1.0094117647058825, "grad_norm": 0.058200380756832155, "learning_rate": 0.00029986534729028444, "loss": 0.044, "step": 2145 }, { "epoch": 1.0098823529411765, "grad_norm": 0.8015072915580104, "learning_rate": 0.00029986348388986333, "loss": 0.6092, "step": 2146 }, { "epoch": 1.0103529411764707, "grad_norm": 0.6599353358921217, "learning_rate": 0.00029986160769036565, "loss": 0.5154, "step": 2147 }, { "epoch": 1.0108235294117647, "grad_norm": 0.5575292670505513, "learning_rate": 0.0002998597186919515, "loss": 0.4779, "step": 2148 }, { "epoch": 1.011294117647059, "grad_norm": 0.1628212704681539, "learning_rate": 0.0002998578168947824, "loss": 0.0582, "step": 2149 }, { "epoch": 1.011764705882353, "grad_norm": 0.14670704931752399, "learning_rate": 0.00029985590229902073, "loss": 0.0618, "step": 2150 }, { "epoch": 1.0122352941176471, "grad_norm": 0.6704182889929426, "learning_rate": 0.0002998539749048299, "loss": 0.617, "step": 2151 }, { "epoch": 1.0127058823529411, "grad_norm": 0.32773818336961175, "learning_rate": 0.0002998520347123746, "loss": 0.3074, "step": 2152 }, { "epoch": 1.0131764705882353, "grad_norm": 0.21533524918912908, "learning_rate": 0.00029985008172182055, "loss": 0.1325, "step": 2153 }, { "epoch": 1.0136470588235293, "grad_norm": 2.027132023065713, "learning_rate": 0.00029984811593333446, "loss": 0.8109, "step": 2154 }, { "epoch": 1.0141176470588236, "grad_norm": 0.8186787265124332, "learning_rate": 0.00029984613734708427, "loss": 0.6769, "step": 2155 }, { "epoch": 1.0145882352941176, "grad_norm": 0.5561117808940803, "learning_rate": 0.00029984414596323896, "loss": 0.2543, "step": 2156 }, { "epoch": 1.0150588235294118, "grad_norm": 1.7010787769258886, "learning_rate": 0.00029984214178196863, "loss": 0.5847, "step": 2157 }, { "epoch": 1.0155294117647058, "grad_norm": 0.4336731292842373, "learning_rate": 0.00029984012480344444, "loss": 0.2119, "step": 2158 }, { "epoch": 1.016, "grad_norm": 0.44067938538062273, "learning_rate": 0.0002998380950278386, "loss": 0.3795, "step": 2159 }, { "epoch": 1.016470588235294, "grad_norm": 0.4534424762704804, "learning_rate": 0.00029983605245532453, "loss": 0.2657, "step": 2160 }, { "epoch": 1.0169411764705882, "grad_norm": 0.5094161835234958, "learning_rate": 0.0002998339970860766, "loss": 0.1941, "step": 2161 }, { "epoch": 1.0174117647058825, "grad_norm": 1.1255209803287383, "learning_rate": 0.00029983192892027037, "loss": 0.7173, "step": 2162 }, { "epoch": 1.0178823529411765, "grad_norm": 0.3596966953060714, "learning_rate": 0.0002998298479580825, "loss": 0.2897, "step": 2163 }, { "epoch": 1.0183529411764707, "grad_norm": 1.7100524640589, "learning_rate": 0.0002998277541996907, "loss": 1.0389, "step": 2164 }, { "epoch": 1.0188235294117647, "grad_norm": 1.5318179518360981, "learning_rate": 0.0002998256476452738, "loss": 0.6229, "step": 2165 }, { "epoch": 1.019294117647059, "grad_norm": 11.103294534806391, "learning_rate": 0.0002998235282950117, "loss": 1.1271, "step": 2166 }, { "epoch": 1.019764705882353, "grad_norm": 0.6555576644446716, "learning_rate": 0.0002998213961490854, "loss": 0.4812, "step": 2167 }, { "epoch": 1.0202352941176471, "grad_norm": 0.21324334877565163, "learning_rate": 0.00029981925120767705, "loss": 0.0846, "step": 2168 }, { "epoch": 1.0207058823529411, "grad_norm": 1.4202707312422282, "learning_rate": 0.00029981709347096975, "loss": 0.4237, "step": 2169 }, { "epoch": 1.0211764705882354, "grad_norm": 3.242776991954895, "learning_rate": 0.00029981492293914777, "loss": 0.5426, "step": 2170 }, { "epoch": 1.0216470588235294, "grad_norm": 1.597936416597401, "learning_rate": 0.0002998127396123966, "loss": 0.5262, "step": 2171 }, { "epoch": 1.0221176470588236, "grad_norm": 1.0012757784447464, "learning_rate": 0.00029981054349090264, "loss": 0.6254, "step": 2172 }, { "epoch": 1.0225882352941176, "grad_norm": 1.6309686573158313, "learning_rate": 0.0002998083345748534, "loss": 0.7423, "step": 2173 }, { "epoch": 1.0230588235294118, "grad_norm": 0.793971571177349, "learning_rate": 0.0002998061128644376, "loss": 0.5728, "step": 2174 }, { "epoch": 1.0235294117647058, "grad_norm": 0.4263161735996386, "learning_rate": 0.00029980387835984494, "loss": 0.1201, "step": 2175 }, { "epoch": 1.024, "grad_norm": 0.15094361351468805, "learning_rate": 0.0002998016310612663, "loss": 0.0432, "step": 2176 }, { "epoch": 1.024470588235294, "grad_norm": 0.6115074207342477, "learning_rate": 0.0002997993709688937, "loss": 0.2118, "step": 2177 }, { "epoch": 1.0249411764705882, "grad_norm": 1.1792320892593344, "learning_rate": 0.0002997970980829199, "loss": 0.5107, "step": 2178 }, { "epoch": 1.0254117647058822, "grad_norm": 0.7387929020549926, "learning_rate": 0.0002997948124035393, "loss": 0.3591, "step": 2179 }, { "epoch": 1.0258823529411765, "grad_norm": 0.4431578384852145, "learning_rate": 0.00029979251393094685, "loss": 0.255, "step": 2180 }, { "epoch": 1.0263529411764707, "grad_norm": 0.31253462716123526, "learning_rate": 0.00029979020266533913, "loss": 0.2051, "step": 2181 }, { "epoch": 1.0268235294117647, "grad_norm": 0.3363502943401131, "learning_rate": 0.0002997878786069133, "loss": 0.2331, "step": 2182 }, { "epoch": 1.027294117647059, "grad_norm": 1.4781954486688411, "learning_rate": 0.0002997855417558679, "loss": 0.7465, "step": 2183 }, { "epoch": 1.027764705882353, "grad_norm": 0.3765209580932073, "learning_rate": 0.0002997831921124026, "loss": 0.2955, "step": 2184 }, { "epoch": 1.0282352941176471, "grad_norm": 0.4711430925109247, "learning_rate": 0.000299780829676718, "loss": 0.3522, "step": 2185 }, { "epoch": 1.0287058823529411, "grad_norm": 0.2223122421969247, "learning_rate": 0.00029977845444901586, "loss": 0.1515, "step": 2186 }, { "epoch": 1.0291764705882354, "grad_norm": 0.9679197426765969, "learning_rate": 0.0002997760664294991, "loss": 0.6834, "step": 2187 }, { "epoch": 1.0296470588235294, "grad_norm": 0.4286597096569831, "learning_rate": 0.00029977366561837154, "loss": 0.4328, "step": 2188 }, { "epoch": 1.0301176470588236, "grad_norm": 0.25218893482113663, "learning_rate": 0.0002997712520158383, "loss": 0.1688, "step": 2189 }, { "epoch": 1.0305882352941176, "grad_norm": 0.8529691145493493, "learning_rate": 0.00029976882562210563, "loss": 1.0294, "step": 2190 }, { "epoch": 1.0310588235294118, "grad_norm": 0.28998524046899643, "learning_rate": 0.00029976638643738055, "loss": 0.2782, "step": 2191 }, { "epoch": 1.0315294117647058, "grad_norm": 0.4285344492840604, "learning_rate": 0.0002997639344618714, "loss": 0.3727, "step": 2192 }, { "epoch": 1.032, "grad_norm": 0.5812695668501601, "learning_rate": 0.0002997614696957878, "loss": 0.3646, "step": 2193 }, { "epoch": 1.032470588235294, "grad_norm": 0.33022702613936783, "learning_rate": 0.00029975899213934004, "loss": 0.1496, "step": 2194 }, { "epoch": 1.0329411764705883, "grad_norm": 0.6835327500873928, "learning_rate": 0.0002997565017927398, "loss": 0.4472, "step": 2195 }, { "epoch": 1.0334117647058823, "grad_norm": 0.704119472889384, "learning_rate": 0.00029975399865619977, "loss": 0.4681, "step": 2196 }, { "epoch": 1.0338823529411765, "grad_norm": 0.28311498355888115, "learning_rate": 0.00029975148272993373, "loss": 0.265, "step": 2197 }, { "epoch": 1.0343529411764707, "grad_norm": 0.4434497497736499, "learning_rate": 0.0002997489540141565, "loss": 0.2095, "step": 2198 }, { "epoch": 1.0348235294117647, "grad_norm": 1.2546393458833107, "learning_rate": 0.00029974641250908406, "loss": 0.6181, "step": 2199 }, { "epoch": 1.035294117647059, "grad_norm": 1.1691161156465188, "learning_rate": 0.0002997438582149335, "loss": 0.7862, "step": 2200 }, { "epoch": 1.035764705882353, "grad_norm": 0.35506860546640806, "learning_rate": 0.000299741291131923, "loss": 0.3418, "step": 2201 }, { "epoch": 1.0362352941176471, "grad_norm": 1.4364263459485, "learning_rate": 0.00029973871126027173, "loss": 0.6205, "step": 2202 }, { "epoch": 1.0367058823529411, "grad_norm": 0.2822502284277718, "learning_rate": 0.00029973611860020013, "loss": 0.163, "step": 2203 }, { "epoch": 1.0371764705882354, "grad_norm": 1.6989228987966527, "learning_rate": 0.0002997335131519294, "loss": 0.8424, "step": 2204 }, { "epoch": 1.0376470588235294, "grad_norm": 0.41337925699369166, "learning_rate": 0.0002997308949156823, "loss": 0.3357, "step": 2205 }, { "epoch": 1.0381176470588236, "grad_norm": 0.8344527873713542, "learning_rate": 0.00029972826389168235, "loss": 0.3242, "step": 2206 }, { "epoch": 1.0385882352941176, "grad_norm": 0.13604631052926885, "learning_rate": 0.00029972562008015425, "loss": 0.0517, "step": 2207 }, { "epoch": 1.0390588235294118, "grad_norm": 0.9293211454730015, "learning_rate": 0.0002997229634813238, "loss": 0.3694, "step": 2208 }, { "epoch": 1.0395294117647058, "grad_norm": 0.4027493370084004, "learning_rate": 0.0002997202940954179, "loss": 0.1715, "step": 2209 }, { "epoch": 1.04, "grad_norm": 0.6098448852058127, "learning_rate": 0.00029971761192266447, "loss": 0.2429, "step": 2210 }, { "epoch": 1.040470588235294, "grad_norm": 0.5616379329839639, "learning_rate": 0.0002997149169632926, "loss": 0.231, "step": 2211 }, { "epoch": 1.0409411764705883, "grad_norm": 0.26752761153552634, "learning_rate": 0.0002997122092175325, "loss": 0.0508, "step": 2212 }, { "epoch": 1.0414117647058823, "grad_norm": 0.95451173025361, "learning_rate": 0.00029970948868561547, "loss": 0.4412, "step": 2213 }, { "epoch": 1.0418823529411765, "grad_norm": 1.282976644594726, "learning_rate": 0.00029970675536777365, "loss": 0.5352, "step": 2214 }, { "epoch": 1.0423529411764705, "grad_norm": 0.8837832894702645, "learning_rate": 0.0002997040092642407, "loss": 0.6341, "step": 2215 }, { "epoch": 1.0428235294117647, "grad_norm": 0.7218831499878569, "learning_rate": 0.00029970125037525106, "loss": 0.4127, "step": 2216 }, { "epoch": 1.043294117647059, "grad_norm": 0.35296705460052874, "learning_rate": 0.00029969847870104033, "loss": 0.1872, "step": 2217 }, { "epoch": 1.043764705882353, "grad_norm": 0.41200464909357215, "learning_rate": 0.00029969569424184527, "loss": 0.3032, "step": 2218 }, { "epoch": 1.0442352941176472, "grad_norm": 1.1126055990630888, "learning_rate": 0.0002996928969979037, "loss": 0.5289, "step": 2219 }, { "epoch": 1.0447058823529412, "grad_norm": 0.9885849805278336, "learning_rate": 0.00029969008696945445, "loss": 0.5624, "step": 2220 }, { "epoch": 1.0451764705882354, "grad_norm": 1.356052170614805, "learning_rate": 0.00029968726415673753, "loss": 1.6956, "step": 2221 }, { "epoch": 1.0456470588235294, "grad_norm": 0.6284168271802983, "learning_rate": 0.00029968442855999406, "loss": 0.388, "step": 2222 }, { "epoch": 1.0461176470588236, "grad_norm": 0.7600007849639433, "learning_rate": 0.0002996815801794662, "loss": 0.4555, "step": 2223 }, { "epoch": 1.0465882352941176, "grad_norm": 1.0926250233936698, "learning_rate": 0.0002996787190153972, "loss": 0.6043, "step": 2224 }, { "epoch": 1.0470588235294118, "grad_norm": 0.631488735427337, "learning_rate": 0.0002996758450680314, "loss": 0.431, "step": 2225 }, { "epoch": 1.0475294117647058, "grad_norm": 0.21098757142936322, "learning_rate": 0.0002996729583376143, "loss": 0.1678, "step": 2226 }, { "epoch": 1.048, "grad_norm": 0.3389326964933691, "learning_rate": 0.00029967005882439244, "loss": 0.2831, "step": 2227 }, { "epoch": 1.048470588235294, "grad_norm": 0.5100004449348965, "learning_rate": 0.0002996671465286134, "loss": 0.3772, "step": 2228 }, { "epoch": 1.0489411764705883, "grad_norm": 0.10173260727117883, "learning_rate": 0.00029966422145052594, "loss": 0.0291, "step": 2229 }, { "epoch": 1.0494117647058823, "grad_norm": 0.7947305717115826, "learning_rate": 0.00029966128359037986, "loss": 0.6325, "step": 2230 }, { "epoch": 1.0498823529411765, "grad_norm": 1.0492516344774396, "learning_rate": 0.0002996583329484261, "loss": 0.5276, "step": 2231 }, { "epoch": 1.0503529411764705, "grad_norm": 0.46982429541682263, "learning_rate": 0.00029965536952491665, "loss": 0.3983, "step": 2232 }, { "epoch": 1.0508235294117647, "grad_norm": 0.7645510623969521, "learning_rate": 0.00029965239332010454, "loss": 0.8084, "step": 2233 }, { "epoch": 1.0512941176470587, "grad_norm": 0.21269246899880803, "learning_rate": 0.000299649404334244, "loss": 0.1107, "step": 2234 }, { "epoch": 1.051764705882353, "grad_norm": 0.7048137750055661, "learning_rate": 0.0002996464025675903, "loss": 0.5353, "step": 2235 }, { "epoch": 1.0522352941176472, "grad_norm": 0.2951035534080629, "learning_rate": 0.00029964338802039987, "loss": 0.1707, "step": 2236 }, { "epoch": 1.0527058823529412, "grad_norm": 0.5622522979775441, "learning_rate": 0.00029964036069293006, "loss": 0.2815, "step": 2237 }, { "epoch": 1.0531764705882354, "grad_norm": 0.3859227341896941, "learning_rate": 0.0002996373205854395, "loss": 0.0552, "step": 2238 }, { "epoch": 1.0536470588235294, "grad_norm": 0.5358174991700081, "learning_rate": 0.00029963426769818774, "loss": 0.2598, "step": 2239 }, { "epoch": 1.0541176470588236, "grad_norm": 1.082924220228683, "learning_rate": 0.0002996312020314356, "loss": 0.7142, "step": 2240 }, { "epoch": 1.0545882352941176, "grad_norm": 0.9064869706810101, "learning_rate": 0.00029962812358544485, "loss": 0.7956, "step": 2241 }, { "epoch": 1.0550588235294118, "grad_norm": 1.084126241158861, "learning_rate": 0.0002996250323604785, "loss": 0.9526, "step": 2242 }, { "epoch": 1.0555294117647058, "grad_norm": 0.00015695005366875362, "learning_rate": 0.00029962192835680037, "loss": 0.0, "step": 2243 }, { "epoch": 1.056, "grad_norm": 0.7754200431592736, "learning_rate": 0.0002996188115746757, "loss": 0.5617, "step": 2244 }, { "epoch": 1.056470588235294, "grad_norm": 0.7217044392382295, "learning_rate": 0.0002996156820143707, "loss": 0.553, "step": 2245 }, { "epoch": 1.0569411764705883, "grad_norm": 0.29826662994724656, "learning_rate": 0.00029961253967615255, "loss": 0.1882, "step": 2246 }, { "epoch": 1.0574117647058823, "grad_norm": 0.9609918439054972, "learning_rate": 0.0002996093845602897, "loss": 0.5381, "step": 2247 }, { "epoch": 1.0578823529411765, "grad_norm": 0.9284906448017886, "learning_rate": 0.00029960621666705155, "loss": 0.5021, "step": 2248 }, { "epoch": 1.0583529411764705, "grad_norm": 0.2075334417889008, "learning_rate": 0.00029960303599670866, "loss": 0.1387, "step": 2249 }, { "epoch": 1.0588235294117647, "grad_norm": 0.3938497400654798, "learning_rate": 0.0002995998425495327, "loss": 0.4285, "step": 2250 }, { "epoch": 1.0592941176470587, "grad_norm": 0.3250692452488272, "learning_rate": 0.0002995966363257964, "loss": 0.1252, "step": 2251 }, { "epoch": 1.059764705882353, "grad_norm": 1.1126642078389632, "learning_rate": 0.0002995934173257736, "loss": 0.7487, "step": 2252 }, { "epoch": 1.0602352941176472, "grad_norm": 1.0832862220204063, "learning_rate": 0.0002995901855497393, "loss": 0.8774, "step": 2253 }, { "epoch": 1.0607058823529412, "grad_norm": 1.0795985392642011, "learning_rate": 0.00029958694099796934, "loss": 0.7335, "step": 2254 }, { "epoch": 1.0611764705882354, "grad_norm": 0.5890159526511389, "learning_rate": 0.0002995836836707409, "loss": 0.4617, "step": 2255 }, { "epoch": 1.0616470588235294, "grad_norm": 0.38218180511677163, "learning_rate": 0.00029958041356833217, "loss": 0.3213, "step": 2256 }, { "epoch": 1.0621176470588236, "grad_norm": 0.5751846027153851, "learning_rate": 0.00029957713069102244, "loss": 0.1616, "step": 2257 }, { "epoch": 1.0625882352941176, "grad_norm": 0.3349919164738495, "learning_rate": 0.00029957383503909204, "loss": 0.3012, "step": 2258 }, { "epoch": 1.0630588235294118, "grad_norm": 0.41753361410392065, "learning_rate": 0.00029957052661282256, "loss": 0.2609, "step": 2259 }, { "epoch": 1.0635294117647058, "grad_norm": 0.6096303579581535, "learning_rate": 0.0002995672054124964, "loss": 0.3576, "step": 2260 }, { "epoch": 1.064, "grad_norm": 0.3024496277023092, "learning_rate": 0.0002995638714383973, "loss": 0.2477, "step": 2261 }, { "epoch": 1.064470588235294, "grad_norm": 0.634379918369373, "learning_rate": 0.00029956052469080994, "loss": 0.4589, "step": 2262 }, { "epoch": 1.0649411764705883, "grad_norm": 0.7843957795533778, "learning_rate": 0.0002995571651700202, "loss": 0.6155, "step": 2263 }, { "epoch": 1.0654117647058823, "grad_norm": 0.43524240764395467, "learning_rate": 0.000299553792876315, "loss": 0.3826, "step": 2264 }, { "epoch": 1.0658823529411765, "grad_norm": 0.05906577611195651, "learning_rate": 0.0002995504078099823, "loss": 0.0149, "step": 2265 }, { "epoch": 1.0663529411764705, "grad_norm": 0.4876368394071849, "learning_rate": 0.00029954700997131126, "loss": 0.2276, "step": 2266 }, { "epoch": 1.0668235294117647, "grad_norm": 0.5283539799149993, "learning_rate": 0.000299543599360592, "loss": 0.306, "step": 2267 }, { "epoch": 1.0672941176470587, "grad_norm": 0.306656364524564, "learning_rate": 0.0002995401759781159, "loss": 0.2534, "step": 2268 }, { "epoch": 1.067764705882353, "grad_norm": 0.635169273152903, "learning_rate": 0.00029953673982417527, "loss": 0.5459, "step": 2269 }, { "epoch": 1.0682352941176472, "grad_norm": 0.691590239896948, "learning_rate": 0.00029953329089906357, "loss": 0.3601, "step": 2270 }, { "epoch": 1.0687058823529412, "grad_norm": 0.6288699209727696, "learning_rate": 0.0002995298292030754, "loss": 0.4806, "step": 2271 }, { "epoch": 1.0691764705882354, "grad_norm": 0.2730600566933975, "learning_rate": 0.00029952635473650633, "loss": 0.2588, "step": 2272 }, { "epoch": 1.0696470588235294, "grad_norm": 0.4531164637427404, "learning_rate": 0.00029952286749965315, "loss": 0.4781, "step": 2273 }, { "epoch": 1.0701176470588236, "grad_norm": 0.2512843678468584, "learning_rate": 0.0002995193674928137, "loss": 0.2387, "step": 2274 }, { "epoch": 1.0705882352941176, "grad_norm": 0.6704019948129952, "learning_rate": 0.0002995158547162869, "loss": 0.6713, "step": 2275 }, { "epoch": 1.0710588235294118, "grad_norm": 0.9440188883027659, "learning_rate": 0.00029951232917037264, "loss": 0.805, "step": 2276 }, { "epoch": 1.0715294117647058, "grad_norm": 0.18787009712269268, "learning_rate": 0.0002995087908553722, "loss": 0.1444, "step": 2277 }, { "epoch": 1.072, "grad_norm": 0.34015378086710074, "learning_rate": 0.00029950523977158766, "loss": 0.0798, "step": 2278 }, { "epoch": 1.072470588235294, "grad_norm": 1.358761360387305, "learning_rate": 0.00029950167591932233, "loss": 0.7904, "step": 2279 }, { "epoch": 1.0729411764705883, "grad_norm": 0.6129457846550697, "learning_rate": 0.0002994980992988806, "loss": 0.5669, "step": 2280 }, { "epoch": 1.0734117647058823, "grad_norm": 0.3278322942136272, "learning_rate": 0.0002994945099105679, "loss": 0.2959, "step": 2281 }, { "epoch": 1.0738823529411765, "grad_norm": 0.300317740061207, "learning_rate": 0.00029949090775469077, "loss": 0.2055, "step": 2282 }, { "epoch": 1.0743529411764705, "grad_norm": 1.1438335020224135, "learning_rate": 0.00029948729283155683, "loss": 0.6486, "step": 2283 }, { "epoch": 1.0748235294117647, "grad_norm": 0.4152780521917976, "learning_rate": 0.0002994836651414749, "loss": 0.2773, "step": 2284 }, { "epoch": 1.0752941176470587, "grad_norm": 1.0273700315666006, "learning_rate": 0.00029948002468475475, "loss": 0.8997, "step": 2285 }, { "epoch": 1.075764705882353, "grad_norm": 0.4930542393796383, "learning_rate": 0.0002994763714617073, "loss": 0.5948, "step": 2286 }, { "epoch": 1.076235294117647, "grad_norm": 0.4625840930601747, "learning_rate": 0.00029947270547264456, "loss": 0.2198, "step": 2287 }, { "epoch": 1.0767058823529412, "grad_norm": 0.2675728166302133, "learning_rate": 0.00029946902671787955, "loss": 0.2306, "step": 2288 }, { "epoch": 1.0771764705882352, "grad_norm": 0.7984253971126934, "learning_rate": 0.00029946533519772657, "loss": 0.5531, "step": 2289 }, { "epoch": 1.0776470588235294, "grad_norm": 0.4488283181988333, "learning_rate": 0.0002994616309125008, "loss": 0.328, "step": 2290 }, { "epoch": 1.0781176470588236, "grad_norm": 0.3997933579348744, "learning_rate": 0.0002994579138625187, "loss": 0.3441, "step": 2291 }, { "epoch": 1.0785882352941176, "grad_norm": 0.5058707052163255, "learning_rate": 0.00029945418404809767, "loss": 0.3736, "step": 2292 }, { "epoch": 1.0790588235294118, "grad_norm": 0.9474462379653211, "learning_rate": 0.0002994504414695562, "loss": 0.6719, "step": 2293 }, { "epoch": 1.0795294117647058, "grad_norm": 0.4908583644992186, "learning_rate": 0.000299446686127214, "loss": 0.5023, "step": 2294 }, { "epoch": 1.08, "grad_norm": 0.32252873455550357, "learning_rate": 0.00029944291802139184, "loss": 0.2374, "step": 2295 }, { "epoch": 1.080470588235294, "grad_norm": 0.8172013279636942, "learning_rate": 0.00029943913715241137, "loss": 0.4748, "step": 2296 }, { "epoch": 1.0809411764705883, "grad_norm": 1.6267942855000221, "learning_rate": 0.0002994353435205956, "loss": 0.8422, "step": 2297 }, { "epoch": 1.0814117647058823, "grad_norm": 0.2062175003556457, "learning_rate": 0.00029943153712626855, "loss": 0.1241, "step": 2298 }, { "epoch": 1.0818823529411765, "grad_norm": 0.24982714726472327, "learning_rate": 0.0002994277179697553, "loss": 0.2129, "step": 2299 }, { "epoch": 1.0823529411764705, "grad_norm": 0.2025184271620619, "learning_rate": 0.000299423886051382, "loss": 0.1379, "step": 2300 }, { "epoch": 1.0828235294117647, "grad_norm": 1.8290863538838664, "learning_rate": 0.00029942004137147586, "loss": 0.737, "step": 2301 }, { "epoch": 1.0832941176470587, "grad_norm": 0.7473258719734914, "learning_rate": 0.0002994161839303653, "loss": 0.2187, "step": 2302 }, { "epoch": 1.083764705882353, "grad_norm": 0.7675042078767573, "learning_rate": 0.00029941231372837974, "loss": 0.4985, "step": 2303 }, { "epoch": 1.084235294117647, "grad_norm": 0.14342430977529239, "learning_rate": 0.0002994084307658497, "loss": 0.0462, "step": 2304 }, { "epoch": 1.0847058823529412, "grad_norm": 0.5708362036034653, "learning_rate": 0.0002994045350431069, "loss": 0.3581, "step": 2305 }, { "epoch": 1.0851764705882352, "grad_norm": 0.9142643236056572, "learning_rate": 0.00029940062656048396, "loss": 0.4053, "step": 2306 }, { "epoch": 1.0856470588235294, "grad_norm": 0.2566764312906741, "learning_rate": 0.0002993967053183147, "loss": 0.1319, "step": 2307 }, { "epoch": 1.0861176470588236, "grad_norm": 0.8824950797285694, "learning_rate": 0.000299392771316934, "loss": 0.3763, "step": 2308 }, { "epoch": 1.0865882352941176, "grad_norm": 0.5924335903309149, "learning_rate": 0.00029938882455667786, "loss": 0.5111, "step": 2309 }, { "epoch": 1.0870588235294119, "grad_norm": 0.4073496314768063, "learning_rate": 0.00029938486503788337, "loss": 0.2873, "step": 2310 }, { "epoch": 1.0875294117647059, "grad_norm": 0.4667802739045278, "learning_rate": 0.0002993808927608887, "loss": 0.4563, "step": 2311 }, { "epoch": 1.088, "grad_norm": 0.5589042311099172, "learning_rate": 0.000299376907726033, "loss": 0.4855, "step": 2312 }, { "epoch": 1.088470588235294, "grad_norm": 0.5169963296794957, "learning_rate": 0.0002993729099336568, "loss": 0.3624, "step": 2313 }, { "epoch": 1.0889411764705883, "grad_norm": 0.2113340551099535, "learning_rate": 0.00029936889938410137, "loss": 0.1255, "step": 2314 }, { "epoch": 1.0894117647058823, "grad_norm": 0.5061837074253649, "learning_rate": 0.0002993648760777092, "loss": 0.4246, "step": 2315 }, { "epoch": 1.0898823529411765, "grad_norm": 0.8759510586909465, "learning_rate": 0.0002993608400148241, "loss": 0.7464, "step": 2316 }, { "epoch": 1.0903529411764705, "grad_norm": 0.3924631896698962, "learning_rate": 0.00029935679119579066, "loss": 0.396, "step": 2317 }, { "epoch": 1.0908235294117647, "grad_norm": 0.3165540620882752, "learning_rate": 0.0002993527296209546, "loss": 0.1978, "step": 2318 }, { "epoch": 1.0912941176470587, "grad_norm": 0.18755276108931648, "learning_rate": 0.0002993486552906629, "loss": 0.1638, "step": 2319 }, { "epoch": 1.091764705882353, "grad_norm": 0.1481374327852422, "learning_rate": 0.00029934456820526344, "loss": 0.0857, "step": 2320 }, { "epoch": 1.092235294117647, "grad_norm": 0.4191794312249747, "learning_rate": 0.00029934046836510537, "loss": 0.3494, "step": 2321 }, { "epoch": 1.0927058823529412, "grad_norm": 0.6129736551252828, "learning_rate": 0.00029933635577053877, "loss": 0.2397, "step": 2322 }, { "epoch": 1.0931764705882352, "grad_norm": 0.633005000050918, "learning_rate": 0.00029933223042191487, "loss": 0.892, "step": 2323 }, { "epoch": 1.0936470588235294, "grad_norm": 0.6071470597747622, "learning_rate": 0.000299328092319586, "loss": 0.6933, "step": 2324 }, { "epoch": 1.0941176470588236, "grad_norm": 0.5675040117842676, "learning_rate": 0.0002993239414639057, "loss": 0.3953, "step": 2325 }, { "epoch": 1.0945882352941176, "grad_norm": 0.6877536601970277, "learning_rate": 0.0002993197778552283, "loss": 0.5103, "step": 2326 }, { "epoch": 1.0950588235294119, "grad_norm": 0.7257751749852572, "learning_rate": 0.0002993156014939095, "loss": 0.768, "step": 2327 }, { "epoch": 1.0955294117647059, "grad_norm": 0.28060108095930525, "learning_rate": 0.00029931141238030587, "loss": 0.2185, "step": 2328 }, { "epoch": 1.096, "grad_norm": 0.12046101132362155, "learning_rate": 0.00029930721051477533, "loss": 0.0662, "step": 2329 }, { "epoch": 1.096470588235294, "grad_norm": 1.4394884059038326, "learning_rate": 0.0002993029958976766, "loss": 0.993, "step": 2330 }, { "epoch": 1.0969411764705883, "grad_norm": 0.27512673664586745, "learning_rate": 0.00029929876852936973, "loss": 0.1293, "step": 2331 }, { "epoch": 1.0974117647058823, "grad_norm": 0.8271456051272957, "learning_rate": 0.00029929452841021573, "loss": 0.5488, "step": 2332 }, { "epoch": 1.0978823529411765, "grad_norm": 0.7086092632982385, "learning_rate": 0.00029929027554057675, "loss": 0.5844, "step": 2333 }, { "epoch": 1.0983529411764705, "grad_norm": 0.8662520444132062, "learning_rate": 0.0002992860099208159, "loss": 0.4959, "step": 2334 }, { "epoch": 1.0988235294117648, "grad_norm": 0.3469272135246945, "learning_rate": 0.0002992817315512976, "loss": 0.3307, "step": 2335 }, { "epoch": 1.0992941176470588, "grad_norm": 0.8929918734912909, "learning_rate": 0.0002992774404323872, "loss": 0.5085, "step": 2336 }, { "epoch": 1.099764705882353, "grad_norm": 0.4038337678290411, "learning_rate": 0.0002992731365644512, "loss": 0.4066, "step": 2337 }, { "epoch": 1.100235294117647, "grad_norm": 0.3745046531121673, "learning_rate": 0.0002992688199478571, "loss": 0.3301, "step": 2338 }, { "epoch": 1.1007058823529412, "grad_norm": 0.4421692854715581, "learning_rate": 0.00029926449058297364, "loss": 0.3602, "step": 2339 }, { "epoch": 1.1011764705882352, "grad_norm": 0.42721941937195895, "learning_rate": 0.0002992601484701706, "loss": 0.2459, "step": 2340 }, { "epoch": 1.1016470588235294, "grad_norm": 0.21249424903249528, "learning_rate": 0.0002992557936098187, "loss": 0.1071, "step": 2341 }, { "epoch": 1.1021176470588236, "grad_norm": 1.04791595682859, "learning_rate": 0.00029925142600229, "loss": 0.8404, "step": 2342 }, { "epoch": 1.1025882352941176, "grad_norm": 0.41079215009247283, "learning_rate": 0.00029924704564795735, "loss": 0.3691, "step": 2343 }, { "epoch": 1.1030588235294116, "grad_norm": 0.18064609098967174, "learning_rate": 0.000299242652547195, "loss": 0.0967, "step": 2344 }, { "epoch": 1.1035294117647059, "grad_norm": 0.4495559146460001, "learning_rate": 0.0002992382467003781, "loss": 0.2523, "step": 2345 }, { "epoch": 1.104, "grad_norm": 1.2307889962449592, "learning_rate": 0.00029923382810788287, "loss": 0.7624, "step": 2346 }, { "epoch": 1.104470588235294, "grad_norm": 0.7830749935958687, "learning_rate": 0.0002992293967700868, "loss": 0.5898, "step": 2347 }, { "epoch": 1.1049411764705883, "grad_norm": 0.5018127412915427, "learning_rate": 0.00029922495268736824, "loss": 0.7013, "step": 2348 }, { "epoch": 1.1054117647058823, "grad_norm": 0.6256293368312631, "learning_rate": 0.0002992204958601068, "loss": 0.5273, "step": 2349 }, { "epoch": 1.1058823529411765, "grad_norm": 0.5108502289575243, "learning_rate": 0.0002992160262886831, "loss": 0.2054, "step": 2350 }, { "epoch": 1.1063529411764705, "grad_norm": 0.1702680071084567, "learning_rate": 0.0002992115439734788, "loss": 0.1066, "step": 2351 }, { "epoch": 1.1068235294117648, "grad_norm": 0.6335974045634822, "learning_rate": 0.0002992070489148768, "loss": 0.4856, "step": 2352 }, { "epoch": 1.1072941176470588, "grad_norm": 0.5326331368165812, "learning_rate": 0.000299202541113261, "loss": 0.3839, "step": 2353 }, { "epoch": 1.107764705882353, "grad_norm": 0.2978932427155028, "learning_rate": 0.0002991980205690163, "loss": 0.186, "step": 2354 }, { "epoch": 1.108235294117647, "grad_norm": 0.747517798581277, "learning_rate": 0.0002991934872825289, "loss": 0.4636, "step": 2355 }, { "epoch": 1.1087058823529412, "grad_norm": 0.5455658504933455, "learning_rate": 0.0002991889412541858, "loss": 0.4358, "step": 2356 }, { "epoch": 1.1091764705882352, "grad_norm": 0.6126718451071556, "learning_rate": 0.0002991843824843754, "loss": 0.5472, "step": 2357 }, { "epoch": 1.1096470588235294, "grad_norm": 0.24415184129439862, "learning_rate": 0.000299179810973487, "loss": 0.1612, "step": 2358 }, { "epoch": 1.1101176470588234, "grad_norm": 0.6942301584192783, "learning_rate": 0.000299175226721911, "loss": 0.476, "step": 2359 }, { "epoch": 1.1105882352941177, "grad_norm": 0.6222402362254839, "learning_rate": 0.00029917062973003897, "loss": 0.5126, "step": 2360 }, { "epoch": 1.1110588235294117, "grad_norm": 0.4339861268375264, "learning_rate": 0.0002991660199982634, "loss": 0.3959, "step": 2361 }, { "epoch": 1.1115294117647059, "grad_norm": 0.997396278802422, "learning_rate": 0.0002991613975269782, "loss": 0.8517, "step": 2362 }, { "epoch": 1.112, "grad_norm": 0.6179656477327957, "learning_rate": 0.0002991567623165779, "loss": 0.6685, "step": 2363 }, { "epoch": 1.112470588235294, "grad_norm": 0.3551361233349753, "learning_rate": 0.00029915211436745855, "loss": 0.5251, "step": 2364 }, { "epoch": 1.1129411764705883, "grad_norm": 0.3696096517152192, "learning_rate": 0.00029914745368001704, "loss": 0.3076, "step": 2365 }, { "epoch": 1.1134117647058823, "grad_norm": 0.6800164892467491, "learning_rate": 0.0002991427802546514, "loss": 0.4759, "step": 2366 }, { "epoch": 1.1138823529411765, "grad_norm": 0.33678028431003587, "learning_rate": 0.0002991380940917608, "loss": 0.1902, "step": 2367 }, { "epoch": 1.1143529411764705, "grad_norm": 0.6904611411533079, "learning_rate": 0.0002991333951917454, "loss": 0.4922, "step": 2368 }, { "epoch": 1.1148235294117648, "grad_norm": 0.2588338151048515, "learning_rate": 0.00029912868355500666, "loss": 0.1619, "step": 2369 }, { "epoch": 1.1152941176470588, "grad_norm": 0.3712508199017058, "learning_rate": 0.0002991239591819468, "loss": 0.2346, "step": 2370 }, { "epoch": 1.115764705882353, "grad_norm": 0.3293179588125251, "learning_rate": 0.00029911922207296937, "loss": 0.1928, "step": 2371 }, { "epoch": 1.116235294117647, "grad_norm": 0.29930605326566223, "learning_rate": 0.00029911447222847893, "loss": 0.3033, "step": 2372 }, { "epoch": 1.1167058823529412, "grad_norm": 0.6838608568451802, "learning_rate": 0.00029910970964888126, "loss": 0.4508, "step": 2373 }, { "epoch": 1.1171764705882352, "grad_norm": 0.8227149083730151, "learning_rate": 0.0002991049343345829, "loss": 0.4863, "step": 2374 }, { "epoch": 1.1176470588235294, "grad_norm": 0.24976750987112858, "learning_rate": 0.00029910014628599184, "loss": 0.1448, "step": 2375 }, { "epoch": 1.1181176470588234, "grad_norm": 0.5962016055745418, "learning_rate": 0.00029909534550351695, "loss": 0.4928, "step": 2376 }, { "epoch": 1.1185882352941177, "grad_norm": 0.8864548427358511, "learning_rate": 0.00029909053198756827, "loss": 0.5046, "step": 2377 }, { "epoch": 1.1190588235294117, "grad_norm": 0.7102428517269095, "learning_rate": 0.00029908570573855683, "loss": 0.5142, "step": 2378 }, { "epoch": 1.1195294117647059, "grad_norm": 0.28738896491456417, "learning_rate": 0.0002990808667568949, "loss": 0.2359, "step": 2379 }, { "epoch": 1.12, "grad_norm": 0.8127208570123878, "learning_rate": 0.0002990760150429957, "loss": 0.6809, "step": 2380 }, { "epoch": 1.120470588235294, "grad_norm": 0.37968711103260544, "learning_rate": 0.0002990711505972736, "loss": 0.2266, "step": 2381 }, { "epoch": 1.1209411764705883, "grad_norm": 0.38218666888371494, "learning_rate": 0.00029906627342014407, "loss": 0.2732, "step": 2382 }, { "epoch": 1.1214117647058823, "grad_norm": 0.5437759050032323, "learning_rate": 0.00029906138351202356, "loss": 0.4281, "step": 2383 }, { "epoch": 1.1218823529411766, "grad_norm": 0.49739366411019403, "learning_rate": 0.0002990564808733298, "loss": 0.3859, "step": 2384 }, { "epoch": 1.1223529411764706, "grad_norm": 0.6662173701663062, "learning_rate": 0.00029905156550448145, "loss": 0.7006, "step": 2385 }, { "epoch": 1.1228235294117648, "grad_norm": 0.4282091881408074, "learning_rate": 0.00029904663740589827, "loss": 0.4235, "step": 2386 }, { "epoch": 1.1232941176470588, "grad_norm": 0.33999990025909765, "learning_rate": 0.0002990416965780012, "loss": 0.3206, "step": 2387 }, { "epoch": 1.123764705882353, "grad_norm": 0.26013899515804195, "learning_rate": 0.0002990367430212123, "loss": 0.1534, "step": 2388 }, { "epoch": 1.124235294117647, "grad_norm": 0.33451068612575, "learning_rate": 0.0002990317767359544, "loss": 0.2016, "step": 2389 }, { "epoch": 1.1247058823529412, "grad_norm": 0.4114152631801773, "learning_rate": 0.0002990267977226518, "loss": 0.193, "step": 2390 }, { "epoch": 1.1251764705882352, "grad_norm": 0.2556829338187391, "learning_rate": 0.00029902180598172977, "loss": 0.1693, "step": 2391 }, { "epoch": 1.1256470588235294, "grad_norm": 0.8271278962904957, "learning_rate": 0.00029901680151361447, "loss": 0.5383, "step": 2392 }, { "epoch": 1.1261176470588234, "grad_norm": 0.574673652189367, "learning_rate": 0.00029901178431873346, "loss": 0.4413, "step": 2393 }, { "epoch": 1.1265882352941177, "grad_norm": 0.6311363298405912, "learning_rate": 0.0002990067543975151, "loss": 0.3712, "step": 2394 }, { "epoch": 1.1270588235294117, "grad_norm": 0.8942842002557794, "learning_rate": 0.0002990017117503891, "loss": 0.668, "step": 2395 }, { "epoch": 1.127529411764706, "grad_norm": 1.0636867817492803, "learning_rate": 0.0002989966563777861, "loss": 0.8864, "step": 2396 }, { "epoch": 1.1280000000000001, "grad_norm": 0.6886927289509905, "learning_rate": 0.00029899158828013774, "loss": 0.655, "step": 2397 }, { "epoch": 1.1284705882352941, "grad_norm": 0.46890240294370145, "learning_rate": 0.000298986507457877, "loss": 0.4276, "step": 2398 }, { "epoch": 1.1289411764705881, "grad_norm": 0.25324649022257223, "learning_rate": 0.0002989814139114377, "loss": 0.163, "step": 2399 }, { "epoch": 1.1294117647058823, "grad_norm": 0.19000064857799467, "learning_rate": 0.0002989763076412549, "loss": 0.0815, "step": 2400 }, { "epoch": 1.1298823529411766, "grad_norm": 0.34889039294422525, "learning_rate": 0.00029897118864776476, "loss": 0.3395, "step": 2401 }, { "epoch": 1.1303529411764706, "grad_norm": 1.332402246262344, "learning_rate": 0.00029896605693140437, "loss": 0.8496, "step": 2402 }, { "epoch": 1.1308235294117648, "grad_norm": 0.16476328531358533, "learning_rate": 0.0002989609124926121, "loss": 0.1165, "step": 2403 }, { "epoch": 1.1312941176470588, "grad_norm": 1.2722728069256346, "learning_rate": 0.0002989557553318272, "loss": 0.7147, "step": 2404 }, { "epoch": 1.131764705882353, "grad_norm": 0.4935613475017764, "learning_rate": 0.0002989505854494902, "loss": 0.3069, "step": 2405 }, { "epoch": 1.132235294117647, "grad_norm": 0.6019834230450888, "learning_rate": 0.00029894540284604255, "loss": 0.3557, "step": 2406 }, { "epoch": 1.1327058823529412, "grad_norm": 0.622997554261647, "learning_rate": 0.000298940207521927, "loss": 0.6413, "step": 2407 }, { "epoch": 1.1331764705882352, "grad_norm": 0.6633362539415131, "learning_rate": 0.00029893499947758706, "loss": 0.5451, "step": 2408 }, { "epoch": 1.1336470588235295, "grad_norm": 0.3072041991080197, "learning_rate": 0.00029892977871346776, "loss": 0.2154, "step": 2409 }, { "epoch": 1.1341176470588235, "grad_norm": 0.42161549369650353, "learning_rate": 0.00029892454523001484, "loss": 0.3066, "step": 2410 }, { "epoch": 1.1345882352941177, "grad_norm": 0.21911986233918979, "learning_rate": 0.00029891929902767526, "loss": 0.0846, "step": 2411 }, { "epoch": 1.1350588235294117, "grad_norm": 0.5032277020579613, "learning_rate": 0.0002989140401068971, "loss": 0.3876, "step": 2412 }, { "epoch": 1.135529411764706, "grad_norm": 0.3910173711940227, "learning_rate": 0.0002989087684681295, "loss": 0.3732, "step": 2413 }, { "epoch": 1.1360000000000001, "grad_norm": 0.30903736796293213, "learning_rate": 0.00029890348411182267, "loss": 0.3377, "step": 2414 }, { "epoch": 1.1364705882352941, "grad_norm": 0.4212245886329147, "learning_rate": 0.00029889818703842796, "loss": 0.3011, "step": 2415 }, { "epoch": 1.1369411764705881, "grad_norm": 1.0499670406929447, "learning_rate": 0.0002988928772483977, "loss": 0.7227, "step": 2416 }, { "epoch": 1.1374117647058823, "grad_norm": 0.35171254560030424, "learning_rate": 0.0002988875547421854, "loss": 0.4143, "step": 2417 }, { "epoch": 1.1378823529411766, "grad_norm": 0.4466345513793331, "learning_rate": 0.00029888221952024567, "loss": 0.4686, "step": 2418 }, { "epoch": 1.1383529411764706, "grad_norm": 0.4040530143816291, "learning_rate": 0.00029887687158303416, "loss": 0.2125, "step": 2419 }, { "epoch": 1.1388235294117648, "grad_norm": 0.4806451469384655, "learning_rate": 0.00029887151093100755, "loss": 0.2256, "step": 2420 }, { "epoch": 1.1392941176470588, "grad_norm": 0.41094865178488826, "learning_rate": 0.00029886613756462365, "loss": 0.319, "step": 2421 }, { "epoch": 1.139764705882353, "grad_norm": 0.3706556690389159, "learning_rate": 0.00029886075148434146, "loss": 0.3961, "step": 2422 }, { "epoch": 1.140235294117647, "grad_norm": 0.49692050773115515, "learning_rate": 0.0002988553526906209, "loss": 0.423, "step": 2423 }, { "epoch": 1.1407058823529412, "grad_norm": 0.5589225782391929, "learning_rate": 0.0002988499411839232, "loss": 0.5386, "step": 2424 }, { "epoch": 1.1411764705882352, "grad_norm": 0.2607703483434621, "learning_rate": 0.0002988445169647103, "loss": 0.1977, "step": 2425 }, { "epoch": 1.1416470588235295, "grad_norm": 0.3505681836678676, "learning_rate": 0.00029883908003344554, "loss": 0.408, "step": 2426 }, { "epoch": 1.1421176470588235, "grad_norm": 0.376770617944233, "learning_rate": 0.00029883363039059343, "loss": 0.2981, "step": 2427 }, { "epoch": 1.1425882352941177, "grad_norm": 0.3294210254734651, "learning_rate": 0.0002988281680366192, "loss": 0.1762, "step": 2428 }, { "epoch": 1.1430588235294117, "grad_norm": 0.3868078432567134, "learning_rate": 0.0002988226929719894, "loss": 0.3422, "step": 2429 }, { "epoch": 1.143529411764706, "grad_norm": 0.31578100142879206, "learning_rate": 0.0002988172051971717, "loss": 0.3285, "step": 2430 }, { "epoch": 1.144, "grad_norm": 0.22360346972502881, "learning_rate": 0.0002988117047126346, "loss": 0.1983, "step": 2431 }, { "epoch": 1.1444705882352941, "grad_norm": 0.45633967201306763, "learning_rate": 0.00029880619151884813, "loss": 0.3821, "step": 2432 }, { "epoch": 1.1449411764705881, "grad_norm": 0.2782794222262196, "learning_rate": 0.000298800665616283, "loss": 0.2177, "step": 2433 }, { "epoch": 1.1454117647058824, "grad_norm": 0.413852445679432, "learning_rate": 0.00029879512700541115, "loss": 0.6381, "step": 2434 }, { "epoch": 1.1458823529411766, "grad_norm": 0.7185725189627621, "learning_rate": 0.0002987895756867056, "loss": 0.4453, "step": 2435 }, { "epoch": 1.1463529411764706, "grad_norm": 0.252586178717413, "learning_rate": 0.00029878401166064046, "loss": 0.209, "step": 2436 }, { "epoch": 1.1468235294117648, "grad_norm": 0.8961383438125071, "learning_rate": 0.000298778434927691, "loss": 0.7564, "step": 2437 }, { "epoch": 1.1472941176470588, "grad_norm": 0.5974032422133654, "learning_rate": 0.0002987728454883334, "loss": 0.5295, "step": 2438 }, { "epoch": 1.147764705882353, "grad_norm": 0.21445573608538102, "learning_rate": 0.00029876724334304506, "loss": 0.1879, "step": 2439 }, { "epoch": 1.148235294117647, "grad_norm": 0.22689566708296172, "learning_rate": 0.00029876162849230446, "loss": 0.2054, "step": 2440 }, { "epoch": 1.1487058823529412, "grad_norm": 0.6628867566853054, "learning_rate": 0.00029875600093659115, "loss": 0.5047, "step": 2441 }, { "epoch": 1.1491764705882352, "grad_norm": 0.34568028530299477, "learning_rate": 0.00029875036067638565, "loss": 0.3357, "step": 2442 }, { "epoch": 1.1496470588235295, "grad_norm": 0.5023662227399349, "learning_rate": 0.00029874470771216983, "loss": 0.5033, "step": 2443 }, { "epoch": 1.1501176470588235, "grad_norm": 0.3336047232634321, "learning_rate": 0.0002987390420444262, "loss": 0.2222, "step": 2444 }, { "epoch": 1.1505882352941177, "grad_norm": 0.5756661194727565, "learning_rate": 0.00029873336367363903, "loss": 0.3882, "step": 2445 }, { "epoch": 1.1510588235294117, "grad_norm": 0.42321699771317656, "learning_rate": 0.00029872767260029294, "loss": 0.4066, "step": 2446 }, { "epoch": 1.151529411764706, "grad_norm": 0.43535604390543, "learning_rate": 0.00029872196882487417, "loss": 0.4056, "step": 2447 }, { "epoch": 1.152, "grad_norm": 0.3225885092125908, "learning_rate": 0.00029871625234786974, "loss": 0.244, "step": 2448 }, { "epoch": 1.1524705882352941, "grad_norm": 0.4528376678052007, "learning_rate": 0.0002987105231697679, "loss": 0.3369, "step": 2449 }, { "epoch": 1.1529411764705881, "grad_norm": 0.25650399294147513, "learning_rate": 0.000298704781291058, "loss": 0.178, "step": 2450 }, { "epoch": 1.1534117647058824, "grad_norm": 0.5304352721813467, "learning_rate": 0.0002986990267122304, "loss": 0.2962, "step": 2451 }, { "epoch": 1.1538823529411766, "grad_norm": 0.8479853920826798, "learning_rate": 0.0002986932594337765, "loss": 0.5826, "step": 2452 }, { "epoch": 1.1543529411764706, "grad_norm": 0.2474018004848891, "learning_rate": 0.0002986874794561889, "loss": 0.2927, "step": 2453 }, { "epoch": 1.1548235294117646, "grad_norm": 0.5350805340047993, "learning_rate": 0.0002986816867799613, "loss": 0.5549, "step": 2454 }, { "epoch": 1.1552941176470588, "grad_norm": 0.36292802102784544, "learning_rate": 0.0002986758814055883, "loss": 0.3287, "step": 2455 }, { "epoch": 1.155764705882353, "grad_norm": 1.0180069254768291, "learning_rate": 0.0002986700633335659, "loss": 0.4551, "step": 2456 }, { "epoch": 1.156235294117647, "grad_norm": 0.6803962385466458, "learning_rate": 0.0002986642325643908, "loss": 0.4508, "step": 2457 }, { "epoch": 1.1567058823529413, "grad_norm": 0.35044304255496156, "learning_rate": 0.00029865838909856103, "loss": 0.2244, "step": 2458 }, { "epoch": 1.1571764705882353, "grad_norm": 0.38657516880748616, "learning_rate": 0.0002986525329365757, "loss": 0.2812, "step": 2459 }, { "epoch": 1.1576470588235295, "grad_norm": 1.0129379516260115, "learning_rate": 0.00029864666407893487, "loss": 0.7848, "step": 2460 }, { "epoch": 1.1581176470588235, "grad_norm": 0.8364481187439852, "learning_rate": 0.00029864078252613983, "loss": 0.6007, "step": 2461 }, { "epoch": 1.1585882352941177, "grad_norm": 0.4387377464839965, "learning_rate": 0.0002986348882786929, "loss": 0.3398, "step": 2462 }, { "epoch": 1.1590588235294117, "grad_norm": 0.20216776670702305, "learning_rate": 0.00029862898133709747, "loss": 0.1537, "step": 2463 }, { "epoch": 1.159529411764706, "grad_norm": 0.1980315006007358, "learning_rate": 0.000298623061701858, "loss": 0.1668, "step": 2464 }, { "epoch": 1.16, "grad_norm": 0.6524241952885268, "learning_rate": 0.00029861712937348007, "loss": 0.4593, "step": 2465 }, { "epoch": 1.1604705882352941, "grad_norm": 1.3725710410792409, "learning_rate": 0.0002986111843524703, "loss": 0.7271, "step": 2466 }, { "epoch": 1.1609411764705881, "grad_norm": 0.4549796497607075, "learning_rate": 0.00029860522663933655, "loss": 0.4168, "step": 2467 }, { "epoch": 1.1614117647058824, "grad_norm": 0.2939558943908123, "learning_rate": 0.00029859925623458745, "loss": 0.2294, "step": 2468 }, { "epoch": 1.1618823529411766, "grad_norm": 0.8399377115492632, "learning_rate": 0.000298593273138733, "loss": 0.5492, "step": 2469 }, { "epoch": 1.1623529411764706, "grad_norm": 0.32569686527320385, "learning_rate": 0.00029858727735228416, "loss": 0.2742, "step": 2470 }, { "epoch": 1.1628235294117646, "grad_norm": 0.2963071226396627, "learning_rate": 0.00029858126887575306, "loss": 0.2622, "step": 2471 }, { "epoch": 1.1632941176470588, "grad_norm": 0.6093026165608109, "learning_rate": 0.00029857524770965285, "loss": 0.452, "step": 2472 }, { "epoch": 1.163764705882353, "grad_norm": 0.42717330005030063, "learning_rate": 0.00029856921385449763, "loss": 0.2646, "step": 2473 }, { "epoch": 1.164235294117647, "grad_norm": 0.6891035564943895, "learning_rate": 0.0002985631673108029, "loss": 0.5104, "step": 2474 }, { "epoch": 1.1647058823529413, "grad_norm": 0.7345869178122224, "learning_rate": 0.00029855710807908493, "loss": 0.3509, "step": 2475 }, { "epoch": 1.1651764705882353, "grad_norm": 0.5070206414250357, "learning_rate": 0.00029855103615986133, "loss": 0.2819, "step": 2476 }, { "epoch": 1.1656470588235295, "grad_norm": 1.2759748212570248, "learning_rate": 0.0002985449515536505, "loss": 0.7247, "step": 2477 }, { "epoch": 1.1661176470588235, "grad_norm": 0.31709138756286015, "learning_rate": 0.0002985388542609723, "loss": 0.3636, "step": 2478 }, { "epoch": 1.1665882352941177, "grad_norm": 0.5648473697995001, "learning_rate": 0.0002985327442823473, "loss": 0.4733, "step": 2479 }, { "epoch": 1.1670588235294117, "grad_norm": 0.39251548398935304, "learning_rate": 0.00029852662161829744, "loss": 0.1792, "step": 2480 }, { "epoch": 1.167529411764706, "grad_norm": 0.7896766813423001, "learning_rate": 0.0002985204862693455, "loss": 0.816, "step": 2481 }, { "epoch": 1.168, "grad_norm": 0.32468812731859187, "learning_rate": 0.00029851433823601563, "loss": 0.2537, "step": 2482 }, { "epoch": 1.1684705882352941, "grad_norm": 0.4250123961911692, "learning_rate": 0.00029850817751883276, "loss": 0.3734, "step": 2483 }, { "epoch": 1.1689411764705882, "grad_norm": 0.537023665193198, "learning_rate": 0.0002985020041183231, "loss": 0.5441, "step": 2484 }, { "epoch": 1.1694117647058824, "grad_norm": 0.13422020442462834, "learning_rate": 0.00029849581803501393, "loss": 0.1388, "step": 2485 }, { "epoch": 1.1698823529411766, "grad_norm": 0.3374839822474612, "learning_rate": 0.0002984896192694335, "loss": 0.2205, "step": 2486 }, { "epoch": 1.1703529411764706, "grad_norm": 0.2157040359678082, "learning_rate": 0.00029848340782211125, "loss": 0.1427, "step": 2487 }, { "epoch": 1.1708235294117646, "grad_norm": 0.6622437077036484, "learning_rate": 0.0002984771836935777, "loss": 0.4666, "step": 2488 }, { "epoch": 1.1712941176470588, "grad_norm": 0.5864998468544225, "learning_rate": 0.0002984709468843644, "loss": 0.4062, "step": 2489 }, { "epoch": 1.171764705882353, "grad_norm": 0.5645248710381566, "learning_rate": 0.0002984646973950039, "loss": 0.4341, "step": 2490 }, { "epoch": 1.172235294117647, "grad_norm": 0.33983611921971324, "learning_rate": 0.00029845843522603006, "loss": 0.2147, "step": 2491 }, { "epoch": 1.1727058823529413, "grad_norm": 0.11754599514181649, "learning_rate": 0.0002984521603779777, "loss": 0.057, "step": 2492 }, { "epoch": 1.1731764705882353, "grad_norm": 0.7599035844369735, "learning_rate": 0.0002984458728513827, "loss": 0.7449, "step": 2493 }, { "epoch": 1.1736470588235295, "grad_norm": 0.8433002834827443, "learning_rate": 0.000298439572646782, "loss": 0.5532, "step": 2494 }, { "epoch": 1.1741176470588235, "grad_norm": 0.5466647029881988, "learning_rate": 0.0002984332597647137, "loss": 0.4257, "step": 2495 }, { "epoch": 1.1745882352941177, "grad_norm": 0.3866085170094972, "learning_rate": 0.000298426934205717, "loss": 0.3676, "step": 2496 }, { "epoch": 1.1750588235294117, "grad_norm": 0.5409366642340978, "learning_rate": 0.00029842059597033204, "loss": 0.4875, "step": 2497 }, { "epoch": 1.175529411764706, "grad_norm": 0.7304239196846285, "learning_rate": 0.0002984142450591002, "loss": 0.6515, "step": 2498 }, { "epoch": 1.176, "grad_norm": 0.43001104212744096, "learning_rate": 0.0002984078814725639, "loss": 0.4258, "step": 2499 }, { "epoch": 1.1764705882352942, "grad_norm": 0.6039879463219505, "learning_rate": 0.0002984015052112665, "loss": 0.4638, "step": 2500 }, { "epoch": 1.1769411764705882, "grad_norm": 0.1425002337733324, "learning_rate": 0.0002983951162757527, "loss": 0.0718, "step": 2501 }, { "epoch": 1.1774117647058824, "grad_norm": 0.7605545579682458, "learning_rate": 0.0002983887146665681, "loss": 0.5761, "step": 2502 }, { "epoch": 1.1778823529411764, "grad_norm": 0.4497378119892917, "learning_rate": 0.00029838230038425944, "loss": 0.402, "step": 2503 }, { "epoch": 1.1783529411764706, "grad_norm": 0.21334325454383535, "learning_rate": 0.0002983758734293745, "loss": 0.1175, "step": 2504 }, { "epoch": 1.1788235294117646, "grad_norm": 0.4626116053934877, "learning_rate": 0.00029836943380246215, "loss": 0.4601, "step": 2505 }, { "epoch": 1.1792941176470588, "grad_norm": 0.7027120936170508, "learning_rate": 0.00029836298150407245, "loss": 0.7799, "step": 2506 }, { "epoch": 1.179764705882353, "grad_norm": 0.3280151262983384, "learning_rate": 0.00029835651653475643, "loss": 0.2729, "step": 2507 }, { "epoch": 1.180235294117647, "grad_norm": 0.1581283629227194, "learning_rate": 0.00029835003889506616, "loss": 0.0954, "step": 2508 }, { "epoch": 1.1807058823529413, "grad_norm": 0.39482415955656536, "learning_rate": 0.000298343548585555, "loss": 0.3633, "step": 2509 }, { "epoch": 1.1811764705882353, "grad_norm": 0.30777369391765913, "learning_rate": 0.0002983370456067771, "loss": 0.1768, "step": 2510 }, { "epoch": 1.1816470588235295, "grad_norm": 0.3380770851224198, "learning_rate": 0.0002983305299592879, "loss": 0.3034, "step": 2511 }, { "epoch": 1.1821176470588235, "grad_norm": 0.8070738929579075, "learning_rate": 0.0002983240016436439, "loss": 0.848, "step": 2512 }, { "epoch": 1.1825882352941177, "grad_norm": 0.24492930546499253, "learning_rate": 0.0002983174606604027, "loss": 0.2314, "step": 2513 }, { "epoch": 1.1830588235294117, "grad_norm": 0.4396741976320989, "learning_rate": 0.0002983109070101228, "loss": 0.5579, "step": 2514 }, { "epoch": 1.183529411764706, "grad_norm": 0.6398710923787566, "learning_rate": 0.00029830434069336403, "loss": 0.7037, "step": 2515 }, { "epoch": 1.184, "grad_norm": 0.2316852501450123, "learning_rate": 0.0002982977617106871, "loss": 0.1283, "step": 2516 }, { "epoch": 1.1844705882352942, "grad_norm": 0.7340277276433883, "learning_rate": 0.0002982911700626539, "loss": 0.518, "step": 2517 }, { "epoch": 1.1849411764705882, "grad_norm": 0.4596203524792539, "learning_rate": 0.0002982845657498274, "loss": 0.3308, "step": 2518 }, { "epoch": 1.1854117647058824, "grad_norm": 0.7124892463873033, "learning_rate": 0.0002982779487727718, "loss": 0.4917, "step": 2519 }, { "epoch": 1.1858823529411764, "grad_norm": 0.2754711525975412, "learning_rate": 0.0002982713191320519, "loss": 0.2705, "step": 2520 }, { "epoch": 1.1863529411764706, "grad_norm": 0.42951795756873956, "learning_rate": 0.00029826467682823415, "loss": 0.452, "step": 2521 }, { "epoch": 1.1868235294117646, "grad_norm": 0.21583164165388563, "learning_rate": 0.0002982580218618858, "loss": 0.0972, "step": 2522 }, { "epoch": 1.1872941176470588, "grad_norm": 0.43685921240994313, "learning_rate": 0.0002982513542335751, "loss": 0.4479, "step": 2523 }, { "epoch": 1.187764705882353, "grad_norm": 0.23071063748236734, "learning_rate": 0.00029824467394387166, "loss": 0.2675, "step": 2524 }, { "epoch": 1.188235294117647, "grad_norm": 0.5196723999072209, "learning_rate": 0.0002982379809933459, "loss": 0.3156, "step": 2525 }, { "epoch": 1.188705882352941, "grad_norm": 0.39340947963726197, "learning_rate": 0.00029823127538256943, "loss": 0.2895, "step": 2526 }, { "epoch": 1.1891764705882353, "grad_norm": 0.5076230686132521, "learning_rate": 0.000298224557112115, "loss": 0.5746, "step": 2527 }, { "epoch": 1.1896470588235295, "grad_norm": 0.29039755189049404, "learning_rate": 0.0002982178261825563, "loss": 0.2713, "step": 2528 }, { "epoch": 1.1901176470588235, "grad_norm": 0.3085779145352659, "learning_rate": 0.0002982110825944683, "loss": 0.281, "step": 2529 }, { "epoch": 1.1905882352941177, "grad_norm": 0.2280006986015345, "learning_rate": 0.0002982043263484269, "loss": 0.1304, "step": 2530 }, { "epoch": 1.1910588235294117, "grad_norm": 0.19550459086075972, "learning_rate": 0.00029819755744500904, "loss": 0.0946, "step": 2531 }, { "epoch": 1.191529411764706, "grad_norm": 0.6629366634672073, "learning_rate": 0.0002981907758847929, "loss": 0.4642, "step": 2532 }, { "epoch": 1.192, "grad_norm": 0.4382761331728096, "learning_rate": 0.0002981839816683576, "loss": 0.4127, "step": 2533 }, { "epoch": 1.1924705882352942, "grad_norm": 0.01837528523518237, "learning_rate": 0.00029817717479628346, "loss": 0.0067, "step": 2534 }, { "epoch": 1.1929411764705882, "grad_norm": 0.32928645296865594, "learning_rate": 0.00029817035526915177, "loss": 0.1312, "step": 2535 }, { "epoch": 1.1934117647058824, "grad_norm": 0.32536994280051, "learning_rate": 0.00029816352308754496, "loss": 0.1794, "step": 2536 }, { "epoch": 1.1938823529411764, "grad_norm": 0.9722257404817546, "learning_rate": 0.00029815667825204654, "loss": 0.4134, "step": 2537 }, { "epoch": 1.1943529411764706, "grad_norm": 0.759611105827424, "learning_rate": 0.0002981498207632411, "loss": 0.5217, "step": 2538 }, { "epoch": 1.1948235294117646, "grad_norm": 0.49680444217423136, "learning_rate": 0.0002981429506217143, "loss": 0.4499, "step": 2539 }, { "epoch": 1.1952941176470588, "grad_norm": 0.4571563579871534, "learning_rate": 0.0002981360678280529, "loss": 0.2143, "step": 2540 }, { "epoch": 1.195764705882353, "grad_norm": 0.32664913822356373, "learning_rate": 0.00029812917238284466, "loss": 0.1245, "step": 2541 }, { "epoch": 1.196235294117647, "grad_norm": 0.2281914347116548, "learning_rate": 0.00029812226428667854, "loss": 0.0874, "step": 2542 }, { "epoch": 1.196705882352941, "grad_norm": 1.0008204322838243, "learning_rate": 0.0002981153435401445, "loss": 0.5182, "step": 2543 }, { "epoch": 1.1971764705882353, "grad_norm": 0.6429314180992477, "learning_rate": 0.00029810841014383366, "loss": 0.5921, "step": 2544 }, { "epoch": 1.1976470588235295, "grad_norm": 0.14903072145681662, "learning_rate": 0.00029810146409833815, "loss": 0.1095, "step": 2545 }, { "epoch": 1.1981176470588235, "grad_norm": 0.7191309137937019, "learning_rate": 0.0002980945054042511, "loss": 0.5372, "step": 2546 }, { "epoch": 1.1985882352941177, "grad_norm": 0.3921936785118397, "learning_rate": 0.00029808753406216693, "loss": 0.1759, "step": 2547 }, { "epoch": 1.1990588235294117, "grad_norm": 0.07660112561960743, "learning_rate": 0.000298080550072681, "loss": 0.0339, "step": 2548 }, { "epoch": 1.199529411764706, "grad_norm": 0.4812078167352552, "learning_rate": 0.0002980735534363897, "loss": 0.2907, "step": 2549 }, { "epoch": 1.2, "grad_norm": 0.2848005365064569, "learning_rate": 0.0002980665441538907, "loss": 0.2081, "step": 2550 }, { "epoch": 1.2004705882352942, "grad_norm": 0.18779976033358498, "learning_rate": 0.00029805952222578257, "loss": 0.0664, "step": 2551 }, { "epoch": 1.2009411764705882, "grad_norm": 0.300581657666466, "learning_rate": 0.000298052487652665, "loss": 0.2468, "step": 2552 }, { "epoch": 1.2014117647058824, "grad_norm": 0.26139265844907916, "learning_rate": 0.0002980454404351388, "loss": 0.1914, "step": 2553 }, { "epoch": 1.2018823529411764, "grad_norm": 0.5252192015757822, "learning_rate": 0.0002980383805738058, "loss": 0.4556, "step": 2554 }, { "epoch": 1.2023529411764706, "grad_norm": 0.26381310059752033, "learning_rate": 0.00029803130806926897, "loss": 0.1817, "step": 2555 }, { "epoch": 1.2028235294117646, "grad_norm": 0.15208207563789905, "learning_rate": 0.0002980242229221324, "loss": 0.1534, "step": 2556 }, { "epoch": 1.2032941176470588, "grad_norm": 0.3261341991241039, "learning_rate": 0.00029801712513300105, "loss": 0.4939, "step": 2557 }, { "epoch": 1.2037647058823528, "grad_norm": 0.34822906571159573, "learning_rate": 0.00029801001470248123, "loss": 0.2908, "step": 2558 }, { "epoch": 1.204235294117647, "grad_norm": 0.4364200819094605, "learning_rate": 0.00029800289163118014, "loss": 0.2736, "step": 2559 }, { "epoch": 1.204705882352941, "grad_norm": 0.5842937790691459, "learning_rate": 0.00029799575591970623, "loss": 0.4125, "step": 2560 }, { "epoch": 1.2051764705882353, "grad_norm": 0.26407718638285754, "learning_rate": 0.00029798860756866883, "loss": 0.0971, "step": 2561 }, { "epoch": 1.2056470588235295, "grad_norm": 0.2953950426056265, "learning_rate": 0.00029798144657867845, "loss": 0.2899, "step": 2562 }, { "epoch": 1.2061176470588235, "grad_norm": 0.8289835618888464, "learning_rate": 0.0002979742729503467, "loss": 0.7959, "step": 2563 }, { "epoch": 1.2065882352941177, "grad_norm": 0.21707871668084486, "learning_rate": 0.0002979670866842862, "loss": 0.2121, "step": 2564 }, { "epoch": 1.2070588235294117, "grad_norm": 0.26858392934306197, "learning_rate": 0.0002979598877811107, "loss": 0.2912, "step": 2565 }, { "epoch": 1.207529411764706, "grad_norm": 0.5508813566874228, "learning_rate": 0.0002979526762414351, "loss": 0.3435, "step": 2566 }, { "epoch": 1.208, "grad_norm": 0.8183793317381939, "learning_rate": 0.0002979454520658752, "loss": 0.6209, "step": 2567 }, { "epoch": 1.2084705882352942, "grad_norm": 0.2949275044340132, "learning_rate": 0.0002979382152550481, "loss": 0.1914, "step": 2568 }, { "epoch": 1.2089411764705882, "grad_norm": 0.6750894968965557, "learning_rate": 0.0002979309658095717, "loss": 0.6237, "step": 2569 }, { "epoch": 1.2094117647058824, "grad_norm": 0.37981361479287584, "learning_rate": 0.0002979237037300653, "loss": 0.4475, "step": 2570 }, { "epoch": 1.2098823529411764, "grad_norm": 0.39306832736192876, "learning_rate": 0.00029791642901714897, "loss": 0.5967, "step": 2571 }, { "epoch": 1.2103529411764706, "grad_norm": 0.29191542634501627, "learning_rate": 0.0002979091416714441, "loss": 0.2181, "step": 2572 }, { "epoch": 1.2108235294117646, "grad_norm": 0.4126825723964215, "learning_rate": 0.00029790184169357307, "loss": 0.3775, "step": 2573 }, { "epoch": 1.2112941176470589, "grad_norm": 0.6388814509679301, "learning_rate": 0.00029789452908415923, "loss": 0.3878, "step": 2574 }, { "epoch": 1.2117647058823529, "grad_norm": 0.3626465932268199, "learning_rate": 0.0002978872038438273, "loss": 0.358, "step": 2575 }, { "epoch": 1.212235294117647, "grad_norm": 0.6000682680358865, "learning_rate": 0.0002978798659732027, "loss": 0.3396, "step": 2576 }, { "epoch": 1.212705882352941, "grad_norm": 0.3517421995336272, "learning_rate": 0.00029787251547291215, "loss": 0.3034, "step": 2577 }, { "epoch": 1.2131764705882353, "grad_norm": 0.36033194184476486, "learning_rate": 0.00029786515234358357, "loss": 0.2432, "step": 2578 }, { "epoch": 1.2136470588235295, "grad_norm": 0.3274225149948871, "learning_rate": 0.0002978577765858457, "loss": 0.3421, "step": 2579 }, { "epoch": 1.2141176470588235, "grad_norm": 0.279553002787556, "learning_rate": 0.00029785038820032844, "loss": 0.1528, "step": 2580 }, { "epoch": 1.2145882352941175, "grad_norm": 0.2849128555856937, "learning_rate": 0.0002978429871876628, "loss": 0.254, "step": 2581 }, { "epoch": 1.2150588235294117, "grad_norm": 0.21940588276019934, "learning_rate": 0.00029783557354848094, "loss": 0.1634, "step": 2582 }, { "epoch": 1.215529411764706, "grad_norm": 0.1569712657690832, "learning_rate": 0.0002978281472834159, "loss": 0.1136, "step": 2583 }, { "epoch": 1.216, "grad_norm": 0.41288298179047744, "learning_rate": 0.00029782070839310206, "loss": 0.3463, "step": 2584 }, { "epoch": 1.2164705882352942, "grad_norm": 0.3087514766413373, "learning_rate": 0.0002978132568781746, "loss": 0.2682, "step": 2585 }, { "epoch": 1.2169411764705882, "grad_norm": 0.38436812433134365, "learning_rate": 0.00029780579273927006, "loss": 0.375, "step": 2586 }, { "epoch": 1.2174117647058824, "grad_norm": 0.4547119620303503, "learning_rate": 0.0002977983159770258, "loss": 0.4001, "step": 2587 }, { "epoch": 1.2178823529411764, "grad_norm": 0.393313263712332, "learning_rate": 0.0002977908265920804, "loss": 0.3466, "step": 2588 }, { "epoch": 1.2183529411764706, "grad_norm": 0.3891245243657899, "learning_rate": 0.00029778332458507353, "loss": 0.4134, "step": 2589 }, { "epoch": 1.2188235294117646, "grad_norm": 0.4870882607370394, "learning_rate": 0.0002977758099566458, "loss": 0.6741, "step": 2590 }, { "epoch": 1.2192941176470589, "grad_norm": 0.13699078423242014, "learning_rate": 0.0002977682827074391, "loss": 0.0949, "step": 2591 }, { "epoch": 1.2197647058823529, "grad_norm": 0.20507088492277634, "learning_rate": 0.0002977607428380963, "loss": 0.1276, "step": 2592 }, { "epoch": 1.220235294117647, "grad_norm": 0.48930086808067613, "learning_rate": 0.00029775319034926125, "loss": 0.4518, "step": 2593 }, { "epoch": 1.220705882352941, "grad_norm": 0.7304523693940221, "learning_rate": 0.000297745625241579, "loss": 0.5004, "step": 2594 }, { "epoch": 1.2211764705882353, "grad_norm": 0.3937064932295796, "learning_rate": 0.0002977380475156958, "loss": 0.4793, "step": 2595 }, { "epoch": 1.2216470588235295, "grad_norm": 0.4522992853292782, "learning_rate": 0.00029773045717225856, "loss": 0.5659, "step": 2596 }, { "epoch": 1.2221176470588235, "grad_norm": 0.5216981901318731, "learning_rate": 0.0002977228542119157, "loss": 0.5486, "step": 2597 }, { "epoch": 1.2225882352941175, "grad_norm": 0.06520457997106625, "learning_rate": 0.0002977152386353165, "loss": 0.0198, "step": 2598 }, { "epoch": 1.2230588235294118, "grad_norm": 0.3320676087991126, "learning_rate": 0.0002977076104431114, "loss": 0.1722, "step": 2599 }, { "epoch": 1.223529411764706, "grad_norm": 0.2895909305087108, "learning_rate": 0.00029769996963595184, "loss": 0.2117, "step": 2600 }, { "epoch": 1.224, "grad_norm": 0.2690072646356862, "learning_rate": 0.0002976923162144904, "loss": 0.3137, "step": 2601 }, { "epoch": 1.2244705882352942, "grad_norm": 0.82079821370531, "learning_rate": 0.0002976846501793808, "loss": 0.5321, "step": 2602 }, { "epoch": 1.2249411764705882, "grad_norm": 0.23682897981183934, "learning_rate": 0.00029767697153127766, "loss": 0.1807, "step": 2603 }, { "epoch": 1.2254117647058824, "grad_norm": 0.5175344250652437, "learning_rate": 0.0002976692802708368, "loss": 0.4278, "step": 2604 }, { "epoch": 1.2258823529411764, "grad_norm": 0.23762791518751983, "learning_rate": 0.000297661576398715, "loss": 0.2445, "step": 2605 }, { "epoch": 1.2263529411764706, "grad_norm": 0.26587783200918225, "learning_rate": 0.0002976538599155704, "loss": 0.1757, "step": 2606 }, { "epoch": 1.2268235294117646, "grad_norm": 0.045837109058340667, "learning_rate": 0.00029764613082206183, "loss": 0.0417, "step": 2607 }, { "epoch": 1.2272941176470589, "grad_norm": 0.06024653698955107, "learning_rate": 0.0002976383891188496, "loss": 0.0219, "step": 2608 }, { "epoch": 1.2277647058823529, "grad_norm": 0.7528280733284374, "learning_rate": 0.0002976306348065947, "loss": 0.7481, "step": 2609 }, { "epoch": 1.228235294117647, "grad_norm": 0.663926635115687, "learning_rate": 0.00029762286788595954, "loss": 0.5502, "step": 2610 }, { "epoch": 1.228705882352941, "grad_norm": 0.1729261750676927, "learning_rate": 0.0002976150883576073, "loss": 0.0791, "step": 2611 }, { "epoch": 1.2291764705882353, "grad_norm": 0.3658262348230341, "learning_rate": 0.0002976072962222025, "loss": 0.358, "step": 2612 }, { "epoch": 1.2296470588235293, "grad_norm": 0.2444255993168782, "learning_rate": 0.0002975994914804106, "loss": 0.2624, "step": 2613 }, { "epoch": 1.2301176470588235, "grad_norm": 0.31916533488069554, "learning_rate": 0.00029759167413289815, "loss": 0.1393, "step": 2614 }, { "epoch": 1.2305882352941175, "grad_norm": 0.7394548813050899, "learning_rate": 0.00029758384418033276, "loss": 0.5237, "step": 2615 }, { "epoch": 1.2310588235294118, "grad_norm": 0.05626908483923904, "learning_rate": 0.00029757600162338326, "loss": 0.0157, "step": 2616 }, { "epoch": 1.231529411764706, "grad_norm": 0.29896879149011735, "learning_rate": 0.0002975681464627193, "loss": 0.2842, "step": 2617 }, { "epoch": 1.232, "grad_norm": 0.4268290036416358, "learning_rate": 0.00029756027869901183, "loss": 0.3356, "step": 2618 }, { "epoch": 1.2324705882352942, "grad_norm": 0.48404294733921455, "learning_rate": 0.0002975523983329328, "loss": 0.3997, "step": 2619 }, { "epoch": 1.2329411764705882, "grad_norm": 0.39797462311061, "learning_rate": 0.00029754450536515517, "loss": 0.2316, "step": 2620 }, { "epoch": 1.2334117647058824, "grad_norm": 0.20689834355773903, "learning_rate": 0.00029753659979635315, "loss": 0.1548, "step": 2621 }, { "epoch": 1.2338823529411764, "grad_norm": 0.21696497001064094, "learning_rate": 0.00029752868162720176, "loss": 0.1815, "step": 2622 }, { "epoch": 1.2343529411764707, "grad_norm": 0.19953920191132068, "learning_rate": 0.0002975207508583774, "loss": 0.1247, "step": 2623 }, { "epoch": 1.2348235294117647, "grad_norm": 0.7488551339894145, "learning_rate": 0.00029751280749055727, "loss": 0.8195, "step": 2624 }, { "epoch": 1.2352941176470589, "grad_norm": 0.18259965016870428, "learning_rate": 0.0002975048515244199, "loss": 0.0688, "step": 2625 }, { "epoch": 1.2357647058823529, "grad_norm": 0.5303789578800371, "learning_rate": 0.0002974968829606447, "loss": 0.3863, "step": 2626 }, { "epoch": 1.236235294117647, "grad_norm": 0.5612741497719763, "learning_rate": 0.0002974889017999122, "loss": 0.3902, "step": 2627 }, { "epoch": 1.236705882352941, "grad_norm": 0.3455359811497816, "learning_rate": 0.00029748090804290403, "loss": 0.3852, "step": 2628 }, { "epoch": 1.2371764705882353, "grad_norm": 0.3634441856729974, "learning_rate": 0.00029747290169030295, "loss": 0.3078, "step": 2629 }, { "epoch": 1.2376470588235293, "grad_norm": 0.4344714163680946, "learning_rate": 0.0002974648827427927, "loss": 0.4004, "step": 2630 }, { "epoch": 1.2381176470588235, "grad_norm": 0.2870187877262733, "learning_rate": 0.0002974568512010582, "loss": 0.3408, "step": 2631 }, { "epoch": 1.2385882352941175, "grad_norm": 0.21261341630280411, "learning_rate": 0.00029744880706578526, "loss": 0.1395, "step": 2632 }, { "epoch": 1.2390588235294118, "grad_norm": 0.27334928555737525, "learning_rate": 0.00029744075033766103, "loss": 0.2739, "step": 2633 }, { "epoch": 1.239529411764706, "grad_norm": 0.3502543186692781, "learning_rate": 0.0002974326810173735, "loss": 0.4973, "step": 2634 }, { "epoch": 1.24, "grad_norm": 0.09962984248397715, "learning_rate": 0.00029742459910561187, "loss": 0.0302, "step": 2635 }, { "epoch": 1.2404705882352942, "grad_norm": 0.5013793162639162, "learning_rate": 0.00029741650460306636, "loss": 0.4336, "step": 2636 }, { "epoch": 1.2409411764705882, "grad_norm": 0.12652083384545096, "learning_rate": 0.0002974083975104283, "loss": 0.0575, "step": 2637 }, { "epoch": 1.2414117647058824, "grad_norm": 0.28984604589572815, "learning_rate": 0.00029740027782839003, "loss": 0.1651, "step": 2638 }, { "epoch": 1.2418823529411764, "grad_norm": 0.45902864055559817, "learning_rate": 0.00029739214555764504, "loss": 0.3739, "step": 2639 }, { "epoch": 1.2423529411764707, "grad_norm": 0.18568803700346545, "learning_rate": 0.00029738400069888787, "loss": 0.1077, "step": 2640 }, { "epoch": 1.2428235294117647, "grad_norm": 0.060558417596777235, "learning_rate": 0.00029737584325281415, "loss": 0.0489, "step": 2641 }, { "epoch": 1.2432941176470589, "grad_norm": 0.539926472250174, "learning_rate": 0.0002973676732201205, "loss": 0.4534, "step": 2642 }, { "epoch": 1.2437647058823529, "grad_norm": 0.3654022109630717, "learning_rate": 0.00029735949060150476, "loss": 0.315, "step": 2643 }, { "epoch": 1.244235294117647, "grad_norm": 0.45528799888995714, "learning_rate": 0.0002973512953976657, "loss": 0.4042, "step": 2644 }, { "epoch": 1.244705882352941, "grad_norm": 0.21572756515662792, "learning_rate": 0.0002973430876093033, "loss": 0.1555, "step": 2645 }, { "epoch": 1.2451764705882353, "grad_norm": 0.23856898360964832, "learning_rate": 0.00029733486723711845, "loss": 0.3242, "step": 2646 }, { "epoch": 1.2456470588235293, "grad_norm": 0.4645105562241315, "learning_rate": 0.0002973266342818133, "loss": 0.4571, "step": 2647 }, { "epoch": 1.2461176470588236, "grad_norm": 0.46592727136573264, "learning_rate": 0.00029731838874409093, "loss": 0.465, "step": 2648 }, { "epoch": 1.2465882352941176, "grad_norm": 0.6014275555421285, "learning_rate": 0.0002973101306246556, "loss": 0.6744, "step": 2649 }, { "epoch": 1.2470588235294118, "grad_norm": 0.10329213895515407, "learning_rate": 0.0002973018599242125, "loss": 0.0541, "step": 2650 }, { "epoch": 1.247529411764706, "grad_norm": 0.4604671360278471, "learning_rate": 0.0002972935766434681, "loss": 0.3461, "step": 2651 }, { "epoch": 1.248, "grad_norm": 0.5158127289331362, "learning_rate": 0.00029728528078312976, "loss": 0.7115, "step": 2652 }, { "epoch": 1.248470588235294, "grad_norm": 0.5144191593042768, "learning_rate": 0.00029727697234390604, "loss": 0.6479, "step": 2653 }, { "epoch": 1.2489411764705882, "grad_norm": 0.427097537715808, "learning_rate": 0.0002972686513265065, "loss": 0.3719, "step": 2654 }, { "epoch": 1.2494117647058824, "grad_norm": 0.35120303979830936, "learning_rate": 0.00029726031773164176, "loss": 0.3418, "step": 2655 }, { "epoch": 1.2498823529411764, "grad_norm": 0.2922358395719514, "learning_rate": 0.00029725197156002357, "loss": 0.1343, "step": 2656 }, { "epoch": 1.2503529411764707, "grad_norm": 0.44138806561098315, "learning_rate": 0.00029724361281236475, "loss": 0.4252, "step": 2657 }, { "epoch": 1.2508235294117647, "grad_norm": 0.34445484981582236, "learning_rate": 0.0002972352414893792, "loss": 0.3133, "step": 2658 }, { "epoch": 1.251294117647059, "grad_norm": 0.5702654455652356, "learning_rate": 0.00029722685759178187, "loss": 0.4952, "step": 2659 }, { "epoch": 1.251764705882353, "grad_norm": 0.21606256475685595, "learning_rate": 0.0002972184611202886, "loss": 0.3428, "step": 2660 }, { "epoch": 1.2522352941176471, "grad_norm": 0.2703219842062109, "learning_rate": 0.00029721005207561683, "loss": 0.2192, "step": 2661 }, { "epoch": 1.2527058823529411, "grad_norm": 0.6250899539448489, "learning_rate": 0.00029720163045848444, "loss": 0.4036, "step": 2662 }, { "epoch": 1.2531764705882353, "grad_norm": 0.37406346215239994, "learning_rate": 0.0002971931962696109, "loss": 0.4575, "step": 2663 }, { "epoch": 1.2536470588235293, "grad_norm": 0.3312289032797347, "learning_rate": 0.00029718474950971626, "loss": 0.3136, "step": 2664 }, { "epoch": 1.2541176470588236, "grad_norm": 0.39460430244219424, "learning_rate": 0.0002971762901795222, "loss": 0.3099, "step": 2665 }, { "epoch": 1.2545882352941176, "grad_norm": 0.39751574841848275, "learning_rate": 0.000297167818279751, "loss": 0.4395, "step": 2666 }, { "epoch": 1.2550588235294118, "grad_norm": 0.5286072294751429, "learning_rate": 0.0002971593338111263, "loss": 0.3864, "step": 2667 }, { "epoch": 1.255529411764706, "grad_norm": 0.41279369353673717, "learning_rate": 0.0002971508367743727, "loss": 0.3405, "step": 2668 }, { "epoch": 1.256, "grad_norm": 0.3031799229988658, "learning_rate": 0.0002971423271702158, "loss": 0.2429, "step": 2669 }, { "epoch": 1.256470588235294, "grad_norm": 0.4773100321866894, "learning_rate": 0.00029713380499938246, "loss": 0.0805, "step": 2670 }, { "epoch": 1.2569411764705882, "grad_norm": 1.0702757802046314, "learning_rate": 0.00029712527026260047, "loss": 1.6237, "step": 2671 }, { "epoch": 1.2574117647058825, "grad_norm": 0.21702675146195596, "learning_rate": 0.0002971167229605988, "loss": 0.189, "step": 2672 }, { "epoch": 1.2578823529411765, "grad_norm": 0.41443011787149875, "learning_rate": 0.00029710816309410737, "loss": 0.2011, "step": 2673 }, { "epoch": 1.2583529411764705, "grad_norm": 0.30234413867673293, "learning_rate": 0.00029709959066385723, "loss": 0.1358, "step": 2674 }, { "epoch": 1.2588235294117647, "grad_norm": 0.4150562147387654, "learning_rate": 0.0002970910056705806, "loss": 0.356, "step": 2675 }, { "epoch": 1.259294117647059, "grad_norm": 0.8588675394307117, "learning_rate": 0.0002970824081150105, "loss": 0.8411, "step": 2676 }, { "epoch": 1.259764705882353, "grad_norm": 0.2548350951847752, "learning_rate": 0.00029707379799788134, "loss": 0.2459, "step": 2677 }, { "epoch": 1.2602352941176471, "grad_norm": 0.0385669998060624, "learning_rate": 0.0002970651753199285, "loss": 0.0107, "step": 2678 }, { "epoch": 1.2607058823529411, "grad_norm": 0.40522619324202336, "learning_rate": 0.0002970565400818883, "loss": 0.1335, "step": 2679 }, { "epoch": 1.2611764705882353, "grad_norm": 2.156611885256392, "learning_rate": 0.0002970478922844983, "loss": 0.8224, "step": 2680 }, { "epoch": 1.2616470588235293, "grad_norm": 0.6313126799278352, "learning_rate": 0.00029703923192849694, "loss": 0.459, "step": 2681 }, { "epoch": 1.2621176470588236, "grad_norm": 0.3276881262325571, "learning_rate": 0.000297030559014624, "loss": 0.2234, "step": 2682 }, { "epoch": 1.2625882352941176, "grad_norm": 0.5422204959105388, "learning_rate": 0.00029702187354362016, "loss": 0.2877, "step": 2683 }, { "epoch": 1.2630588235294118, "grad_norm": 0.4312276070540354, "learning_rate": 0.0002970131755162272, "loss": 0.1942, "step": 2684 }, { "epoch": 1.263529411764706, "grad_norm": 0.859663893765187, "learning_rate": 0.0002970044649331879, "loss": 0.4257, "step": 2685 }, { "epoch": 1.264, "grad_norm": 0.4136434332501217, "learning_rate": 0.0002969957417952463, "loss": 0.2853, "step": 2686 }, { "epoch": 1.264470588235294, "grad_norm": 1.915604902300987, "learning_rate": 0.00029698700610314725, "loss": 0.8439, "step": 2687 }, { "epoch": 1.2649411764705882, "grad_norm": 0.24990498895495356, "learning_rate": 0.000296978257857637, "loss": 0.1474, "step": 2688 }, { "epoch": 1.2654117647058825, "grad_norm": 0.34582509638576686, "learning_rate": 0.00029696949705946255, "loss": 0.4223, "step": 2689 }, { "epoch": 1.2658823529411765, "grad_norm": 0.30777663423187346, "learning_rate": 0.0002969607237093722, "loss": 0.3181, "step": 2690 }, { "epoch": 1.2663529411764705, "grad_norm": 0.8505187133047334, "learning_rate": 0.00029695193780811523, "loss": 0.7961, "step": 2691 }, { "epoch": 1.2668235294117647, "grad_norm": 1.0434507257671946, "learning_rate": 0.0002969431393564419, "loss": 0.3679, "step": 2692 }, { "epoch": 1.267294117647059, "grad_norm": 0.5263492114798138, "learning_rate": 0.0002969343283551038, "loss": 0.343, "step": 2693 }, { "epoch": 1.267764705882353, "grad_norm": 0.9458227322776281, "learning_rate": 0.0002969255048048533, "loss": 0.859, "step": 2694 }, { "epoch": 1.2682352941176471, "grad_norm": 0.29349517694096733, "learning_rate": 0.00029691666870644405, "loss": 0.2616, "step": 2695 }, { "epoch": 1.2687058823529411, "grad_norm": 0.4971357666094454, "learning_rate": 0.0002969078200606307, "loss": 0.2663, "step": 2696 }, { "epoch": 1.2691764705882354, "grad_norm": 0.5415679575425901, "learning_rate": 0.0002968989588681688, "loss": 0.405, "step": 2697 }, { "epoch": 1.2696470588235294, "grad_norm": 0.1350495034588534, "learning_rate": 0.0002968900851298154, "loss": 0.0523, "step": 2698 }, { "epoch": 1.2701176470588236, "grad_norm": 0.20456217394501455, "learning_rate": 0.00029688119884632827, "loss": 0.1654, "step": 2699 }, { "epoch": 1.2705882352941176, "grad_norm": 0.6598199866948524, "learning_rate": 0.0002968723000184662, "loss": 0.3686, "step": 2700 }, { "epoch": 1.2710588235294118, "grad_norm": 0.990148556704317, "learning_rate": 0.0002968633886469894, "loss": 0.4984, "step": 2701 }, { "epoch": 1.271529411764706, "grad_norm": 0.5502874535110568, "learning_rate": 0.00029685446473265876, "loss": 0.4618, "step": 2702 }, { "epoch": 1.272, "grad_norm": 0.8280775788274345, "learning_rate": 0.00029684552827623657, "loss": 0.7106, "step": 2703 }, { "epoch": 1.272470588235294, "grad_norm": 0.13797766660932032, "learning_rate": 0.000296836579278486, "loss": 0.0968, "step": 2704 }, { "epoch": 1.2729411764705882, "grad_norm": 0.3302771622990252, "learning_rate": 0.0002968276177401713, "loss": 0.1489, "step": 2705 }, { "epoch": 1.2734117647058825, "grad_norm": 0.44889495833521775, "learning_rate": 0.00029681864366205786, "loss": 0.2105, "step": 2706 }, { "epoch": 1.2738823529411765, "grad_norm": 0.2153349586758529, "learning_rate": 0.0002968096570449121, "loss": 0.1477, "step": 2707 }, { "epoch": 1.2743529411764705, "grad_norm": 0.6572257368612547, "learning_rate": 0.0002968006578895016, "loss": 0.3116, "step": 2708 }, { "epoch": 1.2748235294117647, "grad_norm": 0.11217700619760604, "learning_rate": 0.0002967916461965948, "loss": 0.0423, "step": 2709 }, { "epoch": 1.275294117647059, "grad_norm": 0.8147888511691228, "learning_rate": 0.00029678262196696143, "loss": 0.4168, "step": 2710 }, { "epoch": 1.275764705882353, "grad_norm": 0.2636793090481927, "learning_rate": 0.0002967735852013722, "loss": 0.1331, "step": 2711 }, { "epoch": 1.2762352941176471, "grad_norm": 0.324111403747283, "learning_rate": 0.0002967645359005988, "loss": 0.2221, "step": 2712 }, { "epoch": 1.2767058823529411, "grad_norm": 0.3076543087575257, "learning_rate": 0.0002967554740654142, "loss": 0.2113, "step": 2713 }, { "epoch": 1.2771764705882354, "grad_norm": 0.9384367669952338, "learning_rate": 0.0002967463996965923, "loss": 0.7152, "step": 2714 }, { "epoch": 1.2776470588235294, "grad_norm": 0.6104793841027423, "learning_rate": 0.0002967373127949081, "loss": 0.4665, "step": 2715 }, { "epoch": 1.2781176470588236, "grad_norm": 0.14848303751276365, "learning_rate": 0.0002967282133611375, "loss": 0.1163, "step": 2716 }, { "epoch": 1.2785882352941176, "grad_norm": 0.22882951852835062, "learning_rate": 0.0002967191013960579, "loss": 0.1331, "step": 2717 }, { "epoch": 1.2790588235294118, "grad_norm": 0.7260465257797073, "learning_rate": 0.00029670997690044745, "loss": 0.7481, "step": 2718 }, { "epoch": 1.2795294117647058, "grad_norm": 0.4591426858359795, "learning_rate": 0.0002967008398750853, "loss": 0.3953, "step": 2719 }, { "epoch": 1.28, "grad_norm": 0.38004124075929807, "learning_rate": 0.00029669169032075186, "loss": 0.3489, "step": 2720 }, { "epoch": 1.280470588235294, "grad_norm": 0.31693089990290296, "learning_rate": 0.0002966825282382285, "loss": 0.395, "step": 2721 }, { "epoch": 1.2809411764705882, "grad_norm": 0.4004915480218354, "learning_rate": 0.0002966733536282978, "loss": 0.2902, "step": 2722 }, { "epoch": 1.2814117647058825, "grad_norm": 0.06599442684593693, "learning_rate": 0.0002966641664917433, "loss": 0.0132, "step": 2723 }, { "epoch": 1.2818823529411765, "grad_norm": 0.5282336039021597, "learning_rate": 0.0002966549668293495, "loss": 0.4199, "step": 2724 }, { "epoch": 1.2823529411764705, "grad_norm": 0.5540178285269307, "learning_rate": 0.0002966457546419023, "loss": 0.4198, "step": 2725 }, { "epoch": 1.2828235294117647, "grad_norm": 0.5071929603320354, "learning_rate": 0.0002966365299301883, "loss": 0.437, "step": 2726 }, { "epoch": 1.283294117647059, "grad_norm": 0.2838267961185546, "learning_rate": 0.00029662729269499546, "loss": 0.2357, "step": 2727 }, { "epoch": 1.283764705882353, "grad_norm": 0.22086907807517647, "learning_rate": 0.00029661804293711254, "loss": 0.2056, "step": 2728 }, { "epoch": 1.2842352941176471, "grad_norm": 0.3374648050197497, "learning_rate": 0.00029660878065732965, "loss": 0.2975, "step": 2729 }, { "epoch": 1.2847058823529411, "grad_norm": 0.5003402885586092, "learning_rate": 0.0002965995058564377, "loss": 0.3792, "step": 2730 }, { "epoch": 1.2851764705882354, "grad_norm": 0.637178650057431, "learning_rate": 0.000296590218535229, "loss": 0.6103, "step": 2731 }, { "epoch": 1.2856470588235294, "grad_norm": 0.17326018080211553, "learning_rate": 0.00029658091869449657, "loss": 0.0998, "step": 2732 }, { "epoch": 1.2861176470588236, "grad_norm": 0.15963041286286228, "learning_rate": 0.0002965716063350347, "loss": 0.1356, "step": 2733 }, { "epoch": 1.2865882352941176, "grad_norm": 0.12748878385761603, "learning_rate": 0.0002965622814576387, "loss": 0.062, "step": 2734 }, { "epoch": 1.2870588235294118, "grad_norm": 0.23245652503066533, "learning_rate": 0.000296552944063105, "loss": 0.1323, "step": 2735 }, { "epoch": 1.2875294117647058, "grad_norm": 0.5813016630542353, "learning_rate": 0.0002965435941522311, "loss": 0.5262, "step": 2736 }, { "epoch": 1.288, "grad_norm": 0.24248507844665032, "learning_rate": 0.00029653423172581534, "loss": 0.1819, "step": 2737 }, { "epoch": 1.288470588235294, "grad_norm": 0.4233179084094795, "learning_rate": 0.0002965248567846575, "loss": 0.4874, "step": 2738 }, { "epoch": 1.2889411764705883, "grad_norm": 0.33248900885698723, "learning_rate": 0.0002965154693295582, "loss": 0.3123, "step": 2739 }, { "epoch": 1.2894117647058825, "grad_norm": 0.31987037670956553, "learning_rate": 0.00029650606936131913, "loss": 0.4717, "step": 2740 }, { "epoch": 1.2898823529411765, "grad_norm": 0.3490231403849886, "learning_rate": 0.0002964966568807432, "loss": 0.3665, "step": 2741 }, { "epoch": 1.2903529411764705, "grad_norm": 0.5880033325636231, "learning_rate": 0.00029648723188863417, "loss": 0.7339, "step": 2742 }, { "epoch": 1.2908235294117647, "grad_norm": 0.1320565191047133, "learning_rate": 0.000296477794385797, "loss": 0.0684, "step": 2743 }, { "epoch": 1.291294117647059, "grad_norm": 0.23917044733966109, "learning_rate": 0.00029646834437303777, "loss": 0.1897, "step": 2744 }, { "epoch": 1.291764705882353, "grad_norm": 0.2571671033502204, "learning_rate": 0.00029645888185116353, "loss": 0.0841, "step": 2745 }, { "epoch": 1.292235294117647, "grad_norm": 0.696815237203467, "learning_rate": 0.00029644940682098234, "loss": 0.5426, "step": 2746 }, { "epoch": 1.2927058823529411, "grad_norm": 0.18595210065824289, "learning_rate": 0.00029643991928330346, "loss": 0.0917, "step": 2747 }, { "epoch": 1.2931764705882354, "grad_norm": 0.24329037297391137, "learning_rate": 0.0002964304192389372, "loss": 0.239, "step": 2748 }, { "epoch": 1.2936470588235294, "grad_norm": 0.1554181854473883, "learning_rate": 0.000296420906688695, "loss": 0.093, "step": 2749 }, { "epoch": 1.2941176470588236, "grad_norm": 0.37085091706924217, "learning_rate": 0.00029641138163338907, "loss": 0.1917, "step": 2750 }, { "epoch": 1.2945882352941176, "grad_norm": 0.5349259771316386, "learning_rate": 0.00029640184407383303, "loss": 0.4709, "step": 2751 }, { "epoch": 1.2950588235294118, "grad_norm": 0.5006674639546876, "learning_rate": 0.00029639229401084137, "loss": 0.5346, "step": 2752 }, { "epoch": 1.2955294117647058, "grad_norm": 0.11750598248321817, "learning_rate": 0.00029638273144522983, "loss": 0.0366, "step": 2753 }, { "epoch": 1.296, "grad_norm": 0.47308219686451697, "learning_rate": 0.00029637315637781494, "loss": 0.4929, "step": 2754 }, { "epoch": 1.296470588235294, "grad_norm": 0.3444946022198599, "learning_rate": 0.0002963635688094146, "loss": 0.3435, "step": 2755 }, { "epoch": 1.2969411764705883, "grad_norm": 0.21603419752761616, "learning_rate": 0.0002963539687408476, "loss": 0.1362, "step": 2756 }, { "epoch": 1.2974117647058825, "grad_norm": 0.3600888222376809, "learning_rate": 0.00029634435617293374, "loss": 0.3635, "step": 2757 }, { "epoch": 1.2978823529411765, "grad_norm": 0.2402183043908981, "learning_rate": 0.00029633473110649407, "loss": 0.2314, "step": 2758 }, { "epoch": 1.2983529411764705, "grad_norm": 0.37395567220649706, "learning_rate": 0.00029632509354235063, "loss": 0.333, "step": 2759 }, { "epoch": 1.2988235294117647, "grad_norm": 0.3993967766966197, "learning_rate": 0.00029631544348132643, "loss": 0.4346, "step": 2760 }, { "epoch": 1.299294117647059, "grad_norm": 0.22653916648766284, "learning_rate": 0.0002963057809242457, "loss": 0.2954, "step": 2761 }, { "epoch": 1.299764705882353, "grad_norm": 0.1574121563922957, "learning_rate": 0.00029629610587193365, "loss": 0.1146, "step": 2762 }, { "epoch": 1.300235294117647, "grad_norm": 0.5052723584222512, "learning_rate": 0.00029628641832521663, "loss": 0.4606, "step": 2763 }, { "epoch": 1.3007058823529412, "grad_norm": 0.6453107172723608, "learning_rate": 0.0002962767182849219, "loss": 0.5646, "step": 2764 }, { "epoch": 1.3011764705882354, "grad_norm": 0.39951617688654883, "learning_rate": 0.00029626700575187794, "loss": 0.383, "step": 2765 }, { "epoch": 1.3016470588235294, "grad_norm": 0.3999995607219263, "learning_rate": 0.0002962572807269143, "loss": 0.4987, "step": 2766 }, { "epoch": 1.3021176470588236, "grad_norm": 0.29833140184094636, "learning_rate": 0.00029624754321086146, "loss": 0.2486, "step": 2767 }, { "epoch": 1.3025882352941176, "grad_norm": 0.35262332713213934, "learning_rate": 0.00029623779320455106, "loss": 0.3784, "step": 2768 }, { "epoch": 1.3030588235294118, "grad_norm": 0.5443128573814635, "learning_rate": 0.00029622803070881586, "loss": 0.3892, "step": 2769 }, { "epoch": 1.3035294117647058, "grad_norm": 0.24053351347225113, "learning_rate": 0.00029621825572448965, "loss": 0.2993, "step": 2770 }, { "epoch": 1.304, "grad_norm": 0.43898285897665057, "learning_rate": 0.0002962084682524071, "loss": 0.59, "step": 2771 }, { "epoch": 1.304470588235294, "grad_norm": 0.528226897426802, "learning_rate": 0.0002961986682934043, "loss": 0.4389, "step": 2772 }, { "epoch": 1.3049411764705883, "grad_norm": 0.25884075543617285, "learning_rate": 0.0002961888558483181, "loss": 0.2044, "step": 2773 }, { "epoch": 1.3054117647058823, "grad_norm": 0.5426569547555246, "learning_rate": 0.0002961790309179866, "loss": 0.6029, "step": 2774 }, { "epoch": 1.3058823529411765, "grad_norm": 0.04916617301001648, "learning_rate": 0.00029616919350324873, "loss": 0.019, "step": 2775 }, { "epoch": 1.3063529411764705, "grad_norm": 0.4547207295943256, "learning_rate": 0.0002961593436049449, "loss": 0.4409, "step": 2776 }, { "epoch": 1.3068235294117647, "grad_norm": 0.24025721142302006, "learning_rate": 0.0002961494812239162, "loss": 0.1603, "step": 2777 }, { "epoch": 1.307294117647059, "grad_norm": 0.10746201652067787, "learning_rate": 0.00029613960636100487, "loss": 0.0586, "step": 2778 }, { "epoch": 1.307764705882353, "grad_norm": 0.19199953735243816, "learning_rate": 0.0002961297190170544, "loss": 0.1118, "step": 2779 }, { "epoch": 1.308235294117647, "grad_norm": 0.4301729262235553, "learning_rate": 0.00029611981919290916, "loss": 0.3011, "step": 2780 }, { "epoch": 1.3087058823529412, "grad_norm": 0.832960652786248, "learning_rate": 0.0002961099068894146, "loss": 0.7277, "step": 2781 }, { "epoch": 1.3091764705882354, "grad_norm": 0.3977737627560336, "learning_rate": 0.00029609998210741744, "loss": 0.4238, "step": 2782 }, { "epoch": 1.3096470588235294, "grad_norm": 0.17723452004394497, "learning_rate": 0.0002960900448477651, "loss": 0.1624, "step": 2783 }, { "epoch": 1.3101176470588236, "grad_norm": 0.6093085776374649, "learning_rate": 0.0002960800951113064, "loss": 0.787, "step": 2784 }, { "epoch": 1.3105882352941176, "grad_norm": 0.7231057984443439, "learning_rate": 0.00029607013289889106, "loss": 0.6998, "step": 2785 }, { "epoch": 1.3110588235294118, "grad_norm": 0.35252352087848327, "learning_rate": 0.0002960601582113699, "loss": 0.2832, "step": 2786 }, { "epoch": 1.3115294117647058, "grad_norm": 0.6432339618510879, "learning_rate": 0.00029605017104959475, "loss": 0.6785, "step": 2787 }, { "epoch": 1.312, "grad_norm": 0.19720358343650995, "learning_rate": 0.00029604017141441865, "loss": 0.1537, "step": 2788 }, { "epoch": 1.312470588235294, "grad_norm": 0.2587360277656072, "learning_rate": 0.0002960301593066956, "loss": 0.1101, "step": 2789 }, { "epoch": 1.3129411764705883, "grad_norm": 0.46331195127858826, "learning_rate": 0.0002960201347272807, "loss": 0.3192, "step": 2790 }, { "epoch": 1.3134117647058823, "grad_norm": 0.6646749791663068, "learning_rate": 0.00029601009767703007, "loss": 0.3774, "step": 2791 }, { "epoch": 1.3138823529411765, "grad_norm": 0.3461458199010586, "learning_rate": 0.0002960000481568009, "loss": 0.4019, "step": 2792 }, { "epoch": 1.3143529411764705, "grad_norm": 0.2100016758077743, "learning_rate": 0.0002959899861674515, "loss": 0.2519, "step": 2793 }, { "epoch": 1.3148235294117647, "grad_norm": 0.2552959459943333, "learning_rate": 0.0002959799117098412, "loss": 0.1449, "step": 2794 }, { "epoch": 1.315294117647059, "grad_norm": 0.29778812215792416, "learning_rate": 0.00029596982478483037, "loss": 0.2674, "step": 2795 }, { "epoch": 1.315764705882353, "grad_norm": 0.5690778961260322, "learning_rate": 0.00029595972539328053, "loss": 0.5825, "step": 2796 }, { "epoch": 1.316235294117647, "grad_norm": 0.6174590877774653, "learning_rate": 0.00029594961353605426, "loss": 0.5256, "step": 2797 }, { "epoch": 1.3167058823529412, "grad_norm": 0.4366713398458223, "learning_rate": 0.0002959394892140151, "loss": 0.3503, "step": 2798 }, { "epoch": 1.3171764705882354, "grad_norm": 0.36226916539669185, "learning_rate": 0.00029592935242802774, "loss": 0.4671, "step": 2799 }, { "epoch": 1.3176470588235294, "grad_norm": 0.1263836232124743, "learning_rate": 0.0002959192031789579, "loss": 0.1146, "step": 2800 }, { "epoch": 1.3181176470588234, "grad_norm": 0.762989007388553, "learning_rate": 0.00029590904146767236, "loss": 0.4561, "step": 2801 }, { "epoch": 1.3185882352941176, "grad_norm": 0.28340856872637693, "learning_rate": 0.000295898867295039, "loss": 0.1442, "step": 2802 }, { "epoch": 1.3190588235294118, "grad_norm": 0.6811721514337395, "learning_rate": 0.0002958886806619268, "loss": 0.5392, "step": 2803 }, { "epoch": 1.3195294117647058, "grad_norm": 0.7213310029427231, "learning_rate": 0.00029587848156920564, "loss": 0.6767, "step": 2804 }, { "epoch": 1.32, "grad_norm": 0.39791859860345824, "learning_rate": 0.0002958682700177466, "loss": 0.4385, "step": 2805 }, { "epoch": 1.320470588235294, "grad_norm": 0.42723346664858697, "learning_rate": 0.00029585804600842187, "loss": 0.3797, "step": 2806 }, { "epoch": 1.3209411764705883, "grad_norm": 0.3562512204122193, "learning_rate": 0.00029584780954210454, "loss": 0.2748, "step": 2807 }, { "epoch": 1.3214117647058823, "grad_norm": 0.45475403416896215, "learning_rate": 0.0002958375606196689, "loss": 0.7504, "step": 2808 }, { "epoch": 1.3218823529411765, "grad_norm": 0.520850717269651, "learning_rate": 0.0002958272992419903, "loss": 0.4367, "step": 2809 }, { "epoch": 1.3223529411764705, "grad_norm": 0.5485207565656481, "learning_rate": 0.000295817025409945, "loss": 0.4053, "step": 2810 }, { "epoch": 1.3228235294117647, "grad_norm": 0.4034647903958428, "learning_rate": 0.00029580673912441057, "loss": 0.2666, "step": 2811 }, { "epoch": 1.323294117647059, "grad_norm": 0.30582841886472967, "learning_rate": 0.00029579644038626546, "loss": 0.197, "step": 2812 }, { "epoch": 1.323764705882353, "grad_norm": 0.28051413831747446, "learning_rate": 0.00029578612919638913, "loss": 0.3283, "step": 2813 }, { "epoch": 1.324235294117647, "grad_norm": 0.5799052596579437, "learning_rate": 0.0002957758055556623, "loss": 0.343, "step": 2814 }, { "epoch": 1.3247058823529412, "grad_norm": 0.31269819045282976, "learning_rate": 0.0002957654694649667, "loss": 0.4302, "step": 2815 }, { "epoch": 1.3251764705882354, "grad_norm": 0.27932124741721576, "learning_rate": 0.000295755120925185, "loss": 0.1646, "step": 2816 }, { "epoch": 1.3256470588235294, "grad_norm": 0.5007368231524911, "learning_rate": 0.0002957447599372011, "loss": 0.7184, "step": 2817 }, { "epoch": 1.3261176470588234, "grad_norm": 0.29167454143872523, "learning_rate": 0.00029573438650189974, "loss": 0.306, "step": 2818 }, { "epoch": 1.3265882352941176, "grad_norm": 0.7184066040319484, "learning_rate": 0.000295724000620167, "loss": 0.8796, "step": 2819 }, { "epoch": 1.3270588235294118, "grad_norm": 0.30487328969533734, "learning_rate": 0.00029571360229288984, "loss": 0.5113, "step": 2820 }, { "epoch": 1.3275294117647058, "grad_norm": 0.24324595669437407, "learning_rate": 0.0002957031915209563, "loss": 0.209, "step": 2821 }, { "epoch": 1.328, "grad_norm": 0.45782331733693704, "learning_rate": 0.00029569276830525553, "loss": 0.4208, "step": 2822 }, { "epoch": 1.328470588235294, "grad_norm": 0.10535974000417725, "learning_rate": 0.00029568233264667775, "loss": 0.05, "step": 2823 }, { "epoch": 1.3289411764705883, "grad_norm": 0.19621933914917752, "learning_rate": 0.00029567188454611415, "loss": 0.1681, "step": 2824 }, { "epoch": 1.3294117647058823, "grad_norm": 0.15705879045141832, "learning_rate": 0.00029566142400445716, "loss": 0.1126, "step": 2825 }, { "epoch": 1.3298823529411765, "grad_norm": 0.44845216590698594, "learning_rate": 0.0002956509510226001, "loss": 0.3878, "step": 2826 }, { "epoch": 1.3303529411764705, "grad_norm": 0.2744704341788826, "learning_rate": 0.00029564046560143735, "loss": 0.4284, "step": 2827 }, { "epoch": 1.3308235294117647, "grad_norm": 0.29787081391824705, "learning_rate": 0.0002956299677418645, "loss": 0.3907, "step": 2828 }, { "epoch": 1.331294117647059, "grad_norm": 0.284578901489107, "learning_rate": 0.000295619457444778, "loss": 0.1801, "step": 2829 }, { "epoch": 1.331764705882353, "grad_norm": 0.3374440238359162, "learning_rate": 0.00029560893471107573, "loss": 0.4131, "step": 2830 }, { "epoch": 1.332235294117647, "grad_norm": 0.3475083952940664, "learning_rate": 0.00029559839954165616, "loss": 0.3234, "step": 2831 }, { "epoch": 1.3327058823529412, "grad_norm": 0.25295991592247735, "learning_rate": 0.00029558785193741914, "loss": 0.1888, "step": 2832 }, { "epoch": 1.3331764705882354, "grad_norm": 0.21059184298267894, "learning_rate": 0.0002955772918992654, "loss": 0.0753, "step": 2833 }, { "epoch": 1.3336470588235294, "grad_norm": 0.3071150772065728, "learning_rate": 0.00029556671942809694, "loss": 0.2302, "step": 2834 }, { "epoch": 1.3341176470588234, "grad_norm": 0.1207563553283702, "learning_rate": 0.00029555613452481657, "loss": 0.0263, "step": 2835 }, { "epoch": 1.3345882352941176, "grad_norm": 0.9101223403661268, "learning_rate": 0.00029554553719032844, "loss": 0.686, "step": 2836 }, { "epoch": 1.3350588235294119, "grad_norm": 0.11775520027657835, "learning_rate": 0.00029553492742553754, "loss": 0.1293, "step": 2837 }, { "epoch": 1.3355294117647059, "grad_norm": 0.21340812108683324, "learning_rate": 0.00029552430523134993, "loss": 0.1644, "step": 2838 }, { "epoch": 1.336, "grad_norm": 0.3314170115736156, "learning_rate": 0.0002955136706086729, "loss": 0.2514, "step": 2839 }, { "epoch": 1.336470588235294, "grad_norm": 0.45421175518310286, "learning_rate": 0.0002955030235584146, "loss": 0.4219, "step": 2840 }, { "epoch": 1.3369411764705883, "grad_norm": 0.160123576631066, "learning_rate": 0.00029549236408148453, "loss": 0.155, "step": 2841 }, { "epoch": 1.3374117647058823, "grad_norm": 0.35236722457612873, "learning_rate": 0.00029548169217879285, "loss": 0.4965, "step": 2842 }, { "epoch": 1.3378823529411765, "grad_norm": 0.3407272086141946, "learning_rate": 0.000295471007851251, "loss": 0.3138, "step": 2843 }, { "epoch": 1.3383529411764705, "grad_norm": 0.2555075114831745, "learning_rate": 0.00029546031109977164, "loss": 0.235, "step": 2844 }, { "epoch": 1.3388235294117647, "grad_norm": 0.40017394754282226, "learning_rate": 0.00029544960192526824, "loss": 0.3717, "step": 2845 }, { "epoch": 1.3392941176470587, "grad_norm": 0.379566061059355, "learning_rate": 0.0002954388803286554, "loss": 0.4152, "step": 2846 }, { "epoch": 1.339764705882353, "grad_norm": 0.46481890740356896, "learning_rate": 0.0002954281463108488, "loss": 0.4244, "step": 2847 }, { "epoch": 1.340235294117647, "grad_norm": 0.6538755192661719, "learning_rate": 0.0002954173998727652, "loss": 0.7065, "step": 2848 }, { "epoch": 1.3407058823529412, "grad_norm": 0.26239423691865604, "learning_rate": 0.00029540664101532227, "loss": 0.1346, "step": 2849 }, { "epoch": 1.3411764705882354, "grad_norm": 0.36367234990978686, "learning_rate": 0.0002953958697394391, "loss": 0.2672, "step": 2850 }, { "epoch": 1.3416470588235294, "grad_norm": 0.3130725457866976, "learning_rate": 0.0002953850860460355, "loss": 0.3208, "step": 2851 }, { "epoch": 1.3421176470588234, "grad_norm": 0.15451010978245194, "learning_rate": 0.00029537428993603237, "loss": 0.1478, "step": 2852 }, { "epoch": 1.3425882352941176, "grad_norm": 0.33021641032848587, "learning_rate": 0.00029536348141035183, "loss": 0.2926, "step": 2853 }, { "epoch": 1.3430588235294119, "grad_norm": 0.38624364941786166, "learning_rate": 0.000295352660469917, "loss": 0.5343, "step": 2854 }, { "epoch": 1.3435294117647059, "grad_norm": 0.39933428737903476, "learning_rate": 0.00029534182711565203, "loss": 0.402, "step": 2855 }, { "epoch": 1.3439999999999999, "grad_norm": 0.44392626074440783, "learning_rate": 0.0002953309813484821, "loss": 0.5377, "step": 2856 }, { "epoch": 1.344470588235294, "grad_norm": 0.3200810848094909, "learning_rate": 0.0002953201231693336, "loss": 0.3523, "step": 2857 }, { "epoch": 1.3449411764705883, "grad_norm": 0.6287771313150773, "learning_rate": 0.00029530925257913366, "loss": 0.6258, "step": 2858 }, { "epoch": 1.3454117647058823, "grad_norm": 0.3156408016708348, "learning_rate": 0.0002952983695788109, "loss": 0.3356, "step": 2859 }, { "epoch": 1.3458823529411765, "grad_norm": 0.3405774690085917, "learning_rate": 0.00029528747416929463, "loss": 0.3453, "step": 2860 }, { "epoch": 1.3463529411764705, "grad_norm": 0.5504000930499816, "learning_rate": 0.0002952765663515155, "loss": 0.345, "step": 2861 }, { "epoch": 1.3468235294117648, "grad_norm": 0.19633633452108515, "learning_rate": 0.000295265646126405, "loss": 0.1294, "step": 2862 }, { "epoch": 1.3472941176470588, "grad_norm": 0.10811381214700874, "learning_rate": 0.00029525471349489576, "loss": 0.0463, "step": 2863 }, { "epoch": 1.347764705882353, "grad_norm": 0.5099459560371257, "learning_rate": 0.00029524376845792153, "loss": 0.4087, "step": 2864 }, { "epoch": 1.348235294117647, "grad_norm": 0.08568628564088941, "learning_rate": 0.0002952328110164171, "loss": 0.0508, "step": 2865 }, { "epoch": 1.3487058823529412, "grad_norm": 0.3237887009095201, "learning_rate": 0.0002952218411713182, "loss": 0.2879, "step": 2866 }, { "epoch": 1.3491764705882354, "grad_norm": 0.5757601450924674, "learning_rate": 0.0002952108589235618, "loss": 0.3632, "step": 2867 }, { "epoch": 1.3496470588235294, "grad_norm": 0.3987915907622555, "learning_rate": 0.0002951998642740858, "loss": 0.3117, "step": 2868 }, { "epoch": 1.3501176470588234, "grad_norm": 0.33586430472841244, "learning_rate": 0.00029518885722382913, "loss": 0.3576, "step": 2869 }, { "epoch": 1.3505882352941176, "grad_norm": 0.3278479738004868, "learning_rate": 0.00029517783777373194, "loss": 0.2845, "step": 2870 }, { "epoch": 1.3510588235294119, "grad_norm": 0.6839046427120556, "learning_rate": 0.0002951668059247353, "loss": 0.4628, "step": 2871 }, { "epoch": 1.3515294117647059, "grad_norm": 0.22116156263337292, "learning_rate": 0.0002951557616777814, "loss": 0.1596, "step": 2872 }, { "epoch": 1.3519999999999999, "grad_norm": 0.14239671583853658, "learning_rate": 0.0002951447050338135, "loss": 0.0607, "step": 2873 }, { "epoch": 1.352470588235294, "grad_norm": 0.758763951282872, "learning_rate": 0.00029513363599377587, "loss": 0.6308, "step": 2874 }, { "epoch": 1.3529411764705883, "grad_norm": 0.43631388383774405, "learning_rate": 0.00029512255455861375, "loss": 0.5202, "step": 2875 }, { "epoch": 1.3534117647058823, "grad_norm": 0.4661312170444493, "learning_rate": 0.0002951114607292737, "loss": 0.4445, "step": 2876 }, { "epoch": 1.3538823529411765, "grad_norm": 0.18417729082042328, "learning_rate": 0.0002951003545067032, "loss": 0.1166, "step": 2877 }, { "epoch": 1.3543529411764705, "grad_norm": 1.0313609424294112, "learning_rate": 0.0002950892358918506, "loss": 0.607, "step": 2878 }, { "epoch": 1.3548235294117648, "grad_norm": 0.5804642712763999, "learning_rate": 0.00029507810488566563, "loss": 0.4428, "step": 2879 }, { "epoch": 1.3552941176470588, "grad_norm": 0.5645570712257593, "learning_rate": 0.000295066961489099, "loss": 0.497, "step": 2880 }, { "epoch": 1.355764705882353, "grad_norm": 0.2349711159914185, "learning_rate": 0.0002950558057031022, "loss": 0.1811, "step": 2881 }, { "epoch": 1.356235294117647, "grad_norm": 0.7218060883902202, "learning_rate": 0.0002950446375286281, "loss": 0.3789, "step": 2882 }, { "epoch": 1.3567058823529412, "grad_norm": 0.9891400205123986, "learning_rate": 0.00029503345696663054, "loss": 0.5213, "step": 2883 }, { "epoch": 1.3571764705882354, "grad_norm": 0.5532432306885188, "learning_rate": 0.0002950222640180644, "loss": 0.4125, "step": 2884 }, { "epoch": 1.3576470588235294, "grad_norm": 0.419009570175098, "learning_rate": 0.0002950110586838855, "loss": 0.4268, "step": 2885 }, { "epoch": 1.3581176470588234, "grad_norm": 0.5411119898014375, "learning_rate": 0.00029499984096505107, "loss": 0.5786, "step": 2886 }, { "epoch": 1.3585882352941177, "grad_norm": 0.3340200230655476, "learning_rate": 0.00029498861086251887, "loss": 0.3323, "step": 2887 }, { "epoch": 1.3590588235294119, "grad_norm": 0.4220266273863993, "learning_rate": 0.00029497736837724817, "loss": 0.4069, "step": 2888 }, { "epoch": 1.3595294117647059, "grad_norm": 0.6440270242278545, "learning_rate": 0.0002949661135101991, "loss": 0.5884, "step": 2889 }, { "epoch": 1.3599999999999999, "grad_norm": 0.31239947763375464, "learning_rate": 0.0002949548462623329, "loss": 0.4305, "step": 2890 }, { "epoch": 1.360470588235294, "grad_norm": 0.4064987985785841, "learning_rate": 0.0002949435666346118, "loss": 0.4436, "step": 2891 }, { "epoch": 1.3609411764705883, "grad_norm": 0.3230888613787402, "learning_rate": 0.00029493227462799914, "loss": 0.3309, "step": 2892 }, { "epoch": 1.3614117647058823, "grad_norm": 0.1958574185606747, "learning_rate": 0.00029492097024345936, "loss": 0.1539, "step": 2893 }, { "epoch": 1.3618823529411765, "grad_norm": 0.19637791854136827, "learning_rate": 0.00029490965348195794, "loss": 0.1216, "step": 2894 }, { "epoch": 1.3623529411764705, "grad_norm": 0.4740810228070558, "learning_rate": 0.00029489832434446126, "loss": 0.4364, "step": 2895 }, { "epoch": 1.3628235294117648, "grad_norm": 0.35690343344874925, "learning_rate": 0.000294886982831937, "loss": 0.4189, "step": 2896 }, { "epoch": 1.3632941176470588, "grad_norm": 0.04190422432253085, "learning_rate": 0.00029487562894535375, "loss": 0.0082, "step": 2897 }, { "epoch": 1.363764705882353, "grad_norm": 0.4551963710388509, "learning_rate": 0.0002948642626856812, "loss": 0.4357, "step": 2898 }, { "epoch": 1.364235294117647, "grad_norm": 0.03633145686604644, "learning_rate": 0.00029485288405389, "loss": 0.0107, "step": 2899 }, { "epoch": 1.3647058823529412, "grad_norm": 0.16156625124240326, "learning_rate": 0.000294841493050952, "loss": 0.0897, "step": 2900 }, { "epoch": 1.3651764705882352, "grad_norm": 0.2754836327471419, "learning_rate": 0.00029483008967784007, "loss": 0.1689, "step": 2901 }, { "epoch": 1.3656470588235294, "grad_norm": 0.3578332382722541, "learning_rate": 0.0002948186739355281, "loss": 0.3297, "step": 2902 }, { "epoch": 1.3661176470588234, "grad_norm": 0.2779668898570167, "learning_rate": 0.00029480724582499107, "loss": 0.2768, "step": 2903 }, { "epoch": 1.3665882352941177, "grad_norm": 0.5484094439219344, "learning_rate": 0.0002947958053472049, "loss": 0.5487, "step": 2904 }, { "epoch": 1.3670588235294119, "grad_norm": 0.1784882185147978, "learning_rate": 0.0002947843525031468, "loss": 0.1296, "step": 2905 }, { "epoch": 1.3675294117647059, "grad_norm": 0.15470471142621164, "learning_rate": 0.0002947728872937948, "loss": 0.0514, "step": 2906 }, { "epoch": 1.3679999999999999, "grad_norm": 0.23085366358705967, "learning_rate": 0.00029476140972012815, "loss": 0.2332, "step": 2907 }, { "epoch": 1.368470588235294, "grad_norm": 0.17436924120506184, "learning_rate": 0.000294749919783127, "loss": 0.1649, "step": 2908 }, { "epoch": 1.3689411764705883, "grad_norm": 0.37571714832084213, "learning_rate": 0.00029473841748377277, "loss": 0.5069, "step": 2909 }, { "epoch": 1.3694117647058823, "grad_norm": 0.31282772382908225, "learning_rate": 0.0002947269028230477, "loss": 0.3294, "step": 2910 }, { "epoch": 1.3698823529411766, "grad_norm": 0.39650175861724535, "learning_rate": 0.0002947153758019353, "loss": 0.3639, "step": 2911 }, { "epoch": 1.3703529411764706, "grad_norm": 0.30168741437030744, "learning_rate": 0.00029470383642142, "loss": 0.2255, "step": 2912 }, { "epoch": 1.3708235294117648, "grad_norm": 0.5087698986334318, "learning_rate": 0.0002946922846824872, "loss": 0.4844, "step": 2913 }, { "epoch": 1.3712941176470588, "grad_norm": 0.41609977694802064, "learning_rate": 0.00029468072058612367, "loss": 0.5923, "step": 2914 }, { "epoch": 1.371764705882353, "grad_norm": 0.28747925406184677, "learning_rate": 0.0002946691441333169, "loss": 0.3099, "step": 2915 }, { "epoch": 1.372235294117647, "grad_norm": 0.2892146413036879, "learning_rate": 0.0002946575553250556, "loss": 0.3995, "step": 2916 }, { "epoch": 1.3727058823529412, "grad_norm": 0.18970526514078692, "learning_rate": 0.00029464595416232956, "loss": 0.1032, "step": 2917 }, { "epoch": 1.3731764705882352, "grad_norm": 0.35649638788264854, "learning_rate": 0.0002946343406461296, "loss": 0.3653, "step": 2918 }, { "epoch": 1.3736470588235294, "grad_norm": 0.30587296905106276, "learning_rate": 0.0002946227147774475, "loss": 0.3102, "step": 2919 }, { "epoch": 1.3741176470588234, "grad_norm": 0.38156161350148504, "learning_rate": 0.0002946110765572761, "loss": 0.414, "step": 2920 }, { "epoch": 1.3745882352941177, "grad_norm": 0.6391160669552063, "learning_rate": 0.00029459942598660953, "loss": 0.6007, "step": 2921 }, { "epoch": 1.3750588235294119, "grad_norm": 0.3510917185684206, "learning_rate": 0.00029458776306644273, "loss": 0.3649, "step": 2922 }, { "epoch": 1.375529411764706, "grad_norm": 0.29947910527652133, "learning_rate": 0.0002945760877977717, "loss": 0.3729, "step": 2923 }, { "epoch": 1.376, "grad_norm": 0.29762516313463655, "learning_rate": 0.0002945644001815937, "loss": 0.2588, "step": 2924 }, { "epoch": 1.3764705882352941, "grad_norm": 0.3894507542220579, "learning_rate": 0.0002945527002189068, "loss": 0.459, "step": 2925 }, { "epoch": 1.3769411764705883, "grad_norm": 0.6580606715253561, "learning_rate": 0.0002945409879107103, "loss": 1.0905, "step": 2926 }, { "epoch": 1.3774117647058823, "grad_norm": 0.16943582714633293, "learning_rate": 0.00029452926325800445, "loss": 0.1346, "step": 2927 }, { "epoch": 1.3778823529411763, "grad_norm": 0.26096632729831915, "learning_rate": 0.0002945175262617906, "loss": 0.1479, "step": 2928 }, { "epoch": 1.3783529411764706, "grad_norm": 0.24420219120819175, "learning_rate": 0.00029450577692307115, "loss": 0.1385, "step": 2929 }, { "epoch": 1.3788235294117648, "grad_norm": 0.16494544282592022, "learning_rate": 0.0002944940152428496, "loss": 0.0936, "step": 2930 }, { "epoch": 1.3792941176470588, "grad_norm": 0.24189314750800736, "learning_rate": 0.0002944822412221303, "loss": 0.2418, "step": 2931 }, { "epoch": 1.379764705882353, "grad_norm": 0.3276403181282415, "learning_rate": 0.00029447045486191904, "loss": 0.2862, "step": 2932 }, { "epoch": 1.380235294117647, "grad_norm": 0.37579076114906546, "learning_rate": 0.0002944586561632223, "loss": 0.2954, "step": 2933 }, { "epoch": 1.3807058823529412, "grad_norm": 0.3191725642235038, "learning_rate": 0.0002944468451270477, "loss": 0.3164, "step": 2934 }, { "epoch": 1.3811764705882352, "grad_norm": 0.12703769559192224, "learning_rate": 0.000294435021754404, "loss": 0.1236, "step": 2935 }, { "epoch": 1.3816470588235295, "grad_norm": 0.44905496521645766, "learning_rate": 0.000294423186046301, "loss": 0.4489, "step": 2936 }, { "epoch": 1.3821176470588235, "grad_norm": 0.37855889246168245, "learning_rate": 0.00029441133800374957, "loss": 0.332, "step": 2937 }, { "epoch": 1.3825882352941177, "grad_norm": 0.344040993480589, "learning_rate": 0.00029439947762776154, "loss": 0.2493, "step": 2938 }, { "epoch": 1.383058823529412, "grad_norm": 0.2702872482789987, "learning_rate": 0.00029438760491934987, "loss": 0.176, "step": 2939 }, { "epoch": 1.383529411764706, "grad_norm": 0.4946819969950085, "learning_rate": 0.0002943757198795284, "loss": 0.475, "step": 2940 }, { "epoch": 1.384, "grad_norm": 0.05587061670409094, "learning_rate": 0.0002943638225093124, "loss": 0.0312, "step": 2941 }, { "epoch": 1.3844705882352941, "grad_norm": 0.33841164084956954, "learning_rate": 0.00029435191280971783, "loss": 0.337, "step": 2942 }, { "epoch": 1.3849411764705883, "grad_norm": 0.12065397784624539, "learning_rate": 0.0002943399907817619, "loss": 0.0982, "step": 2943 }, { "epoch": 1.3854117647058823, "grad_norm": 0.157472775519409, "learning_rate": 0.0002943280564264627, "loss": 0.1696, "step": 2944 }, { "epoch": 1.3858823529411763, "grad_norm": 0.19650064344011517, "learning_rate": 0.0002943161097448396, "loss": 0.1761, "step": 2945 }, { "epoch": 1.3863529411764706, "grad_norm": 0.5608195914400163, "learning_rate": 0.0002943041507379128, "loss": 0.3872, "step": 2946 }, { "epoch": 1.3868235294117648, "grad_norm": 0.28204952590208104, "learning_rate": 0.00029429217940670383, "loss": 0.2565, "step": 2947 }, { "epoch": 1.3872941176470588, "grad_norm": 0.3412821781339178, "learning_rate": 0.0002942801957522349, "loss": 0.2487, "step": 2948 }, { "epoch": 1.387764705882353, "grad_norm": 0.4519256850196684, "learning_rate": 0.00029426819977552965, "loss": 0.4155, "step": 2949 }, { "epoch": 1.388235294117647, "grad_norm": 0.11022698087930523, "learning_rate": 0.0002942561914776124, "loss": 0.0785, "step": 2950 }, { "epoch": 1.3887058823529412, "grad_norm": 0.30947810595952463, "learning_rate": 0.0002942441708595089, "loss": 0.3185, "step": 2951 }, { "epoch": 1.3891764705882352, "grad_norm": 0.3648708865410561, "learning_rate": 0.0002942321379222457, "loss": 0.3449, "step": 2952 }, { "epoch": 1.3896470588235295, "grad_norm": 0.5684991578842263, "learning_rate": 0.0002942200926668504, "loss": 0.6059, "step": 2953 }, { "epoch": 1.3901176470588235, "grad_norm": 0.2380161542132732, "learning_rate": 0.00029420803509435186, "loss": 0.223, "step": 2954 }, { "epoch": 1.3905882352941177, "grad_norm": 0.2585425753631783, "learning_rate": 0.00029419596520577975, "loss": 0.1629, "step": 2955 }, { "epoch": 1.391058823529412, "grad_norm": 0.23532635802094776, "learning_rate": 0.00029418388300216493, "loss": 0.1035, "step": 2956 }, { "epoch": 1.391529411764706, "grad_norm": 0.48867605259324637, "learning_rate": 0.0002941717884845393, "loss": 0.3605, "step": 2957 }, { "epoch": 1.392, "grad_norm": 0.21203682155487566, "learning_rate": 0.00029415968165393577, "loss": 0.1921, "step": 2958 }, { "epoch": 1.3924705882352941, "grad_norm": 0.19309398994871987, "learning_rate": 0.00029414756251138835, "loss": 0.1147, "step": 2959 }, { "epoch": 1.3929411764705883, "grad_norm": 0.2688765993766873, "learning_rate": 0.000294135431057932, "loss": 0.2249, "step": 2960 }, { "epoch": 1.3934117647058823, "grad_norm": 0.4137763337185269, "learning_rate": 0.0002941232872946029, "loss": 0.576, "step": 2961 }, { "epoch": 1.3938823529411764, "grad_norm": 0.3631292535219081, "learning_rate": 0.00029411113122243814, "loss": 0.502, "step": 2962 }, { "epoch": 1.3943529411764706, "grad_norm": 0.22900415112126188, "learning_rate": 0.0002940989628424759, "loss": 0.2203, "step": 2963 }, { "epoch": 1.3948235294117648, "grad_norm": 0.409757601140222, "learning_rate": 0.0002940867821557554, "loss": 0.3957, "step": 2964 }, { "epoch": 1.3952941176470588, "grad_norm": 0.4193581892492248, "learning_rate": 0.00029407458916331705, "loss": 0.5435, "step": 2965 }, { "epoch": 1.395764705882353, "grad_norm": 0.44011555111701295, "learning_rate": 0.00029406238386620205, "loss": 0.4572, "step": 2966 }, { "epoch": 1.396235294117647, "grad_norm": 0.10560862801188955, "learning_rate": 0.0002940501662654528, "loss": 0.0615, "step": 2967 }, { "epoch": 1.3967058823529412, "grad_norm": 0.43503936626254086, "learning_rate": 0.0002940379363621129, "loss": 0.3029, "step": 2968 }, { "epoch": 1.3971764705882352, "grad_norm": 0.7746312952822261, "learning_rate": 0.00029402569415722666, "loss": 0.4382, "step": 2969 }, { "epoch": 1.3976470588235295, "grad_norm": 0.28182899313736076, "learning_rate": 0.00029401343965183965, "loss": 0.2701, "step": 2970 }, { "epoch": 1.3981176470588235, "grad_norm": 0.4012928889267386, "learning_rate": 0.0002940011728469986, "loss": 0.2958, "step": 2971 }, { "epoch": 1.3985882352941177, "grad_norm": 0.624206427371469, "learning_rate": 0.000293988893743751, "loss": 0.4598, "step": 2972 }, { "epoch": 1.3990588235294117, "grad_norm": 0.4619555884414463, "learning_rate": 0.0002939766023431457, "loss": 0.4249, "step": 2973 }, { "epoch": 1.399529411764706, "grad_norm": 0.27796186783391275, "learning_rate": 0.0002939642986462323, "loss": 0.2575, "step": 2974 }, { "epoch": 1.4, "grad_norm": 0.30675297645166266, "learning_rate": 0.0002939519826540616, "loss": 0.371, "step": 2975 }, { "epoch": 1.4004705882352941, "grad_norm": 0.2824041181850221, "learning_rate": 0.0002939396543676856, "loss": 0.231, "step": 2976 }, { "epoch": 1.4009411764705884, "grad_norm": 0.4036523615001926, "learning_rate": 0.00029392731378815706, "loss": 0.465, "step": 2977 }, { "epoch": 1.4014117647058824, "grad_norm": 0.20776421766535433, "learning_rate": 0.00029391496091652993, "loss": 0.1109, "step": 2978 }, { "epoch": 1.4018823529411764, "grad_norm": 0.15665177917766665, "learning_rate": 0.0002939025957538593, "loss": 0.1281, "step": 2979 }, { "epoch": 1.4023529411764706, "grad_norm": 0.2820184729468403, "learning_rate": 0.0002938902183012011, "loss": 0.2678, "step": 2980 }, { "epoch": 1.4028235294117648, "grad_norm": 0.16969005993828626, "learning_rate": 0.00029387782855961253, "loss": 0.1775, "step": 2981 }, { "epoch": 1.4032941176470588, "grad_norm": 0.3280036425154497, "learning_rate": 0.00029386542653015167, "loss": 0.5118, "step": 2982 }, { "epoch": 1.4037647058823528, "grad_norm": 0.32145444650839666, "learning_rate": 0.0002938530122138777, "loss": 0.2896, "step": 2983 }, { "epoch": 1.404235294117647, "grad_norm": 0.36433637862229135, "learning_rate": 0.0002938405856118509, "loss": 0.37, "step": 2984 }, { "epoch": 1.4047058823529412, "grad_norm": 0.2760645213615751, "learning_rate": 0.0002938281467251326, "loss": 0.2696, "step": 2985 }, { "epoch": 1.4051764705882352, "grad_norm": 0.19636971046912646, "learning_rate": 0.00029381569555478505, "loss": 0.1464, "step": 2986 }, { "epoch": 1.4056470588235295, "grad_norm": 0.25990211155090776, "learning_rate": 0.00029380323210187166, "loss": 0.2786, "step": 2987 }, { "epoch": 1.4061176470588235, "grad_norm": 0.04891006138256543, "learning_rate": 0.00029379075636745694, "loss": 0.0174, "step": 2988 }, { "epoch": 1.4065882352941177, "grad_norm": 0.37643009717433396, "learning_rate": 0.0002937782683526063, "loss": 0.3638, "step": 2989 }, { "epoch": 1.4070588235294117, "grad_norm": 0.25773391559853476, "learning_rate": 0.0002937657680583864, "loss": 0.1555, "step": 2990 }, { "epoch": 1.407529411764706, "grad_norm": 0.3542878974264533, "learning_rate": 0.0002937532554858647, "loss": 0.2753, "step": 2991 }, { "epoch": 1.408, "grad_norm": 0.43136801112315115, "learning_rate": 0.0002937407306361099, "loss": 0.3768, "step": 2992 }, { "epoch": 1.4084705882352941, "grad_norm": 0.4554842575985483, "learning_rate": 0.0002937281935101916, "loss": 0.4772, "step": 2993 }, { "epoch": 1.4089411764705884, "grad_norm": 0.4407740396726986, "learning_rate": 0.0002937156441091806, "loss": 0.4525, "step": 2994 }, { "epoch": 1.4094117647058824, "grad_norm": 0.2843659588477465, "learning_rate": 0.0002937030824341487, "loss": 0.2273, "step": 2995 }, { "epoch": 1.4098823529411764, "grad_norm": 0.3518882481227539, "learning_rate": 0.00029369050848616863, "loss": 0.2587, "step": 2996 }, { "epoch": 1.4103529411764706, "grad_norm": 0.3683273192611348, "learning_rate": 0.0002936779222663144, "loss": 0.3505, "step": 2997 }, { "epoch": 1.4108235294117648, "grad_norm": 0.12956364126294428, "learning_rate": 0.0002936653237756609, "loss": 0.0721, "step": 2998 }, { "epoch": 1.4112941176470588, "grad_norm": 0.4964280183857917, "learning_rate": 0.00029365271301528406, "loss": 0.5117, "step": 2999 }, { "epoch": 1.4117647058823528, "grad_norm": 0.36462643660601585, "learning_rate": 0.00029364008998626086, "loss": 0.4243, "step": 3000 }, { "epoch": 1.412235294117647, "grad_norm": 0.45130222353826976, "learning_rate": 0.00029362745468966946, "loss": 0.3772, "step": 3001 }, { "epoch": 1.4127058823529413, "grad_norm": 0.04997048150375971, "learning_rate": 0.0002936148071265889, "loss": 0.0258, "step": 3002 }, { "epoch": 1.4131764705882353, "grad_norm": 0.4433833255927716, "learning_rate": 0.00029360214729809946, "loss": 0.4007, "step": 3003 }, { "epoch": 1.4136470588235295, "grad_norm": 0.47868905265201356, "learning_rate": 0.00029358947520528227, "loss": 0.4268, "step": 3004 }, { "epoch": 1.4141176470588235, "grad_norm": 0.6309168766904224, "learning_rate": 0.0002935767908492196, "loss": 0.4994, "step": 3005 }, { "epoch": 1.4145882352941177, "grad_norm": 0.4881938127369698, "learning_rate": 0.0002935640942309947, "loss": 0.5191, "step": 3006 }, { "epoch": 1.4150588235294117, "grad_norm": 0.23412890509381812, "learning_rate": 0.000293551385351692, "loss": 0.1433, "step": 3007 }, { "epoch": 1.415529411764706, "grad_norm": 0.2571352273344013, "learning_rate": 0.0002935386642123968, "loss": 0.25, "step": 3008 }, { "epoch": 1.416, "grad_norm": 0.21917259842427692, "learning_rate": 0.00029352593081419573, "loss": 0.1574, "step": 3009 }, { "epoch": 1.4164705882352941, "grad_norm": 0.5199766622706221, "learning_rate": 0.0002935131851581762, "loss": 0.5851, "step": 3010 }, { "epoch": 1.4169411764705884, "grad_norm": 0.6786803896781136, "learning_rate": 0.00029350042724542657, "loss": 0.6334, "step": 3011 }, { "epoch": 1.4174117647058824, "grad_norm": 0.4004179520684868, "learning_rate": 0.0002934876570770367, "loss": 0.5732, "step": 3012 }, { "epoch": 1.4178823529411764, "grad_norm": 0.20504337098792483, "learning_rate": 0.00029347487465409714, "loss": 0.0403, "step": 3013 }, { "epoch": 1.4183529411764706, "grad_norm": 0.16357124660912423, "learning_rate": 0.0002934620799776995, "loss": 0.0667, "step": 3014 }, { "epoch": 1.4188235294117648, "grad_norm": 0.679989376628898, "learning_rate": 0.00029344927304893653, "loss": 0.6855, "step": 3015 }, { "epoch": 1.4192941176470588, "grad_norm": 0.513940885249212, "learning_rate": 0.00029343645386890204, "loss": 0.4218, "step": 3016 }, { "epoch": 1.4197647058823528, "grad_norm": 0.451094070776872, "learning_rate": 0.00029342362243869093, "loss": 0.4059, "step": 3017 }, { "epoch": 1.420235294117647, "grad_norm": 0.35357769695177804, "learning_rate": 0.00029341077875939884, "loss": 0.2806, "step": 3018 }, { "epoch": 1.4207058823529413, "grad_norm": 0.2611535853754207, "learning_rate": 0.0002933979228321229, "loss": 0.2626, "step": 3019 }, { "epoch": 1.4211764705882353, "grad_norm": 0.1500721644480386, "learning_rate": 0.00029338505465796095, "loss": 0.1114, "step": 3020 }, { "epoch": 1.4216470588235295, "grad_norm": 0.47829263638940434, "learning_rate": 0.00029337217423801206, "loss": 0.3617, "step": 3021 }, { "epoch": 1.4221176470588235, "grad_norm": 0.26166285044910326, "learning_rate": 0.00029335928157337625, "loss": 0.2262, "step": 3022 }, { "epoch": 1.4225882352941177, "grad_norm": 0.4860628415760249, "learning_rate": 0.00029334637666515457, "loss": 0.4867, "step": 3023 }, { "epoch": 1.4230588235294117, "grad_norm": 0.42343523281318446, "learning_rate": 0.00029333345951444923, "loss": 0.2758, "step": 3024 }, { "epoch": 1.423529411764706, "grad_norm": 0.19709014736448613, "learning_rate": 0.00029332053012236345, "loss": 0.1739, "step": 3025 }, { "epoch": 1.424, "grad_norm": 0.25644466007390876, "learning_rate": 0.0002933075884900014, "loss": 0.2121, "step": 3026 }, { "epoch": 1.4244705882352942, "grad_norm": 0.2019255697594077, "learning_rate": 0.00029329463461846833, "loss": 0.1596, "step": 3027 }, { "epoch": 1.4249411764705882, "grad_norm": 0.3436267579796572, "learning_rate": 0.00029328166850887066, "loss": 0.4417, "step": 3028 }, { "epoch": 1.4254117647058824, "grad_norm": 0.21229795980201424, "learning_rate": 0.00029326869016231566, "loss": 0.2494, "step": 3029 }, { "epoch": 1.4258823529411764, "grad_norm": 0.4979816548723784, "learning_rate": 0.0002932556995799118, "loss": 0.6209, "step": 3030 }, { "epoch": 1.4263529411764706, "grad_norm": 0.36240150360522383, "learning_rate": 0.00029324269676276856, "loss": 0.4399, "step": 3031 }, { "epoch": 1.4268235294117648, "grad_norm": 0.3627050303954664, "learning_rate": 0.0002932296817119964, "loss": 0.3713, "step": 3032 }, { "epoch": 1.4272941176470588, "grad_norm": 0.3786134228397658, "learning_rate": 0.0002932166544287069, "loss": 0.5548, "step": 3033 }, { "epoch": 1.4277647058823528, "grad_norm": 0.3110874997381482, "learning_rate": 0.00029320361491401263, "loss": 0.3318, "step": 3034 }, { "epoch": 1.428235294117647, "grad_norm": 0.1584504191015231, "learning_rate": 0.00029319056316902723, "loss": 0.1341, "step": 3035 }, { "epoch": 1.4287058823529413, "grad_norm": 0.09803608338288501, "learning_rate": 0.0002931774991948654, "loss": 0.0385, "step": 3036 }, { "epoch": 1.4291764705882353, "grad_norm": 0.12489595466054072, "learning_rate": 0.0002931644229926428, "loss": 0.0837, "step": 3037 }, { "epoch": 1.4296470588235295, "grad_norm": 0.3885674917324356, "learning_rate": 0.00029315133456347633, "loss": 0.4494, "step": 3038 }, { "epoch": 1.4301176470588235, "grad_norm": 0.7290217649981016, "learning_rate": 0.0002931382339084837, "loss": 0.6682, "step": 3039 }, { "epoch": 1.4305882352941177, "grad_norm": 0.2705742351063871, "learning_rate": 0.0002931251210287838, "loss": 0.2777, "step": 3040 }, { "epoch": 1.4310588235294117, "grad_norm": 0.18491201239231497, "learning_rate": 0.00029311199592549656, "loss": 0.2313, "step": 3041 }, { "epoch": 1.431529411764706, "grad_norm": 0.23658642397957214, "learning_rate": 0.00029309885859974294, "loss": 0.1267, "step": 3042 }, { "epoch": 1.432, "grad_norm": 1.072737967588725, "learning_rate": 0.00029308570905264484, "loss": 0.7529, "step": 3043 }, { "epoch": 1.4324705882352942, "grad_norm": 0.4748725118343599, "learning_rate": 0.00029307254728532536, "loss": 0.7162, "step": 3044 }, { "epoch": 1.4329411764705882, "grad_norm": 0.14961883983556198, "learning_rate": 0.00029305937329890863, "loss": 0.0894, "step": 3045 }, { "epoch": 1.4334117647058824, "grad_norm": 0.6182088692243474, "learning_rate": 0.00029304618709451966, "loss": 0.8035, "step": 3046 }, { "epoch": 1.4338823529411764, "grad_norm": 0.27400809276754745, "learning_rate": 0.0002930329886732847, "loss": 0.2046, "step": 3047 }, { "epoch": 1.4343529411764706, "grad_norm": 0.2834821026773767, "learning_rate": 0.00029301977803633094, "loss": 0.1691, "step": 3048 }, { "epoch": 1.4348235294117648, "grad_norm": 0.6839107571437663, "learning_rate": 0.0002930065551847866, "loss": 0.4097, "step": 3049 }, { "epoch": 1.4352941176470588, "grad_norm": 0.3874319972133545, "learning_rate": 0.00029299332011978107, "loss": 0.1923, "step": 3050 }, { "epoch": 1.4357647058823528, "grad_norm": 0.15302662492165917, "learning_rate": 0.00029298007284244456, "loss": 0.1568, "step": 3051 }, { "epoch": 1.436235294117647, "grad_norm": 0.4064646695183042, "learning_rate": 0.0002929668133539085, "loss": 0.3956, "step": 3052 }, { "epoch": 1.4367058823529413, "grad_norm": 0.21928234728640894, "learning_rate": 0.0002929535416553054, "loss": 0.1746, "step": 3053 }, { "epoch": 1.4371764705882353, "grad_norm": 0.32317507580338156, "learning_rate": 0.0002929402577477686, "loss": 0.3368, "step": 3054 }, { "epoch": 1.4376470588235293, "grad_norm": 0.4121073178550215, "learning_rate": 0.0002929269616324327, "loss": 0.3389, "step": 3055 }, { "epoch": 1.4381176470588235, "grad_norm": 0.3221751925353723, "learning_rate": 0.0002929136533104332, "loss": 0.2278, "step": 3056 }, { "epoch": 1.4385882352941177, "grad_norm": 0.2711971254621109, "learning_rate": 0.0002929003327829067, "loss": 0.2735, "step": 3057 }, { "epoch": 1.4390588235294117, "grad_norm": 0.14167549323672401, "learning_rate": 0.0002928870000509909, "loss": 0.123, "step": 3058 }, { "epoch": 1.439529411764706, "grad_norm": 0.5215550558931575, "learning_rate": 0.0002928736551158244, "loss": 0.4257, "step": 3059 }, { "epoch": 1.44, "grad_norm": 0.13440842705886516, "learning_rate": 0.00029286029797854704, "loss": 0.0841, "step": 3060 }, { "epoch": 1.4404705882352942, "grad_norm": 0.47035430830790464, "learning_rate": 0.00029284692864029946, "loss": 0.712, "step": 3061 }, { "epoch": 1.4409411764705882, "grad_norm": 0.0773551386595098, "learning_rate": 0.0002928335471022235, "loss": 0.0324, "step": 3062 }, { "epoch": 1.4414117647058824, "grad_norm": 0.6397807510888169, "learning_rate": 0.00029282015336546205, "loss": 0.631, "step": 3063 }, { "epoch": 1.4418823529411764, "grad_norm": 0.49340882813149833, "learning_rate": 0.00029280674743115896, "loss": 0.6319, "step": 3064 }, { "epoch": 1.4423529411764706, "grad_norm": 0.43870031922889596, "learning_rate": 0.0002927933293004592, "loss": 0.4992, "step": 3065 }, { "epoch": 1.4428235294117648, "grad_norm": 0.5376007819364209, "learning_rate": 0.00029277989897450867, "loss": 0.6261, "step": 3066 }, { "epoch": 1.4432941176470588, "grad_norm": 0.4211410639648726, "learning_rate": 0.0002927664564544545, "loss": 0.423, "step": 3067 }, { "epoch": 1.4437647058823528, "grad_norm": 0.3939738916280366, "learning_rate": 0.00029275300174144464, "loss": 0.4002, "step": 3068 }, { "epoch": 1.444235294117647, "grad_norm": 0.44913268661580774, "learning_rate": 0.0002927395348366282, "loss": 0.3808, "step": 3069 }, { "epoch": 1.4447058823529413, "grad_norm": 0.3159775069657792, "learning_rate": 0.00029272605574115544, "loss": 0.3463, "step": 3070 }, { "epoch": 1.4451764705882353, "grad_norm": 0.20266085769347691, "learning_rate": 0.00029271256445617737, "loss": 0.1756, "step": 3071 }, { "epoch": 1.4456470588235293, "grad_norm": 0.17463896959342162, "learning_rate": 0.0002926990609828463, "loss": 0.1249, "step": 3072 }, { "epoch": 1.4461176470588235, "grad_norm": 0.27297583623965443, "learning_rate": 0.00029268554532231556, "loss": 0.2446, "step": 3073 }, { "epoch": 1.4465882352941177, "grad_norm": 0.18224759867757942, "learning_rate": 0.00029267201747573934, "loss": 0.1286, "step": 3074 }, { "epoch": 1.4470588235294117, "grad_norm": 0.1389956644803606, "learning_rate": 0.00029265847744427303, "loss": 0.1199, "step": 3075 }, { "epoch": 1.447529411764706, "grad_norm": 0.5551055302855673, "learning_rate": 0.000292644925229073, "loss": 0.3687, "step": 3076 }, { "epoch": 1.448, "grad_norm": 0.3993504912671494, "learning_rate": 0.00029263136083129666, "loss": 0.3151, "step": 3077 }, { "epoch": 1.4484705882352942, "grad_norm": 0.36508791409593944, "learning_rate": 0.00029261778425210256, "loss": 0.3907, "step": 3078 }, { "epoch": 1.4489411764705882, "grad_norm": 0.43836925825795714, "learning_rate": 0.0002926041954926501, "loss": 0.4472, "step": 3079 }, { "epoch": 1.4494117647058824, "grad_norm": 0.30242666401380663, "learning_rate": 0.00029259059455409987, "loss": 0.3819, "step": 3080 }, { "epoch": 1.4498823529411764, "grad_norm": 0.322749519587469, "learning_rate": 0.00029257698143761354, "loss": 0.3113, "step": 3081 }, { "epoch": 1.4503529411764706, "grad_norm": 0.1945543679338094, "learning_rate": 0.00029256335614435356, "loss": 0.1959, "step": 3082 }, { "epoch": 1.4508235294117646, "grad_norm": 0.45302795079617525, "learning_rate": 0.00029254971867548373, "loss": 0.3696, "step": 3083 }, { "epoch": 1.4512941176470588, "grad_norm": 0.1818564829768233, "learning_rate": 0.00029253606903216874, "loss": 0.1485, "step": 3084 }, { "epoch": 1.4517647058823528, "grad_norm": 0.42251583242126756, "learning_rate": 0.00029252240721557433, "loss": 0.4816, "step": 3085 }, { "epoch": 1.452235294117647, "grad_norm": 0.15329124036620032, "learning_rate": 0.00029250873322686727, "loss": 0.1252, "step": 3086 }, { "epoch": 1.4527058823529413, "grad_norm": 0.08848436184834417, "learning_rate": 0.00029249504706721534, "loss": 0.0525, "step": 3087 }, { "epoch": 1.4531764705882353, "grad_norm": 0.3295731583088289, "learning_rate": 0.00029248134873778754, "loss": 0.2705, "step": 3088 }, { "epoch": 1.4536470588235293, "grad_norm": 0.15945546987170167, "learning_rate": 0.0002924676382397536, "loss": 0.099, "step": 3089 }, { "epoch": 1.4541176470588235, "grad_norm": 0.5228959586883204, "learning_rate": 0.0002924539155742846, "loss": 0.4175, "step": 3090 }, { "epoch": 1.4545882352941177, "grad_norm": 0.30871927325350723, "learning_rate": 0.0002924401807425525, "loss": 0.2503, "step": 3091 }, { "epoch": 1.4550588235294117, "grad_norm": 0.18716990588699575, "learning_rate": 0.0002924264337457303, "loss": 0.1027, "step": 3092 }, { "epoch": 1.455529411764706, "grad_norm": 0.32932875627996255, "learning_rate": 0.000292412674584992, "loss": 0.3893, "step": 3093 }, { "epoch": 1.456, "grad_norm": 0.46129257312886146, "learning_rate": 0.0002923989032615128, "loss": 0.4433, "step": 3094 }, { "epoch": 1.4564705882352942, "grad_norm": 0.26622058152447214, "learning_rate": 0.0002923851197764688, "loss": 0.2076, "step": 3095 }, { "epoch": 1.4569411764705882, "grad_norm": 0.47175343998712127, "learning_rate": 0.00029237132413103715, "loss": 0.5785, "step": 3096 }, { "epoch": 1.4574117647058824, "grad_norm": 0.27196120567233417, "learning_rate": 0.00029235751632639617, "loss": 0.3182, "step": 3097 }, { "epoch": 1.4578823529411764, "grad_norm": 0.41877574642852594, "learning_rate": 0.00029234369636372496, "loss": 0.4644, "step": 3098 }, { "epoch": 1.4583529411764706, "grad_norm": 0.4189330964160751, "learning_rate": 0.0002923298642442039, "loss": 0.3332, "step": 3099 }, { "epoch": 1.4588235294117646, "grad_norm": 0.370817721131375, "learning_rate": 0.00029231601996901433, "loss": 0.3453, "step": 3100 }, { "epoch": 1.4592941176470589, "grad_norm": 0.21240248107527662, "learning_rate": 0.0002923021635393386, "loss": 0.2594, "step": 3101 }, { "epoch": 1.4597647058823529, "grad_norm": 0.2314177681951581, "learning_rate": 0.0002922882949563601, "loss": 0.1911, "step": 3102 }, { "epoch": 1.460235294117647, "grad_norm": 0.13600874507725738, "learning_rate": 0.0002922744142212633, "loss": 0.0629, "step": 3103 }, { "epoch": 1.4607058823529413, "grad_norm": 0.3733443457416637, "learning_rate": 0.00029226052133523363, "loss": 0.322, "step": 3104 }, { "epoch": 1.4611764705882353, "grad_norm": 0.4564820742685653, "learning_rate": 0.0002922466162994577, "loss": 0.3705, "step": 3105 }, { "epoch": 1.4616470588235293, "grad_norm": 0.44254294988925197, "learning_rate": 0.000292232699115123, "loss": 0.3564, "step": 3106 }, { "epoch": 1.4621176470588235, "grad_norm": 0.1951442422820125, "learning_rate": 0.0002922187697834182, "loss": 0.1418, "step": 3107 }, { "epoch": 1.4625882352941177, "grad_norm": 0.36602734402761233, "learning_rate": 0.00029220482830553284, "loss": 0.2689, "step": 3108 }, { "epoch": 1.4630588235294117, "grad_norm": 0.17194018493868818, "learning_rate": 0.0002921908746826576, "loss": 0.1595, "step": 3109 }, { "epoch": 1.4635294117647057, "grad_norm": 0.1842009714439063, "learning_rate": 0.0002921769089159843, "loss": 0.1878, "step": 3110 }, { "epoch": 1.464, "grad_norm": 0.6095748267539566, "learning_rate": 0.00029216293100670554, "loss": 0.4874, "step": 3111 }, { "epoch": 1.4644705882352942, "grad_norm": 0.2667858617203875, "learning_rate": 0.0002921489409560152, "loss": 0.2521, "step": 3112 }, { "epoch": 1.4649411764705882, "grad_norm": 0.1952381009154941, "learning_rate": 0.00029213493876510807, "loss": 0.2105, "step": 3113 }, { "epoch": 1.4654117647058824, "grad_norm": 0.17705498825504168, "learning_rate": 0.0002921209244351799, "loss": 0.1607, "step": 3114 }, { "epoch": 1.4658823529411764, "grad_norm": 0.2603774657775541, "learning_rate": 0.00029210689796742783, "loss": 0.2314, "step": 3115 }, { "epoch": 1.4663529411764706, "grad_norm": 0.2974661905144616, "learning_rate": 0.00029209285936304956, "loss": 0.364, "step": 3116 }, { "epoch": 1.4668235294117646, "grad_norm": 0.3209754001175584, "learning_rate": 0.0002920788086232442, "loss": 0.3117, "step": 3117 }, { "epoch": 1.4672941176470589, "grad_norm": 0.41629634052010206, "learning_rate": 0.00029206474574921165, "loss": 0.4713, "step": 3118 }, { "epoch": 1.4677647058823529, "grad_norm": 0.12402398079134351, "learning_rate": 0.0002920506707421529, "loss": 0.0695, "step": 3119 }, { "epoch": 1.468235294117647, "grad_norm": 0.27149389964407755, "learning_rate": 0.0002920365836032703, "loss": 0.2066, "step": 3120 }, { "epoch": 1.4687058823529413, "grad_norm": 0.2360518204715184, "learning_rate": 0.00029202248433376663, "loss": 0.2063, "step": 3121 }, { "epoch": 1.4691764705882353, "grad_norm": 0.3495434141319128, "learning_rate": 0.00029200837293484627, "loss": 0.2184, "step": 3122 }, { "epoch": 1.4696470588235293, "grad_norm": 0.5292349056140615, "learning_rate": 0.0002919942494077142, "loss": 0.522, "step": 3123 }, { "epoch": 1.4701176470588235, "grad_norm": 0.33236327391239934, "learning_rate": 0.0002919801137535769, "loss": 0.2013, "step": 3124 }, { "epoch": 1.4705882352941178, "grad_norm": 0.154696120094297, "learning_rate": 0.0002919659659736414, "loss": 0.1154, "step": 3125 }, { "epoch": 1.4710588235294118, "grad_norm": 0.20116792255417748, "learning_rate": 0.00029195180606911606, "loss": 0.1534, "step": 3126 }, { "epoch": 1.4715294117647058, "grad_norm": 0.2967737004947424, "learning_rate": 0.0002919376340412102, "loss": 0.3973, "step": 3127 }, { "epoch": 1.472, "grad_norm": 0.4076217234584286, "learning_rate": 0.00029192344989113425, "loss": 0.4084, "step": 3128 }, { "epoch": 1.4724705882352942, "grad_norm": 0.40419917383923876, "learning_rate": 0.0002919092536200995, "loss": 0.3084, "step": 3129 }, { "epoch": 1.4729411764705882, "grad_norm": 0.230611317103838, "learning_rate": 0.00029189504522931847, "loss": 0.3022, "step": 3130 }, { "epoch": 1.4734117647058824, "grad_norm": 0.4094481196688222, "learning_rate": 0.00029188082472000453, "loss": 0.5751, "step": 3131 }, { "epoch": 1.4738823529411764, "grad_norm": 0.37729467823093055, "learning_rate": 0.0002918665920933722, "loss": 0.3045, "step": 3132 }, { "epoch": 1.4743529411764706, "grad_norm": 0.23540600580342472, "learning_rate": 0.0002918523473506372, "loss": 0.1886, "step": 3133 }, { "epoch": 1.4748235294117646, "grad_norm": 0.38415657311734147, "learning_rate": 0.00029183809049301585, "loss": 0.4163, "step": 3134 }, { "epoch": 1.4752941176470589, "grad_norm": 0.3541735839328632, "learning_rate": 0.0002918238215217259, "loss": 0.4874, "step": 3135 }, { "epoch": 1.4757647058823529, "grad_norm": 0.33249607867829534, "learning_rate": 0.00029180954043798595, "loss": 0.3026, "step": 3136 }, { "epoch": 1.476235294117647, "grad_norm": 0.21965722204047622, "learning_rate": 0.00029179524724301563, "loss": 0.2002, "step": 3137 }, { "epoch": 1.4767058823529413, "grad_norm": 0.17463685616857705, "learning_rate": 0.00029178094193803576, "loss": 0.1034, "step": 3138 }, { "epoch": 1.4771764705882353, "grad_norm": 0.11910489150625028, "learning_rate": 0.00029176662452426796, "loss": 0.0841, "step": 3139 }, { "epoch": 1.4776470588235293, "grad_norm": 0.2841968489796847, "learning_rate": 0.0002917522950029351, "loss": 0.2638, "step": 3140 }, { "epoch": 1.4781176470588235, "grad_norm": 0.18501084926038855, "learning_rate": 0.00029173795337526093, "loss": 0.1123, "step": 3141 }, { "epoch": 1.4785882352941178, "grad_norm": 0.47991821894154013, "learning_rate": 0.0002917235996424703, "loss": 0.4244, "step": 3142 }, { "epoch": 1.4790588235294118, "grad_norm": 0.368526278667554, "learning_rate": 0.0002917092338057892, "loss": 0.3492, "step": 3143 }, { "epoch": 1.4795294117647058, "grad_norm": 0.2507250128785583, "learning_rate": 0.00029169485586644434, "loss": 0.2055, "step": 3144 }, { "epoch": 1.48, "grad_norm": 0.4056276527306703, "learning_rate": 0.00029168046582566385, "loss": 0.4391, "step": 3145 }, { "epoch": 1.4804705882352942, "grad_norm": 0.4999928061034032, "learning_rate": 0.00029166606368467667, "loss": 0.6055, "step": 3146 }, { "epoch": 1.4809411764705882, "grad_norm": 0.24126499506122082, "learning_rate": 0.0002916516494447127, "loss": 0.2859, "step": 3147 }, { "epoch": 1.4814117647058824, "grad_norm": 0.633722543311993, "learning_rate": 0.0002916372231070031, "loss": 0.7123, "step": 3148 }, { "epoch": 1.4818823529411764, "grad_norm": 0.4019675356577139, "learning_rate": 0.00029162278467277994, "loss": 0.4263, "step": 3149 }, { "epoch": 1.4823529411764707, "grad_norm": 0.3637976645635073, "learning_rate": 0.0002916083341432763, "loss": 0.4723, "step": 3150 }, { "epoch": 1.4828235294117647, "grad_norm": 0.18832301393645096, "learning_rate": 0.0002915938715197264, "loss": 0.19, "step": 3151 }, { "epoch": 1.4832941176470589, "grad_norm": 0.20815275315552348, "learning_rate": 0.00029157939680336527, "loss": 0.2111, "step": 3152 }, { "epoch": 1.4837647058823529, "grad_norm": 0.2759174950885207, "learning_rate": 0.00029156490999542923, "loss": 0.3164, "step": 3153 }, { "epoch": 1.484235294117647, "grad_norm": 0.3317845601687911, "learning_rate": 0.0002915504110971555, "loss": 0.3005, "step": 3154 }, { "epoch": 1.484705882352941, "grad_norm": 0.29788520495675996, "learning_rate": 0.0002915359001097824, "loss": 0.3863, "step": 3155 }, { "epoch": 1.4851764705882353, "grad_norm": 0.4386484653622655, "learning_rate": 0.0002915213770345492, "loss": 0.2547, "step": 3156 }, { "epoch": 1.4856470588235293, "grad_norm": 0.32039977993805413, "learning_rate": 0.0002915068418726962, "loss": 0.2778, "step": 3157 }, { "epoch": 1.4861176470588235, "grad_norm": 0.20586416887493686, "learning_rate": 0.00029149229462546483, "loss": 0.0819, "step": 3158 }, { "epoch": 1.4865882352941178, "grad_norm": 0.39024198321392417, "learning_rate": 0.0002914777352940975, "loss": 0.4148, "step": 3159 }, { "epoch": 1.4870588235294118, "grad_norm": 0.17651465267997984, "learning_rate": 0.0002914631638798376, "loss": 0.216, "step": 3160 }, { "epoch": 1.4875294117647058, "grad_norm": 0.19733294116414402, "learning_rate": 0.00029144858038392964, "loss": 0.0981, "step": 3161 }, { "epoch": 1.488, "grad_norm": 0.40440559230983697, "learning_rate": 0.00029143398480761915, "loss": 0.2094, "step": 3162 }, { "epoch": 1.4884705882352942, "grad_norm": 0.07641372052042937, "learning_rate": 0.00029141937715215256, "loss": 0.042, "step": 3163 }, { "epoch": 1.4889411764705882, "grad_norm": 0.1321938595231379, "learning_rate": 0.0002914047574187776, "loss": 0.1066, "step": 3164 }, { "epoch": 1.4894117647058824, "grad_norm": 0.04694320987033675, "learning_rate": 0.0002913901256087427, "loss": 0.0176, "step": 3165 }, { "epoch": 1.4898823529411764, "grad_norm": 0.16205766382900375, "learning_rate": 0.00029137548172329756, "loss": 0.1032, "step": 3166 }, { "epoch": 1.4903529411764707, "grad_norm": 0.2089391267642907, "learning_rate": 0.0002913608257636929, "loss": 0.1546, "step": 3167 }, { "epoch": 1.4908235294117647, "grad_norm": 0.02670684223587166, "learning_rate": 0.0002913461577311803, "loss": 0.0047, "step": 3168 }, { "epoch": 1.4912941176470589, "grad_norm": 0.38498926558731894, "learning_rate": 0.00029133147762701246, "loss": 0.3215, "step": 3169 }, { "epoch": 1.4917647058823529, "grad_norm": 0.6240120051045547, "learning_rate": 0.00029131678545244333, "loss": 0.5382, "step": 3170 }, { "epoch": 1.492235294117647, "grad_norm": 0.31435848378998305, "learning_rate": 0.00029130208120872747, "loss": 0.4005, "step": 3171 }, { "epoch": 1.492705882352941, "grad_norm": 0.24317089551900456, "learning_rate": 0.00029128736489712087, "loss": 0.3937, "step": 3172 }, { "epoch": 1.4931764705882353, "grad_norm": 0.2624976459797657, "learning_rate": 0.0002912726365188803, "loss": 0.3366, "step": 3173 }, { "epoch": 1.4936470588235293, "grad_norm": 0.3745397979788921, "learning_rate": 0.0002912578960752636, "loss": 0.3944, "step": 3174 }, { "epoch": 1.4941176470588236, "grad_norm": 0.5620488539117854, "learning_rate": 0.00029124314356752967, "loss": 0.4805, "step": 3175 }, { "epoch": 1.4945882352941178, "grad_norm": 0.21262989283844363, "learning_rate": 0.00029122837899693847, "loss": 0.0944, "step": 3176 }, { "epoch": 1.4950588235294118, "grad_norm": 0.2456643989009959, "learning_rate": 0.0002912136023647511, "loss": 0.156, "step": 3177 }, { "epoch": 1.4955294117647058, "grad_norm": 0.33835271143238804, "learning_rate": 0.0002911988136722293, "loss": 0.1815, "step": 3178 }, { "epoch": 1.496, "grad_norm": 0.2692729443101487, "learning_rate": 0.00029118401292063634, "loss": 0.2095, "step": 3179 }, { "epoch": 1.4964705882352942, "grad_norm": 0.2644848672392768, "learning_rate": 0.00029116920011123615, "loss": 0.2143, "step": 3180 }, { "epoch": 1.4969411764705882, "grad_norm": 0.2638337026714295, "learning_rate": 0.00029115437524529383, "loss": 0.2483, "step": 3181 }, { "epoch": 1.4974117647058822, "grad_norm": 0.3152760717369825, "learning_rate": 0.0002911395383240755, "loss": 0.2845, "step": 3182 }, { "epoch": 1.4978823529411764, "grad_norm": 0.20214940515579352, "learning_rate": 0.0002911246893488483, "loss": 0.1585, "step": 3183 }, { "epoch": 1.4983529411764707, "grad_norm": 0.6302612142139129, "learning_rate": 0.0002911098283208805, "loss": 0.5984, "step": 3184 }, { "epoch": 1.4988235294117647, "grad_norm": 0.2736790599345568, "learning_rate": 0.0002910949552414411, "loss": 0.1982, "step": 3185 }, { "epoch": 1.4992941176470589, "grad_norm": 0.19241009597279116, "learning_rate": 0.0002910800701118005, "loss": 0.1755, "step": 3186 }, { "epoch": 1.4997647058823529, "grad_norm": 0.5985561169605296, "learning_rate": 0.00029106517293323, "loss": 0.6082, "step": 3187 }, { "epoch": 1.5002352941176471, "grad_norm": 0.11923208199274092, "learning_rate": 0.0002910502637070017, "loss": 0.0606, "step": 3188 }, { "epoch": 1.5007058823529413, "grad_norm": 0.1847968777217, "learning_rate": 0.0002910353424343891, "loss": 0.0851, "step": 3189 }, { "epoch": 1.5011764705882353, "grad_norm": 0.1468519715789299, "learning_rate": 0.00029102040911666645, "loss": 0.0884, "step": 3190 }, { "epoch": 1.5016470588235293, "grad_norm": 0.5475809534522054, "learning_rate": 0.0002910054637551092, "loss": 0.3432, "step": 3191 }, { "epoch": 1.5021176470588236, "grad_norm": 0.24496222783327626, "learning_rate": 0.0002909905063509936, "loss": 0.2127, "step": 3192 }, { "epoch": 1.5025882352941178, "grad_norm": 0.6556523952489639, "learning_rate": 0.0002909755369055973, "loss": 0.6128, "step": 3193 }, { "epoch": 1.5030588235294118, "grad_norm": 0.3030809070975141, "learning_rate": 0.00029096055542019865, "loss": 0.4086, "step": 3194 }, { "epoch": 1.5035294117647058, "grad_norm": 0.3743896021758398, "learning_rate": 0.0002909455618960772, "loss": 0.3839, "step": 3195 }, { "epoch": 1.504, "grad_norm": 0.2825602127354741, "learning_rate": 0.0002909305563345134, "loss": 0.1072, "step": 3196 }, { "epoch": 1.5044705882352942, "grad_norm": 0.4286440928663319, "learning_rate": 0.0002909155387367888, "loss": 0.377, "step": 3197 }, { "epoch": 1.5049411764705882, "grad_norm": 0.6018707808950449, "learning_rate": 0.0002909005091041861, "loss": 0.3499, "step": 3198 }, { "epoch": 1.5054117647058822, "grad_norm": 0.3765997375271805, "learning_rate": 0.00029088546743798866, "loss": 0.4516, "step": 3199 }, { "epoch": 1.5058823529411764, "grad_norm": 0.3887793769789747, "learning_rate": 0.00029087041373948135, "loss": 0.4371, "step": 3200 }, { "epoch": 1.5063529411764707, "grad_norm": 0.2473957275856775, "learning_rate": 0.00029085534800994976, "loss": 0.0892, "step": 3201 }, { "epoch": 1.5068235294117647, "grad_norm": 0.22777424484710962, "learning_rate": 0.0002908402702506805, "loss": 0.1555, "step": 3202 }, { "epoch": 1.5072941176470587, "grad_norm": 0.29276109641707143, "learning_rate": 0.0002908251804629614, "loss": 0.4109, "step": 3203 }, { "epoch": 1.507764705882353, "grad_norm": 0.9307030166289362, "learning_rate": 0.0002908100786480811, "loss": 0.6997, "step": 3204 }, { "epoch": 1.5082352941176471, "grad_norm": 0.6986918833987342, "learning_rate": 0.0002907949648073294, "loss": 0.5293, "step": 3205 }, { "epoch": 1.5087058823529413, "grad_norm": 0.1476286159008586, "learning_rate": 0.00029077983894199716, "loss": 0.1498, "step": 3206 }, { "epoch": 1.5091764705882353, "grad_norm": 0.4745912712180165, "learning_rate": 0.00029076470105337615, "loss": 0.4422, "step": 3207 }, { "epoch": 1.5096470588235293, "grad_norm": 0.6828848376174245, "learning_rate": 0.0002907495511427592, "loss": 0.5614, "step": 3208 }, { "epoch": 1.5101176470588236, "grad_norm": 0.16832198381325805, "learning_rate": 0.0002907343892114402, "loss": 0.1208, "step": 3209 }, { "epoch": 1.5105882352941178, "grad_norm": 0.21760153341204025, "learning_rate": 0.0002907192152607141, "loss": 0.2239, "step": 3210 }, { "epoch": 1.5110588235294118, "grad_norm": 0.15710797634220758, "learning_rate": 0.00029070402929187673, "loss": 0.0868, "step": 3211 }, { "epoch": 1.5115294117647058, "grad_norm": 0.36855594514128653, "learning_rate": 0.0002906888313062252, "loss": 0.4638, "step": 3212 }, { "epoch": 1.512, "grad_norm": 0.27571238856808267, "learning_rate": 0.00029067362130505734, "loss": 0.2071, "step": 3213 }, { "epoch": 1.5124705882352942, "grad_norm": 0.2802915709113234, "learning_rate": 0.00029065839928967224, "loss": 0.2801, "step": 3214 }, { "epoch": 1.5129411764705882, "grad_norm": 0.2620122389042848, "learning_rate": 0.0002906431652613699, "loss": 0.2112, "step": 3215 }, { "epoch": 1.5134117647058822, "grad_norm": 0.1899362545995326, "learning_rate": 0.0002906279192214514, "loss": 0.2381, "step": 3216 }, { "epoch": 1.5138823529411765, "grad_norm": 0.04099387609007442, "learning_rate": 0.0002906126611712188, "loss": 0.0082, "step": 3217 }, { "epoch": 1.5143529411764707, "grad_norm": 0.3701589033367112, "learning_rate": 0.00029059739111197524, "loss": 0.4029, "step": 3218 }, { "epoch": 1.5148235294117647, "grad_norm": 0.4142813626769385, "learning_rate": 0.0002905821090450249, "loss": 0.5155, "step": 3219 }, { "epoch": 1.5152941176470587, "grad_norm": 0.23627013667856325, "learning_rate": 0.00029056681497167285, "loss": 0.296, "step": 3220 }, { "epoch": 1.515764705882353, "grad_norm": 0.25578573376392577, "learning_rate": 0.00029055150889322536, "loss": 0.2756, "step": 3221 }, { "epoch": 1.5162352941176471, "grad_norm": 0.39114725684574286, "learning_rate": 0.0002905361908109896, "loss": 0.4561, "step": 3222 }, { "epoch": 1.5167058823529411, "grad_norm": 0.24358649456441836, "learning_rate": 0.0002905208607262738, "loss": 0.2329, "step": 3223 }, { "epoch": 1.5171764705882353, "grad_norm": 0.31596727587390155, "learning_rate": 0.0002905055186403873, "loss": 0.3405, "step": 3224 }, { "epoch": 1.5176470588235293, "grad_norm": 0.27243556769879834, "learning_rate": 0.0002904901645546403, "loss": 0.3022, "step": 3225 }, { "epoch": 1.5181176470588236, "grad_norm": 0.3135686924408339, "learning_rate": 0.0002904747984703441, "loss": 0.3967, "step": 3226 }, { "epoch": 1.5185882352941178, "grad_norm": 0.2758715817232573, "learning_rate": 0.00029045942038881113, "loss": 0.319, "step": 3227 }, { "epoch": 1.5190588235294118, "grad_norm": 0.5340551611524674, "learning_rate": 0.00029044403031135466, "loss": 0.5069, "step": 3228 }, { "epoch": 1.5195294117647058, "grad_norm": 0.31298046345433345, "learning_rate": 0.0002904286282392892, "loss": 0.2782, "step": 3229 }, { "epoch": 1.52, "grad_norm": 0.40269416162398325, "learning_rate": 0.00029041321417393006, "loss": 0.2678, "step": 3230 }, { "epoch": 1.5204705882352942, "grad_norm": 1.4450669934235854, "learning_rate": 0.0002903977881165937, "loss": 1.7509, "step": 3231 }, { "epoch": 1.5209411764705882, "grad_norm": 0.4787617607875538, "learning_rate": 0.0002903823500685976, "loss": 0.3485, "step": 3232 }, { "epoch": 1.5214117647058822, "grad_norm": 0.41297811603026996, "learning_rate": 0.0002903669000312602, "loss": 0.3266, "step": 3233 }, { "epoch": 1.5218823529411765, "grad_norm": 0.28364875420071234, "learning_rate": 0.0002903514380059011, "loss": 0.2279, "step": 3234 }, { "epoch": 1.5223529411764707, "grad_norm": 0.36750409430356573, "learning_rate": 0.0002903359639938407, "loss": 0.3897, "step": 3235 }, { "epoch": 1.5228235294117647, "grad_norm": 0.15298595265309942, "learning_rate": 0.0002903204779964007, "loss": 0.0955, "step": 3236 }, { "epoch": 1.5232941176470587, "grad_norm": 0.4729409542165609, "learning_rate": 0.00029030498001490356, "loss": 0.6624, "step": 3237 }, { "epoch": 1.523764705882353, "grad_norm": 0.5507936553184603, "learning_rate": 0.00029028947005067297, "loss": 0.5964, "step": 3238 }, { "epoch": 1.5242352941176471, "grad_norm": 0.18067801104210826, "learning_rate": 0.00029027394810503346, "loss": 0.1779, "step": 3239 }, { "epoch": 1.5247058823529411, "grad_norm": 0.409819917707078, "learning_rate": 0.0002902584141793108, "loss": 0.5369, "step": 3240 }, { "epoch": 1.5251764705882351, "grad_norm": 0.24921671725643993, "learning_rate": 0.0002902428682748316, "loss": 0.1851, "step": 3241 }, { "epoch": 1.5256470588235294, "grad_norm": 0.34334120280880315, "learning_rate": 0.0002902273103929236, "loss": 0.3763, "step": 3242 }, { "epoch": 1.5261176470588236, "grad_norm": 0.5010737196900307, "learning_rate": 0.00029021174053491547, "loss": 0.5427, "step": 3243 }, { "epoch": 1.5265882352941178, "grad_norm": 0.3106456221316441, "learning_rate": 0.0002901961587021369, "loss": 0.2342, "step": 3244 }, { "epoch": 1.5270588235294118, "grad_norm": 0.3233243251041542, "learning_rate": 0.00029018056489591874, "loss": 0.3503, "step": 3245 }, { "epoch": 1.5275294117647058, "grad_norm": 0.41088744296569135, "learning_rate": 0.0002901649591175928, "loss": 0.3359, "step": 3246 }, { "epoch": 1.528, "grad_norm": 0.39424001673203374, "learning_rate": 0.0002901493413684918, "loss": 0.2643, "step": 3247 }, { "epoch": 1.5284705882352942, "grad_norm": 0.2111447216076576, "learning_rate": 0.0002901337116499496, "loss": 0.1943, "step": 3248 }, { "epoch": 1.5289411764705882, "grad_norm": 0.23252834279677978, "learning_rate": 0.0002901180699633011, "loss": 0.1911, "step": 3249 }, { "epoch": 1.5294117647058822, "grad_norm": 0.6927410050940995, "learning_rate": 0.00029010241630988217, "loss": 0.6764, "step": 3250 }, { "epoch": 1.5298823529411765, "grad_norm": 0.23695441511102233, "learning_rate": 0.0002900867506910297, "loss": 0.318, "step": 3251 }, { "epoch": 1.5303529411764707, "grad_norm": 0.3854000996494527, "learning_rate": 0.00029007107310808153, "loss": 0.3409, "step": 3252 }, { "epoch": 1.5308235294117647, "grad_norm": 0.2748621017358095, "learning_rate": 0.00029005538356237674, "loss": 0.3634, "step": 3253 }, { "epoch": 1.5312941176470587, "grad_norm": 0.4198321502261303, "learning_rate": 0.00029003968205525523, "loss": 0.4402, "step": 3254 }, { "epoch": 1.531764705882353, "grad_norm": 0.3576725040943871, "learning_rate": 0.00029002396858805794, "loss": 0.3732, "step": 3255 }, { "epoch": 1.5322352941176471, "grad_norm": 0.2061039413337315, "learning_rate": 0.0002900082431621269, "loss": 0.156, "step": 3256 }, { "epoch": 1.5327058823529411, "grad_norm": 0.37036396916324393, "learning_rate": 0.0002899925057788052, "loss": 0.3352, "step": 3257 }, { "epoch": 1.5331764705882351, "grad_norm": 0.1701711500217512, "learning_rate": 0.00028997675643943686, "loss": 0.1485, "step": 3258 }, { "epoch": 1.5336470588235294, "grad_norm": 0.4035620877859475, "learning_rate": 0.0002899609951453669, "loss": 0.4292, "step": 3259 }, { "epoch": 1.5341176470588236, "grad_norm": 0.2953870694698706, "learning_rate": 0.0002899452218979415, "loss": 0.32, "step": 3260 }, { "epoch": 1.5345882352941178, "grad_norm": 0.1304373030654985, "learning_rate": 0.0002899294366985076, "loss": 0.097, "step": 3261 }, { "epoch": 1.5350588235294118, "grad_norm": 0.4071976450710939, "learning_rate": 0.0002899136395484136, "loss": 0.4665, "step": 3262 }, { "epoch": 1.5355294117647058, "grad_norm": 0.2706936084412888, "learning_rate": 0.00028989783044900845, "loss": 0.2612, "step": 3263 }, { "epoch": 1.536, "grad_norm": 0.22823330163847153, "learning_rate": 0.0002898820094016424, "loss": 0.2035, "step": 3264 }, { "epoch": 1.5364705882352943, "grad_norm": 0.3183997426641704, "learning_rate": 0.0002898661764076666, "loss": 0.293, "step": 3265 }, { "epoch": 1.5369411764705883, "grad_norm": 0.1935631393248035, "learning_rate": 0.00028985033146843335, "loss": 0.1017, "step": 3266 }, { "epoch": 1.5374117647058823, "grad_norm": 0.3583934974801458, "learning_rate": 0.00028983447458529575, "loss": 0.3316, "step": 3267 }, { "epoch": 1.5378823529411765, "grad_norm": 0.35414325249790507, "learning_rate": 0.0002898186057596082, "loss": 0.3023, "step": 3268 }, { "epoch": 1.5383529411764707, "grad_norm": 0.10252089294010415, "learning_rate": 0.0002898027249927259, "loss": 0.0876, "step": 3269 }, { "epoch": 1.5388235294117647, "grad_norm": 0.3213577637893816, "learning_rate": 0.00028978683228600514, "loss": 0.1669, "step": 3270 }, { "epoch": 1.5392941176470587, "grad_norm": 0.3858357722915075, "learning_rate": 0.0002897709276408032, "loss": 0.4105, "step": 3271 }, { "epoch": 1.539764705882353, "grad_norm": 0.4059426963967318, "learning_rate": 0.00028975501105847856, "loss": 0.238, "step": 3272 }, { "epoch": 1.5402352941176471, "grad_norm": 0.17452267155551054, "learning_rate": 0.00028973908254039044, "loss": 0.0838, "step": 3273 }, { "epoch": 1.5407058823529411, "grad_norm": 0.3686370649760335, "learning_rate": 0.00028972314208789926, "loss": 0.4265, "step": 3274 }, { "epoch": 1.5411764705882351, "grad_norm": 0.334337316440936, "learning_rate": 0.00028970718970236644, "loss": 0.3494, "step": 3275 }, { "epoch": 1.5416470588235294, "grad_norm": 0.30835984525205623, "learning_rate": 0.00028969122538515435, "loss": 0.3154, "step": 3276 }, { "epoch": 1.5421176470588236, "grad_norm": 0.44063309709287846, "learning_rate": 0.00028967524913762636, "loss": 0.4507, "step": 3277 }, { "epoch": 1.5425882352941176, "grad_norm": 0.12846859771855434, "learning_rate": 0.00028965926096114706, "loss": 0.0665, "step": 3278 }, { "epoch": 1.5430588235294118, "grad_norm": 0.2760099182691731, "learning_rate": 0.00028964326085708183, "loss": 0.2397, "step": 3279 }, { "epoch": 1.5435294117647058, "grad_norm": 0.4344676461565325, "learning_rate": 0.00028962724882679715, "loss": 0.4794, "step": 3280 }, { "epoch": 1.544, "grad_norm": 0.18567163306655296, "learning_rate": 0.00028961122487166063, "loss": 0.0966, "step": 3281 }, { "epoch": 1.5444705882352943, "grad_norm": 0.8308474673648972, "learning_rate": 0.00028959518899304064, "loss": 0.7329, "step": 3282 }, { "epoch": 1.5449411764705883, "grad_norm": 0.45339810165990585, "learning_rate": 0.00028957914119230686, "loss": 0.322, "step": 3283 }, { "epoch": 1.5454117647058823, "grad_norm": 0.3377352939048995, "learning_rate": 0.0002895630814708297, "loss": 0.3403, "step": 3284 }, { "epoch": 1.5458823529411765, "grad_norm": 0.39874346789743664, "learning_rate": 0.00028954700982998093, "loss": 0.3603, "step": 3285 }, { "epoch": 1.5463529411764707, "grad_norm": 0.16711099241514454, "learning_rate": 0.000289530926271133, "loss": 0.1147, "step": 3286 }, { "epoch": 1.5468235294117647, "grad_norm": 0.13544487278735465, "learning_rate": 0.0002895148307956596, "loss": 0.1063, "step": 3287 }, { "epoch": 1.5472941176470587, "grad_norm": 0.24900124211432131, "learning_rate": 0.00028949872340493525, "loss": 0.1681, "step": 3288 }, { "epoch": 1.547764705882353, "grad_norm": 0.20671548475375423, "learning_rate": 0.00028948260410033576, "loss": 0.1259, "step": 3289 }, { "epoch": 1.5482352941176472, "grad_norm": 0.30537259159156943, "learning_rate": 0.00028946647288323766, "loss": 0.3487, "step": 3290 }, { "epoch": 1.5487058823529412, "grad_norm": 0.48684742705387124, "learning_rate": 0.00028945032975501877, "loss": 0.5936, "step": 3291 }, { "epoch": 1.5491764705882352, "grad_norm": 0.1819746782208259, "learning_rate": 0.0002894341747170577, "loss": 0.1512, "step": 3292 }, { "epoch": 1.5496470588235294, "grad_norm": 0.14211925816414567, "learning_rate": 0.00028941800777073414, "loss": 0.141, "step": 3293 }, { "epoch": 1.5501176470588236, "grad_norm": 0.3554777327388961, "learning_rate": 0.0002894018289174289, "loss": 0.319, "step": 3294 }, { "epoch": 1.5505882352941176, "grad_norm": 0.5613517113790184, "learning_rate": 0.0002893856381585237, "loss": 0.3818, "step": 3295 }, { "epoch": 1.5510588235294116, "grad_norm": 0.2676742069847251, "learning_rate": 0.0002893694354954014, "loss": 0.1776, "step": 3296 }, { "epoch": 1.5515294117647058, "grad_norm": 1.5152598200008665, "learning_rate": 0.00028935322092944564, "loss": 0.8787, "step": 3297 }, { "epoch": 1.552, "grad_norm": 0.34021282368352096, "learning_rate": 0.00028933699446204126, "loss": 0.2475, "step": 3298 }, { "epoch": 1.5524705882352943, "grad_norm": 0.6032072820078459, "learning_rate": 0.0002893207560945742, "loss": 0.577, "step": 3299 }, { "epoch": 1.5529411764705883, "grad_norm": 0.46165542813442906, "learning_rate": 0.0002893045058284311, "loss": 0.2893, "step": 3300 }, { "epoch": 1.5534117647058823, "grad_norm": 0.5578882646411878, "learning_rate": 0.000289288243665, "loss": 0.2243, "step": 3301 }, { "epoch": 1.5538823529411765, "grad_norm": 0.5235806054748977, "learning_rate": 0.00028927196960566964, "loss": 0.3539, "step": 3302 }, { "epoch": 1.5543529411764707, "grad_norm": 0.3498897107203864, "learning_rate": 0.00028925568365182995, "loss": 0.3809, "step": 3303 }, { "epoch": 1.5548235294117647, "grad_norm": 0.34121587867359565, "learning_rate": 0.00028923938580487183, "loss": 0.3582, "step": 3304 }, { "epoch": 1.5552941176470587, "grad_norm": 0.3191265993738294, "learning_rate": 0.0002892230760661872, "loss": 0.3439, "step": 3305 }, { "epoch": 1.555764705882353, "grad_norm": 0.09310331248774747, "learning_rate": 0.000289206754437169, "loss": 0.0569, "step": 3306 }, { "epoch": 1.5562352941176472, "grad_norm": 0.12600412050889226, "learning_rate": 0.0002891904209192112, "loss": 0.0719, "step": 3307 }, { "epoch": 1.5567058823529412, "grad_norm": 0.36940600985565053, "learning_rate": 0.00028917407551370863, "loss": 0.4203, "step": 3308 }, { "epoch": 1.5571764705882352, "grad_norm": 0.30733966506761584, "learning_rate": 0.00028915771822205747, "loss": 0.3349, "step": 3309 }, { "epoch": 1.5576470588235294, "grad_norm": 0.11855831951508791, "learning_rate": 0.00028914134904565455, "loss": 0.0581, "step": 3310 }, { "epoch": 1.5581176470588236, "grad_norm": 0.08010433681688293, "learning_rate": 0.00028912496798589796, "loss": 0.0358, "step": 3311 }, { "epoch": 1.5585882352941176, "grad_norm": 0.1401779503935949, "learning_rate": 0.0002891085750441867, "loss": 0.0801, "step": 3312 }, { "epoch": 1.5590588235294116, "grad_norm": 0.4301499524197329, "learning_rate": 0.0002890921702219208, "loss": 0.292, "step": 3313 }, { "epoch": 1.5595294117647058, "grad_norm": 0.49825962472128543, "learning_rate": 0.00028907575352050133, "loss": 0.4485, "step": 3314 }, { "epoch": 1.56, "grad_norm": 0.4814270787048861, "learning_rate": 0.00028905932494133034, "loss": 0.2723, "step": 3315 }, { "epoch": 1.5604705882352943, "grad_norm": 0.20446677682059142, "learning_rate": 0.0002890428844858109, "loss": 0.1884, "step": 3316 }, { "epoch": 1.5609411764705883, "grad_norm": 0.24535928081357283, "learning_rate": 0.0002890264321553471, "loss": 0.1726, "step": 3317 }, { "epoch": 1.5614117647058823, "grad_norm": 0.29398990389554863, "learning_rate": 0.00028900996795134414, "loss": 0.2658, "step": 3318 }, { "epoch": 1.5618823529411765, "grad_norm": 0.26152107100498434, "learning_rate": 0.000288993491875208, "loss": 0.2744, "step": 3319 }, { "epoch": 1.5623529411764707, "grad_norm": 0.4133418530917147, "learning_rate": 0.000288977003928346, "loss": 0.2671, "step": 3320 }, { "epoch": 1.5628235294117647, "grad_norm": 0.532423482008353, "learning_rate": 0.0002889605041121661, "loss": 0.6125, "step": 3321 }, { "epoch": 1.5632941176470587, "grad_norm": 0.42510046052613043, "learning_rate": 0.0002889439924280776, "loss": 0.4136, "step": 3322 }, { "epoch": 1.563764705882353, "grad_norm": 0.3354314804033865, "learning_rate": 0.0002889274688774906, "loss": 0.4613, "step": 3323 }, { "epoch": 1.5642352941176472, "grad_norm": 0.4534161244154311, "learning_rate": 0.00028891093346181633, "loss": 0.4136, "step": 3324 }, { "epoch": 1.5647058823529412, "grad_norm": 0.1499221379334014, "learning_rate": 0.000288894386182467, "loss": 0.1201, "step": 3325 }, { "epoch": 1.5651764705882352, "grad_norm": 0.24226660112772988, "learning_rate": 0.0002888778270408558, "loss": 0.255, "step": 3326 }, { "epoch": 1.5656470588235294, "grad_norm": 0.20476134885764571, "learning_rate": 0.000288861256038397, "loss": 0.1444, "step": 3327 }, { "epoch": 1.5661176470588236, "grad_norm": 0.3818436067196998, "learning_rate": 0.0002888446731765058, "loss": 0.3757, "step": 3328 }, { "epoch": 1.5665882352941176, "grad_norm": 0.25604495244321424, "learning_rate": 0.00028882807845659846, "loss": 0.2093, "step": 3329 }, { "epoch": 1.5670588235294116, "grad_norm": 0.45150383554986867, "learning_rate": 0.00028881147188009233, "loss": 0.5239, "step": 3330 }, { "epoch": 1.5675294117647058, "grad_norm": 0.2072925492803478, "learning_rate": 0.00028879485344840563, "loss": 0.102, "step": 3331 }, { "epoch": 1.568, "grad_norm": 0.34344192000559487, "learning_rate": 0.00028877822316295763, "loss": 0.3968, "step": 3332 }, { "epoch": 1.5684705882352943, "grad_norm": 0.3357248201521982, "learning_rate": 0.0002887615810251687, "loss": 0.4688, "step": 3333 }, { "epoch": 1.5689411764705883, "grad_norm": 0.16584152404991756, "learning_rate": 0.0002887449270364601, "loss": 0.1288, "step": 3334 }, { "epoch": 1.5694117647058823, "grad_norm": 0.2709306093128095, "learning_rate": 0.0002887282611982542, "loss": 0.1563, "step": 3335 }, { "epoch": 1.5698823529411765, "grad_norm": 0.38181468572776195, "learning_rate": 0.0002887115835119743, "loss": 0.3581, "step": 3336 }, { "epoch": 1.5703529411764707, "grad_norm": 0.4246868257718763, "learning_rate": 0.00028869489397904486, "loss": 0.3679, "step": 3337 }, { "epoch": 1.5708235294117647, "grad_norm": 0.4322081949593385, "learning_rate": 0.00028867819260089113, "loss": 0.6149, "step": 3338 }, { "epoch": 1.5712941176470587, "grad_norm": 0.40770067986948305, "learning_rate": 0.0002886614793789396, "loss": 0.3817, "step": 3339 }, { "epoch": 1.571764705882353, "grad_norm": 0.3565600760787217, "learning_rate": 0.0002886447543146175, "loss": 0.3735, "step": 3340 }, { "epoch": 1.5722352941176472, "grad_norm": 0.3259546999734731, "learning_rate": 0.0002886280174093534, "loss": 0.2816, "step": 3341 }, { "epoch": 1.5727058823529412, "grad_norm": 0.33022531989494347, "learning_rate": 0.0002886112686645766, "loss": 0.1722, "step": 3342 }, { "epoch": 1.5731764705882352, "grad_norm": 0.6687697896690752, "learning_rate": 0.00028859450808171763, "loss": 0.5375, "step": 3343 }, { "epoch": 1.5736470588235294, "grad_norm": 0.2975686934714623, "learning_rate": 0.00028857773566220784, "loss": 0.3611, "step": 3344 }, { "epoch": 1.5741176470588236, "grad_norm": 0.20338429536764027, "learning_rate": 0.0002885609514074797, "loss": 0.1775, "step": 3345 }, { "epoch": 1.5745882352941176, "grad_norm": 0.5828963918051444, "learning_rate": 0.0002885441553189667, "loss": 0.5537, "step": 3346 }, { "epoch": 1.5750588235294116, "grad_norm": 0.5372268763200861, "learning_rate": 0.00028852734739810324, "loss": 0.4178, "step": 3347 }, { "epoch": 1.5755294117647058, "grad_norm": 0.2461892489497176, "learning_rate": 0.00028851052764632487, "loss": 0.2494, "step": 3348 }, { "epoch": 1.576, "grad_norm": 0.29315915735386905, "learning_rate": 0.00028849369606506807, "loss": 0.3963, "step": 3349 }, { "epoch": 1.576470588235294, "grad_norm": 0.20638391552321886, "learning_rate": 0.0002884768526557703, "loss": 0.3022, "step": 3350 }, { "epoch": 1.5769411764705883, "grad_norm": 0.5588173812787994, "learning_rate": 0.00028845999741987003, "loss": 0.6575, "step": 3351 }, { "epoch": 1.5774117647058823, "grad_norm": 0.2582326835416237, "learning_rate": 0.0002884431303588069, "loss": 0.3203, "step": 3352 }, { "epoch": 1.5778823529411765, "grad_norm": 0.339229620771963, "learning_rate": 0.00028842625147402134, "loss": 0.3084, "step": 3353 }, { "epoch": 1.5783529411764707, "grad_norm": 0.33297129219233057, "learning_rate": 0.00028840936076695493, "loss": 0.3021, "step": 3354 }, { "epoch": 1.5788235294117647, "grad_norm": 0.37306415807529236, "learning_rate": 0.0002883924582390503, "loss": 0.4044, "step": 3355 }, { "epoch": 1.5792941176470587, "grad_norm": 0.3498277344781058, "learning_rate": 0.0002883755438917508, "loss": 0.4404, "step": 3356 }, { "epoch": 1.579764705882353, "grad_norm": 0.1223569952501772, "learning_rate": 0.0002883586177265012, "loss": 0.0398, "step": 3357 }, { "epoch": 1.5802352941176472, "grad_norm": 0.11542002859644329, "learning_rate": 0.000288341679744747, "loss": 0.0621, "step": 3358 }, { "epoch": 1.5807058823529412, "grad_norm": 0.026436890774776546, "learning_rate": 0.0002883247299479348, "loss": 0.016, "step": 3359 }, { "epoch": 1.5811764705882352, "grad_norm": 0.19150030200331133, "learning_rate": 0.0002883077683375121, "loss": 0.0776, "step": 3360 }, { "epoch": 1.5816470588235294, "grad_norm": 0.31796321466803884, "learning_rate": 0.0002882907949149277, "loss": 0.2635, "step": 3361 }, { "epoch": 1.5821176470588236, "grad_norm": 0.22751217657482645, "learning_rate": 0.000288273809681631, "loss": 0.1713, "step": 3362 }, { "epoch": 1.5825882352941176, "grad_norm": 0.48399840601770844, "learning_rate": 0.00028825681263907276, "loss": 0.378, "step": 3363 }, { "epoch": 1.5830588235294116, "grad_norm": 0.13648250037606824, "learning_rate": 0.00028823980378870457, "loss": 0.0958, "step": 3364 }, { "epoch": 1.5835294117647059, "grad_norm": 0.3519833140942352, "learning_rate": 0.0002882227831319791, "loss": 0.3253, "step": 3365 }, { "epoch": 1.584, "grad_norm": 0.25231984798562257, "learning_rate": 0.00028820575067035, "loss": 0.2398, "step": 3366 }, { "epoch": 1.584470588235294, "grad_norm": 0.5908321111153498, "learning_rate": 0.00028818870640527183, "loss": 0.466, "step": 3367 }, { "epoch": 1.584941176470588, "grad_norm": 0.4152892597862898, "learning_rate": 0.0002881716503382004, "loss": 0.518, "step": 3368 }, { "epoch": 1.5854117647058823, "grad_norm": 0.40829602912489965, "learning_rate": 0.0002881545824705922, "loss": 0.4432, "step": 3369 }, { "epoch": 1.5858823529411765, "grad_norm": 0.211200159250056, "learning_rate": 0.00028813750280390506, "loss": 0.1196, "step": 3370 }, { "epoch": 1.5863529411764707, "grad_norm": 0.2790517142674654, "learning_rate": 0.00028812041133959764, "loss": 0.3093, "step": 3371 }, { "epoch": 1.5868235294117647, "grad_norm": 0.31369840779618574, "learning_rate": 0.0002881033080791296, "loss": 0.4173, "step": 3372 }, { "epoch": 1.5872941176470587, "grad_norm": 0.2996279034744644, "learning_rate": 0.0002880861930239616, "loss": 0.3258, "step": 3373 }, { "epoch": 1.587764705882353, "grad_norm": 0.26334329950809743, "learning_rate": 0.0002880690661755555, "loss": 0.2766, "step": 3374 }, { "epoch": 1.5882352941176472, "grad_norm": 0.17128713662509426, "learning_rate": 0.00028805192753537386, "loss": 0.1095, "step": 3375 }, { "epoch": 1.5887058823529412, "grad_norm": 0.21198276978277075, "learning_rate": 0.00028803477710488055, "loss": 0.1463, "step": 3376 }, { "epoch": 1.5891764705882352, "grad_norm": 0.36776137094036665, "learning_rate": 0.00028801761488554017, "loss": 0.3497, "step": 3377 }, { "epoch": 1.5896470588235294, "grad_norm": 0.28043690119194026, "learning_rate": 0.0002880004408788185, "loss": 0.3451, "step": 3378 }, { "epoch": 1.5901176470588236, "grad_norm": 0.18925535736333346, "learning_rate": 0.0002879832550861823, "loss": 0.1695, "step": 3379 }, { "epoch": 1.5905882352941176, "grad_norm": 0.17647034478464163, "learning_rate": 0.0002879660575090993, "loss": 0.1384, "step": 3380 }, { "epoch": 1.5910588235294116, "grad_norm": 0.39014404809632713, "learning_rate": 0.0002879488481490382, "loss": 0.3393, "step": 3381 }, { "epoch": 1.5915294117647059, "grad_norm": 0.1317216344940276, "learning_rate": 0.0002879316270074689, "loss": 0.082, "step": 3382 }, { "epoch": 1.592, "grad_norm": 0.13783026869798515, "learning_rate": 0.00028791439408586215, "loss": 0.0931, "step": 3383 }, { "epoch": 1.592470588235294, "grad_norm": 0.20960526840449267, "learning_rate": 0.00028789714938568965, "loss": 0.2108, "step": 3384 }, { "epoch": 1.592941176470588, "grad_norm": 0.3107590579990092, "learning_rate": 0.0002878798929084242, "loss": 0.4757, "step": 3385 }, { "epoch": 1.5934117647058823, "grad_norm": 0.3492432269438372, "learning_rate": 0.00028786262465553954, "loss": 0.3409, "step": 3386 }, { "epoch": 1.5938823529411765, "grad_norm": 0.3547157890291645, "learning_rate": 0.00028784534462851063, "loss": 0.2934, "step": 3387 }, { "epoch": 1.5943529411764708, "grad_norm": 0.44696831686412825, "learning_rate": 0.0002878280528288131, "loss": 0.6056, "step": 3388 }, { "epoch": 1.5948235294117648, "grad_norm": 0.2935577367082309, "learning_rate": 0.00028781074925792377, "loss": 0.3661, "step": 3389 }, { "epoch": 1.5952941176470588, "grad_norm": 0.394698983295481, "learning_rate": 0.00028779343391732053, "loss": 0.3887, "step": 3390 }, { "epoch": 1.595764705882353, "grad_norm": 0.3984215744232791, "learning_rate": 0.0002877761068084822, "loss": 0.5714, "step": 3391 }, { "epoch": 1.5962352941176472, "grad_norm": 0.16302636015493788, "learning_rate": 0.00028775876793288855, "loss": 0.1499, "step": 3392 }, { "epoch": 1.5967058823529412, "grad_norm": 0.14668585208479284, "learning_rate": 0.0002877414172920204, "loss": 0.0793, "step": 3393 }, { "epoch": 1.5971764705882352, "grad_norm": 0.3023169808646857, "learning_rate": 0.00028772405488735963, "loss": 0.3547, "step": 3394 }, { "epoch": 1.5976470588235294, "grad_norm": 0.23442551304323936, "learning_rate": 0.000287706680720389, "loss": 0.1463, "step": 3395 }, { "epoch": 1.5981176470588236, "grad_norm": 0.41375778404298624, "learning_rate": 0.0002876892947925924, "loss": 0.4759, "step": 3396 }, { "epoch": 1.5985882352941176, "grad_norm": 0.3474580160842615, "learning_rate": 0.0002876718971054547, "loss": 0.3021, "step": 3397 }, { "epoch": 1.5990588235294116, "grad_norm": 0.4461480975727779, "learning_rate": 0.00028765448766046163, "loss": 0.3898, "step": 3398 }, { "epoch": 1.5995294117647059, "grad_norm": 0.3346681323942784, "learning_rate": 0.00028763706645910014, "loss": 0.3352, "step": 3399 }, { "epoch": 1.6, "grad_norm": 0.2323267863840133, "learning_rate": 0.0002876196335028581, "loss": 0.1137, "step": 3400 }, { "epoch": 1.600470588235294, "grad_norm": 0.1931101838395013, "learning_rate": 0.00028760218879322436, "loss": 0.269, "step": 3401 }, { "epoch": 1.600941176470588, "grad_norm": 0.1818373252683165, "learning_rate": 0.00028758473233168874, "loss": 0.1513, "step": 3402 }, { "epoch": 1.6014117647058823, "grad_norm": 0.5017111724852542, "learning_rate": 0.0002875672641197421, "loss": 0.5041, "step": 3403 }, { "epoch": 1.6018823529411765, "grad_norm": 0.15964453096145145, "learning_rate": 0.00028754978415887635, "loss": 0.1431, "step": 3404 }, { "epoch": 1.6023529411764705, "grad_norm": 0.3047888793320893, "learning_rate": 0.0002875322924505844, "loss": 0.3364, "step": 3405 }, { "epoch": 1.6028235294117648, "grad_norm": 0.09902562624205888, "learning_rate": 0.00028751478899636, "loss": 0.0529, "step": 3406 }, { "epoch": 1.6032941176470588, "grad_norm": 0.724348532529016, "learning_rate": 0.00028749727379769817, "loss": 0.6122, "step": 3407 }, { "epoch": 1.603764705882353, "grad_norm": 0.23534160106315483, "learning_rate": 0.0002874797468560948, "loss": 0.2388, "step": 3408 }, { "epoch": 1.6042352941176472, "grad_norm": 0.5001267460837686, "learning_rate": 0.0002874622081730466, "loss": 0.7826, "step": 3409 }, { "epoch": 1.6047058823529412, "grad_norm": 0.3809459175028107, "learning_rate": 0.0002874446577500516, "loss": 0.5257, "step": 3410 }, { "epoch": 1.6051764705882352, "grad_norm": 0.24211571340189958, "learning_rate": 0.0002874270955886087, "loss": 0.3371, "step": 3411 }, { "epoch": 1.6056470588235294, "grad_norm": 0.19859836521713276, "learning_rate": 0.0002874095216902178, "loss": 0.1477, "step": 3412 }, { "epoch": 1.6061176470588236, "grad_norm": 0.25526897968670376, "learning_rate": 0.00028739193605637967, "loss": 0.1395, "step": 3413 }, { "epoch": 1.6065882352941177, "grad_norm": 0.21917889619885458, "learning_rate": 0.00028737433868859634, "loss": 0.1311, "step": 3414 }, { "epoch": 1.6070588235294117, "grad_norm": 0.26896285757641764, "learning_rate": 0.0002873567295883707, "loss": 0.2086, "step": 3415 }, { "epoch": 1.6075294117647059, "grad_norm": 0.5191340560269743, "learning_rate": 0.00028733910875720654, "loss": 0.5227, "step": 3416 }, { "epoch": 1.608, "grad_norm": 0.22747004059664763, "learning_rate": 0.00028732147619660883, "loss": 0.1315, "step": 3417 }, { "epoch": 1.608470588235294, "grad_norm": 0.35468927291972024, "learning_rate": 0.00028730383190808364, "loss": 0.3079, "step": 3418 }, { "epoch": 1.608941176470588, "grad_norm": 0.3236547474903282, "learning_rate": 0.0002872861758931376, "loss": 0.3007, "step": 3419 }, { "epoch": 1.6094117647058823, "grad_norm": 0.5527899259305099, "learning_rate": 0.0002872685081532788, "loss": 0.6353, "step": 3420 }, { "epoch": 1.6098823529411765, "grad_norm": 0.2794233071410728, "learning_rate": 0.00028725082869001615, "loss": 0.3414, "step": 3421 }, { "epoch": 1.6103529411764705, "grad_norm": 0.036775861094529086, "learning_rate": 0.00028723313750485946, "loss": 0.0124, "step": 3422 }, { "epoch": 1.6108235294117645, "grad_norm": 0.16622586878855022, "learning_rate": 0.0002872154345993197, "loss": 0.1046, "step": 3423 }, { "epoch": 1.6112941176470588, "grad_norm": 0.13607610975296072, "learning_rate": 0.0002871977199749088, "loss": 0.0683, "step": 3424 }, { "epoch": 1.611764705882353, "grad_norm": 0.7159408867183117, "learning_rate": 0.00028717999363313967, "loss": 0.6833, "step": 3425 }, { "epoch": 1.6122352941176472, "grad_norm": 0.2350394078629582, "learning_rate": 0.0002871622555755262, "loss": 0.1891, "step": 3426 }, { "epoch": 1.6127058823529412, "grad_norm": 0.6131404083200354, "learning_rate": 0.00028714450580358334, "loss": 0.6861, "step": 3427 }, { "epoch": 1.6131764705882352, "grad_norm": 0.43834816151646605, "learning_rate": 0.000287126744318827, "loss": 0.679, "step": 3428 }, { "epoch": 1.6136470588235294, "grad_norm": 0.14735904898447913, "learning_rate": 0.0002871089711227741, "loss": 0.2091, "step": 3429 }, { "epoch": 1.6141176470588237, "grad_norm": 0.2936077718807252, "learning_rate": 0.0002870911862169426, "loss": 0.3663, "step": 3430 }, { "epoch": 1.6145882352941177, "grad_norm": 0.37077937993189697, "learning_rate": 0.0002870733896028513, "loss": 0.3376, "step": 3431 }, { "epoch": 1.6150588235294117, "grad_norm": 0.12347911861151165, "learning_rate": 0.00028705558128202017, "loss": 0.0852, "step": 3432 }, { "epoch": 1.6155294117647059, "grad_norm": 0.39473105163603395, "learning_rate": 0.0002870377612559702, "loss": 0.4708, "step": 3433 }, { "epoch": 1.616, "grad_norm": 0.4188981046079965, "learning_rate": 0.0002870199295262232, "loss": 0.3625, "step": 3434 }, { "epoch": 1.616470588235294, "grad_norm": 0.34488153009086436, "learning_rate": 0.00028700208609430217, "loss": 0.2913, "step": 3435 }, { "epoch": 1.616941176470588, "grad_norm": 0.3689202618409079, "learning_rate": 0.00028698423096173097, "loss": 0.2701, "step": 3436 }, { "epoch": 1.6174117647058823, "grad_norm": 0.29827315007723404, "learning_rate": 0.0002869663641300346, "loss": 0.4427, "step": 3437 }, { "epoch": 1.6178823529411765, "grad_norm": 0.49040972525978405, "learning_rate": 0.00028694848560073885, "loss": 0.5002, "step": 3438 }, { "epoch": 1.6183529411764705, "grad_norm": 0.2716204692030685, "learning_rate": 0.00028693059537537077, "loss": 0.3503, "step": 3439 }, { "epoch": 1.6188235294117646, "grad_norm": 0.38559388747305556, "learning_rate": 0.00028691269345545816, "loss": 0.3529, "step": 3440 }, { "epoch": 1.6192941176470588, "grad_norm": 0.46043759573282533, "learning_rate": 0.00028689477984252996, "loss": 0.4476, "step": 3441 }, { "epoch": 1.619764705882353, "grad_norm": 0.46902887648495334, "learning_rate": 0.00028687685453811614, "loss": 0.1423, "step": 3442 }, { "epoch": 1.6202352941176472, "grad_norm": 0.2463655481445738, "learning_rate": 0.00028685891754374755, "loss": 0.1278, "step": 3443 }, { "epoch": 1.6207058823529412, "grad_norm": 0.2597641752281796, "learning_rate": 0.0002868409688609561, "loss": 0.2475, "step": 3444 }, { "epoch": 1.6211764705882352, "grad_norm": 0.3407392334643949, "learning_rate": 0.00028682300849127465, "loss": 0.1666, "step": 3445 }, { "epoch": 1.6216470588235294, "grad_norm": 0.7285988806067366, "learning_rate": 0.0002868050364362373, "loss": 0.29, "step": 3446 }, { "epoch": 1.6221176470588237, "grad_norm": 0.398449348713351, "learning_rate": 0.00028678705269737873, "loss": 0.2054, "step": 3447 }, { "epoch": 1.6225882352941177, "grad_norm": 0.39336626580378303, "learning_rate": 0.0002867690572762349, "loss": 0.2743, "step": 3448 }, { "epoch": 1.6230588235294117, "grad_norm": 0.3285074731083739, "learning_rate": 0.0002867510501743428, "loss": 0.2427, "step": 3449 }, { "epoch": 1.6235294117647059, "grad_norm": 0.440988031831508, "learning_rate": 0.0002867330313932402, "loss": 0.3847, "step": 3450 }, { "epoch": 1.624, "grad_norm": 0.46368184691418096, "learning_rate": 0.00028671500093446604, "loss": 0.3405, "step": 3451 }, { "epoch": 1.624470588235294, "grad_norm": 0.07195800861342334, "learning_rate": 0.0002866969587995602, "loss": 0.0212, "step": 3452 }, { "epoch": 1.6249411764705881, "grad_norm": 0.2915025225209558, "learning_rate": 0.00028667890499006364, "loss": 0.2221, "step": 3453 }, { "epoch": 1.6254117647058823, "grad_norm": 0.3729004457639456, "learning_rate": 0.00028666083950751824, "loss": 0.375, "step": 3454 }, { "epoch": 1.6258823529411766, "grad_norm": 0.19182174008969288, "learning_rate": 0.0002866427623534667, "loss": 0.1526, "step": 3455 }, { "epoch": 1.6263529411764706, "grad_norm": 0.391383982390822, "learning_rate": 0.0002866246735294531, "loss": 0.3666, "step": 3456 }, { "epoch": 1.6268235294117646, "grad_norm": 0.291911994641995, "learning_rate": 0.00028660657303702225, "loss": 0.1567, "step": 3457 }, { "epoch": 1.6272941176470588, "grad_norm": 0.21250141865317246, "learning_rate": 0.00028658846087771994, "loss": 0.1421, "step": 3458 }, { "epoch": 1.627764705882353, "grad_norm": 0.27598942420076056, "learning_rate": 0.0002865703370530932, "loss": 0.3104, "step": 3459 }, { "epoch": 1.6282352941176472, "grad_norm": 0.551377130720093, "learning_rate": 0.0002865522015646897, "loss": 0.5279, "step": 3460 }, { "epoch": 1.6287058823529412, "grad_norm": 0.12618484505544592, "learning_rate": 0.0002865340544140585, "loss": 0.0773, "step": 3461 }, { "epoch": 1.6291764705882352, "grad_norm": 0.21625799955513278, "learning_rate": 0.00028651589560274934, "loss": 0.1465, "step": 3462 }, { "epoch": 1.6296470588235294, "grad_norm": 0.25621461669803103, "learning_rate": 0.00028649772513231306, "loss": 0.3353, "step": 3463 }, { "epoch": 1.6301176470588237, "grad_norm": 0.35813226197092185, "learning_rate": 0.00028647954300430154, "loss": 0.4788, "step": 3464 }, { "epoch": 1.6305882352941177, "grad_norm": 0.27486906551173856, "learning_rate": 0.0002864613492202676, "loss": 0.2714, "step": 3465 }, { "epoch": 1.6310588235294117, "grad_norm": 0.46194892727168285, "learning_rate": 0.00028644314378176513, "loss": 0.509, "step": 3466 }, { "epoch": 1.631529411764706, "grad_norm": 0.23566973730351357, "learning_rate": 0.00028642492669034894, "loss": 0.1378, "step": 3467 }, { "epoch": 1.6320000000000001, "grad_norm": 0.2857694976047905, "learning_rate": 0.0002864066979475748, "loss": 0.32, "step": 3468 }, { "epoch": 1.6324705882352941, "grad_norm": 0.22433689988034888, "learning_rate": 0.00028638845755499964, "loss": 0.2865, "step": 3469 }, { "epoch": 1.6329411764705881, "grad_norm": 0.36426736531595244, "learning_rate": 0.00028637020551418123, "loss": 0.3251, "step": 3470 }, { "epoch": 1.6334117647058823, "grad_norm": 0.6843950769929732, "learning_rate": 0.00028635194182667834, "loss": 0.7076, "step": 3471 }, { "epoch": 1.6338823529411766, "grad_norm": 0.329823805092091, "learning_rate": 0.0002863336664940508, "loss": 0.3169, "step": 3472 }, { "epoch": 1.6343529411764706, "grad_norm": 0.18831043967724473, "learning_rate": 0.00028631537951785945, "loss": 0.1513, "step": 3473 }, { "epoch": 1.6348235294117646, "grad_norm": 0.28335016426422655, "learning_rate": 0.000286297080899666, "loss": 0.2197, "step": 3474 }, { "epoch": 1.6352941176470588, "grad_norm": 0.2639201956579521, "learning_rate": 0.0002862787706410334, "loss": 0.2104, "step": 3475 }, { "epoch": 1.635764705882353, "grad_norm": 0.5298131913796679, "learning_rate": 0.00028626044874352534, "loss": 0.4086, "step": 3476 }, { "epoch": 1.636235294117647, "grad_norm": 0.1149823978620186, "learning_rate": 0.0002862421152087066, "loss": 0.1048, "step": 3477 }, { "epoch": 1.636705882352941, "grad_norm": 0.3006851322922738, "learning_rate": 0.00028622377003814295, "loss": 0.3292, "step": 3478 }, { "epoch": 1.6371764705882352, "grad_norm": 0.35864344836830336, "learning_rate": 0.00028620541323340115, "loss": 0.2614, "step": 3479 }, { "epoch": 1.6376470588235295, "grad_norm": 0.2852415340655477, "learning_rate": 0.00028618704479604895, "loss": 0.2613, "step": 3480 }, { "epoch": 1.6381176470588237, "grad_norm": 0.31247098219684594, "learning_rate": 0.00028616866472765514, "loss": 0.3163, "step": 3481 }, { "epoch": 1.6385882352941177, "grad_norm": 0.2058014276711232, "learning_rate": 0.00028615027302978947, "loss": 0.1606, "step": 3482 }, { "epoch": 1.6390588235294117, "grad_norm": 0.43381335005908406, "learning_rate": 0.00028613186970402274, "loss": 0.5093, "step": 3483 }, { "epoch": 1.639529411764706, "grad_norm": 0.28141350238505625, "learning_rate": 0.0002861134547519265, "loss": 0.1675, "step": 3484 }, { "epoch": 1.6400000000000001, "grad_norm": 0.6508677862362431, "learning_rate": 0.0002860950281750737, "loss": 0.5871, "step": 3485 }, { "epoch": 1.6404705882352941, "grad_norm": 0.4259943782259599, "learning_rate": 0.00028607658997503786, "loss": 0.5347, "step": 3486 }, { "epoch": 1.6409411764705881, "grad_norm": 0.20711662632038258, "learning_rate": 0.00028605814015339383, "loss": 0.2015, "step": 3487 }, { "epoch": 1.6414117647058823, "grad_norm": 0.16671074040206146, "learning_rate": 0.0002860396787117173, "loss": 0.0854, "step": 3488 }, { "epoch": 1.6418823529411766, "grad_norm": 0.7594272998463903, "learning_rate": 0.00028602120565158486, "loss": 0.5051, "step": 3489 }, { "epoch": 1.6423529411764706, "grad_norm": 0.43706937307378735, "learning_rate": 0.0002860027209745743, "loss": 0.3175, "step": 3490 }, { "epoch": 1.6428235294117646, "grad_norm": 0.37949203726650743, "learning_rate": 0.00028598422468226425, "loss": 0.4864, "step": 3491 }, { "epoch": 1.6432941176470588, "grad_norm": 0.3108025069828386, "learning_rate": 0.0002859657167762345, "loss": 0.3378, "step": 3492 }, { "epoch": 1.643764705882353, "grad_norm": 0.32803435656774493, "learning_rate": 0.00028594719725806556, "loss": 0.2964, "step": 3493 }, { "epoch": 1.644235294117647, "grad_norm": 0.3722707006805881, "learning_rate": 0.0002859286661293392, "loss": 0.2955, "step": 3494 }, { "epoch": 1.644705882352941, "grad_norm": 0.6493069482020511, "learning_rate": 0.00028591012339163806, "loss": 0.3197, "step": 3495 }, { "epoch": 1.6451764705882352, "grad_norm": 0.30068828275617693, "learning_rate": 0.00028589156904654565, "loss": 0.2234, "step": 3496 }, { "epoch": 1.6456470588235295, "grad_norm": 0.18630998495537002, "learning_rate": 0.00028587300309564675, "loss": 0.1066, "step": 3497 }, { "epoch": 1.6461176470588237, "grad_norm": 0.27718436482318365, "learning_rate": 0.0002858544255405269, "loss": 0.1207, "step": 3498 }, { "epoch": 1.6465882352941177, "grad_norm": 0.5246374593240404, "learning_rate": 0.0002858358363827728, "loss": 0.4902, "step": 3499 }, { "epoch": 1.6470588235294117, "grad_norm": 0.3520040567846815, "learning_rate": 0.000285817235623972, "loss": 0.2607, "step": 3500 }, { "epoch": 1.647529411764706, "grad_norm": 0.5764877222770735, "learning_rate": 0.00028579862326571305, "loss": 0.694, "step": 3501 }, { "epoch": 1.6480000000000001, "grad_norm": 0.3288174601577337, "learning_rate": 0.0002857799993095856, "loss": 0.3654, "step": 3502 }, { "epoch": 1.6484705882352941, "grad_norm": 0.3792575513111529, "learning_rate": 0.0002857613637571802, "loss": 0.3591, "step": 3503 }, { "epoch": 1.6489411764705881, "grad_norm": 0.3480456343917658, "learning_rate": 0.00028574271661008845, "loss": 0.3623, "step": 3504 }, { "epoch": 1.6494117647058824, "grad_norm": 0.6362030125162228, "learning_rate": 0.00028572405786990294, "loss": 0.6087, "step": 3505 }, { "epoch": 1.6498823529411766, "grad_norm": 0.4945951791176474, "learning_rate": 0.00028570538753821704, "loss": 0.5004, "step": 3506 }, { "epoch": 1.6503529411764706, "grad_norm": 0.3382698206949804, "learning_rate": 0.0002856867056166255, "loss": 0.4047, "step": 3507 }, { "epoch": 1.6508235294117646, "grad_norm": 0.20719970540854366, "learning_rate": 0.00028566801210672376, "loss": 0.1203, "step": 3508 }, { "epoch": 1.6512941176470588, "grad_norm": 0.4133878258119107, "learning_rate": 0.0002856493070101084, "loss": 0.4084, "step": 3509 }, { "epoch": 1.651764705882353, "grad_norm": 0.28315731131019384, "learning_rate": 0.0002856305903283767, "loss": 0.2892, "step": 3510 }, { "epoch": 1.652235294117647, "grad_norm": 0.7353269375360698, "learning_rate": 0.00028561186206312744, "loss": 0.6736, "step": 3511 }, { "epoch": 1.652705882352941, "grad_norm": 0.4312518864755086, "learning_rate": 0.00028559312221595996, "loss": 0.3509, "step": 3512 }, { "epoch": 1.6531764705882352, "grad_norm": 0.21443463497033932, "learning_rate": 0.00028557437078847477, "loss": 0.2576, "step": 3513 }, { "epoch": 1.6536470588235295, "grad_norm": 0.35922082749318884, "learning_rate": 0.0002855556077822733, "loss": 0.2772, "step": 3514 }, { "epoch": 1.6541176470588237, "grad_norm": 0.299579936768463, "learning_rate": 0.00028553683319895804, "loss": 0.2235, "step": 3515 }, { "epoch": 1.6545882352941177, "grad_norm": 0.28413008260661515, "learning_rate": 0.00028551804704013244, "loss": 0.2018, "step": 3516 }, { "epoch": 1.6550588235294117, "grad_norm": 0.10375701359645059, "learning_rate": 0.0002854992493074009, "loss": 0.0615, "step": 3517 }, { "epoch": 1.655529411764706, "grad_norm": 0.11316553515105536, "learning_rate": 0.00028548044000236886, "loss": 0.0539, "step": 3518 }, { "epoch": 1.6560000000000001, "grad_norm": 0.3102316183639578, "learning_rate": 0.0002854616191266427, "loss": 0.1717, "step": 3519 }, { "epoch": 1.6564705882352941, "grad_norm": 0.729826162279215, "learning_rate": 0.0002854427866818298, "loss": 0.5227, "step": 3520 }, { "epoch": 1.6569411764705881, "grad_norm": 0.24591504690767502, "learning_rate": 0.00028542394266953865, "loss": 0.1683, "step": 3521 }, { "epoch": 1.6574117647058824, "grad_norm": 0.42196995897351824, "learning_rate": 0.00028540508709137846, "loss": 0.4134, "step": 3522 }, { "epoch": 1.6578823529411766, "grad_norm": 0.2824644694200766, "learning_rate": 0.00028538621994895976, "loss": 0.3483, "step": 3523 }, { "epoch": 1.6583529411764706, "grad_norm": 0.2896384283828535, "learning_rate": 0.00028536734124389373, "loss": 0.2593, "step": 3524 }, { "epoch": 1.6588235294117646, "grad_norm": 0.24983427069600342, "learning_rate": 0.0002853484509777928, "loss": 0.2578, "step": 3525 }, { "epoch": 1.6592941176470588, "grad_norm": 0.44292712755313085, "learning_rate": 0.00028532954915227033, "loss": 0.3702, "step": 3526 }, { "epoch": 1.659764705882353, "grad_norm": 0.11507248307975622, "learning_rate": 0.0002853106357689405, "loss": 0.0526, "step": 3527 }, { "epoch": 1.660235294117647, "grad_norm": 0.5268441851410574, "learning_rate": 0.0002852917108294187, "loss": 0.4875, "step": 3528 }, { "epoch": 1.660705882352941, "grad_norm": 0.30554870517646604, "learning_rate": 0.0002852727743353213, "loss": 0.2475, "step": 3529 }, { "epoch": 1.6611764705882353, "grad_norm": 0.5787365523318121, "learning_rate": 0.00028525382628826536, "loss": 0.6414, "step": 3530 }, { "epoch": 1.6616470588235295, "grad_norm": 0.3306787023228835, "learning_rate": 0.0002852348666898692, "loss": 0.3391, "step": 3531 }, { "epoch": 1.6621176470588235, "grad_norm": 0.35501466330873743, "learning_rate": 0.0002852158955417522, "loss": 0.3966, "step": 3532 }, { "epoch": 1.6625882352941177, "grad_norm": 0.29889970636659147, "learning_rate": 0.0002851969128455345, "loss": 0.2796, "step": 3533 }, { "epoch": 1.6630588235294117, "grad_norm": 0.3226486520878693, "learning_rate": 0.0002851779186028372, "loss": 0.4207, "step": 3534 }, { "epoch": 1.663529411764706, "grad_norm": 0.22654956258274853, "learning_rate": 0.0002851589128152827, "loss": 0.1076, "step": 3535 }, { "epoch": 1.6640000000000001, "grad_norm": 0.17895409975994161, "learning_rate": 0.000285139895484494, "loss": 0.1779, "step": 3536 }, { "epoch": 1.6644705882352941, "grad_norm": 0.0930648420179171, "learning_rate": 0.0002851208666120955, "loss": 0.0522, "step": 3537 }, { "epoch": 1.6649411764705881, "grad_norm": 0.24525502144598302, "learning_rate": 0.0002851018261997121, "loss": 0.3694, "step": 3538 }, { "epoch": 1.6654117647058824, "grad_norm": 0.3697506852345238, "learning_rate": 0.00028508277424897015, "loss": 0.362, "step": 3539 }, { "epoch": 1.6658823529411766, "grad_norm": 0.44484673138219527, "learning_rate": 0.00028506371076149665, "loss": 0.4605, "step": 3540 }, { "epoch": 1.6663529411764706, "grad_norm": 0.33070413405001015, "learning_rate": 0.0002850446357389198, "loss": 0.2083, "step": 3541 }, { "epoch": 1.6668235294117646, "grad_norm": 0.5085933374884313, "learning_rate": 0.00028502554918286867, "loss": 0.4466, "step": 3542 }, { "epoch": 1.6672941176470588, "grad_norm": 0.2975662574509702, "learning_rate": 0.0002850064510949733, "loss": 0.3759, "step": 3543 }, { "epoch": 1.667764705882353, "grad_norm": 0.24446362065288754, "learning_rate": 0.00028498734147686485, "loss": 0.3925, "step": 3544 }, { "epoch": 1.668235294117647, "grad_norm": 0.37400054432450525, "learning_rate": 0.0002849682203301753, "loss": 0.3037, "step": 3545 }, { "epoch": 1.668705882352941, "grad_norm": 0.10790906964356203, "learning_rate": 0.0002849490876565378, "loss": 0.0271, "step": 3546 }, { "epoch": 1.6691764705882353, "grad_norm": 0.2794297665144827, "learning_rate": 0.0002849299434575862, "loss": 0.3156, "step": 3547 }, { "epoch": 1.6696470588235295, "grad_norm": 0.3577185533966699, "learning_rate": 0.00028491078773495564, "loss": 0.2996, "step": 3548 }, { "epoch": 1.6701176470588235, "grad_norm": 0.38107158300367683, "learning_rate": 0.00028489162049028205, "loss": 0.409, "step": 3549 }, { "epoch": 1.6705882352941175, "grad_norm": 0.4699686401089007, "learning_rate": 0.00028487244172520246, "loss": 0.5009, "step": 3550 }, { "epoch": 1.6710588235294117, "grad_norm": 0.09079981435621494, "learning_rate": 0.0002848532514413548, "loss": 0.0442, "step": 3551 }, { "epoch": 1.671529411764706, "grad_norm": 0.19811383385988943, "learning_rate": 0.00028483404964037794, "loss": 0.1497, "step": 3552 }, { "epoch": 1.6720000000000002, "grad_norm": 0.48106385748670083, "learning_rate": 0.000284814836323912, "loss": 0.4515, "step": 3553 }, { "epoch": 1.6724705882352942, "grad_norm": 0.1969995129179624, "learning_rate": 0.0002847956114935977, "loss": 0.1511, "step": 3554 }, { "epoch": 1.6729411764705882, "grad_norm": 0.38750471859545443, "learning_rate": 0.000284776375151077, "loss": 0.3549, "step": 3555 }, { "epoch": 1.6734117647058824, "grad_norm": 0.5804924314781419, "learning_rate": 0.00028475712729799285, "loss": 0.52, "step": 3556 }, { "epoch": 1.6738823529411766, "grad_norm": 0.33853404641755747, "learning_rate": 0.000284737867935989, "loss": 0.3504, "step": 3557 }, { "epoch": 1.6743529411764706, "grad_norm": 0.2902437992025677, "learning_rate": 0.0002847185970667104, "loss": 0.3199, "step": 3558 }, { "epoch": 1.6748235294117646, "grad_norm": 0.32433240337704566, "learning_rate": 0.00028469931469180275, "loss": 0.2099, "step": 3559 }, { "epoch": 1.6752941176470588, "grad_norm": 0.8005015670195684, "learning_rate": 0.00028468002081291296, "loss": 0.4911, "step": 3560 }, { "epoch": 1.675764705882353, "grad_norm": 0.6081783597266277, "learning_rate": 0.0002846607154316888, "loss": 0.3785, "step": 3561 }, { "epoch": 1.676235294117647, "grad_norm": 0.334498690255943, "learning_rate": 0.000284641398549779, "loss": 0.3412, "step": 3562 }, { "epoch": 1.676705882352941, "grad_norm": 0.7197416368395658, "learning_rate": 0.00028462207016883336, "loss": 0.7397, "step": 3563 }, { "epoch": 1.6771764705882353, "grad_norm": 0.46079205203432266, "learning_rate": 0.0002846027302905026, "loss": 0.4016, "step": 3564 }, { "epoch": 1.6776470588235295, "grad_norm": 0.464215485449206, "learning_rate": 0.0002845833789164385, "loss": 0.5981, "step": 3565 }, { "epoch": 1.6781176470588235, "grad_norm": 0.3705205725902509, "learning_rate": 0.00028456401604829367, "loss": 0.258, "step": 3566 }, { "epoch": 1.6785882352941175, "grad_norm": 0.16241542807971032, "learning_rate": 0.00028454464168772186, "loss": 0.0915, "step": 3567 }, { "epoch": 1.6790588235294117, "grad_norm": 0.3508265162999664, "learning_rate": 0.00028452525583637774, "loss": 0.2524, "step": 3568 }, { "epoch": 1.679529411764706, "grad_norm": 0.33175798389470007, "learning_rate": 0.0002845058584959169, "loss": 0.3104, "step": 3569 }, { "epoch": 1.6800000000000002, "grad_norm": 0.2782843144628319, "learning_rate": 0.00028448644966799595, "loss": 0.1693, "step": 3570 }, { "epoch": 1.6804705882352942, "grad_norm": 0.4052488453530742, "learning_rate": 0.0002844670293542726, "loss": 0.4065, "step": 3571 }, { "epoch": 1.6809411764705882, "grad_norm": 0.1879271830303297, "learning_rate": 0.00028444759755640535, "loss": 0.1544, "step": 3572 }, { "epoch": 1.6814117647058824, "grad_norm": 0.13901352033622458, "learning_rate": 0.0002844281542760539, "loss": 0.0945, "step": 3573 }, { "epoch": 1.6818823529411766, "grad_norm": 0.21002143858256062, "learning_rate": 0.0002844086995148786, "loss": 0.1345, "step": 3574 }, { "epoch": 1.6823529411764706, "grad_norm": 0.1338090008715266, "learning_rate": 0.00028438923327454114, "loss": 0.0944, "step": 3575 }, { "epoch": 1.6828235294117646, "grad_norm": 0.15031884455848663, "learning_rate": 0.000284369755556704, "loss": 0.0971, "step": 3576 }, { "epoch": 1.6832941176470588, "grad_norm": 0.4753557685620345, "learning_rate": 0.0002843502663630306, "loss": 0.4278, "step": 3577 }, { "epoch": 1.683764705882353, "grad_norm": 0.3568602948318368, "learning_rate": 0.0002843307656951855, "loss": 0.395, "step": 3578 }, { "epoch": 1.684235294117647, "grad_norm": 0.30996070570147577, "learning_rate": 0.0002843112535548341, "loss": 0.3295, "step": 3579 }, { "epoch": 1.684705882352941, "grad_norm": 0.3726620147502095, "learning_rate": 0.0002842917299436429, "loss": 0.5578, "step": 3580 }, { "epoch": 1.6851764705882353, "grad_norm": 0.4369410601262364, "learning_rate": 0.00028427219486327923, "loss": 0.5999, "step": 3581 }, { "epoch": 1.6856470588235295, "grad_norm": 0.4316874412921562, "learning_rate": 0.00028425264831541153, "loss": 0.3199, "step": 3582 }, { "epoch": 1.6861176470588235, "grad_norm": 0.15692481351697182, "learning_rate": 0.0002842330903017091, "loss": 0.0972, "step": 3583 }, { "epoch": 1.6865882352941175, "grad_norm": 0.3587476926001552, "learning_rate": 0.00028421352082384245, "loss": 0.355, "step": 3584 }, { "epoch": 1.6870588235294117, "grad_norm": 0.18363341647664252, "learning_rate": 0.00028419393988348275, "loss": 0.1678, "step": 3585 }, { "epoch": 1.687529411764706, "grad_norm": 0.5182601174805327, "learning_rate": 0.00028417434748230237, "loss": 0.3843, "step": 3586 }, { "epoch": 1.688, "grad_norm": 0.3931037810778853, "learning_rate": 0.0002841547436219746, "loss": 0.3751, "step": 3587 }, { "epoch": 1.6884705882352942, "grad_norm": 0.07693249879517094, "learning_rate": 0.0002841351283041737, "loss": 0.0413, "step": 3588 }, { "epoch": 1.6889411764705882, "grad_norm": 0.1455494146007021, "learning_rate": 0.0002841155015305749, "loss": 0.0746, "step": 3589 }, { "epoch": 1.6894117647058824, "grad_norm": 0.2844386316836661, "learning_rate": 0.00028409586330285444, "loss": 0.2403, "step": 3590 }, { "epoch": 1.6898823529411766, "grad_norm": 0.3937311908439463, "learning_rate": 0.00028407621362268955, "loss": 0.4684, "step": 3591 }, { "epoch": 1.6903529411764706, "grad_norm": 0.45328926371590555, "learning_rate": 0.0002840565524917583, "loss": 0.4623, "step": 3592 }, { "epoch": 1.6908235294117646, "grad_norm": 0.30375538846811834, "learning_rate": 0.00028403687991174003, "loss": 0.3393, "step": 3593 }, { "epoch": 1.6912941176470588, "grad_norm": 0.4258394086187198, "learning_rate": 0.0002840171958843147, "loss": 0.3687, "step": 3594 }, { "epoch": 1.691764705882353, "grad_norm": 0.29915460702190294, "learning_rate": 0.0002839975004111635, "loss": 0.3715, "step": 3595 }, { "epoch": 1.692235294117647, "grad_norm": 0.32803559475045263, "learning_rate": 0.00028397779349396854, "loss": 0.3502, "step": 3596 }, { "epoch": 1.692705882352941, "grad_norm": 0.3407099366025131, "learning_rate": 0.0002839580751344128, "loss": 0.4069, "step": 3597 }, { "epoch": 1.6931764705882353, "grad_norm": 0.3704611011187358, "learning_rate": 0.00028393834533418043, "loss": 0.5985, "step": 3598 }, { "epoch": 1.6936470588235295, "grad_norm": 0.2600559615216309, "learning_rate": 0.0002839186040949564, "loss": 0.2102, "step": 3599 }, { "epoch": 1.6941176470588235, "grad_norm": 0.3167691752874059, "learning_rate": 0.0002838988514184267, "loss": 0.3215, "step": 3600 }, { "epoch": 1.6945882352941175, "grad_norm": 0.328891447198927, "learning_rate": 0.0002838790873062784, "loss": 0.3877, "step": 3601 }, { "epoch": 1.6950588235294117, "grad_norm": 0.3120341683539293, "learning_rate": 0.0002838593117601992, "loss": 0.3865, "step": 3602 }, { "epoch": 1.695529411764706, "grad_norm": 0.3390045725581036, "learning_rate": 0.0002838395247818783, "loss": 0.2494, "step": 3603 }, { "epoch": 1.696, "grad_norm": 0.27408732283110576, "learning_rate": 0.0002838197263730055, "loss": 0.2356, "step": 3604 }, { "epoch": 1.696470588235294, "grad_norm": 0.35100690383227845, "learning_rate": 0.0002837999165352717, "loss": 0.3432, "step": 3605 }, { "epoch": 1.6969411764705882, "grad_norm": 0.5340200660745333, "learning_rate": 0.0002837800952703687, "loss": 0.364, "step": 3606 }, { "epoch": 1.6974117647058824, "grad_norm": 0.15950997822107063, "learning_rate": 0.00028376026257998947, "loss": 0.1065, "step": 3607 }, { "epoch": 1.6978823529411766, "grad_norm": 0.13233121352640148, "learning_rate": 0.0002837404184658276, "loss": 0.0491, "step": 3608 }, { "epoch": 1.6983529411764706, "grad_norm": 0.3279830880623107, "learning_rate": 0.00028372056292957806, "loss": 0.2686, "step": 3609 }, { "epoch": 1.6988235294117646, "grad_norm": 0.29262183090866956, "learning_rate": 0.0002837006959729366, "loss": 0.1966, "step": 3610 }, { "epoch": 1.6992941176470588, "grad_norm": 0.5373959782244646, "learning_rate": 0.00028368081759759983, "loss": 0.551, "step": 3611 }, { "epoch": 1.699764705882353, "grad_norm": 0.3136241502843492, "learning_rate": 0.00028366092780526554, "loss": 0.262, "step": 3612 }, { "epoch": 1.700235294117647, "grad_norm": 0.1622646713120229, "learning_rate": 0.00028364102659763247, "loss": 0.0767, "step": 3613 }, { "epoch": 1.700705882352941, "grad_norm": 0.35980715462575874, "learning_rate": 0.00028362111397640016, "loss": 0.336, "step": 3614 }, { "epoch": 1.7011764705882353, "grad_norm": 0.209444511596028, "learning_rate": 0.00028360118994326936, "loss": 0.2045, "step": 3615 }, { "epoch": 1.7016470588235295, "grad_norm": 0.2252528858873125, "learning_rate": 0.00028358125449994156, "loss": 0.1827, "step": 3616 }, { "epoch": 1.7021176470588235, "grad_norm": 0.5252694027671281, "learning_rate": 0.0002835613076481195, "loss": 0.4237, "step": 3617 }, { "epoch": 1.7025882352941175, "grad_norm": 0.3704043655521189, "learning_rate": 0.0002835413493895066, "loss": 0.3076, "step": 3618 }, { "epoch": 1.7030588235294117, "grad_norm": 0.27123764657556365, "learning_rate": 0.00028352137972580745, "loss": 0.2361, "step": 3619 }, { "epoch": 1.703529411764706, "grad_norm": 0.3842404613563614, "learning_rate": 0.0002835013986587276, "loss": 0.3848, "step": 3620 }, { "epoch": 1.704, "grad_norm": 0.26971759599592504, "learning_rate": 0.0002834814061899734, "loss": 0.4074, "step": 3621 }, { "epoch": 1.704470588235294, "grad_norm": 0.31705980240867687, "learning_rate": 0.0002834614023212525, "loss": 0.371, "step": 3622 }, { "epoch": 1.7049411764705882, "grad_norm": 0.12830045591872802, "learning_rate": 0.0002834413870542731, "loss": 0.1122, "step": 3623 }, { "epoch": 1.7054117647058824, "grad_norm": 0.20936532933987226, "learning_rate": 0.00028342136039074476, "loss": 0.1564, "step": 3624 }, { "epoch": 1.7058823529411766, "grad_norm": 0.09581925694706696, "learning_rate": 0.0002834013223323778, "loss": 0.0448, "step": 3625 }, { "epoch": 1.7063529411764706, "grad_norm": 0.17500999261640962, "learning_rate": 0.0002833812728808836, "loss": 0.1288, "step": 3626 }, { "epoch": 1.7068235294117646, "grad_norm": 0.045486040065353214, "learning_rate": 0.00028336121203797446, "loss": 0.0165, "step": 3627 }, { "epoch": 1.7072941176470589, "grad_norm": 0.14026004199955758, "learning_rate": 0.0002833411398053637, "loss": 0.0796, "step": 3628 }, { "epoch": 1.707764705882353, "grad_norm": 0.3049691228770123, "learning_rate": 0.00028332105618476556, "loss": 0.3116, "step": 3629 }, { "epoch": 1.708235294117647, "grad_norm": 0.3041865879353827, "learning_rate": 0.0002833009611778953, "loss": 0.1981, "step": 3630 }, { "epoch": 1.708705882352941, "grad_norm": 0.4018270236733309, "learning_rate": 0.0002832808547864691, "loss": 0.3819, "step": 3631 }, { "epoch": 1.7091764705882353, "grad_norm": 0.38440740194885786, "learning_rate": 0.00028326073701220416, "loss": 0.3962, "step": 3632 }, { "epoch": 1.7096470588235295, "grad_norm": 0.21427712388039563, "learning_rate": 0.0002832406078568186, "loss": 0.241, "step": 3633 }, { "epoch": 1.7101176470588235, "grad_norm": 0.3552812716319359, "learning_rate": 0.00028322046732203165, "loss": 0.4933, "step": 3634 }, { "epoch": 1.7105882352941175, "grad_norm": 0.2455085534823582, "learning_rate": 0.0002832003154095633, "loss": 0.2029, "step": 3635 }, { "epoch": 1.7110588235294117, "grad_norm": 0.3649233451723786, "learning_rate": 0.0002831801521211347, "loss": 0.3573, "step": 3636 }, { "epoch": 1.711529411764706, "grad_norm": 0.27527108283715807, "learning_rate": 0.0002831599774584679, "loss": 0.2373, "step": 3637 }, { "epoch": 1.712, "grad_norm": 0.2539165531916147, "learning_rate": 0.0002831397914232858, "loss": 0.1148, "step": 3638 }, { "epoch": 1.712470588235294, "grad_norm": 0.2760656656415857, "learning_rate": 0.0002831195940173125, "loss": 0.2394, "step": 3639 }, { "epoch": 1.7129411764705882, "grad_norm": 0.26046342179998244, "learning_rate": 0.00028309938524227296, "loss": 0.2751, "step": 3640 }, { "epoch": 1.7134117647058824, "grad_norm": 0.26881739690793344, "learning_rate": 0.000283079165099893, "loss": 0.1449, "step": 3641 }, { "epoch": 1.7138823529411766, "grad_norm": 0.21497932385968763, "learning_rate": 0.0002830589335918997, "loss": 0.1398, "step": 3642 }, { "epoch": 1.7143529411764706, "grad_norm": 0.18831046472289303, "learning_rate": 0.0002830386907200207, "loss": 0.1315, "step": 3643 }, { "epoch": 1.7148235294117646, "grad_norm": 0.44647998151064455, "learning_rate": 0.000283018436485985, "loss": 0.4523, "step": 3644 }, { "epoch": 1.7152941176470589, "grad_norm": 0.19892296574182333, "learning_rate": 0.0002829981708915224, "loss": 0.1223, "step": 3645 }, { "epoch": 1.715764705882353, "grad_norm": 0.1754770569415445, "learning_rate": 0.0002829778939383636, "loss": 0.1415, "step": 3646 }, { "epoch": 1.716235294117647, "grad_norm": 0.2662320790566374, "learning_rate": 0.0002829576056282405, "loss": 0.2066, "step": 3647 }, { "epoch": 1.716705882352941, "grad_norm": 0.3862551483952254, "learning_rate": 0.00028293730596288565, "loss": 0.338, "step": 3648 }, { "epoch": 1.7171764705882353, "grad_norm": 0.3606726367695168, "learning_rate": 0.00028291699494403284, "loss": 0.2904, "step": 3649 }, { "epoch": 1.7176470588235295, "grad_norm": 0.22961257158517434, "learning_rate": 0.0002828966725734167, "loss": 0.1681, "step": 3650 }, { "epoch": 1.7181176470588235, "grad_norm": 0.292575644358598, "learning_rate": 0.0002828763388527729, "loss": 0.2552, "step": 3651 }, { "epoch": 1.7185882352941175, "grad_norm": 0.32769358972023294, "learning_rate": 0.000282855993783838, "loss": 0.2631, "step": 3652 }, { "epoch": 1.7190588235294118, "grad_norm": 0.3873139142129574, "learning_rate": 0.0002828356373683495, "loss": 0.3713, "step": 3653 }, { "epoch": 1.719529411764706, "grad_norm": 0.439664298536553, "learning_rate": 0.0002828152696080461, "loss": 0.33, "step": 3654 }, { "epoch": 1.72, "grad_norm": 0.13477602248670975, "learning_rate": 0.0002827948905046672, "loss": 0.1087, "step": 3655 }, { "epoch": 1.720470588235294, "grad_norm": 0.3572150991262217, "learning_rate": 0.00028277450005995326, "loss": 0.3737, "step": 3656 }, { "epoch": 1.7209411764705882, "grad_norm": 0.18608413261636003, "learning_rate": 0.0002827540982756458, "loss": 0.1898, "step": 3657 }, { "epoch": 1.7214117647058824, "grad_norm": 0.33652848858589585, "learning_rate": 0.00028273368515348713, "loss": 0.3536, "step": 3658 }, { "epoch": 1.7218823529411764, "grad_norm": 0.33016652432485744, "learning_rate": 0.00028271326069522076, "loss": 0.4065, "step": 3659 }, { "epoch": 1.7223529411764706, "grad_norm": 0.3295380932230287, "learning_rate": 0.0002826928249025909, "loss": 0.334, "step": 3660 }, { "epoch": 1.7228235294117646, "grad_norm": 0.19622136054687156, "learning_rate": 0.000282672377777343, "loss": 0.1272, "step": 3661 }, { "epoch": 1.7232941176470589, "grad_norm": 0.33070826204620685, "learning_rate": 0.0002826519193212233, "loss": 0.2816, "step": 3662 }, { "epoch": 1.723764705882353, "grad_norm": 0.15795849251938598, "learning_rate": 0.000282631449535979, "loss": 0.0947, "step": 3663 }, { "epoch": 1.724235294117647, "grad_norm": 0.3948770731909715, "learning_rate": 0.00028261096842335836, "loss": 0.4629, "step": 3664 }, { "epoch": 1.724705882352941, "grad_norm": 0.29312828240633604, "learning_rate": 0.00028259047598511053, "loss": 0.2534, "step": 3665 }, { "epoch": 1.7251764705882353, "grad_norm": 0.14944222693845227, "learning_rate": 0.0002825699722229858, "loss": 0.065, "step": 3666 }, { "epoch": 1.7256470588235295, "grad_norm": 0.35986639460372294, "learning_rate": 0.0002825494571387351, "loss": 0.3937, "step": 3667 }, { "epoch": 1.7261176470588235, "grad_norm": 0.46687622018539493, "learning_rate": 0.00028252893073411065, "loss": 0.3576, "step": 3668 }, { "epoch": 1.7265882352941175, "grad_norm": 0.30685788915778206, "learning_rate": 0.00028250839301086544, "loss": 0.3821, "step": 3669 }, { "epoch": 1.7270588235294118, "grad_norm": 0.17891412977892404, "learning_rate": 0.0002824878439707535, "loss": 0.1406, "step": 3670 }, { "epoch": 1.727529411764706, "grad_norm": 0.20454543206757922, "learning_rate": 0.0002824672836155299, "loss": 0.2466, "step": 3671 }, { "epoch": 1.728, "grad_norm": 0.34700284072616433, "learning_rate": 0.0002824467119469505, "loss": 0.3234, "step": 3672 }, { "epoch": 1.728470588235294, "grad_norm": 0.41284638719050437, "learning_rate": 0.0002824261289667722, "loss": 0.3069, "step": 3673 }, { "epoch": 1.7289411764705882, "grad_norm": 0.3239373138068571, "learning_rate": 0.000282405534676753, "loss": 0.5202, "step": 3674 }, { "epoch": 1.7294117647058824, "grad_norm": 0.2692965352707292, "learning_rate": 0.0002823849290786517, "loss": 0.2505, "step": 3675 }, { "epoch": 1.7298823529411764, "grad_norm": 0.24241113659401292, "learning_rate": 0.0002823643121742281, "loss": 0.2306, "step": 3676 }, { "epoch": 1.7303529411764704, "grad_norm": 0.24233775420370565, "learning_rate": 0.000282343683965243, "loss": 0.161, "step": 3677 }, { "epoch": 1.7308235294117646, "grad_norm": 0.16811134970782723, "learning_rate": 0.0002823230444534581, "loss": 0.1568, "step": 3678 }, { "epoch": 1.7312941176470589, "grad_norm": 0.06023949678696225, "learning_rate": 0.0002823023936406362, "loss": 0.0366, "step": 3679 }, { "epoch": 1.731764705882353, "grad_norm": 0.49157723533873887, "learning_rate": 0.0002822817315285409, "loss": 0.5627, "step": 3680 }, { "epoch": 1.732235294117647, "grad_norm": 0.37417726317462663, "learning_rate": 0.000282261058118937, "loss": 0.3119, "step": 3681 }, { "epoch": 1.732705882352941, "grad_norm": 0.19919005147488225, "learning_rate": 0.00028224037341358993, "loss": 0.1165, "step": 3682 }, { "epoch": 1.7331764705882353, "grad_norm": 0.3352859166035637, "learning_rate": 0.0002822196774142663, "loss": 0.3772, "step": 3683 }, { "epoch": 1.7336470588235295, "grad_norm": 0.36100358843882074, "learning_rate": 0.0002821989701227337, "loss": 0.5169, "step": 3684 }, { "epoch": 1.7341176470588235, "grad_norm": 0.2881858192446873, "learning_rate": 0.00028217825154076067, "loss": 0.2862, "step": 3685 }, { "epoch": 1.7345882352941175, "grad_norm": 0.21858245957310138, "learning_rate": 0.0002821575216701166, "loss": 0.1414, "step": 3686 }, { "epoch": 1.7350588235294118, "grad_norm": 0.13377501688076912, "learning_rate": 0.00028213678051257194, "loss": 0.0631, "step": 3687 }, { "epoch": 1.735529411764706, "grad_norm": 0.6395273749473326, "learning_rate": 0.00028211602806989814, "loss": 0.3994, "step": 3688 }, { "epoch": 1.736, "grad_norm": 0.3157073268407831, "learning_rate": 0.00028209526434386744, "loss": 0.4624, "step": 3689 }, { "epoch": 1.736470588235294, "grad_norm": 0.3397106561367648, "learning_rate": 0.0002820744893362533, "loss": 0.3183, "step": 3690 }, { "epoch": 1.7369411764705882, "grad_norm": 0.3130143597073577, "learning_rate": 0.0002820537030488299, "loss": 0.3604, "step": 3691 }, { "epoch": 1.7374117647058824, "grad_norm": 0.2247011573739991, "learning_rate": 0.00028203290548337257, "loss": 0.1714, "step": 3692 }, { "epoch": 1.7378823529411764, "grad_norm": 0.3161102566904776, "learning_rate": 0.0002820120966416575, "loss": 0.1103, "step": 3693 }, { "epoch": 1.7383529411764704, "grad_norm": 0.3139206782992897, "learning_rate": 0.0002819912765254619, "loss": 0.1578, "step": 3694 }, { "epoch": 1.7388235294117647, "grad_norm": 0.2559003588279978, "learning_rate": 0.0002819704451365638, "loss": 0.2955, "step": 3695 }, { "epoch": 1.7392941176470589, "grad_norm": 0.296192130972663, "learning_rate": 0.0002819496024767424, "loss": 0.2399, "step": 3696 }, { "epoch": 1.739764705882353, "grad_norm": 0.3760448604644524, "learning_rate": 0.0002819287485477778, "loss": 0.3415, "step": 3697 }, { "epoch": 1.740235294117647, "grad_norm": 0.3724784409745347, "learning_rate": 0.0002819078833514509, "loss": 0.2259, "step": 3698 }, { "epoch": 1.740705882352941, "grad_norm": 0.19250485774372328, "learning_rate": 0.0002818870068895437, "loss": 0.1393, "step": 3699 }, { "epoch": 1.7411764705882353, "grad_norm": 0.3860331653991044, "learning_rate": 0.0002818661191638393, "loss": 0.4773, "step": 3700 }, { "epoch": 1.7416470588235295, "grad_norm": 0.47287865687213787, "learning_rate": 0.0002818452201761215, "loss": 0.5363, "step": 3701 }, { "epoch": 1.7421176470588235, "grad_norm": 1.2515986539248587, "learning_rate": 0.0002818243099281752, "loss": 0.5239, "step": 3702 }, { "epoch": 1.7425882352941175, "grad_norm": 0.15716134928181277, "learning_rate": 0.00028180338842178623, "loss": 0.1743, "step": 3703 }, { "epoch": 1.7430588235294118, "grad_norm": 0.35468099077721277, "learning_rate": 0.0002817824556587414, "loss": 0.4152, "step": 3704 }, { "epoch": 1.743529411764706, "grad_norm": 0.5996592556739591, "learning_rate": 0.0002817615116408285, "loss": 0.3811, "step": 3705 }, { "epoch": 1.744, "grad_norm": 0.27541401915746677, "learning_rate": 0.0002817405563698362, "loss": 0.2634, "step": 3706 }, { "epoch": 1.744470588235294, "grad_norm": 0.2908116966328846, "learning_rate": 0.0002817195898475542, "loss": 0.3486, "step": 3707 }, { "epoch": 1.7449411764705882, "grad_norm": 0.2908385989902933, "learning_rate": 0.00028169861207577313, "loss": 0.3555, "step": 3708 }, { "epoch": 1.7454117647058824, "grad_norm": 0.17202615322981146, "learning_rate": 0.0002816776230562846, "loss": 0.1018, "step": 3709 }, { "epoch": 1.7458823529411764, "grad_norm": 0.316855556974059, "learning_rate": 0.0002816566227908812, "loss": 0.2005, "step": 3710 }, { "epoch": 1.7463529411764704, "grad_norm": 0.7836565563428363, "learning_rate": 0.00028163561128135645, "loss": 0.6455, "step": 3711 }, { "epoch": 1.7468235294117647, "grad_norm": 0.4020545619248466, "learning_rate": 0.0002816145885295048, "loss": 0.4345, "step": 3712 }, { "epoch": 1.7472941176470589, "grad_norm": 0.2758290367920878, "learning_rate": 0.0002815935545371218, "loss": 0.2502, "step": 3713 }, { "epoch": 1.7477647058823529, "grad_norm": 0.2205344123652232, "learning_rate": 0.00028157250930600374, "loss": 0.2389, "step": 3714 }, { "epoch": 1.748235294117647, "grad_norm": 0.43663739817993985, "learning_rate": 0.00028155145283794797, "loss": 0.2978, "step": 3715 }, { "epoch": 1.748705882352941, "grad_norm": 0.5331618498581967, "learning_rate": 0.0002815303851347529, "loss": 0.1846, "step": 3716 }, { "epoch": 1.7491764705882353, "grad_norm": 0.2966736597060446, "learning_rate": 0.0002815093061982178, "loss": 0.2766, "step": 3717 }, { "epoch": 1.7496470588235296, "grad_norm": 0.6065933930289444, "learning_rate": 0.0002814882160301429, "loss": 0.6945, "step": 3718 }, { "epoch": 1.7501176470588236, "grad_norm": 0.34860575897391205, "learning_rate": 0.00028146711463232943, "loss": 0.3914, "step": 3719 }, { "epoch": 1.7505882352941176, "grad_norm": 0.1978594718971387, "learning_rate": 0.0002814460020065795, "loss": 0.1034, "step": 3720 }, { "epoch": 1.7510588235294118, "grad_norm": 0.30547589577446654, "learning_rate": 0.00028142487815469627, "loss": 0.1961, "step": 3721 }, { "epoch": 1.751529411764706, "grad_norm": 1.06449066937447, "learning_rate": 0.00028140374307848384, "loss": 0.7933, "step": 3722 }, { "epoch": 1.752, "grad_norm": 0.3353684646263928, "learning_rate": 0.00028138259677974714, "loss": 0.2979, "step": 3723 }, { "epoch": 1.752470588235294, "grad_norm": 0.23436972611805737, "learning_rate": 0.0002813614392602923, "loss": 0.3032, "step": 3724 }, { "epoch": 1.7529411764705882, "grad_norm": 0.18096581865217273, "learning_rate": 0.00028134027052192624, "loss": 0.1488, "step": 3725 }, { "epoch": 1.7534117647058824, "grad_norm": 0.46786711885857024, "learning_rate": 0.0002813190905664568, "loss": 0.4349, "step": 3726 }, { "epoch": 1.7538823529411764, "grad_norm": 0.21468260358178587, "learning_rate": 0.0002812978993956929, "loss": 0.1166, "step": 3727 }, { "epoch": 1.7543529411764704, "grad_norm": 0.3380048882435857, "learning_rate": 0.0002812766970114445, "loss": 0.211, "step": 3728 }, { "epoch": 1.7548235294117647, "grad_norm": 0.39435475173870704, "learning_rate": 0.00028125548341552213, "loss": 0.3733, "step": 3729 }, { "epoch": 1.755294117647059, "grad_norm": 0.25411809024201626, "learning_rate": 0.00028123425860973773, "loss": 0.2343, "step": 3730 }, { "epoch": 1.755764705882353, "grad_norm": 0.2715905162728156, "learning_rate": 0.000281213022595904, "loss": 0.3094, "step": 3731 }, { "epoch": 1.756235294117647, "grad_norm": 0.7013102096601534, "learning_rate": 0.0002811917753758345, "loss": 0.7826, "step": 3732 }, { "epoch": 1.7567058823529411, "grad_norm": 0.3006081210619105, "learning_rate": 0.00028117051695134384, "loss": 0.4143, "step": 3733 }, { "epoch": 1.7571764705882353, "grad_norm": 0.329483195829387, "learning_rate": 0.0002811492473242477, "loss": 0.2464, "step": 3734 }, { "epoch": 1.7576470588235296, "grad_norm": 0.30472752871831255, "learning_rate": 0.00028112796649636254, "loss": 0.4019, "step": 3735 }, { "epoch": 1.7581176470588236, "grad_norm": 0.23403700826657275, "learning_rate": 0.0002811066744695059, "loss": 0.0908, "step": 3736 }, { "epoch": 1.7585882352941176, "grad_norm": 0.4366918549048426, "learning_rate": 0.0002810853712454961, "loss": 0.353, "step": 3737 }, { "epoch": 1.7590588235294118, "grad_norm": 0.20802742775286226, "learning_rate": 0.00028106405682615265, "loss": 0.2318, "step": 3738 }, { "epoch": 1.759529411764706, "grad_norm": 0.35812735353434233, "learning_rate": 0.0002810427312132959, "loss": 0.3915, "step": 3739 }, { "epoch": 1.76, "grad_norm": 0.3388574197442859, "learning_rate": 0.00028102139440874715, "loss": 0.3163, "step": 3740 }, { "epoch": 1.760470588235294, "grad_norm": 0.27917904951324274, "learning_rate": 0.0002810000464143287, "loss": 0.2936, "step": 3741 }, { "epoch": 1.7609411764705882, "grad_norm": 0.4341630699867281, "learning_rate": 0.0002809786872318637, "loss": 0.2836, "step": 3742 }, { "epoch": 1.7614117647058825, "grad_norm": 0.3090986205858682, "learning_rate": 0.00028095731686317635, "loss": 0.3103, "step": 3743 }, { "epoch": 1.7618823529411765, "grad_norm": 0.13632856733475304, "learning_rate": 0.0002809359353100918, "loss": 0.0773, "step": 3744 }, { "epoch": 1.7623529411764705, "grad_norm": 0.1357585789045301, "learning_rate": 0.0002809145425744361, "loss": 0.1027, "step": 3745 }, { "epoch": 1.7628235294117647, "grad_norm": 0.4472737321551962, "learning_rate": 0.00028089313865803644, "loss": 0.4394, "step": 3746 }, { "epoch": 1.763294117647059, "grad_norm": 0.31914974922282746, "learning_rate": 0.0002808717235627206, "loss": 0.4152, "step": 3747 }, { "epoch": 1.763764705882353, "grad_norm": 0.418613852405498, "learning_rate": 0.0002808502972903177, "loss": 0.4121, "step": 3748 }, { "epoch": 1.764235294117647, "grad_norm": 0.5030257923847616, "learning_rate": 0.0002808288598426576, "loss": 0.3446, "step": 3749 }, { "epoch": 1.7647058823529411, "grad_norm": 0.21626863493302068, "learning_rate": 0.0002808074112215711, "loss": 0.1997, "step": 3750 }, { "epoch": 1.7651764705882353, "grad_norm": 0.42348840263504134, "learning_rate": 0.00028078595142889013, "loss": 0.5459, "step": 3751 }, { "epoch": 1.7656470588235296, "grad_norm": 0.2137242809047112, "learning_rate": 0.0002807644804664474, "loss": 0.2162, "step": 3752 }, { "epoch": 1.7661176470588236, "grad_norm": 0.2613561202488436, "learning_rate": 0.0002807429983360766, "loss": 0.2428, "step": 3753 }, { "epoch": 1.7665882352941176, "grad_norm": 0.2586820228835942, "learning_rate": 0.0002807215050396125, "loss": 0.2434, "step": 3754 }, { "epoch": 1.7670588235294118, "grad_norm": 0.3428015031209654, "learning_rate": 0.0002807000005788907, "loss": 0.3215, "step": 3755 }, { "epoch": 1.767529411764706, "grad_norm": 0.4862931322677128, "learning_rate": 0.00028067848495574773, "loss": 0.5817, "step": 3756 }, { "epoch": 1.768, "grad_norm": 0.3650664115463596, "learning_rate": 0.0002806569581720212, "loss": 0.4374, "step": 3757 }, { "epoch": 1.768470588235294, "grad_norm": 0.49668680918294544, "learning_rate": 0.00028063542022954956, "loss": 0.5456, "step": 3758 }, { "epoch": 1.7689411764705882, "grad_norm": 0.36561335650604204, "learning_rate": 0.00028061387113017226, "loss": 0.2705, "step": 3759 }, { "epoch": 1.7694117647058825, "grad_norm": 0.35372906341517074, "learning_rate": 0.0002805923108757297, "loss": 0.3424, "step": 3760 }, { "epoch": 1.7698823529411765, "grad_norm": 0.14798422797801852, "learning_rate": 0.00028057073946806323, "loss": 0.1088, "step": 3761 }, { "epoch": 1.7703529411764705, "grad_norm": 0.3608701721493177, "learning_rate": 0.00028054915690901524, "loss": 0.2411, "step": 3762 }, { "epoch": 1.7708235294117647, "grad_norm": 0.13788663274970742, "learning_rate": 0.0002805275632004288, "loss": 0.0803, "step": 3763 }, { "epoch": 1.771294117647059, "grad_norm": 0.2642520424594155, "learning_rate": 0.00028050595834414835, "loss": 0.2779, "step": 3764 }, { "epoch": 1.771764705882353, "grad_norm": 0.18463066647113194, "learning_rate": 0.00028048434234201883, "loss": 0.0862, "step": 3765 }, { "epoch": 1.772235294117647, "grad_norm": 0.14623472876656357, "learning_rate": 0.00028046271519588644, "loss": 0.1038, "step": 3766 }, { "epoch": 1.7727058823529411, "grad_norm": 0.3779445971257875, "learning_rate": 0.0002804410769075984, "loss": 0.4083, "step": 3767 }, { "epoch": 1.7731764705882354, "grad_norm": 0.1552542980012048, "learning_rate": 0.0002804194274790024, "loss": 0.0701, "step": 3768 }, { "epoch": 1.7736470588235296, "grad_norm": 0.26010612542541417, "learning_rate": 0.00028039776691194773, "loss": 0.2568, "step": 3769 }, { "epoch": 1.7741176470588236, "grad_norm": 0.3010395483998841, "learning_rate": 0.0002803760952082841, "loss": 0.2282, "step": 3770 }, { "epoch": 1.7745882352941176, "grad_norm": 0.2358713137018307, "learning_rate": 0.0002803544123698624, "loss": 0.3519, "step": 3771 }, { "epoch": 1.7750588235294118, "grad_norm": 0.27078459540163147, "learning_rate": 0.0002803327183985346, "loss": 0.1693, "step": 3772 }, { "epoch": 1.775529411764706, "grad_norm": 0.26588179147632, "learning_rate": 0.00028031101329615334, "loss": 0.1535, "step": 3773 }, { "epoch": 1.776, "grad_norm": 0.3194364545527635, "learning_rate": 0.00028028929706457235, "loss": 0.3103, "step": 3774 }, { "epoch": 1.776470588235294, "grad_norm": 0.11563232224488955, "learning_rate": 0.0002802675697056463, "loss": 0.0556, "step": 3775 }, { "epoch": 1.7769411764705882, "grad_norm": 0.24288198560277455, "learning_rate": 0.00028024583122123094, "loss": 0.199, "step": 3776 }, { "epoch": 1.7774117647058825, "grad_norm": 0.3879235242683152, "learning_rate": 0.00028022408161318266, "loss": 0.3804, "step": 3777 }, { "epoch": 1.7778823529411765, "grad_norm": 0.34261730260107326, "learning_rate": 0.00028020232088335907, "loss": 0.3, "step": 3778 }, { "epoch": 1.7783529411764705, "grad_norm": 0.5984609966538813, "learning_rate": 0.00028018054903361864, "loss": 0.6138, "step": 3779 }, { "epoch": 1.7788235294117647, "grad_norm": 0.23207012346658418, "learning_rate": 0.00028015876606582083, "loss": 0.1588, "step": 3780 }, { "epoch": 1.779294117647059, "grad_norm": 0.4305761773079365, "learning_rate": 0.0002801369719818259, "loss": 0.4347, "step": 3781 }, { "epoch": 1.779764705882353, "grad_norm": 0.047002454284074006, "learning_rate": 0.0002801151667834953, "loss": 0.0142, "step": 3782 }, { "epoch": 1.780235294117647, "grad_norm": 0.22436478289634368, "learning_rate": 0.00028009335047269126, "loss": 0.1901, "step": 3783 }, { "epoch": 1.7807058823529411, "grad_norm": 0.31101084539699464, "learning_rate": 0.00028007152305127694, "loss": 0.1748, "step": 3784 }, { "epoch": 1.7811764705882354, "grad_norm": 0.18099241484020911, "learning_rate": 0.0002800496845211166, "loss": 0.11, "step": 3785 }, { "epoch": 1.7816470588235294, "grad_norm": 0.4521752990780757, "learning_rate": 0.0002800278348840753, "loss": 0.4925, "step": 3786 }, { "epoch": 1.7821176470588236, "grad_norm": 0.17447641964209662, "learning_rate": 0.00028000597414201906, "loss": 0.1386, "step": 3787 }, { "epoch": 1.7825882352941176, "grad_norm": 0.6125598698660766, "learning_rate": 0.000279984102296815, "loss": 0.4002, "step": 3788 }, { "epoch": 1.7830588235294118, "grad_norm": 0.3680942330816458, "learning_rate": 0.000279962219350331, "loss": 0.3124, "step": 3789 }, { "epoch": 1.783529411764706, "grad_norm": 0.48106036964835486, "learning_rate": 0.00027994032530443606, "loss": 0.4725, "step": 3790 }, { "epoch": 1.784, "grad_norm": 0.28066380743999964, "learning_rate": 0.00027991842016099994, "loss": 0.2128, "step": 3791 }, { "epoch": 1.784470588235294, "grad_norm": 0.14723920733668983, "learning_rate": 0.0002798965039218936, "loss": 0.1045, "step": 3792 }, { "epoch": 1.7849411764705883, "grad_norm": 0.30516531587557905, "learning_rate": 0.0002798745765889886, "loss": 0.2133, "step": 3793 }, { "epoch": 1.7854117647058825, "grad_norm": 0.2223558368108137, "learning_rate": 0.00027985263816415767, "loss": 0.1479, "step": 3794 }, { "epoch": 1.7858823529411765, "grad_norm": 0.04195237767754021, "learning_rate": 0.0002798306886492746, "loss": 0.0101, "step": 3795 }, { "epoch": 1.7863529411764705, "grad_norm": 0.5396677762974691, "learning_rate": 0.00027980872804621385, "loss": 0.612, "step": 3796 }, { "epoch": 1.7868235294117647, "grad_norm": 0.22401922709213415, "learning_rate": 0.0002797867563568511, "loss": 0.182, "step": 3797 }, { "epoch": 1.787294117647059, "grad_norm": 0.48457631973654064, "learning_rate": 0.0002797647735830627, "loss": 0.5922, "step": 3798 }, { "epoch": 1.787764705882353, "grad_norm": 0.12435323269609545, "learning_rate": 0.00027974277972672614, "loss": 0.0485, "step": 3799 }, { "epoch": 1.788235294117647, "grad_norm": 0.3298417619676137, "learning_rate": 0.0002797207747897198, "loss": 0.3401, "step": 3800 }, { "epoch": 1.7887058823529411, "grad_norm": 0.14521052357843192, "learning_rate": 0.000279698758773923, "loss": 0.0745, "step": 3801 }, { "epoch": 1.7891764705882354, "grad_norm": 0.29710481376309344, "learning_rate": 0.0002796767316812161, "loss": 0.2822, "step": 3802 }, { "epoch": 1.7896470588235294, "grad_norm": 0.1780446668000313, "learning_rate": 0.0002796546935134802, "loss": 0.1335, "step": 3803 }, { "epoch": 1.7901176470588234, "grad_norm": 0.3412769644045349, "learning_rate": 0.00027963264427259755, "loss": 0.2678, "step": 3804 }, { "epoch": 1.7905882352941176, "grad_norm": 0.2929066640354593, "learning_rate": 0.00027961058396045124, "loss": 0.3186, "step": 3805 }, { "epoch": 1.7910588235294118, "grad_norm": 0.16273776851774202, "learning_rate": 0.00027958851257892526, "loss": 0.0703, "step": 3806 }, { "epoch": 1.791529411764706, "grad_norm": 0.08577515369429377, "learning_rate": 0.0002795664301299047, "loss": 0.0646, "step": 3807 }, { "epoch": 1.792, "grad_norm": 0.24434399005354748, "learning_rate": 0.00027954433661527554, "loss": 0.2732, "step": 3808 }, { "epoch": 1.792470588235294, "grad_norm": 0.36586993514775684, "learning_rate": 0.0002795222320369246, "loss": 0.4286, "step": 3809 }, { "epoch": 1.7929411764705883, "grad_norm": 0.15753819027101282, "learning_rate": 0.00027950011639673976, "loss": 0.063, "step": 3810 }, { "epoch": 1.7934117647058825, "grad_norm": 0.10581888452387504, "learning_rate": 0.00027947798969660975, "loss": 0.057, "step": 3811 }, { "epoch": 1.7938823529411765, "grad_norm": 0.4039441360346178, "learning_rate": 0.00027945585193842433, "loss": 0.3782, "step": 3812 }, { "epoch": 1.7943529411764705, "grad_norm": 0.1230150446118128, "learning_rate": 0.00027943370312407415, "loss": 0.0733, "step": 3813 }, { "epoch": 1.7948235294117647, "grad_norm": 0.3444833112787447, "learning_rate": 0.00027941154325545096, "loss": 0.3567, "step": 3814 }, { "epoch": 1.795294117647059, "grad_norm": 0.47210431325717817, "learning_rate": 0.00027938937233444713, "loss": 0.4355, "step": 3815 }, { "epoch": 1.795764705882353, "grad_norm": 0.22453658249722952, "learning_rate": 0.00027936719036295623, "loss": 0.1844, "step": 3816 }, { "epoch": 1.796235294117647, "grad_norm": 0.3714606168277369, "learning_rate": 0.0002793449973428728, "loss": 0.3359, "step": 3817 }, { "epoch": 1.7967058823529412, "grad_norm": 0.08001019433804628, "learning_rate": 0.0002793227932760921, "loss": 0.0303, "step": 3818 }, { "epoch": 1.7971764705882354, "grad_norm": 0.4599930301237728, "learning_rate": 0.0002793005781645106, "loss": 0.354, "step": 3819 }, { "epoch": 1.7976470588235294, "grad_norm": 0.5066816540590295, "learning_rate": 0.00027927835201002545, "loss": 0.3221, "step": 3820 }, { "epoch": 1.7981176470588234, "grad_norm": 0.5070120464789506, "learning_rate": 0.00027925611481453497, "loss": 0.461, "step": 3821 }, { "epoch": 1.7985882352941176, "grad_norm": 0.5318036066370265, "learning_rate": 0.0002792338665799383, "loss": 0.5249, "step": 3822 }, { "epoch": 1.7990588235294118, "grad_norm": 0.18486874495447342, "learning_rate": 0.00027921160730813544, "loss": 0.1222, "step": 3823 }, { "epoch": 1.799529411764706, "grad_norm": 0.763798443275989, "learning_rate": 0.0002791893370010276, "loss": 0.674, "step": 3824 }, { "epoch": 1.8, "grad_norm": 0.5533566511626807, "learning_rate": 0.00027916705566051676, "loss": 0.3602, "step": 3825 }, { "epoch": 1.800470588235294, "grad_norm": 0.31077168192828875, "learning_rate": 0.00027914476328850575, "loss": 0.2151, "step": 3826 }, { "epoch": 1.8009411764705883, "grad_norm": 0.23023144530877845, "learning_rate": 0.0002791224598868985, "loss": 0.1501, "step": 3827 }, { "epoch": 1.8014117647058825, "grad_norm": 0.6749865849573308, "learning_rate": 0.00027910014545759986, "loss": 0.6255, "step": 3828 }, { "epoch": 1.8018823529411765, "grad_norm": 0.479697217878809, "learning_rate": 0.00027907782000251557, "loss": 0.4526, "step": 3829 }, { "epoch": 1.8023529411764705, "grad_norm": 0.16420473102830313, "learning_rate": 0.00027905548352355226, "loss": 0.1208, "step": 3830 }, { "epoch": 1.8028235294117647, "grad_norm": 0.29312917356033086, "learning_rate": 0.00027903313602261774, "loss": 0.214, "step": 3831 }, { "epoch": 1.803294117647059, "grad_norm": 0.6400971718714729, "learning_rate": 0.00027901077750162043, "loss": 0.5761, "step": 3832 }, { "epoch": 1.803764705882353, "grad_norm": 0.37374656766504755, "learning_rate": 0.00027898840796246994, "loss": 0.3619, "step": 3833 }, { "epoch": 1.804235294117647, "grad_norm": 0.20896803784998355, "learning_rate": 0.00027896602740707673, "loss": 0.1646, "step": 3834 }, { "epoch": 1.8047058823529412, "grad_norm": 0.3517523667065591, "learning_rate": 0.0002789436358373522, "loss": 0.3806, "step": 3835 }, { "epoch": 1.8051764705882354, "grad_norm": 0.2402794774849289, "learning_rate": 0.0002789212332552088, "loss": 0.1533, "step": 3836 }, { "epoch": 1.8056470588235294, "grad_norm": 0.446127803043293, "learning_rate": 0.0002788988196625596, "loss": 0.4152, "step": 3837 }, { "epoch": 1.8061176470588234, "grad_norm": 0.5919914667458231, "learning_rate": 0.000278876395061319, "loss": 0.4144, "step": 3838 }, { "epoch": 1.8065882352941176, "grad_norm": 0.4645928397608969, "learning_rate": 0.0002788539594534021, "loss": 0.4905, "step": 3839 }, { "epoch": 1.8070588235294118, "grad_norm": 0.18283084256465834, "learning_rate": 0.0002788315128407251, "loss": 0.0958, "step": 3840 }, { "epoch": 1.8075294117647058, "grad_norm": 0.5822827572609037, "learning_rate": 0.0002788090552252049, "loss": 0.5678, "step": 3841 }, { "epoch": 1.808, "grad_norm": 0.4652868457711405, "learning_rate": 0.00027878658660875967, "loss": 0.3845, "step": 3842 }, { "epoch": 1.808470588235294, "grad_norm": 0.40093194891546324, "learning_rate": 0.0002787641069933082, "loss": 0.4616, "step": 3843 }, { "epoch": 1.8089411764705883, "grad_norm": 0.43494156200814293, "learning_rate": 0.0002787416163807704, "loss": 0.524, "step": 3844 }, { "epoch": 1.8094117647058825, "grad_norm": 0.34122816338887535, "learning_rate": 0.0002787191147730671, "loss": 0.3628, "step": 3845 }, { "epoch": 1.8098823529411765, "grad_norm": 0.27356500574008136, "learning_rate": 0.00027869660217212005, "loss": 0.1289, "step": 3846 }, { "epoch": 1.8103529411764705, "grad_norm": 0.28012326718292946, "learning_rate": 0.0002786740785798519, "loss": 0.1968, "step": 3847 }, { "epoch": 1.8108235294117647, "grad_norm": 0.17891803076240378, "learning_rate": 0.00027865154399818627, "loss": 0.1135, "step": 3848 }, { "epoch": 1.811294117647059, "grad_norm": 0.3958078346280242, "learning_rate": 0.00027862899842904783, "loss": 0.2832, "step": 3849 }, { "epoch": 1.811764705882353, "grad_norm": 0.4462564167474234, "learning_rate": 0.00027860644187436195, "loss": 0.3368, "step": 3850 }, { "epoch": 1.812235294117647, "grad_norm": 0.32238053058969623, "learning_rate": 0.0002785838743360551, "loss": 0.3274, "step": 3851 }, { "epoch": 1.8127058823529412, "grad_norm": 0.18113914736822181, "learning_rate": 0.00027856129581605474, "loss": 0.0964, "step": 3852 }, { "epoch": 1.8131764705882354, "grad_norm": 0.15758594257980071, "learning_rate": 0.0002785387063162891, "loss": 0.1403, "step": 3853 }, { "epoch": 1.8136470588235294, "grad_norm": 0.23481833758262224, "learning_rate": 0.00027851610583868745, "loss": 0.139, "step": 3854 }, { "epoch": 1.8141176470588234, "grad_norm": 0.2982110272567404, "learning_rate": 0.00027849349438518003, "loss": 0.3122, "step": 3855 }, { "epoch": 1.8145882352941176, "grad_norm": 0.2453685064167406, "learning_rate": 0.0002784708719576979, "loss": 0.1826, "step": 3856 }, { "epoch": 1.8150588235294118, "grad_norm": 0.4021741831631712, "learning_rate": 0.00027844823855817315, "loss": 0.4229, "step": 3857 }, { "epoch": 1.8155294117647058, "grad_norm": 0.12060090435787355, "learning_rate": 0.0002784255941885388, "loss": 0.0824, "step": 3858 }, { "epoch": 1.8159999999999998, "grad_norm": 0.3902579263647727, "learning_rate": 0.0002784029388507288, "loss": 0.5138, "step": 3859 }, { "epoch": 1.816470588235294, "grad_norm": 0.3422615821968838, "learning_rate": 0.000278380272546678, "loss": 0.2904, "step": 3860 }, { "epoch": 1.8169411764705883, "grad_norm": 0.3759685354820703, "learning_rate": 0.00027835759527832226, "loss": 0.4396, "step": 3861 }, { "epoch": 1.8174117647058825, "grad_norm": 0.14765500099623896, "learning_rate": 0.0002783349070475982, "loss": 0.1088, "step": 3862 }, { "epoch": 1.8178823529411765, "grad_norm": 0.3567060223747459, "learning_rate": 0.00027831220785644366, "loss": 0.3091, "step": 3863 }, { "epoch": 1.8183529411764705, "grad_norm": 0.22761834779821027, "learning_rate": 0.0002782894977067972, "loss": 0.1407, "step": 3864 }, { "epoch": 1.8188235294117647, "grad_norm": 0.2378129365220355, "learning_rate": 0.0002782667766005984, "loss": 0.1135, "step": 3865 }, { "epoch": 1.819294117647059, "grad_norm": 0.39183793205354206, "learning_rate": 0.0002782440445397877, "loss": 0.5256, "step": 3866 }, { "epoch": 1.819764705882353, "grad_norm": 0.38291030046022195, "learning_rate": 0.00027822130152630657, "loss": 0.4458, "step": 3867 }, { "epoch": 1.820235294117647, "grad_norm": 0.22382107302454704, "learning_rate": 0.00027819854756209734, "loss": 0.243, "step": 3868 }, { "epoch": 1.8207058823529412, "grad_norm": 0.23992113661399028, "learning_rate": 0.00027817578264910343, "loss": 0.2, "step": 3869 }, { "epoch": 1.8211764705882354, "grad_norm": 0.2557336630864071, "learning_rate": 0.00027815300678926886, "loss": 0.1994, "step": 3870 }, { "epoch": 1.8216470588235294, "grad_norm": 0.1848158981535463, "learning_rate": 0.000278130219984539, "loss": 0.2256, "step": 3871 }, { "epoch": 1.8221176470588234, "grad_norm": 0.4634414046596183, "learning_rate": 0.00027810742223685983, "loss": 0.5194, "step": 3872 }, { "epoch": 1.8225882352941176, "grad_norm": 0.26005839246085405, "learning_rate": 0.00027808461354817847, "loss": 0.1828, "step": 3873 }, { "epoch": 1.8230588235294118, "grad_norm": 0.20184835252718325, "learning_rate": 0.00027806179392044286, "loss": 0.1089, "step": 3874 }, { "epoch": 1.8235294117647058, "grad_norm": 0.09776302348644791, "learning_rate": 0.0002780389633556019, "loss": 0.031, "step": 3875 }, { "epoch": 1.8239999999999998, "grad_norm": 0.34489002377632344, "learning_rate": 0.00027801612185560554, "loss": 0.3785, "step": 3876 }, { "epoch": 1.824470588235294, "grad_norm": 0.3171852423691069, "learning_rate": 0.00027799326942240434, "loss": 0.2763, "step": 3877 }, { "epoch": 1.8249411764705883, "grad_norm": 0.30995576534318425, "learning_rate": 0.0002779704060579501, "loss": 0.2392, "step": 3878 }, { "epoch": 1.8254117647058825, "grad_norm": 0.33257154362403013, "learning_rate": 0.00027794753176419557, "loss": 0.2865, "step": 3879 }, { "epoch": 1.8258823529411765, "grad_norm": 0.4022878263126746, "learning_rate": 0.00027792464654309427, "loss": 0.6339, "step": 3880 }, { "epoch": 1.8263529411764705, "grad_norm": 0.2392907362743498, "learning_rate": 0.0002779017503966006, "loss": 0.2195, "step": 3881 }, { "epoch": 1.8268235294117647, "grad_norm": 0.1628904738170434, "learning_rate": 0.0002778788433266702, "loss": 0.0985, "step": 3882 }, { "epoch": 1.827294117647059, "grad_norm": 0.2936680435137551, "learning_rate": 0.00027785592533525924, "loss": 0.1819, "step": 3883 }, { "epoch": 1.827764705882353, "grad_norm": 0.6608355830785149, "learning_rate": 0.00027783299642432515, "loss": 0.6506, "step": 3884 }, { "epoch": 1.828235294117647, "grad_norm": 0.21727282084203456, "learning_rate": 0.0002778100565958262, "loss": 0.1188, "step": 3885 }, { "epoch": 1.8287058823529412, "grad_norm": 0.3317605011543527, "learning_rate": 0.0002777871058517215, "loss": 0.3391, "step": 3886 }, { "epoch": 1.8291764705882354, "grad_norm": 0.19446951087814762, "learning_rate": 0.00027776414419397106, "loss": 0.1057, "step": 3887 }, { "epoch": 1.8296470588235294, "grad_norm": 0.4949070416416312, "learning_rate": 0.00027774117162453617, "loss": 0.5218, "step": 3888 }, { "epoch": 1.8301176470588234, "grad_norm": 0.3202158015084854, "learning_rate": 0.00027771818814537855, "loss": 0.2956, "step": 3889 }, { "epoch": 1.8305882352941176, "grad_norm": 0.09795560293977158, "learning_rate": 0.00027769519375846126, "loss": 0.0598, "step": 3890 }, { "epoch": 1.8310588235294118, "grad_norm": 0.3102554647440061, "learning_rate": 0.0002776721884657481, "loss": 0.3316, "step": 3891 }, { "epoch": 1.8315294117647059, "grad_norm": 0.49840067437757435, "learning_rate": 0.00027764917226920377, "loss": 0.4239, "step": 3892 }, { "epoch": 1.8319999999999999, "grad_norm": 0.26059100786946904, "learning_rate": 0.000277626145170794, "loss": 0.3148, "step": 3893 }, { "epoch": 1.832470588235294, "grad_norm": 0.19261779002937532, "learning_rate": 0.0002776031071724855, "loss": 0.1121, "step": 3894 }, { "epoch": 1.8329411764705883, "grad_norm": 0.6424661851102419, "learning_rate": 0.0002775800582762457, "loss": 0.5179, "step": 3895 }, { "epoch": 1.8334117647058825, "grad_norm": 5.205303896018262e-07, "learning_rate": 0.00027755699848404315, "loss": 0.0, "step": 3896 }, { "epoch": 1.8338823529411765, "grad_norm": 0.46421785913950997, "learning_rate": 0.0002775339277978473, "loss": 0.3749, "step": 3897 }, { "epoch": 1.8343529411764705, "grad_norm": 0.420799078229946, "learning_rate": 0.0002775108462196284, "loss": 0.3276, "step": 3898 }, { "epoch": 1.8348235294117647, "grad_norm": 0.19371363179322293, "learning_rate": 0.0002774877537513579, "loss": 0.0891, "step": 3899 }, { "epoch": 1.835294117647059, "grad_norm": 0.1104459874201302, "learning_rate": 0.0002774646503950078, "loss": 0.067, "step": 3900 }, { "epoch": 1.835764705882353, "grad_norm": 0.2988692864666814, "learning_rate": 0.0002774415361525514, "loss": 0.2308, "step": 3901 }, { "epoch": 1.836235294117647, "grad_norm": 0.33659272005778695, "learning_rate": 0.00027741841102596276, "loss": 0.2883, "step": 3902 }, { "epoch": 1.8367058823529412, "grad_norm": 0.5702217303594179, "learning_rate": 0.0002773952750172168, "loss": 0.6106, "step": 3903 }, { "epoch": 1.8371764705882354, "grad_norm": 0.39630912862816364, "learning_rate": 0.0002773721281282894, "loss": 0.3958, "step": 3904 }, { "epoch": 1.8376470588235294, "grad_norm": 0.34190501740191154, "learning_rate": 0.0002773489703611576, "loss": 0.3074, "step": 3905 }, { "epoch": 1.8381176470588234, "grad_norm": 0.3153866362594806, "learning_rate": 0.00027732580171779906, "loss": 0.3752, "step": 3906 }, { "epoch": 1.8385882352941176, "grad_norm": 0.4391650164309115, "learning_rate": 0.0002773026222001925, "loss": 0.3294, "step": 3907 }, { "epoch": 1.8390588235294119, "grad_norm": 0.3433619059995419, "learning_rate": 0.00027727943181031766, "loss": 0.4038, "step": 3908 }, { "epoch": 1.8395294117647059, "grad_norm": 0.49173709046386344, "learning_rate": 0.00027725623055015495, "loss": 0.4158, "step": 3909 }, { "epoch": 1.8399999999999999, "grad_norm": 0.15989696950155202, "learning_rate": 0.000277233018421686, "loss": 0.1474, "step": 3910 }, { "epoch": 1.840470588235294, "grad_norm": 0.46562660347274015, "learning_rate": 0.00027720979542689324, "loss": 0.2372, "step": 3911 }, { "epoch": 1.8409411764705883, "grad_norm": 0.29683234726916913, "learning_rate": 0.0002771865615677599, "loss": 0.3048, "step": 3912 }, { "epoch": 1.8414117647058823, "grad_norm": 0.564032687244247, "learning_rate": 0.00027716331684627044, "loss": 0.4232, "step": 3913 }, { "epoch": 1.8418823529411765, "grad_norm": 0.18207157228409873, "learning_rate": 0.00027714006126441, "loss": 0.0882, "step": 3914 }, { "epoch": 1.8423529411764705, "grad_norm": 0.4396763057135602, "learning_rate": 0.00027711679482416465, "loss": 0.4472, "step": 3915 }, { "epoch": 1.8428235294117647, "grad_norm": 0.15912745261870018, "learning_rate": 0.0002770935175275215, "loss": 0.1431, "step": 3916 }, { "epoch": 1.843294117647059, "grad_norm": 0.3430984135610682, "learning_rate": 0.0002770702293764686, "loss": 0.3183, "step": 3917 }, { "epoch": 1.843764705882353, "grad_norm": 0.21152998301222975, "learning_rate": 0.0002770469303729949, "loss": 0.132, "step": 3918 }, { "epoch": 1.844235294117647, "grad_norm": 0.3754743263443932, "learning_rate": 0.0002770236205190901, "loss": 0.3589, "step": 3919 }, { "epoch": 1.8447058823529412, "grad_norm": 0.2571180151464884, "learning_rate": 0.00027700029981674506, "loss": 0.27, "step": 3920 }, { "epoch": 1.8451764705882354, "grad_norm": 0.3279678075984781, "learning_rate": 0.0002769769682679515, "loss": 0.322, "step": 3921 }, { "epoch": 1.8456470588235294, "grad_norm": 0.16606275520622224, "learning_rate": 0.00027695362587470203, "loss": 0.1334, "step": 3922 }, { "epoch": 1.8461176470588234, "grad_norm": 0.3644638737014053, "learning_rate": 0.0002769302726389902, "loss": 0.3995, "step": 3923 }, { "epoch": 1.8465882352941176, "grad_norm": 0.31397628147370366, "learning_rate": 0.00027690690856281054, "loss": 0.3411, "step": 3924 }, { "epoch": 1.8470588235294119, "grad_norm": 0.29426418405421595, "learning_rate": 0.00027688353364815834, "loss": 0.3379, "step": 3925 }, { "epoch": 1.8475294117647059, "grad_norm": 0.5337911335962853, "learning_rate": 0.0002768601478970301, "loss": 0.7365, "step": 3926 }, { "epoch": 1.8479999999999999, "grad_norm": 0.402287354067072, "learning_rate": 0.00027683675131142294, "loss": 0.451, "step": 3927 }, { "epoch": 1.848470588235294, "grad_norm": 0.1953900100952732, "learning_rate": 0.000276813343893335, "loss": 0.1172, "step": 3928 }, { "epoch": 1.8489411764705883, "grad_norm": 0.38754074727276516, "learning_rate": 0.0002767899256447656, "loss": 0.2902, "step": 3929 }, { "epoch": 1.8494117647058823, "grad_norm": 0.5298630659755924, "learning_rate": 0.0002767664965677146, "loss": 0.6484, "step": 3930 }, { "epoch": 1.8498823529411763, "grad_norm": 0.3441851509311593, "learning_rate": 0.000276743056664183, "loss": 0.3441, "step": 3931 }, { "epoch": 1.8503529411764705, "grad_norm": 0.30928463580740045, "learning_rate": 0.0002767196059361727, "loss": 0.3319, "step": 3932 }, { "epoch": 1.8508235294117648, "grad_norm": 0.5566437686896755, "learning_rate": 0.0002766961443856865, "loss": 0.2853, "step": 3933 }, { "epoch": 1.851294117647059, "grad_norm": 0.18306246125371292, "learning_rate": 0.00027667267201472815, "loss": 0.1092, "step": 3934 }, { "epoch": 1.851764705882353, "grad_norm": 0.14342158076223524, "learning_rate": 0.00027664918882530225, "loss": 0.073, "step": 3935 }, { "epoch": 1.852235294117647, "grad_norm": 0.4563002189461934, "learning_rate": 0.0002766256948194144, "loss": 0.4084, "step": 3936 }, { "epoch": 1.8527058823529412, "grad_norm": 0.4523745503684558, "learning_rate": 0.0002766021899990712, "loss": 0.4374, "step": 3937 }, { "epoch": 1.8531764705882354, "grad_norm": 0.5571071515122079, "learning_rate": 0.0002765786743662799, "loss": 0.5115, "step": 3938 }, { "epoch": 1.8536470588235294, "grad_norm": 0.33776309557724715, "learning_rate": 0.0002765551479230489, "loss": 0.3617, "step": 3939 }, { "epoch": 1.8541176470588234, "grad_norm": 0.34191076543706245, "learning_rate": 0.00027653161067138764, "loss": 0.3871, "step": 3940 }, { "epoch": 1.8545882352941176, "grad_norm": 0.23331492495254855, "learning_rate": 0.00027650806261330616, "loss": 0.1599, "step": 3941 }, { "epoch": 1.8550588235294119, "grad_norm": 0.21301161117183884, "learning_rate": 0.0002764845037508156, "loss": 0.1268, "step": 3942 }, { "epoch": 1.8555294117647059, "grad_norm": 0.35633609390820975, "learning_rate": 0.00027646093408592804, "loss": 0.2659, "step": 3943 }, { "epoch": 1.8559999999999999, "grad_norm": 0.33369765824042374, "learning_rate": 0.0002764373536206564, "loss": 0.2312, "step": 3944 }, { "epoch": 1.856470588235294, "grad_norm": 0.3278187002912551, "learning_rate": 0.0002764137623570146, "loss": 0.4017, "step": 3945 }, { "epoch": 1.8569411764705883, "grad_norm": 0.4921896108447842, "learning_rate": 0.0002763901602970174, "loss": 0.6781, "step": 3946 }, { "epoch": 1.8574117647058823, "grad_norm": 0.2960830309291424, "learning_rate": 0.0002763665474426806, "loss": 0.3175, "step": 3947 }, { "epoch": 1.8578823529411763, "grad_norm": 0.30642805601501255, "learning_rate": 0.00027634292379602086, "loss": 0.3214, "step": 3948 }, { "epoch": 1.8583529411764705, "grad_norm": 0.38814544957255526, "learning_rate": 0.00027631928935905567, "loss": 0.5223, "step": 3949 }, { "epoch": 1.8588235294117648, "grad_norm": 0.29408057286116834, "learning_rate": 0.0002762956441338036, "loss": 0.2746, "step": 3950 }, { "epoch": 1.859294117647059, "grad_norm": 0.3039087432296917, "learning_rate": 0.00027627198812228405, "loss": 0.1969, "step": 3951 }, { "epoch": 1.859764705882353, "grad_norm": 0.34640816341427205, "learning_rate": 0.0002762483213265173, "loss": 0.3531, "step": 3952 }, { "epoch": 1.860235294117647, "grad_norm": 0.37549457098196165, "learning_rate": 0.0002762246437485247, "loss": 0.2231, "step": 3953 }, { "epoch": 1.8607058823529412, "grad_norm": 0.1830447957844186, "learning_rate": 0.0002762009553903284, "loss": 0.1014, "step": 3954 }, { "epoch": 1.8611764705882354, "grad_norm": 0.23798583983301885, "learning_rate": 0.0002761772562539515, "loss": 0.12, "step": 3955 }, { "epoch": 1.8616470588235294, "grad_norm": 0.5247866562325033, "learning_rate": 0.0002761535463414181, "loss": 0.4296, "step": 3956 }, { "epoch": 1.8621176470588234, "grad_norm": 0.45652502892159125, "learning_rate": 0.0002761298256547529, "loss": 0.3195, "step": 3957 }, { "epoch": 1.8625882352941177, "grad_norm": 0.20348335555449942, "learning_rate": 0.00027610609419598204, "loss": 0.2131, "step": 3958 }, { "epoch": 1.8630588235294119, "grad_norm": 0.45916789685309795, "learning_rate": 0.0002760823519671321, "loss": 0.3896, "step": 3959 }, { "epoch": 1.8635294117647059, "grad_norm": 0.2902564457501921, "learning_rate": 0.0002760585989702309, "loss": 0.3053, "step": 3960 }, { "epoch": 1.8639999999999999, "grad_norm": 0.26750953598922284, "learning_rate": 0.0002760348352073071, "loss": 0.279, "step": 3961 }, { "epoch": 1.864470588235294, "grad_norm": 0.254089819398059, "learning_rate": 0.0002760110606803901, "loss": 0.2058, "step": 3962 }, { "epoch": 1.8649411764705883, "grad_norm": 0.1058284640874167, "learning_rate": 0.00027598727539151054, "loss": 0.0636, "step": 3963 }, { "epoch": 1.8654117647058823, "grad_norm": 0.42562991109397635, "learning_rate": 0.0002759634793426996, "loss": 0.395, "step": 3964 }, { "epoch": 1.8658823529411763, "grad_norm": 0.24398013640125496, "learning_rate": 0.00027593967253598974, "loss": 0.1404, "step": 3965 }, { "epoch": 1.8663529411764705, "grad_norm": 0.4595317471355416, "learning_rate": 0.00027591585497341406, "loss": 0.3972, "step": 3966 }, { "epoch": 1.8668235294117648, "grad_norm": 0.24391598377858853, "learning_rate": 0.00027589202665700676, "loss": 0.1675, "step": 3967 }, { "epoch": 1.8672941176470588, "grad_norm": 0.4517928552595612, "learning_rate": 0.0002758681875888029, "loss": 0.5184, "step": 3968 }, { "epoch": 1.867764705882353, "grad_norm": 0.2805581307486651, "learning_rate": 0.00027584433777083845, "loss": 0.1888, "step": 3969 }, { "epoch": 1.868235294117647, "grad_norm": 0.43728220883350716, "learning_rate": 0.0002758204772051503, "loss": 0.4459, "step": 3970 }, { "epoch": 1.8687058823529412, "grad_norm": 0.14667925710756766, "learning_rate": 0.00027579660589377626, "loss": 0.0987, "step": 3971 }, { "epoch": 1.8691764705882354, "grad_norm": 0.44840148689012826, "learning_rate": 0.00027577272383875497, "loss": 0.3998, "step": 3972 }, { "epoch": 1.8696470588235294, "grad_norm": 0.3171423860310188, "learning_rate": 0.0002757488310421262, "loss": 0.2278, "step": 3973 }, { "epoch": 1.8701176470588234, "grad_norm": 0.20783672637661949, "learning_rate": 0.0002757249275059305, "loss": 0.1492, "step": 3974 }, { "epoch": 1.8705882352941177, "grad_norm": 0.3014622399842909, "learning_rate": 0.00027570101323220927, "loss": 0.3206, "step": 3975 }, { "epoch": 1.8710588235294119, "grad_norm": 0.440566534677988, "learning_rate": 0.00027567708822300496, "loss": 0.4901, "step": 3976 }, { "epoch": 1.8715294117647059, "grad_norm": 0.3643402620832481, "learning_rate": 0.0002756531524803609, "loss": 0.3777, "step": 3977 }, { "epoch": 1.8719999999999999, "grad_norm": 0.23868216237998596, "learning_rate": 0.00027562920600632127, "loss": 0.1978, "step": 3978 }, { "epoch": 1.872470588235294, "grad_norm": 0.31475646195791096, "learning_rate": 0.0002756052488029312, "loss": 0.1737, "step": 3979 }, { "epoch": 1.8729411764705883, "grad_norm": 0.4658059420111863, "learning_rate": 0.0002755812808722368, "loss": 0.6097, "step": 3980 }, { "epoch": 1.8734117647058823, "grad_norm": 0.4137360505833882, "learning_rate": 0.0002755573022162851, "loss": 0.339, "step": 3981 }, { "epoch": 1.8738823529411763, "grad_norm": 0.30467406701778066, "learning_rate": 0.0002755333128371239, "loss": 0.2667, "step": 3982 }, { "epoch": 1.8743529411764706, "grad_norm": 0.35671441035740126, "learning_rate": 0.00027550931273680207, "loss": 0.2091, "step": 3983 }, { "epoch": 1.8748235294117648, "grad_norm": 0.3704859120460328, "learning_rate": 0.00027548530191736926, "loss": 0.3553, "step": 3984 }, { "epoch": 1.8752941176470588, "grad_norm": 0.37808384728534866, "learning_rate": 0.0002754612803808762, "loss": 0.4909, "step": 3985 }, { "epoch": 1.8757647058823528, "grad_norm": 0.3224417515356952, "learning_rate": 0.00027543724812937443, "loss": 0.38, "step": 3986 }, { "epoch": 1.876235294117647, "grad_norm": 0.44278263950968894, "learning_rate": 0.0002754132051649164, "loss": 0.3866, "step": 3987 }, { "epoch": 1.8767058823529412, "grad_norm": 0.176222123663167, "learning_rate": 0.00027538915148955545, "loss": 0.0857, "step": 3988 }, { "epoch": 1.8771764705882354, "grad_norm": 0.5279496498266377, "learning_rate": 0.000275365087105346, "loss": 0.3935, "step": 3989 }, { "epoch": 1.8776470588235294, "grad_norm": 0.5701699455231899, "learning_rate": 0.00027534101201434315, "loss": 0.6516, "step": 3990 }, { "epoch": 1.8781176470588234, "grad_norm": 0.36435469704164625, "learning_rate": 0.00027531692621860307, "loss": 0.3379, "step": 3991 }, { "epoch": 1.8785882352941177, "grad_norm": 0.3264164144068691, "learning_rate": 0.00027529282972018287, "loss": 0.2715, "step": 3992 }, { "epoch": 1.879058823529412, "grad_norm": 0.3108675156901407, "learning_rate": 0.0002752687225211404, "loss": 0.2628, "step": 3993 }, { "epoch": 1.879529411764706, "grad_norm": 0.1378197471828131, "learning_rate": 0.0002752446046235346, "loss": 0.0673, "step": 3994 }, { "epoch": 1.88, "grad_norm": 0.22096500542485978, "learning_rate": 0.00027522047602942525, "loss": 0.1466, "step": 3995 }, { "epoch": 1.8804705882352941, "grad_norm": 0.2760381080328071, "learning_rate": 0.00027519633674087305, "loss": 0.2128, "step": 3996 }, { "epoch": 1.8809411764705883, "grad_norm": 0.5241292465529453, "learning_rate": 0.0002751721867599396, "loss": 0.5608, "step": 3997 }, { "epoch": 1.8814117647058823, "grad_norm": 0.14536756988662364, "learning_rate": 0.00027514802608868743, "loss": 0.0496, "step": 3998 }, { "epoch": 1.8818823529411763, "grad_norm": 0.35284196929313927, "learning_rate": 0.00027512385472918, "loss": 0.3906, "step": 3999 }, { "epoch": 1.8823529411764706, "grad_norm": 0.5120842448984471, "learning_rate": 0.0002750996726834817, "loss": 0.6457, "step": 4000 }, { "epoch": 1.8828235294117648, "grad_norm": 0.1627453012834583, "learning_rate": 0.00027507547995365766, "loss": 0.1194, "step": 4001 }, { "epoch": 1.8832941176470588, "grad_norm": 0.2506465766078405, "learning_rate": 0.00027505127654177413, "loss": 0.3535, "step": 4002 }, { "epoch": 1.8837647058823528, "grad_norm": 0.4434367526143825, "learning_rate": 0.0002750270624498983, "loss": 0.5536, "step": 4003 }, { "epoch": 1.884235294117647, "grad_norm": 0.34424405472229386, "learning_rate": 0.000275002837680098, "loss": 0.2908, "step": 4004 }, { "epoch": 1.8847058823529412, "grad_norm": 0.5160488787968872, "learning_rate": 0.0002749786022344422, "loss": 0.3282, "step": 4005 }, { "epoch": 1.8851764705882355, "grad_norm": 0.4085595521679609, "learning_rate": 0.0002749543561150009, "loss": 0.3863, "step": 4006 }, { "epoch": 1.8856470588235295, "grad_norm": 0.43283899753382804, "learning_rate": 0.00027493009932384456, "loss": 0.361, "step": 4007 }, { "epoch": 1.8861176470588235, "grad_norm": 0.26682005496983763, "learning_rate": 0.000274905831863045, "loss": 0.2397, "step": 4008 }, { "epoch": 1.8865882352941177, "grad_norm": 0.2459800200868445, "learning_rate": 0.00027488155373467474, "loss": 0.2493, "step": 4009 }, { "epoch": 1.887058823529412, "grad_norm": 0.44603535819281886, "learning_rate": 0.00027485726494080724, "loss": 0.4808, "step": 4010 }, { "epoch": 1.887529411764706, "grad_norm": 0.36778693507048404, "learning_rate": 0.0002748329654835169, "loss": 0.3609, "step": 4011 }, { "epoch": 1.888, "grad_norm": 0.3581942911199455, "learning_rate": 0.000274808655364879, "loss": 0.3379, "step": 4012 }, { "epoch": 1.8884705882352941, "grad_norm": 0.5396653856847194, "learning_rate": 0.00027478433458696976, "loss": 0.6811, "step": 4013 }, { "epoch": 1.8889411764705883, "grad_norm": 0.2843498992664322, "learning_rate": 0.0002747600031518663, "loss": 0.1263, "step": 4014 }, { "epoch": 1.8894117647058823, "grad_norm": 0.3319531044611061, "learning_rate": 0.0002747356610616466, "loss": 0.3114, "step": 4015 }, { "epoch": 1.8898823529411763, "grad_norm": 0.3087105303081681, "learning_rate": 0.0002747113083183896, "loss": 0.2388, "step": 4016 }, { "epoch": 1.8903529411764706, "grad_norm": 0.2290214862572695, "learning_rate": 0.00027468694492417515, "loss": 0.2539, "step": 4017 }, { "epoch": 1.8908235294117648, "grad_norm": 0.5126110067798243, "learning_rate": 0.0002746625708810841, "loss": 0.4684, "step": 4018 }, { "epoch": 1.8912941176470588, "grad_norm": 0.14594132042586072, "learning_rate": 0.00027463818619119794, "loss": 0.0884, "step": 4019 }, { "epoch": 1.8917647058823528, "grad_norm": 0.809550267705282, "learning_rate": 0.0002746137908565994, "loss": 0.6457, "step": 4020 }, { "epoch": 1.892235294117647, "grad_norm": 0.3988609793213819, "learning_rate": 0.00027458938487937186, "loss": 0.3443, "step": 4021 }, { "epoch": 1.8927058823529412, "grad_norm": 0.36620865722011176, "learning_rate": 0.0002745649682615998, "loss": 0.3111, "step": 4022 }, { "epoch": 1.8931764705882355, "grad_norm": 0.19239092239532232, "learning_rate": 0.00027454054100536836, "loss": 0.1822, "step": 4023 }, { "epoch": 1.8936470588235295, "grad_norm": 0.205997928093307, "learning_rate": 0.00027451610311276397, "loss": 0.2024, "step": 4024 }, { "epoch": 1.8941176470588235, "grad_norm": 0.26099815836734314, "learning_rate": 0.00027449165458587354, "loss": 0.2672, "step": 4025 }, { "epoch": 1.8945882352941177, "grad_norm": 0.26736036152428905, "learning_rate": 0.0002744671954267852, "loss": 0.1817, "step": 4026 }, { "epoch": 1.895058823529412, "grad_norm": 0.28183707061809854, "learning_rate": 0.00027444272563758784, "loss": 0.1928, "step": 4027 }, { "epoch": 1.895529411764706, "grad_norm": 0.32809840080030594, "learning_rate": 0.00027441824522037137, "loss": 0.1909, "step": 4028 }, { "epoch": 1.896, "grad_norm": 0.1812372187426558, "learning_rate": 0.0002743937541772265, "loss": 0.1237, "step": 4029 }, { "epoch": 1.8964705882352941, "grad_norm": 0.2851975201083482, "learning_rate": 0.0002743692525102448, "loss": 0.2113, "step": 4030 }, { "epoch": 1.8969411764705884, "grad_norm": 0.18964191670391603, "learning_rate": 0.000274344740221519, "loss": 0.1183, "step": 4031 }, { "epoch": 1.8974117647058824, "grad_norm": 0.4174558416776498, "learning_rate": 0.0002743202173131424, "loss": 0.4006, "step": 4032 }, { "epoch": 1.8978823529411764, "grad_norm": 0.33701498401915986, "learning_rate": 0.00027429568378720945, "loss": 0.3969, "step": 4033 }, { "epoch": 1.8983529411764706, "grad_norm": 0.17796377918960038, "learning_rate": 0.0002742711396458155, "loss": 0.1256, "step": 4034 }, { "epoch": 1.8988235294117648, "grad_norm": 0.4778735502758157, "learning_rate": 0.00027424658489105667, "loss": 0.3349, "step": 4035 }, { "epoch": 1.8992941176470588, "grad_norm": 0.3217353876501791, "learning_rate": 0.00027422201952503005, "loss": 0.2598, "step": 4036 }, { "epoch": 1.8997647058823528, "grad_norm": 0.315084026909531, "learning_rate": 0.00027419744354983364, "loss": 0.3135, "step": 4037 }, { "epoch": 1.900235294117647, "grad_norm": 0.49485769842887817, "learning_rate": 0.00027417285696756634, "loss": 0.6602, "step": 4038 }, { "epoch": 1.9007058823529412, "grad_norm": 0.35323241146793655, "learning_rate": 0.00027414825978032805, "loss": 0.3949, "step": 4039 }, { "epoch": 1.9011764705882352, "grad_norm": 0.13326320632679592, "learning_rate": 0.0002741236519902194, "loss": 0.1242, "step": 4040 }, { "epoch": 1.9016470588235292, "grad_norm": 0.3626175405459815, "learning_rate": 0.0002740990335993421, "loss": 0.3408, "step": 4041 }, { "epoch": 1.9021176470588235, "grad_norm": 0.4868558933310312, "learning_rate": 0.00027407440460979854, "loss": 0.3023, "step": 4042 }, { "epoch": 1.9025882352941177, "grad_norm": 0.613439506296773, "learning_rate": 0.0002740497650236923, "loss": 1.0301, "step": 4043 }, { "epoch": 1.903058823529412, "grad_norm": 0.24896615285572377, "learning_rate": 0.00027402511484312764, "loss": 0.2235, "step": 4044 }, { "epoch": 1.903529411764706, "grad_norm": 0.1996497427209648, "learning_rate": 0.00027400045407020986, "loss": 0.221, "step": 4045 }, { "epoch": 1.904, "grad_norm": 0.20158164685559107, "learning_rate": 0.00027397578270704504, "loss": 0.1712, "step": 4046 }, { "epoch": 1.9044705882352941, "grad_norm": 0.20809671018587605, "learning_rate": 0.0002739511007557403, "loss": 0.1598, "step": 4047 }, { "epoch": 1.9049411764705884, "grad_norm": 0.10826647971676871, "learning_rate": 0.00027392640821840365, "loss": 0.0472, "step": 4048 }, { "epoch": 1.9054117647058824, "grad_norm": 0.3634925413007315, "learning_rate": 0.0002739017050971438, "loss": 0.2734, "step": 4049 }, { "epoch": 1.9058823529411764, "grad_norm": 0.18403898760324627, "learning_rate": 0.0002738769913940706, "loss": 0.1175, "step": 4050 }, { "epoch": 1.9063529411764706, "grad_norm": 0.7856721691508716, "learning_rate": 0.00027385226711129477, "loss": 0.5653, "step": 4051 }, { "epoch": 1.9068235294117648, "grad_norm": 0.2725680647523748, "learning_rate": 0.00027382753225092777, "loss": 0.2084, "step": 4052 }, { "epoch": 1.9072941176470588, "grad_norm": 0.6072774052999887, "learning_rate": 0.0002738027868150822, "loss": 0.4457, "step": 4053 }, { "epoch": 1.9077647058823528, "grad_norm": 0.5561171930304964, "learning_rate": 0.0002737780308058714, "loss": 0.4564, "step": 4054 }, { "epoch": 1.908235294117647, "grad_norm": 0.281275966647675, "learning_rate": 0.00027375326422540957, "loss": 0.1687, "step": 4055 }, { "epoch": 1.9087058823529413, "grad_norm": 0.3246468051921719, "learning_rate": 0.000273728487075812, "loss": 0.3484, "step": 4056 }, { "epoch": 1.9091764705882353, "grad_norm": 0.3363149197656585, "learning_rate": 0.0002737036993591948, "loss": 0.328, "step": 4057 }, { "epoch": 1.9096470588235293, "grad_norm": 0.19492819483369872, "learning_rate": 0.0002736789010776749, "loss": 0.1276, "step": 4058 }, { "epoch": 1.9101176470588235, "grad_norm": 0.14896471980551607, "learning_rate": 0.00027365409223337016, "loss": 0.0946, "step": 4059 }, { "epoch": 1.9105882352941177, "grad_norm": 0.3120091219555176, "learning_rate": 0.0002736292728283994, "loss": 0.239, "step": 4060 }, { "epoch": 1.911058823529412, "grad_norm": 0.32593954325137053, "learning_rate": 0.0002736044428648824, "loss": 0.2726, "step": 4061 }, { "epoch": 1.911529411764706, "grad_norm": 0.32251491683896744, "learning_rate": 0.0002735796023449397, "loss": 0.2633, "step": 4062 }, { "epoch": 1.912, "grad_norm": 0.3987700240217207, "learning_rate": 0.00027355475127069275, "loss": 0.3473, "step": 4063 }, { "epoch": 1.9124705882352941, "grad_norm": 0.1992529105744146, "learning_rate": 0.0002735298896442641, "loss": 0.1251, "step": 4064 }, { "epoch": 1.9129411764705884, "grad_norm": 0.3456759814572113, "learning_rate": 0.0002735050174677769, "loss": 0.371, "step": 4065 }, { "epoch": 1.9134117647058824, "grad_norm": 0.3418422683435799, "learning_rate": 0.00027348013474335546, "loss": 0.2699, "step": 4066 }, { "epoch": 1.9138823529411764, "grad_norm": 0.3482855520524962, "learning_rate": 0.0002734552414731248, "loss": 0.4196, "step": 4067 }, { "epoch": 1.9143529411764706, "grad_norm": 0.3158721734197892, "learning_rate": 0.000273430337659211, "loss": 0.3035, "step": 4068 }, { "epoch": 1.9148235294117648, "grad_norm": 0.27668242362372214, "learning_rate": 0.0002734054233037409, "loss": 0.3084, "step": 4069 }, { "epoch": 1.9152941176470588, "grad_norm": 0.2471104612980433, "learning_rate": 0.00027338049840884245, "loss": 0.2737, "step": 4070 }, { "epoch": 1.9157647058823528, "grad_norm": 0.3423803035178889, "learning_rate": 0.00027335556297664417, "loss": 0.3962, "step": 4071 }, { "epoch": 1.916235294117647, "grad_norm": 0.3642001963720184, "learning_rate": 0.00027333061700927577, "loss": 0.3918, "step": 4072 }, { "epoch": 1.9167058823529413, "grad_norm": 0.08003825645155055, "learning_rate": 0.00027330566050886776, "loss": 0.0229, "step": 4073 }, { "epoch": 1.9171764705882353, "grad_norm": 0.2539713374829724, "learning_rate": 0.0002732806934775515, "loss": 0.314, "step": 4074 }, { "epoch": 1.9176470588235293, "grad_norm": 0.3813507447018477, "learning_rate": 0.0002732557159174594, "loss": 0.3946, "step": 4075 }, { "epoch": 1.9181176470588235, "grad_norm": 0.19031450071130132, "learning_rate": 0.0002732307278307246, "loss": 0.0823, "step": 4076 }, { "epoch": 1.9185882352941177, "grad_norm": 0.2541508982279964, "learning_rate": 0.0002732057292194811, "loss": 0.174, "step": 4077 }, { "epoch": 1.919058823529412, "grad_norm": 0.30929733505498563, "learning_rate": 0.000273180720085864, "loss": 0.2274, "step": 4078 }, { "epoch": 1.919529411764706, "grad_norm": 0.2528859637473037, "learning_rate": 0.0002731557004320093, "loss": 0.2767, "step": 4079 }, { "epoch": 1.92, "grad_norm": 0.4683055169976377, "learning_rate": 0.0002731306702600537, "loss": 0.3016, "step": 4080 }, { "epoch": 1.9204705882352942, "grad_norm": 0.3258946023363629, "learning_rate": 0.00027310562957213487, "loss": 0.3671, "step": 4081 }, { "epoch": 1.9209411764705884, "grad_norm": 0.41896900014020233, "learning_rate": 0.00027308057837039144, "loss": 0.3864, "step": 4082 }, { "epoch": 1.9214117647058824, "grad_norm": 0.17288667409591854, "learning_rate": 0.0002730555166569629, "loss": 0.1249, "step": 4083 }, { "epoch": 1.9218823529411764, "grad_norm": 0.36413945936891845, "learning_rate": 0.00027303044443398967, "loss": 0.2566, "step": 4084 }, { "epoch": 1.9223529411764706, "grad_norm": 0.07963046576391714, "learning_rate": 0.00027300536170361307, "loss": 0.0395, "step": 4085 }, { "epoch": 1.9228235294117648, "grad_norm": 0.4260732058643789, "learning_rate": 0.00027298026846797515, "loss": 0.4783, "step": 4086 }, { "epoch": 1.9232941176470588, "grad_norm": 0.34051676909594214, "learning_rate": 0.0002729551647292191, "loss": 0.3461, "step": 4087 }, { "epoch": 1.9237647058823528, "grad_norm": 0.5513443563945578, "learning_rate": 0.000272930050489489, "loss": 0.5381, "step": 4088 }, { "epoch": 1.924235294117647, "grad_norm": 0.19313242018971669, "learning_rate": 0.0002729049257509295, "loss": 0.0907, "step": 4089 }, { "epoch": 1.9247058823529413, "grad_norm": 0.031067634732937326, "learning_rate": 0.0002728797905156866, "loss": 0.0095, "step": 4090 }, { "epoch": 1.9251764705882353, "grad_norm": 0.3051317062990115, "learning_rate": 0.0002728546447859068, "loss": 0.3039, "step": 4091 }, { "epoch": 1.9256470588235293, "grad_norm": 0.33769921350378057, "learning_rate": 0.00027282948856373774, "loss": 0.2496, "step": 4092 }, { "epoch": 1.9261176470588235, "grad_norm": 0.3365060100786113, "learning_rate": 0.0002728043218513279, "loss": 0.2829, "step": 4093 }, { "epoch": 1.9265882352941177, "grad_norm": 0.1385780381870695, "learning_rate": 0.0002727791446508266, "loss": 0.0713, "step": 4094 }, { "epoch": 1.9270588235294117, "grad_norm": 0.3086108632626732, "learning_rate": 0.00027275395696438413, "loss": 0.2651, "step": 4095 }, { "epoch": 1.927529411764706, "grad_norm": 0.3704512314594238, "learning_rate": 0.0002727287587941516, "loss": 0.3362, "step": 4096 }, { "epoch": 1.928, "grad_norm": 0.33064054822610084, "learning_rate": 0.00027270355014228113, "loss": 0.346, "step": 4097 }, { "epoch": 1.9284705882352942, "grad_norm": 0.14377546815744247, "learning_rate": 0.00027267833101092564, "loss": 0.1037, "step": 4098 }, { "epoch": 1.9289411764705884, "grad_norm": 0.4576459567826521, "learning_rate": 0.0002726531014022389, "loss": 0.4121, "step": 4099 }, { "epoch": 1.9294117647058824, "grad_norm": 0.17589016841270946, "learning_rate": 0.00027262786131837573, "loss": 0.1126, "step": 4100 }, { "epoch": 1.9298823529411764, "grad_norm": 0.3525155627667901, "learning_rate": 0.00027260261076149166, "loss": 0.3475, "step": 4101 }, { "epoch": 1.9303529411764706, "grad_norm": 0.12688944624258167, "learning_rate": 0.00027257734973374335, "loss": 0.0611, "step": 4102 }, { "epoch": 1.9308235294117648, "grad_norm": 0.36002636607235744, "learning_rate": 0.0002725520782372881, "loss": 0.1995, "step": 4103 }, { "epoch": 1.9312941176470588, "grad_norm": 0.3484689288536441, "learning_rate": 0.00027252679627428423, "loss": 0.2667, "step": 4104 }, { "epoch": 1.9317647058823528, "grad_norm": 0.1541647264306251, "learning_rate": 0.00027250150384689094, "loss": 0.1075, "step": 4105 }, { "epoch": 1.932235294117647, "grad_norm": 0.19732584357554114, "learning_rate": 0.0002724762009572684, "loss": 0.1496, "step": 4106 }, { "epoch": 1.9327058823529413, "grad_norm": 0.27264060988487643, "learning_rate": 0.0002724508876075776, "loss": 0.2503, "step": 4107 }, { "epoch": 1.9331764705882353, "grad_norm": 0.34797068757515087, "learning_rate": 0.0002724255637999803, "loss": 0.3378, "step": 4108 }, { "epoch": 1.9336470588235293, "grad_norm": 0.24028710506123535, "learning_rate": 0.0002724002295366394, "loss": 0.23, "step": 4109 }, { "epoch": 1.9341176470588235, "grad_norm": 0.21569025712200363, "learning_rate": 0.0002723748848197185, "loss": 0.126, "step": 4110 }, { "epoch": 1.9345882352941177, "grad_norm": 0.2572985581279438, "learning_rate": 0.00027234952965138223, "loss": 0.2133, "step": 4111 }, { "epoch": 1.9350588235294117, "grad_norm": 0.2857038793751476, "learning_rate": 0.000272324164033796, "loss": 0.1688, "step": 4112 }, { "epoch": 1.9355294117647057, "grad_norm": 0.5365704677284551, "learning_rate": 0.0002722987879691261, "loss": 0.6275, "step": 4113 }, { "epoch": 1.936, "grad_norm": 0.24894659824338236, "learning_rate": 0.00027227340145953994, "loss": 0.2381, "step": 4114 }, { "epoch": 1.9364705882352942, "grad_norm": 0.4023679916947655, "learning_rate": 0.0002722480045072055, "loss": 0.4744, "step": 4115 }, { "epoch": 1.9369411764705884, "grad_norm": 0.1912898943931408, "learning_rate": 0.0002722225971142918, "loss": 0.1144, "step": 4116 }, { "epoch": 1.9374117647058824, "grad_norm": 0.18854683884528983, "learning_rate": 0.00027219717928296885, "loss": 0.1459, "step": 4117 }, { "epoch": 1.9378823529411764, "grad_norm": 0.3147416488825038, "learning_rate": 0.0002721717510154074, "loss": 0.2316, "step": 4118 }, { "epoch": 1.9383529411764706, "grad_norm": 0.08889131604112308, "learning_rate": 0.0002721463123137792, "loss": 0.0287, "step": 4119 }, { "epoch": 1.9388235294117648, "grad_norm": 0.2583475575544184, "learning_rate": 0.0002721208631802568, "loss": 0.1896, "step": 4120 }, { "epoch": 1.9392941176470588, "grad_norm": 0.20071310064296957, "learning_rate": 0.00027209540361701365, "loss": 0.1071, "step": 4121 }, { "epoch": 1.9397647058823528, "grad_norm": 0.15474110666654672, "learning_rate": 0.0002720699336262242, "loss": 0.0851, "step": 4122 }, { "epoch": 1.940235294117647, "grad_norm": 0.2790306095669235, "learning_rate": 0.00027204445321006365, "loss": 0.2676, "step": 4123 }, { "epoch": 1.9407058823529413, "grad_norm": 0.2171700822064339, "learning_rate": 0.00027201896237070815, "loss": 0.2053, "step": 4124 }, { "epoch": 1.9411764705882353, "grad_norm": 0.32339960847917293, "learning_rate": 0.0002719934611103348, "loss": 0.2533, "step": 4125 }, { "epoch": 1.9416470588235293, "grad_norm": 0.2264159477632554, "learning_rate": 0.00027196794943112147, "loss": 0.1745, "step": 4126 }, { "epoch": 1.9421176470588235, "grad_norm": 0.3003025398263254, "learning_rate": 0.000271942427335247, "loss": 0.2302, "step": 4127 }, { "epoch": 1.9425882352941177, "grad_norm": 0.14467736318540714, "learning_rate": 0.00027191689482489114, "loss": 0.0969, "step": 4128 }, { "epoch": 1.9430588235294117, "grad_norm": 0.21016489756717324, "learning_rate": 0.0002718913519022344, "loss": 0.1704, "step": 4129 }, { "epoch": 1.9435294117647057, "grad_norm": 0.37407253343285357, "learning_rate": 0.00027186579856945843, "loss": 0.376, "step": 4130 }, { "epoch": 1.944, "grad_norm": 0.24005850879623666, "learning_rate": 0.00027184023482874545, "loss": 0.231, "step": 4131 }, { "epoch": 1.9444705882352942, "grad_norm": 0.27372837655108634, "learning_rate": 0.00027181466068227885, "loss": 0.146, "step": 4132 }, { "epoch": 1.9449411764705884, "grad_norm": 0.352869993905262, "learning_rate": 0.00027178907613224264, "loss": 0.3696, "step": 4133 }, { "epoch": 1.9454117647058824, "grad_norm": 0.2451378949510048, "learning_rate": 0.00027176348118082203, "loss": 0.2286, "step": 4134 }, { "epoch": 1.9458823529411764, "grad_norm": 0.4419344724903381, "learning_rate": 0.0002717378758302029, "loss": 0.5332, "step": 4135 }, { "epoch": 1.9463529411764706, "grad_norm": 0.546444484903018, "learning_rate": 0.000271712260082572, "loss": 0.535, "step": 4136 }, { "epoch": 1.9468235294117648, "grad_norm": 0.3316147130834413, "learning_rate": 0.0002716866339401171, "loss": 0.2772, "step": 4137 }, { "epoch": 1.9472941176470588, "grad_norm": 0.22731315853245648, "learning_rate": 0.0002716609974050268, "loss": 0.2349, "step": 4138 }, { "epoch": 1.9477647058823528, "grad_norm": 0.5698150867810371, "learning_rate": 0.0002716353504794906, "loss": 0.5953, "step": 4139 }, { "epoch": 1.948235294117647, "grad_norm": 0.20625426324859547, "learning_rate": 0.0002716096931656988, "loss": 0.1358, "step": 4140 }, { "epoch": 1.9487058823529413, "grad_norm": 0.5457124693525073, "learning_rate": 0.0002715840254658427, "loss": 0.4541, "step": 4141 }, { "epoch": 1.9491764705882353, "grad_norm": 0.403844476173935, "learning_rate": 0.0002715583473821145, "loss": 0.3745, "step": 4142 }, { "epoch": 1.9496470588235293, "grad_norm": 0.24092970802627978, "learning_rate": 0.0002715326589167072, "loss": 0.192, "step": 4143 }, { "epoch": 1.9501176470588235, "grad_norm": 0.3343303629920148, "learning_rate": 0.0002715069600718147, "loss": 0.2625, "step": 4144 }, { "epoch": 1.9505882352941177, "grad_norm": 0.35501626365228545, "learning_rate": 0.0002714812508496319, "loss": 0.369, "step": 4145 }, { "epoch": 1.9510588235294117, "grad_norm": 0.3387456532853103, "learning_rate": 0.0002714555312523543, "loss": 0.2997, "step": 4146 }, { "epoch": 1.9515294117647057, "grad_norm": 0.4657460012269652, "learning_rate": 0.00027142980128217864, "loss": 0.5269, "step": 4147 }, { "epoch": 1.952, "grad_norm": 0.28240945242190846, "learning_rate": 0.00027140406094130234, "loss": 0.2513, "step": 4148 }, { "epoch": 1.9524705882352942, "grad_norm": 0.29467041017893836, "learning_rate": 0.0002713783102319237, "loss": 0.2557, "step": 4149 }, { "epoch": 1.9529411764705882, "grad_norm": 0.38658462713484487, "learning_rate": 0.0002713525491562421, "loss": 0.286, "step": 4150 }, { "epoch": 1.9534117647058824, "grad_norm": 0.2485900233792381, "learning_rate": 0.0002713267777164575, "loss": 0.2356, "step": 4151 }, { "epoch": 1.9538823529411764, "grad_norm": 0.2696689737099364, "learning_rate": 0.00027130099591477104, "loss": 0.1667, "step": 4152 }, { "epoch": 1.9543529411764706, "grad_norm": 0.6160708097124733, "learning_rate": 0.0002712752037533845, "loss": 0.5303, "step": 4153 }, { "epoch": 1.9548235294117648, "grad_norm": 0.593367493495454, "learning_rate": 0.00027124940123450075, "loss": 0.5672, "step": 4154 }, { "epoch": 1.9552941176470588, "grad_norm": 0.5499811932244415, "learning_rate": 0.00027122358836032337, "loss": 0.5511, "step": 4155 }, { "epoch": 1.9557647058823528, "grad_norm": 0.11556031557363211, "learning_rate": 0.000271197765133057, "loss": 0.058, "step": 4156 }, { "epoch": 1.956235294117647, "grad_norm": 0.20384378073503862, "learning_rate": 0.00027117193155490695, "loss": 0.1463, "step": 4157 }, { "epoch": 1.9567058823529413, "grad_norm": 0.6061241256159734, "learning_rate": 0.00027114608762807965, "loss": 0.3794, "step": 4158 }, { "epoch": 1.9571764705882353, "grad_norm": 0.529115582101806, "learning_rate": 0.0002711202333547822, "loss": 0.5059, "step": 4159 }, { "epoch": 1.9576470588235293, "grad_norm": 0.27374585385922257, "learning_rate": 0.00027109436873722277, "loss": 0.3064, "step": 4160 }, { "epoch": 1.9581176470588235, "grad_norm": 0.16723958376802614, "learning_rate": 0.00027106849377761023, "loss": 0.079, "step": 4161 }, { "epoch": 1.9585882352941177, "grad_norm": 0.25924823108228034, "learning_rate": 0.00027104260847815457, "loss": 0.2527, "step": 4162 }, { "epoch": 1.9590588235294117, "grad_norm": 0.3567355888592774, "learning_rate": 0.0002710167128410664, "loss": 0.3364, "step": 4163 }, { "epoch": 1.9595294117647057, "grad_norm": 0.38937926540833023, "learning_rate": 0.00027099080686855735, "loss": 0.368, "step": 4164 }, { "epoch": 1.96, "grad_norm": 0.4549145562599691, "learning_rate": 0.00027096489056283995, "loss": 0.4376, "step": 4165 }, { "epoch": 1.9604705882352942, "grad_norm": 0.24791767957533845, "learning_rate": 0.00027093896392612753, "loss": 0.2632, "step": 4166 }, { "epoch": 1.9609411764705882, "grad_norm": 0.3603581809317886, "learning_rate": 0.0002709130269606345, "loss": 0.4373, "step": 4167 }, { "epoch": 1.9614117647058822, "grad_norm": 0.37997312176734144, "learning_rate": 0.00027088707966857576, "loss": 0.4097, "step": 4168 }, { "epoch": 1.9618823529411764, "grad_norm": 0.34241942783693996, "learning_rate": 0.0002708611220521675, "loss": 0.264, "step": 4169 }, { "epoch": 1.9623529411764706, "grad_norm": 0.30398055912206867, "learning_rate": 0.0002708351541136267, "loss": 0.2646, "step": 4170 }, { "epoch": 1.9628235294117649, "grad_norm": 0.33726032041776993, "learning_rate": 0.0002708091758551709, "loss": 0.381, "step": 4171 }, { "epoch": 1.9632941176470589, "grad_norm": 0.20179811443663478, "learning_rate": 0.000270783187279019, "loss": 0.1527, "step": 4172 }, { "epoch": 1.9637647058823529, "grad_norm": 0.5447925905908442, "learning_rate": 0.00027075718838739043, "loss": 0.5514, "step": 4173 }, { "epoch": 1.964235294117647, "grad_norm": 0.07792853350319305, "learning_rate": 0.0002707311791825057, "loss": 0.0349, "step": 4174 }, { "epoch": 1.9647058823529413, "grad_norm": 0.47862234421876054, "learning_rate": 0.00027070515966658604, "loss": 0.5415, "step": 4175 }, { "epoch": 1.9651764705882353, "grad_norm": 0.26799269471596054, "learning_rate": 0.00027067912984185376, "loss": 0.0898, "step": 4176 }, { "epoch": 1.9656470588235293, "grad_norm": 0.1946841804005412, "learning_rate": 0.0002706530897105318, "loss": 0.0939, "step": 4177 }, { "epoch": 1.9661176470588235, "grad_norm": 0.542745301250271, "learning_rate": 0.0002706270392748442, "loss": 0.4047, "step": 4178 }, { "epoch": 1.9665882352941177, "grad_norm": 0.30161481985246186, "learning_rate": 0.00027060097853701576, "loss": 0.3228, "step": 4179 }, { "epoch": 1.9670588235294117, "grad_norm": 0.28532261189663644, "learning_rate": 0.0002705749074992722, "loss": 0.2109, "step": 4180 }, { "epoch": 1.9675294117647057, "grad_norm": 0.07718628331749912, "learning_rate": 0.00027054882616384015, "loss": 0.0329, "step": 4181 }, { "epoch": 1.968, "grad_norm": 0.4517377095840119, "learning_rate": 0.0002705227345329471, "loss": 0.4431, "step": 4182 }, { "epoch": 1.9684705882352942, "grad_norm": 0.231455942180711, "learning_rate": 0.00027049663260882124, "loss": 0.2042, "step": 4183 }, { "epoch": 1.9689411764705882, "grad_norm": 0.36952440583745727, "learning_rate": 0.00027047052039369204, "loss": 0.2796, "step": 4184 }, { "epoch": 1.9694117647058822, "grad_norm": 0.22401075766004824, "learning_rate": 0.0002704443978897894, "loss": 0.2072, "step": 4185 }, { "epoch": 1.9698823529411764, "grad_norm": 0.23093655535846888, "learning_rate": 0.0002704182650993445, "loss": 0.14, "step": 4186 }, { "epoch": 1.9703529411764706, "grad_norm": 0.3824555189428466, "learning_rate": 0.00027039212202458905, "loss": 0.378, "step": 4187 }, { "epoch": 1.9708235294117649, "grad_norm": 0.249906996461184, "learning_rate": 0.0002703659686677559, "loss": 0.0969, "step": 4188 }, { "epoch": 1.9712941176470589, "grad_norm": 0.18520287505727454, "learning_rate": 0.00027033980503107856, "loss": 0.1555, "step": 4189 }, { "epoch": 1.9717647058823529, "grad_norm": 0.9208973975481995, "learning_rate": 0.00027031363111679164, "loss": 0.5487, "step": 4190 }, { "epoch": 1.972235294117647, "grad_norm": 0.28762981474760485, "learning_rate": 0.0002702874469271305, "loss": 0.2078, "step": 4191 }, { "epoch": 1.9727058823529413, "grad_norm": 0.4125592709489132, "learning_rate": 0.00027026125246433144, "loss": 0.2878, "step": 4192 }, { "epoch": 1.9731764705882353, "grad_norm": 0.6814034690650181, "learning_rate": 0.0002702350477306315, "loss": 0.5395, "step": 4193 }, { "epoch": 1.9736470588235293, "grad_norm": 0.2776275804347747, "learning_rate": 0.0002702088327282687, "loss": 0.1544, "step": 4194 }, { "epoch": 1.9741176470588235, "grad_norm": 0.2887306525981737, "learning_rate": 0.00027018260745948196, "loss": 0.1842, "step": 4195 }, { "epoch": 1.9745882352941178, "grad_norm": 0.22341551627450382, "learning_rate": 0.0002701563719265111, "loss": 0.0954, "step": 4196 }, { "epoch": 1.9750588235294118, "grad_norm": 0.4929304328863682, "learning_rate": 0.0002701301261315967, "loss": 0.3369, "step": 4197 }, { "epoch": 1.9755294117647058, "grad_norm": 0.32372248935363057, "learning_rate": 0.0002701038700769803, "loss": 0.2166, "step": 4198 }, { "epoch": 1.976, "grad_norm": 0.3472683989023232, "learning_rate": 0.00027007760376490425, "loss": 0.389, "step": 4199 }, { "epoch": 1.9764705882352942, "grad_norm": 0.44151632449390005, "learning_rate": 0.0002700513271976119, "loss": 0.5206, "step": 4200 }, { "epoch": 1.9769411764705882, "grad_norm": 0.2261725418897128, "learning_rate": 0.0002700250403773473, "loss": 0.1786, "step": 4201 }, { "epoch": 1.9774117647058822, "grad_norm": 0.7745069162746625, "learning_rate": 0.0002699987433063556, "loss": 0.4358, "step": 4202 }, { "epoch": 1.9778823529411764, "grad_norm": 0.2855372479982766, "learning_rate": 0.0002699724359868826, "loss": 0.3629, "step": 4203 }, { "epoch": 1.9783529411764706, "grad_norm": 0.17727926553925094, "learning_rate": 0.0002699461184211751, "loss": 0.2677, "step": 4204 }, { "epoch": 1.9788235294117649, "grad_norm": 0.33831770658666577, "learning_rate": 0.00026991979061148077, "loss": 0.3596, "step": 4205 }, { "epoch": 1.9792941176470589, "grad_norm": 0.1221318521834891, "learning_rate": 0.00026989345256004814, "loss": 0.0786, "step": 4206 }, { "epoch": 1.9797647058823529, "grad_norm": 0.2994962870697344, "learning_rate": 0.00026986710426912656, "loss": 0.229, "step": 4207 }, { "epoch": 1.980235294117647, "grad_norm": 0.19524370079947664, "learning_rate": 0.0002698407457409663, "loss": 0.1446, "step": 4208 }, { "epoch": 1.9807058823529413, "grad_norm": 0.2579512104230562, "learning_rate": 0.00026981437697781854, "loss": 0.1361, "step": 4209 }, { "epoch": 1.9811764705882353, "grad_norm": 0.3401002973293463, "learning_rate": 0.0002697879979819354, "loss": 0.3718, "step": 4210 }, { "epoch": 1.9816470588235293, "grad_norm": 0.18082627048229888, "learning_rate": 0.0002697616087555696, "loss": 0.1612, "step": 4211 }, { "epoch": 1.9821176470588235, "grad_norm": 0.26887205133851066, "learning_rate": 0.00026973520930097494, "loss": 0.373, "step": 4212 }, { "epoch": 1.9825882352941178, "grad_norm": 0.4304724536144263, "learning_rate": 0.0002697087996204062, "loss": 0.5601, "step": 4213 }, { "epoch": 1.9830588235294118, "grad_norm": 0.3864277300202455, "learning_rate": 0.0002696823797161187, "loss": 0.1095, "step": 4214 }, { "epoch": 1.9835294117647058, "grad_norm": 0.26376895139217044, "learning_rate": 0.00026965594959036906, "loss": 0.2661, "step": 4215 }, { "epoch": 1.984, "grad_norm": 0.43610334154627206, "learning_rate": 0.00026962950924541435, "loss": 0.1284, "step": 4216 }, { "epoch": 1.9844705882352942, "grad_norm": 0.4947759148178746, "learning_rate": 0.0002696030586835128, "loss": 0.3562, "step": 4217 }, { "epoch": 1.9849411764705882, "grad_norm": 0.46700675222979754, "learning_rate": 0.00026957659790692335, "loss": 0.3854, "step": 4218 }, { "epoch": 1.9854117647058822, "grad_norm": 0.09162711307028067, "learning_rate": 0.0002695501269179059, "loss": 0.06, "step": 4219 }, { "epoch": 1.9858823529411764, "grad_norm": 0.19593624736300053, "learning_rate": 0.00026952364571872125, "loss": 0.1057, "step": 4220 }, { "epoch": 1.9863529411764707, "grad_norm": 0.5096296843387168, "learning_rate": 0.000269497154311631, "loss": 0.2095, "step": 4221 }, { "epoch": 1.9868235294117647, "grad_norm": 0.482500878266507, "learning_rate": 0.00026947065269889765, "loss": 0.3756, "step": 4222 }, { "epoch": 1.9872941176470589, "grad_norm": 0.32946661482086126, "learning_rate": 0.00026944414088278453, "loss": 0.2216, "step": 4223 }, { "epoch": 1.9877647058823529, "grad_norm": 0.46829128465315095, "learning_rate": 0.00026941761886555595, "loss": 0.3897, "step": 4224 }, { "epoch": 1.988235294117647, "grad_norm": 0.4090305543616981, "learning_rate": 0.00026939108664947694, "loss": 0.4708, "step": 4225 }, { "epoch": 1.9887058823529413, "grad_norm": 0.34860663199417563, "learning_rate": 0.0002693645442368135, "loss": 0.2863, "step": 4226 }, { "epoch": 1.9891764705882353, "grad_norm": 0.27636807069266484, "learning_rate": 0.0002693379916298326, "loss": 0.313, "step": 4227 }, { "epoch": 1.9896470588235293, "grad_norm": 0.21863818663476053, "learning_rate": 0.00026931142883080176, "loss": 0.0893, "step": 4228 }, { "epoch": 1.9901176470588235, "grad_norm": 0.29328257911118694, "learning_rate": 0.00026928485584198973, "loss": 0.1688, "step": 4229 }, { "epoch": 1.9905882352941178, "grad_norm": 0.260307367381377, "learning_rate": 0.00026925827266566595, "loss": 0.2479, "step": 4230 }, { "epoch": 1.9910588235294118, "grad_norm": 0.3174985304077367, "learning_rate": 0.0002692316793041007, "loss": 0.2232, "step": 4231 }, { "epoch": 1.9915294117647058, "grad_norm": 0.36144255562910227, "learning_rate": 0.0002692050757595652, "loss": 0.3661, "step": 4232 }, { "epoch": 1.992, "grad_norm": 0.33971674906969906, "learning_rate": 0.00026917846203433155, "loss": 0.3422, "step": 4233 }, { "epoch": 1.9924705882352942, "grad_norm": 0.3707755342493682, "learning_rate": 0.00026915183813067264, "loss": 0.5177, "step": 4234 }, { "epoch": 1.9929411764705882, "grad_norm": 0.2527144550671462, "learning_rate": 0.0002691252040508623, "loss": 0.2188, "step": 4235 }, { "epoch": 1.9934117647058822, "grad_norm": 0.3036151566484735, "learning_rate": 0.0002690985597971753, "loss": 0.4608, "step": 4236 }, { "epoch": 1.9938823529411764, "grad_norm": 0.4604959597318192, "learning_rate": 0.00026907190537188707, "loss": 0.4259, "step": 4237 }, { "epoch": 1.9943529411764707, "grad_norm": 0.24893708178632196, "learning_rate": 0.0002690452407772741, "loss": 0.248, "step": 4238 }, { "epoch": 1.9948235294117647, "grad_norm": 0.09809987158016985, "learning_rate": 0.0002690185660156136, "loss": 0.0517, "step": 4239 }, { "epoch": 1.9952941176470587, "grad_norm": 0.428224989226972, "learning_rate": 0.00026899188108918387, "loss": 0.3538, "step": 4240 }, { "epoch": 1.9957647058823529, "grad_norm": 0.12200632454213832, "learning_rate": 0.00026896518600026377, "loss": 0.0659, "step": 4241 }, { "epoch": 1.996235294117647, "grad_norm": 0.2871501047475606, "learning_rate": 0.00026893848075113327, "loss": 0.2652, "step": 4242 }, { "epoch": 1.9967058823529413, "grad_norm": 0.3518602438161703, "learning_rate": 0.00026891176534407315, "loss": 0.2941, "step": 4243 }, { "epoch": 1.9971764705882353, "grad_norm": 0.49995739193606775, "learning_rate": 0.00026888503978136506, "loss": 0.4433, "step": 4244 }, { "epoch": 1.9976470588235293, "grad_norm": 0.03218552173751255, "learning_rate": 0.00026885830406529136, "loss": 0.0067, "step": 4245 }, { "epoch": 1.9981176470588236, "grad_norm": 0.4519650717133047, "learning_rate": 0.0002688315581981355, "loss": 0.3528, "step": 4246 }, { "epoch": 1.9985882352941178, "grad_norm": 0.17765174728798858, "learning_rate": 0.0002688048021821817, "loss": 0.1256, "step": 4247 }, { "epoch": 1.9990588235294118, "grad_norm": 0.2380007002621949, "learning_rate": 0.00026877803601971516, "loss": 0.122, "step": 4248 }, { "epoch": 1.9995294117647058, "grad_norm": 0.1814300611525379, "learning_rate": 0.00026875125971302167, "loss": 0.1499, "step": 4249 }, { "epoch": 2.0, "grad_norm": 0.46948607127207836, "learning_rate": 0.0002687244732643881, "loss": 0.422, "step": 4250 }, { "epoch": 2.0, "eval_loss": 0.2750994563102722, "eval_runtime": 65.3608, "eval_samples_per_second": 11.475, "eval_steps_per_second": 5.737, "step": 4250 }, { "epoch": 2.0004705882352942, "grad_norm": 0.3843992095719931, "learning_rate": 0.0002686976766761022, "loss": 0.3253, "step": 4251 }, { "epoch": 2.0009411764705884, "grad_norm": 0.15717763425421566, "learning_rate": 0.0002686708699504525, "loss": 0.0942, "step": 4252 }, { "epoch": 2.0014117647058822, "grad_norm": 0.29836867761726227, "learning_rate": 0.0002686440530897284, "loss": 0.3274, "step": 4253 }, { "epoch": 2.0018823529411764, "grad_norm": 0.33895397826342827, "learning_rate": 0.00026861722609622027, "loss": 0.2628, "step": 4254 }, { "epoch": 2.0023529411764707, "grad_norm": 0.3614660065403509, "learning_rate": 0.0002685903889722192, "loss": 0.4019, "step": 4255 }, { "epoch": 2.002823529411765, "grad_norm": 0.19182308232480044, "learning_rate": 0.00026856354172001724, "loss": 0.1467, "step": 4256 }, { "epoch": 2.0032941176470587, "grad_norm": 0.44588378944165646, "learning_rate": 0.0002685366843419073, "loss": 0.2846, "step": 4257 }, { "epoch": 2.003764705882353, "grad_norm": 0.45860564201036347, "learning_rate": 0.000268509816840183, "loss": 0.5542, "step": 4258 }, { "epoch": 2.004235294117647, "grad_norm": 0.22505054410042616, "learning_rate": 0.00026848293921713913, "loss": 0.1725, "step": 4259 }, { "epoch": 2.0047058823529413, "grad_norm": 0.4398135129730654, "learning_rate": 0.0002684560514750711, "loss": 0.3598, "step": 4260 }, { "epoch": 2.005176470588235, "grad_norm": 0.29298701245494924, "learning_rate": 0.00026842915361627523, "loss": 0.2491, "step": 4261 }, { "epoch": 2.0056470588235293, "grad_norm": 0.2638337932732402, "learning_rate": 0.0002684022456430487, "loss": 0.2845, "step": 4262 }, { "epoch": 2.0061176470588236, "grad_norm": 0.1548565409503225, "learning_rate": 0.0002683753275576897, "loss": 0.1354, "step": 4263 }, { "epoch": 2.006588235294118, "grad_norm": 0.29475831296380317, "learning_rate": 0.0002683483993624971, "loss": 0.3262, "step": 4264 }, { "epoch": 2.0070588235294116, "grad_norm": 0.1544533077914857, "learning_rate": 0.0002683214610597706, "loss": 0.0531, "step": 4265 }, { "epoch": 2.007529411764706, "grad_norm": 0.3495610978018076, "learning_rate": 0.0002682945126518111, "loss": 0.1561, "step": 4266 }, { "epoch": 2.008, "grad_norm": 0.4303895442940034, "learning_rate": 0.00026826755414091993, "loss": 0.4399, "step": 4267 }, { "epoch": 2.0084705882352942, "grad_norm": 0.38320842547083966, "learning_rate": 0.00026824058552939954, "loss": 0.3278, "step": 4268 }, { "epoch": 2.0089411764705885, "grad_norm": 0.49185056849676345, "learning_rate": 0.00026821360681955316, "loss": 0.4632, "step": 4269 }, { "epoch": 2.0094117647058822, "grad_norm": 0.27426893264737584, "learning_rate": 0.00026818661801368494, "loss": 0.2784, "step": 4270 }, { "epoch": 2.0098823529411765, "grad_norm": 0.12855426289777797, "learning_rate": 0.0002681596191140998, "loss": 0.0638, "step": 4271 }, { "epoch": 2.0103529411764707, "grad_norm": 0.3952126419904165, "learning_rate": 0.00026813261012310365, "loss": 0.237, "step": 4272 }, { "epoch": 2.010823529411765, "grad_norm": 0.5748089500997712, "learning_rate": 0.00026810559104300315, "loss": 0.313, "step": 4273 }, { "epoch": 2.0112941176470587, "grad_norm": 0.11541952501786881, "learning_rate": 0.0002680785618761059, "loss": 0.0621, "step": 4274 }, { "epoch": 2.011764705882353, "grad_norm": 0.17491136302868426, "learning_rate": 0.00026805152262472025, "loss": 0.1147, "step": 4275 }, { "epoch": 2.012235294117647, "grad_norm": 0.46988503848050617, "learning_rate": 0.00026802447329115557, "loss": 0.4129, "step": 4276 }, { "epoch": 2.0127058823529413, "grad_norm": 0.12158352134164561, "learning_rate": 0.00026799741387772193, "loss": 0.0804, "step": 4277 }, { "epoch": 2.013176470588235, "grad_norm": 0.4744488552419081, "learning_rate": 0.0002679703443867304, "loss": 0.3928, "step": 4278 }, { "epoch": 2.0136470588235293, "grad_norm": 0.2711285661110369, "learning_rate": 0.0002679432648204928, "loss": 0.2237, "step": 4279 }, { "epoch": 2.0141176470588236, "grad_norm": 0.23815750976870648, "learning_rate": 0.00026791617518132186, "loss": 0.2557, "step": 4280 }, { "epoch": 2.014588235294118, "grad_norm": 0.31519557467640374, "learning_rate": 0.00026788907547153126, "loss": 0.1896, "step": 4281 }, { "epoch": 2.0150588235294116, "grad_norm": 0.2774499814946569, "learning_rate": 0.00026786196569343525, "loss": 0.2512, "step": 4282 }, { "epoch": 2.015529411764706, "grad_norm": 0.16417336651147998, "learning_rate": 0.0002678348458493494, "loss": 0.0686, "step": 4283 }, { "epoch": 2.016, "grad_norm": 0.2720735597929187, "learning_rate": 0.00026780771594158966, "loss": 0.1988, "step": 4284 }, { "epoch": 2.0164705882352942, "grad_norm": 0.12094734250206755, "learning_rate": 0.0002677805759724731, "loss": 0.0667, "step": 4285 }, { "epoch": 2.016941176470588, "grad_norm": 0.3949473476246422, "learning_rate": 0.0002677534259443177, "loss": 0.2792, "step": 4286 }, { "epoch": 2.0174117647058822, "grad_norm": 0.4205581153245379, "learning_rate": 0.00026772626585944213, "loss": 0.3941, "step": 4287 }, { "epoch": 2.0178823529411765, "grad_norm": 0.26236222975592377, "learning_rate": 0.00026769909572016605, "loss": 0.1788, "step": 4288 }, { "epoch": 2.0183529411764707, "grad_norm": 0.17424757294672413, "learning_rate": 0.0002676719155288098, "loss": 0.0734, "step": 4289 }, { "epoch": 2.018823529411765, "grad_norm": 0.2630975137583509, "learning_rate": 0.00026764472528769485, "loss": 0.2328, "step": 4290 }, { "epoch": 2.0192941176470587, "grad_norm": 0.3200663855661994, "learning_rate": 0.0002676175249991433, "loss": 0.2626, "step": 4291 }, { "epoch": 2.019764705882353, "grad_norm": 0.17090262564017625, "learning_rate": 0.0002675903146654782, "loss": 0.1069, "step": 4292 }, { "epoch": 2.020235294117647, "grad_norm": 0.5253751653352035, "learning_rate": 0.00026756309428902343, "loss": 0.1598, "step": 4293 }, { "epoch": 2.0207058823529414, "grad_norm": 0.5438592467119726, "learning_rate": 0.0002675358638721038, "loss": 0.3541, "step": 4294 }, { "epoch": 2.021176470588235, "grad_norm": 0.3727441919191019, "learning_rate": 0.0002675086234170449, "loss": 0.1953, "step": 4295 }, { "epoch": 2.0216470588235294, "grad_norm": 0.18552132947791697, "learning_rate": 0.0002674813729261731, "loss": 0.1373, "step": 4296 }, { "epoch": 2.0221176470588236, "grad_norm": 0.21009704743406127, "learning_rate": 0.0002674541124018159, "loss": 0.1814, "step": 4297 }, { "epoch": 2.022588235294118, "grad_norm": 0.14110791203607773, "learning_rate": 0.00026742684184630135, "loss": 0.11, "step": 4298 }, { "epoch": 2.0230588235294116, "grad_norm": 0.12461653217352885, "learning_rate": 0.00026739956126195855, "loss": 0.063, "step": 4299 }, { "epoch": 2.023529411764706, "grad_norm": 0.2961386244469217, "learning_rate": 0.0002673722706511174, "loss": 0.1143, "step": 4300 }, { "epoch": 2.024, "grad_norm": 0.44892838018494485, "learning_rate": 0.00026734497001610854, "loss": 0.483, "step": 4301 }, { "epoch": 2.0244705882352942, "grad_norm": 0.4271594506014568, "learning_rate": 0.00026731765935926374, "loss": 0.3769, "step": 4302 }, { "epoch": 2.024941176470588, "grad_norm": 0.1814862939962269, "learning_rate": 0.00026729033868291537, "loss": 0.1711, "step": 4303 }, { "epoch": 2.0254117647058822, "grad_norm": 0.23623611547114678, "learning_rate": 0.00026726300798939675, "loss": 0.1924, "step": 4304 }, { "epoch": 2.0258823529411765, "grad_norm": 0.3104438606062889, "learning_rate": 0.0002672356672810421, "loss": 0.1804, "step": 4305 }, { "epoch": 2.0263529411764707, "grad_norm": 0.40259465598850397, "learning_rate": 0.00026720831656018644, "loss": 0.3021, "step": 4306 }, { "epoch": 2.026823529411765, "grad_norm": 0.1228902545337613, "learning_rate": 0.0002671809558291656, "loss": 0.0431, "step": 4307 }, { "epoch": 2.0272941176470587, "grad_norm": 0.14848925567145066, "learning_rate": 0.00026715358509031643, "loss": 0.0576, "step": 4308 }, { "epoch": 2.027764705882353, "grad_norm": 0.30734047469821135, "learning_rate": 0.00026712620434597636, "loss": 0.2137, "step": 4309 }, { "epoch": 2.028235294117647, "grad_norm": 0.3699133410077011, "learning_rate": 0.00026709881359848406, "loss": 0.2887, "step": 4310 }, { "epoch": 2.0287058823529414, "grad_norm": 0.4093029707421381, "learning_rate": 0.0002670714128501786, "loss": 0.424, "step": 4311 }, { "epoch": 2.029176470588235, "grad_norm": 0.3291836688089953, "learning_rate": 0.0002670440021034003, "loss": 0.3031, "step": 4312 }, { "epoch": 2.0296470588235294, "grad_norm": 0.24333114790812654, "learning_rate": 0.00026701658136049015, "loss": 0.1741, "step": 4313 }, { "epoch": 2.0301176470588236, "grad_norm": 0.4002514281027978, "learning_rate": 0.00026698915062378994, "loss": 0.3131, "step": 4314 }, { "epoch": 2.030588235294118, "grad_norm": 0.7098266678804751, "learning_rate": 0.00026696170989564245, "loss": 0.8648, "step": 4315 }, { "epoch": 2.0310588235294116, "grad_norm": 0.24597898607239424, "learning_rate": 0.0002669342591783913, "loss": 0.232, "step": 4316 }, { "epoch": 2.031529411764706, "grad_norm": 0.3557153264817093, "learning_rate": 0.00026690679847438077, "loss": 0.2866, "step": 4317 }, { "epoch": 2.032, "grad_norm": 0.13677445914691386, "learning_rate": 0.0002668793277859563, "loss": 0.081, "step": 4318 }, { "epoch": 2.0324705882352943, "grad_norm": 0.289311938280233, "learning_rate": 0.0002668518471154639, "loss": 0.2294, "step": 4319 }, { "epoch": 2.032941176470588, "grad_norm": 0.6538844229682851, "learning_rate": 0.0002668243564652506, "loss": 0.528, "step": 4320 }, { "epoch": 2.0334117647058823, "grad_norm": 0.34653060320909995, "learning_rate": 0.0002667968558376642, "loss": 0.3781, "step": 4321 }, { "epoch": 2.0338823529411765, "grad_norm": 0.5767533591017098, "learning_rate": 0.0002667693452350535, "loss": 0.5873, "step": 4322 }, { "epoch": 2.0343529411764707, "grad_norm": 0.30607385352845545, "learning_rate": 0.0002667418246597679, "loss": 0.3484, "step": 4323 }, { "epoch": 2.034823529411765, "grad_norm": 0.11160613430409697, "learning_rate": 0.0002667142941141579, "loss": 0.0383, "step": 4324 }, { "epoch": 2.0352941176470587, "grad_norm": 0.3078357480229034, "learning_rate": 0.0002666867536005747, "loss": 0.3754, "step": 4325 }, { "epoch": 2.035764705882353, "grad_norm": 0.2280185916823153, "learning_rate": 0.00026665920312137045, "loss": 0.1615, "step": 4326 }, { "epoch": 2.036235294117647, "grad_norm": 0.32642427267351465, "learning_rate": 0.000266631642678898, "loss": 0.3196, "step": 4327 }, { "epoch": 2.0367058823529414, "grad_norm": 0.34415484975946253, "learning_rate": 0.00026660407227551113, "loss": 0.2883, "step": 4328 }, { "epoch": 2.037176470588235, "grad_norm": 0.33942549580988984, "learning_rate": 0.0002665764919135646, "loss": 0.2872, "step": 4329 }, { "epoch": 2.0376470588235294, "grad_norm": 0.23904404658501652, "learning_rate": 0.00026654890159541386, "loss": 0.1864, "step": 4330 }, { "epoch": 2.0381176470588236, "grad_norm": 0.772554056141936, "learning_rate": 0.0002665213013234153, "loss": 0.6749, "step": 4331 }, { "epoch": 2.038588235294118, "grad_norm": 0.2696971240399901, "learning_rate": 0.00026649369109992596, "loss": 0.1742, "step": 4332 }, { "epoch": 2.0390588235294116, "grad_norm": 0.19284893958033114, "learning_rate": 0.0002664660709273041, "loss": 0.0855, "step": 4333 }, { "epoch": 2.039529411764706, "grad_norm": 0.15063679673965225, "learning_rate": 0.0002664384408079085, "loss": 0.0435, "step": 4334 }, { "epoch": 2.04, "grad_norm": 0.17352740478272644, "learning_rate": 0.0002664108007440989, "loss": 0.0822, "step": 4335 }, { "epoch": 2.0404705882352943, "grad_norm": 0.15889581774762987, "learning_rate": 0.00026638315073823594, "loss": 0.1006, "step": 4336 }, { "epoch": 2.040941176470588, "grad_norm": 0.2826260665278981, "learning_rate": 0.00026635549079268104, "loss": 0.1632, "step": 4337 }, { "epoch": 2.0414117647058823, "grad_norm": 0.5010071193447092, "learning_rate": 0.00026632782090979653, "loss": 0.3293, "step": 4338 }, { "epoch": 2.0418823529411765, "grad_norm": 0.32652862255307935, "learning_rate": 0.0002663001410919455, "loss": 0.2651, "step": 4339 }, { "epoch": 2.0423529411764707, "grad_norm": 0.22204198276139647, "learning_rate": 0.000266272451341492, "loss": 0.101, "step": 4340 }, { "epoch": 2.0428235294117645, "grad_norm": 0.42164263142525427, "learning_rate": 0.0002662447516608008, "loss": 0.1772, "step": 4341 }, { "epoch": 2.0432941176470587, "grad_norm": 0.5621412023219731, "learning_rate": 0.0002662170420522377, "loss": 0.391, "step": 4342 }, { "epoch": 2.043764705882353, "grad_norm": 0.2246473393601898, "learning_rate": 0.0002661893225181691, "loss": 0.1783, "step": 4343 }, { "epoch": 2.044235294117647, "grad_norm": 0.3039898080661816, "learning_rate": 0.0002661615930609625, "loss": 0.3196, "step": 4344 }, { "epoch": 2.0447058823529414, "grad_norm": 0.35920712621199846, "learning_rate": 0.000266133853682986, "loss": 0.2503, "step": 4345 }, { "epoch": 2.045176470588235, "grad_norm": 0.49238731813137293, "learning_rate": 0.0002661061043866089, "loss": 0.2732, "step": 4346 }, { "epoch": 2.0456470588235294, "grad_norm": 0.44645169867770573, "learning_rate": 0.00026607834517420085, "loss": 0.279, "step": 4347 }, { "epoch": 2.0461176470588236, "grad_norm": 0.2772679223964017, "learning_rate": 0.00026605057604813285, "loss": 0.1576, "step": 4348 }, { "epoch": 2.046588235294118, "grad_norm": 0.20515588054485512, "learning_rate": 0.00026602279701077637, "loss": 0.1437, "step": 4349 }, { "epoch": 2.0470588235294116, "grad_norm": 0.47208575698383426, "learning_rate": 0.000265995008064504, "loss": 0.3595, "step": 4350 }, { "epoch": 2.047529411764706, "grad_norm": 0.2529566118336411, "learning_rate": 0.00026596720921168897, "loss": 0.1692, "step": 4351 }, { "epoch": 2.048, "grad_norm": 0.5318908564074059, "learning_rate": 0.00026593940045470545, "loss": 0.5798, "step": 4352 }, { "epoch": 2.0484705882352943, "grad_norm": 0.3040399522097993, "learning_rate": 0.00026591158179592846, "loss": 0.2236, "step": 4353 }, { "epoch": 2.048941176470588, "grad_norm": 0.2493809764369416, "learning_rate": 0.0002658837532377338, "loss": 0.1445, "step": 4354 }, { "epoch": 2.0494117647058823, "grad_norm": 0.19230689405519746, "learning_rate": 0.00026585591478249824, "loss": 0.1418, "step": 4355 }, { "epoch": 2.0498823529411765, "grad_norm": 0.869328957596308, "learning_rate": 0.0002658280664325993, "loss": 0.5735, "step": 4356 }, { "epoch": 2.0503529411764707, "grad_norm": 0.4309749465781233, "learning_rate": 0.00026580020819041533, "loss": 0.3787, "step": 4357 }, { "epoch": 2.0508235294117645, "grad_norm": 0.2291847800025444, "learning_rate": 0.0002657723400583256, "loss": 0.1709, "step": 4358 }, { "epoch": 2.0512941176470587, "grad_norm": 0.12666484483448995, "learning_rate": 0.00026574446203871013, "loss": 0.041, "step": 4359 }, { "epoch": 2.051764705882353, "grad_norm": 0.4930699898712735, "learning_rate": 0.0002657165741339499, "loss": 0.3711, "step": 4360 }, { "epoch": 2.052235294117647, "grad_norm": 0.8372465001171421, "learning_rate": 0.00026568867634642665, "loss": 0.6488, "step": 4361 }, { "epoch": 2.0527058823529414, "grad_norm": 0.36420138089635895, "learning_rate": 0.00026566076867852304, "loss": 0.2135, "step": 4362 }, { "epoch": 2.053176470588235, "grad_norm": 0.02839819563513812, "learning_rate": 0.0002656328511326224, "loss": 0.006, "step": 4363 }, { "epoch": 2.0536470588235294, "grad_norm": 0.37843485643255004, "learning_rate": 0.00026560492371110913, "loss": 0.2762, "step": 4364 }, { "epoch": 2.0541176470588236, "grad_norm": 0.028014237251217367, "learning_rate": 0.00026557698641636835, "loss": 0.0038, "step": 4365 }, { "epoch": 2.054588235294118, "grad_norm": 0.34541418764167514, "learning_rate": 0.000265549039250786, "loss": 0.2852, "step": 4366 }, { "epoch": 2.0550588235294116, "grad_norm": 0.34337368656407263, "learning_rate": 0.000265521082216749, "loss": 0.1514, "step": 4367 }, { "epoch": 2.055529411764706, "grad_norm": 0.46125366962914666, "learning_rate": 0.0002654931153166449, "loss": 0.3725, "step": 4368 }, { "epoch": 2.056, "grad_norm": 0.08536191675188182, "learning_rate": 0.0002654651385528622, "loss": 0.0406, "step": 4369 }, { "epoch": 2.0564705882352943, "grad_norm": 0.24038764173097704, "learning_rate": 0.00026543715192779043, "loss": 0.2003, "step": 4370 }, { "epoch": 2.056941176470588, "grad_norm": 0.3649030952328348, "learning_rate": 0.0002654091554438196, "loss": 0.3854, "step": 4371 }, { "epoch": 2.0574117647058823, "grad_norm": 0.38246193295247677, "learning_rate": 0.0002653811491033408, "loss": 0.5497, "step": 4372 }, { "epoch": 2.0578823529411765, "grad_norm": 0.17643191847565393, "learning_rate": 0.00026535313290874593, "loss": 0.0854, "step": 4373 }, { "epoch": 2.0583529411764707, "grad_norm": 0.44731822057455933, "learning_rate": 0.0002653251068624278, "loss": 0.2668, "step": 4374 }, { "epoch": 2.0588235294117645, "grad_norm": 0.5399383564635947, "learning_rate": 0.00026529707096677977, "loss": 0.4734, "step": 4375 }, { "epoch": 2.0592941176470587, "grad_norm": 0.19849655361459312, "learning_rate": 0.00026526902522419637, "loss": 0.1816, "step": 4376 }, { "epoch": 2.059764705882353, "grad_norm": 0.20450541574192344, "learning_rate": 0.0002652409696370729, "loss": 0.1566, "step": 4377 }, { "epoch": 2.060235294117647, "grad_norm": 0.2084698876296758, "learning_rate": 0.0002652129042078053, "loss": 0.1619, "step": 4378 }, { "epoch": 2.0607058823529414, "grad_norm": 0.4808179945504206, "learning_rate": 0.00026518482893879056, "loss": 0.5303, "step": 4379 }, { "epoch": 2.061176470588235, "grad_norm": 0.26971682148957793, "learning_rate": 0.0002651567438324265, "loss": 0.1402, "step": 4380 }, { "epoch": 2.0616470588235294, "grad_norm": 0.6901295450968805, "learning_rate": 0.00026512864889111163, "loss": 0.6447, "step": 4381 }, { "epoch": 2.0621176470588236, "grad_norm": 0.2937119820764359, "learning_rate": 0.00026510054411724545, "loss": 0.3975, "step": 4382 }, { "epoch": 2.062588235294118, "grad_norm": 0.29292892390883885, "learning_rate": 0.00026507242951322825, "loss": 0.3711, "step": 4383 }, { "epoch": 2.0630588235294116, "grad_norm": 0.34244846745528795, "learning_rate": 0.00026504430508146116, "loss": 0.3201, "step": 4384 }, { "epoch": 2.063529411764706, "grad_norm": 0.26328866578106225, "learning_rate": 0.00026501617082434607, "loss": 0.249, "step": 4385 }, { "epoch": 2.064, "grad_norm": 0.6580937714350794, "learning_rate": 0.0002649880267442859, "loss": 0.3346, "step": 4386 }, { "epoch": 2.0644705882352943, "grad_norm": 0.3537749600709138, "learning_rate": 0.00026495987284368415, "loss": 0.3438, "step": 4387 }, { "epoch": 2.064941176470588, "grad_norm": 0.37198080194120814, "learning_rate": 0.0002649317091249455, "loss": 0.3482, "step": 4388 }, { "epoch": 2.0654117647058823, "grad_norm": 0.3133110413074568, "learning_rate": 0.00026490353559047505, "loss": 0.3933, "step": 4389 }, { "epoch": 2.0658823529411765, "grad_norm": 0.3665669847346424, "learning_rate": 0.0002648753522426791, "loss": 0.1973, "step": 4390 }, { "epoch": 2.0663529411764707, "grad_norm": 0.3237776252548611, "learning_rate": 0.0002648471590839646, "loss": 0.3584, "step": 4391 }, { "epoch": 2.0668235294117645, "grad_norm": 0.08857716164881382, "learning_rate": 0.00026481895611673944, "loss": 0.0233, "step": 4392 }, { "epoch": 2.0672941176470587, "grad_norm": 0.8171538622902295, "learning_rate": 0.0002647907433434122, "loss": 0.7694, "step": 4393 }, { "epoch": 2.067764705882353, "grad_norm": 0.24827779887255993, "learning_rate": 0.00026476252076639245, "loss": 0.1616, "step": 4394 }, { "epoch": 2.068235294117647, "grad_norm": 0.15589148084999685, "learning_rate": 0.0002647342883880905, "loss": 0.0511, "step": 4395 }, { "epoch": 2.0687058823529414, "grad_norm": 0.42451847651932356, "learning_rate": 0.0002647060462109176, "loss": 0.346, "step": 4396 }, { "epoch": 2.069176470588235, "grad_norm": 0.6011358266354506, "learning_rate": 0.00026467779423728564, "loss": 0.508, "step": 4397 }, { "epoch": 2.0696470588235294, "grad_norm": 0.3427673014415781, "learning_rate": 0.0002646495324696076, "loss": 0.3222, "step": 4398 }, { "epoch": 2.0701176470588236, "grad_norm": 0.1942478021801696, "learning_rate": 0.0002646212609102972, "loss": 0.1477, "step": 4399 }, { "epoch": 2.070588235294118, "grad_norm": 0.2773497222776523, "learning_rate": 0.00026459297956176885, "loss": 0.1888, "step": 4400 }, { "epoch": 2.0710588235294116, "grad_norm": 0.31921895660061744, "learning_rate": 0.000264564688426438, "loss": 0.197, "step": 4401 }, { "epoch": 2.071529411764706, "grad_norm": 0.14849711944954055, "learning_rate": 0.0002645363875067208, "loss": 0.0757, "step": 4402 }, { "epoch": 2.072, "grad_norm": 0.44003025087264414, "learning_rate": 0.0002645080768050342, "loss": 0.2412, "step": 4403 }, { "epoch": 2.0724705882352943, "grad_norm": 0.5429555301721964, "learning_rate": 0.00026447975632379637, "loss": 0.4326, "step": 4404 }, { "epoch": 2.072941176470588, "grad_norm": 0.20641266673797218, "learning_rate": 0.0002644514260654257, "loss": 0.1131, "step": 4405 }, { "epoch": 2.0734117647058823, "grad_norm": 0.2918071266376583, "learning_rate": 0.0002644230860323419, "loss": 0.4715, "step": 4406 }, { "epoch": 2.0738823529411765, "grad_norm": 0.2551520288725708, "learning_rate": 0.00026439473622696534, "loss": 0.167, "step": 4407 }, { "epoch": 2.0743529411764707, "grad_norm": 0.23590644057579876, "learning_rate": 0.0002643663766517172, "loss": 0.1648, "step": 4408 }, { "epoch": 2.0748235294117645, "grad_norm": 0.19602788887736974, "learning_rate": 0.00026433800730901947, "loss": 0.1536, "step": 4409 }, { "epoch": 2.0752941176470587, "grad_norm": 0.47222827589386474, "learning_rate": 0.00026430962820129513, "loss": 0.2672, "step": 4410 }, { "epoch": 2.075764705882353, "grad_norm": 0.4792820214678954, "learning_rate": 0.00026428123933096786, "loss": 0.4135, "step": 4411 }, { "epoch": 2.076235294117647, "grad_norm": 0.37536590816248505, "learning_rate": 0.0002642528407004622, "loss": 0.272, "step": 4412 }, { "epoch": 2.0767058823529414, "grad_norm": 0.24133923669543383, "learning_rate": 0.00026422443231220353, "loss": 0.1388, "step": 4413 }, { "epoch": 2.077176470588235, "grad_norm": 0.5727174238882442, "learning_rate": 0.0002641960141686181, "loss": 0.6079, "step": 4414 }, { "epoch": 2.0776470588235294, "grad_norm": 0.5164653353507636, "learning_rate": 0.0002641675862721329, "loss": 0.6258, "step": 4415 }, { "epoch": 2.0781176470588236, "grad_norm": 0.5345428372379364, "learning_rate": 0.00026413914862517585, "loss": 0.6686, "step": 4416 }, { "epoch": 2.078588235294118, "grad_norm": 0.05811123910119141, "learning_rate": 0.00026411070123017567, "loss": 0.0143, "step": 4417 }, { "epoch": 2.0790588235294116, "grad_norm": 0.24001189162446576, "learning_rate": 0.0002640822440895619, "loss": 0.1816, "step": 4418 }, { "epoch": 2.079529411764706, "grad_norm": 0.4134279297372073, "learning_rate": 0.0002640537772057649, "loss": 0.3127, "step": 4419 }, { "epoch": 2.08, "grad_norm": 0.33531269112663875, "learning_rate": 0.0002640253005812159, "loss": 0.3039, "step": 4420 }, { "epoch": 2.0804705882352943, "grad_norm": 0.5784611635972958, "learning_rate": 0.00026399681421834694, "loss": 0.63, "step": 4421 }, { "epoch": 2.080941176470588, "grad_norm": 0.40016567971586037, "learning_rate": 0.0002639683181195909, "loss": 0.2235, "step": 4422 }, { "epoch": 2.0814117647058823, "grad_norm": 0.33492057946747195, "learning_rate": 0.00026393981228738144, "loss": 0.3182, "step": 4423 }, { "epoch": 2.0818823529411765, "grad_norm": 0.28663797655133966, "learning_rate": 0.0002639112967241532, "loss": 0.2203, "step": 4424 }, { "epoch": 2.0823529411764707, "grad_norm": 0.23335841448110423, "learning_rate": 0.00026388277143234146, "loss": 0.0913, "step": 4425 }, { "epoch": 2.0828235294117645, "grad_norm": 0.23589586258112108, "learning_rate": 0.0002638542364143825, "loss": 0.1979, "step": 4426 }, { "epoch": 2.0832941176470587, "grad_norm": 0.3920746483953552, "learning_rate": 0.00026382569167271317, "loss": 0.385, "step": 4427 }, { "epoch": 2.083764705882353, "grad_norm": 0.3119829288229522, "learning_rate": 0.00026379713720977156, "loss": 0.1748, "step": 4428 }, { "epoch": 2.084235294117647, "grad_norm": 0.1826394992553684, "learning_rate": 0.0002637685730279962, "loss": 0.1071, "step": 4429 }, { "epoch": 2.084705882352941, "grad_norm": 0.34584613820089855, "learning_rate": 0.00026373999912982676, "loss": 0.3298, "step": 4430 }, { "epoch": 2.085176470588235, "grad_norm": 0.2791879520388306, "learning_rate": 0.00026371141551770344, "loss": 0.154, "step": 4431 }, { "epoch": 2.0856470588235294, "grad_norm": 0.19562405662178065, "learning_rate": 0.0002636828221940675, "loss": 0.0981, "step": 4432 }, { "epoch": 2.0861176470588236, "grad_norm": 0.33782248605922743, "learning_rate": 0.00026365421916136093, "loss": 0.4234, "step": 4433 }, { "epoch": 2.086588235294118, "grad_norm": 0.34619758809796203, "learning_rate": 0.00026362560642202655, "loss": 0.3629, "step": 4434 }, { "epoch": 2.0870588235294116, "grad_norm": 0.35690519212398253, "learning_rate": 0.0002635969839785081, "loss": 0.3076, "step": 4435 }, { "epoch": 2.087529411764706, "grad_norm": 0.6405193594463832, "learning_rate": 0.00026356835183325, "loss": 0.4304, "step": 4436 }, { "epoch": 2.088, "grad_norm": 0.21769444450953657, "learning_rate": 0.00026353970998869763, "loss": 0.146, "step": 4437 }, { "epoch": 2.0884705882352943, "grad_norm": 0.4504167327617983, "learning_rate": 0.0002635110584472971, "loss": 0.4358, "step": 4438 }, { "epoch": 2.088941176470588, "grad_norm": 0.1409948075655005, "learning_rate": 0.00026348239721149535, "loss": 0.0438, "step": 4439 }, { "epoch": 2.0894117647058823, "grad_norm": 0.33781569740779976, "learning_rate": 0.0002634537262837403, "loss": 0.2761, "step": 4440 }, { "epoch": 2.0898823529411765, "grad_norm": 0.1781483305268387, "learning_rate": 0.00026342504566648055, "loss": 0.0986, "step": 4441 }, { "epoch": 2.0903529411764707, "grad_norm": 0.26898675342522727, "learning_rate": 0.0002633963553621655, "loss": 0.2636, "step": 4442 }, { "epoch": 2.0908235294117645, "grad_norm": 0.20652208267153752, "learning_rate": 0.0002633676553732456, "loss": 0.1412, "step": 4443 }, { "epoch": 2.0912941176470587, "grad_norm": 0.2554471031228314, "learning_rate": 0.0002633389457021718, "loss": 0.1812, "step": 4444 }, { "epoch": 2.091764705882353, "grad_norm": 0.275317230702228, "learning_rate": 0.0002633102263513961, "loss": 0.3927, "step": 4445 }, { "epoch": 2.092235294117647, "grad_norm": 0.6488786175275616, "learning_rate": 0.0002632814973233713, "loss": 0.6836, "step": 4446 }, { "epoch": 2.092705882352941, "grad_norm": 0.3206221639030903, "learning_rate": 0.00026325275862055104, "loss": 0.4082, "step": 4447 }, { "epoch": 2.093176470588235, "grad_norm": 0.19143311742836688, "learning_rate": 0.0002632240102453896, "loss": 0.0961, "step": 4448 }, { "epoch": 2.0936470588235294, "grad_norm": 0.40906324018833956, "learning_rate": 0.0002631952522003424, "loss": 0.3445, "step": 4449 }, { "epoch": 2.0941176470588236, "grad_norm": 0.4770206155893889, "learning_rate": 0.00026316648448786536, "loss": 0.362, "step": 4450 }, { "epoch": 2.094588235294118, "grad_norm": 0.2624282870173139, "learning_rate": 0.00026313770711041556, "loss": 0.2579, "step": 4451 }, { "epoch": 2.0950588235294116, "grad_norm": 0.1340406979137181, "learning_rate": 0.0002631089200704506, "loss": 0.0816, "step": 4452 }, { "epoch": 2.095529411764706, "grad_norm": 0.34502859561012245, "learning_rate": 0.00026308012337042906, "loss": 0.3029, "step": 4453 }, { "epoch": 2.096, "grad_norm": 0.3295505132956001, "learning_rate": 0.0002630513170128103, "loss": 0.3254, "step": 4454 }, { "epoch": 2.0964705882352943, "grad_norm": 0.32956191512114985, "learning_rate": 0.0002630225010000546, "loss": 0.1698, "step": 4455 }, { "epoch": 2.096941176470588, "grad_norm": 0.18833827800256323, "learning_rate": 0.0002629936753346229, "loss": 0.1163, "step": 4456 }, { "epoch": 2.0974117647058823, "grad_norm": 0.3391478403649201, "learning_rate": 0.00026296484001897715, "loss": 0.3021, "step": 4457 }, { "epoch": 2.0978823529411765, "grad_norm": 0.3015570748358667, "learning_rate": 0.00026293599505558, "loss": 0.2434, "step": 4458 }, { "epoch": 2.0983529411764708, "grad_norm": 0.46287148846187004, "learning_rate": 0.00026290714044689483, "loss": 0.4296, "step": 4459 }, { "epoch": 2.0988235294117645, "grad_norm": 0.10750342911591561, "learning_rate": 0.0002628782761953861, "loss": 0.0485, "step": 4460 }, { "epoch": 2.0992941176470588, "grad_norm": 0.4363752568101263, "learning_rate": 0.0002628494023035189, "loss": 0.5087, "step": 4461 }, { "epoch": 2.099764705882353, "grad_norm": 0.1105846047834698, "learning_rate": 0.00026282051877375923, "loss": 0.0364, "step": 4462 }, { "epoch": 2.100235294117647, "grad_norm": 0.2793693082082926, "learning_rate": 0.0002627916256085739, "loss": 0.2346, "step": 4463 }, { "epoch": 2.100705882352941, "grad_norm": 0.19188979295116165, "learning_rate": 0.0002627627228104305, "loss": 0.1135, "step": 4464 }, { "epoch": 2.101176470588235, "grad_norm": 0.229953946877007, "learning_rate": 0.00026273381038179744, "loss": 0.1008, "step": 4465 }, { "epoch": 2.1016470588235294, "grad_norm": 0.7814113572648265, "learning_rate": 0.000262704888325144, "loss": 0.4749, "step": 4466 }, { "epoch": 2.1021176470588236, "grad_norm": 0.41088995310002835, "learning_rate": 0.0002626759566429404, "loss": 0.3427, "step": 4467 }, { "epoch": 2.1025882352941174, "grad_norm": 0.178970895847542, "learning_rate": 0.0002626470153376573, "loss": 0.1092, "step": 4468 }, { "epoch": 2.1030588235294116, "grad_norm": 0.37400582481962286, "learning_rate": 0.00026261806441176666, "loss": 0.2535, "step": 4469 }, { "epoch": 2.103529411764706, "grad_norm": 0.6691356553065312, "learning_rate": 0.0002625891038677408, "loss": 0.6241, "step": 4470 }, { "epoch": 2.104, "grad_norm": 0.4562879787559052, "learning_rate": 0.0002625601337080533, "loss": 0.3321, "step": 4471 }, { "epoch": 2.1044705882352943, "grad_norm": 0.09314916957317133, "learning_rate": 0.00026253115393517834, "loss": 0.0328, "step": 4472 }, { "epoch": 2.104941176470588, "grad_norm": 0.21033393173682557, "learning_rate": 0.0002625021645515908, "loss": 0.1614, "step": 4473 }, { "epoch": 2.1054117647058823, "grad_norm": 0.25715014434394073, "learning_rate": 0.00026247316555976657, "loss": 0.2665, "step": 4474 }, { "epoch": 2.1058823529411765, "grad_norm": 0.3187443044104742, "learning_rate": 0.00026244415696218237, "loss": 0.2184, "step": 4475 }, { "epoch": 2.1063529411764708, "grad_norm": 0.21188274657781586, "learning_rate": 0.00026241513876131564, "loss": 0.0907, "step": 4476 }, { "epoch": 2.1068235294117645, "grad_norm": 0.3798058588078459, "learning_rate": 0.00026238611095964466, "loss": 0.1998, "step": 4477 }, { "epoch": 2.1072941176470588, "grad_norm": 0.3010195629592632, "learning_rate": 0.0002623570735596485, "loss": 0.2785, "step": 4478 }, { "epoch": 2.107764705882353, "grad_norm": 0.28453300829845773, "learning_rate": 0.0002623280265638072, "loss": 0.3108, "step": 4479 }, { "epoch": 2.108235294117647, "grad_norm": 0.29584672684079427, "learning_rate": 0.0002622989699746014, "loss": 0.2898, "step": 4480 }, { "epoch": 2.108705882352941, "grad_norm": 0.19287948726196052, "learning_rate": 0.00026226990379451284, "loss": 0.142, "step": 4481 }, { "epoch": 2.109176470588235, "grad_norm": 0.4039512418935042, "learning_rate": 0.00026224082802602373, "loss": 0.3307, "step": 4482 }, { "epoch": 2.1096470588235294, "grad_norm": 0.23916572986376036, "learning_rate": 0.00026221174267161744, "loss": 0.1124, "step": 4483 }, { "epoch": 2.1101176470588237, "grad_norm": 0.17486255961505306, "learning_rate": 0.0002621826477337779, "loss": 0.0692, "step": 4484 }, { "epoch": 2.1105882352941174, "grad_norm": 0.48364136939335634, "learning_rate": 0.00026215354321498997, "loss": 0.2612, "step": 4485 }, { "epoch": 2.1110588235294117, "grad_norm": 0.2215936878740989, "learning_rate": 0.00026212442911773936, "loss": 0.1778, "step": 4486 }, { "epoch": 2.111529411764706, "grad_norm": 0.4055395257306334, "learning_rate": 0.0002620953054445125, "loss": 0.4309, "step": 4487 }, { "epoch": 2.112, "grad_norm": 0.2979979559251248, "learning_rate": 0.0002620661721977968, "loss": 0.2979, "step": 4488 }, { "epoch": 2.1124705882352943, "grad_norm": 0.3900398182340769, "learning_rate": 0.00026203702938008025, "loss": 0.3697, "step": 4489 }, { "epoch": 2.112941176470588, "grad_norm": 0.42523668674547876, "learning_rate": 0.00026200787699385186, "loss": 0.2598, "step": 4490 }, { "epoch": 2.1134117647058823, "grad_norm": 0.32672429674690345, "learning_rate": 0.0002619787150416014, "loss": 0.3394, "step": 4491 }, { "epoch": 2.1138823529411765, "grad_norm": 0.38636891652759786, "learning_rate": 0.00026194954352581944, "loss": 0.3948, "step": 4492 }, { "epoch": 2.1143529411764708, "grad_norm": 0.10868064981466811, "learning_rate": 0.0002619203624489973, "loss": 0.0766, "step": 4493 }, { "epoch": 2.1148235294117645, "grad_norm": 0.33307216464717737, "learning_rate": 0.00026189117181362733, "loss": 0.3484, "step": 4494 }, { "epoch": 2.1152941176470588, "grad_norm": 0.34184649135296363, "learning_rate": 0.0002618619716222024, "loss": 0.3545, "step": 4495 }, { "epoch": 2.115764705882353, "grad_norm": 0.5095931561089554, "learning_rate": 0.00026183276187721635, "loss": 0.5171, "step": 4496 }, { "epoch": 2.116235294117647, "grad_norm": 0.30503229746321636, "learning_rate": 0.000261803542581164, "loss": 0.2786, "step": 4497 }, { "epoch": 2.116705882352941, "grad_norm": 0.43502205769167307, "learning_rate": 0.0002617743137365407, "loss": 0.3484, "step": 4498 }, { "epoch": 2.117176470588235, "grad_norm": 0.34779810199106853, "learning_rate": 0.0002617450753458427, "loss": 0.3267, "step": 4499 }, { "epoch": 2.1176470588235294, "grad_norm": 0.48620351249238675, "learning_rate": 0.00026171582741156725, "loss": 0.3076, "step": 4500 }, { "epoch": 2.1181176470588237, "grad_norm": 0.4388145484963979, "learning_rate": 0.0002616865699362121, "loss": 0.4839, "step": 4501 }, { "epoch": 2.1185882352941174, "grad_norm": 0.2810445830408761, "learning_rate": 0.0002616573029222761, "loss": 0.2881, "step": 4502 }, { "epoch": 2.1190588235294117, "grad_norm": 0.10614353259011933, "learning_rate": 0.0002616280263722588, "loss": 0.0681, "step": 4503 }, { "epoch": 2.119529411764706, "grad_norm": 0.36643501804810147, "learning_rate": 0.00026159874028866044, "loss": 0.3259, "step": 4504 }, { "epoch": 2.12, "grad_norm": 0.550919447272843, "learning_rate": 0.0002615694446739823, "loss": 0.329, "step": 4505 }, { "epoch": 2.1204705882352943, "grad_norm": 0.16845111130986273, "learning_rate": 0.0002615401395307263, "loss": 0.0979, "step": 4506 }, { "epoch": 2.120941176470588, "grad_norm": 0.16566612521943255, "learning_rate": 0.00026151082486139534, "loss": 0.1017, "step": 4507 }, { "epoch": 2.1214117647058823, "grad_norm": 0.35181374352470013, "learning_rate": 0.0002614815006684929, "loss": 0.2946, "step": 4508 }, { "epoch": 2.1218823529411766, "grad_norm": 0.31348480285355557, "learning_rate": 0.0002614521669545236, "loss": 0.2058, "step": 4509 }, { "epoch": 2.1223529411764708, "grad_norm": 0.12075615102751992, "learning_rate": 0.0002614228237219925, "loss": 0.0493, "step": 4510 }, { "epoch": 2.1228235294117646, "grad_norm": 0.14007995528340172, "learning_rate": 0.00026139347097340574, "loss": 0.0782, "step": 4511 }, { "epoch": 2.1232941176470588, "grad_norm": 0.4169021751814516, "learning_rate": 0.00026136410871127017, "loss": 0.33, "step": 4512 }, { "epoch": 2.123764705882353, "grad_norm": 0.34110474054145645, "learning_rate": 0.0002613347369380935, "loss": 0.2761, "step": 4513 }, { "epoch": 2.124235294117647, "grad_norm": 0.13176609016329305, "learning_rate": 0.00026130535565638415, "loss": 0.075, "step": 4514 }, { "epoch": 2.124705882352941, "grad_norm": 0.5238716874960694, "learning_rate": 0.00026127596486865155, "loss": 0.6005, "step": 4515 }, { "epoch": 2.1251764705882352, "grad_norm": 0.7337531487355912, "learning_rate": 0.00026124656457740574, "loss": 0.3829, "step": 4516 }, { "epoch": 2.1256470588235294, "grad_norm": 0.35584692670875406, "learning_rate": 0.00026121715478515757, "loss": 0.3208, "step": 4517 }, { "epoch": 2.1261176470588237, "grad_norm": 0.37390773069357863, "learning_rate": 0.0002611877354944189, "loss": 0.3896, "step": 4518 }, { "epoch": 2.1265882352941174, "grad_norm": 0.386569345828159, "learning_rate": 0.00026115830670770225, "loss": 0.3143, "step": 4519 }, { "epoch": 2.1270588235294117, "grad_norm": 0.3487344621879447, "learning_rate": 0.00026112886842752104, "loss": 0.3938, "step": 4520 }, { "epoch": 2.127529411764706, "grad_norm": 0.23172701105255983, "learning_rate": 0.0002610994206563893, "loss": 0.2205, "step": 4521 }, { "epoch": 2.128, "grad_norm": 0.5744446448152272, "learning_rate": 0.0002610699633968222, "loss": 0.575, "step": 4522 }, { "epoch": 2.1284705882352943, "grad_norm": 0.07892237110874337, "learning_rate": 0.00026104049665133533, "loss": 0.0576, "step": 4523 }, { "epoch": 2.128941176470588, "grad_norm": 0.21233206385011916, "learning_rate": 0.00026101102042244543, "loss": 0.1596, "step": 4524 }, { "epoch": 2.1294117647058823, "grad_norm": 0.20246824816885298, "learning_rate": 0.0002609815347126699, "loss": 0.1503, "step": 4525 }, { "epoch": 2.1298823529411766, "grad_norm": 0.31604723571658255, "learning_rate": 0.00026095203952452694, "loss": 0.1502, "step": 4526 }, { "epoch": 2.130352941176471, "grad_norm": 0.377389221750166, "learning_rate": 0.00026092253486053555, "loss": 0.3582, "step": 4527 }, { "epoch": 2.1308235294117646, "grad_norm": 0.47632773589988675, "learning_rate": 0.0002608930207232156, "loss": 0.4026, "step": 4528 }, { "epoch": 2.131294117647059, "grad_norm": 0.21862317377539264, "learning_rate": 0.0002608634971150878, "loss": 0.1052, "step": 4529 }, { "epoch": 2.131764705882353, "grad_norm": 0.4656078933718426, "learning_rate": 0.0002608339640386735, "loss": 0.2787, "step": 4530 }, { "epoch": 2.1322352941176472, "grad_norm": 0.3354967533011007, "learning_rate": 0.00026080442149649516, "loss": 0.2096, "step": 4531 }, { "epoch": 2.132705882352941, "grad_norm": 0.1325943531957621, "learning_rate": 0.0002607748694910756, "loss": 0.0671, "step": 4532 }, { "epoch": 2.1331764705882352, "grad_norm": 0.401806453795846, "learning_rate": 0.0002607453080249389, "loss": 0.2663, "step": 4533 }, { "epoch": 2.1336470588235295, "grad_norm": 0.4768330055436134, "learning_rate": 0.0002607157371006097, "loss": 0.3483, "step": 4534 }, { "epoch": 2.1341176470588237, "grad_norm": 0.3934612634285513, "learning_rate": 0.00026068615672061353, "loss": 0.3913, "step": 4535 }, { "epoch": 2.1345882352941175, "grad_norm": 0.3572344962278902, "learning_rate": 0.0002606565668874766, "loss": 0.3457, "step": 4536 }, { "epoch": 2.1350588235294117, "grad_norm": 0.1283164879141306, "learning_rate": 0.00026062696760372606, "loss": 0.0954, "step": 4537 }, { "epoch": 2.135529411764706, "grad_norm": 0.3606070480996365, "learning_rate": 0.0002605973588718899, "loss": 0.3723, "step": 4538 }, { "epoch": 2.136, "grad_norm": 0.23444907867722922, "learning_rate": 0.00026056774069449685, "loss": 0.2422, "step": 4539 }, { "epoch": 2.1364705882352943, "grad_norm": 0.2121402992941977, "learning_rate": 0.00026053811307407635, "loss": 0.1902, "step": 4540 }, { "epoch": 2.136941176470588, "grad_norm": 0.23738970932278203, "learning_rate": 0.00026050847601315886, "loss": 0.1081, "step": 4541 }, { "epoch": 2.1374117647058823, "grad_norm": 0.3793941831206535, "learning_rate": 0.00026047882951427543, "loss": 0.3864, "step": 4542 }, { "epoch": 2.1378823529411766, "grad_norm": 1.1045623990023876, "learning_rate": 0.00026044917357995813, "loss": 0.3794, "step": 4543 }, { "epoch": 2.138352941176471, "grad_norm": 0.1723933972434641, "learning_rate": 0.00026041950821273965, "loss": 0.1156, "step": 4544 }, { "epoch": 2.1388235294117646, "grad_norm": 0.43641773539811535, "learning_rate": 0.0002603898334151535, "loss": 0.4988, "step": 4545 }, { "epoch": 2.139294117647059, "grad_norm": 0.3973658292892046, "learning_rate": 0.00026036014918973414, "loss": 0.37, "step": 4546 }, { "epoch": 2.139764705882353, "grad_norm": 0.307224464406383, "learning_rate": 0.00026033045553901674, "loss": 0.3627, "step": 4547 }, { "epoch": 2.1402352941176472, "grad_norm": 0.19098834774522383, "learning_rate": 0.0002603007524655373, "loss": 0.0606, "step": 4548 }, { "epoch": 2.140705882352941, "grad_norm": 0.28223963253300294, "learning_rate": 0.0002602710399718325, "loss": 0.244, "step": 4549 }, { "epoch": 2.1411764705882352, "grad_norm": 0.584238654118233, "learning_rate": 0.0002602413180604401, "loss": 0.6049, "step": 4550 }, { "epoch": 2.1416470588235295, "grad_norm": 0.13905746581708778, "learning_rate": 0.0002602115867338983, "loss": 0.0846, "step": 4551 }, { "epoch": 2.1421176470588237, "grad_norm": 0.20644926525358157, "learning_rate": 0.0002601818459947465, "loss": 0.186, "step": 4552 }, { "epoch": 2.1425882352941175, "grad_norm": 0.42347115962587484, "learning_rate": 0.0002601520958455246, "loss": 0.3393, "step": 4553 }, { "epoch": 2.1430588235294117, "grad_norm": 0.3758938129584884, "learning_rate": 0.00026012233628877336, "loss": 0.3008, "step": 4554 }, { "epoch": 2.143529411764706, "grad_norm": 0.2955258132147811, "learning_rate": 0.0002600925673270345, "loss": 0.3311, "step": 4555 }, { "epoch": 2.144, "grad_norm": 0.5149756990317323, "learning_rate": 0.0002600627889628504, "loss": 0.4247, "step": 4556 }, { "epoch": 2.1444705882352944, "grad_norm": 0.27413378706288266, "learning_rate": 0.00026003300119876424, "loss": 0.2745, "step": 4557 }, { "epoch": 2.144941176470588, "grad_norm": 0.16514664070397347, "learning_rate": 0.0002600032040373201, "loss": 0.1423, "step": 4558 }, { "epoch": 2.1454117647058824, "grad_norm": 0.24918496118412983, "learning_rate": 0.00025997339748106275, "loss": 0.2284, "step": 4559 }, { "epoch": 2.1458823529411766, "grad_norm": 0.28863318917715264, "learning_rate": 0.00025994358153253784, "loss": 0.1265, "step": 4560 }, { "epoch": 2.1463529411764704, "grad_norm": 0.6473397325046691, "learning_rate": 0.0002599137561942918, "loss": 0.5377, "step": 4561 }, { "epoch": 2.1468235294117646, "grad_norm": 0.23033385303205425, "learning_rate": 0.00025988392146887183, "loss": 0.1618, "step": 4562 }, { "epoch": 2.147294117647059, "grad_norm": 0.5046673323682036, "learning_rate": 0.00025985407735882606, "loss": 0.5372, "step": 4563 }, { "epoch": 2.147764705882353, "grad_norm": 0.24568111114596716, "learning_rate": 0.0002598242238667032, "loss": 0.2151, "step": 4564 }, { "epoch": 2.1482352941176472, "grad_norm": 0.2087037479625, "learning_rate": 0.00025979436099505296, "loss": 0.106, "step": 4565 }, { "epoch": 2.148705882352941, "grad_norm": 0.2646945434728274, "learning_rate": 0.0002597644887464257, "loss": 0.154, "step": 4566 }, { "epoch": 2.1491764705882352, "grad_norm": 0.37883111760333743, "learning_rate": 0.00025973460712337275, "loss": 0.1861, "step": 4567 }, { "epoch": 2.1496470588235295, "grad_norm": 0.48495090743359576, "learning_rate": 0.0002597047161284461, "loss": 0.6735, "step": 4568 }, { "epoch": 2.1501176470588237, "grad_norm": 0.37498218697577734, "learning_rate": 0.00025967481576419864, "loss": 0.3439, "step": 4569 }, { "epoch": 2.1505882352941175, "grad_norm": 0.7160752862801988, "learning_rate": 0.0002596449060331839, "loss": 0.5736, "step": 4570 }, { "epoch": 2.1510588235294117, "grad_norm": 0.3602414511467744, "learning_rate": 0.0002596149869379564, "loss": 0.2815, "step": 4571 }, { "epoch": 2.151529411764706, "grad_norm": 0.10479613704821387, "learning_rate": 0.00025958505848107133, "loss": 0.082, "step": 4572 }, { "epoch": 2.152, "grad_norm": 0.2836608518701336, "learning_rate": 0.00025955512066508477, "loss": 0.2461, "step": 4573 }, { "epoch": 2.152470588235294, "grad_norm": 0.3088385507233069, "learning_rate": 0.0002595251734925536, "loss": 0.3746, "step": 4574 }, { "epoch": 2.152941176470588, "grad_norm": 0.2292459348650386, "learning_rate": 0.00025949521696603536, "loss": 0.1654, "step": 4575 }, { "epoch": 2.1534117647058824, "grad_norm": 0.22568679681024595, "learning_rate": 0.0002594652510880885, "loss": 0.1274, "step": 4576 }, { "epoch": 2.1538823529411766, "grad_norm": 0.1287434489507802, "learning_rate": 0.0002594352758612723, "loss": 0.0629, "step": 4577 }, { "epoch": 2.1543529411764704, "grad_norm": 0.38727334914204453, "learning_rate": 0.0002594052912881467, "loss": 0.3921, "step": 4578 }, { "epoch": 2.1548235294117646, "grad_norm": 0.2001063572105365, "learning_rate": 0.00025937529737127263, "loss": 0.125, "step": 4579 }, { "epoch": 2.155294117647059, "grad_norm": 0.228924798086124, "learning_rate": 0.0002593452941132117, "loss": 0.1376, "step": 4580 }, { "epoch": 2.155764705882353, "grad_norm": 0.27234754026792185, "learning_rate": 0.0002593152815165263, "loss": 0.3267, "step": 4581 }, { "epoch": 2.1562352941176472, "grad_norm": 0.2814446567102211, "learning_rate": 0.0002592852595837797, "loss": 0.18, "step": 4582 }, { "epoch": 2.156705882352941, "grad_norm": 0.3889673652858673, "learning_rate": 0.00025925522831753584, "loss": 0.2936, "step": 4583 }, { "epoch": 2.1571764705882353, "grad_norm": 0.40502844329026777, "learning_rate": 0.00025922518772035963, "loss": 0.5037, "step": 4584 }, { "epoch": 2.1576470588235295, "grad_norm": 0.6241029966974326, "learning_rate": 0.0002591951377948166, "loss": 0.6354, "step": 4585 }, { "epoch": 2.1581176470588237, "grad_norm": 0.38077464926362486, "learning_rate": 0.00025916507854347327, "loss": 0.2829, "step": 4586 }, { "epoch": 2.1585882352941175, "grad_norm": 0.24857757048660528, "learning_rate": 0.00025913500996889677, "loss": 0.14, "step": 4587 }, { "epoch": 2.1590588235294117, "grad_norm": 0.21350296706740576, "learning_rate": 0.00025910493207365504, "loss": 0.1127, "step": 4588 }, { "epoch": 2.159529411764706, "grad_norm": 0.20847705676387346, "learning_rate": 0.000259074844860317, "loss": 0.1267, "step": 4589 }, { "epoch": 2.16, "grad_norm": 0.3104759684547676, "learning_rate": 0.00025904474833145225, "loss": 0.222, "step": 4590 }, { "epoch": 2.160470588235294, "grad_norm": 0.24325249913900593, "learning_rate": 0.00025901464248963106, "loss": 0.1869, "step": 4591 }, { "epoch": 2.160941176470588, "grad_norm": 0.28917132809836754, "learning_rate": 0.00025898452733742473, "loss": 0.1214, "step": 4592 }, { "epoch": 2.1614117647058824, "grad_norm": 0.2928738830778892, "learning_rate": 0.00025895440287740514, "loss": 0.1511, "step": 4593 }, { "epoch": 2.1618823529411766, "grad_norm": 0.2536330114448007, "learning_rate": 0.00025892426911214517, "loss": 0.1459, "step": 4594 }, { "epoch": 2.1623529411764704, "grad_norm": 0.0713107379854086, "learning_rate": 0.0002588941260442183, "loss": 0.019, "step": 4595 }, { "epoch": 2.1628235294117646, "grad_norm": 0.28161426286049596, "learning_rate": 0.00025886397367619904, "loss": 0.2991, "step": 4596 }, { "epoch": 2.163294117647059, "grad_norm": 0.28989479136940477, "learning_rate": 0.00025883381201066236, "loss": 0.2143, "step": 4597 }, { "epoch": 2.163764705882353, "grad_norm": 0.2996230908849458, "learning_rate": 0.0002588036410501844, "loss": 0.1745, "step": 4598 }, { "epoch": 2.1642352941176473, "grad_norm": 0.2975416914192169, "learning_rate": 0.00025877346079734175, "loss": 0.1704, "step": 4599 }, { "epoch": 2.164705882352941, "grad_norm": 0.37575269009132883, "learning_rate": 0.000258743271254712, "loss": 0.3801, "step": 4600 }, { "epoch": 2.1651764705882353, "grad_norm": 0.3631443075216453, "learning_rate": 0.00025871307242487354, "loss": 0.4756, "step": 4601 }, { "epoch": 2.1656470588235295, "grad_norm": 0.30519488522058486, "learning_rate": 0.00025868286431040546, "loss": 0.4084, "step": 4602 }, { "epoch": 2.1661176470588237, "grad_norm": 0.14338473512740785, "learning_rate": 0.0002586526469138877, "loss": 0.1121, "step": 4603 }, { "epoch": 2.1665882352941175, "grad_norm": 0.20349929495819882, "learning_rate": 0.0002586224202379009, "loss": 0.1872, "step": 4604 }, { "epoch": 2.1670588235294117, "grad_norm": 0.19525802041519008, "learning_rate": 0.00025859218428502667, "loss": 0.1285, "step": 4605 }, { "epoch": 2.167529411764706, "grad_norm": 0.4730327763156603, "learning_rate": 0.00025856193905784725, "loss": 0.2839, "step": 4606 }, { "epoch": 2.168, "grad_norm": 0.4238561838301823, "learning_rate": 0.00025853168455894566, "loss": 0.5376, "step": 4607 }, { "epoch": 2.168470588235294, "grad_norm": 0.49448046608391316, "learning_rate": 0.000258501420790906, "loss": 0.5799, "step": 4608 }, { "epoch": 2.168941176470588, "grad_norm": 0.4829388955368389, "learning_rate": 0.0002584711477563127, "loss": 0.3818, "step": 4609 }, { "epoch": 2.1694117647058824, "grad_norm": 0.2427066713740792, "learning_rate": 0.0002584408654577514, "loss": 0.2047, "step": 4610 }, { "epoch": 2.1698823529411766, "grad_norm": 0.1096176518272893, "learning_rate": 0.0002584105738978083, "loss": 0.0502, "step": 4611 }, { "epoch": 2.1703529411764704, "grad_norm": 0.35811411496997075, "learning_rate": 0.00025838027307907037, "loss": 0.3468, "step": 4612 }, { "epoch": 2.1708235294117646, "grad_norm": 0.3394772470830228, "learning_rate": 0.00025834996300412556, "loss": 0.228, "step": 4613 }, { "epoch": 2.171294117647059, "grad_norm": 0.4321385428712358, "learning_rate": 0.0002583196436755624, "loss": 0.3533, "step": 4614 }, { "epoch": 2.171764705882353, "grad_norm": 0.23113795974404117, "learning_rate": 0.00025828931509597045, "loss": 0.1607, "step": 4615 }, { "epoch": 2.1722352941176473, "grad_norm": 0.4874486820791523, "learning_rate": 0.0002582589772679398, "loss": 0.549, "step": 4616 }, { "epoch": 2.172705882352941, "grad_norm": 0.28702504394776013, "learning_rate": 0.0002582286301940615, "loss": 0.4053, "step": 4617 }, { "epoch": 2.1731764705882353, "grad_norm": 0.10322796190235986, "learning_rate": 0.00025819827387692733, "loss": 0.0555, "step": 4618 }, { "epoch": 2.1736470588235295, "grad_norm": 0.26307794746241114, "learning_rate": 0.0002581679083191298, "loss": 0.2714, "step": 4619 }, { "epoch": 2.1741176470588237, "grad_norm": 0.15616846743978705, "learning_rate": 0.0002581375335232624, "loss": 0.0732, "step": 4620 }, { "epoch": 2.1745882352941175, "grad_norm": 0.6615389149387728, "learning_rate": 0.00025810714949191916, "loss": 0.6317, "step": 4621 }, { "epoch": 2.1750588235294117, "grad_norm": 0.6641367061724124, "learning_rate": 0.00025807675622769514, "loss": 0.4129, "step": 4622 }, { "epoch": 2.175529411764706, "grad_norm": 0.36663434431081604, "learning_rate": 0.00025804635373318603, "loss": 0.3862, "step": 4623 }, { "epoch": 2.176, "grad_norm": 0.37422390836210495, "learning_rate": 0.0002580159420109883, "loss": 0.619, "step": 4624 }, { "epoch": 2.176470588235294, "grad_norm": 0.3785135795638389, "learning_rate": 0.00025798552106369937, "loss": 0.4206, "step": 4625 }, { "epoch": 2.176941176470588, "grad_norm": 0.25217670952544635, "learning_rate": 0.00025795509089391726, "loss": 0.1956, "step": 4626 }, { "epoch": 2.1774117647058824, "grad_norm": 0.45133111260216086, "learning_rate": 0.00025792465150424083, "loss": 0.4605, "step": 4627 }, { "epoch": 2.1778823529411766, "grad_norm": 0.5864151843930643, "learning_rate": 0.00025789420289726983, "loss": 0.589, "step": 4628 }, { "epoch": 2.1783529411764704, "grad_norm": 0.5365778964306087, "learning_rate": 0.00025786374507560467, "loss": 0.3886, "step": 4629 }, { "epoch": 2.1788235294117646, "grad_norm": 0.2524874942940187, "learning_rate": 0.0002578332780418466, "loss": 0.3251, "step": 4630 }, { "epoch": 2.179294117647059, "grad_norm": 0.18757899461988053, "learning_rate": 0.0002578028017985977, "loss": 0.092, "step": 4631 }, { "epoch": 2.179764705882353, "grad_norm": 0.6128059511275578, "learning_rate": 0.00025777231634846074, "loss": 0.6167, "step": 4632 }, { "epoch": 2.1802352941176473, "grad_norm": 0.6141803368085721, "learning_rate": 0.00025774182169403933, "loss": 0.4979, "step": 4633 }, { "epoch": 2.180705882352941, "grad_norm": 0.20469716394448495, "learning_rate": 0.00025771131783793787, "loss": 0.111, "step": 4634 }, { "epoch": 2.1811764705882353, "grad_norm": 0.11823031201684718, "learning_rate": 0.00025768080478276163, "loss": 0.05, "step": 4635 }, { "epoch": 2.1816470588235295, "grad_norm": 0.6072865531765964, "learning_rate": 0.0002576502825311164, "loss": 0.6183, "step": 4636 }, { "epoch": 2.1821176470588237, "grad_norm": 0.120509998502855, "learning_rate": 0.000257619751085609, "loss": 0.0641, "step": 4637 }, { "epoch": 2.1825882352941175, "grad_norm": 0.39994660549715205, "learning_rate": 0.00025758921044884706, "loss": 0.305, "step": 4638 }, { "epoch": 2.1830588235294117, "grad_norm": 0.37284282029496435, "learning_rate": 0.0002575586606234388, "loss": 0.3982, "step": 4639 }, { "epoch": 2.183529411764706, "grad_norm": 0.2582308188051922, "learning_rate": 0.0002575281016119933, "loss": 0.2219, "step": 4640 }, { "epoch": 2.184, "grad_norm": 0.36037742130844663, "learning_rate": 0.00025749753341712056, "loss": 0.3612, "step": 4641 }, { "epoch": 2.184470588235294, "grad_norm": 0.21053454290493684, "learning_rate": 0.0002574669560414312, "loss": 0.1879, "step": 4642 }, { "epoch": 2.184941176470588, "grad_norm": 0.12637477318755197, "learning_rate": 0.0002574363694875367, "loss": 0.0693, "step": 4643 }, { "epoch": 2.1854117647058824, "grad_norm": 0.5155592057596029, "learning_rate": 0.0002574057737580492, "loss": 0.473, "step": 4644 }, { "epoch": 2.1858823529411766, "grad_norm": 0.3616828501719048, "learning_rate": 0.0002573751688555818, "loss": 0.3005, "step": 4645 }, { "epoch": 2.1863529411764704, "grad_norm": 0.3196536195147801, "learning_rate": 0.0002573445547827484, "loss": 0.3523, "step": 4646 }, { "epoch": 2.1868235294117646, "grad_norm": 0.32362706463903773, "learning_rate": 0.0002573139315421634, "loss": 0.3473, "step": 4647 }, { "epoch": 2.187294117647059, "grad_norm": 0.20938174778498841, "learning_rate": 0.0002572832991364423, "loss": 0.1041, "step": 4648 }, { "epoch": 2.187764705882353, "grad_norm": 0.5100683013119072, "learning_rate": 0.00025725265756820125, "loss": 0.3911, "step": 4649 }, { "epoch": 2.1882352941176473, "grad_norm": 0.2750045797162039, "learning_rate": 0.00025722200684005715, "loss": 0.2691, "step": 4650 }, { "epoch": 2.188705882352941, "grad_norm": 0.1344492513862629, "learning_rate": 0.0002571913469546277, "loss": 0.0834, "step": 4651 }, { "epoch": 2.1891764705882353, "grad_norm": 0.5014798344451814, "learning_rate": 0.00025716067791453157, "loss": 0.4892, "step": 4652 }, { "epoch": 2.1896470588235295, "grad_norm": 0.5805001280255881, "learning_rate": 0.0002571299997223878, "loss": 0.4928, "step": 4653 }, { "epoch": 2.1901176470588237, "grad_norm": 0.3292065843152007, "learning_rate": 0.0002570993123808167, "loss": 0.3337, "step": 4654 }, { "epoch": 2.1905882352941175, "grad_norm": 0.5122280408239666, "learning_rate": 0.00025706861589243894, "loss": 0.594, "step": 4655 }, { "epoch": 2.1910588235294117, "grad_norm": 0.32507546103603885, "learning_rate": 0.0002570379102598762, "loss": 0.2954, "step": 4656 }, { "epoch": 2.191529411764706, "grad_norm": 0.5076822556319838, "learning_rate": 0.000257007195485751, "loss": 0.4813, "step": 4657 }, { "epoch": 2.192, "grad_norm": 0.22235637468722666, "learning_rate": 0.00025697647157268635, "loss": 0.2149, "step": 4658 }, { "epoch": 2.192470588235294, "grad_norm": 0.2947387339692602, "learning_rate": 0.0002569457385233063, "loss": 0.3289, "step": 4659 }, { "epoch": 2.192941176470588, "grad_norm": 0.2530517754266309, "learning_rate": 0.00025691499634023564, "loss": 0.241, "step": 4660 }, { "epoch": 2.1934117647058824, "grad_norm": 0.5376365804327156, "learning_rate": 0.0002568842450260999, "loss": 0.5245, "step": 4661 }, { "epoch": 2.1938823529411766, "grad_norm": 0.15355118368981846, "learning_rate": 0.0002568534845835253, "loss": 0.0461, "step": 4662 }, { "epoch": 2.1943529411764704, "grad_norm": 0.21901776689505387, "learning_rate": 0.0002568227150151391, "loss": 0.1266, "step": 4663 }, { "epoch": 2.1948235294117646, "grad_norm": 0.254690767010946, "learning_rate": 0.000256791936323569, "loss": 0.2219, "step": 4664 }, { "epoch": 2.195294117647059, "grad_norm": 0.45791369361395345, "learning_rate": 0.00025676114851144375, "loss": 0.3067, "step": 4665 }, { "epoch": 2.195764705882353, "grad_norm": 0.2391365104942213, "learning_rate": 0.0002567303515813928, "loss": 0.1526, "step": 4666 }, { "epoch": 2.1962352941176473, "grad_norm": 0.330364035823961, "learning_rate": 0.0002566995455360463, "loss": 0.2955, "step": 4667 }, { "epoch": 2.196705882352941, "grad_norm": 0.23097026085223876, "learning_rate": 0.00025666873037803517, "loss": 0.1191, "step": 4668 }, { "epoch": 2.1971764705882353, "grad_norm": 0.3125232031871703, "learning_rate": 0.00025663790610999134, "loss": 0.1908, "step": 4669 }, { "epoch": 2.1976470588235295, "grad_norm": 0.10330624154235517, "learning_rate": 0.0002566070727345473, "loss": 0.0592, "step": 4670 }, { "epoch": 2.1981176470588237, "grad_norm": 0.18865176832641828, "learning_rate": 0.00025657623025433627, "loss": 0.0958, "step": 4671 }, { "epoch": 2.1985882352941175, "grad_norm": 0.3157427184798666, "learning_rate": 0.0002565453786719924, "loss": 0.2931, "step": 4672 }, { "epoch": 2.1990588235294117, "grad_norm": 0.24363737689963189, "learning_rate": 0.00025651451799015067, "loss": 0.2038, "step": 4673 }, { "epoch": 2.199529411764706, "grad_norm": 0.22931153078960412, "learning_rate": 0.0002564836482114466, "loss": 0.1435, "step": 4674 }, { "epoch": 2.2, "grad_norm": 0.7223857058717902, "learning_rate": 0.00025645276933851667, "loss": 0.5302, "step": 4675 }, { "epoch": 2.200470588235294, "grad_norm": 0.37567621825512254, "learning_rate": 0.00025642188137399815, "loss": 0.1921, "step": 4676 }, { "epoch": 2.200941176470588, "grad_norm": 0.2836348384498493, "learning_rate": 0.00025639098432052895, "loss": 0.1683, "step": 4677 }, { "epoch": 2.2014117647058824, "grad_norm": 0.34230202457706643, "learning_rate": 0.0002563600781807478, "loss": 0.2702, "step": 4678 }, { "epoch": 2.2018823529411766, "grad_norm": 0.3291687879141634, "learning_rate": 0.0002563291629572943, "loss": 0.3095, "step": 4679 }, { "epoch": 2.2023529411764704, "grad_norm": 0.311200857662015, "learning_rate": 0.00025629823865280874, "loss": 0.2868, "step": 4680 }, { "epoch": 2.2028235294117646, "grad_norm": 0.40374990951415, "learning_rate": 0.00025626730526993223, "loss": 0.396, "step": 4681 }, { "epoch": 2.203294117647059, "grad_norm": 0.40712161188037693, "learning_rate": 0.00025623636281130657, "loss": 0.4901, "step": 4682 }, { "epoch": 2.203764705882353, "grad_norm": 0.35469721910480834, "learning_rate": 0.0002562054112795745, "loss": 0.3508, "step": 4683 }, { "epoch": 2.2042352941176473, "grad_norm": 0.2720340872330042, "learning_rate": 0.00025617445067737933, "loss": 0.3705, "step": 4684 }, { "epoch": 2.204705882352941, "grad_norm": 0.1721051990716073, "learning_rate": 0.0002561434810073653, "loss": 0.0878, "step": 4685 }, { "epoch": 2.2051764705882353, "grad_norm": 0.340177186481845, "learning_rate": 0.00025611250227217737, "loss": 0.315, "step": 4686 }, { "epoch": 2.2056470588235295, "grad_norm": 0.37300910044165847, "learning_rate": 0.0002560815144744612, "loss": 0.3016, "step": 4687 }, { "epoch": 2.2061176470588233, "grad_norm": 0.2102512881566101, "learning_rate": 0.00025605051761686345, "loss": 0.1293, "step": 4688 }, { "epoch": 2.2065882352941175, "grad_norm": 0.290585027410434, "learning_rate": 0.0002560195117020313, "loss": 0.3283, "step": 4689 }, { "epoch": 2.2070588235294117, "grad_norm": 0.37555564165156635, "learning_rate": 0.0002559884967326128, "loss": 0.212, "step": 4690 }, { "epoch": 2.207529411764706, "grad_norm": 0.30658705186640756, "learning_rate": 0.0002559574727112569, "loss": 0.3427, "step": 4691 }, { "epoch": 2.208, "grad_norm": 0.20656469241777378, "learning_rate": 0.000255926439640613, "loss": 0.1562, "step": 4692 }, { "epoch": 2.208470588235294, "grad_norm": 0.45851480973553405, "learning_rate": 0.00025589539752333165, "loss": 0.3609, "step": 4693 }, { "epoch": 2.208941176470588, "grad_norm": 0.22117024742182412, "learning_rate": 0.0002558643463620639, "loss": 0.2223, "step": 4694 }, { "epoch": 2.2094117647058824, "grad_norm": 0.46918617965104187, "learning_rate": 0.0002558332861594617, "loss": 0.3933, "step": 4695 }, { "epoch": 2.2098823529411766, "grad_norm": 0.15177322396122606, "learning_rate": 0.0002558022169181778, "loss": 0.0858, "step": 4696 }, { "epoch": 2.2103529411764704, "grad_norm": 0.2974988664107675, "learning_rate": 0.00025577113864086557, "loss": 0.2656, "step": 4697 }, { "epoch": 2.2108235294117646, "grad_norm": 0.2764280242361155, "learning_rate": 0.0002557400513301794, "loss": 0.1181, "step": 4698 }, { "epoch": 2.211294117647059, "grad_norm": 0.25720553430581455, "learning_rate": 0.00025570895498877404, "loss": 0.2204, "step": 4699 }, { "epoch": 2.211764705882353, "grad_norm": 0.342688792223768, "learning_rate": 0.00025567784961930546, "loss": 0.3392, "step": 4700 }, { "epoch": 2.212235294117647, "grad_norm": 0.23391875859614977, "learning_rate": 0.0002556467352244302, "loss": 0.2052, "step": 4701 }, { "epoch": 2.212705882352941, "grad_norm": 0.47853532619047334, "learning_rate": 0.00025561561180680563, "loss": 0.3633, "step": 4702 }, { "epoch": 2.2131764705882353, "grad_norm": 0.2601398000567885, "learning_rate": 0.0002555844793690897, "loss": 0.1799, "step": 4703 }, { "epoch": 2.2136470588235295, "grad_norm": 0.500779275220782, "learning_rate": 0.0002555533379139414, "loss": 0.6055, "step": 4704 }, { "epoch": 2.2141176470588233, "grad_norm": 0.26211385945720855, "learning_rate": 0.00025552218744402026, "loss": 0.2231, "step": 4705 }, { "epoch": 2.2145882352941175, "grad_norm": 0.2669900571985707, "learning_rate": 0.00025549102796198673, "loss": 0.2621, "step": 4706 }, { "epoch": 2.2150588235294117, "grad_norm": 0.3793537204719796, "learning_rate": 0.00025545985947050197, "loss": 0.3923, "step": 4707 }, { "epoch": 2.215529411764706, "grad_norm": 0.3239816253460636, "learning_rate": 0.00025542868197222797, "loss": 0.2817, "step": 4708 }, { "epoch": 2.216, "grad_norm": 0.2507888651449082, "learning_rate": 0.00025539749546982736, "loss": 0.1843, "step": 4709 }, { "epoch": 2.216470588235294, "grad_norm": 0.2963120747159017, "learning_rate": 0.00025536629996596367, "loss": 0.2354, "step": 4710 }, { "epoch": 2.216941176470588, "grad_norm": 0.4394700093500952, "learning_rate": 0.0002553350954633012, "loss": 0.6382, "step": 4711 }, { "epoch": 2.2174117647058824, "grad_norm": 0.36743682426030205, "learning_rate": 0.00025530388196450487, "loss": 0.3267, "step": 4712 }, { "epoch": 2.2178823529411766, "grad_norm": 0.4841822223861209, "learning_rate": 0.0002552726594722405, "loss": 0.5588, "step": 4713 }, { "epoch": 2.2183529411764704, "grad_norm": 0.44791999311608455, "learning_rate": 0.00025524142798917466, "loss": 0.3741, "step": 4714 }, { "epoch": 2.2188235294117646, "grad_norm": 0.22744268670611517, "learning_rate": 0.00025521018751797464, "loss": 0.1336, "step": 4715 }, { "epoch": 2.219294117647059, "grad_norm": 0.1890123588139492, "learning_rate": 0.0002551789380613086, "loss": 0.1363, "step": 4716 }, { "epoch": 2.219764705882353, "grad_norm": 0.4010443336634051, "learning_rate": 0.00025514767962184535, "loss": 0.218, "step": 4717 }, { "epoch": 2.220235294117647, "grad_norm": 0.18239849534554783, "learning_rate": 0.0002551164122022545, "loss": 0.0739, "step": 4718 }, { "epoch": 2.220705882352941, "grad_norm": 0.42599861241307896, "learning_rate": 0.00025508513580520644, "loss": 0.3683, "step": 4719 }, { "epoch": 2.2211764705882353, "grad_norm": 0.4112998421178795, "learning_rate": 0.00025505385043337236, "loss": 0.3513, "step": 4720 }, { "epoch": 2.2216470588235295, "grad_norm": 0.4058198224438249, "learning_rate": 0.0002550225560894241, "loss": 0.2826, "step": 4721 }, { "epoch": 2.2221176470588233, "grad_norm": 0.4882750904055555, "learning_rate": 0.00025499125277603446, "loss": 0.3568, "step": 4722 }, { "epoch": 2.2225882352941175, "grad_norm": 0.24385669560506165, "learning_rate": 0.0002549599404958769, "loss": 0.1438, "step": 4723 }, { "epoch": 2.2230588235294118, "grad_norm": 0.4085425021538197, "learning_rate": 0.0002549286192516256, "loss": 0.2424, "step": 4724 }, { "epoch": 2.223529411764706, "grad_norm": 0.24105548159907816, "learning_rate": 0.0002548972890459555, "loss": 0.1469, "step": 4725 }, { "epoch": 2.224, "grad_norm": 0.21076322908106063, "learning_rate": 0.00025486594988154235, "loss": 0.13, "step": 4726 }, { "epoch": 2.224470588235294, "grad_norm": 0.24720017559363955, "learning_rate": 0.0002548346017610628, "loss": 0.1381, "step": 4727 }, { "epoch": 2.224941176470588, "grad_norm": 0.12734289811165536, "learning_rate": 0.000254803244687194, "loss": 0.0971, "step": 4728 }, { "epoch": 2.2254117647058824, "grad_norm": 0.3623029841576748, "learning_rate": 0.0002547718786626141, "loss": 0.317, "step": 4729 }, { "epoch": 2.2258823529411766, "grad_norm": 0.29542291230540624, "learning_rate": 0.0002547405036900018, "loss": 0.2747, "step": 4730 }, { "epoch": 2.2263529411764704, "grad_norm": 0.3318662360942286, "learning_rate": 0.0002547091197720368, "loss": 0.3621, "step": 4731 }, { "epoch": 2.2268235294117646, "grad_norm": 0.3289773039813934, "learning_rate": 0.0002546777269113993, "loss": 0.1849, "step": 4732 }, { "epoch": 2.227294117647059, "grad_norm": 0.37018697002038203, "learning_rate": 0.0002546463251107705, "loss": 0.3209, "step": 4733 }, { "epoch": 2.227764705882353, "grad_norm": 0.26631672351025343, "learning_rate": 0.0002546149143728323, "loss": 0.1781, "step": 4734 }, { "epoch": 2.228235294117647, "grad_norm": 0.3854337654155146, "learning_rate": 0.0002545834947002672, "loss": 0.2999, "step": 4735 }, { "epoch": 2.228705882352941, "grad_norm": 0.46343109690569206, "learning_rate": 0.00025455206609575876, "loss": 0.3282, "step": 4736 }, { "epoch": 2.2291764705882353, "grad_norm": 0.2489211992844377, "learning_rate": 0.000254520628561991, "loss": 0.1721, "step": 4737 }, { "epoch": 2.2296470588235295, "grad_norm": 0.0549364080388569, "learning_rate": 0.00025448918210164886, "loss": 0.018, "step": 4738 }, { "epoch": 2.2301176470588233, "grad_norm": 0.3775224387216654, "learning_rate": 0.00025445772671741816, "loss": 0.3364, "step": 4739 }, { "epoch": 2.2305882352941175, "grad_norm": 0.49379233639679987, "learning_rate": 0.00025442626241198513, "loss": 0.4653, "step": 4740 }, { "epoch": 2.2310588235294118, "grad_norm": 0.31604912689500875, "learning_rate": 0.00025439478918803715, "loss": 0.3498, "step": 4741 }, { "epoch": 2.231529411764706, "grad_norm": 0.26596146939137616, "learning_rate": 0.00025436330704826205, "loss": 0.3638, "step": 4742 }, { "epoch": 2.232, "grad_norm": 0.30387327409282444, "learning_rate": 0.0002543318159953487, "loss": 0.2086, "step": 4743 }, { "epoch": 2.232470588235294, "grad_norm": 0.2576708807678842, "learning_rate": 0.00025430031603198644, "loss": 0.2219, "step": 4744 }, { "epoch": 2.232941176470588, "grad_norm": 0.13625015774077567, "learning_rate": 0.0002542688071608657, "loss": 0.0408, "step": 4745 }, { "epoch": 2.2334117647058824, "grad_norm": 0.27078733748316364, "learning_rate": 0.0002542372893846773, "loss": 0.1912, "step": 4746 }, { "epoch": 2.2338823529411767, "grad_norm": 0.6394449087381028, "learning_rate": 0.00025420576270611313, "loss": 0.5597, "step": 4747 }, { "epoch": 2.2343529411764704, "grad_norm": 0.37571340533743947, "learning_rate": 0.00025417422712786575, "loss": 0.3155, "step": 4748 }, { "epoch": 2.2348235294117647, "grad_norm": 0.4881601497000609, "learning_rate": 0.00025414268265262833, "loss": 0.5109, "step": 4749 }, { "epoch": 2.235294117647059, "grad_norm": 0.2828476989497557, "learning_rate": 0.0002541111292830951, "loss": 0.314, "step": 4750 }, { "epoch": 2.235764705882353, "grad_norm": 0.4586875609820017, "learning_rate": 0.0002540795670219606, "loss": 0.3604, "step": 4751 }, { "epoch": 2.236235294117647, "grad_norm": 0.1708070312441645, "learning_rate": 0.0002540479958719207, "loss": 0.1423, "step": 4752 }, { "epoch": 2.236705882352941, "grad_norm": 0.23409233694937498, "learning_rate": 0.0002540164158356716, "loss": 0.1341, "step": 4753 }, { "epoch": 2.2371764705882353, "grad_norm": 0.43032056711162936, "learning_rate": 0.0002539848269159104, "loss": 0.4986, "step": 4754 }, { "epoch": 2.2376470588235295, "grad_norm": 0.33124882630003183, "learning_rate": 0.00025395322911533484, "loss": 0.2696, "step": 4755 }, { "epoch": 2.2381176470588233, "grad_norm": 0.26775177650451015, "learning_rate": 0.0002539216224366437, "loss": 0.1492, "step": 4756 }, { "epoch": 2.2385882352941175, "grad_norm": 0.2255967033521654, "learning_rate": 0.00025389000688253626, "loss": 0.2108, "step": 4757 }, { "epoch": 2.2390588235294118, "grad_norm": 0.38609791240632063, "learning_rate": 0.00025385838245571264, "loss": 0.4367, "step": 4758 }, { "epoch": 2.239529411764706, "grad_norm": 0.42204600970681905, "learning_rate": 0.00025382674915887374, "loss": 0.3417, "step": 4759 }, { "epoch": 2.24, "grad_norm": 0.4369167819203185, "learning_rate": 0.0002537951069947212, "loss": 0.4793, "step": 4760 }, { "epoch": 2.240470588235294, "grad_norm": 0.19371211498928903, "learning_rate": 0.00025376345596595743, "loss": 0.1336, "step": 4761 }, { "epoch": 2.240941176470588, "grad_norm": 0.2868592926507047, "learning_rate": 0.0002537317960752855, "loss": 0.2771, "step": 4762 }, { "epoch": 2.2414117647058824, "grad_norm": 0.29248491871340143, "learning_rate": 0.0002537001273254094, "loss": 0.2001, "step": 4763 }, { "epoch": 2.2418823529411767, "grad_norm": 0.15812018009472828, "learning_rate": 0.00025366844971903385, "loss": 0.099, "step": 4764 }, { "epoch": 2.2423529411764704, "grad_norm": 0.28489975988493593, "learning_rate": 0.00025363676325886416, "loss": 0.2651, "step": 4765 }, { "epoch": 2.2428235294117647, "grad_norm": 0.34891467449590163, "learning_rate": 0.00025360506794760653, "loss": 0.4279, "step": 4766 }, { "epoch": 2.243294117647059, "grad_norm": 0.3711290426160293, "learning_rate": 0.00025357336378796795, "loss": 0.4373, "step": 4767 }, { "epoch": 2.243764705882353, "grad_norm": 0.22165496145819674, "learning_rate": 0.0002535416507826561, "loss": 0.1669, "step": 4768 }, { "epoch": 2.244235294117647, "grad_norm": 0.29021735197375137, "learning_rate": 0.0002535099289343793, "loss": 0.2027, "step": 4769 }, { "epoch": 2.244705882352941, "grad_norm": 0.19271005328181243, "learning_rate": 0.00025347819824584695, "loss": 0.1499, "step": 4770 }, { "epoch": 2.2451764705882353, "grad_norm": 0.0867751538387381, "learning_rate": 0.0002534464587197689, "loss": 0.0424, "step": 4771 }, { "epoch": 2.2456470588235296, "grad_norm": 0.34666735551570155, "learning_rate": 0.00025341471035885587, "loss": 0.2918, "step": 4772 }, { "epoch": 2.2461176470588233, "grad_norm": 0.5926955367742855, "learning_rate": 0.0002533829531658193, "loss": 0.6591, "step": 4773 }, { "epoch": 2.2465882352941176, "grad_norm": 0.3544354357240242, "learning_rate": 0.00025335118714337146, "loss": 0.4082, "step": 4774 }, { "epoch": 2.2470588235294118, "grad_norm": 0.4580203348129367, "learning_rate": 0.0002533194122942253, "loss": 0.5351, "step": 4775 }, { "epoch": 2.247529411764706, "grad_norm": 0.2646871629561811, "learning_rate": 0.00025328762862109457, "loss": 0.2413, "step": 4776 }, { "epoch": 2.248, "grad_norm": 0.21821892465869536, "learning_rate": 0.00025325583612669373, "loss": 0.1244, "step": 4777 }, { "epoch": 2.248470588235294, "grad_norm": 0.2099349371208721, "learning_rate": 0.00025322403481373797, "loss": 0.142, "step": 4778 }, { "epoch": 2.248941176470588, "grad_norm": 0.3923407772670839, "learning_rate": 0.00025319222468494335, "loss": 0.3165, "step": 4779 }, { "epoch": 2.2494117647058824, "grad_norm": 0.37071152195972273, "learning_rate": 0.00025316040574302663, "loss": 0.2696, "step": 4780 }, { "epoch": 2.2498823529411767, "grad_norm": 0.20379113908986765, "learning_rate": 0.00025312857799070524, "loss": 0.1147, "step": 4781 }, { "epoch": 2.2503529411764704, "grad_norm": 0.28524431244889814, "learning_rate": 0.0002530967414306974, "loss": 0.1842, "step": 4782 }, { "epoch": 2.2508235294117647, "grad_norm": 0.22457487536473195, "learning_rate": 0.0002530648960657222, "loss": 0.1164, "step": 4783 }, { "epoch": 2.251294117647059, "grad_norm": 0.18556650515686196, "learning_rate": 0.0002530330418984993, "loss": 0.0936, "step": 4784 }, { "epoch": 2.251764705882353, "grad_norm": 0.4855529173778703, "learning_rate": 0.00025300117893174934, "loss": 0.5499, "step": 4785 }, { "epoch": 2.252235294117647, "grad_norm": 0.25645182590367394, "learning_rate": 0.0002529693071681934, "loss": 0.1508, "step": 4786 }, { "epoch": 2.252705882352941, "grad_norm": 0.1483534857724999, "learning_rate": 0.00025293742661055364, "loss": 0.0843, "step": 4787 }, { "epoch": 2.2531764705882353, "grad_norm": 0.16568046111336812, "learning_rate": 0.0002529055372615527, "loss": 0.1416, "step": 4788 }, { "epoch": 2.2536470588235296, "grad_norm": 0.48904425061765106, "learning_rate": 0.00025287363912391403, "loss": 0.594, "step": 4789 }, { "epoch": 2.2541176470588233, "grad_norm": 0.29812251579642157, "learning_rate": 0.00025284173220036215, "loss": 0.2609, "step": 4790 }, { "epoch": 2.2545882352941176, "grad_norm": 0.36608877628016206, "learning_rate": 0.00025280981649362184, "loss": 0.3738, "step": 4791 }, { "epoch": 2.255058823529412, "grad_norm": 0.4179643145469768, "learning_rate": 0.0002527778920064189, "loss": 0.1553, "step": 4792 }, { "epoch": 2.255529411764706, "grad_norm": 0.2548487260114977, "learning_rate": 0.00025274595874147984, "loss": 0.2429, "step": 4793 }, { "epoch": 2.2560000000000002, "grad_norm": 0.24848247171272886, "learning_rate": 0.000252714016701532, "loss": 0.1523, "step": 4794 }, { "epoch": 2.256470588235294, "grad_norm": 0.37785582691279573, "learning_rate": 0.0002526820658893033, "loss": 0.3732, "step": 4795 }, { "epoch": 2.2569411764705882, "grad_norm": 0.21055669394102672, "learning_rate": 0.0002526501063075225, "loss": 0.1611, "step": 4796 }, { "epoch": 2.2574117647058825, "grad_norm": 0.3587241094689314, "learning_rate": 0.00025261813795891916, "loss": 0.3024, "step": 4797 }, { "epoch": 2.2578823529411762, "grad_norm": 0.25087093605651783, "learning_rate": 0.00025258616084622345, "loss": 0.2353, "step": 4798 }, { "epoch": 2.2583529411764705, "grad_norm": 0.3717880804894938, "learning_rate": 0.00025255417497216647, "loss": 0.3079, "step": 4799 }, { "epoch": 2.2588235294117647, "grad_norm": 0.3446376267991737, "learning_rate": 0.00025252218033947993, "loss": 0.4101, "step": 4800 }, { "epoch": 2.259294117647059, "grad_norm": 0.3348938884358168, "learning_rate": 0.0002524901769508963, "loss": 0.3354, "step": 4801 }, { "epoch": 2.259764705882353, "grad_norm": 0.4896023343823803, "learning_rate": 0.0002524581648091488, "loss": 0.5521, "step": 4802 }, { "epoch": 2.260235294117647, "grad_norm": 0.33746844699028095, "learning_rate": 0.00025242614391697153, "loss": 0.373, "step": 4803 }, { "epoch": 2.260705882352941, "grad_norm": 0.41013584817089344, "learning_rate": 0.00025239411427709915, "loss": 0.5094, "step": 4804 }, { "epoch": 2.2611764705882353, "grad_norm": 0.29172581025057703, "learning_rate": 0.0002523620758922672, "loss": 0.1268, "step": 4805 }, { "epoch": 2.2616470588235296, "grad_norm": 0.11762795681052743, "learning_rate": 0.00025233002876521183, "loss": 0.0274, "step": 4806 }, { "epoch": 2.2621176470588233, "grad_norm": 0.23594604267395908, "learning_rate": 0.0002522979728986701, "loss": 0.094, "step": 4807 }, { "epoch": 2.2625882352941176, "grad_norm": 0.24311776999634216, "learning_rate": 0.00025226590829537977, "loss": 0.1853, "step": 4808 }, { "epoch": 2.263058823529412, "grad_norm": 0.5799639064579613, "learning_rate": 0.0002522338349580792, "loss": 0.3635, "step": 4809 }, { "epoch": 2.263529411764706, "grad_norm": 0.3103811139823175, "learning_rate": 0.00025220175288950765, "loss": 0.2086, "step": 4810 }, { "epoch": 2.2640000000000002, "grad_norm": 0.36642629181296377, "learning_rate": 0.00025216966209240516, "loss": 0.352, "step": 4811 }, { "epoch": 2.264470588235294, "grad_norm": 0.17421897859753407, "learning_rate": 0.00025213756256951234, "loss": 0.1101, "step": 4812 }, { "epoch": 2.2649411764705882, "grad_norm": 0.46634245624180737, "learning_rate": 0.00025210545432357076, "loss": 0.5056, "step": 4813 }, { "epoch": 2.2654117647058825, "grad_norm": 0.28258844383573684, "learning_rate": 0.0002520733373573225, "loss": 0.2039, "step": 4814 }, { "epoch": 2.2658823529411762, "grad_norm": 0.16493898160863482, "learning_rate": 0.0002520412116735106, "loss": 0.0709, "step": 4815 }, { "epoch": 2.2663529411764705, "grad_norm": 0.17568119612722438, "learning_rate": 0.0002520090772748787, "loss": 0.1389, "step": 4816 }, { "epoch": 2.2668235294117647, "grad_norm": 0.37294766709916566, "learning_rate": 0.00025197693416417126, "loss": 0.3406, "step": 4817 }, { "epoch": 2.267294117647059, "grad_norm": 0.23030678031730364, "learning_rate": 0.0002519447823441334, "loss": 0.1658, "step": 4818 }, { "epoch": 2.267764705882353, "grad_norm": 0.4003529527345472, "learning_rate": 0.00025191262181751113, "loss": 0.2848, "step": 4819 }, { "epoch": 2.268235294117647, "grad_norm": 0.43938264287057394, "learning_rate": 0.00025188045258705107, "loss": 0.4842, "step": 4820 }, { "epoch": 2.268705882352941, "grad_norm": 0.323594299591261, "learning_rate": 0.0002518482746555007, "loss": 0.2747, "step": 4821 }, { "epoch": 2.2691764705882354, "grad_norm": 0.22129080904256687, "learning_rate": 0.000251816088025608, "loss": 0.112, "step": 4822 }, { "epoch": 2.2696470588235296, "grad_norm": 0.336510642570928, "learning_rate": 0.00025178389270012205, "loss": 0.4238, "step": 4823 }, { "epoch": 2.2701176470588234, "grad_norm": 0.24755620768935663, "learning_rate": 0.0002517516886817924, "loss": 0.1063, "step": 4824 }, { "epoch": 2.2705882352941176, "grad_norm": 0.289332253384724, "learning_rate": 0.00025171947597336946, "loss": 0.2372, "step": 4825 }, { "epoch": 2.271058823529412, "grad_norm": 0.2306199950569873, "learning_rate": 0.00025168725457760434, "loss": 0.2207, "step": 4826 }, { "epoch": 2.271529411764706, "grad_norm": 0.1433155034020053, "learning_rate": 0.0002516550244972489, "loss": 0.0732, "step": 4827 }, { "epoch": 2.2720000000000002, "grad_norm": 0.391032947341392, "learning_rate": 0.00025162278573505575, "loss": 0.3899, "step": 4828 }, { "epoch": 2.272470588235294, "grad_norm": 0.17288742635817164, "learning_rate": 0.0002515905382937782, "loss": 0.1067, "step": 4829 }, { "epoch": 2.2729411764705882, "grad_norm": 0.44696646974590665, "learning_rate": 0.0002515582821761704, "loss": 0.4772, "step": 4830 }, { "epoch": 2.2734117647058825, "grad_norm": 0.2557408263909618, "learning_rate": 0.0002515260173849872, "loss": 0.1427, "step": 4831 }, { "epoch": 2.2738823529411762, "grad_norm": 0.5506771559181254, "learning_rate": 0.00025149374392298407, "loss": 0.3506, "step": 4832 }, { "epoch": 2.2743529411764705, "grad_norm": 0.13723191923144187, "learning_rate": 0.00025146146179291746, "loss": 0.1094, "step": 4833 }, { "epoch": 2.2748235294117647, "grad_norm": 0.40900338826289084, "learning_rate": 0.0002514291709975443, "loss": 0.554, "step": 4834 }, { "epoch": 2.275294117647059, "grad_norm": 0.38841744173471904, "learning_rate": 0.00025139687153962246, "loss": 0.2503, "step": 4835 }, { "epoch": 2.275764705882353, "grad_norm": 0.3779087234106391, "learning_rate": 0.00025136456342191043, "loss": 0.2201, "step": 4836 }, { "epoch": 2.276235294117647, "grad_norm": 0.1954075913647251, "learning_rate": 0.0002513322466471675, "loss": 0.1089, "step": 4837 }, { "epoch": 2.276705882352941, "grad_norm": 0.33096249976306524, "learning_rate": 0.00025129992121815365, "loss": 0.2732, "step": 4838 }, { "epoch": 2.2771764705882354, "grad_norm": 0.385297645216811, "learning_rate": 0.00025126758713762964, "loss": 0.3521, "step": 4839 }, { "epoch": 2.2776470588235296, "grad_norm": 0.35298327805250423, "learning_rate": 0.000251235244408357, "loss": 0.3105, "step": 4840 }, { "epoch": 2.2781176470588234, "grad_norm": 0.4545697081524007, "learning_rate": 0.00025120289303309795, "loss": 0.3519, "step": 4841 }, { "epoch": 2.2785882352941176, "grad_norm": 0.13566305880819643, "learning_rate": 0.00025117053301461546, "loss": 0.1013, "step": 4842 }, { "epoch": 2.279058823529412, "grad_norm": 0.35197335784443473, "learning_rate": 0.00025113816435567315, "loss": 0.3692, "step": 4843 }, { "epoch": 2.279529411764706, "grad_norm": 0.2824746441330864, "learning_rate": 0.00025110578705903557, "loss": 0.1639, "step": 4844 }, { "epoch": 2.2800000000000002, "grad_norm": 0.5948836750468691, "learning_rate": 0.0002510734011274678, "loss": 0.5775, "step": 4845 }, { "epoch": 2.280470588235294, "grad_norm": 0.3090033546266535, "learning_rate": 0.00025104100656373585, "loss": 0.1485, "step": 4846 }, { "epoch": 2.2809411764705882, "grad_norm": 0.2357816817468262, "learning_rate": 0.0002510086033706063, "loss": 0.1518, "step": 4847 }, { "epoch": 2.2814117647058825, "grad_norm": 0.2465626180286663, "learning_rate": 0.0002509761915508466, "loss": 0.0993, "step": 4848 }, { "epoch": 2.2818823529411763, "grad_norm": 0.12082682298799477, "learning_rate": 0.0002509437711072249, "loss": 0.0416, "step": 4849 }, { "epoch": 2.2823529411764705, "grad_norm": 0.9567418927652968, "learning_rate": 0.00025091134204250997, "loss": 0.6012, "step": 4850 }, { "epoch": 2.2828235294117647, "grad_norm": 0.29604592381252554, "learning_rate": 0.0002508789043594715, "loss": 0.2632, "step": 4851 }, { "epoch": 2.283294117647059, "grad_norm": 0.43057124986709844, "learning_rate": 0.00025084645806087974, "loss": 0.2515, "step": 4852 }, { "epoch": 2.283764705882353, "grad_norm": 0.10634856675963239, "learning_rate": 0.00025081400314950586, "loss": 0.0251, "step": 4853 }, { "epoch": 2.284235294117647, "grad_norm": 0.18866583237341866, "learning_rate": 0.0002507815396281216, "loss": 0.1269, "step": 4854 }, { "epoch": 2.284705882352941, "grad_norm": 0.22347222880012294, "learning_rate": 0.0002507490674994995, "loss": 0.1187, "step": 4855 }, { "epoch": 2.2851764705882354, "grad_norm": 0.20752909086548507, "learning_rate": 0.0002507165867664129, "loss": 0.1411, "step": 4856 }, { "epoch": 2.2856470588235296, "grad_norm": 0.6532032819720454, "learning_rate": 0.0002506840974316357, "loss": 0.3597, "step": 4857 }, { "epoch": 2.2861176470588234, "grad_norm": 0.15822722643628087, "learning_rate": 0.00025065159949794283, "loss": 0.0697, "step": 4858 }, { "epoch": 2.2865882352941176, "grad_norm": 0.28799266071366597, "learning_rate": 0.00025061909296810966, "loss": 0.2855, "step": 4859 }, { "epoch": 2.287058823529412, "grad_norm": 0.15920603583899026, "learning_rate": 0.0002505865778449124, "loss": 0.1125, "step": 4860 }, { "epoch": 2.287529411764706, "grad_norm": 0.4001893926262672, "learning_rate": 0.000250554054131128, "loss": 0.4167, "step": 4861 }, { "epoch": 2.288, "grad_norm": 0.4825105801046649, "learning_rate": 0.00025052152182953414, "loss": 0.4747, "step": 4862 }, { "epoch": 2.288470588235294, "grad_norm": 0.06872204581185076, "learning_rate": 0.0002504889809429093, "loss": 0.0391, "step": 4863 }, { "epoch": 2.2889411764705883, "grad_norm": 0.1973230477316859, "learning_rate": 0.0002504564314740326, "loss": 0.1406, "step": 4864 }, { "epoch": 2.2894117647058825, "grad_norm": 0.2640364451491238, "learning_rate": 0.0002504238734256839, "loss": 0.17, "step": 4865 }, { "epoch": 2.2898823529411763, "grad_norm": 0.1355974007396973, "learning_rate": 0.0002503913068006438, "loss": 0.077, "step": 4866 }, { "epoch": 2.2903529411764705, "grad_norm": 0.25974557302072565, "learning_rate": 0.00025035873160169373, "loss": 0.1722, "step": 4867 }, { "epoch": 2.2908235294117647, "grad_norm": 0.54033386258407, "learning_rate": 0.0002503261478316157, "loss": 0.5636, "step": 4868 }, { "epoch": 2.291294117647059, "grad_norm": 0.19883387319500995, "learning_rate": 0.0002502935554931926, "loss": 0.1686, "step": 4869 }, { "epoch": 2.291764705882353, "grad_norm": 0.420390496321319, "learning_rate": 0.0002502609545892079, "loss": 0.2472, "step": 4870 }, { "epoch": 2.292235294117647, "grad_norm": 0.43753688402170776, "learning_rate": 0.00025022834512244583, "loss": 0.2459, "step": 4871 }, { "epoch": 2.292705882352941, "grad_norm": 0.3464325299181254, "learning_rate": 0.00025019572709569153, "loss": 0.2992, "step": 4872 }, { "epoch": 2.2931764705882354, "grad_norm": 0.4408708171398003, "learning_rate": 0.0002501631005117307, "loss": 0.3379, "step": 4873 }, { "epoch": 2.2936470588235296, "grad_norm": 0.6115005111914803, "learning_rate": 0.0002501304653733497, "loss": 0.5021, "step": 4874 }, { "epoch": 2.2941176470588234, "grad_norm": 0.2966213533104382, "learning_rate": 0.0002500978216833359, "loss": 0.2562, "step": 4875 }, { "epoch": 2.2945882352941176, "grad_norm": 0.14874651324531066, "learning_rate": 0.0002500651694444771, "loss": 0.0947, "step": 4876 }, { "epoch": 2.295058823529412, "grad_norm": 0.2999150873269076, "learning_rate": 0.00025003250865956203, "loss": 0.2058, "step": 4877 }, { "epoch": 2.295529411764706, "grad_norm": 0.6132062692516187, "learning_rate": 0.0002499998393313801, "loss": 0.427, "step": 4878 }, { "epoch": 2.296, "grad_norm": 0.3956542304729692, "learning_rate": 0.00024996716146272137, "loss": 0.2914, "step": 4879 }, { "epoch": 2.296470588235294, "grad_norm": 0.46479610978689007, "learning_rate": 0.0002499344750563767, "loss": 0.2684, "step": 4880 }, { "epoch": 2.2969411764705883, "grad_norm": 0.2843875296990527, "learning_rate": 0.0002499017801151377, "loss": 0.2043, "step": 4881 }, { "epoch": 2.2974117647058825, "grad_norm": 0.27750775054122884, "learning_rate": 0.0002498690766417967, "loss": 0.2318, "step": 4882 }, { "epoch": 2.2978823529411763, "grad_norm": 0.25719137179263135, "learning_rate": 0.00024983636463914665, "loss": 0.1751, "step": 4883 }, { "epoch": 2.2983529411764705, "grad_norm": 0.54306583405054, "learning_rate": 0.00024980364410998133, "loss": 0.3543, "step": 4884 }, { "epoch": 2.2988235294117647, "grad_norm": 0.65717342377304, "learning_rate": 0.0002497709150570953, "loss": 0.304, "step": 4885 }, { "epoch": 2.299294117647059, "grad_norm": 0.18298603816362854, "learning_rate": 0.0002497381774832838, "loss": 0.1208, "step": 4886 }, { "epoch": 2.299764705882353, "grad_norm": 0.2904904531650668, "learning_rate": 0.00024970543139134275, "loss": 0.2612, "step": 4887 }, { "epoch": 2.300235294117647, "grad_norm": 0.1586495733773666, "learning_rate": 0.0002496726767840687, "loss": 0.0873, "step": 4888 }, { "epoch": 2.300705882352941, "grad_norm": 0.3830913293404635, "learning_rate": 0.0002496399136642592, "loss": 0.2693, "step": 4889 }, { "epoch": 2.3011764705882354, "grad_norm": 0.5466307766001249, "learning_rate": 0.0002496071420347124, "loss": 0.3197, "step": 4890 }, { "epoch": 2.3016470588235296, "grad_norm": 0.42095498678437443, "learning_rate": 0.00024957436189822706, "loss": 0.2493, "step": 4891 }, { "epoch": 2.3021176470588234, "grad_norm": 0.31203208567318824, "learning_rate": 0.0002495415732576028, "loss": 0.2457, "step": 4892 }, { "epoch": 2.3025882352941176, "grad_norm": 0.4710470543965348, "learning_rate": 0.00024950877611564, "loss": 0.3759, "step": 4893 }, { "epoch": 2.303058823529412, "grad_norm": 0.3227667582929457, "learning_rate": 0.0002494759704751395, "loss": 0.2239, "step": 4894 }, { "epoch": 2.303529411764706, "grad_norm": 0.4404267502424254, "learning_rate": 0.00024944315633890327, "loss": 0.2756, "step": 4895 }, { "epoch": 2.304, "grad_norm": 0.2126956660784838, "learning_rate": 0.0002494103337097337, "loss": 0.119, "step": 4896 }, { "epoch": 2.304470588235294, "grad_norm": 0.4945751151039071, "learning_rate": 0.00024937750259043407, "loss": 0.3993, "step": 4897 }, { "epoch": 2.3049411764705883, "grad_norm": 0.24102491451180516, "learning_rate": 0.00024934466298380824, "loss": 0.2002, "step": 4898 }, { "epoch": 2.3054117647058825, "grad_norm": 0.3209460164214311, "learning_rate": 0.0002493118148926609, "loss": 0.2409, "step": 4899 }, { "epoch": 2.3058823529411763, "grad_norm": 0.2581756804503742, "learning_rate": 0.00024927895831979745, "loss": 0.1957, "step": 4900 }, { "epoch": 2.3063529411764705, "grad_norm": 0.5272030805343711, "learning_rate": 0.000249246093268024, "loss": 0.3814, "step": 4901 }, { "epoch": 2.3068235294117647, "grad_norm": 0.39938495107444283, "learning_rate": 0.0002492132197401474, "loss": 0.2027, "step": 4902 }, { "epoch": 2.307294117647059, "grad_norm": 0.3781905368653604, "learning_rate": 0.0002491803377389752, "loss": 0.3427, "step": 4903 }, { "epoch": 2.307764705882353, "grad_norm": 0.442576687570729, "learning_rate": 0.0002491474472673157, "loss": 0.4714, "step": 4904 }, { "epoch": 2.308235294117647, "grad_norm": 0.4600914222297127, "learning_rate": 0.00024911454832797784, "loss": 0.5404, "step": 4905 }, { "epoch": 2.308705882352941, "grad_norm": 0.1994037979528105, "learning_rate": 0.0002490816409237714, "loss": 0.0735, "step": 4906 }, { "epoch": 2.3091764705882354, "grad_norm": 0.41686820986069684, "learning_rate": 0.0002490487250575069, "loss": 0.3899, "step": 4907 }, { "epoch": 2.309647058823529, "grad_norm": 0.09777010998091681, "learning_rate": 0.00024901580073199536, "loss": 0.0637, "step": 4908 }, { "epoch": 2.3101176470588234, "grad_norm": 0.316405213427638, "learning_rate": 0.0002489828679500488, "loss": 0.2112, "step": 4909 }, { "epoch": 2.3105882352941176, "grad_norm": 0.2322095383005852, "learning_rate": 0.00024894992671447983, "loss": 0.1789, "step": 4910 }, { "epoch": 2.311058823529412, "grad_norm": 0.27582084715016936, "learning_rate": 0.0002489169770281018, "loss": 0.2757, "step": 4911 }, { "epoch": 2.311529411764706, "grad_norm": 0.40062702971664815, "learning_rate": 0.0002488840188937287, "loss": 0.3353, "step": 4912 }, { "epoch": 2.312, "grad_norm": 0.32681755473568197, "learning_rate": 0.00024885105231417536, "loss": 0.4185, "step": 4913 }, { "epoch": 2.312470588235294, "grad_norm": 0.2597864645030749, "learning_rate": 0.00024881807729225727, "loss": 0.2348, "step": 4914 }, { "epoch": 2.3129411764705883, "grad_norm": 0.2693269171891923, "learning_rate": 0.00024878509383079075, "loss": 0.2219, "step": 4915 }, { "epoch": 2.3134117647058825, "grad_norm": 0.30066405958774806, "learning_rate": 0.0002487521019325926, "loss": 0.2104, "step": 4916 }, { "epoch": 2.3138823529411763, "grad_norm": 0.280209933192948, "learning_rate": 0.00024871910160048063, "loss": 0.2067, "step": 4917 }, { "epoch": 2.3143529411764705, "grad_norm": 0.3289395604509838, "learning_rate": 0.00024868609283727316, "loss": 0.2763, "step": 4918 }, { "epoch": 2.3148235294117647, "grad_norm": 0.2566856817699401, "learning_rate": 0.00024865307564578927, "loss": 0.1831, "step": 4919 }, { "epoch": 2.315294117647059, "grad_norm": 0.17689172432206837, "learning_rate": 0.00024862005002884886, "loss": 0.0677, "step": 4920 }, { "epoch": 2.315764705882353, "grad_norm": 0.31302785956355333, "learning_rate": 0.0002485870159892724, "loss": 0.1932, "step": 4921 }, { "epoch": 2.316235294117647, "grad_norm": 0.41289749115834895, "learning_rate": 0.00024855397352988127, "loss": 0.4201, "step": 4922 }, { "epoch": 2.316705882352941, "grad_norm": 0.12398960409532295, "learning_rate": 0.0002485209226534974, "loss": 0.044, "step": 4923 }, { "epoch": 2.3171764705882354, "grad_norm": 0.19662495924727263, "learning_rate": 0.00024848786336294346, "loss": 0.1338, "step": 4924 }, { "epoch": 2.317647058823529, "grad_norm": 0.26572343582158925, "learning_rate": 0.0002484547956610429, "loss": 0.2739, "step": 4925 }, { "epoch": 2.3181176470588234, "grad_norm": 0.4525256299167224, "learning_rate": 0.0002484217195506199, "loss": 0.4984, "step": 4926 }, { "epoch": 2.3185882352941176, "grad_norm": 0.3285292031819396, "learning_rate": 0.00024838863503449926, "loss": 0.2792, "step": 4927 }, { "epoch": 2.319058823529412, "grad_norm": 0.5025544084362773, "learning_rate": 0.00024835554211550664, "loss": 0.4541, "step": 4928 }, { "epoch": 2.319529411764706, "grad_norm": 0.29163119783428937, "learning_rate": 0.00024832244079646827, "loss": 0.216, "step": 4929 }, { "epoch": 2.32, "grad_norm": 0.2931735891686066, "learning_rate": 0.00024828933108021116, "loss": 0.301, "step": 4930 }, { "epoch": 2.320470588235294, "grad_norm": 0.17868308030165997, "learning_rate": 0.00024825621296956313, "loss": 0.0751, "step": 4931 }, { "epoch": 2.3209411764705883, "grad_norm": 0.17828608080160485, "learning_rate": 0.00024822308646735254, "loss": 0.0995, "step": 4932 }, { "epoch": 2.3214117647058825, "grad_norm": 0.22946308556627493, "learning_rate": 0.00024818995157640855, "loss": 0.1779, "step": 4933 }, { "epoch": 2.3218823529411763, "grad_norm": 0.3081855238656882, "learning_rate": 0.00024815680829956115, "loss": 0.1801, "step": 4934 }, { "epoch": 2.3223529411764705, "grad_norm": 0.4200254743164678, "learning_rate": 0.0002481236566396408, "loss": 0.5267, "step": 4935 }, { "epoch": 2.3228235294117647, "grad_norm": 0.40248534927852303, "learning_rate": 0.0002480904965994789, "loss": 0.3834, "step": 4936 }, { "epoch": 2.323294117647059, "grad_norm": 0.23330470849460927, "learning_rate": 0.00024805732818190747, "loss": 0.1478, "step": 4937 }, { "epoch": 2.323764705882353, "grad_norm": 0.3903961651047974, "learning_rate": 0.00024802415138975924, "loss": 0.3668, "step": 4938 }, { "epoch": 2.324235294117647, "grad_norm": 0.4668743863508739, "learning_rate": 0.00024799096622586767, "loss": 0.296, "step": 4939 }, { "epoch": 2.324705882352941, "grad_norm": 0.29622839388103056, "learning_rate": 0.00024795777269306696, "loss": 0.2869, "step": 4940 }, { "epoch": 2.3251764705882354, "grad_norm": 0.42299672603047533, "learning_rate": 0.0002479245707941919, "loss": 0.4488, "step": 4941 }, { "epoch": 2.325647058823529, "grad_norm": 0.23542292547212468, "learning_rate": 0.0002478913605320783, "loss": 0.1377, "step": 4942 }, { "epoch": 2.3261176470588234, "grad_norm": 0.5129209087053238, "learning_rate": 0.0002478581419095623, "loss": 0.5417, "step": 4943 }, { "epoch": 2.3265882352941176, "grad_norm": 0.5525303996258912, "learning_rate": 0.0002478249149294809, "loss": 0.3119, "step": 4944 }, { "epoch": 2.327058823529412, "grad_norm": 0.5314589199148015, "learning_rate": 0.00024779167959467207, "loss": 0.4987, "step": 4945 }, { "epoch": 2.327529411764706, "grad_norm": 0.21589734693799845, "learning_rate": 0.0002477584359079741, "loss": 0.1135, "step": 4946 }, { "epoch": 2.328, "grad_norm": 0.28916454863658264, "learning_rate": 0.00024772518387222616, "loss": 0.1469, "step": 4947 }, { "epoch": 2.328470588235294, "grad_norm": 0.2324434526870833, "learning_rate": 0.0002476919234902682, "loss": 0.1245, "step": 4948 }, { "epoch": 2.3289411764705883, "grad_norm": 0.527439108350402, "learning_rate": 0.0002476586547649408, "loss": 0.3555, "step": 4949 }, { "epoch": 2.3294117647058825, "grad_norm": 0.3090687952115445, "learning_rate": 0.00024762537769908535, "loss": 0.1459, "step": 4950 }, { "epoch": 2.3298823529411763, "grad_norm": 0.35874754184013785, "learning_rate": 0.0002475920922955437, "loss": 0.2406, "step": 4951 }, { "epoch": 2.3303529411764705, "grad_norm": 0.2572028302840058, "learning_rate": 0.0002475587985571587, "loss": 0.1609, "step": 4952 }, { "epoch": 2.3308235294117647, "grad_norm": 0.26270900149304866, "learning_rate": 0.0002475254964867738, "loss": 0.1435, "step": 4953 }, { "epoch": 2.331294117647059, "grad_norm": 0.5150177177881841, "learning_rate": 0.0002474921860872331, "loss": 0.3762, "step": 4954 }, { "epoch": 2.331764705882353, "grad_norm": 0.5497786380889927, "learning_rate": 0.00024745886736138154, "loss": 0.3584, "step": 4955 }, { "epoch": 2.332235294117647, "grad_norm": 0.3229810649681659, "learning_rate": 0.0002474255403120647, "loss": 0.2446, "step": 4956 }, { "epoch": 2.332705882352941, "grad_norm": 0.2536309949987645, "learning_rate": 0.0002473922049421288, "loss": 0.2374, "step": 4957 }, { "epoch": 2.3331764705882354, "grad_norm": 0.11887219247581127, "learning_rate": 0.0002473588612544209, "loss": 0.0592, "step": 4958 }, { "epoch": 2.333647058823529, "grad_norm": 0.04195881177273131, "learning_rate": 0.0002473255092517887, "loss": 0.0101, "step": 4959 }, { "epoch": 2.3341176470588234, "grad_norm": 0.18551952444013056, "learning_rate": 0.0002472921489370807, "loss": 0.1393, "step": 4960 }, { "epoch": 2.3345882352941176, "grad_norm": 0.18036001795720172, "learning_rate": 0.0002472587803131459, "loss": 0.1713, "step": 4961 }, { "epoch": 2.335058823529412, "grad_norm": 0.24592642589544184, "learning_rate": 0.0002472254033828342, "loss": 0.1244, "step": 4962 }, { "epoch": 2.335529411764706, "grad_norm": 0.3366233142989288, "learning_rate": 0.00024719201814899617, "loss": 0.3745, "step": 4963 }, { "epoch": 2.336, "grad_norm": 0.44353064868319675, "learning_rate": 0.0002471586246144831, "loss": 0.3547, "step": 4964 }, { "epoch": 2.336470588235294, "grad_norm": 0.46320992337052963, "learning_rate": 0.0002471252227821469, "loss": 0.4917, "step": 4965 }, { "epoch": 2.3369411764705883, "grad_norm": 0.31747759393271463, "learning_rate": 0.0002470918126548403, "loss": 0.2612, "step": 4966 }, { "epoch": 2.3374117647058825, "grad_norm": 0.30379131556893063, "learning_rate": 0.00024705839423541664, "loss": 0.3377, "step": 4967 }, { "epoch": 2.3378823529411763, "grad_norm": 0.3279446219828436, "learning_rate": 0.00024702496752673005, "loss": 0.2679, "step": 4968 }, { "epoch": 2.3383529411764705, "grad_norm": 0.30698082395505055, "learning_rate": 0.00024699153253163533, "loss": 0.3927, "step": 4969 }, { "epoch": 2.3388235294117647, "grad_norm": 0.3601449154054563, "learning_rate": 0.000246958089252988, "loss": 0.3466, "step": 4970 }, { "epoch": 2.339294117647059, "grad_norm": 0.16958791896057926, "learning_rate": 0.0002469246376936443, "loss": 0.123, "step": 4971 }, { "epoch": 2.339764705882353, "grad_norm": 0.05537632280056992, "learning_rate": 0.00024689117785646106, "loss": 0.0144, "step": 4972 }, { "epoch": 2.340235294117647, "grad_norm": 0.28671117114232514, "learning_rate": 0.000246857709744296, "loss": 0.1445, "step": 4973 }, { "epoch": 2.340705882352941, "grad_norm": 0.29779140770151635, "learning_rate": 0.00024682423336000746, "loss": 0.1734, "step": 4974 }, { "epoch": 2.3411764705882354, "grad_norm": 0.18168899441543399, "learning_rate": 0.00024679074870645446, "loss": 0.1137, "step": 4975 }, { "epoch": 2.341647058823529, "grad_norm": 0.29206830561062425, "learning_rate": 0.0002467572557864968, "loss": 0.1367, "step": 4976 }, { "epoch": 2.3421176470588234, "grad_norm": 0.3088223616010606, "learning_rate": 0.0002467237546029949, "loss": 0.2235, "step": 4977 }, { "epoch": 2.3425882352941176, "grad_norm": 0.46315824309475956, "learning_rate": 0.0002466902451588099, "loss": 0.3456, "step": 4978 }, { "epoch": 2.343058823529412, "grad_norm": 0.3173540793847123, "learning_rate": 0.0002466567274568037, "loss": 0.2518, "step": 4979 }, { "epoch": 2.343529411764706, "grad_norm": 0.2819659191789222, "learning_rate": 0.0002466232014998389, "loss": 0.2161, "step": 4980 }, { "epoch": 2.344, "grad_norm": 0.3383027620375803, "learning_rate": 0.00024658966729077875, "loss": 0.3291, "step": 4981 }, { "epoch": 2.344470588235294, "grad_norm": 0.3020232284550719, "learning_rate": 0.00024655612483248727, "loss": 0.2941, "step": 4982 }, { "epoch": 2.3449411764705883, "grad_norm": 0.20490196084284923, "learning_rate": 0.00024652257412782904, "loss": 0.1703, "step": 4983 }, { "epoch": 2.3454117647058825, "grad_norm": 0.32098326985971776, "learning_rate": 0.0002464890151796696, "loss": 0.4013, "step": 4984 }, { "epoch": 2.3458823529411763, "grad_norm": 0.18026646362896978, "learning_rate": 0.000246455447990875, "loss": 0.097, "step": 4985 }, { "epoch": 2.3463529411764705, "grad_norm": 0.2920970448450051, "learning_rate": 0.000246421872564312, "loss": 0.3805, "step": 4986 }, { "epoch": 2.3468235294117648, "grad_norm": 0.29918214755723627, "learning_rate": 0.0002463882889028481, "loss": 0.2608, "step": 4987 }, { "epoch": 2.347294117647059, "grad_norm": 0.3112237879684078, "learning_rate": 0.0002463546970093516, "loss": 0.3904, "step": 4988 }, { "epoch": 2.3477647058823528, "grad_norm": 0.2684014256257073, "learning_rate": 0.00024632109688669135, "loss": 0.2989, "step": 4989 }, { "epoch": 2.348235294117647, "grad_norm": 0.3690713176054641, "learning_rate": 0.000246287488537737, "loss": 0.3767, "step": 4990 }, { "epoch": 2.348705882352941, "grad_norm": 0.2218363698784232, "learning_rate": 0.00024625387196535885, "loss": 0.2171, "step": 4991 }, { "epoch": 2.3491764705882354, "grad_norm": 0.6142327772771573, "learning_rate": 0.0002462202471724278, "loss": 0.759, "step": 4992 }, { "epoch": 2.349647058823529, "grad_norm": 0.5749142362300443, "learning_rate": 0.0002461866141618158, "loss": 0.5667, "step": 4993 }, { "epoch": 2.3501176470588234, "grad_norm": 0.3799609343481172, "learning_rate": 0.0002461529729363951, "loss": 0.3536, "step": 4994 }, { "epoch": 2.3505882352941176, "grad_norm": 0.17573911349311178, "learning_rate": 0.0002461193234990389, "loss": 0.1521, "step": 4995 }, { "epoch": 2.351058823529412, "grad_norm": 0.39745860638058783, "learning_rate": 0.00024608566585262103, "loss": 0.3243, "step": 4996 }, { "epoch": 2.351529411764706, "grad_norm": 0.432340586931658, "learning_rate": 0.000246052000000016, "loss": 0.3742, "step": 4997 }, { "epoch": 2.352, "grad_norm": 0.45965487185215353, "learning_rate": 0.00024601832594409905, "loss": 0.3391, "step": 4998 }, { "epoch": 2.352470588235294, "grad_norm": 0.266985474930483, "learning_rate": 0.00024598464368774604, "loss": 0.3129, "step": 4999 }, { "epoch": 2.3529411764705883, "grad_norm": 0.2593633319327398, "learning_rate": 0.00024595095323383365, "loss": 0.2853, "step": 5000 }, { "epoch": 2.3534117647058825, "grad_norm": 0.509407565854628, "learning_rate": 0.0002459172545852393, "loss": 0.5028, "step": 5001 }, { "epoch": 2.3538823529411763, "grad_norm": 0.2832875410222409, "learning_rate": 0.0002458835477448408, "loss": 0.1927, "step": 5002 }, { "epoch": 2.3543529411764705, "grad_norm": 0.39440788041133845, "learning_rate": 0.00024584983271551717, "loss": 0.5252, "step": 5003 }, { "epoch": 2.3548235294117648, "grad_norm": 0.36892543057853777, "learning_rate": 0.0002458161095001476, "loss": 0.4091, "step": 5004 }, { "epoch": 2.355294117647059, "grad_norm": 0.40262168073616533, "learning_rate": 0.0002457823781016123, "loss": 0.2433, "step": 5005 }, { "epoch": 2.3557647058823528, "grad_norm": 0.17247304483823422, "learning_rate": 0.0002457486385227922, "loss": 0.0661, "step": 5006 }, { "epoch": 2.356235294117647, "grad_norm": 0.18098172995473294, "learning_rate": 0.00024571489076656865, "loss": 0.0686, "step": 5007 }, { "epoch": 2.356705882352941, "grad_norm": 0.14707604614576744, "learning_rate": 0.0002456811348358239, "loss": 0.0796, "step": 5008 }, { "epoch": 2.3571764705882354, "grad_norm": 0.31488828924989876, "learning_rate": 0.000245647370733441, "loss": 0.3042, "step": 5009 }, { "epoch": 2.357647058823529, "grad_norm": 0.2390648865800497, "learning_rate": 0.0002456135984623034, "loss": 0.1485, "step": 5010 }, { "epoch": 2.3581176470588234, "grad_norm": 0.27344761565600073, "learning_rate": 0.0002455798180252956, "loss": 0.1958, "step": 5011 }, { "epoch": 2.3585882352941177, "grad_norm": 0.4893326660805037, "learning_rate": 0.0002455460294253025, "loss": 0.4259, "step": 5012 }, { "epoch": 2.359058823529412, "grad_norm": 0.4730588596689744, "learning_rate": 0.00024551223266520983, "loss": 0.3631, "step": 5013 }, { "epoch": 2.359529411764706, "grad_norm": 0.15305737316392945, "learning_rate": 0.00024547842774790403, "loss": 0.0703, "step": 5014 }, { "epoch": 2.36, "grad_norm": 0.2041744473119423, "learning_rate": 0.0002454446146762721, "loss": 0.0718, "step": 5015 }, { "epoch": 2.360470588235294, "grad_norm": 0.18908662802834483, "learning_rate": 0.000245410793453202, "loss": 0.1038, "step": 5016 }, { "epoch": 2.3609411764705883, "grad_norm": 0.2446112248260109, "learning_rate": 0.00024537696408158215, "loss": 0.135, "step": 5017 }, { "epoch": 2.3614117647058825, "grad_norm": 0.11582560116979965, "learning_rate": 0.0002453431265643017, "loss": 0.0494, "step": 5018 }, { "epoch": 2.3618823529411763, "grad_norm": 0.4045901408485609, "learning_rate": 0.0002453092809042507, "loss": 0.3298, "step": 5019 }, { "epoch": 2.3623529411764705, "grad_norm": 0.38628974674177785, "learning_rate": 0.0002452754271043195, "loss": 0.2984, "step": 5020 }, { "epoch": 2.3628235294117648, "grad_norm": 0.3132934787463057, "learning_rate": 0.0002452415651673995, "loss": 0.2114, "step": 5021 }, { "epoch": 2.363294117647059, "grad_norm": 0.41502398843134414, "learning_rate": 0.00024520769509638274, "loss": 0.5238, "step": 5022 }, { "epoch": 2.3637647058823528, "grad_norm": 0.3155174544685354, "learning_rate": 0.0002451738168941618, "loss": 0.4017, "step": 5023 }, { "epoch": 2.364235294117647, "grad_norm": 0.3112153172778441, "learning_rate": 0.0002451399305636301, "loss": 0.2329, "step": 5024 }, { "epoch": 2.364705882352941, "grad_norm": 0.3391704992170152, "learning_rate": 0.00024510603610768166, "loss": 0.332, "step": 5025 }, { "epoch": 2.3651764705882354, "grad_norm": 0.41875069178760593, "learning_rate": 0.0002450721335292112, "loss": 0.4462, "step": 5026 }, { "epoch": 2.365647058823529, "grad_norm": 0.40842604553012435, "learning_rate": 0.00024503822283111423, "loss": 0.381, "step": 5027 }, { "epoch": 2.3661176470588234, "grad_norm": 0.247280048214407, "learning_rate": 0.0002450043040162869, "loss": 0.1285, "step": 5028 }, { "epoch": 2.3665882352941177, "grad_norm": 0.25407593979254645, "learning_rate": 0.000244970377087626, "loss": 0.199, "step": 5029 }, { "epoch": 2.367058823529412, "grad_norm": 0.26943324081543935, "learning_rate": 0.00024493644204802906, "loss": 0.1733, "step": 5030 }, { "epoch": 2.367529411764706, "grad_norm": 0.13400321375390137, "learning_rate": 0.0002449024989003943, "loss": 0.0601, "step": 5031 }, { "epoch": 2.368, "grad_norm": 0.34963673100428994, "learning_rate": 0.0002448685476476207, "loss": 0.2385, "step": 5032 }, { "epoch": 2.368470588235294, "grad_norm": 0.0505403710851138, "learning_rate": 0.0002448345882926077, "loss": 0.0161, "step": 5033 }, { "epoch": 2.3689411764705883, "grad_norm": 0.3711342452100809, "learning_rate": 0.0002448006208382558, "loss": 0.1782, "step": 5034 }, { "epoch": 2.369411764705882, "grad_norm": 0.35626682088832984, "learning_rate": 0.0002447666452874658, "loss": 0.2162, "step": 5035 }, { "epoch": 2.3698823529411763, "grad_norm": 0.4765477502386524, "learning_rate": 0.00024473266164313953, "loss": 0.3763, "step": 5036 }, { "epoch": 2.3703529411764706, "grad_norm": 0.3088933651272916, "learning_rate": 0.00024469866990817923, "loss": 0.2973, "step": 5037 }, { "epoch": 2.3708235294117648, "grad_norm": 2.000163622528144, "learning_rate": 0.00024466467008548814, "loss": 1.6494, "step": 5038 }, { "epoch": 2.371294117647059, "grad_norm": 0.48170762851971255, "learning_rate": 0.0002446306621779698, "loss": 0.4463, "step": 5039 }, { "epoch": 2.3717647058823528, "grad_norm": 0.3741028266653097, "learning_rate": 0.00024459664618852877, "loss": 0.3698, "step": 5040 }, { "epoch": 2.372235294117647, "grad_norm": 0.2827685775015918, "learning_rate": 0.0002445626221200702, "loss": 0.3083, "step": 5041 }, { "epoch": 2.372705882352941, "grad_norm": 0.22420996269364488, "learning_rate": 0.0002445285899754998, "loss": 0.0965, "step": 5042 }, { "epoch": 2.3731764705882354, "grad_norm": 0.38785293508943125, "learning_rate": 0.0002444945497577242, "loss": 0.3915, "step": 5043 }, { "epoch": 2.373647058823529, "grad_norm": 0.46014567715629695, "learning_rate": 0.0002444605014696506, "loss": 0.4304, "step": 5044 }, { "epoch": 2.3741176470588234, "grad_norm": 0.16582229367797552, "learning_rate": 0.0002444264451141868, "loss": 0.102, "step": 5045 }, { "epoch": 2.3745882352941177, "grad_norm": 0.3658229129512886, "learning_rate": 0.0002443923806942414, "loss": 0.2535, "step": 5046 }, { "epoch": 2.375058823529412, "grad_norm": 0.38256216676518706, "learning_rate": 0.0002443583082127237, "loss": 0.3741, "step": 5047 }, { "epoch": 2.375529411764706, "grad_norm": 1.0556398719025193, "learning_rate": 0.0002443242276725437, "loss": 0.5914, "step": 5048 }, { "epoch": 2.376, "grad_norm": 0.373049822863848, "learning_rate": 0.00024429013907661197, "loss": 0.3199, "step": 5049 }, { "epoch": 2.376470588235294, "grad_norm": 0.29630068860272496, "learning_rate": 0.0002442560424278399, "loss": 0.2881, "step": 5050 }, { "epoch": 2.3769411764705883, "grad_norm": 0.1810329270804091, "learning_rate": 0.0002442219377291395, "loss": 0.0724, "step": 5051 }, { "epoch": 2.377411764705882, "grad_norm": 0.17121396237976236, "learning_rate": 0.00024418782498342336, "loss": 0.1149, "step": 5052 }, { "epoch": 2.3778823529411763, "grad_norm": 0.18114524538842222, "learning_rate": 0.00024415370419360506, "loss": 0.0755, "step": 5053 }, { "epoch": 2.3783529411764706, "grad_norm": 0.18748125302337745, "learning_rate": 0.00024411957536259857, "loss": 0.1159, "step": 5054 }, { "epoch": 2.378823529411765, "grad_norm": 0.2736277192360383, "learning_rate": 0.0002440854384933187, "loss": 0.278, "step": 5055 }, { "epoch": 2.379294117647059, "grad_norm": 0.2436376543758721, "learning_rate": 0.0002440512935886809, "loss": 0.1953, "step": 5056 }, { "epoch": 2.379764705882353, "grad_norm": 0.12994585020301141, "learning_rate": 0.0002440171406516013, "loss": 0.0606, "step": 5057 }, { "epoch": 2.380235294117647, "grad_norm": 0.37371005196387636, "learning_rate": 0.00024398297968499677, "loss": 0.323, "step": 5058 }, { "epoch": 2.3807058823529412, "grad_norm": 0.30246477647056536, "learning_rate": 0.00024394881069178477, "loss": 0.3602, "step": 5059 }, { "epoch": 2.3811764705882354, "grad_norm": 0.15277844845747657, "learning_rate": 0.00024391463367488348, "loss": 0.0606, "step": 5060 }, { "epoch": 2.3816470588235292, "grad_norm": 0.4564816186140499, "learning_rate": 0.00024388044863721188, "loss": 0.5682, "step": 5061 }, { "epoch": 2.3821176470588235, "grad_norm": 0.3117795071524464, "learning_rate": 0.00024384625558168948, "loss": 0.2962, "step": 5062 }, { "epoch": 2.3825882352941177, "grad_norm": 0.24376504995228784, "learning_rate": 0.0002438120545112365, "loss": 0.1306, "step": 5063 }, { "epoch": 2.383058823529412, "grad_norm": 0.3079870169495136, "learning_rate": 0.000243777845428774, "loss": 0.2675, "step": 5064 }, { "epoch": 2.383529411764706, "grad_norm": 0.4044908345521096, "learning_rate": 0.00024374362833722344, "loss": 0.3338, "step": 5065 }, { "epoch": 2.384, "grad_norm": 0.3889557374599168, "learning_rate": 0.00024370940323950724, "loss": 0.493, "step": 5066 }, { "epoch": 2.384470588235294, "grad_norm": 0.5474850343932894, "learning_rate": 0.0002436751701385484, "loss": 0.5, "step": 5067 }, { "epoch": 2.3849411764705883, "grad_norm": 0.4111621933957658, "learning_rate": 0.0002436409290372705, "loss": 0.353, "step": 5068 }, { "epoch": 2.385411764705882, "grad_norm": 0.3851425222332026, "learning_rate": 0.000243606679938598, "loss": 0.4297, "step": 5069 }, { "epoch": 2.3858823529411763, "grad_norm": 0.2622702899671757, "learning_rate": 0.00024357242284545587, "loss": 0.2226, "step": 5070 }, { "epoch": 2.3863529411764706, "grad_norm": 0.2003250573295963, "learning_rate": 0.00024353815776076985, "loss": 0.1393, "step": 5071 }, { "epoch": 2.386823529411765, "grad_norm": 0.36918970663287537, "learning_rate": 0.00024350388468746635, "loss": 0.434, "step": 5072 }, { "epoch": 2.387294117647059, "grad_norm": 0.5181674783116115, "learning_rate": 0.00024346960362847247, "loss": 0.4266, "step": 5073 }, { "epoch": 2.387764705882353, "grad_norm": 0.379902060437149, "learning_rate": 0.000243435314586716, "loss": 0.3642, "step": 5074 }, { "epoch": 2.388235294117647, "grad_norm": 0.38512692877362614, "learning_rate": 0.00024340101756512532, "loss": 0.3037, "step": 5075 }, { "epoch": 2.3887058823529412, "grad_norm": 0.33259203758192285, "learning_rate": 0.00024336671256662963, "loss": 0.4531, "step": 5076 }, { "epoch": 2.3891764705882355, "grad_norm": 0.46946836689374405, "learning_rate": 0.00024333239959415873, "loss": 0.4077, "step": 5077 }, { "epoch": 2.3896470588235292, "grad_norm": 0.1618205669548557, "learning_rate": 0.0002432980786506431, "loss": 0.0891, "step": 5078 }, { "epoch": 2.3901176470588235, "grad_norm": 0.39048787441083804, "learning_rate": 0.00024326374973901395, "loss": 0.3548, "step": 5079 }, { "epoch": 2.3905882352941177, "grad_norm": 0.22039894151866068, "learning_rate": 0.0002432294128622031, "loss": 0.2128, "step": 5080 }, { "epoch": 2.391058823529412, "grad_norm": 0.1841631517283094, "learning_rate": 0.0002431950680231431, "loss": 0.0788, "step": 5081 }, { "epoch": 2.391529411764706, "grad_norm": 0.1780655443698855, "learning_rate": 0.00024316071522476718, "loss": 0.1065, "step": 5082 }, { "epoch": 2.392, "grad_norm": 0.48320745671788096, "learning_rate": 0.0002431263544700093, "loss": 0.5297, "step": 5083 }, { "epoch": 2.392470588235294, "grad_norm": 0.4011126001332104, "learning_rate": 0.00024309198576180388, "loss": 0.2208, "step": 5084 }, { "epoch": 2.3929411764705883, "grad_norm": 0.48086631115039247, "learning_rate": 0.00024305760910308627, "loss": 0.3915, "step": 5085 }, { "epoch": 2.393411764705882, "grad_norm": 0.5603932236460567, "learning_rate": 0.00024302322449679245, "loss": 0.6368, "step": 5086 }, { "epoch": 2.3938823529411764, "grad_norm": 0.20973030270403717, "learning_rate": 0.0002429888319458589, "loss": 0.1565, "step": 5087 }, { "epoch": 2.3943529411764706, "grad_norm": 0.37838899383827523, "learning_rate": 0.00024295443145322303, "loss": 0.3015, "step": 5088 }, { "epoch": 2.394823529411765, "grad_norm": 0.35331868746743, "learning_rate": 0.0002429200230218228, "loss": 0.4689, "step": 5089 }, { "epoch": 2.395294117647059, "grad_norm": 0.28048390975069193, "learning_rate": 0.00024288560665459679, "loss": 0.2483, "step": 5090 }, { "epoch": 2.395764705882353, "grad_norm": 0.3726568467016559, "learning_rate": 0.0002428511823544844, "loss": 0.3541, "step": 5091 }, { "epoch": 2.396235294117647, "grad_norm": 0.3634002877240533, "learning_rate": 0.00024281675012442563, "loss": 0.2083, "step": 5092 }, { "epoch": 2.3967058823529412, "grad_norm": 0.3339450384190516, "learning_rate": 0.00024278230996736109, "loss": 0.2457, "step": 5093 }, { "epoch": 2.3971764705882355, "grad_norm": 0.4070770420535774, "learning_rate": 0.0002427478618862322, "loss": 0.3595, "step": 5094 }, { "epoch": 2.3976470588235292, "grad_norm": 0.20526666072229394, "learning_rate": 0.00024271340588398097, "loss": 0.1759, "step": 5095 }, { "epoch": 2.3981176470588235, "grad_norm": 0.24080425651012716, "learning_rate": 0.00024267894196355015, "loss": 0.1766, "step": 5096 }, { "epoch": 2.3985882352941177, "grad_norm": 0.17662744957328586, "learning_rate": 0.00024264447012788308, "loss": 0.0853, "step": 5097 }, { "epoch": 2.399058823529412, "grad_norm": 0.36116990907180463, "learning_rate": 0.00024260999037992384, "loss": 0.4266, "step": 5098 }, { "epoch": 2.3995294117647057, "grad_norm": 0.4861170902153208, "learning_rate": 0.00024257550272261714, "loss": 0.2696, "step": 5099 }, { "epoch": 2.4, "grad_norm": 0.4019969505224742, "learning_rate": 0.00024254100715890846, "loss": 0.2901, "step": 5100 }, { "epoch": 2.400470588235294, "grad_norm": 0.34675038076680315, "learning_rate": 0.00024250650369174386, "loss": 0.3628, "step": 5101 }, { "epoch": 2.4009411764705884, "grad_norm": 0.23427458952654115, "learning_rate": 0.0002424719923240701, "loss": 0.1802, "step": 5102 }, { "epoch": 2.401411764705882, "grad_norm": 0.49992261880698735, "learning_rate": 0.00024243747305883461, "loss": 0.378, "step": 5103 }, { "epoch": 2.4018823529411764, "grad_norm": 0.3011415017721021, "learning_rate": 0.00024240294589898554, "loss": 0.2528, "step": 5104 }, { "epoch": 2.4023529411764706, "grad_norm": 0.49675675417517723, "learning_rate": 0.00024236841084747165, "loss": 0.438, "step": 5105 }, { "epoch": 2.402823529411765, "grad_norm": 0.3046738615789889, "learning_rate": 0.00024233386790724242, "loss": 0.2294, "step": 5106 }, { "epoch": 2.403294117647059, "grad_norm": 0.32066376843487837, "learning_rate": 0.000242299317081248, "loss": 0.3041, "step": 5107 }, { "epoch": 2.403764705882353, "grad_norm": 0.350887322383729, "learning_rate": 0.00024226475837243908, "loss": 0.3419, "step": 5108 }, { "epoch": 2.404235294117647, "grad_norm": 0.20049160545001796, "learning_rate": 0.00024223019178376732, "loss": 0.0991, "step": 5109 }, { "epoch": 2.4047058823529412, "grad_norm": 0.19197932783951738, "learning_rate": 0.00024219561731818475, "loss": 0.1126, "step": 5110 }, { "epoch": 2.4051764705882355, "grad_norm": 0.5197378372389624, "learning_rate": 0.00024216103497864426, "loss": 0.332, "step": 5111 }, { "epoch": 2.4056470588235292, "grad_norm": 0.16518381151259176, "learning_rate": 0.00024212644476809934, "loss": 0.1393, "step": 5112 }, { "epoch": 2.4061176470588235, "grad_norm": 0.31299928507821756, "learning_rate": 0.0002420918466895042, "loss": 0.2553, "step": 5113 }, { "epoch": 2.4065882352941177, "grad_norm": 0.4105704170159088, "learning_rate": 0.0002420572407458136, "loss": 0.2697, "step": 5114 }, { "epoch": 2.407058823529412, "grad_norm": 0.36274228492519495, "learning_rate": 0.0002420226269399831, "loss": 0.3033, "step": 5115 }, { "epoch": 2.4075294117647057, "grad_norm": 0.22912567254882454, "learning_rate": 0.00024198800527496889, "loss": 0.1895, "step": 5116 }, { "epoch": 2.408, "grad_norm": 0.2470920243509173, "learning_rate": 0.00024195337575372788, "loss": 0.2215, "step": 5117 }, { "epoch": 2.408470588235294, "grad_norm": 0.3197848538304853, "learning_rate": 0.00024191873837921748, "loss": 0.2835, "step": 5118 }, { "epoch": 2.4089411764705884, "grad_norm": 0.33801512197171873, "learning_rate": 0.000241884093154396, "loss": 0.2592, "step": 5119 }, { "epoch": 2.409411764705882, "grad_norm": 0.24161851458065883, "learning_rate": 0.00024184944008222227, "loss": 0.1733, "step": 5120 }, { "epoch": 2.4098823529411764, "grad_norm": 0.2667921001583677, "learning_rate": 0.00024181477916565582, "loss": 0.2554, "step": 5121 }, { "epoch": 2.4103529411764706, "grad_norm": 0.16435971343561995, "learning_rate": 0.0002417801104076569, "loss": 0.0737, "step": 5122 }, { "epoch": 2.410823529411765, "grad_norm": 0.3817945383194009, "learning_rate": 0.00024174543381118638, "loss": 0.3266, "step": 5123 }, { "epoch": 2.411294117647059, "grad_norm": 0.2925112512820596, "learning_rate": 0.00024171074937920579, "loss": 0.1986, "step": 5124 }, { "epoch": 2.411764705882353, "grad_norm": 0.2528556355572129, "learning_rate": 0.00024167605711467738, "loss": 0.27, "step": 5125 }, { "epoch": 2.412235294117647, "grad_norm": 0.3086429586910072, "learning_rate": 0.000241641357020564, "loss": 0.1699, "step": 5126 }, { "epoch": 2.4127058823529413, "grad_norm": 0.3688082905658057, "learning_rate": 0.00024160664909982928, "loss": 0.2457, "step": 5127 }, { "epoch": 2.4131764705882355, "grad_norm": 0.5334631484444252, "learning_rate": 0.00024157193335543734, "loss": 0.494, "step": 5128 }, { "epoch": 2.4136470588235293, "grad_norm": 0.4793353360422452, "learning_rate": 0.00024153720979035317, "loss": 0.4126, "step": 5129 }, { "epoch": 2.4141176470588235, "grad_norm": 0.23442763808995337, "learning_rate": 0.00024150247840754228, "loss": 0.1124, "step": 5130 }, { "epoch": 2.4145882352941177, "grad_norm": 0.2018397405997861, "learning_rate": 0.00024146773920997091, "loss": 0.1061, "step": 5131 }, { "epoch": 2.415058823529412, "grad_norm": 0.16133834952175732, "learning_rate": 0.000241432992200606, "loss": 0.0822, "step": 5132 }, { "epoch": 2.4155294117647057, "grad_norm": 0.3256115516938197, "learning_rate": 0.000241398237382415, "loss": 0.2279, "step": 5133 }, { "epoch": 2.416, "grad_norm": 0.31862207332661485, "learning_rate": 0.00024136347475836633, "loss": 0.1719, "step": 5134 }, { "epoch": 2.416470588235294, "grad_norm": 0.16356776280201643, "learning_rate": 0.0002413287043314287, "loss": 0.0826, "step": 5135 }, { "epoch": 2.4169411764705884, "grad_norm": 0.3247386133654554, "learning_rate": 0.00024129392610457176, "loss": 0.1731, "step": 5136 }, { "epoch": 2.417411764705882, "grad_norm": 0.3010496805109437, "learning_rate": 0.0002412591400807658, "loss": 0.2614, "step": 5137 }, { "epoch": 2.4178823529411764, "grad_norm": 0.2593205075336458, "learning_rate": 0.0002412243462629816, "loss": 0.1718, "step": 5138 }, { "epoch": 2.4183529411764706, "grad_norm": 0.25236253191069796, "learning_rate": 0.0002411895446541908, "loss": 0.1811, "step": 5139 }, { "epoch": 2.418823529411765, "grad_norm": 0.3356152075889552, "learning_rate": 0.00024115473525736561, "loss": 0.3493, "step": 5140 }, { "epoch": 2.419294117647059, "grad_norm": 0.5429141413301006, "learning_rate": 0.0002411199180754789, "loss": 0.5552, "step": 5141 }, { "epoch": 2.419764705882353, "grad_norm": 0.3296627677559648, "learning_rate": 0.00024108509311150422, "loss": 0.2824, "step": 5142 }, { "epoch": 2.420235294117647, "grad_norm": 0.4728382303089758, "learning_rate": 0.00024105026036841582, "loss": 0.5818, "step": 5143 }, { "epoch": 2.4207058823529413, "grad_norm": 0.836837830730149, "learning_rate": 0.00024101541984918864, "loss": 0.6175, "step": 5144 }, { "epoch": 2.4211764705882355, "grad_norm": 0.272500568346757, "learning_rate": 0.00024098057155679813, "loss": 0.2423, "step": 5145 }, { "epoch": 2.4216470588235293, "grad_norm": 0.19587749328208817, "learning_rate": 0.0002409457154942206, "loss": 0.2041, "step": 5146 }, { "epoch": 2.4221176470588235, "grad_norm": 0.19615673069601583, "learning_rate": 0.00024091085166443285, "loss": 0.133, "step": 5147 }, { "epoch": 2.4225882352941177, "grad_norm": 0.18579930084319918, "learning_rate": 0.00024087598007041243, "loss": 0.1353, "step": 5148 }, { "epoch": 2.423058823529412, "grad_norm": 0.3573281774350118, "learning_rate": 0.0002408411007151376, "loss": 0.2686, "step": 5149 }, { "epoch": 2.4235294117647057, "grad_norm": 0.37784298173204456, "learning_rate": 0.00024080621360158717, "loss": 0.2597, "step": 5150 }, { "epoch": 2.424, "grad_norm": 0.20926417268253114, "learning_rate": 0.00024077131873274072, "loss": 0.111, "step": 5151 }, { "epoch": 2.424470588235294, "grad_norm": 0.24357842238317126, "learning_rate": 0.0002407364161115784, "loss": 0.2281, "step": 5152 }, { "epoch": 2.4249411764705884, "grad_norm": 0.2858658151616761, "learning_rate": 0.00024070150574108113, "loss": 0.2718, "step": 5153 }, { "epoch": 2.425411764705882, "grad_norm": 0.4268240569501004, "learning_rate": 0.00024066658762423034, "loss": 0.3731, "step": 5154 }, { "epoch": 2.4258823529411764, "grad_norm": 0.3034148280619247, "learning_rate": 0.00024063166176400825, "loss": 0.1821, "step": 5155 }, { "epoch": 2.4263529411764706, "grad_norm": 0.23687946257328005, "learning_rate": 0.0002405967281633977, "loss": 0.1044, "step": 5156 }, { "epoch": 2.426823529411765, "grad_norm": 0.36003990497467536, "learning_rate": 0.00024056178682538222, "loss": 0.3305, "step": 5157 }, { "epoch": 2.427294117647059, "grad_norm": 0.4665747951193076, "learning_rate": 0.00024052683775294592, "loss": 0.4355, "step": 5158 }, { "epoch": 2.427764705882353, "grad_norm": 0.320727862807659, "learning_rate": 0.00024049188094907363, "loss": 0.2724, "step": 5159 }, { "epoch": 2.428235294117647, "grad_norm": 0.15249879203691957, "learning_rate": 0.00024045691641675086, "loss": 0.1147, "step": 5160 }, { "epoch": 2.4287058823529413, "grad_norm": 0.35169779374438814, "learning_rate": 0.00024042194415896376, "loss": 0.3982, "step": 5161 }, { "epoch": 2.429176470588235, "grad_norm": 0.5013077242148227, "learning_rate": 0.00024038696417869908, "loss": 0.4538, "step": 5162 }, { "epoch": 2.4296470588235293, "grad_norm": 0.3412339720114793, "learning_rate": 0.00024035197647894434, "loss": 0.3984, "step": 5163 }, { "epoch": 2.4301176470588235, "grad_norm": 0.19873554638074928, "learning_rate": 0.00024031698106268763, "loss": 0.0999, "step": 5164 }, { "epoch": 2.4305882352941177, "grad_norm": 0.3405585671114223, "learning_rate": 0.0002402819779329177, "loss": 0.3409, "step": 5165 }, { "epoch": 2.431058823529412, "grad_norm": 0.2381275509112344, "learning_rate": 0.00024024696709262406, "loss": 0.1704, "step": 5166 }, { "epoch": 2.4315294117647057, "grad_norm": 0.2572714456791618, "learning_rate": 0.00024021194854479674, "loss": 0.2238, "step": 5167 }, { "epoch": 2.432, "grad_norm": 0.49606954154360394, "learning_rate": 0.00024017692229242655, "loss": 0.5046, "step": 5168 }, { "epoch": 2.432470588235294, "grad_norm": 0.5038628923054148, "learning_rate": 0.00024014188833850486, "loss": 0.5617, "step": 5169 }, { "epoch": 2.4329411764705884, "grad_norm": 0.267801480585755, "learning_rate": 0.0002401068466860238, "loss": 0.2595, "step": 5170 }, { "epoch": 2.433411764705882, "grad_norm": 0.2801068310263726, "learning_rate": 0.00024007179733797604, "loss": 0.266, "step": 5171 }, { "epoch": 2.4338823529411764, "grad_norm": 0.3250691753185171, "learning_rate": 0.00024003674029735497, "loss": 0.2993, "step": 5172 }, { "epoch": 2.4343529411764706, "grad_norm": 0.44081956015495527, "learning_rate": 0.00024000167556715468, "loss": 0.3809, "step": 5173 }, { "epoch": 2.434823529411765, "grad_norm": 0.2939016663876918, "learning_rate": 0.00023996660315036986, "loss": 0.226, "step": 5174 }, { "epoch": 2.435294117647059, "grad_norm": 0.20044571355165136, "learning_rate": 0.00023993152304999582, "loss": 0.1298, "step": 5175 }, { "epoch": 2.435764705882353, "grad_norm": 0.39247914181288507, "learning_rate": 0.0002398964352690286, "loss": 0.3564, "step": 5176 }, { "epoch": 2.436235294117647, "grad_norm": 0.5575080199937724, "learning_rate": 0.00023986133981046485, "loss": 0.519, "step": 5177 }, { "epoch": 2.4367058823529413, "grad_norm": 0.3247377887104566, "learning_rate": 0.00023982623667730195, "loss": 0.3149, "step": 5178 }, { "epoch": 2.437176470588235, "grad_norm": 0.1628553693983453, "learning_rate": 0.0002397911258725378, "loss": 0.0924, "step": 5179 }, { "epoch": 2.4376470588235293, "grad_norm": 0.2736098858991241, "learning_rate": 0.00023975600739917113, "loss": 0.3347, "step": 5180 }, { "epoch": 2.4381176470588235, "grad_norm": 0.3172849417061642, "learning_rate": 0.00023972088126020114, "loss": 0.4153, "step": 5181 }, { "epoch": 2.4385882352941177, "grad_norm": 0.3212895229045304, "learning_rate": 0.00023968574745862782, "loss": 0.2631, "step": 5182 }, { "epoch": 2.439058823529412, "grad_norm": 0.3866931951890522, "learning_rate": 0.0002396506059974518, "loss": 0.4659, "step": 5183 }, { "epoch": 2.4395294117647057, "grad_norm": 0.06845891342275232, "learning_rate": 0.00023961545687967422, "loss": 0.0227, "step": 5184 }, { "epoch": 2.44, "grad_norm": 0.22086700130629494, "learning_rate": 0.0002395803001082971, "loss": 0.1694, "step": 5185 }, { "epoch": 2.440470588235294, "grad_norm": 0.33221659283160443, "learning_rate": 0.00023954513568632296, "loss": 0.1864, "step": 5186 }, { "epoch": 2.4409411764705884, "grad_norm": 0.3336501218722701, "learning_rate": 0.000239509963616755, "loss": 0.2658, "step": 5187 }, { "epoch": 2.441411764705882, "grad_norm": 0.28459400253980033, "learning_rate": 0.00023947478390259712, "loss": 0.2587, "step": 5188 }, { "epoch": 2.4418823529411764, "grad_norm": 0.24280384610625705, "learning_rate": 0.00023943959654685383, "loss": 0.2192, "step": 5189 }, { "epoch": 2.4423529411764706, "grad_norm": 0.06759576576536319, "learning_rate": 0.00023940440155253024, "loss": 0.0196, "step": 5190 }, { "epoch": 2.442823529411765, "grad_norm": 0.13780943234071594, "learning_rate": 0.0002393691989226323, "loss": 0.0761, "step": 5191 }, { "epoch": 2.443294117647059, "grad_norm": 0.4530736982189825, "learning_rate": 0.00023933398866016642, "loss": 0.626, "step": 5192 }, { "epoch": 2.443764705882353, "grad_norm": 0.5787749536003604, "learning_rate": 0.0002392987707681397, "loss": 0.5864, "step": 5193 }, { "epoch": 2.444235294117647, "grad_norm": 0.34544967168733043, "learning_rate": 0.00023926354524955994, "loss": 0.3123, "step": 5194 }, { "epoch": 2.4447058823529413, "grad_norm": 0.3723547485851614, "learning_rate": 0.0002392283121074356, "loss": 0.3777, "step": 5195 }, { "epoch": 2.445176470588235, "grad_norm": 0.22159330079569817, "learning_rate": 0.00023919307134477572, "loss": 0.1507, "step": 5196 }, { "epoch": 2.4456470588235293, "grad_norm": 0.21395032002543682, "learning_rate": 0.00023915782296459007, "loss": 0.0719, "step": 5197 }, { "epoch": 2.4461176470588235, "grad_norm": 0.31845581857006267, "learning_rate": 0.00023912256696988904, "loss": 0.3241, "step": 5198 }, { "epoch": 2.4465882352941177, "grad_norm": 0.30437606418998714, "learning_rate": 0.00023908730336368364, "loss": 0.3477, "step": 5199 }, { "epoch": 2.447058823529412, "grad_norm": 0.3322040751285332, "learning_rate": 0.00023905203214898558, "loss": 0.2608, "step": 5200 }, { "epoch": 2.4475294117647057, "grad_norm": 0.2035251418872783, "learning_rate": 0.00023901675332880715, "loss": 0.189, "step": 5201 }, { "epoch": 2.448, "grad_norm": 0.23487847913080226, "learning_rate": 0.00023898146690616134, "loss": 0.2058, "step": 5202 }, { "epoch": 2.448470588235294, "grad_norm": 0.3335078733323479, "learning_rate": 0.0002389461728840618, "loss": 0.3664, "step": 5203 }, { "epoch": 2.4489411764705884, "grad_norm": 0.34042347294753356, "learning_rate": 0.00023891087126552292, "loss": 0.2249, "step": 5204 }, { "epoch": 2.449411764705882, "grad_norm": 0.15304626953323644, "learning_rate": 0.00023887556205355945, "loss": 0.0754, "step": 5205 }, { "epoch": 2.4498823529411764, "grad_norm": 0.24982454653224057, "learning_rate": 0.00023884024525118702, "loss": 0.3054, "step": 5206 }, { "epoch": 2.4503529411764706, "grad_norm": 0.38848515121804034, "learning_rate": 0.00023880492086142193, "loss": 0.3666, "step": 5207 }, { "epoch": 2.450823529411765, "grad_norm": 0.2861856775320704, "learning_rate": 0.00023876958888728098, "loss": 0.2116, "step": 5208 }, { "epoch": 2.451294117647059, "grad_norm": 0.24629146393737467, "learning_rate": 0.00023873424933178178, "loss": 0.1756, "step": 5209 }, { "epoch": 2.451764705882353, "grad_norm": 0.1938727847147083, "learning_rate": 0.00023869890219794238, "loss": 0.146, "step": 5210 }, { "epoch": 2.452235294117647, "grad_norm": 0.21017015919926121, "learning_rate": 0.0002386635474887817, "loss": 0.144, "step": 5211 }, { "epoch": 2.4527058823529413, "grad_norm": 0.48891265285266367, "learning_rate": 0.0002386281852073192, "loss": 0.5856, "step": 5212 }, { "epoch": 2.453176470588235, "grad_norm": 0.1477914576644396, "learning_rate": 0.00023859281535657496, "loss": 0.0385, "step": 5213 }, { "epoch": 2.4536470588235293, "grad_norm": 0.08908328152321918, "learning_rate": 0.00023855743793956967, "loss": 0.0608, "step": 5214 }, { "epoch": 2.4541176470588235, "grad_norm": 0.32937856054401227, "learning_rate": 0.0002385220529593249, "loss": 0.3648, "step": 5215 }, { "epoch": 2.4545882352941177, "grad_norm": 0.34830216348411236, "learning_rate": 0.00023848666041886258, "loss": 0.2723, "step": 5216 }, { "epoch": 2.455058823529412, "grad_norm": 0.3622195680453333, "learning_rate": 0.00023845126032120539, "loss": 0.2028, "step": 5217 }, { "epoch": 2.4555294117647057, "grad_norm": 0.14210597767594993, "learning_rate": 0.00023841585266937677, "loss": 0.0792, "step": 5218 }, { "epoch": 2.456, "grad_norm": 0.5216238456187927, "learning_rate": 0.00023838043746640065, "loss": 0.4498, "step": 5219 }, { "epoch": 2.456470588235294, "grad_norm": 0.5589491544298972, "learning_rate": 0.00023834501471530165, "loss": 0.3479, "step": 5220 }, { "epoch": 2.4569411764705884, "grad_norm": 0.625827278406552, "learning_rate": 0.0002383095844191051, "loss": 0.482, "step": 5221 }, { "epoch": 2.457411764705882, "grad_norm": 0.2701870319602725, "learning_rate": 0.0002382741465808368, "loss": 0.2525, "step": 5222 }, { "epoch": 2.4578823529411764, "grad_norm": 0.23134721568025565, "learning_rate": 0.0002382387012035235, "loss": 0.2287, "step": 5223 }, { "epoch": 2.4583529411764706, "grad_norm": 0.4682348274160137, "learning_rate": 0.0002382032482901922, "loss": 0.3015, "step": 5224 }, { "epoch": 2.458823529411765, "grad_norm": 0.13186206971309408, "learning_rate": 0.00023816778784387094, "loss": 0.0492, "step": 5225 }, { "epoch": 2.4592941176470586, "grad_norm": 0.36243150352171133, "learning_rate": 0.00023813231986758815, "loss": 0.3441, "step": 5226 }, { "epoch": 2.459764705882353, "grad_norm": 0.1912898442821494, "learning_rate": 0.00023809684436437294, "loss": 0.0885, "step": 5227 }, { "epoch": 2.460235294117647, "grad_norm": 0.6700232762832956, "learning_rate": 0.00023806136133725506, "loss": 0.4693, "step": 5228 }, { "epoch": 2.4607058823529413, "grad_norm": 0.31187889123813983, "learning_rate": 0.00023802587078926505, "loss": 0.2431, "step": 5229 }, { "epoch": 2.461176470588235, "grad_norm": 0.4932828072191362, "learning_rate": 0.00023799037272343387, "loss": 0.4823, "step": 5230 }, { "epoch": 2.4616470588235293, "grad_norm": 0.27177823940201806, "learning_rate": 0.00023795486714279327, "loss": 0.1475, "step": 5231 }, { "epoch": 2.4621176470588235, "grad_norm": 0.27002457805201, "learning_rate": 0.0002379193540503756, "loss": 0.1293, "step": 5232 }, { "epoch": 2.4625882352941177, "grad_norm": 0.4650604241630205, "learning_rate": 0.00023788383344921384, "loss": 0.4523, "step": 5233 }, { "epoch": 2.463058823529412, "grad_norm": 0.5796805730104138, "learning_rate": 0.0002378483053423417, "loss": 0.3594, "step": 5234 }, { "epoch": 2.4635294117647057, "grad_norm": 0.3834662137210741, "learning_rate": 0.00023781276973279335, "loss": 0.4382, "step": 5235 }, { "epoch": 2.464, "grad_norm": 0.36475085257123246, "learning_rate": 0.0002377772266236037, "loss": 0.2801, "step": 5236 }, { "epoch": 2.464470588235294, "grad_norm": 0.12551845843083925, "learning_rate": 0.0002377416760178084, "loss": 0.0638, "step": 5237 }, { "epoch": 2.4649411764705884, "grad_norm": 0.30611561699955764, "learning_rate": 0.00023770611791844364, "loss": 0.2601, "step": 5238 }, { "epoch": 2.465411764705882, "grad_norm": 0.3455027287770886, "learning_rate": 0.00023767055232854615, "loss": 0.2127, "step": 5239 }, { "epoch": 2.4658823529411764, "grad_norm": 0.5917096586493081, "learning_rate": 0.00023763497925115347, "loss": 0.6499, "step": 5240 }, { "epoch": 2.4663529411764706, "grad_norm": 0.40827578519164354, "learning_rate": 0.00023759939868930373, "loss": 0.3065, "step": 5241 }, { "epoch": 2.466823529411765, "grad_norm": 0.5285304150377793, "learning_rate": 0.00023756381064603564, "loss": 0.6118, "step": 5242 }, { "epoch": 2.4672941176470586, "grad_norm": 0.3035072615343353, "learning_rate": 0.0002375282151243887, "loss": 0.3009, "step": 5243 }, { "epoch": 2.467764705882353, "grad_norm": 0.35667077830938404, "learning_rate": 0.00023749261212740278, "loss": 0.3597, "step": 5244 }, { "epoch": 2.468235294117647, "grad_norm": 0.340890125668263, "learning_rate": 0.00023745700165811868, "loss": 0.1313, "step": 5245 }, { "epoch": 2.4687058823529413, "grad_norm": 0.5461435254832957, "learning_rate": 0.00023742138371957765, "loss": 0.2629, "step": 5246 }, { "epoch": 2.469176470588235, "grad_norm": 0.4992408915094788, "learning_rate": 0.00023738575831482164, "loss": 0.5695, "step": 5247 }, { "epoch": 2.4696470588235293, "grad_norm": 0.4065510213335687, "learning_rate": 0.0002373501254468933, "loss": 0.3779, "step": 5248 }, { "epoch": 2.4701176470588235, "grad_norm": 0.19357455987968103, "learning_rate": 0.00023731448511883575, "loss": 0.1384, "step": 5249 }, { "epoch": 2.4705882352941178, "grad_norm": 0.20164918392845496, "learning_rate": 0.00023727883733369292, "loss": 0.1018, "step": 5250 }, { "epoch": 2.471058823529412, "grad_norm": 0.2685984930554374, "learning_rate": 0.00023724318209450925, "loss": 0.2448, "step": 5251 }, { "epoch": 2.4715294117647058, "grad_norm": 0.22278068947954555, "learning_rate": 0.00023720751940432992, "loss": 0.1165, "step": 5252 }, { "epoch": 2.472, "grad_norm": 0.40482395894400414, "learning_rate": 0.0002371718492662007, "loss": 0.2942, "step": 5253 }, { "epoch": 2.472470588235294, "grad_norm": 0.3082863897438412, "learning_rate": 0.00023713617168316797, "loss": 0.2087, "step": 5254 }, { "epoch": 2.4729411764705884, "grad_norm": 0.7960886365674162, "learning_rate": 0.00023710048665827872, "loss": 0.5355, "step": 5255 }, { "epoch": 2.473411764705882, "grad_norm": 0.20603461979397306, "learning_rate": 0.00023706479419458075, "loss": 0.1979, "step": 5256 }, { "epoch": 2.4738823529411764, "grad_norm": 0.15091474620652484, "learning_rate": 0.00023702909429512228, "loss": 0.0851, "step": 5257 }, { "epoch": 2.4743529411764706, "grad_norm": 0.4990930699322159, "learning_rate": 0.00023699338696295228, "loss": 0.5069, "step": 5258 }, { "epoch": 2.474823529411765, "grad_norm": 0.2618201226790193, "learning_rate": 0.00023695767220112033, "loss": 0.1456, "step": 5259 }, { "epoch": 2.4752941176470586, "grad_norm": 0.2779123956399183, "learning_rate": 0.0002369219500126767, "loss": 0.2001, "step": 5260 }, { "epoch": 2.475764705882353, "grad_norm": 0.37200969079644414, "learning_rate": 0.00023688622040067217, "loss": 0.3252, "step": 5261 }, { "epoch": 2.476235294117647, "grad_norm": 0.4652643739491704, "learning_rate": 0.0002368504833681582, "loss": 0.2438, "step": 5262 }, { "epoch": 2.4767058823529413, "grad_norm": 0.48747839284926414, "learning_rate": 0.00023681473891818699, "loss": 0.3569, "step": 5263 }, { "epoch": 2.477176470588235, "grad_norm": 0.44850983664607796, "learning_rate": 0.00023677898705381123, "loss": 0.4468, "step": 5264 }, { "epoch": 2.4776470588235293, "grad_norm": 0.25834671550858807, "learning_rate": 0.00023674322777808432, "loss": 0.1759, "step": 5265 }, { "epoch": 2.4781176470588235, "grad_norm": 0.5163179946156117, "learning_rate": 0.0002367074610940603, "loss": 0.3422, "step": 5266 }, { "epoch": 2.4785882352941178, "grad_norm": 0.28722397312983244, "learning_rate": 0.00023667168700479383, "loss": 0.2232, "step": 5267 }, { "epoch": 2.479058823529412, "grad_norm": 0.37945623264793754, "learning_rate": 0.0002366359055133401, "loss": 0.3378, "step": 5268 }, { "epoch": 2.4795294117647058, "grad_norm": 0.19407893482918823, "learning_rate": 0.00023660011662275516, "loss": 0.1191, "step": 5269 }, { "epoch": 2.48, "grad_norm": 0.14654675660953687, "learning_rate": 0.00023656432033609544, "loss": 0.0211, "step": 5270 }, { "epoch": 2.480470588235294, "grad_norm": 0.434276989078998, "learning_rate": 0.0002365285166564182, "loss": 0.3073, "step": 5271 }, { "epoch": 2.4809411764705884, "grad_norm": 0.5078118908289698, "learning_rate": 0.00023649270558678123, "loss": 0.5286, "step": 5272 }, { "epoch": 2.481411764705882, "grad_norm": 0.20906528270048688, "learning_rate": 0.00023645688713024295, "loss": 0.1832, "step": 5273 }, { "epoch": 2.4818823529411764, "grad_norm": 0.5755326128342391, "learning_rate": 0.0002364210612898624, "loss": 0.6739, "step": 5274 }, { "epoch": 2.4823529411764707, "grad_norm": 0.29446337893006813, "learning_rate": 0.00023638522806869935, "loss": 0.2847, "step": 5275 }, { "epoch": 2.482823529411765, "grad_norm": 0.5379833329546677, "learning_rate": 0.0002363493874698141, "loss": 0.4731, "step": 5276 }, { "epoch": 2.4832941176470587, "grad_norm": 0.5561232942824909, "learning_rate": 0.0002363135394962676, "loss": 0.5516, "step": 5277 }, { "epoch": 2.483764705882353, "grad_norm": 0.3616181235581402, "learning_rate": 0.00023627768415112154, "loss": 0.405, "step": 5278 }, { "epoch": 2.484235294117647, "grad_norm": 0.2802426676660523, "learning_rate": 0.00023624182143743797, "loss": 0.2677, "step": 5279 }, { "epoch": 2.4847058823529413, "grad_norm": 0.3826727733534067, "learning_rate": 0.00023620595135827985, "loss": 0.311, "step": 5280 }, { "epoch": 2.485176470588235, "grad_norm": 0.5522892127203164, "learning_rate": 0.0002361700739167107, "loss": 0.4692, "step": 5281 }, { "epoch": 2.4856470588235293, "grad_norm": 0.40629416525121165, "learning_rate": 0.00023613418911579454, "loss": 0.3078, "step": 5282 }, { "epoch": 2.4861176470588235, "grad_norm": 0.22657481156103482, "learning_rate": 0.00023609829695859618, "loss": 0.1446, "step": 5283 }, { "epoch": 2.4865882352941178, "grad_norm": 0.38270303469072003, "learning_rate": 0.00023606239744818094, "loss": 0.2661, "step": 5284 }, { "epoch": 2.487058823529412, "grad_norm": 0.3235800603402073, "learning_rate": 0.0002360264905876148, "loss": 0.1809, "step": 5285 }, { "epoch": 2.4875294117647058, "grad_norm": 0.08444827953107736, "learning_rate": 0.0002359905763799644, "loss": 0.0476, "step": 5286 }, { "epoch": 2.488, "grad_norm": 0.22549796916024056, "learning_rate": 0.00023595465482829697, "loss": 0.1329, "step": 5287 }, { "epoch": 2.488470588235294, "grad_norm": 0.257228942030384, "learning_rate": 0.0002359187259356805, "loss": 0.2514, "step": 5288 }, { "epoch": 2.488941176470588, "grad_norm": 0.26835379105489277, "learning_rate": 0.00023588278970518333, "loss": 0.2059, "step": 5289 }, { "epoch": 2.489411764705882, "grad_norm": 0.3627162727340388, "learning_rate": 0.00023584684613987472, "loss": 0.3635, "step": 5290 }, { "epoch": 2.4898823529411764, "grad_norm": 0.19677834672197445, "learning_rate": 0.00023581089524282432, "loss": 0.1447, "step": 5291 }, { "epoch": 2.4903529411764707, "grad_norm": 0.21509391564674882, "learning_rate": 0.0002357749370171026, "loss": 0.1303, "step": 5292 }, { "epoch": 2.490823529411765, "grad_norm": 0.33269852155540375, "learning_rate": 0.00023573897146578053, "loss": 0.2368, "step": 5293 }, { "epoch": 2.4912941176470587, "grad_norm": 0.13941931961724252, "learning_rate": 0.00023570299859192976, "loss": 0.0617, "step": 5294 }, { "epoch": 2.491764705882353, "grad_norm": 0.13488914067245933, "learning_rate": 0.00023566701839862253, "loss": 0.0677, "step": 5295 }, { "epoch": 2.492235294117647, "grad_norm": 0.6723908032321475, "learning_rate": 0.00023563103088893172, "loss": 0.5652, "step": 5296 }, { "epoch": 2.4927058823529413, "grad_norm": 0.49149005432919424, "learning_rate": 0.00023559503606593088, "loss": 0.4846, "step": 5297 }, { "epoch": 2.493176470588235, "grad_norm": 0.11228197680507358, "learning_rate": 0.0002355590339326941, "loss": 0.0441, "step": 5298 }, { "epoch": 2.4936470588235293, "grad_norm": 0.35597280402261855, "learning_rate": 0.00023552302449229612, "loss": 0.3691, "step": 5299 }, { "epoch": 2.4941176470588236, "grad_norm": 0.18372320633886144, "learning_rate": 0.00023548700774781242, "loss": 0.1102, "step": 5300 }, { "epoch": 2.4945882352941178, "grad_norm": 0.44858973696288834, "learning_rate": 0.00023545098370231893, "loss": 0.4786, "step": 5301 }, { "epoch": 2.495058823529412, "grad_norm": 0.3014830112144279, "learning_rate": 0.00023541495235889227, "loss": 0.2355, "step": 5302 }, { "epoch": 2.4955294117647058, "grad_norm": 0.12584355566117614, "learning_rate": 0.0002353789137206098, "loss": 0.0528, "step": 5303 }, { "epoch": 2.496, "grad_norm": 0.4448077917208118, "learning_rate": 0.00023534286779054923, "loss": 0.4953, "step": 5304 }, { "epoch": 2.496470588235294, "grad_norm": 0.3277043234343936, "learning_rate": 0.00023530681457178916, "loss": 0.2302, "step": 5305 }, { "epoch": 2.496941176470588, "grad_norm": 0.39531774837734474, "learning_rate": 0.00023527075406740876, "loss": 0.3695, "step": 5306 }, { "epoch": 2.497411764705882, "grad_norm": 0.185720554552728, "learning_rate": 0.00023523468628048764, "loss": 0.1568, "step": 5307 }, { "epoch": 2.4978823529411764, "grad_norm": 0.4435552738247089, "learning_rate": 0.0002351986112141063, "loss": 0.2496, "step": 5308 }, { "epoch": 2.4983529411764707, "grad_norm": 0.2785870422733963, "learning_rate": 0.00023516252887134558, "loss": 0.2675, "step": 5309 }, { "epoch": 2.498823529411765, "grad_norm": 0.119899408033394, "learning_rate": 0.00023512643925528725, "loss": 0.0446, "step": 5310 }, { "epoch": 2.4992941176470587, "grad_norm": 0.16321240799692854, "learning_rate": 0.00023509034236901348, "loss": 0.1311, "step": 5311 }, { "epoch": 2.499764705882353, "grad_norm": 0.49492083596645536, "learning_rate": 0.00023505423821560705, "loss": 0.3892, "step": 5312 }, { "epoch": 2.500235294117647, "grad_norm": 0.3812351293719962, "learning_rate": 0.00023501812679815152, "loss": 0.2828, "step": 5313 }, { "epoch": 2.5007058823529413, "grad_norm": 0.35773596399904867, "learning_rate": 0.00023498200811973092, "loss": 0.3268, "step": 5314 }, { "epoch": 2.501176470588235, "grad_norm": 0.2745955864025251, "learning_rate": 0.00023494588218343003, "loss": 0.1623, "step": 5315 }, { "epoch": 2.5016470588235293, "grad_norm": 0.1804369973297509, "learning_rate": 0.0002349097489923341, "loss": 0.0788, "step": 5316 }, { "epoch": 2.5021176470588236, "grad_norm": 0.1486976716834443, "learning_rate": 0.00023487360854952915, "loss": 0.0615, "step": 5317 }, { "epoch": 2.502588235294118, "grad_norm": 0.2025030536710642, "learning_rate": 0.00023483746085810176, "loss": 0.0841, "step": 5318 }, { "epoch": 2.503058823529412, "grad_norm": 0.47191621156149155, "learning_rate": 0.00023480130592113904, "loss": 0.3342, "step": 5319 }, { "epoch": 2.503529411764706, "grad_norm": 0.692064349687824, "learning_rate": 0.00023476514374172887, "loss": 0.4121, "step": 5320 }, { "epoch": 2.504, "grad_norm": 0.2656308025647586, "learning_rate": 0.0002347289743229596, "loss": 0.1165, "step": 5321 }, { "epoch": 2.5044705882352942, "grad_norm": 0.4160137551264288, "learning_rate": 0.00023469279766792037, "loss": 0.2965, "step": 5322 }, { "epoch": 2.504941176470588, "grad_norm": 0.33365062518843325, "learning_rate": 0.0002346566137797008, "loss": 0.3435, "step": 5323 }, { "epoch": 2.5054117647058822, "grad_norm": 0.1553193305761654, "learning_rate": 0.0002346204226613912, "loss": 0.1116, "step": 5324 }, { "epoch": 2.5058823529411764, "grad_norm": 0.2995028294737103, "learning_rate": 0.0002345842243160824, "loss": 0.243, "step": 5325 }, { "epoch": 2.5063529411764707, "grad_norm": 0.48650826978780704, "learning_rate": 0.00023454801874686596, "loss": 0.2109, "step": 5326 }, { "epoch": 2.506823529411765, "grad_norm": 0.6749687197461313, "learning_rate": 0.00023451180595683409, "loss": 0.3259, "step": 5327 }, { "epoch": 2.5072941176470587, "grad_norm": 0.32616988957197546, "learning_rate": 0.0002344755859490794, "loss": 0.1906, "step": 5328 }, { "epoch": 2.507764705882353, "grad_norm": 0.2572843598890886, "learning_rate": 0.00023443935872669534, "loss": 0.1717, "step": 5329 }, { "epoch": 2.508235294117647, "grad_norm": 0.28113950646224806, "learning_rate": 0.00023440312429277583, "loss": 0.2547, "step": 5330 }, { "epoch": 2.5087058823529413, "grad_norm": 0.24426521613132815, "learning_rate": 0.00023436688265041555, "loss": 0.2086, "step": 5331 }, { "epoch": 2.509176470588235, "grad_norm": 0.06592571569699471, "learning_rate": 0.0002343306338027097, "loss": 0.0199, "step": 5332 }, { "epoch": 2.5096470588235293, "grad_norm": 0.20468181134671665, "learning_rate": 0.00023429437775275406, "loss": 0.1045, "step": 5333 }, { "epoch": 2.5101176470588236, "grad_norm": 0.45748117855135906, "learning_rate": 0.00023425811450364508, "loss": 0.2499, "step": 5334 }, { "epoch": 2.510588235294118, "grad_norm": 0.25585061459707953, "learning_rate": 0.00023422184405847985, "loss": 0.155, "step": 5335 }, { "epoch": 2.511058823529412, "grad_norm": 0.18118593713023476, "learning_rate": 0.00023418556642035602, "loss": 0.1076, "step": 5336 }, { "epoch": 2.511529411764706, "grad_norm": 0.4511706469771155, "learning_rate": 0.0002341492815923719, "loss": 0.3246, "step": 5337 }, { "epoch": 2.512, "grad_norm": 0.2536112952674027, "learning_rate": 0.00023411298957762636, "loss": 0.1482, "step": 5338 }, { "epoch": 2.5124705882352942, "grad_norm": 0.43856921720560915, "learning_rate": 0.00023407669037921902, "loss": 0.5116, "step": 5339 }, { "epoch": 2.512941176470588, "grad_norm": 0.204544066054927, "learning_rate": 0.00023404038400024984, "loss": 0.1166, "step": 5340 }, { "epoch": 2.5134117647058822, "grad_norm": 0.3148301134495848, "learning_rate": 0.00023400407044381965, "loss": 0.2948, "step": 5341 }, { "epoch": 2.5138823529411765, "grad_norm": 0.23560196467440545, "learning_rate": 0.00023396774971302985, "loss": 0.1412, "step": 5342 }, { "epoch": 2.5143529411764707, "grad_norm": 0.5333722593228837, "learning_rate": 0.00023393142181098235, "loss": 0.5203, "step": 5343 }, { "epoch": 2.514823529411765, "grad_norm": 0.2008637804017884, "learning_rate": 0.00023389508674077973, "loss": 0.0652, "step": 5344 }, { "epoch": 2.5152941176470587, "grad_norm": 0.4910028462327897, "learning_rate": 0.00023385874450552522, "loss": 0.472, "step": 5345 }, { "epoch": 2.515764705882353, "grad_norm": 0.3754202741095511, "learning_rate": 0.00023382239510832262, "loss": 0.2641, "step": 5346 }, { "epoch": 2.516235294117647, "grad_norm": 0.19558489825961298, "learning_rate": 0.0002337860385522763, "loss": 0.1581, "step": 5347 }, { "epoch": 2.516705882352941, "grad_norm": 0.16318955981327835, "learning_rate": 0.00023374967484049132, "loss": 0.0919, "step": 5348 }, { "epoch": 2.517176470588235, "grad_norm": 0.1477395721947332, "learning_rate": 0.00023371330397607335, "loss": 0.0886, "step": 5349 }, { "epoch": 2.5176470588235293, "grad_norm": 0.4536868008594657, "learning_rate": 0.00023367692596212858, "loss": 0.3091, "step": 5350 }, { "epoch": 2.5181176470588236, "grad_norm": 0.49722302289927717, "learning_rate": 0.0002336405408017639, "loss": 0.3376, "step": 5351 }, { "epoch": 2.518588235294118, "grad_norm": 0.5335029843175161, "learning_rate": 0.0002336041484980868, "loss": 0.3613, "step": 5352 }, { "epoch": 2.519058823529412, "grad_norm": 0.3227250946572619, "learning_rate": 0.0002335677490542053, "loss": 0.3059, "step": 5353 }, { "epoch": 2.519529411764706, "grad_norm": 0.2990759013313029, "learning_rate": 0.0002335313424732282, "loss": 0.2676, "step": 5354 }, { "epoch": 2.52, "grad_norm": 0.3966576701047312, "learning_rate": 0.0002334949287582647, "loss": 0.0902, "step": 5355 }, { "epoch": 2.5204705882352942, "grad_norm": 0.3205244546939481, "learning_rate": 0.00023345850791242477, "loss": 0.2159, "step": 5356 }, { "epoch": 2.520941176470588, "grad_norm": 0.21863779771972885, "learning_rate": 0.00023342207993881883, "loss": 0.1361, "step": 5357 }, { "epoch": 2.5214117647058822, "grad_norm": 0.2910829076381313, "learning_rate": 0.0002333856448405581, "loss": 0.169, "step": 5358 }, { "epoch": 2.5218823529411765, "grad_norm": 0.2108580871152502, "learning_rate": 0.00023334920262075437, "loss": 0.0827, "step": 5359 }, { "epoch": 2.5223529411764707, "grad_norm": 0.291188304806656, "learning_rate": 0.00023331275328251983, "loss": 0.0774, "step": 5360 }, { "epoch": 2.522823529411765, "grad_norm": 0.5895630325506344, "learning_rate": 0.00023327629682896758, "loss": 0.4141, "step": 5361 }, { "epoch": 2.5232941176470587, "grad_norm": 0.3917746054248645, "learning_rate": 0.00023323983326321106, "loss": 0.306, "step": 5362 }, { "epoch": 2.523764705882353, "grad_norm": 0.37979656076976287, "learning_rate": 0.0002332033625883645, "loss": 0.2967, "step": 5363 }, { "epoch": 2.524235294117647, "grad_norm": 0.3558363840139796, "learning_rate": 0.00023316688480754267, "loss": 0.3719, "step": 5364 }, { "epoch": 2.524705882352941, "grad_norm": 0.25975519404320785, "learning_rate": 0.00023313039992386094, "loss": 0.1112, "step": 5365 }, { "epoch": 2.525176470588235, "grad_norm": 0.31107545161178196, "learning_rate": 0.00023309390794043532, "loss": 0.185, "step": 5366 }, { "epoch": 2.5256470588235294, "grad_norm": 0.481082152998056, "learning_rate": 0.00023305740886038235, "loss": 0.2488, "step": 5367 }, { "epoch": 2.5261176470588236, "grad_norm": 0.15226310798568665, "learning_rate": 0.00023302090268681932, "loss": 0.0843, "step": 5368 }, { "epoch": 2.526588235294118, "grad_norm": 0.4604998421616377, "learning_rate": 0.00023298438942286395, "loss": 0.2481, "step": 5369 }, { "epoch": 2.527058823529412, "grad_norm": 0.47137092993181523, "learning_rate": 0.00023294786907163473, "loss": 0.2807, "step": 5370 }, { "epoch": 2.527529411764706, "grad_norm": 0.3182479771996685, "learning_rate": 0.0002329113416362506, "loss": 0.2738, "step": 5371 }, { "epoch": 2.528, "grad_norm": 0.4002786142763672, "learning_rate": 0.00023287480711983125, "loss": 0.3718, "step": 5372 }, { "epoch": 2.5284705882352942, "grad_norm": 0.3036438464205468, "learning_rate": 0.00023283826552549682, "loss": 0.2689, "step": 5373 }, { "epoch": 2.528941176470588, "grad_norm": 0.34977354999237975, "learning_rate": 0.00023280171685636822, "loss": 0.1796, "step": 5374 }, { "epoch": 2.5294117647058822, "grad_norm": 0.5635955198377395, "learning_rate": 0.0002327651611155669, "loss": 0.4149, "step": 5375 }, { "epoch": 2.5298823529411765, "grad_norm": 0.23725930576196266, "learning_rate": 0.0002327285983062148, "loss": 0.1275, "step": 5376 }, { "epoch": 2.5303529411764707, "grad_norm": 0.42765198991600156, "learning_rate": 0.00023269202843143462, "loss": 0.3196, "step": 5377 }, { "epoch": 2.530823529411765, "grad_norm": 0.19446128272038088, "learning_rate": 0.00023265545149434968, "loss": 0.1155, "step": 5378 }, { "epoch": 2.5312941176470587, "grad_norm": 0.35411191160818417, "learning_rate": 0.0002326188674980837, "loss": 0.2697, "step": 5379 }, { "epoch": 2.531764705882353, "grad_norm": 0.5230827642100022, "learning_rate": 0.0002325822764457612, "loss": 0.546, "step": 5380 }, { "epoch": 2.532235294117647, "grad_norm": 0.4586211600965953, "learning_rate": 0.0002325456783405072, "loss": 0.6035, "step": 5381 }, { "epoch": 2.532705882352941, "grad_norm": 0.5865578537817507, "learning_rate": 0.00023250907318544744, "loss": 0.498, "step": 5382 }, { "epoch": 2.533176470588235, "grad_norm": 0.3598313941298094, "learning_rate": 0.0002324724609837081, "loss": 0.3191, "step": 5383 }, { "epoch": 2.5336470588235294, "grad_norm": 0.3427871395221994, "learning_rate": 0.0002324358417384161, "loss": 0.4257, "step": 5384 }, { "epoch": 2.5341176470588236, "grad_norm": 0.37918114128002967, "learning_rate": 0.00023239921545269884, "loss": 0.3345, "step": 5385 }, { "epoch": 2.534588235294118, "grad_norm": 0.3152060423417744, "learning_rate": 0.00023236258212968445, "loss": 0.2582, "step": 5386 }, { "epoch": 2.535058823529412, "grad_norm": 0.21024811228034415, "learning_rate": 0.00023232594177250148, "loss": 0.1433, "step": 5387 }, { "epoch": 2.535529411764706, "grad_norm": 0.28007067368180966, "learning_rate": 0.00023228929438427932, "loss": 0.3443, "step": 5388 }, { "epoch": 2.536, "grad_norm": 0.25708954298088904, "learning_rate": 0.0002322526399681478, "loss": 0.207, "step": 5389 }, { "epoch": 2.5364705882352943, "grad_norm": 0.2533217202894044, "learning_rate": 0.0002322159785272374, "loss": 0.1595, "step": 5390 }, { "epoch": 2.536941176470588, "grad_norm": 0.35880044840134145, "learning_rate": 0.00023217931006467912, "loss": 0.313, "step": 5391 }, { "epoch": 2.5374117647058823, "grad_norm": 0.2850248217654285, "learning_rate": 0.00023214263458360467, "loss": 0.2759, "step": 5392 }, { "epoch": 2.5378823529411765, "grad_norm": 0.4039846494662139, "learning_rate": 0.00023210595208714635, "loss": 0.4036, "step": 5393 }, { "epoch": 2.5383529411764707, "grad_norm": 0.0717257347097812, "learning_rate": 0.00023206926257843702, "loss": 0.0184, "step": 5394 }, { "epoch": 2.538823529411765, "grad_norm": 0.39876494988087363, "learning_rate": 0.00023203256606061008, "loss": 0.2786, "step": 5395 }, { "epoch": 2.5392941176470587, "grad_norm": 0.26787468519867164, "learning_rate": 0.00023199586253679964, "loss": 0.1967, "step": 5396 }, { "epoch": 2.539764705882353, "grad_norm": 0.506195315322371, "learning_rate": 0.00023195915201014035, "loss": 0.4439, "step": 5397 }, { "epoch": 2.540235294117647, "grad_norm": 0.1924288256708935, "learning_rate": 0.00023192243448376745, "loss": 0.0996, "step": 5398 }, { "epoch": 2.540705882352941, "grad_norm": 0.32744666721049953, "learning_rate": 0.00023188570996081687, "loss": 0.3169, "step": 5399 }, { "epoch": 2.541176470588235, "grad_norm": 0.4201764966151249, "learning_rate": 0.00023184897844442495, "loss": 0.3279, "step": 5400 }, { "epoch": 2.5416470588235294, "grad_norm": 0.5233769790983939, "learning_rate": 0.00023181223993772886, "loss": 0.457, "step": 5401 }, { "epoch": 2.5421176470588236, "grad_norm": 0.49370061992852576, "learning_rate": 0.00023177549444386618, "loss": 0.4098, "step": 5402 }, { "epoch": 2.542588235294118, "grad_norm": 0.36587676670839026, "learning_rate": 0.00023173874196597515, "loss": 0.287, "step": 5403 }, { "epoch": 2.543058823529412, "grad_norm": 0.4358553405161352, "learning_rate": 0.00023170198250719464, "loss": 0.3276, "step": 5404 }, { "epoch": 2.543529411764706, "grad_norm": 0.4718998790089953, "learning_rate": 0.00023166521607066409, "loss": 0.5566, "step": 5405 }, { "epoch": 2.544, "grad_norm": 0.3172135838555579, "learning_rate": 0.00023162844265952348, "loss": 0.2133, "step": 5406 }, { "epoch": 2.5444705882352943, "grad_norm": 0.4392613351977296, "learning_rate": 0.0002315916622769135, "loss": 0.63, "step": 5407 }, { "epoch": 2.544941176470588, "grad_norm": 0.4203136046674954, "learning_rate": 0.00023155487492597536, "loss": 0.2099, "step": 5408 }, { "epoch": 2.5454117647058823, "grad_norm": 0.38217244618200086, "learning_rate": 0.0002315180806098509, "loss": 0.3088, "step": 5409 }, { "epoch": 2.5458823529411765, "grad_norm": 0.31812309693413166, "learning_rate": 0.00023148127933168245, "loss": 0.2299, "step": 5410 }, { "epoch": 2.5463529411764707, "grad_norm": 0.3377311507670496, "learning_rate": 0.00023144447109461314, "loss": 0.3349, "step": 5411 }, { "epoch": 2.546823529411765, "grad_norm": 0.36111507988093783, "learning_rate": 0.0002314076559017865, "loss": 0.3245, "step": 5412 }, { "epoch": 2.5472941176470587, "grad_norm": 0.5153053681115035, "learning_rate": 0.00023137083375634676, "loss": 0.3956, "step": 5413 }, { "epoch": 2.547764705882353, "grad_norm": 0.5584700679933942, "learning_rate": 0.00023133400466143867, "loss": 0.5725, "step": 5414 }, { "epoch": 2.548235294117647, "grad_norm": 0.11235282549584091, "learning_rate": 0.00023129716862020766, "loss": 0.063, "step": 5415 }, { "epoch": 2.548705882352941, "grad_norm": 0.24808232730372204, "learning_rate": 0.0002312603256357997, "loss": 0.1543, "step": 5416 }, { "epoch": 2.549176470588235, "grad_norm": 0.37278469209618326, "learning_rate": 0.00023122347571136127, "loss": 0.4107, "step": 5417 }, { "epoch": 2.5496470588235294, "grad_norm": 0.4762630183331209, "learning_rate": 0.00023118661885003969, "loss": 0.593, "step": 5418 }, { "epoch": 2.5501176470588236, "grad_norm": 0.4000085609886227, "learning_rate": 0.00023114975505498264, "loss": 0.341, "step": 5419 }, { "epoch": 2.550588235294118, "grad_norm": 0.6191566262284893, "learning_rate": 0.00023111288432933847, "loss": 0.5137, "step": 5420 }, { "epoch": 2.5510588235294116, "grad_norm": 0.3186378941185016, "learning_rate": 0.0002310760066762561, "loss": 0.1351, "step": 5421 }, { "epoch": 2.551529411764706, "grad_norm": 0.39425812724582565, "learning_rate": 0.00023103912209888512, "loss": 0.2654, "step": 5422 }, { "epoch": 2.552, "grad_norm": 0.28141997413732406, "learning_rate": 0.00023100223060037558, "loss": 0.2168, "step": 5423 }, { "epoch": 2.5524705882352943, "grad_norm": 0.48258952808882205, "learning_rate": 0.00023096533218387827, "loss": 0.4377, "step": 5424 }, { "epoch": 2.552941176470588, "grad_norm": 0.19307205786313597, "learning_rate": 0.00023092842685254442, "loss": 0.1734, "step": 5425 }, { "epoch": 2.5534117647058823, "grad_norm": 0.40688559885549613, "learning_rate": 0.000230891514609526, "loss": 0.3147, "step": 5426 }, { "epoch": 2.5538823529411765, "grad_norm": 0.3951275719799203, "learning_rate": 0.00023085459545797542, "loss": 0.427, "step": 5427 }, { "epoch": 2.5543529411764707, "grad_norm": 0.3042663759665769, "learning_rate": 0.00023081766940104585, "loss": 0.3567, "step": 5428 }, { "epoch": 2.554823529411765, "grad_norm": 0.1916356485961053, "learning_rate": 0.00023078073644189085, "loss": 0.0721, "step": 5429 }, { "epoch": 2.5552941176470587, "grad_norm": 0.5195233872362293, "learning_rate": 0.00023074379658366478, "loss": 0.2781, "step": 5430 }, { "epoch": 2.555764705882353, "grad_norm": 0.506259040857141, "learning_rate": 0.0002307068498295224, "loss": 0.5265, "step": 5431 }, { "epoch": 2.556235294117647, "grad_norm": 0.3062236203286039, "learning_rate": 0.00023066989618261919, "loss": 0.2543, "step": 5432 }, { "epoch": 2.556705882352941, "grad_norm": 0.18902317264514837, "learning_rate": 0.00023063293564611114, "loss": 0.1648, "step": 5433 }, { "epoch": 2.557176470588235, "grad_norm": 0.37744084697143904, "learning_rate": 0.00023059596822315491, "loss": 0.389, "step": 5434 }, { "epoch": 2.5576470588235294, "grad_norm": 0.3322164561617665, "learning_rate": 0.00023055899391690766, "loss": 0.274, "step": 5435 }, { "epoch": 2.5581176470588236, "grad_norm": 0.26104291321791706, "learning_rate": 0.00023052201273052723, "loss": 0.1169, "step": 5436 }, { "epoch": 2.558588235294118, "grad_norm": 0.1704906442030127, "learning_rate": 0.00023048502466717192, "loss": 0.1289, "step": 5437 }, { "epoch": 2.5590588235294116, "grad_norm": 0.48509153859269527, "learning_rate": 0.00023044802973000073, "loss": 0.284, "step": 5438 }, { "epoch": 2.559529411764706, "grad_norm": 0.27742188952033336, "learning_rate": 0.00023041102792217323, "loss": 0.2025, "step": 5439 }, { "epoch": 2.56, "grad_norm": 0.31930309242284005, "learning_rate": 0.00023037401924684946, "loss": 0.1455, "step": 5440 }, { "epoch": 2.5604705882352943, "grad_norm": 0.3424749055999614, "learning_rate": 0.00023033700370719026, "loss": 0.3499, "step": 5441 }, { "epoch": 2.560941176470588, "grad_norm": 0.29877994033874195, "learning_rate": 0.00023029998130635692, "loss": 0.2674, "step": 5442 }, { "epoch": 2.5614117647058823, "grad_norm": 0.5930506347495271, "learning_rate": 0.00023026295204751126, "loss": 0.5788, "step": 5443 }, { "epoch": 2.5618823529411765, "grad_norm": 0.3343404029925483, "learning_rate": 0.00023022591593381582, "loss": 0.192, "step": 5444 }, { "epoch": 2.5623529411764707, "grad_norm": 0.17375184118122763, "learning_rate": 0.0002301888729684337, "loss": 0.1443, "step": 5445 }, { "epoch": 2.562823529411765, "grad_norm": 0.42754373404068924, "learning_rate": 0.00023015182315452843, "loss": 0.2699, "step": 5446 }, { "epoch": 2.5632941176470587, "grad_norm": 0.25419280919507814, "learning_rate": 0.00023011476649526438, "loss": 0.1522, "step": 5447 }, { "epoch": 2.563764705882353, "grad_norm": 0.4093849362179796, "learning_rate": 0.0002300777029938063, "loss": 0.1205, "step": 5448 }, { "epoch": 2.564235294117647, "grad_norm": 0.4152140116327804, "learning_rate": 0.0002300406326533196, "loss": 0.3495, "step": 5449 }, { "epoch": 2.564705882352941, "grad_norm": 0.31750872189355567, "learning_rate": 0.00023000355547697027, "loss": 0.1706, "step": 5450 }, { "epoch": 2.565176470588235, "grad_norm": 0.5927238660985327, "learning_rate": 0.00022996647146792488, "loss": 0.479, "step": 5451 }, { "epoch": 2.5656470588235294, "grad_norm": 0.24965573843162503, "learning_rate": 0.0002299293806293506, "loss": 0.1552, "step": 5452 }, { "epoch": 2.5661176470588236, "grad_norm": 0.45952740697624544, "learning_rate": 0.00022989228296441515, "loss": 0.2912, "step": 5453 }, { "epoch": 2.566588235294118, "grad_norm": 0.3903018838329916, "learning_rate": 0.00022985517847628686, "loss": 0.3999, "step": 5454 }, { "epoch": 2.5670588235294116, "grad_norm": 0.420929006895542, "learning_rate": 0.0002298180671681347, "loss": 0.3321, "step": 5455 }, { "epoch": 2.567529411764706, "grad_norm": 0.3506328668651954, "learning_rate": 0.00022978094904312803, "loss": 0.1971, "step": 5456 }, { "epoch": 2.568, "grad_norm": 0.36117740350561084, "learning_rate": 0.000229743824104437, "loss": 0.3027, "step": 5457 }, { "epoch": 2.5684705882352943, "grad_norm": 0.14650400137751166, "learning_rate": 0.00022970669235523224, "loss": 0.0627, "step": 5458 }, { "epoch": 2.568941176470588, "grad_norm": 0.09735415538901115, "learning_rate": 0.00022966955379868502, "loss": 0.0188, "step": 5459 }, { "epoch": 2.5694117647058823, "grad_norm": 0.4220049688488414, "learning_rate": 0.00022963240843796714, "loss": 0.2607, "step": 5460 }, { "epoch": 2.5698823529411765, "grad_norm": 0.35941431735884954, "learning_rate": 0.000229595256276251, "loss": 0.2688, "step": 5461 }, { "epoch": 2.5703529411764707, "grad_norm": 0.15301515708776692, "learning_rate": 0.00022955809731670955, "loss": 0.0612, "step": 5462 }, { "epoch": 2.570823529411765, "grad_norm": 0.38195016632493894, "learning_rate": 0.00022952093156251635, "loss": 0.2917, "step": 5463 }, { "epoch": 2.5712941176470587, "grad_norm": 0.2519886673484159, "learning_rate": 0.0002294837590168455, "loss": 0.1757, "step": 5464 }, { "epoch": 2.571764705882353, "grad_norm": 0.46571782010912005, "learning_rate": 0.00022944657968287184, "loss": 0.3402, "step": 5465 }, { "epoch": 2.572235294117647, "grad_norm": 0.29521202904666544, "learning_rate": 0.00022940939356377052, "loss": 0.3105, "step": 5466 }, { "epoch": 2.572705882352941, "grad_norm": 0.2802724457576281, "learning_rate": 0.00022937220066271752, "loss": 0.2455, "step": 5467 }, { "epoch": 2.573176470588235, "grad_norm": 0.3039500672028715, "learning_rate": 0.0002293350009828893, "loss": 0.1683, "step": 5468 }, { "epoch": 2.5736470588235294, "grad_norm": 0.35433028912792564, "learning_rate": 0.00022929779452746283, "loss": 0.2993, "step": 5469 }, { "epoch": 2.5741176470588236, "grad_norm": 0.3058797185893273, "learning_rate": 0.0002292605812996157, "loss": 0.2203, "step": 5470 }, { "epoch": 2.574588235294118, "grad_norm": 0.18689469558614905, "learning_rate": 0.00022922336130252622, "loss": 0.0739, "step": 5471 }, { "epoch": 2.5750588235294116, "grad_norm": 0.633589097417056, "learning_rate": 0.00022918613453937314, "loss": 0.6831, "step": 5472 }, { "epoch": 2.575529411764706, "grad_norm": 0.19622919201410258, "learning_rate": 0.00022914890101333573, "loss": 0.1442, "step": 5473 }, { "epoch": 2.576, "grad_norm": 0.38300304123150497, "learning_rate": 0.00022911166072759386, "loss": 0.217, "step": 5474 }, { "epoch": 2.576470588235294, "grad_norm": 0.4238499347782306, "learning_rate": 0.00022907441368532822, "loss": 0.2322, "step": 5475 }, { "epoch": 2.576941176470588, "grad_norm": 0.3148677612298798, "learning_rate": 0.00022903715988971973, "loss": 0.1434, "step": 5476 }, { "epoch": 2.5774117647058823, "grad_norm": 0.3028928904664001, "learning_rate": 0.0002289998993439502, "loss": 0.1781, "step": 5477 }, { "epoch": 2.5778823529411765, "grad_norm": 0.16896513366159144, "learning_rate": 0.00022896263205120167, "loss": 0.1046, "step": 5478 }, { "epoch": 2.5783529411764707, "grad_norm": 0.33224649342785834, "learning_rate": 0.0002289253580146571, "loss": 0.2456, "step": 5479 }, { "epoch": 2.578823529411765, "grad_norm": 0.3216034220084752, "learning_rate": 0.00022888807723749988, "loss": 0.2513, "step": 5480 }, { "epoch": 2.5792941176470587, "grad_norm": 0.39750809734294346, "learning_rate": 0.0002288507897229138, "loss": 0.1794, "step": 5481 }, { "epoch": 2.579764705882353, "grad_norm": 0.20702795438599897, "learning_rate": 0.00022881349547408363, "loss": 0.1175, "step": 5482 }, { "epoch": 2.580235294117647, "grad_norm": 0.5557593765726425, "learning_rate": 0.00022877619449419436, "loss": 0.5062, "step": 5483 }, { "epoch": 2.580705882352941, "grad_norm": 0.2141397081105366, "learning_rate": 0.00022873888678643164, "loss": 0.1112, "step": 5484 }, { "epoch": 2.581176470588235, "grad_norm": 0.48405102880368, "learning_rate": 0.0002287015723539818, "loss": 0.3435, "step": 5485 }, { "epoch": 2.5816470588235294, "grad_norm": 0.17252476288763363, "learning_rate": 0.00022866425120003165, "loss": 0.1294, "step": 5486 }, { "epoch": 2.5821176470588236, "grad_norm": 0.20818720820968767, "learning_rate": 0.0002286269233277686, "loss": 0.113, "step": 5487 }, { "epoch": 2.582588235294118, "grad_norm": 0.46883607361924373, "learning_rate": 0.0002285895887403806, "loss": 0.3645, "step": 5488 }, { "epoch": 2.5830588235294116, "grad_norm": 0.3358842039603478, "learning_rate": 0.00022855224744105625, "loss": 0.2722, "step": 5489 }, { "epoch": 2.583529411764706, "grad_norm": 0.41322612582985524, "learning_rate": 0.00022851489943298472, "loss": 0.2186, "step": 5490 }, { "epoch": 2.584, "grad_norm": 0.20162021136423933, "learning_rate": 0.0002284775447193556, "loss": 0.11, "step": 5491 }, { "epoch": 2.584470588235294, "grad_norm": 0.20138281129386856, "learning_rate": 0.00022844018330335928, "loss": 0.1391, "step": 5492 }, { "epoch": 2.584941176470588, "grad_norm": 0.5936809944919131, "learning_rate": 0.00022840281518818657, "loss": 0.6526, "step": 5493 }, { "epoch": 2.5854117647058823, "grad_norm": 0.4189409539292712, "learning_rate": 0.00022836544037702883, "loss": 0.3044, "step": 5494 }, { "epoch": 2.5858823529411765, "grad_norm": 0.589026759381405, "learning_rate": 0.00022832805887307815, "loss": 0.3732, "step": 5495 }, { "epoch": 2.5863529411764707, "grad_norm": 0.30972685258265026, "learning_rate": 0.00022829067067952702, "loss": 0.2447, "step": 5496 }, { "epoch": 2.586823529411765, "grad_norm": 0.5336058340027218, "learning_rate": 0.00022825327579956863, "loss": 0.6024, "step": 5497 }, { "epoch": 2.5872941176470587, "grad_norm": 0.3021982111716024, "learning_rate": 0.0002282158742363966, "loss": 0.1795, "step": 5498 }, { "epoch": 2.587764705882353, "grad_norm": 0.3909506799672361, "learning_rate": 0.00022817846599320528, "loss": 0.3475, "step": 5499 }, { "epoch": 2.588235294117647, "grad_norm": 0.4054012672914366, "learning_rate": 0.00022814105107318952, "loss": 0.3438, "step": 5500 }, { "epoch": 2.588705882352941, "grad_norm": 0.20348408827828368, "learning_rate": 0.00022810362947954475, "loss": 0.1867, "step": 5501 }, { "epoch": 2.589176470588235, "grad_norm": 0.2647873552651724, "learning_rate": 0.00022806620121546684, "loss": 0.1167, "step": 5502 }, { "epoch": 2.5896470588235294, "grad_norm": 0.20987016625854588, "learning_rate": 0.0002280287662841525, "loss": 0.239, "step": 5503 }, { "epoch": 2.5901176470588236, "grad_norm": 0.23833750252960012, "learning_rate": 0.00022799132468879878, "loss": 0.1465, "step": 5504 }, { "epoch": 2.590588235294118, "grad_norm": 0.29508597220115984, "learning_rate": 0.00022795387643260336, "loss": 0.2941, "step": 5505 }, { "epoch": 2.5910588235294116, "grad_norm": 0.3199849872973361, "learning_rate": 0.00022791642151876455, "loss": 0.3111, "step": 5506 }, { "epoch": 2.591529411764706, "grad_norm": 0.21991959835965705, "learning_rate": 0.00022787895995048118, "loss": 0.1594, "step": 5507 }, { "epoch": 2.592, "grad_norm": 0.3498269104029512, "learning_rate": 0.00022784149173095262, "loss": 0.3133, "step": 5508 }, { "epoch": 2.592470588235294, "grad_norm": 0.25813365266107485, "learning_rate": 0.00022780401686337885, "loss": 0.1365, "step": 5509 }, { "epoch": 2.592941176470588, "grad_norm": 0.2956854065496514, "learning_rate": 0.00022776653535096044, "loss": 0.2082, "step": 5510 }, { "epoch": 2.5934117647058823, "grad_norm": 0.270545204900747, "learning_rate": 0.00022772904719689844, "loss": 0.2558, "step": 5511 }, { "epoch": 2.5938823529411765, "grad_norm": 0.29192345340983034, "learning_rate": 0.00022769155240439455, "loss": 0.2329, "step": 5512 }, { "epoch": 2.5943529411764708, "grad_norm": 0.2391972204762227, "learning_rate": 0.00022765405097665103, "loss": 0.1391, "step": 5513 }, { "epoch": 2.594823529411765, "grad_norm": 0.3282502287079089, "learning_rate": 0.00022761654291687065, "loss": 0.2314, "step": 5514 }, { "epoch": 2.5952941176470588, "grad_norm": 0.4103198916199713, "learning_rate": 0.00022757902822825683, "loss": 0.2283, "step": 5515 }, { "epoch": 2.595764705882353, "grad_norm": 0.35751042536946315, "learning_rate": 0.00022754150691401345, "loss": 0.3365, "step": 5516 }, { "epoch": 2.596235294117647, "grad_norm": 0.41631368039373173, "learning_rate": 0.00022750397897734504, "loss": 0.3376, "step": 5517 }, { "epoch": 2.596705882352941, "grad_norm": 0.26250803537022915, "learning_rate": 0.00022746644442145665, "loss": 0.0948, "step": 5518 }, { "epoch": 2.597176470588235, "grad_norm": 0.38753324332106037, "learning_rate": 0.00022742890324955396, "loss": 0.37, "step": 5519 }, { "epoch": 2.5976470588235294, "grad_norm": 0.23733856662073097, "learning_rate": 0.00022739135546484315, "loss": 0.1094, "step": 5520 }, { "epoch": 2.5981176470588236, "grad_norm": 0.30006916349767, "learning_rate": 0.00022735380107053094, "loss": 0.2872, "step": 5521 }, { "epoch": 2.598588235294118, "grad_norm": 0.28930198465788265, "learning_rate": 0.00022731624006982473, "loss": 0.1705, "step": 5522 }, { "epoch": 2.5990588235294116, "grad_norm": 0.30377458076066977, "learning_rate": 0.0002272786724659324, "loss": 0.2051, "step": 5523 }, { "epoch": 2.599529411764706, "grad_norm": 0.3954227269662395, "learning_rate": 0.00022724109826206235, "loss": 0.2237, "step": 5524 }, { "epoch": 2.6, "grad_norm": 0.20462326149901777, "learning_rate": 0.00022720351746142366, "loss": 0.0902, "step": 5525 }, { "epoch": 2.600470588235294, "grad_norm": 0.24882336805203462, "learning_rate": 0.00022716593006722594, "loss": 0.1549, "step": 5526 }, { "epoch": 2.600941176470588, "grad_norm": 0.6961534265767553, "learning_rate": 0.0002271283360826793, "loss": 0.8123, "step": 5527 }, { "epoch": 2.6014117647058823, "grad_norm": 0.2727144807034017, "learning_rate": 0.0002270907355109944, "loss": 0.1727, "step": 5528 }, { "epoch": 2.6018823529411765, "grad_norm": 0.5432302236021578, "learning_rate": 0.00022705312835538257, "loss": 0.6716, "step": 5529 }, { "epoch": 2.6023529411764708, "grad_norm": 1.3859646497286668, "learning_rate": 0.00022701551461905563, "loss": 1.6608, "step": 5530 }, { "epoch": 2.602823529411765, "grad_norm": 0.29080155548379655, "learning_rate": 0.00022697789430522602, "loss": 0.1498, "step": 5531 }, { "epoch": 2.6032941176470588, "grad_norm": 0.21370735965314414, "learning_rate": 0.00022694026741710657, "loss": 0.1031, "step": 5532 }, { "epoch": 2.603764705882353, "grad_norm": 0.1194811070996535, "learning_rate": 0.00022690263395791095, "loss": 0.0673, "step": 5533 }, { "epoch": 2.604235294117647, "grad_norm": 0.2251903375270536, "learning_rate": 0.00022686499393085322, "loss": 0.1743, "step": 5534 }, { "epoch": 2.604705882352941, "grad_norm": 0.3063834561579126, "learning_rate": 0.00022682734733914792, "loss": 0.2546, "step": 5535 }, { "epoch": 2.605176470588235, "grad_norm": 0.7740408098232444, "learning_rate": 0.00022678969418601038, "loss": 0.4637, "step": 5536 }, { "epoch": 2.6056470588235294, "grad_norm": 0.31772807655244584, "learning_rate": 0.0002267520344746563, "loss": 0.2677, "step": 5537 }, { "epoch": 2.6061176470588236, "grad_norm": 0.1473983079507113, "learning_rate": 0.000226714368208302, "loss": 0.0608, "step": 5538 }, { "epoch": 2.606588235294118, "grad_norm": 0.23138524847289293, "learning_rate": 0.00022667669539016435, "loss": 0.1786, "step": 5539 }, { "epoch": 2.6070588235294117, "grad_norm": 0.5594449031319572, "learning_rate": 0.00022663901602346082, "loss": 0.5587, "step": 5540 }, { "epoch": 2.607529411764706, "grad_norm": 0.21507467764366509, "learning_rate": 0.00022660133011140942, "loss": 0.1372, "step": 5541 }, { "epoch": 2.608, "grad_norm": 0.6457612360559987, "learning_rate": 0.00022656363765722872, "loss": 0.6587, "step": 5542 }, { "epoch": 2.608470588235294, "grad_norm": 0.4053225944910266, "learning_rate": 0.00022652593866413775, "loss": 0.3144, "step": 5543 }, { "epoch": 2.608941176470588, "grad_norm": 0.31315189299723933, "learning_rate": 0.00022648823313535635, "loss": 0.1535, "step": 5544 }, { "epoch": 2.6094117647058823, "grad_norm": 0.24383542090663496, "learning_rate": 0.00022645052107410463, "loss": 0.0944, "step": 5545 }, { "epoch": 2.6098823529411765, "grad_norm": 0.13501545842914023, "learning_rate": 0.0002264128024836034, "loss": 0.0494, "step": 5546 }, { "epoch": 2.6103529411764708, "grad_norm": 0.5019337947007797, "learning_rate": 0.00022637507736707408, "loss": 0.2552, "step": 5547 }, { "epoch": 2.6108235294117645, "grad_norm": 0.4854847983915139, "learning_rate": 0.00022633734572773852, "loss": 0.3052, "step": 5548 }, { "epoch": 2.6112941176470588, "grad_norm": 0.6060412689891644, "learning_rate": 0.00022629960756881916, "loss": 0.4697, "step": 5549 }, { "epoch": 2.611764705882353, "grad_norm": 0.1997674982968379, "learning_rate": 0.00022626186289353913, "loss": 0.0911, "step": 5550 }, { "epoch": 2.612235294117647, "grad_norm": 0.2699691109859485, "learning_rate": 0.00022622411170512193, "loss": 0.245, "step": 5551 }, { "epoch": 2.612705882352941, "grad_norm": 0.29873247599061725, "learning_rate": 0.0002261863540067917, "loss": 0.2052, "step": 5552 }, { "epoch": 2.613176470588235, "grad_norm": 0.25024778545935217, "learning_rate": 0.00022614858980177314, "loss": 0.1516, "step": 5553 }, { "epoch": 2.6136470588235294, "grad_norm": 0.16967073191928694, "learning_rate": 0.00022611081909329153, "loss": 0.0538, "step": 5554 }, { "epoch": 2.6141176470588237, "grad_norm": 0.7487316584811183, "learning_rate": 0.00022607304188457264, "loss": 0.8995, "step": 5555 }, { "epoch": 2.614588235294118, "grad_norm": 0.42734883057190787, "learning_rate": 0.0002260352581788428, "loss": 0.3125, "step": 5556 }, { "epoch": 2.6150588235294117, "grad_norm": 0.19322974518771846, "learning_rate": 0.000225997467979329, "loss": 0.0925, "step": 5557 }, { "epoch": 2.615529411764706, "grad_norm": 0.26480795288406195, "learning_rate": 0.00022595967128925866, "loss": 0.1178, "step": 5558 }, { "epoch": 2.616, "grad_norm": 0.2949753008847512, "learning_rate": 0.0002259218681118598, "loss": 0.0636, "step": 5559 }, { "epoch": 2.616470588235294, "grad_norm": 0.36100349006000243, "learning_rate": 0.000225884058450361, "loss": 0.2286, "step": 5560 }, { "epoch": 2.616941176470588, "grad_norm": 0.3045276383606303, "learning_rate": 0.00022584624230799144, "loss": 0.2662, "step": 5561 }, { "epoch": 2.6174117647058823, "grad_norm": 0.4688287151696779, "learning_rate": 0.00022580841968798072, "loss": 0.2812, "step": 5562 }, { "epoch": 2.6178823529411765, "grad_norm": 0.28409076881409784, "learning_rate": 0.00022577059059355914, "loss": 0.2042, "step": 5563 }, { "epoch": 2.6183529411764708, "grad_norm": 0.6707044015119338, "learning_rate": 0.00022573275502795747, "loss": 0.6065, "step": 5564 }, { "epoch": 2.6188235294117646, "grad_norm": 0.2736540737862513, "learning_rate": 0.00022569491299440704, "loss": 0.2118, "step": 5565 }, { "epoch": 2.6192941176470588, "grad_norm": 0.5030786360387828, "learning_rate": 0.00022565706449613978, "loss": 0.3164, "step": 5566 }, { "epoch": 2.619764705882353, "grad_norm": 0.4389987344928062, "learning_rate": 0.0002256192095363881, "loss": 0.153, "step": 5567 }, { "epoch": 2.620235294117647, "grad_norm": 0.682448465021886, "learning_rate": 0.00022558134811838504, "loss": 0.1612, "step": 5568 }, { "epoch": 2.620705882352941, "grad_norm": 0.24629788708711275, "learning_rate": 0.00022554348024536413, "loss": 0.1738, "step": 5569 }, { "epoch": 2.621176470588235, "grad_norm": 0.3788351268059603, "learning_rate": 0.00022550560592055942, "loss": 0.3556, "step": 5570 }, { "epoch": 2.6216470588235294, "grad_norm": 0.380281187901144, "learning_rate": 0.00022546772514720566, "loss": 0.2178, "step": 5571 }, { "epoch": 2.6221176470588237, "grad_norm": 0.2343132221150397, "learning_rate": 0.00022542983792853797, "loss": 0.1808, "step": 5572 }, { "epoch": 2.622588235294118, "grad_norm": 0.31669269959390384, "learning_rate": 0.00022539194426779216, "loss": 0.1929, "step": 5573 }, { "epoch": 2.6230588235294117, "grad_norm": 0.18139535154774145, "learning_rate": 0.00022535404416820453, "loss": 0.0657, "step": 5574 }, { "epoch": 2.623529411764706, "grad_norm": 0.4326556434803, "learning_rate": 0.0002253161376330119, "loss": 0.3114, "step": 5575 }, { "epoch": 2.624, "grad_norm": 0.34215111477488913, "learning_rate": 0.00022527822466545164, "loss": 0.2113, "step": 5576 }, { "epoch": 2.624470588235294, "grad_norm": 0.3242681428519615, "learning_rate": 0.00022524030526876183, "loss": 0.2159, "step": 5577 }, { "epoch": 2.624941176470588, "grad_norm": 0.36529679324691877, "learning_rate": 0.0002252023794461809, "loss": 0.3963, "step": 5578 }, { "epoch": 2.6254117647058823, "grad_norm": 0.6263026212112598, "learning_rate": 0.00022516444720094785, "loss": 0.5534, "step": 5579 }, { "epoch": 2.6258823529411766, "grad_norm": 0.38208613390181195, "learning_rate": 0.00022512650853630234, "loss": 0.3503, "step": 5580 }, { "epoch": 2.626352941176471, "grad_norm": 0.32831044091755346, "learning_rate": 0.00022508856345548455, "loss": 0.1855, "step": 5581 }, { "epoch": 2.6268235294117646, "grad_norm": 0.4936470370335613, "learning_rate": 0.00022505061196173513, "loss": 0.2062, "step": 5582 }, { "epoch": 2.627294117647059, "grad_norm": 0.3464337319958351, "learning_rate": 0.00022501265405829526, "loss": 0.186, "step": 5583 }, { "epoch": 2.627764705882353, "grad_norm": 0.29638768334635734, "learning_rate": 0.00022497468974840687, "loss": 0.3354, "step": 5584 }, { "epoch": 2.6282352941176472, "grad_norm": 0.391475413328892, "learning_rate": 0.00022493671903531218, "loss": 0.3374, "step": 5585 }, { "epoch": 2.628705882352941, "grad_norm": 0.6215218807699431, "learning_rate": 0.00022489874192225414, "loss": 0.5237, "step": 5586 }, { "epoch": 2.6291764705882352, "grad_norm": 0.38755909642979874, "learning_rate": 0.00022486075841247617, "loss": 0.343, "step": 5587 }, { "epoch": 2.6296470588235294, "grad_norm": 0.1963345716105872, "learning_rate": 0.0002248227685092222, "loss": 0.1515, "step": 5588 }, { "epoch": 2.6301176470588237, "grad_norm": 0.46915008785695994, "learning_rate": 0.00022478477221573687, "loss": 0.4775, "step": 5589 }, { "epoch": 2.630588235294118, "grad_norm": 0.17834031793986374, "learning_rate": 0.00022474676953526508, "loss": 0.1085, "step": 5590 }, { "epoch": 2.6310588235294117, "grad_norm": 0.3093014639600467, "learning_rate": 0.00022470876047105262, "loss": 0.2974, "step": 5591 }, { "epoch": 2.631529411764706, "grad_norm": 0.2956453723139773, "learning_rate": 0.00022467074502634558, "loss": 0.2077, "step": 5592 }, { "epoch": 2.632, "grad_norm": 0.41200924117104465, "learning_rate": 0.00022463272320439063, "loss": 0.2927, "step": 5593 }, { "epoch": 2.632470588235294, "grad_norm": 0.4819517053764907, "learning_rate": 0.00022459469500843502, "loss": 0.3632, "step": 5594 }, { "epoch": 2.632941176470588, "grad_norm": 0.02702939237443022, "learning_rate": 0.0002245566604417266, "loss": 0.0049, "step": 5595 }, { "epoch": 2.6334117647058823, "grad_norm": 0.9062517872194586, "learning_rate": 0.0002245186195075137, "loss": 0.9065, "step": 5596 }, { "epoch": 2.6338823529411766, "grad_norm": 0.2377818189134823, "learning_rate": 0.0002244805722090451, "loss": 0.146, "step": 5597 }, { "epoch": 2.634352941176471, "grad_norm": 0.25111115621545443, "learning_rate": 0.0002244425185495704, "loss": 0.1142, "step": 5598 }, { "epoch": 2.6348235294117646, "grad_norm": 0.3794370133703001, "learning_rate": 0.00022440445853233945, "loss": 0.2187, "step": 5599 }, { "epoch": 2.635294117647059, "grad_norm": 0.4112828657485391, "learning_rate": 0.00022436639216060275, "loss": 0.2237, "step": 5600 }, { "epoch": 2.635764705882353, "grad_norm": 0.362870980974315, "learning_rate": 0.00022432831943761146, "loss": 0.3169, "step": 5601 }, { "epoch": 2.636235294117647, "grad_norm": 0.2692144902348936, "learning_rate": 0.00022429024036661708, "loss": 0.1822, "step": 5602 }, { "epoch": 2.636705882352941, "grad_norm": 0.3688809952403735, "learning_rate": 0.00022425215495087177, "loss": 0.3702, "step": 5603 }, { "epoch": 2.6371764705882352, "grad_norm": 0.4169085611173269, "learning_rate": 0.00022421406319362823, "loss": 0.2039, "step": 5604 }, { "epoch": 2.6376470588235295, "grad_norm": 0.13580320811074617, "learning_rate": 0.00022417596509813968, "loss": 0.0668, "step": 5605 }, { "epoch": 2.6381176470588237, "grad_norm": 0.03991560604147765, "learning_rate": 0.0002241378606676599, "loss": 0.0065, "step": 5606 }, { "epoch": 2.638588235294118, "grad_norm": 0.21846552822842796, "learning_rate": 0.00022409974990544312, "loss": 0.1593, "step": 5607 }, { "epoch": 2.6390588235294117, "grad_norm": 0.2986431452312309, "learning_rate": 0.0002240616328147443, "loss": 0.3662, "step": 5608 }, { "epoch": 2.639529411764706, "grad_norm": 0.4310186227040844, "learning_rate": 0.00022402350939881873, "loss": 0.4758, "step": 5609 }, { "epoch": 2.64, "grad_norm": 0.4340937867549348, "learning_rate": 0.00022398537966092236, "loss": 0.3418, "step": 5610 }, { "epoch": 2.640470588235294, "grad_norm": 0.2847649183799182, "learning_rate": 0.00022394724360431167, "loss": 0.3363, "step": 5611 }, { "epoch": 2.640941176470588, "grad_norm": 0.17369352941707425, "learning_rate": 0.00022390910123224373, "loss": 0.1, "step": 5612 }, { "epoch": 2.6414117647058823, "grad_norm": 0.3169251704291532, "learning_rate": 0.00022387095254797597, "loss": 0.3175, "step": 5613 }, { "epoch": 2.6418823529411766, "grad_norm": 0.16900681756251537, "learning_rate": 0.00022383279755476651, "loss": 0.092, "step": 5614 }, { "epoch": 2.642352941176471, "grad_norm": 0.2935812221400834, "learning_rate": 0.00022379463625587406, "loss": 0.3675, "step": 5615 }, { "epoch": 2.6428235294117646, "grad_norm": 0.5991162139968985, "learning_rate": 0.00022375646865455766, "loss": 0.4779, "step": 5616 }, { "epoch": 2.643294117647059, "grad_norm": 0.05277838063912425, "learning_rate": 0.00022371829475407713, "loss": 0.0083, "step": 5617 }, { "epoch": 2.643764705882353, "grad_norm": 0.3747331713885044, "learning_rate": 0.00022368011455769251, "loss": 0.3632, "step": 5618 }, { "epoch": 2.644235294117647, "grad_norm": 0.38845563533173255, "learning_rate": 0.00022364192806866481, "loss": 0.3362, "step": 5619 }, { "epoch": 2.644705882352941, "grad_norm": 0.35872301169722753, "learning_rate": 0.00022360373529025523, "loss": 0.3184, "step": 5620 }, { "epoch": 2.6451764705882352, "grad_norm": 0.4261118874081857, "learning_rate": 0.00022356553622572564, "loss": 0.3939, "step": 5621 }, { "epoch": 2.6456470588235295, "grad_norm": 0.33641373578951667, "learning_rate": 0.00022352733087833838, "loss": 0.3105, "step": 5622 }, { "epoch": 2.6461176470588237, "grad_norm": 0.35954620816396177, "learning_rate": 0.00022348911925135646, "loss": 0.3383, "step": 5623 }, { "epoch": 2.646588235294118, "grad_norm": 0.33855595587915577, "learning_rate": 0.00022345090134804325, "loss": 0.265, "step": 5624 }, { "epoch": 2.6470588235294117, "grad_norm": 0.21761650630461737, "learning_rate": 0.0002234126771716628, "loss": 0.1237, "step": 5625 }, { "epoch": 2.647529411764706, "grad_norm": 0.4490201862762724, "learning_rate": 0.00022337444672547967, "loss": 0.3826, "step": 5626 }, { "epoch": 2.648, "grad_norm": 0.45060940829848184, "learning_rate": 0.00022333621001275884, "loss": 0.3455, "step": 5627 }, { "epoch": 2.648470588235294, "grad_norm": 0.30216517625319383, "learning_rate": 0.00022329796703676597, "loss": 0.1913, "step": 5628 }, { "epoch": 2.648941176470588, "grad_norm": 0.39584895370205914, "learning_rate": 0.00022325971780076721, "loss": 0.3279, "step": 5629 }, { "epoch": 2.6494117647058824, "grad_norm": 0.3848701334888477, "learning_rate": 0.00022322146230802924, "loss": 0.3106, "step": 5630 }, { "epoch": 2.6498823529411766, "grad_norm": 0.38202048579079945, "learning_rate": 0.0002231832005618192, "loss": 0.3314, "step": 5631 }, { "epoch": 2.650352941176471, "grad_norm": 0.2688614094167543, "learning_rate": 0.00022314493256540485, "loss": 0.1917, "step": 5632 }, { "epoch": 2.6508235294117646, "grad_norm": 0.4080820362628382, "learning_rate": 0.0002231066583220545, "loss": 0.2675, "step": 5633 }, { "epoch": 2.651294117647059, "grad_norm": 0.4267002186572907, "learning_rate": 0.00022306837783503698, "loss": 0.3078, "step": 5634 }, { "epoch": 2.651764705882353, "grad_norm": 0.3925889864720041, "learning_rate": 0.00022303009110762154, "loss": 0.3054, "step": 5635 }, { "epoch": 2.652235294117647, "grad_norm": 0.44329241565417177, "learning_rate": 0.00022299179814307818, "loss": 0.2774, "step": 5636 }, { "epoch": 2.652705882352941, "grad_norm": 0.3776827303073911, "learning_rate": 0.0002229534989446772, "loss": 0.2258, "step": 5637 }, { "epoch": 2.6531764705882352, "grad_norm": 0.4876027626152273, "learning_rate": 0.00022291519351568955, "loss": 0.3964, "step": 5638 }, { "epoch": 2.6536470588235295, "grad_norm": 0.33573088346696395, "learning_rate": 0.00022287688185938673, "loss": 0.2231, "step": 5639 }, { "epoch": 2.6541176470588237, "grad_norm": 0.5059275617828519, "learning_rate": 0.00022283856397904077, "loss": 0.4756, "step": 5640 }, { "epoch": 2.654588235294118, "grad_norm": 0.21988661120723255, "learning_rate": 0.00022280023987792416, "loss": 0.1577, "step": 5641 }, { "epoch": 2.6550588235294117, "grad_norm": 0.2866403656327071, "learning_rate": 0.00022276190955930998, "loss": 0.2456, "step": 5642 }, { "epoch": 2.655529411764706, "grad_norm": 0.5713640385358009, "learning_rate": 0.00022272357302647187, "loss": 0.469, "step": 5643 }, { "epoch": 2.656, "grad_norm": 0.35478328935289954, "learning_rate": 0.00022268523028268385, "loss": 0.2841, "step": 5644 }, { "epoch": 2.656470588235294, "grad_norm": 0.21202712507227847, "learning_rate": 0.00022264688133122069, "loss": 0.1467, "step": 5645 }, { "epoch": 2.656941176470588, "grad_norm": 0.3144536562864387, "learning_rate": 0.00022260852617535752, "loss": 0.2396, "step": 5646 }, { "epoch": 2.6574117647058824, "grad_norm": 0.21972181608354416, "learning_rate": 0.00022257016481837006, "loss": 0.2848, "step": 5647 }, { "epoch": 2.6578823529411766, "grad_norm": 0.38353515755044243, "learning_rate": 0.00022253179726353458, "loss": 0.2956, "step": 5648 }, { "epoch": 2.658352941176471, "grad_norm": 0.36189504643012715, "learning_rate": 0.00022249342351412783, "loss": 0.2137, "step": 5649 }, { "epoch": 2.6588235294117646, "grad_norm": 0.2734435991899589, "learning_rate": 0.00022245504357342716, "loss": 0.1544, "step": 5650 }, { "epoch": 2.659294117647059, "grad_norm": 0.32187702008650604, "learning_rate": 0.00022241665744471037, "loss": 0.2943, "step": 5651 }, { "epoch": 2.659764705882353, "grad_norm": 0.1345159462865122, "learning_rate": 0.00022237826513125584, "loss": 0.0657, "step": 5652 }, { "epoch": 2.660235294117647, "grad_norm": 0.4813600478313249, "learning_rate": 0.00022233986663634241, "loss": 0.2684, "step": 5653 }, { "epoch": 2.660705882352941, "grad_norm": 0.37524415053929305, "learning_rate": 0.0002223014619632496, "loss": 0.2499, "step": 5654 }, { "epoch": 2.6611764705882353, "grad_norm": 0.031749793664709355, "learning_rate": 0.00022226305111525726, "loss": 0.004, "step": 5655 }, { "epoch": 2.6616470588235295, "grad_norm": 0.39783462177332424, "learning_rate": 0.00022222463409564595, "loss": 0.3408, "step": 5656 }, { "epoch": 2.6621176470588237, "grad_norm": 0.38546592793773393, "learning_rate": 0.0002221862109076966, "loss": 0.4006, "step": 5657 }, { "epoch": 2.662588235294118, "grad_norm": 0.10665008092876706, "learning_rate": 0.00022214778155469077, "loss": 0.0299, "step": 5658 }, { "epoch": 2.6630588235294117, "grad_norm": 0.1399344265377674, "learning_rate": 0.0002221093460399105, "loss": 0.0805, "step": 5659 }, { "epoch": 2.663529411764706, "grad_norm": 0.4492012624584326, "learning_rate": 0.0002220709043666384, "loss": 0.3165, "step": 5660 }, { "epoch": 2.664, "grad_norm": 0.24269318217402167, "learning_rate": 0.0002220324565381576, "loss": 0.0963, "step": 5661 }, { "epoch": 2.664470588235294, "grad_norm": 0.4944012654290118, "learning_rate": 0.0002219940025577516, "loss": 0.4226, "step": 5662 }, { "epoch": 2.664941176470588, "grad_norm": 0.16843303599726708, "learning_rate": 0.00022195554242870473, "loss": 0.056, "step": 5663 }, { "epoch": 2.6654117647058824, "grad_norm": 0.3799316753042287, "learning_rate": 0.0002219170761543016, "loss": 0.3631, "step": 5664 }, { "epoch": 2.6658823529411766, "grad_norm": 0.26454107324561366, "learning_rate": 0.00022187860373782737, "loss": 0.2263, "step": 5665 }, { "epoch": 2.666352941176471, "grad_norm": 0.43177567979084586, "learning_rate": 0.00022184012518256785, "loss": 0.4365, "step": 5666 }, { "epoch": 2.6668235294117646, "grad_norm": 0.32947594160278576, "learning_rate": 0.00022180164049180926, "loss": 0.3035, "step": 5667 }, { "epoch": 2.667294117647059, "grad_norm": 0.2813784318965471, "learning_rate": 0.00022176314966883837, "loss": 0.1856, "step": 5668 }, { "epoch": 2.667764705882353, "grad_norm": 0.21198215350624533, "learning_rate": 0.0002217246527169425, "loss": 0.0889, "step": 5669 }, { "epoch": 2.668235294117647, "grad_norm": 0.372236036261671, "learning_rate": 0.00022168614963940952, "loss": 0.3016, "step": 5670 }, { "epoch": 2.668705882352941, "grad_norm": 0.3063854924392309, "learning_rate": 0.00022164764043952772, "loss": 0.2137, "step": 5671 }, { "epoch": 2.6691764705882353, "grad_norm": 0.4276221285018988, "learning_rate": 0.00022160912512058598, "loss": 0.49, "step": 5672 }, { "epoch": 2.6696470588235295, "grad_norm": 0.14232775455523625, "learning_rate": 0.0002215706036858737, "loss": 0.0589, "step": 5673 }, { "epoch": 2.6701176470588237, "grad_norm": 0.23409452383926346, "learning_rate": 0.00022153207613868087, "loss": 0.214, "step": 5674 }, { "epoch": 2.6705882352941175, "grad_norm": 0.07806716706593927, "learning_rate": 0.00022149354248229784, "loss": 0.0322, "step": 5675 }, { "epoch": 2.6710588235294117, "grad_norm": 0.13912260988407435, "learning_rate": 0.00022145500272001558, "loss": 0.0505, "step": 5676 }, { "epoch": 2.671529411764706, "grad_norm": 0.2028079068287125, "learning_rate": 0.00022141645685512564, "loss": 0.113, "step": 5677 }, { "epoch": 2.672, "grad_norm": 0.34644280154761403, "learning_rate": 0.00022137790489091999, "loss": 0.3204, "step": 5678 }, { "epoch": 2.672470588235294, "grad_norm": 0.3731346256845982, "learning_rate": 0.00022133934683069115, "loss": 0.2342, "step": 5679 }, { "epoch": 2.672941176470588, "grad_norm": 0.4484604080082598, "learning_rate": 0.00022130078267773218, "loss": 0.2168, "step": 5680 }, { "epoch": 2.6734117647058824, "grad_norm": 0.3416881940698174, "learning_rate": 0.00022126221243533664, "loss": 0.3687, "step": 5681 }, { "epoch": 2.6738823529411766, "grad_norm": 0.24682757233967156, "learning_rate": 0.00022122363610679863, "loss": 0.1725, "step": 5682 }, { "epoch": 2.674352941176471, "grad_norm": 0.2325270077347151, "learning_rate": 0.0002211850536954127, "loss": 0.1724, "step": 5683 }, { "epoch": 2.6748235294117646, "grad_norm": 0.4017343804181566, "learning_rate": 0.00022114646520447406, "loss": 0.3276, "step": 5684 }, { "epoch": 2.675294117647059, "grad_norm": 0.18413296297507273, "learning_rate": 0.00022110787063727832, "loss": 0.1001, "step": 5685 }, { "epoch": 2.675764705882353, "grad_norm": 0.5132908122819836, "learning_rate": 0.00022106926999712165, "loss": 0.4507, "step": 5686 }, { "epoch": 2.676235294117647, "grad_norm": 0.5843801875259097, "learning_rate": 0.0002210306632873007, "loss": 0.4115, "step": 5687 }, { "epoch": 2.676705882352941, "grad_norm": 0.1745775431515521, "learning_rate": 0.00022099205051111272, "loss": 0.044, "step": 5688 }, { "epoch": 2.6771764705882353, "grad_norm": 0.3131230669049384, "learning_rate": 0.0002209534316718554, "loss": 0.1932, "step": 5689 }, { "epoch": 2.6776470588235295, "grad_norm": 0.2055239482970952, "learning_rate": 0.000220914806772827, "loss": 0.096, "step": 5690 }, { "epoch": 2.6781176470588237, "grad_norm": 0.19352234395652804, "learning_rate": 0.0002208761758173263, "loss": 0.1225, "step": 5691 }, { "epoch": 2.6785882352941175, "grad_norm": 0.7125853278358931, "learning_rate": 0.0002208375388086525, "loss": 0.5227, "step": 5692 }, { "epoch": 2.6790588235294117, "grad_norm": 0.23743201125783875, "learning_rate": 0.00022079889575010543, "loss": 0.1099, "step": 5693 }, { "epoch": 2.679529411764706, "grad_norm": 0.3021789951627655, "learning_rate": 0.0002207602466449854, "loss": 0.2579, "step": 5694 }, { "epoch": 2.68, "grad_norm": 0.3396060568521696, "learning_rate": 0.00022072159149659327, "loss": 0.2211, "step": 5695 }, { "epoch": 2.680470588235294, "grad_norm": 0.30590185762112876, "learning_rate": 0.00022068293030823026, "loss": 0.1547, "step": 5696 }, { "epoch": 2.680941176470588, "grad_norm": 0.2139908515998044, "learning_rate": 0.00022064426308319838, "loss": 0.1223, "step": 5697 }, { "epoch": 2.6814117647058824, "grad_norm": 0.31716544192279705, "learning_rate": 0.0002206055898247999, "loss": 0.2283, "step": 5698 }, { "epoch": 2.6818823529411766, "grad_norm": 0.2713359281025259, "learning_rate": 0.0002205669105363377, "loss": 0.1958, "step": 5699 }, { "epoch": 2.682352941176471, "grad_norm": 0.22412137394497372, "learning_rate": 0.00022052822522111522, "loss": 0.1495, "step": 5700 }, { "epoch": 2.6828235294117646, "grad_norm": 0.29688040953701744, "learning_rate": 0.0002204895338824364, "loss": 0.2059, "step": 5701 }, { "epoch": 2.683294117647059, "grad_norm": 0.510693771391505, "learning_rate": 0.0002204508365236056, "loss": 0.4108, "step": 5702 }, { "epoch": 2.683764705882353, "grad_norm": 0.20293853645326024, "learning_rate": 0.0002204121331479278, "loss": 0.0979, "step": 5703 }, { "epoch": 2.684235294117647, "grad_norm": 0.4662853819881928, "learning_rate": 0.0002203734237587085, "loss": 0.3886, "step": 5704 }, { "epoch": 2.684705882352941, "grad_norm": 0.29232970177862555, "learning_rate": 0.00022033470835925361, "loss": 0.2369, "step": 5705 }, { "epoch": 2.6851764705882353, "grad_norm": 0.32934121729347193, "learning_rate": 0.00022029598695286963, "loss": 0.1738, "step": 5706 }, { "epoch": 2.6856470588235295, "grad_norm": 0.16063665563115861, "learning_rate": 0.00022025725954286356, "loss": 0.0777, "step": 5707 }, { "epoch": 2.6861176470588237, "grad_norm": 0.5106249465481898, "learning_rate": 0.00022021852613254297, "loss": 0.5584, "step": 5708 }, { "epoch": 2.6865882352941175, "grad_norm": 0.38981571610211985, "learning_rate": 0.00022017978672521577, "loss": 0.3147, "step": 5709 }, { "epoch": 2.6870588235294117, "grad_norm": 0.5396568598410229, "learning_rate": 0.00022014104132419055, "loss": 0.679, "step": 5710 }, { "epoch": 2.687529411764706, "grad_norm": 0.35157118707065776, "learning_rate": 0.0002201022899327764, "loss": 0.3111, "step": 5711 }, { "epoch": 2.6879999999999997, "grad_norm": 0.35380110836861506, "learning_rate": 0.00022006353255428281, "loss": 0.2208, "step": 5712 }, { "epoch": 2.688470588235294, "grad_norm": 0.3549274024936489, "learning_rate": 0.00022002476919201987, "loss": 0.2757, "step": 5713 }, { "epoch": 2.688941176470588, "grad_norm": 0.20094935202698158, "learning_rate": 0.00021998599984929818, "loss": 0.1096, "step": 5714 }, { "epoch": 2.6894117647058824, "grad_norm": 0.36302123231708433, "learning_rate": 0.00021994722452942886, "loss": 0.2885, "step": 5715 }, { "epoch": 2.6898823529411766, "grad_norm": 0.3516614247376759, "learning_rate": 0.00021990844323572346, "loss": 0.3453, "step": 5716 }, { "epoch": 2.690352941176471, "grad_norm": 0.20721681459198812, "learning_rate": 0.0002198696559714941, "loss": 0.078, "step": 5717 }, { "epoch": 2.6908235294117646, "grad_norm": 0.3819619993589351, "learning_rate": 0.0002198308627400534, "loss": 0.2184, "step": 5718 }, { "epoch": 2.691294117647059, "grad_norm": 0.2551093082304713, "learning_rate": 0.00021979206354471453, "loss": 0.1436, "step": 5719 }, { "epoch": 2.691764705882353, "grad_norm": 0.5567513907815125, "learning_rate": 0.00021975325838879107, "loss": 0.4701, "step": 5720 }, { "epoch": 2.692235294117647, "grad_norm": 0.36265139176845496, "learning_rate": 0.0002197144472755972, "loss": 0.3033, "step": 5721 }, { "epoch": 2.692705882352941, "grad_norm": 0.23197498170841094, "learning_rate": 0.00021967563020844762, "loss": 0.1214, "step": 5722 }, { "epoch": 2.6931764705882353, "grad_norm": 0.22791950622140866, "learning_rate": 0.00021963680719065746, "loss": 0.0952, "step": 5723 }, { "epoch": 2.6936470588235295, "grad_norm": 0.2527085650004206, "learning_rate": 0.0002195979782255424, "loss": 0.2846, "step": 5724 }, { "epoch": 2.6941176470588237, "grad_norm": 0.3775462101883935, "learning_rate": 0.0002195591433164186, "loss": 0.2393, "step": 5725 }, { "epoch": 2.6945882352941175, "grad_norm": 0.37369935824808787, "learning_rate": 0.00021952030246660283, "loss": 0.2333, "step": 5726 }, { "epoch": 2.6950588235294117, "grad_norm": 0.5024256669635135, "learning_rate": 0.00021948145567941217, "loss": 0.3062, "step": 5727 }, { "epoch": 2.695529411764706, "grad_norm": 0.3541702748562639, "learning_rate": 0.0002194426029581644, "loss": 0.2187, "step": 5728 }, { "epoch": 2.6959999999999997, "grad_norm": 0.6936277541422503, "learning_rate": 0.00021940374430617777, "loss": 0.7161, "step": 5729 }, { "epoch": 2.696470588235294, "grad_norm": 0.23146990660038733, "learning_rate": 0.0002193648797267709, "loss": 0.1027, "step": 5730 }, { "epoch": 2.696941176470588, "grad_norm": 0.34911781542938086, "learning_rate": 0.0002193260092232631, "loss": 0.1953, "step": 5731 }, { "epoch": 2.6974117647058824, "grad_norm": 0.28331385445901675, "learning_rate": 0.00021928713279897408, "loss": 0.1474, "step": 5732 }, { "epoch": 2.6978823529411766, "grad_norm": 0.5387672673710855, "learning_rate": 0.00021924825045722404, "loss": 0.4509, "step": 5733 }, { "epoch": 2.698352941176471, "grad_norm": 0.4742743274027747, "learning_rate": 0.00021920936220133373, "loss": 0.3943, "step": 5734 }, { "epoch": 2.6988235294117646, "grad_norm": 0.23841404613170317, "learning_rate": 0.0002191704680346245, "loss": 0.1269, "step": 5735 }, { "epoch": 2.699294117647059, "grad_norm": 0.3438768722540836, "learning_rate": 0.00021913156796041794, "loss": 0.2083, "step": 5736 }, { "epoch": 2.699764705882353, "grad_norm": 0.4481181635732915, "learning_rate": 0.0002190926619820364, "loss": 0.282, "step": 5737 }, { "epoch": 2.700235294117647, "grad_norm": 0.29249983904514704, "learning_rate": 0.00021905375010280266, "loss": 0.1689, "step": 5738 }, { "epoch": 2.700705882352941, "grad_norm": 0.40548073129678297, "learning_rate": 0.00021901483232603995, "loss": 0.267, "step": 5739 }, { "epoch": 2.7011764705882353, "grad_norm": 0.24506593597163256, "learning_rate": 0.000218975908655072, "loss": 0.1406, "step": 5740 }, { "epoch": 2.7016470588235295, "grad_norm": 0.32239256226170687, "learning_rate": 0.0002189369790932232, "loss": 0.2694, "step": 5741 }, { "epoch": 2.7021176470588237, "grad_norm": 0.3527765393594347, "learning_rate": 0.0002188980436438182, "loss": 0.254, "step": 5742 }, { "epoch": 2.7025882352941175, "grad_norm": 0.492312221846654, "learning_rate": 0.0002188591023101823, "loss": 0.2434, "step": 5743 }, { "epoch": 2.7030588235294117, "grad_norm": 0.2800693094366365, "learning_rate": 0.00021882015509564132, "loss": 0.1969, "step": 5744 }, { "epoch": 2.703529411764706, "grad_norm": 0.2848522191923192, "learning_rate": 0.00021878120200352158, "loss": 0.1813, "step": 5745 }, { "epoch": 2.7039999999999997, "grad_norm": 0.18870985845453928, "learning_rate": 0.00021874224303714983, "loss": 0.124, "step": 5746 }, { "epoch": 2.704470588235294, "grad_norm": 0.5465892641065143, "learning_rate": 0.00021870327819985331, "loss": 0.5259, "step": 5747 }, { "epoch": 2.704941176470588, "grad_norm": 0.33531848489021177, "learning_rate": 0.00021866430749495987, "loss": 0.3052, "step": 5748 }, { "epoch": 2.7054117647058824, "grad_norm": 0.29362257793367147, "learning_rate": 0.00021862533092579774, "loss": 0.1756, "step": 5749 }, { "epoch": 2.7058823529411766, "grad_norm": 0.4058385858120252, "learning_rate": 0.00021858634849569576, "loss": 0.5036, "step": 5750 }, { "epoch": 2.706352941176471, "grad_norm": 0.5372650158493123, "learning_rate": 0.00021854736020798319, "loss": 0.5431, "step": 5751 }, { "epoch": 2.7068235294117646, "grad_norm": 0.41458849152947197, "learning_rate": 0.00021850836606598987, "loss": 0.4499, "step": 5752 }, { "epoch": 2.707294117647059, "grad_norm": 0.21773476278189077, "learning_rate": 0.00021846936607304603, "loss": 0.136, "step": 5753 }, { "epoch": 2.707764705882353, "grad_norm": 0.19056033521125335, "learning_rate": 0.00021843036023248247, "loss": 0.0821, "step": 5754 }, { "epoch": 2.708235294117647, "grad_norm": 0.3741541970913023, "learning_rate": 0.00021839134854763055, "loss": 0.3087, "step": 5755 }, { "epoch": 2.708705882352941, "grad_norm": 0.25367017186730806, "learning_rate": 0.00021835233102182199, "loss": 0.2044, "step": 5756 }, { "epoch": 2.7091764705882353, "grad_norm": 0.11007398735493922, "learning_rate": 0.00021831330765838903, "loss": 0.019, "step": 5757 }, { "epoch": 2.7096470588235295, "grad_norm": 0.35471743566008207, "learning_rate": 0.00021827427846066462, "loss": 0.3222, "step": 5758 }, { "epoch": 2.7101176470588237, "grad_norm": 0.34345459769747866, "learning_rate": 0.0002182352434319819, "loss": 0.1638, "step": 5759 }, { "epoch": 2.7105882352941175, "grad_norm": 0.23586778430895622, "learning_rate": 0.0002181962025756747, "loss": 0.1674, "step": 5760 }, { "epoch": 2.7110588235294117, "grad_norm": 0.2887854604858209, "learning_rate": 0.00021815715589507728, "loss": 0.1911, "step": 5761 }, { "epoch": 2.711529411764706, "grad_norm": 0.4998481803035955, "learning_rate": 0.00021811810339352446, "loss": 0.2285, "step": 5762 }, { "epoch": 2.7119999999999997, "grad_norm": 0.33200983928984246, "learning_rate": 0.0002180790450743515, "loss": 0.2416, "step": 5763 }, { "epoch": 2.712470588235294, "grad_norm": 0.17206953555723425, "learning_rate": 0.0002180399809408941, "loss": 0.0936, "step": 5764 }, { "epoch": 2.712941176470588, "grad_norm": 0.13746053645153447, "learning_rate": 0.00021800091099648865, "loss": 0.0538, "step": 5765 }, { "epoch": 2.7134117647058824, "grad_norm": 0.36717825432705864, "learning_rate": 0.0002179618352444719, "loss": 0.3427, "step": 5766 }, { "epoch": 2.7138823529411766, "grad_norm": 0.5373535404479869, "learning_rate": 0.00021792275368818102, "loss": 0.5362, "step": 5767 }, { "epoch": 2.714352941176471, "grad_norm": 0.19946217802722344, "learning_rate": 0.00021788366633095375, "loss": 0.0944, "step": 5768 }, { "epoch": 2.7148235294117646, "grad_norm": 0.3604634184772994, "learning_rate": 0.00021784457317612852, "loss": 0.3423, "step": 5769 }, { "epoch": 2.715294117647059, "grad_norm": 0.20308953308511532, "learning_rate": 0.0002178054742270439, "loss": 0.0948, "step": 5770 }, { "epoch": 2.715764705882353, "grad_norm": 0.4224682792515772, "learning_rate": 0.0002177663694870392, "loss": 0.5194, "step": 5771 }, { "epoch": 2.716235294117647, "grad_norm": 0.26697388139904155, "learning_rate": 0.00021772725895945422, "loss": 0.176, "step": 5772 }, { "epoch": 2.716705882352941, "grad_norm": 0.38048163310252725, "learning_rate": 0.0002176881426476291, "loss": 0.4782, "step": 5773 }, { "epoch": 2.7171764705882353, "grad_norm": 0.1710196994687181, "learning_rate": 0.00021764902055490456, "loss": 0.1123, "step": 5774 }, { "epoch": 2.7176470588235295, "grad_norm": 0.20373832530367503, "learning_rate": 0.00021760989268462182, "loss": 0.1221, "step": 5775 }, { "epoch": 2.7181176470588237, "grad_norm": 0.19203356276337705, "learning_rate": 0.00021757075904012266, "loss": 0.0944, "step": 5776 }, { "epoch": 2.7185882352941175, "grad_norm": 0.20965338487625793, "learning_rate": 0.00021753161962474924, "loss": 0.1388, "step": 5777 }, { "epoch": 2.7190588235294118, "grad_norm": 0.39496644851299134, "learning_rate": 0.00021749247444184422, "loss": 0.2873, "step": 5778 }, { "epoch": 2.719529411764706, "grad_norm": 0.48322264243410434, "learning_rate": 0.00021745332349475092, "loss": 0.3399, "step": 5779 }, { "epoch": 2.7199999999999998, "grad_norm": 0.40110804391910504, "learning_rate": 0.0002174141667868129, "loss": 0.3919, "step": 5780 }, { "epoch": 2.720470588235294, "grad_norm": 0.1886424948881129, "learning_rate": 0.00021737500432137435, "loss": 0.1299, "step": 5781 }, { "epoch": 2.720941176470588, "grad_norm": 0.42247423315533517, "learning_rate": 0.00021733583610177997, "loss": 0.4617, "step": 5782 }, { "epoch": 2.7214117647058824, "grad_norm": 0.4458196370799483, "learning_rate": 0.00021729666213137494, "loss": 0.3693, "step": 5783 }, { "epoch": 2.7218823529411766, "grad_norm": 0.22429655151142303, "learning_rate": 0.00021725748241350486, "loss": 0.0899, "step": 5784 }, { "epoch": 2.722352941176471, "grad_norm": 0.06465899547528393, "learning_rate": 0.00021721829695151585, "loss": 0.023, "step": 5785 }, { "epoch": 2.7228235294117646, "grad_norm": 0.8005085741994636, "learning_rate": 0.00021717910574875458, "loss": 0.5942, "step": 5786 }, { "epoch": 2.723294117647059, "grad_norm": 0.20677931923678533, "learning_rate": 0.00021713990880856818, "loss": 0.1177, "step": 5787 }, { "epoch": 2.723764705882353, "grad_norm": 0.334716123882785, "learning_rate": 0.0002171007061343043, "loss": 0.3901, "step": 5788 }, { "epoch": 2.724235294117647, "grad_norm": 0.4370889708169247, "learning_rate": 0.0002170614977293109, "loss": 0.3374, "step": 5789 }, { "epoch": 2.724705882352941, "grad_norm": 0.36102269020083083, "learning_rate": 0.00021702228359693672, "loss": 0.3521, "step": 5790 }, { "epoch": 2.7251764705882353, "grad_norm": 0.25786249575153225, "learning_rate": 0.00021698306374053075, "loss": 0.0935, "step": 5791 }, { "epoch": 2.7256470588235295, "grad_norm": 0.3019328890310067, "learning_rate": 0.00021694383816344258, "loss": 0.2093, "step": 5792 }, { "epoch": 2.7261176470588238, "grad_norm": 0.2565059951284732, "learning_rate": 0.0002169046068690223, "loss": 0.2546, "step": 5793 }, { "epoch": 2.7265882352941175, "grad_norm": 0.1655015464291859, "learning_rate": 0.00021686536986062043, "loss": 0.0517, "step": 5794 }, { "epoch": 2.7270588235294118, "grad_norm": 0.3504332578256942, "learning_rate": 0.000216826127141588, "loss": 0.3068, "step": 5795 }, { "epoch": 2.727529411764706, "grad_norm": 0.5688046729950524, "learning_rate": 0.0002167868787152765, "loss": 0.4007, "step": 5796 }, { "epoch": 2.7279999999999998, "grad_norm": 0.3460995897212271, "learning_rate": 0.000216747624585038, "loss": 0.2476, "step": 5797 }, { "epoch": 2.728470588235294, "grad_norm": 0.6358990154839018, "learning_rate": 0.00021670836475422498, "loss": 0.387, "step": 5798 }, { "epoch": 2.728941176470588, "grad_norm": 0.32687657804235204, "learning_rate": 0.00021666909922619037, "loss": 0.2614, "step": 5799 }, { "epoch": 2.7294117647058824, "grad_norm": 0.3393674223506057, "learning_rate": 0.0002166298280042877, "loss": 0.2943, "step": 5800 }, { "epoch": 2.7298823529411766, "grad_norm": 0.2748027889774491, "learning_rate": 0.00021659055109187088, "loss": 0.1875, "step": 5801 }, { "epoch": 2.7303529411764704, "grad_norm": 0.4267545485511188, "learning_rate": 0.0002165512684922944, "loss": 0.4825, "step": 5802 }, { "epoch": 2.7308235294117646, "grad_norm": 0.2692596054089221, "learning_rate": 0.0002165119802089131, "loss": 0.1613, "step": 5803 }, { "epoch": 2.731294117647059, "grad_norm": 0.4231917242881979, "learning_rate": 0.00021647268624508254, "loss": 0.4124, "step": 5804 }, { "epoch": 2.731764705882353, "grad_norm": 0.5175892704431547, "learning_rate": 0.00021643338660415844, "loss": 0.359, "step": 5805 }, { "epoch": 2.732235294117647, "grad_norm": 0.4702987780682605, "learning_rate": 0.0002163940812894973, "loss": 0.3941, "step": 5806 }, { "epoch": 2.732705882352941, "grad_norm": 0.30266434081492655, "learning_rate": 0.00021635477030445597, "loss": 0.2863, "step": 5807 }, { "epoch": 2.7331764705882353, "grad_norm": 0.27343010670990514, "learning_rate": 0.0002163154536523918, "loss": 0.2278, "step": 5808 }, { "epoch": 2.7336470588235295, "grad_norm": 0.2682329211813494, "learning_rate": 0.0002162761313366626, "loss": 0.1823, "step": 5809 }, { "epoch": 2.7341176470588238, "grad_norm": 0.04313822432652707, "learning_rate": 0.0002162368033606267, "loss": 0.0113, "step": 5810 }, { "epoch": 2.7345882352941175, "grad_norm": 0.15593316992428988, "learning_rate": 0.00021619746972764288, "loss": 0.1194, "step": 5811 }, { "epoch": 2.7350588235294118, "grad_norm": 0.6442485837991776, "learning_rate": 0.00021615813044107045, "loss": 0.4247, "step": 5812 }, { "epoch": 2.735529411764706, "grad_norm": 0.4191220693664444, "learning_rate": 0.00021611878550426922, "loss": 0.2578, "step": 5813 }, { "epoch": 2.7359999999999998, "grad_norm": 0.1727891331112452, "learning_rate": 0.00021607943492059936, "loss": 0.085, "step": 5814 }, { "epoch": 2.736470588235294, "grad_norm": 0.04790711772050509, "learning_rate": 0.00021604007869342168, "loss": 0.0133, "step": 5815 }, { "epoch": 2.736941176470588, "grad_norm": 0.23829755466200167, "learning_rate": 0.00021600071682609733, "loss": 0.145, "step": 5816 }, { "epoch": 2.7374117647058824, "grad_norm": 0.24647639612973687, "learning_rate": 0.00021596134932198803, "loss": 0.1199, "step": 5817 }, { "epoch": 2.7378823529411767, "grad_norm": 0.7470115354057779, "learning_rate": 0.00021592197618445596, "loss": 0.4518, "step": 5818 }, { "epoch": 2.7383529411764704, "grad_norm": 0.2728351423328962, "learning_rate": 0.00021588259741686373, "loss": 0.1503, "step": 5819 }, { "epoch": 2.7388235294117647, "grad_norm": 0.4541538743648188, "learning_rate": 0.00021584321302257464, "loss": 0.3652, "step": 5820 }, { "epoch": 2.739294117647059, "grad_norm": 0.24709127325839358, "learning_rate": 0.00021580382300495215, "loss": 0.1716, "step": 5821 }, { "epoch": 2.739764705882353, "grad_norm": 0.3283460163437219, "learning_rate": 0.00021576442736736032, "loss": 0.2728, "step": 5822 }, { "epoch": 2.740235294117647, "grad_norm": 0.20349693581900138, "learning_rate": 0.00021572502611316393, "loss": 0.1057, "step": 5823 }, { "epoch": 2.740705882352941, "grad_norm": 0.16843390637481245, "learning_rate": 0.0002156856192457279, "loss": 0.0719, "step": 5824 }, { "epoch": 2.7411764705882353, "grad_norm": 0.333718375947939, "learning_rate": 0.00021564620676841782, "loss": 0.2689, "step": 5825 }, { "epoch": 2.7416470588235295, "grad_norm": 0.16388181812271088, "learning_rate": 0.0002156067886845996, "loss": 0.0627, "step": 5826 }, { "epoch": 2.7421176470588238, "grad_norm": 0.4613236140744269, "learning_rate": 0.00021556736499763992, "loss": 0.3519, "step": 5827 }, { "epoch": 2.7425882352941175, "grad_norm": 0.5992379400992668, "learning_rate": 0.00021552793571090568, "loss": 0.4777, "step": 5828 }, { "epoch": 2.7430588235294118, "grad_norm": 0.38890148676338226, "learning_rate": 0.0002154885008277642, "loss": 0.3376, "step": 5829 }, { "epoch": 2.743529411764706, "grad_norm": 0.07478105510267502, "learning_rate": 0.0002154490603515836, "loss": 0.0444, "step": 5830 }, { "epoch": 2.7439999999999998, "grad_norm": 0.18298080232044026, "learning_rate": 0.0002154096142857323, "loss": 0.0835, "step": 5831 }, { "epoch": 2.744470588235294, "grad_norm": 0.1496333624083958, "learning_rate": 0.000215370162633579, "loss": 0.0747, "step": 5832 }, { "epoch": 2.744941176470588, "grad_norm": 0.3312795336004489, "learning_rate": 0.00021533070539849316, "loss": 0.1931, "step": 5833 }, { "epoch": 2.7454117647058824, "grad_norm": 0.18987136745872127, "learning_rate": 0.00021529124258384466, "loss": 0.0734, "step": 5834 }, { "epoch": 2.7458823529411767, "grad_norm": 0.4728534337762171, "learning_rate": 0.00021525177419300383, "loss": 0.2922, "step": 5835 }, { "epoch": 2.7463529411764704, "grad_norm": 0.2123555698018331, "learning_rate": 0.0002152123002293414, "loss": 0.0839, "step": 5836 }, { "epoch": 2.7468235294117647, "grad_norm": 0.5955294231204998, "learning_rate": 0.0002151728206962287, "loss": 0.3431, "step": 5837 }, { "epoch": 2.747294117647059, "grad_norm": 0.4511301344232949, "learning_rate": 0.00021513333559703738, "loss": 0.331, "step": 5838 }, { "epoch": 2.7477647058823527, "grad_norm": 0.7948816345319354, "learning_rate": 0.00021509384493513978, "loss": 0.5453, "step": 5839 }, { "epoch": 2.748235294117647, "grad_norm": 0.26935591699929134, "learning_rate": 0.00021505434871390848, "loss": 0.1285, "step": 5840 }, { "epoch": 2.748705882352941, "grad_norm": 0.2770695998942719, "learning_rate": 0.00021501484693671677, "loss": 0.146, "step": 5841 }, { "epoch": 2.7491764705882353, "grad_norm": 0.277664069237073, "learning_rate": 0.00021497533960693823, "loss": 0.2659, "step": 5842 }, { "epoch": 2.7496470588235296, "grad_norm": 0.17233148348110328, "learning_rate": 0.00021493582672794693, "loss": 0.0821, "step": 5843 }, { "epoch": 2.7501176470588238, "grad_norm": 0.2080050924750174, "learning_rate": 0.00021489630830311758, "loss": 0.0987, "step": 5844 }, { "epoch": 2.7505882352941176, "grad_norm": 0.5264959922972743, "learning_rate": 0.00021485678433582516, "loss": 0.4616, "step": 5845 }, { "epoch": 2.751058823529412, "grad_norm": 0.517038576126558, "learning_rate": 0.00021481725482944518, "loss": 0.3325, "step": 5846 }, { "epoch": 2.751529411764706, "grad_norm": 0.2571003970127026, "learning_rate": 0.0002147777197873538, "loss": 0.1037, "step": 5847 }, { "epoch": 2.752, "grad_norm": 0.5303644562574256, "learning_rate": 0.0002147381792129274, "loss": 0.5236, "step": 5848 }, { "epoch": 2.752470588235294, "grad_norm": 0.5576243886327971, "learning_rate": 0.00021469863310954296, "loss": 0.4727, "step": 5849 }, { "epoch": 2.7529411764705882, "grad_norm": 0.3890304970783495, "learning_rate": 0.00021465908148057787, "loss": 0.3654, "step": 5850 }, { "epoch": 2.7534117647058824, "grad_norm": 0.3231324694980351, "learning_rate": 0.00021461952432941006, "loss": 0.3102, "step": 5851 }, { "epoch": 2.7538823529411767, "grad_norm": 0.7520063627519494, "learning_rate": 0.00021457996165941798, "loss": 0.7762, "step": 5852 }, { "epoch": 2.7543529411764704, "grad_norm": 8.142771559484862, "learning_rate": 0.00021454039347398033, "loss": 3.1375, "step": 5853 }, { "epoch": 2.7548235294117647, "grad_norm": 26.077480924522472, "learning_rate": 0.00021450081977647658, "loss": 2.315, "step": 5854 }, { "epoch": 2.755294117647059, "grad_norm": 0.29919854233897847, "learning_rate": 0.00021446124057028642, "loss": 0.3548, "step": 5855 }, { "epoch": 2.7557647058823527, "grad_norm": 0.5123831424212947, "learning_rate": 0.00021442165585879014, "loss": 0.4188, "step": 5856 }, { "epoch": 2.756235294117647, "grad_norm": 0.2092556855542328, "learning_rate": 0.00021438206564536845, "loss": 0.1572, "step": 5857 }, { "epoch": 2.756705882352941, "grad_norm": 0.45712334713430985, "learning_rate": 0.00021434246993340256, "loss": 0.0763, "step": 5858 }, { "epoch": 2.7571764705882353, "grad_norm": 2.3007061254683685, "learning_rate": 0.00021430286872627417, "loss": 0.2427, "step": 5859 }, { "epoch": 2.7576470588235296, "grad_norm": 6.197467748951957, "learning_rate": 0.00021426326202736532, "loss": 0.6296, "step": 5860 }, { "epoch": 2.758117647058824, "grad_norm": 1.6474629526470135, "learning_rate": 0.00021422364984005876, "loss": 0.3797, "step": 5861 }, { "epoch": 2.7585882352941176, "grad_norm": 0.21128884976186113, "learning_rate": 0.00021418403216773746, "loss": 0.0673, "step": 5862 }, { "epoch": 2.759058823529412, "grad_norm": 0.8537360790655721, "learning_rate": 0.000214144409013785, "loss": 0.3738, "step": 5863 }, { "epoch": 2.759529411764706, "grad_norm": 0.7901517928854892, "learning_rate": 0.00021410478038158537, "loss": 0.3312, "step": 5864 }, { "epoch": 2.76, "grad_norm": 0.6387878797229121, "learning_rate": 0.00021406514627452306, "loss": 0.2508, "step": 5865 }, { "epoch": 2.760470588235294, "grad_norm": 0.2699219469517985, "learning_rate": 0.00021402550669598303, "loss": 0.0986, "step": 5866 }, { "epoch": 2.7609411764705882, "grad_norm": 0.3386906667162184, "learning_rate": 0.00021398586164935067, "loss": 0.1469, "step": 5867 }, { "epoch": 2.7614117647058825, "grad_norm": 0.9840792537561909, "learning_rate": 0.0002139462111380119, "loss": 0.5219, "step": 5868 }, { "epoch": 2.7618823529411767, "grad_norm": 0.6707686282980242, "learning_rate": 0.00021390655516535304, "loss": 0.3544, "step": 5869 }, { "epoch": 2.7623529411764705, "grad_norm": 0.2911871105163618, "learning_rate": 0.00021386689373476087, "loss": 0.2272, "step": 5870 }, { "epoch": 2.7628235294117647, "grad_norm": 0.12350834340741261, "learning_rate": 0.00021382722684962276, "loss": 0.04, "step": 5871 }, { "epoch": 2.763294117647059, "grad_norm": 0.33366755377308543, "learning_rate": 0.00021378755451332637, "loss": 0.1057, "step": 5872 }, { "epoch": 2.7637647058823527, "grad_norm": 1.3031292824403937, "learning_rate": 0.00021374787672926, "loss": 0.5671, "step": 5873 }, { "epoch": 2.764235294117647, "grad_norm": 1.228031507035687, "learning_rate": 0.00021370819350081222, "loss": 0.39, "step": 5874 }, { "epoch": 2.764705882352941, "grad_norm": 0.29552330198608295, "learning_rate": 0.00021366850483137226, "loss": 0.2298, "step": 5875 }, { "epoch": 2.7651764705882353, "grad_norm": 0.148426229147104, "learning_rate": 0.00021362881072432965, "loss": 0.0658, "step": 5876 }, { "epoch": 2.7656470588235296, "grad_norm": 0.7673693120508834, "learning_rate": 0.00021358911118307455, "loss": 0.663, "step": 5877 }, { "epoch": 2.766117647058824, "grad_norm": 0.2829347090308931, "learning_rate": 0.00021354940621099745, "loss": 0.2443, "step": 5878 }, { "epoch": 2.7665882352941176, "grad_norm": 0.26236176223271257, "learning_rate": 0.00021350969581148934, "loss": 0.1328, "step": 5879 }, { "epoch": 2.767058823529412, "grad_norm": 0.3680225378479445, "learning_rate": 0.00021346997998794172, "loss": 0.2694, "step": 5880 }, { "epoch": 2.767529411764706, "grad_norm": 0.36574508999803995, "learning_rate": 0.00021343025874374645, "loss": 0.3091, "step": 5881 }, { "epoch": 2.768, "grad_norm": 0.7982513062425479, "learning_rate": 0.000213390532082296, "loss": 0.3314, "step": 5882 }, { "epoch": 2.768470588235294, "grad_norm": 0.695997266328832, "learning_rate": 0.00021335080000698317, "loss": 0.3596, "step": 5883 }, { "epoch": 2.7689411764705882, "grad_norm": 0.20270355810652846, "learning_rate": 0.00021331106252120126, "loss": 0.1146, "step": 5884 }, { "epoch": 2.7694117647058825, "grad_norm": 0.33817302737200644, "learning_rate": 0.00021327131962834407, "loss": 0.1439, "step": 5885 }, { "epoch": 2.7698823529411767, "grad_norm": 0.4729492414509164, "learning_rate": 0.00021323157133180587, "loss": 0.2999, "step": 5886 }, { "epoch": 2.7703529411764705, "grad_norm": 0.5837144310312454, "learning_rate": 0.0002131918176349813, "loss": 0.3219, "step": 5887 }, { "epoch": 2.7708235294117647, "grad_norm": 0.3379398210462244, "learning_rate": 0.0002131520585412656, "loss": 0.2387, "step": 5888 }, { "epoch": 2.771294117647059, "grad_norm": 0.06507301564408226, "learning_rate": 0.0002131122940540543, "loss": 0.016, "step": 5889 }, { "epoch": 2.7717647058823527, "grad_norm": 0.25374358920662765, "learning_rate": 0.00021307252417674356, "loss": 0.2191, "step": 5890 }, { "epoch": 2.772235294117647, "grad_norm": 0.6411450420367756, "learning_rate": 0.0002130327489127298, "loss": 0.6973, "step": 5891 }, { "epoch": 2.772705882352941, "grad_norm": 0.6120629286292051, "learning_rate": 0.0002129929682654102, "loss": 0.4867, "step": 5892 }, { "epoch": 2.7731764705882354, "grad_norm": 0.32840994359164893, "learning_rate": 0.0002129531822381821, "loss": 0.3105, "step": 5893 }, { "epoch": 2.7736470588235296, "grad_norm": 0.26648213441516033, "learning_rate": 0.00021291339083444344, "loss": 0.1415, "step": 5894 }, { "epoch": 2.774117647058824, "grad_norm": 0.3047874876844381, "learning_rate": 0.00021287359405759265, "loss": 0.1599, "step": 5895 }, { "epoch": 2.7745882352941176, "grad_norm": 0.34910491186642345, "learning_rate": 0.0002128337919110285, "loss": 0.2332, "step": 5896 }, { "epoch": 2.775058823529412, "grad_norm": 0.4543338185758962, "learning_rate": 0.00021279398439815034, "loss": 0.3742, "step": 5897 }, { "epoch": 2.775529411764706, "grad_norm": 0.3542750723652182, "learning_rate": 0.00021275417152235787, "loss": 0.3015, "step": 5898 }, { "epoch": 2.776, "grad_norm": 0.630165168260436, "learning_rate": 0.00021271435328705138, "loss": 0.6075, "step": 5899 }, { "epoch": 2.776470588235294, "grad_norm": 0.477703130372392, "learning_rate": 0.00021267452969563153, "loss": 0.3384, "step": 5900 }, { "epoch": 2.7769411764705882, "grad_norm": 0.5420791600996914, "learning_rate": 0.00021263470075149937, "loss": 0.2685, "step": 5901 }, { "epoch": 2.7774117647058825, "grad_norm": 0.2570032865913702, "learning_rate": 0.00021259486645805662, "loss": 0.127, "step": 5902 }, { "epoch": 2.7778823529411767, "grad_norm": 0.2710414232575608, "learning_rate": 0.00021255502681870518, "loss": 0.1407, "step": 5903 }, { "epoch": 2.7783529411764705, "grad_norm": 0.25465407211015884, "learning_rate": 0.00021251518183684767, "loss": 0.2493, "step": 5904 }, { "epoch": 2.7788235294117647, "grad_norm": 0.3961894809069048, "learning_rate": 0.00021247533151588693, "loss": 0.3634, "step": 5905 }, { "epoch": 2.779294117647059, "grad_norm": 0.21082666615594878, "learning_rate": 0.0002124354758592265, "loss": 0.0946, "step": 5906 }, { "epoch": 2.7797647058823527, "grad_norm": 0.6855272612569203, "learning_rate": 0.00021239561487027017, "loss": 0.4791, "step": 5907 }, { "epoch": 2.780235294117647, "grad_norm": 0.3392373404998002, "learning_rate": 0.00021235574855242226, "loss": 0.3441, "step": 5908 }, { "epoch": 2.780705882352941, "grad_norm": 0.6536190916836669, "learning_rate": 0.00021231587690908758, "loss": 0.6397, "step": 5909 }, { "epoch": 2.7811764705882354, "grad_norm": 0.3265316666132328, "learning_rate": 0.00021227599994367137, "loss": 0.258, "step": 5910 }, { "epoch": 2.7816470588235296, "grad_norm": 0.3539266384447113, "learning_rate": 0.0002122361176595793, "loss": 0.213, "step": 5911 }, { "epoch": 2.782117647058824, "grad_norm": 0.2639826805723604, "learning_rate": 0.0002121962300602175, "loss": 0.1437, "step": 5912 }, { "epoch": 2.7825882352941176, "grad_norm": 0.7093939295961011, "learning_rate": 0.00021215633714899264, "loss": 0.576, "step": 5913 }, { "epoch": 2.783058823529412, "grad_norm": 0.6778561335866826, "learning_rate": 0.00021211643892931167, "loss": 0.5425, "step": 5914 }, { "epoch": 2.783529411764706, "grad_norm": 0.5644238616130944, "learning_rate": 0.00021207653540458209, "loss": 0.5162, "step": 5915 }, { "epoch": 2.784, "grad_norm": 0.354235082670806, "learning_rate": 0.00021203662657821198, "loss": 0.2068, "step": 5916 }, { "epoch": 2.784470588235294, "grad_norm": 0.6377716633231915, "learning_rate": 0.00021199671245360964, "loss": 0.6174, "step": 5917 }, { "epoch": 2.7849411764705883, "grad_norm": 0.3036685772994696, "learning_rate": 0.0002119567930341839, "loss": 0.1066, "step": 5918 }, { "epoch": 2.7854117647058825, "grad_norm": 0.2962376686465532, "learning_rate": 0.00021191686832334423, "loss": 0.195, "step": 5919 }, { "epoch": 2.7858823529411767, "grad_norm": 0.47036027580141676, "learning_rate": 0.0002118769383245003, "loss": 0.4319, "step": 5920 }, { "epoch": 2.7863529411764705, "grad_norm": 0.4384866111390983, "learning_rate": 0.0002118370030410623, "loss": 0.3829, "step": 5921 }, { "epoch": 2.7868235294117647, "grad_norm": 0.28733309145542785, "learning_rate": 0.0002117970624764409, "loss": 0.2035, "step": 5922 }, { "epoch": 2.787294117647059, "grad_norm": 0.1433919157601082, "learning_rate": 0.00021175711663404728, "loss": 0.0758, "step": 5923 }, { "epoch": 2.7877647058823527, "grad_norm": 0.7331249642036088, "learning_rate": 0.000211717165517293, "loss": 0.541, "step": 5924 }, { "epoch": 2.788235294117647, "grad_norm": 0.3752624328463542, "learning_rate": 0.00021167720912959004, "loss": 0.3543, "step": 5925 }, { "epoch": 2.788705882352941, "grad_norm": 0.2994167901385454, "learning_rate": 0.00021163724747435091, "loss": 0.2743, "step": 5926 }, { "epoch": 2.7891764705882354, "grad_norm": 0.12458506107128271, "learning_rate": 0.00021159728055498856, "loss": 0.0498, "step": 5927 }, { "epoch": 2.7896470588235296, "grad_norm": 0.401046392641192, "learning_rate": 0.00021155730837491625, "loss": 0.3485, "step": 5928 }, { "epoch": 2.7901176470588234, "grad_norm": 0.10271145939547102, "learning_rate": 0.0002115173309375479, "loss": 0.0453, "step": 5929 }, { "epoch": 2.7905882352941176, "grad_norm": 0.46226934586657586, "learning_rate": 0.00021147734824629772, "loss": 0.2831, "step": 5930 }, { "epoch": 2.791058823529412, "grad_norm": 0.34212227045119165, "learning_rate": 0.0002114373603045805, "loss": 0.236, "step": 5931 }, { "epoch": 2.791529411764706, "grad_norm": 0.3026620325454854, "learning_rate": 0.00021139736711581132, "loss": 0.2288, "step": 5932 }, { "epoch": 2.792, "grad_norm": 0.4259230343142024, "learning_rate": 0.00021135736868340587, "loss": 0.3431, "step": 5933 }, { "epoch": 2.792470588235294, "grad_norm": 0.5431689893644581, "learning_rate": 0.00021131736501078018, "loss": 0.6684, "step": 5934 }, { "epoch": 2.7929411764705883, "grad_norm": 0.35913730862406296, "learning_rate": 0.0002112773561013507, "loss": 0.3162, "step": 5935 }, { "epoch": 2.7934117647058825, "grad_norm": 0.2517139724568644, "learning_rate": 0.0002112373419585345, "loss": 0.2533, "step": 5936 }, { "epoch": 2.7938823529411767, "grad_norm": 0.3840941308768681, "learning_rate": 0.00021119732258574893, "loss": 0.2464, "step": 5937 }, { "epoch": 2.7943529411764705, "grad_norm": 0.3845851338278745, "learning_rate": 0.0002111572979864118, "loss": 0.3219, "step": 5938 }, { "epoch": 2.7948235294117647, "grad_norm": 0.3908740999074167, "learning_rate": 0.00021111726816394147, "loss": 0.3693, "step": 5939 }, { "epoch": 2.795294117647059, "grad_norm": 0.2691210280312973, "learning_rate": 0.00021107723312175665, "loss": 0.087, "step": 5940 }, { "epoch": 2.7957647058823527, "grad_norm": 0.09369056650495941, "learning_rate": 0.00021103719286327658, "loss": 0.0337, "step": 5941 }, { "epoch": 2.796235294117647, "grad_norm": 0.38433749174912724, "learning_rate": 0.0002109971473919208, "loss": 0.2816, "step": 5942 }, { "epoch": 2.796705882352941, "grad_norm": 0.35827165152304546, "learning_rate": 0.00021095709671110942, "loss": 0.236, "step": 5943 }, { "epoch": 2.7971764705882354, "grad_norm": 0.24124468867360888, "learning_rate": 0.00021091704082426308, "loss": 0.1291, "step": 5944 }, { "epoch": 2.7976470588235296, "grad_norm": 0.4644811704634537, "learning_rate": 0.0002108769797348026, "loss": 0.3791, "step": 5945 }, { "epoch": 2.7981176470588234, "grad_norm": 0.3293479086290006, "learning_rate": 0.00021083691344614947, "loss": 0.0871, "step": 5946 }, { "epoch": 2.7985882352941176, "grad_norm": 0.5560399806626231, "learning_rate": 0.00021079684196172554, "loss": 0.3319, "step": 5947 }, { "epoch": 2.799058823529412, "grad_norm": 0.39427223313668974, "learning_rate": 0.00021075676528495307, "loss": 0.2715, "step": 5948 }, { "epoch": 2.799529411764706, "grad_norm": 0.3069791792079133, "learning_rate": 0.00021071668341925482, "loss": 0.1374, "step": 5949 }, { "epoch": 2.8, "grad_norm": 0.11063372779943381, "learning_rate": 0.00021067659636805403, "loss": 0.0362, "step": 5950 }, { "epoch": 2.800470588235294, "grad_norm": 0.19893426029909556, "learning_rate": 0.00021063650413477428, "loss": 0.1005, "step": 5951 }, { "epoch": 2.8009411764705883, "grad_norm": 0.45217597248005614, "learning_rate": 0.00021059640672283964, "loss": 0.2425, "step": 5952 }, { "epoch": 2.8014117647058825, "grad_norm": 0.44466951447212577, "learning_rate": 0.0002105563041356747, "loss": 0.3232, "step": 5953 }, { "epoch": 2.8018823529411767, "grad_norm": 0.4599177915248965, "learning_rate": 0.00021051619637670431, "loss": 0.2854, "step": 5954 }, { "epoch": 2.8023529411764705, "grad_norm": 0.1879125110700302, "learning_rate": 0.00021047608344935395, "loss": 0.1221, "step": 5955 }, { "epoch": 2.8028235294117647, "grad_norm": 0.3841085339443097, "learning_rate": 0.0002104359653570494, "loss": 0.2912, "step": 5956 }, { "epoch": 2.803294117647059, "grad_norm": 0.4652396534428968, "learning_rate": 0.00021039584210321699, "loss": 0.5253, "step": 5957 }, { "epoch": 2.8037647058823527, "grad_norm": 0.3973029480681908, "learning_rate": 0.00021035571369128341, "loss": 0.3377, "step": 5958 }, { "epoch": 2.804235294117647, "grad_norm": 0.6005866107243258, "learning_rate": 0.00021031558012467582, "loss": 0.6718, "step": 5959 }, { "epoch": 2.804705882352941, "grad_norm": 0.3542247013624326, "learning_rate": 0.00021027544140682188, "loss": 0.3117, "step": 5960 }, { "epoch": 2.8051764705882354, "grad_norm": 0.3893453975019429, "learning_rate": 0.00021023529754114956, "loss": 0.2478, "step": 5961 }, { "epoch": 2.8056470588235296, "grad_norm": 0.21636543964086313, "learning_rate": 0.0002101951485310874, "loss": 0.1195, "step": 5962 }, { "epoch": 2.8061176470588234, "grad_norm": 0.39966868904072456, "learning_rate": 0.0002101549943800643, "loss": 0.2729, "step": 5963 }, { "epoch": 2.8065882352941176, "grad_norm": 0.46567542041301846, "learning_rate": 0.00021011483509150963, "loss": 0.3949, "step": 5964 }, { "epoch": 2.807058823529412, "grad_norm": 0.5080418517595742, "learning_rate": 0.00021007467066885314, "loss": 0.4587, "step": 5965 }, { "epoch": 2.8075294117647056, "grad_norm": 0.27811460672867677, "learning_rate": 0.00021003450111552514, "loss": 0.2501, "step": 5966 }, { "epoch": 2.808, "grad_norm": 0.26747532250460715, "learning_rate": 0.00020999432643495626, "loss": 0.1217, "step": 5967 }, { "epoch": 2.808470588235294, "grad_norm": 0.2333933962492308, "learning_rate": 0.00020995414663057768, "loss": 0.105, "step": 5968 }, { "epoch": 2.8089411764705883, "grad_norm": 0.761687467228625, "learning_rate": 0.00020991396170582088, "loss": 0.4424, "step": 5969 }, { "epoch": 2.8094117647058825, "grad_norm": 0.4181439889354299, "learning_rate": 0.00020987377166411785, "loss": 0.4165, "step": 5970 }, { "epoch": 2.8098823529411767, "grad_norm": 0.3869553836143336, "learning_rate": 0.00020983357650890106, "loss": 0.3526, "step": 5971 }, { "epoch": 2.8103529411764705, "grad_norm": 0.5121611387099384, "learning_rate": 0.00020979337624360336, "loss": 0.4907, "step": 5972 }, { "epoch": 2.8108235294117647, "grad_norm": 0.3818639183954832, "learning_rate": 0.000209753170871658, "loss": 0.3359, "step": 5973 }, { "epoch": 2.811294117647059, "grad_norm": 0.12841814358061188, "learning_rate": 0.00020971296039649882, "loss": 0.0543, "step": 5974 }, { "epoch": 2.8117647058823527, "grad_norm": 0.33359576755182657, "learning_rate": 0.00020967274482155992, "loss": 0.2685, "step": 5975 }, { "epoch": 2.812235294117647, "grad_norm": 0.2470714790985088, "learning_rate": 0.00020963252415027588, "loss": 0.142, "step": 5976 }, { "epoch": 2.812705882352941, "grad_norm": 0.4850631865312818, "learning_rate": 0.0002095922983860818, "loss": 0.3153, "step": 5977 }, { "epoch": 2.8131764705882354, "grad_norm": 0.38851420626755695, "learning_rate": 0.00020955206753241315, "loss": 0.2389, "step": 5978 }, { "epoch": 2.8136470588235296, "grad_norm": 0.49690651353412923, "learning_rate": 0.00020951183159270586, "loss": 0.2853, "step": 5979 }, { "epoch": 2.8141176470588234, "grad_norm": 0.6374744017319008, "learning_rate": 0.0002094715905703962, "loss": 0.4033, "step": 5980 }, { "epoch": 2.8145882352941176, "grad_norm": 0.4454809412955944, "learning_rate": 0.00020943134446892102, "loss": 0.2309, "step": 5981 }, { "epoch": 2.815058823529412, "grad_norm": 0.2981208941290238, "learning_rate": 0.00020939109329171752, "loss": 0.2587, "step": 5982 }, { "epoch": 2.8155294117647056, "grad_norm": 0.25563141094697484, "learning_rate": 0.00020935083704222331, "loss": 0.1086, "step": 5983 }, { "epoch": 2.816, "grad_norm": 0.22827622623260593, "learning_rate": 0.00020931057572387655, "loss": 0.1077, "step": 5984 }, { "epoch": 2.816470588235294, "grad_norm": 0.37208470435834157, "learning_rate": 0.00020927030934011568, "loss": 0.2217, "step": 5985 }, { "epoch": 2.8169411764705883, "grad_norm": 0.14400173522291482, "learning_rate": 0.00020923003789437967, "loss": 0.0599, "step": 5986 }, { "epoch": 2.8174117647058825, "grad_norm": 0.7730414949078586, "learning_rate": 0.00020918976139010784, "loss": 0.4141, "step": 5987 }, { "epoch": 2.8178823529411767, "grad_norm": 0.45638335845880756, "learning_rate": 0.00020914947983074016, "loss": 0.3413, "step": 5988 }, { "epoch": 2.8183529411764705, "grad_norm": 0.6425863919369009, "learning_rate": 0.00020910919321971675, "loss": 0.5417, "step": 5989 }, { "epoch": 2.8188235294117647, "grad_norm": 0.3964945683253239, "learning_rate": 0.00020906890156047826, "loss": 0.2525, "step": 5990 }, { "epoch": 2.819294117647059, "grad_norm": 0.4004954481980305, "learning_rate": 0.0002090286048564659, "loss": 0.2219, "step": 5991 }, { "epoch": 2.8197647058823527, "grad_norm": 0.23675797029115195, "learning_rate": 0.00020898830311112118, "loss": 0.2118, "step": 5992 }, { "epoch": 2.820235294117647, "grad_norm": 0.3120983948825851, "learning_rate": 0.00020894799632788596, "loss": 0.2203, "step": 5993 }, { "epoch": 2.820705882352941, "grad_norm": 0.4964959396509231, "learning_rate": 0.0002089076845102028, "loss": 0.4801, "step": 5994 }, { "epoch": 2.8211764705882354, "grad_norm": 0.4524291766920233, "learning_rate": 0.00020886736766151438, "loss": 0.1872, "step": 5995 }, { "epoch": 2.8216470588235296, "grad_norm": 0.3575465474957256, "learning_rate": 0.00020882704578526406, "loss": 0.2745, "step": 5996 }, { "epoch": 2.8221176470588234, "grad_norm": 0.6698725877273217, "learning_rate": 0.0002087867188848954, "loss": 0.4926, "step": 5997 }, { "epoch": 2.8225882352941176, "grad_norm": 0.3670518018281137, "learning_rate": 0.0002087463869638527, "loss": 0.3083, "step": 5998 }, { "epoch": 2.823058823529412, "grad_norm": 0.20713082538961258, "learning_rate": 0.00020870605002558038, "loss": 0.1067, "step": 5999 }, { "epoch": 2.8235294117647056, "grad_norm": 0.32663956006236977, "learning_rate": 0.00020866570807352337, "loss": 0.3061, "step": 6000 }, { "epoch": 2.824, "grad_norm": 0.3653452378319818, "learning_rate": 0.0002086253611111272, "loss": 0.3507, "step": 6001 }, { "epoch": 2.824470588235294, "grad_norm": 0.2613734488689672, "learning_rate": 0.00020858500914183763, "loss": 0.1284, "step": 6002 }, { "epoch": 2.8249411764705883, "grad_norm": 0.2737547161785731, "learning_rate": 0.0002085446521691009, "loss": 0.1607, "step": 6003 }, { "epoch": 2.8254117647058825, "grad_norm": 0.3118178613879124, "learning_rate": 0.00020850429019636372, "loss": 0.2584, "step": 6004 }, { "epoch": 2.8258823529411767, "grad_norm": 0.20044942492267387, "learning_rate": 0.00020846392322707323, "loss": 0.1211, "step": 6005 }, { "epoch": 2.8263529411764705, "grad_norm": 0.30857076776451225, "learning_rate": 0.00020842355126467693, "loss": 0.188, "step": 6006 }, { "epoch": 2.8268235294117647, "grad_norm": 0.2979993450771898, "learning_rate": 0.00020838317431262274, "loss": 0.2101, "step": 6007 }, { "epoch": 2.827294117647059, "grad_norm": 0.5057490802974606, "learning_rate": 0.0002083427923743591, "loss": 0.5938, "step": 6008 }, { "epoch": 2.8277647058823527, "grad_norm": 0.2808898668259971, "learning_rate": 0.00020830240545333489, "loss": 0.1826, "step": 6009 }, { "epoch": 2.828235294117647, "grad_norm": 0.3427456357618049, "learning_rate": 0.00020826201355299918, "loss": 0.2201, "step": 6010 }, { "epoch": 2.828705882352941, "grad_norm": 0.3985947605390751, "learning_rate": 0.0002082216166768018, "loss": 0.1855, "step": 6011 }, { "epoch": 2.8291764705882354, "grad_norm": 0.4129655759241457, "learning_rate": 0.00020818121482819277, "loss": 0.2321, "step": 6012 }, { "epoch": 2.8296470588235296, "grad_norm": 0.22551047516318937, "learning_rate": 0.00020814080801062263, "loss": 0.1051, "step": 6013 }, { "epoch": 2.8301176470588234, "grad_norm": 0.3591165920421293, "learning_rate": 0.00020810039622754228, "loss": 0.3828, "step": 6014 }, { "epoch": 2.8305882352941176, "grad_norm": 0.26755530042130826, "learning_rate": 0.00020805997948240314, "loss": 0.2005, "step": 6015 }, { "epoch": 2.831058823529412, "grad_norm": 0.2646535160467725, "learning_rate": 0.00020801955777865694, "loss": 0.2942, "step": 6016 }, { "epoch": 2.8315294117647056, "grad_norm": 0.5522789729419031, "learning_rate": 0.00020797913111975593, "loss": 0.609, "step": 6017 }, { "epoch": 2.832, "grad_norm": 0.2360243943925265, "learning_rate": 0.00020793869950915276, "loss": 0.1526, "step": 6018 }, { "epoch": 2.832470588235294, "grad_norm": 0.4051995666927667, "learning_rate": 0.00020789826295030046, "loss": 0.2716, "step": 6019 }, { "epoch": 2.8329411764705883, "grad_norm": 0.20368360499964466, "learning_rate": 0.0002078578214466525, "loss": 0.1669, "step": 6020 }, { "epoch": 2.8334117647058825, "grad_norm": 0.33124466193018376, "learning_rate": 0.00020781737500166279, "loss": 0.174, "step": 6021 }, { "epoch": 2.8338823529411767, "grad_norm": 0.2679399115384325, "learning_rate": 0.00020777692361878569, "loss": 0.1317, "step": 6022 }, { "epoch": 2.8343529411764705, "grad_norm": 0.3416523125502077, "learning_rate": 0.0002077364673014759, "loss": 0.3036, "step": 6023 }, { "epoch": 2.8348235294117647, "grad_norm": 0.45215389618708357, "learning_rate": 0.00020769600605318857, "loss": 0.5291, "step": 6024 }, { "epoch": 2.835294117647059, "grad_norm": 0.23420364342717032, "learning_rate": 0.0002076555398773794, "loss": 0.1758, "step": 6025 }, { "epoch": 2.8357647058823527, "grad_norm": 0.24381878987696398, "learning_rate": 0.00020761506877750428, "loss": 0.1277, "step": 6026 }, { "epoch": 2.836235294117647, "grad_norm": 0.2923375402500779, "learning_rate": 0.00020757459275701966, "loss": 0.2052, "step": 6027 }, { "epoch": 2.836705882352941, "grad_norm": 0.2367620789118709, "learning_rate": 0.0002075341118193825, "loss": 0.1019, "step": 6028 }, { "epoch": 2.8371764705882354, "grad_norm": 0.39357295958872746, "learning_rate": 0.00020749362596804994, "loss": 0.459, "step": 6029 }, { "epoch": 2.8376470588235296, "grad_norm": 0.2861171053447979, "learning_rate": 0.00020745313520647971, "loss": 0.1657, "step": 6030 }, { "epoch": 2.8381176470588234, "grad_norm": 0.2321510406262621, "learning_rate": 0.00020741263953812992, "loss": 0.1231, "step": 6031 }, { "epoch": 2.8385882352941176, "grad_norm": 0.394646998059736, "learning_rate": 0.0002073721389664591, "loss": 0.293, "step": 6032 }, { "epoch": 2.839058823529412, "grad_norm": 0.4858218280176134, "learning_rate": 0.00020733163349492625, "loss": 0.483, "step": 6033 }, { "epoch": 2.8395294117647056, "grad_norm": 0.399396838610817, "learning_rate": 0.00020729112312699066, "loss": 0.3382, "step": 6034 }, { "epoch": 2.84, "grad_norm": 0.36904579048851033, "learning_rate": 0.00020725060786611212, "loss": 0.4464, "step": 6035 }, { "epoch": 2.840470588235294, "grad_norm": 0.4043858608914552, "learning_rate": 0.0002072100877157509, "loss": 0.4682, "step": 6036 }, { "epoch": 2.8409411764705883, "grad_norm": 0.4209613697047848, "learning_rate": 0.00020716956267936759, "loss": 0.2812, "step": 6037 }, { "epoch": 2.8414117647058825, "grad_norm": 0.36615805642186, "learning_rate": 0.0002071290327604231, "loss": 0.3656, "step": 6038 }, { "epoch": 2.8418823529411767, "grad_norm": 0.15722933711506254, "learning_rate": 0.00020708849796237914, "loss": 0.0833, "step": 6039 }, { "epoch": 2.8423529411764705, "grad_norm": 0.14734888699728513, "learning_rate": 0.00020704795828869736, "loss": 0.0723, "step": 6040 }, { "epoch": 2.8428235294117647, "grad_norm": 0.3294506803279619, "learning_rate": 0.00020700741374284014, "loss": 0.3171, "step": 6041 }, { "epoch": 2.843294117647059, "grad_norm": 0.3238027493514581, "learning_rate": 0.00020696686432827017, "loss": 0.2034, "step": 6042 }, { "epoch": 2.8437647058823528, "grad_norm": 0.21834786157919372, "learning_rate": 0.00020692631004845055, "loss": 0.0692, "step": 6043 }, { "epoch": 2.844235294117647, "grad_norm": 0.2087395665229225, "learning_rate": 0.00020688575090684484, "loss": 0.1001, "step": 6044 }, { "epoch": 2.844705882352941, "grad_norm": 0.44844605385744535, "learning_rate": 0.00020684518690691696, "loss": 0.2852, "step": 6045 }, { "epoch": 2.8451764705882354, "grad_norm": 0.4327812677938988, "learning_rate": 0.00020680461805213133, "loss": 0.2617, "step": 6046 }, { "epoch": 2.8456470588235296, "grad_norm": 0.30798439880662265, "learning_rate": 0.00020676404434595265, "loss": 0.3061, "step": 6047 }, { "epoch": 2.8461176470588234, "grad_norm": 0.318913020658719, "learning_rate": 0.00020672346579184617, "loss": 0.2621, "step": 6048 }, { "epoch": 2.8465882352941176, "grad_norm": 0.3639811379029706, "learning_rate": 0.00020668288239327753, "loss": 0.2565, "step": 6049 }, { "epoch": 2.847058823529412, "grad_norm": 0.4865880906096324, "learning_rate": 0.00020664229415371266, "loss": 0.5555, "step": 6050 }, { "epoch": 2.8475294117647056, "grad_norm": 0.28244102936382703, "learning_rate": 0.00020660170107661803, "loss": 0.2431, "step": 6051 }, { "epoch": 2.848, "grad_norm": 0.5283112650558778, "learning_rate": 0.00020656110316546054, "loss": 0.4834, "step": 6052 }, { "epoch": 2.848470588235294, "grad_norm": 0.16257186364252613, "learning_rate": 0.00020652050042370743, "loss": 0.0864, "step": 6053 }, { "epoch": 2.8489411764705883, "grad_norm": 0.18640456513654036, "learning_rate": 0.0002064798928548263, "loss": 0.0655, "step": 6054 }, { "epoch": 2.8494117647058825, "grad_norm": 0.42940729482022955, "learning_rate": 0.00020643928046228525, "loss": 0.2798, "step": 6055 }, { "epoch": 2.8498823529411763, "grad_norm": 0.4389926885356886, "learning_rate": 0.00020639866324955285, "loss": 0.181, "step": 6056 }, { "epoch": 2.8503529411764705, "grad_norm": 0.3492675942398658, "learning_rate": 0.00020635804122009802, "loss": 0.2345, "step": 6057 }, { "epoch": 2.8508235294117648, "grad_norm": 0.27726239185222756, "learning_rate": 0.00020631741437738998, "loss": 0.1793, "step": 6058 }, { "epoch": 2.851294117647059, "grad_norm": 0.5445969646572162, "learning_rate": 0.00020627678272489854, "loss": 0.4054, "step": 6059 }, { "epoch": 2.8517647058823528, "grad_norm": 0.41917404822429627, "learning_rate": 0.00020623614626609385, "loss": 0.2683, "step": 6060 }, { "epoch": 2.852235294117647, "grad_norm": 0.05747158001718416, "learning_rate": 0.0002061955050044464, "loss": 0.0123, "step": 6061 }, { "epoch": 2.852705882352941, "grad_norm": 0.21149702880659738, "learning_rate": 0.00020615485894342716, "loss": 0.1088, "step": 6062 }, { "epoch": 2.8531764705882354, "grad_norm": 0.17039365875250126, "learning_rate": 0.00020611420808650758, "loss": 0.1283, "step": 6063 }, { "epoch": 2.8536470588235296, "grad_norm": 0.5395794246553816, "learning_rate": 0.00020607355243715936, "loss": 0.429, "step": 6064 }, { "epoch": 2.8541176470588234, "grad_norm": 0.3310894435481534, "learning_rate": 0.00020603289199885473, "loss": 0.2749, "step": 6065 }, { "epoch": 2.8545882352941176, "grad_norm": 0.29572424284549576, "learning_rate": 0.00020599222677506628, "loss": 0.2082, "step": 6066 }, { "epoch": 2.855058823529412, "grad_norm": 0.3820588432830117, "learning_rate": 0.00020595155676926704, "loss": 0.3373, "step": 6067 }, { "epoch": 2.8555294117647056, "grad_norm": 0.1473235778242011, "learning_rate": 0.0002059108819849304, "loss": 0.0488, "step": 6068 }, { "epoch": 2.856, "grad_norm": 0.41501130437497036, "learning_rate": 0.0002058702024255302, "loss": 0.2086, "step": 6069 }, { "epoch": 2.856470588235294, "grad_norm": 0.5387897636536211, "learning_rate": 0.00020582951809454067, "loss": 0.4281, "step": 6070 }, { "epoch": 2.8569411764705883, "grad_norm": 0.2328715239932245, "learning_rate": 0.00020578882899543644, "loss": 0.1506, "step": 6071 }, { "epoch": 2.8574117647058825, "grad_norm": 0.39717599904194606, "learning_rate": 0.0002057481351316926, "loss": 0.3246, "step": 6072 }, { "epoch": 2.8578823529411763, "grad_norm": 0.2490041610357104, "learning_rate": 0.00020570743650678454, "loss": 0.2317, "step": 6073 }, { "epoch": 2.8583529411764705, "grad_norm": 0.22216963538019802, "learning_rate": 0.00020566673312418817, "loss": 0.1117, "step": 6074 }, { "epoch": 2.8588235294117648, "grad_norm": 0.6473497168139715, "learning_rate": 0.00020562602498737975, "loss": 0.4793, "step": 6075 }, { "epoch": 2.859294117647059, "grad_norm": 0.3353464086858056, "learning_rate": 0.00020558531209983593, "loss": 0.2739, "step": 6076 }, { "epoch": 2.8597647058823528, "grad_norm": 0.3408389603029807, "learning_rate": 0.00020554459446503386, "loss": 0.3074, "step": 6077 }, { "epoch": 2.860235294117647, "grad_norm": 0.18414669484657248, "learning_rate": 0.00020550387208645093, "loss": 0.0924, "step": 6078 }, { "epoch": 2.860705882352941, "grad_norm": 0.2224352805580128, "learning_rate": 0.0002054631449675651, "loss": 0.1377, "step": 6079 }, { "epoch": 2.8611764705882354, "grad_norm": 0.4072464577528546, "learning_rate": 0.00020542241311185467, "loss": 0.1963, "step": 6080 }, { "epoch": 2.8616470588235297, "grad_norm": 0.6383181962299771, "learning_rate": 0.0002053816765227983, "loss": 0.2643, "step": 6081 }, { "epoch": 2.8621176470588234, "grad_norm": 0.4371589557241938, "learning_rate": 0.0002053409352038751, "loss": 0.4585, "step": 6082 }, { "epoch": 2.8625882352941177, "grad_norm": 0.3732287035932156, "learning_rate": 0.00020530018915856457, "loss": 0.3106, "step": 6083 }, { "epoch": 2.863058823529412, "grad_norm": 0.21029265364379526, "learning_rate": 0.00020525943839034668, "loss": 0.0795, "step": 6084 }, { "epoch": 2.8635294117647057, "grad_norm": 0.5300425045561861, "learning_rate": 0.0002052186829027017, "loss": 0.3817, "step": 6085 }, { "epoch": 2.864, "grad_norm": 0.3362703162119605, "learning_rate": 0.00020517792269911033, "loss": 0.1311, "step": 6086 }, { "epoch": 2.864470588235294, "grad_norm": 0.676626811807939, "learning_rate": 0.0002051371577830537, "loss": 0.4451, "step": 6087 }, { "epoch": 2.8649411764705883, "grad_norm": 0.35045628362424724, "learning_rate": 0.00020509638815801342, "loss": 0.3319, "step": 6088 }, { "epoch": 2.8654117647058825, "grad_norm": 0.550556905796301, "learning_rate": 0.00020505561382747127, "loss": 0.5977, "step": 6089 }, { "epoch": 2.8658823529411763, "grad_norm": 0.17154489822314514, "learning_rate": 0.00020501483479490968, "loss": 0.0653, "step": 6090 }, { "epoch": 2.8663529411764705, "grad_norm": 0.790061575066423, "learning_rate": 0.00020497405106381138, "loss": 0.5007, "step": 6091 }, { "epoch": 2.8668235294117648, "grad_norm": 0.20973549132918878, "learning_rate": 0.0002049332626376594, "loss": 0.1206, "step": 6092 }, { "epoch": 2.8672941176470585, "grad_norm": 0.19534739992626945, "learning_rate": 0.00020489246951993742, "loss": 0.1004, "step": 6093 }, { "epoch": 2.8677647058823528, "grad_norm": 0.4288026489176524, "learning_rate": 0.00020485167171412926, "loss": 0.4101, "step": 6094 }, { "epoch": 2.868235294117647, "grad_norm": 0.09339759158532811, "learning_rate": 0.00020481086922371935, "loss": 0.0121, "step": 6095 }, { "epoch": 2.868705882352941, "grad_norm": 0.339399296080805, "learning_rate": 0.00020477006205219228, "loss": 0.3053, "step": 6096 }, { "epoch": 2.8691764705882354, "grad_norm": 0.1684912628018048, "learning_rate": 0.0002047292502030333, "loss": 0.0722, "step": 6097 }, { "epoch": 2.8696470588235297, "grad_norm": 0.4586516182929989, "learning_rate": 0.00020468843367972786, "loss": 0.4036, "step": 6098 }, { "epoch": 2.8701176470588234, "grad_norm": 0.35977221062999737, "learning_rate": 0.00020464761248576203, "loss": 0.2844, "step": 6099 }, { "epoch": 2.8705882352941177, "grad_norm": 0.255408186717509, "learning_rate": 0.00020460678662462194, "loss": 0.2755, "step": 6100 }, { "epoch": 2.871058823529412, "grad_norm": 0.36962737000264817, "learning_rate": 0.0002045659560997945, "loss": 0.2799, "step": 6101 }, { "epoch": 2.8715294117647057, "grad_norm": 0.3841559819243124, "learning_rate": 0.00020452512091476678, "loss": 0.3546, "step": 6102 }, { "epoch": 2.872, "grad_norm": 0.4303721159461924, "learning_rate": 0.0002044842810730262, "loss": 0.4218, "step": 6103 }, { "epoch": 2.872470588235294, "grad_norm": 0.6343652063171603, "learning_rate": 0.00020444343657806087, "loss": 0.3369, "step": 6104 }, { "epoch": 2.8729411764705883, "grad_norm": 0.3498211097204774, "learning_rate": 0.000204402587433359, "loss": 0.3072, "step": 6105 }, { "epoch": 2.8734117647058826, "grad_norm": 0.3375135100690242, "learning_rate": 0.0002043617336424093, "loss": 0.238, "step": 6106 }, { "epoch": 2.8738823529411763, "grad_norm": 0.3496573791812957, "learning_rate": 0.00020432087520870086, "loss": 0.1261, "step": 6107 }, { "epoch": 2.8743529411764706, "grad_norm": 0.4273855756120285, "learning_rate": 0.0002042800121357233, "loss": 0.2768, "step": 6108 }, { "epoch": 2.8748235294117648, "grad_norm": 0.2593340224874622, "learning_rate": 0.00020423914442696642, "loss": 0.0684, "step": 6109 }, { "epoch": 2.8752941176470586, "grad_norm": 0.4646924733786116, "learning_rate": 0.00020419827208592058, "loss": 0.4941, "step": 6110 }, { "epoch": 2.875764705882353, "grad_norm": 0.5236795462843895, "learning_rate": 0.00020415739511607653, "loss": 0.252, "step": 6111 }, { "epoch": 2.876235294117647, "grad_norm": 0.7515794966166297, "learning_rate": 0.0002041165135209252, "loss": 0.4109, "step": 6112 }, { "epoch": 2.8767058823529412, "grad_norm": 0.4534745998313096, "learning_rate": 0.0002040756273039582, "loss": 0.2668, "step": 6113 }, { "epoch": 2.8771764705882354, "grad_norm": 0.20073628110749592, "learning_rate": 0.0002040347364686674, "loss": 0.086, "step": 6114 }, { "epoch": 2.8776470588235297, "grad_norm": 0.3828649834694294, "learning_rate": 0.00020399384101854506, "loss": 0.3267, "step": 6115 }, { "epoch": 2.8781176470588234, "grad_norm": 0.7060760410400073, "learning_rate": 0.00020395294095708387, "loss": 0.5672, "step": 6116 }, { "epoch": 2.8785882352941177, "grad_norm": 0.31268110290600626, "learning_rate": 0.00020391203628777687, "loss": 0.1431, "step": 6117 }, { "epoch": 2.879058823529412, "grad_norm": 0.26249140308957625, "learning_rate": 0.00020387112701411754, "loss": 0.1526, "step": 6118 }, { "epoch": 2.8795294117647057, "grad_norm": 0.2373881484632613, "learning_rate": 0.00020383021313959972, "loss": 0.1129, "step": 6119 }, { "epoch": 2.88, "grad_norm": 0.16993128111488126, "learning_rate": 0.00020378929466771767, "loss": 0.0972, "step": 6120 }, { "epoch": 2.880470588235294, "grad_norm": 0.3460291997183686, "learning_rate": 0.000203748371601966, "loss": 0.2359, "step": 6121 }, { "epoch": 2.8809411764705883, "grad_norm": 0.13357768938088777, "learning_rate": 0.0002037074439458398, "loss": 0.0592, "step": 6122 }, { "epoch": 2.8814117647058826, "grad_norm": 0.3644150046364143, "learning_rate": 0.0002036665117028344, "loss": 0.2989, "step": 6123 }, { "epoch": 2.8818823529411763, "grad_norm": 0.19017669937484458, "learning_rate": 0.0002036255748764457, "loss": 0.0838, "step": 6124 }, { "epoch": 2.8823529411764706, "grad_norm": 0.2905088389842114, "learning_rate": 0.00020358463347016988, "loss": 0.1433, "step": 6125 }, { "epoch": 2.882823529411765, "grad_norm": 0.3433217745245269, "learning_rate": 0.00020354368748750357, "loss": 0.1189, "step": 6126 }, { "epoch": 2.8832941176470586, "grad_norm": 0.6401987393305252, "learning_rate": 0.00020350273693194363, "loss": 0.4469, "step": 6127 }, { "epoch": 2.883764705882353, "grad_norm": 0.22882466879035268, "learning_rate": 0.00020346178180698758, "loss": 0.0959, "step": 6128 }, { "epoch": 2.884235294117647, "grad_norm": 0.3671103413439487, "learning_rate": 0.00020342082211613312, "loss": 0.2997, "step": 6129 }, { "epoch": 2.8847058823529412, "grad_norm": 0.4537902594295555, "learning_rate": 0.00020337985786287842, "loss": 0.4708, "step": 6130 }, { "epoch": 2.8851764705882355, "grad_norm": 0.31143876729820996, "learning_rate": 0.00020333888905072208, "loss": 0.1732, "step": 6131 }, { "epoch": 2.8856470588235297, "grad_norm": 0.44631386729523165, "learning_rate": 0.00020329791568316294, "loss": 0.213, "step": 6132 }, { "epoch": 2.8861176470588235, "grad_norm": 0.3193994397397916, "learning_rate": 0.00020325693776370037, "loss": 0.2797, "step": 6133 }, { "epoch": 2.8865882352941177, "grad_norm": 0.42974233052859223, "learning_rate": 0.0002032159552958341, "loss": 0.4177, "step": 6134 }, { "epoch": 2.887058823529412, "grad_norm": 0.3342793573458239, "learning_rate": 0.00020317496828306427, "loss": 0.2667, "step": 6135 }, { "epoch": 2.8875294117647057, "grad_norm": 0.357530330077439, "learning_rate": 0.00020313397672889132, "loss": 0.2792, "step": 6136 }, { "epoch": 2.888, "grad_norm": 0.5355129191114025, "learning_rate": 0.00020309298063681607, "loss": 0.3262, "step": 6137 }, { "epoch": 2.888470588235294, "grad_norm": 0.4427546871549631, "learning_rate": 0.00020305198001033988, "loss": 0.2536, "step": 6138 }, { "epoch": 2.8889411764705883, "grad_norm": 0.6905242128807667, "learning_rate": 0.00020301097485296443, "loss": 0.586, "step": 6139 }, { "epoch": 2.8894117647058826, "grad_norm": 0.17983459166639015, "learning_rate": 0.00020296996516819163, "loss": 0.1161, "step": 6140 }, { "epoch": 2.8898823529411763, "grad_norm": 0.20589567904786096, "learning_rate": 0.00020292895095952402, "loss": 0.1127, "step": 6141 }, { "epoch": 2.8903529411764706, "grad_norm": 0.1744173621694062, "learning_rate": 0.0002028879322304644, "loss": 0.1203, "step": 6142 }, { "epoch": 2.890823529411765, "grad_norm": 0.43061520645698415, "learning_rate": 0.00020284690898451593, "loss": 0.3225, "step": 6143 }, { "epoch": 2.8912941176470586, "grad_norm": 0.5744105702442169, "learning_rate": 0.00020280588122518223, "loss": 0.5412, "step": 6144 }, { "epoch": 2.891764705882353, "grad_norm": 0.26658029745775136, "learning_rate": 0.00020276484895596726, "loss": 0.3081, "step": 6145 }, { "epoch": 2.892235294117647, "grad_norm": 0.23596489175487675, "learning_rate": 0.00020272381218037542, "loss": 0.1376, "step": 6146 }, { "epoch": 2.8927058823529412, "grad_norm": 0.23381345609821166, "learning_rate": 0.00020268277090191133, "loss": 0.0633, "step": 6147 }, { "epoch": 2.8931764705882355, "grad_norm": 0.3294113391499901, "learning_rate": 0.00020264172512408026, "loss": 0.18, "step": 6148 }, { "epoch": 2.8936470588235297, "grad_norm": 0.6007299495920009, "learning_rate": 0.00020260067485038767, "loss": 0.4046, "step": 6149 }, { "epoch": 2.8941176470588235, "grad_norm": 0.5174833250504275, "learning_rate": 0.0002025596200843394, "loss": 0.3004, "step": 6150 }, { "epoch": 2.8945882352941177, "grad_norm": 0.36867354684598985, "learning_rate": 0.00020251856082944178, "loss": 0.3205, "step": 6151 }, { "epoch": 2.895058823529412, "grad_norm": 0.2454367997602342, "learning_rate": 0.00020247749708920146, "loss": 0.2323, "step": 6152 }, { "epoch": 2.8955294117647057, "grad_norm": 0.4275199550386487, "learning_rate": 0.00020243642886712553, "loss": 0.2252, "step": 6153 }, { "epoch": 2.896, "grad_norm": 0.3906121983172532, "learning_rate": 0.0002023953561667213, "loss": 0.2586, "step": 6154 }, { "epoch": 2.896470588235294, "grad_norm": 0.29627660251710314, "learning_rate": 0.0002023542789914967, "loss": 0.2362, "step": 6155 }, { "epoch": 2.8969411764705884, "grad_norm": 0.5955251376359285, "learning_rate": 0.0002023131973449599, "loss": 0.4168, "step": 6156 }, { "epoch": 2.8974117647058826, "grad_norm": 0.14133168880982117, "learning_rate": 0.00020227211123061942, "loss": 0.0532, "step": 6157 }, { "epoch": 2.8978823529411764, "grad_norm": 0.1990449791621335, "learning_rate": 0.00020223102065198426, "loss": 0.1473, "step": 6158 }, { "epoch": 2.8983529411764706, "grad_norm": 0.584912327679505, "learning_rate": 0.00020218992561256375, "loss": 0.5835, "step": 6159 }, { "epoch": 2.898823529411765, "grad_norm": 0.36648700221257824, "learning_rate": 0.0002021488261158676, "loss": 0.2561, "step": 6160 }, { "epoch": 2.8992941176470586, "grad_norm": 0.3933264168632504, "learning_rate": 0.0002021077221654059, "loss": 0.1505, "step": 6161 }, { "epoch": 2.899764705882353, "grad_norm": 0.2176541899224822, "learning_rate": 0.00020206661376468912, "loss": 0.0763, "step": 6162 }, { "epoch": 2.900235294117647, "grad_norm": 0.3972024830684802, "learning_rate": 0.00020202550091722818, "loss": 0.2343, "step": 6163 }, { "epoch": 2.9007058823529412, "grad_norm": 0.38632516896642105, "learning_rate": 0.00020198438362653426, "loss": 0.24, "step": 6164 }, { "epoch": 2.9011764705882355, "grad_norm": 0.2568607235549829, "learning_rate": 0.00020194326189611897, "loss": 0.1322, "step": 6165 }, { "epoch": 2.9016470588235292, "grad_norm": 0.22098206899896478, "learning_rate": 0.00020190213572949435, "loss": 0.0892, "step": 6166 }, { "epoch": 2.9021176470588235, "grad_norm": 0.41690423978427826, "learning_rate": 0.0002018610051301728, "loss": 0.252, "step": 6167 }, { "epoch": 2.9025882352941177, "grad_norm": 0.44171498962414596, "learning_rate": 0.00020181987010166699, "loss": 0.2668, "step": 6168 }, { "epoch": 2.903058823529412, "grad_norm": 0.3981732285622792, "learning_rate": 0.0002017787306474901, "loss": 0.3868, "step": 6169 }, { "epoch": 2.9035294117647057, "grad_norm": 0.37705455562807305, "learning_rate": 0.00020173758677115568, "loss": 0.1885, "step": 6170 }, { "epoch": 2.904, "grad_norm": 0.4626592538225563, "learning_rate": 0.00020169643847617753, "loss": 0.3743, "step": 6171 }, { "epoch": 2.904470588235294, "grad_norm": 0.15468189525395687, "learning_rate": 0.00020165528576607, "loss": 0.1356, "step": 6172 }, { "epoch": 2.9049411764705884, "grad_norm": 0.15621622465776325, "learning_rate": 0.0002016141286443477, "loss": 0.0559, "step": 6173 }, { "epoch": 2.9054117647058826, "grad_norm": 0.6822249611024724, "learning_rate": 0.0002015729671145256, "loss": 0.4774, "step": 6174 }, { "epoch": 2.9058823529411764, "grad_norm": 0.31727293967334375, "learning_rate": 0.0002015318011801192, "loss": 0.1643, "step": 6175 }, { "epoch": 2.9063529411764706, "grad_norm": 0.2224837322322435, "learning_rate": 0.0002014906308446442, "loss": 0.142, "step": 6176 }, { "epoch": 2.906823529411765, "grad_norm": 0.34702757949208196, "learning_rate": 0.0002014494561116168, "loss": 0.1736, "step": 6177 }, { "epoch": 2.9072941176470586, "grad_norm": 0.45398027824647663, "learning_rate": 0.00020140827698455338, "loss": 0.2975, "step": 6178 }, { "epoch": 2.907764705882353, "grad_norm": 0.19940833576714107, "learning_rate": 0.00020136709346697107, "loss": 0.1105, "step": 6179 }, { "epoch": 2.908235294117647, "grad_norm": 0.5144372731794467, "learning_rate": 0.00020132590556238696, "loss": 0.3154, "step": 6180 }, { "epoch": 2.9087058823529413, "grad_norm": 0.3305434784784096, "learning_rate": 0.00020128471327431877, "loss": 0.2621, "step": 6181 }, { "epoch": 2.9091764705882355, "grad_norm": 0.2556984082057433, "learning_rate": 0.0002012435166062845, "loss": 0.168, "step": 6182 }, { "epoch": 2.9096470588235293, "grad_norm": 0.5861357080997491, "learning_rate": 0.00020120231556180262, "loss": 0.4356, "step": 6183 }, { "epoch": 2.9101176470588235, "grad_norm": 0.3158220367667805, "learning_rate": 0.00020116111014439178, "loss": 0.2761, "step": 6184 }, { "epoch": 2.9105882352941177, "grad_norm": 0.061602675202400745, "learning_rate": 0.00020111990035757125, "loss": 0.0225, "step": 6185 }, { "epoch": 2.911058823529412, "grad_norm": 0.05164517196891064, "learning_rate": 0.00020107868620486047, "loss": 0.0124, "step": 6186 }, { "epoch": 2.9115294117647057, "grad_norm": 0.377198109650893, "learning_rate": 0.0002010374676897793, "loss": 0.1482, "step": 6187 }, { "epoch": 2.912, "grad_norm": 1.1232238124439053, "learning_rate": 0.00020099624481584807, "loss": 0.4804, "step": 6188 }, { "epoch": 2.912470588235294, "grad_norm": 0.4374069923141651, "learning_rate": 0.0002009550175865874, "loss": 0.2464, "step": 6189 }, { "epoch": 2.9129411764705884, "grad_norm": 0.5377676961701175, "learning_rate": 0.00020091378600551834, "loss": 0.5447, "step": 6190 }, { "epoch": 2.9134117647058826, "grad_norm": 0.323171692004185, "learning_rate": 0.00020087255007616222, "loss": 0.3401, "step": 6191 }, { "epoch": 2.9138823529411764, "grad_norm": 0.3659419159636616, "learning_rate": 0.00020083130980204074, "loss": 0.2616, "step": 6192 }, { "epoch": 2.9143529411764706, "grad_norm": 0.4579189645522887, "learning_rate": 0.0002007900651866761, "loss": 0.3527, "step": 6193 }, { "epoch": 2.914823529411765, "grad_norm": 0.8672049070333945, "learning_rate": 0.00020074881623359083, "loss": 0.8508, "step": 6194 }, { "epoch": 2.9152941176470586, "grad_norm": 0.3072826174217709, "learning_rate": 0.00020070756294630765, "loss": 0.1339, "step": 6195 }, { "epoch": 2.915764705882353, "grad_norm": 0.5512612714042705, "learning_rate": 0.0002006663053283499, "loss": 0.4915, "step": 6196 }, { "epoch": 2.916235294117647, "grad_norm": 0.7778910372272659, "learning_rate": 0.00020062504338324116, "loss": 0.3285, "step": 6197 }, { "epoch": 2.9167058823529413, "grad_norm": 0.46418416849999455, "learning_rate": 0.0002005837771145054, "loss": 0.2447, "step": 6198 }, { "epoch": 2.9171764705882355, "grad_norm": 0.5856385277410546, "learning_rate": 0.00020054250652566696, "loss": 0.4306, "step": 6199 }, { "epoch": 2.9176470588235293, "grad_norm": 0.30681375228523605, "learning_rate": 0.0002005012316202506, "loss": 0.2744, "step": 6200 }, { "epoch": 2.9181176470588235, "grad_norm": 0.5424004634663557, "learning_rate": 0.00020045995240178133, "loss": 0.2663, "step": 6201 }, { "epoch": 2.9185882352941177, "grad_norm": 0.43295904193017976, "learning_rate": 0.00020041866887378457, "loss": 0.2087, "step": 6202 }, { "epoch": 2.919058823529412, "grad_norm": 0.5964853927191248, "learning_rate": 0.0002003773810397862, "loss": 0.5076, "step": 6203 }, { "epoch": 2.9195294117647057, "grad_norm": 0.4934349425565499, "learning_rate": 0.00020033608890331246, "loss": 0.3799, "step": 6204 }, { "epoch": 2.92, "grad_norm": 0.4139801421993407, "learning_rate": 0.00020029479246788973, "loss": 0.4402, "step": 6205 }, { "epoch": 2.920470588235294, "grad_norm": 0.3154107158070739, "learning_rate": 0.00020025349173704508, "loss": 0.3492, "step": 6206 }, { "epoch": 2.9209411764705884, "grad_norm": 0.163646365912034, "learning_rate": 0.00020021218671430573, "loss": 0.0665, "step": 6207 }, { "epoch": 2.9214117647058826, "grad_norm": 0.3741951520793159, "learning_rate": 0.00020017087740319936, "loss": 0.2397, "step": 6208 }, { "epoch": 2.9218823529411764, "grad_norm": 0.10185195650115252, "learning_rate": 0.00020012956380725393, "loss": 0.0422, "step": 6209 }, { "epoch": 2.9223529411764706, "grad_norm": 0.7090134178976878, "learning_rate": 0.00020008824592999795, "loss": 0.4027, "step": 6210 }, { "epoch": 2.922823529411765, "grad_norm": 0.5030274368788693, "learning_rate": 0.00020004692377496004, "loss": 0.2437, "step": 6211 }, { "epoch": 2.9232941176470586, "grad_norm": 0.45025472587350496, "learning_rate": 0.00020000559734566933, "loss": 0.3005, "step": 6212 }, { "epoch": 2.923764705882353, "grad_norm": 0.34830562868127357, "learning_rate": 0.00019996426664565538, "loss": 0.2089, "step": 6213 }, { "epoch": 2.924235294117647, "grad_norm": 0.40306190233221023, "learning_rate": 0.000199922931678448, "loss": 0.2435, "step": 6214 }, { "epoch": 2.9247058823529413, "grad_norm": 0.4069763302253985, "learning_rate": 0.00019988159244757736, "loss": 0.2094, "step": 6215 }, { "epoch": 2.9251764705882355, "grad_norm": 0.3347100737645077, "learning_rate": 0.000199840248956574, "loss": 0.3177, "step": 6216 }, { "epoch": 2.9256470588235293, "grad_norm": 0.31141907480742786, "learning_rate": 0.000199798901208969, "loss": 0.1769, "step": 6217 }, { "epoch": 2.9261176470588235, "grad_norm": 0.24435089065424886, "learning_rate": 0.0001997575492082935, "loss": 0.1632, "step": 6218 }, { "epoch": 2.9265882352941177, "grad_norm": 0.7432245517474743, "learning_rate": 0.0001997161929580793, "loss": 0.4139, "step": 6219 }, { "epoch": 2.9270588235294115, "grad_norm": 0.29328223963694755, "learning_rate": 0.00019967483246185834, "loss": 0.1366, "step": 6220 }, { "epoch": 2.9275294117647057, "grad_norm": 0.4606348677212745, "learning_rate": 0.00019963346772316307, "loss": 0.399, "step": 6221 }, { "epoch": 2.928, "grad_norm": 0.33596719369771927, "learning_rate": 0.00019959209874552612, "loss": 0.2979, "step": 6222 }, { "epoch": 2.928470588235294, "grad_norm": 0.6770950594545052, "learning_rate": 0.00019955072553248075, "loss": 0.6269, "step": 6223 }, { "epoch": 2.9289411764705884, "grad_norm": 0.1880663583819199, "learning_rate": 0.0001995093480875604, "loss": 0.0755, "step": 6224 }, { "epoch": 2.9294117647058826, "grad_norm": 0.36668294369430704, "learning_rate": 0.00019946796641429884, "loss": 0.0843, "step": 6225 }, { "epoch": 2.9298823529411764, "grad_norm": 0.3019572626279145, "learning_rate": 0.0001994265805162303, "loss": 0.1389, "step": 6226 }, { "epoch": 2.9303529411764706, "grad_norm": 0.6107161850356441, "learning_rate": 0.00019938519039688937, "loss": 0.6471, "step": 6227 }, { "epoch": 2.930823529411765, "grad_norm": 0.5748468243392072, "learning_rate": 0.00019934379605981097, "loss": 0.2512, "step": 6228 }, { "epoch": 2.9312941176470586, "grad_norm": 0.5404985942626719, "learning_rate": 0.0001993023975085303, "loss": 0.4204, "step": 6229 }, { "epoch": 2.931764705882353, "grad_norm": 0.2818180672313043, "learning_rate": 0.00019926099474658305, "loss": 0.1313, "step": 6230 }, { "epoch": 2.932235294117647, "grad_norm": 0.37747988543304223, "learning_rate": 0.0001992195877775053, "loss": 0.2543, "step": 6231 }, { "epoch": 2.9327058823529413, "grad_norm": 0.537726536164467, "learning_rate": 0.00019917817660483327, "loss": 0.5238, "step": 6232 }, { "epoch": 2.9331764705882355, "grad_norm": 0.2995866441105536, "learning_rate": 0.00019913676123210373, "loss": 0.1368, "step": 6233 }, { "epoch": 2.9336470588235293, "grad_norm": 0.3683804728923013, "learning_rate": 0.00019909534166285382, "loss": 0.2717, "step": 6234 }, { "epoch": 2.9341176470588235, "grad_norm": 0.4476247200556376, "learning_rate": 0.00019905391790062092, "loss": 0.4824, "step": 6235 }, { "epoch": 2.9345882352941177, "grad_norm": 0.4162264303573545, "learning_rate": 0.00019901248994894276, "loss": 0.3824, "step": 6236 }, { "epoch": 2.9350588235294115, "grad_norm": 0.43674094416734255, "learning_rate": 0.0001989710578113576, "loss": 0.2987, "step": 6237 }, { "epoch": 2.9355294117647057, "grad_norm": 0.2691162320063435, "learning_rate": 0.00019892962149140392, "loss": 0.0882, "step": 6238 }, { "epoch": 2.936, "grad_norm": 0.2577433839812706, "learning_rate": 0.0001988881809926205, "loss": 0.1549, "step": 6239 }, { "epoch": 2.936470588235294, "grad_norm": 0.33385036277410357, "learning_rate": 0.00019884673631854672, "loss": 0.234, "step": 6240 }, { "epoch": 2.9369411764705884, "grad_norm": 0.4603093874629453, "learning_rate": 0.000198805287472722, "loss": 0.2138, "step": 6241 }, { "epoch": 2.9374117647058826, "grad_norm": 0.1301558764229476, "learning_rate": 0.0001987638344586864, "loss": 0.0304, "step": 6242 }, { "epoch": 2.9378823529411764, "grad_norm": 0.2444995140177691, "learning_rate": 0.00019872237727998004, "loss": 0.2474, "step": 6243 }, { "epoch": 2.9383529411764706, "grad_norm": 0.47297122581505024, "learning_rate": 0.0001986809159401438, "loss": 0.1896, "step": 6244 }, { "epoch": 2.938823529411765, "grad_norm": 0.18886759098717015, "learning_rate": 0.00019863945044271846, "loss": 0.0816, "step": 6245 }, { "epoch": 2.9392941176470586, "grad_norm": 0.4507626023988674, "learning_rate": 0.00019859798079124554, "loss": 0.3442, "step": 6246 }, { "epoch": 2.939764705882353, "grad_norm": 0.23831388009711496, "learning_rate": 0.00019855650698926666, "loss": 0.1522, "step": 6247 }, { "epoch": 2.940235294117647, "grad_norm": 0.3297216173796831, "learning_rate": 0.00019851502904032396, "loss": 0.2625, "step": 6248 }, { "epoch": 2.9407058823529413, "grad_norm": 0.3896147410974866, "learning_rate": 0.00019847354694795978, "loss": 0.2032, "step": 6249 }, { "epoch": 2.9411764705882355, "grad_norm": 0.6919275582194546, "learning_rate": 0.00019843206071571692, "loss": 0.4393, "step": 6250 }, { "epoch": 2.9416470588235293, "grad_norm": 0.0324452142998621, "learning_rate": 0.00019839057034713858, "loss": 0.0036, "step": 6251 }, { "epoch": 2.9421176470588235, "grad_norm": 0.9458829800852003, "learning_rate": 0.00019834907584576812, "loss": 1.1383, "step": 6252 }, { "epoch": 2.9425882352941177, "grad_norm": 0.4117296521567231, "learning_rate": 0.00019830757721514943, "loss": 0.3511, "step": 6253 }, { "epoch": 2.9430588235294115, "grad_norm": 0.27812874006836746, "learning_rate": 0.00019826607445882667, "loss": 0.2245, "step": 6254 }, { "epoch": 2.9435294117647057, "grad_norm": 0.2887321617009633, "learning_rate": 0.00019822456758034445, "loss": 0.1962, "step": 6255 }, { "epoch": 2.944, "grad_norm": 0.3712430087530576, "learning_rate": 0.0001981830565832476, "loss": 0.264, "step": 6256 }, { "epoch": 2.944470588235294, "grad_norm": 0.16586610693006862, "learning_rate": 0.0001981415414710814, "loss": 0.0451, "step": 6257 }, { "epoch": 2.9449411764705884, "grad_norm": 0.09324178362850297, "learning_rate": 0.00019810002224739142, "loss": 0.0485, "step": 6258 }, { "epoch": 2.9454117647058826, "grad_norm": 0.3162701327873165, "learning_rate": 0.00019805849891572363, "loss": 0.1859, "step": 6259 }, { "epoch": 2.9458823529411764, "grad_norm": 0.36361825700687983, "learning_rate": 0.00019801697147962428, "loss": 0.2017, "step": 6260 }, { "epoch": 2.9463529411764706, "grad_norm": 0.4603443519124081, "learning_rate": 0.00019797543994264004, "loss": 0.3816, "step": 6261 }, { "epoch": 2.946823529411765, "grad_norm": 0.348860023144619, "learning_rate": 0.00019793390430831794, "loss": 0.2976, "step": 6262 }, { "epoch": 2.9472941176470586, "grad_norm": 0.331495746113047, "learning_rate": 0.00019789236458020527, "loss": 0.3735, "step": 6263 }, { "epoch": 2.947764705882353, "grad_norm": 0.4823641907303198, "learning_rate": 0.00019785082076184977, "loss": 0.3981, "step": 6264 }, { "epoch": 2.948235294117647, "grad_norm": 0.2640754963741523, "learning_rate": 0.00019780927285679945, "loss": 0.1367, "step": 6265 }, { "epoch": 2.9487058823529413, "grad_norm": 0.24438991113755434, "learning_rate": 0.00019776772086860277, "loss": 0.1373, "step": 6266 }, { "epoch": 2.9491764705882355, "grad_norm": 0.2570416513591807, "learning_rate": 0.0001977261648008084, "loss": 0.2122, "step": 6267 }, { "epoch": 2.9496470588235293, "grad_norm": 0.18203817046147566, "learning_rate": 0.00019768460465696545, "loss": 0.0471, "step": 6268 }, { "epoch": 2.9501176470588235, "grad_norm": 0.37703930172211264, "learning_rate": 0.0001976430404406234, "loss": 0.297, "step": 6269 }, { "epoch": 2.9505882352941177, "grad_norm": 0.49981723766370917, "learning_rate": 0.00019760147215533197, "loss": 0.479, "step": 6270 }, { "epoch": 2.9510588235294115, "grad_norm": 0.4757227450485595, "learning_rate": 0.00019755989980464137, "loss": 0.2778, "step": 6271 }, { "epoch": 2.9515294117647057, "grad_norm": 0.37830891362172747, "learning_rate": 0.00019751832339210204, "loss": 0.2496, "step": 6272 }, { "epoch": 2.952, "grad_norm": 0.22095560670301867, "learning_rate": 0.00019747674292126477, "loss": 0.1035, "step": 6273 }, { "epoch": 2.952470588235294, "grad_norm": 0.4047233787067612, "learning_rate": 0.00019743515839568083, "loss": 0.3618, "step": 6274 }, { "epoch": 2.9529411764705884, "grad_norm": 0.22231536852563122, "learning_rate": 0.00019739356981890167, "loss": 0.0862, "step": 6275 }, { "epoch": 2.9534117647058826, "grad_norm": 0.41046560317696584, "learning_rate": 0.00019735197719447916, "loss": 0.3616, "step": 6276 }, { "epoch": 2.9538823529411764, "grad_norm": 0.5480595104694856, "learning_rate": 0.00019731038052596558, "loss": 0.3652, "step": 6277 }, { "epoch": 2.9543529411764706, "grad_norm": 0.2502095636013249, "learning_rate": 0.00019726877981691347, "loss": 0.0998, "step": 6278 }, { "epoch": 2.954823529411765, "grad_norm": 0.3789422759331044, "learning_rate": 0.0001972271750708757, "loss": 0.3613, "step": 6279 }, { "epoch": 2.9552941176470586, "grad_norm": 0.3219525871095766, "learning_rate": 0.0001971855662914055, "loss": 0.2542, "step": 6280 }, { "epoch": 2.955764705882353, "grad_norm": 0.39602991977550633, "learning_rate": 0.00019714395348205649, "loss": 0.2781, "step": 6281 }, { "epoch": 2.956235294117647, "grad_norm": 0.5267722512725566, "learning_rate": 0.00019710233664638264, "loss": 0.3816, "step": 6282 }, { "epoch": 2.9567058823529413, "grad_norm": 0.4980268317147878, "learning_rate": 0.00019706071578793819, "loss": 0.3157, "step": 6283 }, { "epoch": 2.9571764705882355, "grad_norm": 0.5718745996723608, "learning_rate": 0.00019701909091027778, "loss": 0.4633, "step": 6284 }, { "epoch": 2.9576470588235293, "grad_norm": 0.404486973573483, "learning_rate": 0.0001969774620169564, "loss": 0.3846, "step": 6285 }, { "epoch": 2.9581176470588235, "grad_norm": 0.39580419422569346, "learning_rate": 0.00019693582911152935, "loss": 0.2741, "step": 6286 }, { "epoch": 2.9585882352941177, "grad_norm": 0.3299611954269457, "learning_rate": 0.0001968941921975522, "loss": 0.2834, "step": 6287 }, { "epoch": 2.9590588235294115, "grad_norm": 0.43353313387458453, "learning_rate": 0.00019685255127858107, "loss": 0.3318, "step": 6288 }, { "epoch": 2.9595294117647057, "grad_norm": 0.35332488349283314, "learning_rate": 0.00019681090635817228, "loss": 0.231, "step": 6289 }, { "epoch": 2.96, "grad_norm": 0.3058796654423251, "learning_rate": 0.00019676925743988244, "loss": 0.1093, "step": 6290 }, { "epoch": 2.960470588235294, "grad_norm": 0.31774915047484154, "learning_rate": 0.00019672760452726863, "loss": 0.167, "step": 6291 }, { "epoch": 2.9609411764705884, "grad_norm": 0.3151731876239984, "learning_rate": 0.0001966859476238882, "loss": 0.1641, "step": 6292 }, { "epoch": 2.961411764705882, "grad_norm": 0.47412626922919693, "learning_rate": 0.00019664428673329883, "loss": 0.3474, "step": 6293 }, { "epoch": 2.9618823529411764, "grad_norm": 0.3282874003004878, "learning_rate": 0.00019660262185905856, "loss": 0.0968, "step": 6294 }, { "epoch": 2.9623529411764706, "grad_norm": 0.6117390925513395, "learning_rate": 0.0001965609530047259, "loss": 0.375, "step": 6295 }, { "epoch": 2.962823529411765, "grad_norm": 0.45165125919196114, "learning_rate": 0.0001965192801738594, "loss": 0.4608, "step": 6296 }, { "epoch": 2.9632941176470586, "grad_norm": 0.15415065028389283, "learning_rate": 0.0001964776033700182, "loss": 0.0649, "step": 6297 }, { "epoch": 2.963764705882353, "grad_norm": 0.2811650726839177, "learning_rate": 0.00019643592259676164, "loss": 0.1976, "step": 6298 }, { "epoch": 2.964235294117647, "grad_norm": 0.10665296629516198, "learning_rate": 0.0001963942378576496, "loss": 0.0378, "step": 6299 }, { "epoch": 2.9647058823529413, "grad_norm": 0.560003476625799, "learning_rate": 0.0001963525491562421, "loss": 0.4871, "step": 6300 }, { "epoch": 2.9651764705882355, "grad_norm": 0.3886185036964095, "learning_rate": 0.0001963108564960995, "loss": 0.1604, "step": 6301 }, { "epoch": 2.9656470588235293, "grad_norm": 0.3021986962857455, "learning_rate": 0.00019626915988078263, "loss": 0.3016, "step": 6302 }, { "epoch": 2.9661176470588235, "grad_norm": 0.25816616463808373, "learning_rate": 0.00019622745931385254, "loss": 0.2381, "step": 6303 }, { "epoch": 2.9665882352941177, "grad_norm": 0.3214621695801598, "learning_rate": 0.00019618575479887066, "loss": 0.213, "step": 6304 }, { "epoch": 2.9670588235294115, "grad_norm": 0.21489626484260554, "learning_rate": 0.00019614404633939886, "loss": 0.1005, "step": 6305 }, { "epoch": 2.9675294117647057, "grad_norm": 0.2656381993034533, "learning_rate": 0.0001961023339389991, "loss": 0.1679, "step": 6306 }, { "epoch": 2.968, "grad_norm": 0.38425245426320903, "learning_rate": 0.00019606061760123392, "loss": 0.3134, "step": 6307 }, { "epoch": 2.968470588235294, "grad_norm": 0.4615772864566995, "learning_rate": 0.00019601889732966604, "loss": 0.2743, "step": 6308 }, { "epoch": 2.9689411764705884, "grad_norm": 0.21216741555583166, "learning_rate": 0.0001959771731278586, "loss": 0.1044, "step": 6309 }, { "epoch": 2.969411764705882, "grad_norm": 0.24795092532114918, "learning_rate": 0.0001959354449993751, "loss": 0.0921, "step": 6310 }, { "epoch": 2.9698823529411764, "grad_norm": 0.18968473792731644, "learning_rate": 0.00019589371294777925, "loss": 0.0797, "step": 6311 }, { "epoch": 2.9703529411764706, "grad_norm": 0.3463463051109008, "learning_rate": 0.00019585197697663526, "loss": 0.3222, "step": 6312 }, { "epoch": 2.970823529411765, "grad_norm": 0.2982951791776332, "learning_rate": 0.0001958102370895075, "loss": 0.2411, "step": 6313 }, { "epoch": 2.9712941176470586, "grad_norm": 0.365495809016603, "learning_rate": 0.00019576849328996075, "loss": 0.282, "step": 6314 }, { "epoch": 2.971764705882353, "grad_norm": 0.3501033884683989, "learning_rate": 0.00019572674558156024, "loss": 0.208, "step": 6315 }, { "epoch": 2.972235294117647, "grad_norm": 0.31907937387695684, "learning_rate": 0.00019568499396787134, "loss": 0.251, "step": 6316 }, { "epoch": 2.9727058823529413, "grad_norm": 0.515262925195226, "learning_rate": 0.0001956432384524599, "loss": 0.5746, "step": 6317 }, { "epoch": 2.9731764705882355, "grad_norm": 0.3599880918763829, "learning_rate": 0.000195601479038892, "loss": 0.1523, "step": 6318 }, { "epoch": 2.9736470588235293, "grad_norm": 0.21844161226177505, "learning_rate": 0.0001955597157307341, "loss": 0.1517, "step": 6319 }, { "epoch": 2.9741176470588235, "grad_norm": 0.3865418160719566, "learning_rate": 0.000195517948531553, "loss": 0.3381, "step": 6320 }, { "epoch": 2.9745882352941178, "grad_norm": 0.5920691609824236, "learning_rate": 0.00019547617744491587, "loss": 0.5925, "step": 6321 }, { "epoch": 2.9750588235294115, "grad_norm": 0.23480137990465683, "learning_rate": 0.00019543440247439008, "loss": 0.156, "step": 6322 }, { "epoch": 2.9755294117647058, "grad_norm": 0.12587014302049804, "learning_rate": 0.00019539262362354353, "loss": 0.0491, "step": 6323 }, { "epoch": 2.976, "grad_norm": 0.5118063470085324, "learning_rate": 0.00019535084089594422, "loss": 0.5016, "step": 6324 }, { "epoch": 2.976470588235294, "grad_norm": 0.565415743200641, "learning_rate": 0.00019530905429516066, "loss": 0.3142, "step": 6325 }, { "epoch": 2.9769411764705884, "grad_norm": 0.21605909603701448, "learning_rate": 0.00019526726382476165, "loss": 0.1259, "step": 6326 }, { "epoch": 2.977411764705882, "grad_norm": 0.14423912409246958, "learning_rate": 0.00019522546948831627, "loss": 0.0926, "step": 6327 }, { "epoch": 2.9778823529411764, "grad_norm": 0.4268260098728063, "learning_rate": 0.00019518367128939392, "loss": 0.401, "step": 6328 }, { "epoch": 2.9783529411764706, "grad_norm": 0.29793024298197684, "learning_rate": 0.0001951418692315645, "loss": 0.2267, "step": 6329 }, { "epoch": 2.978823529411765, "grad_norm": 0.32655686132346495, "learning_rate": 0.00019510006331839794, "loss": 0.2157, "step": 6330 }, { "epoch": 2.9792941176470586, "grad_norm": 0.18677266167929538, "learning_rate": 0.00019505825355346483, "loss": 0.068, "step": 6331 }, { "epoch": 2.979764705882353, "grad_norm": 0.4821870651765613, "learning_rate": 0.0001950164399403358, "loss": 0.4914, "step": 6332 }, { "epoch": 2.980235294117647, "grad_norm": 0.6301280438777986, "learning_rate": 0.00019497462248258202, "loss": 0.5722, "step": 6333 }, { "epoch": 2.9807058823529413, "grad_norm": 0.296148393005077, "learning_rate": 0.0001949328011837749, "loss": 0.1963, "step": 6334 }, { "epoch": 2.9811764705882355, "grad_norm": 0.4432246150788323, "learning_rate": 0.00019489097604748613, "loss": 0.3242, "step": 6335 }, { "epoch": 2.9816470588235293, "grad_norm": 0.3393085725149803, "learning_rate": 0.00019484914707728785, "loss": 0.2237, "step": 6336 }, { "epoch": 2.9821176470588235, "grad_norm": 0.31312201860750205, "learning_rate": 0.00019480731427675238, "loss": 0.3191, "step": 6337 }, { "epoch": 2.9825882352941178, "grad_norm": 0.2825793696737476, "learning_rate": 0.0001947654776494525, "loss": 0.1724, "step": 6338 }, { "epoch": 2.9830588235294115, "grad_norm": 0.14444349937568493, "learning_rate": 0.0001947236371989613, "loss": 0.0586, "step": 6339 }, { "epoch": 2.9835294117647058, "grad_norm": 0.5346975834329325, "learning_rate": 0.0001946817929288521, "loss": 0.3994, "step": 6340 }, { "epoch": 2.984, "grad_norm": 0.3924889597461805, "learning_rate": 0.0001946399448426986, "loss": 0.3723, "step": 6341 }, { "epoch": 2.984470588235294, "grad_norm": 0.4310524412939383, "learning_rate": 0.00019459809294407486, "loss": 0.309, "step": 6342 }, { "epoch": 2.9849411764705884, "grad_norm": 0.2416619706360588, "learning_rate": 0.00019455623723655521, "loss": 0.1118, "step": 6343 }, { "epoch": 2.985411764705882, "grad_norm": 0.5226172596314059, "learning_rate": 0.00019451437772371438, "loss": 0.4877, "step": 6344 }, { "epoch": 2.9858823529411764, "grad_norm": 0.44160088688149746, "learning_rate": 0.00019447251440912736, "loss": 0.4986, "step": 6345 }, { "epoch": 2.9863529411764707, "grad_norm": 0.32836608937146966, "learning_rate": 0.00019443064729636947, "loss": 0.1997, "step": 6346 }, { "epoch": 2.9868235294117644, "grad_norm": 0.29393010090380517, "learning_rate": 0.00019438877638901635, "loss": 0.1532, "step": 6347 }, { "epoch": 2.9872941176470587, "grad_norm": 0.35545395830546683, "learning_rate": 0.00019434690169064402, "loss": 0.1829, "step": 6348 }, { "epoch": 2.987764705882353, "grad_norm": 0.24829469720680222, "learning_rate": 0.00019430502320482871, "loss": 0.1885, "step": 6349 }, { "epoch": 2.988235294117647, "grad_norm": 0.39404017198299246, "learning_rate": 0.00019426314093514717, "loss": 0.2789, "step": 6350 }, { "epoch": 2.9887058823529413, "grad_norm": 0.3283197528552297, "learning_rate": 0.0001942212548851763, "loss": 0.1753, "step": 6351 }, { "epoch": 2.9891764705882355, "grad_norm": 0.7186136092124289, "learning_rate": 0.00019417936505849328, "loss": 0.3643, "step": 6352 }, { "epoch": 2.9896470588235293, "grad_norm": 0.5115555669774722, "learning_rate": 0.00019413747145867583, "loss": 0.383, "step": 6353 }, { "epoch": 2.9901176470588235, "grad_norm": 0.7434983733062646, "learning_rate": 0.00019409557408930184, "loss": 0.3935, "step": 6354 }, { "epoch": 2.9905882352941178, "grad_norm": 0.3827473746168448, "learning_rate": 0.0001940536729539495, "loss": 0.2752, "step": 6355 }, { "epoch": 2.9910588235294115, "grad_norm": 0.22321031003402786, "learning_rate": 0.00019401176805619746, "loss": 0.1022, "step": 6356 }, { "epoch": 2.9915294117647058, "grad_norm": 0.30963740664889305, "learning_rate": 0.0001939698593996245, "loss": 0.1404, "step": 6357 }, { "epoch": 2.992, "grad_norm": 0.16485690599304653, "learning_rate": 0.00019392794698780988, "loss": 0.0672, "step": 6358 }, { "epoch": 2.992470588235294, "grad_norm": 0.46768475543295623, "learning_rate": 0.00019388603082433315, "loss": 0.4999, "step": 6359 }, { "epoch": 2.9929411764705884, "grad_norm": 0.11471659701380249, "learning_rate": 0.00019384411091277412, "loss": 0.0495, "step": 6360 }, { "epoch": 2.993411764705882, "grad_norm": 0.24286420917824342, "learning_rate": 0.00019380218725671298, "loss": 0.1536, "step": 6361 }, { "epoch": 2.9938823529411764, "grad_norm": 0.3779214399546118, "learning_rate": 0.00019376025985973013, "loss": 0.2492, "step": 6362 }, { "epoch": 2.9943529411764707, "grad_norm": 0.17395624279283572, "learning_rate": 0.00019371832872540652, "loss": 0.0953, "step": 6363 }, { "epoch": 2.9948235294117644, "grad_norm": 0.36723508689180523, "learning_rate": 0.00019367639385732316, "loss": 0.1845, "step": 6364 }, { "epoch": 2.9952941176470587, "grad_norm": 0.33248871813849934, "learning_rate": 0.00019363445525906152, "loss": 0.2309, "step": 6365 }, { "epoch": 2.995764705882353, "grad_norm": 0.3986686709217931, "learning_rate": 0.00019359251293420336, "loss": 0.3334, "step": 6366 }, { "epoch": 2.996235294117647, "grad_norm": 0.4829465680669444, "learning_rate": 0.00019355056688633086, "loss": 0.4539, "step": 6367 }, { "epoch": 2.9967058823529413, "grad_norm": 0.10967275090865361, "learning_rate": 0.0001935086171190263, "loss": 0.0268, "step": 6368 }, { "epoch": 2.9971764705882356, "grad_norm": 0.34919817194059, "learning_rate": 0.00019346666363587238, "loss": 0.2845, "step": 6369 }, { "epoch": 2.9976470588235293, "grad_norm": 0.46556628840360154, "learning_rate": 0.00019342470644045223, "loss": 0.5536, "step": 6370 }, { "epoch": 2.9981176470588236, "grad_norm": 0.2426727391878, "learning_rate": 0.00019338274553634913, "loss": 0.1793, "step": 6371 }, { "epoch": 2.9985882352941178, "grad_norm": 0.19588372752011943, "learning_rate": 0.00019334078092714683, "loss": 0.1059, "step": 6372 }, { "epoch": 2.9990588235294116, "grad_norm": 0.09923032195936211, "learning_rate": 0.00019329881261642918, "loss": 0.0289, "step": 6373 }, { "epoch": 2.9995294117647058, "grad_norm": 0.15779781534449913, "learning_rate": 0.00019325684060778056, "loss": 0.1349, "step": 6374 }, { "epoch": 3.0, "grad_norm": 0.27224891731273504, "learning_rate": 0.00019321486490478563, "loss": 0.1404, "step": 6375 }, { "epoch": 3.0, "eval_loss": 0.24881701171398163, "eval_runtime": 65.1934, "eval_samples_per_second": 11.504, "eval_steps_per_second": 5.752, "step": 6375 }, { "epoch": 3.0004705882352942, "grad_norm": 0.41744807752320556, "learning_rate": 0.0001931728855110292, "loss": 0.3513, "step": 6376 }, { "epoch": 3.0009411764705884, "grad_norm": 0.2577212997201391, "learning_rate": 0.00019313090243009665, "loss": 0.2024, "step": 6377 }, { "epoch": 3.0014117647058822, "grad_norm": 0.32216921941676846, "learning_rate": 0.00019308891566557345, "loss": 0.2076, "step": 6378 }, { "epoch": 3.0018823529411764, "grad_norm": 0.3893284075144416, "learning_rate": 0.0001930469252210455, "loss": 0.3144, "step": 6379 }, { "epoch": 3.0023529411764707, "grad_norm": 0.236990625547593, "learning_rate": 0.000193004931100099, "loss": 0.123, "step": 6380 }, { "epoch": 3.002823529411765, "grad_norm": 0.6475109823805276, "learning_rate": 0.00019296293330632047, "loss": 0.6058, "step": 6381 }, { "epoch": 3.0032941176470587, "grad_norm": 0.22876853158930927, "learning_rate": 0.0001929209318432967, "loss": 0.179, "step": 6382 }, { "epoch": 3.003764705882353, "grad_norm": 0.4260558465815508, "learning_rate": 0.00019287892671461475, "loss": 0.2884, "step": 6383 }, { "epoch": 3.004235294117647, "grad_norm": 0.3709050345765531, "learning_rate": 0.0001928369179238622, "loss": 0.3449, "step": 6384 }, { "epoch": 3.0047058823529413, "grad_norm": 0.35217732696419685, "learning_rate": 0.00019279490547462674, "loss": 0.292, "step": 6385 }, { "epoch": 3.005176470588235, "grad_norm": 0.33205680360847434, "learning_rate": 0.0001927528893704964, "loss": 0.2305, "step": 6386 }, { "epoch": 3.0056470588235293, "grad_norm": 0.2580388442848365, "learning_rate": 0.00019271086961505964, "loss": 0.1067, "step": 6387 }, { "epoch": 3.0061176470588236, "grad_norm": 0.5004680264566644, "learning_rate": 0.0001926688462119051, "loss": 0.3139, "step": 6388 }, { "epoch": 3.006588235294118, "grad_norm": 0.6454542828692306, "learning_rate": 0.00019262681916462187, "loss": 0.541, "step": 6389 }, { "epoch": 3.0070588235294116, "grad_norm": 0.23281471785061275, "learning_rate": 0.00019258478847679914, "loss": 0.1538, "step": 6390 }, { "epoch": 3.007529411764706, "grad_norm": 0.28706696224678785, "learning_rate": 0.0001925427541520266, "loss": 0.1335, "step": 6391 }, { "epoch": 3.008, "grad_norm": 0.4553673661884408, "learning_rate": 0.00019250071619389422, "loss": 0.3017, "step": 6392 }, { "epoch": 3.0084705882352942, "grad_norm": 0.32317189471871355, "learning_rate": 0.00019245867460599217, "loss": 0.2063, "step": 6393 }, { "epoch": 3.0089411764705885, "grad_norm": 0.6236815806496017, "learning_rate": 0.00019241662939191106, "loss": 0.3325, "step": 6394 }, { "epoch": 3.0094117647058822, "grad_norm": 0.2911807045475685, "learning_rate": 0.00019237458055524178, "loss": 0.1194, "step": 6395 }, { "epoch": 3.0098823529411765, "grad_norm": 0.5721362319868357, "learning_rate": 0.00019233252809957546, "loss": 0.4645, "step": 6396 }, { "epoch": 3.0103529411764707, "grad_norm": 0.19111623465410402, "learning_rate": 0.00019229047202850357, "loss": 0.0874, "step": 6397 }, { "epoch": 3.010823529411765, "grad_norm": 0.38466936528889417, "learning_rate": 0.00019224841234561797, "loss": 0.2722, "step": 6398 }, { "epoch": 3.0112941176470587, "grad_norm": 0.7046355438659904, "learning_rate": 0.00019220634905451074, "loss": 0.5059, "step": 6399 }, { "epoch": 3.011764705882353, "grad_norm": 0.29735683085930253, "learning_rate": 0.00019216428215877425, "loss": 0.1474, "step": 6400 }, { "epoch": 3.012235294117647, "grad_norm": 0.5046770414049289, "learning_rate": 0.00019212221166200124, "loss": 0.4452, "step": 6401 }, { "epoch": 3.0127058823529413, "grad_norm": 0.2505611480342397, "learning_rate": 0.00019208013756778476, "loss": 0.0779, "step": 6402 }, { "epoch": 3.013176470588235, "grad_norm": 0.20505296360061145, "learning_rate": 0.0001920380598797181, "loss": 0.0791, "step": 6403 }, { "epoch": 3.0136470588235293, "grad_norm": 0.13247216673329812, "learning_rate": 0.000191995978601395, "loss": 0.0325, "step": 6404 }, { "epoch": 3.0141176470588236, "grad_norm": 0.5313165967039712, "learning_rate": 0.0001919538937364093, "loss": 0.3021, "step": 6405 }, { "epoch": 3.014588235294118, "grad_norm": 0.4607266039472261, "learning_rate": 0.00019191180528835527, "loss": 0.3134, "step": 6406 }, { "epoch": 3.0150588235294116, "grad_norm": 0.3623577452571169, "learning_rate": 0.00019186971326082752, "loss": 0.3133, "step": 6407 }, { "epoch": 3.015529411764706, "grad_norm": 0.45042490708603145, "learning_rate": 0.0001918276176574209, "loss": 0.3101, "step": 6408 }, { "epoch": 3.016, "grad_norm": 0.5796309769382951, "learning_rate": 0.00019178551848173054, "loss": 0.4713, "step": 6409 }, { "epoch": 3.0164705882352942, "grad_norm": 0.409261122948106, "learning_rate": 0.00019174341573735196, "loss": 0.2724, "step": 6410 }, { "epoch": 3.016941176470588, "grad_norm": 0.41349500356207025, "learning_rate": 0.0001917013094278809, "loss": 0.2993, "step": 6411 }, { "epoch": 3.0174117647058822, "grad_norm": 0.5949605217350291, "learning_rate": 0.0001916591995569135, "loss": 0.5435, "step": 6412 }, { "epoch": 3.0178823529411765, "grad_norm": 0.36482120734865287, "learning_rate": 0.00019161708612804615, "loss": 0.3004, "step": 6413 }, { "epoch": 3.0183529411764707, "grad_norm": 0.5854073167588335, "learning_rate": 0.00019157496914487545, "loss": 0.4181, "step": 6414 }, { "epoch": 3.018823529411765, "grad_norm": 0.6325793444445116, "learning_rate": 0.00019153284861099848, "loss": 0.4446, "step": 6415 }, { "epoch": 3.0192941176470587, "grad_norm": 0.24159366333681437, "learning_rate": 0.00019149072453001254, "loss": 0.1275, "step": 6416 }, { "epoch": 3.019764705882353, "grad_norm": 0.5228379276423187, "learning_rate": 0.00019144859690551518, "loss": 0.3647, "step": 6417 }, { "epoch": 3.020235294117647, "grad_norm": 0.37852849247441195, "learning_rate": 0.00019140646574110438, "loss": 0.1861, "step": 6418 }, { "epoch": 3.0207058823529414, "grad_norm": 0.2687148829951538, "learning_rate": 0.0001913643310403783, "loss": 0.1307, "step": 6419 }, { "epoch": 3.021176470588235, "grad_norm": 0.3003059022726449, "learning_rate": 0.00019132219280693544, "loss": 0.1063, "step": 6420 }, { "epoch": 3.0216470588235294, "grad_norm": 0.24827409732285813, "learning_rate": 0.00019128005104437462, "loss": 0.1202, "step": 6421 }, { "epoch": 3.0221176470588236, "grad_norm": 0.5394711527855904, "learning_rate": 0.00019123790575629503, "loss": 0.3479, "step": 6422 }, { "epoch": 3.022588235294118, "grad_norm": 0.13743806312657722, "learning_rate": 0.00019119575694629594, "loss": 0.0483, "step": 6423 }, { "epoch": 3.0230588235294116, "grad_norm": 0.3768772609038541, "learning_rate": 0.00019115360461797716, "loss": 0.2648, "step": 6424 }, { "epoch": 3.023529411764706, "grad_norm": 14.93605754074195, "learning_rate": 0.00019111144877493873, "loss": 0.9074, "step": 6425 }, { "epoch": 3.024, "grad_norm": 0.47422214335390994, "learning_rate": 0.0001910692894207809, "loss": 0.3589, "step": 6426 }, { "epoch": 3.0244705882352942, "grad_norm": 0.18397639860295806, "learning_rate": 0.00019102712655910426, "loss": 0.0612, "step": 6427 }, { "epoch": 3.024941176470588, "grad_norm": 0.25787639775256926, "learning_rate": 0.00019098496019350985, "loss": 0.0493, "step": 6428 }, { "epoch": 3.0254117647058822, "grad_norm": 10.553717940004839, "learning_rate": 0.0001909427903275988, "loss": 0.5995, "step": 6429 }, { "epoch": 3.0258823529411765, "grad_norm": 0.5073656806994705, "learning_rate": 0.00019090061696497258, "loss": 0.3742, "step": 6430 }, { "epoch": 3.0263529411764707, "grad_norm": 0.5182157633338463, "learning_rate": 0.0001908584401092331, "loss": 0.427, "step": 6431 }, { "epoch": 3.026823529411765, "grad_norm": 0.38073625016945073, "learning_rate": 0.0001908162597639824, "loss": 0.2312, "step": 6432 }, { "epoch": 3.0272941176470587, "grad_norm": 0.5294079372633785, "learning_rate": 0.00019077407593282294, "loss": 0.271, "step": 6433 }, { "epoch": 3.027764705882353, "grad_norm": 0.2990839955514404, "learning_rate": 0.00019073188861935735, "loss": 0.1884, "step": 6434 }, { "epoch": 3.028235294117647, "grad_norm": 0.31633777586536105, "learning_rate": 0.00019068969782718874, "loss": 0.2031, "step": 6435 }, { "epoch": 3.0287058823529414, "grad_norm": 0.9346090564574321, "learning_rate": 0.00019064750355992036, "loss": 0.4766, "step": 6436 }, { "epoch": 3.029176470588235, "grad_norm": 0.41279693170290993, "learning_rate": 0.00019060530582115578, "loss": 0.307, "step": 6437 }, { "epoch": 3.0296470588235294, "grad_norm": 0.31865804477961224, "learning_rate": 0.00019056310461449885, "loss": 0.2533, "step": 6438 }, { "epoch": 3.0301176470588236, "grad_norm": 0.5777165251263486, "learning_rate": 0.00019052089994355394, "loss": 0.3169, "step": 6439 }, { "epoch": 3.030588235294118, "grad_norm": 0.5297421418204108, "learning_rate": 0.00019047869181192535, "loss": 0.4152, "step": 6440 }, { "epoch": 3.0310588235294116, "grad_norm": 0.36008536653866713, "learning_rate": 0.00019043648022321793, "loss": 0.1637, "step": 6441 }, { "epoch": 3.031529411764706, "grad_norm": 0.19537726461262397, "learning_rate": 0.00019039426518103678, "loss": 0.0655, "step": 6442 }, { "epoch": 3.032, "grad_norm": 0.6484962580389507, "learning_rate": 0.00019035204668898725, "loss": 0.4927, "step": 6443 }, { "epoch": 3.0324705882352943, "grad_norm": 0.5226289381369633, "learning_rate": 0.00019030982475067498, "loss": 0.3063, "step": 6444 }, { "epoch": 3.032941176470588, "grad_norm": 0.2271126681721761, "learning_rate": 0.00019026759936970598, "loss": 0.1082, "step": 6445 }, { "epoch": 3.0334117647058823, "grad_norm": 0.5248970560593925, "learning_rate": 0.0001902253705496865, "loss": 0.3612, "step": 6446 }, { "epoch": 3.0338823529411765, "grad_norm": 0.559451545577526, "learning_rate": 0.00019018313829422307, "loss": 0.404, "step": 6447 }, { "epoch": 3.0343529411764707, "grad_norm": 0.22798937575971903, "learning_rate": 0.0001901409026069225, "loss": 0.1568, "step": 6448 }, { "epoch": 3.034823529411765, "grad_norm": 0.41475697110154935, "learning_rate": 0.000190098663491392, "loss": 0.344, "step": 6449 }, { "epoch": 3.0352941176470587, "grad_norm": 0.4914571901152938, "learning_rate": 0.00019005642095123895, "loss": 0.3797, "step": 6450 }, { "epoch": 3.035764705882353, "grad_norm": 0.2363046325197706, "learning_rate": 0.00019001417499007103, "loss": 0.1238, "step": 6451 }, { "epoch": 3.036235294117647, "grad_norm": 0.43691949001334374, "learning_rate": 0.0001899719256114963, "loss": 0.3217, "step": 6452 }, { "epoch": 3.0367058823529414, "grad_norm": 0.3725962670779063, "learning_rate": 0.00018992967281912312, "loss": 0.2079, "step": 6453 }, { "epoch": 3.037176470588235, "grad_norm": 0.4141976280245599, "learning_rate": 0.00018988741661655996, "loss": 0.2822, "step": 6454 }, { "epoch": 3.0376470588235294, "grad_norm": 0.46160211532749595, "learning_rate": 0.00018984515700741579, "loss": 0.525, "step": 6455 }, { "epoch": 3.0381176470588236, "grad_norm": 0.30089804442490214, "learning_rate": 0.0001898028939952998, "loss": 0.257, "step": 6456 }, { "epoch": 3.038588235294118, "grad_norm": 0.4945708638200833, "learning_rate": 0.00018976062758382144, "loss": 0.3605, "step": 6457 }, { "epoch": 3.0390588235294116, "grad_norm": 0.3937700503244403, "learning_rate": 0.0001897183577765904, "loss": 0.2739, "step": 6458 }, { "epoch": 3.039529411764706, "grad_norm": 0.1824198933646547, "learning_rate": 0.00018967608457721684, "loss": 0.14, "step": 6459 }, { "epoch": 3.04, "grad_norm": 0.27156372710815224, "learning_rate": 0.00018963380798931102, "loss": 0.111, "step": 6460 }, { "epoch": 3.0404705882352943, "grad_norm": 0.36031465721928363, "learning_rate": 0.00018959152801648364, "loss": 0.1152, "step": 6461 }, { "epoch": 3.040941176470588, "grad_norm": 0.29600781442243584, "learning_rate": 0.00018954924466234554, "loss": 0.2424, "step": 6462 }, { "epoch": 3.0414117647058823, "grad_norm": 0.4475075619658828, "learning_rate": 0.00018950695793050796, "loss": 0.2122, "step": 6463 }, { "epoch": 3.0418823529411765, "grad_norm": 0.33596702813302187, "learning_rate": 0.0001894646678245824, "loss": 0.1522, "step": 6464 }, { "epoch": 3.0423529411764707, "grad_norm": 0.5032598297300368, "learning_rate": 0.00018942237434818062, "loss": 0.3493, "step": 6465 }, { "epoch": 3.0428235294117645, "grad_norm": 0.3354856311625533, "learning_rate": 0.00018938007750491468, "loss": 0.2073, "step": 6466 }, { "epoch": 3.0432941176470587, "grad_norm": 0.4518078609563663, "learning_rate": 0.00018933777729839701, "loss": 0.2354, "step": 6467 }, { "epoch": 3.043764705882353, "grad_norm": 0.5303579053513967, "learning_rate": 0.0001892954737322402, "loss": 0.2847, "step": 6468 }, { "epoch": 3.044235294117647, "grad_norm": 0.5434048621832522, "learning_rate": 0.00018925316681005718, "loss": 0.3787, "step": 6469 }, { "epoch": 3.0447058823529414, "grad_norm": 0.3108515944252475, "learning_rate": 0.0001892108565354612, "loss": 0.2959, "step": 6470 }, { "epoch": 3.045176470588235, "grad_norm": 0.3447814533708418, "learning_rate": 0.0001891685429120657, "loss": 0.3066, "step": 6471 }, { "epoch": 3.0456470588235294, "grad_norm": 0.6200458316739625, "learning_rate": 0.00018912622594348454, "loss": 0.5586, "step": 6472 }, { "epoch": 3.0461176470588236, "grad_norm": 8.69128367253282, "learning_rate": 0.00018908390563333175, "loss": 0.5958, "step": 6473 }, { "epoch": 3.046588235294118, "grad_norm": 0.47731992452379385, "learning_rate": 0.00018904158198522177, "loss": 0.4105, "step": 6474 }, { "epoch": 3.0470588235294116, "grad_norm": 0.3282006443406858, "learning_rate": 0.00018899925500276912, "loss": 0.2219, "step": 6475 }, { "epoch": 3.047529411764706, "grad_norm": 0.29087695978616707, "learning_rate": 0.00018895692468958884, "loss": 0.1848, "step": 6476 }, { "epoch": 3.048, "grad_norm": 0.3770208652081194, "learning_rate": 0.00018891459104929608, "loss": 0.1494, "step": 6477 }, { "epoch": 3.0484705882352943, "grad_norm": 0.3357964597436875, "learning_rate": 0.0001888722540855064, "loss": 0.1998, "step": 6478 }, { "epoch": 3.048941176470588, "grad_norm": 0.383650853921726, "learning_rate": 0.00018882991380183556, "loss": 0.1938, "step": 6479 }, { "epoch": 3.0494117647058823, "grad_norm": 0.600934891136812, "learning_rate": 0.0001887875702018996, "loss": 0.3592, "step": 6480 }, { "epoch": 3.0498823529411765, "grad_norm": 0.5105620120564837, "learning_rate": 0.00018874522328931492, "loss": 0.5478, "step": 6481 }, { "epoch": 3.0503529411764707, "grad_norm": 0.36907275320186106, "learning_rate": 0.0001887028730676981, "loss": 0.2389, "step": 6482 }, { "epoch": 3.0508235294117645, "grad_norm": 0.320758099823578, "learning_rate": 0.00018866051954066608, "loss": 0.1233, "step": 6483 }, { "epoch": 3.0512941176470587, "grad_norm": 0.23303731882257225, "learning_rate": 0.00018861816271183613, "loss": 0.1269, "step": 6484 }, { "epoch": 3.051764705882353, "grad_norm": 0.4074642530664053, "learning_rate": 0.00018857580258482557, "loss": 0.379, "step": 6485 }, { "epoch": 3.052235294117647, "grad_norm": 0.6503700696019822, "learning_rate": 0.0001885334391632523, "loss": 0.3452, "step": 6486 }, { "epoch": 3.0527058823529414, "grad_norm": 0.3765147263815219, "learning_rate": 0.00018849107245073432, "loss": 0.3147, "step": 6487 }, { "epoch": 3.053176470588235, "grad_norm": 0.5804907157415841, "learning_rate": 0.00018844870245088996, "loss": 0.4451, "step": 6488 }, { "epoch": 3.0536470588235294, "grad_norm": 0.47834824528462727, "learning_rate": 0.00018840632916733777, "loss": 0.2752, "step": 6489 }, { "epoch": 3.0541176470588236, "grad_norm": 0.19273227217043346, "learning_rate": 0.00018836395260369676, "loss": 0.0897, "step": 6490 }, { "epoch": 3.054588235294118, "grad_norm": 0.22534078320028172, "learning_rate": 0.000188321572763586, "loss": 0.0813, "step": 6491 }, { "epoch": 3.0550588235294116, "grad_norm": 0.17432407287906737, "learning_rate": 0.00018827918965062492, "loss": 0.0611, "step": 6492 }, { "epoch": 3.055529411764706, "grad_norm": 0.35507144840414273, "learning_rate": 0.00018823680326843334, "loss": 0.2828, "step": 6493 }, { "epoch": 3.056, "grad_norm": 0.20707817967548647, "learning_rate": 0.00018819441362063122, "loss": 0.1524, "step": 6494 }, { "epoch": 3.0564705882352943, "grad_norm": 0.19885439599378305, "learning_rate": 0.00018815202071083878, "loss": 0.0785, "step": 6495 }, { "epoch": 3.056941176470588, "grad_norm": 0.42828593884966487, "learning_rate": 0.00018810962454267666, "loss": 0.2377, "step": 6496 }, { "epoch": 3.0574117647058823, "grad_norm": 0.30795683220006176, "learning_rate": 0.00018806722511976567, "loss": 0.1871, "step": 6497 }, { "epoch": 3.0578823529411765, "grad_norm": 0.46574255085790817, "learning_rate": 0.00018802482244572693, "loss": 0.3506, "step": 6498 }, { "epoch": 3.0583529411764707, "grad_norm": 0.2932074162487903, "learning_rate": 0.00018798241652418182, "loss": 0.1051, "step": 6499 }, { "epoch": 3.0588235294117645, "grad_norm": 0.39144310040501334, "learning_rate": 0.00018794000735875208, "loss": 0.2201, "step": 6500 }, { "epoch": 3.0592941176470587, "grad_norm": 0.4293012874705562, "learning_rate": 0.00018789759495305963, "loss": 0.4115, "step": 6501 }, { "epoch": 3.059764705882353, "grad_norm": 0.36206092524309513, "learning_rate": 0.00018785517931072668, "loss": 0.2834, "step": 6502 }, { "epoch": 3.060235294117647, "grad_norm": 0.30325816604189454, "learning_rate": 0.0001878127604353757, "loss": 0.233, "step": 6503 }, { "epoch": 3.0607058823529414, "grad_norm": 0.43186829166755925, "learning_rate": 0.00018777033833062954, "loss": 0.3204, "step": 6504 }, { "epoch": 3.061176470588235, "grad_norm": 0.3514357829660448, "learning_rate": 0.00018772791300011122, "loss": 0.1858, "step": 6505 }, { "epoch": 3.0616470588235294, "grad_norm": 0.2741648465779214, "learning_rate": 0.00018768548444744404, "loss": 0.1024, "step": 6506 }, { "epoch": 3.0621176470588236, "grad_norm": 0.42662070040737865, "learning_rate": 0.00018764305267625169, "loss": 0.2796, "step": 6507 }, { "epoch": 3.062588235294118, "grad_norm": 0.17169895195188437, "learning_rate": 0.000187600617690158, "loss": 0.092, "step": 6508 }, { "epoch": 3.0630588235294116, "grad_norm": 0.3162951486072918, "learning_rate": 0.00018755817949278706, "loss": 0.1216, "step": 6509 }, { "epoch": 3.063529411764706, "grad_norm": 0.4983720746358517, "learning_rate": 0.00018751573808776346, "loss": 0.4966, "step": 6510 }, { "epoch": 3.064, "grad_norm": 0.5286130816586803, "learning_rate": 0.00018747329347871178, "loss": 0.4342, "step": 6511 }, { "epoch": 3.0644705882352943, "grad_norm": 0.3632262336987869, "learning_rate": 0.00018743084566925702, "loss": 0.2365, "step": 6512 }, { "epoch": 3.064941176470588, "grad_norm": 0.4932387588016492, "learning_rate": 0.00018738839466302443, "loss": 0.3461, "step": 6513 }, { "epoch": 3.0654117647058823, "grad_norm": 0.28092247410024385, "learning_rate": 0.00018734594046363956, "loss": 0.1338, "step": 6514 }, { "epoch": 3.0658823529411765, "grad_norm": 0.27814571143650346, "learning_rate": 0.00018730348307472824, "loss": 0.2395, "step": 6515 }, { "epoch": 3.0663529411764707, "grad_norm": 0.23923029488161657, "learning_rate": 0.0001872610224999164, "loss": 0.1575, "step": 6516 }, { "epoch": 3.0668235294117645, "grad_norm": 0.4414827571261567, "learning_rate": 0.00018721855874283058, "loss": 0.3793, "step": 6517 }, { "epoch": 3.0672941176470587, "grad_norm": 0.43958956592650905, "learning_rate": 0.00018717609180709721, "loss": 0.3314, "step": 6518 }, { "epoch": 3.067764705882353, "grad_norm": 0.43999675900136426, "learning_rate": 0.00018713362169634325, "loss": 0.3031, "step": 6519 }, { "epoch": 3.068235294117647, "grad_norm": 0.3992302419964135, "learning_rate": 0.00018709114841419588, "loss": 0.3474, "step": 6520 }, { "epoch": 3.0687058823529414, "grad_norm": 0.33475754389532064, "learning_rate": 0.0001870486719642825, "loss": 0.2019, "step": 6521 }, { "epoch": 3.069176470588235, "grad_norm": 0.38729410390711144, "learning_rate": 0.00018700619235023085, "loss": 0.3335, "step": 6522 }, { "epoch": 3.0696470588235294, "grad_norm": 0.3723451026422911, "learning_rate": 0.00018696370957566882, "loss": 0.3091, "step": 6523 }, { "epoch": 3.0701176470588236, "grad_norm": 0.2180943536231947, "learning_rate": 0.0001869212236442247, "loss": 0.1696, "step": 6524 }, { "epoch": 3.070588235294118, "grad_norm": 0.5024024570426424, "learning_rate": 0.00018687873455952697, "loss": 0.3169, "step": 6525 }, { "epoch": 3.0710588235294116, "grad_norm": 0.2013039693837278, "learning_rate": 0.00018683624232520446, "loss": 0.0951, "step": 6526 }, { "epoch": 3.071529411764706, "grad_norm": 0.2888993507060973, "learning_rate": 0.00018679374694488608, "loss": 0.1522, "step": 6527 }, { "epoch": 3.072, "grad_norm": 0.3416754253240858, "learning_rate": 0.00018675124842220135, "loss": 0.3533, "step": 6528 }, { "epoch": 3.0724705882352943, "grad_norm": 0.29510780048557056, "learning_rate": 0.0001867087467607797, "loss": 0.1403, "step": 6529 }, { "epoch": 3.072941176470588, "grad_norm": 0.4888656138290324, "learning_rate": 0.00018666624196425098, "loss": 0.2781, "step": 6530 }, { "epoch": 3.0734117647058823, "grad_norm": 0.5175839754333158, "learning_rate": 0.00018662373403624538, "loss": 0.3044, "step": 6531 }, { "epoch": 3.0738823529411765, "grad_norm": 0.371063574284946, "learning_rate": 0.00018658122298039327, "loss": 0.341, "step": 6532 }, { "epoch": 3.0743529411764707, "grad_norm": 0.16084420923828105, "learning_rate": 0.00018653870880032527, "loss": 0.0637, "step": 6533 }, { "epoch": 3.0748235294117645, "grad_norm": 0.3614397114879638, "learning_rate": 0.00018649619149967232, "loss": 0.2854, "step": 6534 }, { "epoch": 3.0752941176470587, "grad_norm": 0.4105777709842464, "learning_rate": 0.00018645367108206565, "loss": 0.3487, "step": 6535 }, { "epoch": 3.075764705882353, "grad_norm": 0.3519948543906951, "learning_rate": 0.00018641114755113658, "loss": 0.2801, "step": 6536 }, { "epoch": 3.076235294117647, "grad_norm": 0.3039079927252951, "learning_rate": 0.00018636862091051697, "loss": 0.096, "step": 6537 }, { "epoch": 3.0767058823529414, "grad_norm": 0.299443807355517, "learning_rate": 0.00018632609116383874, "loss": 0.2162, "step": 6538 }, { "epoch": 3.077176470588235, "grad_norm": 0.20450823111891825, "learning_rate": 0.00018628355831473416, "loss": 0.0904, "step": 6539 }, { "epoch": 3.0776470588235294, "grad_norm": 0.43084968660791023, "learning_rate": 0.00018624102236683567, "loss": 0.306, "step": 6540 }, { "epoch": 3.0781176470588236, "grad_norm": 0.5409217652365909, "learning_rate": 0.00018619848332377615, "loss": 0.4595, "step": 6541 }, { "epoch": 3.078588235294118, "grad_norm": 0.36640075989956067, "learning_rate": 0.00018615594118918862, "loss": 0.2643, "step": 6542 }, { "epoch": 3.0790588235294116, "grad_norm": 0.2599379621155461, "learning_rate": 0.00018611339596670632, "loss": 0.241, "step": 6543 }, { "epoch": 3.079529411764706, "grad_norm": 0.3785037057626756, "learning_rate": 0.00018607084765996285, "loss": 0.2239, "step": 6544 }, { "epoch": 3.08, "grad_norm": 0.46264550408018384, "learning_rate": 0.00018602829627259213, "loss": 0.2704, "step": 6545 }, { "epoch": 3.0804705882352943, "grad_norm": 0.29958870322023373, "learning_rate": 0.00018598574180822818, "loss": 0.1847, "step": 6546 }, { "epoch": 3.080941176470588, "grad_norm": 0.37970125258418697, "learning_rate": 0.00018594318427050537, "loss": 0.2457, "step": 6547 }, { "epoch": 3.0814117647058823, "grad_norm": 0.2119597093145403, "learning_rate": 0.00018590062366305828, "loss": 0.1455, "step": 6548 }, { "epoch": 3.0818823529411765, "grad_norm": 0.2784386283851651, "learning_rate": 0.0001858580599895219, "loss": 0.1698, "step": 6549 }, { "epoch": 3.0823529411764707, "grad_norm": 0.22840385885234268, "learning_rate": 0.00018581549325353126, "loss": 0.1532, "step": 6550 }, { "epoch": 3.0828235294117645, "grad_norm": 0.2318289491544347, "learning_rate": 0.00018577292345872188, "loss": 0.1228, "step": 6551 }, { "epoch": 3.0832941176470587, "grad_norm": 0.44626021667497284, "learning_rate": 0.00018573035060872933, "loss": 0.3149, "step": 6552 }, { "epoch": 3.083764705882353, "grad_norm": 0.3049240012472342, "learning_rate": 0.0001856877747071896, "loss": 0.1911, "step": 6553 }, { "epoch": 3.084235294117647, "grad_norm": 0.38818494348271443, "learning_rate": 0.00018564519575773884, "loss": 0.2691, "step": 6554 }, { "epoch": 3.084705882352941, "grad_norm": 0.2749014075468508, "learning_rate": 0.00018560261376401356, "loss": 0.2727, "step": 6555 }, { "epoch": 3.085176470588235, "grad_norm": 0.4498858280566651, "learning_rate": 0.0001855600287296504, "loss": 0.4139, "step": 6556 }, { "epoch": 3.0856470588235294, "grad_norm": 0.45355288100510177, "learning_rate": 0.00018551744065828634, "loss": 0.3449, "step": 6557 }, { "epoch": 3.0861176470588236, "grad_norm": 0.5216882966526565, "learning_rate": 0.0001854748495535587, "loss": 0.383, "step": 6558 }, { "epoch": 3.086588235294118, "grad_norm": 0.29652178881360974, "learning_rate": 0.0001854322554191049, "loss": 0.1382, "step": 6559 }, { "epoch": 3.0870588235294116, "grad_norm": 0.5396366702860756, "learning_rate": 0.00018538965825856264, "loss": 0.5176, "step": 6560 }, { "epoch": 3.087529411764706, "grad_norm": 0.29908348813230134, "learning_rate": 0.00018534705807557003, "loss": 0.2004, "step": 6561 }, { "epoch": 3.088, "grad_norm": 0.3166018834664014, "learning_rate": 0.00018530445487376524, "loss": 0.1653, "step": 6562 }, { "epoch": 3.0884705882352943, "grad_norm": 0.5221377045470573, "learning_rate": 0.0001852618486567869, "loss": 0.4757, "step": 6563 }, { "epoch": 3.088941176470588, "grad_norm": 0.3434372085844414, "learning_rate": 0.00018521923942827362, "loss": 0.1915, "step": 6564 }, { "epoch": 3.0894117647058823, "grad_norm": 0.4356281345779731, "learning_rate": 0.00018517662719186464, "loss": 0.3663, "step": 6565 }, { "epoch": 3.0898823529411765, "grad_norm": 0.3752768193340773, "learning_rate": 0.0001851340119511991, "loss": 0.3193, "step": 6566 }, { "epoch": 3.0903529411764707, "grad_norm": 0.3785245218451673, "learning_rate": 0.00018509139370991666, "loss": 0.3335, "step": 6567 }, { "epoch": 3.0908235294117645, "grad_norm": 0.23277848511812665, "learning_rate": 0.000185048772471657, "loss": 0.1151, "step": 6568 }, { "epoch": 3.0912941176470587, "grad_norm": 0.4115130723834498, "learning_rate": 0.00018500614824006035, "loss": 0.2658, "step": 6569 }, { "epoch": 3.091764705882353, "grad_norm": 0.37800703757281146, "learning_rate": 0.0001849635210187669, "loss": 0.4001, "step": 6570 }, { "epoch": 3.092235294117647, "grad_norm": 0.22342603275646877, "learning_rate": 0.00018492089081141724, "loss": 0.1492, "step": 6571 }, { "epoch": 3.092705882352941, "grad_norm": 0.2518828379812898, "learning_rate": 0.00018487825762165223, "loss": 0.1171, "step": 6572 }, { "epoch": 3.093176470588235, "grad_norm": 0.5223038756491826, "learning_rate": 0.00018483562145311296, "loss": 0.4435, "step": 6573 }, { "epoch": 3.0936470588235294, "grad_norm": 0.43973850892386884, "learning_rate": 0.00018479298230944071, "loss": 0.1884, "step": 6574 }, { "epoch": 3.0941176470588236, "grad_norm": 0.37325811772285605, "learning_rate": 0.00018475034019427717, "loss": 0.189, "step": 6575 }, { "epoch": 3.094588235294118, "grad_norm": 0.3869867356831351, "learning_rate": 0.00018470769511126409, "loss": 0.307, "step": 6576 }, { "epoch": 3.0950588235294116, "grad_norm": 0.26346208979759017, "learning_rate": 0.00018466504706404362, "loss": 0.1525, "step": 6577 }, { "epoch": 3.095529411764706, "grad_norm": 0.3082224441211216, "learning_rate": 0.00018462239605625808, "loss": 0.2199, "step": 6578 }, { "epoch": 3.096, "grad_norm": 0.7311194566586148, "learning_rate": 0.0001845797420915501, "loss": 0.527, "step": 6579 }, { "epoch": 3.0964705882352943, "grad_norm": 0.5992516761506075, "learning_rate": 0.00018453708517356256, "loss": 0.7732, "step": 6580 }, { "epoch": 3.096941176470588, "grad_norm": 0.3955143659291468, "learning_rate": 0.0001844944253059385, "loss": 0.3611, "step": 6581 }, { "epoch": 3.0974117647058823, "grad_norm": 0.0758944104031964, "learning_rate": 0.00018445176249232134, "loss": 0.0145, "step": 6582 }, { "epoch": 3.0978823529411765, "grad_norm": 0.38101865085953346, "learning_rate": 0.0001844090967363547, "loss": 0.2642, "step": 6583 }, { "epoch": 3.0983529411764708, "grad_norm": 0.4318836543423894, "learning_rate": 0.0001843664280416824, "loss": 0.2736, "step": 6584 }, { "epoch": 3.0988235294117645, "grad_norm": 0.31075574370471587, "learning_rate": 0.0001843237564119486, "loss": 0.1602, "step": 6585 }, { "epoch": 3.0992941176470588, "grad_norm": 0.6495189788751184, "learning_rate": 0.00018428108185079763, "loss": 0.5069, "step": 6586 }, { "epoch": 3.099764705882353, "grad_norm": 0.38005921302545426, "learning_rate": 0.00018423840436187412, "loss": 0.1528, "step": 6587 }, { "epoch": 3.100235294117647, "grad_norm": 0.33307278900147547, "learning_rate": 0.0001841957239488229, "loss": 0.2396, "step": 6588 }, { "epoch": 3.100705882352941, "grad_norm": 0.34676004415652684, "learning_rate": 0.00018415304061528918, "loss": 0.2678, "step": 6589 }, { "epoch": 3.101176470588235, "grad_norm": 0.31943659509856953, "learning_rate": 0.00018411035436491827, "loss": 0.1944, "step": 6590 }, { "epoch": 3.1016470588235294, "grad_norm": 0.4175525474113326, "learning_rate": 0.0001840676652013558, "loss": 0.3066, "step": 6591 }, { "epoch": 3.1021176470588236, "grad_norm": 0.3899998759341908, "learning_rate": 0.00018402497312824758, "loss": 0.1965, "step": 6592 }, { "epoch": 3.1025882352941174, "grad_norm": 0.37130339989780764, "learning_rate": 0.0001839822781492398, "loss": 0.3357, "step": 6593 }, { "epoch": 3.1030588235294116, "grad_norm": 0.22462837671275623, "learning_rate": 0.00018393958026797878, "loss": 0.1204, "step": 6594 }, { "epoch": 3.103529411764706, "grad_norm": 0.2538968355659979, "learning_rate": 0.0001838968794881111, "loss": 0.164, "step": 6595 }, { "epoch": 3.104, "grad_norm": 0.9804308541899314, "learning_rate": 0.00018385417581328364, "loss": 0.4551, "step": 6596 }, { "epoch": 3.1044705882352943, "grad_norm": 0.26087872112452104, "learning_rate": 0.00018381146924714356, "loss": 0.1206, "step": 6597 }, { "epoch": 3.104941176470588, "grad_norm": 0.5344131493041709, "learning_rate": 0.00018376875979333812, "loss": 0.2506, "step": 6598 }, { "epoch": 3.1054117647058823, "grad_norm": 0.37077087379735846, "learning_rate": 0.000183726047455515, "loss": 0.1618, "step": 6599 }, { "epoch": 3.1058823529411765, "grad_norm": 0.9717389468275521, "learning_rate": 0.000183683332237322, "loss": 0.4071, "step": 6600 }, { "epoch": 3.1063529411764708, "grad_norm": 0.3258220452202024, "learning_rate": 0.00018364061414240717, "loss": 0.1521, "step": 6601 }, { "epoch": 3.1068235294117645, "grad_norm": 0.3090867118082994, "learning_rate": 0.00018359789317441892, "loss": 0.1326, "step": 6602 }, { "epoch": 3.1072941176470588, "grad_norm": 0.3804620140208724, "learning_rate": 0.00018355516933700578, "loss": 0.3068, "step": 6603 }, { "epoch": 3.107764705882353, "grad_norm": 0.35599832782341734, "learning_rate": 0.0001835124426338166, "loss": 0.2894, "step": 6604 }, { "epoch": 3.108235294117647, "grad_norm": 0.3312935920852753, "learning_rate": 0.00018346971306850045, "loss": 0.1708, "step": 6605 }, { "epoch": 3.108705882352941, "grad_norm": 0.2641637393861711, "learning_rate": 0.00018342698064470662, "loss": 0.0819, "step": 6606 }, { "epoch": 3.109176470588235, "grad_norm": 0.5625548682788172, "learning_rate": 0.0001833842453660847, "loss": 0.2531, "step": 6607 }, { "epoch": 3.1096470588235294, "grad_norm": 0.20406784871558659, "learning_rate": 0.00018334150723628442, "loss": 0.0524, "step": 6608 }, { "epoch": 3.1101176470588237, "grad_norm": 0.12142658498430357, "learning_rate": 0.00018329876625895592, "loss": 0.0437, "step": 6609 }, { "epoch": 3.1105882352941174, "grad_norm": 0.32213871258429555, "learning_rate": 0.0001832560224377495, "loss": 0.2454, "step": 6610 }, { "epoch": 3.1110588235294117, "grad_norm": 0.49830115342738807, "learning_rate": 0.0001832132757763156, "loss": 0.3513, "step": 6611 }, { "epoch": 3.111529411764706, "grad_norm": 0.35928656583708435, "learning_rate": 0.000183170526278305, "loss": 0.2358, "step": 6612 }, { "epoch": 3.112, "grad_norm": 0.2261073712910074, "learning_rate": 0.0001831277739473688, "loss": 0.1267, "step": 6613 }, { "epoch": 3.1124705882352943, "grad_norm": 0.36254621798751674, "learning_rate": 0.00018308501878715823, "loss": 0.3546, "step": 6614 }, { "epoch": 3.112941176470588, "grad_norm": 0.6350461517764109, "learning_rate": 0.00018304226080132472, "loss": 0.5372, "step": 6615 }, { "epoch": 3.1134117647058823, "grad_norm": 0.373860685025843, "learning_rate": 0.00018299949999352013, "loss": 0.2671, "step": 6616 }, { "epoch": 3.1138823529411765, "grad_norm": 0.3209048112502129, "learning_rate": 0.00018295673636739635, "loss": 0.2265, "step": 6617 }, { "epoch": 3.1143529411764708, "grad_norm": 0.3579075128470308, "learning_rate": 0.00018291396992660565, "loss": 0.253, "step": 6618 }, { "epoch": 3.1148235294117645, "grad_norm": 0.3648774053797063, "learning_rate": 0.00018287120067480046, "loss": 0.249, "step": 6619 }, { "epoch": 3.1152941176470588, "grad_norm": 0.39007299012241253, "learning_rate": 0.0001828284286156335, "loss": 0.4103, "step": 6620 }, { "epoch": 3.115764705882353, "grad_norm": 0.3001783569090046, "learning_rate": 0.0001827856537527577, "loss": 0.1749, "step": 6621 }, { "epoch": 3.116235294117647, "grad_norm": 0.4336358634450322, "learning_rate": 0.00018274287608982626, "loss": 0.3786, "step": 6622 }, { "epoch": 3.116705882352941, "grad_norm": 0.2712060337022034, "learning_rate": 0.00018270009563049264, "loss": 0.1258, "step": 6623 }, { "epoch": 3.117176470588235, "grad_norm": 0.34538606490840057, "learning_rate": 0.00018265731237841043, "loss": 0.1508, "step": 6624 }, { "epoch": 3.1176470588235294, "grad_norm": 0.44388595093325006, "learning_rate": 0.00018261452633723356, "loss": 0.2522, "step": 6625 }, { "epoch": 3.1181176470588237, "grad_norm": 0.2667198273287338, "learning_rate": 0.00018257173751061613, "loss": 0.1981, "step": 6626 }, { "epoch": 3.1185882352941174, "grad_norm": 0.8427268556055493, "learning_rate": 0.00018252894590221262, "loss": 0.3063, "step": 6627 }, { "epoch": 3.1190588235294117, "grad_norm": 0.32586391475037296, "learning_rate": 0.00018248615151567754, "loss": 0.2657, "step": 6628 }, { "epoch": 3.119529411764706, "grad_norm": 0.3907042371224951, "learning_rate": 0.00018244335435466577, "loss": 0.41, "step": 6629 }, { "epoch": 3.12, "grad_norm": 0.3596127632099503, "learning_rate": 0.00018240055442283242, "loss": 0.3789, "step": 6630 }, { "epoch": 3.1204705882352943, "grad_norm": 0.3899863398393278, "learning_rate": 0.00018235775172383277, "loss": 0.3547, "step": 6631 }, { "epoch": 3.120941176470588, "grad_norm": 0.4276614193518877, "learning_rate": 0.00018231494626132243, "loss": 0.3355, "step": 6632 }, { "epoch": 3.1214117647058823, "grad_norm": 0.25692977183636384, "learning_rate": 0.00018227213803895717, "loss": 0.093, "step": 6633 }, { "epoch": 3.1218823529411766, "grad_norm": 0.2801385322473976, "learning_rate": 0.00018222932706039307, "loss": 0.1876, "step": 6634 }, { "epoch": 3.1223529411764708, "grad_norm": 0.4553711194760462, "learning_rate": 0.0001821865133292863, "loss": 0.3753, "step": 6635 }, { "epoch": 3.1228235294117646, "grad_norm": 0.3194948402509522, "learning_rate": 0.00018214369684929347, "loss": 0.2835, "step": 6636 }, { "epoch": 3.1232941176470588, "grad_norm": 0.36912838835196116, "learning_rate": 0.00018210087762407122, "loss": 0.3355, "step": 6637 }, { "epoch": 3.123764705882353, "grad_norm": 0.303898298044846, "learning_rate": 0.00018205805565727658, "loss": 0.1705, "step": 6638 }, { "epoch": 3.124235294117647, "grad_norm": 0.4019018448282908, "learning_rate": 0.00018201523095256672, "loss": 0.3378, "step": 6639 }, { "epoch": 3.124705882352941, "grad_norm": 0.3721011635427063, "learning_rate": 0.00018197240351359913, "loss": 0.3303, "step": 6640 }, { "epoch": 3.1251764705882352, "grad_norm": 0.21097481048102898, "learning_rate": 0.0001819295733440315, "loss": 0.0454, "step": 6641 }, { "epoch": 3.1256470588235294, "grad_norm": 0.3735983589047529, "learning_rate": 0.00018188674044752163, "loss": 0.3165, "step": 6642 }, { "epoch": 3.1261176470588237, "grad_norm": 0.29989310416163084, "learning_rate": 0.00018184390482772774, "loss": 0.2545, "step": 6643 }, { "epoch": 3.1265882352941174, "grad_norm": 0.439424823381995, "learning_rate": 0.0001818010664883082, "loss": 0.3659, "step": 6644 }, { "epoch": 3.1270588235294117, "grad_norm": 0.4170205432678292, "learning_rate": 0.0001817582254329216, "loss": 0.2319, "step": 6645 }, { "epoch": 3.127529411764706, "grad_norm": 0.20544277066038233, "learning_rate": 0.0001817153816652267, "loss": 0.0688, "step": 6646 }, { "epoch": 3.128, "grad_norm": 0.27887975655306957, "learning_rate": 0.0001816725351888827, "loss": 0.1092, "step": 6647 }, { "epoch": 3.1284705882352943, "grad_norm": 0.4066013738981765, "learning_rate": 0.00018162968600754888, "loss": 0.263, "step": 6648 }, { "epoch": 3.128941176470588, "grad_norm": 0.27583627910379305, "learning_rate": 0.00018158683412488466, "loss": 0.1522, "step": 6649 }, { "epoch": 3.1294117647058823, "grad_norm": 0.8847566843797249, "learning_rate": 0.00018154397954454993, "loss": 0.4641, "step": 6650 }, { "epoch": 3.1298823529411766, "grad_norm": 0.7477311799758479, "learning_rate": 0.0001815011222702046, "loss": 0.4088, "step": 6651 }, { "epoch": 3.130352941176471, "grad_norm": 0.284824325446686, "learning_rate": 0.00018145826230550892, "loss": 0.1881, "step": 6652 }, { "epoch": 3.1308235294117646, "grad_norm": 0.06621456968924627, "learning_rate": 0.00018141539965412334, "loss": 0.0181, "step": 6653 }, { "epoch": 3.131294117647059, "grad_norm": 0.3866833298613768, "learning_rate": 0.00018137253431970853, "loss": 0.3512, "step": 6654 }, { "epoch": 3.131764705882353, "grad_norm": 0.20956312201081312, "learning_rate": 0.0001813296663059254, "loss": 0.1692, "step": 6655 }, { "epoch": 3.1322352941176472, "grad_norm": 0.5183889083574922, "learning_rate": 0.00018128679561643508, "loss": 0.4832, "step": 6656 }, { "epoch": 3.132705882352941, "grad_norm": 0.43198239946054356, "learning_rate": 0.000181243922254899, "loss": 0.2596, "step": 6657 }, { "epoch": 3.1331764705882352, "grad_norm": 0.32719518467247505, "learning_rate": 0.0001812010462249787, "loss": 0.2462, "step": 6658 }, { "epoch": 3.1336470588235295, "grad_norm": 0.17496636249653033, "learning_rate": 0.00018115816753033601, "loss": 0.0709, "step": 6659 }, { "epoch": 3.1341176470588237, "grad_norm": 0.4623260140911043, "learning_rate": 0.00018111528617463296, "loss": 0.3453, "step": 6660 }, { "epoch": 3.1345882352941175, "grad_norm": 0.6282417807792987, "learning_rate": 0.00018107240216153184, "loss": 0.3099, "step": 6661 }, { "epoch": 3.1350588235294117, "grad_norm": 0.4087552520102273, "learning_rate": 0.00018102951549469522, "loss": 0.3898, "step": 6662 }, { "epoch": 3.135529411764706, "grad_norm": 0.44496432452842566, "learning_rate": 0.00018098662617778573, "loss": 0.2928, "step": 6663 }, { "epoch": 3.136, "grad_norm": 0.2741285088885958, "learning_rate": 0.0001809437342144664, "loss": 0.1452, "step": 6664 }, { "epoch": 3.1364705882352943, "grad_norm": 0.2529174726099116, "learning_rate": 0.00018090083960840042, "loss": 0.1063, "step": 6665 }, { "epoch": 3.136941176470588, "grad_norm": 0.34457067122868507, "learning_rate": 0.0001808579423632511, "loss": 0.2954, "step": 6666 }, { "epoch": 3.1374117647058823, "grad_norm": 0.49237279558359315, "learning_rate": 0.0001808150424826822, "loss": 0.2968, "step": 6667 }, { "epoch": 3.1378823529411766, "grad_norm": 0.4553979139754484, "learning_rate": 0.0001807721399703575, "loss": 0.2131, "step": 6668 }, { "epoch": 3.138352941176471, "grad_norm": 0.3172162374805196, "learning_rate": 0.0001807292348299411, "loss": 0.2057, "step": 6669 }, { "epoch": 3.1388235294117646, "grad_norm": 0.5387588323983495, "learning_rate": 0.00018068632706509733, "loss": 0.3749, "step": 6670 }, { "epoch": 3.139294117647059, "grad_norm": 0.39692020207963913, "learning_rate": 0.00018064341667949074, "loss": 0.2829, "step": 6671 }, { "epoch": 3.139764705882353, "grad_norm": 0.29698497408370567, "learning_rate": 0.00018060050367678607, "loss": 0.2209, "step": 6672 }, { "epoch": 3.1402352941176472, "grad_norm": 0.19681668506623445, "learning_rate": 0.00018055758806064822, "loss": 0.0851, "step": 6673 }, { "epoch": 3.140705882352941, "grad_norm": 0.1684534053984616, "learning_rate": 0.00018051466983474252, "loss": 0.0481, "step": 6674 }, { "epoch": 3.1411764705882352, "grad_norm": 0.2647766605518334, "learning_rate": 0.00018047174900273435, "loss": 0.0895, "step": 6675 }, { "epoch": 3.1416470588235295, "grad_norm": 0.5337662237895588, "learning_rate": 0.00018042882556828927, "loss": 0.378, "step": 6676 }, { "epoch": 3.1421176470588237, "grad_norm": 0.2680679819266379, "learning_rate": 0.0001803858995350733, "loss": 0.1951, "step": 6677 }, { "epoch": 3.1425882352941175, "grad_norm": 0.30549244717357005, "learning_rate": 0.00018034297090675246, "loss": 0.2482, "step": 6678 }, { "epoch": 3.1430588235294117, "grad_norm": 0.43900184416884863, "learning_rate": 0.00018030003968699308, "loss": 0.2011, "step": 6679 }, { "epoch": 3.143529411764706, "grad_norm": 0.27635018496155145, "learning_rate": 0.00018025710587946166, "loss": 0.1892, "step": 6680 }, { "epoch": 3.144, "grad_norm": 0.27490923577344883, "learning_rate": 0.00018021416948782503, "loss": 0.1773, "step": 6681 }, { "epoch": 3.1444705882352944, "grad_norm": 0.21141221848135586, "learning_rate": 0.00018017123051575012, "loss": 0.0822, "step": 6682 }, { "epoch": 3.144941176470588, "grad_norm": 0.43699563287924953, "learning_rate": 0.0001801282889669041, "loss": 0.2902, "step": 6683 }, { "epoch": 3.1454117647058824, "grad_norm": 0.42791952479541834, "learning_rate": 0.00018008534484495445, "loss": 0.4053, "step": 6684 }, { "epoch": 3.1458823529411766, "grad_norm": 0.4079280681611571, "learning_rate": 0.0001800423981535688, "loss": 0.2752, "step": 6685 }, { "epoch": 3.1463529411764704, "grad_norm": 0.3824864055196328, "learning_rate": 0.000179999448896415, "loss": 0.3264, "step": 6686 }, { "epoch": 3.1468235294117646, "grad_norm": 0.3776959521324152, "learning_rate": 0.00017995649707716103, "loss": 0.3159, "step": 6687 }, { "epoch": 3.147294117647059, "grad_norm": 0.31176698705819333, "learning_rate": 0.00017991354269947533, "loss": 0.1641, "step": 6688 }, { "epoch": 3.147764705882353, "grad_norm": 0.1989887955478689, "learning_rate": 0.00017987058576702638, "loss": 0.1102, "step": 6689 }, { "epoch": 3.1482352941176472, "grad_norm": 0.36313978614760495, "learning_rate": 0.00017982762628348286, "loss": 0.256, "step": 6690 }, { "epoch": 3.148705882352941, "grad_norm": 0.3738622905302329, "learning_rate": 0.00017978466425251375, "loss": 0.1572, "step": 6691 }, { "epoch": 3.1491764705882352, "grad_norm": 0.4875802641847935, "learning_rate": 0.00017974169967778825, "loss": 0.2539, "step": 6692 }, { "epoch": 3.1496470588235295, "grad_norm": 0.2445625326895167, "learning_rate": 0.00017969873256297572, "loss": 0.0964, "step": 6693 }, { "epoch": 3.1501176470588237, "grad_norm": 0.2587880794497508, "learning_rate": 0.0001796557629117457, "loss": 0.1477, "step": 6694 }, { "epoch": 3.1505882352941175, "grad_norm": 0.5831482140083732, "learning_rate": 0.00017961279072776812, "loss": 0.3415, "step": 6695 }, { "epoch": 3.1510588235294117, "grad_norm": 0.47757253795888693, "learning_rate": 0.00017956981601471294, "loss": 0.3228, "step": 6696 }, { "epoch": 3.151529411764706, "grad_norm": 0.39185495365219475, "learning_rate": 0.00017952683877625042, "loss": 0.2253, "step": 6697 }, { "epoch": 3.152, "grad_norm": 0.21534859890355587, "learning_rate": 0.00017948385901605106, "loss": 0.062, "step": 6698 }, { "epoch": 3.152470588235294, "grad_norm": 0.38832365651168155, "learning_rate": 0.0001794408767377855, "loss": 0.3541, "step": 6699 }, { "epoch": 3.152941176470588, "grad_norm": 0.40463345906230197, "learning_rate": 0.00017939789194512472, "loss": 0.3053, "step": 6700 }, { "epoch": 3.1534117647058824, "grad_norm": 0.44874556115691777, "learning_rate": 0.0001793549046417397, "loss": 0.3018, "step": 6701 }, { "epoch": 3.1538823529411766, "grad_norm": 0.13807718354422363, "learning_rate": 0.00017931191483130186, "loss": 0.0505, "step": 6702 }, { "epoch": 3.1543529411764704, "grad_norm": 0.3007501555976774, "learning_rate": 0.00017926892251748274, "loss": 0.1838, "step": 6703 }, { "epoch": 3.1548235294117646, "grad_norm": 0.3146410320018305, "learning_rate": 0.00017922592770395405, "loss": 0.1787, "step": 6704 }, { "epoch": 3.155294117647059, "grad_norm": 0.2571686096569322, "learning_rate": 0.00017918293039438781, "loss": 0.1364, "step": 6705 }, { "epoch": 3.155764705882353, "grad_norm": 0.3647430845812846, "learning_rate": 0.00017913993059245616, "loss": 0.197, "step": 6706 }, { "epoch": 3.1562352941176472, "grad_norm": 0.2915561151942652, "learning_rate": 0.00017909692830183152, "loss": 0.1324, "step": 6707 }, { "epoch": 3.156705882352941, "grad_norm": 0.557488417869788, "learning_rate": 0.00017905392352618647, "loss": 0.4959, "step": 6708 }, { "epoch": 3.1571764705882353, "grad_norm": 0.42688535336551414, "learning_rate": 0.00017901091626919383, "loss": 0.2188, "step": 6709 }, { "epoch": 3.1576470588235295, "grad_norm": 0.5625328374251704, "learning_rate": 0.00017896790653452668, "loss": 0.4733, "step": 6710 }, { "epoch": 3.1581176470588237, "grad_norm": 0.1893458323439295, "learning_rate": 0.00017892489432585817, "loss": 0.0539, "step": 6711 }, { "epoch": 3.1585882352941175, "grad_norm": 0.22634747527106097, "learning_rate": 0.00017888187964686182, "loss": 0.0879, "step": 6712 }, { "epoch": 3.1590588235294117, "grad_norm": 0.3156502537092593, "learning_rate": 0.00017883886250121133, "loss": 0.1904, "step": 6713 }, { "epoch": 3.159529411764706, "grad_norm": 0.48600155069161677, "learning_rate": 0.00017879584289258048, "loss": 0.2538, "step": 6714 }, { "epoch": 3.16, "grad_norm": 0.40767937387394676, "learning_rate": 0.00017875282082464343, "loss": 0.2422, "step": 6715 }, { "epoch": 3.160470588235294, "grad_norm": 0.4172450917641063, "learning_rate": 0.00017870979630107446, "loss": 0.2136, "step": 6716 }, { "epoch": 3.160941176470588, "grad_norm": 0.3096965047509148, "learning_rate": 0.00017866676932554802, "loss": 0.1557, "step": 6717 }, { "epoch": 3.1614117647058824, "grad_norm": 0.35860349990160084, "learning_rate": 0.0001786237399017389, "loss": 0.2063, "step": 6718 }, { "epoch": 3.1618823529411766, "grad_norm": 0.2971180548915532, "learning_rate": 0.000178580708033322, "loss": 0.1903, "step": 6719 }, { "epoch": 3.1623529411764704, "grad_norm": 0.10773082523075685, "learning_rate": 0.00017853767372397243, "loss": 0.0355, "step": 6720 }, { "epoch": 3.1628235294117646, "grad_norm": 0.3262778826025918, "learning_rate": 0.00017849463697736554, "loss": 0.2437, "step": 6721 }, { "epoch": 3.163294117647059, "grad_norm": 0.41399415431675224, "learning_rate": 0.0001784515977971769, "loss": 0.3329, "step": 6722 }, { "epoch": 3.163764705882353, "grad_norm": 0.11383011784981822, "learning_rate": 0.00017840855618708227, "loss": 0.0436, "step": 6723 }, { "epoch": 3.1642352941176473, "grad_norm": 0.42706133777632355, "learning_rate": 0.00017836551215075758, "loss": 0.2767, "step": 6724 }, { "epoch": 3.164705882352941, "grad_norm": 0.4013546447680049, "learning_rate": 0.000178322465691879, "loss": 0.3364, "step": 6725 }, { "epoch": 3.1651764705882353, "grad_norm": 0.37912465329577766, "learning_rate": 0.00017827941681412298, "loss": 0.231, "step": 6726 }, { "epoch": 3.1656470588235295, "grad_norm": 0.21828307923999293, "learning_rate": 0.00017823636552116603, "loss": 0.111, "step": 6727 }, { "epoch": 3.1661176470588237, "grad_norm": 0.3185559087726698, "learning_rate": 0.00017819331181668496, "loss": 0.1768, "step": 6728 }, { "epoch": 3.1665882352941175, "grad_norm": 0.31975629829600644, "learning_rate": 0.00017815025570435677, "loss": 0.2088, "step": 6729 }, { "epoch": 3.1670588235294117, "grad_norm": 0.5245519209663017, "learning_rate": 0.0001781071971878587, "loss": 0.3966, "step": 6730 }, { "epoch": 3.167529411764706, "grad_norm": 0.2474526164737337, "learning_rate": 0.00017806413627086807, "loss": 0.1125, "step": 6731 }, { "epoch": 3.168, "grad_norm": 0.2805177320754892, "learning_rate": 0.0001780210729570626, "loss": 0.1303, "step": 6732 }, { "epoch": 3.168470588235294, "grad_norm": 0.5686916520419186, "learning_rate": 0.00017797800725012002, "loss": 0.3424, "step": 6733 }, { "epoch": 3.168941176470588, "grad_norm": 0.37910362985084817, "learning_rate": 0.00017793493915371842, "loss": 0.2827, "step": 6734 }, { "epoch": 3.1694117647058824, "grad_norm": 0.6582008384182392, "learning_rate": 0.00017789186867153594, "loss": 0.2684, "step": 6735 }, { "epoch": 3.1698823529411766, "grad_norm": 0.30522538145845035, "learning_rate": 0.00017784879580725113, "loss": 0.2538, "step": 6736 }, { "epoch": 3.1703529411764704, "grad_norm": 0.4472702093178092, "learning_rate": 0.00017780572056454256, "loss": 0.3385, "step": 6737 }, { "epoch": 3.1708235294117646, "grad_norm": 0.26702998753673884, "learning_rate": 0.00017776264294708905, "loss": 0.1732, "step": 6738 }, { "epoch": 3.171294117647059, "grad_norm": 0.32251889854807086, "learning_rate": 0.00017771956295856964, "loss": 0.2548, "step": 6739 }, { "epoch": 3.171764705882353, "grad_norm": 0.38567476673912443, "learning_rate": 0.0001776764806026636, "loss": 0.2865, "step": 6740 }, { "epoch": 3.1722352941176473, "grad_norm": 0.19258336987125463, "learning_rate": 0.00017763339588305037, "loss": 0.0632, "step": 6741 }, { "epoch": 3.172705882352941, "grad_norm": 0.6210278165732942, "learning_rate": 0.00017759030880340956, "loss": 0.3187, "step": 6742 }, { "epoch": 3.1731764705882353, "grad_norm": 0.057586138409943324, "learning_rate": 0.00017754721936742108, "loss": 0.0098, "step": 6743 }, { "epoch": 3.1736470588235295, "grad_norm": 0.5413917099395474, "learning_rate": 0.00017750412757876492, "loss": 0.3711, "step": 6744 }, { "epoch": 3.1741176470588237, "grad_norm": 0.26447309521465084, "learning_rate": 0.00017746103344112136, "loss": 0.099, "step": 6745 }, { "epoch": 3.1745882352941175, "grad_norm": 0.4186115747116018, "learning_rate": 0.00017741793695817093, "loss": 0.289, "step": 6746 }, { "epoch": 3.1750588235294117, "grad_norm": 0.43125149692031073, "learning_rate": 0.00017737483813359413, "loss": 0.1535, "step": 6747 }, { "epoch": 3.175529411764706, "grad_norm": 0.2901915712257674, "learning_rate": 0.00017733173697107187, "loss": 0.0921, "step": 6748 }, { "epoch": 3.176, "grad_norm": 0.2128214235021299, "learning_rate": 0.00017728863347428524, "loss": 0.1482, "step": 6749 }, { "epoch": 3.176470588235294, "grad_norm": 0.11446439305540128, "learning_rate": 0.00017724552764691545, "loss": 0.0481, "step": 6750 }, { "epoch": 3.176941176470588, "grad_norm": 0.3384509657349339, "learning_rate": 0.000177202419492644, "loss": 0.2243, "step": 6751 }, { "epoch": 3.1774117647058824, "grad_norm": 0.4888773409832752, "learning_rate": 0.00017715930901515242, "loss": 0.2401, "step": 6752 }, { "epoch": 3.1778823529411766, "grad_norm": 0.41502731125416115, "learning_rate": 0.0001771161962181227, "loss": 0.3488, "step": 6753 }, { "epoch": 3.1783529411764704, "grad_norm": 0.44500885573307036, "learning_rate": 0.0001770730811052368, "loss": 0.3389, "step": 6754 }, { "epoch": 3.1788235294117646, "grad_norm": 0.43005932108777445, "learning_rate": 0.00017702996368017693, "loss": 0.2884, "step": 6755 }, { "epoch": 3.179294117647059, "grad_norm": 0.16900403274671064, "learning_rate": 0.00017698684394662563, "loss": 0.0621, "step": 6756 }, { "epoch": 3.179764705882353, "grad_norm": 0.5906217678553972, "learning_rate": 0.0001769437219082655, "loss": 0.4559, "step": 6757 }, { "epoch": 3.1802352941176473, "grad_norm": 0.6887548049672774, "learning_rate": 0.00017690059756877933, "loss": 0.4288, "step": 6758 }, { "epoch": 3.180705882352941, "grad_norm": 0.29894895854979925, "learning_rate": 0.00017685747093185017, "loss": 0.1668, "step": 6759 }, { "epoch": 3.1811764705882353, "grad_norm": 0.4690431835503764, "learning_rate": 0.00017681434200116132, "loss": 0.3198, "step": 6760 }, { "epoch": 3.1816470588235295, "grad_norm": 0.33154823812202044, "learning_rate": 0.0001767712107803961, "loss": 0.1811, "step": 6761 }, { "epoch": 3.1821176470588237, "grad_norm": 0.2686386230561004, "learning_rate": 0.00017672807727323818, "loss": 0.1826, "step": 6762 }, { "epoch": 3.1825882352941175, "grad_norm": 0.060769613093341515, "learning_rate": 0.00017668494148337133, "loss": 0.006, "step": 6763 }, { "epoch": 3.1830588235294117, "grad_norm": 0.5714230134867214, "learning_rate": 0.00017664180341447964, "loss": 0.4043, "step": 6764 }, { "epoch": 3.183529411764706, "grad_norm": 0.2780446861332768, "learning_rate": 0.0001765986630702472, "loss": 0.1521, "step": 6765 }, { "epoch": 3.184, "grad_norm": 0.2925652726416706, "learning_rate": 0.00017655552045435846, "loss": 0.1668, "step": 6766 }, { "epoch": 3.184470588235294, "grad_norm": 0.39527519487221957, "learning_rate": 0.00017651237557049808, "loss": 0.2979, "step": 6767 }, { "epoch": 3.184941176470588, "grad_norm": 0.4966769740813768, "learning_rate": 0.00017646922842235076, "loss": 0.4188, "step": 6768 }, { "epoch": 3.1854117647058824, "grad_norm": 0.35703512233895707, "learning_rate": 0.00017642607901360147, "loss": 0.203, "step": 6769 }, { "epoch": 3.1858823529411766, "grad_norm": 0.6378416598585994, "learning_rate": 0.00017638292734793547, "loss": 0.5406, "step": 6770 }, { "epoch": 3.1863529411764704, "grad_norm": 0.5424708744145721, "learning_rate": 0.00017633977342903807, "loss": 0.3191, "step": 6771 }, { "epoch": 3.1868235294117646, "grad_norm": 0.355273302845989, "learning_rate": 0.0001762966172605948, "loss": 0.2756, "step": 6772 }, { "epoch": 3.187294117647059, "grad_norm": 0.3014728973225883, "learning_rate": 0.00017625345884629141, "loss": 0.2239, "step": 6773 }, { "epoch": 3.187764705882353, "grad_norm": 0.37936207936004973, "learning_rate": 0.00017621029818981387, "loss": 0.2829, "step": 6774 }, { "epoch": 3.1882352941176473, "grad_norm": 0.30714204768160547, "learning_rate": 0.00017616713529484835, "loss": 0.1053, "step": 6775 }, { "epoch": 3.188705882352941, "grad_norm": 0.23849054619731008, "learning_rate": 0.00017612397016508104, "loss": 0.1057, "step": 6776 }, { "epoch": 3.1891764705882353, "grad_norm": 0.9733527264043172, "learning_rate": 0.0001760808028041986, "loss": 0.7062, "step": 6777 }, { "epoch": 3.1896470588235295, "grad_norm": 0.42491444379705345, "learning_rate": 0.0001760376332158877, "loss": 0.2523, "step": 6778 }, { "epoch": 3.1901176470588237, "grad_norm": 0.19955434917892892, "learning_rate": 0.00017599446140383512, "loss": 0.0715, "step": 6779 }, { "epoch": 3.1905882352941175, "grad_norm": 0.44547116705050344, "learning_rate": 0.00017595128737172812, "loss": 0.2541, "step": 6780 }, { "epoch": 3.1910588235294117, "grad_norm": 0.20226950629793466, "learning_rate": 0.00017590811112325384, "loss": 0.0848, "step": 6781 }, { "epoch": 3.191529411764706, "grad_norm": 0.25060301257436285, "learning_rate": 0.00017586493266209978, "loss": 0.13, "step": 6782 }, { "epoch": 3.192, "grad_norm": 0.4359639651937591, "learning_rate": 0.00017582175199195365, "loss": 0.2272, "step": 6783 }, { "epoch": 3.192470588235294, "grad_norm": 0.48304147108313494, "learning_rate": 0.00017577856911650324, "loss": 0.416, "step": 6784 }, { "epoch": 3.192941176470588, "grad_norm": 0.349580059994057, "learning_rate": 0.00017573538403943656, "loss": 0.1452, "step": 6785 }, { "epoch": 3.1934117647058824, "grad_norm": 0.2999797145155727, "learning_rate": 0.00017569219676444187, "loss": 0.1633, "step": 6786 }, { "epoch": 3.1938823529411766, "grad_norm": 0.36824362327762467, "learning_rate": 0.00017564900729520751, "loss": 0.4043, "step": 6787 }, { "epoch": 3.1943529411764704, "grad_norm": 0.46191378723022286, "learning_rate": 0.0001756058156354222, "loss": 0.3942, "step": 6788 }, { "epoch": 3.1948235294117646, "grad_norm": 0.5729678611752844, "learning_rate": 0.00017556262178877459, "loss": 0.677, "step": 6789 }, { "epoch": 3.195294117647059, "grad_norm": 0.3256814529355082, "learning_rate": 0.00017551942575895368, "loss": 0.3428, "step": 6790 }, { "epoch": 3.195764705882353, "grad_norm": 0.3160544624402937, "learning_rate": 0.00017547622754964868, "loss": 0.134, "step": 6791 }, { "epoch": 3.1962352941176473, "grad_norm": 0.08584381808510588, "learning_rate": 0.0001754330271645489, "loss": 0.021, "step": 6792 }, { "epoch": 3.196705882352941, "grad_norm": 0.13992628023744302, "learning_rate": 0.00017538982460734375, "loss": 0.0585, "step": 6793 }, { "epoch": 3.1971764705882353, "grad_norm": 0.515800284038666, "learning_rate": 0.00017534661988172315, "loss": 0.409, "step": 6794 }, { "epoch": 3.1976470588235295, "grad_norm": 0.326422144563091, "learning_rate": 0.00017530341299137685, "loss": 0.2145, "step": 6795 }, { "epoch": 3.1981176470588237, "grad_norm": 0.6121172120346889, "learning_rate": 0.00017526020393999496, "loss": 0.5877, "step": 6796 }, { "epoch": 3.1985882352941175, "grad_norm": 0.30002277007123807, "learning_rate": 0.00017521699273126778, "loss": 0.1138, "step": 6797 }, { "epoch": 3.1990588235294117, "grad_norm": 0.4748415519875851, "learning_rate": 0.00017517377936888568, "loss": 0.3601, "step": 6798 }, { "epoch": 3.199529411764706, "grad_norm": 0.3054673591938512, "learning_rate": 0.0001751305638565394, "loss": 0.1377, "step": 6799 }, { "epoch": 3.2, "grad_norm": 0.4688339390117099, "learning_rate": 0.00017508734619791966, "loss": 0.4476, "step": 6800 }, { "epoch": 3.200470588235294, "grad_norm": 0.14740477848951322, "learning_rate": 0.00017504412639671754, "loss": 0.0461, "step": 6801 }, { "epoch": 3.200941176470588, "grad_norm": 0.06180365323603712, "learning_rate": 0.00017500090445662415, "loss": 0.0099, "step": 6802 }, { "epoch": 3.2014117647058824, "grad_norm": 0.30818760669457673, "learning_rate": 0.0001749576803813309, "loss": 0.231, "step": 6803 }, { "epoch": 3.2018823529411766, "grad_norm": 0.41634685732652515, "learning_rate": 0.00017491445417452934, "loss": 0.3, "step": 6804 }, { "epoch": 3.2023529411764704, "grad_norm": 0.4211025696676291, "learning_rate": 0.00017487122583991123, "loss": 0.404, "step": 6805 }, { "epoch": 3.2028235294117646, "grad_norm": 0.2458009923497951, "learning_rate": 0.00017482799538116834, "loss": 0.1049, "step": 6806 }, { "epoch": 3.203294117647059, "grad_norm": 0.398038978654855, "learning_rate": 0.00017478476280199294, "loss": 0.1849, "step": 6807 }, { "epoch": 3.203764705882353, "grad_norm": 0.6971205576551259, "learning_rate": 0.0001747415281060772, "loss": 0.5208, "step": 6808 }, { "epoch": 3.2042352941176473, "grad_norm": 0.557105513320977, "learning_rate": 0.00017469829129711361, "loss": 0.4745, "step": 6809 }, { "epoch": 3.204705882352941, "grad_norm": 0.1812041656790127, "learning_rate": 0.00017465505237879476, "loss": 0.0582, "step": 6810 }, { "epoch": 3.2051764705882353, "grad_norm": 0.41742728843076526, "learning_rate": 0.00017461181135481355, "loss": 0.2874, "step": 6811 }, { "epoch": 3.2056470588235295, "grad_norm": 0.2973592235218322, "learning_rate": 0.00017456856822886297, "loss": 0.2227, "step": 6812 }, { "epoch": 3.2061176470588233, "grad_norm": 0.6284935423367984, "learning_rate": 0.00017452532300463609, "loss": 0.5296, "step": 6813 }, { "epoch": 3.2065882352941175, "grad_norm": 0.3866662703334861, "learning_rate": 0.00017448207568582634, "loss": 0.2665, "step": 6814 }, { "epoch": 3.2070588235294117, "grad_norm": 0.41659328795122064, "learning_rate": 0.00017443882627612726, "loss": 0.2273, "step": 6815 }, { "epoch": 3.207529411764706, "grad_norm": 0.41474203055606446, "learning_rate": 0.00017439557477923255, "loss": 0.1619, "step": 6816 }, { "epoch": 3.208, "grad_norm": 0.3035562832668746, "learning_rate": 0.00017435232119883604, "loss": 0.2192, "step": 6817 }, { "epoch": 3.208470588235294, "grad_norm": 0.2439011613646967, "learning_rate": 0.0001743090655386319, "loss": 0.0977, "step": 6818 }, { "epoch": 3.208941176470588, "grad_norm": 0.6416747335709435, "learning_rate": 0.00017426580780231435, "loss": 0.5531, "step": 6819 }, { "epoch": 3.2094117647058824, "grad_norm": 0.5075033511564723, "learning_rate": 0.0001742225479935777, "loss": 0.4383, "step": 6820 }, { "epoch": 3.2098823529411766, "grad_norm": 0.3171222145702744, "learning_rate": 0.00017417928611611675, "loss": 0.2291, "step": 6821 }, { "epoch": 3.2103529411764704, "grad_norm": 0.3675234491180208, "learning_rate": 0.0001741360221736261, "loss": 0.2706, "step": 6822 }, { "epoch": 3.2108235294117646, "grad_norm": 0.2529071097192416, "learning_rate": 0.0001740927561698008, "loss": 0.1238, "step": 6823 }, { "epoch": 3.211294117647059, "grad_norm": 0.4576541929605935, "learning_rate": 0.00017404948810833594, "loss": 0.3736, "step": 6824 }, { "epoch": 3.211764705882353, "grad_norm": 0.5195602368615434, "learning_rate": 0.0001740062179929268, "loss": 0.4472, "step": 6825 }, { "epoch": 3.212235294117647, "grad_norm": 0.27242204560629185, "learning_rate": 0.00017396294582726896, "loss": 0.1338, "step": 6826 }, { "epoch": 3.212705882352941, "grad_norm": 0.40407832006814864, "learning_rate": 0.00017391967161505795, "loss": 0.2452, "step": 6827 }, { "epoch": 3.2131764705882353, "grad_norm": 0.1829901069295186, "learning_rate": 0.00017387639535998967, "loss": 0.0658, "step": 6828 }, { "epoch": 3.2136470588235295, "grad_norm": 0.34284502007520895, "learning_rate": 0.00017383311706576016, "loss": 0.2046, "step": 6829 }, { "epoch": 3.2141176470588233, "grad_norm": 0.4928090871770452, "learning_rate": 0.00017378983673606548, "loss": 0.3409, "step": 6830 }, { "epoch": 3.2145882352941175, "grad_norm": 0.39977359628994524, "learning_rate": 0.00017374655437460206, "loss": 0.3181, "step": 6831 }, { "epoch": 3.2150588235294117, "grad_norm": 0.5767915763281142, "learning_rate": 0.00017370326998506646, "loss": 0.3386, "step": 6832 }, { "epoch": 3.215529411764706, "grad_norm": 0.39311309326265664, "learning_rate": 0.00017365998357115535, "loss": 0.1667, "step": 6833 }, { "epoch": 3.216, "grad_norm": 0.5562048084835203, "learning_rate": 0.00017361669513656558, "loss": 0.1777, "step": 6834 }, { "epoch": 3.216470588235294, "grad_norm": 0.33619406085043213, "learning_rate": 0.0001735734046849942, "loss": 0.3378, "step": 6835 }, { "epoch": 3.216941176470588, "grad_norm": 0.35172253434904627, "learning_rate": 0.0001735301122201384, "loss": 0.2856, "step": 6836 }, { "epoch": 3.2174117647058824, "grad_norm": 0.34207752280762127, "learning_rate": 0.00017348681774569563, "loss": 0.2918, "step": 6837 }, { "epoch": 3.2178823529411766, "grad_norm": 0.3772154332016851, "learning_rate": 0.00017344352126536343, "loss": 0.2177, "step": 6838 }, { "epoch": 3.2183529411764704, "grad_norm": 0.20768406312002047, "learning_rate": 0.00017340022278283953, "loss": 0.1101, "step": 6839 }, { "epoch": 3.2188235294117646, "grad_norm": 0.26304206530928315, "learning_rate": 0.00017335692230182183, "loss": 0.1798, "step": 6840 }, { "epoch": 3.219294117647059, "grad_norm": 0.5023527748389983, "learning_rate": 0.00017331361982600836, "loss": 0.3408, "step": 6841 }, { "epoch": 3.219764705882353, "grad_norm": 0.3258388135610386, "learning_rate": 0.00017327031535909745, "loss": 0.2234, "step": 6842 }, { "epoch": 3.220235294117647, "grad_norm": 0.306865630357837, "learning_rate": 0.00017322700890478744, "loss": 0.1974, "step": 6843 }, { "epoch": 3.220705882352941, "grad_norm": 0.4408181912364455, "learning_rate": 0.00017318370046677693, "loss": 0.3472, "step": 6844 }, { "epoch": 3.2211764705882353, "grad_norm": 0.3492789211240375, "learning_rate": 0.00017314039004876473, "loss": 0.1987, "step": 6845 }, { "epoch": 3.2216470588235295, "grad_norm": 0.2412048816656334, "learning_rate": 0.0001730970776544497, "loss": 0.1744, "step": 6846 }, { "epoch": 3.2221176470588233, "grad_norm": 0.484124224453392, "learning_rate": 0.00017305376328753092, "loss": 0.4142, "step": 6847 }, { "epoch": 3.2225882352941175, "grad_norm": 0.22763232719310197, "learning_rate": 0.00017301044695170769, "loss": 0.1041, "step": 6848 }, { "epoch": 3.2230588235294118, "grad_norm": 0.07720278283050162, "learning_rate": 0.00017296712865067944, "loss": 0.0181, "step": 6849 }, { "epoch": 3.223529411764706, "grad_norm": 0.387240317321238, "learning_rate": 0.00017292380838814577, "loss": 0.2553, "step": 6850 }, { "epoch": 3.224, "grad_norm": 0.40542018705791205, "learning_rate": 0.00017288048616780636, "loss": 0.3319, "step": 6851 }, { "epoch": 3.224470588235294, "grad_norm": 0.4791195987106091, "learning_rate": 0.00017283716199336123, "loss": 0.2873, "step": 6852 }, { "epoch": 3.224941176470588, "grad_norm": 0.27300044242460336, "learning_rate": 0.0001727938358685105, "loss": 0.1319, "step": 6853 }, { "epoch": 3.2254117647058824, "grad_norm": 0.737842971198939, "learning_rate": 0.0001727505077969543, "loss": 0.4312, "step": 6854 }, { "epoch": 3.2258823529411766, "grad_norm": 0.3916712921308357, "learning_rate": 0.00017270717778239316, "loss": 0.3148, "step": 6855 }, { "epoch": 3.2263529411764704, "grad_norm": 0.36732507949321114, "learning_rate": 0.0001726638458285277, "loss": 0.2426, "step": 6856 }, { "epoch": 3.2268235294117646, "grad_norm": 0.5863049387889963, "learning_rate": 0.00017262051193905865, "loss": 0.5029, "step": 6857 }, { "epoch": 3.227294117647059, "grad_norm": 0.21121353742568852, "learning_rate": 0.00017257717611768685, "loss": 0.1406, "step": 6858 }, { "epoch": 3.227764705882353, "grad_norm": 0.20198536210398463, "learning_rate": 0.00017253383836811356, "loss": 0.0996, "step": 6859 }, { "epoch": 3.228235294117647, "grad_norm": 0.09517574146765011, "learning_rate": 0.0001724904986940399, "loss": 0.0364, "step": 6860 }, { "epoch": 3.228705882352941, "grad_norm": 0.2202117213171291, "learning_rate": 0.00017244715709916733, "loss": 0.1078, "step": 6861 }, { "epoch": 3.2291764705882353, "grad_norm": 0.4435120618961333, "learning_rate": 0.00017240381358719746, "loss": 0.3358, "step": 6862 }, { "epoch": 3.2296470588235295, "grad_norm": 0.0580152511638444, "learning_rate": 0.000172360468161832, "loss": 0.0094, "step": 6863 }, { "epoch": 3.2301176470588233, "grad_norm": 0.29859303427470196, "learning_rate": 0.00017231712082677291, "loss": 0.1336, "step": 6864 }, { "epoch": 3.2305882352941175, "grad_norm": 0.469807504468399, "learning_rate": 0.0001722737715857222, "loss": 0.3114, "step": 6865 }, { "epoch": 3.2310588235294118, "grad_norm": 0.1753049788014701, "learning_rate": 0.00017223042044238219, "loss": 0.0541, "step": 6866 }, { "epoch": 3.231529411764706, "grad_norm": 0.25418108172342685, "learning_rate": 0.0001721870674004552, "loss": 0.0588, "step": 6867 }, { "epoch": 3.232, "grad_norm": 0.5157161644400974, "learning_rate": 0.00017214371246364384, "loss": 0.3569, "step": 6868 }, { "epoch": 3.232470588235294, "grad_norm": 0.29094194945302154, "learning_rate": 0.00017210035563565082, "loss": 0.1424, "step": 6869 }, { "epoch": 3.232941176470588, "grad_norm": 0.27969637744409, "learning_rate": 0.00017205699692017902, "loss": 0.1107, "step": 6870 }, { "epoch": 3.2334117647058824, "grad_norm": 0.37111306702003394, "learning_rate": 0.0001720136363209315, "loss": 0.3803, "step": 6871 }, { "epoch": 3.2338823529411767, "grad_norm": 0.3375144355031858, "learning_rate": 0.0001719702738416115, "loss": 0.1417, "step": 6872 }, { "epoch": 3.2343529411764704, "grad_norm": 0.5474773612462733, "learning_rate": 0.0001719269094859223, "loss": 0.54, "step": 6873 }, { "epoch": 3.2348235294117647, "grad_norm": 0.24273155810972788, "learning_rate": 0.0001718835432575675, "loss": 0.1809, "step": 6874 }, { "epoch": 3.235294117647059, "grad_norm": 0.22847353418338917, "learning_rate": 0.00017184017516025076, "loss": 0.1387, "step": 6875 }, { "epoch": 3.235764705882353, "grad_norm": 0.28277969107238404, "learning_rate": 0.0001717968051976759, "loss": 0.1922, "step": 6876 }, { "epoch": 3.236235294117647, "grad_norm": 0.3379167871825588, "learning_rate": 0.00017175343337354698, "loss": 0.2932, "step": 6877 }, { "epoch": 3.236705882352941, "grad_norm": 0.2968366866597626, "learning_rate": 0.00017171005969156817, "loss": 0.0842, "step": 6878 }, { "epoch": 3.2371764705882353, "grad_norm": 0.49618218574217154, "learning_rate": 0.00017166668415544376, "loss": 0.2557, "step": 6879 }, { "epoch": 3.2376470588235295, "grad_norm": 0.3991475939371831, "learning_rate": 0.0001716233067688783, "loss": 0.2159, "step": 6880 }, { "epoch": 3.2381176470588233, "grad_norm": 0.4075246886456125, "learning_rate": 0.00017157992753557633, "loss": 0.2898, "step": 6881 }, { "epoch": 3.2385882352941175, "grad_norm": 0.425457740218744, "learning_rate": 0.0001715365464592427, "loss": 0.2975, "step": 6882 }, { "epoch": 3.2390588235294118, "grad_norm": 0.2692461416685318, "learning_rate": 0.00017149316354358237, "loss": 0.1731, "step": 6883 }, { "epoch": 3.239529411764706, "grad_norm": 0.36006104331484945, "learning_rate": 0.00017144977879230046, "loss": 0.1556, "step": 6884 }, { "epoch": 3.24, "grad_norm": 0.314693261292855, "learning_rate": 0.00017140639220910222, "loss": 0.1635, "step": 6885 }, { "epoch": 3.240470588235294, "grad_norm": 0.3330290505225648, "learning_rate": 0.00017136300379769312, "loss": 0.1794, "step": 6886 }, { "epoch": 3.240941176470588, "grad_norm": 0.3737793380657937, "learning_rate": 0.00017131961356177869, "loss": 0.2337, "step": 6887 }, { "epoch": 3.2414117647058824, "grad_norm": 0.38777232510036186, "learning_rate": 0.00017127622150506472, "loss": 0.2199, "step": 6888 }, { "epoch": 3.2418823529411767, "grad_norm": 0.34087600941862145, "learning_rate": 0.000171232827631257, "loss": 0.2283, "step": 6889 }, { "epoch": 3.2423529411764704, "grad_norm": 0.494330599615667, "learning_rate": 0.00017118943194406169, "loss": 0.3146, "step": 6890 }, { "epoch": 3.2428235294117647, "grad_norm": 0.4644918523299326, "learning_rate": 0.00017114603444718496, "loss": 0.306, "step": 6891 }, { "epoch": 3.243294117647059, "grad_norm": 0.5585534766464295, "learning_rate": 0.00017110263514433317, "loss": 0.3104, "step": 6892 }, { "epoch": 3.243764705882353, "grad_norm": 0.3090985653611173, "learning_rate": 0.00017105923403921283, "loss": 0.2027, "step": 6893 }, { "epoch": 3.244235294117647, "grad_norm": 0.6413092339376981, "learning_rate": 0.00017101583113553063, "loss": 0.3235, "step": 6894 }, { "epoch": 3.244705882352941, "grad_norm": 0.3620742603863029, "learning_rate": 0.00017097242643699333, "loss": 0.2601, "step": 6895 }, { "epoch": 3.2451764705882353, "grad_norm": 0.5175829123843305, "learning_rate": 0.00017092901994730797, "loss": 0.2965, "step": 6896 }, { "epoch": 3.2456470588235296, "grad_norm": 0.2754755271986787, "learning_rate": 0.00017088561167018164, "loss": 0.1536, "step": 6897 }, { "epoch": 3.2461176470588233, "grad_norm": 0.33457597986244636, "learning_rate": 0.0001708422016093216, "loss": 0.1267, "step": 6898 }, { "epoch": 3.2465882352941176, "grad_norm": 0.4602656802024471, "learning_rate": 0.00017079878976843533, "loss": 0.3741, "step": 6899 }, { "epoch": 3.2470588235294118, "grad_norm": 0.5139494011148824, "learning_rate": 0.00017075537615123042, "loss": 0.3259, "step": 6900 }, { "epoch": 3.247529411764706, "grad_norm": 0.20852461930669836, "learning_rate": 0.0001707119607614146, "loss": 0.075, "step": 6901 }, { "epoch": 3.248, "grad_norm": 0.5346374931203798, "learning_rate": 0.0001706685436026957, "loss": 0.2319, "step": 6902 }, { "epoch": 3.248470588235294, "grad_norm": 0.29850037979550037, "learning_rate": 0.0001706251246787818, "loss": 0.147, "step": 6903 }, { "epoch": 3.248941176470588, "grad_norm": 0.37871713856527844, "learning_rate": 0.0001705817039933811, "loss": 0.3247, "step": 6904 }, { "epoch": 3.2494117647058824, "grad_norm": 0.5134452558743323, "learning_rate": 0.00017053828155020193, "loss": 0.4646, "step": 6905 }, { "epoch": 3.2498823529411767, "grad_norm": 0.2747464542961122, "learning_rate": 0.00017049485735295275, "loss": 0.2256, "step": 6906 }, { "epoch": 3.2503529411764704, "grad_norm": 0.19867423780625343, "learning_rate": 0.00017045143140534226, "loss": 0.077, "step": 6907 }, { "epoch": 3.2508235294117647, "grad_norm": 0.3383398000567427, "learning_rate": 0.0001704080037110792, "loss": 0.1942, "step": 6908 }, { "epoch": 3.251294117647059, "grad_norm": 0.4413510443164798, "learning_rate": 0.00017036457427387248, "loss": 0.3855, "step": 6909 }, { "epoch": 3.251764705882353, "grad_norm": 0.36089138424294587, "learning_rate": 0.00017032114309743128, "loss": 0.3366, "step": 6910 }, { "epoch": 3.252235294117647, "grad_norm": 0.25682431337114053, "learning_rate": 0.0001702777101854648, "loss": 0.179, "step": 6911 }, { "epoch": 3.252705882352941, "grad_norm": 0.16234383354745907, "learning_rate": 0.0001702342755416824, "loss": 0.0768, "step": 6912 }, { "epoch": 3.2531764705882353, "grad_norm": 0.37655025438593925, "learning_rate": 0.00017019083916979357, "loss": 0.1108, "step": 6913 }, { "epoch": 3.2536470588235296, "grad_norm": 0.3121677308227036, "learning_rate": 0.0001701474010735081, "loss": 0.2305, "step": 6914 }, { "epoch": 3.2541176470588233, "grad_norm": 0.820922026011668, "learning_rate": 0.00017010396125653572, "loss": 0.4812, "step": 6915 }, { "epoch": 3.2545882352941176, "grad_norm": 0.39844379232636756, "learning_rate": 0.00017006051972258645, "loss": 0.3332, "step": 6916 }, { "epoch": 3.255058823529412, "grad_norm": 0.3937527678975444, "learning_rate": 0.0001700170764753704, "loss": 0.2944, "step": 6917 }, { "epoch": 3.255529411764706, "grad_norm": 0.6154173787874986, "learning_rate": 0.00016997363151859785, "loss": 0.6179, "step": 6918 }, { "epoch": 3.2560000000000002, "grad_norm": 0.2522559963046217, "learning_rate": 0.00016993018485597915, "loss": 0.2566, "step": 6919 }, { "epoch": 3.256470588235294, "grad_norm": 0.30712285756898233, "learning_rate": 0.00016988673649122492, "loss": 0.1379, "step": 6920 }, { "epoch": 3.2569411764705882, "grad_norm": 0.2012152245189915, "learning_rate": 0.00016984328642804587, "loss": 0.0493, "step": 6921 }, { "epoch": 3.2574117647058825, "grad_norm": 0.28283025580907917, "learning_rate": 0.00016979983467015284, "loss": 0.1532, "step": 6922 }, { "epoch": 3.2578823529411762, "grad_norm": 0.19050061208421515, "learning_rate": 0.00016975638122125678, "loss": 0.1292, "step": 6923 }, { "epoch": 3.2583529411764705, "grad_norm": 0.2891247534093535, "learning_rate": 0.00016971292608506888, "loss": 0.1549, "step": 6924 }, { "epoch": 3.2588235294117647, "grad_norm": 0.4026308601449429, "learning_rate": 0.0001696694692653004, "loss": 0.2451, "step": 6925 }, { "epoch": 3.259294117647059, "grad_norm": 0.29934893402454366, "learning_rate": 0.00016962601076566275, "loss": 0.1369, "step": 6926 }, { "epoch": 3.259764705882353, "grad_norm": 0.3492687572656281, "learning_rate": 0.00016958255058986753, "loss": 0.3212, "step": 6927 }, { "epoch": 3.260235294117647, "grad_norm": 0.6049377182252755, "learning_rate": 0.00016953908874162645, "loss": 0.5387, "step": 6928 }, { "epoch": 3.260705882352941, "grad_norm": 0.45315669534350816, "learning_rate": 0.00016949562522465135, "loss": 0.3765, "step": 6929 }, { "epoch": 3.2611764705882353, "grad_norm": 0.8890302867617876, "learning_rate": 0.00016945216004265418, "loss": 0.3449, "step": 6930 }, { "epoch": 3.2616470588235296, "grad_norm": 0.30444173262298985, "learning_rate": 0.0001694086931993472, "loss": 0.1407, "step": 6931 }, { "epoch": 3.2621176470588233, "grad_norm": 0.2674263095716429, "learning_rate": 0.0001693652246984426, "loss": 0.1644, "step": 6932 }, { "epoch": 3.2625882352941176, "grad_norm": 0.3260540123529906, "learning_rate": 0.0001693217545436528, "loss": 0.1227, "step": 6933 }, { "epoch": 3.263058823529412, "grad_norm": 0.32737503718861716, "learning_rate": 0.0001692782827386904, "loss": 0.2057, "step": 6934 }, { "epoch": 3.263529411764706, "grad_norm": 0.154536566562669, "learning_rate": 0.00016923480928726813, "loss": 0.0677, "step": 6935 }, { "epoch": 3.2640000000000002, "grad_norm": 0.1818315989044733, "learning_rate": 0.00016919133419309875, "loss": 0.0518, "step": 6936 }, { "epoch": 3.264470588235294, "grad_norm": 0.46914370478431583, "learning_rate": 0.00016914785745989532, "loss": 0.3263, "step": 6937 }, { "epoch": 3.2649411764705882, "grad_norm": 0.2238481804906079, "learning_rate": 0.00016910437909137098, "loss": 0.0738, "step": 6938 }, { "epoch": 3.2654117647058825, "grad_norm": 0.34910439945928534, "learning_rate": 0.00016906089909123892, "loss": 0.3024, "step": 6939 }, { "epoch": 3.2658823529411762, "grad_norm": 0.4470243080959553, "learning_rate": 0.00016901741746321253, "loss": 0.3149, "step": 6940 }, { "epoch": 3.2663529411764705, "grad_norm": 0.3370477130858682, "learning_rate": 0.0001689739342110055, "loss": 0.2875, "step": 6941 }, { "epoch": 3.2668235294117647, "grad_norm": 0.30257593151311835, "learning_rate": 0.00016893044933833135, "loss": 0.2141, "step": 6942 }, { "epoch": 3.267294117647059, "grad_norm": 0.3500644073461708, "learning_rate": 0.00016888696284890402, "loss": 0.2143, "step": 6943 }, { "epoch": 3.267764705882353, "grad_norm": 0.3990471276744831, "learning_rate": 0.0001688434747464374, "loss": 0.349, "step": 6944 }, { "epoch": 3.268235294117647, "grad_norm": 0.3955525098480127, "learning_rate": 0.00016879998503464561, "loss": 0.3482, "step": 6945 }, { "epoch": 3.268705882352941, "grad_norm": 0.40374451702013187, "learning_rate": 0.00016875649371724292, "loss": 0.3057, "step": 6946 }, { "epoch": 3.2691764705882354, "grad_norm": 0.11697927534398724, "learning_rate": 0.00016871300079794363, "loss": 0.025, "step": 6947 }, { "epoch": 3.2696470588235296, "grad_norm": 0.3757798797862377, "learning_rate": 0.0001686695062804623, "loss": 0.3389, "step": 6948 }, { "epoch": 3.2701176470588234, "grad_norm": 0.27156109333440126, "learning_rate": 0.00016862601016851358, "loss": 0.1871, "step": 6949 }, { "epoch": 3.2705882352941176, "grad_norm": 0.2944421030635976, "learning_rate": 0.00016858251246581216, "loss": 0.2059, "step": 6950 }, { "epoch": 3.271058823529412, "grad_norm": 0.4686858868005503, "learning_rate": 0.0001685390131760731, "loss": 0.3267, "step": 6951 }, { "epoch": 3.271529411764706, "grad_norm": 0.37137099822544944, "learning_rate": 0.00016849551230301135, "loss": 0.2979, "step": 6952 }, { "epoch": 3.2720000000000002, "grad_norm": 0.4802560176029697, "learning_rate": 0.00016845200985034215, "loss": 0.3057, "step": 6953 }, { "epoch": 3.272470588235294, "grad_norm": 0.5865022620032865, "learning_rate": 0.0001684085058217808, "loss": 0.4222, "step": 6954 }, { "epoch": 3.2729411764705882, "grad_norm": 0.5988176597027771, "learning_rate": 0.00016836500022104274, "loss": 0.3957, "step": 6955 }, { "epoch": 3.2734117647058825, "grad_norm": 0.13372968389510057, "learning_rate": 0.00016832149305184363, "loss": 0.0303, "step": 6956 }, { "epoch": 3.2738823529411762, "grad_norm": 0.31396009927483515, "learning_rate": 0.00016827798431789907, "loss": 0.1873, "step": 6957 }, { "epoch": 3.2743529411764705, "grad_norm": 0.3674186265297699, "learning_rate": 0.00016823447402292504, "loss": 0.234, "step": 6958 }, { "epoch": 3.2748235294117647, "grad_norm": 0.38690596647515907, "learning_rate": 0.00016819096217063752, "loss": 0.2401, "step": 6959 }, { "epoch": 3.275294117647059, "grad_norm": 0.4271804419186593, "learning_rate": 0.00016814744876475252, "loss": 0.259, "step": 6960 }, { "epoch": 3.275764705882353, "grad_norm": 0.287253095953972, "learning_rate": 0.00016810393380898645, "loss": 0.2447, "step": 6961 }, { "epoch": 3.276235294117647, "grad_norm": 0.2500420886874819, "learning_rate": 0.0001680604173070556, "loss": 0.18, "step": 6962 }, { "epoch": 3.276705882352941, "grad_norm": 0.3040593303135578, "learning_rate": 0.00016801689926267653, "loss": 0.1955, "step": 6963 }, { "epoch": 3.2771764705882354, "grad_norm": 0.5372559620167261, "learning_rate": 0.00016797337967956592, "loss": 0.5759, "step": 6964 }, { "epoch": 3.2776470588235296, "grad_norm": 0.3640880862254217, "learning_rate": 0.00016792985856144049, "loss": 0.2717, "step": 6965 }, { "epoch": 3.2781176470588234, "grad_norm": 0.25361654897095653, "learning_rate": 0.00016788633591201723, "loss": 0.2139, "step": 6966 }, { "epoch": 3.2785882352941176, "grad_norm": 0.23384638463001067, "learning_rate": 0.0001678428117350131, "loss": 0.1357, "step": 6967 }, { "epoch": 3.279058823529412, "grad_norm": 0.39245859877973044, "learning_rate": 0.0001677992860341454, "loss": 0.1795, "step": 6968 }, { "epoch": 3.279529411764706, "grad_norm": 0.6253001104406904, "learning_rate": 0.00016775575881313132, "loss": 0.4182, "step": 6969 }, { "epoch": 3.2800000000000002, "grad_norm": 0.5012707505752886, "learning_rate": 0.00016771223007568836, "loss": 0.3331, "step": 6970 }, { "epoch": 3.280470588235294, "grad_norm": 0.17685948429665937, "learning_rate": 0.00016766869982553405, "loss": 0.0422, "step": 6971 }, { "epoch": 3.2809411764705882, "grad_norm": 0.308246057874883, "learning_rate": 0.0001676251680663861, "loss": 0.1427, "step": 6972 }, { "epoch": 3.2814117647058825, "grad_norm": 0.2349652835713559, "learning_rate": 0.0001675816348019624, "loss": 0.1111, "step": 6973 }, { "epoch": 3.2818823529411763, "grad_norm": 0.27187387059485285, "learning_rate": 0.0001675381000359808, "loss": 0.1948, "step": 6974 }, { "epoch": 3.2823529411764705, "grad_norm": 0.4032146623286002, "learning_rate": 0.0001674945637721595, "loss": 0.1756, "step": 6975 }, { "epoch": 3.2828235294117647, "grad_norm": 0.2060373181946975, "learning_rate": 0.00016745102601421663, "loss": 0.1252, "step": 6976 }, { "epoch": 3.283294117647059, "grad_norm": 0.4066419224973115, "learning_rate": 0.00016740748676587045, "loss": 0.2651, "step": 6977 }, { "epoch": 3.283764705882353, "grad_norm": 0.5752199787730561, "learning_rate": 0.0001673639460308396, "loss": 0.6222, "step": 6978 }, { "epoch": 3.284235294117647, "grad_norm": 0.29423064862451065, "learning_rate": 0.00016732040381284258, "loss": 0.2206, "step": 6979 }, { "epoch": 3.284705882352941, "grad_norm": 0.7676735029550044, "learning_rate": 0.00016727686011559813, "loss": 0.6587, "step": 6980 }, { "epoch": 3.2851764705882354, "grad_norm": 0.18680667818005514, "learning_rate": 0.00016723331494282502, "loss": 0.0999, "step": 6981 }, { "epoch": 3.2856470588235296, "grad_norm": 0.6390206351304816, "learning_rate": 0.00016718976829824234, "loss": 0.5908, "step": 6982 }, { "epoch": 3.2861176470588234, "grad_norm": 0.3481561776171253, "learning_rate": 0.00016714622018556915, "loss": 0.2801, "step": 6983 }, { "epoch": 3.2865882352941176, "grad_norm": 0.2577533861280022, "learning_rate": 0.0001671026706085246, "loss": 0.1815, "step": 6984 }, { "epoch": 3.287058823529412, "grad_norm": 0.17558925677016318, "learning_rate": 0.0001670591195708281, "loss": 0.0959, "step": 6985 }, { "epoch": 3.287529411764706, "grad_norm": 0.2676687611673121, "learning_rate": 0.00016701556707619914, "loss": 0.1738, "step": 6986 }, { "epoch": 3.288, "grad_norm": 0.19614925772225134, "learning_rate": 0.00016697201312835727, "loss": 0.1175, "step": 6987 }, { "epoch": 3.288470588235294, "grad_norm": 0.5312315842259945, "learning_rate": 0.00016692845773102222, "loss": 0.3324, "step": 6988 }, { "epoch": 3.2889411764705883, "grad_norm": 0.6090334832326317, "learning_rate": 0.0001668849008879139, "loss": 0.512, "step": 6989 }, { "epoch": 3.2894117647058825, "grad_norm": 0.15444173230091351, "learning_rate": 0.00016684134260275217, "loss": 0.0776, "step": 6990 }, { "epoch": 3.2898823529411763, "grad_norm": 0.45014318508081286, "learning_rate": 0.0001667977828792572, "loss": 0.2923, "step": 6991 }, { "epoch": 3.2903529411764705, "grad_norm": 0.8313806326864114, "learning_rate": 0.00016675422172114917, "loss": 0.6155, "step": 6992 }, { "epoch": 3.2908235294117647, "grad_norm": 0.3505198927174196, "learning_rate": 0.00016671065913214845, "loss": 0.1601, "step": 6993 }, { "epoch": 3.291294117647059, "grad_norm": 0.4680026459964188, "learning_rate": 0.00016666709511597546, "loss": 0.5702, "step": 6994 }, { "epoch": 3.291764705882353, "grad_norm": 0.31438339685504435, "learning_rate": 0.0001666235296763508, "loss": 0.1579, "step": 6995 }, { "epoch": 3.292235294117647, "grad_norm": 0.2959489320960553, "learning_rate": 0.00016657996281699514, "loss": 0.2276, "step": 6996 }, { "epoch": 3.292705882352941, "grad_norm": 0.2871334004352889, "learning_rate": 0.00016653639454162938, "loss": 0.1185, "step": 6997 }, { "epoch": 3.2931764705882354, "grad_norm": 0.5454906186029, "learning_rate": 0.00016649282485397437, "loss": 0.2907, "step": 6998 }, { "epoch": 3.2936470588235296, "grad_norm": 0.44884957053028157, "learning_rate": 0.00016644925375775124, "loss": 0.1842, "step": 6999 }, { "epoch": 3.2941176470588234, "grad_norm": 0.4532783327953635, "learning_rate": 0.00016640568125668117, "loss": 0.2072, "step": 7000 }, { "epoch": 3.2945882352941176, "grad_norm": 0.24431157801303635, "learning_rate": 0.00016636210735448539, "loss": 0.1024, "step": 7001 }, { "epoch": 3.295058823529412, "grad_norm": 0.2378672969433444, "learning_rate": 0.00016631853205488546, "loss": 0.0873, "step": 7002 }, { "epoch": 3.295529411764706, "grad_norm": 0.5679040785409457, "learning_rate": 0.00016627495536160284, "loss": 0.1098, "step": 7003 }, { "epoch": 3.296, "grad_norm": 0.19349914412576769, "learning_rate": 0.0001662313772783592, "loss": 0.0986, "step": 7004 }, { "epoch": 3.296470588235294, "grad_norm": 0.3446025580414779, "learning_rate": 0.00016618779780887627, "loss": 0.2311, "step": 7005 }, { "epoch": 3.2969411764705883, "grad_norm": 0.1915886264128002, "learning_rate": 0.00016614421695687606, "loss": 0.1021, "step": 7006 }, { "epoch": 3.2974117647058825, "grad_norm": 0.10616532550957107, "learning_rate": 0.00016610063472608053, "loss": 0.0225, "step": 7007 }, { "epoch": 3.2978823529411763, "grad_norm": 0.4351968930223973, "learning_rate": 0.0001660570511202118, "loss": 0.2868, "step": 7008 }, { "epoch": 3.2983529411764705, "grad_norm": 0.537057807059848, "learning_rate": 0.00016601346614299214, "loss": 0.4369, "step": 7009 }, { "epoch": 3.2988235294117647, "grad_norm": 0.2082153949182009, "learning_rate": 0.00016596987979814395, "loss": 0.0664, "step": 7010 }, { "epoch": 3.299294117647059, "grad_norm": 0.090772589181073, "learning_rate": 0.0001659262920893897, "loss": 0.0391, "step": 7011 }, { "epoch": 3.299764705882353, "grad_norm": 0.9404136229301686, "learning_rate": 0.00016588270302045197, "loss": 0.7463, "step": 7012 }, { "epoch": 3.300235294117647, "grad_norm": 0.30943954265279344, "learning_rate": 0.00016583911259505352, "loss": 0.2474, "step": 7013 }, { "epoch": 3.300705882352941, "grad_norm": 0.29771178334558396, "learning_rate": 0.00016579552081691715, "loss": 0.0913, "step": 7014 }, { "epoch": 3.3011764705882354, "grad_norm": 0.42025229278088694, "learning_rate": 0.0001657519276897658, "loss": 0.3267, "step": 7015 }, { "epoch": 3.3016470588235296, "grad_norm": 0.4456637570462965, "learning_rate": 0.0001657083332173226, "loss": 0.36, "step": 7016 }, { "epoch": 3.3021176470588234, "grad_norm": 0.30884295482351554, "learning_rate": 0.0001656647374033107, "loss": 0.2186, "step": 7017 }, { "epoch": 3.3025882352941176, "grad_norm": 0.20325432239016467, "learning_rate": 0.00016562114025145332, "loss": 0.0652, "step": 7018 }, { "epoch": 3.303058823529412, "grad_norm": 0.24123247597978945, "learning_rate": 0.00016557754176547404, "loss": 0.0746, "step": 7019 }, { "epoch": 3.303529411764706, "grad_norm": 0.29768305092570807, "learning_rate": 0.0001655339419490962, "loss": 0.1954, "step": 7020 }, { "epoch": 3.304, "grad_norm": 0.32252001994937296, "learning_rate": 0.00016549034080604359, "loss": 0.2938, "step": 7021 }, { "epoch": 3.304470588235294, "grad_norm": 0.22311307467008207, "learning_rate": 0.00016544673834003985, "loss": 0.1235, "step": 7022 }, { "epoch": 3.3049411764705883, "grad_norm": 0.23998634289041132, "learning_rate": 0.00016540313455480895, "loss": 0.1662, "step": 7023 }, { "epoch": 3.3054117647058825, "grad_norm": 0.23379968073446789, "learning_rate": 0.00016535952945407473, "loss": 0.1368, "step": 7024 }, { "epoch": 3.3058823529411763, "grad_norm": 0.2631794642912977, "learning_rate": 0.00016531592304156137, "loss": 0.2317, "step": 7025 }, { "epoch": 3.3063529411764705, "grad_norm": 0.15007028937579076, "learning_rate": 0.0001652723153209931, "loss": 0.0781, "step": 7026 }, { "epoch": 3.3068235294117647, "grad_norm": 0.24827204110347953, "learning_rate": 0.0001652287062960942, "loss": 0.1621, "step": 7027 }, { "epoch": 3.307294117647059, "grad_norm": 0.5116138158233756, "learning_rate": 0.00016518509597058902, "loss": 0.4581, "step": 7028 }, { "epoch": 3.307764705882353, "grad_norm": 0.35106382035846007, "learning_rate": 0.0001651414843482022, "loss": 0.2391, "step": 7029 }, { "epoch": 3.308235294117647, "grad_norm": 0.38820949805269156, "learning_rate": 0.00016509787143265836, "loss": 0.1962, "step": 7030 }, { "epoch": 3.308705882352941, "grad_norm": 0.17663354932802683, "learning_rate": 0.0001650542572276822, "loss": 0.0654, "step": 7031 }, { "epoch": 3.3091764705882354, "grad_norm": 0.20694374071094265, "learning_rate": 0.00016501064173699866, "loss": 0.1309, "step": 7032 }, { "epoch": 3.309647058823529, "grad_norm": 0.49983694420230773, "learning_rate": 0.0001649670249643327, "loss": 0.3726, "step": 7033 }, { "epoch": 3.3101176470588234, "grad_norm": 0.2208865460113967, "learning_rate": 0.0001649234069134094, "loss": 0.1118, "step": 7034 }, { "epoch": 3.3105882352941176, "grad_norm": 0.1948785267067883, "learning_rate": 0.00016487978758795397, "loss": 0.0997, "step": 7035 }, { "epoch": 3.311058823529412, "grad_norm": 0.42376067314087756, "learning_rate": 0.00016483616699169163, "loss": 0.4982, "step": 7036 }, { "epoch": 3.311529411764706, "grad_norm": 0.326221471713756, "learning_rate": 0.0001647925451283479, "loss": 0.2095, "step": 7037 }, { "epoch": 3.312, "grad_norm": 0.41501716467196015, "learning_rate": 0.00016474892200164825, "loss": 0.1795, "step": 7038 }, { "epoch": 3.312470588235294, "grad_norm": 0.40221549212758556, "learning_rate": 0.0001647052976153183, "loss": 0.2637, "step": 7039 }, { "epoch": 3.3129411764705883, "grad_norm": 0.4524100171622535, "learning_rate": 0.00016466167197308384, "loss": 0.2168, "step": 7040 }, { "epoch": 3.3134117647058825, "grad_norm": 0.28941796849841395, "learning_rate": 0.00016461804507867067, "loss": 0.2111, "step": 7041 }, { "epoch": 3.3138823529411763, "grad_norm": 0.36884640552753417, "learning_rate": 0.00016457441693580472, "loss": 0.2249, "step": 7042 }, { "epoch": 3.3143529411764705, "grad_norm": 0.3175974919303717, "learning_rate": 0.0001645307875482121, "loss": 0.1549, "step": 7043 }, { "epoch": 3.3148235294117647, "grad_norm": 0.31681023449358675, "learning_rate": 0.00016448715691961898, "loss": 0.178, "step": 7044 }, { "epoch": 3.315294117647059, "grad_norm": 0.28906852558242296, "learning_rate": 0.00016444352505375155, "loss": 0.156, "step": 7045 }, { "epoch": 3.315764705882353, "grad_norm": 0.23268251631222908, "learning_rate": 0.00016439989195433626, "loss": 0.0998, "step": 7046 }, { "epoch": 3.316235294117647, "grad_norm": 0.6014846482606823, "learning_rate": 0.00016435625762509955, "loss": 0.3839, "step": 7047 }, { "epoch": 3.316705882352941, "grad_norm": 0.4044337421414201, "learning_rate": 0.00016431262206976805, "loss": 0.1911, "step": 7048 }, { "epoch": 3.3171764705882354, "grad_norm": 0.44714401716731905, "learning_rate": 0.00016426898529206836, "loss": 0.2991, "step": 7049 }, { "epoch": 3.317647058823529, "grad_norm": 0.22516396717927004, "learning_rate": 0.00016422534729572738, "loss": 0.1388, "step": 7050 }, { "epoch": 3.3181176470588234, "grad_norm": 0.2734112015599577, "learning_rate": 0.00016418170808447198, "loss": 0.1752, "step": 7051 }, { "epoch": 3.3185882352941176, "grad_norm": 0.14921275514572788, "learning_rate": 0.00016413806766202912, "loss": 0.0469, "step": 7052 }, { "epoch": 3.319058823529412, "grad_norm": 0.7244630355207367, "learning_rate": 0.00016409442603212594, "loss": 0.3612, "step": 7053 }, { "epoch": 3.319529411764706, "grad_norm": 0.2557629324601194, "learning_rate": 0.00016405078319848965, "loss": 0.0998, "step": 7054 }, { "epoch": 3.32, "grad_norm": 0.22692884996069837, "learning_rate": 0.0001640071391648476, "loss": 0.1202, "step": 7055 }, { "epoch": 3.320470588235294, "grad_norm": 0.2988547951929932, "learning_rate": 0.0001639634939349271, "loss": 0.2488, "step": 7056 }, { "epoch": 3.3209411764705883, "grad_norm": 0.643098198125993, "learning_rate": 0.00016391984751245572, "loss": 0.3136, "step": 7057 }, { "epoch": 3.3214117647058825, "grad_norm": 0.22641314226408252, "learning_rate": 0.00016387619990116116, "loss": 0.0822, "step": 7058 }, { "epoch": 3.3218823529411763, "grad_norm": 0.5215851951001759, "learning_rate": 0.00016383255110477102, "loss": 0.2991, "step": 7059 }, { "epoch": 3.3223529411764705, "grad_norm": 0.2726286002609566, "learning_rate": 0.00016378890112701312, "loss": 0.112, "step": 7060 }, { "epoch": 3.3228235294117647, "grad_norm": 0.500866948965806, "learning_rate": 0.0001637452499716155, "loss": 0.2896, "step": 7061 }, { "epoch": 3.323294117647059, "grad_norm": 0.1754305735340004, "learning_rate": 0.0001637015976423061, "loss": 0.0567, "step": 7062 }, { "epoch": 3.323764705882353, "grad_norm": 0.2975819519688705, "learning_rate": 0.00016365794414281303, "loss": 0.2199, "step": 7063 }, { "epoch": 3.324235294117647, "grad_norm": 0.25594336333102186, "learning_rate": 0.00016361428947686457, "loss": 0.1969, "step": 7064 }, { "epoch": 3.324705882352941, "grad_norm": 0.1444152764882329, "learning_rate": 0.00016357063364818902, "loss": 0.0458, "step": 7065 }, { "epoch": 3.3251764705882354, "grad_norm": 0.2337574762778872, "learning_rate": 0.00016352697666051474, "loss": 0.136, "step": 7066 }, { "epoch": 3.325647058823529, "grad_norm": 0.20888235227750987, "learning_rate": 0.0001634833185175703, "loss": 0.1252, "step": 7067 }, { "epoch": 3.3261176470588234, "grad_norm": 0.39979180673781894, "learning_rate": 0.00016343965922308438, "loss": 0.3497, "step": 7068 }, { "epoch": 3.3265882352941176, "grad_norm": 0.2257126973051158, "learning_rate": 0.00016339599878078563, "loss": 0.1076, "step": 7069 }, { "epoch": 3.327058823529412, "grad_norm": 0.1326245863652446, "learning_rate": 0.00016335233719440282, "loss": 0.0483, "step": 7070 }, { "epoch": 3.327529411764706, "grad_norm": 0.37863160082040637, "learning_rate": 0.000163308674467665, "loss": 0.3374, "step": 7071 }, { "epoch": 3.328, "grad_norm": 0.4612465502723321, "learning_rate": 0.00016326501060430105, "loss": 0.3973, "step": 7072 }, { "epoch": 3.328470588235294, "grad_norm": 0.5685003108923395, "learning_rate": 0.00016322134560804017, "loss": 0.4514, "step": 7073 }, { "epoch": 3.3289411764705883, "grad_norm": 0.35125273510669314, "learning_rate": 0.00016317767948261148, "loss": 0.2521, "step": 7074 }, { "epoch": 3.3294117647058825, "grad_norm": 0.40497159475862726, "learning_rate": 0.00016313401223174434, "loss": 0.2841, "step": 7075 }, { "epoch": 3.3298823529411763, "grad_norm": 0.6464952808656304, "learning_rate": 0.00016309034385916815, "loss": 0.5968, "step": 7076 }, { "epoch": 3.3303529411764705, "grad_norm": 0.14160860109888251, "learning_rate": 0.00016304667436861233, "loss": 0.0389, "step": 7077 }, { "epoch": 3.3308235294117647, "grad_norm": 0.3237923871176505, "learning_rate": 0.00016300300376380657, "loss": 0.3011, "step": 7078 }, { "epoch": 3.331294117647059, "grad_norm": 0.3056517504019126, "learning_rate": 0.00016295933204848052, "loss": 0.125, "step": 7079 }, { "epoch": 3.331764705882353, "grad_norm": 0.296257083442741, "learning_rate": 0.00016291565922636392, "loss": 0.1541, "step": 7080 }, { "epoch": 3.332235294117647, "grad_norm": 0.4171824575993192, "learning_rate": 0.0001628719853011867, "loss": 0.2491, "step": 7081 }, { "epoch": 3.332705882352941, "grad_norm": 0.2974564411779902, "learning_rate": 0.00016282831027667881, "loss": 0.168, "step": 7082 }, { "epoch": 3.3331764705882354, "grad_norm": 0.3912308768538553, "learning_rate": 0.00016278463415657027, "loss": 0.1747, "step": 7083 }, { "epoch": 3.333647058823529, "grad_norm": 1.651753434754446e-05, "learning_rate": 0.00016274095694459128, "loss": 0.0, "step": 7084 }, { "epoch": 3.3341176470588234, "grad_norm": 0.2626360737633639, "learning_rate": 0.00016269727864447206, "loss": 0.1695, "step": 7085 }, { "epoch": 3.3345882352941176, "grad_norm": 0.4194481246109853, "learning_rate": 0.00016265359925994298, "loss": 0.3474, "step": 7086 }, { "epoch": 3.335058823529412, "grad_norm": 0.31129691804578496, "learning_rate": 0.00016260991879473441, "loss": 0.1843, "step": 7087 }, { "epoch": 3.335529411764706, "grad_norm": 0.4504727657195008, "learning_rate": 0.000162566237252577, "loss": 0.2549, "step": 7088 }, { "epoch": 3.336, "grad_norm": 0.3440405662235936, "learning_rate": 0.00016252255463720125, "loss": 0.1855, "step": 7089 }, { "epoch": 3.336470588235294, "grad_norm": 0.32259765406489826, "learning_rate": 0.0001624788709523379, "loss": 0.1322, "step": 7090 }, { "epoch": 3.3369411764705883, "grad_norm": 0.20027066067955043, "learning_rate": 0.00016243518620171778, "loss": 0.1014, "step": 7091 }, { "epoch": 3.3374117647058825, "grad_norm": 0.39168234095637733, "learning_rate": 0.00016239150038907177, "loss": 0.2645, "step": 7092 }, { "epoch": 3.3378823529411763, "grad_norm": 0.3295062086899315, "learning_rate": 0.00016234781351813079, "loss": 0.264, "step": 7093 }, { "epoch": 3.3383529411764705, "grad_norm": 0.34670639504467404, "learning_rate": 0.00016230412559262606, "loss": 0.3166, "step": 7094 }, { "epoch": 3.3388235294117647, "grad_norm": 0.14143202109537098, "learning_rate": 0.00016226043661628862, "loss": 0.0668, "step": 7095 }, { "epoch": 3.339294117647059, "grad_norm": 0.303239138485577, "learning_rate": 0.00016221674659284976, "loss": 0.1928, "step": 7096 }, { "epoch": 3.339764705882353, "grad_norm": 0.4771453177339286, "learning_rate": 0.00016217305552604076, "loss": 0.4293, "step": 7097 }, { "epoch": 3.340235294117647, "grad_norm": 0.15145265994937712, "learning_rate": 0.00016212936341959312, "loss": 0.0488, "step": 7098 }, { "epoch": 3.340705882352941, "grad_norm": 0.322745903844112, "learning_rate": 0.0001620856702772384, "loss": 0.2206, "step": 7099 }, { "epoch": 3.3411764705882354, "grad_norm": 0.37151161006208633, "learning_rate": 0.00016204197610270816, "loss": 0.307, "step": 7100 }, { "epoch": 3.341647058823529, "grad_norm": 0.7411173735221823, "learning_rate": 0.00016199828089973406, "loss": 0.5516, "step": 7101 }, { "epoch": 3.3421176470588234, "grad_norm": 0.4213087445336398, "learning_rate": 0.00016195458467204793, "loss": 0.3157, "step": 7102 }, { "epoch": 3.3425882352941176, "grad_norm": 0.22893957198880763, "learning_rate": 0.00016191088742338166, "loss": 0.0922, "step": 7103 }, { "epoch": 3.343058823529412, "grad_norm": 0.5719733976273295, "learning_rate": 0.00016186718915746714, "loss": 0.3871, "step": 7104 }, { "epoch": 3.343529411764706, "grad_norm": 0.36911815979713264, "learning_rate": 0.00016182348987803648, "loss": 0.4109, "step": 7105 }, { "epoch": 3.344, "grad_norm": 0.28094767437646495, "learning_rate": 0.0001617797895888218, "loss": 0.3126, "step": 7106 }, { "epoch": 3.344470588235294, "grad_norm": 0.42896574936018467, "learning_rate": 0.00016173608829355527, "loss": 0.2742, "step": 7107 }, { "epoch": 3.3449411764705883, "grad_norm": 0.40606316257537783, "learning_rate": 0.00016169238599596927, "loss": 0.3679, "step": 7108 }, { "epoch": 3.3454117647058825, "grad_norm": 0.43542317450943946, "learning_rate": 0.00016164868269979619, "loss": 0.287, "step": 7109 }, { "epoch": 3.3458823529411763, "grad_norm": 0.2700984209948891, "learning_rate": 0.0001616049784087684, "loss": 0.2039, "step": 7110 }, { "epoch": 3.3463529411764705, "grad_norm": 0.36408945586283725, "learning_rate": 0.00016156127312661855, "loss": 0.239, "step": 7111 }, { "epoch": 3.3468235294117648, "grad_norm": 0.49285084236423005, "learning_rate": 0.0001615175668570793, "loss": 0.4401, "step": 7112 }, { "epoch": 3.347294117647059, "grad_norm": 0.3274246097847283, "learning_rate": 0.0001614738596038833, "loss": 0.23, "step": 7113 }, { "epoch": 3.3477647058823528, "grad_norm": 0.2625229812469338, "learning_rate": 0.00016143015137076345, "loss": 0.1311, "step": 7114 }, { "epoch": 3.348235294117647, "grad_norm": 0.2166266557015221, "learning_rate": 0.0001613864421614526, "loss": 0.0918, "step": 7115 }, { "epoch": 3.348705882352941, "grad_norm": 0.4668817789044859, "learning_rate": 0.00016134273197968374, "loss": 0.237, "step": 7116 }, { "epoch": 3.3491764705882354, "grad_norm": 0.1769075232542796, "learning_rate": 0.0001612990208291899, "loss": 0.1127, "step": 7117 }, { "epoch": 3.349647058823529, "grad_norm": 0.1815833596873622, "learning_rate": 0.0001612553087137043, "loss": 0.0963, "step": 7118 }, { "epoch": 3.3501176470588234, "grad_norm": 0.4478241521092322, "learning_rate": 0.0001612115956369601, "loss": 0.2956, "step": 7119 }, { "epoch": 3.3505882352941176, "grad_norm": 0.6023476103351211, "learning_rate": 0.00016116788160269065, "loss": 0.4076, "step": 7120 }, { "epoch": 3.351058823529412, "grad_norm": 0.46913918825748724, "learning_rate": 0.0001611241666146293, "loss": 0.3136, "step": 7121 }, { "epoch": 3.351529411764706, "grad_norm": 0.39781800378596904, "learning_rate": 0.00016108045067650962, "loss": 0.1786, "step": 7122 }, { "epoch": 3.352, "grad_norm": 0.2768920613956153, "learning_rate": 0.0001610367337920651, "loss": 0.2521, "step": 7123 }, { "epoch": 3.352470588235294, "grad_norm": 0.3877324858918899, "learning_rate": 0.00016099301596502936, "loss": 0.3233, "step": 7124 }, { "epoch": 3.3529411764705883, "grad_norm": 0.29719603526945493, "learning_rate": 0.00016094929719913612, "loss": 0.2062, "step": 7125 }, { "epoch": 3.3534117647058825, "grad_norm": 0.3565800333073179, "learning_rate": 0.0001609055774981192, "loss": 0.2915, "step": 7126 }, { "epoch": 3.3538823529411763, "grad_norm": 0.2504814387353851, "learning_rate": 0.0001608618568657125, "loss": 0.1425, "step": 7127 }, { "epoch": 3.3543529411764705, "grad_norm": 0.4468879274331284, "learning_rate": 0.0001608181353056499, "loss": 0.4103, "step": 7128 }, { "epoch": 3.3548235294117648, "grad_norm": 0.5751999160629206, "learning_rate": 0.0001607744128216655, "loss": 0.5469, "step": 7129 }, { "epoch": 3.355294117647059, "grad_norm": 0.3091974504567729, "learning_rate": 0.00016073068941749343, "loss": 0.2577, "step": 7130 }, { "epoch": 3.3557647058823528, "grad_norm": 0.30127514741851735, "learning_rate": 0.0001606869650968678, "loss": 0.1662, "step": 7131 }, { "epoch": 3.356235294117647, "grad_norm": 0.2850787207292117, "learning_rate": 0.00016064323986352295, "loss": 0.1772, "step": 7132 }, { "epoch": 3.356705882352941, "grad_norm": 0.1250441706638147, "learning_rate": 0.00016059951372119323, "loss": 0.0336, "step": 7133 }, { "epoch": 3.3571764705882354, "grad_norm": 0.1415645003250034, "learning_rate": 0.000160555786673613, "loss": 0.0386, "step": 7134 }, { "epoch": 3.357647058823529, "grad_norm": 0.12080467521091913, "learning_rate": 0.00016051205872451686, "loss": 0.0487, "step": 7135 }, { "epoch": 3.3581176470588234, "grad_norm": 0.34450695859935465, "learning_rate": 0.00016046832987763935, "loss": 0.1801, "step": 7136 }, { "epoch": 3.3585882352941177, "grad_norm": 0.2957036742324408, "learning_rate": 0.0001604246001367151, "loss": 0.146, "step": 7137 }, { "epoch": 3.359058823529412, "grad_norm": 0.5527889699508052, "learning_rate": 0.00016038086950547885, "loss": 0.4392, "step": 7138 }, { "epoch": 3.359529411764706, "grad_norm": 0.5545242267375873, "learning_rate": 0.00016033713798766546, "loss": 0.2562, "step": 7139 }, { "epoch": 3.36, "grad_norm": 0.22611551881667602, "learning_rate": 0.00016029340558700977, "loss": 0.0978, "step": 7140 }, { "epoch": 3.360470588235294, "grad_norm": 0.44199503963775444, "learning_rate": 0.00016024967230724672, "loss": 0.3387, "step": 7141 }, { "epoch": 3.3609411764705883, "grad_norm": 0.27185561650542467, "learning_rate": 0.00016020593815211138, "loss": 0.152, "step": 7142 }, { "epoch": 3.3614117647058825, "grad_norm": 0.23793403644233946, "learning_rate": 0.00016016220312533888, "loss": 0.1078, "step": 7143 }, { "epoch": 3.3618823529411763, "grad_norm": 0.31274178428445293, "learning_rate": 0.00016011846723066445, "loss": 0.1255, "step": 7144 }, { "epoch": 3.3623529411764705, "grad_norm": 0.3043390102775579, "learning_rate": 0.0001600747304718232, "loss": 0.1549, "step": 7145 }, { "epoch": 3.3628235294117648, "grad_norm": 0.13188643285225213, "learning_rate": 0.00016003099285255058, "loss": 0.0466, "step": 7146 }, { "epoch": 3.363294117647059, "grad_norm": 0.4186294085478646, "learning_rate": 0.000159987254376582, "loss": 0.3111, "step": 7147 }, { "epoch": 3.3637647058823528, "grad_norm": 0.18552399725633012, "learning_rate": 0.00015994351504765287, "loss": 0.1005, "step": 7148 }, { "epoch": 3.364235294117647, "grad_norm": 0.6328556845885213, "learning_rate": 0.00015989977486949883, "loss": 0.5077, "step": 7149 }, { "epoch": 3.364705882352941, "grad_norm": 0.38591408453679826, "learning_rate": 0.00015985603384585542, "loss": 0.3296, "step": 7150 }, { "epoch": 3.3651764705882354, "grad_norm": 0.21073649984609258, "learning_rate": 0.0001598122919804584, "loss": 0.0752, "step": 7151 }, { "epoch": 3.365647058823529, "grad_norm": 0.31124169880103303, "learning_rate": 0.0001597685492770435, "loss": 0.1746, "step": 7152 }, { "epoch": 3.3661176470588234, "grad_norm": 0.10527343369124655, "learning_rate": 0.00015972480573934656, "loss": 0.0244, "step": 7153 }, { "epoch": 3.3665882352941177, "grad_norm": 0.9266062315526743, "learning_rate": 0.00015968106137110352, "loss": 0.5615, "step": 7154 }, { "epoch": 3.367058823529412, "grad_norm": 0.39773603129373136, "learning_rate": 0.00015963731617605034, "loss": 0.2639, "step": 7155 }, { "epoch": 3.367529411764706, "grad_norm": 0.27560596989735586, "learning_rate": 0.00015959357015792313, "loss": 0.1771, "step": 7156 }, { "epoch": 3.368, "grad_norm": 0.32846256021321096, "learning_rate": 0.00015954982332045794, "loss": 0.3225, "step": 7157 }, { "epoch": 3.368470588235294, "grad_norm": 0.16424817460456456, "learning_rate": 0.000159506075667391, "loss": 0.0391, "step": 7158 }, { "epoch": 3.3689411764705883, "grad_norm": 0.16736874559515227, "learning_rate": 0.0001594623272024586, "loss": 0.0712, "step": 7159 }, { "epoch": 3.369411764705882, "grad_norm": 0.41010210421722365, "learning_rate": 0.000159418577929397, "loss": 0.3123, "step": 7160 }, { "epoch": 3.3698823529411763, "grad_norm": 0.28773252432137963, "learning_rate": 0.0001593748278519427, "loss": 0.1398, "step": 7161 }, { "epoch": 3.3703529411764706, "grad_norm": 0.21552957766290382, "learning_rate": 0.00015933107697383206, "loss": 0.0986, "step": 7162 }, { "epoch": 3.3708235294117648, "grad_norm": 0.3745893265035678, "learning_rate": 0.0001592873252988017, "loss": 0.1516, "step": 7163 }, { "epoch": 3.371294117647059, "grad_norm": 0.23741079037356588, "learning_rate": 0.0001592435728305882, "loss": 0.1105, "step": 7164 }, { "epoch": 3.3717647058823528, "grad_norm": 0.33318399100319557, "learning_rate": 0.00015919981957292825, "loss": 0.2303, "step": 7165 }, { "epoch": 3.372235294117647, "grad_norm": 0.38684331930074134, "learning_rate": 0.00015915606552955856, "loss": 0.2681, "step": 7166 }, { "epoch": 3.372705882352941, "grad_norm": 0.2300868888240794, "learning_rate": 0.000159112310704216, "loss": 0.1008, "step": 7167 }, { "epoch": 3.3731764705882354, "grad_norm": 0.25243484674401295, "learning_rate": 0.00015906855510063744, "loss": 0.1172, "step": 7168 }, { "epoch": 3.373647058823529, "grad_norm": 0.5451965180257095, "learning_rate": 0.0001590247987225597, "loss": 0.4087, "step": 7169 }, { "epoch": 3.3741176470588234, "grad_norm": 0.6106560694003805, "learning_rate": 0.00015898104157371993, "loss": 0.51, "step": 7170 }, { "epoch": 3.3745882352941177, "grad_norm": 0.40164532631069266, "learning_rate": 0.00015893728365785518, "loss": 0.3361, "step": 7171 }, { "epoch": 3.375058823529412, "grad_norm": 0.21340016957342597, "learning_rate": 0.00015889352497870253, "loss": 0.0819, "step": 7172 }, { "epoch": 3.375529411764706, "grad_norm": 0.3095846304748837, "learning_rate": 0.00015884976553999925, "loss": 0.1988, "step": 7173 }, { "epoch": 3.376, "grad_norm": 0.16938396106948594, "learning_rate": 0.0001588060053454826, "loss": 0.0621, "step": 7174 }, { "epoch": 3.376470588235294, "grad_norm": 0.8896186353842813, "learning_rate": 0.0001587622443988899, "loss": 0.7205, "step": 7175 }, { "epoch": 3.3769411764705883, "grad_norm": 0.4730168181093669, "learning_rate": 0.0001587184827039585, "loss": 0.247, "step": 7176 }, { "epoch": 3.377411764705882, "grad_norm": 0.2566598735418769, "learning_rate": 0.00015867472026442596, "loss": 0.1122, "step": 7177 }, { "epoch": 3.3778823529411763, "grad_norm": 0.43538457664321606, "learning_rate": 0.00015863095708402972, "loss": 0.3286, "step": 7178 }, { "epoch": 3.3783529411764706, "grad_norm": 0.4484814717752218, "learning_rate": 0.00015858719316650744, "loss": 0.4565, "step": 7179 }, { "epoch": 3.378823529411765, "grad_norm": 0.3178353989483293, "learning_rate": 0.00015854342851559673, "loss": 0.2564, "step": 7180 }, { "epoch": 3.379294117647059, "grad_norm": 0.10164780151991053, "learning_rate": 0.00015849966313503537, "loss": 0.0325, "step": 7181 }, { "epoch": 3.379764705882353, "grad_norm": 0.27027832035076016, "learning_rate": 0.00015845589702856102, "loss": 0.0975, "step": 7182 }, { "epoch": 3.380235294117647, "grad_norm": 0.37221920668125824, "learning_rate": 0.0001584121301999116, "loss": 0.2259, "step": 7183 }, { "epoch": 3.3807058823529412, "grad_norm": 0.5236315359256912, "learning_rate": 0.00015836836265282502, "loss": 0.3474, "step": 7184 }, { "epoch": 3.3811764705882354, "grad_norm": 0.5467746335163146, "learning_rate": 0.0001583245943910392, "loss": 0.3076, "step": 7185 }, { "epoch": 3.3816470588235292, "grad_norm": 0.4032286033262, "learning_rate": 0.0001582808254182922, "loss": 0.3804, "step": 7186 }, { "epoch": 3.3821176470588235, "grad_norm": 0.3800484618535002, "learning_rate": 0.0001582370557383221, "loss": 0.1875, "step": 7187 }, { "epoch": 3.3825882352941177, "grad_norm": 0.4528117081629833, "learning_rate": 0.00015819328535486704, "loss": 0.515, "step": 7188 }, { "epoch": 3.383058823529412, "grad_norm": 0.3326950201429194, "learning_rate": 0.0001581495142716652, "loss": 0.211, "step": 7189 }, { "epoch": 3.383529411764706, "grad_norm": 0.4471627809736816, "learning_rate": 0.00015810574249245488, "loss": 0.4467, "step": 7190 }, { "epoch": 3.384, "grad_norm": 0.28626856233535175, "learning_rate": 0.00015806197002097438, "loss": 0.2524, "step": 7191 }, { "epoch": 3.384470588235294, "grad_norm": 0.2585638510790312, "learning_rate": 0.00015801819686096213, "loss": 0.1007, "step": 7192 }, { "epoch": 3.3849411764705883, "grad_norm": 0.42104630864735976, "learning_rate": 0.00015797442301615652, "loss": 0.3117, "step": 7193 }, { "epoch": 3.385411764705882, "grad_norm": 0.37668917202094643, "learning_rate": 0.00015793064849029608, "loss": 0.2073, "step": 7194 }, { "epoch": 3.3858823529411763, "grad_norm": 0.2862182344279979, "learning_rate": 0.00015788687328711933, "loss": 0.2675, "step": 7195 }, { "epoch": 3.3863529411764706, "grad_norm": 0.2643478136119016, "learning_rate": 0.00015784309741036492, "loss": 0.0867, "step": 7196 }, { "epoch": 3.386823529411765, "grad_norm": 0.44359520493372195, "learning_rate": 0.00015779932086377153, "loss": 0.3002, "step": 7197 }, { "epoch": 3.387294117647059, "grad_norm": 0.2808206694885822, "learning_rate": 0.00015775554365107791, "loss": 0.1662, "step": 7198 }, { "epoch": 3.387764705882353, "grad_norm": 0.4245861227884116, "learning_rate": 0.0001577117657760228, "loss": 0.3323, "step": 7199 }, { "epoch": 3.388235294117647, "grad_norm": 0.531794247404235, "learning_rate": 0.00015766798724234506, "loss": 0.412, "step": 7200 }, { "epoch": 3.3887058823529412, "grad_norm": 0.3006777280638244, "learning_rate": 0.00015762420805378363, "loss": 0.2906, "step": 7201 }, { "epoch": 3.3891764705882355, "grad_norm": 0.22126052728867202, "learning_rate": 0.00015758042821407742, "loss": 0.0966, "step": 7202 }, { "epoch": 3.3896470588235292, "grad_norm": 0.4476867486045969, "learning_rate": 0.00015753664772696544, "loss": 0.3961, "step": 7203 }, { "epoch": 3.3901176470588235, "grad_norm": 0.4481540711186053, "learning_rate": 0.00015749286659618678, "loss": 0.2462, "step": 7204 }, { "epoch": 3.3905882352941177, "grad_norm": 0.30410673115947773, "learning_rate": 0.00015744908482548059, "loss": 0.2195, "step": 7205 }, { "epoch": 3.391058823529412, "grad_norm": 0.28194377644775864, "learning_rate": 0.00015740530241858597, "loss": 0.1222, "step": 7206 }, { "epoch": 3.391529411764706, "grad_norm": 0.22828750931265024, "learning_rate": 0.00015736151937924224, "loss": 0.1358, "step": 7207 }, { "epoch": 3.392, "grad_norm": 0.4291199537962168, "learning_rate": 0.00015731773571118864, "loss": 0.298, "step": 7208 }, { "epoch": 3.392470588235294, "grad_norm": 0.4139714825583532, "learning_rate": 0.00015727395141816448, "loss": 0.305, "step": 7209 }, { "epoch": 3.3929411764705883, "grad_norm": 0.5083589063200284, "learning_rate": 0.00015723016650390922, "loss": 0.2628, "step": 7210 }, { "epoch": 3.393411764705882, "grad_norm": 0.3404989066444066, "learning_rate": 0.0001571863809721623, "loss": 0.2187, "step": 7211 }, { "epoch": 3.3938823529411764, "grad_norm": 0.26255064265193734, "learning_rate": 0.00015714259482666317, "loss": 0.1243, "step": 7212 }, { "epoch": 3.3943529411764706, "grad_norm": 0.6215704470615674, "learning_rate": 0.00015709880807115138, "loss": 0.4849, "step": 7213 }, { "epoch": 3.394823529411765, "grad_norm": 0.20072274667097487, "learning_rate": 0.0001570550207093666, "loss": 0.0899, "step": 7214 }, { "epoch": 3.395294117647059, "grad_norm": 0.3966384845609658, "learning_rate": 0.00015701123274504845, "loss": 0.283, "step": 7215 }, { "epoch": 3.395764705882353, "grad_norm": 0.19411019832892487, "learning_rate": 0.00015696744418193664, "loss": 0.0731, "step": 7216 }, { "epoch": 3.396235294117647, "grad_norm": 0.3593560405738249, "learning_rate": 0.00015692365502377088, "loss": 0.2502, "step": 7217 }, { "epoch": 3.3967058823529412, "grad_norm": 0.4443405634349097, "learning_rate": 0.00015687986527429103, "loss": 0.336, "step": 7218 }, { "epoch": 3.3971764705882355, "grad_norm": 0.7196769757554531, "learning_rate": 0.00015683607493723699, "loss": 0.4995, "step": 7219 }, { "epoch": 3.3976470588235292, "grad_norm": 0.3817667543266105, "learning_rate": 0.00015679228401634855, "loss": 0.3, "step": 7220 }, { "epoch": 3.3981176470588235, "grad_norm": 0.47824800158515063, "learning_rate": 0.0001567484925153658, "loss": 0.4557, "step": 7221 }, { "epoch": 3.3985882352941177, "grad_norm": 0.30744878012762283, "learning_rate": 0.00015670470043802867, "loss": 0.1374, "step": 7222 }, { "epoch": 3.399058823529412, "grad_norm": 0.22484687449297072, "learning_rate": 0.0001566609077880772, "loss": 0.0945, "step": 7223 }, { "epoch": 3.3995294117647057, "grad_norm": 0.15034737463796574, "learning_rate": 0.00015661711456925158, "loss": 0.0773, "step": 7224 }, { "epoch": 3.4, "grad_norm": 0.32853274868575316, "learning_rate": 0.0001565733207852919, "loss": 0.2437, "step": 7225 }, { "epoch": 3.400470588235294, "grad_norm": 0.24397772782689683, "learning_rate": 0.00015652952643993841, "loss": 0.1279, "step": 7226 }, { "epoch": 3.4009411764705884, "grad_norm": 0.2437043262135692, "learning_rate": 0.00015648573153693128, "loss": 0.2222, "step": 7227 }, { "epoch": 3.401411764705882, "grad_norm": 0.2165067078088061, "learning_rate": 0.0001564419360800109, "loss": 0.1261, "step": 7228 }, { "epoch": 3.4018823529411764, "grad_norm": 0.293241700526244, "learning_rate": 0.0001563981400729176, "loss": 0.1654, "step": 7229 }, { "epoch": 3.4023529411764706, "grad_norm": 0.5404574581734685, "learning_rate": 0.0001563543435193917, "loss": 0.3965, "step": 7230 }, { "epoch": 3.402823529411765, "grad_norm": 0.41686313925193974, "learning_rate": 0.0001563105464231737, "loss": 0.3002, "step": 7231 }, { "epoch": 3.403294117647059, "grad_norm": 0.45754441354423453, "learning_rate": 0.0001562667487880041, "loss": 0.3264, "step": 7232 }, { "epoch": 3.403764705882353, "grad_norm": 0.2506217336204624, "learning_rate": 0.00015622295061762346, "loss": 0.1343, "step": 7233 }, { "epoch": 3.404235294117647, "grad_norm": 0.3230637089541007, "learning_rate": 0.00015617915191577222, "loss": 0.1983, "step": 7234 }, { "epoch": 3.4047058823529412, "grad_norm": 0.09175442049026106, "learning_rate": 0.0001561353526861912, "loss": 0.029, "step": 7235 }, { "epoch": 3.4051764705882355, "grad_norm": 0.28395911442451444, "learning_rate": 0.0001560915529326209, "loss": 0.1779, "step": 7236 }, { "epoch": 3.4056470588235292, "grad_norm": 0.37323628652537094, "learning_rate": 0.00015604775265880212, "loss": 0.318, "step": 7237 }, { "epoch": 3.4061176470588235, "grad_norm": 0.168685742130338, "learning_rate": 0.0001560039518684756, "loss": 0.0946, "step": 7238 }, { "epoch": 3.4065882352941177, "grad_norm": 0.22958748524936484, "learning_rate": 0.00015596015056538215, "loss": 0.1268, "step": 7239 }, { "epoch": 3.407058823529412, "grad_norm": 0.5197477432543535, "learning_rate": 0.0001559163487532626, "loss": 0.4417, "step": 7240 }, { "epoch": 3.4075294117647057, "grad_norm": 0.28434761744505693, "learning_rate": 0.00015587254643585784, "loss": 0.1486, "step": 7241 }, { "epoch": 3.408, "grad_norm": 0.20863318800516792, "learning_rate": 0.00015582874361690885, "loss": 0.0887, "step": 7242 }, { "epoch": 3.408470588235294, "grad_norm": 0.5501579619937566, "learning_rate": 0.00015578494030015652, "loss": 0.5448, "step": 7243 }, { "epoch": 3.4089411764705884, "grad_norm": 0.545209480729958, "learning_rate": 0.00015574113648934192, "loss": 0.3215, "step": 7244 }, { "epoch": 3.409411764705882, "grad_norm": 0.2783755309403142, "learning_rate": 0.0001556973321882061, "loss": 0.1647, "step": 7245 }, { "epoch": 3.4098823529411764, "grad_norm": 0.3498860158538791, "learning_rate": 0.0001556535274004902, "loss": 0.1508, "step": 7246 }, { "epoch": 3.4103529411764706, "grad_norm": 0.20984644853622542, "learning_rate": 0.00015560972212993527, "loss": 0.108, "step": 7247 }, { "epoch": 3.410823529411765, "grad_norm": 0.40320149795203875, "learning_rate": 0.0001555659163802826, "loss": 0.2105, "step": 7248 }, { "epoch": 3.411294117647059, "grad_norm": 0.4440892159827733, "learning_rate": 0.00015552211015527336, "loss": 0.3585, "step": 7249 }, { "epoch": 3.411764705882353, "grad_norm": 0.28106335295524437, "learning_rate": 0.00015547830345864885, "loss": 0.0449, "step": 7250 }, { "epoch": 3.412235294117647, "grad_norm": 0.1732164390875284, "learning_rate": 0.00015543449629415028, "loss": 0.0668, "step": 7251 }, { "epoch": 3.4127058823529413, "grad_norm": 0.2895461238463543, "learning_rate": 0.00015539068866551916, "loss": 0.1808, "step": 7252 }, { "epoch": 3.4131764705882355, "grad_norm": 0.1856922366610948, "learning_rate": 0.00015534688057649673, "loss": 0.0485, "step": 7253 }, { "epoch": 3.4136470588235293, "grad_norm": 0.3879316542864649, "learning_rate": 0.0001553030720308245, "loss": 0.3561, "step": 7254 }, { "epoch": 3.4141176470588235, "grad_norm": 0.35522468380971445, "learning_rate": 0.00015525926303224387, "loss": 0.1811, "step": 7255 }, { "epoch": 3.4145882352941177, "grad_norm": 0.4441672884387748, "learning_rate": 0.0001552154535844964, "loss": 0.3338, "step": 7256 }, { "epoch": 3.415058823529412, "grad_norm": 0.5416707050338976, "learning_rate": 0.00015517164369132364, "loss": 0.1879, "step": 7257 }, { "epoch": 3.4155294117647057, "grad_norm": 0.49704864351349215, "learning_rate": 0.0001551278333564671, "loss": 0.3925, "step": 7258 }, { "epoch": 3.416, "grad_norm": 0.3502839026330885, "learning_rate": 0.00015508402258366846, "loss": 0.224, "step": 7259 }, { "epoch": 3.416470588235294, "grad_norm": 0.3473587977996994, "learning_rate": 0.00015504021137666938, "loss": 0.2918, "step": 7260 }, { "epoch": 3.4169411764705884, "grad_norm": 0.47870266848391185, "learning_rate": 0.00015499639973921148, "loss": 0.3638, "step": 7261 }, { "epoch": 3.417411764705882, "grad_norm": 0.06366782496877144, "learning_rate": 0.00015495258767503657, "loss": 0.022, "step": 7262 }, { "epoch": 3.4178823529411764, "grad_norm": 0.28202894254528926, "learning_rate": 0.0001549087751878864, "loss": 0.1823, "step": 7263 }, { "epoch": 3.4183529411764706, "grad_norm": 0.24805947877116324, "learning_rate": 0.00015486496228150266, "loss": 0.0705, "step": 7264 }, { "epoch": 3.418823529411765, "grad_norm": 0.3409551677552689, "learning_rate": 0.00015482114895962735, "loss": 0.2646, "step": 7265 }, { "epoch": 3.419294117647059, "grad_norm": 0.7371555637563808, "learning_rate": 0.00015477733522600227, "loss": 0.6351, "step": 7266 }, { "epoch": 3.419764705882353, "grad_norm": 0.3177223596387886, "learning_rate": 0.00015473352108436934, "loss": 0.1696, "step": 7267 }, { "epoch": 3.420235294117647, "grad_norm": 0.4598964936982903, "learning_rate": 0.00015468970653847043, "loss": 0.3346, "step": 7268 }, { "epoch": 3.4207058823529413, "grad_norm": 0.45066249881037446, "learning_rate": 0.00015464589159204764, "loss": 0.3153, "step": 7269 }, { "epoch": 3.4211764705882355, "grad_norm": 0.23336830056342, "learning_rate": 0.0001546020762488429, "loss": 0.1113, "step": 7270 }, { "epoch": 3.4216470588235293, "grad_norm": 0.35462157373476233, "learning_rate": 0.00015455826051259826, "loss": 0.1595, "step": 7271 }, { "epoch": 3.4221176470588235, "grad_norm": 0.2628587169448892, "learning_rate": 0.00015451444438705584, "loss": 0.1491, "step": 7272 }, { "epoch": 3.4225882352941177, "grad_norm": 0.2885630516938848, "learning_rate": 0.00015447062787595772, "loss": 0.0996, "step": 7273 }, { "epoch": 3.423058823529412, "grad_norm": 0.39886213847051055, "learning_rate": 0.00015442681098304603, "loss": 0.2803, "step": 7274 }, { "epoch": 3.4235294117647057, "grad_norm": 0.4301887101611416, "learning_rate": 0.000154382993712063, "loss": 0.2616, "step": 7275 }, { "epoch": 3.424, "grad_norm": 0.37731198579304165, "learning_rate": 0.00015433917606675075, "loss": 0.3092, "step": 7276 }, { "epoch": 3.424470588235294, "grad_norm": 0.2593820375119645, "learning_rate": 0.00015429535805085162, "loss": 0.1345, "step": 7277 }, { "epoch": 3.4249411764705884, "grad_norm": 0.5239091380018642, "learning_rate": 0.00015425153966810783, "loss": 0.4584, "step": 7278 }, { "epoch": 3.425411764705882, "grad_norm": 0.22118095432019305, "learning_rate": 0.00015420772092226173, "loss": 0.1011, "step": 7279 }, { "epoch": 3.4258823529411764, "grad_norm": 0.35153332488811073, "learning_rate": 0.00015416390181705562, "loss": 0.3619, "step": 7280 }, { "epoch": 3.4263529411764706, "grad_norm": 0.2944062925934922, "learning_rate": 0.00015412008235623186, "loss": 0.2244, "step": 7281 }, { "epoch": 3.426823529411765, "grad_norm": 0.3536142661896626, "learning_rate": 0.00015407626254353282, "loss": 0.3172, "step": 7282 }, { "epoch": 3.427294117647059, "grad_norm": 0.1575828625707199, "learning_rate": 0.00015403244238270102, "loss": 0.0726, "step": 7283 }, { "epoch": 3.427764705882353, "grad_norm": 0.26773510360272784, "learning_rate": 0.00015398862187747888, "loss": 0.2748, "step": 7284 }, { "epoch": 3.428235294117647, "grad_norm": 0.38389851093977384, "learning_rate": 0.0001539448010316088, "loss": 0.1611, "step": 7285 }, { "epoch": 3.4287058823529413, "grad_norm": 0.46207057845237653, "learning_rate": 0.0001539009798488334, "loss": 0.3064, "step": 7286 }, { "epoch": 3.429176470588235, "grad_norm": 0.33706779665910974, "learning_rate": 0.0001538571583328952, "loss": 0.2181, "step": 7287 }, { "epoch": 3.4296470588235293, "grad_norm": 0.10828476331126097, "learning_rate": 0.0001538133364875367, "loss": 0.0229, "step": 7288 }, { "epoch": 3.4301176470588235, "grad_norm": 0.6217979359222593, "learning_rate": 0.00015376951431650063, "loss": 0.4699, "step": 7289 }, { "epoch": 3.4305882352941177, "grad_norm": 0.6198628977951286, "learning_rate": 0.00015372569182352952, "loss": 0.5336, "step": 7290 }, { "epoch": 3.431058823529412, "grad_norm": 0.3779580764316771, "learning_rate": 0.00015368186901236605, "loss": 0.2853, "step": 7291 }, { "epoch": 3.4315294117647057, "grad_norm": 0.449502196087152, "learning_rate": 0.00015363804588675292, "loss": 0.1502, "step": 7292 }, { "epoch": 3.432, "grad_norm": 0.0637913569792831, "learning_rate": 0.00015359422245043278, "loss": 0.01, "step": 7293 }, { "epoch": 3.432470588235294, "grad_norm": 0.2756916633467224, "learning_rate": 0.00015355039870714848, "loss": 0.2155, "step": 7294 }, { "epoch": 3.4329411764705884, "grad_norm": 0.23426164738096106, "learning_rate": 0.00015350657466064264, "loss": 0.1191, "step": 7295 }, { "epoch": 3.433411764705882, "grad_norm": 0.22600195744520754, "learning_rate": 0.00015346275031465818, "loss": 0.1214, "step": 7296 }, { "epoch": 3.4338823529411764, "grad_norm": 0.24997725892953274, "learning_rate": 0.00015341892567293785, "loss": 0.1568, "step": 7297 }, { "epoch": 3.4343529411764706, "grad_norm": 0.38933863371150895, "learning_rate": 0.0001533751007392245, "loss": 0.2885, "step": 7298 }, { "epoch": 3.434823529411765, "grad_norm": 0.4798034209032733, "learning_rate": 0.00015333127551726096, "loss": 0.4031, "step": 7299 }, { "epoch": 3.435294117647059, "grad_norm": 0.6358418593044061, "learning_rate": 0.0001532874500107902, "loss": 0.608, "step": 7300 }, { "epoch": 3.435764705882353, "grad_norm": 0.293014823984612, "learning_rate": 0.00015324362422355507, "loss": 0.1329, "step": 7301 }, { "epoch": 3.436235294117647, "grad_norm": 0.2788004663343579, "learning_rate": 0.00015319979815929847, "loss": 0.1386, "step": 7302 }, { "epoch": 3.4367058823529413, "grad_norm": 0.35687896069718894, "learning_rate": 0.00015315597182176348, "loss": 0.2188, "step": 7303 }, { "epoch": 3.437176470588235, "grad_norm": 0.22523815492321217, "learning_rate": 0.000153112145214693, "loss": 0.2244, "step": 7304 }, { "epoch": 3.4376470588235293, "grad_norm": 0.6297469499271634, "learning_rate": 0.00015306831834183008, "loss": 0.5508, "step": 7305 }, { "epoch": 3.4381176470588235, "grad_norm": 0.3598983455734933, "learning_rate": 0.00015302449120691765, "loss": 0.1741, "step": 7306 }, { "epoch": 3.4385882352941177, "grad_norm": 0.4026703932707832, "learning_rate": 0.00015298066381369888, "loss": 0.332, "step": 7307 }, { "epoch": 3.439058823529412, "grad_norm": 0.2565929101460593, "learning_rate": 0.0001529368361659168, "loss": 0.0966, "step": 7308 }, { "epoch": 3.4395294117647057, "grad_norm": 0.3298931204122272, "learning_rate": 0.0001528930082673145, "loss": 0.2215, "step": 7309 }, { "epoch": 3.44, "grad_norm": 0.41019575474695863, "learning_rate": 0.0001528491801216351, "loss": 0.308, "step": 7310 }, { "epoch": 3.440470588235294, "grad_norm": 0.22353374724002087, "learning_rate": 0.00015280535173262176, "loss": 0.1577, "step": 7311 }, { "epoch": 3.4409411764705884, "grad_norm": 0.33528982132420004, "learning_rate": 0.0001527615231040176, "loss": 0.256, "step": 7312 }, { "epoch": 3.441411764705882, "grad_norm": 0.5121934737883007, "learning_rate": 0.00015271769423956584, "loss": 0.3112, "step": 7313 }, { "epoch": 3.4418823529411764, "grad_norm": 0.41110598618251076, "learning_rate": 0.00015267386514300965, "loss": 0.4018, "step": 7314 }, { "epoch": 3.4423529411764706, "grad_norm": 0.6587047917603684, "learning_rate": 0.00015263003581809226, "loss": 0.4275, "step": 7315 }, { "epoch": 3.442823529411765, "grad_norm": 0.34284310093854947, "learning_rate": 0.00015258620626855691, "loss": 0.144, "step": 7316 }, { "epoch": 3.443294117647059, "grad_norm": 0.4833565164219002, "learning_rate": 0.00015254237649814687, "loss": 0.3774, "step": 7317 }, { "epoch": 3.443764705882353, "grad_norm": 0.4217075801279621, "learning_rate": 0.00015249854651060542, "loss": 0.2698, "step": 7318 }, { "epoch": 3.444235294117647, "grad_norm": 0.21534456606782187, "learning_rate": 0.00015245471630967583, "loss": 0.0679, "step": 7319 }, { "epoch": 3.4447058823529413, "grad_norm": 0.5087067847172073, "learning_rate": 0.00015241088589910138, "loss": 0.3274, "step": 7320 }, { "epoch": 3.445176470588235, "grad_norm": 0.2882020484763863, "learning_rate": 0.0001523670552826255, "loss": 0.1409, "step": 7321 }, { "epoch": 3.4456470588235293, "grad_norm": 0.03572249741199343, "learning_rate": 0.00015232322446399147, "loss": 0.0032, "step": 7322 }, { "epoch": 3.4461176470588235, "grad_norm": 0.8938693535289667, "learning_rate": 0.00015227939344694268, "loss": 0.5089, "step": 7323 }, { "epoch": 3.4465882352941177, "grad_norm": 0.746934470474596, "learning_rate": 0.0001522355622352225, "loss": 0.5627, "step": 7324 }, { "epoch": 3.447058823529412, "grad_norm": 0.23767016253532153, "learning_rate": 0.00015219173083257437, "loss": 0.0658, "step": 7325 }, { "epoch": 3.4475294117647057, "grad_norm": 0.39737319705862356, "learning_rate": 0.00015214789924274168, "loss": 0.2915, "step": 7326 }, { "epoch": 3.448, "grad_norm": 1.595928267947035, "learning_rate": 0.00015210406746946783, "loss": 1.5036, "step": 7327 }, { "epoch": 3.448470588235294, "grad_norm": 0.18979382881555965, "learning_rate": 0.0001520602355164963, "loss": 0.0906, "step": 7328 }, { "epoch": 3.4489411764705884, "grad_norm": 0.34309144363032906, "learning_rate": 0.00015201640338757054, "loss": 0.2844, "step": 7329 }, { "epoch": 3.449411764705882, "grad_norm": 0.59226719619925, "learning_rate": 0.00015197257108643406, "loss": 0.4446, "step": 7330 }, { "epoch": 3.4498823529411764, "grad_norm": 0.20642599487689306, "learning_rate": 0.00015192873861683036, "loss": 0.0776, "step": 7331 }, { "epoch": 3.4503529411764706, "grad_norm": 0.4498612994693006, "learning_rate": 0.00015188490598250288, "loss": 0.1938, "step": 7332 }, { "epoch": 3.450823529411765, "grad_norm": 0.8034196992005305, "learning_rate": 0.0001518410731871952, "loss": 0.5086, "step": 7333 }, { "epoch": 3.451294117647059, "grad_norm": 0.24495860220110177, "learning_rate": 0.00015179724023465084, "loss": 0.1145, "step": 7334 }, { "epoch": 3.451764705882353, "grad_norm": 0.4098154756881936, "learning_rate": 0.00015175340712861337, "loss": 0.1459, "step": 7335 }, { "epoch": 3.452235294117647, "grad_norm": 0.36549917007203114, "learning_rate": 0.00015170957387282627, "loss": 0.2948, "step": 7336 }, { "epoch": 3.4527058823529413, "grad_norm": 0.3715325325609653, "learning_rate": 0.00015166574047103322, "loss": 0.1592, "step": 7337 }, { "epoch": 3.453176470588235, "grad_norm": 0.3911920381121777, "learning_rate": 0.00015162190692697778, "loss": 0.297, "step": 7338 }, { "epoch": 3.4536470588235293, "grad_norm": 0.5312821290000689, "learning_rate": 0.00015157807324440347, "loss": 0.3845, "step": 7339 }, { "epoch": 3.4541176470588235, "grad_norm": 0.13969632575188365, "learning_rate": 0.00015153423942705402, "loss": 0.0702, "step": 7340 }, { "epoch": 3.4545882352941177, "grad_norm": 0.2917667098682691, "learning_rate": 0.000151490405478673, "loss": 0.1788, "step": 7341 }, { "epoch": 3.455058823529412, "grad_norm": 0.5835695196533932, "learning_rate": 0.000151446571403004, "loss": 0.5123, "step": 7342 }, { "epoch": 3.4555294117647057, "grad_norm": 0.35369815701650104, "learning_rate": 0.0001514027372037907, "loss": 0.2124, "step": 7343 }, { "epoch": 3.456, "grad_norm": 0.2751790606729813, "learning_rate": 0.0001513589028847768, "loss": 0.1578, "step": 7344 }, { "epoch": 3.456470588235294, "grad_norm": 0.538175740466606, "learning_rate": 0.00015131506844970596, "loss": 0.5064, "step": 7345 }, { "epoch": 3.4569411764705884, "grad_norm": 0.12301955682439776, "learning_rate": 0.00015127123390232177, "loss": 0.0282, "step": 7346 }, { "epoch": 3.457411764705882, "grad_norm": 0.5373569642884297, "learning_rate": 0.00015122739924636796, "loss": 0.483, "step": 7347 }, { "epoch": 3.4578823529411764, "grad_norm": 0.4029114562316123, "learning_rate": 0.00015118356448558828, "loss": 0.2697, "step": 7348 }, { "epoch": 3.4583529411764706, "grad_norm": 0.2622731457473826, "learning_rate": 0.00015113972962372635, "loss": 0.1142, "step": 7349 }, { "epoch": 3.458823529411765, "grad_norm": 0.15245280445491438, "learning_rate": 0.00015109589466452594, "loss": 0.0308, "step": 7350 }, { "epoch": 3.4592941176470586, "grad_norm": 0.22487148909309945, "learning_rate": 0.00015105205961173074, "loss": 0.1103, "step": 7351 }, { "epoch": 3.459764705882353, "grad_norm": 0.2238787498925804, "learning_rate": 0.00015100822446908452, "loss": 0.0779, "step": 7352 }, { "epoch": 3.460235294117647, "grad_norm": 0.2359307825753, "learning_rate": 0.00015096438924033095, "loss": 0.0919, "step": 7353 }, { "epoch": 3.4607058823529413, "grad_norm": 0.5083138176893889, "learning_rate": 0.00015092055392921383, "loss": 0.4218, "step": 7354 }, { "epoch": 3.461176470588235, "grad_norm": 0.279812214478575, "learning_rate": 0.00015087671853947684, "loss": 0.1522, "step": 7355 }, { "epoch": 3.4616470588235293, "grad_norm": 0.41139415842615323, "learning_rate": 0.00015083288307486388, "loss": 0.4082, "step": 7356 }, { "epoch": 3.4621176470588235, "grad_norm": 0.23245184207436198, "learning_rate": 0.00015078904753911853, "loss": 0.1608, "step": 7357 }, { "epoch": 3.4625882352941177, "grad_norm": 0.38069061989649544, "learning_rate": 0.0001507452119359847, "loss": 0.2288, "step": 7358 }, { "epoch": 3.463058823529412, "grad_norm": 0.2682051692849319, "learning_rate": 0.0001507013762692061, "loss": 0.0882, "step": 7359 }, { "epoch": 3.4635294117647057, "grad_norm": 0.32305447939407206, "learning_rate": 0.0001506575405425265, "loss": 0.2166, "step": 7360 }, { "epoch": 3.464, "grad_norm": 0.13911356821182133, "learning_rate": 0.00015061370475968975, "loss": 0.0317, "step": 7361 }, { "epoch": 3.464470588235294, "grad_norm": 0.27647420455586486, "learning_rate": 0.0001505698689244396, "loss": 0.2539, "step": 7362 }, { "epoch": 3.4649411764705884, "grad_norm": 0.30982830639333214, "learning_rate": 0.0001505260330405198, "loss": 0.2313, "step": 7363 }, { "epoch": 3.465411764705882, "grad_norm": 0.5177062391483631, "learning_rate": 0.00015048219711167422, "loss": 0.217, "step": 7364 }, { "epoch": 3.4658823529411764, "grad_norm": 0.11133947865066016, "learning_rate": 0.00015043836114164667, "loss": 0.031, "step": 7365 }, { "epoch": 3.4663529411764706, "grad_norm": 0.25407710655816973, "learning_rate": 0.00015039452513418094, "loss": 0.1108, "step": 7366 }, { "epoch": 3.466823529411765, "grad_norm": 0.44094368202777656, "learning_rate": 0.00015035068909302079, "loss": 0.2495, "step": 7367 }, { "epoch": 3.4672941176470586, "grad_norm": 0.4610597839966695, "learning_rate": 0.00015030685302191011, "loss": 0.2735, "step": 7368 }, { "epoch": 3.467764705882353, "grad_norm": 0.29470164610988814, "learning_rate": 0.00015026301692459265, "loss": 0.161, "step": 7369 }, { "epoch": 3.468235294117647, "grad_norm": 0.2713763714261472, "learning_rate": 0.00015021918080481225, "loss": 0.1538, "step": 7370 }, { "epoch": 3.4687058823529413, "grad_norm": 0.3182610267440145, "learning_rate": 0.00015017534466631272, "loss": 0.1732, "step": 7371 }, { "epoch": 3.469176470588235, "grad_norm": 0.5241001633769398, "learning_rate": 0.00015013150851283795, "loss": 0.5608, "step": 7372 }, { "epoch": 3.4696470588235293, "grad_norm": 0.3006425285860296, "learning_rate": 0.0001500876723481317, "loss": 0.1265, "step": 7373 }, { "epoch": 3.4701176470588235, "grad_norm": 0.43972036998659714, "learning_rate": 0.00015004383617593772, "loss": 0.3198, "step": 7374 }, { "epoch": 3.4705882352941178, "grad_norm": 0.3427997225623673, "learning_rate": 0.00015, "loss": 0.2456, "step": 7375 }, { "epoch": 3.471058823529412, "grad_norm": 0.34803515745324687, "learning_rate": 0.00014995616382406225, "loss": 0.2096, "step": 7376 }, { "epoch": 3.4715294117647058, "grad_norm": 0.2665340008692338, "learning_rate": 0.00014991232765186832, "loss": 0.1544, "step": 7377 }, { "epoch": 3.472, "grad_norm": 0.5593936026536781, "learning_rate": 0.00014986849148716205, "loss": 0.2907, "step": 7378 }, { "epoch": 3.472470588235294, "grad_norm": 0.26029547411563486, "learning_rate": 0.00014982465533368725, "loss": 0.2033, "step": 7379 }, { "epoch": 3.4729411764705884, "grad_norm": 0.5556535555907892, "learning_rate": 0.00014978081919518772, "loss": 0.4911, "step": 7380 }, { "epoch": 3.473411764705882, "grad_norm": 0.25076365749491075, "learning_rate": 0.00014973698307540735, "loss": 0.1766, "step": 7381 }, { "epoch": 3.4738823529411764, "grad_norm": 0.42348220309736145, "learning_rate": 0.00014969314697808991, "loss": 0.2703, "step": 7382 }, { "epoch": 3.4743529411764706, "grad_norm": 0.6078473986409013, "learning_rate": 0.00014964931090697922, "loss": 0.3406, "step": 7383 }, { "epoch": 3.474823529411765, "grad_norm": 0.29895825691281686, "learning_rate": 0.00014960547486581906, "loss": 0.1534, "step": 7384 }, { "epoch": 3.4752941176470586, "grad_norm": 0.3859699500566222, "learning_rate": 0.0001495616388583533, "loss": 0.3087, "step": 7385 }, { "epoch": 3.475764705882353, "grad_norm": 0.320984467891895, "learning_rate": 0.00014951780288832575, "loss": 0.166, "step": 7386 }, { "epoch": 3.476235294117647, "grad_norm": 0.2458053052191914, "learning_rate": 0.0001494739669594802, "loss": 0.127, "step": 7387 }, { "epoch": 3.4767058823529413, "grad_norm": 0.3498099719125904, "learning_rate": 0.00014943013107556044, "loss": 0.2538, "step": 7388 }, { "epoch": 3.477176470588235, "grad_norm": 0.27552385923600925, "learning_rate": 0.00014938629524031028, "loss": 0.1475, "step": 7389 }, { "epoch": 3.4776470588235293, "grad_norm": 0.2212899843576442, "learning_rate": 0.00014934245945747349, "loss": 0.0967, "step": 7390 }, { "epoch": 3.4781176470588235, "grad_norm": 0.5365271520581111, "learning_rate": 0.00014929862373079392, "loss": 0.2409, "step": 7391 }, { "epoch": 3.4785882352941178, "grad_norm": 0.28847587697541566, "learning_rate": 0.00014925478806401533, "loss": 0.1884, "step": 7392 }, { "epoch": 3.479058823529412, "grad_norm": 0.32062392449879784, "learning_rate": 0.0001492109524608815, "loss": 0.1105, "step": 7393 }, { "epoch": 3.4795294117647058, "grad_norm": 0.27439925475686266, "learning_rate": 0.00014916711692513618, "loss": 0.0784, "step": 7394 }, { "epoch": 3.48, "grad_norm": 0.18267898888790768, "learning_rate": 0.00014912328146052316, "loss": 0.0921, "step": 7395 }, { "epoch": 3.480470588235294, "grad_norm": 0.3522432452652955, "learning_rate": 0.00014907944607078623, "loss": 0.1661, "step": 7396 }, { "epoch": 3.4809411764705884, "grad_norm": 0.33474006515160376, "learning_rate": 0.00014903561075966908, "loss": 0.3935, "step": 7397 }, { "epoch": 3.481411764705882, "grad_norm": 0.3554899872973533, "learning_rate": 0.00014899177553091543, "loss": 0.2146, "step": 7398 }, { "epoch": 3.4818823529411764, "grad_norm": 0.3093080251082735, "learning_rate": 0.0001489479403882692, "loss": 0.2363, "step": 7399 }, { "epoch": 3.4823529411764707, "grad_norm": 0.21667534264809749, "learning_rate": 0.00014890410533547404, "loss": 0.1097, "step": 7400 }, { "epoch": 3.482823529411765, "grad_norm": 0.2568044444226004, "learning_rate": 0.00014886027037627362, "loss": 0.1219, "step": 7401 }, { "epoch": 3.4832941176470587, "grad_norm": 0.25043777077447765, "learning_rate": 0.0001488164355144117, "loss": 0.1085, "step": 7402 }, { "epoch": 3.483764705882353, "grad_norm": 0.4806485007929501, "learning_rate": 0.00014877260075363199, "loss": 0.3348, "step": 7403 }, { "epoch": 3.484235294117647, "grad_norm": 0.328797319658774, "learning_rate": 0.00014872876609767823, "loss": 0.16, "step": 7404 }, { "epoch": 3.4847058823529413, "grad_norm": 0.30252077348572504, "learning_rate": 0.00014868493155029402, "loss": 0.2144, "step": 7405 }, { "epoch": 3.485176470588235, "grad_norm": 0.41987536325013236, "learning_rate": 0.00014864109711522315, "loss": 0.3429, "step": 7406 }, { "epoch": 3.4856470588235293, "grad_norm": 0.3442048770190783, "learning_rate": 0.00014859726279620926, "loss": 0.2322, "step": 7407 }, { "epoch": 3.4861176470588235, "grad_norm": 0.5300468288818069, "learning_rate": 0.00014855342859699597, "loss": 0.4287, "step": 7408 }, { "epoch": 3.4865882352941178, "grad_norm": 0.36265245875985325, "learning_rate": 0.000148509594521327, "loss": 0.2869, "step": 7409 }, { "epoch": 3.487058823529412, "grad_norm": 0.27153437438590355, "learning_rate": 0.00014846576057294596, "loss": 0.095, "step": 7410 }, { "epoch": 3.4875294117647058, "grad_norm": 0.27233008921932417, "learning_rate": 0.0001484219267555965, "loss": 0.1532, "step": 7411 }, { "epoch": 3.488, "grad_norm": 0.17955400539924973, "learning_rate": 0.0001483780930730222, "loss": 0.0459, "step": 7412 }, { "epoch": 3.488470588235294, "grad_norm": 0.15343555062350092, "learning_rate": 0.00014833425952896675, "loss": 0.0481, "step": 7413 }, { "epoch": 3.488941176470588, "grad_norm": 0.3091029988398599, "learning_rate": 0.00014829042612717373, "loss": 0.1701, "step": 7414 }, { "epoch": 3.489411764705882, "grad_norm": 0.757605538724557, "learning_rate": 0.00014824659287138663, "loss": 0.5418, "step": 7415 }, { "epoch": 3.4898823529411764, "grad_norm": 0.2915603057733016, "learning_rate": 0.00014820275976534916, "loss": 0.1664, "step": 7416 }, { "epoch": 3.4903529411764707, "grad_norm": 0.18293755454875363, "learning_rate": 0.0001481589268128048, "loss": 0.0842, "step": 7417 }, { "epoch": 3.490823529411765, "grad_norm": 0.17978698151160158, "learning_rate": 0.0001481150940174971, "loss": 0.0904, "step": 7418 }, { "epoch": 3.4912941176470587, "grad_norm": 0.4415337072961243, "learning_rate": 0.00014807126138316964, "loss": 0.3482, "step": 7419 }, { "epoch": 3.491764705882353, "grad_norm": 0.4136060928497585, "learning_rate": 0.00014802742891356592, "loss": 0.1865, "step": 7420 }, { "epoch": 3.492235294117647, "grad_norm": 0.4573417357098273, "learning_rate": 0.00014798359661242943, "loss": 0.3093, "step": 7421 }, { "epoch": 3.4927058823529413, "grad_norm": 0.34603896995560124, "learning_rate": 0.00014793976448350366, "loss": 0.1277, "step": 7422 }, { "epoch": 3.493176470588235, "grad_norm": 0.5575634143249653, "learning_rate": 0.00014789593253053217, "loss": 0.334, "step": 7423 }, { "epoch": 3.4936470588235293, "grad_norm": 0.12070027998969948, "learning_rate": 0.00014785210075725835, "loss": 0.0488, "step": 7424 }, { "epoch": 3.4941176470588236, "grad_norm": 0.44461713282165605, "learning_rate": 0.0001478082691674256, "loss": 0.2958, "step": 7425 }, { "epoch": 3.4945882352941178, "grad_norm": 0.612154757350026, "learning_rate": 0.00014776443776477746, "loss": 0.274, "step": 7426 }, { "epoch": 3.495058823529412, "grad_norm": 0.4301093599467932, "learning_rate": 0.00014772060655305732, "loss": 0.3176, "step": 7427 }, { "epoch": 3.4955294117647058, "grad_norm": 0.31187779856258674, "learning_rate": 0.0001476767755360085, "loss": 0.2545, "step": 7428 }, { "epoch": 3.496, "grad_norm": 0.4213070462464961, "learning_rate": 0.00014763294471737448, "loss": 0.275, "step": 7429 }, { "epoch": 3.496470588235294, "grad_norm": 0.49898055420432813, "learning_rate": 0.00014758911410089862, "loss": 0.3818, "step": 7430 }, { "epoch": 3.496941176470588, "grad_norm": 0.40501808742711776, "learning_rate": 0.0001475452836903242, "loss": 0.3313, "step": 7431 }, { "epoch": 3.497411764705882, "grad_norm": 0.17222008540734393, "learning_rate": 0.0001475014534893946, "loss": 0.0611, "step": 7432 }, { "epoch": 3.4978823529411764, "grad_norm": 0.5719838604130795, "learning_rate": 0.00014745762350185316, "loss": 0.414, "step": 7433 }, { "epoch": 3.4983529411764707, "grad_norm": 0.35204558004051195, "learning_rate": 0.0001474137937314431, "loss": 0.266, "step": 7434 }, { "epoch": 3.498823529411765, "grad_norm": 0.37860969105157277, "learning_rate": 0.00014736996418190774, "loss": 0.238, "step": 7435 }, { "epoch": 3.4992941176470587, "grad_norm": 0.3788746204526922, "learning_rate": 0.00014732613485699035, "loss": 0.3655, "step": 7436 }, { "epoch": 3.499764705882353, "grad_norm": 0.4059695754095944, "learning_rate": 0.0001472823057604342, "loss": 0.1751, "step": 7437 }, { "epoch": 3.500235294117647, "grad_norm": 0.35847613580082577, "learning_rate": 0.0001472384768959824, "loss": 0.2069, "step": 7438 }, { "epoch": 3.5007058823529413, "grad_norm": 0.19801840268239884, "learning_rate": 0.00014719464826737827, "loss": 0.1248, "step": 7439 }, { "epoch": 3.501176470588235, "grad_norm": 0.21266363094027324, "learning_rate": 0.0001471508198783649, "loss": 0.1089, "step": 7440 }, { "epoch": 3.5016470588235293, "grad_norm": 0.14815609005530156, "learning_rate": 0.0001471069917326855, "loss": 0.0723, "step": 7441 }, { "epoch": 3.5021176470588236, "grad_norm": 0.4072285110845951, "learning_rate": 0.0001470631638340832, "loss": 0.3189, "step": 7442 }, { "epoch": 3.502588235294118, "grad_norm": 0.40456917134360154, "learning_rate": 0.00014701933618630107, "loss": 0.3422, "step": 7443 }, { "epoch": 3.503058823529412, "grad_norm": 0.44947963400018154, "learning_rate": 0.00014697550879308233, "loss": 0.3843, "step": 7444 }, { "epoch": 3.503529411764706, "grad_norm": 0.43244710045198476, "learning_rate": 0.00014693168165816995, "loss": 0.2881, "step": 7445 }, { "epoch": 3.504, "grad_norm": 0.14894075689498154, "learning_rate": 0.00014688785478530697, "loss": 0.0387, "step": 7446 }, { "epoch": 3.5044705882352942, "grad_norm": 0.2687304413622271, "learning_rate": 0.00014684402817823647, "loss": 0.1515, "step": 7447 }, { "epoch": 3.504941176470588, "grad_norm": 0.2259640334798594, "learning_rate": 0.00014680020184070148, "loss": 0.0645, "step": 7448 }, { "epoch": 3.5054117647058822, "grad_norm": 0.16025816329961443, "learning_rate": 0.0001467563757764449, "loss": 0.1162, "step": 7449 }, { "epoch": 3.5058823529411764, "grad_norm": 0.24637870011185187, "learning_rate": 0.00014671254998920976, "loss": 0.1295, "step": 7450 }, { "epoch": 3.5063529411764707, "grad_norm": 0.4595038339618329, "learning_rate": 0.000146668724482739, "loss": 0.2807, "step": 7451 }, { "epoch": 3.506823529411765, "grad_norm": 0.493564312435584, "learning_rate": 0.0001466248992607755, "loss": 0.3398, "step": 7452 }, { "epoch": 3.5072941176470587, "grad_norm": 0.1836762124915304, "learning_rate": 0.00014658107432706212, "loss": 0.062, "step": 7453 }, { "epoch": 3.507764705882353, "grad_norm": 0.7512733809169244, "learning_rate": 0.0001465372496853418, "loss": 0.3309, "step": 7454 }, { "epoch": 3.508235294117647, "grad_norm": 0.1225559409519088, "learning_rate": 0.00014649342533935733, "loss": 0.0331, "step": 7455 }, { "epoch": 3.5087058823529413, "grad_norm": 0.4193847692659356, "learning_rate": 0.00014644960129285152, "loss": 0.2501, "step": 7456 }, { "epoch": 3.509176470588235, "grad_norm": 0.1954251237763496, "learning_rate": 0.0001464057775495672, "loss": 0.0862, "step": 7457 }, { "epoch": 3.5096470588235293, "grad_norm": 0.43921348567045393, "learning_rate": 0.0001463619541132471, "loss": 0.4634, "step": 7458 }, { "epoch": 3.5101176470588236, "grad_norm": 0.30803183590821315, "learning_rate": 0.00014631813098763392, "loss": 0.2512, "step": 7459 }, { "epoch": 3.510588235294118, "grad_norm": 0.6797696848739065, "learning_rate": 0.00014627430817647048, "loss": 0.3866, "step": 7460 }, { "epoch": 3.511058823529412, "grad_norm": 0.18102184606687377, "learning_rate": 0.00014623048568349937, "loss": 0.053, "step": 7461 }, { "epoch": 3.511529411764706, "grad_norm": 0.09742964894038336, "learning_rate": 0.0001461866635124633, "loss": 0.0256, "step": 7462 }, { "epoch": 3.512, "grad_norm": 0.3784127426630067, "learning_rate": 0.0001461428416671048, "loss": 0.2138, "step": 7463 }, { "epoch": 3.5124705882352942, "grad_norm": 0.4470632753284192, "learning_rate": 0.0001460990201511666, "loss": 0.293, "step": 7464 }, { "epoch": 3.512941176470588, "grad_norm": 0.20520983965747633, "learning_rate": 0.0001460551989683912, "loss": 0.07, "step": 7465 }, { "epoch": 3.5134117647058822, "grad_norm": 0.2824869867847493, "learning_rate": 0.00014601137812252112, "loss": 0.1742, "step": 7466 }, { "epoch": 3.5138823529411765, "grad_norm": 0.41920694697234173, "learning_rate": 0.00014596755761729895, "loss": 0.2619, "step": 7467 }, { "epoch": 3.5143529411764707, "grad_norm": 0.6566055171688243, "learning_rate": 0.00014592373745646715, "loss": 0.5938, "step": 7468 }, { "epoch": 3.514823529411765, "grad_norm": 0.3063065050517796, "learning_rate": 0.00014587991764376812, "loss": 0.1853, "step": 7469 }, { "epoch": 3.5152941176470587, "grad_norm": 0.281389981958685, "learning_rate": 0.00014583609818294438, "loss": 0.1459, "step": 7470 }, { "epoch": 3.515764705882353, "grad_norm": 0.34150157373025786, "learning_rate": 0.00014579227907773827, "loss": 0.2032, "step": 7471 }, { "epoch": 3.516235294117647, "grad_norm": 0.2938278263962255, "learning_rate": 0.00014574846033189217, "loss": 0.2332, "step": 7472 }, { "epoch": 3.516705882352941, "grad_norm": 0.39712491780421777, "learning_rate": 0.00014570464194914835, "loss": 0.351, "step": 7473 }, { "epoch": 3.517176470588235, "grad_norm": 0.24420264520852297, "learning_rate": 0.00014566082393324925, "loss": 0.1539, "step": 7474 }, { "epoch": 3.5176470588235293, "grad_norm": 0.2828651668148354, "learning_rate": 0.00014561700628793706, "loss": 0.2434, "step": 7475 }, { "epoch": 3.5181176470588236, "grad_norm": 0.18233467246186175, "learning_rate": 0.000145573189016954, "loss": 0.0739, "step": 7476 }, { "epoch": 3.518588235294118, "grad_norm": 0.4811966237587478, "learning_rate": 0.0001455293721240423, "loss": 0.4583, "step": 7477 }, { "epoch": 3.519058823529412, "grad_norm": 0.39665778233354365, "learning_rate": 0.0001454855556129442, "loss": 0.2524, "step": 7478 }, { "epoch": 3.519529411764706, "grad_norm": 0.322992707797697, "learning_rate": 0.00014544173948740174, "loss": 0.2885, "step": 7479 }, { "epoch": 3.52, "grad_norm": 0.2611402453085833, "learning_rate": 0.00014539792375115713, "loss": 0.1358, "step": 7480 }, { "epoch": 3.5204705882352942, "grad_norm": 0.32071104509365117, "learning_rate": 0.0001453541084079524, "loss": 0.1902, "step": 7481 }, { "epoch": 3.520941176470588, "grad_norm": 0.3778706785186694, "learning_rate": 0.0001453102934615296, "loss": 0.2439, "step": 7482 }, { "epoch": 3.5214117647058822, "grad_norm": 0.2854813401122938, "learning_rate": 0.0001452664789156307, "loss": 0.1125, "step": 7483 }, { "epoch": 3.5218823529411765, "grad_norm": 0.3422226250488346, "learning_rate": 0.00014522266477399776, "loss": 0.3243, "step": 7484 }, { "epoch": 3.5223529411764707, "grad_norm": 0.1781939931811693, "learning_rate": 0.00014517885104037268, "loss": 0.0794, "step": 7485 }, { "epoch": 3.522823529411765, "grad_norm": 0.4415981141570924, "learning_rate": 0.00014513503771849734, "loss": 0.1981, "step": 7486 }, { "epoch": 3.5232941176470587, "grad_norm": 0.34733663694227107, "learning_rate": 0.00014509122481211358, "loss": 0.1177, "step": 7487 }, { "epoch": 3.523764705882353, "grad_norm": 0.22025256874513335, "learning_rate": 0.0001450474123249634, "loss": 0.1157, "step": 7488 }, { "epoch": 3.524235294117647, "grad_norm": 0.230022522129841, "learning_rate": 0.00014500360026078847, "loss": 0.1234, "step": 7489 }, { "epoch": 3.524705882352941, "grad_norm": 0.7870339164408043, "learning_rate": 0.00014495978862333063, "loss": 0.5387, "step": 7490 }, { "epoch": 3.525176470588235, "grad_norm": 0.29178319536086317, "learning_rate": 0.00014491597741633148, "loss": 0.1572, "step": 7491 }, { "epoch": 3.5256470588235294, "grad_norm": 0.37916898896146994, "learning_rate": 0.00014487216664353287, "loss": 0.2463, "step": 7492 }, { "epoch": 3.5261176470588236, "grad_norm": 0.4758359661486548, "learning_rate": 0.00014482835630867636, "loss": 0.1813, "step": 7493 }, { "epoch": 3.526588235294118, "grad_norm": 0.3543503394307356, "learning_rate": 0.00014478454641550354, "loss": 0.1691, "step": 7494 }, { "epoch": 3.527058823529412, "grad_norm": 0.2092834826923528, "learning_rate": 0.0001447407369677561, "loss": 0.1041, "step": 7495 }, { "epoch": 3.527529411764706, "grad_norm": 0.523511450222369, "learning_rate": 0.0001446969279691755, "loss": 0.1673, "step": 7496 }, { "epoch": 3.528, "grad_norm": 0.44250087758747797, "learning_rate": 0.00014465311942350324, "loss": 0.3464, "step": 7497 }, { "epoch": 3.5284705882352942, "grad_norm": 0.4296924899800624, "learning_rate": 0.00014460931133448084, "loss": 0.2351, "step": 7498 }, { "epoch": 3.528941176470588, "grad_norm": 0.49694093695646474, "learning_rate": 0.0001445655037058497, "loss": 0.4434, "step": 7499 }, { "epoch": 3.5294117647058822, "grad_norm": 0.43834287553920287, "learning_rate": 0.00014452169654135115, "loss": 0.21, "step": 7500 }, { "epoch": 3.5298823529411765, "grad_norm": 0.40503913940465147, "learning_rate": 0.0001444778898447266, "loss": 0.3674, "step": 7501 }, { "epoch": 3.5303529411764707, "grad_norm": 0.3117056820082854, "learning_rate": 0.00014443408361971739, "loss": 0.2773, "step": 7502 }, { "epoch": 3.530823529411765, "grad_norm": 0.3363477203223054, "learning_rate": 0.0001443902778700647, "loss": 0.2113, "step": 7503 }, { "epoch": 3.5312941176470587, "grad_norm": 0.47670565812965876, "learning_rate": 0.0001443464725995098, "loss": 0.359, "step": 7504 }, { "epoch": 3.531764705882353, "grad_norm": 0.4654213266840677, "learning_rate": 0.0001443026678117939, "loss": 0.2553, "step": 7505 }, { "epoch": 3.532235294117647, "grad_norm": 0.3483955354108432, "learning_rate": 0.00014425886351065808, "loss": 0.2642, "step": 7506 }, { "epoch": 3.532705882352941, "grad_norm": 0.2340290975185438, "learning_rate": 0.00014421505969984346, "loss": 0.1262, "step": 7507 }, { "epoch": 3.533176470588235, "grad_norm": 0.11101411214064666, "learning_rate": 0.00014417125638309115, "loss": 0.0301, "step": 7508 }, { "epoch": 3.5336470588235294, "grad_norm": 0.2983840148671788, "learning_rate": 0.00014412745356414213, "loss": 0.1209, "step": 7509 }, { "epoch": 3.5341176470588236, "grad_norm": 0.42423716070564277, "learning_rate": 0.00014408365124673737, "loss": 0.2857, "step": 7510 }, { "epoch": 3.534588235294118, "grad_norm": 0.3297091655555434, "learning_rate": 0.00014403984943461783, "loss": 0.2019, "step": 7511 }, { "epoch": 3.535058823529412, "grad_norm": 0.08317152182770055, "learning_rate": 0.0001439960481315244, "loss": 0.022, "step": 7512 }, { "epoch": 3.535529411764706, "grad_norm": 0.40231234716451486, "learning_rate": 0.00014395224734119788, "loss": 0.2715, "step": 7513 }, { "epoch": 3.536, "grad_norm": 0.2163531782672119, "learning_rate": 0.00014390844706737907, "loss": 0.0567, "step": 7514 }, { "epoch": 3.5364705882352943, "grad_norm": 0.4753045835843946, "learning_rate": 0.00014386464731380881, "loss": 0.3905, "step": 7515 }, { "epoch": 3.536941176470588, "grad_norm": 0.2532059597592927, "learning_rate": 0.00014382084808422775, "loss": 0.1864, "step": 7516 }, { "epoch": 3.5374117647058823, "grad_norm": 0.30571810087026674, "learning_rate": 0.00014377704938237654, "loss": 0.1208, "step": 7517 }, { "epoch": 3.5378823529411765, "grad_norm": 0.27438235601511035, "learning_rate": 0.00014373325121199587, "loss": 0.1816, "step": 7518 }, { "epoch": 3.5383529411764707, "grad_norm": 0.39705595024123264, "learning_rate": 0.00014368945357682628, "loss": 0.4099, "step": 7519 }, { "epoch": 3.538823529411765, "grad_norm": 0.36585274120191996, "learning_rate": 0.0001436456564806083, "loss": 0.3044, "step": 7520 }, { "epoch": 3.5392941176470587, "grad_norm": 0.43692738823228283, "learning_rate": 0.00014360185992708243, "loss": 0.4154, "step": 7521 }, { "epoch": 3.539764705882353, "grad_norm": 0.32375784801537594, "learning_rate": 0.00014355806391998913, "loss": 0.2071, "step": 7522 }, { "epoch": 3.540235294117647, "grad_norm": 0.10742566244570162, "learning_rate": 0.00014351426846306875, "loss": 0.0383, "step": 7523 }, { "epoch": 3.540705882352941, "grad_norm": 0.3812202701072833, "learning_rate": 0.00014347047356006161, "loss": 0.287, "step": 7524 }, { "epoch": 3.541176470588235, "grad_norm": 1.1768674420469492, "learning_rate": 0.00014342667921470812, "loss": 0.7769, "step": 7525 }, { "epoch": 3.5416470588235294, "grad_norm": 0.5231845110556409, "learning_rate": 0.00014338288543074845, "loss": 0.3675, "step": 7526 }, { "epoch": 3.5421176470588236, "grad_norm": 0.32868705582305435, "learning_rate": 0.00014333909221192278, "loss": 0.1483, "step": 7527 }, { "epoch": 3.542588235294118, "grad_norm": 0.39189971920439365, "learning_rate": 0.00014329529956197135, "loss": 0.1488, "step": 7528 }, { "epoch": 3.543058823529412, "grad_norm": 0.7871146757339127, "learning_rate": 0.00014325150748463423, "loss": 0.2969, "step": 7529 }, { "epoch": 3.543529411764706, "grad_norm": 0.3390417923028141, "learning_rate": 0.00014320771598365142, "loss": 0.1931, "step": 7530 }, { "epoch": 3.544, "grad_norm": 0.29668039754780806, "learning_rate": 0.00014316392506276302, "loss": 0.1699, "step": 7531 }, { "epoch": 3.5444705882352943, "grad_norm": 0.26267593743362766, "learning_rate": 0.0001431201347257089, "loss": 0.1586, "step": 7532 }, { "epoch": 3.544941176470588, "grad_norm": 0.2597841571150847, "learning_rate": 0.00014307634497622907, "loss": 0.1187, "step": 7533 }, { "epoch": 3.5454117647058823, "grad_norm": 0.36587169833628025, "learning_rate": 0.00014303255581806336, "loss": 0.2655, "step": 7534 }, { "epoch": 3.5458823529411765, "grad_norm": 0.2982524833310802, "learning_rate": 0.0001429887672549515, "loss": 0.2057, "step": 7535 }, { "epoch": 3.5463529411764707, "grad_norm": 0.35027049930850607, "learning_rate": 0.00014294497929063336, "loss": 0.2169, "step": 7536 }, { "epoch": 3.546823529411765, "grad_norm": 0.3890736129723009, "learning_rate": 0.0001429011919288486, "loss": 0.2501, "step": 7537 }, { "epoch": 3.5472941176470587, "grad_norm": 0.2834536256985031, "learning_rate": 0.0001428574051733368, "loss": 0.1084, "step": 7538 }, { "epoch": 3.547764705882353, "grad_norm": 0.23158669024655396, "learning_rate": 0.0001428136190278377, "loss": 0.1174, "step": 7539 }, { "epoch": 3.548235294117647, "grad_norm": 0.39963265935992814, "learning_rate": 0.00014276983349609075, "loss": 0.2641, "step": 7540 }, { "epoch": 3.548705882352941, "grad_norm": 0.46567434182198625, "learning_rate": 0.0001427260485818355, "loss": 0.3066, "step": 7541 }, { "epoch": 3.549176470588235, "grad_norm": 0.750744766484674, "learning_rate": 0.00014268226428881136, "loss": 0.562, "step": 7542 }, { "epoch": 3.5496470588235294, "grad_norm": 0.6931322738848266, "learning_rate": 0.00014263848062075776, "loss": 0.5185, "step": 7543 }, { "epoch": 3.5501176470588236, "grad_norm": 0.24371539408836843, "learning_rate": 0.00014259469758141403, "loss": 0.0787, "step": 7544 }, { "epoch": 3.550588235294118, "grad_norm": 0.3370701456288076, "learning_rate": 0.00014255091517451941, "loss": 0.3043, "step": 7545 }, { "epoch": 3.5510588235294116, "grad_norm": 0.33069905983583026, "learning_rate": 0.0001425071334038132, "loss": 0.1791, "step": 7546 }, { "epoch": 3.551529411764706, "grad_norm": 0.24732529626146849, "learning_rate": 0.00014246335227303456, "loss": 0.1641, "step": 7547 }, { "epoch": 3.552, "grad_norm": 0.471171485877146, "learning_rate": 0.00014241957178592258, "loss": 0.2736, "step": 7548 }, { "epoch": 3.5524705882352943, "grad_norm": 0.5164398328836878, "learning_rate": 0.00014237579194621637, "loss": 0.3096, "step": 7549 }, { "epoch": 3.552941176470588, "grad_norm": 0.23076140177505325, "learning_rate": 0.00014233201275765494, "loss": 0.1051, "step": 7550 }, { "epoch": 3.5534117647058823, "grad_norm": 0.14104077368147622, "learning_rate": 0.0001422882342239772, "loss": 0.0602, "step": 7551 }, { "epoch": 3.5538823529411765, "grad_norm": 0.37969045556980463, "learning_rate": 0.00014224445634892206, "loss": 0.2358, "step": 7552 }, { "epoch": 3.5543529411764707, "grad_norm": 0.24795726585334757, "learning_rate": 0.00014220067913622844, "loss": 0.1693, "step": 7553 }, { "epoch": 3.554823529411765, "grad_norm": 0.33698378036237686, "learning_rate": 0.00014215690258963508, "loss": 0.2621, "step": 7554 }, { "epoch": 3.5552941176470587, "grad_norm": 0.24051938697627054, "learning_rate": 0.00014211312671288064, "loss": 0.0528, "step": 7555 }, { "epoch": 3.555764705882353, "grad_norm": 0.27897530014088323, "learning_rate": 0.00014206935150970392, "loss": 0.1106, "step": 7556 }, { "epoch": 3.556235294117647, "grad_norm": 0.5484115307228262, "learning_rate": 0.00014202557698384348, "loss": 0.4333, "step": 7557 }, { "epoch": 3.556705882352941, "grad_norm": 0.48927130621512843, "learning_rate": 0.00014198180313903784, "loss": 0.3759, "step": 7558 }, { "epoch": 3.557176470588235, "grad_norm": 0.18726959514465408, "learning_rate": 0.0001419380299790256, "loss": 0.075, "step": 7559 }, { "epoch": 3.5576470588235294, "grad_norm": 0.3482992723446622, "learning_rate": 0.00014189425750754512, "loss": 0.2383, "step": 7560 }, { "epoch": 3.5581176470588236, "grad_norm": 0.4532675449392613, "learning_rate": 0.00014185048572833478, "loss": 0.4335, "step": 7561 }, { "epoch": 3.558588235294118, "grad_norm": 0.25315095616500044, "learning_rate": 0.00014180671464513296, "loss": 0.1198, "step": 7562 }, { "epoch": 3.5590588235294116, "grad_norm": 0.33272286023408043, "learning_rate": 0.0001417629442616779, "loss": 0.2128, "step": 7563 }, { "epoch": 3.559529411764706, "grad_norm": 0.43586873283304706, "learning_rate": 0.00014171917458170782, "loss": 0.3305, "step": 7564 }, { "epoch": 3.56, "grad_norm": 0.29758182879348233, "learning_rate": 0.0001416754056089608, "loss": 0.2191, "step": 7565 }, { "epoch": 3.5604705882352943, "grad_norm": 0.17017850918957406, "learning_rate": 0.000141631637347175, "loss": 0.0377, "step": 7566 }, { "epoch": 3.560941176470588, "grad_norm": 0.3435402100374185, "learning_rate": 0.0001415878698000884, "loss": 0.232, "step": 7567 }, { "epoch": 3.5614117647058823, "grad_norm": 0.5429180717959579, "learning_rate": 0.00014154410297143898, "loss": 0.2225, "step": 7568 }, { "epoch": 3.5618823529411765, "grad_norm": 0.22502248837530353, "learning_rate": 0.00014150033686496469, "loss": 0.0996, "step": 7569 }, { "epoch": 3.5623529411764707, "grad_norm": 0.17333634275725837, "learning_rate": 0.0001414565714844033, "loss": 0.0554, "step": 7570 }, { "epoch": 3.562823529411765, "grad_norm": 0.28533468880857227, "learning_rate": 0.00014141280683349253, "loss": 0.1519, "step": 7571 }, { "epoch": 3.5632941176470587, "grad_norm": 0.362474508674598, "learning_rate": 0.00014136904291597028, "loss": 0.2844, "step": 7572 }, { "epoch": 3.563764705882353, "grad_norm": 0.3894071239158175, "learning_rate": 0.00014132527973557407, "loss": 0.2099, "step": 7573 }, { "epoch": 3.564235294117647, "grad_norm": 0.35424455402858024, "learning_rate": 0.00014128151729604152, "loss": 0.181, "step": 7574 }, { "epoch": 3.564705882352941, "grad_norm": 0.1648350610721086, "learning_rate": 0.00014123775560111009, "loss": 0.1005, "step": 7575 }, { "epoch": 3.565176470588235, "grad_norm": 0.25297616278595253, "learning_rate": 0.00014119399465451738, "loss": 0.103, "step": 7576 }, { "epoch": 3.5656470588235294, "grad_norm": 0.4065001813405191, "learning_rate": 0.0001411502344600007, "loss": 0.3067, "step": 7577 }, { "epoch": 3.5661176470588236, "grad_norm": 0.30753979800180553, "learning_rate": 0.00014110647502129742, "loss": 0.1385, "step": 7578 }, { "epoch": 3.566588235294118, "grad_norm": 0.4088745376156818, "learning_rate": 0.00014106271634214477, "loss": 0.3477, "step": 7579 }, { "epoch": 3.5670588235294116, "grad_norm": 0.4027029088105102, "learning_rate": 0.00014101895842628001, "loss": 0.3093, "step": 7580 }, { "epoch": 3.567529411764706, "grad_norm": 0.46859140388280396, "learning_rate": 0.00014097520127744024, "loss": 0.3602, "step": 7581 }, { "epoch": 3.568, "grad_norm": 0.31494851249858596, "learning_rate": 0.00014093144489936259, "loss": 0.1009, "step": 7582 }, { "epoch": 3.5684705882352943, "grad_norm": 0.3322204567898513, "learning_rate": 0.00014088768929578397, "loss": 0.2292, "step": 7583 }, { "epoch": 3.568941176470588, "grad_norm": 0.21200621579206858, "learning_rate": 0.00014084393447044138, "loss": 0.0911, "step": 7584 }, { "epoch": 3.5694117647058823, "grad_norm": 0.20693102514829723, "learning_rate": 0.00014080018042707173, "loss": 0.1366, "step": 7585 }, { "epoch": 3.5698823529411765, "grad_norm": 0.5634578885391893, "learning_rate": 0.00014075642716941176, "loss": 0.4474, "step": 7586 }, { "epoch": 3.5703529411764707, "grad_norm": 0.5163913910432905, "learning_rate": 0.0001407126747011983, "loss": 0.4631, "step": 7587 }, { "epoch": 3.570823529411765, "grad_norm": 0.6296560788512885, "learning_rate": 0.00014066892302616794, "loss": 0.5479, "step": 7588 }, { "epoch": 3.5712941176470587, "grad_norm": 0.232991958556948, "learning_rate": 0.0001406251721480573, "loss": 0.1658, "step": 7589 }, { "epoch": 3.571764705882353, "grad_norm": 0.39561482080779226, "learning_rate": 0.000140581422070603, "loss": 0.3021, "step": 7590 }, { "epoch": 3.572235294117647, "grad_norm": 0.2154778689231346, "learning_rate": 0.0001405376727975414, "loss": 0.0705, "step": 7591 }, { "epoch": 3.572705882352941, "grad_norm": 0.35136358311350263, "learning_rate": 0.000140493924332609, "loss": 0.1301, "step": 7592 }, { "epoch": 3.573176470588235, "grad_norm": 0.48811059259735845, "learning_rate": 0.00014045017667954203, "loss": 0.3254, "step": 7593 }, { "epoch": 3.5736470588235294, "grad_norm": 0.41436511798200604, "learning_rate": 0.00014040642984207687, "loss": 0.182, "step": 7594 }, { "epoch": 3.5741176470588236, "grad_norm": 0.28921754110784925, "learning_rate": 0.00014036268382394964, "loss": 0.1957, "step": 7595 }, { "epoch": 3.574588235294118, "grad_norm": 0.33973674723959085, "learning_rate": 0.00014031893862889646, "loss": 0.1407, "step": 7596 }, { "epoch": 3.5750588235294116, "grad_norm": 0.40896937571697417, "learning_rate": 0.00014027519426065344, "loss": 0.2275, "step": 7597 }, { "epoch": 3.575529411764706, "grad_norm": 0.20783592476149693, "learning_rate": 0.00014023145072295654, "loss": 0.0773, "step": 7598 }, { "epoch": 3.576, "grad_norm": 0.48125226441555363, "learning_rate": 0.0001401877080195416, "loss": 0.2445, "step": 7599 }, { "epoch": 3.576470588235294, "grad_norm": 0.4799970170685716, "learning_rate": 0.00014014396615414458, "loss": 0.3847, "step": 7600 }, { "epoch": 3.576941176470588, "grad_norm": 0.2629579589804913, "learning_rate": 0.0001401002251305012, "loss": 0.1125, "step": 7601 }, { "epoch": 3.5774117647058823, "grad_norm": 0.031063665070588074, "learning_rate": 0.0001400564849523471, "loss": 0.0025, "step": 7602 }, { "epoch": 3.5778823529411765, "grad_norm": 0.062403305221481034, "learning_rate": 0.00014001274562341798, "loss": 0.0091, "step": 7603 }, { "epoch": 3.5783529411764707, "grad_norm": 0.3547070601763109, "learning_rate": 0.0001399690071474494, "loss": 0.1903, "step": 7604 }, { "epoch": 3.578823529411765, "grad_norm": 0.49299961568655076, "learning_rate": 0.0001399252695281768, "loss": 0.357, "step": 7605 }, { "epoch": 3.5792941176470587, "grad_norm": 0.45192207020681685, "learning_rate": 0.00013988153276933555, "loss": 0.2879, "step": 7606 }, { "epoch": 3.579764705882353, "grad_norm": 0.5658818179462112, "learning_rate": 0.0001398377968746611, "loss": 0.2202, "step": 7607 }, { "epoch": 3.580235294117647, "grad_norm": 0.3301618776951094, "learning_rate": 0.0001397940618478886, "loss": 0.1591, "step": 7608 }, { "epoch": 3.580705882352941, "grad_norm": 0.2336910333743382, "learning_rate": 0.00013975032769275326, "loss": 0.1208, "step": 7609 }, { "epoch": 3.581176470588235, "grad_norm": 0.33342545597586304, "learning_rate": 0.00013970659441299026, "loss": 0.192, "step": 7610 }, { "epoch": 3.5816470588235294, "grad_norm": 0.3195318990377626, "learning_rate": 0.00013966286201233457, "loss": 0.2198, "step": 7611 }, { "epoch": 3.5821176470588236, "grad_norm": 0.26942101374840793, "learning_rate": 0.00013961913049452115, "loss": 0.1328, "step": 7612 }, { "epoch": 3.582588235294118, "grad_norm": 0.465129140694307, "learning_rate": 0.00013957539986328492, "loss": 0.3078, "step": 7613 }, { "epoch": 3.5830588235294116, "grad_norm": 0.1926424075901963, "learning_rate": 0.00013953167012236068, "loss": 0.1013, "step": 7614 }, { "epoch": 3.583529411764706, "grad_norm": 0.4708062097183356, "learning_rate": 0.00013948794127548316, "loss": 0.3456, "step": 7615 }, { "epoch": 3.584, "grad_norm": 0.4958652538711665, "learning_rate": 0.00013944421332638698, "loss": 0.278, "step": 7616 }, { "epoch": 3.584470588235294, "grad_norm": 0.3453679855891507, "learning_rate": 0.0001394004862788068, "loss": 0.2584, "step": 7617 }, { "epoch": 3.584941176470588, "grad_norm": 0.6045773409190016, "learning_rate": 0.00013935676013647708, "loss": 0.3549, "step": 7618 }, { "epoch": 3.5854117647058823, "grad_norm": 0.23155397286051038, "learning_rate": 0.0001393130349031322, "loss": 0.1457, "step": 7619 }, { "epoch": 3.5858823529411765, "grad_norm": 0.3484486283005314, "learning_rate": 0.00013926931058250657, "loss": 0.224, "step": 7620 }, { "epoch": 3.5863529411764707, "grad_norm": 0.40620136658570327, "learning_rate": 0.00013922558717833445, "loss": 0.2654, "step": 7621 }, { "epoch": 3.586823529411765, "grad_norm": 0.42447766256166697, "learning_rate": 0.00013918186469435005, "loss": 0.2965, "step": 7622 }, { "epoch": 3.5872941176470587, "grad_norm": 0.470574814381085, "learning_rate": 0.00013913814313428748, "loss": 0.3368, "step": 7623 }, { "epoch": 3.587764705882353, "grad_norm": 0.4391099515455031, "learning_rate": 0.00013909442250188074, "loss": 0.4014, "step": 7624 }, { "epoch": 3.588235294117647, "grad_norm": 0.2753095692275904, "learning_rate": 0.00013905070280086386, "loss": 0.1206, "step": 7625 }, { "epoch": 3.588705882352941, "grad_norm": 0.3810741182981147, "learning_rate": 0.00013900698403497062, "loss": 0.2758, "step": 7626 }, { "epoch": 3.589176470588235, "grad_norm": 0.21968628524578765, "learning_rate": 0.00013896326620793488, "loss": 0.0926, "step": 7627 }, { "epoch": 3.5896470588235294, "grad_norm": 0.3675486920571789, "learning_rate": 0.00013891954932349033, "loss": 0.2893, "step": 7628 }, { "epoch": 3.5901176470588236, "grad_norm": 0.7315047338827563, "learning_rate": 0.00013887583338537064, "loss": 0.636, "step": 7629 }, { "epoch": 3.590588235294118, "grad_norm": 0.2899482593404231, "learning_rate": 0.0001388321183973093, "loss": 0.218, "step": 7630 }, { "epoch": 3.5910588235294116, "grad_norm": 0.2552104145803105, "learning_rate": 0.0001387884043630399, "loss": 0.1165, "step": 7631 }, { "epoch": 3.591529411764706, "grad_norm": 0.37333858201737014, "learning_rate": 0.0001387446912862957, "loss": 0.2408, "step": 7632 }, { "epoch": 3.592, "grad_norm": 0.3549349164696366, "learning_rate": 0.0001387009791708101, "loss": 0.3175, "step": 7633 }, { "epoch": 3.592470588235294, "grad_norm": 0.3699857905143578, "learning_rate": 0.00013865726802031627, "loss": 0.2886, "step": 7634 }, { "epoch": 3.592941176470588, "grad_norm": 0.4773946696605903, "learning_rate": 0.0001386135578385474, "loss": 0.5033, "step": 7635 }, { "epoch": 3.5934117647058823, "grad_norm": 0.4432993992884707, "learning_rate": 0.00013856984862923655, "loss": 0.1415, "step": 7636 }, { "epoch": 3.5938823529411765, "grad_norm": 0.2328794049481987, "learning_rate": 0.00013852614039611667, "loss": 0.0855, "step": 7637 }, { "epoch": 3.5943529411764708, "grad_norm": 0.3506093947110573, "learning_rate": 0.0001384824331429207, "loss": 0.258, "step": 7638 }, { "epoch": 3.594823529411765, "grad_norm": 0.3284113399400117, "learning_rate": 0.00013843872687338142, "loss": 0.2107, "step": 7639 }, { "epoch": 3.5952941176470588, "grad_norm": 0.20540164101705677, "learning_rate": 0.00013839502159123158, "loss": 0.081, "step": 7640 }, { "epoch": 3.595764705882353, "grad_norm": 0.28715523841718343, "learning_rate": 0.00013835131730020382, "loss": 0.1168, "step": 7641 }, { "epoch": 3.596235294117647, "grad_norm": 0.19499303932182346, "learning_rate": 0.0001383076140040307, "loss": 0.0959, "step": 7642 }, { "epoch": 3.596705882352941, "grad_norm": 0.5575883170042375, "learning_rate": 0.0001382639117064447, "loss": 0.3322, "step": 7643 }, { "epoch": 3.597176470588235, "grad_norm": 0.2819014094456665, "learning_rate": 0.0001382202104111782, "loss": 0.2155, "step": 7644 }, { "epoch": 3.5976470588235294, "grad_norm": 0.6840001983400705, "learning_rate": 0.0001381765101219635, "loss": 0.5422, "step": 7645 }, { "epoch": 3.5981176470588236, "grad_norm": 0.2715497662508158, "learning_rate": 0.00013813281084253286, "loss": 0.1347, "step": 7646 }, { "epoch": 3.598588235294118, "grad_norm": 0.3381623308053299, "learning_rate": 0.00013808911257661834, "loss": 0.1856, "step": 7647 }, { "epoch": 3.5990588235294116, "grad_norm": 0.29265466417579455, "learning_rate": 0.00013804541532795204, "loss": 0.1544, "step": 7648 }, { "epoch": 3.599529411764706, "grad_norm": 0.33837925690627185, "learning_rate": 0.00013800171910026594, "loss": 0.2323, "step": 7649 }, { "epoch": 3.6, "grad_norm": 0.36810289636181615, "learning_rate": 0.00013795802389729184, "loss": 0.1935, "step": 7650 }, { "epoch": 3.600470588235294, "grad_norm": 0.2246006071676623, "learning_rate": 0.0001379143297227616, "loss": 0.1022, "step": 7651 }, { "epoch": 3.600941176470588, "grad_norm": 0.20233412311972954, "learning_rate": 0.00013787063658040686, "loss": 0.1465, "step": 7652 }, { "epoch": 3.6014117647058823, "grad_norm": 0.2741189991412712, "learning_rate": 0.00013782694447395926, "loss": 0.1246, "step": 7653 }, { "epoch": 3.6018823529411765, "grad_norm": 0.4138110333370557, "learning_rate": 0.0001377832534071503, "loss": 0.2824, "step": 7654 }, { "epoch": 3.6023529411764708, "grad_norm": 0.2178815500900353, "learning_rate": 0.0001377395633837114, "loss": 0.0946, "step": 7655 }, { "epoch": 3.602823529411765, "grad_norm": 0.42058423800551803, "learning_rate": 0.000137695874407374, "loss": 0.4398, "step": 7656 }, { "epoch": 3.6032941176470588, "grad_norm": 0.8387639194359064, "learning_rate": 0.0001376521864818692, "loss": 0.5507, "step": 7657 }, { "epoch": 3.603764705882353, "grad_norm": 0.2735939363341517, "learning_rate": 0.00013760849961092826, "loss": 0.2619, "step": 7658 }, { "epoch": 3.604235294117647, "grad_norm": 0.5844291558946771, "learning_rate": 0.00013756481379828225, "loss": 0.47, "step": 7659 }, { "epoch": 3.604705882352941, "grad_norm": 0.5130656520533352, "learning_rate": 0.00013752112904766208, "loss": 0.4639, "step": 7660 }, { "epoch": 3.605176470588235, "grad_norm": 0.484005729219481, "learning_rate": 0.00013747744536279878, "loss": 0.2823, "step": 7661 }, { "epoch": 3.6056470588235294, "grad_norm": 0.42064661300252426, "learning_rate": 0.00013743376274742304, "loss": 0.3699, "step": 7662 }, { "epoch": 3.6061176470588236, "grad_norm": 0.27872549606059877, "learning_rate": 0.00013739008120526559, "loss": 0.1385, "step": 7663 }, { "epoch": 3.606588235294118, "grad_norm": 0.5409341619143158, "learning_rate": 0.00013734640074005702, "loss": 0.3267, "step": 7664 }, { "epoch": 3.6070588235294117, "grad_norm": 0.3056891147747053, "learning_rate": 0.0001373027213555279, "loss": 0.2153, "step": 7665 }, { "epoch": 3.607529411764706, "grad_norm": 0.2836462905662945, "learning_rate": 0.0001372590430554087, "loss": 0.1029, "step": 7666 }, { "epoch": 3.608, "grad_norm": 0.4148544612426444, "learning_rate": 0.0001372153658434297, "loss": 0.3428, "step": 7667 }, { "epoch": 3.608470588235294, "grad_norm": 0.31493929709757496, "learning_rate": 0.00013717168972332116, "loss": 0.2333, "step": 7668 }, { "epoch": 3.608941176470588, "grad_norm": 0.49576990255457154, "learning_rate": 0.00013712801469881324, "loss": 0.4619, "step": 7669 }, { "epoch": 3.6094117647058823, "grad_norm": 0.3216654616190871, "learning_rate": 0.00013708434077363603, "loss": 0.2113, "step": 7670 }, { "epoch": 3.6098823529411765, "grad_norm": 0.3047653966396119, "learning_rate": 0.00013704066795151946, "loss": 0.1973, "step": 7671 }, { "epoch": 3.6103529411764708, "grad_norm": 0.3627094655584777, "learning_rate": 0.00013699699623619338, "loss": 0.2359, "step": 7672 }, { "epoch": 3.6108235294117645, "grad_norm": 0.5005763975016353, "learning_rate": 0.00013695332563138762, "loss": 0.2167, "step": 7673 }, { "epoch": 3.6112941176470588, "grad_norm": 0.7308455943887177, "learning_rate": 0.00013690965614083185, "loss": 0.4522, "step": 7674 }, { "epoch": 3.611764705882353, "grad_norm": 0.4161203129776847, "learning_rate": 0.00013686598776825563, "loss": 0.2723, "step": 7675 }, { "epoch": 3.612235294117647, "grad_norm": 0.2996458784926046, "learning_rate": 0.00013682232051738852, "loss": 0.1609, "step": 7676 }, { "epoch": 3.612705882352941, "grad_norm": 0.31079707867240586, "learning_rate": 0.00013677865439195986, "loss": 0.1185, "step": 7677 }, { "epoch": 3.613176470588235, "grad_norm": 0.3555083323827483, "learning_rate": 0.00013673498939569892, "loss": 0.2661, "step": 7678 }, { "epoch": 3.6136470588235294, "grad_norm": 0.43083938429605034, "learning_rate": 0.000136691325532335, "loss": 0.3302, "step": 7679 }, { "epoch": 3.6141176470588237, "grad_norm": 0.19891702475214532, "learning_rate": 0.00013664766280559716, "loss": 0.0656, "step": 7680 }, { "epoch": 3.614588235294118, "grad_norm": 0.3172297822273401, "learning_rate": 0.00013660400121921434, "loss": 0.1831, "step": 7681 }, { "epoch": 3.6150588235294117, "grad_norm": 0.21054861599847888, "learning_rate": 0.00013656034077691562, "loss": 0.0757, "step": 7682 }, { "epoch": 3.615529411764706, "grad_norm": 0.5442673368857353, "learning_rate": 0.00013651668148242967, "loss": 0.3834, "step": 7683 }, { "epoch": 3.616, "grad_norm": 0.41344685944160126, "learning_rate": 0.00013647302333948528, "loss": 0.2466, "step": 7684 }, { "epoch": 3.616470588235294, "grad_norm": 0.393513903847359, "learning_rate": 0.000136429366351811, "loss": 0.3253, "step": 7685 }, { "epoch": 3.616941176470588, "grad_norm": 0.7930036101922506, "learning_rate": 0.00013638571052313543, "loss": 0.416, "step": 7686 }, { "epoch": 3.6174117647058823, "grad_norm": 0.43620248141141005, "learning_rate": 0.00013634205585718698, "loss": 0.2812, "step": 7687 }, { "epoch": 3.6178823529411765, "grad_norm": 0.3438886328197425, "learning_rate": 0.00013629840235769388, "loss": 0.1992, "step": 7688 }, { "epoch": 3.6183529411764708, "grad_norm": 0.221704961590562, "learning_rate": 0.0001362547500283845, "loss": 0.1518, "step": 7689 }, { "epoch": 3.6188235294117646, "grad_norm": 0.35612544617924247, "learning_rate": 0.00013621109887298685, "loss": 0.2451, "step": 7690 }, { "epoch": 3.6192941176470588, "grad_norm": 0.41685877530288834, "learning_rate": 0.00013616744889522898, "loss": 0.3602, "step": 7691 }, { "epoch": 3.619764705882353, "grad_norm": 0.5123098710536576, "learning_rate": 0.00013612380009883886, "loss": 0.3937, "step": 7692 }, { "epoch": 3.620235294117647, "grad_norm": 0.48872006132521234, "learning_rate": 0.00013608015248754425, "loss": 0.2373, "step": 7693 }, { "epoch": 3.620705882352941, "grad_norm": 0.22724137235906683, "learning_rate": 0.00013603650606507293, "loss": 0.1043, "step": 7694 }, { "epoch": 3.621176470588235, "grad_norm": 0.43721518406027154, "learning_rate": 0.0001359928608351524, "loss": 0.3536, "step": 7695 }, { "epoch": 3.6216470588235294, "grad_norm": 0.2836312684351209, "learning_rate": 0.00013594921680151032, "loss": 0.129, "step": 7696 }, { "epoch": 3.6221176470588237, "grad_norm": 0.46400977289447326, "learning_rate": 0.00013590557396787406, "loss": 0.4, "step": 7697 }, { "epoch": 3.622588235294118, "grad_norm": 0.45783064370513377, "learning_rate": 0.00013586193233797086, "loss": 0.4186, "step": 7698 }, { "epoch": 3.6230588235294117, "grad_norm": 0.3516353810124195, "learning_rate": 0.00013581829191552802, "loss": 0.1586, "step": 7699 }, { "epoch": 3.623529411764706, "grad_norm": 0.3952248410961963, "learning_rate": 0.00013577465270427262, "loss": 0.2463, "step": 7700 }, { "epoch": 3.624, "grad_norm": 0.40526780360016246, "learning_rate": 0.00013573101470793161, "loss": 0.165, "step": 7701 }, { "epoch": 3.624470588235294, "grad_norm": 0.46793615075004874, "learning_rate": 0.00013568737793023198, "loss": 0.1961, "step": 7702 }, { "epoch": 3.624941176470588, "grad_norm": 0.30849880320831163, "learning_rate": 0.00013564374237490047, "loss": 0.1459, "step": 7703 }, { "epoch": 3.6254117647058823, "grad_norm": 0.30351683071286845, "learning_rate": 0.0001356001080456638, "loss": 0.2091, "step": 7704 }, { "epoch": 3.6258823529411766, "grad_norm": 0.6445027688014784, "learning_rate": 0.00013555647494624846, "loss": 0.4689, "step": 7705 }, { "epoch": 3.626352941176471, "grad_norm": 0.5632456639237792, "learning_rate": 0.00013551284308038105, "loss": 0.3917, "step": 7706 }, { "epoch": 3.6268235294117646, "grad_norm": 0.43776377393548305, "learning_rate": 0.00013546921245178792, "loss": 0.1883, "step": 7707 }, { "epoch": 3.627294117647059, "grad_norm": 0.31851379877983343, "learning_rate": 0.00013542558306419526, "loss": 0.1961, "step": 7708 }, { "epoch": 3.627764705882353, "grad_norm": 0.8454115410688648, "learning_rate": 0.00013538195492132928, "loss": 0.4021, "step": 7709 }, { "epoch": 3.6282352941176472, "grad_norm": 0.35156115171873165, "learning_rate": 0.00013533832802691613, "loss": 0.2975, "step": 7710 }, { "epoch": 3.628705882352941, "grad_norm": 0.3752724185073319, "learning_rate": 0.00013529470238468166, "loss": 0.2573, "step": 7711 }, { "epoch": 3.6291764705882352, "grad_norm": 0.5670586652130419, "learning_rate": 0.00013525107799835172, "loss": 0.4705, "step": 7712 }, { "epoch": 3.6296470588235294, "grad_norm": 0.28271061110783685, "learning_rate": 0.00013520745487165207, "loss": 0.2077, "step": 7713 }, { "epoch": 3.6301176470588237, "grad_norm": 0.3682039988027689, "learning_rate": 0.00013516383300830835, "loss": 0.1954, "step": 7714 }, { "epoch": 3.630588235294118, "grad_norm": 0.4587803526930821, "learning_rate": 0.00013512021241204604, "loss": 0.3787, "step": 7715 }, { "epoch": 3.6310588235294117, "grad_norm": 0.5606822291928466, "learning_rate": 0.00013507659308659053, "loss": 0.3328, "step": 7716 }, { "epoch": 3.631529411764706, "grad_norm": 0.298478135103624, "learning_rate": 0.00013503297503566727, "loss": 0.1061, "step": 7717 }, { "epoch": 3.632, "grad_norm": 0.5794130994365199, "learning_rate": 0.00013498935826300132, "loss": 0.4623, "step": 7718 }, { "epoch": 3.632470588235294, "grad_norm": 0.33408031921353254, "learning_rate": 0.00013494574277231772, "loss": 0.0869, "step": 7719 }, { "epoch": 3.632941176470588, "grad_norm": 0.4570666536109246, "learning_rate": 0.00013490212856734162, "loss": 0.361, "step": 7720 }, { "epoch": 3.6334117647058823, "grad_norm": 0.44561010849309657, "learning_rate": 0.00013485851565179778, "loss": 0.3599, "step": 7721 }, { "epoch": 3.6338823529411766, "grad_norm": 0.5511155629997142, "learning_rate": 0.00013481490402941098, "loss": 0.3494, "step": 7722 }, { "epoch": 3.634352941176471, "grad_norm": 0.40177303235492723, "learning_rate": 0.00013477129370390582, "loss": 0.2694, "step": 7723 }, { "epoch": 3.6348235294117646, "grad_norm": 0.4810115844177368, "learning_rate": 0.00013472768467900686, "loss": 0.3981, "step": 7724 }, { "epoch": 3.635294117647059, "grad_norm": 0.3863384072427635, "learning_rate": 0.0001346840769584386, "loss": 0.2913, "step": 7725 }, { "epoch": 3.635764705882353, "grad_norm": 0.44008270132998767, "learning_rate": 0.00013464047054592525, "loss": 0.2458, "step": 7726 }, { "epoch": 3.636235294117647, "grad_norm": 0.3151876944220696, "learning_rate": 0.00013459686544519108, "loss": 0.2029, "step": 7727 }, { "epoch": 3.636705882352941, "grad_norm": 0.1525410037450352, "learning_rate": 0.00013455326165996015, "loss": 0.0306, "step": 7728 }, { "epoch": 3.6371764705882352, "grad_norm": 0.27151772843557825, "learning_rate": 0.0001345096591939564, "loss": 0.1219, "step": 7729 }, { "epoch": 3.6376470588235295, "grad_norm": 0.42095034124145164, "learning_rate": 0.00013446605805090376, "loss": 0.2901, "step": 7730 }, { "epoch": 3.6381176470588237, "grad_norm": 0.38263872725034437, "learning_rate": 0.000134422458234526, "loss": 0.2257, "step": 7731 }, { "epoch": 3.638588235294118, "grad_norm": 0.3257134043879855, "learning_rate": 0.00013437885974854665, "loss": 0.2086, "step": 7732 }, { "epoch": 3.6390588235294117, "grad_norm": 0.35449783509026384, "learning_rate": 0.00013433526259668931, "loss": 0.1902, "step": 7733 }, { "epoch": 3.639529411764706, "grad_norm": 0.2957050990131144, "learning_rate": 0.0001342916667826774, "loss": 0.1371, "step": 7734 }, { "epoch": 3.64, "grad_norm": 0.4097801514554806, "learning_rate": 0.0001342480723102342, "loss": 0.2847, "step": 7735 }, { "epoch": 3.640470588235294, "grad_norm": 0.32514534358676167, "learning_rate": 0.00013420447918308285, "loss": 0.1694, "step": 7736 }, { "epoch": 3.640941176470588, "grad_norm": 0.41319248873262043, "learning_rate": 0.00013416088740494649, "loss": 0.3629, "step": 7737 }, { "epoch": 3.6414117647058823, "grad_norm": 0.3202563181440041, "learning_rate": 0.00013411729697954803, "loss": 0.195, "step": 7738 }, { "epoch": 3.6418823529411766, "grad_norm": 0.48057241353266084, "learning_rate": 0.00013407370791061027, "loss": 0.2949, "step": 7739 }, { "epoch": 3.642352941176471, "grad_norm": 0.34988306314188844, "learning_rate": 0.00013403012020185603, "loss": 0.086, "step": 7740 }, { "epoch": 3.6428235294117646, "grad_norm": 0.3558568343886898, "learning_rate": 0.00013398653385700784, "loss": 0.16, "step": 7741 }, { "epoch": 3.643294117647059, "grad_norm": 0.3287093230900141, "learning_rate": 0.00013394294887978818, "loss": 0.16, "step": 7742 }, { "epoch": 3.643764705882353, "grad_norm": 0.1611906331728521, "learning_rate": 0.0001338993652739195, "loss": 0.1005, "step": 7743 }, { "epoch": 3.644235294117647, "grad_norm": 0.726785579901326, "learning_rate": 0.00013385578304312397, "loss": 0.2386, "step": 7744 }, { "epoch": 3.644705882352941, "grad_norm": 0.4137452596815051, "learning_rate": 0.00013381220219112373, "loss": 0.2098, "step": 7745 }, { "epoch": 3.6451764705882352, "grad_norm": 0.4671663389847656, "learning_rate": 0.00013376862272164084, "loss": 0.2818, "step": 7746 }, { "epoch": 3.6456470588235295, "grad_norm": 0.43822343539161995, "learning_rate": 0.00013372504463839718, "loss": 0.2032, "step": 7747 }, { "epoch": 3.6461176470588237, "grad_norm": 0.5668059890758339, "learning_rate": 0.00013368146794511457, "loss": 0.5749, "step": 7748 }, { "epoch": 3.646588235294118, "grad_norm": 0.443463450715964, "learning_rate": 0.0001336378926455146, "loss": 0.2884, "step": 7749 }, { "epoch": 3.6470588235294117, "grad_norm": 0.3491824065092756, "learning_rate": 0.00013359431874331886, "loss": 0.2654, "step": 7750 }, { "epoch": 3.647529411764706, "grad_norm": 0.2577836466449292, "learning_rate": 0.0001335507462422488, "loss": 0.1097, "step": 7751 }, { "epoch": 3.648, "grad_norm": 0.2797699460492023, "learning_rate": 0.00013350717514602563, "loss": 0.1671, "step": 7752 }, { "epoch": 3.648470588235294, "grad_norm": 0.1718472540591475, "learning_rate": 0.00013346360545837062, "loss": 0.0819, "step": 7753 }, { "epoch": 3.648941176470588, "grad_norm": 0.37353622489807053, "learning_rate": 0.0001334200371830048, "loss": 0.148, "step": 7754 }, { "epoch": 3.6494117647058824, "grad_norm": 0.2674911489624438, "learning_rate": 0.00013337647032364919, "loss": 0.1601, "step": 7755 }, { "epoch": 3.6498823529411766, "grad_norm": 0.6067583370433959, "learning_rate": 0.0001333329048840245, "loss": 0.4089, "step": 7756 }, { "epoch": 3.650352941176471, "grad_norm": 0.28568118459296565, "learning_rate": 0.0001332893408678515, "loss": 0.1983, "step": 7757 }, { "epoch": 3.6508235294117646, "grad_norm": 0.4463444660301295, "learning_rate": 0.00013324577827885078, "loss": 0.1712, "step": 7758 }, { "epoch": 3.651294117647059, "grad_norm": 0.4875976365330267, "learning_rate": 0.00013320221712074276, "loss": 0.2148, "step": 7759 }, { "epoch": 3.651764705882353, "grad_norm": 0.993151143655389, "learning_rate": 0.00013315865739724778, "loss": 0.6845, "step": 7760 }, { "epoch": 3.652235294117647, "grad_norm": 0.3663067862836546, "learning_rate": 0.00013311509911208608, "loss": 0.3223, "step": 7761 }, { "epoch": 3.652705882352941, "grad_norm": 0.2780956409318289, "learning_rate": 0.00013307154226897773, "loss": 0.1673, "step": 7762 }, { "epoch": 3.6531764705882352, "grad_norm": 0.5444850721368646, "learning_rate": 0.0001330279868716427, "loss": 0.2875, "step": 7763 }, { "epoch": 3.6536470588235295, "grad_norm": 0.46570909891635937, "learning_rate": 0.00013298443292380084, "loss": 0.2925, "step": 7764 }, { "epoch": 3.6541176470588237, "grad_norm": 0.29748353356711177, "learning_rate": 0.00013294088042917188, "loss": 0.1194, "step": 7765 }, { "epoch": 3.654588235294118, "grad_norm": 0.2639693649076732, "learning_rate": 0.0001328973293914754, "loss": 0.1173, "step": 7766 }, { "epoch": 3.6550588235294117, "grad_norm": 0.4753707861826605, "learning_rate": 0.00013285377981443085, "loss": 0.285, "step": 7767 }, { "epoch": 3.655529411764706, "grad_norm": 0.8983362033819775, "learning_rate": 0.00013281023170175766, "loss": 0.5135, "step": 7768 }, { "epoch": 3.656, "grad_norm": 0.5763491485249455, "learning_rate": 0.00013276668505717498, "loss": 0.3912, "step": 7769 }, { "epoch": 3.656470588235294, "grad_norm": 0.4860963422448824, "learning_rate": 0.00013272313988440187, "loss": 0.3164, "step": 7770 }, { "epoch": 3.656941176470588, "grad_norm": 0.42524243920659466, "learning_rate": 0.0001326795961871574, "loss": 0.3053, "step": 7771 }, { "epoch": 3.6574117647058824, "grad_norm": 0.3371087763262457, "learning_rate": 0.00013263605396916038, "loss": 0.1257, "step": 7772 }, { "epoch": 3.6578823529411766, "grad_norm": 0.19496881149373832, "learning_rate": 0.00013259251323412952, "loss": 0.0686, "step": 7773 }, { "epoch": 3.658352941176471, "grad_norm": 0.20526451549237668, "learning_rate": 0.0001325489739857834, "loss": 0.1113, "step": 7774 }, { "epoch": 3.6588235294117646, "grad_norm": 0.6160807460351027, "learning_rate": 0.0001325054362278405, "loss": 0.2711, "step": 7775 }, { "epoch": 3.659294117647059, "grad_norm": 0.3494205487300246, "learning_rate": 0.00013246189996401916, "loss": 0.1764, "step": 7776 }, { "epoch": 3.659764705882353, "grad_norm": 0.5923478364977471, "learning_rate": 0.00013241836519803756, "loss": 0.4231, "step": 7777 }, { "epoch": 3.660235294117647, "grad_norm": 0.565656014896998, "learning_rate": 0.00013237483193361387, "loss": 0.3235, "step": 7778 }, { "epoch": 3.660705882352941, "grad_norm": 0.19244023848442215, "learning_rate": 0.00013233130017446595, "loss": 0.1317, "step": 7779 }, { "epoch": 3.6611764705882353, "grad_norm": 0.5090096682115939, "learning_rate": 0.00013228776992431165, "loss": 0.4739, "step": 7780 }, { "epoch": 3.6616470588235295, "grad_norm": 0.4185233528688906, "learning_rate": 0.00013224424118686868, "loss": 0.3295, "step": 7781 }, { "epoch": 3.6621176470588237, "grad_norm": 0.3908268227684962, "learning_rate": 0.00013220071396585463, "loss": 0.2236, "step": 7782 }, { "epoch": 3.662588235294118, "grad_norm": 0.37670005622967323, "learning_rate": 0.0001321571882649869, "loss": 0.2512, "step": 7783 }, { "epoch": 3.6630588235294117, "grad_norm": 0.43515642766909024, "learning_rate": 0.00013211366408798277, "loss": 0.2643, "step": 7784 }, { "epoch": 3.663529411764706, "grad_norm": 0.3688348534239938, "learning_rate": 0.00013207014143855951, "loss": 0.263, "step": 7785 }, { "epoch": 3.664, "grad_norm": 0.5615203199217023, "learning_rate": 0.00013202662032043408, "loss": 0.2412, "step": 7786 }, { "epoch": 3.664470588235294, "grad_norm": 0.270953366694382, "learning_rate": 0.00013198310073732344, "loss": 0.1017, "step": 7787 }, { "epoch": 3.664941176470588, "grad_norm": 0.2724176028636436, "learning_rate": 0.0001319395826929444, "loss": 0.2013, "step": 7788 }, { "epoch": 3.6654117647058824, "grad_norm": 0.3135337807624623, "learning_rate": 0.00013189606619101358, "loss": 0.2152, "step": 7789 }, { "epoch": 3.6658823529411766, "grad_norm": 0.16786931251063655, "learning_rate": 0.00013185255123524745, "loss": 0.0484, "step": 7790 }, { "epoch": 3.666352941176471, "grad_norm": 0.498410370035348, "learning_rate": 0.00013180903782936254, "loss": 0.4413, "step": 7791 }, { "epoch": 3.6668235294117646, "grad_norm": 0.5434314292889422, "learning_rate": 0.000131765525977075, "loss": 0.4352, "step": 7792 }, { "epoch": 3.667294117647059, "grad_norm": 0.22531948099199317, "learning_rate": 0.0001317220156821009, "loss": 0.1428, "step": 7793 }, { "epoch": 3.667764705882353, "grad_norm": 0.32263274162202843, "learning_rate": 0.00013167850694815642, "loss": 0.2309, "step": 7794 }, { "epoch": 3.668235294117647, "grad_norm": 0.33904164421367156, "learning_rate": 0.0001316349997789573, "loss": 0.2186, "step": 7795 }, { "epoch": 3.668705882352941, "grad_norm": 0.22483684943846188, "learning_rate": 0.00013159149417821923, "loss": 0.1353, "step": 7796 }, { "epoch": 3.6691764705882353, "grad_norm": 1.3158703583657567, "learning_rate": 0.00013154799014965785, "loss": 0.7666, "step": 7797 }, { "epoch": 3.6696470588235295, "grad_norm": 0.13176231705540084, "learning_rate": 0.00013150448769698857, "loss": 0.0713, "step": 7798 }, { "epoch": 3.6701176470588237, "grad_norm": 0.1954595782870741, "learning_rate": 0.00013146098682392685, "loss": 0.068, "step": 7799 }, { "epoch": 3.6705882352941175, "grad_norm": 0.16262425872768296, "learning_rate": 0.0001314174875341878, "loss": 0.0881, "step": 7800 }, { "epoch": 3.6710588235294117, "grad_norm": 0.2512369583862865, "learning_rate": 0.00013137398983148642, "loss": 0.1139, "step": 7801 }, { "epoch": 3.671529411764706, "grad_norm": 0.6002292279788636, "learning_rate": 0.00013133049371953767, "loss": 0.3702, "step": 7802 }, { "epoch": 3.672, "grad_norm": 0.3990033853515003, "learning_rate": 0.00013128699920205634, "loss": 0.2002, "step": 7803 }, { "epoch": 3.672470588235294, "grad_norm": 0.5633571190152452, "learning_rate": 0.00013124350628275708, "loss": 0.3864, "step": 7804 }, { "epoch": 3.672941176470588, "grad_norm": 0.3453835752827489, "learning_rate": 0.00013120001496535433, "loss": 0.279, "step": 7805 }, { "epoch": 3.6734117647058824, "grad_norm": 0.4451853943761037, "learning_rate": 0.00013115652525356257, "loss": 0.2925, "step": 7806 }, { "epoch": 3.6738823529411766, "grad_norm": 0.1772136471611208, "learning_rate": 0.00013111303715109598, "loss": 0.0803, "step": 7807 }, { "epoch": 3.674352941176471, "grad_norm": 0.10351754694183052, "learning_rate": 0.0001310695506616686, "loss": 0.0375, "step": 7808 }, { "epoch": 3.6748235294117646, "grad_norm": 0.1815881129550629, "learning_rate": 0.0001310260657889945, "loss": 0.0698, "step": 7809 }, { "epoch": 3.675294117647059, "grad_norm": 0.19478019862878893, "learning_rate": 0.00013098258253678744, "loss": 0.1023, "step": 7810 }, { "epoch": 3.675764705882353, "grad_norm": 0.19290343910009736, "learning_rate": 0.00013093910090876108, "loss": 0.0833, "step": 7811 }, { "epoch": 3.676235294117647, "grad_norm": 0.2892977951563181, "learning_rate": 0.00013089562090862905, "loss": 0.2202, "step": 7812 }, { "epoch": 3.676705882352941, "grad_norm": 0.3003059029670459, "learning_rate": 0.00013085214254010465, "loss": 0.1777, "step": 7813 }, { "epoch": 3.6771764705882353, "grad_norm": 0.5238354030929153, "learning_rate": 0.00013080866580690126, "loss": 0.3244, "step": 7814 }, { "epoch": 3.6776470588235295, "grad_norm": 0.40986945963137883, "learning_rate": 0.00013076519071273187, "loss": 0.3352, "step": 7815 }, { "epoch": 3.6781176470588237, "grad_norm": 0.283253777609644, "learning_rate": 0.00013072171726130957, "loss": 0.1691, "step": 7816 }, { "epoch": 3.6785882352941175, "grad_norm": 0.5043282128684051, "learning_rate": 0.00013067824545634718, "loss": 0.3775, "step": 7817 }, { "epoch": 3.6790588235294117, "grad_norm": 0.20713119657452525, "learning_rate": 0.00013063477530155738, "loss": 0.0533, "step": 7818 }, { "epoch": 3.679529411764706, "grad_norm": 0.4134467175827058, "learning_rate": 0.00013059130680065278, "loss": 0.3514, "step": 7819 }, { "epoch": 3.68, "grad_norm": 0.8077959247525132, "learning_rate": 0.0001305478399573458, "loss": 0.2923, "step": 7820 }, { "epoch": 3.680470588235294, "grad_norm": 0.22709908103602083, "learning_rate": 0.00013050437477534862, "loss": 0.0946, "step": 7821 }, { "epoch": 3.680941176470588, "grad_norm": 0.3839906592952088, "learning_rate": 0.00013046091125837352, "loss": 0.1298, "step": 7822 }, { "epoch": 3.6814117647058824, "grad_norm": 0.28776778333787434, "learning_rate": 0.00013041744941013245, "loss": 0.179, "step": 7823 }, { "epoch": 3.6818823529411766, "grad_norm": 0.2546799858522931, "learning_rate": 0.00013037398923433725, "loss": 0.1207, "step": 7824 }, { "epoch": 3.682352941176471, "grad_norm": 0.13690883314244406, "learning_rate": 0.00013033053073469958, "loss": 0.0464, "step": 7825 }, { "epoch": 3.6828235294117646, "grad_norm": 0.3601913628742346, "learning_rate": 0.00013028707391493112, "loss": 0.2679, "step": 7826 }, { "epoch": 3.683294117647059, "grad_norm": 0.39892227585819856, "learning_rate": 0.00013024361877874324, "loss": 0.1532, "step": 7827 }, { "epoch": 3.683764705882353, "grad_norm": 0.7090164447713806, "learning_rate": 0.00013020016532984714, "loss": 0.3593, "step": 7828 }, { "epoch": 3.684235294117647, "grad_norm": 0.3048265789157166, "learning_rate": 0.00013015671357195413, "loss": 0.1331, "step": 7829 }, { "epoch": 3.684705882352941, "grad_norm": 0.558748882346546, "learning_rate": 0.00013011326350877508, "loss": 0.5001, "step": 7830 }, { "epoch": 3.6851764705882353, "grad_norm": 0.30143273884835364, "learning_rate": 0.00013006981514402085, "loss": 0.2415, "step": 7831 }, { "epoch": 3.6856470588235295, "grad_norm": 0.36701788172373445, "learning_rate": 0.00013002636848140217, "loss": 0.2988, "step": 7832 }, { "epoch": 3.6861176470588237, "grad_norm": 0.33948918107454995, "learning_rate": 0.00012998292352462963, "loss": 0.321, "step": 7833 }, { "epoch": 3.6865882352941175, "grad_norm": 0.23290636096302733, "learning_rate": 0.00012993948027741357, "loss": 0.1192, "step": 7834 }, { "epoch": 3.6870588235294117, "grad_norm": 0.3429176831186486, "learning_rate": 0.0001298960387434643, "loss": 0.2241, "step": 7835 }, { "epoch": 3.687529411764706, "grad_norm": 0.19983342376759236, "learning_rate": 0.00012985259892649191, "loss": 0.0664, "step": 7836 }, { "epoch": 3.6879999999999997, "grad_norm": 0.1681854402669644, "learning_rate": 0.00012980916083020646, "loss": 0.0908, "step": 7837 }, { "epoch": 3.688470588235294, "grad_norm": 0.41421886875336766, "learning_rate": 0.0001297657244583176, "loss": 0.2142, "step": 7838 }, { "epoch": 3.688941176470588, "grad_norm": 0.3373865246319564, "learning_rate": 0.00012972228981453522, "loss": 0.2479, "step": 7839 }, { "epoch": 3.6894117647058824, "grad_norm": 0.24561622296930793, "learning_rate": 0.00012967885690256872, "loss": 0.1147, "step": 7840 }, { "epoch": 3.6898823529411766, "grad_norm": 0.4597478033547869, "learning_rate": 0.0001296354257261275, "loss": 0.3949, "step": 7841 }, { "epoch": 3.690352941176471, "grad_norm": 0.3719698331738357, "learning_rate": 0.0001295919962889208, "loss": 0.3105, "step": 7842 }, { "epoch": 3.6908235294117646, "grad_norm": 0.6345264325196117, "learning_rate": 0.0001295485685946577, "loss": 0.5929, "step": 7843 }, { "epoch": 3.691294117647059, "grad_norm": 0.4245705286008463, "learning_rate": 0.0001295051426470472, "loss": 0.3131, "step": 7844 }, { "epoch": 3.691764705882353, "grad_norm": 0.2996487785084394, "learning_rate": 0.00012946171844979804, "loss": 0.194, "step": 7845 }, { "epoch": 3.692235294117647, "grad_norm": 0.17281447134897174, "learning_rate": 0.00012941829600661886, "loss": 0.0352, "step": 7846 }, { "epoch": 3.692705882352941, "grad_norm": 0.31987644234615936, "learning_rate": 0.00012937487532121818, "loss": 0.2974, "step": 7847 }, { "epoch": 3.6931764705882353, "grad_norm": 0.5135048486744631, "learning_rate": 0.00012933145639730428, "loss": 0.5042, "step": 7848 }, { "epoch": 3.6936470588235295, "grad_norm": 0.5662338682754584, "learning_rate": 0.00012928803923858536, "loss": 0.3017, "step": 7849 }, { "epoch": 3.6941176470588237, "grad_norm": 0.27829364420535313, "learning_rate": 0.00012924462384876953, "loss": 0.0945, "step": 7850 }, { "epoch": 3.6945882352941175, "grad_norm": 0.3290746510286145, "learning_rate": 0.00012920121023156462, "loss": 0.2313, "step": 7851 }, { "epoch": 3.6950588235294117, "grad_norm": 0.4529172398718315, "learning_rate": 0.00012915779839067838, "loss": 0.3276, "step": 7852 }, { "epoch": 3.695529411764706, "grad_norm": 0.3313533066275173, "learning_rate": 0.00012911438832981836, "loss": 0.1759, "step": 7853 }, { "epoch": 3.6959999999999997, "grad_norm": 0.2526751725620692, "learning_rate": 0.00012907098005269203, "loss": 0.1122, "step": 7854 }, { "epoch": 3.696470588235294, "grad_norm": 0.41764845691220664, "learning_rate": 0.00012902757356300667, "loss": 0.338, "step": 7855 }, { "epoch": 3.696941176470588, "grad_norm": 0.5338960463522632, "learning_rate": 0.00012898416886446937, "loss": 0.447, "step": 7856 }, { "epoch": 3.6974117647058824, "grad_norm": 0.5592882998000931, "learning_rate": 0.00012894076596078717, "loss": 0.3593, "step": 7857 }, { "epoch": 3.6978823529411766, "grad_norm": 0.4708697524379397, "learning_rate": 0.00012889736485566683, "loss": 0.288, "step": 7858 }, { "epoch": 3.698352941176471, "grad_norm": 0.43567338722421234, "learning_rate": 0.000128853965552815, "loss": 0.2972, "step": 7859 }, { "epoch": 3.6988235294117646, "grad_norm": 0.4254374750278421, "learning_rate": 0.0001288105680559383, "loss": 0.2893, "step": 7860 }, { "epoch": 3.699294117647059, "grad_norm": 0.4354755621665817, "learning_rate": 0.000128767172368743, "loss": 0.3225, "step": 7861 }, { "epoch": 3.699764705882353, "grad_norm": 0.3542074106516374, "learning_rate": 0.0001287237784949353, "loss": 0.2584, "step": 7862 }, { "epoch": 3.700235294117647, "grad_norm": 0.33204569847315474, "learning_rate": 0.0001286803864382213, "loss": 0.1366, "step": 7863 }, { "epoch": 3.700705882352941, "grad_norm": 0.30022813857904995, "learning_rate": 0.00012863699620230688, "loss": 0.1944, "step": 7864 }, { "epoch": 3.7011764705882353, "grad_norm": 0.3506621600941569, "learning_rate": 0.00012859360779089775, "loss": 0.2175, "step": 7865 }, { "epoch": 3.7016470588235295, "grad_norm": 0.15189606988342436, "learning_rate": 0.00012855022120769952, "loss": 0.0437, "step": 7866 }, { "epoch": 3.7021176470588237, "grad_norm": 0.37674950161017673, "learning_rate": 0.0001285068364564176, "loss": 0.2047, "step": 7867 }, { "epoch": 3.7025882352941175, "grad_norm": 0.48570621183925927, "learning_rate": 0.0001284634535407573, "loss": 0.3194, "step": 7868 }, { "epoch": 3.7030588235294117, "grad_norm": 0.21412214680883884, "learning_rate": 0.00012842007246442367, "loss": 0.0956, "step": 7869 }, { "epoch": 3.703529411764706, "grad_norm": 0.37828068368639634, "learning_rate": 0.0001283766932311217, "loss": 0.1806, "step": 7870 }, { "epoch": 3.7039999999999997, "grad_norm": 0.29072652753985356, "learning_rate": 0.00012833331584455622, "loss": 0.1569, "step": 7871 }, { "epoch": 3.704470588235294, "grad_norm": 0.4964358103491723, "learning_rate": 0.0001282899403084318, "loss": 0.308, "step": 7872 }, { "epoch": 3.704941176470588, "grad_norm": 0.29808365375574253, "learning_rate": 0.000128246566626453, "loss": 0.124, "step": 7873 }, { "epoch": 3.7054117647058824, "grad_norm": 0.42282696637908984, "learning_rate": 0.0001282031948023241, "loss": 0.259, "step": 7874 }, { "epoch": 3.7058823529411766, "grad_norm": 0.41554594519288235, "learning_rate": 0.0001281598248397493, "loss": 0.4346, "step": 7875 }, { "epoch": 3.706352941176471, "grad_norm": 0.0701917729704216, "learning_rate": 0.00012811645674243254, "loss": 0.0172, "step": 7876 }, { "epoch": 3.7068235294117646, "grad_norm": 0.3935585279957979, "learning_rate": 0.00012807309051407774, "loss": 0.2139, "step": 7877 }, { "epoch": 3.707294117647059, "grad_norm": 1.061618907831858, "learning_rate": 0.00012802972615838856, "loss": 0.9926, "step": 7878 }, { "epoch": 3.707764705882353, "grad_norm": 0.1751660291992852, "learning_rate": 0.00012798636367906849, "loss": 0.0408, "step": 7879 }, { "epoch": 3.708235294117647, "grad_norm": 0.42564096242157645, "learning_rate": 0.00012794300307982098, "loss": 0.2924, "step": 7880 }, { "epoch": 3.708705882352941, "grad_norm": 0.5034545902538913, "learning_rate": 0.0001278996443643492, "loss": 0.266, "step": 7881 }, { "epoch": 3.7091764705882353, "grad_norm": 0.17163686337305828, "learning_rate": 0.00012785628753635617, "loss": 0.0705, "step": 7882 }, { "epoch": 3.7096470588235295, "grad_norm": 0.27598649034758593, "learning_rate": 0.0001278129325995448, "loss": 0.1956, "step": 7883 }, { "epoch": 3.7101176470588237, "grad_norm": 0.8157146272303046, "learning_rate": 0.00012776957955761784, "loss": 0.3475, "step": 7884 }, { "epoch": 3.7105882352941175, "grad_norm": 0.33995129348401076, "learning_rate": 0.0001277262284142778, "loss": 0.2082, "step": 7885 }, { "epoch": 3.7110588235294117, "grad_norm": 0.31700675100499953, "learning_rate": 0.00012768287917322706, "loss": 0.2139, "step": 7886 }, { "epoch": 3.711529411764706, "grad_norm": 0.33494193289206925, "learning_rate": 0.00012763953183816794, "loss": 0.2202, "step": 7887 }, { "epoch": 3.7119999999999997, "grad_norm": 0.5403687382254512, "learning_rate": 0.00012759618641280251, "loss": 0.2194, "step": 7888 }, { "epoch": 3.712470588235294, "grad_norm": 0.30750978894338166, "learning_rate": 0.00012755284290083264, "loss": 0.1125, "step": 7889 }, { "epoch": 3.712941176470588, "grad_norm": 0.30848644931141017, "learning_rate": 0.00012750950130596006, "loss": 0.121, "step": 7890 }, { "epoch": 3.7134117647058824, "grad_norm": 0.7588374306853221, "learning_rate": 0.0001274661616318864, "loss": 0.5156, "step": 7891 }, { "epoch": 3.7138823529411766, "grad_norm": 0.7387442058237561, "learning_rate": 0.0001274228238823131, "loss": 0.2786, "step": 7892 }, { "epoch": 3.714352941176471, "grad_norm": 0.4536988255115725, "learning_rate": 0.00012737948806094135, "loss": 0.2318, "step": 7893 }, { "epoch": 3.7148235294117646, "grad_norm": 0.48098542251867765, "learning_rate": 0.00012733615417147228, "loss": 0.2105, "step": 7894 }, { "epoch": 3.715294117647059, "grad_norm": 0.5686764289885188, "learning_rate": 0.0001272928222176068, "loss": 0.5158, "step": 7895 }, { "epoch": 3.715764705882353, "grad_norm": 0.34397169827614954, "learning_rate": 0.0001272494922030457, "loss": 0.1455, "step": 7896 }, { "epoch": 3.716235294117647, "grad_norm": 0.36690954931159614, "learning_rate": 0.0001272061641314895, "loss": 0.1854, "step": 7897 }, { "epoch": 3.716705882352941, "grad_norm": 0.42686490450091075, "learning_rate": 0.00012716283800663874, "loss": 0.3195, "step": 7898 }, { "epoch": 3.7171764705882353, "grad_norm": 0.48055129342155817, "learning_rate": 0.00012711951383219362, "loss": 0.2491, "step": 7899 }, { "epoch": 3.7176470588235295, "grad_norm": 0.3562661399609016, "learning_rate": 0.00012707619161185423, "loss": 0.2637, "step": 7900 }, { "epoch": 3.7181176470588237, "grad_norm": 0.25001037141600524, "learning_rate": 0.00012703287134932054, "loss": 0.1192, "step": 7901 }, { "epoch": 3.7185882352941175, "grad_norm": 0.3147460533196467, "learning_rate": 0.0001269895530482923, "loss": 0.137, "step": 7902 }, { "epoch": 3.7190588235294118, "grad_norm": 0.3827418319777715, "learning_rate": 0.00012694623671246908, "loss": 0.3066, "step": 7903 }, { "epoch": 3.719529411764706, "grad_norm": 0.4716564773226344, "learning_rate": 0.0001269029223455503, "loss": 0.3878, "step": 7904 }, { "epoch": 3.7199999999999998, "grad_norm": 0.41509929861999345, "learning_rate": 0.00012685960995123527, "loss": 0.2164, "step": 7905 }, { "epoch": 3.720470588235294, "grad_norm": 0.49440928867439526, "learning_rate": 0.00012681629953322307, "loss": 0.3303, "step": 7906 }, { "epoch": 3.720941176470588, "grad_norm": 0.457537621309603, "learning_rate": 0.00012677299109521256, "loss": 0.3543, "step": 7907 }, { "epoch": 3.7214117647058824, "grad_norm": 0.31428102787835654, "learning_rate": 0.00012672968464090255, "loss": 0.1514, "step": 7908 }, { "epoch": 3.7218823529411766, "grad_norm": 0.09209243517602944, "learning_rate": 0.00012668638017399164, "loss": 0.0193, "step": 7909 }, { "epoch": 3.722352941176471, "grad_norm": 0.5505929594929028, "learning_rate": 0.00012664307769817815, "loss": 0.5051, "step": 7910 }, { "epoch": 3.7228235294117646, "grad_norm": 3.4293040246112607, "learning_rate": 0.00012659977721716045, "loss": 1.3321, "step": 7911 }, { "epoch": 3.723294117647059, "grad_norm": 0.45102696937857456, "learning_rate": 0.00012655647873463657, "loss": 0.2572, "step": 7912 }, { "epoch": 3.723764705882353, "grad_norm": 0.2604486104669232, "learning_rate": 0.00012651318225430437, "loss": 0.1309, "step": 7913 }, { "epoch": 3.724235294117647, "grad_norm": 0.523465223170347, "learning_rate": 0.00012646988777986157, "loss": 0.418, "step": 7914 }, { "epoch": 3.724705882352941, "grad_norm": 0.1968063576979468, "learning_rate": 0.00012642659531500583, "loss": 0.0608, "step": 7915 }, { "epoch": 3.7251764705882353, "grad_norm": 0.33810681323391756, "learning_rate": 0.00012638330486343445, "loss": 0.15, "step": 7916 }, { "epoch": 3.7256470588235295, "grad_norm": 0.3629148722996588, "learning_rate": 0.00012634001642884465, "loss": 0.2371, "step": 7917 }, { "epoch": 3.7261176470588238, "grad_norm": 0.37200261153404507, "learning_rate": 0.0001262967300149335, "loss": 0.204, "step": 7918 }, { "epoch": 3.7265882352941175, "grad_norm": 0.8940389515877922, "learning_rate": 0.0001262534456253979, "loss": 0.3406, "step": 7919 }, { "epoch": 3.7270588235294118, "grad_norm": 0.3951940429082629, "learning_rate": 0.0001262101632639345, "loss": 0.3309, "step": 7920 }, { "epoch": 3.727529411764706, "grad_norm": 0.21351335157448192, "learning_rate": 0.00012616688293423987, "loss": 0.0977, "step": 7921 }, { "epoch": 3.7279999999999998, "grad_norm": 0.1763824437254144, "learning_rate": 0.00012612360464001033, "loss": 0.0892, "step": 7922 }, { "epoch": 3.728470588235294, "grad_norm": 0.5726868688505851, "learning_rate": 0.00012608032838494205, "loss": 0.6721, "step": 7923 }, { "epoch": 3.728941176470588, "grad_norm": 0.4050103696911144, "learning_rate": 0.00012603705417273107, "loss": 0.2117, "step": 7924 }, { "epoch": 3.7294117647058824, "grad_norm": 0.26481445369468276, "learning_rate": 0.0001259937820070732, "loss": 0.183, "step": 7925 }, { "epoch": 3.7298823529411766, "grad_norm": 0.20936365872169055, "learning_rate": 0.0001259505118916641, "loss": 0.1284, "step": 7926 }, { "epoch": 3.7303529411764704, "grad_norm": 0.5011887216746786, "learning_rate": 0.0001259072438301992, "loss": 0.5077, "step": 7927 }, { "epoch": 3.7308235294117646, "grad_norm": 0.44781335735115485, "learning_rate": 0.00012586397782637392, "loss": 0.2778, "step": 7928 }, { "epoch": 3.731294117647059, "grad_norm": 0.4193998562958937, "learning_rate": 0.0001258207138838833, "loss": 0.2107, "step": 7929 }, { "epoch": 3.731764705882353, "grad_norm": 0.442205652880372, "learning_rate": 0.00012577745200642227, "loss": 0.3018, "step": 7930 }, { "epoch": 3.732235294117647, "grad_norm": 0.5416189079427601, "learning_rate": 0.00012573419219768562, "loss": 0.4112, "step": 7931 }, { "epoch": 3.732705882352941, "grad_norm": 0.3811970429295145, "learning_rate": 0.00012569093446136806, "loss": 0.2814, "step": 7932 }, { "epoch": 3.7331764705882353, "grad_norm": 0.21799966906663526, "learning_rate": 0.0001256476788011639, "loss": 0.0809, "step": 7933 }, { "epoch": 3.7336470588235295, "grad_norm": 0.2551192973727383, "learning_rate": 0.00012560442522076745, "loss": 0.1228, "step": 7934 }, { "epoch": 3.7341176470588238, "grad_norm": 0.16871951937645993, "learning_rate": 0.00012556117372387272, "loss": 0.0519, "step": 7935 }, { "epoch": 3.7345882352941175, "grad_norm": 0.37508616186732313, "learning_rate": 0.00012551792431417363, "loss": 0.2288, "step": 7936 }, { "epoch": 3.7350588235294118, "grad_norm": 0.2892102603451945, "learning_rate": 0.0001254746769953639, "loss": 0.1353, "step": 7937 }, { "epoch": 3.735529411764706, "grad_norm": 0.3121681645019337, "learning_rate": 0.000125431431771137, "loss": 0.1477, "step": 7938 }, { "epoch": 3.7359999999999998, "grad_norm": 0.4997058549062567, "learning_rate": 0.0001253881886451864, "loss": 0.2636, "step": 7939 }, { "epoch": 3.736470588235294, "grad_norm": 0.2581371724792003, "learning_rate": 0.0001253449476212052, "loss": 0.1573, "step": 7940 }, { "epoch": 3.736941176470588, "grad_norm": 0.3805966421478141, "learning_rate": 0.00012530170870288636, "loss": 0.221, "step": 7941 }, { "epoch": 3.7374117647058824, "grad_norm": 0.3258917975720982, "learning_rate": 0.00012525847189392278, "loss": 0.2747, "step": 7942 }, { "epoch": 3.7378823529411767, "grad_norm": 0.3593561317594438, "learning_rate": 0.00012521523719800706, "loss": 0.1442, "step": 7943 }, { "epoch": 3.7383529411764704, "grad_norm": 0.22396603779503416, "learning_rate": 0.00012517200461883163, "loss": 0.1186, "step": 7944 }, { "epoch": 3.7388235294117647, "grad_norm": 0.1768667909969878, "learning_rate": 0.00012512877416008878, "loss": 0.0947, "step": 7945 }, { "epoch": 3.739294117647059, "grad_norm": 0.3625705887791623, "learning_rate": 0.00012508554582547064, "loss": 0.33, "step": 7946 }, { "epoch": 3.739764705882353, "grad_norm": 0.3760051386224014, "learning_rate": 0.00012504231961866908, "loss": 0.4394, "step": 7947 }, { "epoch": 3.740235294117647, "grad_norm": 0.4413583251131774, "learning_rate": 0.00012499909554337582, "loss": 0.4242, "step": 7948 }, { "epoch": 3.740705882352941, "grad_norm": 0.35764480972687673, "learning_rate": 0.00012495587360328246, "loss": 0.2315, "step": 7949 }, { "epoch": 3.7411764705882353, "grad_norm": 0.21523828515460391, "learning_rate": 0.00012491265380208032, "loss": 0.0755, "step": 7950 }, { "epoch": 3.7416470588235295, "grad_norm": 0.2261069225301566, "learning_rate": 0.0001248694361434606, "loss": 0.1076, "step": 7951 }, { "epoch": 3.7421176470588238, "grad_norm": 0.41371013483768937, "learning_rate": 0.0001248262206311143, "loss": 0.3661, "step": 7952 }, { "epoch": 3.7425882352941175, "grad_norm": 0.7986740330447128, "learning_rate": 0.00012478300726873225, "loss": 0.2832, "step": 7953 }, { "epoch": 3.7430588235294118, "grad_norm": 0.2491181250901987, "learning_rate": 0.00012473979606000504, "loss": 0.1117, "step": 7954 }, { "epoch": 3.743529411764706, "grad_norm": 0.29080313209697256, "learning_rate": 0.00012469658700862315, "loss": 0.2148, "step": 7955 }, { "epoch": 3.7439999999999998, "grad_norm": 0.47823123767322495, "learning_rate": 0.00012465338011827685, "loss": 0.3135, "step": 7956 }, { "epoch": 3.744470588235294, "grad_norm": 0.15829944501914814, "learning_rate": 0.00012461017539265622, "loss": 0.0463, "step": 7957 }, { "epoch": 3.744941176470588, "grad_norm": 0.6776721386973185, "learning_rate": 0.00012456697283545114, "loss": 0.4306, "step": 7958 }, { "epoch": 3.7454117647058824, "grad_norm": 0.34259124986188294, "learning_rate": 0.00012452377245035132, "loss": 0.3116, "step": 7959 }, { "epoch": 3.7458823529411767, "grad_norm": 0.486632106710934, "learning_rate": 0.00012448057424104632, "loss": 0.2115, "step": 7960 }, { "epoch": 3.7463529411764704, "grad_norm": 0.39579055310575545, "learning_rate": 0.00012443737821122541, "loss": 0.3156, "step": 7961 }, { "epoch": 3.7468235294117647, "grad_norm": 0.2765285375520829, "learning_rate": 0.0001243941843645778, "loss": 0.0951, "step": 7962 }, { "epoch": 3.747294117647059, "grad_norm": 0.4534735699808478, "learning_rate": 0.00012435099270479246, "loss": 0.2368, "step": 7963 }, { "epoch": 3.7477647058823527, "grad_norm": 0.41326566403046777, "learning_rate": 0.00012430780323555816, "loss": 0.2902, "step": 7964 }, { "epoch": 3.748235294117647, "grad_norm": 0.35057856503864726, "learning_rate": 0.00012426461596056344, "loss": 0.2069, "step": 7965 }, { "epoch": 3.748705882352941, "grad_norm": 0.3923378267107986, "learning_rate": 0.00012422143088349679, "loss": 0.2615, "step": 7966 }, { "epoch": 3.7491764705882353, "grad_norm": 0.798593111592012, "learning_rate": 0.00012417824800804638, "loss": 0.4229, "step": 7967 }, { "epoch": 3.7496470588235296, "grad_norm": 0.47254376389207375, "learning_rate": 0.0001241350673379002, "loss": 0.2416, "step": 7968 }, { "epoch": 3.7501176470588238, "grad_norm": 0.28218776010601576, "learning_rate": 0.0001240918888767462, "loss": 0.1427, "step": 7969 }, { "epoch": 3.7505882352941176, "grad_norm": 0.4072111325839789, "learning_rate": 0.00012404871262827193, "loss": 0.2905, "step": 7970 }, { "epoch": 3.751058823529412, "grad_norm": 0.33317228019725564, "learning_rate": 0.00012400553859616488, "loss": 0.1692, "step": 7971 }, { "epoch": 3.751529411764706, "grad_norm": 0.3345323328504318, "learning_rate": 0.00012396236678411236, "loss": 0.269, "step": 7972 }, { "epoch": 3.752, "grad_norm": 0.32639790220733894, "learning_rate": 0.00012391919719580144, "loss": 0.2595, "step": 7973 }, { "epoch": 3.752470588235294, "grad_norm": 0.1851185662299799, "learning_rate": 0.00012387602983491894, "loss": 0.0532, "step": 7974 }, { "epoch": 3.7529411764705882, "grad_norm": 0.7045705954759113, "learning_rate": 0.00012383286470515166, "loss": 0.5602, "step": 7975 }, { "epoch": 3.7534117647058824, "grad_norm": 0.30894922324559787, "learning_rate": 0.0001237897018101861, "loss": 0.0945, "step": 7976 }, { "epoch": 3.7538823529411767, "grad_norm": 0.5000889728785193, "learning_rate": 0.00012374654115370856, "loss": 0.2925, "step": 7977 }, { "epoch": 3.7543529411764704, "grad_norm": 0.21385746910206863, "learning_rate": 0.0001237033827394052, "loss": 0.155, "step": 7978 }, { "epoch": 3.7548235294117647, "grad_norm": 0.4777619267154603, "learning_rate": 0.0001236602265709619, "loss": 0.487, "step": 7979 }, { "epoch": 3.755294117647059, "grad_norm": 0.3569303300141174, "learning_rate": 0.00012361707265206448, "loss": 0.2035, "step": 7980 }, { "epoch": 3.7557647058823527, "grad_norm": 0.400167089825291, "learning_rate": 0.00012357392098639848, "loss": 0.1914, "step": 7981 }, { "epoch": 3.756235294117647, "grad_norm": 0.29931834733123225, "learning_rate": 0.00012353077157764922, "loss": 0.2156, "step": 7982 }, { "epoch": 3.756705882352941, "grad_norm": 0.47954913470040045, "learning_rate": 0.00012348762442950187, "loss": 0.3353, "step": 7983 }, { "epoch": 3.7571764705882353, "grad_norm": 0.40486154899298354, "learning_rate": 0.00012344447954564148, "loss": 0.2171, "step": 7984 }, { "epoch": 3.7576470588235296, "grad_norm": 0.20707002526010576, "learning_rate": 0.0001234013369297528, "loss": 0.0696, "step": 7985 }, { "epoch": 3.758117647058824, "grad_norm": 0.5341591660775357, "learning_rate": 0.00012335819658552036, "loss": 0.4059, "step": 7986 }, { "epoch": 3.7585882352941176, "grad_norm": 0.5899126406015875, "learning_rate": 0.00012331505851662864, "loss": 0.4824, "step": 7987 }, { "epoch": 3.759058823529412, "grad_norm": 0.39631782006288135, "learning_rate": 0.00012327192272676183, "loss": 0.276, "step": 7988 }, { "epoch": 3.759529411764706, "grad_norm": 0.44976297965252027, "learning_rate": 0.0001232287892196039, "loss": 0.2699, "step": 7989 }, { "epoch": 3.76, "grad_norm": 0.2994994995041866, "learning_rate": 0.00012318565799883865, "loss": 0.1509, "step": 7990 }, { "epoch": 3.760470588235294, "grad_norm": 0.16345842257117124, "learning_rate": 0.0001231425290681498, "loss": 0.0846, "step": 7991 }, { "epoch": 3.7609411764705882, "grad_norm": 0.3320622144242474, "learning_rate": 0.00012309940243122064, "loss": 0.1852, "step": 7992 }, { "epoch": 3.7614117647058825, "grad_norm": 0.35855998583127474, "learning_rate": 0.00012305627809173448, "loss": 0.2107, "step": 7993 }, { "epoch": 3.7618823529411767, "grad_norm": 0.25381584341128594, "learning_rate": 0.00012301315605337435, "loss": 0.1396, "step": 7994 }, { "epoch": 3.7623529411764705, "grad_norm": 0.42368915425034226, "learning_rate": 0.00012297003631982304, "loss": 0.2154, "step": 7995 }, { "epoch": 3.7628235294117647, "grad_norm": 0.547563906465951, "learning_rate": 0.0001229269188947632, "loss": 0.3686, "step": 7996 }, { "epoch": 3.763294117647059, "grad_norm": 0.5078902484921654, "learning_rate": 0.0001228838037818773, "loss": 0.2637, "step": 7997 }, { "epoch": 3.7637647058823527, "grad_norm": 0.47677662535824566, "learning_rate": 0.00012284069098484758, "loss": 0.3387, "step": 7998 }, { "epoch": 3.764235294117647, "grad_norm": 0.3279558133995702, "learning_rate": 0.000122797580507356, "loss": 0.3322, "step": 7999 }, { "epoch": 3.764705882352941, "grad_norm": 0.43678535098354987, "learning_rate": 0.00012275447235308453, "loss": 0.2727, "step": 8000 }, { "epoch": 3.7651764705882353, "grad_norm": 0.45321119476356536, "learning_rate": 0.00012271136652571474, "loss": 0.2446, "step": 8001 }, { "epoch": 3.7656470588235296, "grad_norm": 0.1437006448481283, "learning_rate": 0.0001226682630289281, "loss": 0.0381, "step": 8002 }, { "epoch": 3.766117647058824, "grad_norm": 0.38221816785345974, "learning_rate": 0.00012262516186640587, "loss": 0.27, "step": 8003 }, { "epoch": 3.7665882352941176, "grad_norm": 0.36739604557955075, "learning_rate": 0.0001225820630418291, "loss": 0.2335, "step": 8004 }, { "epoch": 3.767058823529412, "grad_norm": 0.2653147643000375, "learning_rate": 0.00012253896655887861, "loss": 0.1257, "step": 8005 }, { "epoch": 3.767529411764706, "grad_norm": 0.3574014168394384, "learning_rate": 0.00012249587242123505, "loss": 0.1778, "step": 8006 }, { "epoch": 3.768, "grad_norm": 0.44196686788609135, "learning_rate": 0.00012245278063257892, "loss": 0.2271, "step": 8007 }, { "epoch": 3.768470588235294, "grad_norm": 0.18734814789973903, "learning_rate": 0.00012240969119659044, "loss": 0.0677, "step": 8008 }, { "epoch": 3.7689411764705882, "grad_norm": 0.3383070239586301, "learning_rate": 0.00012236660411694964, "loss": 0.2446, "step": 8009 }, { "epoch": 3.7694117647058825, "grad_norm": 0.23991202017504779, "learning_rate": 0.00012232351939733642, "loss": 0.0965, "step": 8010 }, { "epoch": 3.7698823529411767, "grad_norm": 0.540602121087919, "learning_rate": 0.0001222804370414304, "loss": 0.3339, "step": 8011 }, { "epoch": 3.7703529411764705, "grad_norm": 0.5192296089153604, "learning_rate": 0.00012223735705291098, "loss": 0.4631, "step": 8012 }, { "epoch": 3.7708235294117647, "grad_norm": 0.27197828940458035, "learning_rate": 0.00012219427943545744, "loss": 0.1663, "step": 8013 }, { "epoch": 3.771294117647059, "grad_norm": 0.35433430034400004, "learning_rate": 0.00012215120419274887, "loss": 0.1994, "step": 8014 }, { "epoch": 3.7717647058823527, "grad_norm": 0.32830223623019783, "learning_rate": 0.00012210813132846406, "loss": 0.2334, "step": 8015 }, { "epoch": 3.772235294117647, "grad_norm": 0.6913879006484053, "learning_rate": 0.0001220650608462816, "loss": 0.5345, "step": 8016 }, { "epoch": 3.772705882352941, "grad_norm": 0.3090740223077699, "learning_rate": 0.00012202199274987999, "loss": 0.2191, "step": 8017 }, { "epoch": 3.7731764705882354, "grad_norm": 0.23643024627013048, "learning_rate": 0.00012197892704293743, "loss": 0.0784, "step": 8018 }, { "epoch": 3.7736470588235296, "grad_norm": 0.2479653389387593, "learning_rate": 0.00012193586372913192, "loss": 0.1537, "step": 8019 }, { "epoch": 3.774117647058824, "grad_norm": 0.3010780252986841, "learning_rate": 0.00012189280281214126, "loss": 0.1319, "step": 8020 }, { "epoch": 3.7745882352941176, "grad_norm": 0.29769619611449394, "learning_rate": 0.00012184974429564318, "loss": 0.1979, "step": 8021 }, { "epoch": 3.775058823529412, "grad_norm": 0.18506103224378512, "learning_rate": 0.000121806688183315, "loss": 0.0514, "step": 8022 }, { "epoch": 3.775529411764706, "grad_norm": 0.4980942110184772, "learning_rate": 0.00012176363447883397, "loss": 0.4035, "step": 8023 }, { "epoch": 3.776, "grad_norm": 0.3572713354393498, "learning_rate": 0.00012172058318587698, "loss": 0.2947, "step": 8024 }, { "epoch": 3.776470588235294, "grad_norm": 0.44578102310660783, "learning_rate": 0.00012167753430812093, "loss": 0.3079, "step": 8025 }, { "epoch": 3.7769411764705882, "grad_norm": 0.31022765474025604, "learning_rate": 0.0001216344878492424, "loss": 0.2415, "step": 8026 }, { "epoch": 3.7774117647058825, "grad_norm": 0.4942956452904708, "learning_rate": 0.0001215914438129177, "loss": 0.2478, "step": 8027 }, { "epoch": 3.7778823529411767, "grad_norm": 0.33695140500150855, "learning_rate": 0.00012154840220282306, "loss": 0.2498, "step": 8028 }, { "epoch": 3.7783529411764705, "grad_norm": 0.23542215834568314, "learning_rate": 0.00012150536302263444, "loss": 0.1257, "step": 8029 }, { "epoch": 3.7788235294117647, "grad_norm": 0.24273387664775495, "learning_rate": 0.00012146232627602753, "loss": 0.0872, "step": 8030 }, { "epoch": 3.779294117647059, "grad_norm": 0.37379376936093034, "learning_rate": 0.00012141929196667798, "loss": 0.281, "step": 8031 }, { "epoch": 3.7797647058823527, "grad_norm": 0.30179236867663134, "learning_rate": 0.00012137626009826108, "loss": 0.26, "step": 8032 }, { "epoch": 3.780235294117647, "grad_norm": 0.3516072531027345, "learning_rate": 0.00012133323067445198, "loss": 0.2622, "step": 8033 }, { "epoch": 3.780705882352941, "grad_norm": 0.497697369917999, "learning_rate": 0.00012129020369892555, "loss": 0.4117, "step": 8034 }, { "epoch": 3.7811764705882354, "grad_norm": 0.3783369356812623, "learning_rate": 0.00012124717917535656, "loss": 0.2398, "step": 8035 }, { "epoch": 3.7816470588235296, "grad_norm": 0.47679277892220895, "learning_rate": 0.00012120415710741951, "loss": 0.3166, "step": 8036 }, { "epoch": 3.782117647058824, "grad_norm": 0.4882525931134095, "learning_rate": 0.00012116113749878867, "loss": 0.4658, "step": 8037 }, { "epoch": 3.7825882352941176, "grad_norm": 0.41180087411957966, "learning_rate": 0.00012111812035313814, "loss": 0.3055, "step": 8038 }, { "epoch": 3.783058823529412, "grad_norm": 0.29318791390662025, "learning_rate": 0.00012107510567414183, "loss": 0.1215, "step": 8039 }, { "epoch": 3.783529411764706, "grad_norm": 0.12992486128841182, "learning_rate": 0.00012103209346547332, "loss": 0.0391, "step": 8040 }, { "epoch": 3.784, "grad_norm": 0.4986590778724883, "learning_rate": 0.00012098908373080617, "loss": 0.2877, "step": 8041 }, { "epoch": 3.784470588235294, "grad_norm": 0.25153866452159385, "learning_rate": 0.00012094607647381355, "loss": 0.1283, "step": 8042 }, { "epoch": 3.7849411764705883, "grad_norm": 0.18493705074344438, "learning_rate": 0.00012090307169816847, "loss": 0.0823, "step": 8043 }, { "epoch": 3.7854117647058825, "grad_norm": 0.4122527392626898, "learning_rate": 0.00012086006940754382, "loss": 0.2616, "step": 8044 }, { "epoch": 3.7858823529411767, "grad_norm": 0.30209078144812923, "learning_rate": 0.00012081706960561217, "loss": 0.1193, "step": 8045 }, { "epoch": 3.7863529411764705, "grad_norm": 0.4101730118705069, "learning_rate": 0.00012077407229604593, "loss": 0.2597, "step": 8046 }, { "epoch": 3.7868235294117647, "grad_norm": 0.24694039930079692, "learning_rate": 0.00012073107748251724, "loss": 0.1219, "step": 8047 }, { "epoch": 3.787294117647059, "grad_norm": 0.39687500970179007, "learning_rate": 0.00012068808516869812, "loss": 0.2526, "step": 8048 }, { "epoch": 3.7877647058823527, "grad_norm": 0.42565978192899706, "learning_rate": 0.00012064509535826029, "loss": 0.3596, "step": 8049 }, { "epoch": 3.788235294117647, "grad_norm": 0.29295588163433006, "learning_rate": 0.00012060210805487529, "loss": 0.1035, "step": 8050 }, { "epoch": 3.788705882352941, "grad_norm": 0.42211604852347095, "learning_rate": 0.00012055912326221448, "loss": 0.3371, "step": 8051 }, { "epoch": 3.7891764705882354, "grad_norm": 0.24196819174948062, "learning_rate": 0.00012051614098394894, "loss": 0.1679, "step": 8052 }, { "epoch": 3.7896470588235296, "grad_norm": 0.39902836958062027, "learning_rate": 0.00012047316122374957, "loss": 0.2969, "step": 8053 }, { "epoch": 3.7901176470588234, "grad_norm": 0.355269843445846, "learning_rate": 0.00012043018398528707, "loss": 0.3166, "step": 8054 }, { "epoch": 3.7905882352941176, "grad_norm": 0.25763044691003734, "learning_rate": 0.0001203872092722319, "loss": 0.1138, "step": 8055 }, { "epoch": 3.791058823529412, "grad_norm": 0.8111224099940045, "learning_rate": 0.00012034423708825431, "loss": 0.5578, "step": 8056 }, { "epoch": 3.791529411764706, "grad_norm": 0.2515132444080284, "learning_rate": 0.00012030126743702431, "loss": 0.1135, "step": 8057 }, { "epoch": 3.792, "grad_norm": 0.3406030437000537, "learning_rate": 0.00012025830032221177, "loss": 0.129, "step": 8058 }, { "epoch": 3.792470588235294, "grad_norm": 0.5368315186074967, "learning_rate": 0.00012021533574748625, "loss": 0.3932, "step": 8059 }, { "epoch": 3.7929411764705883, "grad_norm": 0.36003132980863756, "learning_rate": 0.00012017237371651714, "loss": 0.2423, "step": 8060 }, { "epoch": 3.7934117647058825, "grad_norm": 0.71149288095615, "learning_rate": 0.00012012941423297363, "loss": 0.3663, "step": 8061 }, { "epoch": 3.7938823529411767, "grad_norm": 0.3857503313597318, "learning_rate": 0.00012008645730052468, "loss": 0.2548, "step": 8062 }, { "epoch": 3.7943529411764705, "grad_norm": 0.0982233696316718, "learning_rate": 0.00012004350292283896, "loss": 0.0297, "step": 8063 }, { "epoch": 3.7948235294117647, "grad_norm": 0.44350617517584306, "learning_rate": 0.00012000055110358502, "loss": 0.2769, "step": 8064 }, { "epoch": 3.795294117647059, "grad_norm": 0.49905754796450413, "learning_rate": 0.00011995760184643118, "loss": 0.3951, "step": 8065 }, { "epoch": 3.7957647058823527, "grad_norm": 0.37940088617491646, "learning_rate": 0.00011991465515504551, "loss": 0.2689, "step": 8066 }, { "epoch": 3.796235294117647, "grad_norm": 0.22319754498122252, "learning_rate": 0.00011987171103309586, "loss": 0.0889, "step": 8067 }, { "epoch": 3.796705882352941, "grad_norm": 0.3164823997015967, "learning_rate": 0.00011982876948424986, "loss": 0.2193, "step": 8068 }, { "epoch": 3.7971764705882354, "grad_norm": 0.5949471421398275, "learning_rate": 0.00011978583051217493, "loss": 0.3799, "step": 8069 }, { "epoch": 3.7976470588235296, "grad_norm": 0.4159421870266204, "learning_rate": 0.0001197428941205383, "loss": 0.2414, "step": 8070 }, { "epoch": 3.7981176470588234, "grad_norm": 0.40067751607167984, "learning_rate": 0.00011969996031300687, "loss": 0.3023, "step": 8071 }, { "epoch": 3.7985882352941176, "grad_norm": 0.5109817716247979, "learning_rate": 0.00011965702909324751, "loss": 0.3658, "step": 8072 }, { "epoch": 3.799058823529412, "grad_norm": 0.15039586361077573, "learning_rate": 0.00011961410046492667, "loss": 0.0826, "step": 8073 }, { "epoch": 3.799529411764706, "grad_norm": 0.23415348396527913, "learning_rate": 0.0001195711744317107, "loss": 0.1174, "step": 8074 }, { "epoch": 3.8, "grad_norm": 0.09355642521554534, "learning_rate": 0.00011952825099726567, "loss": 0.027, "step": 8075 }, { "epoch": 3.800470588235294, "grad_norm": 0.15322190077392378, "learning_rate": 0.00011948533016525749, "loss": 0.0848, "step": 8076 }, { "epoch": 3.8009411764705883, "grad_norm": 0.4269404110838942, "learning_rate": 0.00011944241193935178, "loss": 0.225, "step": 8077 }, { "epoch": 3.8014117647058825, "grad_norm": 0.3802677894316398, "learning_rate": 0.00011939949632321395, "loss": 0.1974, "step": 8078 }, { "epoch": 3.8018823529411767, "grad_norm": 0.1873423009532814, "learning_rate": 0.00011935658332050925, "loss": 0.071, "step": 8079 }, { "epoch": 3.8023529411764705, "grad_norm": 0.14073663898886557, "learning_rate": 0.00011931367293490265, "loss": 0.0449, "step": 8080 }, { "epoch": 3.8028235294117647, "grad_norm": 0.13814648918891603, "learning_rate": 0.00011927076517005885, "loss": 0.0499, "step": 8081 }, { "epoch": 3.803294117647059, "grad_norm": 0.2642626185239083, "learning_rate": 0.0001192278600296425, "loss": 0.1625, "step": 8082 }, { "epoch": 3.8037647058823527, "grad_norm": 0.07100179522741314, "learning_rate": 0.00011918495751731781, "loss": 0.0118, "step": 8083 }, { "epoch": 3.804235294117647, "grad_norm": 0.38350861755609017, "learning_rate": 0.0001191420576367489, "loss": 0.2569, "step": 8084 }, { "epoch": 3.804705882352941, "grad_norm": 0.5819656823874612, "learning_rate": 0.00011909916039159958, "loss": 0.2913, "step": 8085 }, { "epoch": 3.8051764705882354, "grad_norm": 0.522093610963858, "learning_rate": 0.00011905626578553359, "loss": 0.3253, "step": 8086 }, { "epoch": 3.8056470588235296, "grad_norm": 0.452236570617167, "learning_rate": 0.00011901337382221425, "loss": 0.1952, "step": 8087 }, { "epoch": 3.8061176470588234, "grad_norm": 0.5209008531846583, "learning_rate": 0.00011897048450530477, "loss": 0.4303, "step": 8088 }, { "epoch": 3.8065882352941176, "grad_norm": 0.21385218955445331, "learning_rate": 0.00011892759783846812, "loss": 0.0599, "step": 8089 }, { "epoch": 3.807058823529412, "grad_norm": 0.5574311632171738, "learning_rate": 0.00011888471382536704, "loss": 0.2686, "step": 8090 }, { "epoch": 3.8075294117647056, "grad_norm": 0.2732443249926049, "learning_rate": 0.00011884183246966399, "loss": 0.1418, "step": 8091 }, { "epoch": 3.808, "grad_norm": 0.4125799482283586, "learning_rate": 0.00011879895377502129, "loss": 0.3971, "step": 8092 }, { "epoch": 3.808470588235294, "grad_norm": 0.19360396398837082, "learning_rate": 0.00011875607774510099, "loss": 0.0612, "step": 8093 }, { "epoch": 3.8089411764705883, "grad_norm": 0.2031705646649333, "learning_rate": 0.00011871320438356491, "loss": 0.0773, "step": 8094 }, { "epoch": 3.8094117647058825, "grad_norm": 0.3472036821678149, "learning_rate": 0.0001186703336940746, "loss": 0.2358, "step": 8095 }, { "epoch": 3.8098823529411767, "grad_norm": 0.30036494828152455, "learning_rate": 0.00011862746568029149, "loss": 0.1345, "step": 8096 }, { "epoch": 3.8103529411764705, "grad_norm": 0.41432792410634023, "learning_rate": 0.00011858460034587669, "loss": 0.3105, "step": 8097 }, { "epoch": 3.8108235294117647, "grad_norm": 0.20021495745106876, "learning_rate": 0.0001185417376944911, "loss": 0.0725, "step": 8098 }, { "epoch": 3.811294117647059, "grad_norm": 0.37208630849444424, "learning_rate": 0.00011849887772979541, "loss": 0.3173, "step": 8099 }, { "epoch": 3.8117647058823527, "grad_norm": 0.2315133926904643, "learning_rate": 0.00011845602045545008, "loss": 0.1004, "step": 8100 }, { "epoch": 3.812235294117647, "grad_norm": 0.21449309629147445, "learning_rate": 0.00011841316587511531, "loss": 0.0908, "step": 8101 }, { "epoch": 3.812705882352941, "grad_norm": 0.25996240811057336, "learning_rate": 0.00011837031399245115, "loss": 0.1767, "step": 8102 }, { "epoch": 3.8131764705882354, "grad_norm": 0.20306055400431575, "learning_rate": 0.00011832746481111728, "loss": 0.0918, "step": 8103 }, { "epoch": 3.8136470588235296, "grad_norm": 0.37740199807081637, "learning_rate": 0.00011828461833477326, "loss": 0.1218, "step": 8104 }, { "epoch": 3.8141176470588234, "grad_norm": 0.13311879559841228, "learning_rate": 0.00011824177456707844, "loss": 0.0507, "step": 8105 }, { "epoch": 3.8145882352941176, "grad_norm": 0.40192113906905763, "learning_rate": 0.00011819893351169183, "loss": 0.2503, "step": 8106 }, { "epoch": 3.815058823529412, "grad_norm": 0.39234290060571053, "learning_rate": 0.00011815609517227228, "loss": 0.3, "step": 8107 }, { "epoch": 3.8155294117647056, "grad_norm": 0.3733416440644312, "learning_rate": 0.00011811325955247833, "loss": 0.3027, "step": 8108 }, { "epoch": 3.816, "grad_norm": 0.36166133163451675, "learning_rate": 0.00011807042665596846, "loss": 0.1494, "step": 8109 }, { "epoch": 3.816470588235294, "grad_norm": 0.3122086729397077, "learning_rate": 0.0001180275964864008, "loss": 0.1604, "step": 8110 }, { "epoch": 3.8169411764705883, "grad_norm": 0.334784830049783, "learning_rate": 0.00011798476904743321, "loss": 0.2401, "step": 8111 }, { "epoch": 3.8174117647058825, "grad_norm": 0.2744806403510385, "learning_rate": 0.00011794194434272336, "loss": 0.1128, "step": 8112 }, { "epoch": 3.8178823529411767, "grad_norm": 0.37649303014878827, "learning_rate": 0.00011789912237592874, "loss": 0.1711, "step": 8113 }, { "epoch": 3.8183529411764705, "grad_norm": 0.42420424820323055, "learning_rate": 0.00011785630315070651, "loss": 0.2425, "step": 8114 }, { "epoch": 3.8188235294117647, "grad_norm": 0.2709867438053639, "learning_rate": 0.00011781348667071365, "loss": 0.1288, "step": 8115 }, { "epoch": 3.819294117647059, "grad_norm": 0.3359296599085628, "learning_rate": 0.0001177706729396069, "loss": 0.2244, "step": 8116 }, { "epoch": 3.8197647058823527, "grad_norm": 0.5196173444053276, "learning_rate": 0.00011772786196104276, "loss": 0.3821, "step": 8117 }, { "epoch": 3.820235294117647, "grad_norm": 0.32675979779913533, "learning_rate": 0.00011768505373867753, "loss": 0.3174, "step": 8118 }, { "epoch": 3.820705882352941, "grad_norm": 0.8694296441385068, "learning_rate": 0.00011764224827616717, "loss": 0.3761, "step": 8119 }, { "epoch": 3.8211764705882354, "grad_norm": 0.42479449014872916, "learning_rate": 0.00011759944557716755, "loss": 0.2808, "step": 8120 }, { "epoch": 3.8216470588235296, "grad_norm": 0.5523809581374605, "learning_rate": 0.00011755664564533422, "loss": 0.3014, "step": 8121 }, { "epoch": 3.8221176470588234, "grad_norm": 0.4201146486780291, "learning_rate": 0.00011751384848432245, "loss": 0.3788, "step": 8122 }, { "epoch": 3.8225882352941176, "grad_norm": 0.34834233772339407, "learning_rate": 0.00011747105409778738, "loss": 0.1187, "step": 8123 }, { "epoch": 3.823058823529412, "grad_norm": 0.40743641642818856, "learning_rate": 0.00011742826248938384, "loss": 0.3222, "step": 8124 }, { "epoch": 3.8235294117647056, "grad_norm": 0.2716866072708952, "learning_rate": 0.00011738547366276645, "loss": 0.1489, "step": 8125 }, { "epoch": 3.824, "grad_norm": 0.3368356433942026, "learning_rate": 0.00011734268762158957, "loss": 0.254, "step": 8126 }, { "epoch": 3.824470588235294, "grad_norm": 0.3902252442067047, "learning_rate": 0.00011729990436950737, "loss": 0.2831, "step": 8127 }, { "epoch": 3.8249411764705883, "grad_norm": 0.35495223787153646, "learning_rate": 0.00011725712391017374, "loss": 0.3156, "step": 8128 }, { "epoch": 3.8254117647058825, "grad_norm": 0.40470260601766095, "learning_rate": 0.00011721434624724227, "loss": 0.2882, "step": 8129 }, { "epoch": 3.8258823529411767, "grad_norm": 0.1715549099281471, "learning_rate": 0.0001171715713843665, "loss": 0.0809, "step": 8130 }, { "epoch": 3.8263529411764705, "grad_norm": 0.2877706907420081, "learning_rate": 0.00011712879932519954, "loss": 0.2111, "step": 8131 }, { "epoch": 3.8268235294117647, "grad_norm": 0.26473525187339086, "learning_rate": 0.00011708603007339433, "loss": 0.0842, "step": 8132 }, { "epoch": 3.827294117647059, "grad_norm": 0.3675376094005924, "learning_rate": 0.00011704326363260365, "loss": 0.2074, "step": 8133 }, { "epoch": 3.8277647058823527, "grad_norm": 0.4213975475403511, "learning_rate": 0.00011700050000647986, "loss": 0.3018, "step": 8134 }, { "epoch": 3.828235294117647, "grad_norm": 0.7049779665833391, "learning_rate": 0.00011695773919867525, "loss": 0.5604, "step": 8135 }, { "epoch": 3.828705882352941, "grad_norm": 0.24351378772341642, "learning_rate": 0.00011691498121284176, "loss": 0.0997, "step": 8136 }, { "epoch": 3.8291764705882354, "grad_norm": 0.24778100709026374, "learning_rate": 0.00011687222605263118, "loss": 0.1183, "step": 8137 }, { "epoch": 3.8296470588235296, "grad_norm": 0.3226330094618365, "learning_rate": 0.000116829473721695, "loss": 0.1988, "step": 8138 }, { "epoch": 3.8301176470588234, "grad_norm": 0.3314905149162078, "learning_rate": 0.00011678672422368442, "loss": 0.2852, "step": 8139 }, { "epoch": 3.8305882352941176, "grad_norm": 0.4283186032498768, "learning_rate": 0.00011674397756225053, "loss": 0.3539, "step": 8140 }, { "epoch": 3.831058823529412, "grad_norm": 0.42322796779544397, "learning_rate": 0.00011670123374104409, "loss": 0.3352, "step": 8141 }, { "epoch": 3.8315294117647056, "grad_norm": 0.37967204264138393, "learning_rate": 0.00011665849276371557, "loss": 0.1836, "step": 8142 }, { "epoch": 3.832, "grad_norm": 0.4569139664018221, "learning_rate": 0.00011661575463391534, "loss": 0.3496, "step": 8143 }, { "epoch": 3.832470588235294, "grad_norm": 0.1766551506863542, "learning_rate": 0.0001165730193552934, "loss": 0.0442, "step": 8144 }, { "epoch": 3.8329411764705883, "grad_norm": 0.6126402356003592, "learning_rate": 0.00011653028693149961, "loss": 0.5179, "step": 8145 }, { "epoch": 3.8334117647058825, "grad_norm": 0.6022946193672198, "learning_rate": 0.00011648755736618342, "loss": 0.486, "step": 8146 }, { "epoch": 3.8338823529411767, "grad_norm": 0.44698019758360186, "learning_rate": 0.00011644483066299424, "loss": 0.3473, "step": 8147 }, { "epoch": 3.8343529411764705, "grad_norm": 0.3090032533343729, "learning_rate": 0.00011640210682558112, "loss": 0.1423, "step": 8148 }, { "epoch": 3.8348235294117647, "grad_norm": 0.3817326665748638, "learning_rate": 0.00011635938585759284, "loss": 0.3695, "step": 8149 }, { "epoch": 3.835294117647059, "grad_norm": 0.39876125210848296, "learning_rate": 0.00011631666776267803, "loss": 0.2768, "step": 8150 }, { "epoch": 3.8357647058823527, "grad_norm": 0.3763248596216116, "learning_rate": 0.00011627395254448504, "loss": 0.2145, "step": 8151 }, { "epoch": 3.836235294117647, "grad_norm": 0.37956901538741405, "learning_rate": 0.00011623124020666186, "loss": 0.2257, "step": 8152 }, { "epoch": 3.836705882352941, "grad_norm": 0.4988869698123706, "learning_rate": 0.00011618853075285644, "loss": 0.2593, "step": 8153 }, { "epoch": 3.8371764705882354, "grad_norm": 0.510377976723695, "learning_rate": 0.00011614582418671629, "loss": 0.4874, "step": 8154 }, { "epoch": 3.8376470588235296, "grad_norm": 0.3695922177753802, "learning_rate": 0.00011610312051188888, "loss": 0.1776, "step": 8155 }, { "epoch": 3.8381176470588234, "grad_norm": 0.3267704103929392, "learning_rate": 0.00011606041973202123, "loss": 0.2936, "step": 8156 }, { "epoch": 3.8385882352941176, "grad_norm": 0.3039377247740679, "learning_rate": 0.00011601772185076018, "loss": 0.2224, "step": 8157 }, { "epoch": 3.839058823529412, "grad_norm": 0.47801279382747286, "learning_rate": 0.00011597502687175238, "loss": 0.3945, "step": 8158 }, { "epoch": 3.8395294117647056, "grad_norm": 0.302212093558646, "learning_rate": 0.0001159323347986442, "loss": 0.1355, "step": 8159 }, { "epoch": 3.84, "grad_norm": 0.24761627816963272, "learning_rate": 0.00011588964563508167, "loss": 0.0597, "step": 8160 }, { "epoch": 3.840470588235294, "grad_norm": 0.31640580239469435, "learning_rate": 0.00011584695938471077, "loss": 0.1432, "step": 8161 }, { "epoch": 3.8409411764705883, "grad_norm": 0.31000533573964817, "learning_rate": 0.00011580427605117705, "loss": 0.1285, "step": 8162 }, { "epoch": 3.8414117647058825, "grad_norm": 0.3722766850920652, "learning_rate": 0.00011576159563812589, "loss": 0.1604, "step": 8163 }, { "epoch": 3.8418823529411767, "grad_norm": 0.25022837434273365, "learning_rate": 0.00011571891814920236, "loss": 0.1251, "step": 8164 }, { "epoch": 3.8423529411764705, "grad_norm": 0.45668836975516874, "learning_rate": 0.00011567624358805139, "loss": 0.3554, "step": 8165 }, { "epoch": 3.8428235294117647, "grad_norm": 0.5569395031668918, "learning_rate": 0.0001156335719583176, "loss": 0.3474, "step": 8166 }, { "epoch": 3.843294117647059, "grad_norm": 0.33108545798033934, "learning_rate": 0.00011559090326364527, "loss": 0.1214, "step": 8167 }, { "epoch": 3.8437647058823528, "grad_norm": 0.4021825316630115, "learning_rate": 0.00011554823750767863, "loss": 0.2621, "step": 8168 }, { "epoch": 3.844235294117647, "grad_norm": 0.46883116895159926, "learning_rate": 0.0001155055746940615, "loss": 0.4438, "step": 8169 }, { "epoch": 3.844705882352941, "grad_norm": 0.37714699676557234, "learning_rate": 0.00011546291482643744, "loss": 0.2331, "step": 8170 }, { "epoch": 3.8451764705882354, "grad_norm": 0.37558379239477646, "learning_rate": 0.00011542025790844988, "loss": 0.1992, "step": 8171 }, { "epoch": 3.8456470588235296, "grad_norm": 0.5202250334656243, "learning_rate": 0.00011537760394374192, "loss": 0.3847, "step": 8172 }, { "epoch": 3.8461176470588234, "grad_norm": 0.333880500663696, "learning_rate": 0.00011533495293595637, "loss": 0.2547, "step": 8173 }, { "epoch": 3.8465882352941176, "grad_norm": 0.2802674095249335, "learning_rate": 0.0001152923048887359, "loss": 0.1839, "step": 8174 }, { "epoch": 3.847058823529412, "grad_norm": 0.3023292456108049, "learning_rate": 0.00011524965980572284, "loss": 0.2403, "step": 8175 }, { "epoch": 3.8475294117647056, "grad_norm": 0.44691198377567387, "learning_rate": 0.00011520701769055927, "loss": 0.2765, "step": 8176 }, { "epoch": 3.848, "grad_norm": 0.1813752137174218, "learning_rate": 0.00011516437854688703, "loss": 0.0935, "step": 8177 }, { "epoch": 3.848470588235294, "grad_norm": 0.4308293041713348, "learning_rate": 0.00011512174237834776, "loss": 0.2997, "step": 8178 }, { "epoch": 3.8489411764705883, "grad_norm": 0.29049031415090526, "learning_rate": 0.00011507910918858276, "loss": 0.1827, "step": 8179 }, { "epoch": 3.8494117647058825, "grad_norm": 0.2835630651793703, "learning_rate": 0.0001150364789812331, "loss": 0.1786, "step": 8180 }, { "epoch": 3.8498823529411763, "grad_norm": 0.22308255865315813, "learning_rate": 0.00011499385175993965, "loss": 0.0901, "step": 8181 }, { "epoch": 3.8503529411764705, "grad_norm": 0.45385519391668283, "learning_rate": 0.00011495122752834296, "loss": 0.382, "step": 8182 }, { "epoch": 3.8508235294117648, "grad_norm": 0.13363943504718367, "learning_rate": 0.00011490860629008333, "loss": 0.0263, "step": 8183 }, { "epoch": 3.851294117647059, "grad_norm": 0.3900343076128436, "learning_rate": 0.00011486598804880088, "loss": 0.2686, "step": 8184 }, { "epoch": 3.8517647058823528, "grad_norm": 0.3904227179581711, "learning_rate": 0.00011482337280813536, "loss": 0.2593, "step": 8185 }, { "epoch": 3.852235294117647, "grad_norm": 0.2877380907251103, "learning_rate": 0.00011478076057172638, "loss": 0.1689, "step": 8186 }, { "epoch": 3.852705882352941, "grad_norm": 0.3457121115836732, "learning_rate": 0.00011473815134321315, "loss": 0.1258, "step": 8187 }, { "epoch": 3.8531764705882354, "grad_norm": 0.5685967562354717, "learning_rate": 0.00011469554512623477, "loss": 0.4267, "step": 8188 }, { "epoch": 3.8536470588235296, "grad_norm": 0.502890154215127, "learning_rate": 0.00011465294192443001, "loss": 0.3286, "step": 8189 }, { "epoch": 3.8541176470588234, "grad_norm": 0.6403875413038675, "learning_rate": 0.00011461034174143736, "loss": 0.411, "step": 8190 }, { "epoch": 3.8545882352941176, "grad_norm": 0.23931026053650217, "learning_rate": 0.00011456774458089515, "loss": 0.0925, "step": 8191 }, { "epoch": 3.855058823529412, "grad_norm": 0.38914121349469744, "learning_rate": 0.00011452515044644132, "loss": 0.2048, "step": 8192 }, { "epoch": 3.8555294117647056, "grad_norm": 0.5569405072349384, "learning_rate": 0.00011448255934171363, "loss": 0.4224, "step": 8193 }, { "epoch": 3.856, "grad_norm": 0.1541997897288286, "learning_rate": 0.00011443997127034962, "loss": 0.028, "step": 8194 }, { "epoch": 3.856470588235294, "grad_norm": 0.7834471918359387, "learning_rate": 0.0001143973862359865, "loss": 0.3662, "step": 8195 }, { "epoch": 3.8569411764705883, "grad_norm": 0.45547573302844596, "learning_rate": 0.0001143548042422612, "loss": 0.3105, "step": 8196 }, { "epoch": 3.8574117647058825, "grad_norm": 0.334694515426817, "learning_rate": 0.0001143122252928104, "loss": 0.2601, "step": 8197 }, { "epoch": 3.8578823529411763, "grad_norm": 0.4357119437764841, "learning_rate": 0.00011426964939127063, "loss": 0.2238, "step": 8198 }, { "epoch": 3.8583529411764705, "grad_norm": 0.4428880699367713, "learning_rate": 0.00011422707654127809, "loss": 0.421, "step": 8199 }, { "epoch": 3.8588235294117648, "grad_norm": 0.2949422959808317, "learning_rate": 0.00011418450674646868, "loss": 0.1158, "step": 8200 }, { "epoch": 3.859294117647059, "grad_norm": 0.36776087707228594, "learning_rate": 0.00011414194001047806, "loss": 0.2068, "step": 8201 }, { "epoch": 3.8597647058823528, "grad_norm": 0.3735148762197487, "learning_rate": 0.00011409937633694165, "loss": 0.192, "step": 8202 }, { "epoch": 3.860235294117647, "grad_norm": 0.4426071220400605, "learning_rate": 0.0001140568157294946, "loss": 0.1945, "step": 8203 }, { "epoch": 3.860705882352941, "grad_norm": 0.48717188392595806, "learning_rate": 0.00011401425819177179, "loss": 0.4925, "step": 8204 }, { "epoch": 3.8611764705882354, "grad_norm": 0.3498623702322267, "learning_rate": 0.00011397170372740782, "loss": 0.2065, "step": 8205 }, { "epoch": 3.8616470588235297, "grad_norm": 0.5356606165344152, "learning_rate": 0.0001139291523400371, "loss": 0.2849, "step": 8206 }, { "epoch": 3.8621176470588234, "grad_norm": 0.24958607120613346, "learning_rate": 0.00011388660403329368, "loss": 0.1496, "step": 8207 }, { "epoch": 3.8625882352941177, "grad_norm": 0.6115273747803396, "learning_rate": 0.0001138440588108114, "loss": 0.4298, "step": 8208 }, { "epoch": 3.863058823529412, "grad_norm": 0.4487122515505186, "learning_rate": 0.00011380151667622384, "loss": 0.327, "step": 8209 }, { "epoch": 3.8635294117647057, "grad_norm": 0.39153196867378315, "learning_rate": 0.00011375897763316433, "loss": 0.113, "step": 8210 }, { "epoch": 3.864, "grad_norm": 0.3405982133994297, "learning_rate": 0.00011371644168526586, "loss": 0.2953, "step": 8211 }, { "epoch": 3.864470588235294, "grad_norm": 0.3785960460754901, "learning_rate": 0.00011367390883616126, "loss": 0.171, "step": 8212 }, { "epoch": 3.8649411764705883, "grad_norm": 0.5923619108940807, "learning_rate": 0.00011363137908948303, "loss": 0.2659, "step": 8213 }, { "epoch": 3.8654117647058825, "grad_norm": 0.2548332322186804, "learning_rate": 0.00011358885244886342, "loss": 0.1345, "step": 8214 }, { "epoch": 3.8658823529411763, "grad_norm": 0.37519932219326035, "learning_rate": 0.00011354632891793437, "loss": 0.2413, "step": 8215 }, { "epoch": 3.8663529411764705, "grad_norm": 0.23568603104647273, "learning_rate": 0.00011350380850032766, "loss": 0.1075, "step": 8216 }, { "epoch": 3.8668235294117648, "grad_norm": 0.37847411631954947, "learning_rate": 0.00011346129119967473, "loss": 0.2345, "step": 8217 }, { "epoch": 3.8672941176470585, "grad_norm": 0.20059450843657614, "learning_rate": 0.0001134187770196067, "loss": 0.0786, "step": 8218 }, { "epoch": 3.8677647058823528, "grad_norm": 0.28702862444980043, "learning_rate": 0.00011337626596375458, "loss": 0.1279, "step": 8219 }, { "epoch": 3.868235294117647, "grad_norm": 0.22183374895659133, "learning_rate": 0.000113333758035749, "loss": 0.0806, "step": 8220 }, { "epoch": 3.868705882352941, "grad_norm": 0.31263865022871895, "learning_rate": 0.00011329125323922031, "loss": 0.2306, "step": 8221 }, { "epoch": 3.8691764705882354, "grad_norm": 0.21020211026900423, "learning_rate": 0.00011324875157779865, "loss": 0.0742, "step": 8222 }, { "epoch": 3.8696470588235297, "grad_norm": 0.3418312137491032, "learning_rate": 0.00011320625305511388, "loss": 0.157, "step": 8223 }, { "epoch": 3.8701176470588234, "grad_norm": 0.813566298577104, "learning_rate": 0.00011316375767479554, "loss": 0.3968, "step": 8224 }, { "epoch": 3.8705882352941177, "grad_norm": 0.21264012645678254, "learning_rate": 0.00011312126544047302, "loss": 0.0838, "step": 8225 }, { "epoch": 3.871058823529412, "grad_norm": 0.47560924890997874, "learning_rate": 0.0001130787763557753, "loss": 0.1911, "step": 8226 }, { "epoch": 3.8715294117647057, "grad_norm": 0.2916739205096126, "learning_rate": 0.00011303629042433118, "loss": 0.1962, "step": 8227 }, { "epoch": 3.872, "grad_norm": 0.3042981518604644, "learning_rate": 0.00011299380764976915, "loss": 0.1331, "step": 8228 }, { "epoch": 3.872470588235294, "grad_norm": 0.4799127037974841, "learning_rate": 0.00011295132803571747, "loss": 0.3156, "step": 8229 }, { "epoch": 3.8729411764705883, "grad_norm": 0.27477883487566324, "learning_rate": 0.00011290885158580411, "loss": 0.1922, "step": 8230 }, { "epoch": 3.8734117647058826, "grad_norm": 0.11042867163057564, "learning_rate": 0.00011286637830365672, "loss": 0.0353, "step": 8231 }, { "epoch": 3.8738823529411763, "grad_norm": 0.1798534679169742, "learning_rate": 0.0001128239081929028, "loss": 0.0741, "step": 8232 }, { "epoch": 3.8743529411764706, "grad_norm": 0.4677876510135216, "learning_rate": 0.00011278144125716946, "loss": 0.2987, "step": 8233 }, { "epoch": 3.8748235294117648, "grad_norm": 0.14474792365586595, "learning_rate": 0.00011273897750008358, "loss": 0.0455, "step": 8234 }, { "epoch": 3.8752941176470586, "grad_norm": 0.36829833435821774, "learning_rate": 0.0001126965169252718, "loss": 0.2396, "step": 8235 }, { "epoch": 3.875764705882353, "grad_norm": 0.4936881114284402, "learning_rate": 0.00011265405953636044, "loss": 0.3218, "step": 8236 }, { "epoch": 3.876235294117647, "grad_norm": 0.29337576116823233, "learning_rate": 0.0001126116053369756, "loss": 0.1483, "step": 8237 }, { "epoch": 3.8767058823529412, "grad_norm": 0.4403688852048514, "learning_rate": 0.000112569154330743, "loss": 0.2554, "step": 8238 }, { "epoch": 3.8771764705882354, "grad_norm": 0.5840226152541216, "learning_rate": 0.00011252670652128825, "loss": 0.5216, "step": 8239 }, { "epoch": 3.8776470588235297, "grad_norm": 0.3757507957874842, "learning_rate": 0.00011248426191223658, "loss": 0.2491, "step": 8240 }, { "epoch": 3.8781176470588234, "grad_norm": 0.45918210772975054, "learning_rate": 0.00011244182050721293, "loss": 0.2638, "step": 8241 }, { "epoch": 3.8785882352941177, "grad_norm": 0.3538860721216733, "learning_rate": 0.00011239938230984198, "loss": 0.1954, "step": 8242 }, { "epoch": 3.879058823529412, "grad_norm": 0.5725442290426556, "learning_rate": 0.00011235694732374829, "loss": 0.3776, "step": 8243 }, { "epoch": 3.8795294117647057, "grad_norm": 0.40937179513399624, "learning_rate": 0.00011231451555255592, "loss": 0.2109, "step": 8244 }, { "epoch": 3.88, "grad_norm": 0.40220703452013135, "learning_rate": 0.00011227208699988879, "loss": 0.2288, "step": 8245 }, { "epoch": 3.880470588235294, "grad_norm": 0.29998740852023814, "learning_rate": 0.00011222966166937044, "loss": 0.2353, "step": 8246 }, { "epoch": 3.8809411764705883, "grad_norm": 0.3044624403516303, "learning_rate": 0.00011218723956462427, "loss": 0.1202, "step": 8247 }, { "epoch": 3.8814117647058826, "grad_norm": 0.32378925570495554, "learning_rate": 0.00011214482068927332, "loss": 0.2558, "step": 8248 }, { "epoch": 3.8818823529411763, "grad_norm": 0.2869461217464782, "learning_rate": 0.00011210240504694033, "loss": 0.2367, "step": 8249 }, { "epoch": 3.8823529411764706, "grad_norm": 0.4486783620608769, "learning_rate": 0.00011205999264124786, "loss": 0.3458, "step": 8250 }, { "epoch": 3.882823529411765, "grad_norm": 0.19128692245548345, "learning_rate": 0.00011201758347581814, "loss": 0.0639, "step": 8251 }, { "epoch": 3.8832941176470586, "grad_norm": 0.4181311957518571, "learning_rate": 0.00011197517755427303, "loss": 0.326, "step": 8252 }, { "epoch": 3.883764705882353, "grad_norm": 0.4398859434087353, "learning_rate": 0.00011193277488023432, "loss": 0.3223, "step": 8253 }, { "epoch": 3.884235294117647, "grad_norm": 0.40507399612202616, "learning_rate": 0.00011189037545732335, "loss": 0.1718, "step": 8254 }, { "epoch": 3.8847058823529412, "grad_norm": 0.6297605696894902, "learning_rate": 0.00011184797928916124, "loss": 0.4892, "step": 8255 }, { "epoch": 3.8851764705882355, "grad_norm": 0.452160091169685, "learning_rate": 0.0001118055863793688, "loss": 0.2394, "step": 8256 }, { "epoch": 3.8856470588235297, "grad_norm": 0.3640265799259328, "learning_rate": 0.00011176319673156666, "loss": 0.2196, "step": 8257 }, { "epoch": 3.8861176470588235, "grad_norm": 0.32386113979391795, "learning_rate": 0.00011172081034937507, "loss": 0.1312, "step": 8258 }, { "epoch": 3.8865882352941177, "grad_norm": 0.3767730268683949, "learning_rate": 0.00011167842723641398, "loss": 0.2398, "step": 8259 }, { "epoch": 3.887058823529412, "grad_norm": 0.33512427739146883, "learning_rate": 0.00011163604739630324, "loss": 0.0764, "step": 8260 }, { "epoch": 3.8875294117647057, "grad_norm": 0.3881068016478923, "learning_rate": 0.00011159367083266219, "loss": 0.2978, "step": 8261 }, { "epoch": 3.888, "grad_norm": 0.19536442125124256, "learning_rate": 0.00011155129754911002, "loss": 0.1024, "step": 8262 }, { "epoch": 3.888470588235294, "grad_norm": 0.2973945784689125, "learning_rate": 0.00011150892754926568, "loss": 0.2108, "step": 8263 }, { "epoch": 3.8889411764705883, "grad_norm": 0.4035385865945946, "learning_rate": 0.00011146656083674768, "loss": 0.2658, "step": 8264 }, { "epoch": 3.8894117647058826, "grad_norm": 0.14737018574881006, "learning_rate": 0.00011142419741517443, "loss": 0.0587, "step": 8265 }, { "epoch": 3.8898823529411763, "grad_norm": 0.6441073714987416, "learning_rate": 0.00011138183728816389, "loss": 0.525, "step": 8266 }, { "epoch": 3.8903529411764706, "grad_norm": 0.2939440753656234, "learning_rate": 0.0001113394804593339, "loss": 0.1942, "step": 8267 }, { "epoch": 3.890823529411765, "grad_norm": 0.3296957148622878, "learning_rate": 0.00011129712693230191, "loss": 0.2385, "step": 8268 }, { "epoch": 3.8912941176470586, "grad_norm": 0.23244221167828436, "learning_rate": 0.00011125477671068506, "loss": 0.0977, "step": 8269 }, { "epoch": 3.891764705882353, "grad_norm": 0.27959665299529224, "learning_rate": 0.0001112124297981004, "loss": 0.1314, "step": 8270 }, { "epoch": 3.892235294117647, "grad_norm": 0.1753277300735991, "learning_rate": 0.00011117008619816445, "loss": 0.0603, "step": 8271 }, { "epoch": 3.8927058823529412, "grad_norm": 0.4278988016799633, "learning_rate": 0.00011112774591449356, "loss": 0.2767, "step": 8272 }, { "epoch": 3.8931764705882355, "grad_norm": 0.33256923827566465, "learning_rate": 0.00011108540895070388, "loss": 0.092, "step": 8273 }, { "epoch": 3.8936470588235297, "grad_norm": 0.2035705936621554, "learning_rate": 0.00011104307531041117, "loss": 0.0733, "step": 8274 }, { "epoch": 3.8941176470588235, "grad_norm": 0.3722197616584057, "learning_rate": 0.0001110007449972309, "loss": 0.1435, "step": 8275 }, { "epoch": 3.8945882352941177, "grad_norm": 0.7009994971043892, "learning_rate": 0.00011095841801477826, "loss": 0.3672, "step": 8276 }, { "epoch": 3.895058823529412, "grad_norm": 0.3264816625696727, "learning_rate": 0.00011091609436666825, "loss": 0.1484, "step": 8277 }, { "epoch": 3.8955294117647057, "grad_norm": 0.37053401305764067, "learning_rate": 0.00011087377405651549, "loss": 0.134, "step": 8278 }, { "epoch": 3.896, "grad_norm": 0.5430539368752805, "learning_rate": 0.00011083145708793431, "loss": 0.4078, "step": 8279 }, { "epoch": 3.896470588235294, "grad_norm": 0.4690897077887806, "learning_rate": 0.00011078914346453883, "loss": 0.2777, "step": 8280 }, { "epoch": 3.8969411764705884, "grad_norm": 0.12470147346179991, "learning_rate": 0.00011074683318994286, "loss": 0.0219, "step": 8281 }, { "epoch": 3.8974117647058826, "grad_norm": 1.0803503309190698, "learning_rate": 0.00011070452626775981, "loss": 0.5904, "step": 8282 }, { "epoch": 3.8978823529411764, "grad_norm": 0.7310057805539651, "learning_rate": 0.000110662222701603, "loss": 0.3276, "step": 8283 }, { "epoch": 3.8983529411764706, "grad_norm": 0.2829551570941015, "learning_rate": 0.00011061992249508533, "loss": 0.1602, "step": 8284 }, { "epoch": 3.898823529411765, "grad_norm": 0.3588504382947893, "learning_rate": 0.0001105776256518194, "loss": 0.1903, "step": 8285 }, { "epoch": 3.8992941176470586, "grad_norm": 0.30713104198035057, "learning_rate": 0.0001105353321754176, "loss": 0.1635, "step": 8286 }, { "epoch": 3.899764705882353, "grad_norm": 0.532625743573966, "learning_rate": 0.000110493042069492, "loss": 0.3523, "step": 8287 }, { "epoch": 3.900235294117647, "grad_norm": 0.3655201942046458, "learning_rate": 0.00011045075533765444, "loss": 0.1344, "step": 8288 }, { "epoch": 3.9007058823529412, "grad_norm": 0.4695946994102809, "learning_rate": 0.00011040847198351635, "loss": 0.1992, "step": 8289 }, { "epoch": 3.9011764705882355, "grad_norm": 0.5243464396557989, "learning_rate": 0.00011036619201068892, "loss": 0.3902, "step": 8290 }, { "epoch": 3.9016470588235292, "grad_norm": 0.5044425546162219, "learning_rate": 0.00011032391542278313, "loss": 0.3005, "step": 8291 }, { "epoch": 3.9021176470588235, "grad_norm": 0.5058222950395512, "learning_rate": 0.00011028164222340957, "loss": 0.5309, "step": 8292 }, { "epoch": 3.9025882352941177, "grad_norm": 0.20187225862189837, "learning_rate": 0.00011023937241617853, "loss": 0.1022, "step": 8293 }, { "epoch": 3.903058823529412, "grad_norm": 0.3211429221708891, "learning_rate": 0.00011019710600470016, "loss": 0.2196, "step": 8294 }, { "epoch": 3.9035294117647057, "grad_norm": 0.3656270885440691, "learning_rate": 0.00011015484299258417, "loss": 0.2591, "step": 8295 }, { "epoch": 3.904, "grad_norm": 0.2905457086842631, "learning_rate": 0.00011011258338344002, "loss": 0.1121, "step": 8296 }, { "epoch": 3.904470588235294, "grad_norm": 0.39986542504702344, "learning_rate": 0.00011007032718087688, "loss": 0.2813, "step": 8297 }, { "epoch": 3.9049411764705884, "grad_norm": 0.249379008640429, "learning_rate": 0.00011002807438850367, "loss": 0.0815, "step": 8298 }, { "epoch": 3.9054117647058826, "grad_norm": 0.20471567847369423, "learning_rate": 0.00010998582500992897, "loss": 0.059, "step": 8299 }, { "epoch": 3.9058823529411764, "grad_norm": 0.5511526080713337, "learning_rate": 0.00010994357904876106, "loss": 0.2356, "step": 8300 }, { "epoch": 3.9063529411764706, "grad_norm": 0.41479162646750284, "learning_rate": 0.000109901336508608, "loss": 0.1779, "step": 8301 }, { "epoch": 3.906823529411765, "grad_norm": 0.27101922371061904, "learning_rate": 0.00010985909739307748, "loss": 0.1254, "step": 8302 }, { "epoch": 3.9072941176470586, "grad_norm": 0.24308381977985535, "learning_rate": 0.0001098168617057769, "loss": 0.1211, "step": 8303 }, { "epoch": 3.907764705882353, "grad_norm": 0.37975012866779545, "learning_rate": 0.00010977462945031347, "loss": 0.2679, "step": 8304 }, { "epoch": 3.908235294117647, "grad_norm": 0.8726052605894334, "learning_rate": 0.00010973240063029398, "loss": 0.3293, "step": 8305 }, { "epoch": 3.9087058823529413, "grad_norm": 0.2563694703690478, "learning_rate": 0.000109690175249325, "loss": 0.1247, "step": 8306 }, { "epoch": 3.9091764705882355, "grad_norm": 0.29512049119861183, "learning_rate": 0.00010964795331101273, "loss": 0.1557, "step": 8307 }, { "epoch": 3.9096470588235293, "grad_norm": 0.3164303670647106, "learning_rate": 0.00010960573481896322, "loss": 0.1313, "step": 8308 }, { "epoch": 3.9101176470588235, "grad_norm": 0.23298412979616545, "learning_rate": 0.00010956351977678206, "loss": 0.0583, "step": 8309 }, { "epoch": 3.9105882352941177, "grad_norm": 0.36715416925148403, "learning_rate": 0.00010952130818807465, "loss": 0.269, "step": 8310 }, { "epoch": 3.911058823529412, "grad_norm": 0.37397926777333157, "learning_rate": 0.00010947910005644608, "loss": 0.1738, "step": 8311 }, { "epoch": 3.9115294117647057, "grad_norm": 0.2949953269299573, "learning_rate": 0.00010943689538550111, "loss": 0.1659, "step": 8312 }, { "epoch": 3.912, "grad_norm": 0.24431933284423923, "learning_rate": 0.00010939469417884424, "loss": 0.0708, "step": 8313 }, { "epoch": 3.912470588235294, "grad_norm": 0.4163957546345931, "learning_rate": 0.00010935249644007966, "loss": 0.2917, "step": 8314 }, { "epoch": 3.9129411764705884, "grad_norm": 0.3897720112673272, "learning_rate": 0.00010931030217281125, "loss": 0.2217, "step": 8315 }, { "epoch": 3.9134117647058826, "grad_norm": 0.16653782846179735, "learning_rate": 0.00010926811138064264, "loss": 0.0429, "step": 8316 }, { "epoch": 3.9138823529411764, "grad_norm": 0.30099588738074323, "learning_rate": 0.00010922592406717706, "loss": 0.1386, "step": 8317 }, { "epoch": 3.9143529411764706, "grad_norm": 0.3648062259341376, "learning_rate": 0.0001091837402360176, "loss": 0.2341, "step": 8318 }, { "epoch": 3.914823529411765, "grad_norm": 0.4297943934113239, "learning_rate": 0.00010914155989076692, "loss": 0.2971, "step": 8319 }, { "epoch": 3.9152941176470586, "grad_norm": 0.1343138852997656, "learning_rate": 0.00010909938303502741, "loss": 0.0625, "step": 8320 }, { "epoch": 3.915764705882353, "grad_norm": 0.2758213539458435, "learning_rate": 0.00010905720967240122, "loss": 0.0713, "step": 8321 }, { "epoch": 3.916235294117647, "grad_norm": 0.36202856948801043, "learning_rate": 0.00010901503980649017, "loss": 0.2792, "step": 8322 }, { "epoch": 3.9167058823529413, "grad_norm": 0.4399688967967281, "learning_rate": 0.0001089728734408957, "loss": 0.2182, "step": 8323 }, { "epoch": 3.9171764705882355, "grad_norm": 0.44879772403275475, "learning_rate": 0.00010893071057921911, "loss": 0.2302, "step": 8324 }, { "epoch": 3.9176470588235293, "grad_norm": 0.32699617186309643, "learning_rate": 0.0001088885512250613, "loss": 0.1574, "step": 8325 }, { "epoch": 3.9181176470588235, "grad_norm": 0.45940521637573706, "learning_rate": 0.00010884639538202285, "loss": 0.3085, "step": 8326 }, { "epoch": 3.9185882352941177, "grad_norm": 0.49852621093855576, "learning_rate": 0.00010880424305370406, "loss": 0.3633, "step": 8327 }, { "epoch": 3.919058823529412, "grad_norm": 0.14174390774469986, "learning_rate": 0.00010876209424370502, "loss": 0.0404, "step": 8328 }, { "epoch": 3.9195294117647057, "grad_norm": 0.3748547346006112, "learning_rate": 0.00010871994895562538, "loss": 0.1012, "step": 8329 }, { "epoch": 3.92, "grad_norm": 0.3725653991245885, "learning_rate": 0.00010867780719306458, "loss": 0.2729, "step": 8330 }, { "epoch": 3.920470588235294, "grad_norm": 0.4773513414584278, "learning_rate": 0.00010863566895962166, "loss": 0.3576, "step": 8331 }, { "epoch": 3.9209411764705884, "grad_norm": 0.1567140957331483, "learning_rate": 0.0001085935342588956, "loss": 0.063, "step": 8332 }, { "epoch": 3.9214117647058826, "grad_norm": 0.2925974927093288, "learning_rate": 0.00010855140309448479, "loss": 0.107, "step": 8333 }, { "epoch": 3.9218823529411764, "grad_norm": 0.4635832402429818, "learning_rate": 0.00010850927546998744, "loss": 0.3064, "step": 8334 }, { "epoch": 3.9223529411764706, "grad_norm": 0.34012779487386646, "learning_rate": 0.00010846715138900148, "loss": 0.2834, "step": 8335 }, { "epoch": 3.922823529411765, "grad_norm": 0.2764766899580957, "learning_rate": 0.00010842503085512453, "loss": 0.1242, "step": 8336 }, { "epoch": 3.9232941176470586, "grad_norm": 0.37484419373441996, "learning_rate": 0.00010838291387195387, "loss": 0.2462, "step": 8337 }, { "epoch": 3.923764705882353, "grad_norm": 0.27427599037917627, "learning_rate": 0.00010834080044308646, "loss": 0.1729, "step": 8338 }, { "epoch": 3.924235294117647, "grad_norm": 0.3017164983337372, "learning_rate": 0.00010829869057211906, "loss": 0.1839, "step": 8339 }, { "epoch": 3.9247058823529413, "grad_norm": 0.38877080853102014, "learning_rate": 0.00010825658426264803, "loss": 0.2875, "step": 8340 }, { "epoch": 3.9251764705882355, "grad_norm": 0.3291561527874878, "learning_rate": 0.00010821448151826941, "loss": 0.2083, "step": 8341 }, { "epoch": 3.9256470588235293, "grad_norm": 0.41814000697802056, "learning_rate": 0.00010817238234257907, "loss": 0.3122, "step": 8342 }, { "epoch": 3.9261176470588235, "grad_norm": 0.2817899146275099, "learning_rate": 0.00010813028673917245, "loss": 0.1994, "step": 8343 }, { "epoch": 3.9265882352941177, "grad_norm": 0.5305660222268206, "learning_rate": 0.0001080881947116447, "loss": 0.4928, "step": 8344 }, { "epoch": 3.9270588235294115, "grad_norm": 0.3671244233253564, "learning_rate": 0.00010804610626359068, "loss": 0.2661, "step": 8345 }, { "epoch": 3.9275294117647057, "grad_norm": 0.39632288758641054, "learning_rate": 0.000108004021398605, "loss": 0.2251, "step": 8346 }, { "epoch": 3.928, "grad_norm": 0.3487056426975573, "learning_rate": 0.00010796194012028185, "loss": 0.221, "step": 8347 }, { "epoch": 3.928470588235294, "grad_norm": 0.31902487608274227, "learning_rate": 0.00010791986243221521, "loss": 0.104, "step": 8348 }, { "epoch": 3.9289411764705884, "grad_norm": 0.3253435528489508, "learning_rate": 0.00010787778833799875, "loss": 0.1794, "step": 8349 }, { "epoch": 3.9294117647058826, "grad_norm": 0.5577907969689635, "learning_rate": 0.00010783571784122577, "loss": 0.2766, "step": 8350 }, { "epoch": 3.9298823529411764, "grad_norm": 0.3378739567089844, "learning_rate": 0.00010779365094548926, "loss": 0.2396, "step": 8351 }, { "epoch": 3.9303529411764706, "grad_norm": 0.2639336980404282, "learning_rate": 0.00010775158765438202, "loss": 0.1261, "step": 8352 }, { "epoch": 3.930823529411765, "grad_norm": 0.05258625349431785, "learning_rate": 0.0001077095279714964, "loss": 0.0075, "step": 8353 }, { "epoch": 3.9312941176470586, "grad_norm": 0.2783059561726339, "learning_rate": 0.00010766747190042452, "loss": 0.1192, "step": 8354 }, { "epoch": 3.931764705882353, "grad_norm": 0.21886490083328172, "learning_rate": 0.00010762541944475821, "loss": 0.1188, "step": 8355 }, { "epoch": 3.932235294117647, "grad_norm": 0.5782500292158834, "learning_rate": 0.0001075833706080889, "loss": 0.359, "step": 8356 }, { "epoch": 3.9327058823529413, "grad_norm": 0.39397011205532867, "learning_rate": 0.00010754132539400782, "loss": 0.2693, "step": 8357 }, { "epoch": 3.9331764705882355, "grad_norm": 0.5138835908468483, "learning_rate": 0.00010749928380610577, "loss": 0.3739, "step": 8358 }, { "epoch": 3.9336470588235293, "grad_norm": 0.47545530027423616, "learning_rate": 0.00010745724584797338, "loss": 0.3711, "step": 8359 }, { "epoch": 3.9341176470588235, "grad_norm": 0.3856104496347617, "learning_rate": 0.00010741521152320088, "loss": 0.1714, "step": 8360 }, { "epoch": 3.9345882352941177, "grad_norm": 0.2639533882830063, "learning_rate": 0.00010737318083537813, "loss": 0.0986, "step": 8361 }, { "epoch": 3.9350588235294115, "grad_norm": 0.28949943519502336, "learning_rate": 0.00010733115378809485, "loss": 0.1536, "step": 8362 }, { "epoch": 3.9355294117647057, "grad_norm": 0.2754525016380361, "learning_rate": 0.00010728913038494038, "loss": 0.1103, "step": 8363 }, { "epoch": 3.936, "grad_norm": 0.3521536331525601, "learning_rate": 0.00010724711062950358, "loss": 0.1209, "step": 8364 }, { "epoch": 3.936470588235294, "grad_norm": 0.4875532110583817, "learning_rate": 0.00010720509452537329, "loss": 0.3031, "step": 8365 }, { "epoch": 3.9369411764705884, "grad_norm": 0.3210715908390735, "learning_rate": 0.00010716308207613784, "loss": 0.1895, "step": 8366 }, { "epoch": 3.9374117647058826, "grad_norm": 0.5180600541218661, "learning_rate": 0.00010712107328538528, "loss": 0.4102, "step": 8367 }, { "epoch": 3.9378823529411764, "grad_norm": 0.5295106938546027, "learning_rate": 0.00010707906815670337, "loss": 0.3933, "step": 8368 }, { "epoch": 3.9383529411764706, "grad_norm": 0.27750580562239785, "learning_rate": 0.00010703706669367956, "loss": 0.1185, "step": 8369 }, { "epoch": 3.938823529411765, "grad_norm": 0.44764926916704273, "learning_rate": 0.00010699506889990102, "loss": 0.2286, "step": 8370 }, { "epoch": 3.9392941176470586, "grad_norm": 0.38793679248528623, "learning_rate": 0.00010695307477895449, "loss": 0.2781, "step": 8371 }, { "epoch": 3.939764705882353, "grad_norm": 0.4066587886950299, "learning_rate": 0.00010691108433442656, "loss": 0.3699, "step": 8372 }, { "epoch": 3.940235294117647, "grad_norm": 0.3326297794426668, "learning_rate": 0.00010686909756990337, "loss": 0.2614, "step": 8373 }, { "epoch": 3.9407058823529413, "grad_norm": 0.5911993090985354, "learning_rate": 0.00010682711448897077, "loss": 0.4377, "step": 8374 }, { "epoch": 3.9411764705882355, "grad_norm": 0.5159444389666843, "learning_rate": 0.00010678513509521435, "loss": 0.4895, "step": 8375 }, { "epoch": 3.9416470588235293, "grad_norm": 0.3388427907237416, "learning_rate": 0.00010674315939221937, "loss": 0.2063, "step": 8376 }, { "epoch": 3.9421176470588235, "grad_norm": 0.2394870277970043, "learning_rate": 0.00010670118738357078, "loss": 0.0778, "step": 8377 }, { "epoch": 3.9425882352941177, "grad_norm": 0.18531847377907903, "learning_rate": 0.00010665921907285317, "loss": 0.0536, "step": 8378 }, { "epoch": 3.9430588235294115, "grad_norm": 0.7119946491326887, "learning_rate": 0.0001066172544636508, "loss": 0.3913, "step": 8379 }, { "epoch": 3.9435294117647057, "grad_norm": 0.44261981769227915, "learning_rate": 0.00010657529355954771, "loss": 0.4224, "step": 8380 }, { "epoch": 3.944, "grad_norm": 0.396628255359557, "learning_rate": 0.00010653333636412758, "loss": 0.38, "step": 8381 }, { "epoch": 3.944470588235294, "grad_norm": 0.3177585271672112, "learning_rate": 0.00010649138288097366, "loss": 0.1228, "step": 8382 }, { "epoch": 3.9449411764705884, "grad_norm": 0.1615055006528415, "learning_rate": 0.00010644943311366911, "loss": 0.0659, "step": 8383 }, { "epoch": 3.9454117647058826, "grad_norm": 0.3011556296726576, "learning_rate": 0.00010640748706579658, "loss": 0.1463, "step": 8384 }, { "epoch": 3.9458823529411764, "grad_norm": 0.3514057048584211, "learning_rate": 0.00010636554474093846, "loss": 0.1836, "step": 8385 }, { "epoch": 3.9463529411764706, "grad_norm": 0.4807553226542312, "learning_rate": 0.00010632360614267683, "loss": 0.3705, "step": 8386 }, { "epoch": 3.946823529411765, "grad_norm": 0.33144910491841556, "learning_rate": 0.0001062816712745935, "loss": 0.1723, "step": 8387 }, { "epoch": 3.9472941176470586, "grad_norm": 0.24551328727338523, "learning_rate": 0.00010623974014026986, "loss": 0.079, "step": 8388 }, { "epoch": 3.947764705882353, "grad_norm": 0.29912501999635166, "learning_rate": 0.00010619781274328703, "loss": 0.143, "step": 8389 }, { "epoch": 3.948235294117647, "grad_norm": 0.2900698182694152, "learning_rate": 0.00010615588908722588, "loss": 0.1278, "step": 8390 }, { "epoch": 3.9487058823529413, "grad_norm": 0.33366088855506415, "learning_rate": 0.00010611396917566685, "loss": 0.1746, "step": 8391 }, { "epoch": 3.9491764705882355, "grad_norm": 0.22902052533755377, "learning_rate": 0.00010607205301219008, "loss": 0.067, "step": 8392 }, { "epoch": 3.9496470588235293, "grad_norm": 0.922598209107867, "learning_rate": 0.00010603014060037549, "loss": 0.5202, "step": 8393 }, { "epoch": 3.9501176470588235, "grad_norm": 0.4105063850466711, "learning_rate": 0.00010598823194380256, "loss": 0.2974, "step": 8394 }, { "epoch": 3.9505882352941177, "grad_norm": 0.7498605823810058, "learning_rate": 0.0001059463270460505, "loss": 0.5159, "step": 8395 }, { "epoch": 3.9510588235294115, "grad_norm": 0.3359918990517711, "learning_rate": 0.00010590442591069816, "loss": 0.29, "step": 8396 }, { "epoch": 3.9515294117647057, "grad_norm": 0.18710901942755, "learning_rate": 0.00010586252854132415, "loss": 0.0605, "step": 8397 }, { "epoch": 3.952, "grad_norm": 0.4700462286904066, "learning_rate": 0.0001058206349415067, "loss": 0.2509, "step": 8398 }, { "epoch": 3.952470588235294, "grad_norm": 0.2716602192184326, "learning_rate": 0.00010577874511482371, "loss": 0.1716, "step": 8399 }, { "epoch": 3.9529411764705884, "grad_norm": 0.31246783322450883, "learning_rate": 0.00010573685906485282, "loss": 0.1855, "step": 8400 }, { "epoch": 3.9534117647058826, "grad_norm": 0.5600448008422769, "learning_rate": 0.00010569497679517126, "loss": 0.2539, "step": 8401 }, { "epoch": 3.9538823529411764, "grad_norm": 0.37274727325934004, "learning_rate": 0.00010565309830935599, "loss": 0.2285, "step": 8402 }, { "epoch": 3.9543529411764706, "grad_norm": 0.4402001502049259, "learning_rate": 0.00010561122361098365, "loss": 0.1973, "step": 8403 }, { "epoch": 3.954823529411765, "grad_norm": 0.3102789609389747, "learning_rate": 0.00010556935270363055, "loss": 0.1046, "step": 8404 }, { "epoch": 3.9552941176470586, "grad_norm": 0.2504728346538282, "learning_rate": 0.00010552748559087261, "loss": 0.1102, "step": 8405 }, { "epoch": 3.955764705882353, "grad_norm": 0.38337693683372714, "learning_rate": 0.00010548562227628559, "loss": 0.2384, "step": 8406 }, { "epoch": 3.956235294117647, "grad_norm": 0.2769703280292645, "learning_rate": 0.00010544376276344477, "loss": 0.222, "step": 8407 }, { "epoch": 3.9567058823529413, "grad_norm": 0.37032009400146654, "learning_rate": 0.00010540190705592516, "loss": 0.1279, "step": 8408 }, { "epoch": 3.9571764705882355, "grad_norm": 0.445862712153612, "learning_rate": 0.00010536005515730141, "loss": 0.3258, "step": 8409 }, { "epoch": 3.9576470588235293, "grad_norm": 0.266570977233048, "learning_rate": 0.00010531820707114793, "loss": 0.1068, "step": 8410 }, { "epoch": 3.9581176470588235, "grad_norm": 0.25522656497123636, "learning_rate": 0.00010527636280103872, "loss": 0.088, "step": 8411 }, { "epoch": 3.9585882352941177, "grad_norm": 0.4376240415767509, "learning_rate": 0.00010523452235054746, "loss": 0.3156, "step": 8412 }, { "epoch": 3.9590588235294115, "grad_norm": 0.15991063873761, "learning_rate": 0.00010519268572324763, "loss": 0.047, "step": 8413 }, { "epoch": 3.9595294117647057, "grad_norm": 0.5188183804419528, "learning_rate": 0.00010515085292271219, "loss": 0.296, "step": 8414 }, { "epoch": 3.96, "grad_norm": 0.40533247923779103, "learning_rate": 0.00010510902395251387, "loss": 0.2348, "step": 8415 }, { "epoch": 3.960470588235294, "grad_norm": 0.5639395410870547, "learning_rate": 0.00010506719881622514, "loss": 0.5073, "step": 8416 }, { "epoch": 3.9609411764705884, "grad_norm": 0.43473038461140073, "learning_rate": 0.00010502537751741801, "loss": 0.2035, "step": 8417 }, { "epoch": 3.961411764705882, "grad_norm": 0.24043414626452492, "learning_rate": 0.00010498356005966422, "loss": 0.1401, "step": 8418 }, { "epoch": 3.9618823529411764, "grad_norm": 0.393124955445886, "learning_rate": 0.00010494174644653518, "loss": 0.2091, "step": 8419 }, { "epoch": 3.9623529411764706, "grad_norm": 0.38778704780393813, "learning_rate": 0.000104899936681602, "loss": 0.2628, "step": 8420 }, { "epoch": 3.962823529411765, "grad_norm": 0.4400624243060837, "learning_rate": 0.0001048581307684355, "loss": 0.2868, "step": 8421 }, { "epoch": 3.9632941176470586, "grad_norm": 0.37098707641657874, "learning_rate": 0.00010481632871060605, "loss": 0.1312, "step": 8422 }, { "epoch": 3.963764705882353, "grad_norm": 0.38516189961715336, "learning_rate": 0.0001047745305116837, "loss": 0.1368, "step": 8423 }, { "epoch": 3.964235294117647, "grad_norm": 0.393695017565702, "learning_rate": 0.00010473273617523831, "loss": 0.206, "step": 8424 }, { "epoch": 3.9647058823529413, "grad_norm": 0.2706553173708368, "learning_rate": 0.00010469094570483928, "loss": 0.0622, "step": 8425 }, { "epoch": 3.9651764705882355, "grad_norm": 0.25165620894482355, "learning_rate": 0.00010464915910405576, "loss": 0.0794, "step": 8426 }, { "epoch": 3.9656470588235293, "grad_norm": 0.46168820805365596, "learning_rate": 0.00010460737637645645, "loss": 0.3888, "step": 8427 }, { "epoch": 3.9661176470588235, "grad_norm": 0.2922380325550821, "learning_rate": 0.00010456559752560987, "loss": 0.1146, "step": 8428 }, { "epoch": 3.9665882352941177, "grad_norm": 0.27624116508293084, "learning_rate": 0.00010452382255508412, "loss": 0.2209, "step": 8429 }, { "epoch": 3.9670588235294115, "grad_norm": 0.23640566680823571, "learning_rate": 0.00010448205146844696, "loss": 0.0815, "step": 8430 }, { "epoch": 3.9675294117647057, "grad_norm": 0.3393241587034134, "learning_rate": 0.00010444028426926588, "loss": 0.2845, "step": 8431 }, { "epoch": 3.968, "grad_norm": 0.2115945404413875, "learning_rate": 0.000104398520961108, "loss": 0.0661, "step": 8432 }, { "epoch": 3.968470588235294, "grad_norm": 0.555449045453766, "learning_rate": 0.00010435676154754009, "loss": 0.2008, "step": 8433 }, { "epoch": 3.9689411764705884, "grad_norm": 0.24663795753936016, "learning_rate": 0.00010431500603212863, "loss": 0.1086, "step": 8434 }, { "epoch": 3.969411764705882, "grad_norm": 0.295197951603568, "learning_rate": 0.00010427325441843976, "loss": 0.1461, "step": 8435 }, { "epoch": 3.9698823529411764, "grad_norm": 0.47623558148697925, "learning_rate": 0.00010423150671003924, "loss": 0.3126, "step": 8436 }, { "epoch": 3.9703529411764706, "grad_norm": 0.3530080804131532, "learning_rate": 0.00010418976291049251, "loss": 0.2228, "step": 8437 }, { "epoch": 3.970823529411765, "grad_norm": 0.2673363218084547, "learning_rate": 0.00010414802302336474, "loss": 0.1505, "step": 8438 }, { "epoch": 3.9712941176470586, "grad_norm": 0.3163443713578462, "learning_rate": 0.00010410628705222073, "loss": 0.2352, "step": 8439 }, { "epoch": 3.971764705882353, "grad_norm": 0.258364174115461, "learning_rate": 0.00010406455500062487, "loss": 0.0949, "step": 8440 }, { "epoch": 3.972235294117647, "grad_norm": 0.34717366692890045, "learning_rate": 0.00010402282687214137, "loss": 0.1883, "step": 8441 }, { "epoch": 3.9727058823529413, "grad_norm": 0.3160091258032069, "learning_rate": 0.00010398110267033396, "loss": 0.1809, "step": 8442 }, { "epoch": 3.9731764705882355, "grad_norm": 0.304988279291341, "learning_rate": 0.00010393938239876608, "loss": 0.1463, "step": 8443 }, { "epoch": 3.9736470588235293, "grad_norm": 0.4192298663161631, "learning_rate": 0.0001038976660610009, "loss": 0.2303, "step": 8444 }, { "epoch": 3.9741176470588235, "grad_norm": 0.5019027031299537, "learning_rate": 0.00010385595366060116, "loss": 0.3386, "step": 8445 }, { "epoch": 3.9745882352941178, "grad_norm": 0.44033899376465935, "learning_rate": 0.0001038142452011293, "loss": 0.3092, "step": 8446 }, { "epoch": 3.9750588235294115, "grad_norm": 0.7220690666934421, "learning_rate": 0.00010377254068614743, "loss": 0.406, "step": 8447 }, { "epoch": 3.9755294117647058, "grad_norm": 0.8637856576056882, "learning_rate": 0.00010373084011921736, "loss": 0.4894, "step": 8448 }, { "epoch": 3.976, "grad_norm": 0.3570164141731024, "learning_rate": 0.00010368914350390048, "loss": 0.0979, "step": 8449 }, { "epoch": 3.976470588235294, "grad_norm": 0.3357042646515495, "learning_rate": 0.0001036474508437579, "loss": 0.2005, "step": 8450 }, { "epoch": 3.9769411764705884, "grad_norm": 0.39012019622719046, "learning_rate": 0.00010360576214235036, "loss": 0.2036, "step": 8451 }, { "epoch": 3.977411764705882, "grad_norm": 0.4131786853692831, "learning_rate": 0.00010356407740323832, "loss": 0.1763, "step": 8452 }, { "epoch": 3.9778823529411764, "grad_norm": 0.5866738511442469, "learning_rate": 0.00010352239662998182, "loss": 0.4982, "step": 8453 }, { "epoch": 3.9783529411764706, "grad_norm": 0.32529684295417866, "learning_rate": 0.00010348071982614063, "loss": 0.2265, "step": 8454 }, { "epoch": 3.978823529411765, "grad_norm": 0.31452986530296084, "learning_rate": 0.00010343904699527415, "loss": 0.1199, "step": 8455 }, { "epoch": 3.9792941176470586, "grad_norm": 0.0581967375884714, "learning_rate": 0.00010339737814094144, "loss": 0.0097, "step": 8456 }, { "epoch": 3.979764705882353, "grad_norm": 0.6109462830374615, "learning_rate": 0.00010335571326670117, "loss": 0.4367, "step": 8457 }, { "epoch": 3.980235294117647, "grad_norm": 0.38938853710705496, "learning_rate": 0.00010331405237611182, "loss": 0.2178, "step": 8458 }, { "epoch": 3.9807058823529413, "grad_norm": 0.29020737640684974, "learning_rate": 0.00010327239547273138, "loss": 0.1039, "step": 8459 }, { "epoch": 3.9811764705882355, "grad_norm": 0.48433575638167164, "learning_rate": 0.00010323074256011756, "loss": 0.3073, "step": 8460 }, { "epoch": 3.9816470588235293, "grad_norm": 0.23600888259411856, "learning_rate": 0.00010318909364182774, "loss": 0.1768, "step": 8461 }, { "epoch": 3.9821176470588235, "grad_norm": 0.1351091133834172, "learning_rate": 0.00010314744872141893, "loss": 0.0347, "step": 8462 }, { "epoch": 3.9825882352941178, "grad_norm": 0.3078843441217446, "learning_rate": 0.00010310580780244779, "loss": 0.2007, "step": 8463 }, { "epoch": 3.9830588235294115, "grad_norm": 0.23762572032458143, "learning_rate": 0.00010306417088847067, "loss": 0.0701, "step": 8464 }, { "epoch": 3.9835294117647058, "grad_norm": 0.34971467643228277, "learning_rate": 0.00010302253798304356, "loss": 0.2841, "step": 8465 }, { "epoch": 3.984, "grad_norm": 0.336285832704636, "learning_rate": 0.00010298090908972217, "loss": 0.1911, "step": 8466 }, { "epoch": 3.984470588235294, "grad_norm": 0.307363336009625, "learning_rate": 0.00010293928421206179, "loss": 0.1716, "step": 8467 }, { "epoch": 3.9849411764705884, "grad_norm": 0.45780471501230885, "learning_rate": 0.00010289766335361733, "loss": 0.4079, "step": 8468 }, { "epoch": 3.985411764705882, "grad_norm": 0.2791096539301848, "learning_rate": 0.00010285604651794347, "loss": 0.1487, "step": 8469 }, { "epoch": 3.9858823529411764, "grad_norm": 0.26046851980118846, "learning_rate": 0.00010281443370859449, "loss": 0.1686, "step": 8470 }, { "epoch": 3.9863529411764707, "grad_norm": 0.23701762131342777, "learning_rate": 0.00010277282492912429, "loss": 0.1078, "step": 8471 }, { "epoch": 3.9868235294117644, "grad_norm": 0.49028466028773304, "learning_rate": 0.0001027312201830865, "loss": 0.382, "step": 8472 }, { "epoch": 3.9872941176470587, "grad_norm": 0.3211201870315484, "learning_rate": 0.00010268961947403437, "loss": 0.1679, "step": 8473 }, { "epoch": 3.987764705882353, "grad_norm": 0.3583064536208049, "learning_rate": 0.00010264802280552077, "loss": 0.2313, "step": 8474 }, { "epoch": 3.988235294117647, "grad_norm": 0.24247157602382147, "learning_rate": 0.0001026064301810983, "loss": 0.0819, "step": 8475 }, { "epoch": 3.9887058823529413, "grad_norm": 0.14563439848858417, "learning_rate": 0.00010256484160431917, "loss": 0.0504, "step": 8476 }, { "epoch": 3.9891764705882355, "grad_norm": 0.3519318539701292, "learning_rate": 0.00010252325707873523, "loss": 0.245, "step": 8477 }, { "epoch": 3.9896470588235293, "grad_norm": 0.20213175072361259, "learning_rate": 0.00010248167660789797, "loss": 0.0911, "step": 8478 }, { "epoch": 3.9901176470588235, "grad_norm": 0.4011942655485969, "learning_rate": 0.00010244010019535864, "loss": 0.2296, "step": 8479 }, { "epoch": 3.9905882352941178, "grad_norm": 0.35130042287766455, "learning_rate": 0.00010239852784466803, "loss": 0.1901, "step": 8480 }, { "epoch": 3.9910588235294115, "grad_norm": 0.2007551747283113, "learning_rate": 0.00010235695955937661, "loss": 0.078, "step": 8481 }, { "epoch": 3.9915294117647058, "grad_norm": 0.24290620915552233, "learning_rate": 0.00010231539534303454, "loss": 0.1156, "step": 8482 }, { "epoch": 3.992, "grad_norm": 0.2902561125292638, "learning_rate": 0.00010227383519919161, "loss": 0.1588, "step": 8483 }, { "epoch": 3.992470588235294, "grad_norm": 0.3650731202954205, "learning_rate": 0.00010223227913139722, "loss": 0.1181, "step": 8484 }, { "epoch": 3.9929411764705884, "grad_norm": 0.3815802910634102, "learning_rate": 0.0001021907271432005, "loss": 0.2734, "step": 8485 }, { "epoch": 3.993411764705882, "grad_norm": 0.34792121258925673, "learning_rate": 0.00010214917923815022, "loss": 0.1836, "step": 8486 }, { "epoch": 3.9938823529411764, "grad_norm": 0.3184863268161506, "learning_rate": 0.00010210763541979473, "loss": 0.1436, "step": 8487 }, { "epoch": 3.9943529411764707, "grad_norm": 0.28718156249568494, "learning_rate": 0.00010206609569168204, "loss": 0.1455, "step": 8488 }, { "epoch": 3.9948235294117644, "grad_norm": 0.23045198463902059, "learning_rate": 0.00010202456005735993, "loss": 0.1242, "step": 8489 }, { "epoch": 3.9952941176470587, "grad_norm": 0.28376717638161086, "learning_rate": 0.00010198302852037573, "loss": 0.1032, "step": 8490 }, { "epoch": 3.995764705882353, "grad_norm": 0.5811973113441423, "learning_rate": 0.00010194150108427637, "loss": 0.5351, "step": 8491 }, { "epoch": 3.996235294117647, "grad_norm": 0.40748580736216927, "learning_rate": 0.00010189997775260857, "loss": 0.2826, "step": 8492 }, { "epoch": 3.9967058823529413, "grad_norm": 0.3659827514624549, "learning_rate": 0.00010185845852891858, "loss": 0.18, "step": 8493 }, { "epoch": 3.9971764705882356, "grad_norm": 0.23900483249803287, "learning_rate": 0.00010181694341675235, "loss": 0.0932, "step": 8494 }, { "epoch": 3.9976470588235293, "grad_norm": 0.16300769855970443, "learning_rate": 0.00010177543241965555, "loss": 0.0391, "step": 8495 }, { "epoch": 3.9981176470588236, "grad_norm": 0.4719769231334745, "learning_rate": 0.00010173392554117333, "loss": 0.3423, "step": 8496 }, { "epoch": 3.9985882352941178, "grad_norm": 0.4158980453166547, "learning_rate": 0.00010169242278485063, "loss": 0.2857, "step": 8497 }, { "epoch": 3.9990588235294116, "grad_norm": 0.42670375107280417, "learning_rate": 0.00010165092415423193, "loss": 0.2799, "step": 8498 }, { "epoch": 3.9995294117647058, "grad_norm": 0.31434157018698605, "learning_rate": 0.00010160942965286148, "loss": 0.1931, "step": 8499 }, { "epoch": 4.0, "grad_norm": 0.29673432956102175, "learning_rate": 0.0001015679392842831, "loss": 0.1235, "step": 8500 }, { "epoch": 4.0, "eval_loss": 0.2215253859758377, "eval_runtime": 65.1712, "eval_samples_per_second": 11.508, "eval_steps_per_second": 5.754, "step": 8500 }, { "epoch": 4.000470588235294, "grad_norm": 0.45086650092220376, "learning_rate": 0.00010152645305204022, "loss": 0.3073, "step": 8501 }, { "epoch": 4.0009411764705884, "grad_norm": 0.7528285064002601, "learning_rate": 0.00010148497095967606, "loss": 0.4204, "step": 8502 }, { "epoch": 4.001411764705883, "grad_norm": 2.781076137551428, "learning_rate": 0.00010144349301073334, "loss": 1.4167, "step": 8503 }, { "epoch": 4.001882352941177, "grad_norm": 0.3327278101693948, "learning_rate": 0.00010140201920875445, "loss": 0.2749, "step": 8504 }, { "epoch": 4.00235294117647, "grad_norm": 0.26111275165689873, "learning_rate": 0.00010136054955728154, "loss": 0.1216, "step": 8505 }, { "epoch": 4.0028235294117644, "grad_norm": 0.43415106085653393, "learning_rate": 0.00010131908405985625, "loss": 0.3405, "step": 8506 }, { "epoch": 4.003294117647059, "grad_norm": 0.2536934719349309, "learning_rate": 0.00010127762272001996, "loss": 0.1707, "step": 8507 }, { "epoch": 4.003764705882353, "grad_norm": 0.047057687408990065, "learning_rate": 0.00010123616554131363, "loss": 0.0087, "step": 8508 }, { "epoch": 4.004235294117647, "grad_norm": 0.413200409339565, "learning_rate": 0.00010119471252727796, "loss": 0.2671, "step": 8509 }, { "epoch": 4.004705882352941, "grad_norm": 0.24882848685763062, "learning_rate": 0.00010115326368145327, "loss": 0.1219, "step": 8510 }, { "epoch": 4.005176470588236, "grad_norm": 0.2503729479637105, "learning_rate": 0.00010111181900737945, "loss": 0.0649, "step": 8511 }, { "epoch": 4.00564705882353, "grad_norm": 0.32730423961202343, "learning_rate": 0.00010107037850859605, "loss": 0.195, "step": 8512 }, { "epoch": 4.006117647058823, "grad_norm": 0.4758917267206466, "learning_rate": 0.00010102894218864235, "loss": 0.2331, "step": 8513 }, { "epoch": 4.006588235294117, "grad_norm": 0.4061785036920495, "learning_rate": 0.00010098751005105719, "loss": 0.2708, "step": 8514 }, { "epoch": 4.007058823529412, "grad_norm": 0.2878724357407421, "learning_rate": 0.0001009460820993791, "loss": 0.1528, "step": 8515 }, { "epoch": 4.007529411764706, "grad_norm": 0.18734235881097547, "learning_rate": 0.00010090465833714616, "loss": 0.0641, "step": 8516 }, { "epoch": 4.008, "grad_norm": 0.20016448042454, "learning_rate": 0.00010086323876789621, "loss": 0.1072, "step": 8517 }, { "epoch": 4.008470588235294, "grad_norm": 0.400190254579706, "learning_rate": 0.00010082182339516674, "loss": 0.1842, "step": 8518 }, { "epoch": 4.0089411764705885, "grad_norm": 0.47421138676517266, "learning_rate": 0.00010078041222249469, "loss": 0.2898, "step": 8519 }, { "epoch": 4.009411764705883, "grad_norm": 0.1975498261852336, "learning_rate": 0.00010073900525341692, "loss": 0.0602, "step": 8520 }, { "epoch": 4.009882352941177, "grad_norm": 0.3837584211241528, "learning_rate": 0.00010069760249146972, "loss": 0.086, "step": 8521 }, { "epoch": 4.01035294117647, "grad_norm": 0.5944275121705817, "learning_rate": 0.00010065620394018904, "loss": 0.2759, "step": 8522 }, { "epoch": 4.0108235294117645, "grad_norm": 0.3978933562789847, "learning_rate": 0.00010061480960311061, "loss": 0.2402, "step": 8523 }, { "epoch": 4.011294117647059, "grad_norm": 0.5590819776087345, "learning_rate": 0.0001005734194837697, "loss": 0.2606, "step": 8524 }, { "epoch": 4.011764705882353, "grad_norm": 0.21627558292878749, "learning_rate": 0.00010053203358570115, "loss": 0.0876, "step": 8525 }, { "epoch": 4.012235294117647, "grad_norm": 0.5398417374680116, "learning_rate": 0.00010049065191243961, "loss": 0.3023, "step": 8526 }, { "epoch": 4.012705882352941, "grad_norm": 0.35110673215120014, "learning_rate": 0.00010044927446751923, "loss": 0.2788, "step": 8527 }, { "epoch": 4.013176470588236, "grad_norm": 0.526394249028231, "learning_rate": 0.00010040790125447386, "loss": 0.5194, "step": 8528 }, { "epoch": 4.01364705882353, "grad_norm": 0.3098150290707764, "learning_rate": 0.00010036653227683694, "loss": 0.1778, "step": 8529 }, { "epoch": 4.014117647058823, "grad_norm": 0.18319490718750098, "learning_rate": 0.00010032516753814166, "loss": 0.08, "step": 8530 }, { "epoch": 4.014588235294117, "grad_norm": 0.4336660575334198, "learning_rate": 0.0001002838070419207, "loss": 0.1946, "step": 8531 }, { "epoch": 4.015058823529412, "grad_norm": 0.5196326160686172, "learning_rate": 0.00010024245079170645, "loss": 0.3117, "step": 8532 }, { "epoch": 4.015529411764706, "grad_norm": 0.22940389491684873, "learning_rate": 0.000100201098791031, "loss": 0.0749, "step": 8533 }, { "epoch": 4.016, "grad_norm": 0.34057192600509517, "learning_rate": 0.00010015975104342598, "loss": 0.1622, "step": 8534 }, { "epoch": 4.016470588235294, "grad_norm": 0.5705023384983078, "learning_rate": 0.00010011840755242266, "loss": 0.3415, "step": 8535 }, { "epoch": 4.0169411764705885, "grad_norm": 0.4642747146258997, "learning_rate": 0.00010007706832155201, "loss": 0.2673, "step": 8536 }, { "epoch": 4.017411764705883, "grad_norm": 0.5789510838010812, "learning_rate": 0.00010003573335434462, "loss": 0.402, "step": 8537 }, { "epoch": 4.017882352941177, "grad_norm": 0.3267065114381866, "learning_rate": 9.999440265433065e-05, "loss": 0.214, "step": 8538 }, { "epoch": 4.01835294117647, "grad_norm": 0.36169028651430785, "learning_rate": 9.995307622503993e-05, "loss": 0.2675, "step": 8539 }, { "epoch": 4.0188235294117645, "grad_norm": 0.32440366063127773, "learning_rate": 9.991175407000206e-05, "loss": 0.1868, "step": 8540 }, { "epoch": 4.019294117647059, "grad_norm": 0.5626879192268438, "learning_rate": 9.987043619274604e-05, "loss": 0.333, "step": 8541 }, { "epoch": 4.019764705882353, "grad_norm": 0.39453876484406253, "learning_rate": 9.982912259680063e-05, "loss": 0.1512, "step": 8542 }, { "epoch": 4.020235294117647, "grad_norm": 0.38039384261868486, "learning_rate": 9.978781328569427e-05, "loss": 0.209, "step": 8543 }, { "epoch": 4.020705882352941, "grad_norm": 0.8643964354490653, "learning_rate": 9.974650826295493e-05, "loss": 0.5809, "step": 8544 }, { "epoch": 4.021176470588236, "grad_norm": 0.36658261070688114, "learning_rate": 9.970520753211027e-05, "loss": 0.3055, "step": 8545 }, { "epoch": 4.02164705882353, "grad_norm": 0.32369316932154524, "learning_rate": 9.966391109668758e-05, "loss": 0.1921, "step": 8546 }, { "epoch": 4.022117647058823, "grad_norm": 0.38197593927065837, "learning_rate": 9.96226189602138e-05, "loss": 0.3351, "step": 8547 }, { "epoch": 4.022588235294117, "grad_norm": 0.10147470325561835, "learning_rate": 9.958133112621547e-05, "loss": 0.0215, "step": 8548 }, { "epoch": 4.023058823529412, "grad_norm": 0.22727615559363815, "learning_rate": 9.954004759821871e-05, "loss": 0.0962, "step": 8549 }, { "epoch": 4.023529411764706, "grad_norm": 0.4545070512935477, "learning_rate": 9.949876837974944e-05, "loss": 0.3234, "step": 8550 }, { "epoch": 4.024, "grad_norm": 0.3819778846093913, "learning_rate": 9.945749347433304e-05, "loss": 0.2529, "step": 8551 }, { "epoch": 4.024470588235294, "grad_norm": 0.3369398671399105, "learning_rate": 9.941622288549458e-05, "loss": 0.2021, "step": 8552 }, { "epoch": 4.0249411764705885, "grad_norm": 0.43789710795704734, "learning_rate": 9.937495661675878e-05, "loss": 0.2696, "step": 8553 }, { "epoch": 4.025411764705883, "grad_norm": 0.6038460331899748, "learning_rate": 9.933369467165004e-05, "loss": 0.2073, "step": 8554 }, { "epoch": 4.025882352941177, "grad_norm": 0.42097015455626113, "learning_rate": 9.929243705369232e-05, "loss": 0.348, "step": 8555 }, { "epoch": 4.02635294117647, "grad_norm": 0.3978619462662213, "learning_rate": 9.925118376640918e-05, "loss": 0.233, "step": 8556 }, { "epoch": 4.0268235294117645, "grad_norm": 0.37411172064577397, "learning_rate": 9.920993481332383e-05, "loss": 0.2554, "step": 8557 }, { "epoch": 4.027294117647059, "grad_norm": 0.23238426422202688, "learning_rate": 9.916869019795922e-05, "loss": 0.1185, "step": 8558 }, { "epoch": 4.027764705882353, "grad_norm": 0.33200592803604734, "learning_rate": 9.912744992383777e-05, "loss": 0.1814, "step": 8559 }, { "epoch": 4.028235294117647, "grad_norm": 0.34010863900403493, "learning_rate": 9.90862139944816e-05, "loss": 0.2136, "step": 8560 }, { "epoch": 4.028705882352941, "grad_norm": 0.42184835648805374, "learning_rate": 9.904498241341252e-05, "loss": 0.3358, "step": 8561 }, { "epoch": 4.029176470588236, "grad_norm": 0.29139828933168793, "learning_rate": 9.900375518415188e-05, "loss": 0.1332, "step": 8562 }, { "epoch": 4.02964705882353, "grad_norm": 0.32384975456414133, "learning_rate": 9.896253231022066e-05, "loss": 0.0628, "step": 8563 }, { "epoch": 4.030117647058823, "grad_norm": 0.4729009272469021, "learning_rate": 9.892131379513953e-05, "loss": 0.3601, "step": 8564 }, { "epoch": 4.030588235294117, "grad_norm": 0.3592200289095218, "learning_rate": 9.888009964242876e-05, "loss": 0.2573, "step": 8565 }, { "epoch": 4.031058823529412, "grad_norm": 0.36866460287067776, "learning_rate": 9.883888985560822e-05, "loss": 0.1435, "step": 8566 }, { "epoch": 4.031529411764706, "grad_norm": 0.2747298469356767, "learning_rate": 9.879768443819737e-05, "loss": 0.0603, "step": 8567 }, { "epoch": 4.032, "grad_norm": 0.36145899487480115, "learning_rate": 9.875648339371546e-05, "loss": 0.2479, "step": 8568 }, { "epoch": 4.032470588235294, "grad_norm": 0.43867039907996636, "learning_rate": 9.871528672568124e-05, "loss": 0.252, "step": 8569 }, { "epoch": 4.0329411764705885, "grad_norm": 0.5086220699332884, "learning_rate": 9.867409443761304e-05, "loss": 0.338, "step": 8570 }, { "epoch": 4.033411764705883, "grad_norm": 0.479253380569732, "learning_rate": 9.863290653302894e-05, "loss": 0.338, "step": 8571 }, { "epoch": 4.033882352941176, "grad_norm": 0.07525225831949876, "learning_rate": 9.859172301544659e-05, "loss": 0.0152, "step": 8572 }, { "epoch": 4.03435294117647, "grad_norm": 0.5857738986364269, "learning_rate": 9.855054388838322e-05, "loss": 0.3108, "step": 8573 }, { "epoch": 4.0348235294117645, "grad_norm": 0.24269471860769934, "learning_rate": 9.85093691553558e-05, "loss": 0.1014, "step": 8574 }, { "epoch": 4.035294117647059, "grad_norm": 0.3258886405833715, "learning_rate": 9.84681988198808e-05, "loss": 0.1679, "step": 8575 }, { "epoch": 4.035764705882353, "grad_norm": 0.28508927409769275, "learning_rate": 9.842703288547438e-05, "loss": 0.1787, "step": 8576 }, { "epoch": 4.036235294117647, "grad_norm": 0.29773887030503204, "learning_rate": 9.838587135565229e-05, "loss": 0.1071, "step": 8577 }, { "epoch": 4.036705882352941, "grad_norm": 0.5049745837801622, "learning_rate": 9.834471423393e-05, "loss": 0.3473, "step": 8578 }, { "epoch": 4.037176470588236, "grad_norm": 0.31474264278869574, "learning_rate": 9.830356152382245e-05, "loss": 0.207, "step": 8579 }, { "epoch": 4.03764705882353, "grad_norm": 0.30476259980027454, "learning_rate": 9.826241322884432e-05, "loss": 0.121, "step": 8580 }, { "epoch": 4.038117647058823, "grad_norm": 0.329557713514677, "learning_rate": 9.822126935250987e-05, "loss": 0.1515, "step": 8581 }, { "epoch": 4.038588235294117, "grad_norm": 0.2874962357021846, "learning_rate": 9.818012989833301e-05, "loss": 0.1789, "step": 8582 }, { "epoch": 4.039058823529412, "grad_norm": 0.4687784312301478, "learning_rate": 9.81389948698272e-05, "loss": 0.2936, "step": 8583 }, { "epoch": 4.039529411764706, "grad_norm": 0.6375707224942337, "learning_rate": 9.809786427050562e-05, "loss": 0.4833, "step": 8584 }, { "epoch": 4.04, "grad_norm": 0.2820357767379994, "learning_rate": 9.805673810388103e-05, "loss": 0.1822, "step": 8585 }, { "epoch": 4.040470588235294, "grad_norm": 0.43319357647653867, "learning_rate": 9.801561637346575e-05, "loss": 0.3769, "step": 8586 }, { "epoch": 4.0409411764705885, "grad_norm": 0.22964081790700075, "learning_rate": 9.797449908277184e-05, "loss": 0.1465, "step": 8587 }, { "epoch": 4.041411764705883, "grad_norm": 0.484864518358692, "learning_rate": 9.793338623531089e-05, "loss": 0.2776, "step": 8588 }, { "epoch": 4.041882352941176, "grad_norm": 0.442019928065791, "learning_rate": 9.789227783459415e-05, "loss": 0.3194, "step": 8589 }, { "epoch": 4.04235294117647, "grad_norm": 0.3008792755821084, "learning_rate": 9.785117388413244e-05, "loss": 0.0897, "step": 8590 }, { "epoch": 4.0428235294117645, "grad_norm": 0.518949924103154, "learning_rate": 9.781007438743629e-05, "loss": 0.3765, "step": 8591 }, { "epoch": 4.043294117647059, "grad_norm": 0.4477376666169104, "learning_rate": 9.776897934801577e-05, "loss": 0.1876, "step": 8592 }, { "epoch": 4.043764705882353, "grad_norm": 0.36669236435632085, "learning_rate": 9.77278887693806e-05, "loss": 0.1151, "step": 8593 }, { "epoch": 4.044235294117647, "grad_norm": 0.4257072099704093, "learning_rate": 9.768680265504013e-05, "loss": 0.1568, "step": 8594 }, { "epoch": 4.044705882352941, "grad_norm": 0.2973500147236581, "learning_rate": 9.764572100850331e-05, "loss": 0.1082, "step": 8595 }, { "epoch": 4.045176470588236, "grad_norm": 0.3889491931784586, "learning_rate": 9.760464383327868e-05, "loss": 0.1868, "step": 8596 }, { "epoch": 4.04564705882353, "grad_norm": 0.25386968250790576, "learning_rate": 9.756357113287449e-05, "loss": 0.0789, "step": 8597 }, { "epoch": 4.046117647058823, "grad_norm": 0.29706840131244516, "learning_rate": 9.75225029107985e-05, "loss": 0.1066, "step": 8598 }, { "epoch": 4.046588235294117, "grad_norm": 0.39901625087023546, "learning_rate": 9.74814391705582e-05, "loss": 0.2292, "step": 8599 }, { "epoch": 4.047058823529412, "grad_norm": 0.2517167798692807, "learning_rate": 9.744037991566058e-05, "loss": 0.137, "step": 8600 }, { "epoch": 4.047529411764706, "grad_norm": 1.0303221146549932, "learning_rate": 9.739932514961232e-05, "loss": 0.5889, "step": 8601 }, { "epoch": 4.048, "grad_norm": 0.5991390238198235, "learning_rate": 9.73582748759197e-05, "loss": 0.2828, "step": 8602 }, { "epoch": 4.048470588235294, "grad_norm": 0.31060651012721313, "learning_rate": 9.731722909808863e-05, "loss": 0.0766, "step": 8603 }, { "epoch": 4.0489411764705885, "grad_norm": 0.4140168214689735, "learning_rate": 9.727618781962457e-05, "loss": 0.2589, "step": 8604 }, { "epoch": 4.049411764705883, "grad_norm": 0.449327545697667, "learning_rate": 9.723515104403268e-05, "loss": 0.2666, "step": 8605 }, { "epoch": 4.049882352941176, "grad_norm": 0.27603022829816526, "learning_rate": 9.719411877481776e-05, "loss": 0.0996, "step": 8606 }, { "epoch": 4.05035294117647, "grad_norm": 0.3396263108077545, "learning_rate": 9.715309101548406e-05, "loss": 0.2702, "step": 8607 }, { "epoch": 4.0508235294117645, "grad_norm": 0.24053944966737464, "learning_rate": 9.711206776953559e-05, "loss": 0.0986, "step": 8608 }, { "epoch": 4.051294117647059, "grad_norm": 0.3270393817265561, "learning_rate": 9.707104904047596e-05, "loss": 0.2587, "step": 8609 }, { "epoch": 4.051764705882353, "grad_norm": 0.3720254918876598, "learning_rate": 9.703003483180836e-05, "loss": 0.2429, "step": 8610 }, { "epoch": 4.052235294117647, "grad_norm": 0.4213752476152518, "learning_rate": 9.698902514703558e-05, "loss": 0.2271, "step": 8611 }, { "epoch": 4.052705882352941, "grad_norm": 0.2695772969179787, "learning_rate": 9.694801998966011e-05, "loss": 0.0893, "step": 8612 }, { "epoch": 4.053176470588236, "grad_norm": 0.3640176190621353, "learning_rate": 9.690701936318395e-05, "loss": 0.2227, "step": 8613 }, { "epoch": 4.05364705882353, "grad_norm": 0.37676588651412557, "learning_rate": 9.686602327110869e-05, "loss": 0.1262, "step": 8614 }, { "epoch": 4.054117647058823, "grad_norm": 0.32720235760008914, "learning_rate": 9.682503171693573e-05, "loss": 0.167, "step": 8615 }, { "epoch": 4.054588235294117, "grad_norm": 0.4373044254139869, "learning_rate": 9.678404470416587e-05, "loss": 0.291, "step": 8616 }, { "epoch": 4.055058823529412, "grad_norm": 0.5684902575005171, "learning_rate": 9.674306223629963e-05, "loss": 0.3876, "step": 8617 }, { "epoch": 4.055529411764706, "grad_norm": 0.21538118834658931, "learning_rate": 9.670208431683704e-05, "loss": 0.0699, "step": 8618 }, { "epoch": 4.056, "grad_norm": 0.1085363165736603, "learning_rate": 9.666111094927792e-05, "loss": 0.0314, "step": 8619 }, { "epoch": 4.056470588235294, "grad_norm": 0.3370363926955828, "learning_rate": 9.662014213712155e-05, "loss": 0.1214, "step": 8620 }, { "epoch": 4.0569411764705885, "grad_norm": 0.39865006547274656, "learning_rate": 9.657917788386684e-05, "loss": 0.2533, "step": 8621 }, { "epoch": 4.057411764705883, "grad_norm": 0.3262996269179888, "learning_rate": 9.65382181930124e-05, "loss": 0.2319, "step": 8622 }, { "epoch": 4.057882352941176, "grad_norm": 0.5309702448905634, "learning_rate": 9.649726306805635e-05, "loss": 0.3707, "step": 8623 }, { "epoch": 4.05835294117647, "grad_norm": 0.2798430154808735, "learning_rate": 9.645631251249643e-05, "loss": 0.1512, "step": 8624 }, { "epoch": 4.0588235294117645, "grad_norm": 0.42289414760928135, "learning_rate": 9.641536652983008e-05, "loss": 0.2375, "step": 8625 }, { "epoch": 4.059294117647059, "grad_norm": 0.4747264509904446, "learning_rate": 9.637442512355428e-05, "loss": 0.2307, "step": 8626 }, { "epoch": 4.059764705882353, "grad_norm": 0.43133221131006333, "learning_rate": 9.63334882971656e-05, "loss": 0.2085, "step": 8627 }, { "epoch": 4.060235294117647, "grad_norm": 0.40305263502929706, "learning_rate": 9.629255605416021e-05, "loss": 0.2185, "step": 8628 }, { "epoch": 4.060705882352941, "grad_norm": 0.37820495632567575, "learning_rate": 9.6251628398034e-05, "loss": 0.2344, "step": 8629 }, { "epoch": 4.061176470588236, "grad_norm": 0.33927765571593327, "learning_rate": 9.621070533228236e-05, "loss": 0.1348, "step": 8630 }, { "epoch": 4.06164705882353, "grad_norm": 0.1933232121216386, "learning_rate": 9.616978686040028e-05, "loss": 0.0774, "step": 8631 }, { "epoch": 4.062117647058823, "grad_norm": 0.41442915820269693, "learning_rate": 9.612887298588246e-05, "loss": 0.3295, "step": 8632 }, { "epoch": 4.062588235294117, "grad_norm": 0.3817210677554566, "learning_rate": 9.608796371222316e-05, "loss": 0.262, "step": 8633 }, { "epoch": 4.063058823529412, "grad_norm": 0.43234455204133027, "learning_rate": 9.604705904291614e-05, "loss": 0.3003, "step": 8634 }, { "epoch": 4.063529411764706, "grad_norm": 0.423091014102357, "learning_rate": 9.600615898145496e-05, "loss": 0.2183, "step": 8635 }, { "epoch": 4.064, "grad_norm": 0.28158716526978317, "learning_rate": 9.596526353133261e-05, "loss": 0.122, "step": 8636 }, { "epoch": 4.064470588235294, "grad_norm": 0.5439851246618516, "learning_rate": 9.59243726960418e-05, "loss": 0.3955, "step": 8637 }, { "epoch": 4.0649411764705885, "grad_norm": 0.41750202656956553, "learning_rate": 9.588348647907481e-05, "loss": 0.368, "step": 8638 }, { "epoch": 4.065411764705883, "grad_norm": 0.31334504511874245, "learning_rate": 9.584260488392353e-05, "loss": 0.1406, "step": 8639 }, { "epoch": 4.065882352941176, "grad_norm": 0.30755660814692315, "learning_rate": 9.580172791407942e-05, "loss": 0.1456, "step": 8640 }, { "epoch": 4.06635294117647, "grad_norm": 0.3395902958424559, "learning_rate": 9.576085557303355e-05, "loss": 0.1684, "step": 8641 }, { "epoch": 4.0668235294117645, "grad_norm": 0.22957009276212112, "learning_rate": 9.571998786427665e-05, "loss": 0.1022, "step": 8642 }, { "epoch": 4.067294117647059, "grad_norm": 0.33782493816868936, "learning_rate": 9.567912479129907e-05, "loss": 0.2228, "step": 8643 }, { "epoch": 4.067764705882353, "grad_norm": 0.36594504034986514, "learning_rate": 9.563826635759068e-05, "loss": 0.1487, "step": 8644 }, { "epoch": 4.068235294117647, "grad_norm": 0.20540437716151724, "learning_rate": 9.559741256664098e-05, "loss": 0.0676, "step": 8645 }, { "epoch": 4.068705882352941, "grad_norm": 0.5903822491278253, "learning_rate": 9.555656342193907e-05, "loss": 0.4325, "step": 8646 }, { "epoch": 4.069176470588236, "grad_norm": 0.25588792758818407, "learning_rate": 9.551571892697372e-05, "loss": 0.0833, "step": 8647 }, { "epoch": 4.06964705882353, "grad_norm": 0.33417632112747103, "learning_rate": 9.547487908523322e-05, "loss": 0.2098, "step": 8648 }, { "epoch": 4.070117647058823, "grad_norm": 0.4844681030349146, "learning_rate": 9.543404390020543e-05, "loss": 0.1698, "step": 8649 }, { "epoch": 4.070588235294117, "grad_norm": 0.2526295932451519, "learning_rate": 9.5393213375378e-05, "loss": 0.0663, "step": 8650 }, { "epoch": 4.071058823529412, "grad_norm": 0.34915516890989207, "learning_rate": 9.535238751423799e-05, "loss": 0.2086, "step": 8651 }, { "epoch": 4.071529411764706, "grad_norm": 0.266617892661925, "learning_rate": 9.531156632027209e-05, "loss": 0.0745, "step": 8652 }, { "epoch": 4.072, "grad_norm": 0.27749103782582624, "learning_rate": 9.52707497969667e-05, "loss": 0.1313, "step": 8653 }, { "epoch": 4.072470588235294, "grad_norm": 0.2520321090530596, "learning_rate": 9.522993794780773e-05, "loss": 0.0945, "step": 8654 }, { "epoch": 4.0729411764705885, "grad_norm": 0.43627597200152174, "learning_rate": 9.518913077628066e-05, "loss": 0.2546, "step": 8655 }, { "epoch": 4.073411764705883, "grad_norm": 0.5366046143785236, "learning_rate": 9.514832828587071e-05, "loss": 0.2929, "step": 8656 }, { "epoch": 4.073882352941176, "grad_norm": 0.31283654704096137, "learning_rate": 9.510753048006257e-05, "loss": 0.1637, "step": 8657 }, { "epoch": 4.07435294117647, "grad_norm": 0.3251011938844394, "learning_rate": 9.506673736234057e-05, "loss": 0.1383, "step": 8658 }, { "epoch": 4.0748235294117645, "grad_norm": 0.38995657604776024, "learning_rate": 9.502594893618861e-05, "loss": 0.2209, "step": 8659 }, { "epoch": 4.075294117647059, "grad_norm": 0.36225308446073207, "learning_rate": 9.49851652050903e-05, "loss": 0.1598, "step": 8660 }, { "epoch": 4.075764705882353, "grad_norm": 0.22026759726749495, "learning_rate": 9.494438617252875e-05, "loss": 0.083, "step": 8661 }, { "epoch": 4.076235294117647, "grad_norm": 0.42368086103055275, "learning_rate": 9.490361184198659e-05, "loss": 0.2967, "step": 8662 }, { "epoch": 4.076705882352941, "grad_norm": 0.33633671598651227, "learning_rate": 9.486284221694626e-05, "loss": 0.171, "step": 8663 }, { "epoch": 4.077176470588236, "grad_norm": 0.33748072229885423, "learning_rate": 9.482207730088968e-05, "loss": 0.2118, "step": 8664 }, { "epoch": 4.07764705882353, "grad_norm": 0.2514634272981591, "learning_rate": 9.47813170972983e-05, "loss": 0.0822, "step": 8665 }, { "epoch": 4.078117647058823, "grad_norm": 0.49055764940828794, "learning_rate": 9.474056160965332e-05, "loss": 0.3251, "step": 8666 }, { "epoch": 4.078588235294117, "grad_norm": 0.4985714652628491, "learning_rate": 9.469981084143541e-05, "loss": 0.358, "step": 8667 }, { "epoch": 4.079058823529412, "grad_norm": 0.04656036005755833, "learning_rate": 9.465906479612492e-05, "loss": 0.0075, "step": 8668 }, { "epoch": 4.079529411764706, "grad_norm": 0.625722975010142, "learning_rate": 9.461832347720168e-05, "loss": 0.3163, "step": 8669 }, { "epoch": 4.08, "grad_norm": 0.32089131531666765, "learning_rate": 9.457758688814533e-05, "loss": 0.1263, "step": 8670 }, { "epoch": 4.080470588235294, "grad_norm": 0.2929492854470175, "learning_rate": 9.453685503243489e-05, "loss": 0.1223, "step": 8671 }, { "epoch": 4.0809411764705885, "grad_norm": 0.8709695212083912, "learning_rate": 9.449612791354904e-05, "loss": 0.3205, "step": 8672 }, { "epoch": 4.081411764705883, "grad_norm": 0.17744275434135806, "learning_rate": 9.445540553496614e-05, "loss": 0.0547, "step": 8673 }, { "epoch": 4.081882352941176, "grad_norm": 0.24638508049405503, "learning_rate": 9.441468790016406e-05, "loss": 0.09, "step": 8674 }, { "epoch": 4.08235294117647, "grad_norm": 0.3919484193335352, "learning_rate": 9.437397501262026e-05, "loss": 0.1416, "step": 8675 }, { "epoch": 4.0828235294117645, "grad_norm": 0.4808929591536881, "learning_rate": 9.433326687581183e-05, "loss": 0.1246, "step": 8676 }, { "epoch": 4.083294117647059, "grad_norm": 0.2842726572512093, "learning_rate": 9.429256349321548e-05, "loss": 0.1838, "step": 8677 }, { "epoch": 4.083764705882353, "grad_norm": 0.4962460311155359, "learning_rate": 9.425186486830745e-05, "loss": 0.3888, "step": 8678 }, { "epoch": 4.084235294117647, "grad_norm": 0.342565140334494, "learning_rate": 9.421117100456356e-05, "loss": 0.1641, "step": 8679 }, { "epoch": 4.084705882352941, "grad_norm": 0.28930211415729573, "learning_rate": 9.417048190545935e-05, "loss": 0.1026, "step": 8680 }, { "epoch": 4.085176470588236, "grad_norm": 0.30656858786961727, "learning_rate": 9.412979757446983e-05, "loss": 0.1213, "step": 8681 }, { "epoch": 4.085647058823529, "grad_norm": 0.2483531629964398, "learning_rate": 9.40891180150696e-05, "loss": 0.1097, "step": 8682 }, { "epoch": 4.086117647058823, "grad_norm": 0.18862922322339798, "learning_rate": 9.404844323073296e-05, "loss": 0.052, "step": 8683 }, { "epoch": 4.086588235294117, "grad_norm": 0.5223538195419386, "learning_rate": 9.400777322493373e-05, "loss": 0.3765, "step": 8684 }, { "epoch": 4.087058823529412, "grad_norm": 2.677840312211502, "learning_rate": 9.396710800114526e-05, "loss": 1.4298, "step": 8685 }, { "epoch": 4.087529411764706, "grad_norm": 0.4092274424616445, "learning_rate": 9.39264475628406e-05, "loss": 0.284, "step": 8686 }, { "epoch": 4.088, "grad_norm": 0.5503704139091539, "learning_rate": 9.388579191349237e-05, "loss": 0.3481, "step": 8687 }, { "epoch": 4.088470588235294, "grad_norm": 0.28525520725759757, "learning_rate": 9.384514105657277e-05, "loss": 0.111, "step": 8688 }, { "epoch": 4.0889411764705885, "grad_norm": 0.51313004743482, "learning_rate": 9.380449499555358e-05, "loss": 0.2945, "step": 8689 }, { "epoch": 4.089411764705883, "grad_norm": 0.276194395751427, "learning_rate": 9.376385373390611e-05, "loss": 0.1561, "step": 8690 }, { "epoch": 4.089882352941176, "grad_norm": 0.6323894555694713, "learning_rate": 9.372321727510139e-05, "loss": 0.3447, "step": 8691 }, { "epoch": 4.09035294117647, "grad_norm": 0.21442075088993015, "learning_rate": 9.368258562260999e-05, "loss": 0.0942, "step": 8692 }, { "epoch": 4.0908235294117645, "grad_norm": 0.5767122663780373, "learning_rate": 9.364195877990195e-05, "loss": 0.3293, "step": 8693 }, { "epoch": 4.091294117647059, "grad_norm": 0.41911134841228886, "learning_rate": 9.360133675044708e-05, "loss": 0.2354, "step": 8694 }, { "epoch": 4.091764705882353, "grad_norm": 0.3591467389254274, "learning_rate": 9.356071953771472e-05, "loss": 0.2302, "step": 8695 }, { "epoch": 4.092235294117647, "grad_norm": 0.33243794017333594, "learning_rate": 9.352010714517371e-05, "loss": 0.1752, "step": 8696 }, { "epoch": 4.092705882352941, "grad_norm": 0.3980688110770197, "learning_rate": 9.347949957629258e-05, "loss": 0.2691, "step": 8697 }, { "epoch": 4.093176470588236, "grad_norm": 0.18189036470534511, "learning_rate": 9.343889683453945e-05, "loss": 0.0504, "step": 8698 }, { "epoch": 4.09364705882353, "grad_norm": 0.4064268564437253, "learning_rate": 9.339829892338193e-05, "loss": 0.1351, "step": 8699 }, { "epoch": 4.094117647058823, "grad_norm": 0.5077610918088562, "learning_rate": 9.33577058462873e-05, "loss": 0.3627, "step": 8700 }, { "epoch": 4.094588235294117, "grad_norm": 0.3337669933499305, "learning_rate": 9.331711760672246e-05, "loss": 0.2959, "step": 8701 }, { "epoch": 4.095058823529412, "grad_norm": 0.45261715123507557, "learning_rate": 9.32765342081538e-05, "loss": 0.3589, "step": 8702 }, { "epoch": 4.095529411764706, "grad_norm": 0.12583245867941031, "learning_rate": 9.323595565404731e-05, "loss": 0.0351, "step": 8703 }, { "epoch": 4.096, "grad_norm": 0.23954427202363282, "learning_rate": 9.319538194786867e-05, "loss": 0.0973, "step": 8704 }, { "epoch": 4.096470588235294, "grad_norm": 0.09383646886436463, "learning_rate": 9.315481309308303e-05, "loss": 0.0262, "step": 8705 }, { "epoch": 4.0969411764705885, "grad_norm": 0.3185926341614758, "learning_rate": 9.311424909315514e-05, "loss": 0.3148, "step": 8706 }, { "epoch": 4.097411764705883, "grad_norm": 0.3215583122090861, "learning_rate": 9.307368995154945e-05, "loss": 0.2587, "step": 8707 }, { "epoch": 4.097882352941176, "grad_norm": 0.33044690269563265, "learning_rate": 9.303313567172985e-05, "loss": 0.2465, "step": 8708 }, { "epoch": 4.09835294117647, "grad_norm": 0.4509833430232151, "learning_rate": 9.299258625715988e-05, "loss": 0.1696, "step": 8709 }, { "epoch": 4.0988235294117645, "grad_norm": 0.32997028551552116, "learning_rate": 9.295204171130264e-05, "loss": 0.2024, "step": 8710 }, { "epoch": 4.099294117647059, "grad_norm": 0.42578770672106653, "learning_rate": 9.291150203762087e-05, "loss": 0.2503, "step": 8711 }, { "epoch": 4.099764705882353, "grad_norm": 0.39658340618434845, "learning_rate": 9.287096723957685e-05, "loss": 0.2645, "step": 8712 }, { "epoch": 4.100235294117647, "grad_norm": 0.10670165018817099, "learning_rate": 9.283043732063243e-05, "loss": 0.0275, "step": 8713 }, { "epoch": 4.100705882352941, "grad_norm": 0.4794352936140378, "learning_rate": 9.27899122842491e-05, "loss": 0.2156, "step": 8714 }, { "epoch": 4.101176470588236, "grad_norm": 0.48990665522776417, "learning_rate": 9.274939213388785e-05, "loss": 0.1472, "step": 8715 }, { "epoch": 4.101647058823529, "grad_norm": 0.47695717724129894, "learning_rate": 9.270887687300932e-05, "loss": 0.4049, "step": 8716 }, { "epoch": 4.102117647058823, "grad_norm": 0.3823639624270179, "learning_rate": 9.266836650507375e-05, "loss": 0.2552, "step": 8717 }, { "epoch": 4.102588235294117, "grad_norm": 0.3050389879623219, "learning_rate": 9.26278610335409e-05, "loss": 0.1166, "step": 8718 }, { "epoch": 4.103058823529412, "grad_norm": 0.5023556467764638, "learning_rate": 9.25873604618701e-05, "loss": 0.4162, "step": 8719 }, { "epoch": 4.103529411764706, "grad_norm": 0.32312110511695363, "learning_rate": 9.254686479352029e-05, "loss": 0.1643, "step": 8720 }, { "epoch": 4.104, "grad_norm": 0.2819891994166324, "learning_rate": 9.250637403195008e-05, "loss": 0.1616, "step": 8721 }, { "epoch": 4.104470588235294, "grad_norm": 0.2933448078111075, "learning_rate": 9.246588818061753e-05, "loss": 0.1319, "step": 8722 }, { "epoch": 4.1049411764705885, "grad_norm": 0.36107648142212884, "learning_rate": 9.242540724298032e-05, "loss": 0.2338, "step": 8723 }, { "epoch": 4.105411764705883, "grad_norm": 0.3846800473584523, "learning_rate": 9.238493122249574e-05, "loss": 0.2454, "step": 8724 }, { "epoch": 4.105882352941176, "grad_norm": 1.188077769807458, "learning_rate": 9.234446012262065e-05, "loss": 0.2313, "step": 8725 }, { "epoch": 4.10635294117647, "grad_norm": 0.3724832525605513, "learning_rate": 9.230399394681141e-05, "loss": 0.2394, "step": 8726 }, { "epoch": 4.1068235294117645, "grad_norm": 0.26967492132108584, "learning_rate": 9.226353269852413e-05, "loss": 0.0739, "step": 8727 }, { "epoch": 4.107294117647059, "grad_norm": 0.22565455814321303, "learning_rate": 9.222307638121436e-05, "loss": 0.0681, "step": 8728 }, { "epoch": 4.107764705882353, "grad_norm": 0.21717545634034752, "learning_rate": 9.218262499833723e-05, "loss": 0.0754, "step": 8729 }, { "epoch": 4.108235294117647, "grad_norm": 0.410985293174091, "learning_rate": 9.21421785533475e-05, "loss": 0.3348, "step": 8730 }, { "epoch": 4.108705882352941, "grad_norm": 0.2281731162955723, "learning_rate": 9.210173704969952e-05, "loss": 0.0993, "step": 8731 }, { "epoch": 4.109176470588236, "grad_norm": 0.18323850587962692, "learning_rate": 9.206130049084721e-05, "loss": 0.0839, "step": 8732 }, { "epoch": 4.109647058823529, "grad_norm": 0.6400379452096777, "learning_rate": 9.202086888024404e-05, "loss": 0.3406, "step": 8733 }, { "epoch": 4.110117647058823, "grad_norm": 0.20104104443467943, "learning_rate": 9.1980442221343e-05, "loss": 0.0785, "step": 8734 }, { "epoch": 4.110588235294117, "grad_norm": 0.16220555097310835, "learning_rate": 9.194002051759685e-05, "loss": 0.0349, "step": 8735 }, { "epoch": 4.111058823529412, "grad_norm": 0.4716761857134908, "learning_rate": 9.189960377245768e-05, "loss": 0.395, "step": 8736 }, { "epoch": 4.111529411764706, "grad_norm": 0.32176931869333736, "learning_rate": 9.185919198937736e-05, "loss": 0.1286, "step": 8737 }, { "epoch": 4.112, "grad_norm": 0.31016375244192324, "learning_rate": 9.181878517180719e-05, "loss": 0.0598, "step": 8738 }, { "epoch": 4.112470588235294, "grad_norm": 0.2604550602669142, "learning_rate": 9.177838332319818e-05, "loss": 0.1534, "step": 8739 }, { "epoch": 4.1129411764705885, "grad_norm": 0.21612414746961767, "learning_rate": 9.173798644700079e-05, "loss": 0.0854, "step": 8740 }, { "epoch": 4.113411764705883, "grad_norm": 0.45439242444153055, "learning_rate": 9.169759454666511e-05, "loss": 0.3572, "step": 8741 }, { "epoch": 4.113882352941176, "grad_norm": 0.40507805934617347, "learning_rate": 9.165720762564086e-05, "loss": 0.1782, "step": 8742 }, { "epoch": 4.11435294117647, "grad_norm": 0.33535179902559337, "learning_rate": 9.161682568737726e-05, "loss": 0.2014, "step": 8743 }, { "epoch": 4.1148235294117645, "grad_norm": 0.1472206368248521, "learning_rate": 9.157644873532308e-05, "loss": 0.0413, "step": 8744 }, { "epoch": 4.115294117647059, "grad_norm": 0.43468429032072115, "learning_rate": 9.153607677292676e-05, "loss": 0.2624, "step": 8745 }, { "epoch": 4.115764705882353, "grad_norm": 0.21970929774422618, "learning_rate": 9.149570980363624e-05, "loss": 0.1067, "step": 8746 }, { "epoch": 4.116235294117647, "grad_norm": 0.36068760259989363, "learning_rate": 9.145534783089909e-05, "loss": 0.2205, "step": 8747 }, { "epoch": 4.116705882352941, "grad_norm": 0.6773354344084715, "learning_rate": 9.141499085816237e-05, "loss": 0.3573, "step": 8748 }, { "epoch": 4.117176470588236, "grad_norm": 0.3769754375568564, "learning_rate": 9.137463888887279e-05, "loss": 0.2011, "step": 8749 }, { "epoch": 4.117647058823529, "grad_norm": 0.39084552511498766, "learning_rate": 9.133429192647661e-05, "loss": 0.1623, "step": 8750 }, { "epoch": 4.118117647058823, "grad_norm": 0.7285356714468147, "learning_rate": 9.129394997441963e-05, "loss": 0.4799, "step": 8751 }, { "epoch": 4.118588235294117, "grad_norm": 0.40562779083715045, "learning_rate": 9.12536130361473e-05, "loss": 0.2806, "step": 8752 }, { "epoch": 4.119058823529412, "grad_norm": 0.3159236984483025, "learning_rate": 9.121328111510457e-05, "loss": 0.1269, "step": 8753 }, { "epoch": 4.119529411764706, "grad_norm": 0.43807504350458204, "learning_rate": 9.117295421473595e-05, "loss": 0.1611, "step": 8754 }, { "epoch": 4.12, "grad_norm": 0.5303610276787454, "learning_rate": 9.113263233848562e-05, "loss": 0.4536, "step": 8755 }, { "epoch": 4.120470588235294, "grad_norm": 0.24349721971776483, "learning_rate": 9.109231548979722e-05, "loss": 0.079, "step": 8756 }, { "epoch": 4.1209411764705886, "grad_norm": 0.3765433291862785, "learning_rate": 9.105200367211401e-05, "loss": 0.0813, "step": 8757 }, { "epoch": 4.121411764705883, "grad_norm": 0.3253151670942287, "learning_rate": 9.101169688887882e-05, "loss": 0.1206, "step": 8758 }, { "epoch": 4.121882352941176, "grad_norm": 0.46249120836521074, "learning_rate": 9.097139514353407e-05, "loss": 0.2589, "step": 8759 }, { "epoch": 4.12235294117647, "grad_norm": 0.4580764746837961, "learning_rate": 9.093109843952171e-05, "loss": 0.3333, "step": 8760 }, { "epoch": 4.1228235294117646, "grad_norm": 0.33688366645181267, "learning_rate": 9.089080678028325e-05, "loss": 0.2074, "step": 8761 }, { "epoch": 4.123294117647059, "grad_norm": 0.5740569038527992, "learning_rate": 9.085052016925984e-05, "loss": 0.2733, "step": 8762 }, { "epoch": 4.123764705882353, "grad_norm": 0.19059746619273216, "learning_rate": 9.081023860989213e-05, "loss": 0.0552, "step": 8763 }, { "epoch": 4.124235294117647, "grad_norm": 0.6153746526477903, "learning_rate": 9.076996210562035e-05, "loss": 0.2752, "step": 8764 }, { "epoch": 4.124705882352941, "grad_norm": 0.28684135054217214, "learning_rate": 9.072969065988435e-05, "loss": 0.1176, "step": 8765 }, { "epoch": 4.125176470588236, "grad_norm": 0.3723927981472946, "learning_rate": 9.06894242761235e-05, "loss": 0.1743, "step": 8766 }, { "epoch": 4.125647058823529, "grad_norm": 0.33513997664170947, "learning_rate": 9.064916295777669e-05, "loss": 0.1528, "step": 8767 }, { "epoch": 4.126117647058823, "grad_norm": 0.2387558063210818, "learning_rate": 9.06089067082825e-05, "loss": 0.1639, "step": 8768 }, { "epoch": 4.1265882352941174, "grad_norm": 0.47977237870530764, "learning_rate": 9.0568655531079e-05, "loss": 0.3117, "step": 8769 }, { "epoch": 4.127058823529412, "grad_norm": 0.17166573759690476, "learning_rate": 9.052840942960383e-05, "loss": 0.0571, "step": 8770 }, { "epoch": 4.127529411764706, "grad_norm": 0.45275339665475806, "learning_rate": 9.048816840729415e-05, "loss": 0.182, "step": 8771 }, { "epoch": 4.128, "grad_norm": 0.15986518718880244, "learning_rate": 9.044793246758685e-05, "loss": 0.0192, "step": 8772 }, { "epoch": 4.128470588235294, "grad_norm": 0.5529911293804606, "learning_rate": 9.04077016139182e-05, "loss": 0.3914, "step": 8773 }, { "epoch": 4.128941176470589, "grad_norm": 0.4073854538386617, "learning_rate": 9.03674758497241e-05, "loss": 0.3163, "step": 8774 }, { "epoch": 4.129411764705883, "grad_norm": 0.3777125117155228, "learning_rate": 9.032725517844004e-05, "loss": 0.2553, "step": 8775 }, { "epoch": 4.129882352941176, "grad_norm": 0.32239119741620104, "learning_rate": 9.028703960350113e-05, "loss": 0.1414, "step": 8776 }, { "epoch": 4.13035294117647, "grad_norm": 0.27024087735455576, "learning_rate": 9.024682912834196e-05, "loss": 0.1064, "step": 8777 }, { "epoch": 4.130823529411765, "grad_norm": 0.4179855785764662, "learning_rate": 9.020662375639663e-05, "loss": 0.2577, "step": 8778 }, { "epoch": 4.131294117647059, "grad_norm": 0.3341452850126539, "learning_rate": 9.016642349109892e-05, "loss": 0.23, "step": 8779 }, { "epoch": 4.131764705882353, "grad_norm": 0.37581480731235145, "learning_rate": 9.012622833588213e-05, "loss": 0.1561, "step": 8780 }, { "epoch": 4.132235294117647, "grad_norm": 0.3418045757144591, "learning_rate": 9.008603829417913e-05, "loss": 0.1332, "step": 8781 }, { "epoch": 4.1327058823529415, "grad_norm": 0.38085083061642566, "learning_rate": 9.00458533694223e-05, "loss": 0.2835, "step": 8782 }, { "epoch": 4.133176470588236, "grad_norm": 0.4038915904080585, "learning_rate": 9.000567356504368e-05, "loss": 0.3107, "step": 8783 }, { "epoch": 4.133647058823529, "grad_norm": 0.24114297123147166, "learning_rate": 8.996549888447484e-05, "loss": 0.1034, "step": 8784 }, { "epoch": 4.134117647058823, "grad_norm": 0.40013628322156936, "learning_rate": 8.99253293311468e-05, "loss": 0.2904, "step": 8785 }, { "epoch": 4.1345882352941175, "grad_norm": 0.36313845576437753, "learning_rate": 8.988516490849037e-05, "loss": 0.2745, "step": 8786 }, { "epoch": 4.135058823529412, "grad_norm": 0.31605819441240707, "learning_rate": 8.984500561993569e-05, "loss": 0.1083, "step": 8787 }, { "epoch": 4.135529411764706, "grad_norm": 0.4576219329657582, "learning_rate": 8.98048514689126e-05, "loss": 0.3017, "step": 8788 }, { "epoch": 4.136, "grad_norm": 0.25024084958767023, "learning_rate": 8.976470245885041e-05, "loss": 0.1199, "step": 8789 }, { "epoch": 4.136470588235294, "grad_norm": 0.37322977510961075, "learning_rate": 8.972455859317811e-05, "loss": 0.2751, "step": 8790 }, { "epoch": 4.136941176470589, "grad_norm": 0.34264951692124196, "learning_rate": 8.968441987532417e-05, "loss": 0.1925, "step": 8791 }, { "epoch": 4.137411764705883, "grad_norm": 0.3710993781944125, "learning_rate": 8.964428630871657e-05, "loss": 0.1062, "step": 8792 }, { "epoch": 4.137882352941176, "grad_norm": 0.4403948108290417, "learning_rate": 8.960415789678303e-05, "loss": 0.2878, "step": 8793 }, { "epoch": 4.13835294117647, "grad_norm": 0.48227253512028867, "learning_rate": 8.95640346429506e-05, "loss": 0.2925, "step": 8794 }, { "epoch": 4.138823529411765, "grad_norm": 0.2782080820468067, "learning_rate": 8.952391655064605e-05, "loss": 0.1609, "step": 8795 }, { "epoch": 4.139294117647059, "grad_norm": 0.8713992515869768, "learning_rate": 8.948380362329567e-05, "loss": 0.581, "step": 8796 }, { "epoch": 4.139764705882353, "grad_norm": 0.3646738128799032, "learning_rate": 8.944369586432531e-05, "loss": 0.1886, "step": 8797 }, { "epoch": 4.140235294117647, "grad_norm": 0.305786685476477, "learning_rate": 8.940359327716032e-05, "loss": 0.0763, "step": 8798 }, { "epoch": 4.1407058823529415, "grad_norm": 0.13505880001121626, "learning_rate": 8.936349586522569e-05, "loss": 0.0257, "step": 8799 }, { "epoch": 4.141176470588236, "grad_norm": 0.28775632207064683, "learning_rate": 8.932340363194595e-05, "loss": 0.1211, "step": 8800 }, { "epoch": 4.141647058823529, "grad_norm": 0.22344732851578666, "learning_rate": 8.928331658074516e-05, "loss": 0.0638, "step": 8801 }, { "epoch": 4.142117647058823, "grad_norm": 0.4153660464130414, "learning_rate": 8.92432347150469e-05, "loss": 0.253, "step": 8802 }, { "epoch": 4.1425882352941175, "grad_norm": 0.2998702854728435, "learning_rate": 8.920315803827446e-05, "loss": 0.1361, "step": 8803 }, { "epoch": 4.143058823529412, "grad_norm": 0.43398383310459815, "learning_rate": 8.916308655385054e-05, "loss": 0.2523, "step": 8804 }, { "epoch": 4.143529411764706, "grad_norm": 0.42134221762245977, "learning_rate": 8.912302026519738e-05, "loss": 0.171, "step": 8805 }, { "epoch": 4.144, "grad_norm": 0.37745536865748136, "learning_rate": 8.908295917573692e-05, "loss": 0.3184, "step": 8806 }, { "epoch": 4.144470588235294, "grad_norm": 0.49752639311855684, "learning_rate": 8.904290328889055e-05, "loss": 0.3813, "step": 8807 }, { "epoch": 4.144941176470589, "grad_norm": 0.41093232831244286, "learning_rate": 8.900285260807922e-05, "loss": 0.2752, "step": 8808 }, { "epoch": 4.145411764705882, "grad_norm": 0.20891178910147157, "learning_rate": 8.896280713672344e-05, "loss": 0.0635, "step": 8809 }, { "epoch": 4.145882352941176, "grad_norm": 0.8527990568140824, "learning_rate": 8.892276687824335e-05, "loss": 0.395, "step": 8810 }, { "epoch": 4.14635294117647, "grad_norm": 0.4050217350084456, "learning_rate": 8.888273183605856e-05, "loss": 0.2694, "step": 8811 }, { "epoch": 4.146823529411765, "grad_norm": 0.20594089314314343, "learning_rate": 8.884270201358821e-05, "loss": 0.0649, "step": 8812 }, { "epoch": 4.147294117647059, "grad_norm": 0.2935812095163071, "learning_rate": 8.880267741425111e-05, "loss": 0.1136, "step": 8813 }, { "epoch": 4.147764705882353, "grad_norm": 0.4374265611854043, "learning_rate": 8.876265804146554e-05, "loss": 0.2539, "step": 8814 }, { "epoch": 4.148235294117647, "grad_norm": 0.4361773329985168, "learning_rate": 8.872264389864929e-05, "loss": 0.2681, "step": 8815 }, { "epoch": 4.1487058823529415, "grad_norm": 0.08609383040796297, "learning_rate": 8.868263498921987e-05, "loss": 0.0032, "step": 8816 }, { "epoch": 4.149176470588236, "grad_norm": 0.8103755411557628, "learning_rate": 8.864263131659417e-05, "loss": 0.5707, "step": 8817 }, { "epoch": 4.149647058823529, "grad_norm": 0.538314489817438, "learning_rate": 8.860263288418866e-05, "loss": 0.3778, "step": 8818 }, { "epoch": 4.150117647058823, "grad_norm": 0.30336110905692143, "learning_rate": 8.856263969541948e-05, "loss": 0.0927, "step": 8819 }, { "epoch": 4.1505882352941175, "grad_norm": 0.2998673647432849, "learning_rate": 8.852265175370223e-05, "loss": 0.1868, "step": 8820 }, { "epoch": 4.151058823529412, "grad_norm": 0.20173502009029898, "learning_rate": 8.848266906245207e-05, "loss": 0.1011, "step": 8821 }, { "epoch": 4.151529411764706, "grad_norm": 0.46413894324540605, "learning_rate": 8.844269162508373e-05, "loss": 0.3839, "step": 8822 }, { "epoch": 4.152, "grad_norm": 0.43631544869515465, "learning_rate": 8.840271944501143e-05, "loss": 0.352, "step": 8823 }, { "epoch": 4.152470588235294, "grad_norm": 0.18766896012294362, "learning_rate": 8.836275252564903e-05, "loss": 0.0912, "step": 8824 }, { "epoch": 4.152941176470589, "grad_norm": 0.28886559622726754, "learning_rate": 8.832279087040992e-05, "loss": 0.2746, "step": 8825 }, { "epoch": 4.153411764705883, "grad_norm": 0.33731687791077186, "learning_rate": 8.828283448270698e-05, "loss": 0.2503, "step": 8826 }, { "epoch": 4.153882352941176, "grad_norm": 0.43831151168975274, "learning_rate": 8.824288336595267e-05, "loss": 0.2422, "step": 8827 }, { "epoch": 4.15435294117647, "grad_norm": 0.2700072483498461, "learning_rate": 8.820293752355906e-05, "loss": 0.0852, "step": 8828 }, { "epoch": 4.154823529411765, "grad_norm": 0.6123658396524949, "learning_rate": 8.816299695893773e-05, "loss": 0.2621, "step": 8829 }, { "epoch": 4.155294117647059, "grad_norm": 0.4866810556687028, "learning_rate": 8.81230616754997e-05, "loss": 0.374, "step": 8830 }, { "epoch": 4.155764705882353, "grad_norm": 0.44207286005100954, "learning_rate": 8.808313167665575e-05, "loss": 0.225, "step": 8831 }, { "epoch": 4.156235294117647, "grad_norm": 0.35445248900240567, "learning_rate": 8.804320696581606e-05, "loss": 0.1967, "step": 8832 }, { "epoch": 4.1567058823529415, "grad_norm": 0.2763503344314028, "learning_rate": 8.800328754639038e-05, "loss": 0.1636, "step": 8833 }, { "epoch": 4.157176470588236, "grad_norm": 0.32974460009360973, "learning_rate": 8.796337342178804e-05, "loss": 0.167, "step": 8834 }, { "epoch": 4.157647058823529, "grad_norm": 0.24012119355500297, "learning_rate": 8.792346459541789e-05, "loss": 0.125, "step": 8835 }, { "epoch": 4.158117647058823, "grad_norm": 0.6221954995051804, "learning_rate": 8.788356107068833e-05, "loss": 0.4538, "step": 8836 }, { "epoch": 4.1585882352941175, "grad_norm": 0.38130154261651733, "learning_rate": 8.784366285100738e-05, "loss": 0.231, "step": 8837 }, { "epoch": 4.159058823529412, "grad_norm": 0.28243999982807466, "learning_rate": 8.780376993978246e-05, "loss": 0.1228, "step": 8838 }, { "epoch": 4.159529411764706, "grad_norm": 0.41442619117122464, "learning_rate": 8.77638823404207e-05, "loss": 0.2577, "step": 8839 }, { "epoch": 4.16, "grad_norm": 0.5619802894604844, "learning_rate": 8.772400005632859e-05, "loss": 0.4662, "step": 8840 }, { "epoch": 4.160470588235294, "grad_norm": 0.3559295792597348, "learning_rate": 8.768412309091238e-05, "loss": 0.0796, "step": 8841 }, { "epoch": 4.160941176470589, "grad_norm": 0.14557237806145262, "learning_rate": 8.764425144757771e-05, "loss": 0.0435, "step": 8842 }, { "epoch": 4.161411764705882, "grad_norm": 0.31596773350405544, "learning_rate": 8.760438512972982e-05, "loss": 0.2237, "step": 8843 }, { "epoch": 4.161882352941176, "grad_norm": 0.397067567100581, "learning_rate": 8.756452414077351e-05, "loss": 0.1688, "step": 8844 }, { "epoch": 4.16235294117647, "grad_norm": 0.3180656190994277, "learning_rate": 8.752466848411305e-05, "loss": 0.2319, "step": 8845 }, { "epoch": 4.162823529411765, "grad_norm": 0.6774875448185824, "learning_rate": 8.748481816315233e-05, "loss": 0.2932, "step": 8846 }, { "epoch": 4.163294117647059, "grad_norm": 0.22411440520991782, "learning_rate": 8.744497318129483e-05, "loss": 0.1014, "step": 8847 }, { "epoch": 4.163764705882353, "grad_norm": 0.391228612963175, "learning_rate": 8.74051335419434e-05, "loss": 0.2407, "step": 8848 }, { "epoch": 4.164235294117647, "grad_norm": 0.3091524498796127, "learning_rate": 8.736529924850057e-05, "loss": 0.1822, "step": 8849 }, { "epoch": 4.1647058823529415, "grad_norm": 0.36409391795140894, "learning_rate": 8.73254703043685e-05, "loss": 0.2595, "step": 8850 }, { "epoch": 4.165176470588236, "grad_norm": 0.36075823986188316, "learning_rate": 8.72856467129486e-05, "loss": 0.1909, "step": 8851 }, { "epoch": 4.165647058823529, "grad_norm": 0.4523707498557826, "learning_rate": 8.72458284776421e-05, "loss": 0.2502, "step": 8852 }, { "epoch": 4.166117647058823, "grad_norm": 0.3619624558710853, "learning_rate": 8.72060156018497e-05, "loss": 0.1852, "step": 8853 }, { "epoch": 4.1665882352941175, "grad_norm": 0.5174014556594154, "learning_rate": 8.716620808897151e-05, "loss": 0.4012, "step": 8854 }, { "epoch": 4.167058823529412, "grad_norm": 0.6011232223299772, "learning_rate": 8.712640594240735e-05, "loss": 0.4766, "step": 8855 }, { "epoch": 4.167529411764706, "grad_norm": 0.3482961808940234, "learning_rate": 8.708660916555657e-05, "loss": 0.161, "step": 8856 }, { "epoch": 4.168, "grad_norm": 0.5782612933454423, "learning_rate": 8.70468177618179e-05, "loss": 0.4185, "step": 8857 }, { "epoch": 4.168470588235294, "grad_norm": 0.2602457803770594, "learning_rate": 8.700703173458978e-05, "loss": 0.1444, "step": 8858 }, { "epoch": 4.168941176470589, "grad_norm": 0.240586911229387, "learning_rate": 8.69672510872702e-05, "loss": 0.0923, "step": 8859 }, { "epoch": 4.169411764705882, "grad_norm": 0.6815583883566154, "learning_rate": 8.692747582325646e-05, "loss": 0.3241, "step": 8860 }, { "epoch": 4.169882352941176, "grad_norm": 0.3112536055480308, "learning_rate": 8.68877059459457e-05, "loss": 0.1481, "step": 8861 }, { "epoch": 4.17035294117647, "grad_norm": 0.3414280955295875, "learning_rate": 8.684794145873443e-05, "loss": 0.2271, "step": 8862 }, { "epoch": 4.170823529411765, "grad_norm": 0.4376557884769243, "learning_rate": 8.680818236501867e-05, "loss": 0.2806, "step": 8863 }, { "epoch": 4.171294117647059, "grad_norm": 0.3664349517149423, "learning_rate": 8.67684286681941e-05, "loss": 0.1673, "step": 8864 }, { "epoch": 4.171764705882353, "grad_norm": 0.538049759959461, "learning_rate": 8.672868037165587e-05, "loss": 0.4156, "step": 8865 }, { "epoch": 4.172235294117647, "grad_norm": 0.4320976578268264, "learning_rate": 8.668893747879869e-05, "loss": 0.2648, "step": 8866 }, { "epoch": 4.1727058823529415, "grad_norm": 0.4995568054289237, "learning_rate": 8.664919999301683e-05, "loss": 0.3743, "step": 8867 }, { "epoch": 4.173176470588236, "grad_norm": 0.3949580480212317, "learning_rate": 8.660946791770397e-05, "loss": 0.3918, "step": 8868 }, { "epoch": 4.173647058823529, "grad_norm": 0.4306084170381473, "learning_rate": 8.656974125625349e-05, "loss": 0.247, "step": 8869 }, { "epoch": 4.174117647058823, "grad_norm": 0.4453349196786276, "learning_rate": 8.653002001205827e-05, "loss": 0.2539, "step": 8870 }, { "epoch": 4.1745882352941175, "grad_norm": 0.4707389646939912, "learning_rate": 8.649030418851061e-05, "loss": 0.4376, "step": 8871 }, { "epoch": 4.175058823529412, "grad_norm": 0.47345545106268655, "learning_rate": 8.645059378900255e-05, "loss": 0.286, "step": 8872 }, { "epoch": 4.175529411764706, "grad_norm": 0.37365221482790245, "learning_rate": 8.641088881692542e-05, "loss": 0.1217, "step": 8873 }, { "epoch": 4.176, "grad_norm": 0.3299508438123599, "learning_rate": 8.637118927567028e-05, "loss": 0.2368, "step": 8874 }, { "epoch": 4.176470588235294, "grad_norm": 0.4296313904673469, "learning_rate": 8.633149516862775e-05, "loss": 0.1965, "step": 8875 }, { "epoch": 4.176941176470589, "grad_norm": 0.5454155059568152, "learning_rate": 8.629180649918775e-05, "loss": 0.3651, "step": 8876 }, { "epoch": 4.177411764705882, "grad_norm": 0.21334522940166864, "learning_rate": 8.625212327073998e-05, "loss": 0.0844, "step": 8877 }, { "epoch": 4.177882352941176, "grad_norm": 0.5587418324046026, "learning_rate": 8.621244548667363e-05, "loss": 0.3675, "step": 8878 }, { "epoch": 4.17835294117647, "grad_norm": 0.33758238440626703, "learning_rate": 8.617277315037723e-05, "loss": 0.2039, "step": 8879 }, { "epoch": 4.178823529411765, "grad_norm": 1.0089112756412646, "learning_rate": 8.613310626523909e-05, "loss": 0.5788, "step": 8880 }, { "epoch": 4.179294117647059, "grad_norm": 0.4847475725643619, "learning_rate": 8.609344483464697e-05, "loss": 0.4421, "step": 8881 }, { "epoch": 4.179764705882353, "grad_norm": 0.6056516797967209, "learning_rate": 8.60537888619881e-05, "loss": 0.5318, "step": 8882 }, { "epoch": 4.180235294117647, "grad_norm": 0.9896414569413945, "learning_rate": 8.601413835064929e-05, "loss": 0.5684, "step": 8883 }, { "epoch": 4.1807058823529415, "grad_norm": 0.49731347876135196, "learning_rate": 8.597449330401698e-05, "loss": 0.4037, "step": 8884 }, { "epoch": 4.181176470588236, "grad_norm": 0.22980607930114058, "learning_rate": 8.593485372547693e-05, "loss": 0.0861, "step": 8885 }, { "epoch": 4.181647058823529, "grad_norm": 0.2670159186192433, "learning_rate": 8.589521961841462e-05, "loss": 0.1001, "step": 8886 }, { "epoch": 4.182117647058823, "grad_norm": 0.45182125452341654, "learning_rate": 8.585559098621503e-05, "loss": 0.2726, "step": 8887 }, { "epoch": 4.1825882352941175, "grad_norm": 0.32718378118508523, "learning_rate": 8.581596783226254e-05, "loss": 0.2074, "step": 8888 }, { "epoch": 4.183058823529412, "grad_norm": 0.6467836222413045, "learning_rate": 8.577635015994122e-05, "loss": 0.5609, "step": 8889 }, { "epoch": 4.183529411764706, "grad_norm": 0.3617503413619758, "learning_rate": 8.573673797263466e-05, "loss": 0.2518, "step": 8890 }, { "epoch": 4.184, "grad_norm": 0.3661550570908529, "learning_rate": 8.569713127372583e-05, "loss": 0.2625, "step": 8891 }, { "epoch": 4.184470588235294, "grad_norm": 0.3970017564084234, "learning_rate": 8.565753006659741e-05, "loss": 0.1361, "step": 8892 }, { "epoch": 4.184941176470589, "grad_norm": 0.401539993208564, "learning_rate": 8.561793435463157e-05, "loss": 0.3205, "step": 8893 }, { "epoch": 4.185411764705882, "grad_norm": 0.7194886523797832, "learning_rate": 8.557834414120986e-05, "loss": 0.4246, "step": 8894 }, { "epoch": 4.185882352941176, "grad_norm": 0.3395164816078734, "learning_rate": 8.553875942971362e-05, "loss": 0.1376, "step": 8895 }, { "epoch": 4.18635294117647, "grad_norm": 0.38821919441326314, "learning_rate": 8.549918022352343e-05, "loss": 0.2375, "step": 8896 }, { "epoch": 4.186823529411765, "grad_norm": 0.31442247490829117, "learning_rate": 8.545960652601964e-05, "loss": 0.092, "step": 8897 }, { "epoch": 4.187294117647059, "grad_norm": 0.40901358713035607, "learning_rate": 8.542003834058207e-05, "loss": 0.2608, "step": 8898 }, { "epoch": 4.187764705882353, "grad_norm": 0.3481596011050729, "learning_rate": 8.538047567058994e-05, "loss": 0.1143, "step": 8899 }, { "epoch": 4.188235294117647, "grad_norm": 0.5899462777357053, "learning_rate": 8.534091851942214e-05, "loss": 0.4621, "step": 8900 }, { "epoch": 4.1887058823529415, "grad_norm": 0.13453237496005746, "learning_rate": 8.530136689045712e-05, "loss": 0.0476, "step": 8901 }, { "epoch": 4.189176470588236, "grad_norm": 0.3967913478248885, "learning_rate": 8.526182078707261e-05, "loss": 0.2268, "step": 8902 }, { "epoch": 4.189647058823529, "grad_norm": 0.4799092463706297, "learning_rate": 8.522228021264618e-05, "loss": 0.4635, "step": 8903 }, { "epoch": 4.190117647058823, "grad_norm": 0.37030811062744046, "learning_rate": 8.518274517055482e-05, "loss": 0.1443, "step": 8904 }, { "epoch": 4.1905882352941175, "grad_norm": 0.3866723182649236, "learning_rate": 8.514321566417488e-05, "loss": 0.2164, "step": 8905 }, { "epoch": 4.191058823529412, "grad_norm": 0.3282457714682659, "learning_rate": 8.510369169688243e-05, "loss": 0.264, "step": 8906 }, { "epoch": 4.191529411764706, "grad_norm": 0.3990409909760635, "learning_rate": 8.506417327205309e-05, "loss": 0.1374, "step": 8907 }, { "epoch": 4.192, "grad_norm": 0.2325593396060646, "learning_rate": 8.502466039306175e-05, "loss": 0.0871, "step": 8908 }, { "epoch": 4.192470588235294, "grad_norm": 0.3146663489951544, "learning_rate": 8.498515306328323e-05, "loss": 0.1079, "step": 8909 }, { "epoch": 4.192941176470589, "grad_norm": 0.43123191387365445, "learning_rate": 8.494565128609146e-05, "loss": 0.2231, "step": 8910 }, { "epoch": 4.193411764705882, "grad_norm": 0.43528730505263086, "learning_rate": 8.490615506486018e-05, "loss": 0.2253, "step": 8911 }, { "epoch": 4.193882352941176, "grad_norm": 0.5148813749775777, "learning_rate": 8.486666440296259e-05, "loss": 0.3671, "step": 8912 }, { "epoch": 4.19435294117647, "grad_norm": 0.6497526213558502, "learning_rate": 8.482717930377129e-05, "loss": 0.3493, "step": 8913 }, { "epoch": 4.194823529411765, "grad_norm": 0.124361608575456, "learning_rate": 8.478769977065853e-05, "loss": 0.0341, "step": 8914 }, { "epoch": 4.195294117647059, "grad_norm": 0.5003999606399318, "learning_rate": 8.474822580699614e-05, "loss": 0.453, "step": 8915 }, { "epoch": 4.195764705882353, "grad_norm": 0.6608932523462548, "learning_rate": 8.470875741615527e-05, "loss": 0.3012, "step": 8916 }, { "epoch": 4.196235294117647, "grad_norm": 0.2556512404029159, "learning_rate": 8.466929460150678e-05, "loss": 0.0818, "step": 8917 }, { "epoch": 4.1967058823529415, "grad_norm": 0.3070430193507266, "learning_rate": 8.462983736642101e-05, "loss": 0.1527, "step": 8918 }, { "epoch": 4.197176470588236, "grad_norm": 0.3631477668709245, "learning_rate": 8.459038571426772e-05, "loss": 0.2988, "step": 8919 }, { "epoch": 4.197647058823529, "grad_norm": 0.16218181315424549, "learning_rate": 8.455093964841632e-05, "loss": 0.0509, "step": 8920 }, { "epoch": 4.198117647058823, "grad_norm": 0.11670413571376666, "learning_rate": 8.451149917223576e-05, "loss": 0.029, "step": 8921 }, { "epoch": 4.1985882352941175, "grad_norm": 0.4631785076596522, "learning_rate": 8.447206428909434e-05, "loss": 0.2414, "step": 8922 }, { "epoch": 4.199058823529412, "grad_norm": 0.43699724125838685, "learning_rate": 8.44326350023601e-05, "loss": 0.214, "step": 8923 }, { "epoch": 4.199529411764706, "grad_norm": 0.3893402810065613, "learning_rate": 8.439321131540036e-05, "loss": 0.2066, "step": 8924 }, { "epoch": 4.2, "grad_norm": 0.52579565840261, "learning_rate": 8.435379323158218e-05, "loss": 0.1214, "step": 8925 }, { "epoch": 4.200470588235294, "grad_norm": 0.4429377158160139, "learning_rate": 8.431438075427213e-05, "loss": 0.3231, "step": 8926 }, { "epoch": 4.200941176470589, "grad_norm": 0.35202073656749533, "learning_rate": 8.427497388683607e-05, "loss": 0.2417, "step": 8927 }, { "epoch": 4.201411764705882, "grad_norm": 0.21836864002914888, "learning_rate": 8.423557263263961e-05, "loss": 0.0691, "step": 8928 }, { "epoch": 4.201882352941176, "grad_norm": 0.41135114452574334, "learning_rate": 8.419617699504791e-05, "loss": 0.2794, "step": 8929 }, { "epoch": 4.20235294117647, "grad_norm": 0.48193003534040024, "learning_rate": 8.415678697742538e-05, "loss": 0.3057, "step": 8930 }, { "epoch": 4.202823529411765, "grad_norm": 0.36249881894156555, "learning_rate": 8.41174025831362e-05, "loss": 0.2622, "step": 8931 }, { "epoch": 4.203294117647059, "grad_norm": 0.38358478915323074, "learning_rate": 8.407802381554406e-05, "loss": 0.2729, "step": 8932 }, { "epoch": 4.203764705882353, "grad_norm": 0.778285312337035, "learning_rate": 8.403865067801197e-05, "loss": 0.563, "step": 8933 }, { "epoch": 4.204235294117647, "grad_norm": 0.5569015774350428, "learning_rate": 8.399928317390265e-05, "loss": 0.491, "step": 8934 }, { "epoch": 4.2047058823529415, "grad_norm": 0.43473488038796226, "learning_rate": 8.395992130657834e-05, "loss": 0.1901, "step": 8935 }, { "epoch": 4.205176470588235, "grad_norm": 0.47047273738724515, "learning_rate": 8.392056507940061e-05, "loss": 0.2795, "step": 8936 }, { "epoch": 4.205647058823529, "grad_norm": 0.2935569526854513, "learning_rate": 8.388121449573075e-05, "loss": 0.1895, "step": 8937 }, { "epoch": 4.206117647058823, "grad_norm": 0.4115005348578964, "learning_rate": 8.384186955892954e-05, "loss": 0.3244, "step": 8938 }, { "epoch": 4.2065882352941175, "grad_norm": 0.33368018579080977, "learning_rate": 8.380253027235709e-05, "loss": 0.2223, "step": 8939 }, { "epoch": 4.207058823529412, "grad_norm": 0.39369289921578704, "learning_rate": 8.37631966393733e-05, "loss": 0.2073, "step": 8940 }, { "epoch": 4.207529411764706, "grad_norm": 1.3111548513722484, "learning_rate": 8.372386866333743e-05, "loss": 0.4464, "step": 8941 }, { "epoch": 4.208, "grad_norm": 0.16925812551920674, "learning_rate": 8.36845463476082e-05, "loss": 0.0695, "step": 8942 }, { "epoch": 4.208470588235294, "grad_norm": 0.16450143548391932, "learning_rate": 8.364522969554399e-05, "loss": 0.0416, "step": 8943 }, { "epoch": 4.208941176470589, "grad_norm": 0.2198841203527595, "learning_rate": 8.36059187105027e-05, "loss": 0.114, "step": 8944 }, { "epoch": 4.209411764705882, "grad_norm": 0.509378294956643, "learning_rate": 8.356661339584156e-05, "loss": 0.3176, "step": 8945 }, { "epoch": 4.209882352941176, "grad_norm": 0.5171138414140611, "learning_rate": 8.352731375491754e-05, "loss": 0.1842, "step": 8946 }, { "epoch": 4.21035294117647, "grad_norm": 0.26603404067410236, "learning_rate": 8.348801979108691e-05, "loss": 0.097, "step": 8947 }, { "epoch": 4.210823529411765, "grad_norm": 0.4903305907336342, "learning_rate": 8.344873150770562e-05, "loss": 0.3558, "step": 8948 }, { "epoch": 4.211294117647059, "grad_norm": 0.3292313192570679, "learning_rate": 8.340944890812916e-05, "loss": 0.1911, "step": 8949 }, { "epoch": 4.211764705882353, "grad_norm": 0.3436648248752441, "learning_rate": 8.337017199571235e-05, "loss": 0.118, "step": 8950 }, { "epoch": 4.212235294117647, "grad_norm": 0.4301393990305778, "learning_rate": 8.333090077380964e-05, "loss": 0.0613, "step": 8951 }, { "epoch": 4.2127058823529415, "grad_norm": 0.3855617263636292, "learning_rate": 8.329163524577504e-05, "loss": 0.2829, "step": 8952 }, { "epoch": 4.213176470588236, "grad_norm": 0.4082077598368258, "learning_rate": 8.325237541496197e-05, "loss": 0.3205, "step": 8953 }, { "epoch": 4.213647058823529, "grad_norm": 0.28602679158357325, "learning_rate": 8.32131212847235e-05, "loss": 0.1607, "step": 8954 }, { "epoch": 4.214117647058823, "grad_norm": 0.41843937442679174, "learning_rate": 8.3173872858412e-05, "loss": 0.1862, "step": 8955 }, { "epoch": 4.2145882352941175, "grad_norm": 0.3371834134039902, "learning_rate": 8.313463013937953e-05, "loss": 0.1157, "step": 8956 }, { "epoch": 4.215058823529412, "grad_norm": 0.3903059022692013, "learning_rate": 8.309539313097769e-05, "loss": 0.1708, "step": 8957 }, { "epoch": 4.215529411764706, "grad_norm": 0.40701055676020886, "learning_rate": 8.305616183655738e-05, "loss": 0.1743, "step": 8958 }, { "epoch": 4.216, "grad_norm": 0.39009970763866736, "learning_rate": 8.30169362594692e-05, "loss": 0.2609, "step": 8959 }, { "epoch": 4.216470588235294, "grad_norm": 0.4276967766690534, "learning_rate": 8.29777164030633e-05, "loss": 0.3051, "step": 8960 }, { "epoch": 4.216941176470589, "grad_norm": 0.2992096925891447, "learning_rate": 8.293850227068908e-05, "loss": 0.2051, "step": 8961 }, { "epoch": 4.217411764705882, "grad_norm": 0.39065733255066964, "learning_rate": 8.289929386569569e-05, "loss": 0.207, "step": 8962 }, { "epoch": 4.217882352941176, "grad_norm": 0.33242501703047367, "learning_rate": 8.28600911914318e-05, "loss": 0.0978, "step": 8963 }, { "epoch": 4.21835294117647, "grad_norm": 0.20068963996731065, "learning_rate": 8.282089425124538e-05, "loss": 0.0557, "step": 8964 }, { "epoch": 4.218823529411765, "grad_norm": 0.38399567686681907, "learning_rate": 8.278170304848413e-05, "loss": 0.2482, "step": 8965 }, { "epoch": 4.219294117647059, "grad_norm": 0.39959228822174836, "learning_rate": 8.274251758649518e-05, "loss": 0.2991, "step": 8966 }, { "epoch": 4.219764705882353, "grad_norm": 0.48332352675260226, "learning_rate": 8.270333786862506e-05, "loss": 0.3896, "step": 8967 }, { "epoch": 4.220235294117647, "grad_norm": 0.48646488138659433, "learning_rate": 8.266416389821997e-05, "loss": 0.3707, "step": 8968 }, { "epoch": 4.2207058823529415, "grad_norm": 0.20494693653592086, "learning_rate": 8.262499567862565e-05, "loss": 0.0682, "step": 8969 }, { "epoch": 4.221176470588235, "grad_norm": 0.4919367144375493, "learning_rate": 8.258583321318708e-05, "loss": 0.3595, "step": 8970 }, { "epoch": 4.221647058823529, "grad_norm": 0.3576651466909188, "learning_rate": 8.254667650524904e-05, "loss": 0.2991, "step": 8971 }, { "epoch": 4.222117647058823, "grad_norm": 0.3891871244990216, "learning_rate": 8.250752555815574e-05, "loss": 0.2181, "step": 8972 }, { "epoch": 4.2225882352941175, "grad_norm": 0.4303847531658638, "learning_rate": 8.246838037525073e-05, "loss": 0.2406, "step": 8973 }, { "epoch": 4.223058823529412, "grad_norm": 0.22584715293069305, "learning_rate": 8.242924095987736e-05, "loss": 0.0672, "step": 8974 }, { "epoch": 4.223529411764706, "grad_norm": 0.4237916407162298, "learning_rate": 8.239010731537816e-05, "loss": 0.3725, "step": 8975 }, { "epoch": 4.224, "grad_norm": 0.40421724713573953, "learning_rate": 8.235097944509545e-05, "loss": 0.2965, "step": 8976 }, { "epoch": 4.224470588235294, "grad_norm": 0.28676021903382665, "learning_rate": 8.231185735237096e-05, "loss": 0.0949, "step": 8977 }, { "epoch": 4.224941176470589, "grad_norm": 0.3652777967639665, "learning_rate": 8.22727410405458e-05, "loss": 0.2895, "step": 8978 }, { "epoch": 4.225411764705882, "grad_norm": 0.4623250336949744, "learning_rate": 8.223363051296074e-05, "loss": 0.3443, "step": 8979 }, { "epoch": 4.225882352941176, "grad_norm": 0.42272489770680377, "learning_rate": 8.21945257729561e-05, "loss": 0.265, "step": 8980 }, { "epoch": 4.22635294117647, "grad_norm": 0.23229351338671583, "learning_rate": 8.215542682387148e-05, "loss": 0.1031, "step": 8981 }, { "epoch": 4.226823529411765, "grad_norm": 0.28574772153440353, "learning_rate": 8.211633366904618e-05, "loss": 0.1963, "step": 8982 }, { "epoch": 4.227294117647059, "grad_norm": 0.3164391250136957, "learning_rate": 8.207724631181904e-05, "loss": 0.1553, "step": 8983 }, { "epoch": 4.227764705882353, "grad_norm": 0.46585532881890324, "learning_rate": 8.203816475552813e-05, "loss": 0.2422, "step": 8984 }, { "epoch": 4.228235294117647, "grad_norm": 0.1879240334757001, "learning_rate": 8.19990890035113e-05, "loss": 0.054, "step": 8985 }, { "epoch": 4.2287058823529415, "grad_norm": 0.6623240797933746, "learning_rate": 8.196001905910589e-05, "loss": 0.4795, "step": 8986 }, { "epoch": 4.229176470588235, "grad_norm": 0.5349812835449486, "learning_rate": 8.192095492564851e-05, "loss": 0.3181, "step": 8987 }, { "epoch": 4.229647058823529, "grad_norm": 0.46310263174173627, "learning_rate": 8.188189660647554e-05, "loss": 0.1403, "step": 8988 }, { "epoch": 4.230117647058823, "grad_norm": 0.4033141731949913, "learning_rate": 8.184284410492274e-05, "loss": 0.3131, "step": 8989 }, { "epoch": 4.2305882352941175, "grad_norm": 0.457055149750501, "learning_rate": 8.180379742432533e-05, "loss": 0.3493, "step": 8990 }, { "epoch": 4.231058823529412, "grad_norm": 0.3750875209133344, "learning_rate": 8.176475656801811e-05, "loss": 0.1761, "step": 8991 }, { "epoch": 4.231529411764706, "grad_norm": 0.24203215220976929, "learning_rate": 8.172572153933541e-05, "loss": 0.0539, "step": 8992 }, { "epoch": 4.232, "grad_norm": 0.26413770999789005, "learning_rate": 8.168669234161094e-05, "loss": 0.1156, "step": 8993 }, { "epoch": 4.232470588235294, "grad_norm": 0.43750996838527423, "learning_rate": 8.164766897817806e-05, "loss": 0.2405, "step": 8994 }, { "epoch": 4.232941176470589, "grad_norm": 0.3045247680994167, "learning_rate": 8.160865145236946e-05, "loss": 0.1198, "step": 8995 }, { "epoch": 4.233411764705882, "grad_norm": 0.3538174015088335, "learning_rate": 8.15696397675175e-05, "loss": 0.2309, "step": 8996 }, { "epoch": 4.233882352941176, "grad_norm": 0.396104556033529, "learning_rate": 8.153063392695394e-05, "loss": 0.1516, "step": 8997 }, { "epoch": 4.23435294117647, "grad_norm": 0.4247944155963734, "learning_rate": 8.14916339340101e-05, "loss": 0.2517, "step": 8998 }, { "epoch": 4.234823529411765, "grad_norm": 0.8745188125237061, "learning_rate": 8.145263979201672e-05, "loss": 0.4608, "step": 8999 }, { "epoch": 4.235294117647059, "grad_norm": 0.6269186952136604, "learning_rate": 8.141365150430421e-05, "loss": 0.4578, "step": 9000 }, { "epoch": 4.235764705882353, "grad_norm": 0.19716712195753613, "learning_rate": 8.137466907420222e-05, "loss": 0.0434, "step": 9001 }, { "epoch": 4.236235294117647, "grad_norm": 0.19877668178498792, "learning_rate": 8.133569250504013e-05, "loss": 0.0696, "step": 9002 }, { "epoch": 4.2367058823529415, "grad_norm": 0.08666817127192668, "learning_rate": 8.129672180014666e-05, "loss": 0.0137, "step": 9003 }, { "epoch": 4.237176470588235, "grad_norm": 0.3859469959009899, "learning_rate": 8.125775696285013e-05, "loss": 0.2232, "step": 9004 }, { "epoch": 4.237647058823529, "grad_norm": 0.4359152280470805, "learning_rate": 8.121879799647839e-05, "loss": 0.1689, "step": 9005 }, { "epoch": 4.238117647058823, "grad_norm": 0.4671361037592942, "learning_rate": 8.117984490435861e-05, "loss": 0.2429, "step": 9006 }, { "epoch": 4.2385882352941175, "grad_norm": 0.34697122112484424, "learning_rate": 8.114089768981765e-05, "loss": 0.2158, "step": 9007 }, { "epoch": 4.239058823529412, "grad_norm": 0.24936213538877516, "learning_rate": 8.110195635618184e-05, "loss": 0.1105, "step": 9008 }, { "epoch": 4.239529411764706, "grad_norm": 0.3918640959792843, "learning_rate": 8.106302090677682e-05, "loss": 0.1592, "step": 9009 }, { "epoch": 4.24, "grad_norm": 0.338886075722994, "learning_rate": 8.102409134492795e-05, "loss": 0.1842, "step": 9010 }, { "epoch": 4.240470588235294, "grad_norm": 0.2929942016936643, "learning_rate": 8.098516767396008e-05, "loss": 0.1723, "step": 9011 }, { "epoch": 4.240941176470589, "grad_norm": 0.29833573427196897, "learning_rate": 8.094624989719734e-05, "loss": 0.1109, "step": 9012 }, { "epoch": 4.241411764705882, "grad_norm": 0.2604546785343567, "learning_rate": 8.090733801796356e-05, "loss": 0.0891, "step": 9013 }, { "epoch": 4.241882352941176, "grad_norm": 0.10948840990319982, "learning_rate": 8.086843203958207e-05, "loss": 0.0218, "step": 9014 }, { "epoch": 4.24235294117647, "grad_norm": 0.6817409841910533, "learning_rate": 8.082953196537553e-05, "loss": 0.3933, "step": 9015 }, { "epoch": 4.242823529411765, "grad_norm": 0.38270023806132225, "learning_rate": 8.079063779866622e-05, "loss": 0.1929, "step": 9016 }, { "epoch": 4.243294117647059, "grad_norm": 0.3510615523794485, "learning_rate": 8.075174954277598e-05, "loss": 0.2342, "step": 9017 }, { "epoch": 4.243764705882353, "grad_norm": 0.31365601395360637, "learning_rate": 8.071286720102594e-05, "loss": 0.1054, "step": 9018 }, { "epoch": 4.244235294117647, "grad_norm": 0.6146025327984186, "learning_rate": 8.067399077673688e-05, "loss": 0.3437, "step": 9019 }, { "epoch": 4.2447058823529416, "grad_norm": 0.4713437380662563, "learning_rate": 8.06351202732291e-05, "loss": 0.1763, "step": 9020 }, { "epoch": 4.245176470588235, "grad_norm": 0.4333145716088812, "learning_rate": 8.059625569382223e-05, "loss": 0.1632, "step": 9021 }, { "epoch": 4.245647058823529, "grad_norm": 0.4526614580829217, "learning_rate": 8.055739704183553e-05, "loss": 0.1107, "step": 9022 }, { "epoch": 4.246117647058823, "grad_norm": 0.319469392756492, "learning_rate": 8.051854432058785e-05, "loss": 0.1948, "step": 9023 }, { "epoch": 4.2465882352941176, "grad_norm": 0.5155924069974884, "learning_rate": 8.047969753339717e-05, "loss": 0.216, "step": 9024 }, { "epoch": 4.247058823529412, "grad_norm": 0.4630929473338336, "learning_rate": 8.04408566835814e-05, "loss": 0.2871, "step": 9025 }, { "epoch": 4.247529411764706, "grad_norm": 0.38159477240027484, "learning_rate": 8.040202177445761e-05, "loss": 0.2365, "step": 9026 }, { "epoch": 4.248, "grad_norm": 0.4193395113688255, "learning_rate": 8.036319280934251e-05, "loss": 0.2006, "step": 9027 }, { "epoch": 4.248470588235294, "grad_norm": 0.33611347344579584, "learning_rate": 8.032436979155237e-05, "loss": 0.1458, "step": 9028 }, { "epoch": 4.248941176470589, "grad_norm": 0.452627906046195, "learning_rate": 8.02855527244028e-05, "loss": 0.2323, "step": 9029 }, { "epoch": 4.249411764705882, "grad_norm": 0.34256244712629996, "learning_rate": 8.024674161120892e-05, "loss": 0.1205, "step": 9030 }, { "epoch": 4.249882352941176, "grad_norm": 0.2915327549357652, "learning_rate": 8.020793645528552e-05, "loss": 0.1534, "step": 9031 }, { "epoch": 4.2503529411764704, "grad_norm": 0.34765772354310465, "learning_rate": 8.016913725994661e-05, "loss": 0.125, "step": 9032 }, { "epoch": 4.250823529411765, "grad_norm": 0.1985757257166082, "learning_rate": 8.01303440285059e-05, "loss": 0.0636, "step": 9033 }, { "epoch": 4.251294117647059, "grad_norm": 0.3373713349502837, "learning_rate": 8.009155676427659e-05, "loss": 0.1902, "step": 9034 }, { "epoch": 4.251764705882353, "grad_norm": 0.41145053158792, "learning_rate": 8.005277547057114e-05, "loss": 0.1765, "step": 9035 }, { "epoch": 4.252235294117647, "grad_norm": 0.408418005705066, "learning_rate": 8.001400015070178e-05, "loss": 0.2797, "step": 9036 }, { "epoch": 4.252705882352942, "grad_norm": 0.5501864722594274, "learning_rate": 7.997523080798014e-05, "loss": 0.404, "step": 9037 }, { "epoch": 4.253176470588235, "grad_norm": 0.3074356061943121, "learning_rate": 7.993646744571719e-05, "loss": 0.1373, "step": 9038 }, { "epoch": 4.253647058823529, "grad_norm": 0.23713031329558554, "learning_rate": 7.98977100672236e-05, "loss": 0.0681, "step": 9039 }, { "epoch": 4.254117647058823, "grad_norm": 0.4321791335972469, "learning_rate": 7.985895867580948e-05, "loss": 0.2991, "step": 9040 }, { "epoch": 4.254588235294118, "grad_norm": 0.49689257107609225, "learning_rate": 7.982021327478419e-05, "loss": 0.4486, "step": 9041 }, { "epoch": 4.255058823529412, "grad_norm": 0.2925183996992368, "learning_rate": 7.978147386745705e-05, "loss": 0.094, "step": 9042 }, { "epoch": 4.255529411764706, "grad_norm": 0.4067681835694501, "learning_rate": 7.974274045713638e-05, "loss": 0.2625, "step": 9043 }, { "epoch": 4.256, "grad_norm": 0.29355569016299127, "learning_rate": 7.97040130471303e-05, "loss": 0.1542, "step": 9044 }, { "epoch": 4.2564705882352944, "grad_norm": 0.4534263348848469, "learning_rate": 7.966529164074636e-05, "loss": 0.263, "step": 9045 }, { "epoch": 4.256941176470589, "grad_norm": 0.44131944273263385, "learning_rate": 7.962657624129147e-05, "loss": 0.332, "step": 9046 }, { "epoch": 4.257411764705882, "grad_norm": 0.6347130813392722, "learning_rate": 7.958786685207212e-05, "loss": 0.2872, "step": 9047 }, { "epoch": 4.257882352941176, "grad_norm": 0.5597737437713692, "learning_rate": 7.954916347639438e-05, "loss": 0.2895, "step": 9048 }, { "epoch": 4.2583529411764705, "grad_norm": 0.1473533455576961, "learning_rate": 7.951046611756358e-05, "loss": 0.0422, "step": 9049 }, { "epoch": 4.258823529411765, "grad_norm": 0.2659967820029561, "learning_rate": 7.947177477888472e-05, "loss": 0.0548, "step": 9050 }, { "epoch": 4.259294117647059, "grad_norm": 0.3560349051386165, "learning_rate": 7.94330894636623e-05, "loss": 0.1028, "step": 9051 }, { "epoch": 4.259764705882353, "grad_norm": 0.28160610408057296, "learning_rate": 7.939441017520011e-05, "loss": 0.128, "step": 9052 }, { "epoch": 4.260235294117647, "grad_norm": 0.19226452833420668, "learning_rate": 7.935573691680166e-05, "loss": 0.0474, "step": 9053 }, { "epoch": 4.260705882352942, "grad_norm": 0.5354624939029538, "learning_rate": 7.931706969176972e-05, "loss": 0.3264, "step": 9054 }, { "epoch": 4.261176470588235, "grad_norm": 0.272772447456813, "learning_rate": 7.927840850340673e-05, "loss": 0.1359, "step": 9055 }, { "epoch": 4.261647058823529, "grad_norm": 0.3115344618023847, "learning_rate": 7.92397533550146e-05, "loss": 0.1618, "step": 9056 }, { "epoch": 4.262117647058823, "grad_norm": 0.6107829089311918, "learning_rate": 7.920110424989455e-05, "loss": 0.3465, "step": 9057 }, { "epoch": 4.262588235294118, "grad_norm": 0.4107160985814294, "learning_rate": 7.916246119134748e-05, "loss": 0.1039, "step": 9058 }, { "epoch": 4.263058823529412, "grad_norm": 0.43988969204522144, "learning_rate": 7.912382418267374e-05, "loss": 0.1719, "step": 9059 }, { "epoch": 4.263529411764706, "grad_norm": 0.35910185567884695, "learning_rate": 7.908519322717297e-05, "loss": 0.2108, "step": 9060 }, { "epoch": 4.264, "grad_norm": 0.9794897405188898, "learning_rate": 7.904656832814457e-05, "loss": 0.2399, "step": 9061 }, { "epoch": 4.2644705882352945, "grad_norm": 0.39780577042636855, "learning_rate": 7.900794948888728e-05, "loss": 0.2863, "step": 9062 }, { "epoch": 4.264941176470588, "grad_norm": 0.38077012006096833, "learning_rate": 7.896933671269929e-05, "loss": 0.2351, "step": 9063 }, { "epoch": 4.265411764705882, "grad_norm": 0.23631936218822805, "learning_rate": 7.893073000287834e-05, "loss": 0.0789, "step": 9064 }, { "epoch": 4.265882352941176, "grad_norm": 0.2976129320038237, "learning_rate": 7.88921293627217e-05, "loss": 0.224, "step": 9065 }, { "epoch": 4.2663529411764705, "grad_norm": 0.569011553323068, "learning_rate": 7.885353479552592e-05, "loss": 0.2543, "step": 9066 }, { "epoch": 4.266823529411765, "grad_norm": 0.2968300452286028, "learning_rate": 7.881494630458726e-05, "loss": 0.1574, "step": 9067 }, { "epoch": 4.267294117647059, "grad_norm": 0.30591984478215395, "learning_rate": 7.87763638932014e-05, "loss": 0.1966, "step": 9068 }, { "epoch": 4.267764705882353, "grad_norm": 0.35309522789375203, "learning_rate": 7.873778756466335e-05, "loss": 0.2144, "step": 9069 }, { "epoch": 4.268235294117647, "grad_norm": 0.630557861568397, "learning_rate": 7.869921732226779e-05, "loss": 0.3069, "step": 9070 }, { "epoch": 4.268705882352942, "grad_norm": 0.5053921788952329, "learning_rate": 7.866065316930886e-05, "loss": 0.2521, "step": 9071 }, { "epoch": 4.269176470588235, "grad_norm": 0.3046338317526292, "learning_rate": 7.862209510907999e-05, "loss": 0.2191, "step": 9072 }, { "epoch": 4.269647058823529, "grad_norm": 0.5945606444385302, "learning_rate": 7.858354314487432e-05, "loss": 0.1757, "step": 9073 }, { "epoch": 4.270117647058823, "grad_norm": 0.18444792650123865, "learning_rate": 7.854499727998442e-05, "loss": 0.0691, "step": 9074 }, { "epoch": 4.270588235294118, "grad_norm": 0.4761799147283932, "learning_rate": 7.850645751770217e-05, "loss": 0.2453, "step": 9075 }, { "epoch": 4.271058823529412, "grad_norm": 0.45404424393560383, "learning_rate": 7.846792386131917e-05, "loss": 0.4293, "step": 9076 }, { "epoch": 4.271529411764706, "grad_norm": 0.47695598770590725, "learning_rate": 7.84293963141263e-05, "loss": 0.3733, "step": 9077 }, { "epoch": 4.272, "grad_norm": 0.21009117678097658, "learning_rate": 7.839087487941403e-05, "loss": 0.1104, "step": 9078 }, { "epoch": 4.2724705882352945, "grad_norm": 0.35858141753113654, "learning_rate": 7.835235956047234e-05, "loss": 0.3047, "step": 9079 }, { "epoch": 4.272941176470589, "grad_norm": 0.4414414809543515, "learning_rate": 7.83138503605905e-05, "loss": 0.176, "step": 9080 }, { "epoch": 4.273411764705882, "grad_norm": 0.32991509825942933, "learning_rate": 7.827534728305747e-05, "loss": 0.2603, "step": 9081 }, { "epoch": 4.273882352941176, "grad_norm": 0.33871637710523916, "learning_rate": 7.823685033116166e-05, "loss": 0.2351, "step": 9082 }, { "epoch": 4.2743529411764705, "grad_norm": 0.3713165977786681, "learning_rate": 7.819835950819078e-05, "loss": 0.1801, "step": 9083 }, { "epoch": 4.274823529411765, "grad_norm": 0.36254396598516214, "learning_rate": 7.815987481743215e-05, "loss": 0.1611, "step": 9084 }, { "epoch": 4.275294117647059, "grad_norm": 0.6447679690666343, "learning_rate": 7.812139626217266e-05, "loss": 0.3658, "step": 9085 }, { "epoch": 4.275764705882353, "grad_norm": 0.48437980755487514, "learning_rate": 7.808292384569838e-05, "loss": 0.2579, "step": 9086 }, { "epoch": 4.276235294117647, "grad_norm": 0.2077083730314419, "learning_rate": 7.804445757129525e-05, "loss": 0.0756, "step": 9087 }, { "epoch": 4.276705882352942, "grad_norm": 0.2679626673821309, "learning_rate": 7.800599744224834e-05, "loss": 0.139, "step": 9088 }, { "epoch": 4.277176470588235, "grad_norm": 0.881854149619733, "learning_rate": 7.796754346184238e-05, "loss": 0.6188, "step": 9089 }, { "epoch": 4.277647058823529, "grad_norm": 0.783059971680758, "learning_rate": 7.792909563336157e-05, "loss": 0.4382, "step": 9090 }, { "epoch": 4.278117647058823, "grad_norm": 0.44864881157543635, "learning_rate": 7.789065396008945e-05, "loss": 0.2751, "step": 9091 }, { "epoch": 4.278588235294118, "grad_norm": 0.35713094460861144, "learning_rate": 7.785221844530918e-05, "loss": 0.2612, "step": 9092 }, { "epoch": 4.279058823529412, "grad_norm": 0.32862429676135235, "learning_rate": 7.78137890923034e-05, "loss": 0.144, "step": 9093 }, { "epoch": 4.279529411764706, "grad_norm": 0.13297228842454634, "learning_rate": 7.777536590435404e-05, "loss": 0.0358, "step": 9094 }, { "epoch": 4.28, "grad_norm": 0.5387840830062777, "learning_rate": 7.773694888474267e-05, "loss": 0.37, "step": 9095 }, { "epoch": 4.2804705882352945, "grad_norm": 0.2424515162673219, "learning_rate": 7.769853803675041e-05, "loss": 0.1025, "step": 9096 }, { "epoch": 4.280941176470588, "grad_norm": 0.3464755273793385, "learning_rate": 7.766013336365755e-05, "loss": 0.1535, "step": 9097 }, { "epoch": 4.281411764705882, "grad_norm": 0.38961504402763725, "learning_rate": 7.762173486874414e-05, "loss": 0.239, "step": 9098 }, { "epoch": 4.281882352941176, "grad_norm": 0.20417359803479893, "learning_rate": 7.758334255528963e-05, "loss": 0.1115, "step": 9099 }, { "epoch": 4.2823529411764705, "grad_norm": 0.4020862113941361, "learning_rate": 7.754495642657282e-05, "loss": 0.1689, "step": 9100 }, { "epoch": 4.282823529411765, "grad_norm": 0.48428949438539914, "learning_rate": 7.750657648587213e-05, "loss": 0.2956, "step": 9101 }, { "epoch": 4.283294117647059, "grad_norm": 0.2601446580946935, "learning_rate": 7.746820273646542e-05, "loss": 0.1, "step": 9102 }, { "epoch": 4.283764705882353, "grad_norm": 0.45818359724647767, "learning_rate": 7.742983518162993e-05, "loss": 0.236, "step": 9103 }, { "epoch": 4.284235294117647, "grad_norm": 0.30642417920225484, "learning_rate": 7.73914738246425e-05, "loss": 0.1883, "step": 9104 }, { "epoch": 4.284705882352942, "grad_norm": 0.2527206159287033, "learning_rate": 7.735311866877933e-05, "loss": 0.0669, "step": 9105 }, { "epoch": 4.285176470588235, "grad_norm": 0.4241641081785786, "learning_rate": 7.731476971731612e-05, "loss": 0.2373, "step": 9106 }, { "epoch": 4.285647058823529, "grad_norm": 0.2840741558316325, "learning_rate": 7.727642697352817e-05, "loss": 0.0604, "step": 9107 }, { "epoch": 4.286117647058823, "grad_norm": 0.2748269255093979, "learning_rate": 7.723809044069e-05, "loss": 0.0942, "step": 9108 }, { "epoch": 4.286588235294118, "grad_norm": 0.3905922724110503, "learning_rate": 7.719976012207579e-05, "loss": 0.2192, "step": 9109 }, { "epoch": 4.287058823529412, "grad_norm": 0.31397684053822933, "learning_rate": 7.716143602095924e-05, "loss": 0.0705, "step": 9110 }, { "epoch": 4.287529411764706, "grad_norm": 0.34611924515930836, "learning_rate": 7.712311814061323e-05, "loss": 0.1089, "step": 9111 }, { "epoch": 4.288, "grad_norm": 0.2873529483209792, "learning_rate": 7.708480648431043e-05, "loss": 0.1277, "step": 9112 }, { "epoch": 4.2884705882352945, "grad_norm": 0.5793202993059513, "learning_rate": 7.704650105532284e-05, "loss": 0.3968, "step": 9113 }, { "epoch": 4.288941176470589, "grad_norm": 0.41403292997933894, "learning_rate": 7.700820185692184e-05, "loss": 0.2224, "step": 9114 }, { "epoch": 4.289411764705882, "grad_norm": 0.2921021459320323, "learning_rate": 7.69699088923784e-05, "loss": 0.1853, "step": 9115 }, { "epoch": 4.289882352941176, "grad_norm": 0.17852130708161867, "learning_rate": 7.693162216496303e-05, "loss": 0.0514, "step": 9116 }, { "epoch": 4.2903529411764705, "grad_norm": 0.9223270780601039, "learning_rate": 7.689334167794547e-05, "loss": 0.225, "step": 9117 }, { "epoch": 4.290823529411765, "grad_norm": 0.44206242098755316, "learning_rate": 7.685506743459511e-05, "loss": 0.2956, "step": 9118 }, { "epoch": 4.291294117647059, "grad_norm": 0.30100392432479556, "learning_rate": 7.681679943818082e-05, "loss": 0.0878, "step": 9119 }, { "epoch": 4.291764705882353, "grad_norm": 0.1659872223033511, "learning_rate": 7.677853769197078e-05, "loss": 0.0521, "step": 9120 }, { "epoch": 4.292235294117647, "grad_norm": 0.6730412235217402, "learning_rate": 7.674028219923275e-05, "loss": 0.2108, "step": 9121 }, { "epoch": 4.292705882352941, "grad_norm": 0.5892553061618838, "learning_rate": 7.670203296323403e-05, "loss": 0.4693, "step": 9122 }, { "epoch": 4.293176470588235, "grad_norm": 0.5121597769521075, "learning_rate": 7.666378998724116e-05, "loss": 0.3135, "step": 9123 }, { "epoch": 4.293647058823529, "grad_norm": 0.41361998045751996, "learning_rate": 7.662555327452031e-05, "loss": 0.2799, "step": 9124 }, { "epoch": 4.294117647058823, "grad_norm": 0.2902955565483539, "learning_rate": 7.65873228283372e-05, "loss": 0.157, "step": 9125 }, { "epoch": 4.294588235294118, "grad_norm": 0.2829657298362969, "learning_rate": 7.654909865195675e-05, "loss": 0.2139, "step": 9126 }, { "epoch": 4.295058823529412, "grad_norm": 0.0963969208754933, "learning_rate": 7.65108807486436e-05, "loss": 0.0156, "step": 9127 }, { "epoch": 4.295529411764706, "grad_norm": 0.2548633997706831, "learning_rate": 7.647266912166162e-05, "loss": 0.0752, "step": 9128 }, { "epoch": 4.296, "grad_norm": 0.34267485423356925, "learning_rate": 7.643446377427438e-05, "loss": 0.1292, "step": 9129 }, { "epoch": 4.2964705882352945, "grad_norm": 0.5365781947282666, "learning_rate": 7.639626470974474e-05, "loss": 0.1545, "step": 9130 }, { "epoch": 4.296941176470588, "grad_norm": 0.23314247553981643, "learning_rate": 7.635807193133512e-05, "loss": 0.1046, "step": 9131 }, { "epoch": 4.297411764705882, "grad_norm": 0.18788357305378875, "learning_rate": 7.631988544230745e-05, "loss": 0.0642, "step": 9132 }, { "epoch": 4.297882352941176, "grad_norm": 0.39131576055054773, "learning_rate": 7.628170524592288e-05, "loss": 0.2542, "step": 9133 }, { "epoch": 4.2983529411764705, "grad_norm": 0.26832146960753533, "learning_rate": 7.624353134544229e-05, "loss": 0.2034, "step": 9134 }, { "epoch": 4.298823529411765, "grad_norm": 0.3304830206864925, "learning_rate": 7.620536374412594e-05, "loss": 0.0982, "step": 9135 }, { "epoch": 4.299294117647059, "grad_norm": 0.2416378819938613, "learning_rate": 7.616720244523342e-05, "loss": 0.0956, "step": 9136 }, { "epoch": 4.299764705882353, "grad_norm": 0.37606056229429896, "learning_rate": 7.612904745202399e-05, "loss": 0.2752, "step": 9137 }, { "epoch": 4.300235294117647, "grad_norm": 0.27643339849829857, "learning_rate": 7.609089876775628e-05, "loss": 0.1714, "step": 9138 }, { "epoch": 4.300705882352942, "grad_norm": 0.28656332916094235, "learning_rate": 7.605275639568828e-05, "loss": 0.1, "step": 9139 }, { "epoch": 4.301176470588235, "grad_norm": 0.5805718137660443, "learning_rate": 7.60146203390776e-05, "loss": 0.3024, "step": 9140 }, { "epoch": 4.301647058823529, "grad_norm": 0.19589840836186903, "learning_rate": 7.59764906011813e-05, "loss": 0.0585, "step": 9141 }, { "epoch": 4.302117647058823, "grad_norm": 0.666807318637585, "learning_rate": 7.593836718525572e-05, "loss": 0.3873, "step": 9142 }, { "epoch": 4.302588235294118, "grad_norm": 0.27261415785280285, "learning_rate": 7.590025009455684e-05, "loss": 0.1336, "step": 9143 }, { "epoch": 4.303058823529412, "grad_norm": 0.49498576343591605, "learning_rate": 7.586213933234013e-05, "loss": 0.3211, "step": 9144 }, { "epoch": 4.303529411764706, "grad_norm": 0.5728331233145203, "learning_rate": 7.58240349018603e-05, "loss": 0.1838, "step": 9145 }, { "epoch": 4.304, "grad_norm": 0.38606646726593274, "learning_rate": 7.578593680637173e-05, "loss": 0.2935, "step": 9146 }, { "epoch": 4.3044705882352945, "grad_norm": 0.2532665625955975, "learning_rate": 7.574784504912823e-05, "loss": 0.1337, "step": 9147 }, { "epoch": 4.304941176470588, "grad_norm": 0.1637604376953355, "learning_rate": 7.57097596333829e-05, "loss": 0.0692, "step": 9148 }, { "epoch": 4.305411764705882, "grad_norm": 0.11247330238900144, "learning_rate": 7.567168056238851e-05, "loss": 0.0217, "step": 9149 }, { "epoch": 4.305882352941176, "grad_norm": 0.4542685160802802, "learning_rate": 7.563360783939722e-05, "loss": 0.2308, "step": 9150 }, { "epoch": 4.3063529411764705, "grad_norm": 0.3252705002504598, "learning_rate": 7.559554146766054e-05, "loss": 0.234, "step": 9151 }, { "epoch": 4.306823529411765, "grad_norm": 0.44617537865972134, "learning_rate": 7.555748145042957e-05, "loss": 0.2897, "step": 9152 }, { "epoch": 4.307294117647059, "grad_norm": 0.3207197771890839, "learning_rate": 7.551942779095486e-05, "loss": 0.1992, "step": 9153 }, { "epoch": 4.307764705882353, "grad_norm": 0.5184192200234802, "learning_rate": 7.54813804924863e-05, "loss": 0.3097, "step": 9154 }, { "epoch": 4.308235294117647, "grad_norm": 0.3851065184773232, "learning_rate": 7.544333955827342e-05, "loss": 0.1549, "step": 9155 }, { "epoch": 4.308705882352941, "grad_norm": 0.5504567302358004, "learning_rate": 7.540530499156496e-05, "loss": 0.1891, "step": 9156 }, { "epoch": 4.309176470588235, "grad_norm": 0.25410517105325375, "learning_rate": 7.536727679560937e-05, "loss": 0.0876, "step": 9157 }, { "epoch": 4.309647058823529, "grad_norm": 0.3829142732431265, "learning_rate": 7.532925497365444e-05, "loss": 0.3064, "step": 9158 }, { "epoch": 4.310117647058823, "grad_norm": 0.2903848403089903, "learning_rate": 7.529123952894734e-05, "loss": 0.1957, "step": 9159 }, { "epoch": 4.310588235294118, "grad_norm": 0.4425532646107854, "learning_rate": 7.525323046473485e-05, "loss": 0.2459, "step": 9160 }, { "epoch": 4.311058823529412, "grad_norm": 0.3102421615371518, "learning_rate": 7.521522778426315e-05, "loss": 0.1326, "step": 9161 }, { "epoch": 4.311529411764706, "grad_norm": 0.262854732424133, "learning_rate": 7.517723149077776e-05, "loss": 0.1016, "step": 9162 }, { "epoch": 4.312, "grad_norm": 0.4119221654616353, "learning_rate": 7.513924158752383e-05, "loss": 0.2068, "step": 9163 }, { "epoch": 4.3124705882352945, "grad_norm": 0.5006249617737022, "learning_rate": 7.510125807774589e-05, "loss": 0.3418, "step": 9164 }, { "epoch": 4.312941176470588, "grad_norm": 0.35875303706801437, "learning_rate": 7.506328096468782e-05, "loss": 0.1727, "step": 9165 }, { "epoch": 4.313411764705882, "grad_norm": 0.4205007774331941, "learning_rate": 7.502531025159313e-05, "loss": 0.2364, "step": 9166 }, { "epoch": 4.313882352941176, "grad_norm": 0.1858529643660762, "learning_rate": 7.498734594170474e-05, "loss": 0.063, "step": 9167 }, { "epoch": 4.3143529411764705, "grad_norm": 0.3022206179165827, "learning_rate": 7.49493880382649e-05, "loss": 0.1083, "step": 9168 }, { "epoch": 4.314823529411765, "grad_norm": 0.5445453868282814, "learning_rate": 7.491143654451544e-05, "loss": 0.2772, "step": 9169 }, { "epoch": 4.315294117647059, "grad_norm": 0.6162415674544852, "learning_rate": 7.487349146369766e-05, "loss": 0.2167, "step": 9170 }, { "epoch": 4.315764705882353, "grad_norm": 0.3334212500090324, "learning_rate": 7.483555279905215e-05, "loss": 0.1713, "step": 9171 }, { "epoch": 4.316235294117647, "grad_norm": 0.3209150600515632, "learning_rate": 7.479762055381909e-05, "loss": 0.111, "step": 9172 }, { "epoch": 4.316705882352942, "grad_norm": 0.5105492022163292, "learning_rate": 7.475969473123819e-05, "loss": 0.2492, "step": 9173 }, { "epoch": 4.317176470588235, "grad_norm": 0.00023625889094948182, "learning_rate": 7.472177533454833e-05, "loss": 0.0, "step": 9174 }, { "epoch": 4.317647058823529, "grad_norm": 0.33604934547169474, "learning_rate": 7.46838623669881e-05, "loss": 0.2449, "step": 9175 }, { "epoch": 4.318117647058823, "grad_norm": 0.4910178323245378, "learning_rate": 7.464595583179544e-05, "loss": 0.4126, "step": 9176 }, { "epoch": 4.318588235294118, "grad_norm": 0.45358696140044386, "learning_rate": 7.460805573220779e-05, "loss": 0.2108, "step": 9177 }, { "epoch": 4.319058823529412, "grad_norm": 0.49349580050235703, "learning_rate": 7.4570162071462e-05, "loss": 0.3517, "step": 9178 }, { "epoch": 4.319529411764706, "grad_norm": 0.8613512516994839, "learning_rate": 7.453227485279432e-05, "loss": 0.1665, "step": 9179 }, { "epoch": 4.32, "grad_norm": 0.35183695695288075, "learning_rate": 7.449439407944052e-05, "loss": 0.1853, "step": 9180 }, { "epoch": 4.3204705882352945, "grad_norm": 0.5336678942161374, "learning_rate": 7.445651975463588e-05, "loss": 0.3559, "step": 9181 }, { "epoch": 4.320941176470588, "grad_norm": 0.26107119292794706, "learning_rate": 7.441865188161494e-05, "loss": 0.0967, "step": 9182 }, { "epoch": 4.321411764705882, "grad_norm": 0.2752804763091589, "learning_rate": 7.43807904636119e-05, "loss": 0.0766, "step": 9183 }, { "epoch": 4.321882352941176, "grad_norm": 0.4924808572305183, "learning_rate": 7.434293550386021e-05, "loss": 0.3263, "step": 9184 }, { "epoch": 4.3223529411764705, "grad_norm": 0.3900809193275679, "learning_rate": 7.430508700559292e-05, "loss": 0.0711, "step": 9185 }, { "epoch": 4.322823529411765, "grad_norm": 0.47274690809760117, "learning_rate": 7.426724497204254e-05, "loss": 0.3, "step": 9186 }, { "epoch": 4.323294117647059, "grad_norm": 0.4333957835042665, "learning_rate": 7.422940940644085e-05, "loss": 0.19, "step": 9187 }, { "epoch": 4.323764705882353, "grad_norm": 0.3163964346316881, "learning_rate": 7.419158031201924e-05, "loss": 0.0863, "step": 9188 }, { "epoch": 4.324235294117647, "grad_norm": 0.3447740056645644, "learning_rate": 7.415375769200859e-05, "loss": 0.074, "step": 9189 }, { "epoch": 4.324705882352941, "grad_norm": 0.3924086153971744, "learning_rate": 7.411594154963897e-05, "loss": 0.2979, "step": 9190 }, { "epoch": 4.325176470588235, "grad_norm": 0.4540950537227373, "learning_rate": 7.407813188814017e-05, "loss": 0.3486, "step": 9191 }, { "epoch": 4.325647058823529, "grad_norm": 0.34217711829841785, "learning_rate": 7.404032871074135e-05, "loss": 0.1686, "step": 9192 }, { "epoch": 4.326117647058823, "grad_norm": 0.3753019249697412, "learning_rate": 7.400253202067101e-05, "loss": 0.2993, "step": 9193 }, { "epoch": 4.326588235294118, "grad_norm": 0.4350490283853139, "learning_rate": 7.396474182115717e-05, "loss": 0.193, "step": 9194 }, { "epoch": 4.327058823529412, "grad_norm": 0.24336964531873928, "learning_rate": 7.392695811542739e-05, "loss": 0.1041, "step": 9195 }, { "epoch": 4.327529411764706, "grad_norm": 0.3891336885862207, "learning_rate": 7.388918090670847e-05, "loss": 0.189, "step": 9196 }, { "epoch": 4.328, "grad_norm": 0.6061826783515953, "learning_rate": 7.385141019822683e-05, "loss": 0.2574, "step": 9197 }, { "epoch": 4.3284705882352945, "grad_norm": 0.2680530203130005, "learning_rate": 7.381364599320831e-05, "loss": 0.0923, "step": 9198 }, { "epoch": 4.328941176470588, "grad_norm": 0.47299765707115937, "learning_rate": 7.377588829487806e-05, "loss": 0.3184, "step": 9199 }, { "epoch": 4.329411764705882, "grad_norm": 0.3103325594566574, "learning_rate": 7.373813710646083e-05, "loss": 0.1251, "step": 9200 }, { "epoch": 4.329882352941176, "grad_norm": 0.21799945757271774, "learning_rate": 7.370039243118082e-05, "loss": 0.0796, "step": 9201 }, { "epoch": 4.3303529411764705, "grad_norm": 0.47329970227291107, "learning_rate": 7.366265427226148e-05, "loss": 0.2254, "step": 9202 }, { "epoch": 4.330823529411765, "grad_norm": 0.4283046164962629, "learning_rate": 7.36249226329259e-05, "loss": 0.3544, "step": 9203 }, { "epoch": 4.331294117647059, "grad_norm": 0.42461451390638394, "learning_rate": 7.358719751639659e-05, "loss": 0.2276, "step": 9204 }, { "epoch": 4.331764705882353, "grad_norm": 0.3171507533108276, "learning_rate": 7.354947892589536e-05, "loss": 0.1358, "step": 9205 }, { "epoch": 4.332235294117647, "grad_norm": 0.5549608323848725, "learning_rate": 7.351176686464368e-05, "loss": 0.214, "step": 9206 }, { "epoch": 4.332705882352942, "grad_norm": 0.8684779265599093, "learning_rate": 7.347406133586223e-05, "loss": 0.3444, "step": 9207 }, { "epoch": 4.333176470588235, "grad_norm": 0.37363992139726393, "learning_rate": 7.34363623427713e-05, "loss": 0.2737, "step": 9208 }, { "epoch": 4.333647058823529, "grad_norm": 0.36857335615087833, "learning_rate": 7.339866988859061e-05, "loss": 0.2763, "step": 9209 }, { "epoch": 4.334117647058823, "grad_norm": 0.36422979270162875, "learning_rate": 7.33609839765392e-05, "loss": 0.1915, "step": 9210 }, { "epoch": 4.334588235294118, "grad_norm": 0.16935703849379166, "learning_rate": 7.332330460983565e-05, "loss": 0.0447, "step": 9211 }, { "epoch": 4.335058823529412, "grad_norm": 0.18236422977828443, "learning_rate": 7.328563179169806e-05, "loss": 0.0634, "step": 9212 }, { "epoch": 4.335529411764706, "grad_norm": 0.4772299289400878, "learning_rate": 7.324796552534373e-05, "loss": 0.354, "step": 9213 }, { "epoch": 4.336, "grad_norm": 0.261959511456514, "learning_rate": 7.321030581398961e-05, "loss": 0.1115, "step": 9214 }, { "epoch": 4.3364705882352945, "grad_norm": 0.42380786107978646, "learning_rate": 7.317265266085209e-05, "loss": 0.2346, "step": 9215 }, { "epoch": 4.336941176470588, "grad_norm": 0.3001371876705887, "learning_rate": 7.31350060691468e-05, "loss": 0.1627, "step": 9216 }, { "epoch": 4.337411764705882, "grad_norm": 0.6338438648408989, "learning_rate": 7.309736604208902e-05, "loss": 0.4524, "step": 9217 }, { "epoch": 4.337882352941176, "grad_norm": 0.17974434623545116, "learning_rate": 7.305973258289343e-05, "loss": 0.0391, "step": 9218 }, { "epoch": 4.3383529411764705, "grad_norm": 0.3177577140782787, "learning_rate": 7.302210569477397e-05, "loss": 0.1312, "step": 9219 }, { "epoch": 4.338823529411765, "grad_norm": 0.7647968346449819, "learning_rate": 7.298448538094436e-05, "loss": 0.5437, "step": 9220 }, { "epoch": 4.339294117647059, "grad_norm": 0.5115682661546929, "learning_rate": 7.294687164461741e-05, "loss": 0.2428, "step": 9221 }, { "epoch": 4.339764705882353, "grad_norm": 0.5062624727870335, "learning_rate": 7.290926448900557e-05, "loss": 0.4247, "step": 9222 }, { "epoch": 4.340235294117647, "grad_norm": 0.5323131978929614, "learning_rate": 7.287166391732071e-05, "loss": 0.2829, "step": 9223 }, { "epoch": 4.340705882352941, "grad_norm": 0.567720970516075, "learning_rate": 7.283406993277401e-05, "loss": 0.3319, "step": 9224 }, { "epoch": 4.341176470588235, "grad_norm": 0.4141135807253662, "learning_rate": 7.279648253857626e-05, "loss": 0.2798, "step": 9225 }, { "epoch": 4.341647058823529, "grad_norm": 0.34581544364681394, "learning_rate": 7.275890173793761e-05, "loss": 0.1997, "step": 9226 }, { "epoch": 4.342117647058823, "grad_norm": 0.6253789585654371, "learning_rate": 7.272132753406757e-05, "loss": 0.4566, "step": 9227 }, { "epoch": 4.342588235294118, "grad_norm": 0.28064351408755517, "learning_rate": 7.26837599301752e-05, "loss": 0.1409, "step": 9228 }, { "epoch": 4.343058823529412, "grad_norm": 0.18883283436625073, "learning_rate": 7.264619892946904e-05, "loss": 0.0745, "step": 9229 }, { "epoch": 4.343529411764706, "grad_norm": 0.4022965890293831, "learning_rate": 7.260864453515683e-05, "loss": 0.3252, "step": 9230 }, { "epoch": 4.344, "grad_norm": 0.29407952265149817, "learning_rate": 7.2571096750446e-05, "loss": 0.1753, "step": 9231 }, { "epoch": 4.3444705882352945, "grad_norm": 0.5794711811801193, "learning_rate": 7.253355557854334e-05, "loss": 0.2227, "step": 9232 }, { "epoch": 4.344941176470588, "grad_norm": 0.2892887367452355, "learning_rate": 7.249602102265496e-05, "loss": 0.0726, "step": 9233 }, { "epoch": 4.345411764705882, "grad_norm": 0.34832031933817964, "learning_rate": 7.245849308598658e-05, "loss": 0.214, "step": 9234 }, { "epoch": 4.345882352941176, "grad_norm": 0.39064109333103003, "learning_rate": 7.242097177174318e-05, "loss": 0.2263, "step": 9235 }, { "epoch": 4.3463529411764705, "grad_norm": 0.1149800699325941, "learning_rate": 7.238345708312931e-05, "loss": 0.0444, "step": 9236 }, { "epoch": 4.346823529411765, "grad_norm": 0.3038789997479818, "learning_rate": 7.234594902334897e-05, "loss": 0.1906, "step": 9237 }, { "epoch": 4.347294117647059, "grad_norm": 0.2617176885133183, "learning_rate": 7.230844759560544e-05, "loss": 0.1543, "step": 9238 }, { "epoch": 4.347764705882353, "grad_norm": 0.20416315135777827, "learning_rate": 7.227095280310152e-05, "loss": 0.0961, "step": 9239 }, { "epoch": 4.348235294117647, "grad_norm": 0.4389381467210363, "learning_rate": 7.223346464903958e-05, "loss": 0.3681, "step": 9240 }, { "epoch": 4.348705882352942, "grad_norm": 0.5808678069653032, "learning_rate": 7.219598313662113e-05, "loss": 0.5503, "step": 9241 }, { "epoch": 4.349176470588235, "grad_norm": 0.23894327686931224, "learning_rate": 7.215850826904734e-05, "loss": 0.0691, "step": 9242 }, { "epoch": 4.349647058823529, "grad_norm": 0.17777907127128786, "learning_rate": 7.212104004951883e-05, "loss": 0.063, "step": 9243 }, { "epoch": 4.350117647058823, "grad_norm": 0.32651429400440807, "learning_rate": 7.208357848123543e-05, "loss": 0.1842, "step": 9244 }, { "epoch": 4.350588235294118, "grad_norm": 0.3264654356145811, "learning_rate": 7.204612356739658e-05, "loss": 0.1983, "step": 9245 }, { "epoch": 4.351058823529412, "grad_norm": 0.4498605228753802, "learning_rate": 7.200867531120124e-05, "loss": 0.3114, "step": 9246 }, { "epoch": 4.351529411764706, "grad_norm": 0.1667224541872359, "learning_rate": 7.197123371584749e-05, "loss": 0.0499, "step": 9247 }, { "epoch": 4.352, "grad_norm": 0.19801532903910396, "learning_rate": 7.193379878453312e-05, "loss": 0.0591, "step": 9248 }, { "epoch": 4.352470588235294, "grad_norm": 0.5910554935639193, "learning_rate": 7.189637052045529e-05, "loss": 0.3403, "step": 9249 }, { "epoch": 4.352941176470588, "grad_norm": 0.09334547088065925, "learning_rate": 7.185894892681048e-05, "loss": 0.016, "step": 9250 }, { "epoch": 4.353411764705882, "grad_norm": 0.4103451361604251, "learning_rate": 7.182153400679468e-05, "loss": 0.2417, "step": 9251 }, { "epoch": 4.353882352941176, "grad_norm": 0.10503846194609144, "learning_rate": 7.17841257636034e-05, "loss": 0.0158, "step": 9252 }, { "epoch": 4.3543529411764705, "grad_norm": 0.31140021530766737, "learning_rate": 7.17467242004314e-05, "loss": 0.1708, "step": 9253 }, { "epoch": 4.354823529411765, "grad_norm": 0.29848062747364895, "learning_rate": 7.170932932047296e-05, "loss": 0.1762, "step": 9254 }, { "epoch": 4.355294117647059, "grad_norm": 0.5969281696156001, "learning_rate": 7.167194112692186e-05, "loss": 0.4426, "step": 9255 }, { "epoch": 4.355764705882353, "grad_norm": 0.5143098539791577, "learning_rate": 7.163455962297116e-05, "loss": 0.434, "step": 9256 }, { "epoch": 4.356235294117647, "grad_norm": 0.4287501455194709, "learning_rate": 7.159718481181346e-05, "loss": 0.2601, "step": 9257 }, { "epoch": 4.356705882352941, "grad_norm": 0.2570046389739899, "learning_rate": 7.155981669664072e-05, "loss": 0.106, "step": 9258 }, { "epoch": 4.357176470588235, "grad_norm": 0.3239325018499754, "learning_rate": 7.152245528064437e-05, "loss": 0.178, "step": 9259 }, { "epoch": 4.357647058823529, "grad_norm": 0.7748859140257874, "learning_rate": 7.148510056701532e-05, "loss": 0.3884, "step": 9260 }, { "epoch": 4.358117647058823, "grad_norm": 0.4853278379191561, "learning_rate": 7.144775255894375e-05, "loss": 0.1985, "step": 9261 }, { "epoch": 4.358588235294118, "grad_norm": 0.24485616554968326, "learning_rate": 7.14104112596194e-05, "loss": 0.0701, "step": 9262 }, { "epoch": 4.359058823529412, "grad_norm": 0.37524716682773546, "learning_rate": 7.137307667223141e-05, "loss": 0.215, "step": 9263 }, { "epoch": 4.359529411764706, "grad_norm": 0.35601046533910535, "learning_rate": 7.133574879996833e-05, "loss": 0.1389, "step": 9264 }, { "epoch": 4.36, "grad_norm": 0.3012009092451043, "learning_rate": 7.129842764601821e-05, "loss": 0.1491, "step": 9265 }, { "epoch": 4.3604705882352945, "grad_norm": 0.3950040889540827, "learning_rate": 7.126111321356834e-05, "loss": 0.2805, "step": 9266 }, { "epoch": 4.360941176470588, "grad_norm": 0.5946491868031066, "learning_rate": 7.122380550580562e-05, "loss": 0.206, "step": 9267 }, { "epoch": 4.361411764705882, "grad_norm": 0.49209436274881363, "learning_rate": 7.118650452591634e-05, "loss": 0.3993, "step": 9268 }, { "epoch": 4.361882352941176, "grad_norm": 0.33778523969995966, "learning_rate": 7.114921027708613e-05, "loss": 0.1895, "step": 9269 }, { "epoch": 4.3623529411764705, "grad_norm": 0.5466759450959054, "learning_rate": 7.111192276250011e-05, "loss": 0.3645, "step": 9270 }, { "epoch": 4.362823529411765, "grad_norm": 0.3639946758250032, "learning_rate": 7.107464198534287e-05, "loss": 0.2566, "step": 9271 }, { "epoch": 4.363294117647059, "grad_norm": 0.4721630362314604, "learning_rate": 7.103736794879829e-05, "loss": 0.2038, "step": 9272 }, { "epoch": 4.363764705882353, "grad_norm": 0.2990151754843816, "learning_rate": 7.100010065604978e-05, "loss": 0.1235, "step": 9273 }, { "epoch": 4.364235294117647, "grad_norm": 0.4297739770976193, "learning_rate": 7.096284011028024e-05, "loss": 0.1327, "step": 9274 }, { "epoch": 4.364705882352941, "grad_norm": 0.3435133628973851, "learning_rate": 7.092558631467177e-05, "loss": 0.2535, "step": 9275 }, { "epoch": 4.365176470588235, "grad_norm": 0.4990288092495262, "learning_rate": 7.088833927240608e-05, "loss": 0.2085, "step": 9276 }, { "epoch": 4.365647058823529, "grad_norm": 0.6039484740510674, "learning_rate": 7.085109898666433e-05, "loss": 0.3941, "step": 9277 }, { "epoch": 4.366117647058823, "grad_norm": 0.4382576464951002, "learning_rate": 7.081386546062687e-05, "loss": 0.3107, "step": 9278 }, { "epoch": 4.366588235294118, "grad_norm": 0.27654885569218424, "learning_rate": 7.077663869747372e-05, "loss": 0.1329, "step": 9279 }, { "epoch": 4.367058823529412, "grad_norm": 0.4197554803041881, "learning_rate": 7.073941870038427e-05, "loss": 0.2753, "step": 9280 }, { "epoch": 4.367529411764706, "grad_norm": 0.365709445977552, "learning_rate": 7.070220547253717e-05, "loss": 0.223, "step": 9281 }, { "epoch": 4.368, "grad_norm": 0.501606490129437, "learning_rate": 7.066499901711068e-05, "loss": 0.2008, "step": 9282 }, { "epoch": 4.368470588235294, "grad_norm": 1.0037344868206048, "learning_rate": 7.062779933728247e-05, "loss": 0.6051, "step": 9283 }, { "epoch": 4.368941176470588, "grad_norm": 0.4100111960145308, "learning_rate": 7.059060643622944e-05, "loss": 0.3026, "step": 9284 }, { "epoch": 4.369411764705882, "grad_norm": 0.5398278925750948, "learning_rate": 7.05534203171282e-05, "loss": 0.4338, "step": 9285 }, { "epoch": 4.369882352941176, "grad_norm": 0.232839390364319, "learning_rate": 7.051624098315449e-05, "loss": 0.079, "step": 9286 }, { "epoch": 4.3703529411764706, "grad_norm": 0.40610847822032325, "learning_rate": 7.047906843748365e-05, "loss": 0.1654, "step": 9287 }, { "epoch": 4.370823529411765, "grad_norm": 0.31540493156471777, "learning_rate": 7.044190268329049e-05, "loss": 0.2309, "step": 9288 }, { "epoch": 4.371294117647059, "grad_norm": 0.14957481639834308, "learning_rate": 7.0404743723749e-05, "loss": 0.0386, "step": 9289 }, { "epoch": 4.371764705882353, "grad_norm": 0.43169276862223993, "learning_rate": 7.036759156203281e-05, "loss": 0.1844, "step": 9290 }, { "epoch": 4.372235294117647, "grad_norm": 0.25959343899977355, "learning_rate": 7.033044620131495e-05, "loss": 0.1011, "step": 9291 }, { "epoch": 4.372705882352941, "grad_norm": 0.37592742358068687, "learning_rate": 7.02933076447677e-05, "loss": 0.1443, "step": 9292 }, { "epoch": 4.373176470588235, "grad_norm": 0.24931700120491554, "learning_rate": 7.025617589556297e-05, "loss": 0.0946, "step": 9293 }, { "epoch": 4.373647058823529, "grad_norm": 0.3528480363891976, "learning_rate": 7.021905095687198e-05, "loss": 0.2074, "step": 9294 }, { "epoch": 4.374117647058823, "grad_norm": 0.5893949257757056, "learning_rate": 7.018193283186531e-05, "loss": 0.4254, "step": 9295 }, { "epoch": 4.374588235294118, "grad_norm": 0.30527173330777047, "learning_rate": 7.01448215237131e-05, "loss": 0.1504, "step": 9296 }, { "epoch": 4.375058823529412, "grad_norm": 0.3735680440087188, "learning_rate": 7.010771703558488e-05, "loss": 0.094, "step": 9297 }, { "epoch": 4.375529411764706, "grad_norm": 0.24467278852075677, "learning_rate": 7.007061937064941e-05, "loss": 0.0385, "step": 9298 }, { "epoch": 4.376, "grad_norm": 0.4660945561625529, "learning_rate": 7.00335285320751e-05, "loss": 0.2674, "step": 9299 }, { "epoch": 4.376470588235295, "grad_norm": 0.24166285575008484, "learning_rate": 6.999644452302975e-05, "loss": 0.1058, "step": 9300 }, { "epoch": 4.376941176470588, "grad_norm": 0.4857913120719804, "learning_rate": 6.995936734668042e-05, "loss": 0.3252, "step": 9301 }, { "epoch": 4.377411764705882, "grad_norm": 0.30980097375875704, "learning_rate": 6.992229700619369e-05, "loss": 0.1437, "step": 9302 }, { "epoch": 4.377882352941176, "grad_norm": 0.2961094787041119, "learning_rate": 6.988523350473563e-05, "loss": 0.0895, "step": 9303 }, { "epoch": 4.378352941176471, "grad_norm": 0.31792741972985405, "learning_rate": 6.984817684547155e-05, "loss": 0.1539, "step": 9304 }, { "epoch": 4.378823529411765, "grad_norm": 0.29311477741730946, "learning_rate": 6.981112703156628e-05, "loss": 0.1495, "step": 9305 }, { "epoch": 4.379294117647059, "grad_norm": 0.5165008792528458, "learning_rate": 6.977408406618418e-05, "loss": 0.2655, "step": 9306 }, { "epoch": 4.379764705882353, "grad_norm": 0.41745578310446024, "learning_rate": 6.973704795248871e-05, "loss": 0.2327, "step": 9307 }, { "epoch": 4.3802352941176474, "grad_norm": 0.28057234443518675, "learning_rate": 6.970001869364306e-05, "loss": 0.0711, "step": 9308 }, { "epoch": 4.380705882352941, "grad_norm": 0.1557341614986796, "learning_rate": 6.966299629280968e-05, "loss": 0.0474, "step": 9309 }, { "epoch": 4.381176470588235, "grad_norm": 0.4879147008413534, "learning_rate": 6.962598075315046e-05, "loss": 0.3868, "step": 9310 }, { "epoch": 4.381647058823529, "grad_norm": 0.4264308400639079, "learning_rate": 6.958897207782677e-05, "loss": 0.2592, "step": 9311 }, { "epoch": 4.3821176470588235, "grad_norm": 0.43730146799298875, "learning_rate": 6.955197026999925e-05, "loss": 0.1808, "step": 9312 }, { "epoch": 4.382588235294118, "grad_norm": 0.36794601021331735, "learning_rate": 6.951497533282809e-05, "loss": 0.1078, "step": 9313 }, { "epoch": 4.383058823529412, "grad_norm": 0.6556458907227666, "learning_rate": 6.947798726947277e-05, "loss": 0.3506, "step": 9314 }, { "epoch": 4.383529411764706, "grad_norm": 0.4251110981877861, "learning_rate": 6.94410060830923e-05, "loss": 0.2801, "step": 9315 }, { "epoch": 4.384, "grad_norm": 0.3525893652857641, "learning_rate": 6.940403177684509e-05, "loss": 0.1703, "step": 9316 }, { "epoch": 4.384470588235294, "grad_norm": 0.4623958348838427, "learning_rate": 6.936706435388882e-05, "loss": 0.2377, "step": 9317 }, { "epoch": 4.384941176470588, "grad_norm": 0.2878912864810637, "learning_rate": 6.93301038173808e-05, "loss": 0.1512, "step": 9318 }, { "epoch": 4.385411764705882, "grad_norm": 0.2502933650920456, "learning_rate": 6.929315017047762e-05, "loss": 0.1037, "step": 9319 }, { "epoch": 4.385882352941176, "grad_norm": 0.0860840192643513, "learning_rate": 6.925620341633522e-05, "loss": 0.0186, "step": 9320 }, { "epoch": 4.386352941176471, "grad_norm": 0.4656934325136332, "learning_rate": 6.921926355810911e-05, "loss": 0.1874, "step": 9321 }, { "epoch": 4.386823529411765, "grad_norm": 0.2598779639598526, "learning_rate": 6.918233059895418e-05, "loss": 0.0936, "step": 9322 }, { "epoch": 4.387294117647059, "grad_norm": 0.2542694681118305, "learning_rate": 6.914540454202457e-05, "loss": 0.0595, "step": 9323 }, { "epoch": 4.387764705882353, "grad_norm": 0.35377235077292496, "learning_rate": 6.910848539047399e-05, "loss": 0.2716, "step": 9324 }, { "epoch": 4.3882352941176475, "grad_norm": 0.0594217898962022, "learning_rate": 6.907157314745558e-05, "loss": 0.0077, "step": 9325 }, { "epoch": 4.388705882352941, "grad_norm": 0.3932988176353832, "learning_rate": 6.903466781612173e-05, "loss": 0.2719, "step": 9326 }, { "epoch": 4.389176470588235, "grad_norm": 0.21628345531872692, "learning_rate": 6.899776939962438e-05, "loss": 0.0451, "step": 9327 }, { "epoch": 4.389647058823529, "grad_norm": 0.34483533287229756, "learning_rate": 6.89608779011149e-05, "loss": 0.1539, "step": 9328 }, { "epoch": 4.3901176470588235, "grad_norm": 0.42227865256043323, "learning_rate": 6.892399332374387e-05, "loss": 0.2171, "step": 9329 }, { "epoch": 4.390588235294118, "grad_norm": 0.340017993536581, "learning_rate": 6.888711567066149e-05, "loss": 0.1187, "step": 9330 }, { "epoch": 4.391058823529412, "grad_norm": 0.7485423390979936, "learning_rate": 6.885024494501735e-05, "loss": 0.246, "step": 9331 }, { "epoch": 4.391529411764706, "grad_norm": 0.31550494421833275, "learning_rate": 6.881338114996027e-05, "loss": 0.1573, "step": 9332 }, { "epoch": 4.392, "grad_norm": 0.24078917609189646, "learning_rate": 6.877652428863866e-05, "loss": 0.0727, "step": 9333 }, { "epoch": 4.392470588235295, "grad_norm": 0.4168082051762734, "learning_rate": 6.873967436420033e-05, "loss": 0.2226, "step": 9334 }, { "epoch": 4.392941176470588, "grad_norm": 0.5315551497602501, "learning_rate": 6.870283137979233e-05, "loss": 0.2257, "step": 9335 }, { "epoch": 4.393411764705882, "grad_norm": 0.24399400324630183, "learning_rate": 6.866599533856135e-05, "loss": 0.0682, "step": 9336 }, { "epoch": 4.393882352941176, "grad_norm": 0.42424188803422586, "learning_rate": 6.862916624365325e-05, "loss": 0.2275, "step": 9337 }, { "epoch": 4.394352941176471, "grad_norm": 0.14416867918296744, "learning_rate": 6.859234409821347e-05, "loss": 0.0312, "step": 9338 }, { "epoch": 4.394823529411765, "grad_norm": 0.12810539488832628, "learning_rate": 6.855552890538688e-05, "loss": 0.0253, "step": 9339 }, { "epoch": 4.395294117647059, "grad_norm": 0.4635347740374137, "learning_rate": 6.851872066831751e-05, "loss": 0.2575, "step": 9340 }, { "epoch": 4.395764705882353, "grad_norm": 0.4396819193852414, "learning_rate": 6.848191939014909e-05, "loss": 0.3176, "step": 9341 }, { "epoch": 4.3962352941176475, "grad_norm": 0.24935202474090676, "learning_rate": 6.844512507402465e-05, "loss": 0.1022, "step": 9342 }, { "epoch": 4.396705882352941, "grad_norm": 0.24928233762882623, "learning_rate": 6.84083377230865e-05, "loss": 0.0286, "step": 9343 }, { "epoch": 4.397176470588235, "grad_norm": 0.27828886258241997, "learning_rate": 6.837155734047652e-05, "loss": 0.0781, "step": 9344 }, { "epoch": 4.397647058823529, "grad_norm": 0.31328101059458574, "learning_rate": 6.833478392933597e-05, "loss": 0.1463, "step": 9345 }, { "epoch": 4.3981176470588235, "grad_norm": 0.1786327120520047, "learning_rate": 6.829801749280537e-05, "loss": 0.0911, "step": 9346 }, { "epoch": 4.398588235294118, "grad_norm": 0.41220060533038366, "learning_rate": 6.826125803402484e-05, "loss": 0.2984, "step": 9347 }, { "epoch": 4.399058823529412, "grad_norm": 0.16190613648157512, "learning_rate": 6.822450555613385e-05, "loss": 0.0302, "step": 9348 }, { "epoch": 4.399529411764706, "grad_norm": 0.4485753966642052, "learning_rate": 6.818776006227114e-05, "loss": 0.2221, "step": 9349 }, { "epoch": 4.4, "grad_norm": 0.366234708473822, "learning_rate": 6.815102155557501e-05, "loss": 0.2374, "step": 9350 }, { "epoch": 4.400470588235294, "grad_norm": 0.4270083197359331, "learning_rate": 6.811429003918317e-05, "loss": 0.2886, "step": 9351 }, { "epoch": 4.400941176470588, "grad_norm": 0.2223164903266245, "learning_rate": 6.807756551623248e-05, "loss": 0.0863, "step": 9352 }, { "epoch": 4.401411764705882, "grad_norm": 0.44948391340817584, "learning_rate": 6.804084798985964e-05, "loss": 0.3428, "step": 9353 }, { "epoch": 4.401882352941176, "grad_norm": 0.13229949169578442, "learning_rate": 6.800413746320034e-05, "loss": 0.0367, "step": 9354 }, { "epoch": 4.402352941176471, "grad_norm": 0.34785137580936315, "learning_rate": 6.796743393938987e-05, "loss": 0.207, "step": 9355 }, { "epoch": 4.402823529411765, "grad_norm": 0.33443697719490223, "learning_rate": 6.793073742156298e-05, "loss": 0.2327, "step": 9356 }, { "epoch": 4.403294117647059, "grad_norm": 0.5765769500080089, "learning_rate": 6.78940479128536e-05, "loss": 0.3337, "step": 9357 }, { "epoch": 4.403764705882353, "grad_norm": 0.396606631086207, "learning_rate": 6.785736541639526e-05, "loss": 0.0991, "step": 9358 }, { "epoch": 4.4042352941176475, "grad_norm": 0.24112598462561957, "learning_rate": 6.782068993532085e-05, "loss": 0.0762, "step": 9359 }, { "epoch": 4.404705882352941, "grad_norm": 0.21362944061410566, "learning_rate": 6.778402147276259e-05, "loss": 0.0898, "step": 9360 }, { "epoch": 4.405176470588235, "grad_norm": 0.4290197159684095, "learning_rate": 6.774736003185215e-05, "loss": 0.2963, "step": 9361 }, { "epoch": 4.405647058823529, "grad_norm": 0.560632930879558, "learning_rate": 6.771070561572067e-05, "loss": 0.4015, "step": 9362 }, { "epoch": 4.4061176470588235, "grad_norm": 0.2549634076528438, "learning_rate": 6.767405822749848e-05, "loss": 0.137, "step": 9363 }, { "epoch": 4.406588235294118, "grad_norm": 0.35832401595765784, "learning_rate": 6.76374178703156e-05, "loss": 0.2406, "step": 9364 }, { "epoch": 4.407058823529412, "grad_norm": 0.2083398758989964, "learning_rate": 6.760078454730115e-05, "loss": 0.0543, "step": 9365 }, { "epoch": 4.407529411764706, "grad_norm": 0.6026260619279857, "learning_rate": 6.756415826158388e-05, "loss": 0.4731, "step": 9366 }, { "epoch": 4.408, "grad_norm": 0.3360156731680111, "learning_rate": 6.752753901629189e-05, "loss": 0.2079, "step": 9367 }, { "epoch": 4.408470588235295, "grad_norm": 0.28493290317386977, "learning_rate": 6.749092681455255e-05, "loss": 0.0959, "step": 9368 }, { "epoch": 4.408941176470588, "grad_norm": 0.3806311069255406, "learning_rate": 6.745432165949274e-05, "loss": 0.2753, "step": 9369 }, { "epoch": 4.409411764705882, "grad_norm": 0.4473255255772565, "learning_rate": 6.741772355423882e-05, "loss": 0.2646, "step": 9370 }, { "epoch": 4.409882352941176, "grad_norm": 0.26917019994612873, "learning_rate": 6.73811325019163e-05, "loss": 0.1458, "step": 9371 }, { "epoch": 4.410352941176471, "grad_norm": 0.22195882787990806, "learning_rate": 6.734454850565031e-05, "loss": 0.0644, "step": 9372 }, { "epoch": 4.410823529411765, "grad_norm": 0.37601296209847573, "learning_rate": 6.730797156856537e-05, "loss": 0.21, "step": 9373 }, { "epoch": 4.411294117647059, "grad_norm": 0.40713147257037063, "learning_rate": 6.727140169378521e-05, "loss": 0.2925, "step": 9374 }, { "epoch": 4.411764705882353, "grad_norm": 0.28611325498845885, "learning_rate": 6.72348388844331e-05, "loss": 0.1864, "step": 9375 }, { "epoch": 4.412235294117647, "grad_norm": 0.4733601256754119, "learning_rate": 6.719828314363178e-05, "loss": 0.3342, "step": 9376 }, { "epoch": 4.412705882352941, "grad_norm": 0.39941978079951734, "learning_rate": 6.716173447450317e-05, "loss": 0.2484, "step": 9377 }, { "epoch": 4.413176470588235, "grad_norm": 0.4040724211083026, "learning_rate": 6.712519288016875e-05, "loss": 0.2677, "step": 9378 }, { "epoch": 4.413647058823529, "grad_norm": 0.21807176904994832, "learning_rate": 6.70886583637494e-05, "loss": 0.0803, "step": 9379 }, { "epoch": 4.4141176470588235, "grad_norm": 0.36481537159780697, "learning_rate": 6.705213092836527e-05, "loss": 0.1444, "step": 9380 }, { "epoch": 4.414588235294118, "grad_norm": 0.1935528459594326, "learning_rate": 6.701561057713599e-05, "loss": 0.0587, "step": 9381 }, { "epoch": 4.415058823529412, "grad_norm": 0.3984144684359909, "learning_rate": 6.697909731318069e-05, "loss": 0.1687, "step": 9382 }, { "epoch": 4.415529411764706, "grad_norm": 0.5111470956734829, "learning_rate": 6.694259113961761e-05, "loss": 0.2502, "step": 9383 }, { "epoch": 4.416, "grad_norm": 0.28326772098760045, "learning_rate": 6.690609205956464e-05, "loss": 0.1253, "step": 9384 }, { "epoch": 4.416470588235294, "grad_norm": 0.3827073006605048, "learning_rate": 6.686960007613906e-05, "loss": 0.2431, "step": 9385 }, { "epoch": 4.416941176470588, "grad_norm": 0.5006482709342218, "learning_rate": 6.683311519245733e-05, "loss": 0.3631, "step": 9386 }, { "epoch": 4.417411764705882, "grad_norm": 0.460416313871562, "learning_rate": 6.679663741163553e-05, "loss": 0.3389, "step": 9387 }, { "epoch": 4.417882352941176, "grad_norm": 0.5131819598880042, "learning_rate": 6.676016673678896e-05, "loss": 0.4084, "step": 9388 }, { "epoch": 4.418352941176471, "grad_norm": 0.3738765023172249, "learning_rate": 6.672370317103245e-05, "loss": 0.1219, "step": 9389 }, { "epoch": 4.418823529411765, "grad_norm": 0.6303889763858891, "learning_rate": 6.668724671748018e-05, "loss": 0.1539, "step": 9390 }, { "epoch": 4.419294117647059, "grad_norm": 0.7327746039941505, "learning_rate": 6.665079737924566e-05, "loss": 0.6526, "step": 9391 }, { "epoch": 4.419764705882353, "grad_norm": 0.6581619704932556, "learning_rate": 6.661435515944186e-05, "loss": 0.3319, "step": 9392 }, { "epoch": 4.4202352941176475, "grad_norm": 0.587949892950433, "learning_rate": 6.65779200611812e-05, "loss": 0.377, "step": 9393 }, { "epoch": 4.420705882352941, "grad_norm": 0.3360250535001454, "learning_rate": 6.654149208757529e-05, "loss": 0.1622, "step": 9394 }, { "epoch": 4.421176470588235, "grad_norm": 0.38207836642163684, "learning_rate": 6.65050712417353e-05, "loss": 0.194, "step": 9395 }, { "epoch": 4.421647058823529, "grad_norm": 0.21133861440005378, "learning_rate": 6.646865752677186e-05, "loss": 0.0956, "step": 9396 }, { "epoch": 4.4221176470588235, "grad_norm": 0.3358486954812281, "learning_rate": 6.643225094579462e-05, "loss": 0.1296, "step": 9397 }, { "epoch": 4.422588235294118, "grad_norm": 0.5761275589291678, "learning_rate": 6.63958515019132e-05, "loss": 0.2716, "step": 9398 }, { "epoch": 4.423058823529412, "grad_norm": 0.6781418148258747, "learning_rate": 6.635945919823607e-05, "loss": 0.4201, "step": 9399 }, { "epoch": 4.423529411764706, "grad_norm": 0.13501754864154852, "learning_rate": 6.632307403787138e-05, "loss": 0.0384, "step": 9400 }, { "epoch": 4.424, "grad_norm": 0.3049547877742487, "learning_rate": 6.628669602392665e-05, "loss": 0.1263, "step": 9401 }, { "epoch": 4.424470588235294, "grad_norm": 0.4210280344255968, "learning_rate": 6.625032515950865e-05, "loss": 0.2549, "step": 9402 }, { "epoch": 4.424941176470588, "grad_norm": 0.3319399959427973, "learning_rate": 6.621396144772365e-05, "loss": 0.132, "step": 9403 }, { "epoch": 4.425411764705882, "grad_norm": 0.39149512645657303, "learning_rate": 6.617760489167738e-05, "loss": 0.2179, "step": 9404 }, { "epoch": 4.425882352941176, "grad_norm": 0.40880339146486283, "learning_rate": 6.614125549447475e-05, "loss": 0.2483, "step": 9405 }, { "epoch": 4.426352941176471, "grad_norm": 0.33788259605830645, "learning_rate": 6.610491325922022e-05, "loss": 0.1529, "step": 9406 }, { "epoch": 4.426823529411765, "grad_norm": 0.4253341159774873, "learning_rate": 6.606857818901765e-05, "loss": 0.1332, "step": 9407 }, { "epoch": 4.427294117647059, "grad_norm": 0.929348920280671, "learning_rate": 6.603225028697013e-05, "loss": 0.6859, "step": 9408 }, { "epoch": 4.427764705882353, "grad_norm": 0.15635644391228648, "learning_rate": 6.59959295561803e-05, "loss": 0.051, "step": 9409 }, { "epoch": 4.428235294117647, "grad_norm": 0.3987598551846669, "learning_rate": 6.595961599975016e-05, "loss": 0.2425, "step": 9410 }, { "epoch": 4.428705882352941, "grad_norm": 0.2910935879908244, "learning_rate": 6.5923309620781e-05, "loss": 0.1376, "step": 9411 }, { "epoch": 4.429176470588235, "grad_norm": 0.3579175323371435, "learning_rate": 6.588701042237358e-05, "loss": 0.1822, "step": 9412 }, { "epoch": 4.429647058823529, "grad_norm": 0.32682652116870314, "learning_rate": 6.585071840762809e-05, "loss": 0.1754, "step": 9413 }, { "epoch": 4.4301176470588235, "grad_norm": 0.3356368783872655, "learning_rate": 6.581443357964395e-05, "loss": 0.1383, "step": 9414 }, { "epoch": 4.430588235294118, "grad_norm": 0.28385052135463174, "learning_rate": 6.577815594152017e-05, "loss": 0.1044, "step": 9415 }, { "epoch": 4.431058823529412, "grad_norm": 0.3765338556451384, "learning_rate": 6.574188549635493e-05, "loss": 0.1918, "step": 9416 }, { "epoch": 4.431529411764706, "grad_norm": 0.40199010397164425, "learning_rate": 6.570562224724593e-05, "loss": 0.2243, "step": 9417 }, { "epoch": 4.432, "grad_norm": 0.5715265831223163, "learning_rate": 6.566936619729033e-05, "loss": 0.2561, "step": 9418 }, { "epoch": 4.432470588235294, "grad_norm": 0.5016785998613699, "learning_rate": 6.563311734958444e-05, "loss": 0.3526, "step": 9419 }, { "epoch": 4.432941176470588, "grad_norm": 0.20964647624958968, "learning_rate": 6.559687570722413e-05, "loss": 0.0709, "step": 9420 }, { "epoch": 4.433411764705882, "grad_norm": 0.2385519995291404, "learning_rate": 6.556064127330467e-05, "loss": 0.1168, "step": 9421 }, { "epoch": 4.433882352941176, "grad_norm": 0.32258369767366246, "learning_rate": 6.552441405092058e-05, "loss": 0.1158, "step": 9422 }, { "epoch": 4.434352941176471, "grad_norm": 0.34269924032114185, "learning_rate": 6.54881940431659e-05, "loss": 0.1466, "step": 9423 }, { "epoch": 4.434823529411765, "grad_norm": 0.4665886194883694, "learning_rate": 6.5451981253134e-05, "loss": 0.1265, "step": 9424 }, { "epoch": 4.435294117647059, "grad_norm": 0.30044526280892264, "learning_rate": 6.541577568391758e-05, "loss": 0.0832, "step": 9425 }, { "epoch": 4.435764705882353, "grad_norm": 0.3342086900703397, "learning_rate": 6.537957733860879e-05, "loss": 0.1236, "step": 9426 }, { "epoch": 4.4362352941176475, "grad_norm": 0.5856252017122399, "learning_rate": 6.53433862202992e-05, "loss": 0.3315, "step": 9427 }, { "epoch": 4.436705882352941, "grad_norm": 0.1596013579201676, "learning_rate": 6.530720233207962e-05, "loss": 0.0552, "step": 9428 }, { "epoch": 4.437176470588235, "grad_norm": 0.36411691414405956, "learning_rate": 6.527102567704036e-05, "loss": 0.1132, "step": 9429 }, { "epoch": 4.437647058823529, "grad_norm": 0.0856014702796779, "learning_rate": 6.523485625827119e-05, "loss": 0.0206, "step": 9430 }, { "epoch": 4.4381176470588235, "grad_norm": 0.3122009319237332, "learning_rate": 6.519869407886098e-05, "loss": 0.1589, "step": 9431 }, { "epoch": 4.438588235294118, "grad_norm": 0.3872405718524053, "learning_rate": 6.516253914189825e-05, "loss": 0.3254, "step": 9432 }, { "epoch": 4.439058823529412, "grad_norm": 0.539391673117376, "learning_rate": 6.512639145047086e-05, "loss": 0.3093, "step": 9433 }, { "epoch": 4.439529411764706, "grad_norm": 0.5920843492742411, "learning_rate": 6.509025100766591e-05, "loss": 0.4311, "step": 9434 }, { "epoch": 4.44, "grad_norm": 0.39261153790200554, "learning_rate": 6.505411781656997e-05, "loss": 0.3619, "step": 9435 }, { "epoch": 4.440470588235294, "grad_norm": 0.2111929011968434, "learning_rate": 6.50179918802691e-05, "loss": 0.1152, "step": 9436 }, { "epoch": 4.440941176470588, "grad_norm": 0.4882153008320113, "learning_rate": 6.49818732018485e-05, "loss": 0.2571, "step": 9437 }, { "epoch": 4.441411764705882, "grad_norm": 0.24362868804652155, "learning_rate": 6.494576178439298e-05, "loss": 0.0733, "step": 9438 }, { "epoch": 4.441882352941176, "grad_norm": 0.3217154475558286, "learning_rate": 6.490965763098654e-05, "loss": 0.2585, "step": 9439 }, { "epoch": 4.442352941176471, "grad_norm": 0.21121985253368328, "learning_rate": 6.487356074471272e-05, "loss": 0.0556, "step": 9440 }, { "epoch": 4.442823529411765, "grad_norm": 0.34926748174412126, "learning_rate": 6.483747112865436e-05, "loss": 0.1831, "step": 9441 }, { "epoch": 4.443294117647059, "grad_norm": 0.1917488302884867, "learning_rate": 6.480138878589368e-05, "loss": 0.0491, "step": 9442 }, { "epoch": 4.443764705882353, "grad_norm": 0.4961222858374899, "learning_rate": 6.476531371951229e-05, "loss": 0.2806, "step": 9443 }, { "epoch": 4.444235294117647, "grad_norm": 0.29012104502631525, "learning_rate": 6.472924593259122e-05, "loss": 0.0979, "step": 9444 }, { "epoch": 4.444705882352941, "grad_norm": 0.5997264227538087, "learning_rate": 6.469318542821078e-05, "loss": 0.3236, "step": 9445 }, { "epoch": 4.445176470588235, "grad_norm": 0.3794781416200036, "learning_rate": 6.465713220945075e-05, "loss": 0.2556, "step": 9446 }, { "epoch": 4.445647058823529, "grad_norm": 0.49924416906316244, "learning_rate": 6.46210862793902e-05, "loss": 0.2073, "step": 9447 }, { "epoch": 4.4461176470588235, "grad_norm": 0.4800354308354845, "learning_rate": 6.458504764110764e-05, "loss": 0.307, "step": 9448 }, { "epoch": 4.446588235294118, "grad_norm": 0.30930869337761296, "learning_rate": 6.454901629768105e-05, "loss": 0.1197, "step": 9449 }, { "epoch": 4.447058823529412, "grad_norm": 0.584903183974139, "learning_rate": 6.451299225218754e-05, "loss": 0.3956, "step": 9450 }, { "epoch": 4.447529411764706, "grad_norm": 0.6178027860688005, "learning_rate": 6.447697550770381e-05, "loss": 0.2609, "step": 9451 }, { "epoch": 4.448, "grad_norm": 0.4398897500104653, "learning_rate": 6.44409660673059e-05, "loss": 0.1932, "step": 9452 }, { "epoch": 4.448470588235294, "grad_norm": 0.16242220013351377, "learning_rate": 6.440496393406912e-05, "loss": 0.0401, "step": 9453 }, { "epoch": 4.448941176470588, "grad_norm": 0.32858848792122236, "learning_rate": 6.436896911106824e-05, "loss": 0.1855, "step": 9454 }, { "epoch": 4.449411764705882, "grad_norm": 0.5793422117222803, "learning_rate": 6.433298160137749e-05, "loss": 0.432, "step": 9455 }, { "epoch": 4.449882352941176, "grad_norm": 0.30180374687679784, "learning_rate": 6.429700140807023e-05, "loss": 0.1653, "step": 9456 }, { "epoch": 4.450352941176471, "grad_norm": 0.5738345439353287, "learning_rate": 6.426102853421942e-05, "loss": 0.5369, "step": 9457 }, { "epoch": 4.450823529411765, "grad_norm": 0.30336983150441577, "learning_rate": 6.42250629828974e-05, "loss": 0.2091, "step": 9458 }, { "epoch": 4.451294117647059, "grad_norm": 0.49423068245424795, "learning_rate": 6.418910475717565e-05, "loss": 0.2789, "step": 9459 }, { "epoch": 4.451764705882353, "grad_norm": 0.40187914948199105, "learning_rate": 6.415315386012525e-05, "loss": 0.3241, "step": 9460 }, { "epoch": 4.4522352941176475, "grad_norm": 0.27805797108292396, "learning_rate": 6.411721029481665e-05, "loss": 0.1265, "step": 9461 }, { "epoch": 4.452705882352941, "grad_norm": 0.1998920161007016, "learning_rate": 6.40812740643195e-05, "loss": 0.0679, "step": 9462 }, { "epoch": 4.453176470588235, "grad_norm": 0.6225670875109499, "learning_rate": 6.404534517170296e-05, "loss": 0.287, "step": 9463 }, { "epoch": 4.453647058823529, "grad_norm": 0.4479330828087951, "learning_rate": 6.40094236200356e-05, "loss": 0.2323, "step": 9464 }, { "epoch": 4.4541176470588235, "grad_norm": 0.8967224983084368, "learning_rate": 6.397350941238519e-05, "loss": 0.2973, "step": 9465 }, { "epoch": 4.454588235294118, "grad_norm": 0.3057854655490065, "learning_rate": 6.39376025518191e-05, "loss": 0.1989, "step": 9466 }, { "epoch": 4.455058823529412, "grad_norm": 0.3576482965295583, "learning_rate": 6.390170304140382e-05, "loss": 0.1688, "step": 9467 }, { "epoch": 4.455529411764706, "grad_norm": 0.27027195560782447, "learning_rate": 6.386581088420542e-05, "loss": 0.0609, "step": 9468 }, { "epoch": 4.456, "grad_norm": 0.25314360261865765, "learning_rate": 6.382992608328929e-05, "loss": 0.0808, "step": 9469 }, { "epoch": 4.456470588235294, "grad_norm": 0.9202500771052379, "learning_rate": 6.37940486417201e-05, "loss": 0.4833, "step": 9470 }, { "epoch": 4.456941176470588, "grad_norm": 0.5846864148573233, "learning_rate": 6.375817856256199e-05, "loss": 0.3856, "step": 9471 }, { "epoch": 4.457411764705882, "grad_norm": 1.3422361901585267, "learning_rate": 6.37223158488785e-05, "loss": 0.6573, "step": 9472 }, { "epoch": 4.457882352941176, "grad_norm": 0.27153888244091945, "learning_rate": 6.368646050373236e-05, "loss": 0.1115, "step": 9473 }, { "epoch": 4.458352941176471, "grad_norm": 1.72761089935141, "learning_rate": 6.365061253018586e-05, "loss": 0.8845, "step": 9474 }, { "epoch": 4.458823529411765, "grad_norm": 0.4589058466691402, "learning_rate": 6.361477193130065e-05, "loss": 0.207, "step": 9475 }, { "epoch": 4.459294117647059, "grad_norm": 0.2262551162898333, "learning_rate": 6.35789387101376e-05, "loss": 0.0651, "step": 9476 }, { "epoch": 4.459764705882353, "grad_norm": 0.31034930253701243, "learning_rate": 6.354311286975705e-05, "loss": 0.091, "step": 9477 }, { "epoch": 4.460235294117647, "grad_norm": 0.3763312169282521, "learning_rate": 6.35072944132188e-05, "loss": 0.2113, "step": 9478 }, { "epoch": 4.460705882352941, "grad_norm": 0.20159629184610114, "learning_rate": 6.34714833435818e-05, "loss": 0.0561, "step": 9479 }, { "epoch": 4.461176470588235, "grad_norm": 0.41672803061872293, "learning_rate": 6.343567966390452e-05, "loss": 0.2636, "step": 9480 }, { "epoch": 4.461647058823529, "grad_norm": 0.5287826563865229, "learning_rate": 6.339988337724487e-05, "loss": 0.2707, "step": 9481 }, { "epoch": 4.4621176470588235, "grad_norm": 0.43567812118759947, "learning_rate": 6.336409448665988e-05, "loss": 0.1661, "step": 9482 }, { "epoch": 4.462588235294118, "grad_norm": 0.28731693959528626, "learning_rate": 6.332831299520617e-05, "loss": 0.1938, "step": 9483 }, { "epoch": 4.463058823529412, "grad_norm": 0.42020384699142566, "learning_rate": 6.32925389059397e-05, "loss": 0.1407, "step": 9484 }, { "epoch": 4.463529411764706, "grad_norm": 0.31242999717686115, "learning_rate": 6.325677222191568e-05, "loss": 0.2, "step": 9485 }, { "epoch": 4.464, "grad_norm": 0.40337427598670017, "learning_rate": 6.322101294618875e-05, "loss": 0.3121, "step": 9486 }, { "epoch": 4.464470588235294, "grad_norm": 0.5268343165583083, "learning_rate": 6.318526108181299e-05, "loss": 0.3643, "step": 9487 }, { "epoch": 4.464941176470588, "grad_norm": 0.3722240608426159, "learning_rate": 6.314951663184175e-05, "loss": 0.2367, "step": 9488 }, { "epoch": 4.465411764705882, "grad_norm": 0.6321179910142848, "learning_rate": 6.311377959932783e-05, "loss": 0.295, "step": 9489 }, { "epoch": 4.465882352941176, "grad_norm": 0.33949768927313734, "learning_rate": 6.307804998732328e-05, "loss": 0.1047, "step": 9490 }, { "epoch": 4.466352941176471, "grad_norm": 0.3501287471210273, "learning_rate": 6.304232779887958e-05, "loss": 0.1227, "step": 9491 }, { "epoch": 4.466823529411765, "grad_norm": 0.2546250519133782, "learning_rate": 6.300661303704768e-05, "loss": 0.1168, "step": 9492 }, { "epoch": 4.467294117647059, "grad_norm": 0.3163339286254538, "learning_rate": 6.297090570487766e-05, "loss": 0.0871, "step": 9493 }, { "epoch": 4.467764705882353, "grad_norm": 0.7117877831630713, "learning_rate": 6.293520580541925e-05, "loss": 0.4871, "step": 9494 }, { "epoch": 4.4682352941176475, "grad_norm": 0.4882153105647524, "learning_rate": 6.289951334172123e-05, "loss": 0.2191, "step": 9495 }, { "epoch": 4.468705882352941, "grad_norm": 0.5293624694384876, "learning_rate": 6.2863828316832e-05, "loss": 0.4029, "step": 9496 }, { "epoch": 4.469176470588235, "grad_norm": 0.23364966685749655, "learning_rate": 6.282815073379932e-05, "loss": 0.0998, "step": 9497 }, { "epoch": 4.469647058823529, "grad_norm": 0.5027050636226768, "learning_rate": 6.279248059567006e-05, "loss": 0.3012, "step": 9498 }, { "epoch": 4.4701176470588235, "grad_norm": 0.42524357778290306, "learning_rate": 6.275681790549072e-05, "loss": 0.2691, "step": 9499 }, { "epoch": 4.470588235294118, "grad_norm": 0.1192787942783487, "learning_rate": 6.27211626663071e-05, "loss": 0.0198, "step": 9500 }, { "epoch": 4.471058823529412, "grad_norm": 0.3896583593203409, "learning_rate": 6.268551488116423e-05, "loss": 0.308, "step": 9501 }, { "epoch": 4.471529411764706, "grad_norm": 0.22292962416715556, "learning_rate": 6.264987455310667e-05, "loss": 0.0803, "step": 9502 }, { "epoch": 4.4719999999999995, "grad_norm": 0.5716543562132788, "learning_rate": 6.261424168517832e-05, "loss": 0.4372, "step": 9503 }, { "epoch": 4.472470588235294, "grad_norm": 0.2660429120108794, "learning_rate": 6.257861628042233e-05, "loss": 0.1067, "step": 9504 }, { "epoch": 4.472941176470588, "grad_norm": 0.437191075828688, "learning_rate": 6.254299834188128e-05, "loss": 0.3182, "step": 9505 }, { "epoch": 4.473411764705882, "grad_norm": 0.22780223206897166, "learning_rate": 6.25073878725972e-05, "loss": 0.1193, "step": 9506 }, { "epoch": 4.473882352941176, "grad_norm": 0.5975254350227496, "learning_rate": 6.24717848756113e-05, "loss": 0.3491, "step": 9507 }, { "epoch": 4.474352941176471, "grad_norm": 0.457496744650917, "learning_rate": 6.24361893539643e-05, "loss": 0.347, "step": 9508 }, { "epoch": 4.474823529411765, "grad_norm": 0.5569867659854653, "learning_rate": 6.240060131069628e-05, "loss": 0.3926, "step": 9509 }, { "epoch": 4.475294117647059, "grad_norm": 0.44330326599856124, "learning_rate": 6.23650207488465e-05, "loss": 0.3464, "step": 9510 }, { "epoch": 4.475764705882353, "grad_norm": 0.39949844113015465, "learning_rate": 6.232944767145384e-05, "loss": 0.2613, "step": 9511 }, { "epoch": 4.476235294117647, "grad_norm": 0.10494158959940238, "learning_rate": 6.229388208155639e-05, "loss": 0.0265, "step": 9512 }, { "epoch": 4.476705882352941, "grad_norm": 0.46680199929791755, "learning_rate": 6.225832398219158e-05, "loss": 0.4265, "step": 9513 }, { "epoch": 4.477176470588235, "grad_norm": 0.17620697193422646, "learning_rate": 6.222277337639624e-05, "loss": 0.0608, "step": 9514 }, { "epoch": 4.477647058823529, "grad_norm": 0.37311367624414665, "learning_rate": 6.218723026720668e-05, "loss": 0.1289, "step": 9515 }, { "epoch": 4.4781176470588235, "grad_norm": 0.399492482109356, "learning_rate": 6.21516946576583e-05, "loss": 0.2388, "step": 9516 }, { "epoch": 4.478588235294118, "grad_norm": 0.40302956567905734, "learning_rate": 6.211616655078616e-05, "loss": 0.2884, "step": 9517 }, { "epoch": 4.479058823529412, "grad_norm": 0.2688736469191164, "learning_rate": 6.208064594962441e-05, "loss": 0.0814, "step": 9518 }, { "epoch": 4.479529411764706, "grad_norm": 0.5402144744616991, "learning_rate": 6.204513285720673e-05, "loss": 0.2243, "step": 9519 }, { "epoch": 4.48, "grad_norm": 0.3813108318682352, "learning_rate": 6.200962727656616e-05, "loss": 0.3246, "step": 9520 }, { "epoch": 4.480470588235294, "grad_norm": 0.33274155837118785, "learning_rate": 6.197412921073498e-05, "loss": 0.1315, "step": 9521 }, { "epoch": 4.480941176470588, "grad_norm": 0.38018122634745627, "learning_rate": 6.193863866274492e-05, "loss": 0.245, "step": 9522 }, { "epoch": 4.481411764705882, "grad_norm": 0.430198903550421, "learning_rate": 6.190315563562711e-05, "loss": 0.1232, "step": 9523 }, { "epoch": 4.481882352941176, "grad_norm": 0.7188663424466385, "learning_rate": 6.186768013241188e-05, "loss": 0.3362, "step": 9524 }, { "epoch": 4.482352941176471, "grad_norm": 0.556481891309855, "learning_rate": 6.183221215612904e-05, "loss": 0.3151, "step": 9525 }, { "epoch": 4.482823529411765, "grad_norm": 0.16167092799481667, "learning_rate": 6.179675170980779e-05, "loss": 0.0595, "step": 9526 }, { "epoch": 4.483294117647059, "grad_norm": 0.25547903720674836, "learning_rate": 6.176129879647654e-05, "loss": 0.151, "step": 9527 }, { "epoch": 4.483764705882353, "grad_norm": 0.3377883183499125, "learning_rate": 6.172585341916317e-05, "loss": 0.108, "step": 9528 }, { "epoch": 4.484235294117647, "grad_norm": 0.17688603207629497, "learning_rate": 6.169041558089496e-05, "loss": 0.0373, "step": 9529 }, { "epoch": 4.484705882352941, "grad_norm": 0.5124386789383326, "learning_rate": 6.16549852846983e-05, "loss": 0.434, "step": 9530 }, { "epoch": 4.485176470588235, "grad_norm": 0.36670856827063336, "learning_rate": 6.161956253359935e-05, "loss": 0.228, "step": 9531 }, { "epoch": 4.485647058823529, "grad_norm": 0.38502489410685076, "learning_rate": 6.15841473306232e-05, "loss": 0.1755, "step": 9532 }, { "epoch": 4.4861176470588235, "grad_norm": 0.3517686935621558, "learning_rate": 6.154873967879455e-05, "loss": 0.2354, "step": 9533 }, { "epoch": 4.486588235294118, "grad_norm": 0.19785575193140278, "learning_rate": 6.151333958113743e-05, "loss": 0.0524, "step": 9534 }, { "epoch": 4.487058823529412, "grad_norm": 0.4133700725225527, "learning_rate": 6.147794704067508e-05, "loss": 0.2949, "step": 9535 }, { "epoch": 4.487529411764706, "grad_norm": 0.47679901264399255, "learning_rate": 6.144256206043025e-05, "loss": 0.2865, "step": 9536 }, { "epoch": 4.4879999999999995, "grad_norm": 0.1959315410782236, "learning_rate": 6.140718464342505e-05, "loss": 0.0497, "step": 9537 }, { "epoch": 4.488470588235294, "grad_norm": 0.31100962554629, "learning_rate": 6.137181479268077e-05, "loss": 0.1217, "step": 9538 }, { "epoch": 4.488941176470588, "grad_norm": 0.5597507253190785, "learning_rate": 6.133645251121823e-05, "loss": 0.1156, "step": 9539 }, { "epoch": 4.489411764705882, "grad_norm": 0.3517388451794829, "learning_rate": 6.130109780205759e-05, "loss": 0.1223, "step": 9540 }, { "epoch": 4.489882352941176, "grad_norm": 0.5394362227563136, "learning_rate": 6.126575066821821e-05, "loss": 0.3058, "step": 9541 }, { "epoch": 4.490352941176471, "grad_norm": 0.3096837142007012, "learning_rate": 6.123041111271896e-05, "loss": 0.0986, "step": 9542 }, { "epoch": 4.490823529411765, "grad_norm": 0.4863197185380549, "learning_rate": 6.119507913857807e-05, "loss": 0.2766, "step": 9543 }, { "epoch": 4.491294117647059, "grad_norm": 0.39937018776232064, "learning_rate": 6.115975474881296e-05, "loss": 0.1662, "step": 9544 }, { "epoch": 4.491764705882353, "grad_norm": 0.23787688019750827, "learning_rate": 6.11244379464406e-05, "loss": 0.0954, "step": 9545 }, { "epoch": 4.492235294117647, "grad_norm": 0.2619275266011201, "learning_rate": 6.108912873447712e-05, "loss": 0.091, "step": 9546 }, { "epoch": 4.492705882352941, "grad_norm": 0.41824106833239116, "learning_rate": 6.105382711593814e-05, "loss": 0.2863, "step": 9547 }, { "epoch": 4.493176470588235, "grad_norm": 0.3937932441078377, "learning_rate": 6.101853309383866e-05, "loss": 0.259, "step": 9548 }, { "epoch": 4.493647058823529, "grad_norm": 0.5941431893656973, "learning_rate": 6.0983246671192866e-05, "loss": 0.3402, "step": 9549 }, { "epoch": 4.4941176470588236, "grad_norm": 0.4791284766575449, "learning_rate": 6.0947967851014405e-05, "loss": 0.288, "step": 9550 }, { "epoch": 4.494588235294118, "grad_norm": 0.289518701716131, "learning_rate": 6.091269663631636e-05, "loss": 0.1594, "step": 9551 }, { "epoch": 4.495058823529412, "grad_norm": 0.4320866001922811, "learning_rate": 6.0877433030110935e-05, "loss": 0.3141, "step": 9552 }, { "epoch": 4.495529411764706, "grad_norm": 0.23359605315065252, "learning_rate": 6.084217703540988e-05, "loss": 0.0864, "step": 9553 }, { "epoch": 4.496, "grad_norm": 0.29293461166470475, "learning_rate": 6.0806928655224266e-05, "loss": 0.2135, "step": 9554 }, { "epoch": 4.496470588235294, "grad_norm": 0.340900354974821, "learning_rate": 6.0771687892564394e-05, "loss": 0.2019, "step": 9555 }, { "epoch": 4.496941176470588, "grad_norm": 0.23350373205450853, "learning_rate": 6.0736454750440025e-05, "loss": 0.1002, "step": 9556 }, { "epoch": 4.497411764705882, "grad_norm": 0.21977781507906088, "learning_rate": 6.070122923186032e-05, "loss": 0.0895, "step": 9557 }, { "epoch": 4.497882352941176, "grad_norm": 0.25728404432165713, "learning_rate": 6.0666011339833584e-05, "loss": 0.1254, "step": 9558 }, { "epoch": 4.498352941176471, "grad_norm": 0.4648993758257409, "learning_rate": 6.063080107736767e-05, "loss": 0.2715, "step": 9559 }, { "epoch": 4.498823529411765, "grad_norm": 0.08237770319693594, "learning_rate": 6.0595598447469726e-05, "loss": 0.0107, "step": 9560 }, { "epoch": 4.499294117647059, "grad_norm": 1.2860760333256114, "learning_rate": 6.056040345314618e-05, "loss": 0.4138, "step": 9561 }, { "epoch": 4.499764705882353, "grad_norm": 0.35577553321609584, "learning_rate": 6.0525216097402865e-05, "loss": 0.2002, "step": 9562 }, { "epoch": 4.5002352941176476, "grad_norm": 0.44399888776523283, "learning_rate": 6.049003638324501e-05, "loss": 0.2707, "step": 9563 }, { "epoch": 4.500705882352941, "grad_norm": 0.40388513554156674, "learning_rate": 6.0454864313677054e-05, "loss": 0.2705, "step": 9564 }, { "epoch": 4.501176470588235, "grad_norm": 0.3158451428757248, "learning_rate": 6.041969989170289e-05, "loss": 0.182, "step": 9565 }, { "epoch": 4.501647058823529, "grad_norm": 0.4607761253073971, "learning_rate": 6.0384543120325796e-05, "loss": 0.2909, "step": 9566 }, { "epoch": 4.502117647058824, "grad_norm": 0.3446768235495825, "learning_rate": 6.034939400254824e-05, "loss": 0.1238, "step": 9567 }, { "epoch": 4.502588235294118, "grad_norm": 0.6653550926292285, "learning_rate": 6.031425254137222e-05, "loss": 0.2884, "step": 9568 }, { "epoch": 4.503058823529412, "grad_norm": 0.14610695049750932, "learning_rate": 6.0279118739798874e-05, "loss": 0.0366, "step": 9569 }, { "epoch": 4.503529411764706, "grad_norm": 0.21847031675316447, "learning_rate": 6.0243992600828875e-05, "loss": 0.0808, "step": 9570 }, { "epoch": 4.504, "grad_norm": 0.28413199793792543, "learning_rate": 6.020887412746221e-05, "loss": 0.1097, "step": 9571 }, { "epoch": 4.504470588235294, "grad_norm": 0.19510096496775484, "learning_rate": 6.017376332269807e-05, "loss": 0.0309, "step": 9572 }, { "epoch": 4.504941176470588, "grad_norm": 0.3607817286259719, "learning_rate": 6.0138660189535126e-05, "loss": 0.1241, "step": 9573 }, { "epoch": 4.505411764705882, "grad_norm": 0.2897134972312739, "learning_rate": 6.0103564730971375e-05, "loss": 0.1503, "step": 9574 }, { "epoch": 4.5058823529411764, "grad_norm": 0.4428894246075164, "learning_rate": 6.006847695000415e-05, "loss": 0.2626, "step": 9575 }, { "epoch": 4.506352941176471, "grad_norm": 0.2865476916074861, "learning_rate": 6.0033396849630136e-05, "loss": 0.1423, "step": 9576 }, { "epoch": 4.506823529411765, "grad_norm": 0.46593334746590087, "learning_rate": 5.999832443284527e-05, "loss": 0.3326, "step": 9577 }, { "epoch": 4.507294117647059, "grad_norm": 0.3299174977289126, "learning_rate": 5.9963259702644963e-05, "loss": 0.1525, "step": 9578 }, { "epoch": 4.507764705882353, "grad_norm": 0.3581228493952125, "learning_rate": 5.992820266202394e-05, "loss": 0.192, "step": 9579 }, { "epoch": 4.508235294117647, "grad_norm": 0.3237663281166194, "learning_rate": 5.989315331397616e-05, "loss": 0.1503, "step": 9580 }, { "epoch": 4.508705882352941, "grad_norm": 0.292232982865403, "learning_rate": 5.985811166149507e-05, "loss": 0.1241, "step": 9581 }, { "epoch": 4.509176470588235, "grad_norm": 0.23134032094847537, "learning_rate": 5.982307770757343e-05, "loss": 0.0439, "step": 9582 }, { "epoch": 4.509647058823529, "grad_norm": 0.7171901369786511, "learning_rate": 5.978805145520322e-05, "loss": 0.42, "step": 9583 }, { "epoch": 4.510117647058824, "grad_norm": 0.47667687119352853, "learning_rate": 5.97530329073759e-05, "loss": 0.2997, "step": 9584 }, { "epoch": 4.510588235294118, "grad_norm": 0.30834893306541966, "learning_rate": 5.971802206708229e-05, "loss": 0.1709, "step": 9585 }, { "epoch": 4.511058823529412, "grad_norm": 0.8081760868688864, "learning_rate": 5.968301893731237e-05, "loss": 0.6563, "step": 9586 }, { "epoch": 4.511529411764706, "grad_norm": 0.4474508938338464, "learning_rate": 5.964802352105564e-05, "loss": 0.2401, "step": 9587 }, { "epoch": 4.5120000000000005, "grad_norm": 0.35251455490667366, "learning_rate": 5.961303582130092e-05, "loss": 0.1469, "step": 9588 }, { "epoch": 4.512470588235294, "grad_norm": 0.40735993449269925, "learning_rate": 5.957805584103623e-05, "loss": 0.2222, "step": 9589 }, { "epoch": 4.512941176470588, "grad_norm": 0.518253119232536, "learning_rate": 5.954308358324909e-05, "loss": 0.2493, "step": 9590 }, { "epoch": 4.513411764705882, "grad_norm": 0.2876293540121791, "learning_rate": 5.950811905092636e-05, "loss": 0.176, "step": 9591 }, { "epoch": 4.5138823529411765, "grad_norm": 0.5223013840892026, "learning_rate": 5.947316224705407e-05, "loss": 0.2119, "step": 9592 }, { "epoch": 4.514352941176471, "grad_norm": 0.41392146414764724, "learning_rate": 5.943821317461775e-05, "loss": 0.1269, "step": 9593 }, { "epoch": 4.514823529411765, "grad_norm": 0.4616124366476143, "learning_rate": 5.940327183660228e-05, "loss": 0.271, "step": 9594 }, { "epoch": 4.515294117647059, "grad_norm": 0.402482018864745, "learning_rate": 5.936833823599172e-05, "loss": 0.2043, "step": 9595 }, { "epoch": 4.5157647058823525, "grad_norm": 0.33744696118415934, "learning_rate": 5.933341237576968e-05, "loss": 0.195, "step": 9596 }, { "epoch": 4.516235294117647, "grad_norm": 0.24820960571400363, "learning_rate": 5.9298494258918875e-05, "loss": 0.0628, "step": 9597 }, { "epoch": 4.516705882352941, "grad_norm": 0.3444460716480954, "learning_rate": 5.926358388842156e-05, "loss": 0.2084, "step": 9598 }, { "epoch": 4.517176470588235, "grad_norm": 0.592872835126872, "learning_rate": 5.922868126725929e-05, "loss": 0.2495, "step": 9599 }, { "epoch": 4.517647058823529, "grad_norm": 0.5194399954699596, "learning_rate": 5.919378639841281e-05, "loss": 0.2094, "step": 9600 }, { "epoch": 4.518117647058824, "grad_norm": 0.3340435207001814, "learning_rate": 5.915889928486239e-05, "loss": 0.1222, "step": 9601 }, { "epoch": 4.518588235294118, "grad_norm": 0.3842681934080257, "learning_rate": 5.912401992958757e-05, "loss": 0.2643, "step": 9602 }, { "epoch": 4.519058823529412, "grad_norm": 0.3024750153073328, "learning_rate": 5.908914833556715e-05, "loss": 0.156, "step": 9603 }, { "epoch": 4.519529411764706, "grad_norm": 0.4749572110306544, "learning_rate": 5.9054284505779396e-05, "loss": 0.3926, "step": 9604 }, { "epoch": 4.52, "grad_norm": 0.20405841220897067, "learning_rate": 5.901942844320187e-05, "loss": 0.0822, "step": 9605 }, { "epoch": 4.520470588235294, "grad_norm": 0.32085041932189073, "learning_rate": 5.898458015081136e-05, "loss": 0.1955, "step": 9606 }, { "epoch": 4.520941176470588, "grad_norm": 0.6453287138756391, "learning_rate": 5.8949739631584136e-05, "loss": 0.2358, "step": 9607 }, { "epoch": 4.521411764705882, "grad_norm": 0.5414948286878788, "learning_rate": 5.8914906888495797e-05, "loss": 0.4537, "step": 9608 }, { "epoch": 4.5218823529411765, "grad_norm": 0.4076561262917914, "learning_rate": 5.888008192452112e-05, "loss": 0.1977, "step": 9609 }, { "epoch": 4.522352941176471, "grad_norm": 0.7563470560503235, "learning_rate": 5.8845264742634394e-05, "loss": 0.3947, "step": 9610 }, { "epoch": 4.522823529411765, "grad_norm": 0.4943067999500725, "learning_rate": 5.881045534580922e-05, "loss": 0.1737, "step": 9611 }, { "epoch": 4.523294117647059, "grad_norm": 0.45026699906652873, "learning_rate": 5.87756537370184e-05, "loss": 0.2757, "step": 9612 }, { "epoch": 4.523764705882353, "grad_norm": 0.3996293507615973, "learning_rate": 5.874085991923419e-05, "loss": 0.2308, "step": 9613 }, { "epoch": 4.524235294117647, "grad_norm": 0.31278094287335956, "learning_rate": 5.8706073895428225e-05, "loss": 0.1304, "step": 9614 }, { "epoch": 4.524705882352941, "grad_norm": 0.3068612253721963, "learning_rate": 5.867129566857129e-05, "loss": 0.1022, "step": 9615 }, { "epoch": 4.525176470588235, "grad_norm": 0.16478436238077498, "learning_rate": 5.8636525241633673e-05, "loss": 0.0377, "step": 9616 }, { "epoch": 4.525647058823529, "grad_norm": 0.19082538482208622, "learning_rate": 5.860176261758499e-05, "loss": 0.0681, "step": 9617 }, { "epoch": 4.526117647058824, "grad_norm": 0.42915621100576384, "learning_rate": 5.8567007799394015e-05, "loss": 0.2432, "step": 9618 }, { "epoch": 4.526588235294118, "grad_norm": 0.6172715982456346, "learning_rate": 5.853226079002906e-05, "loss": 0.3169, "step": 9619 }, { "epoch": 4.527058823529412, "grad_norm": 0.32725728453165254, "learning_rate": 5.849752159245769e-05, "loss": 0.1238, "step": 9620 }, { "epoch": 4.527529411764706, "grad_norm": 0.3134700594165365, "learning_rate": 5.84627902096468e-05, "loss": 0.1367, "step": 9621 }, { "epoch": 4.5280000000000005, "grad_norm": 0.4590993967979882, "learning_rate": 5.842806664456264e-05, "loss": 0.3234, "step": 9622 }, { "epoch": 4.528470588235294, "grad_norm": 1.2176158268569857, "learning_rate": 5.839335090017072e-05, "loss": 0.4428, "step": 9623 }, { "epoch": 4.528941176470588, "grad_norm": 0.3952207961731945, "learning_rate": 5.835864297943594e-05, "loss": 0.114, "step": 9624 }, { "epoch": 4.529411764705882, "grad_norm": 0.5760975159254893, "learning_rate": 5.8323942885322605e-05, "loss": 0.3942, "step": 9625 }, { "epoch": 4.5298823529411765, "grad_norm": 0.6636675280653519, "learning_rate": 5.828925062079416e-05, "loss": 0.4092, "step": 9626 }, { "epoch": 4.530352941176471, "grad_norm": 0.47268253030910473, "learning_rate": 5.8254566188813615e-05, "loss": 0.1103, "step": 9627 }, { "epoch": 4.530823529411765, "grad_norm": 0.2849629083532791, "learning_rate": 5.821988959234308e-05, "loss": 0.1316, "step": 9628 }, { "epoch": 4.531294117647059, "grad_norm": 0.36042795437050135, "learning_rate": 5.8185220834344124e-05, "loss": 0.1852, "step": 9629 }, { "epoch": 4.5317647058823525, "grad_norm": 0.46500757873205406, "learning_rate": 5.815055991777773e-05, "loss": 0.2667, "step": 9630 }, { "epoch": 4.532235294117647, "grad_norm": 0.2227112548122344, "learning_rate": 5.811590684560398e-05, "loss": 0.1151, "step": 9631 }, { "epoch": 4.532705882352941, "grad_norm": 0.4591186535685799, "learning_rate": 5.808126162078249e-05, "loss": 0.3399, "step": 9632 }, { "epoch": 4.533176470588235, "grad_norm": 0.27615102016317356, "learning_rate": 5.8046624246272146e-05, "loss": 0.1435, "step": 9633 }, { "epoch": 4.533647058823529, "grad_norm": 0.1649294308474599, "learning_rate": 5.801199472503109e-05, "loss": 0.0379, "step": 9634 }, { "epoch": 4.534117647058824, "grad_norm": 0.6087221032926166, "learning_rate": 5.797737306001686e-05, "loss": 0.3503, "step": 9635 }, { "epoch": 4.534588235294118, "grad_norm": 0.5730142726962816, "learning_rate": 5.794275925418641e-05, "loss": 0.2832, "step": 9636 }, { "epoch": 4.535058823529412, "grad_norm": 0.5993199773297185, "learning_rate": 5.79081533104958e-05, "loss": 0.4275, "step": 9637 }, { "epoch": 4.535529411764706, "grad_norm": 0.1432077253146772, "learning_rate": 5.787355523190059e-05, "loss": 0.0643, "step": 9638 }, { "epoch": 4.536, "grad_norm": 0.5882053560553859, "learning_rate": 5.783896502135572e-05, "loss": 0.3684, "step": 9639 }, { "epoch": 4.536470588235294, "grad_norm": 0.3359104939893666, "learning_rate": 5.780438268181522e-05, "loss": 0.1703, "step": 9640 }, { "epoch": 4.536941176470588, "grad_norm": 0.5086404464090317, "learning_rate": 5.776980821623264e-05, "loss": 0.2236, "step": 9641 }, { "epoch": 4.537411764705882, "grad_norm": 0.3081964776983434, "learning_rate": 5.7735241627560905e-05, "loss": 0.097, "step": 9642 }, { "epoch": 4.5378823529411765, "grad_norm": 0.48573853285778323, "learning_rate": 5.770068291875202e-05, "loss": 0.2084, "step": 9643 }, { "epoch": 4.538352941176471, "grad_norm": 0.43185230005576264, "learning_rate": 5.766613209275754e-05, "loss": 0.2291, "step": 9644 }, { "epoch": 4.538823529411765, "grad_norm": 0.16913795392242703, "learning_rate": 5.763158915252833e-05, "loss": 0.0367, "step": 9645 }, { "epoch": 4.539294117647059, "grad_norm": 0.7441123545247262, "learning_rate": 5.759705410101443e-05, "loss": 0.1889, "step": 9646 }, { "epoch": 4.539764705882353, "grad_norm": 0.41333777385813886, "learning_rate": 5.756252694116538e-05, "loss": 0.2734, "step": 9647 }, { "epoch": 4.540235294117647, "grad_norm": 0.5198494122875988, "learning_rate": 5.752800767592988e-05, "loss": 0.3053, "step": 9648 }, { "epoch": 4.540705882352941, "grad_norm": 0.26622929693472785, "learning_rate": 5.74934963082561e-05, "loss": 0.0804, "step": 9649 }, { "epoch": 4.541176470588235, "grad_norm": 0.313619544473162, "learning_rate": 5.745899284109154e-05, "loss": 0.1756, "step": 9650 }, { "epoch": 4.541647058823529, "grad_norm": 0.38196574521573723, "learning_rate": 5.742449727738285e-05, "loss": 0.1795, "step": 9651 }, { "epoch": 4.542117647058824, "grad_norm": 0.39965331229126483, "learning_rate": 5.739000962007616e-05, "loss": 0.2221, "step": 9652 }, { "epoch": 4.542588235294118, "grad_norm": 0.36683280556934617, "learning_rate": 5.7355529872116957e-05, "loss": 0.2219, "step": 9653 }, { "epoch": 4.543058823529412, "grad_norm": 0.29831651678685944, "learning_rate": 5.732105803644986e-05, "loss": 0.1548, "step": 9654 }, { "epoch": 4.543529411764706, "grad_norm": 0.3174557536083277, "learning_rate": 5.7286594116019e-05, "loss": 0.1442, "step": 9655 }, { "epoch": 4.5440000000000005, "grad_norm": 0.4226564951938525, "learning_rate": 5.725213811376782e-05, "loss": 0.1716, "step": 9656 }, { "epoch": 4.544470588235294, "grad_norm": 0.3878029961376224, "learning_rate": 5.7217690032638916e-05, "loss": 0.1598, "step": 9657 }, { "epoch": 4.544941176470588, "grad_norm": 0.2950406764125572, "learning_rate": 5.718324987557438e-05, "loss": 0.2045, "step": 9658 }, { "epoch": 4.545411764705882, "grad_norm": 0.526467488106314, "learning_rate": 5.714881764551561e-05, "loss": 0.1889, "step": 9659 }, { "epoch": 4.5458823529411765, "grad_norm": 0.41734392516954943, "learning_rate": 5.7114393345403194e-05, "loss": 0.1613, "step": 9660 }, { "epoch": 4.546352941176471, "grad_norm": 0.31964834490618443, "learning_rate": 5.707997697817719e-05, "loss": 0.0937, "step": 9661 }, { "epoch": 4.546823529411765, "grad_norm": 0.39922526970302663, "learning_rate": 5.704556854677698e-05, "loss": 0.1464, "step": 9662 }, { "epoch": 4.547294117647059, "grad_norm": 0.5006907836423771, "learning_rate": 5.701116805414104e-05, "loss": 0.2831, "step": 9663 }, { "epoch": 4.5477647058823525, "grad_norm": 0.6039152612650508, "learning_rate": 5.6976775503207574e-05, "loss": 0.431, "step": 9664 }, { "epoch": 4.548235294117647, "grad_norm": 0.24949900483141058, "learning_rate": 5.6942390896913686e-05, "loss": 0.0502, "step": 9665 }, { "epoch": 4.548705882352941, "grad_norm": 0.4554574131571732, "learning_rate": 5.690801423819608e-05, "loss": 0.2467, "step": 9666 }, { "epoch": 4.549176470588235, "grad_norm": 0.6146961202915094, "learning_rate": 5.687364552999071e-05, "loss": 0.3064, "step": 9667 }, { "epoch": 4.549647058823529, "grad_norm": 0.5102091718155534, "learning_rate": 5.683928477523275e-05, "loss": 0.3626, "step": 9668 }, { "epoch": 4.550117647058824, "grad_norm": 0.33854559008904284, "learning_rate": 5.680493197685681e-05, "loss": 0.0951, "step": 9669 }, { "epoch": 4.550588235294118, "grad_norm": 0.2640129003376415, "learning_rate": 5.677058713779687e-05, "loss": 0.0853, "step": 9670 }, { "epoch": 4.551058823529412, "grad_norm": 0.42575616713555564, "learning_rate": 5.6736250260986e-05, "loss": 0.3086, "step": 9671 }, { "epoch": 4.551529411764706, "grad_norm": 0.3076638144373396, "learning_rate": 5.6701921349356826e-05, "loss": 0.1536, "step": 9672 }, { "epoch": 4.552, "grad_norm": 0.5097756554763759, "learning_rate": 5.666760040584126e-05, "loss": 0.1765, "step": 9673 }, { "epoch": 4.552470588235294, "grad_norm": 0.17762437929868705, "learning_rate": 5.663328743337034e-05, "loss": 0.0455, "step": 9674 }, { "epoch": 4.552941176470588, "grad_norm": 0.37703013295451754, "learning_rate": 5.659898243487463e-05, "loss": 0.1631, "step": 9675 }, { "epoch": 4.553411764705882, "grad_norm": 0.38586237219957176, "learning_rate": 5.6564685413284014e-05, "loss": 0.183, "step": 9676 }, { "epoch": 4.5538823529411765, "grad_norm": 0.15941762709327567, "learning_rate": 5.6530396371527515e-05, "loss": 0.0438, "step": 9677 }, { "epoch": 4.554352941176471, "grad_norm": 0.400296922665892, "learning_rate": 5.649611531253366e-05, "loss": 0.1793, "step": 9678 }, { "epoch": 4.554823529411765, "grad_norm": 0.41428318448792273, "learning_rate": 5.646184223923017e-05, "loss": 0.2754, "step": 9679 }, { "epoch": 4.555294117647059, "grad_norm": 0.4083434977839461, "learning_rate": 5.642757715454413e-05, "loss": 0.2472, "step": 9680 }, { "epoch": 4.555764705882353, "grad_norm": 0.29882132395529926, "learning_rate": 5.639332006140203e-05, "loss": 0.1526, "step": 9681 }, { "epoch": 4.556235294117647, "grad_norm": 0.4578022555632267, "learning_rate": 5.635907096272949e-05, "loss": 0.2932, "step": 9682 }, { "epoch": 4.556705882352941, "grad_norm": 0.3518108274776055, "learning_rate": 5.632482986145159e-05, "loss": 0.1544, "step": 9683 }, { "epoch": 4.557176470588235, "grad_norm": 0.5337976616786775, "learning_rate": 5.6290596760492745e-05, "loss": 0.2476, "step": 9684 }, { "epoch": 4.557647058823529, "grad_norm": 0.4227991793455953, "learning_rate": 5.625637166277654e-05, "loss": 0.3215, "step": 9685 }, { "epoch": 4.558117647058824, "grad_norm": 0.3433799993248422, "learning_rate": 5.622215457122598e-05, "loss": 0.1327, "step": 9686 }, { "epoch": 4.558588235294118, "grad_norm": 0.4342058332027129, "learning_rate": 5.618794548876346e-05, "loss": 0.2937, "step": 9687 }, { "epoch": 4.559058823529412, "grad_norm": 0.1355674472329552, "learning_rate": 5.6153744418310505e-05, "loss": 0.038, "step": 9688 }, { "epoch": 4.559529411764705, "grad_norm": 0.3544451997524012, "learning_rate": 5.611955136278808e-05, "loss": 0.1569, "step": 9689 }, { "epoch": 4.5600000000000005, "grad_norm": 0.3444451099305725, "learning_rate": 5.608536632511649e-05, "loss": 0.2256, "step": 9690 }, { "epoch": 4.560470588235294, "grad_norm": 0.3373315688240375, "learning_rate": 5.6051189308215225e-05, "loss": 0.1114, "step": 9691 }, { "epoch": 4.560941176470588, "grad_norm": 0.336783728131253, "learning_rate": 5.6017020315003195e-05, "loss": 0.2051, "step": 9692 }, { "epoch": 4.561411764705882, "grad_norm": 0.4613610182880015, "learning_rate": 5.598285934839868e-05, "loss": 0.3041, "step": 9693 }, { "epoch": 4.5618823529411765, "grad_norm": 0.31475483021174205, "learning_rate": 5.594870641131907e-05, "loss": 0.1328, "step": 9694 }, { "epoch": 4.562352941176471, "grad_norm": 0.42541600048171946, "learning_rate": 5.591456150668126e-05, "loss": 0.3016, "step": 9695 }, { "epoch": 4.562823529411765, "grad_norm": 0.3582491808786695, "learning_rate": 5.588042463740143e-05, "loss": 0.138, "step": 9696 }, { "epoch": 4.563294117647059, "grad_norm": 0.5224035448640082, "learning_rate": 5.584629580639494e-05, "loss": 0.3754, "step": 9697 }, { "epoch": 4.5637647058823525, "grad_norm": 0.4579593957082214, "learning_rate": 5.581217501657666e-05, "loss": 0.1561, "step": 9698 }, { "epoch": 4.564235294117647, "grad_norm": 0.31131518616825715, "learning_rate": 5.577806227086057e-05, "loss": 0.1394, "step": 9699 }, { "epoch": 4.564705882352941, "grad_norm": 0.33807147961558665, "learning_rate": 5.57439575721601e-05, "loss": 0.203, "step": 9700 }, { "epoch": 4.565176470588235, "grad_norm": 0.5039986819850063, "learning_rate": 5.570986092338806e-05, "loss": 0.1582, "step": 9701 }, { "epoch": 4.565647058823529, "grad_norm": 0.3502517531807317, "learning_rate": 5.5675772327456313e-05, "loss": 0.1778, "step": 9702 }, { "epoch": 4.566117647058824, "grad_norm": 0.6308246847211921, "learning_rate": 5.5641691787276286e-05, "loss": 0.4253, "step": 9703 }, { "epoch": 4.566588235294118, "grad_norm": 0.16422330172233457, "learning_rate": 5.560761930575864e-05, "loss": 0.0339, "step": 9704 }, { "epoch": 4.567058823529412, "grad_norm": 0.6203873467956146, "learning_rate": 5.557355488581325e-05, "loss": 0.3491, "step": 9705 }, { "epoch": 4.567529411764706, "grad_norm": 0.3525544780388869, "learning_rate": 5.553949853034943e-05, "loss": 0.1682, "step": 9706 }, { "epoch": 4.568, "grad_norm": 0.4192329864157742, "learning_rate": 5.550545024227582e-05, "loss": 0.1044, "step": 9707 }, { "epoch": 4.568470588235294, "grad_norm": 0.22017207129852523, "learning_rate": 5.5471410024500156e-05, "loss": 0.0399, "step": 9708 }, { "epoch": 4.568941176470588, "grad_norm": 0.17620138889000597, "learning_rate": 5.5437377879929826e-05, "loss": 0.0459, "step": 9709 }, { "epoch": 4.569411764705882, "grad_norm": 0.7295977248887023, "learning_rate": 5.5403353811471194e-05, "loss": 0.4385, "step": 9710 }, { "epoch": 4.5698823529411765, "grad_norm": 0.6518219508548537, "learning_rate": 5.5369337822030154e-05, "loss": 0.3411, "step": 9711 }, { "epoch": 4.570352941176471, "grad_norm": 0.22273860890616604, "learning_rate": 5.533532991451187e-05, "loss": 0.0911, "step": 9712 }, { "epoch": 4.570823529411765, "grad_norm": 0.2871966388852429, "learning_rate": 5.53013300918207e-05, "loss": 0.1606, "step": 9713 }, { "epoch": 4.571294117647059, "grad_norm": 0.16646548018681084, "learning_rate": 5.526733835686041e-05, "loss": 0.0552, "step": 9714 }, { "epoch": 4.571764705882353, "grad_norm": 0.4553095826582596, "learning_rate": 5.523335471253416e-05, "loss": 0.1707, "step": 9715 }, { "epoch": 4.572235294117647, "grad_norm": 0.5839590535221448, "learning_rate": 5.519937916174419e-05, "loss": 0.1365, "step": 9716 }, { "epoch": 4.572705882352941, "grad_norm": 0.3773880513315709, "learning_rate": 5.516541170739223e-05, "loss": 0.1314, "step": 9717 }, { "epoch": 4.573176470588235, "grad_norm": 0.5109735400805266, "learning_rate": 5.5131452352379314e-05, "loss": 0.2788, "step": 9718 }, { "epoch": 4.573647058823529, "grad_norm": 0.3330308926021101, "learning_rate": 5.509750109960566e-05, "loss": 0.1181, "step": 9719 }, { "epoch": 4.574117647058824, "grad_norm": 0.44209333481040447, "learning_rate": 5.5063557951970885e-05, "loss": 0.2008, "step": 9720 }, { "epoch": 4.574588235294118, "grad_norm": 0.49715300294619597, "learning_rate": 5.502962291237399e-05, "loss": 0.2183, "step": 9721 }, { "epoch": 4.575058823529412, "grad_norm": 0.39219301584421107, "learning_rate": 5.4995695983713076e-05, "loss": 0.2302, "step": 9722 }, { "epoch": 4.575529411764705, "grad_norm": 0.3675368162766334, "learning_rate": 5.496177716888571e-05, "loss": 0.2617, "step": 9723 }, { "epoch": 4.576, "grad_norm": 0.38220180475909576, "learning_rate": 5.4927866470788786e-05, "loss": 0.2277, "step": 9724 }, { "epoch": 4.576470588235294, "grad_norm": 0.3708790483418446, "learning_rate": 5.489396389231834e-05, "loss": 0.1691, "step": 9725 }, { "epoch": 4.576941176470588, "grad_norm": 0.3462767066869165, "learning_rate": 5.4860069436369916e-05, "loss": 0.1801, "step": 9726 }, { "epoch": 4.577411764705882, "grad_norm": 0.2437973708900926, "learning_rate": 5.4826183105838184e-05, "loss": 0.0872, "step": 9727 }, { "epoch": 4.5778823529411765, "grad_norm": 0.33829303531859267, "learning_rate": 5.4792304903617226e-05, "loss": 0.1787, "step": 9728 }, { "epoch": 4.578352941176471, "grad_norm": 0.3531238994373274, "learning_rate": 5.4758434832600475e-05, "loss": 0.1144, "step": 9729 }, { "epoch": 4.578823529411765, "grad_norm": 0.4657874725155716, "learning_rate": 5.47245728956805e-05, "loss": 0.2447, "step": 9730 }, { "epoch": 4.579294117647059, "grad_norm": 0.37076121350195723, "learning_rate": 5.469071909574933e-05, "loss": 0.1474, "step": 9731 }, { "epoch": 4.5797647058823525, "grad_norm": 0.479390490389071, "learning_rate": 5.465687343569828e-05, "loss": 0.4131, "step": 9732 }, { "epoch": 4.580235294117647, "grad_norm": 0.30515275645065404, "learning_rate": 5.462303591841783e-05, "loss": 0.2011, "step": 9733 }, { "epoch": 4.580705882352941, "grad_norm": 0.3790745914819297, "learning_rate": 5.458920654679795e-05, "loss": 0.3028, "step": 9734 }, { "epoch": 4.581176470588235, "grad_norm": 0.41116046081129726, "learning_rate": 5.4555385323727876e-05, "loss": 0.2229, "step": 9735 }, { "epoch": 4.581647058823529, "grad_norm": 0.20377618224240576, "learning_rate": 5.4521572252095985e-05, "loss": 0.1149, "step": 9736 }, { "epoch": 4.582117647058824, "grad_norm": 0.2746708877403809, "learning_rate": 5.4487767334790145e-05, "loss": 0.215, "step": 9737 }, { "epoch": 4.582588235294118, "grad_norm": 0.48433387626575625, "learning_rate": 5.445397057469751e-05, "loss": 0.3391, "step": 9738 }, { "epoch": 4.583058823529412, "grad_norm": 0.6782177573901703, "learning_rate": 5.4420181974704386e-05, "loss": 0.3418, "step": 9739 }, { "epoch": 4.583529411764706, "grad_norm": 0.284592641170087, "learning_rate": 5.4386401537696536e-05, "loss": 0.194, "step": 9740 }, { "epoch": 4.584, "grad_norm": 0.29950615698015204, "learning_rate": 5.435262926655903e-05, "loss": 0.12, "step": 9741 }, { "epoch": 4.584470588235294, "grad_norm": 0.3667398476794601, "learning_rate": 5.43188651641761e-05, "loss": 0.1773, "step": 9742 }, { "epoch": 4.584941176470588, "grad_norm": 0.43836541559730746, "learning_rate": 5.428510923343136e-05, "loss": 0.2877, "step": 9743 }, { "epoch": 4.585411764705882, "grad_norm": 0.4893405113056331, "learning_rate": 5.425136147720785e-05, "loss": 0.3055, "step": 9744 }, { "epoch": 4.5858823529411765, "grad_norm": 0.2609374144419266, "learning_rate": 5.421762189838765e-05, "loss": 0.0879, "step": 9745 }, { "epoch": 4.586352941176471, "grad_norm": 0.5304164245716735, "learning_rate": 5.4183890499852364e-05, "loss": 0.3378, "step": 9746 }, { "epoch": 4.586823529411765, "grad_norm": 0.3660270905060104, "learning_rate": 5.4150167284482844e-05, "loss": 0.2071, "step": 9747 }, { "epoch": 4.587294117647059, "grad_norm": 0.4866770271808284, "learning_rate": 5.4116452255159126e-05, "loss": 0.2349, "step": 9748 }, { "epoch": 4.587764705882353, "grad_norm": 0.40075671179224637, "learning_rate": 5.408274541476075e-05, "loss": 0.2428, "step": 9749 }, { "epoch": 4.588235294117647, "grad_norm": 0.06111549063756677, "learning_rate": 5.4049046766166335e-05, "loss": 0.0081, "step": 9750 }, { "epoch": 4.588705882352941, "grad_norm": 0.513162305682377, "learning_rate": 5.401535631225396e-05, "loss": 0.3285, "step": 9751 }, { "epoch": 4.589176470588235, "grad_norm": 0.3442081077605309, "learning_rate": 5.398167405590096e-05, "loss": 0.1731, "step": 9752 }, { "epoch": 4.589647058823529, "grad_norm": 0.11768142278556062, "learning_rate": 5.3947999999983967e-05, "loss": 0.0409, "step": 9753 }, { "epoch": 4.590117647058824, "grad_norm": 0.7453997656582456, "learning_rate": 5.3914334147378894e-05, "loss": 0.404, "step": 9754 }, { "epoch": 4.590588235294118, "grad_norm": 0.40736981112561665, "learning_rate": 5.388067650096106e-05, "loss": 0.2577, "step": 9755 }, { "epoch": 4.591058823529412, "grad_norm": 0.29845539200716414, "learning_rate": 5.384702706360483e-05, "loss": 0.077, "step": 9756 }, { "epoch": 4.591529411764705, "grad_norm": 0.39041133769883823, "learning_rate": 5.3813385838184185e-05, "loss": 0.1523, "step": 9757 }, { "epoch": 4.592, "grad_norm": 0.49383490373235456, "learning_rate": 5.3779752827572126e-05, "loss": 0.2262, "step": 9758 }, { "epoch": 4.592470588235294, "grad_norm": 0.2794623409684175, "learning_rate": 5.374612803464115e-05, "loss": 0.0858, "step": 9759 }, { "epoch": 4.592941176470588, "grad_norm": 0.3861858258691325, "learning_rate": 5.3712511462262994e-05, "loss": 0.2013, "step": 9760 }, { "epoch": 4.593411764705882, "grad_norm": 0.35463573572620294, "learning_rate": 5.36789031133086e-05, "loss": 0.1224, "step": 9761 }, { "epoch": 4.5938823529411765, "grad_norm": 0.38087266747466014, "learning_rate": 5.364530299064834e-05, "loss": 0.2425, "step": 9762 }, { "epoch": 4.594352941176471, "grad_norm": 0.27710994290348734, "learning_rate": 5.361171109715186e-05, "loss": 0.1451, "step": 9763 }, { "epoch": 4.594823529411765, "grad_norm": 0.4364765085878699, "learning_rate": 5.357812743568799e-05, "loss": 0.2398, "step": 9764 }, { "epoch": 4.595294117647059, "grad_norm": 0.28195694384354414, "learning_rate": 5.354455200912498e-05, "loss": 0.1981, "step": 9765 }, { "epoch": 4.5957647058823525, "grad_norm": 0.3594760403308885, "learning_rate": 5.35109848203304e-05, "loss": 0.173, "step": 9766 }, { "epoch": 4.596235294117647, "grad_norm": 0.619518439354919, "learning_rate": 5.3477425872170926e-05, "loss": 0.4361, "step": 9767 }, { "epoch": 4.596705882352941, "grad_norm": 0.34794698206289626, "learning_rate": 5.3443875167512736e-05, "loss": 0.1392, "step": 9768 }, { "epoch": 4.597176470588235, "grad_norm": 0.30606719905967655, "learning_rate": 5.341033270922126e-05, "loss": 0.1712, "step": 9769 }, { "epoch": 4.597647058823529, "grad_norm": 0.3184091408798645, "learning_rate": 5.337679850016109e-05, "loss": 0.086, "step": 9770 }, { "epoch": 4.598117647058824, "grad_norm": 0.38156797797339925, "learning_rate": 5.3343272543196264e-05, "loss": 0.2178, "step": 9771 }, { "epoch": 4.598588235294118, "grad_norm": 0.13791671070488196, "learning_rate": 5.33097548411901e-05, "loss": 0.0311, "step": 9772 }, { "epoch": 4.599058823529412, "grad_norm": 0.5264527610517337, "learning_rate": 5.3276245397005094e-05, "loss": 0.2913, "step": 9773 }, { "epoch": 4.599529411764706, "grad_norm": 0.4707049667832219, "learning_rate": 5.324274421350316e-05, "loss": 0.3339, "step": 9774 }, { "epoch": 4.6, "grad_norm": 0.07338515979238192, "learning_rate": 5.3209251293545514e-05, "loss": 0.0058, "step": 9775 }, { "epoch": 4.600470588235294, "grad_norm": 0.40873961510026885, "learning_rate": 5.317576663999251e-05, "loss": 0.2643, "step": 9776 }, { "epoch": 4.600941176470588, "grad_norm": 0.21950196810641218, "learning_rate": 5.314229025570401e-05, "loss": 0.0551, "step": 9777 }, { "epoch": 4.601411764705882, "grad_norm": 0.2392466384792494, "learning_rate": 5.3108822143538944e-05, "loss": 0.0548, "step": 9778 }, { "epoch": 4.6018823529411765, "grad_norm": 0.3960077295689723, "learning_rate": 5.307536230635571e-05, "loss": 0.1576, "step": 9779 }, { "epoch": 4.602352941176471, "grad_norm": 0.3088742792895999, "learning_rate": 5.304191074701201e-05, "loss": 0.1235, "step": 9780 }, { "epoch": 4.602823529411765, "grad_norm": 0.22877947024886033, "learning_rate": 5.3008467468364675e-05, "loss": 0.0735, "step": 9781 }, { "epoch": 4.603294117647059, "grad_norm": 0.6157748598707505, "learning_rate": 5.297503247326993e-05, "loss": 0.3565, "step": 9782 }, { "epoch": 4.603764705882353, "grad_norm": 0.43640686807408974, "learning_rate": 5.294160576458339e-05, "loss": 0.1368, "step": 9783 }, { "epoch": 4.604235294117647, "grad_norm": 0.40943150206374535, "learning_rate": 5.290818734515972e-05, "loss": 0.1988, "step": 9784 }, { "epoch": 4.604705882352941, "grad_norm": 0.1588086399299787, "learning_rate": 5.287477721785309e-05, "loss": 0.0355, "step": 9785 }, { "epoch": 4.605176470588235, "grad_norm": 0.3825804032391767, "learning_rate": 5.284137538551693e-05, "loss": 0.1802, "step": 9786 }, { "epoch": 4.605647058823529, "grad_norm": 0.38492543797079576, "learning_rate": 5.2807981851003825e-05, "loss": 0.1857, "step": 9787 }, { "epoch": 4.606117647058824, "grad_norm": 0.5891151839646546, "learning_rate": 5.2774596617165793e-05, "loss": 0.303, "step": 9788 }, { "epoch": 4.606588235294118, "grad_norm": 0.29072220311694674, "learning_rate": 5.2741219686854146e-05, "loss": 0.1511, "step": 9789 }, { "epoch": 4.607058823529412, "grad_norm": 0.47645651269768075, "learning_rate": 5.270785106291934e-05, "loss": 0.2351, "step": 9790 }, { "epoch": 4.607529411764705, "grad_norm": 0.2592000636843403, "learning_rate": 5.2674490748211265e-05, "loss": 0.0712, "step": 9791 }, { "epoch": 4.608, "grad_norm": 0.4562441110988213, "learning_rate": 5.2641138745579126e-05, "loss": 0.3047, "step": 9792 }, { "epoch": 4.608470588235294, "grad_norm": 0.5740782806433269, "learning_rate": 5.2607795057871216e-05, "loss": 0.3084, "step": 9793 }, { "epoch": 4.608941176470588, "grad_norm": 0.6456644835404395, "learning_rate": 5.257445968793531e-05, "loss": 0.3725, "step": 9794 }, { "epoch": 4.609411764705882, "grad_norm": 0.16844230399437987, "learning_rate": 5.254113263861848e-05, "loss": 0.0329, "step": 9795 }, { "epoch": 4.6098823529411765, "grad_norm": 0.4589060639628941, "learning_rate": 5.250781391276684e-05, "loss": 0.2261, "step": 9796 }, { "epoch": 4.610352941176471, "grad_norm": 0.34946996681603, "learning_rate": 5.24745035132262e-05, "loss": 0.0869, "step": 9797 }, { "epoch": 4.610823529411765, "grad_norm": 0.24040718881278766, "learning_rate": 5.2441201442841264e-05, "loss": 0.1221, "step": 9798 }, { "epoch": 4.611294117647059, "grad_norm": 0.18114852023979697, "learning_rate": 5.240790770445625e-05, "loss": 0.0615, "step": 9799 }, { "epoch": 4.6117647058823525, "grad_norm": 0.4260813018170083, "learning_rate": 5.237462230091467e-05, "loss": 0.3229, "step": 9800 }, { "epoch": 4.612235294117647, "grad_norm": 0.8212144158388885, "learning_rate": 5.234134523505912e-05, "loss": 0.3129, "step": 9801 }, { "epoch": 4.612705882352941, "grad_norm": 0.5437693324545401, "learning_rate": 5.230807650973172e-05, "loss": 0.3611, "step": 9802 }, { "epoch": 4.613176470588235, "grad_norm": 0.5485221494431138, "learning_rate": 5.227481612777379e-05, "loss": 0.192, "step": 9803 }, { "epoch": 4.613647058823529, "grad_norm": 0.5468651013370279, "learning_rate": 5.224156409202587e-05, "loss": 0.3434, "step": 9804 }, { "epoch": 4.614117647058824, "grad_norm": 0.22126491068441323, "learning_rate": 5.2208320405327865e-05, "loss": 0.0943, "step": 9805 }, { "epoch": 4.614588235294118, "grad_norm": 0.3759968946755141, "learning_rate": 5.2175085070519024e-05, "loss": 0.1496, "step": 9806 }, { "epoch": 4.615058823529412, "grad_norm": 0.34642633692461533, "learning_rate": 5.2141858090437696e-05, "loss": 0.2431, "step": 9807 }, { "epoch": 4.615529411764706, "grad_norm": 0.27892624802683896, "learning_rate": 5.210863946792172e-05, "loss": 0.1049, "step": 9808 }, { "epoch": 4.616, "grad_norm": 0.2785604092737027, "learning_rate": 5.207542920580805e-05, "loss": 0.0922, "step": 9809 }, { "epoch": 4.616470588235294, "grad_norm": 0.37346160930667616, "learning_rate": 5.204222730693303e-05, "loss": 0.1811, "step": 9810 }, { "epoch": 4.616941176470588, "grad_norm": 0.15572625899648412, "learning_rate": 5.2009033774132345e-05, "loss": 0.0539, "step": 9811 }, { "epoch": 4.617411764705882, "grad_norm": 0.390267632557944, "learning_rate": 5.197584861024076e-05, "loss": 0.2408, "step": 9812 }, { "epoch": 4.6178823529411765, "grad_norm": 0.8935615548866752, "learning_rate": 5.194267181809251e-05, "loss": 0.3851, "step": 9813 }, { "epoch": 4.618352941176471, "grad_norm": 0.3509622764842525, "learning_rate": 5.1909503400521104e-05, "loss": 0.1628, "step": 9814 }, { "epoch": 4.618823529411765, "grad_norm": 0.6195122676717274, "learning_rate": 5.187634336035918e-05, "loss": 0.2166, "step": 9815 }, { "epoch": 4.619294117647058, "grad_norm": 0.23156435570887438, "learning_rate": 5.184319170043884e-05, "loss": 0.0697, "step": 9816 }, { "epoch": 4.619764705882353, "grad_norm": 0.5666920436511604, "learning_rate": 5.1810048423591424e-05, "loss": 0.2418, "step": 9817 }, { "epoch": 4.620235294117647, "grad_norm": 0.1507738292069964, "learning_rate": 5.177691353264746e-05, "loss": 0.0321, "step": 9818 }, { "epoch": 4.620705882352941, "grad_norm": 0.1952556019715859, "learning_rate": 5.174378703043684e-05, "loss": 0.0517, "step": 9819 }, { "epoch": 4.621176470588235, "grad_norm": 0.13034068688577574, "learning_rate": 5.171066891978881e-05, "loss": 0.021, "step": 9820 }, { "epoch": 4.621647058823529, "grad_norm": 0.49513028581724555, "learning_rate": 5.167755920353171e-05, "loss": 0.2015, "step": 9821 }, { "epoch": 4.622117647058824, "grad_norm": 0.4587120765934857, "learning_rate": 5.164445788449334e-05, "loss": 0.213, "step": 9822 }, { "epoch": 4.622588235294118, "grad_norm": 0.46102660144509944, "learning_rate": 5.161136496550072e-05, "loss": 0.2166, "step": 9823 }, { "epoch": 4.623058823529412, "grad_norm": 0.23861550339627058, "learning_rate": 5.15782804493801e-05, "loss": 0.09, "step": 9824 }, { "epoch": 4.623529411764705, "grad_norm": 0.24691433511554742, "learning_rate": 5.1545204338957066e-05, "loss": 0.1232, "step": 9825 }, { "epoch": 4.624, "grad_norm": 0.3541305219415661, "learning_rate": 5.151213663705655e-05, "loss": 0.103, "step": 9826 }, { "epoch": 4.624470588235294, "grad_norm": 0.5153940772179609, "learning_rate": 5.14790773465026e-05, "loss": 0.3742, "step": 9827 }, { "epoch": 4.624941176470588, "grad_norm": 0.44581337902404805, "learning_rate": 5.144602647011874e-05, "loss": 0.3205, "step": 9828 }, { "epoch": 4.625411764705882, "grad_norm": 0.4336327090673481, "learning_rate": 5.141298401072758e-05, "loss": 0.3154, "step": 9829 }, { "epoch": 4.625882352941177, "grad_norm": 0.3672555749896108, "learning_rate": 5.137994997115115e-05, "loss": 0.229, "step": 9830 }, { "epoch": 4.626352941176471, "grad_norm": 0.47325738400347866, "learning_rate": 5.134692435421077e-05, "loss": 0.2657, "step": 9831 }, { "epoch": 4.626823529411765, "grad_norm": 0.4108606597176942, "learning_rate": 5.1313907162726883e-05, "loss": 0.2012, "step": 9832 }, { "epoch": 4.627294117647059, "grad_norm": 0.3161498592121057, "learning_rate": 5.128089839951937e-05, "loss": 0.1079, "step": 9833 }, { "epoch": 4.627764705882353, "grad_norm": 0.6298189315114371, "learning_rate": 5.124789806740742e-05, "loss": 0.4654, "step": 9834 }, { "epoch": 4.628235294117647, "grad_norm": 0.1660612995302782, "learning_rate": 5.12149061692093e-05, "loss": 0.0225, "step": 9835 }, { "epoch": 4.628705882352941, "grad_norm": 0.3089424489407268, "learning_rate": 5.118192270774271e-05, "loss": 0.0938, "step": 9836 }, { "epoch": 4.629176470588235, "grad_norm": 0.39626314576187305, "learning_rate": 5.114894768582468e-05, "loss": 0.1598, "step": 9837 }, { "epoch": 4.6296470588235294, "grad_norm": 0.7663864672680262, "learning_rate": 5.111598110627133e-05, "loss": 0.4445, "step": 9838 }, { "epoch": 4.630117647058824, "grad_norm": 0.3642637176043255, "learning_rate": 5.108302297189822e-05, "loss": 0.1778, "step": 9839 }, { "epoch": 4.630588235294118, "grad_norm": 0.07281068618350248, "learning_rate": 5.105007328552019e-05, "loss": 0.0164, "step": 9840 }, { "epoch": 4.631058823529412, "grad_norm": 0.4088469871092087, "learning_rate": 5.101713204995114e-05, "loss": 0.2557, "step": 9841 }, { "epoch": 4.631529411764706, "grad_norm": 0.4031567652914063, "learning_rate": 5.0984199268004615e-05, "loss": 0.0984, "step": 9842 }, { "epoch": 4.632, "grad_norm": 0.5560575206547176, "learning_rate": 5.095127494249309e-05, "loss": 0.1854, "step": 9843 }, { "epoch": 4.632470588235294, "grad_norm": 0.04461552703892845, "learning_rate": 5.091835907622854e-05, "loss": 0.0047, "step": 9844 }, { "epoch": 4.632941176470588, "grad_norm": 0.4039143056363369, "learning_rate": 5.0885451672022156e-05, "loss": 0.2134, "step": 9845 }, { "epoch": 4.633411764705882, "grad_norm": 0.6383605840259277, "learning_rate": 5.0852552732684296e-05, "loss": 0.3196, "step": 9846 }, { "epoch": 4.633882352941177, "grad_norm": 0.3854181801309888, "learning_rate": 5.081966226102474e-05, "loss": 0.236, "step": 9847 }, { "epoch": 4.634352941176471, "grad_norm": 0.24112184970173745, "learning_rate": 5.078678025985257e-05, "loss": 0.1167, "step": 9848 }, { "epoch": 4.634823529411765, "grad_norm": 0.39731793476731675, "learning_rate": 5.0753906731975954e-05, "loss": 0.2077, "step": 9849 }, { "epoch": 4.635294117647058, "grad_norm": 0.3602251580616986, "learning_rate": 5.07210416802025e-05, "loss": 0.1525, "step": 9850 }, { "epoch": 4.635764705882353, "grad_norm": 0.15624445874266343, "learning_rate": 5.0688185107339084e-05, "loss": 0.0187, "step": 9851 }, { "epoch": 4.636235294117647, "grad_norm": 0.3289326501843935, "learning_rate": 5.065533701619175e-05, "loss": 0.1046, "step": 9852 }, { "epoch": 4.636705882352941, "grad_norm": 0.2943489569694727, "learning_rate": 5.06224974095659e-05, "loss": 0.0936, "step": 9853 }, { "epoch": 4.637176470588235, "grad_norm": 0.47650571892309534, "learning_rate": 5.058966629026629e-05, "loss": 0.2816, "step": 9854 }, { "epoch": 4.6376470588235295, "grad_norm": 0.373822361334689, "learning_rate": 5.0556843661096714e-05, "loss": 0.1785, "step": 9855 }, { "epoch": 4.638117647058824, "grad_norm": 0.34889960330629116, "learning_rate": 5.052402952486046e-05, "loss": 0.1296, "step": 9856 }, { "epoch": 4.638588235294118, "grad_norm": 0.5958261258407868, "learning_rate": 5.0491223884360054e-05, "loss": 0.3541, "step": 9857 }, { "epoch": 4.639058823529412, "grad_norm": 0.584970923344568, "learning_rate": 5.0458426742397174e-05, "loss": 0.2323, "step": 9858 }, { "epoch": 4.6395294117647055, "grad_norm": 0.5080842774141624, "learning_rate": 5.042563810177296e-05, "loss": 0.3011, "step": 9859 }, { "epoch": 4.64, "grad_norm": 0.39880118953298277, "learning_rate": 5.039285796528759e-05, "loss": 0.2009, "step": 9860 }, { "epoch": 4.640470588235294, "grad_norm": 0.2613854219729079, "learning_rate": 5.0360086335740734e-05, "loss": 0.0996, "step": 9861 }, { "epoch": 4.640941176470588, "grad_norm": 0.5249788032059145, "learning_rate": 5.032732321593128e-05, "loss": 0.0812, "step": 9862 }, { "epoch": 4.641411764705882, "grad_norm": 0.39491276441604456, "learning_rate": 5.029456860865727e-05, "loss": 0.1475, "step": 9863 }, { "epoch": 4.641882352941177, "grad_norm": 0.2091342613548217, "learning_rate": 5.0261822516716146e-05, "loss": 0.0683, "step": 9864 }, { "epoch": 4.642352941176471, "grad_norm": 0.20679225109078558, "learning_rate": 5.022908494290465e-05, "loss": 0.0514, "step": 9865 }, { "epoch": 4.642823529411765, "grad_norm": 0.3707438697249213, "learning_rate": 5.0196355890018605e-05, "loss": 0.1069, "step": 9866 }, { "epoch": 4.643294117647059, "grad_norm": 0.2813004227968876, "learning_rate": 5.016363536085333e-05, "loss": 0.1514, "step": 9867 }, { "epoch": 4.643764705882353, "grad_norm": 0.4546530504593565, "learning_rate": 5.0130923358203316e-05, "loss": 0.3031, "step": 9868 }, { "epoch": 4.644235294117647, "grad_norm": 0.5001357051516875, "learning_rate": 5.0098219884862265e-05, "loss": 0.1206, "step": 9869 }, { "epoch": 4.644705882352941, "grad_norm": 0.05852719081306149, "learning_rate": 5.006552494362325e-05, "loss": 0.0072, "step": 9870 }, { "epoch": 4.645176470588235, "grad_norm": 0.3250181701663816, "learning_rate": 5.0032838537278627e-05, "loss": 0.1202, "step": 9871 }, { "epoch": 4.6456470588235295, "grad_norm": 0.4140653212321906, "learning_rate": 5.000016066861988e-05, "loss": 0.2119, "step": 9872 }, { "epoch": 4.646117647058824, "grad_norm": 0.48003570086141756, "learning_rate": 4.9967491340437904e-05, "loss": 0.2813, "step": 9873 }, { "epoch": 4.646588235294118, "grad_norm": 0.41681702926106556, "learning_rate": 4.993483055552289e-05, "loss": 0.1947, "step": 9874 }, { "epoch": 4.647058823529412, "grad_norm": 0.27928449804674804, "learning_rate": 4.990217831666409e-05, "loss": 0.0851, "step": 9875 }, { "epoch": 4.647529411764706, "grad_norm": 0.5327538503397801, "learning_rate": 4.986953462665026e-05, "loss": 0.2388, "step": 9876 }, { "epoch": 4.648, "grad_norm": 0.40099631527706053, "learning_rate": 4.9836899488269345e-05, "loss": 0.1452, "step": 9877 }, { "epoch": 4.648470588235294, "grad_norm": 0.2544756471828298, "learning_rate": 4.9804272904308465e-05, "loss": 0.0937, "step": 9878 }, { "epoch": 4.648941176470588, "grad_norm": 1.2417783878040778, "learning_rate": 4.97716548775542e-05, "loss": 0.159, "step": 9879 }, { "epoch": 4.649411764705882, "grad_norm": 0.6137018594464113, "learning_rate": 4.9739045410792164e-05, "loss": 0.3756, "step": 9880 }, { "epoch": 4.649882352941177, "grad_norm": 0.2708973833468867, "learning_rate": 4.970644450680742e-05, "loss": 0.0785, "step": 9881 }, { "epoch": 4.650352941176471, "grad_norm": 0.35997306944490864, "learning_rate": 4.9673852168384316e-05, "loss": 0.2259, "step": 9882 }, { "epoch": 4.650823529411765, "grad_norm": 0.4194384500503605, "learning_rate": 4.964126839830628e-05, "loss": 0.2242, "step": 9883 }, { "epoch": 4.651294117647058, "grad_norm": 0.4145638539338187, "learning_rate": 4.960869319935616e-05, "loss": 0.2801, "step": 9884 }, { "epoch": 4.651764705882353, "grad_norm": 0.5247401609909471, "learning_rate": 4.9576126574316086e-05, "loss": 0.2203, "step": 9885 }, { "epoch": 4.652235294117647, "grad_norm": 0.33297672397239114, "learning_rate": 4.9543568525967376e-05, "loss": 0.2051, "step": 9886 }, { "epoch": 4.652705882352941, "grad_norm": 0.4080762494973298, "learning_rate": 4.9511019057090684e-05, "loss": 0.1748, "step": 9887 }, { "epoch": 4.653176470588235, "grad_norm": 0.3492710354406761, "learning_rate": 4.947847817046582e-05, "loss": 0.1186, "step": 9888 }, { "epoch": 4.6536470588235295, "grad_norm": 0.43448779640773827, "learning_rate": 4.9445945868871974e-05, "loss": 0.2176, "step": 9889 }, { "epoch": 4.654117647058824, "grad_norm": 0.44518595086097346, "learning_rate": 4.941342215508761e-05, "loss": 0.2728, "step": 9890 }, { "epoch": 4.654588235294118, "grad_norm": 0.3285413460115125, "learning_rate": 4.9380907031890315e-05, "loss": 0.1309, "step": 9891 }, { "epoch": 4.655058823529412, "grad_norm": 0.3204267978896303, "learning_rate": 4.934840050205711e-05, "loss": 0.1839, "step": 9892 }, { "epoch": 4.6555294117647055, "grad_norm": 0.4983442321095853, "learning_rate": 4.931590256836424e-05, "loss": 0.3784, "step": 9893 }, { "epoch": 4.656, "grad_norm": 0.3364982564671835, "learning_rate": 4.928341323358708e-05, "loss": 0.219, "step": 9894 }, { "epoch": 4.656470588235294, "grad_norm": 0.2954284048405545, "learning_rate": 4.925093250050045e-05, "loss": 0.1464, "step": 9895 }, { "epoch": 4.656941176470588, "grad_norm": 0.4451521537575521, "learning_rate": 4.921846037187841e-05, "loss": 0.3531, "step": 9896 }, { "epoch": 4.657411764705882, "grad_norm": 0.19149160293725118, "learning_rate": 4.918599685049414e-05, "loss": 0.0837, "step": 9897 }, { "epoch": 4.657882352941177, "grad_norm": 0.48922155766786857, "learning_rate": 4.915354193912021e-05, "loss": 0.3266, "step": 9898 }, { "epoch": 4.658352941176471, "grad_norm": 1.653951111807803, "learning_rate": 4.9121095640528516e-05, "loss": 0.5147, "step": 9899 }, { "epoch": 4.658823529411765, "grad_norm": 0.4206730318800821, "learning_rate": 4.908865795748999e-05, "loss": 0.3039, "step": 9900 }, { "epoch": 4.659294117647059, "grad_norm": 0.3129293593187972, "learning_rate": 4.905622889277506e-05, "loss": 0.1828, "step": 9901 }, { "epoch": 4.659764705882353, "grad_norm": 0.2526830031859977, "learning_rate": 4.902380844915337e-05, "loss": 0.1064, "step": 9902 }, { "epoch": 4.660235294117647, "grad_norm": 0.5565508236996035, "learning_rate": 4.899139662939366e-05, "loss": 0.4265, "step": 9903 }, { "epoch": 4.660705882352941, "grad_norm": 0.16147606823390973, "learning_rate": 4.895899343626412e-05, "loss": 0.0532, "step": 9904 }, { "epoch": 4.661176470588235, "grad_norm": 0.4285780655641153, "learning_rate": 4.892659887253219e-05, "loss": 0.1923, "step": 9905 }, { "epoch": 4.6616470588235295, "grad_norm": 0.3415159740338085, "learning_rate": 4.8894212940964445e-05, "loss": 0.1947, "step": 9906 }, { "epoch": 4.662117647058824, "grad_norm": 0.7083281581739285, "learning_rate": 4.8861835644326866e-05, "loss": 0.4192, "step": 9907 }, { "epoch": 4.662588235294118, "grad_norm": 0.38963801522005015, "learning_rate": 4.882946698538457e-05, "loss": 0.2142, "step": 9908 }, { "epoch": 4.663058823529412, "grad_norm": 0.4707508647407432, "learning_rate": 4.879710696690202e-05, "loss": 0.3125, "step": 9909 }, { "epoch": 4.663529411764706, "grad_norm": 0.21837393957243628, "learning_rate": 4.876475559164299e-05, "loss": 0.0906, "step": 9910 }, { "epoch": 4.664, "grad_norm": 0.3981990158751158, "learning_rate": 4.873241286237033e-05, "loss": 0.184, "step": 9911 }, { "epoch": 4.664470588235294, "grad_norm": 0.2897003081127668, "learning_rate": 4.8700078781846326e-05, "loss": 0.1179, "step": 9912 }, { "epoch": 4.664941176470588, "grad_norm": 0.3414348084814198, "learning_rate": 4.866775335283253e-05, "loss": 0.1715, "step": 9913 }, { "epoch": 4.665411764705882, "grad_norm": 0.30109441759178884, "learning_rate": 4.863543657808956e-05, "loss": 0.1252, "step": 9914 }, { "epoch": 4.665882352941177, "grad_norm": 0.5011801491795188, "learning_rate": 4.860312846037753e-05, "loss": 0.3284, "step": 9915 }, { "epoch": 4.666352941176471, "grad_norm": 0.5468364333668028, "learning_rate": 4.857082900245569e-05, "loss": 0.2476, "step": 9916 }, { "epoch": 4.666823529411765, "grad_norm": 0.25798634556361516, "learning_rate": 4.8538538207082534e-05, "loss": 0.105, "step": 9917 }, { "epoch": 4.667294117647058, "grad_norm": 0.2947526188244581, "learning_rate": 4.850625607701588e-05, "loss": 0.1261, "step": 9918 }, { "epoch": 4.667764705882353, "grad_norm": 0.2680696940945275, "learning_rate": 4.8473982615012816e-05, "loss": 0.129, "step": 9919 }, { "epoch": 4.668235294117647, "grad_norm": 0.2911841227807385, "learning_rate": 4.844171782382959e-05, "loss": 0.1099, "step": 9920 }, { "epoch": 4.668705882352941, "grad_norm": 6.047106615219814e-05, "learning_rate": 4.840946170622179e-05, "loss": 0.0, "step": 9921 }, { "epoch": 4.669176470588235, "grad_norm": 0.440337143452538, "learning_rate": 4.8377214264944296e-05, "loss": 0.2632, "step": 9922 }, { "epoch": 4.6696470588235295, "grad_norm": 0.35589529765866823, "learning_rate": 4.834497550275114e-05, "loss": 0.1902, "step": 9923 }, { "epoch": 4.670117647058824, "grad_norm": 0.2518062499818046, "learning_rate": 4.831274542239566e-05, "loss": 0.0895, "step": 9924 }, { "epoch": 4.670588235294118, "grad_norm": 0.26656136455407997, "learning_rate": 4.8280524026630565e-05, "loss": 0.0589, "step": 9925 }, { "epoch": 4.671058823529412, "grad_norm": 0.43253105621724225, "learning_rate": 4.824831131820759e-05, "loss": 0.2466, "step": 9926 }, { "epoch": 4.6715294117647055, "grad_norm": 0.42051354448564143, "learning_rate": 4.821610729987792e-05, "loss": 0.3032, "step": 9927 }, { "epoch": 4.672, "grad_norm": 0.3087033827051453, "learning_rate": 4.818391197439199e-05, "loss": 0.1791, "step": 9928 }, { "epoch": 4.672470588235294, "grad_norm": 0.2886984744734186, "learning_rate": 4.815172534449933e-05, "loss": 0.0998, "step": 9929 }, { "epoch": 4.672941176470588, "grad_norm": 0.45938598822354404, "learning_rate": 4.811954741294889e-05, "loss": 0.273, "step": 9930 }, { "epoch": 4.673411764705882, "grad_norm": 0.2212243480702187, "learning_rate": 4.808737818248882e-05, "loss": 0.0758, "step": 9931 }, { "epoch": 4.673882352941177, "grad_norm": 0.35623013263055403, "learning_rate": 4.805521765586653e-05, "loss": 0.2277, "step": 9932 }, { "epoch": 4.674352941176471, "grad_norm": 0.31560540790906516, "learning_rate": 4.802306583582873e-05, "loss": 0.1178, "step": 9933 }, { "epoch": 4.674823529411765, "grad_norm": 0.42340022853254305, "learning_rate": 4.7990922725121266e-05, "loss": 0.1962, "step": 9934 }, { "epoch": 4.675294117647059, "grad_norm": 0.39392115703850833, "learning_rate": 4.795878832648935e-05, "loss": 0.2123, "step": 9935 }, { "epoch": 4.675764705882353, "grad_norm": 0.49831652908760826, "learning_rate": 4.7926662642677466e-05, "loss": 0.4359, "step": 9936 }, { "epoch": 4.676235294117647, "grad_norm": 0.1551606046967066, "learning_rate": 4.7894545676429206e-05, "loss": 0.0315, "step": 9937 }, { "epoch": 4.676705882352941, "grad_norm": 0.1304630888297053, "learning_rate": 4.7862437430487635e-05, "loss": 0.0377, "step": 9938 }, { "epoch": 4.677176470588235, "grad_norm": 0.03074356327036578, "learning_rate": 4.783033790759481e-05, "loss": 0.0022, "step": 9939 }, { "epoch": 4.6776470588235295, "grad_norm": 0.3565456604749199, "learning_rate": 4.77982471104923e-05, "loss": 0.1453, "step": 9940 }, { "epoch": 4.678117647058824, "grad_norm": 0.3954631764968829, "learning_rate": 4.776616504192082e-05, "loss": 0.2204, "step": 9941 }, { "epoch": 4.678588235294118, "grad_norm": 0.4096766126581352, "learning_rate": 4.7734091704620244e-05, "loss": 0.1396, "step": 9942 }, { "epoch": 4.679058823529411, "grad_norm": 0.3487307819624178, "learning_rate": 4.7702027101329846e-05, "loss": 0.1798, "step": 9943 }, { "epoch": 4.679529411764706, "grad_norm": 0.3745222234129374, "learning_rate": 4.7669971234788165e-05, "loss": 0.1713, "step": 9944 }, { "epoch": 4.68, "grad_norm": 0.39002197225701735, "learning_rate": 4.763792410773281e-05, "loss": 0.1436, "step": 9945 }, { "epoch": 4.680470588235294, "grad_norm": 0.3965183753815438, "learning_rate": 4.760588572290082e-05, "loss": 0.2674, "step": 9946 }, { "epoch": 4.680941176470588, "grad_norm": 0.1712834669900534, "learning_rate": 4.7573856083028475e-05, "loss": 0.0547, "step": 9947 }, { "epoch": 4.681411764705882, "grad_norm": 0.43016873677470163, "learning_rate": 4.7541835190851175e-05, "loss": 0.2585, "step": 9948 }, { "epoch": 4.681882352941177, "grad_norm": 0.5341753515210498, "learning_rate": 4.7509823049103706e-05, "loss": 0.3284, "step": 9949 }, { "epoch": 4.682352941176471, "grad_norm": 0.3939805228456799, "learning_rate": 4.74778196605201e-05, "loss": 0.2294, "step": 9950 }, { "epoch": 4.682823529411765, "grad_norm": 0.3986578129944384, "learning_rate": 4.744582502783352e-05, "loss": 0.3385, "step": 9951 }, { "epoch": 4.683294117647058, "grad_norm": 1.3309353916760205, "learning_rate": 4.7413839153776504e-05, "loss": 0.737, "step": 9952 }, { "epoch": 4.683764705882353, "grad_norm": 0.1259313739148576, "learning_rate": 4.7381862041080865e-05, "loss": 0.0234, "step": 9953 }, { "epoch": 4.684235294117647, "grad_norm": 0.5801125106410205, "learning_rate": 4.7349893692477495e-05, "loss": 0.3978, "step": 9954 }, { "epoch": 4.684705882352941, "grad_norm": 0.5914617177142223, "learning_rate": 4.7317934110696685e-05, "loss": 0.4046, "step": 9955 }, { "epoch": 4.685176470588235, "grad_norm": 0.4123738210873601, "learning_rate": 4.728598329846802e-05, "loss": 0.184, "step": 9956 }, { "epoch": 4.6856470588235295, "grad_norm": 0.2440248742917708, "learning_rate": 4.725404125852013e-05, "loss": 0.1204, "step": 9957 }, { "epoch": 4.686117647058824, "grad_norm": 0.4604969045716509, "learning_rate": 4.7222107993581135e-05, "loss": 0.2645, "step": 9958 }, { "epoch": 4.686588235294118, "grad_norm": 0.5095958212805238, "learning_rate": 4.7190183506378205e-05, "loss": 0.4307, "step": 9959 }, { "epoch": 4.687058823529412, "grad_norm": 0.48385639566946254, "learning_rate": 4.7158267799637864e-05, "loss": 0.304, "step": 9960 }, { "epoch": 4.6875294117647055, "grad_norm": 0.4104123527576861, "learning_rate": 4.712636087608594e-05, "loss": 0.2681, "step": 9961 }, { "epoch": 4.688, "grad_norm": 0.20121591704639252, "learning_rate": 4.709446273844736e-05, "loss": 0.0789, "step": 9962 }, { "epoch": 4.688470588235294, "grad_norm": 0.32359451255325267, "learning_rate": 4.7062573389446395e-05, "loss": 0.1577, "step": 9963 }, { "epoch": 4.688941176470588, "grad_norm": 0.3419927344429019, "learning_rate": 4.70306928318066e-05, "loss": 0.1026, "step": 9964 }, { "epoch": 4.689411764705882, "grad_norm": 0.42678279164224536, "learning_rate": 4.6998821068250675e-05, "loss": 0.1849, "step": 9965 }, { "epoch": 4.689882352941177, "grad_norm": 0.4597909845751141, "learning_rate": 4.696695810150065e-05, "loss": 0.2768, "step": 9966 }, { "epoch": 4.690352941176471, "grad_norm": 0.35450080828178, "learning_rate": 4.693510393427781e-05, "loss": 0.1865, "step": 9967 }, { "epoch": 4.690823529411765, "grad_norm": 0.2243626397255751, "learning_rate": 4.690325856930258e-05, "loss": 0.0729, "step": 9968 }, { "epoch": 4.691294117647059, "grad_norm": 0.3334480721184569, "learning_rate": 4.6871422009294764e-05, "loss": 0.1188, "step": 9969 }, { "epoch": 4.691764705882353, "grad_norm": 0.3730717304564639, "learning_rate": 4.6839594256973406e-05, "loss": 0.0567, "step": 9970 }, { "epoch": 4.692235294117647, "grad_norm": 1.1940513073458618, "learning_rate": 4.680777531505662e-05, "loss": 0.4583, "step": 9971 }, { "epoch": 4.692705882352941, "grad_norm": 0.45047433311503843, "learning_rate": 4.6775965186262e-05, "loss": 0.2197, "step": 9972 }, { "epoch": 4.693176470588235, "grad_norm": 0.5016724474727889, "learning_rate": 4.6744163873306314e-05, "loss": 0.3827, "step": 9973 }, { "epoch": 4.6936470588235295, "grad_norm": 0.21417204426559494, "learning_rate": 4.671237137890539e-05, "loss": 0.0414, "step": 9974 }, { "epoch": 4.694117647058824, "grad_norm": 0.3741705599525189, "learning_rate": 4.668058770577468e-05, "loss": 0.3149, "step": 9975 }, { "epoch": 4.694588235294118, "grad_norm": 0.4485386903853037, "learning_rate": 4.6648812856628515e-05, "loss": 0.2087, "step": 9976 }, { "epoch": 4.695058823529411, "grad_norm": 0.4729420671538785, "learning_rate": 4.6617046834180656e-05, "loss": 0.3508, "step": 9977 }, { "epoch": 4.6955294117647055, "grad_norm": 0.5537533994606078, "learning_rate": 4.658528964114413e-05, "loss": 0.1731, "step": 9978 }, { "epoch": 4.696, "grad_norm": 0.4228234502924856, "learning_rate": 4.655354128023107e-05, "loss": 0.2615, "step": 9979 }, { "epoch": 4.696470588235294, "grad_norm": 0.33580122481371216, "learning_rate": 4.6521801754152995e-05, "loss": 0.1356, "step": 9980 }, { "epoch": 4.696941176470588, "grad_norm": 0.2961863779810404, "learning_rate": 4.649007106562066e-05, "loss": 0.1922, "step": 9981 }, { "epoch": 4.697411764705882, "grad_norm": 0.4891125660054, "learning_rate": 4.645834921734392e-05, "loss": 0.3184, "step": 9982 }, { "epoch": 4.697882352941177, "grad_norm": 0.5033903877698429, "learning_rate": 4.642663621203201e-05, "loss": 0.1816, "step": 9983 }, { "epoch": 4.698352941176471, "grad_norm": 0.5419446164612876, "learning_rate": 4.639493205239346e-05, "loss": 0.2061, "step": 9984 }, { "epoch": 4.698823529411765, "grad_norm": 0.21795301995722102, "learning_rate": 4.636323674113583e-05, "loss": 0.0619, "step": 9985 }, { "epoch": 4.699294117647058, "grad_norm": 0.35337243566167154, "learning_rate": 4.6331550280966126e-05, "loss": 0.1114, "step": 9986 }, { "epoch": 4.699764705882353, "grad_norm": 0.7397843143102925, "learning_rate": 4.629987267459056e-05, "loss": 0.2572, "step": 9987 }, { "epoch": 4.700235294117647, "grad_norm": 0.30971925414726675, "learning_rate": 4.626820392471446e-05, "loss": 0.0628, "step": 9988 }, { "epoch": 4.700705882352941, "grad_norm": 0.2895303137509899, "learning_rate": 4.6236544034042604e-05, "loss": 0.0644, "step": 9989 }, { "epoch": 4.701176470588235, "grad_norm": 0.4518512076607143, "learning_rate": 4.62048930052788e-05, "loss": 0.2948, "step": 9990 }, { "epoch": 4.7016470588235295, "grad_norm": 0.4021660560046377, "learning_rate": 4.6173250841126245e-05, "loss": 0.2715, "step": 9991 }, { "epoch": 4.702117647058824, "grad_norm": 0.5177793599947312, "learning_rate": 4.614161754428737e-05, "loss": 0.2121, "step": 9992 }, { "epoch": 4.702588235294118, "grad_norm": 0.33494447834673147, "learning_rate": 4.6109993117463756e-05, "loss": 0.1858, "step": 9993 }, { "epoch": 4.703058823529412, "grad_norm": 0.3507721861317219, "learning_rate": 4.607837756335629e-05, "loss": 0.1172, "step": 9994 }, { "epoch": 4.7035294117647055, "grad_norm": 0.45289339201260953, "learning_rate": 4.604677088466516e-05, "loss": 0.2199, "step": 9995 }, { "epoch": 4.704, "grad_norm": 1.202089558224737, "learning_rate": 4.6015173084089635e-05, "loss": 0.5494, "step": 9996 }, { "epoch": 4.704470588235294, "grad_norm": 0.18803035158949255, "learning_rate": 4.598358416432838e-05, "loss": 0.0746, "step": 9997 }, { "epoch": 4.704941176470588, "grad_norm": 0.36348421269689996, "learning_rate": 4.595200412807927e-05, "loss": 0.2258, "step": 9998 }, { "epoch": 4.705411764705882, "grad_norm": 0.4295558746318011, "learning_rate": 4.592043297803932e-05, "loss": 0.2741, "step": 9999 }, { "epoch": 4.705882352941177, "grad_norm": 0.7942416699596576, "learning_rate": 4.58888707169049e-05, "loss": 0.4461, "step": 10000 }, { "epoch": 4.706352941176471, "grad_norm": 0.5021262677737539, "learning_rate": 4.585731734737164e-05, "loss": 0.4516, "step": 10001 }, { "epoch": 4.706823529411765, "grad_norm": 0.32566998225264715, "learning_rate": 4.5825772872134234e-05, "loss": 0.1241, "step": 10002 }, { "epoch": 4.707294117647059, "grad_norm": 0.4152620101144116, "learning_rate": 4.579423729388681e-05, "loss": 0.2579, "step": 10003 }, { "epoch": 4.707764705882353, "grad_norm": 0.4778980903579011, "learning_rate": 4.576271061532268e-05, "loss": 0.2741, "step": 10004 }, { "epoch": 4.708235294117647, "grad_norm": 0.24605363753348797, "learning_rate": 4.573119283913431e-05, "loss": 0.0747, "step": 10005 }, { "epoch": 4.708705882352941, "grad_norm": 0.5313935542447681, "learning_rate": 4.569968396801351e-05, "loss": 0.2929, "step": 10006 }, { "epoch": 4.709176470588235, "grad_norm": 0.545364521044486, "learning_rate": 4.566818400465133e-05, "loss": 0.447, "step": 10007 }, { "epoch": 4.7096470588235295, "grad_norm": 0.5258911178894619, "learning_rate": 4.563669295173793e-05, "loss": 0.4022, "step": 10008 }, { "epoch": 4.710117647058824, "grad_norm": 0.25052236238826886, "learning_rate": 4.560521081196289e-05, "loss": 0.0549, "step": 10009 }, { "epoch": 4.710588235294118, "grad_norm": 0.4255832367302741, "learning_rate": 4.5573737588014875e-05, "loss": 0.2601, "step": 10010 }, { "epoch": 4.711058823529411, "grad_norm": 0.6047736694040361, "learning_rate": 4.554227328258185e-05, "loss": 0.3273, "step": 10011 }, { "epoch": 4.7115294117647055, "grad_norm": 0.25719443577509576, "learning_rate": 4.551081789835113e-05, "loss": 0.0908, "step": 10012 }, { "epoch": 4.712, "grad_norm": 0.5173720002888637, "learning_rate": 4.547937143800901e-05, "loss": 0.2777, "step": 10013 }, { "epoch": 4.712470588235294, "grad_norm": 0.3989265123704767, "learning_rate": 4.544793390424124e-05, "loss": 0.1888, "step": 10014 }, { "epoch": 4.712941176470588, "grad_norm": 0.13020339038142859, "learning_rate": 4.541650529973279e-05, "loss": 0.0326, "step": 10015 }, { "epoch": 4.713411764705882, "grad_norm": 0.4622922417149946, "learning_rate": 4.5385085627167717e-05, "loss": 0.1583, "step": 10016 }, { "epoch": 4.713882352941177, "grad_norm": 0.334032946962786, "learning_rate": 4.535367488922947e-05, "loss": 0.176, "step": 10017 }, { "epoch": 4.714352941176471, "grad_norm": 0.5666195541580346, "learning_rate": 4.5322273088600714e-05, "loss": 0.3883, "step": 10018 }, { "epoch": 4.714823529411765, "grad_norm": 0.264425265186806, "learning_rate": 4.5290880227963186e-05, "loss": 0.0702, "step": 10019 }, { "epoch": 4.715294117647058, "grad_norm": 0.3950418330209938, "learning_rate": 4.525949630999818e-05, "loss": 0.2096, "step": 10020 }, { "epoch": 4.715764705882353, "grad_norm": 0.27857255598050135, "learning_rate": 4.522812133738589e-05, "loss": 0.0936, "step": 10021 }, { "epoch": 4.716235294117647, "grad_norm": 0.47429020308342285, "learning_rate": 4.519675531280594e-05, "loss": 0.1899, "step": 10022 }, { "epoch": 4.716705882352941, "grad_norm": 0.3941274855151763, "learning_rate": 4.5165398238937194e-05, "loss": 0.1892, "step": 10023 }, { "epoch": 4.717176470588235, "grad_norm": 0.6389270223581831, "learning_rate": 4.51340501184576e-05, "loss": 0.3408, "step": 10024 }, { "epoch": 4.7176470588235295, "grad_norm": 0.25735329126776263, "learning_rate": 4.510271095404449e-05, "loss": 0.1045, "step": 10025 }, { "epoch": 4.718117647058824, "grad_norm": 0.3538348719591191, "learning_rate": 4.507138074837443e-05, "loss": 0.1894, "step": 10026 }, { "epoch": 4.718588235294118, "grad_norm": 0.45582683746156416, "learning_rate": 4.504005950412308e-05, "loss": 0.2646, "step": 10027 }, { "epoch": 4.719058823529412, "grad_norm": 0.7143301762074746, "learning_rate": 4.5008747223965456e-05, "loss": 0.2365, "step": 10028 }, { "epoch": 4.7195294117647055, "grad_norm": 0.26281254096796053, "learning_rate": 4.4977443910575866e-05, "loss": 0.0928, "step": 10029 }, { "epoch": 4.72, "grad_norm": 0.3266267649330609, "learning_rate": 4.494614956662765e-05, "loss": 0.1644, "step": 10030 }, { "epoch": 4.720470588235294, "grad_norm": 0.5000357260291199, "learning_rate": 4.491486419479355e-05, "loss": 0.3071, "step": 10031 }, { "epoch": 4.720941176470588, "grad_norm": 0.5593294490153884, "learning_rate": 4.488358779774551e-05, "loss": 0.3713, "step": 10032 }, { "epoch": 4.721411764705882, "grad_norm": 0.3676157427453878, "learning_rate": 4.485232037815465e-05, "loss": 0.1216, "step": 10033 }, { "epoch": 4.721882352941177, "grad_norm": 0.19832391246016406, "learning_rate": 4.4821061938691355e-05, "loss": 0.0406, "step": 10034 }, { "epoch": 4.722352941176471, "grad_norm": 0.5537952841586404, "learning_rate": 4.478981248202533e-05, "loss": 0.2232, "step": 10035 }, { "epoch": 4.722823529411765, "grad_norm": 0.7417763055345521, "learning_rate": 4.4758572010825324e-05, "loss": 0.2549, "step": 10036 }, { "epoch": 4.723294117647059, "grad_norm": 0.23542307188515785, "learning_rate": 4.472734052775947e-05, "loss": 0.0985, "step": 10037 }, { "epoch": 4.723764705882353, "grad_norm": 0.28140334037495857, "learning_rate": 4.4696118035495136e-05, "loss": 0.0471, "step": 10038 }, { "epoch": 4.724235294117647, "grad_norm": 0.533942832808445, "learning_rate": 4.4664904536698796e-05, "loss": 0.2742, "step": 10039 }, { "epoch": 4.724705882352941, "grad_norm": 0.41694087968518545, "learning_rate": 4.463370003403632e-05, "loss": 0.1789, "step": 10040 }, { "epoch": 4.725176470588235, "grad_norm": 0.42863016050704766, "learning_rate": 4.4602504530172636e-05, "loss": 0.2379, "step": 10041 }, { "epoch": 4.7256470588235295, "grad_norm": 0.359608671964593, "learning_rate": 4.457131802777202e-05, "loss": 0.1344, "step": 10042 }, { "epoch": 4.726117647058824, "grad_norm": 0.69581307847967, "learning_rate": 4.454014052949804e-05, "loss": 0.3523, "step": 10043 }, { "epoch": 4.726588235294118, "grad_norm": 0.3948074965178209, "learning_rate": 4.4508972038013274e-05, "loss": 0.1659, "step": 10044 }, { "epoch": 4.727058823529411, "grad_norm": 0.6198224524427591, "learning_rate": 4.4477812555979724e-05, "loss": 0.2225, "step": 10045 }, { "epoch": 4.7275294117647055, "grad_norm": 0.11916418448994127, "learning_rate": 4.4446662086058634e-05, "loss": 0.0261, "step": 10046 }, { "epoch": 4.728, "grad_norm": 0.644544456890162, "learning_rate": 4.4415520630910265e-05, "loss": 0.3707, "step": 10047 }, { "epoch": 4.728470588235294, "grad_norm": 0.4542394849819812, "learning_rate": 4.438438819319434e-05, "loss": 0.2315, "step": 10048 }, { "epoch": 4.728941176470588, "grad_norm": 0.19540855951135705, "learning_rate": 4.435326477556975e-05, "loss": 0.0644, "step": 10049 }, { "epoch": 4.729411764705882, "grad_norm": 0.5124886663734131, "learning_rate": 4.432215038069449e-05, "loss": 0.2934, "step": 10050 }, { "epoch": 4.729882352941177, "grad_norm": 0.44053520327177287, "learning_rate": 4.4291045011225934e-05, "loss": 0.2951, "step": 10051 }, { "epoch": 4.730352941176471, "grad_norm": 0.6765590330437233, "learning_rate": 4.4259948669820676e-05, "loss": 0.3797, "step": 10052 }, { "epoch": 4.730823529411765, "grad_norm": 0.23635557750809696, "learning_rate": 4.422886135913441e-05, "loss": 0.1039, "step": 10053 }, { "epoch": 4.731294117647058, "grad_norm": 0.5514257860518773, "learning_rate": 4.419778308182219e-05, "loss": 0.2799, "step": 10054 }, { "epoch": 4.731764705882353, "grad_norm": 0.6381032713246536, "learning_rate": 4.41667138405383e-05, "loss": 0.4884, "step": 10055 }, { "epoch": 4.732235294117647, "grad_norm": 0.387262286728041, "learning_rate": 4.413565363793611e-05, "loss": 0.2397, "step": 10056 }, { "epoch": 4.732705882352941, "grad_norm": 0.5111149262745124, "learning_rate": 4.4104602476668364e-05, "loss": 0.4017, "step": 10057 }, { "epoch": 4.733176470588235, "grad_norm": 0.49378269869595576, "learning_rate": 4.4073560359387035e-05, "loss": 0.2128, "step": 10058 }, { "epoch": 4.7336470588235295, "grad_norm": 0.474283441903047, "learning_rate": 4.4042527288743155e-05, "loss": 0.1454, "step": 10059 }, { "epoch": 4.734117647058824, "grad_norm": 0.1131479820189961, "learning_rate": 4.401150326738722e-05, "loss": 0.0379, "step": 10060 }, { "epoch": 4.734588235294118, "grad_norm": 0.5107356082462425, "learning_rate": 4.3980488297968726e-05, "loss": 0.2843, "step": 10061 }, { "epoch": 4.735058823529412, "grad_norm": 0.4315647072365066, "learning_rate": 4.394948238313655e-05, "loss": 0.2664, "step": 10062 }, { "epoch": 4.7355294117647055, "grad_norm": 0.10658679964382944, "learning_rate": 4.391848552553875e-05, "loss": 0.0298, "step": 10063 }, { "epoch": 4.736, "grad_norm": 0.17768990018027817, "learning_rate": 4.3887497727822616e-05, "loss": 0.0487, "step": 10064 }, { "epoch": 4.736470588235294, "grad_norm": 0.40762814933313235, "learning_rate": 4.385651899263467e-05, "loss": 0.2295, "step": 10065 }, { "epoch": 4.736941176470588, "grad_norm": 0.3253112358171402, "learning_rate": 4.3825549322620657e-05, "loss": 0.1953, "step": 10066 }, { "epoch": 4.737411764705882, "grad_norm": 0.6561971899870682, "learning_rate": 4.3794588720425486e-05, "loss": 0.372, "step": 10067 }, { "epoch": 4.737882352941177, "grad_norm": 0.46346853586514547, "learning_rate": 4.3763637188693404e-05, "loss": 0.2842, "step": 10068 }, { "epoch": 4.738352941176471, "grad_norm": 0.28199861299809226, "learning_rate": 4.3732694730067766e-05, "loss": 0.1032, "step": 10069 }, { "epoch": 4.738823529411764, "grad_norm": 0.4825849257395115, "learning_rate": 4.370176134719121e-05, "loss": 0.2981, "step": 10070 }, { "epoch": 4.739294117647058, "grad_norm": 0.294362946244081, "learning_rate": 4.36708370427057e-05, "loss": 0.1317, "step": 10071 }, { "epoch": 4.739764705882353, "grad_norm": 0.20584778833336123, "learning_rate": 4.363992181925218e-05, "loss": 0.0676, "step": 10072 }, { "epoch": 4.740235294117647, "grad_norm": 0.19596961508331842, "learning_rate": 4.3609015679471036e-05, "loss": 0.0587, "step": 10073 }, { "epoch": 4.740705882352941, "grad_norm": 0.3995689278746473, "learning_rate": 4.357811862600185e-05, "loss": 0.2266, "step": 10074 }, { "epoch": 4.741176470588235, "grad_norm": 0.824662631307053, "learning_rate": 4.3547230661483276e-05, "loss": 0.4444, "step": 10075 }, { "epoch": 4.7416470588235295, "grad_norm": 0.5722441482749465, "learning_rate": 4.351635178855336e-05, "loss": 0.2793, "step": 10076 }, { "epoch": 4.742117647058824, "grad_norm": 0.2010013963876618, "learning_rate": 4.348548200984934e-05, "loss": 0.0709, "step": 10077 }, { "epoch": 4.742588235294118, "grad_norm": 0.4696893964521437, "learning_rate": 4.345462132800756e-05, "loss": 0.1565, "step": 10078 }, { "epoch": 4.743058823529411, "grad_norm": 0.10418043626686967, "learning_rate": 4.342376974566372e-05, "loss": 0.0205, "step": 10079 }, { "epoch": 4.7435294117647056, "grad_norm": 0.413443478429639, "learning_rate": 4.339292726545275e-05, "loss": 0.252, "step": 10080 }, { "epoch": 4.744, "grad_norm": 0.4010406121800187, "learning_rate": 4.3362093890008645e-05, "loss": 0.2297, "step": 10081 }, { "epoch": 4.744470588235294, "grad_norm": 0.31389170196592847, "learning_rate": 4.333126962196478e-05, "loss": 0.0935, "step": 10082 }, { "epoch": 4.744941176470588, "grad_norm": 0.33341587071388507, "learning_rate": 4.330045446395374e-05, "loss": 0.1752, "step": 10083 }, { "epoch": 4.745411764705882, "grad_norm": 0.2425457421840934, "learning_rate": 4.3269648418607194e-05, "loss": 0.1723, "step": 10084 }, { "epoch": 4.745882352941177, "grad_norm": 0.42698668177979093, "learning_rate": 4.32388514885562e-05, "loss": 0.2138, "step": 10085 }, { "epoch": 4.746352941176471, "grad_norm": 0.35539870050982647, "learning_rate": 4.320806367643098e-05, "loss": 0.2136, "step": 10086 }, { "epoch": 4.746823529411765, "grad_norm": 0.2501946133051113, "learning_rate": 4.3177284984860895e-05, "loss": 0.0984, "step": 10087 }, { "epoch": 4.747294117647058, "grad_norm": 0.697093192518301, "learning_rate": 4.314651541647468e-05, "loss": 0.1079, "step": 10088 }, { "epoch": 4.747764705882353, "grad_norm": 0.3825629301317535, "learning_rate": 4.3115754973900104e-05, "loss": 0.1934, "step": 10089 }, { "epoch": 4.748235294117647, "grad_norm": 0.5030505762234481, "learning_rate": 4.308500365976434e-05, "loss": 0.284, "step": 10090 }, { "epoch": 4.748705882352941, "grad_norm": 0.3591590197226992, "learning_rate": 4.305426147669371e-05, "loss": 0.2106, "step": 10091 }, { "epoch": 4.749176470588235, "grad_norm": 0.5190637279375127, "learning_rate": 4.302352842731368e-05, "loss": 0.404, "step": 10092 }, { "epoch": 4.7496470588235296, "grad_norm": 0.40215155753571424, "learning_rate": 4.299280451424903e-05, "loss": 0.1275, "step": 10093 }, { "epoch": 4.750117647058824, "grad_norm": 0.40156070171411185, "learning_rate": 4.296208974012379e-05, "loss": 0.1298, "step": 10094 }, { "epoch": 4.750588235294118, "grad_norm": 0.35639186419058966, "learning_rate": 4.2931384107561065e-05, "loss": 0.2311, "step": 10095 }, { "epoch": 4.751058823529412, "grad_norm": 0.4908261197231149, "learning_rate": 4.290068761918329e-05, "loss": 0.4355, "step": 10096 }, { "epoch": 4.751529411764706, "grad_norm": 0.24884174955011937, "learning_rate": 4.287000027761217e-05, "loss": 0.1337, "step": 10097 }, { "epoch": 4.752, "grad_norm": 0.5442568778327124, "learning_rate": 4.2839322085468445e-05, "loss": 0.352, "step": 10098 }, { "epoch": 4.752470588235294, "grad_norm": 0.4995748046495653, "learning_rate": 4.280865304537223e-05, "loss": 0.2514, "step": 10099 }, { "epoch": 4.752941176470588, "grad_norm": 0.3333438498700232, "learning_rate": 4.277799315994286e-05, "loss": 0.2285, "step": 10100 }, { "epoch": 4.7534117647058824, "grad_norm": 0.5560278537429078, "learning_rate": 4.2747342431798767e-05, "loss": 0.3331, "step": 10101 }, { "epoch": 4.753882352941177, "grad_norm": 0.3119658868981576, "learning_rate": 4.271670086355768e-05, "loss": 0.1129, "step": 10102 }, { "epoch": 4.754352941176471, "grad_norm": 0.2324115300750434, "learning_rate": 4.268606845783663e-05, "loss": 0.1207, "step": 10103 }, { "epoch": 4.754823529411764, "grad_norm": 0.40764318864239873, "learning_rate": 4.265544521725165e-05, "loss": 0.327, "step": 10104 }, { "epoch": 4.7552941176470584, "grad_norm": 0.43246396825456984, "learning_rate": 4.2624831144418166e-05, "loss": 0.2491, "step": 10105 }, { "epoch": 4.755764705882353, "grad_norm": 0.11986154497243841, "learning_rate": 4.259422624195082e-05, "loss": 0.0245, "step": 10106 }, { "epoch": 4.756235294117647, "grad_norm": 0.415794573919403, "learning_rate": 4.2563630512463284e-05, "loss": 0.1657, "step": 10107 }, { "epoch": 4.756705882352941, "grad_norm": 0.41481977113554225, "learning_rate": 4.253304395856876e-05, "loss": 0.1358, "step": 10108 }, { "epoch": 4.757176470588235, "grad_norm": 0.3728106603191613, "learning_rate": 4.2502466582879375e-05, "loss": 0.2479, "step": 10109 }, { "epoch": 4.75764705882353, "grad_norm": 0.37398519719103623, "learning_rate": 4.2471898388006606e-05, "loss": 0.2649, "step": 10110 }, { "epoch": 4.758117647058824, "grad_norm": 0.47143031991641776, "learning_rate": 4.244133937656118e-05, "loss": 0.1825, "step": 10111 }, { "epoch": 4.758588235294118, "grad_norm": 0.34981824478422374, "learning_rate": 4.2410789551152894e-05, "loss": 0.2495, "step": 10112 }, { "epoch": 4.759058823529411, "grad_norm": 0.4408644210225979, "learning_rate": 4.238024891439092e-05, "loss": 0.2307, "step": 10113 }, { "epoch": 4.759529411764706, "grad_norm": 0.4906040610999272, "learning_rate": 4.234971746888359e-05, "loss": 0.3833, "step": 10114 }, { "epoch": 4.76, "grad_norm": 0.4290648829489906, "learning_rate": 4.231919521723836e-05, "loss": 0.2019, "step": 10115 }, { "epoch": 4.760470588235294, "grad_norm": 0.7637339131544724, "learning_rate": 4.228868216206206e-05, "loss": 0.1771, "step": 10116 }, { "epoch": 4.760941176470588, "grad_norm": 0.39940960211030413, "learning_rate": 4.225817830596064e-05, "loss": 0.1479, "step": 10117 }, { "epoch": 4.7614117647058825, "grad_norm": 0.4288275444920085, "learning_rate": 4.2227683651539236e-05, "loss": 0.2478, "step": 10118 }, { "epoch": 4.761882352941177, "grad_norm": 0.4626755529865638, "learning_rate": 4.21971982014023e-05, "loss": 0.1151, "step": 10119 }, { "epoch": 4.762352941176471, "grad_norm": 0.37921174018914955, "learning_rate": 4.2166721958153365e-05, "loss": 0.2355, "step": 10120 }, { "epoch": 4.762823529411765, "grad_norm": 0.46987123758248783, "learning_rate": 4.2136254924395305e-05, "loss": 0.3231, "step": 10121 }, { "epoch": 4.7632941176470585, "grad_norm": 0.2628840908285636, "learning_rate": 4.210579710273018e-05, "loss": 0.1679, "step": 10122 }, { "epoch": 4.763764705882353, "grad_norm": 0.38331337772754925, "learning_rate": 4.207534849575916e-05, "loss": 0.2345, "step": 10123 }, { "epoch": 4.764235294117647, "grad_norm": 0.3301029231989519, "learning_rate": 4.204490910608274e-05, "loss": 0.1721, "step": 10124 }, { "epoch": 4.764705882352941, "grad_norm": 0.5590057911019118, "learning_rate": 4.201447893630065e-05, "loss": 0.3596, "step": 10125 }, { "epoch": 4.765176470588235, "grad_norm": 0.45213233347976434, "learning_rate": 4.1984057989011674e-05, "loss": 0.2323, "step": 10126 }, { "epoch": 4.76564705882353, "grad_norm": 0.2789241846967319, "learning_rate": 4.195364626681396e-05, "loss": 0.1356, "step": 10127 }, { "epoch": 4.766117647058824, "grad_norm": 0.14252196893955374, "learning_rate": 4.192324377230487e-05, "loss": 0.0344, "step": 10128 }, { "epoch": 4.766588235294118, "grad_norm": 0.9193319310665014, "learning_rate": 4.1892850508080824e-05, "loss": 0.3969, "step": 10129 }, { "epoch": 4.767058823529412, "grad_norm": 0.22581938611887348, "learning_rate": 4.18624664767376e-05, "loss": 0.0957, "step": 10130 }, { "epoch": 4.767529411764706, "grad_norm": 0.11136456850852794, "learning_rate": 4.1832091680870214e-05, "loss": 0.0276, "step": 10131 }, { "epoch": 4.768, "grad_norm": 0.3654880088266905, "learning_rate": 4.1801726123072696e-05, "loss": 0.3265, "step": 10132 }, { "epoch": 4.768470588235294, "grad_norm": 0.9043005616573316, "learning_rate": 4.1771369805938486e-05, "loss": 0.2996, "step": 10133 }, { "epoch": 4.768941176470588, "grad_norm": 0.29914619075559845, "learning_rate": 4.1741022732060204e-05, "loss": 0.1037, "step": 10134 }, { "epoch": 4.7694117647058825, "grad_norm": 0.2489265577536215, "learning_rate": 4.171068490402954e-05, "loss": 0.104, "step": 10135 }, { "epoch": 4.769882352941177, "grad_norm": 0.30035685456098715, "learning_rate": 4.168035632443754e-05, "loss": 0.1142, "step": 10136 }, { "epoch": 4.770352941176471, "grad_norm": 0.2893114733920837, "learning_rate": 4.165003699587445e-05, "loss": 0.1726, "step": 10137 }, { "epoch": 4.770823529411764, "grad_norm": 0.3528907096156089, "learning_rate": 4.161972692092962e-05, "loss": 0.2196, "step": 10138 }, { "epoch": 4.7712941176470585, "grad_norm": 0.42140672777576577, "learning_rate": 4.158942610219175e-05, "loss": 0.2424, "step": 10139 }, { "epoch": 4.771764705882353, "grad_norm": 0.35591325235187216, "learning_rate": 4.155913454224862e-05, "loss": 0.0827, "step": 10140 }, { "epoch": 4.772235294117647, "grad_norm": 0.3215064033675726, "learning_rate": 4.1528852243687276e-05, "loss": 0.1534, "step": 10141 }, { "epoch": 4.772705882352941, "grad_norm": 0.17740070197533336, "learning_rate": 4.1498579209094046e-05, "loss": 0.0345, "step": 10142 }, { "epoch": 4.773176470588235, "grad_norm": 1.08354339275599, "learning_rate": 4.146831544105431e-05, "loss": 0.2472, "step": 10143 }, { "epoch": 4.77364705882353, "grad_norm": 0.40552512499734045, "learning_rate": 4.143806094215277e-05, "loss": 0.2686, "step": 10144 }, { "epoch": 4.774117647058824, "grad_norm": 0.4043602223541921, "learning_rate": 4.140781571497336e-05, "loss": 0.1433, "step": 10145 }, { "epoch": 4.774588235294118, "grad_norm": 0.26711038849610647, "learning_rate": 4.137757976209911e-05, "loss": 0.1067, "step": 10146 }, { "epoch": 4.775058823529411, "grad_norm": 0.3630002481688881, "learning_rate": 4.134735308611232e-05, "loss": 0.2453, "step": 10147 }, { "epoch": 4.775529411764706, "grad_norm": 0.3238262182013424, "learning_rate": 4.131713568959457e-05, "loss": 0.1491, "step": 10148 }, { "epoch": 4.776, "grad_norm": 0.3501358715558878, "learning_rate": 4.128692757512646e-05, "loss": 0.1248, "step": 10149 }, { "epoch": 4.776470588235294, "grad_norm": 0.553355249541162, "learning_rate": 4.125672874528797e-05, "loss": 0.1247, "step": 10150 }, { "epoch": 4.776941176470588, "grad_norm": 0.267301493504011, "learning_rate": 4.122653920265828e-05, "loss": 0.105, "step": 10151 }, { "epoch": 4.7774117647058825, "grad_norm": 0.4284390511362937, "learning_rate": 4.119635894981558e-05, "loss": 0.1633, "step": 10152 }, { "epoch": 4.777882352941177, "grad_norm": 0.45084307623816594, "learning_rate": 4.1166187989337596e-05, "loss": 0.2225, "step": 10153 }, { "epoch": 4.778352941176471, "grad_norm": 0.1819508747469168, "learning_rate": 4.113602632380094e-05, "loss": 0.0372, "step": 10154 }, { "epoch": 4.778823529411765, "grad_norm": 0.465754506671075, "learning_rate": 4.110587395578161e-05, "loss": 0.2991, "step": 10155 }, { "epoch": 4.7792941176470585, "grad_norm": 0.2508818183584752, "learning_rate": 4.1075730887854805e-05, "loss": 0.085, "step": 10156 }, { "epoch": 4.779764705882353, "grad_norm": 0.36314405873477157, "learning_rate": 4.104559712259482e-05, "loss": 0.1237, "step": 10157 }, { "epoch": 4.780235294117647, "grad_norm": 0.42963182246506587, "learning_rate": 4.101547266257526e-05, "loss": 0.1996, "step": 10158 }, { "epoch": 4.780705882352941, "grad_norm": 0.3649000864058858, "learning_rate": 4.098535751036894e-05, "loss": 0.2079, "step": 10159 }, { "epoch": 4.781176470588235, "grad_norm": 0.21919376994604375, "learning_rate": 4.0955251668547754e-05, "loss": 0.0652, "step": 10160 }, { "epoch": 4.78164705882353, "grad_norm": 0.6071319997003415, "learning_rate": 4.092515513968294e-05, "loss": 0.2672, "step": 10161 }, { "epoch": 4.782117647058824, "grad_norm": 1.9103776095502447, "learning_rate": 4.0895067926344935e-05, "loss": 0.5803, "step": 10162 }, { "epoch": 4.782588235294118, "grad_norm": 0.354957390901273, "learning_rate": 4.0864990031103235e-05, "loss": 0.1034, "step": 10163 }, { "epoch": 4.783058823529412, "grad_norm": 0.2792209283809079, "learning_rate": 4.0834921456526684e-05, "loss": 0.0927, "step": 10164 }, { "epoch": 4.783529411764706, "grad_norm": 0.5050772175654006, "learning_rate": 4.080486220518336e-05, "loss": 0.2927, "step": 10165 }, { "epoch": 4.784, "grad_norm": 0.20842273160909394, "learning_rate": 4.0774812279640325e-05, "loss": 0.0668, "step": 10166 }, { "epoch": 4.784470588235294, "grad_norm": 0.5979515639947297, "learning_rate": 4.074477168246409e-05, "loss": 0.377, "step": 10167 }, { "epoch": 4.784941176470588, "grad_norm": 0.4855139495262406, "learning_rate": 4.071474041622029e-05, "loss": 0.3675, "step": 10168 }, { "epoch": 4.7854117647058825, "grad_norm": 0.23098095940357052, "learning_rate": 4.0684718483473656e-05, "loss": 0.0614, "step": 10169 }, { "epoch": 4.785882352941177, "grad_norm": 0.44720869002385316, "learning_rate": 4.06547058867883e-05, "loss": 0.2298, "step": 10170 }, { "epoch": 4.786352941176471, "grad_norm": 0.5316612814336013, "learning_rate": 4.0624702628727345e-05, "loss": 0.249, "step": 10171 }, { "epoch": 4.786823529411764, "grad_norm": 0.8913364302772998, "learning_rate": 4.0594708711853267e-05, "loss": 0.4898, "step": 10172 }, { "epoch": 4.7872941176470585, "grad_norm": 1.291098409045499, "learning_rate": 4.0564724138727734e-05, "loss": 0.2189, "step": 10173 }, { "epoch": 4.787764705882353, "grad_norm": 0.5563182604280679, "learning_rate": 4.0534748911911505e-05, "loss": 0.414, "step": 10174 }, { "epoch": 4.788235294117647, "grad_norm": 0.2532295542140102, "learning_rate": 4.0504783033964645e-05, "loss": 0.126, "step": 10175 }, { "epoch": 4.788705882352941, "grad_norm": 0.244474093617152, "learning_rate": 4.047482650744641e-05, "loss": 0.0298, "step": 10176 }, { "epoch": 4.789176470588235, "grad_norm": 0.5065787298956365, "learning_rate": 4.044487933491518e-05, "loss": 0.2311, "step": 10177 }, { "epoch": 4.78964705882353, "grad_norm": 0.3367349212139635, "learning_rate": 4.041494151892862e-05, "loss": 0.095, "step": 10178 }, { "epoch": 4.790117647058824, "grad_norm": 0.2488617289009583, "learning_rate": 4.038501306204361e-05, "loss": 0.117, "step": 10179 }, { "epoch": 4.790588235294118, "grad_norm": 0.4528685805013257, "learning_rate": 4.0355093966816095e-05, "loss": 0.2589, "step": 10180 }, { "epoch": 4.791058823529411, "grad_norm": 0.40980600664290173, "learning_rate": 4.032518423580135e-05, "loss": 0.2427, "step": 10181 }, { "epoch": 4.791529411764706, "grad_norm": 0.47441603300655516, "learning_rate": 4.029528387155389e-05, "loss": 0.2625, "step": 10182 }, { "epoch": 4.792, "grad_norm": 0.6249029145837187, "learning_rate": 4.026539287662722e-05, "loss": 0.3884, "step": 10183 }, { "epoch": 4.792470588235294, "grad_norm": 0.3573428949551969, "learning_rate": 4.0235511253574256e-05, "loss": 0.1303, "step": 10184 }, { "epoch": 4.792941176470588, "grad_norm": 0.5051975048461009, "learning_rate": 4.020563900494707e-05, "loss": 0.3125, "step": 10185 }, { "epoch": 4.7934117647058825, "grad_norm": 0.4456792122400068, "learning_rate": 4.01757761332968e-05, "loss": 0.2176, "step": 10186 }, { "epoch": 4.793882352941177, "grad_norm": 0.3542616598700833, "learning_rate": 4.0145922641173934e-05, "loss": 0.1336, "step": 10187 }, { "epoch": 4.794352941176471, "grad_norm": 0.45485438217938756, "learning_rate": 4.011607853112815e-05, "loss": 0.1731, "step": 10188 }, { "epoch": 4.794823529411765, "grad_norm": 0.37992070678167594, "learning_rate": 4.008624380570819e-05, "loss": 0.206, "step": 10189 }, { "epoch": 4.7952941176470585, "grad_norm": 0.28987882648353497, "learning_rate": 4.0056418467462184e-05, "loss": 0.1163, "step": 10190 }, { "epoch": 4.795764705882353, "grad_norm": 0.38115978703578646, "learning_rate": 4.002660251893727e-05, "loss": 0.2081, "step": 10191 }, { "epoch": 4.796235294117647, "grad_norm": 0.3563567324841685, "learning_rate": 3.9996795962679896e-05, "loss": 0.1984, "step": 10192 }, { "epoch": 4.796705882352941, "grad_norm": 0.31985103223776057, "learning_rate": 3.9966998801235766e-05, "loss": 0.0996, "step": 10193 }, { "epoch": 4.797176470588235, "grad_norm": 0.5273926532874449, "learning_rate": 3.9937211037149616e-05, "loss": 0.2744, "step": 10194 }, { "epoch": 4.79764705882353, "grad_norm": 0.3920538056299784, "learning_rate": 3.9907432672965485e-05, "loss": 0.1852, "step": 10195 }, { "epoch": 4.798117647058824, "grad_norm": 0.4423380168699396, "learning_rate": 3.987766371122661e-05, "loss": 0.1982, "step": 10196 }, { "epoch": 4.798588235294117, "grad_norm": 0.42860977990394317, "learning_rate": 3.984790415447539e-05, "loss": 0.1653, "step": 10197 }, { "epoch": 4.799058823529411, "grad_norm": 0.5344932607278782, "learning_rate": 3.9818154005253514e-05, "loss": 0.1728, "step": 10198 }, { "epoch": 4.799529411764706, "grad_norm": 0.4441006159049107, "learning_rate": 3.978841326610166e-05, "loss": 0.1419, "step": 10199 }, { "epoch": 4.8, "grad_norm": 0.8708548502386401, "learning_rate": 3.97586819395599e-05, "loss": 0.5715, "step": 10200 }, { "epoch": 4.800470588235294, "grad_norm": 0.24763076213127147, "learning_rate": 3.972896002816749e-05, "loss": 0.1078, "step": 10201 }, { "epoch": 4.800941176470588, "grad_norm": 0.4479401500177132, "learning_rate": 3.969924753446271e-05, "loss": 0.0895, "step": 10202 }, { "epoch": 4.8014117647058825, "grad_norm": 0.40855483505585877, "learning_rate": 3.966954446098322e-05, "loss": 0.2438, "step": 10203 }, { "epoch": 4.801882352941177, "grad_norm": 0.4854973631233768, "learning_rate": 3.9639850810265845e-05, "loss": 0.244, "step": 10204 }, { "epoch": 4.802352941176471, "grad_norm": 0.31372381540741495, "learning_rate": 3.961016658484648e-05, "loss": 0.2149, "step": 10205 }, { "epoch": 4.802823529411764, "grad_norm": 0.34492041191401107, "learning_rate": 3.9580491787260346e-05, "loss": 0.1196, "step": 10206 }, { "epoch": 4.8032941176470585, "grad_norm": 0.29797915590039, "learning_rate": 3.955082642004187e-05, "loss": 0.1494, "step": 10207 }, { "epoch": 4.803764705882353, "grad_norm": 0.10326250304667507, "learning_rate": 3.952117048572452e-05, "loss": 0.0209, "step": 10208 }, { "epoch": 4.804235294117647, "grad_norm": 0.3520812798220239, "learning_rate": 3.94915239868411e-05, "loss": 0.2504, "step": 10209 }, { "epoch": 4.804705882352941, "grad_norm": 0.5640934649384783, "learning_rate": 3.946188692592362e-05, "loss": 0.2355, "step": 10210 }, { "epoch": 4.805176470588235, "grad_norm": 0.587215971685553, "learning_rate": 3.943225930550315e-05, "loss": 0.3153, "step": 10211 }, { "epoch": 4.80564705882353, "grad_norm": 0.6319062341054503, "learning_rate": 3.9402641128110054e-05, "loss": 0.2859, "step": 10212 }, { "epoch": 4.806117647058824, "grad_norm": 0.29864915148588794, "learning_rate": 3.937303239627392e-05, "loss": 0.1552, "step": 10213 }, { "epoch": 4.806588235294118, "grad_norm": 0.3670801034376345, "learning_rate": 3.934343311252341e-05, "loss": 0.2585, "step": 10214 }, { "epoch": 4.807058823529411, "grad_norm": 0.684312615099875, "learning_rate": 3.9313843279386486e-05, "loss": 0.3114, "step": 10215 }, { "epoch": 4.807529411764706, "grad_norm": 0.6886248517269944, "learning_rate": 3.9284262899390285e-05, "loss": 0.4139, "step": 10216 }, { "epoch": 4.808, "grad_norm": 0.3499774650957969, "learning_rate": 3.925469197506107e-05, "loss": 0.1008, "step": 10217 }, { "epoch": 4.808470588235294, "grad_norm": 1.0072156556064114, "learning_rate": 3.922513050892434e-05, "loss": 0.3637, "step": 10218 }, { "epoch": 4.808941176470588, "grad_norm": 0.2901528661423993, "learning_rate": 3.919557850350486e-05, "loss": 0.1349, "step": 10219 }, { "epoch": 4.8094117647058825, "grad_norm": 0.2324170707055016, "learning_rate": 3.916603596132642e-05, "loss": 0.0886, "step": 10220 }, { "epoch": 4.809882352941177, "grad_norm": 0.7068490463157604, "learning_rate": 3.91365028849122e-05, "loss": 0.1963, "step": 10221 }, { "epoch": 4.810352941176471, "grad_norm": 0.4294449085567541, "learning_rate": 3.910697927678437e-05, "loss": 0.2381, "step": 10222 }, { "epoch": 4.810823529411765, "grad_norm": 0.28312930231152816, "learning_rate": 3.9077465139464446e-05, "loss": 0.1947, "step": 10223 }, { "epoch": 4.8112941176470585, "grad_norm": 0.24377243577740804, "learning_rate": 3.9047960475473095e-05, "loss": 0.0946, "step": 10224 }, { "epoch": 4.811764705882353, "grad_norm": 0.24841955213066305, "learning_rate": 3.901846528733011e-05, "loss": 0.0718, "step": 10225 }, { "epoch": 4.812235294117647, "grad_norm": 0.3721782220686379, "learning_rate": 3.898897957755454e-05, "loss": 0.3027, "step": 10226 }, { "epoch": 4.812705882352941, "grad_norm": 0.37109873896591206, "learning_rate": 3.895950334866468e-05, "loss": 0.2388, "step": 10227 }, { "epoch": 4.813176470588235, "grad_norm": 0.27023424193504647, "learning_rate": 3.893003660317783e-05, "loss": 0.1292, "step": 10228 }, { "epoch": 4.81364705882353, "grad_norm": 0.37469874167894174, "learning_rate": 3.8900579343610646e-05, "loss": 0.1473, "step": 10229 }, { "epoch": 4.814117647058824, "grad_norm": 0.379108185677502, "learning_rate": 3.887113157247897e-05, "loss": 0.0904, "step": 10230 }, { "epoch": 4.814588235294117, "grad_norm": 0.4377100110431768, "learning_rate": 3.884169329229769e-05, "loss": 0.1179, "step": 10231 }, { "epoch": 4.815058823529411, "grad_norm": 0.4176175151894559, "learning_rate": 3.881226450558104e-05, "loss": 0.2109, "step": 10232 }, { "epoch": 4.815529411764706, "grad_norm": 0.08210921721980227, "learning_rate": 3.878284521484242e-05, "loss": 0.0189, "step": 10233 }, { "epoch": 4.816, "grad_norm": 0.2815751088255435, "learning_rate": 3.8753435422594295e-05, "loss": 0.0402, "step": 10234 }, { "epoch": 4.816470588235294, "grad_norm": 0.26615232347215456, "learning_rate": 3.8724035131348437e-05, "loss": 0.0762, "step": 10235 }, { "epoch": 4.816941176470588, "grad_norm": 0.34624902006573116, "learning_rate": 3.8694644343615836e-05, "loss": 0.1333, "step": 10236 }, { "epoch": 4.8174117647058825, "grad_norm": 0.36119325705940253, "learning_rate": 3.866526306190651e-05, "loss": 0.1256, "step": 10237 }, { "epoch": 4.817882352941177, "grad_norm": 0.28910675732235347, "learning_rate": 3.863589128872981e-05, "loss": 0.1557, "step": 10238 }, { "epoch": 4.818352941176471, "grad_norm": 0.6159396626669178, "learning_rate": 3.8606529026594276e-05, "loss": 0.3427, "step": 10239 }, { "epoch": 4.818823529411764, "grad_norm": 0.537643844936979, "learning_rate": 3.85771762780075e-05, "loss": 0.2916, "step": 10240 }, { "epoch": 4.8192941176470585, "grad_norm": 0.40757670650408995, "learning_rate": 3.854783304547639e-05, "loss": 0.1847, "step": 10241 }, { "epoch": 4.819764705882353, "grad_norm": 0.109377619680957, "learning_rate": 3.851849933150703e-05, "loss": 0.0214, "step": 10242 }, { "epoch": 4.820235294117647, "grad_norm": 0.3610446908640306, "learning_rate": 3.848917513860463e-05, "loss": 0.1323, "step": 10243 }, { "epoch": 4.820705882352941, "grad_norm": 0.3666708345101597, "learning_rate": 3.8459860469273666e-05, "loss": 0.131, "step": 10244 }, { "epoch": 4.821176470588235, "grad_norm": 0.660582414432715, "learning_rate": 3.8430555326017684e-05, "loss": 0.35, "step": 10245 }, { "epoch": 4.82164705882353, "grad_norm": 0.3579607745129847, "learning_rate": 3.840125971133952e-05, "loss": 0.0965, "step": 10246 }, { "epoch": 4.822117647058824, "grad_norm": 0.32645167844424616, "learning_rate": 3.8371973627741204e-05, "loss": 0.175, "step": 10247 }, { "epoch": 4.822588235294118, "grad_norm": 0.31022537037247133, "learning_rate": 3.8342697077723845e-05, "loss": 0.1121, "step": 10248 }, { "epoch": 4.823058823529411, "grad_norm": 0.1402237698932521, "learning_rate": 3.8313430063787856e-05, "loss": 0.0465, "step": 10249 }, { "epoch": 4.823529411764706, "grad_norm": 0.30710654719501596, "learning_rate": 3.8284172588432716e-05, "loss": 0.1991, "step": 10250 }, { "epoch": 4.824, "grad_norm": 0.42009559543131064, "learning_rate": 3.8254924654157214e-05, "loss": 0.2168, "step": 10251 }, { "epoch": 4.824470588235294, "grad_norm": 0.7506769284398291, "learning_rate": 3.8225686263459296e-05, "loss": 0.3871, "step": 10252 }, { "epoch": 4.824941176470588, "grad_norm": 0.33661352735140837, "learning_rate": 3.819645741883597e-05, "loss": 0.1703, "step": 10253 }, { "epoch": 4.8254117647058825, "grad_norm": 0.42489527035238084, "learning_rate": 3.816723812278358e-05, "loss": 0.1732, "step": 10254 }, { "epoch": 4.825882352941177, "grad_norm": 0.3532414447990254, "learning_rate": 3.813802837779762e-05, "loss": 0.0938, "step": 10255 }, { "epoch": 4.826352941176471, "grad_norm": 0.37362601599574535, "learning_rate": 3.810882818637268e-05, "loss": 0.2476, "step": 10256 }, { "epoch": 4.826823529411765, "grad_norm": 0.6315703429923986, "learning_rate": 3.807963755100263e-05, "loss": 0.3208, "step": 10257 }, { "epoch": 4.8272941176470585, "grad_norm": 0.4882243771327362, "learning_rate": 3.8050456474180554e-05, "loss": 0.2697, "step": 10258 }, { "epoch": 4.827764705882353, "grad_norm": 0.5717072801623273, "learning_rate": 3.802128495839856e-05, "loss": 0.1466, "step": 10259 }, { "epoch": 4.828235294117647, "grad_norm": 0.6212218757163321, "learning_rate": 3.799212300614809e-05, "loss": 0.5121, "step": 10260 }, { "epoch": 4.828705882352941, "grad_norm": 0.4097157257790999, "learning_rate": 3.7962970619919745e-05, "loss": 0.1867, "step": 10261 }, { "epoch": 4.829176470588235, "grad_norm": 0.9190066199774635, "learning_rate": 3.7933827802203207e-05, "loss": 0.3971, "step": 10262 }, { "epoch": 4.82964705882353, "grad_norm": 0.4081856208051943, "learning_rate": 3.790469455548745e-05, "loss": 0.1496, "step": 10263 }, { "epoch": 4.830117647058824, "grad_norm": 0.5694608099192839, "learning_rate": 3.787557088226065e-05, "loss": 0.359, "step": 10264 }, { "epoch": 4.830588235294117, "grad_norm": 0.6420389467167084, "learning_rate": 3.7846456785010014e-05, "loss": 0.3922, "step": 10265 }, { "epoch": 4.831058823529411, "grad_norm": 0.44694618587525803, "learning_rate": 3.7817352266222086e-05, "loss": 0.2184, "step": 10266 }, { "epoch": 4.831529411764706, "grad_norm": 0.6408768833804466, "learning_rate": 3.778825732838257e-05, "loss": 0.275, "step": 10267 }, { "epoch": 4.832, "grad_norm": 0.22544588295894596, "learning_rate": 3.775917197397622e-05, "loss": 0.0679, "step": 10268 }, { "epoch": 4.832470588235294, "grad_norm": 0.36006596745590624, "learning_rate": 3.773009620548718e-05, "loss": 0.1029, "step": 10269 }, { "epoch": 4.832941176470588, "grad_norm": 0.49303246662946526, "learning_rate": 3.770103002539857e-05, "loss": 0.3551, "step": 10270 }, { "epoch": 4.8334117647058825, "grad_norm": 0.42110616039835597, "learning_rate": 3.767197343619279e-05, "loss": 0.2614, "step": 10271 }, { "epoch": 4.833882352941177, "grad_norm": 0.5181271828554997, "learning_rate": 3.7642926440351514e-05, "loss": 0.1559, "step": 10272 }, { "epoch": 4.834352941176471, "grad_norm": 0.2623513842014949, "learning_rate": 3.761388904035537e-05, "loss": 0.0601, "step": 10273 }, { "epoch": 4.834823529411764, "grad_norm": 0.3696959629959838, "learning_rate": 3.758486123868435e-05, "loss": 0.1849, "step": 10274 }, { "epoch": 4.8352941176470585, "grad_norm": 0.28554806348731776, "learning_rate": 3.755584303781763e-05, "loss": 0.0877, "step": 10275 }, { "epoch": 4.835764705882353, "grad_norm": 0.37561834586357445, "learning_rate": 3.7526834440233395e-05, "loss": 0.2485, "step": 10276 }, { "epoch": 4.836235294117647, "grad_norm": 0.426071607855318, "learning_rate": 3.749783544840918e-05, "loss": 0.1158, "step": 10277 }, { "epoch": 4.836705882352941, "grad_norm": 0.519188560830939, "learning_rate": 3.746884606482169e-05, "loss": 0.3554, "step": 10278 }, { "epoch": 4.837176470588235, "grad_norm": 0.05328079554334359, "learning_rate": 3.743986629194666e-05, "loss": 0.0053, "step": 10279 }, { "epoch": 4.83764705882353, "grad_norm": 0.5553714076553471, "learning_rate": 3.741089613225914e-05, "loss": 0.2389, "step": 10280 }, { "epoch": 4.838117647058824, "grad_norm": 0.39627671666355885, "learning_rate": 3.7381935588233385e-05, "loss": 0.1934, "step": 10281 }, { "epoch": 4.838588235294118, "grad_norm": 0.3520181879018188, "learning_rate": 3.735298466234269e-05, "loss": 0.1625, "step": 10282 }, { "epoch": 4.839058823529411, "grad_norm": 0.360952616418711, "learning_rate": 3.732404335705962e-05, "loss": 0.1699, "step": 10283 }, { "epoch": 4.839529411764706, "grad_norm": 0.35312496664322923, "learning_rate": 3.729511167485597e-05, "loss": 0.1728, "step": 10284 }, { "epoch": 4.84, "grad_norm": 0.3478054757628461, "learning_rate": 3.72661896182025e-05, "loss": 0.2027, "step": 10285 }, { "epoch": 4.840470588235294, "grad_norm": 0.47634190889367706, "learning_rate": 3.7237277189569505e-05, "loss": 0.39, "step": 10286 }, { "epoch": 4.840941176470588, "grad_norm": 0.5287329590884157, "learning_rate": 3.720837439142608e-05, "loss": 0.2492, "step": 10287 }, { "epoch": 4.8414117647058825, "grad_norm": 0.4165937532629883, "learning_rate": 3.717948122624072e-05, "loss": 0.2938, "step": 10288 }, { "epoch": 4.841882352941177, "grad_norm": 0.4506395368029781, "learning_rate": 3.7150597696481076e-05, "loss": 0.2877, "step": 10289 }, { "epoch": 4.842352941176471, "grad_norm": 0.45008514861057003, "learning_rate": 3.712172380461387e-05, "loss": 0.3342, "step": 10290 }, { "epoch": 4.842823529411765, "grad_norm": 0.3615307765811504, "learning_rate": 3.709285955310513e-05, "loss": 0.2331, "step": 10291 }, { "epoch": 4.8432941176470585, "grad_norm": 0.3750193774994405, "learning_rate": 3.7064004944420026e-05, "loss": 0.2253, "step": 10292 }, { "epoch": 4.843764705882353, "grad_norm": 0.10048225800499676, "learning_rate": 3.703515998102281e-05, "loss": 0.0227, "step": 10293 }, { "epoch": 4.844235294117647, "grad_norm": 0.321698590699851, "learning_rate": 3.7006324665377025e-05, "loss": 0.1847, "step": 10294 }, { "epoch": 4.844705882352941, "grad_norm": 0.34374736965879243, "learning_rate": 3.6977498999945384e-05, "loss": 0.2001, "step": 10295 }, { "epoch": 4.845176470588235, "grad_norm": 0.3597547703100247, "learning_rate": 3.694868298718965e-05, "loss": 0.2381, "step": 10296 }, { "epoch": 4.84564705882353, "grad_norm": 0.7593017929915199, "learning_rate": 3.6919876629570905e-05, "loss": 0.2761, "step": 10297 }, { "epoch": 4.846117647058824, "grad_norm": 0.18426492745621884, "learning_rate": 3.68910799295494e-05, "loss": 0.0444, "step": 10298 }, { "epoch": 4.846588235294117, "grad_norm": 0.953872687203756, "learning_rate": 3.686229288958442e-05, "loss": 0.3699, "step": 10299 }, { "epoch": 4.847058823529411, "grad_norm": 0.1898829708298661, "learning_rate": 3.6833515512134606e-05, "loss": 0.0499, "step": 10300 }, { "epoch": 4.847529411764706, "grad_norm": 0.5331496254422085, "learning_rate": 3.6804747799657594e-05, "loss": 0.2976, "step": 10301 }, { "epoch": 4.848, "grad_norm": 0.2522941971207824, "learning_rate": 3.677598975461036e-05, "loss": 0.078, "step": 10302 }, { "epoch": 4.848470588235294, "grad_norm": 0.47904205003388095, "learning_rate": 3.6747241379448986e-05, "loss": 0.1984, "step": 10303 }, { "epoch": 4.848941176470588, "grad_norm": 0.3870383515851386, "learning_rate": 3.671850267662866e-05, "loss": 0.2701, "step": 10304 }, { "epoch": 4.8494117647058825, "grad_norm": 0.2292262149587426, "learning_rate": 3.668977364860386e-05, "loss": 0.0718, "step": 10305 }, { "epoch": 4.849882352941177, "grad_norm": 0.29483710887461373, "learning_rate": 3.666105429782821e-05, "loss": 0.1494, "step": 10306 }, { "epoch": 4.850352941176471, "grad_norm": 0.20421151848649172, "learning_rate": 3.6632344626754405e-05, "loss": 0.051, "step": 10307 }, { "epoch": 4.850823529411764, "grad_norm": 0.25061228649967887, "learning_rate": 3.660364463783443e-05, "loss": 0.0769, "step": 10308 }, { "epoch": 4.8512941176470585, "grad_norm": 0.9856760754379732, "learning_rate": 3.657495433351945e-05, "loss": 0.3054, "step": 10309 }, { "epoch": 4.851764705882353, "grad_norm": 0.37764250246785247, "learning_rate": 3.6546273716259685e-05, "loss": 0.2111, "step": 10310 }, { "epoch": 4.852235294117647, "grad_norm": 0.46658148372595715, "learning_rate": 3.6517602788504613e-05, "loss": 0.3331, "step": 10311 }, { "epoch": 4.852705882352941, "grad_norm": 0.25100390909534986, "learning_rate": 3.648894155270295e-05, "loss": 0.0934, "step": 10312 }, { "epoch": 4.853176470588235, "grad_norm": 0.37507517707012705, "learning_rate": 3.64602900113024e-05, "loss": 0.1003, "step": 10313 }, { "epoch": 4.85364705882353, "grad_norm": 0.9264657749386105, "learning_rate": 3.643164816674999e-05, "loss": 0.2168, "step": 10314 }, { "epoch": 4.854117647058824, "grad_norm": 0.5957785831891814, "learning_rate": 3.6403016021491924e-05, "loss": 0.5186, "step": 10315 }, { "epoch": 4.854588235294118, "grad_norm": 0.434945764579378, "learning_rate": 3.637439357797343e-05, "loss": 0.2182, "step": 10316 }, { "epoch": 4.855058823529411, "grad_norm": 0.369786935557799, "learning_rate": 3.634578083863905e-05, "loss": 0.172, "step": 10317 }, { "epoch": 4.855529411764706, "grad_norm": 0.5517902215921925, "learning_rate": 3.631717780593252e-05, "loss": 0.3706, "step": 10318 }, { "epoch": 4.856, "grad_norm": 0.3591869441597302, "learning_rate": 3.628858448229655e-05, "loss": 0.1808, "step": 10319 }, { "epoch": 4.856470588235294, "grad_norm": 0.377929666683761, "learning_rate": 3.6260000870173266e-05, "loss": 0.222, "step": 10320 }, { "epoch": 4.856941176470588, "grad_norm": 0.3920031186525971, "learning_rate": 3.6231426972003766e-05, "loss": 0.2069, "step": 10321 }, { "epoch": 4.8574117647058825, "grad_norm": 0.18747321825093424, "learning_rate": 3.620286279022843e-05, "loss": 0.0513, "step": 10322 }, { "epoch": 4.857882352941177, "grad_norm": 0.3342365710409483, "learning_rate": 3.617430832728682e-05, "loss": 0.1293, "step": 10323 }, { "epoch": 4.85835294117647, "grad_norm": 0.5299823432195828, "learning_rate": 3.614576358561757e-05, "loss": 0.3289, "step": 10324 }, { "epoch": 4.858823529411764, "grad_norm": 0.12299992842241766, "learning_rate": 3.611722856765855e-05, "loss": 0.0249, "step": 10325 }, { "epoch": 4.8592941176470585, "grad_norm": 0.4285061549055985, "learning_rate": 3.608870327584683e-05, "loss": 0.3446, "step": 10326 }, { "epoch": 4.859764705882353, "grad_norm": 0.6144697372915071, "learning_rate": 3.606018771261856e-05, "loss": 0.3787, "step": 10327 }, { "epoch": 4.860235294117647, "grad_norm": 0.36198284114075974, "learning_rate": 3.6031681880409106e-05, "loss": 0.2264, "step": 10328 }, { "epoch": 4.860705882352941, "grad_norm": 0.6071371616976065, "learning_rate": 3.6003185781653097e-05, "loss": 0.3453, "step": 10329 }, { "epoch": 4.861176470588235, "grad_norm": 0.20516854217089472, "learning_rate": 3.5974699418784084e-05, "loss": 0.0625, "step": 10330 }, { "epoch": 4.86164705882353, "grad_norm": 0.3311758210872765, "learning_rate": 3.5946222794235105e-05, "loss": 0.1794, "step": 10331 }, { "epoch": 4.862117647058824, "grad_norm": 0.3434993215098957, "learning_rate": 3.5917755910438094e-05, "loss": 0.2603, "step": 10332 }, { "epoch": 4.862588235294117, "grad_norm": 0.4998087071816665, "learning_rate": 3.58892987698243e-05, "loss": 0.3293, "step": 10333 }, { "epoch": 4.863058823529411, "grad_norm": 0.23843319616363967, "learning_rate": 3.5860851374824134e-05, "loss": 0.0652, "step": 10334 }, { "epoch": 4.863529411764706, "grad_norm": 0.4325835740979089, "learning_rate": 3.5832413727867075e-05, "loss": 0.2626, "step": 10335 }, { "epoch": 4.864, "grad_norm": 0.22925169128995826, "learning_rate": 3.580398583138188e-05, "loss": 0.0956, "step": 10336 }, { "epoch": 4.864470588235294, "grad_norm": 0.49192293180949753, "learning_rate": 3.577556768779646e-05, "loss": 0.2468, "step": 10337 }, { "epoch": 4.864941176470588, "grad_norm": 0.3553945233802481, "learning_rate": 3.574715929953778e-05, "loss": 0.1197, "step": 10338 }, { "epoch": 4.8654117647058825, "grad_norm": 0.5145819953068489, "learning_rate": 3.571876066903211e-05, "loss": 0.2788, "step": 10339 }, { "epoch": 4.865882352941177, "grad_norm": 0.29289085979675766, "learning_rate": 3.569037179870487e-05, "loss": 0.1025, "step": 10340 }, { "epoch": 4.866352941176471, "grad_norm": 0.8846975420858837, "learning_rate": 3.5661992690980506e-05, "loss": 0.169, "step": 10341 }, { "epoch": 4.866823529411764, "grad_norm": 0.4486822370379079, "learning_rate": 3.56336233482828e-05, "loss": 0.2745, "step": 10342 }, { "epoch": 4.8672941176470585, "grad_norm": 0.1876582571199204, "learning_rate": 3.560526377303467e-05, "loss": 0.0471, "step": 10343 }, { "epoch": 4.867764705882353, "grad_norm": 0.5454782958260701, "learning_rate": 3.557691396765807e-05, "loss": 0.1778, "step": 10344 }, { "epoch": 4.868235294117647, "grad_norm": 0.10163299248971235, "learning_rate": 3.554857393457426e-05, "loss": 0.0213, "step": 10345 }, { "epoch": 4.868705882352941, "grad_norm": 0.3341597086866154, "learning_rate": 3.5520243676203664e-05, "loss": 0.1447, "step": 10346 }, { "epoch": 4.869176470588235, "grad_norm": 0.4333736743551181, "learning_rate": 3.5491923194965735e-05, "loss": 0.1922, "step": 10347 }, { "epoch": 4.86964705882353, "grad_norm": 0.36771760475842746, "learning_rate": 3.546361249327922e-05, "loss": 0.174, "step": 10348 }, { "epoch": 4.870117647058824, "grad_norm": 0.3745494634053055, "learning_rate": 3.5435311573562024e-05, "loss": 0.196, "step": 10349 }, { "epoch": 4.870588235294118, "grad_norm": 0.32561902891307143, "learning_rate": 3.540702043823113e-05, "loss": 0.1362, "step": 10350 }, { "epoch": 4.871058823529411, "grad_norm": 0.9721079261651746, "learning_rate": 3.537873908970282e-05, "loss": 0.4298, "step": 10351 }, { "epoch": 4.871529411764706, "grad_norm": 0.48806098347627297, "learning_rate": 3.535046753039235e-05, "loss": 0.2182, "step": 10352 }, { "epoch": 4.872, "grad_norm": 0.582556427746516, "learning_rate": 3.53222057627143e-05, "loss": 0.281, "step": 10353 }, { "epoch": 4.872470588235294, "grad_norm": 0.38650164316674596, "learning_rate": 3.5293953789082425e-05, "loss": 0.1696, "step": 10354 }, { "epoch": 4.872941176470588, "grad_norm": 0.4619809807689601, "learning_rate": 3.5265711611909476e-05, "loss": 0.2715, "step": 10355 }, { "epoch": 4.8734117647058826, "grad_norm": 0.3312909015272678, "learning_rate": 3.5237479233607525e-05, "loss": 0.1273, "step": 10356 }, { "epoch": 4.873882352941177, "grad_norm": 0.4099075756193177, "learning_rate": 3.520925665658779e-05, "loss": 0.2647, "step": 10357 }, { "epoch": 4.87435294117647, "grad_norm": 0.3606877721485955, "learning_rate": 3.518104388326055e-05, "loss": 0.1926, "step": 10358 }, { "epoch": 4.874823529411764, "grad_norm": 0.3208682995273084, "learning_rate": 3.5152840916035344e-05, "loss": 0.1673, "step": 10359 }, { "epoch": 4.875294117647059, "grad_norm": 0.6678149776463854, "learning_rate": 3.5124647757320883e-05, "loss": 0.2314, "step": 10360 }, { "epoch": 4.875764705882353, "grad_norm": 0.4805430618497564, "learning_rate": 3.509646440952492e-05, "loss": 0.3008, "step": 10361 }, { "epoch": 4.876235294117647, "grad_norm": 0.6598883880468595, "learning_rate": 3.506829087505451e-05, "loss": 0.3692, "step": 10362 }, { "epoch": 4.876705882352941, "grad_norm": 0.3398322126295075, "learning_rate": 3.504012715631583e-05, "loss": 0.1902, "step": 10363 }, { "epoch": 4.8771764705882354, "grad_norm": 0.587015922820418, "learning_rate": 3.501197325571411e-05, "loss": 0.215, "step": 10364 }, { "epoch": 4.87764705882353, "grad_norm": 0.21342175882709796, "learning_rate": 3.49838291756539e-05, "loss": 0.0408, "step": 10365 }, { "epoch": 4.878117647058824, "grad_norm": 0.49005144446307375, "learning_rate": 3.495569491853887e-05, "loss": 0.2128, "step": 10366 }, { "epoch": 4.878588235294117, "grad_norm": 0.1746799419235917, "learning_rate": 3.492757048677175e-05, "loss": 0.0514, "step": 10367 }, { "epoch": 4.8790588235294114, "grad_norm": 0.40310558936012275, "learning_rate": 3.4899455882754526e-05, "loss": 0.2692, "step": 10368 }, { "epoch": 4.879529411764706, "grad_norm": 0.5082902418273627, "learning_rate": 3.4871351108888375e-05, "loss": 0.2653, "step": 10369 }, { "epoch": 4.88, "grad_norm": 0.3347945022383648, "learning_rate": 3.4843256167573514e-05, "loss": 0.0825, "step": 10370 }, { "epoch": 4.880470588235294, "grad_norm": 0.42385841016088, "learning_rate": 3.481517106120944e-05, "loss": 0.1529, "step": 10371 }, { "epoch": 4.880941176470588, "grad_norm": 0.30247195536178784, "learning_rate": 3.478709579219472e-05, "loss": 0.1275, "step": 10372 }, { "epoch": 4.881411764705883, "grad_norm": 0.3323727050529072, "learning_rate": 3.4759030362927115e-05, "loss": 0.1407, "step": 10373 }, { "epoch": 4.881882352941177, "grad_norm": 0.3020657965215814, "learning_rate": 3.4730974775803574e-05, "loss": 0.2267, "step": 10374 }, { "epoch": 4.882352941176471, "grad_norm": 0.20049202150240614, "learning_rate": 3.4702929033220174e-05, "loss": 0.033, "step": 10375 }, { "epoch": 4.882823529411764, "grad_norm": 0.281123716495929, "learning_rate": 3.467489313757217e-05, "loss": 0.0823, "step": 10376 }, { "epoch": 4.883294117647059, "grad_norm": 0.5052380530756564, "learning_rate": 3.4646867091254e-05, "loss": 0.2562, "step": 10377 }, { "epoch": 4.883764705882353, "grad_norm": 0.40000574627692054, "learning_rate": 3.4618850896659135e-05, "loss": 0.2061, "step": 10378 }, { "epoch": 4.884235294117647, "grad_norm": 0.3975006560397426, "learning_rate": 3.4590844556180395e-05, "loss": 0.2283, "step": 10379 }, { "epoch": 4.884705882352941, "grad_norm": 0.2127668900400521, "learning_rate": 3.456284807220958e-05, "loss": 0.1003, "step": 10380 }, { "epoch": 4.8851764705882355, "grad_norm": 0.28298061870858865, "learning_rate": 3.453486144713772e-05, "loss": 0.1277, "step": 10381 }, { "epoch": 4.88564705882353, "grad_norm": 0.2369987809811172, "learning_rate": 3.450688468335512e-05, "loss": 0.1208, "step": 10382 }, { "epoch": 4.886117647058824, "grad_norm": 0.3877181093022807, "learning_rate": 3.4478917783251e-05, "loss": 0.227, "step": 10383 }, { "epoch": 4.886588235294118, "grad_norm": 0.3337753108054552, "learning_rate": 3.445096074921393e-05, "loss": 0.1351, "step": 10384 }, { "epoch": 4.8870588235294115, "grad_norm": 0.3972786801903994, "learning_rate": 3.442301358363163e-05, "loss": 0.2526, "step": 10385 }, { "epoch": 4.887529411764706, "grad_norm": 0.33386085506162017, "learning_rate": 3.4395076288890825e-05, "loss": 0.1833, "step": 10386 }, { "epoch": 4.888, "grad_norm": 0.4503786791367274, "learning_rate": 3.436714886737753e-05, "loss": 0.3105, "step": 10387 }, { "epoch": 4.888470588235294, "grad_norm": 0.4459003212506234, "learning_rate": 3.4339231321476965e-05, "loss": 0.1834, "step": 10388 }, { "epoch": 4.888941176470588, "grad_norm": 0.30756691470523334, "learning_rate": 3.43113236535733e-05, "loss": 0.1359, "step": 10389 }, { "epoch": 4.889411764705883, "grad_norm": 0.21322989280348045, "learning_rate": 3.428342586605005e-05, "loss": 0.0589, "step": 10390 }, { "epoch": 4.889882352941177, "grad_norm": 0.507238078505365, "learning_rate": 3.425553796128985e-05, "loss": 0.2841, "step": 10391 }, { "epoch": 4.89035294117647, "grad_norm": 0.2925824795404461, "learning_rate": 3.42276599416744e-05, "loss": 0.1332, "step": 10392 }, { "epoch": 4.890823529411764, "grad_norm": 0.33972549352103565, "learning_rate": 3.419979180958464e-05, "loss": 0.201, "step": 10393 }, { "epoch": 4.891294117647059, "grad_norm": 0.6186509027114553, "learning_rate": 3.417193356740071e-05, "loss": 0.2043, "step": 10394 }, { "epoch": 4.891764705882353, "grad_norm": 0.3067219879469917, "learning_rate": 3.414408521750173e-05, "loss": 0.082, "step": 10395 }, { "epoch": 4.892235294117647, "grad_norm": 0.5278834020437599, "learning_rate": 3.411624676226616e-05, "loss": 0.1051, "step": 10396 }, { "epoch": 4.892705882352941, "grad_norm": 0.11434167276395164, "learning_rate": 3.408841820407157e-05, "loss": 0.0289, "step": 10397 }, { "epoch": 4.8931764705882355, "grad_norm": 0.1975249177800811, "learning_rate": 3.406059954529454e-05, "loss": 0.081, "step": 10398 }, { "epoch": 4.89364705882353, "grad_norm": 0.3171941865143686, "learning_rate": 3.403279078831101e-05, "loss": 0.1264, "step": 10399 }, { "epoch": 4.894117647058824, "grad_norm": 0.6048010874048193, "learning_rate": 3.4004991935496004e-05, "loss": 0.4029, "step": 10400 }, { "epoch": 4.894588235294117, "grad_norm": 0.31289090299225314, "learning_rate": 3.397720298922358e-05, "loss": 0.1408, "step": 10401 }, { "epoch": 4.8950588235294115, "grad_norm": 0.6915432043056858, "learning_rate": 3.394942395186716e-05, "loss": 0.5661, "step": 10402 }, { "epoch": 4.895529411764706, "grad_norm": 0.34776014445343667, "learning_rate": 3.392165482579912e-05, "loss": 0.1156, "step": 10403 }, { "epoch": 4.896, "grad_norm": 0.3925069352672933, "learning_rate": 3.389389561339112e-05, "loss": 0.1197, "step": 10404 }, { "epoch": 4.896470588235294, "grad_norm": 0.49151063506048676, "learning_rate": 3.3866146317013976e-05, "loss": 0.1922, "step": 10405 }, { "epoch": 4.896941176470588, "grad_norm": 0.541558533969749, "learning_rate": 3.3838406939037525e-05, "loss": 0.2501, "step": 10406 }, { "epoch": 4.897411764705883, "grad_norm": 0.35339085676833487, "learning_rate": 3.381067748183087e-05, "loss": 0.1384, "step": 10407 }, { "epoch": 4.897882352941177, "grad_norm": 0.5924827680832018, "learning_rate": 3.3782957947762336e-05, "loss": 0.3396, "step": 10408 }, { "epoch": 4.898352941176471, "grad_norm": 0.3810886036695429, "learning_rate": 3.375524833919918e-05, "loss": 0.1897, "step": 10409 }, { "epoch": 4.898823529411764, "grad_norm": 0.31120877042072254, "learning_rate": 3.372754865850799e-05, "loss": 0.1242, "step": 10410 }, { "epoch": 4.899294117647059, "grad_norm": 0.2520076749070582, "learning_rate": 3.3699858908054506e-05, "loss": 0.1047, "step": 10411 }, { "epoch": 4.899764705882353, "grad_norm": 0.4279574725648799, "learning_rate": 3.367217909020348e-05, "loss": 0.2474, "step": 10412 }, { "epoch": 4.900235294117647, "grad_norm": 0.3056520993801183, "learning_rate": 3.364450920731894e-05, "loss": 0.1633, "step": 10413 }, { "epoch": 4.900705882352941, "grad_norm": 0.5173592695235338, "learning_rate": 3.361684926176408e-05, "loss": 0.2834, "step": 10414 }, { "epoch": 4.9011764705882355, "grad_norm": 0.4555686746633172, "learning_rate": 3.35891992559011e-05, "loss": 0.3127, "step": 10415 }, { "epoch": 4.90164705882353, "grad_norm": 0.10086181359594668, "learning_rate": 3.35615591920915e-05, "loss": 0.0131, "step": 10416 }, { "epoch": 4.902117647058824, "grad_norm": 0.2797694340096931, "learning_rate": 3.353392907269591e-05, "loss": 0.0769, "step": 10417 }, { "epoch": 4.902588235294118, "grad_norm": 0.4106607472198827, "learning_rate": 3.3506308900073964e-05, "loss": 0.2425, "step": 10418 }, { "epoch": 4.9030588235294115, "grad_norm": 0.37617538038087783, "learning_rate": 3.3478698676584714e-05, "loss": 0.1981, "step": 10419 }, { "epoch": 4.903529411764706, "grad_norm": 0.3250951773358256, "learning_rate": 3.34510984045861e-05, "loss": 0.1859, "step": 10420 }, { "epoch": 4.904, "grad_norm": 0.8204200074846408, "learning_rate": 3.3423508086435326e-05, "loss": 0.3156, "step": 10421 }, { "epoch": 4.904470588235294, "grad_norm": 0.2142405370853946, "learning_rate": 3.339592772448883e-05, "loss": 0.0546, "step": 10422 }, { "epoch": 4.904941176470588, "grad_norm": 0.15911811392695435, "learning_rate": 3.3368357321102e-05, "loss": 0.0504, "step": 10423 }, { "epoch": 4.905411764705883, "grad_norm": 0.08996763333876132, "learning_rate": 3.334079687862952e-05, "loss": 0.0108, "step": 10424 }, { "epoch": 4.905882352941177, "grad_norm": 0.2572833771847059, "learning_rate": 3.331324639942526e-05, "loss": 0.0817, "step": 10425 }, { "epoch": 4.90635294117647, "grad_norm": 0.2819535106181408, "learning_rate": 3.328570588584204e-05, "loss": 0.1189, "step": 10426 }, { "epoch": 4.906823529411764, "grad_norm": 0.3044941686125376, "learning_rate": 3.325817534023202e-05, "loss": 0.1015, "step": 10427 }, { "epoch": 4.907294117647059, "grad_norm": 0.38627380784028714, "learning_rate": 3.32306547649465e-05, "loss": 0.2669, "step": 10428 }, { "epoch": 4.907764705882353, "grad_norm": 0.3659938736415211, "learning_rate": 3.320314416233574e-05, "loss": 0.116, "step": 10429 }, { "epoch": 4.908235294117647, "grad_norm": 0.3368915502270372, "learning_rate": 3.317564353474943e-05, "loss": 0.1231, "step": 10430 }, { "epoch": 4.908705882352941, "grad_norm": 0.3507547903646935, "learning_rate": 3.314815288453613e-05, "loss": 0.1982, "step": 10431 }, { "epoch": 4.9091764705882355, "grad_norm": 0.5364089576353158, "learning_rate": 3.312067221404371e-05, "loss": 0.3674, "step": 10432 }, { "epoch": 4.90964705882353, "grad_norm": 0.3512783535479048, "learning_rate": 3.309320152561923e-05, "loss": 0.2021, "step": 10433 }, { "epoch": 4.910117647058824, "grad_norm": 0.36467568538660133, "learning_rate": 3.3065740821608725e-05, "loss": 0.2095, "step": 10434 }, { "epoch": 4.910588235294117, "grad_norm": 0.4623353070840861, "learning_rate": 3.303829010435751e-05, "loss": 0.2692, "step": 10435 }, { "epoch": 4.9110588235294115, "grad_norm": 0.347400098607573, "learning_rate": 3.3010849376210056e-05, "loss": 0.2176, "step": 10436 }, { "epoch": 4.911529411764706, "grad_norm": 0.3753481057151986, "learning_rate": 3.298341863950986e-05, "loss": 0.1367, "step": 10437 }, { "epoch": 4.912, "grad_norm": 0.5576906750423787, "learning_rate": 3.295599789659967e-05, "loss": 0.2873, "step": 10438 }, { "epoch": 4.912470588235294, "grad_norm": 0.4879996212866734, "learning_rate": 3.2928587149821385e-05, "loss": 0.3246, "step": 10439 }, { "epoch": 4.912941176470588, "grad_norm": 0.372533455452589, "learning_rate": 3.290118640151597e-05, "loss": 0.242, "step": 10440 }, { "epoch": 4.913411764705883, "grad_norm": 0.28888998810282585, "learning_rate": 3.287379565402358e-05, "loss": 0.0817, "step": 10441 }, { "epoch": 4.913882352941177, "grad_norm": 0.16339644837244957, "learning_rate": 3.284641490968359e-05, "loss": 0.0376, "step": 10442 }, { "epoch": 4.914352941176471, "grad_norm": 0.3350647499871051, "learning_rate": 3.281904417083436e-05, "loss": 0.1487, "step": 10443 }, { "epoch": 4.914823529411764, "grad_norm": 0.509294136218984, "learning_rate": 3.279168343981355e-05, "loss": 0.2672, "step": 10444 }, { "epoch": 4.915294117647059, "grad_norm": 0.33043685176295506, "learning_rate": 3.2764332718957894e-05, "loss": 0.1828, "step": 10445 }, { "epoch": 4.915764705882353, "grad_norm": 0.9181759231861116, "learning_rate": 3.2736992010603226e-05, "loss": 0.1726, "step": 10446 }, { "epoch": 4.916235294117647, "grad_norm": 0.5002080793550958, "learning_rate": 3.270966131708461e-05, "loss": 0.2966, "step": 10447 }, { "epoch": 4.916705882352941, "grad_norm": 0.4739192800061739, "learning_rate": 3.268234064073628e-05, "loss": 0.2784, "step": 10448 }, { "epoch": 4.9171764705882355, "grad_norm": 0.292731396235224, "learning_rate": 3.265502998389143e-05, "loss": 0.0813, "step": 10449 }, { "epoch": 4.91764705882353, "grad_norm": 0.8296412629039035, "learning_rate": 3.262772934888265e-05, "loss": 0.487, "step": 10450 }, { "epoch": 4.918117647058823, "grad_norm": 0.366430754815527, "learning_rate": 3.260043873804147e-05, "loss": 0.1157, "step": 10451 }, { "epoch": 4.918588235294117, "grad_norm": 0.6213703379968515, "learning_rate": 3.257315815369863e-05, "loss": 0.3513, "step": 10452 }, { "epoch": 4.9190588235294115, "grad_norm": 0.233638513025159, "learning_rate": 3.2545887598184114e-05, "loss": 0.092, "step": 10453 }, { "epoch": 4.919529411764706, "grad_norm": 0.6080947963300982, "learning_rate": 3.251862707382686e-05, "loss": 0.2241, "step": 10454 }, { "epoch": 4.92, "grad_norm": 0.5547092905448013, "learning_rate": 3.249137658295511e-05, "loss": 0.3202, "step": 10455 }, { "epoch": 4.920470588235294, "grad_norm": 0.30223492923092427, "learning_rate": 3.246413612789622e-05, "loss": 0.111, "step": 10456 }, { "epoch": 4.920941176470588, "grad_norm": 0.33378753551843376, "learning_rate": 3.243690571097656e-05, "loss": 0.1314, "step": 10457 }, { "epoch": 4.921411764705883, "grad_norm": 0.5088063027455286, "learning_rate": 3.24096853345218e-05, "loss": 0.2158, "step": 10458 }, { "epoch": 4.921882352941177, "grad_norm": 0.2656825596785148, "learning_rate": 3.2382475000856725e-05, "loss": 0.1161, "step": 10459 }, { "epoch": 4.92235294117647, "grad_norm": 0.3269729881728671, "learning_rate": 3.2355274712305154e-05, "loss": 0.2163, "step": 10460 }, { "epoch": 4.922823529411764, "grad_norm": 0.5476662334496439, "learning_rate": 3.232808447119018e-05, "loss": 0.2466, "step": 10461 }, { "epoch": 4.923294117647059, "grad_norm": 0.3819913257968928, "learning_rate": 3.2300904279834e-05, "loss": 0.1888, "step": 10462 }, { "epoch": 4.923764705882353, "grad_norm": 0.2360268767237336, "learning_rate": 3.227373414055782e-05, "loss": 0.0715, "step": 10463 }, { "epoch": 4.924235294117647, "grad_norm": 0.21402340255926205, "learning_rate": 3.224657405568229e-05, "loss": 0.0573, "step": 10464 }, { "epoch": 4.924705882352941, "grad_norm": 0.21712260463877503, "learning_rate": 3.221942402752685e-05, "loss": 0.0527, "step": 10465 }, { "epoch": 4.9251764705882355, "grad_norm": 0.41336443378981264, "learning_rate": 3.2192284058410324e-05, "loss": 0.2834, "step": 10466 }, { "epoch": 4.92564705882353, "grad_norm": 0.2992538049803922, "learning_rate": 3.216515415065061e-05, "loss": 0.1201, "step": 10467 }, { "epoch": 4.926117647058824, "grad_norm": 0.8260430417999274, "learning_rate": 3.213803430656468e-05, "loss": 0.5923, "step": 10468 }, { "epoch": 4.926588235294117, "grad_norm": 0.3955609693206914, "learning_rate": 3.2110924528468725e-05, "loss": 0.1152, "step": 10469 }, { "epoch": 4.9270588235294115, "grad_norm": 0.23783800460393545, "learning_rate": 3.208382481867809e-05, "loss": 0.0939, "step": 10470 }, { "epoch": 4.927529411764706, "grad_norm": 0.314670099151752, "learning_rate": 3.205673517950716e-05, "loss": 0.0473, "step": 10471 }, { "epoch": 4.928, "grad_norm": 0.5355166951895217, "learning_rate": 3.202965561326955e-05, "loss": 0.1703, "step": 10472 }, { "epoch": 4.928470588235294, "grad_norm": 0.3815070052351628, "learning_rate": 3.200258612227803e-05, "loss": 0.1863, "step": 10473 }, { "epoch": 4.928941176470588, "grad_norm": 0.4299490551732266, "learning_rate": 3.19755267088444e-05, "loss": 0.2274, "step": 10474 }, { "epoch": 4.929411764705883, "grad_norm": 0.460440644954492, "learning_rate": 3.194847737527969e-05, "loss": 0.3148, "step": 10475 }, { "epoch": 4.929882352941177, "grad_norm": 0.4181653438149089, "learning_rate": 3.1921438123894086e-05, "loss": 0.2048, "step": 10476 }, { "epoch": 4.930352941176471, "grad_norm": 0.48400732617427306, "learning_rate": 3.189440895699681e-05, "loss": 0.3125, "step": 10477 }, { "epoch": 4.930823529411764, "grad_norm": 0.44295465360864567, "learning_rate": 3.1867389876896296e-05, "loss": 0.1697, "step": 10478 }, { "epoch": 4.931294117647059, "grad_norm": 0.37335591448065675, "learning_rate": 3.184038088590019e-05, "loss": 0.0835, "step": 10479 }, { "epoch": 4.931764705882353, "grad_norm": 1.4071114712536936, "learning_rate": 3.181338198631506e-05, "loss": 0.682, "step": 10480 }, { "epoch": 4.932235294117647, "grad_norm": 0.3545395155224383, "learning_rate": 3.178639318044687e-05, "loss": 0.1035, "step": 10481 }, { "epoch": 4.932705882352941, "grad_norm": 0.2653533236308347, "learning_rate": 3.175941447060048e-05, "loss": 0.1145, "step": 10482 }, { "epoch": 4.9331764705882355, "grad_norm": 0.7228530599671468, "learning_rate": 3.173244585908007e-05, "loss": 0.3625, "step": 10483 }, { "epoch": 4.93364705882353, "grad_norm": 0.5536757068634832, "learning_rate": 3.170548734818893e-05, "loss": 0.3809, "step": 10484 }, { "epoch": 4.934117647058823, "grad_norm": 0.4238685410916647, "learning_rate": 3.1678538940229356e-05, "loss": 0.1919, "step": 10485 }, { "epoch": 4.934588235294117, "grad_norm": 0.38467068834451, "learning_rate": 3.16516006375029e-05, "loss": 0.2088, "step": 10486 }, { "epoch": 4.9350588235294115, "grad_norm": 0.38049972218477623, "learning_rate": 3.1624672442310304e-05, "loss": 0.1924, "step": 10487 }, { "epoch": 4.935529411764706, "grad_norm": 0.33298035623168193, "learning_rate": 3.159775435695126e-05, "loss": 0.2125, "step": 10488 }, { "epoch": 4.936, "grad_norm": 0.4259874836055417, "learning_rate": 3.1570846383724775e-05, "loss": 0.2172, "step": 10489 }, { "epoch": 4.936470588235294, "grad_norm": 0.4750139780926214, "learning_rate": 3.154394852492892e-05, "loss": 0.274, "step": 10490 }, { "epoch": 4.936941176470588, "grad_norm": 0.7128085197838417, "learning_rate": 3.1517060782860865e-05, "loss": 0.4191, "step": 10491 }, { "epoch": 4.937411764705883, "grad_norm": 0.23899067570058005, "learning_rate": 3.1490183159816956e-05, "loss": 0.0814, "step": 10492 }, { "epoch": 4.937882352941177, "grad_norm": 0.22770262021783153, "learning_rate": 3.146331565809273e-05, "loss": 0.0956, "step": 10493 }, { "epoch": 4.93835294117647, "grad_norm": 0.38403757084016626, "learning_rate": 3.143645827998274e-05, "loss": 0.2476, "step": 10494 }, { "epoch": 4.938823529411764, "grad_norm": 0.6487804571886991, "learning_rate": 3.140961102778075e-05, "loss": 0.4248, "step": 10495 }, { "epoch": 4.939294117647059, "grad_norm": 0.37163412715340854, "learning_rate": 3.138277390377971e-05, "loss": 0.2448, "step": 10496 }, { "epoch": 4.939764705882353, "grad_norm": 1.0460680051099558, "learning_rate": 3.135594691027155e-05, "loss": 0.3391, "step": 10497 }, { "epoch": 4.940235294117647, "grad_norm": 0.27848005912715845, "learning_rate": 3.132913004954747e-05, "loss": 0.1244, "step": 10498 }, { "epoch": 4.940705882352941, "grad_norm": 0.6057607692257824, "learning_rate": 3.130232332389781e-05, "loss": 0.4851, "step": 10499 }, { "epoch": 4.9411764705882355, "grad_norm": 0.37384624196890553, "learning_rate": 3.1275526735611896e-05, "loss": 0.1673, "step": 10500 }, { "epoch": 4.94164705882353, "grad_norm": 0.5649011698924258, "learning_rate": 3.1248740286978375e-05, "loss": 0.4103, "step": 10501 }, { "epoch": 4.942117647058824, "grad_norm": 0.3470023680333888, "learning_rate": 3.1221963980284876e-05, "loss": 0.1356, "step": 10502 }, { "epoch": 4.942588235294117, "grad_norm": 0.3724764096201154, "learning_rate": 3.119519781781825e-05, "loss": 0.1681, "step": 10503 }, { "epoch": 4.9430588235294115, "grad_norm": 0.5436645424392256, "learning_rate": 3.1168441801864505e-05, "loss": 0.3395, "step": 10504 }, { "epoch": 4.943529411764706, "grad_norm": 0.3380621248334419, "learning_rate": 3.1141695934708664e-05, "loss": 0.1237, "step": 10505 }, { "epoch": 4.944, "grad_norm": 0.35475042373043386, "learning_rate": 3.1114960218634986e-05, "loss": 0.1596, "step": 10506 }, { "epoch": 4.944470588235294, "grad_norm": 0.16229901270662006, "learning_rate": 3.1088234655926815e-05, "loss": 0.0358, "step": 10507 }, { "epoch": 4.944941176470588, "grad_norm": 0.5017520355533998, "learning_rate": 3.106151924886668e-05, "loss": 0.3835, "step": 10508 }, { "epoch": 4.945411764705883, "grad_norm": 0.27270296699345953, "learning_rate": 3.103481399973622e-05, "loss": 0.0715, "step": 10509 }, { "epoch": 4.945882352941177, "grad_norm": 0.35582332181323034, "learning_rate": 3.100811891081614e-05, "loss": 0.1464, "step": 10510 }, { "epoch": 4.946352941176471, "grad_norm": 0.3536121460111363, "learning_rate": 3.098143398438635e-05, "loss": 0.2128, "step": 10511 }, { "epoch": 4.946823529411764, "grad_norm": 0.6205596834267122, "learning_rate": 3.0954759222725916e-05, "loss": 0.3439, "step": 10512 }, { "epoch": 4.947294117647059, "grad_norm": 0.3649452069203106, "learning_rate": 3.092809462811292e-05, "loss": 0.151, "step": 10513 }, { "epoch": 4.947764705882353, "grad_norm": 0.38986478787404427, "learning_rate": 3.090144020282469e-05, "loss": 0.2523, "step": 10514 }, { "epoch": 4.948235294117647, "grad_norm": 0.38720732810732184, "learning_rate": 3.087479594913767e-05, "loss": 0.2389, "step": 10515 }, { "epoch": 4.948705882352941, "grad_norm": 0.41277759777537776, "learning_rate": 3.0848161869327356e-05, "loss": 0.2829, "step": 10516 }, { "epoch": 4.9491764705882355, "grad_norm": 0.27339489367218484, "learning_rate": 3.0821537965668454e-05, "loss": 0.0961, "step": 10517 }, { "epoch": 4.94964705882353, "grad_norm": 0.7683264701752817, "learning_rate": 3.079492424043481e-05, "loss": 0.3171, "step": 10518 }, { "epoch": 4.950117647058823, "grad_norm": 0.5496325925791219, "learning_rate": 3.0768320695899294e-05, "loss": 0.3933, "step": 10519 }, { "epoch": 4.950588235294117, "grad_norm": 0.32025247837221776, "learning_rate": 3.0741727334334026e-05, "loss": 0.0858, "step": 10520 }, { "epoch": 4.9510588235294115, "grad_norm": 0.37752966218322387, "learning_rate": 3.071514415801024e-05, "loss": 0.2262, "step": 10521 }, { "epoch": 4.951529411764706, "grad_norm": 0.5126557330045269, "learning_rate": 3.068857116919819e-05, "loss": 0.1794, "step": 10522 }, { "epoch": 4.952, "grad_norm": 0.3481116524560776, "learning_rate": 3.066200837016737e-05, "loss": 0.1023, "step": 10523 }, { "epoch": 4.952470588235294, "grad_norm": 0.3801091604699569, "learning_rate": 3.063545576318644e-05, "loss": 0.1139, "step": 10524 }, { "epoch": 4.952941176470588, "grad_norm": 0.4714494544147263, "learning_rate": 3.060891335052302e-05, "loss": 0.203, "step": 10525 }, { "epoch": 4.953411764705883, "grad_norm": 0.5546358040899991, "learning_rate": 3.058238113444401e-05, "loss": 0.3179, "step": 10526 }, { "epoch": 4.953882352941177, "grad_norm": 0.22133116959846594, "learning_rate": 3.0555859117215446e-05, "loss": 0.0653, "step": 10527 }, { "epoch": 4.95435294117647, "grad_norm": 0.40167483757776135, "learning_rate": 3.052934730110232e-05, "loss": 0.1188, "step": 10528 }, { "epoch": 4.954823529411764, "grad_norm": 0.39578762495105735, "learning_rate": 3.050284568836895e-05, "loss": 0.1368, "step": 10529 }, { "epoch": 4.955294117647059, "grad_norm": 0.4957526002617255, "learning_rate": 3.0476354281278726e-05, "loss": 0.1707, "step": 10530 }, { "epoch": 4.955764705882353, "grad_norm": 0.5787110597262817, "learning_rate": 3.044987308209407e-05, "loss": 0.3349, "step": 10531 }, { "epoch": 4.956235294117647, "grad_norm": 0.300333557357027, "learning_rate": 3.0423402093076683e-05, "loss": 0.1799, "step": 10532 }, { "epoch": 4.956705882352941, "grad_norm": 0.36763085966120085, "learning_rate": 3.039694131648723e-05, "loss": 0.1355, "step": 10533 }, { "epoch": 4.9571764705882355, "grad_norm": 0.2733805703529933, "learning_rate": 3.0370490754585636e-05, "loss": 0.0922, "step": 10534 }, { "epoch": 4.95764705882353, "grad_norm": 0.42697414999921346, "learning_rate": 3.034405040963095e-05, "loss": 0.1417, "step": 10535 }, { "epoch": 4.958117647058824, "grad_norm": 0.32123134049486424, "learning_rate": 3.0317620283881243e-05, "loss": 0.1695, "step": 10536 }, { "epoch": 4.958588235294117, "grad_norm": 0.386717080741166, "learning_rate": 3.0291200379593782e-05, "loss": 0.1854, "step": 10537 }, { "epoch": 4.9590588235294115, "grad_norm": 0.6166980849918094, "learning_rate": 3.0264790699025027e-05, "loss": 0.442, "step": 10538 }, { "epoch": 4.959529411764706, "grad_norm": 0.33197219665399663, "learning_rate": 3.0238391244430405e-05, "loss": 0.1113, "step": 10539 }, { "epoch": 4.96, "grad_norm": 0.417615134844097, "learning_rate": 3.0212002018064614e-05, "loss": 0.2023, "step": 10540 }, { "epoch": 4.960470588235294, "grad_norm": 0.7657955394540563, "learning_rate": 3.0185623022181415e-05, "loss": 0.4284, "step": 10541 }, { "epoch": 4.960941176470588, "grad_norm": 0.4028096427932891, "learning_rate": 3.015925425903368e-05, "loss": 0.2015, "step": 10542 }, { "epoch": 4.961411764705883, "grad_norm": 0.49136751630823267, "learning_rate": 3.0132895730873448e-05, "loss": 0.2401, "step": 10543 }, { "epoch": 4.961882352941177, "grad_norm": 0.4558530015868791, "learning_rate": 3.010654743995189e-05, "loss": 0.2404, "step": 10544 }, { "epoch": 4.962352941176471, "grad_norm": 0.6025313866941808, "learning_rate": 3.0080209388519222e-05, "loss": 0.3616, "step": 10545 }, { "epoch": 4.962823529411764, "grad_norm": 0.28618465173016777, "learning_rate": 3.005388157882488e-05, "loss": 0.1521, "step": 10546 }, { "epoch": 4.963294117647059, "grad_norm": 0.5405350712561839, "learning_rate": 3.002756401311741e-05, "loss": 0.2622, "step": 10547 }, { "epoch": 4.963764705882353, "grad_norm": 0.49040967228200943, "learning_rate": 3.0001256693644405e-05, "loss": 0.2811, "step": 10548 }, { "epoch": 4.964235294117647, "grad_norm": 0.6069143599785893, "learning_rate": 2.997495962265266e-05, "loss": 0.3464, "step": 10549 }, { "epoch": 4.964705882352941, "grad_norm": 0.24651405463034323, "learning_rate": 2.9948672802388135e-05, "loss": 0.0958, "step": 10550 }, { "epoch": 4.9651764705882355, "grad_norm": 0.5060832039411945, "learning_rate": 2.9922396235095762e-05, "loss": 0.3369, "step": 10551 }, { "epoch": 4.96564705882353, "grad_norm": 0.19556044662252933, "learning_rate": 2.9896129923019717e-05, "loss": 0.0796, "step": 10552 }, { "epoch": 4.966117647058823, "grad_norm": 0.3672745855105463, "learning_rate": 2.9869873868403287e-05, "loss": 0.1789, "step": 10553 }, { "epoch": 4.966588235294117, "grad_norm": 0.4105352147031191, "learning_rate": 2.9843628073488863e-05, "loss": 0.1801, "step": 10554 }, { "epoch": 4.9670588235294115, "grad_norm": 0.25184317899957703, "learning_rate": 2.9817392540518014e-05, "loss": 0.0683, "step": 10555 }, { "epoch": 4.967529411764706, "grad_norm": 0.430926703665171, "learning_rate": 2.9791167271731287e-05, "loss": 0.2635, "step": 10556 }, { "epoch": 4.968, "grad_norm": 0.22293411128002835, "learning_rate": 2.9764952269368487e-05, "loss": 0.0395, "step": 10557 }, { "epoch": 4.968470588235294, "grad_norm": 0.5094673069908048, "learning_rate": 2.973874753566857e-05, "loss": 0.387, "step": 10558 }, { "epoch": 4.968941176470588, "grad_norm": 0.6400714834702389, "learning_rate": 2.9712553072869437e-05, "loss": 0.1479, "step": 10559 }, { "epoch": 4.969411764705883, "grad_norm": 0.42154462726462405, "learning_rate": 2.968636888320833e-05, "loss": 0.2808, "step": 10560 }, { "epoch": 4.969882352941177, "grad_norm": 0.1515663579283699, "learning_rate": 2.9660194968921403e-05, "loss": 0.0175, "step": 10561 }, { "epoch": 4.97035294117647, "grad_norm": 0.3689069196383349, "learning_rate": 2.963403133224409e-05, "loss": 0.21, "step": 10562 }, { "epoch": 4.970823529411764, "grad_norm": 0.4687217659801645, "learning_rate": 2.9607877975410936e-05, "loss": 0.2504, "step": 10563 }, { "epoch": 4.971294117647059, "grad_norm": 0.22337980728331666, "learning_rate": 2.958173490065549e-05, "loss": 0.0344, "step": 10564 }, { "epoch": 4.971764705882353, "grad_norm": 0.24046827483484862, "learning_rate": 2.955560211021053e-05, "loss": 0.0786, "step": 10565 }, { "epoch": 4.972235294117647, "grad_norm": 0.4895641521503106, "learning_rate": 2.952947960630797e-05, "loss": 0.1804, "step": 10566 }, { "epoch": 4.972705882352941, "grad_norm": 0.4366198785664804, "learning_rate": 2.9503367391178713e-05, "loss": 0.2506, "step": 10567 }, { "epoch": 4.9731764705882355, "grad_norm": 0.28472007800435306, "learning_rate": 2.94772654670529e-05, "loss": 0.128, "step": 10568 }, { "epoch": 4.97364705882353, "grad_norm": 0.363256614049592, "learning_rate": 2.9451173836159826e-05, "loss": 0.2626, "step": 10569 }, { "epoch": 4.974117647058824, "grad_norm": 0.44344494510245674, "learning_rate": 2.9425092500727764e-05, "loss": 0.2412, "step": 10570 }, { "epoch": 4.974588235294117, "grad_norm": 0.3266312202602474, "learning_rate": 2.9399021462984206e-05, "loss": 0.1718, "step": 10571 }, { "epoch": 4.9750588235294115, "grad_norm": 0.4489779288249205, "learning_rate": 2.9372960725155804e-05, "loss": 0.1922, "step": 10572 }, { "epoch": 4.975529411764706, "grad_norm": 0.4657564867553107, "learning_rate": 2.934691028946819e-05, "loss": 0.1242, "step": 10573 }, { "epoch": 4.976, "grad_norm": 0.4724212715373485, "learning_rate": 2.9320870158146235e-05, "loss": 0.2924, "step": 10574 }, { "epoch": 4.976470588235294, "grad_norm": 0.21904372719208207, "learning_rate": 2.9294840333413932e-05, "loss": 0.095, "step": 10575 }, { "epoch": 4.976941176470588, "grad_norm": 0.2673266024326884, "learning_rate": 2.926882081749428e-05, "loss": 0.1278, "step": 10576 }, { "epoch": 4.977411764705883, "grad_norm": 0.3359011800107963, "learning_rate": 2.9242811612609522e-05, "loss": 0.1831, "step": 10577 }, { "epoch": 4.977882352941176, "grad_norm": 0.3421417268722325, "learning_rate": 2.9216812720980998e-05, "loss": 0.1703, "step": 10578 }, { "epoch": 4.97835294117647, "grad_norm": 0.20660004498372658, "learning_rate": 2.919082414482908e-05, "loss": 0.0518, "step": 10579 }, { "epoch": 4.978823529411764, "grad_norm": 0.17708283125506735, "learning_rate": 2.9164845886373323e-05, "loss": 0.047, "step": 10580 }, { "epoch": 4.979294117647059, "grad_norm": 0.2661972043977025, "learning_rate": 2.9138877947832474e-05, "loss": 0.1349, "step": 10581 }, { "epoch": 4.979764705882353, "grad_norm": 0.39533893720088525, "learning_rate": 2.911292033142422e-05, "loss": 0.206, "step": 10582 }, { "epoch": 4.980235294117647, "grad_norm": 0.8108184506440923, "learning_rate": 2.9086973039365562e-05, "loss": 0.3078, "step": 10583 }, { "epoch": 4.980705882352941, "grad_norm": 0.3609095162217287, "learning_rate": 2.906103607387244e-05, "loss": 0.0931, "step": 10584 }, { "epoch": 4.9811764705882355, "grad_norm": 0.21332847419927733, "learning_rate": 2.9035109437160047e-05, "loss": 0.0525, "step": 10585 }, { "epoch": 4.98164705882353, "grad_norm": 0.22735280266373573, "learning_rate": 2.9009193131442665e-05, "loss": 0.0535, "step": 10586 }, { "epoch": 4.982117647058823, "grad_norm": 0.3685469946407654, "learning_rate": 2.8983287158933622e-05, "loss": 0.1039, "step": 10587 }, { "epoch": 4.982588235294117, "grad_norm": 0.1934157028527918, "learning_rate": 2.895739152184543e-05, "loss": 0.0585, "step": 10588 }, { "epoch": 4.9830588235294115, "grad_norm": 0.7091179085616769, "learning_rate": 2.8931506222389756e-05, "loss": 0.236, "step": 10589 }, { "epoch": 4.983529411764706, "grad_norm": 0.1670301193830465, "learning_rate": 2.8905631262777245e-05, "loss": 0.0227, "step": 10590 }, { "epoch": 4.984, "grad_norm": 0.25188649668911123, "learning_rate": 2.8879766645217784e-05, "loss": 0.0991, "step": 10591 }, { "epoch": 4.984470588235294, "grad_norm": 0.27964750843903896, "learning_rate": 2.8853912371920384e-05, "loss": 0.1145, "step": 10592 }, { "epoch": 4.984941176470588, "grad_norm": 0.3746433645786192, "learning_rate": 2.882806844509305e-05, "loss": 0.2396, "step": 10593 }, { "epoch": 4.985411764705883, "grad_norm": 0.4459704773039038, "learning_rate": 2.8802234866943012e-05, "loss": 0.2442, "step": 10594 }, { "epoch": 4.985882352941177, "grad_norm": 0.5787665680699168, "learning_rate": 2.8776411639676644e-05, "loss": 0.2348, "step": 10595 }, { "epoch": 4.98635294117647, "grad_norm": 0.4164658011751518, "learning_rate": 2.8750598765499226e-05, "loss": 0.1122, "step": 10596 }, { "epoch": 4.986823529411764, "grad_norm": 0.3029711140668151, "learning_rate": 2.8724796246615485e-05, "loss": 0.0976, "step": 10597 }, { "epoch": 4.987294117647059, "grad_norm": 0.23584549604020547, "learning_rate": 2.8699004085228948e-05, "loss": 0.1109, "step": 10598 }, { "epoch": 4.987764705882353, "grad_norm": 0.2656519419471349, "learning_rate": 2.867322228354243e-05, "loss": 0.0745, "step": 10599 }, { "epoch": 4.988235294117647, "grad_norm": 0.4938494944353137, "learning_rate": 2.8647450843757897e-05, "loss": 0.3222, "step": 10600 }, { "epoch": 4.988705882352941, "grad_norm": 0.4971765090364214, "learning_rate": 2.862168976807624e-05, "loss": 0.3748, "step": 10601 }, { "epoch": 4.9891764705882355, "grad_norm": 0.21950386954376858, "learning_rate": 2.859593905869763e-05, "loss": 0.0912, "step": 10602 }, { "epoch": 4.98964705882353, "grad_norm": 0.4416217149709826, "learning_rate": 2.8570198717821362e-05, "loss": 0.2233, "step": 10603 }, { "epoch": 4.990117647058824, "grad_norm": 0.3715100583595862, "learning_rate": 2.8544468747645676e-05, "loss": 0.1662, "step": 10604 }, { "epoch": 4.990588235294117, "grad_norm": 0.2710416749273398, "learning_rate": 2.8518749150368114e-05, "loss": 0.1295, "step": 10605 }, { "epoch": 4.9910588235294115, "grad_norm": 0.17184629432307139, "learning_rate": 2.849303992818527e-05, "loss": 0.0382, "step": 10606 }, { "epoch": 4.991529411764706, "grad_norm": 0.37524903076548227, "learning_rate": 2.846734108329275e-05, "loss": 0.1399, "step": 10607 }, { "epoch": 4.992, "grad_norm": 0.36560856434723715, "learning_rate": 2.8441652617885437e-05, "loss": 0.102, "step": 10608 }, { "epoch": 4.992470588235294, "grad_norm": 0.3066628214688468, "learning_rate": 2.8415974534157255e-05, "loss": 0.1438, "step": 10609 }, { "epoch": 4.992941176470588, "grad_norm": 0.9205339975502209, "learning_rate": 2.8390306834301186e-05, "loss": 0.5571, "step": 10610 }, { "epoch": 4.993411764705883, "grad_norm": 0.5117660135002077, "learning_rate": 2.836464952050944e-05, "loss": 0.2655, "step": 10611 }, { "epoch": 4.993882352941176, "grad_norm": 0.25534064007092305, "learning_rate": 2.8339002594973204e-05, "loss": 0.0765, "step": 10612 }, { "epoch": 4.99435294117647, "grad_norm": 0.25820293134180744, "learning_rate": 2.8313366059882887e-05, "loss": 0.0925, "step": 10613 }, { "epoch": 4.994823529411764, "grad_norm": 0.38416416382248275, "learning_rate": 2.828773991742802e-05, "loss": 0.1773, "step": 10614 }, { "epoch": 4.995294117647059, "grad_norm": 0.4917325406374177, "learning_rate": 2.8262124169797106e-05, "loss": 0.3232, "step": 10615 }, { "epoch": 4.995764705882353, "grad_norm": 0.3827929224894731, "learning_rate": 2.823651881917793e-05, "loss": 0.1637, "step": 10616 }, { "epoch": 4.996235294117647, "grad_norm": 0.14923607606522749, "learning_rate": 2.8210923867757317e-05, "loss": 0.0393, "step": 10617 }, { "epoch": 4.996705882352941, "grad_norm": 0.23981866544116834, "learning_rate": 2.8185339317721146e-05, "loss": 0.0909, "step": 10618 }, { "epoch": 4.9971764705882356, "grad_norm": 0.4011399939453106, "learning_rate": 2.8159765171254495e-05, "loss": 0.2853, "step": 10619 }, { "epoch": 4.99764705882353, "grad_norm": 0.6186729159089236, "learning_rate": 2.813420143054157e-05, "loss": 0.1545, "step": 10620 }, { "epoch": 4.998117647058823, "grad_norm": 0.3553347149620053, "learning_rate": 2.810864809776554e-05, "loss": 0.1614, "step": 10621 }, { "epoch": 4.998588235294117, "grad_norm": 0.36075226931276916, "learning_rate": 2.8083105175108845e-05, "loss": 0.2082, "step": 10622 }, { "epoch": 4.9990588235294116, "grad_norm": 0.4051076792909761, "learning_rate": 2.8057572664752986e-05, "loss": 0.134, "step": 10623 }, { "epoch": 4.999529411764706, "grad_norm": 0.4695760788318255, "learning_rate": 2.8032050568878524e-05, "loss": 0.2031, "step": 10624 }, { "epoch": 5.0, "grad_norm": 0.582696305966886, "learning_rate": 2.800653888966519e-05, "loss": 0.2455, "step": 10625 }, { "epoch": 5.0, "eval_loss": 0.20219503343105316, "eval_runtime": 65.3924, "eval_samples_per_second": 11.469, "eval_steps_per_second": 5.735, "step": 10625 }, { "epoch": 5.000470588235294, "grad_norm": 0.680954520403659, "learning_rate": 2.7981037629291852e-05, "loss": 0.2052, "step": 10626 }, { "epoch": 5.0009411764705884, "grad_norm": 0.540604590291474, "learning_rate": 2.7955546789936363e-05, "loss": 0.1302, "step": 10627 }, { "epoch": 5.001411764705883, "grad_norm": 0.5483671504688407, "learning_rate": 2.7930066373775794e-05, "loss": 0.4233, "step": 10628 }, { "epoch": 5.001882352941177, "grad_norm": 0.22891095273180687, "learning_rate": 2.7904596382986343e-05, "loss": 0.0964, "step": 10629 }, { "epoch": 5.00235294117647, "grad_norm": 0.3348918621427304, "learning_rate": 2.7879136819743203e-05, "loss": 0.0858, "step": 10630 }, { "epoch": 5.0028235294117644, "grad_norm": 0.3517936008600509, "learning_rate": 2.7853687686220766e-05, "loss": 0.2101, "step": 10631 }, { "epoch": 5.003294117647059, "grad_norm": 0.5521344584664327, "learning_rate": 2.782824898459258e-05, "loss": 0.2864, "step": 10632 }, { "epoch": 5.003764705882353, "grad_norm": 0.2329825448671833, "learning_rate": 2.7802820717031132e-05, "loss": 0.0647, "step": 10633 }, { "epoch": 5.004235294117647, "grad_norm": 0.40614421028334047, "learning_rate": 2.7777402885708212e-05, "loss": 0.2551, "step": 10634 }, { "epoch": 5.004705882352941, "grad_norm": 0.7588914227255568, "learning_rate": 2.7751995492794544e-05, "loss": 0.2918, "step": 10635 }, { "epoch": 5.005176470588236, "grad_norm": 0.29764599428702676, "learning_rate": 2.7726598540460076e-05, "loss": 0.1211, "step": 10636 }, { "epoch": 5.00564705882353, "grad_norm": 0.4409261608936938, "learning_rate": 2.770121203087389e-05, "loss": 0.2097, "step": 10637 }, { "epoch": 5.006117647058823, "grad_norm": 0.24670514723991688, "learning_rate": 2.7675835966204024e-05, "loss": 0.0798, "step": 10638 }, { "epoch": 5.006588235294117, "grad_norm": 0.12220215054578915, "learning_rate": 2.7650470348617772e-05, "loss": 0.0244, "step": 10639 }, { "epoch": 5.007058823529412, "grad_norm": 0.7211625116501037, "learning_rate": 2.7625115180281464e-05, "loss": 0.4119, "step": 10640 }, { "epoch": 5.007529411764706, "grad_norm": 0.3444110275954441, "learning_rate": 2.7599770463360576e-05, "loss": 0.2506, "step": 10641 }, { "epoch": 5.008, "grad_norm": 0.6452601375831943, "learning_rate": 2.7574436200019688e-05, "loss": 0.3512, "step": 10642 }, { "epoch": 5.008470588235294, "grad_norm": 0.2122438433591581, "learning_rate": 2.7549112392422405e-05, "loss": 0.0462, "step": 10643 }, { "epoch": 5.0089411764705885, "grad_norm": 0.6099299515867259, "learning_rate": 2.7523799042731548e-05, "loss": 0.2181, "step": 10644 }, { "epoch": 5.009411764705883, "grad_norm": 0.430916464900252, "learning_rate": 2.749849615310902e-05, "loss": 0.213, "step": 10645 }, { "epoch": 5.009882352941177, "grad_norm": 0.3461498107459436, "learning_rate": 2.7473203725715754e-05, "loss": 0.2096, "step": 10646 }, { "epoch": 5.01035294117647, "grad_norm": 0.40468044063758346, "learning_rate": 2.7447921762711894e-05, "loss": 0.1747, "step": 10647 }, { "epoch": 5.0108235294117645, "grad_norm": 0.41042109679699534, "learning_rate": 2.7422650266256663e-05, "loss": 0.1323, "step": 10648 }, { "epoch": 5.011294117647059, "grad_norm": 0.296846691979072, "learning_rate": 2.7397389238508282e-05, "loss": 0.1234, "step": 10649 }, { "epoch": 5.011764705882353, "grad_norm": 0.12714786552312066, "learning_rate": 2.7372138681624244e-05, "loss": 0.0253, "step": 10650 }, { "epoch": 5.012235294117647, "grad_norm": 0.33324309025769255, "learning_rate": 2.7346898597761073e-05, "loss": 0.131, "step": 10651 }, { "epoch": 5.012705882352941, "grad_norm": 0.45579437836670705, "learning_rate": 2.7321668989074347e-05, "loss": 0.2681, "step": 10652 }, { "epoch": 5.013176470588236, "grad_norm": 0.5204826834611616, "learning_rate": 2.7296449857718806e-05, "loss": 0.424, "step": 10653 }, { "epoch": 5.01364705882353, "grad_norm": 0.5064607288280026, "learning_rate": 2.727124120584836e-05, "loss": 0.2466, "step": 10654 }, { "epoch": 5.014117647058823, "grad_norm": 0.39902318962477923, "learning_rate": 2.724604303561584e-05, "loss": 0.0832, "step": 10655 }, { "epoch": 5.014588235294117, "grad_norm": 0.35289494500029245, "learning_rate": 2.7220855349173353e-05, "loss": 0.1659, "step": 10656 }, { "epoch": 5.015058823529412, "grad_norm": 0.4307195519500462, "learning_rate": 2.71956781486721e-05, "loss": 0.2253, "step": 10657 }, { "epoch": 5.015529411764706, "grad_norm": 0.5199415389085954, "learning_rate": 2.7170511436262243e-05, "loss": 0.3061, "step": 10658 }, { "epoch": 5.016, "grad_norm": 0.12963581091848542, "learning_rate": 2.7145355214093178e-05, "loss": 0.0223, "step": 10659 }, { "epoch": 5.016470588235294, "grad_norm": 0.3996639695006852, "learning_rate": 2.7120209484313422e-05, "loss": 0.1207, "step": 10660 }, { "epoch": 5.0169411764705885, "grad_norm": 0.6734292666713318, "learning_rate": 2.7095074249070455e-05, "loss": 0.2118, "step": 10661 }, { "epoch": 5.017411764705883, "grad_norm": 0.15711669729057692, "learning_rate": 2.7069949510511023e-05, "loss": 0.0281, "step": 10662 }, { "epoch": 5.017882352941177, "grad_norm": 0.4004231916847804, "learning_rate": 2.7044835270780836e-05, "loss": 0.1932, "step": 10663 }, { "epoch": 5.01835294117647, "grad_norm": 0.38099911377552526, "learning_rate": 2.701973153202482e-05, "loss": 0.1649, "step": 10664 }, { "epoch": 5.0188235294117645, "grad_norm": 0.5649986062069051, "learning_rate": 2.6994638296386977e-05, "loss": 0.3046, "step": 10665 }, { "epoch": 5.019294117647059, "grad_norm": 0.562267997648, "learning_rate": 2.6969555566010316e-05, "loss": 0.3155, "step": 10666 }, { "epoch": 5.019764705882353, "grad_norm": 0.5020972051311995, "learning_rate": 2.694448334303708e-05, "loss": 0.1737, "step": 10667 }, { "epoch": 5.020235294117647, "grad_norm": 0.49231504328460723, "learning_rate": 2.6919421629608584e-05, "loss": 0.2784, "step": 10668 }, { "epoch": 5.020705882352941, "grad_norm": 0.5204251905402574, "learning_rate": 2.6894370427865157e-05, "loss": 0.273, "step": 10669 }, { "epoch": 5.021176470588236, "grad_norm": 0.39445859127577104, "learning_rate": 2.6869329739946312e-05, "loss": 0.2281, "step": 10670 }, { "epoch": 5.02164705882353, "grad_norm": 0.2699246474074968, "learning_rate": 2.6844299567990713e-05, "loss": 0.117, "step": 10671 }, { "epoch": 5.022117647058823, "grad_norm": 0.3952092059001414, "learning_rate": 2.6819279914135956e-05, "loss": 0.1644, "step": 10672 }, { "epoch": 5.022588235294117, "grad_norm": 0.4808205807696882, "learning_rate": 2.6794270780518895e-05, "loss": 0.2434, "step": 10673 }, { "epoch": 5.023058823529412, "grad_norm": 0.04726341973482318, "learning_rate": 2.6769272169275474e-05, "loss": 0.0061, "step": 10674 }, { "epoch": 5.023529411764706, "grad_norm": 0.43579085986751437, "learning_rate": 2.67442840825406e-05, "loss": 0.2406, "step": 10675 }, { "epoch": 5.024, "grad_norm": 0.5303621518783931, "learning_rate": 2.671930652244845e-05, "loss": 0.2263, "step": 10676 }, { "epoch": 5.024470588235294, "grad_norm": 0.15810692335446164, "learning_rate": 2.6694339491132237e-05, "loss": 0.0339, "step": 10677 }, { "epoch": 5.0249411764705885, "grad_norm": 0.38442912048690103, "learning_rate": 2.666938299072422e-05, "loss": 0.2331, "step": 10678 }, { "epoch": 5.025411764705883, "grad_norm": 0.9292914649839928, "learning_rate": 2.664443702335581e-05, "loss": 0.3641, "step": 10679 }, { "epoch": 5.025882352941177, "grad_norm": 0.36394752351388493, "learning_rate": 2.6619501591157583e-05, "loss": 0.2219, "step": 10680 }, { "epoch": 5.02635294117647, "grad_norm": 0.6085793876129655, "learning_rate": 2.6594576696259055e-05, "loss": 0.2961, "step": 10681 }, { "epoch": 5.0268235294117645, "grad_norm": 0.2723296610287722, "learning_rate": 2.6569662340789012e-05, "loss": 0.1187, "step": 10682 }, { "epoch": 5.027294117647059, "grad_norm": 0.3203701140678099, "learning_rate": 2.6544758526875193e-05, "loss": 0.187, "step": 10683 }, { "epoch": 5.027764705882353, "grad_norm": 0.22727132366356392, "learning_rate": 2.651986525664455e-05, "loss": 0.086, "step": 10684 }, { "epoch": 5.028235294117647, "grad_norm": 0.18006077892837127, "learning_rate": 2.649498253222307e-05, "loss": 0.0367, "step": 10685 }, { "epoch": 5.028705882352941, "grad_norm": 0.5695009505142827, "learning_rate": 2.647011035573588e-05, "loss": 0.276, "step": 10686 }, { "epoch": 5.029176470588236, "grad_norm": 0.2675731627526499, "learning_rate": 2.6445248729307165e-05, "loss": 0.0748, "step": 10687 }, { "epoch": 5.02964705882353, "grad_norm": 0.26065978150530755, "learning_rate": 2.6420397655060282e-05, "loss": 0.0606, "step": 10688 }, { "epoch": 5.030117647058823, "grad_norm": 0.3627045917616826, "learning_rate": 2.639555713511756e-05, "loss": 0.2273, "step": 10689 }, { "epoch": 5.030588235294117, "grad_norm": 0.3860628110497501, "learning_rate": 2.637072717160056e-05, "loss": 0.1752, "step": 10690 }, { "epoch": 5.031058823529412, "grad_norm": 0.5259327495109194, "learning_rate": 2.6345907766629843e-05, "loss": 0.1992, "step": 10691 }, { "epoch": 5.031529411764706, "grad_norm": 0.42602296192008615, "learning_rate": 2.6321098922325102e-05, "loss": 0.1771, "step": 10692 }, { "epoch": 5.032, "grad_norm": 0.4156625672824071, "learning_rate": 2.6296300640805206e-05, "loss": 0.1624, "step": 10693 }, { "epoch": 5.032470588235294, "grad_norm": 0.6899412447254429, "learning_rate": 2.627151292418795e-05, "loss": 0.246, "step": 10694 }, { "epoch": 5.0329411764705885, "grad_norm": 0.3332210992648431, "learning_rate": 2.6246735774590383e-05, "loss": 0.0822, "step": 10695 }, { "epoch": 5.033411764705883, "grad_norm": 0.5425665429137857, "learning_rate": 2.6221969194128617e-05, "loss": 0.1888, "step": 10696 }, { "epoch": 5.033882352941176, "grad_norm": 0.2741445003773395, "learning_rate": 2.6197213184917775e-05, "loss": 0.089, "step": 10697 }, { "epoch": 5.03435294117647, "grad_norm": 0.3379566548422801, "learning_rate": 2.6172467749072185e-05, "loss": 0.1122, "step": 10698 }, { "epoch": 5.0348235294117645, "grad_norm": 0.12380250971969954, "learning_rate": 2.6147732888705235e-05, "loss": 0.036, "step": 10699 }, { "epoch": 5.035294117647059, "grad_norm": 1.3469816240837535, "learning_rate": 2.6123008605929375e-05, "loss": 0.8249, "step": 10700 }, { "epoch": 5.035764705882353, "grad_norm": 0.5496003220828747, "learning_rate": 2.6098294902856177e-05, "loss": 0.3502, "step": 10701 }, { "epoch": 5.036235294117647, "grad_norm": 0.23267172316522589, "learning_rate": 2.607359178159637e-05, "loss": 0.0639, "step": 10702 }, { "epoch": 5.036705882352941, "grad_norm": 0.5250626246739452, "learning_rate": 2.604889924425965e-05, "loss": 0.1738, "step": 10703 }, { "epoch": 5.037176470588236, "grad_norm": 0.16473573434910338, "learning_rate": 2.6024217292954917e-05, "loss": 0.0459, "step": 10704 }, { "epoch": 5.03764705882353, "grad_norm": 0.5770364056413961, "learning_rate": 2.599954592979015e-05, "loss": 0.2622, "step": 10705 }, { "epoch": 5.038117647058823, "grad_norm": 0.2254849781813657, "learning_rate": 2.5974885156872345e-05, "loss": 0.0923, "step": 10706 }, { "epoch": 5.038588235294117, "grad_norm": 0.30735080845691937, "learning_rate": 2.5950234976307682e-05, "loss": 0.0829, "step": 10707 }, { "epoch": 5.039058823529412, "grad_norm": 0.3645409181356054, "learning_rate": 2.5925595390201466e-05, "loss": 0.1558, "step": 10708 }, { "epoch": 5.039529411764706, "grad_norm": 0.1870812751912668, "learning_rate": 2.5900966400657923e-05, "loss": 0.032, "step": 10709 }, { "epoch": 5.04, "grad_norm": 0.8913493981762124, "learning_rate": 2.5876348009780564e-05, "loss": 0.5191, "step": 10710 }, { "epoch": 5.040470588235294, "grad_norm": 0.5066627451964798, "learning_rate": 2.5851740219671945e-05, "loss": 0.224, "step": 10711 }, { "epoch": 5.0409411764705885, "grad_norm": 0.3986021790835113, "learning_rate": 2.5827143032433627e-05, "loss": 0.1578, "step": 10712 }, { "epoch": 5.041411764705883, "grad_norm": 1.4589847000214973, "learning_rate": 2.5802556450166396e-05, "loss": 0.3602, "step": 10713 }, { "epoch": 5.041882352941176, "grad_norm": 0.35450088922378903, "learning_rate": 2.5777980474969983e-05, "loss": 0.1428, "step": 10714 }, { "epoch": 5.04235294117647, "grad_norm": 0.721249169629243, "learning_rate": 2.5753415108943332e-05, "loss": 0.415, "step": 10715 }, { "epoch": 5.0428235294117645, "grad_norm": 0.42069586371617024, "learning_rate": 2.572886035418451e-05, "loss": 0.1186, "step": 10716 }, { "epoch": 5.043294117647059, "grad_norm": 0.42309562453586125, "learning_rate": 2.5704316212790527e-05, "loss": 0.127, "step": 10717 }, { "epoch": 5.043764705882353, "grad_norm": 0.15819864934873534, "learning_rate": 2.5679782686857586e-05, "loss": 0.022, "step": 10718 }, { "epoch": 5.044235294117647, "grad_norm": 0.3998373666078936, "learning_rate": 2.5655259778481045e-05, "loss": 0.2388, "step": 10719 }, { "epoch": 5.044705882352941, "grad_norm": 0.5475626765971959, "learning_rate": 2.5630747489755182e-05, "loss": 0.2721, "step": 10720 }, { "epoch": 5.045176470588236, "grad_norm": 0.7163695203902031, "learning_rate": 2.560624582277352e-05, "loss": 0.3031, "step": 10721 }, { "epoch": 5.04564705882353, "grad_norm": 0.21697913455323656, "learning_rate": 2.558175477962865e-05, "loss": 0.0496, "step": 10722 }, { "epoch": 5.046117647058823, "grad_norm": 0.9277168059940492, "learning_rate": 2.5557274362412132e-05, "loss": 0.2918, "step": 10723 }, { "epoch": 5.046588235294117, "grad_norm": 0.33581330293838146, "learning_rate": 2.55328045732148e-05, "loss": 0.1074, "step": 10724 }, { "epoch": 5.047058823529412, "grad_norm": 0.45721100021158784, "learning_rate": 2.5508345414126474e-05, "loss": 0.1574, "step": 10725 }, { "epoch": 5.047529411764706, "grad_norm": 0.6531622410590512, "learning_rate": 2.5483896887236056e-05, "loss": 0.1134, "step": 10726 }, { "epoch": 5.048, "grad_norm": 0.8404564191433607, "learning_rate": 2.5459458994631604e-05, "loss": 0.5518, "step": 10727 }, { "epoch": 5.048470588235294, "grad_norm": 0.11747083471141585, "learning_rate": 2.5435031738400234e-05, "loss": 0.0337, "step": 10728 }, { "epoch": 5.0489411764705885, "grad_norm": 0.49105538196147436, "learning_rate": 2.541061512062808e-05, "loss": 0.2333, "step": 10729 }, { "epoch": 5.049411764705883, "grad_norm": 0.21465261734671345, "learning_rate": 2.538620914340057e-05, "loss": 0.0449, "step": 10730 }, { "epoch": 5.049882352941176, "grad_norm": 0.40369328509245034, "learning_rate": 2.5361813808802007e-05, "loss": 0.2898, "step": 10731 }, { "epoch": 5.05035294117647, "grad_norm": 0.13616146112969932, "learning_rate": 2.533742911891587e-05, "loss": 0.0343, "step": 10732 }, { "epoch": 5.0508235294117645, "grad_norm": 0.28431331725662723, "learning_rate": 2.531305507582481e-05, "loss": 0.1017, "step": 10733 }, { "epoch": 5.051294117647059, "grad_norm": 0.335518595597084, "learning_rate": 2.5288691681610383e-05, "loss": 0.1955, "step": 10734 }, { "epoch": 5.051764705882353, "grad_norm": 0.16921100038552286, "learning_rate": 2.526433893835339e-05, "loss": 0.0343, "step": 10735 }, { "epoch": 5.052235294117647, "grad_norm": 0.31111873801923323, "learning_rate": 2.5239996848133715e-05, "loss": 0.1008, "step": 10736 }, { "epoch": 5.052705882352941, "grad_norm": 0.3474050209737729, "learning_rate": 2.5215665413030213e-05, "loss": 0.1058, "step": 10737 }, { "epoch": 5.053176470588236, "grad_norm": 0.3205821846867815, "learning_rate": 2.519134463512095e-05, "loss": 0.0904, "step": 10738 }, { "epoch": 5.05364705882353, "grad_norm": 0.4941197233366053, "learning_rate": 2.516703451648309e-05, "loss": 0.2131, "step": 10739 }, { "epoch": 5.054117647058823, "grad_norm": 0.6584945827931129, "learning_rate": 2.5142735059192737e-05, "loss": 0.4367, "step": 10740 }, { "epoch": 5.054588235294117, "grad_norm": 0.31692616374363475, "learning_rate": 2.511844626532526e-05, "loss": 0.1914, "step": 10741 }, { "epoch": 5.055058823529412, "grad_norm": 0.37565558768828067, "learning_rate": 2.5094168136954985e-05, "loss": 0.1606, "step": 10742 }, { "epoch": 5.055529411764706, "grad_norm": 0.37338564124898665, "learning_rate": 2.506990067615542e-05, "loss": 0.1571, "step": 10743 }, { "epoch": 5.056, "grad_norm": 0.505923505552797, "learning_rate": 2.504564388499914e-05, "loss": 0.2627, "step": 10744 }, { "epoch": 5.056470588235294, "grad_norm": 0.34682050753795857, "learning_rate": 2.502139776555772e-05, "loss": 0.0874, "step": 10745 }, { "epoch": 5.0569411764705885, "grad_norm": 0.47767419443537723, "learning_rate": 2.4997162319901977e-05, "loss": 0.2573, "step": 10746 }, { "epoch": 5.057411764705883, "grad_norm": 0.3635890041789764, "learning_rate": 2.497293755010173e-05, "loss": 0.1824, "step": 10747 }, { "epoch": 5.057882352941176, "grad_norm": 0.2279843196792608, "learning_rate": 2.4948723458225827e-05, "loss": 0.0547, "step": 10748 }, { "epoch": 5.05835294117647, "grad_norm": 0.22608066378687383, "learning_rate": 2.4924520046342317e-05, "loss": 0.0621, "step": 10749 }, { "epoch": 5.0588235294117645, "grad_norm": 0.3534394194017556, "learning_rate": 2.4900327316518326e-05, "loss": 0.1308, "step": 10750 }, { "epoch": 5.059294117647059, "grad_norm": 0.2548986505364129, "learning_rate": 2.4876145270819954e-05, "loss": 0.0673, "step": 10751 }, { "epoch": 5.059764705882353, "grad_norm": 0.20507738171189197, "learning_rate": 2.485197391131251e-05, "loss": 0.0577, "step": 10752 }, { "epoch": 5.060235294117647, "grad_norm": 0.237278978689838, "learning_rate": 2.4827813240060367e-05, "loss": 0.0645, "step": 10753 }, { "epoch": 5.060705882352941, "grad_norm": 0.49357716090870046, "learning_rate": 2.480366325912692e-05, "loss": 0.2144, "step": 10754 }, { "epoch": 5.061176470588236, "grad_norm": 0.3421780598455031, "learning_rate": 2.4779523970574703e-05, "loss": 0.1723, "step": 10755 }, { "epoch": 5.06164705882353, "grad_norm": 0.43330473866798846, "learning_rate": 2.4755395376465386e-05, "loss": 0.1841, "step": 10756 }, { "epoch": 5.062117647058823, "grad_norm": 0.35036022082620366, "learning_rate": 2.4731277478859586e-05, "loss": 0.2178, "step": 10757 }, { "epoch": 5.062588235294117, "grad_norm": 0.3169983367423155, "learning_rate": 2.4707170279817118e-05, "loss": 0.1438, "step": 10758 }, { "epoch": 5.063058823529412, "grad_norm": 0.35098574994066345, "learning_rate": 2.4683073781396924e-05, "loss": 0.1445, "step": 10759 }, { "epoch": 5.063529411764706, "grad_norm": 0.2871305793679576, "learning_rate": 2.465898798565685e-05, "loss": 0.1146, "step": 10760 }, { "epoch": 5.064, "grad_norm": 0.41536923759461747, "learning_rate": 2.4634912894653996e-05, "loss": 0.1615, "step": 10761 }, { "epoch": 5.064470588235294, "grad_norm": 0.20256682302727166, "learning_rate": 2.461084851044454e-05, "loss": 0.0498, "step": 10762 }, { "epoch": 5.0649411764705885, "grad_norm": 0.5553525811195213, "learning_rate": 2.4586794835083613e-05, "loss": 0.2549, "step": 10763 }, { "epoch": 5.065411764705883, "grad_norm": 0.19688010487485205, "learning_rate": 2.4562751870625597e-05, "loss": 0.0458, "step": 10764 }, { "epoch": 5.065882352941176, "grad_norm": 0.41536162415623246, "learning_rate": 2.4538719619123786e-05, "loss": 0.2329, "step": 10765 }, { "epoch": 5.06635294117647, "grad_norm": 0.5532424690753937, "learning_rate": 2.4514698082630706e-05, "loss": 0.3754, "step": 10766 }, { "epoch": 5.0668235294117645, "grad_norm": 0.3833652670237656, "learning_rate": 2.449068726319795e-05, "loss": 0.0937, "step": 10767 }, { "epoch": 5.067294117647059, "grad_norm": 0.42377923409346885, "learning_rate": 2.4466687162876093e-05, "loss": 0.1658, "step": 10768 }, { "epoch": 5.067764705882353, "grad_norm": 0.3070566477099989, "learning_rate": 2.4442697783714894e-05, "loss": 0.1252, "step": 10769 }, { "epoch": 5.068235294117647, "grad_norm": 0.24350615891410937, "learning_rate": 2.4418719127763183e-05, "loss": 0.0801, "step": 10770 }, { "epoch": 5.068705882352941, "grad_norm": 0.4326973215521739, "learning_rate": 2.43947511970688e-05, "loss": 0.1254, "step": 10771 }, { "epoch": 5.069176470588236, "grad_norm": 0.2029142981132546, "learning_rate": 2.437079399367875e-05, "loss": 0.0743, "step": 10772 }, { "epoch": 5.06964705882353, "grad_norm": 0.4247157675548701, "learning_rate": 2.434684751963914e-05, "loss": 0.2466, "step": 10773 }, { "epoch": 5.070117647058823, "grad_norm": 0.5745795496522371, "learning_rate": 2.4322911776995002e-05, "loss": 0.3104, "step": 10774 }, { "epoch": 5.070588235294117, "grad_norm": 0.34402536822453716, "learning_rate": 2.429898676779072e-05, "loss": 0.0796, "step": 10775 }, { "epoch": 5.071058823529412, "grad_norm": 0.367113851273492, "learning_rate": 2.4275072494069487e-05, "loss": 0.1534, "step": 10776 }, { "epoch": 5.071529411764706, "grad_norm": 0.7202817700993763, "learning_rate": 2.425116895787374e-05, "loss": 0.371, "step": 10777 }, { "epoch": 5.072, "grad_norm": 0.41654561016324304, "learning_rate": 2.4227276161244995e-05, "loss": 0.2158, "step": 10778 }, { "epoch": 5.072470588235294, "grad_norm": 0.3857542096987638, "learning_rate": 2.4203394106223755e-05, "loss": 0.1506, "step": 10779 }, { "epoch": 5.0729411764705885, "grad_norm": 0.40613821611002027, "learning_rate": 2.417952279484967e-05, "loss": 0.2159, "step": 10780 }, { "epoch": 5.073411764705883, "grad_norm": 0.37056031814328644, "learning_rate": 2.4155662229161527e-05, "loss": 0.143, "step": 10781 }, { "epoch": 5.073882352941176, "grad_norm": 0.6545945841869323, "learning_rate": 2.4131812411197045e-05, "loss": 0.2805, "step": 10782 }, { "epoch": 5.07435294117647, "grad_norm": 0.4618048016623052, "learning_rate": 2.4107973342993174e-05, "loss": 0.2731, "step": 10783 }, { "epoch": 5.0748235294117645, "grad_norm": 0.5218731826770411, "learning_rate": 2.408414502658591e-05, "loss": 0.2116, "step": 10784 }, { "epoch": 5.075294117647059, "grad_norm": 0.5648820398166202, "learning_rate": 2.4060327464010254e-05, "loss": 0.2879, "step": 10785 }, { "epoch": 5.075764705882353, "grad_norm": 0.3003048656702646, "learning_rate": 2.4036520657300346e-05, "loss": 0.0611, "step": 10786 }, { "epoch": 5.076235294117647, "grad_norm": 0.14402117448582694, "learning_rate": 2.401272460848946e-05, "loss": 0.0352, "step": 10787 }, { "epoch": 5.076705882352941, "grad_norm": 0.24777501709171543, "learning_rate": 2.3988939319609817e-05, "loss": 0.0543, "step": 10788 }, { "epoch": 5.077176470588236, "grad_norm": 0.43731609395912385, "learning_rate": 2.3965164792692833e-05, "loss": 0.2296, "step": 10789 }, { "epoch": 5.07764705882353, "grad_norm": 0.28881859730048737, "learning_rate": 2.3941401029769024e-05, "loss": 0.0552, "step": 10790 }, { "epoch": 5.078117647058823, "grad_norm": 0.4841638935670621, "learning_rate": 2.3917648032867844e-05, "loss": 0.1883, "step": 10791 }, { "epoch": 5.078588235294117, "grad_norm": 0.15345242853761623, "learning_rate": 2.3893905804017977e-05, "loss": 0.0421, "step": 10792 }, { "epoch": 5.079058823529412, "grad_norm": 0.43243055762345917, "learning_rate": 2.387017434524707e-05, "loss": 0.1633, "step": 10793 }, { "epoch": 5.079529411764706, "grad_norm": 0.4112679728557049, "learning_rate": 2.3846453658581927e-05, "loss": 0.1119, "step": 10794 }, { "epoch": 5.08, "grad_norm": 0.33914120044304036, "learning_rate": 2.382274374604847e-05, "loss": 0.1543, "step": 10795 }, { "epoch": 5.080470588235294, "grad_norm": 0.446470924604425, "learning_rate": 2.3799044609671558e-05, "loss": 0.2548, "step": 10796 }, { "epoch": 5.0809411764705885, "grad_norm": 0.25089243321611615, "learning_rate": 2.3775356251475248e-05, "loss": 0.0714, "step": 10797 }, { "epoch": 5.081411764705883, "grad_norm": 0.28102157164315816, "learning_rate": 2.375167867348268e-05, "loss": 0.1084, "step": 10798 }, { "epoch": 5.081882352941176, "grad_norm": 0.6100950271726517, "learning_rate": 2.3728011877715947e-05, "loss": 0.2961, "step": 10799 }, { "epoch": 5.08235294117647, "grad_norm": 0.48529997779846956, "learning_rate": 2.3704355866196373e-05, "loss": 0.3329, "step": 10800 }, { "epoch": 5.0828235294117645, "grad_norm": 0.436557814983425, "learning_rate": 2.3680710640944328e-05, "loss": 0.1269, "step": 10801 }, { "epoch": 5.083294117647059, "grad_norm": 0.5496203625439374, "learning_rate": 2.3657076203979132e-05, "loss": 0.2951, "step": 10802 }, { "epoch": 5.083764705882353, "grad_norm": 0.406747671870648, "learning_rate": 2.363345255731935e-05, "loss": 0.2936, "step": 10803 }, { "epoch": 5.084235294117647, "grad_norm": 0.26512224045059185, "learning_rate": 2.360983970298257e-05, "loss": 0.1389, "step": 10804 }, { "epoch": 5.084705882352941, "grad_norm": 0.3643477358916192, "learning_rate": 2.3586237642985396e-05, "loss": 0.105, "step": 10805 }, { "epoch": 5.085176470588236, "grad_norm": 0.3563494112893341, "learning_rate": 2.356264637934358e-05, "loss": 0.183, "step": 10806 }, { "epoch": 5.085647058823529, "grad_norm": 0.29397256319523846, "learning_rate": 2.3539065914071963e-05, "loss": 0.1302, "step": 10807 }, { "epoch": 5.086117647058823, "grad_norm": 0.6636819354294596, "learning_rate": 2.351549624918437e-05, "loss": 0.4539, "step": 10808 }, { "epoch": 5.086588235294117, "grad_norm": 0.5175849878738912, "learning_rate": 2.34919373866938e-05, "loss": 0.1263, "step": 10809 }, { "epoch": 5.087058823529412, "grad_norm": 0.5043116238588152, "learning_rate": 2.3468389328612348e-05, "loss": 0.2572, "step": 10810 }, { "epoch": 5.087529411764706, "grad_norm": 0.37866494035262244, "learning_rate": 2.344485207695102e-05, "loss": 0.2087, "step": 10811 }, { "epoch": 5.088, "grad_norm": 0.5458790330749375, "learning_rate": 2.3421325633720094e-05, "loss": 0.3651, "step": 10812 }, { "epoch": 5.088470588235294, "grad_norm": 0.32867844260541507, "learning_rate": 2.3397810000928845e-05, "loss": 0.0914, "step": 10813 }, { "epoch": 5.0889411764705885, "grad_norm": 0.4003491128686473, "learning_rate": 2.337430518058557e-05, "loss": 0.2109, "step": 10814 }, { "epoch": 5.089411764705883, "grad_norm": 0.4288344813806051, "learning_rate": 2.335081117469777e-05, "loss": 0.0766, "step": 10815 }, { "epoch": 5.089882352941176, "grad_norm": 0.48902967666789615, "learning_rate": 2.3327327985271865e-05, "loss": 0.3139, "step": 10816 }, { "epoch": 5.09035294117647, "grad_norm": 0.484543050467604, "learning_rate": 2.3303855614313483e-05, "loss": 0.2378, "step": 10817 }, { "epoch": 5.0908235294117645, "grad_norm": 0.3407863319314678, "learning_rate": 2.328039406382727e-05, "loss": 0.1229, "step": 10818 }, { "epoch": 5.091294117647059, "grad_norm": 0.2934338981910716, "learning_rate": 2.325694333581695e-05, "loss": 0.1162, "step": 10819 }, { "epoch": 5.091764705882353, "grad_norm": 0.29007419066966406, "learning_rate": 2.32335034322854e-05, "loss": 0.1121, "step": 10820 }, { "epoch": 5.092235294117647, "grad_norm": 0.48051754035645144, "learning_rate": 2.3210074355234392e-05, "loss": 0.3323, "step": 10821 }, { "epoch": 5.092705882352941, "grad_norm": 0.27984167105315466, "learning_rate": 2.318665610666493e-05, "loss": 0.0557, "step": 10822 }, { "epoch": 5.093176470588236, "grad_norm": 0.4820001605350041, "learning_rate": 2.3163248688577098e-05, "loss": 0.2428, "step": 10823 }, { "epoch": 5.09364705882353, "grad_norm": 0.46952110913156003, "learning_rate": 2.3139852102969906e-05, "loss": 0.2295, "step": 10824 }, { "epoch": 5.094117647058823, "grad_norm": 0.24418217957647498, "learning_rate": 2.31164663518416e-05, "loss": 0.0973, "step": 10825 }, { "epoch": 5.094588235294117, "grad_norm": 0.900511633870272, "learning_rate": 2.3093091437189476e-05, "loss": 0.2312, "step": 10826 }, { "epoch": 5.095058823529412, "grad_norm": 0.4590500043214801, "learning_rate": 2.306972736100977e-05, "loss": 0.281, "step": 10827 }, { "epoch": 5.095529411764706, "grad_norm": 0.3359528282751316, "learning_rate": 2.304637412529794e-05, "loss": 0.1357, "step": 10828 }, { "epoch": 5.096, "grad_norm": 0.30232976989192556, "learning_rate": 2.3023031732048504e-05, "loss": 0.1277, "step": 10829 }, { "epoch": 5.096470588235294, "grad_norm": 0.3299295436282832, "learning_rate": 2.2999700183254934e-05, "loss": 0.1414, "step": 10830 }, { "epoch": 5.0969411764705885, "grad_norm": 0.5091913559456654, "learning_rate": 2.29763794809099e-05, "loss": 0.1827, "step": 10831 }, { "epoch": 5.097411764705883, "grad_norm": 0.4243907000803771, "learning_rate": 2.2953069627005132e-05, "loss": 0.2451, "step": 10832 }, { "epoch": 5.097882352941176, "grad_norm": 0.485688606748997, "learning_rate": 2.2929770623531368e-05, "loss": 0.304, "step": 10833 }, { "epoch": 5.09835294117647, "grad_norm": 0.28854779519470436, "learning_rate": 2.290648247247845e-05, "loss": 0.1028, "step": 10834 }, { "epoch": 5.0988235294117645, "grad_norm": 0.5200570782916875, "learning_rate": 2.2883205175835368e-05, "loss": 0.2825, "step": 10835 }, { "epoch": 5.099294117647059, "grad_norm": 0.4132779319547243, "learning_rate": 2.2859938735590017e-05, "loss": 0.2076, "step": 10836 }, { "epoch": 5.099764705882353, "grad_norm": 0.5221923500061655, "learning_rate": 2.2836683153729522e-05, "loss": 0.1944, "step": 10837 }, { "epoch": 5.100235294117647, "grad_norm": 0.5786426001063273, "learning_rate": 2.2813438432240067e-05, "loss": 0.3408, "step": 10838 }, { "epoch": 5.100705882352941, "grad_norm": 0.6008660802087399, "learning_rate": 2.2790204573106756e-05, "loss": 0.4022, "step": 10839 }, { "epoch": 5.101176470588236, "grad_norm": 0.336032680506036, "learning_rate": 2.276698157831396e-05, "loss": 0.116, "step": 10840 }, { "epoch": 5.101647058823529, "grad_norm": 0.1913437190284976, "learning_rate": 2.2743769449845035e-05, "loss": 0.0325, "step": 10841 }, { "epoch": 5.102117647058823, "grad_norm": 0.5855331071043346, "learning_rate": 2.272056818968235e-05, "loss": 0.2695, "step": 10842 }, { "epoch": 5.102588235294117, "grad_norm": 0.12652362606048995, "learning_rate": 2.269737779980748e-05, "loss": 0.0218, "step": 10843 }, { "epoch": 5.103058823529412, "grad_norm": 0.48669255030437203, "learning_rate": 2.2674198282200934e-05, "loss": 0.2338, "step": 10844 }, { "epoch": 5.103529411764706, "grad_norm": 0.34810525370287615, "learning_rate": 2.2651029638842372e-05, "loss": 0.2105, "step": 10845 }, { "epoch": 5.104, "grad_norm": 0.30188546456589643, "learning_rate": 2.262787187171058e-05, "loss": 0.1131, "step": 10846 }, { "epoch": 5.104470588235294, "grad_norm": 0.0001981143043630161, "learning_rate": 2.2604724982783224e-05, "loss": 0.0, "step": 10847 }, { "epoch": 5.1049411764705885, "grad_norm": 0.4349905327422216, "learning_rate": 2.258158897403724e-05, "loss": 0.1253, "step": 10848 }, { "epoch": 5.105411764705883, "grad_norm": 0.44672287152423173, "learning_rate": 2.2558463847448566e-05, "loss": 0.1937, "step": 10849 }, { "epoch": 5.105882352941176, "grad_norm": 0.5874017256675088, "learning_rate": 2.2535349604992153e-05, "loss": 0.2549, "step": 10850 }, { "epoch": 5.10635294117647, "grad_norm": 0.7407575726444579, "learning_rate": 2.251224624864209e-05, "loss": 0.4275, "step": 10851 }, { "epoch": 5.1068235294117645, "grad_norm": 0.37168391015726243, "learning_rate": 2.2489153780371566e-05, "loss": 0.1841, "step": 10852 }, { "epoch": 5.107294117647059, "grad_norm": 0.9978943049825099, "learning_rate": 2.2466072202152702e-05, "loss": 0.3862, "step": 10853 }, { "epoch": 5.107764705882353, "grad_norm": 0.3016516714809791, "learning_rate": 2.2443001515956823e-05, "loss": 0.1053, "step": 10854 }, { "epoch": 5.108235294117647, "grad_norm": 0.20205659086129593, "learning_rate": 2.24199417237543e-05, "loss": 0.0383, "step": 10855 }, { "epoch": 5.108705882352941, "grad_norm": 0.4372927171421085, "learning_rate": 2.2396892827514507e-05, "loss": 0.295, "step": 10856 }, { "epoch": 5.109176470588236, "grad_norm": 0.2669108303067813, "learning_rate": 2.2373854829205957e-05, "loss": 0.0873, "step": 10857 }, { "epoch": 5.109647058823529, "grad_norm": 0.36580877589725996, "learning_rate": 2.235082773079624e-05, "loss": 0.1869, "step": 10858 }, { "epoch": 5.110117647058823, "grad_norm": 0.3423427577481525, "learning_rate": 2.2327811534251933e-05, "loss": 0.1318, "step": 10859 }, { "epoch": 5.110588235294117, "grad_norm": 0.41061529417540304, "learning_rate": 2.230480624153873e-05, "loss": 0.2497, "step": 10860 }, { "epoch": 5.111058823529412, "grad_norm": 0.4444241646390166, "learning_rate": 2.2281811854621445e-05, "loss": 0.1898, "step": 10861 }, { "epoch": 5.111529411764706, "grad_norm": 0.46597111423038784, "learning_rate": 2.2258828375463855e-05, "loss": 0.2898, "step": 10862 }, { "epoch": 5.112, "grad_norm": 0.4620438703147822, "learning_rate": 2.223585580602889e-05, "loss": 0.3045, "step": 10863 }, { "epoch": 5.112470588235294, "grad_norm": 0.22832247542525488, "learning_rate": 2.221289414827853e-05, "loss": 0.0911, "step": 10864 }, { "epoch": 5.1129411764705885, "grad_norm": 0.47629882643857885, "learning_rate": 2.218994340417378e-05, "loss": 0.2812, "step": 10865 }, { "epoch": 5.113411764705883, "grad_norm": 0.3871003065379155, "learning_rate": 2.2167003575674824e-05, "loss": 0.1979, "step": 10866 }, { "epoch": 5.113882352941176, "grad_norm": 0.34618500142273073, "learning_rate": 2.2144074664740723e-05, "loss": 0.1658, "step": 10867 }, { "epoch": 5.11435294117647, "grad_norm": 0.3240485437599046, "learning_rate": 2.2121156673329794e-05, "loss": 0.0977, "step": 10868 }, { "epoch": 5.1148235294117645, "grad_norm": 0.40317871690425156, "learning_rate": 2.2098249603399348e-05, "loss": 0.2502, "step": 10869 }, { "epoch": 5.115294117647059, "grad_norm": 0.5139250913006187, "learning_rate": 2.2075353456905727e-05, "loss": 0.1586, "step": 10870 }, { "epoch": 5.115764705882353, "grad_norm": 0.5882614912336055, "learning_rate": 2.2052468235804393e-05, "loss": 0.1653, "step": 10871 }, { "epoch": 5.116235294117647, "grad_norm": 0.39202832530254805, "learning_rate": 2.2029593942049838e-05, "loss": 0.1455, "step": 10872 }, { "epoch": 5.116705882352941, "grad_norm": 0.7484758323331385, "learning_rate": 2.200673057759565e-05, "loss": 0.5079, "step": 10873 }, { "epoch": 5.117176470588236, "grad_norm": 0.4520639995425334, "learning_rate": 2.1983878144394497e-05, "loss": 0.3204, "step": 10874 }, { "epoch": 5.117647058823529, "grad_norm": 0.2432429493267954, "learning_rate": 2.1961036644398035e-05, "loss": 0.0828, "step": 10875 }, { "epoch": 5.118117647058823, "grad_norm": 0.21464988058036324, "learning_rate": 2.193820607955707e-05, "loss": 0.0544, "step": 10876 }, { "epoch": 5.118588235294117, "grad_norm": 0.38662682866260895, "learning_rate": 2.1915386451821494e-05, "loss": 0.0999, "step": 10877 }, { "epoch": 5.119058823529412, "grad_norm": 0.4468150406007739, "learning_rate": 2.1892577763140112e-05, "loss": 0.285, "step": 10878 }, { "epoch": 5.119529411764706, "grad_norm": 0.35764851391044966, "learning_rate": 2.186978001546097e-05, "loss": 0.1153, "step": 10879 }, { "epoch": 5.12, "grad_norm": 0.6959660490030091, "learning_rate": 2.1846993210731108e-05, "loss": 0.3691, "step": 10880 }, { "epoch": 5.120470588235294, "grad_norm": 0.3474402625616565, "learning_rate": 2.1824217350896585e-05, "loss": 0.2011, "step": 10881 }, { "epoch": 5.1209411764705886, "grad_norm": 0.5980972064569662, "learning_rate": 2.180145243790261e-05, "loss": 0.3096, "step": 10882 }, { "epoch": 5.121411764705883, "grad_norm": 0.4184952629126197, "learning_rate": 2.1778698473693435e-05, "loss": 0.205, "step": 10883 }, { "epoch": 5.121882352941176, "grad_norm": 0.3814852943558357, "learning_rate": 2.1755955460212282e-05, "loss": 0.1339, "step": 10884 }, { "epoch": 5.12235294117647, "grad_norm": 0.702130982713795, "learning_rate": 2.1733223399401585e-05, "loss": 0.3493, "step": 10885 }, { "epoch": 5.1228235294117646, "grad_norm": 0.326468683447361, "learning_rate": 2.1710502293202786e-05, "loss": 0.0925, "step": 10886 }, { "epoch": 5.123294117647059, "grad_norm": 0.2459875277092059, "learning_rate": 2.1687792143556303e-05, "loss": 0.065, "step": 10887 }, { "epoch": 5.123764705882353, "grad_norm": 0.4952277508760737, "learning_rate": 2.1665092952401754e-05, "loss": 0.2519, "step": 10888 }, { "epoch": 5.124235294117647, "grad_norm": 0.27370690147342047, "learning_rate": 2.1642404721677765e-05, "loss": 0.0811, "step": 10889 }, { "epoch": 5.124705882352941, "grad_norm": 0.5153858113183083, "learning_rate": 2.1619727453321974e-05, "loss": 0.3477, "step": 10890 }, { "epoch": 5.125176470588236, "grad_norm": 0.46542525439887583, "learning_rate": 2.159706114927116e-05, "loss": 0.3371, "step": 10891 }, { "epoch": 5.125647058823529, "grad_norm": 0.34898074129516915, "learning_rate": 2.157440581146118e-05, "loss": 0.1402, "step": 10892 }, { "epoch": 5.126117647058823, "grad_norm": 0.2669848310547078, "learning_rate": 2.155176144182682e-05, "loss": 0.1171, "step": 10893 }, { "epoch": 5.1265882352941174, "grad_norm": 1.0100758492477129, "learning_rate": 2.1529128042302113e-05, "loss": 0.2345, "step": 10894 }, { "epoch": 5.127058823529412, "grad_norm": 0.5564522595938676, "learning_rate": 2.150650561481998e-05, "loss": 0.2195, "step": 10895 }, { "epoch": 5.127529411764706, "grad_norm": 0.4676122788257326, "learning_rate": 2.1483894161312525e-05, "loss": 0.2153, "step": 10896 }, { "epoch": 5.128, "grad_norm": 0.43345085588283455, "learning_rate": 2.1461293683710905e-05, "loss": 0.2619, "step": 10897 }, { "epoch": 5.128470588235294, "grad_norm": 0.40571581755551844, "learning_rate": 2.143870418394526e-05, "loss": 0.1295, "step": 10898 }, { "epoch": 5.128941176470589, "grad_norm": 0.6547640768901939, "learning_rate": 2.141612566394486e-05, "loss": 0.4117, "step": 10899 }, { "epoch": 5.129411764705883, "grad_norm": 0.2963517238950402, "learning_rate": 2.1393558125638066e-05, "loss": 0.1617, "step": 10900 }, { "epoch": 5.129882352941176, "grad_norm": 0.8363931958479768, "learning_rate": 2.1371001570952186e-05, "loss": 0.3448, "step": 10901 }, { "epoch": 5.13035294117647, "grad_norm": 0.1591338028859555, "learning_rate": 2.1348456001813674e-05, "loss": 0.0427, "step": 10902 }, { "epoch": 5.130823529411765, "grad_norm": 0.5327908641109619, "learning_rate": 2.132592142014811e-05, "loss": 0.2333, "step": 10903 }, { "epoch": 5.131294117647059, "grad_norm": 0.6006167117629468, "learning_rate": 2.130339782787996e-05, "loss": 0.4243, "step": 10904 }, { "epoch": 5.131764705882353, "grad_norm": 0.28032196122093384, "learning_rate": 2.1280885226932877e-05, "loss": 0.1249, "step": 10905 }, { "epoch": 5.132235294117647, "grad_norm": 0.2981605984976159, "learning_rate": 2.125838361922961e-05, "loss": 0.0972, "step": 10906 }, { "epoch": 5.1327058823529415, "grad_norm": 0.3495433277497778, "learning_rate": 2.1235893006691786e-05, "loss": 0.1227, "step": 10907 }, { "epoch": 5.133176470588236, "grad_norm": 0.5703361086931238, "learning_rate": 2.1213413391240335e-05, "loss": 0.4316, "step": 10908 }, { "epoch": 5.133647058823529, "grad_norm": 0.7152516932108183, "learning_rate": 2.119094477479506e-05, "loss": 0.3345, "step": 10909 }, { "epoch": 5.134117647058823, "grad_norm": 0.5901641896969968, "learning_rate": 2.116848715927489e-05, "loss": 0.1588, "step": 10910 }, { "epoch": 5.1345882352941175, "grad_norm": 0.26816422174752474, "learning_rate": 2.114604054659787e-05, "loss": 0.0916, "step": 10911 }, { "epoch": 5.135058823529412, "grad_norm": 0.8133520743376906, "learning_rate": 2.1123604938680987e-05, "loss": 0.4254, "step": 10912 }, { "epoch": 5.135529411764706, "grad_norm": 0.5572536963464985, "learning_rate": 2.1101180337440378e-05, "loss": 0.3183, "step": 10913 }, { "epoch": 5.136, "grad_norm": 0.5020960249264058, "learning_rate": 2.1078766744791238e-05, "loss": 0.2388, "step": 10914 }, { "epoch": 5.136470588235294, "grad_norm": 0.19085550396036768, "learning_rate": 2.1056364162647736e-05, "loss": 0.0357, "step": 10915 }, { "epoch": 5.136941176470589, "grad_norm": 0.3094402776805618, "learning_rate": 2.1033972592923214e-05, "loss": 0.1163, "step": 10916 }, { "epoch": 5.137411764705883, "grad_norm": 0.3902552510038607, "learning_rate": 2.1011592037530023e-05, "loss": 0.1958, "step": 10917 }, { "epoch": 5.137882352941176, "grad_norm": 0.37359822594202147, "learning_rate": 2.098922249837953e-05, "loss": 0.0994, "step": 10918 }, { "epoch": 5.13835294117647, "grad_norm": 0.37783919888745365, "learning_rate": 2.096686397738223e-05, "loss": 0.1881, "step": 10919 }, { "epoch": 5.138823529411765, "grad_norm": 0.4634418267785735, "learning_rate": 2.094451647644769e-05, "loss": 0.3456, "step": 10920 }, { "epoch": 5.139294117647059, "grad_norm": 0.4424704188050999, "learning_rate": 2.0922179997484424e-05, "loss": 0.2626, "step": 10921 }, { "epoch": 5.139764705882353, "grad_norm": 0.3993056974336225, "learning_rate": 2.0899854542400147e-05, "loss": 0.1276, "step": 10922 }, { "epoch": 5.140235294117647, "grad_norm": 0.5706407448289746, "learning_rate": 2.087754011310147e-05, "loss": 0.1362, "step": 10923 }, { "epoch": 5.1407058823529415, "grad_norm": 0.5641181775257722, "learning_rate": 2.0855236711494223e-05, "loss": 0.4272, "step": 10924 }, { "epoch": 5.141176470588236, "grad_norm": 0.5883415436561625, "learning_rate": 2.083294433948324e-05, "loss": 0.1157, "step": 10925 }, { "epoch": 5.141647058823529, "grad_norm": 0.6920247172064151, "learning_rate": 2.0810662998972345e-05, "loss": 0.227, "step": 10926 }, { "epoch": 5.142117647058823, "grad_norm": 0.4032843798643679, "learning_rate": 2.0788392691864504e-05, "loss": 0.1184, "step": 10927 }, { "epoch": 5.1425882352941175, "grad_norm": 0.3408286379063278, "learning_rate": 2.0766133420061732e-05, "loss": 0.1212, "step": 10928 }, { "epoch": 5.143058823529412, "grad_norm": 0.33754676256062993, "learning_rate": 2.0743885185465025e-05, "loss": 0.0798, "step": 10929 }, { "epoch": 5.143529411764706, "grad_norm": 0.254834032933083, "learning_rate": 2.0721647989974514e-05, "loss": 0.1372, "step": 10930 }, { "epoch": 5.144, "grad_norm": 0.5107902568262845, "learning_rate": 2.0699421835489417e-05, "loss": 0.2541, "step": 10931 }, { "epoch": 5.144470588235294, "grad_norm": 0.30771803234091083, "learning_rate": 2.0677206723907857e-05, "loss": 0.0945, "step": 10932 }, { "epoch": 5.144941176470589, "grad_norm": 0.37127431669056443, "learning_rate": 2.0655002657127188e-05, "loss": 0.1651, "step": 10933 }, { "epoch": 5.145411764705882, "grad_norm": 0.2946167266452303, "learning_rate": 2.0632809637043747e-05, "loss": 0.1422, "step": 10934 }, { "epoch": 5.145882352941176, "grad_norm": 0.582489462296417, "learning_rate": 2.0610627665552875e-05, "loss": 0.2613, "step": 10935 }, { "epoch": 5.14635294117647, "grad_norm": 0.588255059257769, "learning_rate": 2.058845674454905e-05, "loss": 0.3596, "step": 10936 }, { "epoch": 5.146823529411765, "grad_norm": 0.3456015339708058, "learning_rate": 2.0566296875925813e-05, "loss": 0.1133, "step": 10937 }, { "epoch": 5.147294117647059, "grad_norm": 0.31586663034438817, "learning_rate": 2.054414806157566e-05, "loss": 0.1494, "step": 10938 }, { "epoch": 5.147764705882353, "grad_norm": 0.5172796763913587, "learning_rate": 2.052201030339024e-05, "loss": 0.1634, "step": 10939 }, { "epoch": 5.148235294117647, "grad_norm": 0.6409304481228508, "learning_rate": 2.0499883603260274e-05, "loss": 0.3394, "step": 10940 }, { "epoch": 5.1487058823529415, "grad_norm": 0.14978768866084297, "learning_rate": 2.0477767963075386e-05, "loss": 0.0238, "step": 10941 }, { "epoch": 5.149176470588236, "grad_norm": 0.6507569239434542, "learning_rate": 2.045566338472443e-05, "loss": 0.4646, "step": 10942 }, { "epoch": 5.149647058823529, "grad_norm": 0.687781829131408, "learning_rate": 2.043356987009526e-05, "loss": 0.2566, "step": 10943 }, { "epoch": 5.150117647058823, "grad_norm": 0.41965720176120563, "learning_rate": 2.041148742107471e-05, "loss": 0.1973, "step": 10944 }, { "epoch": 5.1505882352941175, "grad_norm": 0.44037185604375617, "learning_rate": 2.038941603954879e-05, "loss": 0.1051, "step": 10945 }, { "epoch": 5.151058823529412, "grad_norm": 0.39083307224126485, "learning_rate": 2.036735572740245e-05, "loss": 0.1711, "step": 10946 }, { "epoch": 5.151529411764706, "grad_norm": 0.6131655588838838, "learning_rate": 2.0345306486519787e-05, "loss": 0.3763, "step": 10947 }, { "epoch": 5.152, "grad_norm": 0.31594404289387046, "learning_rate": 2.032326831878392e-05, "loss": 0.111, "step": 10948 }, { "epoch": 5.152470588235294, "grad_norm": 0.5883481343563565, "learning_rate": 2.0301241226076963e-05, "loss": 0.2602, "step": 10949 }, { "epoch": 5.152941176470589, "grad_norm": 0.46719479991432633, "learning_rate": 2.027922521028018e-05, "loss": 0.1092, "step": 10950 }, { "epoch": 5.153411764705883, "grad_norm": 0.49524355169741086, "learning_rate": 2.025722027327385e-05, "loss": 0.1673, "step": 10951 }, { "epoch": 5.153882352941176, "grad_norm": 0.3599335784998269, "learning_rate": 2.0235226416937284e-05, "loss": 0.1723, "step": 10952 }, { "epoch": 5.15435294117647, "grad_norm": 0.716915163391799, "learning_rate": 2.0213243643148914e-05, "loss": 0.3881, "step": 10953 }, { "epoch": 5.154823529411765, "grad_norm": 0.41955888752975834, "learning_rate": 2.0191271953786104e-05, "loss": 0.3103, "step": 10954 }, { "epoch": 5.155294117647059, "grad_norm": 0.34847820554223485, "learning_rate": 2.016931135072537e-05, "loss": 0.148, "step": 10955 }, { "epoch": 5.155764705882353, "grad_norm": 0.3112099130864145, "learning_rate": 2.0147361835842296e-05, "loss": 0.1045, "step": 10956 }, { "epoch": 5.156235294117647, "grad_norm": 0.5428090849379454, "learning_rate": 2.0125423411011416e-05, "loss": 0.3274, "step": 10957 }, { "epoch": 5.1567058823529415, "grad_norm": 0.31411136185303445, "learning_rate": 2.0103496078106396e-05, "loss": 0.073, "step": 10958 }, { "epoch": 5.157176470588236, "grad_norm": 0.4449811766805127, "learning_rate": 2.0081579839000005e-05, "loss": 0.2764, "step": 10959 }, { "epoch": 5.157647058823529, "grad_norm": 0.2961295717934059, "learning_rate": 2.00596746955639e-05, "loss": 0.0787, "step": 10960 }, { "epoch": 5.158117647058823, "grad_norm": 0.1684642194845119, "learning_rate": 2.0037780649668912e-05, "loss": 0.0527, "step": 10961 }, { "epoch": 5.1585882352941175, "grad_norm": 0.30690066238646785, "learning_rate": 2.001589770318498e-05, "loss": 0.1306, "step": 10962 }, { "epoch": 5.159058823529412, "grad_norm": 0.34862220489500995, "learning_rate": 1.9994025857980896e-05, "loss": 0.1718, "step": 10963 }, { "epoch": 5.159529411764706, "grad_norm": 0.5541054942335709, "learning_rate": 1.9972165115924695e-05, "loss": 0.2815, "step": 10964 }, { "epoch": 5.16, "grad_norm": 0.47925993409706247, "learning_rate": 1.99503154788834e-05, "loss": 0.2482, "step": 10965 }, { "epoch": 5.160470588235294, "grad_norm": 0.4018628401870125, "learning_rate": 1.9928476948723022e-05, "loss": 0.2215, "step": 10966 }, { "epoch": 5.160941176470589, "grad_norm": 0.43310921131300395, "learning_rate": 1.99066495273087e-05, "loss": 0.2446, "step": 10967 }, { "epoch": 5.161411764705882, "grad_norm": 0.6610432969979333, "learning_rate": 1.9884833216504657e-05, "loss": 0.2409, "step": 10968 }, { "epoch": 5.161882352941176, "grad_norm": 0.4173550803458064, "learning_rate": 1.9863028018174043e-05, "loss": 0.1861, "step": 10969 }, { "epoch": 5.16235294117647, "grad_norm": 0.40899218658650427, "learning_rate": 1.9841233934179145e-05, "loss": 0.2722, "step": 10970 }, { "epoch": 5.162823529411765, "grad_norm": 0.5640615854974405, "learning_rate": 1.9819450966381328e-05, "loss": 0.2797, "step": 10971 }, { "epoch": 5.163294117647059, "grad_norm": 0.30581492909361596, "learning_rate": 1.9797679116640903e-05, "loss": 0.1698, "step": 10972 }, { "epoch": 5.163764705882353, "grad_norm": 0.8920595771002392, "learning_rate": 1.9775918386817346e-05, "loss": 0.496, "step": 10973 }, { "epoch": 5.164235294117647, "grad_norm": 0.47921170295831617, "learning_rate": 1.9754168778769058e-05, "loss": 0.193, "step": 10974 }, { "epoch": 5.1647058823529415, "grad_norm": 0.25858437330785705, "learning_rate": 1.9732430294353636e-05, "loss": 0.0875, "step": 10975 }, { "epoch": 5.165176470588236, "grad_norm": 0.29716634535272435, "learning_rate": 1.9710702935427642e-05, "loss": 0.1008, "step": 10976 }, { "epoch": 5.165647058823529, "grad_norm": 0.3203466796584818, "learning_rate": 1.9688986703846658e-05, "loss": 0.1786, "step": 10977 }, { "epoch": 5.166117647058823, "grad_norm": 0.4996541602576373, "learning_rate": 1.9667281601465372e-05, "loss": 0.227, "step": 10978 }, { "epoch": 5.1665882352941175, "grad_norm": 0.48209613859154593, "learning_rate": 1.964558763013755e-05, "loss": 0.1385, "step": 10979 }, { "epoch": 5.167058823529412, "grad_norm": 0.41771216953321594, "learning_rate": 1.9623904791715907e-05, "loss": 0.3283, "step": 10980 }, { "epoch": 5.167529411764706, "grad_norm": 0.5058593493966904, "learning_rate": 1.9602233088052277e-05, "loss": 0.2331, "step": 10981 }, { "epoch": 5.168, "grad_norm": 0.38573501377916947, "learning_rate": 1.9580572520997566e-05, "loss": 0.1034, "step": 10982 }, { "epoch": 5.168470588235294, "grad_norm": 0.4230926284044933, "learning_rate": 1.9558923092401645e-05, "loss": 0.2169, "step": 10983 }, { "epoch": 5.168941176470589, "grad_norm": 0.3001264123741159, "learning_rate": 1.95372848041135e-05, "loss": 0.0782, "step": 10984 }, { "epoch": 5.169411764705882, "grad_norm": 0.5512244076581014, "learning_rate": 1.951565765798117e-05, "loss": 0.2556, "step": 10985 }, { "epoch": 5.169882352941176, "grad_norm": 0.4675109082950825, "learning_rate": 1.9494041655851677e-05, "loss": 0.2766, "step": 10986 }, { "epoch": 5.17035294117647, "grad_norm": 0.36075669494559787, "learning_rate": 1.9472436799571145e-05, "loss": 0.1809, "step": 10987 }, { "epoch": 5.170823529411765, "grad_norm": 0.6115094459469357, "learning_rate": 1.9450843090984797e-05, "loss": 0.2184, "step": 10988 }, { "epoch": 5.171294117647059, "grad_norm": 0.1692972553501042, "learning_rate": 1.9429260531936743e-05, "loss": 0.0333, "step": 10989 }, { "epoch": 5.171764705882353, "grad_norm": 0.3376982848063362, "learning_rate": 1.940768912427027e-05, "loss": 0.1433, "step": 10990 }, { "epoch": 5.172235294117647, "grad_norm": 0.3675290915864998, "learning_rate": 1.9386128869827743e-05, "loss": 0.1007, "step": 10991 }, { "epoch": 5.1727058823529415, "grad_norm": 0.35750687778504386, "learning_rate": 1.936457977045044e-05, "loss": 0.1755, "step": 10992 }, { "epoch": 5.173176470588236, "grad_norm": 0.26933893015344007, "learning_rate": 1.9343041827978785e-05, "loss": 0.1071, "step": 10993 }, { "epoch": 5.173647058823529, "grad_norm": 0.4462988525852531, "learning_rate": 1.9321515044252273e-05, "loss": 0.211, "step": 10994 }, { "epoch": 5.174117647058823, "grad_norm": 0.34971288641017384, "learning_rate": 1.9299999421109303e-05, "loss": 0.1194, "step": 10995 }, { "epoch": 5.1745882352941175, "grad_norm": 0.5304932830621508, "learning_rate": 1.9278494960387452e-05, "loss": 0.2818, "step": 10996 }, { "epoch": 5.175058823529412, "grad_norm": 0.2819299071313077, "learning_rate": 1.9257001663923338e-05, "loss": 0.0859, "step": 10997 }, { "epoch": 5.175529411764706, "grad_norm": 0.3609661426748662, "learning_rate": 1.9235519533552556e-05, "loss": 0.092, "step": 10998 }, { "epoch": 5.176, "grad_norm": 0.349757936792781, "learning_rate": 1.921404857110984e-05, "loss": 0.2032, "step": 10999 }, { "epoch": 5.176470588235294, "grad_norm": 0.3938530591346248, "learning_rate": 1.9192588778428842e-05, "loss": 0.2365, "step": 11000 }, { "epoch": 5.176941176470589, "grad_norm": 0.23157409955197003, "learning_rate": 1.9171140157342407e-05, "loss": 0.098, "step": 11001 }, { "epoch": 5.177411764705882, "grad_norm": 0.49539909145127015, "learning_rate": 1.9149702709682292e-05, "loss": 0.2786, "step": 11002 }, { "epoch": 5.177882352941176, "grad_norm": 0.41383043396257096, "learning_rate": 1.9128276437279362e-05, "loss": 0.2093, "step": 11003 }, { "epoch": 5.17835294117647, "grad_norm": 0.4280804684844604, "learning_rate": 1.9106861341963587e-05, "loss": 0.2256, "step": 11004 }, { "epoch": 5.178823529411765, "grad_norm": 0.22587731525840585, "learning_rate": 1.9085457425563855e-05, "loss": 0.0685, "step": 11005 }, { "epoch": 5.179294117647059, "grad_norm": 0.3127800931053032, "learning_rate": 1.9064064689908187e-05, "loss": 0.0744, "step": 11006 }, { "epoch": 5.179764705882353, "grad_norm": 0.16461037549972243, "learning_rate": 1.9042683136823654e-05, "loss": 0.0307, "step": 11007 }, { "epoch": 5.180235294117647, "grad_norm": 0.07627672844241479, "learning_rate": 1.9021312768136314e-05, "loss": 0.01, "step": 11008 }, { "epoch": 5.1807058823529415, "grad_norm": 0.28384303232032926, "learning_rate": 1.899995358567129e-05, "loss": 0.1705, "step": 11009 }, { "epoch": 5.181176470588236, "grad_norm": 0.4843741808625652, "learning_rate": 1.897860559125282e-05, "loss": 0.3344, "step": 11010 }, { "epoch": 5.181647058823529, "grad_norm": 0.44372949869719625, "learning_rate": 1.895726878670405e-05, "loss": 0.2676, "step": 11011 }, { "epoch": 5.182117647058823, "grad_norm": 0.2995570292264209, "learning_rate": 1.8935943173847302e-05, "loss": 0.095, "step": 11012 }, { "epoch": 5.1825882352941175, "grad_norm": 0.31939435383019527, "learning_rate": 1.8914628754503885e-05, "loss": 0.1438, "step": 11013 }, { "epoch": 5.183058823529412, "grad_norm": 0.5466658099596008, "learning_rate": 1.889332553049413e-05, "loss": 0.2344, "step": 11014 }, { "epoch": 5.183529411764706, "grad_norm": 0.7243143983089926, "learning_rate": 1.887203350363743e-05, "loss": 0.3666, "step": 11015 }, { "epoch": 5.184, "grad_norm": 0.31207512678609733, "learning_rate": 1.8850752675752295e-05, "loss": 0.1163, "step": 11016 }, { "epoch": 5.184470588235294, "grad_norm": 0.2598013084033923, "learning_rate": 1.8829483048656124e-05, "loss": 0.0658, "step": 11017 }, { "epoch": 5.184941176470589, "grad_norm": 0.2649768824284918, "learning_rate": 1.8808224624165497e-05, "loss": 0.0763, "step": 11018 }, { "epoch": 5.185411764705882, "grad_norm": 0.3158625976197293, "learning_rate": 1.878697740409601e-05, "loss": 0.1201, "step": 11019 }, { "epoch": 5.185882352941176, "grad_norm": 0.27573111361438574, "learning_rate": 1.8765741390262212e-05, "loss": 0.0847, "step": 11020 }, { "epoch": 5.18635294117647, "grad_norm": 0.390078438325273, "learning_rate": 1.8744516584477803e-05, "loss": 0.1862, "step": 11021 }, { "epoch": 5.186823529411765, "grad_norm": 0.184041207784778, "learning_rate": 1.87233029885555e-05, "loss": 0.0392, "step": 11022 }, { "epoch": 5.187294117647059, "grad_norm": 0.709163558317527, "learning_rate": 1.8702100604307018e-05, "loss": 0.1864, "step": 11023 }, { "epoch": 5.187764705882353, "grad_norm": 0.47299025606306205, "learning_rate": 1.868090943354318e-05, "loss": 0.2552, "step": 11024 }, { "epoch": 5.188235294117647, "grad_norm": 0.42524635716938813, "learning_rate": 1.8659729478073764e-05, "loss": 0.2514, "step": 11025 }, { "epoch": 5.1887058823529415, "grad_norm": 0.4436928092972297, "learning_rate": 1.8638560739707664e-05, "loss": 0.2072, "step": 11026 }, { "epoch": 5.189176470588236, "grad_norm": 0.4125875427217261, "learning_rate": 1.8617403220252846e-05, "loss": 0.1078, "step": 11027 }, { "epoch": 5.189647058823529, "grad_norm": 0.4178342124508013, "learning_rate": 1.8596256921516166e-05, "loss": 0.2445, "step": 11028 }, { "epoch": 5.190117647058823, "grad_norm": 0.4844591223356022, "learning_rate": 1.8575121845303698e-05, "loss": 0.1089, "step": 11029 }, { "epoch": 5.1905882352941175, "grad_norm": 0.3509749820702183, "learning_rate": 1.8553997993420495e-05, "loss": 0.154, "step": 11030 }, { "epoch": 5.191058823529412, "grad_norm": 0.4112252268134064, "learning_rate": 1.8532885367670547e-05, "loss": 0.1171, "step": 11031 }, { "epoch": 5.191529411764706, "grad_norm": 0.4391574763627959, "learning_rate": 1.8511783969857046e-05, "loss": 0.228, "step": 11032 }, { "epoch": 5.192, "grad_norm": 0.6484103623388698, "learning_rate": 1.849069380178217e-05, "loss": 0.1416, "step": 11033 }, { "epoch": 5.192470588235294, "grad_norm": 0.5261239106627751, "learning_rate": 1.8469614865247075e-05, "loss": 0.2476, "step": 11034 }, { "epoch": 5.192941176470589, "grad_norm": 0.47970879114636195, "learning_rate": 1.8448547162052002e-05, "loss": 0.2912, "step": 11035 }, { "epoch": 5.193411764705882, "grad_norm": 0.3062163901397128, "learning_rate": 1.8427490693996298e-05, "loss": 0.1009, "step": 11036 }, { "epoch": 5.193882352941176, "grad_norm": 0.6775005480262145, "learning_rate": 1.840644546287821e-05, "loss": 0.3056, "step": 11037 }, { "epoch": 5.19435294117647, "grad_norm": 0.2765336571198937, "learning_rate": 1.838541147049515e-05, "loss": 0.0686, "step": 11038 }, { "epoch": 5.194823529411765, "grad_norm": 0.5361712315264053, "learning_rate": 1.8364388718643542e-05, "loss": 0.2023, "step": 11039 }, { "epoch": 5.195294117647059, "grad_norm": 0.42797664145876946, "learning_rate": 1.8343377209118755e-05, "loss": 0.1755, "step": 11040 }, { "epoch": 5.195764705882353, "grad_norm": 0.49312202714186026, "learning_rate": 1.8322376943715382e-05, "loss": 0.3043, "step": 11041 }, { "epoch": 5.196235294117647, "grad_norm": 0.23554088275442467, "learning_rate": 1.8301387924226863e-05, "loss": 0.0936, "step": 11042 }, { "epoch": 5.1967058823529415, "grad_norm": 0.4265141967394384, "learning_rate": 1.828041015244579e-05, "loss": 0.1982, "step": 11043 }, { "epoch": 5.197176470588236, "grad_norm": 0.40503650990518253, "learning_rate": 1.82594436301638e-05, "loss": 0.1518, "step": 11044 }, { "epoch": 5.197647058823529, "grad_norm": 0.6689009600269069, "learning_rate": 1.8238488359171476e-05, "loss": 0.2767, "step": 11045 }, { "epoch": 5.198117647058823, "grad_norm": 0.5852157272951597, "learning_rate": 1.821754434125854e-05, "loss": 0.116, "step": 11046 }, { "epoch": 5.1985882352941175, "grad_norm": 0.2582843720977608, "learning_rate": 1.8196611578213738e-05, "loss": 0.0514, "step": 11047 }, { "epoch": 5.199058823529412, "grad_norm": 0.928972825941162, "learning_rate": 1.817569007182476e-05, "loss": 0.2019, "step": 11048 }, { "epoch": 5.199529411764706, "grad_norm": 0.34818792794562103, "learning_rate": 1.815477982387846e-05, "loss": 0.0879, "step": 11049 }, { "epoch": 5.2, "grad_norm": 0.5963317051435287, "learning_rate": 1.813388083616068e-05, "loss": 0.2814, "step": 11050 }, { "epoch": 5.200470588235294, "grad_norm": 0.4678628728143303, "learning_rate": 1.8112993110456235e-05, "loss": 0.1801, "step": 11051 }, { "epoch": 5.200941176470589, "grad_norm": 0.407161841024337, "learning_rate": 1.8092116648549136e-05, "loss": 0.1756, "step": 11052 }, { "epoch": 5.201411764705882, "grad_norm": 0.34920193528639043, "learning_rate": 1.8071251452222224e-05, "loss": 0.1713, "step": 11053 }, { "epoch": 5.201882352941176, "grad_norm": 0.5828897405016805, "learning_rate": 1.8050397523257555e-05, "loss": 0.3493, "step": 11054 }, { "epoch": 5.20235294117647, "grad_norm": 0.35773452120843474, "learning_rate": 1.8029554863436186e-05, "loss": 0.1741, "step": 11055 }, { "epoch": 5.202823529411765, "grad_norm": 0.3904909780949048, "learning_rate": 1.8008723474538095e-05, "loss": 0.1804, "step": 11056 }, { "epoch": 5.203294117647059, "grad_norm": 0.3191965987499659, "learning_rate": 1.798790335834244e-05, "loss": 0.1124, "step": 11057 }, { "epoch": 5.203764705882353, "grad_norm": 0.4040001916154317, "learning_rate": 1.796709451662739e-05, "loss": 0.2055, "step": 11058 }, { "epoch": 5.204235294117647, "grad_norm": 0.14527821568411953, "learning_rate": 1.7946296951170065e-05, "loss": 0.0196, "step": 11059 }, { "epoch": 5.2047058823529415, "grad_norm": 0.5055349325075805, "learning_rate": 1.792551066374668e-05, "loss": 0.1962, "step": 11060 }, { "epoch": 5.205176470588235, "grad_norm": 0.7401787716842527, "learning_rate": 1.7904735656132545e-05, "loss": 0.3378, "step": 11061 }, { "epoch": 5.205647058823529, "grad_norm": 0.4928478541217036, "learning_rate": 1.7883971930101883e-05, "loss": 0.3116, "step": 11062 }, { "epoch": 5.206117647058823, "grad_norm": 0.5201435155163081, "learning_rate": 1.786321948742803e-05, "loss": 0.3134, "step": 11063 }, { "epoch": 5.2065882352941175, "grad_norm": 0.22952480115311982, "learning_rate": 1.78424783298834e-05, "loss": 0.0549, "step": 11064 }, { "epoch": 5.207058823529412, "grad_norm": 0.399757070590028, "learning_rate": 1.782174845923931e-05, "loss": 0.0875, "step": 11065 }, { "epoch": 5.207529411764706, "grad_norm": 0.594200494586859, "learning_rate": 1.7801029877266237e-05, "loss": 0.2888, "step": 11066 }, { "epoch": 5.208, "grad_norm": 0.5860932879704305, "learning_rate": 1.778032258573368e-05, "loss": 0.408, "step": 11067 }, { "epoch": 5.208470588235294, "grad_norm": 0.2836159403068285, "learning_rate": 1.7759626586410074e-05, "loss": 0.0821, "step": 11068 }, { "epoch": 5.208941176470589, "grad_norm": 0.4397426346070711, "learning_rate": 1.7738941881063007e-05, "loss": 0.1854, "step": 11069 }, { "epoch": 5.209411764705882, "grad_norm": 0.7291345144828665, "learning_rate": 1.7718268471459045e-05, "loss": 0.4371, "step": 11070 }, { "epoch": 5.209882352941176, "grad_norm": 0.2650966661306651, "learning_rate": 1.769760635936378e-05, "loss": 0.0651, "step": 11071 }, { "epoch": 5.21035294117647, "grad_norm": 0.3741659471359341, "learning_rate": 1.7676955546541867e-05, "loss": 0.2564, "step": 11072 }, { "epoch": 5.210823529411765, "grad_norm": 0.429870192866169, "learning_rate": 1.7656316034757024e-05, "loss": 0.21, "step": 11073 }, { "epoch": 5.211294117647059, "grad_norm": 0.4330325128639432, "learning_rate": 1.7635687825771894e-05, "loss": 0.1991, "step": 11074 }, { "epoch": 5.211764705882353, "grad_norm": 0.3769235152588239, "learning_rate": 1.7615070921348316e-05, "loss": 0.2149, "step": 11075 }, { "epoch": 5.212235294117647, "grad_norm": 0.795761214303735, "learning_rate": 1.7594465323246987e-05, "loss": 0.2222, "step": 11076 }, { "epoch": 5.2127058823529415, "grad_norm": 0.4073635320803616, "learning_rate": 1.757387103322776e-05, "loss": 0.167, "step": 11077 }, { "epoch": 5.213176470588236, "grad_norm": 0.439509923573149, "learning_rate": 1.7553288053049515e-05, "loss": 0.2976, "step": 11078 }, { "epoch": 5.213647058823529, "grad_norm": 0.4748634826470388, "learning_rate": 1.7532716384470108e-05, "loss": 0.2275, "step": 11079 }, { "epoch": 5.214117647058823, "grad_norm": 0.3244404932898591, "learning_rate": 1.7512156029246453e-05, "loss": 0.2112, "step": 11080 }, { "epoch": 5.2145882352941175, "grad_norm": 0.38232708845487723, "learning_rate": 1.7491606989134566e-05, "loss": 0.2241, "step": 11081 }, { "epoch": 5.215058823529412, "grad_norm": 0.22556781336260193, "learning_rate": 1.7471069265889353e-05, "loss": 0.0576, "step": 11082 }, { "epoch": 5.215529411764706, "grad_norm": 0.34683626537331796, "learning_rate": 1.7450542861264883e-05, "loss": 0.1351, "step": 11083 }, { "epoch": 5.216, "grad_norm": 0.5729423422941982, "learning_rate": 1.7430027777014217e-05, "loss": 0.2863, "step": 11084 }, { "epoch": 5.216470588235294, "grad_norm": 0.33265468294972933, "learning_rate": 1.7409524014889388e-05, "loss": 0.1535, "step": 11085 }, { "epoch": 5.216941176470589, "grad_norm": 0.5824175245909878, "learning_rate": 1.738903157664161e-05, "loss": 0.2939, "step": 11086 }, { "epoch": 5.217411764705882, "grad_norm": 0.5301277741054821, "learning_rate": 1.736855046402097e-05, "loss": 0.3351, "step": 11087 }, { "epoch": 5.217882352941176, "grad_norm": 0.19211206360266958, "learning_rate": 1.7348080678776665e-05, "loss": 0.0429, "step": 11088 }, { "epoch": 5.21835294117647, "grad_norm": 0.63414018366709, "learning_rate": 1.732762222265695e-05, "loss": 0.2978, "step": 11089 }, { "epoch": 5.218823529411765, "grad_norm": 0.33015285089183327, "learning_rate": 1.7307175097409044e-05, "loss": 0.1068, "step": 11090 }, { "epoch": 5.219294117647059, "grad_norm": 0.24320029474513363, "learning_rate": 1.7286739304779214e-05, "loss": 0.0821, "step": 11091 }, { "epoch": 5.219764705882353, "grad_norm": 0.7373159187876771, "learning_rate": 1.7266314846512848e-05, "loss": 0.1669, "step": 11092 }, { "epoch": 5.220235294117647, "grad_norm": 0.5386785628545588, "learning_rate": 1.7245901724354218e-05, "loss": 0.2743, "step": 11093 }, { "epoch": 5.2207058823529415, "grad_norm": 0.41212391332191184, "learning_rate": 1.7225499940046726e-05, "loss": 0.2199, "step": 11094 }, { "epoch": 5.221176470588235, "grad_norm": 0.19726188964451719, "learning_rate": 1.720510949533282e-05, "loss": 0.051, "step": 11095 }, { "epoch": 5.221647058823529, "grad_norm": 0.3364033570820994, "learning_rate": 1.71847303919539e-05, "loss": 0.2029, "step": 11096 }, { "epoch": 5.222117647058823, "grad_norm": 1.0078869565462367, "learning_rate": 1.7164362631650457e-05, "loss": 0.3647, "step": 11097 }, { "epoch": 5.2225882352941175, "grad_norm": 0.2724482838395058, "learning_rate": 1.7144006216162036e-05, "loss": 0.068, "step": 11098 }, { "epoch": 5.223058823529412, "grad_norm": 0.4915094271407177, "learning_rate": 1.7123661147227096e-05, "loss": 0.2637, "step": 11099 }, { "epoch": 5.223529411764706, "grad_norm": 0.324138751968887, "learning_rate": 1.7103327426583265e-05, "loss": 0.1042, "step": 11100 }, { "epoch": 5.224, "grad_norm": 0.3377679759487948, "learning_rate": 1.7083005055967146e-05, "loss": 0.0989, "step": 11101 }, { "epoch": 5.224470588235294, "grad_norm": 0.8353522830536384, "learning_rate": 1.7062694037114326e-05, "loss": 0.3202, "step": 11102 }, { "epoch": 5.224941176470589, "grad_norm": 0.33360580403194934, "learning_rate": 1.7042394371759516e-05, "loss": 0.1466, "step": 11103 }, { "epoch": 5.225411764705882, "grad_norm": 0.4673546192737102, "learning_rate": 1.7022106061636342e-05, "loss": 0.2661, "step": 11104 }, { "epoch": 5.225882352941176, "grad_norm": 0.3561023855714766, "learning_rate": 1.700182910847757e-05, "loss": 0.0811, "step": 11105 }, { "epoch": 5.22635294117647, "grad_norm": 0.9888176238120324, "learning_rate": 1.6981563514014967e-05, "loss": 0.2388, "step": 11106 }, { "epoch": 5.226823529411765, "grad_norm": 0.5239092675984397, "learning_rate": 1.6961309279979278e-05, "loss": 0.2601, "step": 11107 }, { "epoch": 5.227294117647059, "grad_norm": 0.5640159260001901, "learning_rate": 1.6941066408100308e-05, "loss": 0.3329, "step": 11108 }, { "epoch": 5.227764705882353, "grad_norm": 0.41987459018544476, "learning_rate": 1.6920834900106967e-05, "loss": 0.1875, "step": 11109 }, { "epoch": 5.228235294117647, "grad_norm": 0.4180415857841557, "learning_rate": 1.690061475772703e-05, "loss": 0.2179, "step": 11110 }, { "epoch": 5.2287058823529415, "grad_norm": 0.2937680372413238, "learning_rate": 1.6880405982687455e-05, "loss": 0.0992, "step": 11111 }, { "epoch": 5.229176470588235, "grad_norm": 0.4137900692219721, "learning_rate": 1.6860208576714173e-05, "loss": 0.1901, "step": 11112 }, { "epoch": 5.229647058823529, "grad_norm": 0.6013421632727773, "learning_rate": 1.68400225415321e-05, "loss": 0.3081, "step": 11113 }, { "epoch": 5.230117647058823, "grad_norm": 0.6299708813304684, "learning_rate": 1.6819847878865255e-05, "loss": 0.3829, "step": 11114 }, { "epoch": 5.2305882352941175, "grad_norm": 0.36658647369497, "learning_rate": 1.6799684590436663e-05, "loss": 0.1309, "step": 11115 }, { "epoch": 5.231058823529412, "grad_norm": 0.35340087435485623, "learning_rate": 1.6779532677968327e-05, "loss": 0.1599, "step": 11116 }, { "epoch": 5.231529411764706, "grad_norm": 0.5837574403255371, "learning_rate": 1.6759392143181354e-05, "loss": 0.442, "step": 11117 }, { "epoch": 5.232, "grad_norm": 0.23889403806032622, "learning_rate": 1.6739262987795854e-05, "loss": 0.0586, "step": 11118 }, { "epoch": 5.232470588235294, "grad_norm": 0.32826778523823835, "learning_rate": 1.6719145213530898e-05, "loss": 0.1141, "step": 11119 }, { "epoch": 5.232941176470589, "grad_norm": 0.43099774868301816, "learning_rate": 1.669903882210468e-05, "loss": 0.2382, "step": 11120 }, { "epoch": 5.233411764705882, "grad_norm": 0.371550873933665, "learning_rate": 1.667894381523443e-05, "loss": 0.1317, "step": 11121 }, { "epoch": 5.233882352941176, "grad_norm": 0.2936413803755311, "learning_rate": 1.665886019463627e-05, "loss": 0.1325, "step": 11122 }, { "epoch": 5.23435294117647, "grad_norm": 0.4275347525179762, "learning_rate": 1.6638787962025495e-05, "loss": 0.2168, "step": 11123 }, { "epoch": 5.234823529411765, "grad_norm": 0.48404115401257836, "learning_rate": 1.6618727119116388e-05, "loss": 0.2983, "step": 11124 }, { "epoch": 5.235294117647059, "grad_norm": 0.41938207238556546, "learning_rate": 1.6598677667622175e-05, "loss": 0.2211, "step": 11125 }, { "epoch": 5.235764705882353, "grad_norm": 0.318621515052607, "learning_rate": 1.657863960925525e-05, "loss": 0.1628, "step": 11126 }, { "epoch": 5.236235294117647, "grad_norm": 0.3621351730370466, "learning_rate": 1.655861294572692e-05, "loss": 0.2183, "step": 11127 }, { "epoch": 5.2367058823529415, "grad_norm": 0.2524972136898557, "learning_rate": 1.6538597678747555e-05, "loss": 0.0435, "step": 11128 }, { "epoch": 5.237176470588235, "grad_norm": 0.49745474539068346, "learning_rate": 1.651859381002658e-05, "loss": 0.2918, "step": 11129 }, { "epoch": 5.237647058823529, "grad_norm": 0.27062632294310884, "learning_rate": 1.649860134127241e-05, "loss": 0.0689, "step": 11130 }, { "epoch": 5.238117647058823, "grad_norm": 0.5848935715061677, "learning_rate": 1.647862027419251e-05, "loss": 0.403, "step": 11131 }, { "epoch": 5.2385882352941175, "grad_norm": 0.2678883784804134, "learning_rate": 1.6458650610493385e-05, "loss": 0.071, "step": 11132 }, { "epoch": 5.239058823529412, "grad_norm": 0.7992970060237715, "learning_rate": 1.6438692351880482e-05, "loss": 0.2369, "step": 11133 }, { "epoch": 5.239529411764706, "grad_norm": 0.11807527213537865, "learning_rate": 1.6418745500058395e-05, "loss": 0.0148, "step": 11134 }, { "epoch": 5.24, "grad_norm": 0.46183812433300614, "learning_rate": 1.639881005673064e-05, "loss": 0.3182, "step": 11135 }, { "epoch": 5.240470588235294, "grad_norm": 0.4116580886860972, "learning_rate": 1.6378886023599803e-05, "loss": 0.1208, "step": 11136 }, { "epoch": 5.240941176470589, "grad_norm": 0.5823512016449514, "learning_rate": 1.6358973402367524e-05, "loss": 0.2411, "step": 11137 }, { "epoch": 5.241411764705882, "grad_norm": 0.31902789555090144, "learning_rate": 1.6339072194734414e-05, "loss": 0.1542, "step": 11138 }, { "epoch": 5.241882352941176, "grad_norm": 0.15544258944077868, "learning_rate": 1.6319182402400144e-05, "loss": 0.0221, "step": 11139 }, { "epoch": 5.24235294117647, "grad_norm": 0.6946294559040226, "learning_rate": 1.629930402706342e-05, "loss": 0.2975, "step": 11140 }, { "epoch": 5.242823529411765, "grad_norm": 0.3592617751488429, "learning_rate": 1.6279437070421885e-05, "loss": 0.2107, "step": 11141 }, { "epoch": 5.243294117647059, "grad_norm": 0.3597044793193479, "learning_rate": 1.625958153417235e-05, "loss": 0.1318, "step": 11142 }, { "epoch": 5.243764705882353, "grad_norm": 0.2810438928270346, "learning_rate": 1.6239737420010558e-05, "loss": 0.0819, "step": 11143 }, { "epoch": 5.244235294117647, "grad_norm": 0.27198068178049956, "learning_rate": 1.621990472963125e-05, "loss": 0.0467, "step": 11144 }, { "epoch": 5.2447058823529416, "grad_norm": 0.44335540610917734, "learning_rate": 1.620008346472826e-05, "loss": 0.1974, "step": 11145 }, { "epoch": 5.245176470588235, "grad_norm": 0.5505371335838439, "learning_rate": 1.618027362699446e-05, "loss": 0.3564, "step": 11146 }, { "epoch": 5.245647058823529, "grad_norm": 0.24168556659369553, "learning_rate": 1.6160475218121654e-05, "loss": 0.045, "step": 11147 }, { "epoch": 5.246117647058823, "grad_norm": 0.5950503061452053, "learning_rate": 1.6140688239800736e-05, "loss": 0.4029, "step": 11148 }, { "epoch": 5.2465882352941176, "grad_norm": 1.697063073173196, "learning_rate": 1.6120912693721654e-05, "loss": 0.7577, "step": 11149 }, { "epoch": 5.247058823529412, "grad_norm": 0.5193465665392483, "learning_rate": 1.6101148581573274e-05, "loss": 0.2284, "step": 11150 }, { "epoch": 5.247529411764706, "grad_norm": 0.4282035229582423, "learning_rate": 1.6081395905043565e-05, "loss": 0.0971, "step": 11151 }, { "epoch": 5.248, "grad_norm": 0.3919423792521138, "learning_rate": 1.606165466581954e-05, "loss": 0.1042, "step": 11152 }, { "epoch": 5.248470588235294, "grad_norm": 0.3476130353731919, "learning_rate": 1.6041924865587152e-05, "loss": 0.0793, "step": 11153 }, { "epoch": 5.248941176470589, "grad_norm": 0.45019628648247234, "learning_rate": 1.602220650603147e-05, "loss": 0.2897, "step": 11154 }, { "epoch": 5.249411764705882, "grad_norm": 0.6420207646457243, "learning_rate": 1.600249958883648e-05, "loss": 0.1372, "step": 11155 }, { "epoch": 5.249882352941176, "grad_norm": 0.6154928456440976, "learning_rate": 1.5982804115685273e-05, "loss": 0.34, "step": 11156 }, { "epoch": 5.2503529411764704, "grad_norm": 0.39910266208282713, "learning_rate": 1.5963120088259977e-05, "loss": 0.1485, "step": 11157 }, { "epoch": 5.250823529411765, "grad_norm": 0.6515995931387546, "learning_rate": 1.594344750824164e-05, "loss": 0.3421, "step": 11158 }, { "epoch": 5.251294117647059, "grad_norm": 0.4972198998383463, "learning_rate": 1.5923786377310433e-05, "loss": 0.2332, "step": 11159 }, { "epoch": 5.251764705882353, "grad_norm": 0.5541537265806652, "learning_rate": 1.5904136697145542e-05, "loss": 0.2466, "step": 11160 }, { "epoch": 5.252235294117647, "grad_norm": 0.351774005847728, "learning_rate": 1.5884498469425078e-05, "loss": 0.0913, "step": 11161 }, { "epoch": 5.252705882352942, "grad_norm": 0.26621767212842695, "learning_rate": 1.5864871695826282e-05, "loss": 0.0683, "step": 11162 }, { "epoch": 5.253176470588235, "grad_norm": 0.38260413261138865, "learning_rate": 1.5845256378025395e-05, "loss": 0.2731, "step": 11163 }, { "epoch": 5.253647058823529, "grad_norm": 0.35211770102139783, "learning_rate": 1.582565251769761e-05, "loss": 0.0989, "step": 11164 }, { "epoch": 5.254117647058823, "grad_norm": 0.643840743517289, "learning_rate": 1.5806060116517227e-05, "loss": 0.2287, "step": 11165 }, { "epoch": 5.254588235294118, "grad_norm": 0.37357643193032763, "learning_rate": 1.578647917615757e-05, "loss": 0.1759, "step": 11166 }, { "epoch": 5.255058823529412, "grad_norm": 0.2117448682240489, "learning_rate": 1.5766909698290863e-05, "loss": 0.0675, "step": 11167 }, { "epoch": 5.255529411764706, "grad_norm": 0.32936641705703446, "learning_rate": 1.574735168458846e-05, "loss": 0.1514, "step": 11168 }, { "epoch": 5.256, "grad_norm": 0.30207045938143723, "learning_rate": 1.572780513672079e-05, "loss": 0.0963, "step": 11169 }, { "epoch": 5.2564705882352944, "grad_norm": 0.27948636260006454, "learning_rate": 1.5708270056357115e-05, "loss": 0.0669, "step": 11170 }, { "epoch": 5.256941176470589, "grad_norm": 0.37107375323575553, "learning_rate": 1.5688746445165867e-05, "loss": 0.2011, "step": 11171 }, { "epoch": 5.257411764705882, "grad_norm": 0.4988333802097123, "learning_rate": 1.5669234304814515e-05, "loss": 0.248, "step": 11172 }, { "epoch": 5.257882352941176, "grad_norm": 0.858065330043827, "learning_rate": 1.564973363696939e-05, "loss": 0.2859, "step": 11173 }, { "epoch": 5.2583529411764705, "grad_norm": 0.45078638029200974, "learning_rate": 1.5630244443296013e-05, "loss": 0.2365, "step": 11174 }, { "epoch": 5.258823529411765, "grad_norm": 0.3674584200716493, "learning_rate": 1.5610766725458834e-05, "loss": 0.206, "step": 11175 }, { "epoch": 5.259294117647059, "grad_norm": 0.16574430726018552, "learning_rate": 1.5591300485121354e-05, "loss": 0.0434, "step": 11176 }, { "epoch": 5.259764705882353, "grad_norm": 0.2909275571939458, "learning_rate": 1.5571845723946113e-05, "loss": 0.1499, "step": 11177 }, { "epoch": 5.260235294117647, "grad_norm": 0.45812550630065585, "learning_rate": 1.555240244359458e-05, "loss": 0.2592, "step": 11178 }, { "epoch": 5.260705882352942, "grad_norm": 0.8219616431286917, "learning_rate": 1.5532970645727344e-05, "loss": 0.4705, "step": 11179 }, { "epoch": 5.261176470588235, "grad_norm": 0.6633559900386211, "learning_rate": 1.551355033200401e-05, "loss": 0.1987, "step": 11180 }, { "epoch": 5.261647058823529, "grad_norm": 0.3313678845811519, "learning_rate": 1.5494141504083087e-05, "loss": 0.1551, "step": 11181 }, { "epoch": 5.262117647058823, "grad_norm": 0.39507823535014885, "learning_rate": 1.547474416362226e-05, "loss": 0.148, "step": 11182 }, { "epoch": 5.262588235294118, "grad_norm": 0.14717773136909246, "learning_rate": 1.5455358312278093e-05, "loss": 0.0213, "step": 11183 }, { "epoch": 5.263058823529412, "grad_norm": 0.3594016793805639, "learning_rate": 1.5435983951706276e-05, "loss": 0.1135, "step": 11184 }, { "epoch": 5.263529411764706, "grad_norm": 0.37761241793554423, "learning_rate": 1.5416621083561486e-05, "loss": 0.1907, "step": 11185 }, { "epoch": 5.264, "grad_norm": 0.3853818463084546, "learning_rate": 1.5397269709497352e-05, "loss": 0.1975, "step": 11186 }, { "epoch": 5.2644705882352945, "grad_norm": 0.13563954308462686, "learning_rate": 1.5377929831166597e-05, "loss": 0.0143, "step": 11187 }, { "epoch": 5.264941176470588, "grad_norm": 0.6693515051518371, "learning_rate": 1.5358601450220988e-05, "loss": 0.3905, "step": 11188 }, { "epoch": 5.265411764705882, "grad_norm": 0.2523221704098221, "learning_rate": 1.53392845683112e-05, "loss": 0.0998, "step": 11189 }, { "epoch": 5.265882352941176, "grad_norm": 0.35383331718704486, "learning_rate": 1.5319979187087018e-05, "loss": 0.0832, "step": 11190 }, { "epoch": 5.2663529411764705, "grad_norm": 0.2373630330388125, "learning_rate": 1.5300685308197254e-05, "loss": 0.1047, "step": 11191 }, { "epoch": 5.266823529411765, "grad_norm": 0.5434886777646868, "learning_rate": 1.528140293328962e-05, "loss": 0.1893, "step": 11192 }, { "epoch": 5.267294117647059, "grad_norm": 0.35196939023835233, "learning_rate": 1.5262132064010956e-05, "loss": 0.1115, "step": 11193 }, { "epoch": 5.267764705882353, "grad_norm": 0.512450108792976, "learning_rate": 1.5242872702007142e-05, "loss": 0.2855, "step": 11194 }, { "epoch": 5.268235294117647, "grad_norm": 0.5566759012093804, "learning_rate": 1.522362484892296e-05, "loss": 0.199, "step": 11195 }, { "epoch": 5.268705882352942, "grad_norm": 0.47279285484682493, "learning_rate": 1.5204388506402281e-05, "loss": 0.2437, "step": 11196 }, { "epoch": 5.269176470588235, "grad_norm": 0.15520141778625182, "learning_rate": 1.5185163676088007e-05, "loss": 0.0352, "step": 11197 }, { "epoch": 5.269647058823529, "grad_norm": 0.6338743487237031, "learning_rate": 1.5165950359622009e-05, "loss": 0.386, "step": 11198 }, { "epoch": 5.270117647058823, "grad_norm": 0.6435007055744941, "learning_rate": 1.5146748558645205e-05, "loss": 0.2897, "step": 11199 }, { "epoch": 5.270588235294118, "grad_norm": 0.3138197282020151, "learning_rate": 1.5127558274797535e-05, "loss": 0.129, "step": 11200 }, { "epoch": 5.271058823529412, "grad_norm": 0.42298884569356965, "learning_rate": 1.5108379509717922e-05, "loss": 0.2767, "step": 11201 }, { "epoch": 5.271529411764706, "grad_norm": 0.24489768828563147, "learning_rate": 1.5089212265044336e-05, "loss": 0.0879, "step": 11202 }, { "epoch": 5.272, "grad_norm": 0.3645948375366426, "learning_rate": 1.5070056542413788e-05, "loss": 0.1797, "step": 11203 }, { "epoch": 5.2724705882352945, "grad_norm": 0.28781396062886716, "learning_rate": 1.5050912343462202e-05, "loss": 0.0708, "step": 11204 }, { "epoch": 5.272941176470589, "grad_norm": 0.4562579657824929, "learning_rate": 1.5031779669824668e-05, "loss": 0.3135, "step": 11205 }, { "epoch": 5.273411764705882, "grad_norm": 0.46301720111459826, "learning_rate": 1.5012658523135113e-05, "loss": 0.1991, "step": 11206 }, { "epoch": 5.273882352941176, "grad_norm": 0.4146148517096667, "learning_rate": 1.4993548905026648e-05, "loss": 0.1885, "step": 11207 }, { "epoch": 5.2743529411764705, "grad_norm": 0.6484784563184506, "learning_rate": 1.4974450817131334e-05, "loss": 0.1359, "step": 11208 }, { "epoch": 5.274823529411765, "grad_norm": 0.35808839228387607, "learning_rate": 1.4955364261080183e-05, "loss": 0.1378, "step": 11209 }, { "epoch": 5.275294117647059, "grad_norm": 0.29677298893122345, "learning_rate": 1.4936289238503324e-05, "loss": 0.0936, "step": 11210 }, { "epoch": 5.275764705882353, "grad_norm": 0.4390175375281368, "learning_rate": 1.491722575102987e-05, "loss": 0.1265, "step": 11211 }, { "epoch": 5.276235294117647, "grad_norm": 0.32856972427786035, "learning_rate": 1.4898173800287888e-05, "loss": 0.118, "step": 11212 }, { "epoch": 5.276705882352942, "grad_norm": 0.294341665106014, "learning_rate": 1.4879133387904523e-05, "loss": 0.0762, "step": 11213 }, { "epoch": 5.277176470588235, "grad_norm": 0.34549821475489473, "learning_rate": 1.4860104515505977e-05, "loss": 0.065, "step": 11214 }, { "epoch": 5.277647058823529, "grad_norm": 0.25025778813868504, "learning_rate": 1.4841087184717315e-05, "loss": 0.0997, "step": 11215 }, { "epoch": 5.278117647058823, "grad_norm": 0.4268010218232155, "learning_rate": 1.4822081397162771e-05, "loss": 0.1865, "step": 11216 }, { "epoch": 5.278588235294118, "grad_norm": 0.4908013893336951, "learning_rate": 1.4803087154465548e-05, "loss": 0.316, "step": 11217 }, { "epoch": 5.279058823529412, "grad_norm": 0.3479206447042343, "learning_rate": 1.4784104458247764e-05, "loss": 0.1307, "step": 11218 }, { "epoch": 5.279529411764706, "grad_norm": 0.45803618662969, "learning_rate": 1.476513331013074e-05, "loss": 0.2728, "step": 11219 }, { "epoch": 5.28, "grad_norm": 0.4144371706051676, "learning_rate": 1.4746173711734627e-05, "loss": 0.2166, "step": 11220 }, { "epoch": 5.2804705882352945, "grad_norm": 0.4061979946931576, "learning_rate": 1.4727225664678699e-05, "loss": 0.1364, "step": 11221 }, { "epoch": 5.280941176470588, "grad_norm": 0.3226752049061953, "learning_rate": 1.4708289170581227e-05, "loss": 0.1189, "step": 11222 }, { "epoch": 5.281411764705882, "grad_norm": 0.5402653985845921, "learning_rate": 1.4689364231059436e-05, "loss": 0.358, "step": 11223 }, { "epoch": 5.281882352941176, "grad_norm": 0.2669139966169366, "learning_rate": 1.4670450847729648e-05, "loss": 0.0736, "step": 11224 }, { "epoch": 5.2823529411764705, "grad_norm": 0.30919479548183815, "learning_rate": 1.465154902220717e-05, "loss": 0.1596, "step": 11225 }, { "epoch": 5.282823529411765, "grad_norm": 0.4536960125076591, "learning_rate": 1.4632658756106247e-05, "loss": 0.2611, "step": 11226 }, { "epoch": 5.283294117647059, "grad_norm": 0.49488206630387616, "learning_rate": 1.4613780051040236e-05, "loss": 0.1872, "step": 11227 }, { "epoch": 5.283764705882353, "grad_norm": 0.4604363859464677, "learning_rate": 1.4594912908621515e-05, "loss": 0.2305, "step": 11228 }, { "epoch": 5.284235294117647, "grad_norm": 0.5213935802123459, "learning_rate": 1.4576057330461361e-05, "loss": 0.2895, "step": 11229 }, { "epoch": 5.284705882352942, "grad_norm": 0.3270406500329489, "learning_rate": 1.4557213318170152e-05, "loss": 0.1414, "step": 11230 }, { "epoch": 5.285176470588235, "grad_norm": 0.31476208807924033, "learning_rate": 1.453838087335732e-05, "loss": 0.1459, "step": 11231 }, { "epoch": 5.285647058823529, "grad_norm": 0.423640285074371, "learning_rate": 1.4519559997631142e-05, "loss": 0.2285, "step": 11232 }, { "epoch": 5.286117647058823, "grad_norm": 0.8127605553655208, "learning_rate": 1.45007506925991e-05, "loss": 0.3276, "step": 11233 }, { "epoch": 5.286588235294118, "grad_norm": 0.14736606018711626, "learning_rate": 1.448195295986756e-05, "loss": 0.0223, "step": 11234 }, { "epoch": 5.287058823529412, "grad_norm": 0.4391415118899875, "learning_rate": 1.4463166801041936e-05, "loss": 0.2075, "step": 11235 }, { "epoch": 5.287529411764706, "grad_norm": 0.31129430190417784, "learning_rate": 1.4444392217726697e-05, "loss": 0.1715, "step": 11236 }, { "epoch": 5.288, "grad_norm": 0.31577775144056847, "learning_rate": 1.442562921152524e-05, "loss": 0.1346, "step": 11237 }, { "epoch": 5.2884705882352945, "grad_norm": 0.34250601122632557, "learning_rate": 1.4406877784040021e-05, "loss": 0.0941, "step": 11238 }, { "epoch": 5.288941176470589, "grad_norm": 0.30941200562215077, "learning_rate": 1.4388137936872556e-05, "loss": 0.1506, "step": 11239 }, { "epoch": 5.289411764705882, "grad_norm": 0.438823297830055, "learning_rate": 1.4369409671623267e-05, "loss": 0.19, "step": 11240 }, { "epoch": 5.289882352941176, "grad_norm": 0.3737689664555424, "learning_rate": 1.4350692989891638e-05, "loss": 0.1997, "step": 11241 }, { "epoch": 5.2903529411764705, "grad_norm": 0.39553635088307915, "learning_rate": 1.4331987893276225e-05, "loss": 0.1684, "step": 11242 }, { "epoch": 5.290823529411765, "grad_norm": 0.3829833611934329, "learning_rate": 1.4313294383374468e-05, "loss": 0.1836, "step": 11243 }, { "epoch": 5.291294117647059, "grad_norm": 0.416133355271638, "learning_rate": 1.429461246178289e-05, "loss": 0.1686, "step": 11244 }, { "epoch": 5.291764705882353, "grad_norm": 0.3725234243372054, "learning_rate": 1.4275942130097096e-05, "loss": 0.0966, "step": 11245 }, { "epoch": 5.292235294117647, "grad_norm": 0.29328627444915933, "learning_rate": 1.4257283389911512e-05, "loss": 0.0349, "step": 11246 }, { "epoch": 5.292705882352941, "grad_norm": 0.35144161571189086, "learning_rate": 1.4238636242819761e-05, "loss": 0.1826, "step": 11247 }, { "epoch": 5.293176470588235, "grad_norm": 0.5520211088285785, "learning_rate": 1.422000069041439e-05, "loss": 0.3029, "step": 11248 }, { "epoch": 5.293647058823529, "grad_norm": 0.3927969877589292, "learning_rate": 1.420137673428694e-05, "loss": 0.1298, "step": 11249 }, { "epoch": 5.294117647058823, "grad_norm": 0.4618788175214689, "learning_rate": 1.4182764376028006e-05, "loss": 0.2408, "step": 11250 }, { "epoch": 5.294588235294118, "grad_norm": 0.4524091159466276, "learning_rate": 1.4164163617227197e-05, "loss": 0.1598, "step": 11251 }, { "epoch": 5.295058823529412, "grad_norm": 0.4416555344104688, "learning_rate": 1.4145574459473064e-05, "loss": 0.1473, "step": 11252 }, { "epoch": 5.295529411764706, "grad_norm": 0.22310221075995817, "learning_rate": 1.4126996904353233e-05, "loss": 0.0611, "step": 11253 }, { "epoch": 5.296, "grad_norm": 0.31393777436623177, "learning_rate": 1.4108430953454352e-05, "loss": 0.1531, "step": 11254 }, { "epoch": 5.2964705882352945, "grad_norm": 0.6199970293385613, "learning_rate": 1.4089876608361971e-05, "loss": 0.4169, "step": 11255 }, { "epoch": 5.296941176470588, "grad_norm": 0.2550008704650713, "learning_rate": 1.4071333870660807e-05, "loss": 0.0628, "step": 11256 }, { "epoch": 5.297411764705882, "grad_norm": 0.23970187789596878, "learning_rate": 1.4052802741934404e-05, "loss": 0.0746, "step": 11257 }, { "epoch": 5.297882352941176, "grad_norm": 1.259992462988324, "learning_rate": 1.4034283223765485e-05, "loss": 0.493, "step": 11258 }, { "epoch": 5.2983529411764705, "grad_norm": 0.12942663530807025, "learning_rate": 1.4015775317735717e-05, "loss": 0.0275, "step": 11259 }, { "epoch": 5.298823529411765, "grad_norm": 0.4305470742592095, "learning_rate": 1.3997279025425701e-05, "loss": 0.2097, "step": 11260 }, { "epoch": 5.299294117647059, "grad_norm": 0.4626849357545475, "learning_rate": 1.397879434841514e-05, "loss": 0.2823, "step": 11261 }, { "epoch": 5.299764705882353, "grad_norm": 0.3611170981169735, "learning_rate": 1.3960321288282722e-05, "loss": 0.2054, "step": 11262 }, { "epoch": 5.300235294117647, "grad_norm": 0.4130871095483341, "learning_rate": 1.3941859846606152e-05, "loss": 0.1872, "step": 11263 }, { "epoch": 5.300705882352942, "grad_norm": 0.3112014959238422, "learning_rate": 1.3923410024962117e-05, "loss": 0.1415, "step": 11264 }, { "epoch": 5.301176470588235, "grad_norm": 0.28839984629974874, "learning_rate": 1.390497182492631e-05, "loss": 0.1329, "step": 11265 }, { "epoch": 5.301647058823529, "grad_norm": 0.5633738871160621, "learning_rate": 1.388654524807345e-05, "loss": 0.3325, "step": 11266 }, { "epoch": 5.302117647058823, "grad_norm": 0.356721573181612, "learning_rate": 1.3868130295977282e-05, "loss": 0.1372, "step": 11267 }, { "epoch": 5.302588235294118, "grad_norm": 0.4601266224818548, "learning_rate": 1.3849726970210479e-05, "loss": 0.1929, "step": 11268 }, { "epoch": 5.303058823529412, "grad_norm": 0.4150257460182534, "learning_rate": 1.3831335272344801e-05, "loss": 0.2303, "step": 11269 }, { "epoch": 5.303529411764706, "grad_norm": 0.3943408552155359, "learning_rate": 1.3812955203951026e-05, "loss": 0.2058, "step": 11270 }, { "epoch": 5.304, "grad_norm": 0.2731043257204989, "learning_rate": 1.3794586766598848e-05, "loss": 0.0512, "step": 11271 }, { "epoch": 5.3044705882352945, "grad_norm": 0.26614477642528844, "learning_rate": 1.3776229961857043e-05, "loss": 0.1193, "step": 11272 }, { "epoch": 5.304941176470588, "grad_norm": 0.20946478575957916, "learning_rate": 1.3757884791293394e-05, "loss": 0.0308, "step": 11273 }, { "epoch": 5.305411764705882, "grad_norm": 0.46230620021626134, "learning_rate": 1.373955125647463e-05, "loss": 0.2749, "step": 11274 }, { "epoch": 5.305882352941176, "grad_norm": 0.4917079563545382, "learning_rate": 1.3721229358966545e-05, "loss": 0.2137, "step": 11275 }, { "epoch": 5.3063529411764705, "grad_norm": 0.3509209762778147, "learning_rate": 1.3702919100333925e-05, "loss": 0.1448, "step": 11276 }, { "epoch": 5.306823529411765, "grad_norm": 0.37957638638134655, "learning_rate": 1.3684620482140534e-05, "loss": 0.1977, "step": 11277 }, { "epoch": 5.307294117647059, "grad_norm": 0.33782688374990905, "learning_rate": 1.3666333505949173e-05, "loss": 0.1463, "step": 11278 }, { "epoch": 5.307764705882353, "grad_norm": 0.35556983554096916, "learning_rate": 1.3648058173321675e-05, "loss": 0.1766, "step": 11279 }, { "epoch": 5.308235294117647, "grad_norm": 0.494642059999892, "learning_rate": 1.3629794485818791e-05, "loss": 0.2824, "step": 11280 }, { "epoch": 5.308705882352941, "grad_norm": 0.5558738651518823, "learning_rate": 1.361154244500034e-05, "loss": 0.2729, "step": 11281 }, { "epoch": 5.309176470588235, "grad_norm": 0.49128931511503937, "learning_rate": 1.3593302052425159e-05, "loss": 0.135, "step": 11282 }, { "epoch": 5.309647058823529, "grad_norm": 1.0308903338369484, "learning_rate": 1.3575073309651052e-05, "loss": 0.5214, "step": 11283 }, { "epoch": 5.310117647058823, "grad_norm": 0.5347613207405162, "learning_rate": 1.3556856218234858e-05, "loss": 0.1849, "step": 11284 }, { "epoch": 5.310588235294118, "grad_norm": 0.501476256492666, "learning_rate": 1.3538650779732362e-05, "loss": 0.1536, "step": 11285 }, { "epoch": 5.311058823529412, "grad_norm": 0.18490332289426054, "learning_rate": 1.3520456995698443e-05, "loss": 0.0552, "step": 11286 }, { "epoch": 5.311529411764706, "grad_norm": 0.5457007320460658, "learning_rate": 1.3502274867686936e-05, "loss": 0.3691, "step": 11287 }, { "epoch": 5.312, "grad_norm": 0.44457436300261666, "learning_rate": 1.348410439725065e-05, "loss": 0.2837, "step": 11288 }, { "epoch": 5.3124705882352945, "grad_norm": 0.3869086048584093, "learning_rate": 1.3465945585941462e-05, "loss": 0.1384, "step": 11289 }, { "epoch": 5.312941176470588, "grad_norm": 0.2610446113772853, "learning_rate": 1.3447798435310247e-05, "loss": 0.0656, "step": 11290 }, { "epoch": 5.313411764705882, "grad_norm": 0.28956387051982263, "learning_rate": 1.3429662946906799e-05, "loss": 0.1142, "step": 11291 }, { "epoch": 5.313882352941176, "grad_norm": 0.41571630814188965, "learning_rate": 1.3411539122279996e-05, "loss": 0.2463, "step": 11292 }, { "epoch": 5.3143529411764705, "grad_norm": 0.16995573206115397, "learning_rate": 1.3393426962977765e-05, "loss": 0.0487, "step": 11293 }, { "epoch": 5.314823529411765, "grad_norm": 0.37092620961595574, "learning_rate": 1.3375326470546872e-05, "loss": 0.2125, "step": 11294 }, { "epoch": 5.315294117647059, "grad_norm": 0.4136957829837764, "learning_rate": 1.3357237646533258e-05, "loss": 0.0721, "step": 11295 }, { "epoch": 5.315764705882353, "grad_norm": 0.5359742038670864, "learning_rate": 1.3339160492481793e-05, "loss": 0.2525, "step": 11296 }, { "epoch": 5.316235294117647, "grad_norm": 0.5991280130361096, "learning_rate": 1.3321095009936321e-05, "loss": 0.2423, "step": 11297 }, { "epoch": 5.316705882352942, "grad_norm": 0.3025656957753684, "learning_rate": 1.3303041200439729e-05, "loss": 0.1159, "step": 11298 }, { "epoch": 5.317176470588235, "grad_norm": 0.40970786715587093, "learning_rate": 1.3284999065533946e-05, "loss": 0.1387, "step": 11299 }, { "epoch": 5.317647058823529, "grad_norm": 0.3165870838921326, "learning_rate": 1.326696860675981e-05, "loss": 0.1039, "step": 11300 }, { "epoch": 5.318117647058823, "grad_norm": 0.3299818730669785, "learning_rate": 1.3248949825657217e-05, "loss": 0.1462, "step": 11301 }, { "epoch": 5.318588235294118, "grad_norm": 0.598341980249323, "learning_rate": 1.3230942723765092e-05, "loss": 0.3797, "step": 11302 }, { "epoch": 5.319058823529412, "grad_norm": 0.971001083800062, "learning_rate": 1.3212947302621284e-05, "loss": 0.2438, "step": 11303 }, { "epoch": 5.319529411764706, "grad_norm": 0.2297759483675684, "learning_rate": 1.3194963563762718e-05, "loss": 0.0511, "step": 11304 }, { "epoch": 5.32, "grad_norm": 0.38479898528222706, "learning_rate": 1.317699150872531e-05, "loss": 0.136, "step": 11305 }, { "epoch": 5.3204705882352945, "grad_norm": 0.3351618916290446, "learning_rate": 1.3159031139043918e-05, "loss": 0.1681, "step": 11306 }, { "epoch": 5.320941176470588, "grad_norm": 0.2899215485307357, "learning_rate": 1.3141082456252467e-05, "loss": 0.0924, "step": 11307 }, { "epoch": 5.321411764705882, "grad_norm": 0.421461599459863, "learning_rate": 1.312314546188386e-05, "loss": 0.1595, "step": 11308 }, { "epoch": 5.321882352941176, "grad_norm": 0.4653688829736152, "learning_rate": 1.3105220157470009e-05, "loss": 0.2099, "step": 11309 }, { "epoch": 5.3223529411764705, "grad_norm": 0.6511054563821067, "learning_rate": 1.3087306544541836e-05, "loss": 0.4272, "step": 11310 }, { "epoch": 5.322823529411765, "grad_norm": 0.3420881626011725, "learning_rate": 1.306940462462922e-05, "loss": 0.1595, "step": 11311 }, { "epoch": 5.323294117647059, "grad_norm": 0.43875216153299484, "learning_rate": 1.3051514399261104e-05, "loss": 0.1454, "step": 11312 }, { "epoch": 5.323764705882353, "grad_norm": 0.3142431817408246, "learning_rate": 1.3033635869965398e-05, "loss": 0.1173, "step": 11313 }, { "epoch": 5.324235294117647, "grad_norm": 0.46632401189717676, "learning_rate": 1.3015769038268981e-05, "loss": 0.2056, "step": 11314 }, { "epoch": 5.324705882352941, "grad_norm": 0.4472028501750313, "learning_rate": 1.2997913905697816e-05, "loss": 0.2769, "step": 11315 }, { "epoch": 5.325176470588235, "grad_norm": 0.3148751171275556, "learning_rate": 1.2980070473776782e-05, "loss": 0.0779, "step": 11316 }, { "epoch": 5.325647058823529, "grad_norm": 0.6597720317714129, "learning_rate": 1.2962238744029795e-05, "loss": 0.3014, "step": 11317 }, { "epoch": 5.326117647058823, "grad_norm": 0.3734264305500225, "learning_rate": 1.2944418717979804e-05, "loss": 0.0852, "step": 11318 }, { "epoch": 5.326588235294118, "grad_norm": 0.468378498757032, "learning_rate": 1.2926610397148707e-05, "loss": 0.1831, "step": 11319 }, { "epoch": 5.327058823529412, "grad_norm": 1.127145930771782, "learning_rate": 1.2908813783057404e-05, "loss": 0.4852, "step": 11320 }, { "epoch": 5.327529411764706, "grad_norm": 0.3048717331037709, "learning_rate": 1.2891028877225863e-05, "loss": 0.0753, "step": 11321 }, { "epoch": 5.328, "grad_norm": 0.09226942181430593, "learning_rate": 1.287325568117295e-05, "loss": 0.0166, "step": 11322 }, { "epoch": 5.3284705882352945, "grad_norm": 0.5342016775108003, "learning_rate": 1.2855494196416604e-05, "loss": 0.2855, "step": 11323 }, { "epoch": 5.328941176470588, "grad_norm": 0.4021176261668332, "learning_rate": 1.2837744424473761e-05, "loss": 0.2362, "step": 11324 }, { "epoch": 5.329411764705882, "grad_norm": 0.3686790533868626, "learning_rate": 1.2820006366860292e-05, "loss": 0.1145, "step": 11325 }, { "epoch": 5.329882352941176, "grad_norm": 0.17954047151021024, "learning_rate": 1.2802280025091165e-05, "loss": 0.0337, "step": 11326 }, { "epoch": 5.3303529411764705, "grad_norm": 0.4748036483833591, "learning_rate": 1.2784565400680291e-05, "loss": 0.1901, "step": 11327 }, { "epoch": 5.330823529411765, "grad_norm": 0.5543572109015464, "learning_rate": 1.276686249514054e-05, "loss": 0.2683, "step": 11328 }, { "epoch": 5.331294117647059, "grad_norm": 0.4678784444096865, "learning_rate": 1.2749171309983852e-05, "loss": 0.1959, "step": 11329 }, { "epoch": 5.331764705882353, "grad_norm": 0.4500309812487589, "learning_rate": 1.273149184672117e-05, "loss": 0.2121, "step": 11330 }, { "epoch": 5.332235294117647, "grad_norm": 0.31607113910178486, "learning_rate": 1.2713824106862369e-05, "loss": 0.1024, "step": 11331 }, { "epoch": 5.332705882352942, "grad_norm": 0.49219785433386576, "learning_rate": 1.2696168091916376e-05, "loss": 0.182, "step": 11332 }, { "epoch": 5.333176470588235, "grad_norm": 0.33293872639290256, "learning_rate": 1.2678523803391116e-05, "loss": 0.1627, "step": 11333 }, { "epoch": 5.333647058823529, "grad_norm": 0.6487493028855977, "learning_rate": 1.2660891242793453e-05, "loss": 0.3598, "step": 11334 }, { "epoch": 5.334117647058823, "grad_norm": 0.3354430375862571, "learning_rate": 1.2643270411629346e-05, "loss": 0.0727, "step": 11335 }, { "epoch": 5.334588235294118, "grad_norm": 0.5548282698577806, "learning_rate": 1.2625661311403662e-05, "loss": 0.3038, "step": 11336 }, { "epoch": 5.335058823529412, "grad_norm": 0.33742108956445205, "learning_rate": 1.2608063943620312e-05, "loss": 0.1648, "step": 11337 }, { "epoch": 5.335529411764706, "grad_norm": 0.46026760414075113, "learning_rate": 1.2590478309782243e-05, "loss": 0.2782, "step": 11338 }, { "epoch": 5.336, "grad_norm": 0.4682481214720018, "learning_rate": 1.2572904411391272e-05, "loss": 0.2818, "step": 11339 }, { "epoch": 5.3364705882352945, "grad_norm": 0.5534240997574588, "learning_rate": 1.2555342249948347e-05, "loss": 0.3354, "step": 11340 }, { "epoch": 5.336941176470588, "grad_norm": 0.2135143074161095, "learning_rate": 1.2537791826953387e-05, "loss": 0.0644, "step": 11341 }, { "epoch": 5.337411764705882, "grad_norm": 0.1945519372419229, "learning_rate": 1.2520253143905223e-05, "loss": 0.0396, "step": 11342 }, { "epoch": 5.337882352941176, "grad_norm": 0.0868502765648663, "learning_rate": 1.2502726202301777e-05, "loss": 0.0102, "step": 11343 }, { "epoch": 5.3383529411764705, "grad_norm": 1.0256801937895645, "learning_rate": 1.2485211003639967e-05, "loss": 0.1385, "step": 11344 }, { "epoch": 5.338823529411765, "grad_norm": 0.5410112552855736, "learning_rate": 1.2467707549415612e-05, "loss": 0.2704, "step": 11345 }, { "epoch": 5.339294117647059, "grad_norm": 0.34098312982048984, "learning_rate": 1.2450215841123618e-05, "loss": 0.1589, "step": 11346 }, { "epoch": 5.339764705882353, "grad_norm": 0.4040531236385972, "learning_rate": 1.2432735880257904e-05, "loss": 0.2074, "step": 11347 }, { "epoch": 5.340235294117647, "grad_norm": 0.3810399483512721, "learning_rate": 1.2415267668311263e-05, "loss": 0.1218, "step": 11348 }, { "epoch": 5.340705882352941, "grad_norm": 0.37295529238600433, "learning_rate": 1.2397811206775632e-05, "loss": 0.1413, "step": 11349 }, { "epoch": 5.341176470588235, "grad_norm": 0.3177751734091197, "learning_rate": 1.2380366497141886e-05, "loss": 0.1113, "step": 11350 }, { "epoch": 5.341647058823529, "grad_norm": 0.3516002472587834, "learning_rate": 1.23629335408998e-05, "loss": 0.1835, "step": 11351 }, { "epoch": 5.342117647058823, "grad_norm": 1.1722522452293185, "learning_rate": 1.234551233953835e-05, "loss": 0.3346, "step": 11352 }, { "epoch": 5.342588235294118, "grad_norm": 0.4752343178688172, "learning_rate": 1.232810289454531e-05, "loss": 0.1807, "step": 11353 }, { "epoch": 5.343058823529412, "grad_norm": 0.3309659807457115, "learning_rate": 1.2310705207407561e-05, "loss": 0.134, "step": 11354 }, { "epoch": 5.343529411764706, "grad_norm": 0.1829819799655552, "learning_rate": 1.229331927961098e-05, "loss": 0.0356, "step": 11355 }, { "epoch": 5.344, "grad_norm": 0.33817360115785505, "learning_rate": 1.2275945112640363e-05, "loss": 0.214, "step": 11356 }, { "epoch": 5.3444705882352945, "grad_norm": 0.42971613169505307, "learning_rate": 1.2258582707979554e-05, "loss": 0.2105, "step": 11357 }, { "epoch": 5.344941176470588, "grad_norm": 0.2993947417376117, "learning_rate": 1.2241232067111418e-05, "loss": 0.1316, "step": 11358 }, { "epoch": 5.345411764705882, "grad_norm": 0.42487000973894623, "learning_rate": 1.2223893191517753e-05, "loss": 0.185, "step": 11359 }, { "epoch": 5.345882352941176, "grad_norm": 0.24748455815917939, "learning_rate": 1.2206566082679392e-05, "loss": 0.0528, "step": 11360 }, { "epoch": 5.3463529411764705, "grad_norm": 0.36130830219545257, "learning_rate": 1.2189250742076184e-05, "loss": 0.1944, "step": 11361 }, { "epoch": 5.346823529411765, "grad_norm": 0.5113994873415223, "learning_rate": 1.2171947171186897e-05, "loss": 0.1824, "step": 11362 }, { "epoch": 5.347294117647059, "grad_norm": 0.2108189044450434, "learning_rate": 1.2154655371489397e-05, "loss": 0.0548, "step": 11363 }, { "epoch": 5.347764705882353, "grad_norm": 0.3720702159412738, "learning_rate": 1.213737534446042e-05, "loss": 0.1397, "step": 11364 }, { "epoch": 5.348235294117647, "grad_norm": 0.48113991420916813, "learning_rate": 1.2120107091575803e-05, "loss": 0.1228, "step": 11365 }, { "epoch": 5.348705882352942, "grad_norm": 0.3898537354365244, "learning_rate": 1.2102850614310366e-05, "loss": 0.102, "step": 11366 }, { "epoch": 5.349176470588235, "grad_norm": 0.3680686135998815, "learning_rate": 1.2085605914137847e-05, "loss": 0.1688, "step": 11367 }, { "epoch": 5.349647058823529, "grad_norm": 0.17470476674111127, "learning_rate": 1.206837299253105e-05, "loss": 0.0279, "step": 11368 }, { "epoch": 5.350117647058823, "grad_norm": 0.5166885160070008, "learning_rate": 1.2051151850961766e-05, "loss": 0.2272, "step": 11369 }, { "epoch": 5.350588235294118, "grad_norm": 0.4592001401257559, "learning_rate": 1.2033942490900716e-05, "loss": 0.2426, "step": 11370 }, { "epoch": 5.351058823529412, "grad_norm": 0.23487568881948728, "learning_rate": 1.201674491381771e-05, "loss": 0.0822, "step": 11371 }, { "epoch": 5.351529411764706, "grad_norm": 0.18599354254111572, "learning_rate": 1.1999559121181523e-05, "loss": 0.0548, "step": 11372 }, { "epoch": 5.352, "grad_norm": 0.5724678307114021, "learning_rate": 1.1982385114459846e-05, "loss": 0.3912, "step": 11373 }, { "epoch": 5.352470588235294, "grad_norm": 0.5640051364578575, "learning_rate": 1.1965222895119442e-05, "loss": 0.2898, "step": 11374 }, { "epoch": 5.352941176470588, "grad_norm": 0.333494990818666, "learning_rate": 1.1948072464626101e-05, "loss": 0.1614, "step": 11375 }, { "epoch": 5.353411764705882, "grad_norm": 0.5679960572633296, "learning_rate": 1.1930933824444472e-05, "loss": 0.2462, "step": 11376 }, { "epoch": 5.353882352941176, "grad_norm": 0.5441194492009117, "learning_rate": 1.1913806976038331e-05, "loss": 0.2935, "step": 11377 }, { "epoch": 5.3543529411764705, "grad_norm": 0.38719603317428375, "learning_rate": 1.1896691920870394e-05, "loss": 0.227, "step": 11378 }, { "epoch": 5.354823529411765, "grad_norm": 0.44271330841298057, "learning_rate": 1.1879588660402356e-05, "loss": 0.213, "step": 11379 }, { "epoch": 5.355294117647059, "grad_norm": 0.45727059849615337, "learning_rate": 1.18624971960949e-05, "loss": 0.2251, "step": 11380 }, { "epoch": 5.355764705882353, "grad_norm": 0.24371650001789108, "learning_rate": 1.1845417529407791e-05, "loss": 0.0452, "step": 11381 }, { "epoch": 5.356235294117647, "grad_norm": 0.5931124017633564, "learning_rate": 1.1828349661799629e-05, "loss": 0.3735, "step": 11382 }, { "epoch": 5.356705882352941, "grad_norm": 0.3957848755685006, "learning_rate": 1.1811293594728133e-05, "loss": 0.2095, "step": 11383 }, { "epoch": 5.357176470588235, "grad_norm": 0.7064391927353645, "learning_rate": 1.179424932965002e-05, "loss": 0.1921, "step": 11384 }, { "epoch": 5.357647058823529, "grad_norm": 0.2823107070189204, "learning_rate": 1.1777216868020872e-05, "loss": 0.0765, "step": 11385 }, { "epoch": 5.358117647058823, "grad_norm": 1.328249962726917, "learning_rate": 1.1760196211295398e-05, "loss": 0.3828, "step": 11386 }, { "epoch": 5.358588235294118, "grad_norm": 0.2808574252299725, "learning_rate": 1.1743187360927214e-05, "loss": 0.107, "step": 11387 }, { "epoch": 5.359058823529412, "grad_norm": 0.1792032332236969, "learning_rate": 1.1726190318368977e-05, "loss": 0.061, "step": 11388 }, { "epoch": 5.359529411764706, "grad_norm": 0.4606007334085261, "learning_rate": 1.1709205085072344e-05, "loss": 0.1174, "step": 11389 }, { "epoch": 5.36, "grad_norm": 0.09802934875139624, "learning_rate": 1.1692231662487866e-05, "loss": 0.0158, "step": 11390 }, { "epoch": 5.3604705882352945, "grad_norm": 0.35926454067965663, "learning_rate": 1.1675270052065222e-05, "loss": 0.1423, "step": 11391 }, { "epoch": 5.360941176470588, "grad_norm": 0.5715212545887005, "learning_rate": 1.1658320255253e-05, "loss": 0.2419, "step": 11392 }, { "epoch": 5.361411764705882, "grad_norm": 0.5664934815091333, "learning_rate": 1.1641382273498777e-05, "loss": 0.3029, "step": 11393 }, { "epoch": 5.361882352941176, "grad_norm": 0.5126721341667076, "learning_rate": 1.1624456108249164e-05, "loss": 0.2245, "step": 11394 }, { "epoch": 5.3623529411764705, "grad_norm": 0.3388153666702032, "learning_rate": 1.1607541760949735e-05, "loss": 0.0873, "step": 11395 }, { "epoch": 5.362823529411765, "grad_norm": 0.36702118454786953, "learning_rate": 1.1590639233045002e-05, "loss": 0.1155, "step": 11396 }, { "epoch": 5.363294117647059, "grad_norm": 0.5642420394921539, "learning_rate": 1.1573748525978626e-05, "loss": 0.3386, "step": 11397 }, { "epoch": 5.363764705882353, "grad_norm": 0.3530293798041227, "learning_rate": 1.155686964119309e-05, "loss": 0.1766, "step": 11398 }, { "epoch": 5.364235294117647, "grad_norm": 0.5524040307811506, "learning_rate": 1.154000258012992e-05, "loss": 0.157, "step": 11399 }, { "epoch": 5.364705882352941, "grad_norm": 0.2737435085126173, "learning_rate": 1.1523147344229716e-05, "loss": 0.052, "step": 11400 }, { "epoch": 5.365176470588235, "grad_norm": 0.32183857230398755, "learning_rate": 1.1506303934931926e-05, "loss": 0.09, "step": 11401 }, { "epoch": 5.365647058823529, "grad_norm": 0.4302509551371074, "learning_rate": 1.1489472353675083e-05, "loss": 0.206, "step": 11402 }, { "epoch": 5.366117647058823, "grad_norm": 0.5611342257767798, "learning_rate": 1.147265260189672e-05, "loss": 0.3614, "step": 11403 }, { "epoch": 5.366588235294118, "grad_norm": 0.21782622931147355, "learning_rate": 1.1455844681033272e-05, "loss": 0.046, "step": 11404 }, { "epoch": 5.367058823529412, "grad_norm": 0.1947885180172967, "learning_rate": 1.1439048592520256e-05, "loss": 0.0694, "step": 11405 }, { "epoch": 5.367529411764706, "grad_norm": 0.36711295676878575, "learning_rate": 1.1422264337792125e-05, "loss": 0.1699, "step": 11406 }, { "epoch": 5.368, "grad_norm": 0.5060062703840277, "learning_rate": 1.1405491918282333e-05, "loss": 0.2486, "step": 11407 }, { "epoch": 5.368470588235294, "grad_norm": 0.24827676848242808, "learning_rate": 1.1388731335423334e-05, "loss": 0.069, "step": 11408 }, { "epoch": 5.368941176470588, "grad_norm": 0.7303704762611433, "learning_rate": 1.1371982590646584e-05, "loss": 0.2641, "step": 11409 }, { "epoch": 5.369411764705882, "grad_norm": 0.35053495556726333, "learning_rate": 1.135524568538247e-05, "loss": 0.1379, "step": 11410 }, { "epoch": 5.369882352941176, "grad_norm": 0.33264747385820875, "learning_rate": 1.1338520621060415e-05, "loss": 0.1776, "step": 11411 }, { "epoch": 5.3703529411764706, "grad_norm": 0.8229498074322086, "learning_rate": 1.1321807399108845e-05, "loss": 0.176, "step": 11412 }, { "epoch": 5.370823529411765, "grad_norm": 1.43236944925238, "learning_rate": 1.1305106020955135e-05, "loss": 0.5453, "step": 11413 }, { "epoch": 5.371294117647059, "grad_norm": 0.08302197344595497, "learning_rate": 1.1288416488025675e-05, "loss": 0.0113, "step": 11414 }, { "epoch": 5.371764705882353, "grad_norm": 0.40916251075803634, "learning_rate": 1.1271738801745795e-05, "loss": 0.0583, "step": 11415 }, { "epoch": 5.372235294117647, "grad_norm": 0.18903582128814098, "learning_rate": 1.1255072963539885e-05, "loss": 0.0417, "step": 11416 }, { "epoch": 5.372705882352941, "grad_norm": 0.4057009765911151, "learning_rate": 1.123841897483131e-05, "loss": 0.2442, "step": 11417 }, { "epoch": 5.373176470588235, "grad_norm": 0.3897549897883742, "learning_rate": 1.1221776837042345e-05, "loss": 0.1717, "step": 11418 }, { "epoch": 5.373647058823529, "grad_norm": 0.2832774593979208, "learning_rate": 1.1205146551594357e-05, "loss": 0.1072, "step": 11419 }, { "epoch": 5.374117647058823, "grad_norm": 0.2935233787727932, "learning_rate": 1.1188528119907642e-05, "loss": 0.1051, "step": 11420 }, { "epoch": 5.374588235294118, "grad_norm": 0.3814789075251805, "learning_rate": 1.117192154340148e-05, "loss": 0.2127, "step": 11421 }, { "epoch": 5.375058823529412, "grad_norm": 0.2067691027972098, "learning_rate": 1.115532682349417e-05, "loss": 0.0382, "step": 11422 }, { "epoch": 5.375529411764706, "grad_norm": 0.29798867488567865, "learning_rate": 1.1138743961602998e-05, "loss": 0.1543, "step": 11423 }, { "epoch": 5.376, "grad_norm": 0.455194270652013, "learning_rate": 1.1122172959144177e-05, "loss": 0.2603, "step": 11424 }, { "epoch": 5.376470588235295, "grad_norm": 0.3918974136403947, "learning_rate": 1.1105613817532976e-05, "loss": 0.2001, "step": 11425 }, { "epoch": 5.376941176470588, "grad_norm": 0.6864747414755038, "learning_rate": 1.108906653818365e-05, "loss": 0.3983, "step": 11426 }, { "epoch": 5.377411764705882, "grad_norm": 0.5081863258466177, "learning_rate": 1.1072531122509381e-05, "loss": 0.3288, "step": 11427 }, { "epoch": 5.377882352941176, "grad_norm": 0.38098247471298236, "learning_rate": 1.1056007571922377e-05, "loss": 0.1909, "step": 11428 }, { "epoch": 5.378352941176471, "grad_norm": 0.44160039495731435, "learning_rate": 1.1039495887833872e-05, "loss": 0.3307, "step": 11429 }, { "epoch": 5.378823529411765, "grad_norm": 0.3433593363890336, "learning_rate": 1.1022996071653988e-05, "loss": 0.1489, "step": 11430 }, { "epoch": 5.379294117647059, "grad_norm": 0.3193762329210564, "learning_rate": 1.1006508124791935e-05, "loss": 0.0952, "step": 11431 }, { "epoch": 5.379764705882353, "grad_norm": 0.2571134189819364, "learning_rate": 1.099003204865585e-05, "loss": 0.096, "step": 11432 }, { "epoch": 5.3802352941176474, "grad_norm": 0.43674161352859353, "learning_rate": 1.0973567844652858e-05, "loss": 0.1407, "step": 11433 }, { "epoch": 5.380705882352941, "grad_norm": 0.3856825923005027, "learning_rate": 1.0957115514189086e-05, "loss": 0.1985, "step": 11434 }, { "epoch": 5.381176470588235, "grad_norm": 0.5343145450999716, "learning_rate": 1.0940675058669673e-05, "loss": 0.2488, "step": 11435 }, { "epoch": 5.381647058823529, "grad_norm": 0.4727111823411054, "learning_rate": 1.0924246479498682e-05, "loss": 0.1849, "step": 11436 }, { "epoch": 5.3821176470588235, "grad_norm": 0.6473030629291713, "learning_rate": 1.0907829778079207e-05, "loss": 0.4081, "step": 11437 }, { "epoch": 5.382588235294118, "grad_norm": 0.4440375410911387, "learning_rate": 1.089142495581331e-05, "loss": 0.2113, "step": 11438 }, { "epoch": 5.383058823529412, "grad_norm": 0.2502487212574838, "learning_rate": 1.0875032014102036e-05, "loss": 0.1065, "step": 11439 }, { "epoch": 5.383529411764706, "grad_norm": 0.38118427870288435, "learning_rate": 1.0858650954345433e-05, "loss": 0.1838, "step": 11440 }, { "epoch": 5.384, "grad_norm": 0.3473350854241585, "learning_rate": 1.0842281777942514e-05, "loss": 0.1526, "step": 11441 }, { "epoch": 5.384470588235294, "grad_norm": 0.5787246844255347, "learning_rate": 1.082592448629131e-05, "loss": 0.2619, "step": 11442 }, { "epoch": 5.384941176470588, "grad_norm": 0.2695019030026734, "learning_rate": 1.0809579080788805e-05, "loss": 0.0646, "step": 11443 }, { "epoch": 5.385411764705882, "grad_norm": 0.26134272767281647, "learning_rate": 1.0793245562830966e-05, "loss": 0.0799, "step": 11444 }, { "epoch": 5.385882352941176, "grad_norm": 0.2753941661672308, "learning_rate": 1.0776923933812775e-05, "loss": 0.1364, "step": 11445 }, { "epoch": 5.386352941176471, "grad_norm": 0.3702410767406176, "learning_rate": 1.0760614195128153e-05, "loss": 0.2062, "step": 11446 }, { "epoch": 5.386823529411765, "grad_norm": 0.2849006042281679, "learning_rate": 1.0744316348170035e-05, "loss": 0.0773, "step": 11447 }, { "epoch": 5.387294117647059, "grad_norm": 0.3727655321683575, "learning_rate": 1.0728030394330373e-05, "loss": 0.1378, "step": 11448 }, { "epoch": 5.387764705882353, "grad_norm": 0.7729351611952241, "learning_rate": 1.0711756335000005e-05, "loss": 0.2824, "step": 11449 }, { "epoch": 5.3882352941176475, "grad_norm": 0.5377203100730279, "learning_rate": 1.069549417156887e-05, "loss": 0.2172, "step": 11450 }, { "epoch": 5.388705882352941, "grad_norm": 0.4212936695979358, "learning_rate": 1.0679243905425822e-05, "loss": 0.2165, "step": 11451 }, { "epoch": 5.389176470588235, "grad_norm": 0.5054398550967174, "learning_rate": 1.0663005537958702e-05, "loss": 0.1493, "step": 11452 }, { "epoch": 5.389647058823529, "grad_norm": 1.0367530545979027, "learning_rate": 1.0646779070554335e-05, "loss": 0.5952, "step": 11453 }, { "epoch": 5.3901176470588235, "grad_norm": 0.1754388434536575, "learning_rate": 1.0630564504598593e-05, "loss": 0.0266, "step": 11454 }, { "epoch": 5.390588235294118, "grad_norm": 0.3289981918666486, "learning_rate": 1.0614361841476238e-05, "loss": 0.0754, "step": 11455 }, { "epoch": 5.391058823529412, "grad_norm": 0.5221381887215795, "learning_rate": 1.0598171082571044e-05, "loss": 0.3553, "step": 11456 }, { "epoch": 5.391529411764706, "grad_norm": 0.4383448544217741, "learning_rate": 1.0581992229265823e-05, "loss": 0.1579, "step": 11457 }, { "epoch": 5.392, "grad_norm": 0.42476875204548675, "learning_rate": 1.0565825282942303e-05, "loss": 0.1721, "step": 11458 }, { "epoch": 5.392470588235295, "grad_norm": 0.4883160386537151, "learning_rate": 1.0549670244981211e-05, "loss": 0.3041, "step": 11459 }, { "epoch": 5.392941176470588, "grad_norm": 0.4140429210897304, "learning_rate": 1.0533527116762296e-05, "loss": 0.203, "step": 11460 }, { "epoch": 5.393411764705882, "grad_norm": 0.5812321816852747, "learning_rate": 1.0517395899664239e-05, "loss": 0.3797, "step": 11461 }, { "epoch": 5.393882352941176, "grad_norm": 0.35225316625393305, "learning_rate": 1.0501276595064716e-05, "loss": 0.1161, "step": 11462 }, { "epoch": 5.394352941176471, "grad_norm": 0.5731545430039672, "learning_rate": 1.0485169204340432e-05, "loss": 0.3413, "step": 11463 }, { "epoch": 5.394823529411765, "grad_norm": 0.5672407811630872, "learning_rate": 1.0469073728867e-05, "loss": 0.2618, "step": 11464 }, { "epoch": 5.395294117647059, "grad_norm": 0.15912463942956823, "learning_rate": 1.0452990170019088e-05, "loss": 0.0241, "step": 11465 }, { "epoch": 5.395764705882353, "grad_norm": 0.41839839365454146, "learning_rate": 1.0436918529170264e-05, "loss": 0.2094, "step": 11466 }, { "epoch": 5.3962352941176475, "grad_norm": 0.40448153816968796, "learning_rate": 1.0420858807693162e-05, "loss": 0.1937, "step": 11467 }, { "epoch": 5.396705882352941, "grad_norm": 0.16765880991327128, "learning_rate": 1.0404811006959352e-05, "loss": 0.0369, "step": 11468 }, { "epoch": 5.397176470588235, "grad_norm": 0.42219763677448935, "learning_rate": 1.0388775128339389e-05, "loss": 0.161, "step": 11469 }, { "epoch": 5.397647058823529, "grad_norm": 0.3552081446577601, "learning_rate": 1.037275117320281e-05, "loss": 0.1805, "step": 11470 }, { "epoch": 5.3981176470588235, "grad_norm": 0.396244863468822, "learning_rate": 1.0356739142918169e-05, "loss": 0.1701, "step": 11471 }, { "epoch": 5.398588235294118, "grad_norm": 0.4768250139064371, "learning_rate": 1.0340739038852924e-05, "loss": 0.2382, "step": 11472 }, { "epoch": 5.399058823529412, "grad_norm": 0.6835342667355901, "learning_rate": 1.0324750862373598e-05, "loss": 0.5087, "step": 11473 }, { "epoch": 5.399529411764706, "grad_norm": 0.2727398530445805, "learning_rate": 1.0308774614845683e-05, "loss": 0.0964, "step": 11474 }, { "epoch": 5.4, "grad_norm": 0.49176010046595886, "learning_rate": 1.0292810297633569e-05, "loss": 0.2088, "step": 11475 }, { "epoch": 5.400470588235294, "grad_norm": 0.2714229180958639, "learning_rate": 1.02768579121007e-05, "loss": 0.0557, "step": 11476 }, { "epoch": 5.400941176470588, "grad_norm": 0.3310413184489143, "learning_rate": 1.0260917459609535e-05, "loss": 0.1372, "step": 11477 }, { "epoch": 5.401411764705882, "grad_norm": 0.43564364872142197, "learning_rate": 1.024498894152142e-05, "loss": 0.1992, "step": 11478 }, { "epoch": 5.401882352941176, "grad_norm": 0.6719158396323927, "learning_rate": 1.0229072359196749e-05, "loss": 0.3422, "step": 11479 }, { "epoch": 5.402352941176471, "grad_norm": 0.40788273414754433, "learning_rate": 1.0213167713994885e-05, "loss": 0.1716, "step": 11480 }, { "epoch": 5.402823529411765, "grad_norm": 0.6662392976088948, "learning_rate": 1.0197275007274126e-05, "loss": 0.4188, "step": 11481 }, { "epoch": 5.403294117647059, "grad_norm": 0.2598721450904264, "learning_rate": 1.0181394240391798e-05, "loss": 0.0953, "step": 11482 }, { "epoch": 5.403764705882353, "grad_norm": 0.424099039453818, "learning_rate": 1.016552541470424e-05, "loss": 0.1751, "step": 11483 }, { "epoch": 5.4042352941176475, "grad_norm": 0.30392297311918964, "learning_rate": 1.0149668531566662e-05, "loss": 0.0352, "step": 11484 }, { "epoch": 5.404705882352941, "grad_norm": 0.6501255136675534, "learning_rate": 1.0133823592333385e-05, "loss": 0.533, "step": 11485 }, { "epoch": 5.405176470588235, "grad_norm": 0.5031741000930973, "learning_rate": 1.0117990598357589e-05, "loss": 0.2585, "step": 11486 }, { "epoch": 5.405647058823529, "grad_norm": 0.5388334684152261, "learning_rate": 1.0102169550991529e-05, "loss": 0.184, "step": 11487 }, { "epoch": 5.4061176470588235, "grad_norm": 0.3667398850342157, "learning_rate": 1.008636045158639e-05, "loss": 0.1408, "step": 11488 }, { "epoch": 5.406588235294118, "grad_norm": 0.7736862345867288, "learning_rate": 1.0070563301492323e-05, "loss": 0.0755, "step": 11489 }, { "epoch": 5.407058823529412, "grad_norm": 0.4215749611539361, "learning_rate": 1.0054778102058502e-05, "loss": 0.2335, "step": 11490 }, { "epoch": 5.407529411764706, "grad_norm": 0.38939449774792123, "learning_rate": 1.0039004854633082e-05, "loss": 0.172, "step": 11491 }, { "epoch": 5.408, "grad_norm": 0.22917574403189808, "learning_rate": 1.0023243560563132e-05, "loss": 0.0502, "step": 11492 }, { "epoch": 5.408470588235295, "grad_norm": 0.37522993989682063, "learning_rate": 1.0007494221194762e-05, "loss": 0.1626, "step": 11493 }, { "epoch": 5.408941176470588, "grad_norm": 0.441405638106572, "learning_rate": 9.991756837873061e-06, "loss": 0.327, "step": 11494 }, { "epoch": 5.409411764705882, "grad_norm": 0.3839918199089626, "learning_rate": 9.976031411942054e-06, "loss": 0.1446, "step": 11495 }, { "epoch": 5.409882352941176, "grad_norm": 0.33400651595711, "learning_rate": 9.960317944744783e-06, "loss": 0.152, "step": 11496 }, { "epoch": 5.410352941176471, "grad_norm": 0.7425085056961243, "learning_rate": 9.944616437623243e-06, "loss": 0.355, "step": 11497 }, { "epoch": 5.410823529411765, "grad_norm": 0.27659846250149955, "learning_rate": 9.928926891918426e-06, "loss": 0.1019, "step": 11498 }, { "epoch": 5.411294117647059, "grad_norm": 0.3609579876220128, "learning_rate": 9.913249308970312e-06, "loss": 0.1707, "step": 11499 }, { "epoch": 5.411764705882353, "grad_norm": 0.7466093327028382, "learning_rate": 9.89758369011781e-06, "loss": 0.1735, "step": 11500 }, { "epoch": 5.412235294117647, "grad_norm": 0.5365864998943557, "learning_rate": 9.881930036698854e-06, "loss": 0.2992, "step": 11501 }, { "epoch": 5.412705882352941, "grad_norm": 0.49074291638652584, "learning_rate": 9.866288350050372e-06, "loss": 0.2383, "step": 11502 }, { "epoch": 5.413176470588235, "grad_norm": 0.44907792569631805, "learning_rate": 9.850658631508196e-06, "loss": 0.2855, "step": 11503 }, { "epoch": 5.413647058823529, "grad_norm": 0.3464557405299405, "learning_rate": 9.835040882407207e-06, "loss": 0.0833, "step": 11504 }, { "epoch": 5.4141176470588235, "grad_norm": 0.3596643219237782, "learning_rate": 9.819435104081242e-06, "loss": 0.1775, "step": 11505 }, { "epoch": 5.414588235294118, "grad_norm": 0.4117297513042804, "learning_rate": 9.803841297863096e-06, "loss": 0.2085, "step": 11506 }, { "epoch": 5.415058823529412, "grad_norm": 0.7236871864272446, "learning_rate": 9.788259465084541e-06, "loss": 0.395, "step": 11507 }, { "epoch": 5.415529411764706, "grad_norm": 0.25764110620257263, "learning_rate": 9.772689607076394e-06, "loss": 0.0711, "step": 11508 }, { "epoch": 5.416, "grad_norm": 0.7714558303412349, "learning_rate": 9.757131725168343e-06, "loss": 0.487, "step": 11509 }, { "epoch": 5.416470588235294, "grad_norm": 0.1987193495690625, "learning_rate": 9.74158582068914e-06, "loss": 0.0489, "step": 11510 }, { "epoch": 5.416941176470588, "grad_norm": 0.5025509764475019, "learning_rate": 9.726051894966492e-06, "loss": 0.144, "step": 11511 }, { "epoch": 5.417411764705882, "grad_norm": 0.6195019890922425, "learning_rate": 9.710529949327018e-06, "loss": 0.246, "step": 11512 }, { "epoch": 5.417882352941176, "grad_norm": 0.4338400074983988, "learning_rate": 9.695019985096409e-06, "loss": 0.2448, "step": 11513 }, { "epoch": 5.418352941176471, "grad_norm": 0.1097654540439061, "learning_rate": 9.679522003599305e-06, "loss": 0.0189, "step": 11514 }, { "epoch": 5.418823529411765, "grad_norm": 0.35923537257868227, "learning_rate": 9.664036006159266e-06, "loss": 0.1426, "step": 11515 }, { "epoch": 5.419294117647059, "grad_norm": 0.24704134050713386, "learning_rate": 9.648561994098913e-06, "loss": 0.0597, "step": 11516 }, { "epoch": 5.419764705882353, "grad_norm": 0.3220053904024125, "learning_rate": 9.633099968739777e-06, "loss": 0.085, "step": 11517 }, { "epoch": 5.4202352941176475, "grad_norm": 0.3014911450191627, "learning_rate": 9.617649931402399e-06, "loss": 0.1178, "step": 11518 }, { "epoch": 5.420705882352941, "grad_norm": 0.48461544861156824, "learning_rate": 9.602211883406308e-06, "loss": 0.2124, "step": 11519 }, { "epoch": 5.421176470588235, "grad_norm": 0.08502004163670222, "learning_rate": 9.586785826069932e-06, "loss": 0.0163, "step": 11520 }, { "epoch": 5.421647058823529, "grad_norm": 0.358815224190557, "learning_rate": 9.571371760710783e-06, "loss": 0.1524, "step": 11521 }, { "epoch": 5.4221176470588235, "grad_norm": 0.1779638803402008, "learning_rate": 9.55596968864531e-06, "loss": 0.0474, "step": 11522 }, { "epoch": 5.422588235294118, "grad_norm": 0.47587758833461435, "learning_rate": 9.540579611188875e-06, "loss": 0.2596, "step": 11523 }, { "epoch": 5.423058823529412, "grad_norm": 0.36510881283418845, "learning_rate": 9.52520152965588e-06, "loss": 0.1559, "step": 11524 }, { "epoch": 5.423529411764706, "grad_norm": 0.2833071608861282, "learning_rate": 9.509835445359736e-06, "loss": 0.0906, "step": 11525 }, { "epoch": 5.424, "grad_norm": 0.3402730839952212, "learning_rate": 9.494481359612715e-06, "loss": 0.1308, "step": 11526 }, { "epoch": 5.424470588235294, "grad_norm": 0.39435117375086215, "learning_rate": 9.479139273726166e-06, "loss": 0.1632, "step": 11527 }, { "epoch": 5.424941176470588, "grad_norm": 0.7974079650560117, "learning_rate": 9.46380918901039e-06, "loss": 0.5508, "step": 11528 }, { "epoch": 5.425411764705882, "grad_norm": 0.472899425850491, "learning_rate": 9.448491106774591e-06, "loss": 0.2104, "step": 11529 }, { "epoch": 5.425882352941176, "grad_norm": 0.406673022756616, "learning_rate": 9.433185028327106e-06, "loss": 0.2083, "step": 11530 }, { "epoch": 5.426352941176471, "grad_norm": 0.29505098646775757, "learning_rate": 9.417890954975071e-06, "loss": 0.1204, "step": 11531 }, { "epoch": 5.426823529411765, "grad_norm": 0.42055842421656514, "learning_rate": 9.402608888024692e-06, "loss": 0.1948, "step": 11532 }, { "epoch": 5.427294117647059, "grad_norm": 0.26025639519458, "learning_rate": 9.38733882878116e-06, "loss": 0.1291, "step": 11533 }, { "epoch": 5.427764705882353, "grad_norm": 0.473091343123325, "learning_rate": 9.37208077854858e-06, "loss": 0.1369, "step": 11534 }, { "epoch": 5.428235294117647, "grad_norm": 0.4401926283520227, "learning_rate": 9.356834738630075e-06, "loss": 0.2238, "step": 11535 }, { "epoch": 5.428705882352941, "grad_norm": 0.7095646003780622, "learning_rate": 9.341600710327773e-06, "loss": 0.1505, "step": 11536 }, { "epoch": 5.429176470588235, "grad_norm": 0.26159557546875617, "learning_rate": 9.32637869494265e-06, "loss": 0.0853, "step": 11537 }, { "epoch": 5.429647058823529, "grad_norm": 0.38543085037731223, "learning_rate": 9.311168693774796e-06, "loss": 0.1818, "step": 11538 }, { "epoch": 5.4301176470588235, "grad_norm": 0.43180906789038775, "learning_rate": 9.295970708123223e-06, "loss": 0.1231, "step": 11539 }, { "epoch": 5.430588235294118, "grad_norm": 0.4272436527046367, "learning_rate": 9.280784739285896e-06, "loss": 0.1731, "step": 11540 }, { "epoch": 5.431058823529412, "grad_norm": 0.21381235907575363, "learning_rate": 9.265610788559758e-06, "loss": 0.0534, "step": 11541 }, { "epoch": 5.431529411764706, "grad_norm": 0.39718970807615556, "learning_rate": 9.250448857240789e-06, "loss": 0.1458, "step": 11542 }, { "epoch": 5.432, "grad_norm": 0.3981052093031816, "learning_rate": 9.235298946623838e-06, "loss": 0.1387, "step": 11543 }, { "epoch": 5.432470588235294, "grad_norm": 0.5784309100552367, "learning_rate": 9.220161058002834e-06, "loss": 0.2925, "step": 11544 }, { "epoch": 5.432941176470588, "grad_norm": 0.7804388771216639, "learning_rate": 9.205035192670563e-06, "loss": 0.4954, "step": 11545 }, { "epoch": 5.433411764705882, "grad_norm": 0.38124753763136293, "learning_rate": 9.189921351918888e-06, "loss": 0.1999, "step": 11546 }, { "epoch": 5.433882352941176, "grad_norm": 0.714699873149776, "learning_rate": 9.174819537038614e-06, "loss": 0.4112, "step": 11547 }, { "epoch": 5.434352941176471, "grad_norm": 0.5530635329348575, "learning_rate": 9.159729749319473e-06, "loss": 0.2956, "step": 11548 }, { "epoch": 5.434823529411765, "grad_norm": 0.9237899299817034, "learning_rate": 9.144651990050234e-06, "loss": 0.5552, "step": 11549 }, { "epoch": 5.435294117647059, "grad_norm": 0.4868828809054107, "learning_rate": 9.129586260518634e-06, "loss": 0.3111, "step": 11550 }, { "epoch": 5.435764705882353, "grad_norm": 0.38962240463092124, "learning_rate": 9.114532562011312e-06, "loss": 0.1091, "step": 11551 }, { "epoch": 5.4362352941176475, "grad_norm": 0.4622823823957066, "learning_rate": 9.099490895813938e-06, "loss": 0.2173, "step": 11552 }, { "epoch": 5.436705882352941, "grad_norm": 0.2125313676904935, "learning_rate": 9.084461263211168e-06, "loss": 0.0384, "step": 11553 }, { "epoch": 5.437176470588235, "grad_norm": 0.3844981061935216, "learning_rate": 9.069443665486593e-06, "loss": 0.2217, "step": 11554 }, { "epoch": 5.437647058823529, "grad_norm": 0.49138135529992266, "learning_rate": 9.054438103922767e-06, "loss": 0.0682, "step": 11555 }, { "epoch": 5.4381176470588235, "grad_norm": 0.21063263101849444, "learning_rate": 9.039444579801302e-06, "loss": 0.0405, "step": 11556 }, { "epoch": 5.438588235294118, "grad_norm": 0.359881368375663, "learning_rate": 9.024463094402656e-06, "loss": 0.117, "step": 11557 }, { "epoch": 5.439058823529412, "grad_norm": 0.2877403930669461, "learning_rate": 9.009493649006322e-06, "loss": 0.1072, "step": 11558 }, { "epoch": 5.439529411764706, "grad_norm": 0.3050864802039857, "learning_rate": 8.994536244890827e-06, "loss": 0.0668, "step": 11559 }, { "epoch": 5.44, "grad_norm": 0.6905097069202617, "learning_rate": 8.979590883333532e-06, "loss": 0.1899, "step": 11560 }, { "epoch": 5.440470588235294, "grad_norm": 0.4733629518830766, "learning_rate": 8.964657565610884e-06, "loss": 0.1488, "step": 11561 }, { "epoch": 5.440941176470588, "grad_norm": 0.25204643553425643, "learning_rate": 8.949736292998277e-06, "loss": 0.0467, "step": 11562 }, { "epoch": 5.441411764705882, "grad_norm": 0.40001621132013787, "learning_rate": 8.934827066770012e-06, "loss": 0.111, "step": 11563 }, { "epoch": 5.441882352941176, "grad_norm": 0.33239356422551014, "learning_rate": 8.919929888199435e-06, "loss": 0.1818, "step": 11564 }, { "epoch": 5.442352941176471, "grad_norm": 0.29073080431014026, "learning_rate": 8.90504475855886e-06, "loss": 0.0728, "step": 11565 }, { "epoch": 5.442823529411765, "grad_norm": 0.3862414360332829, "learning_rate": 8.890171679119506e-06, "loss": 0.204, "step": 11566 }, { "epoch": 5.443294117647059, "grad_norm": 0.4851780783877553, "learning_rate": 8.875310651151657e-06, "loss": 0.3383, "step": 11567 }, { "epoch": 5.443764705882353, "grad_norm": 0.5215579410900527, "learning_rate": 8.86046167592448e-06, "loss": 0.2036, "step": 11568 }, { "epoch": 5.444235294117647, "grad_norm": 0.7295667147615025, "learning_rate": 8.845624754706144e-06, "loss": 0.3769, "step": 11569 }, { "epoch": 5.444705882352941, "grad_norm": 0.36830684659415786, "learning_rate": 8.830799888763834e-06, "loss": 0.1517, "step": 11570 }, { "epoch": 5.445176470588235, "grad_norm": 0.3867099199172953, "learning_rate": 8.81598707936364e-06, "loss": 0.1699, "step": 11571 }, { "epoch": 5.445647058823529, "grad_norm": 0.35523181498995476, "learning_rate": 8.801186327770632e-06, "loss": 0.1655, "step": 11572 }, { "epoch": 5.4461176470588235, "grad_norm": 0.37300919088139256, "learning_rate": 8.7863976352489e-06, "loss": 0.2061, "step": 11573 }, { "epoch": 5.446588235294118, "grad_norm": 0.5670852496272624, "learning_rate": 8.771621003061463e-06, "loss": 0.1319, "step": 11574 }, { "epoch": 5.447058823529412, "grad_norm": 0.3673883947640132, "learning_rate": 8.756856432470315e-06, "loss": 0.2058, "step": 11575 }, { "epoch": 5.447529411764706, "grad_norm": 0.501271978322239, "learning_rate": 8.742103924736416e-06, "loss": 0.2298, "step": 11576 }, { "epoch": 5.448, "grad_norm": 0.387840101613679, "learning_rate": 8.727363481119704e-06, "loss": 0.2172, "step": 11577 }, { "epoch": 5.448470588235294, "grad_norm": 0.40049539609985124, "learning_rate": 8.712635102879124e-06, "loss": 0.2268, "step": 11578 }, { "epoch": 5.448941176470588, "grad_norm": 0.35614399247834905, "learning_rate": 8.697918791272485e-06, "loss": 0.1833, "step": 11579 }, { "epoch": 5.449411764705882, "grad_norm": 0.42626959664975067, "learning_rate": 8.68321454755665e-06, "loss": 0.1719, "step": 11580 }, { "epoch": 5.449882352941176, "grad_norm": 0.2932536771859773, "learning_rate": 8.668522372987479e-06, "loss": 0.1159, "step": 11581 }, { "epoch": 5.450352941176471, "grad_norm": 0.3401350387473817, "learning_rate": 8.653842268819723e-06, "loss": 0.0596, "step": 11582 }, { "epoch": 5.450823529411765, "grad_norm": 0.28765928337900365, "learning_rate": 8.639174236307106e-06, "loss": 0.0675, "step": 11583 }, { "epoch": 5.451294117647059, "grad_norm": 0.48733242629529633, "learning_rate": 8.624518276702414e-06, "loss": 0.186, "step": 11584 }, { "epoch": 5.451764705882353, "grad_norm": 0.41881178611733677, "learning_rate": 8.609874391257293e-06, "loss": 0.2168, "step": 11585 }, { "epoch": 5.4522352941176475, "grad_norm": 0.463582892625349, "learning_rate": 8.595242581222394e-06, "loss": 0.1361, "step": 11586 }, { "epoch": 5.452705882352941, "grad_norm": 0.4062103127147958, "learning_rate": 8.580622847847402e-06, "loss": 0.1003, "step": 11587 }, { "epoch": 5.453176470588235, "grad_norm": 0.4731149340293153, "learning_rate": 8.566015192380849e-06, "loss": 0.1324, "step": 11588 }, { "epoch": 5.453647058823529, "grad_norm": 0.45762460807571026, "learning_rate": 8.551419616070322e-06, "loss": 0.1975, "step": 11589 }, { "epoch": 5.4541176470588235, "grad_norm": 0.686247416511546, "learning_rate": 8.536836120162389e-06, "loss": 0.3607, "step": 11590 }, { "epoch": 5.454588235294118, "grad_norm": 0.4348741117071093, "learning_rate": 8.522264705902504e-06, "loss": 0.29, "step": 11591 }, { "epoch": 5.455058823529412, "grad_norm": 0.32620514272239887, "learning_rate": 8.507705374535157e-06, "loss": 0.1666, "step": 11592 }, { "epoch": 5.455529411764706, "grad_norm": 0.32602842682745353, "learning_rate": 8.4931581273038e-06, "loss": 0.1489, "step": 11593 }, { "epoch": 5.456, "grad_norm": 0.650643321436358, "learning_rate": 8.47862296545081e-06, "loss": 0.2343, "step": 11594 }, { "epoch": 5.456470588235294, "grad_norm": 0.4381600340671323, "learning_rate": 8.464099890217607e-06, "loss": 0.2337, "step": 11595 }, { "epoch": 5.456941176470588, "grad_norm": 0.42636551802523776, "learning_rate": 8.449588902844484e-06, "loss": 0.2494, "step": 11596 }, { "epoch": 5.457411764705882, "grad_norm": 0.3266235506584944, "learning_rate": 8.43509000457075e-06, "loss": 0.1241, "step": 11597 }, { "epoch": 5.457882352941176, "grad_norm": 0.3210993412555706, "learning_rate": 8.420603196634745e-06, "loss": 0.0957, "step": 11598 }, { "epoch": 5.458352941176471, "grad_norm": 0.31991857511188, "learning_rate": 8.406128480273633e-06, "loss": 0.124, "step": 11599 }, { "epoch": 5.458823529411765, "grad_norm": 0.3913537053947848, "learning_rate": 8.391665856723655e-06, "loss": 0.1619, "step": 11600 }, { "epoch": 5.459294117647059, "grad_norm": 0.4012608884293867, "learning_rate": 8.377215327220043e-06, "loss": 0.2321, "step": 11601 }, { "epoch": 5.459764705882353, "grad_norm": 0.8159730785623048, "learning_rate": 8.362776892996858e-06, "loss": 0.4247, "step": 11602 }, { "epoch": 5.460235294117647, "grad_norm": 0.4870937496241432, "learning_rate": 8.348350555287264e-06, "loss": 0.2858, "step": 11603 }, { "epoch": 5.460705882352941, "grad_norm": 0.30799366457979827, "learning_rate": 8.333936315323343e-06, "loss": 0.1155, "step": 11604 }, { "epoch": 5.461176470588235, "grad_norm": 0.23034433202161247, "learning_rate": 8.319534174336106e-06, "loss": 0.0737, "step": 11605 }, { "epoch": 5.461647058823529, "grad_norm": 0.37363783945583867, "learning_rate": 8.305144133555608e-06, "loss": 0.1912, "step": 11606 }, { "epoch": 5.4621176470588235, "grad_norm": 0.46942401543878315, "learning_rate": 8.290766194210813e-06, "loss": 0.2417, "step": 11607 }, { "epoch": 5.462588235294118, "grad_norm": 0.5432082877510515, "learning_rate": 8.276400357529655e-06, "loss": 0.322, "step": 11608 }, { "epoch": 5.463058823529412, "grad_norm": 0.4201891098192555, "learning_rate": 8.262046624739038e-06, "loss": 0.1923, "step": 11609 }, { "epoch": 5.463529411764706, "grad_norm": 0.4564048181909712, "learning_rate": 8.247704997064896e-06, "loss": 0.1418, "step": 11610 }, { "epoch": 5.464, "grad_norm": 0.1936766137415025, "learning_rate": 8.233375475732018e-06, "loss": 0.055, "step": 11611 }, { "epoch": 5.464470588235294, "grad_norm": 0.4532055186247031, "learning_rate": 8.21905806196424e-06, "loss": 0.3245, "step": 11612 }, { "epoch": 5.464941176470588, "grad_norm": 0.09553556707470945, "learning_rate": 8.204752756984351e-06, "loss": 0.0161, "step": 11613 }, { "epoch": 5.465411764705882, "grad_norm": 0.4216075799051418, "learning_rate": 8.19045956201404e-06, "loss": 0.1682, "step": 11614 }, { "epoch": 5.465882352941176, "grad_norm": 0.6112434920269727, "learning_rate": 8.176178478274082e-06, "loss": 0.194, "step": 11615 }, { "epoch": 5.466352941176471, "grad_norm": 0.3000513615968212, "learning_rate": 8.161909506984132e-06, "loss": 0.135, "step": 11616 }, { "epoch": 5.466823529411765, "grad_norm": 0.40174988774898995, "learning_rate": 8.147652649362802e-06, "loss": 0.1396, "step": 11617 }, { "epoch": 5.467294117647059, "grad_norm": 0.45230641129911003, "learning_rate": 8.133407906627715e-06, "loss": 0.191, "step": 11618 }, { "epoch": 5.467764705882353, "grad_norm": 0.4290166509361052, "learning_rate": 8.119175279995448e-06, "loss": 0.1209, "step": 11619 }, { "epoch": 5.4682352941176475, "grad_norm": 0.16228966267555972, "learning_rate": 8.104954770681532e-06, "loss": 0.0305, "step": 11620 }, { "epoch": 5.468705882352941, "grad_norm": 0.3889213166489552, "learning_rate": 8.090746379900475e-06, "loss": 0.1744, "step": 11621 }, { "epoch": 5.469176470588235, "grad_norm": 0.2306051752851168, "learning_rate": 8.076550108865743e-06, "loss": 0.0497, "step": 11622 }, { "epoch": 5.469647058823529, "grad_norm": 0.4057546739783162, "learning_rate": 8.062365958789746e-06, "loss": 0.1775, "step": 11623 }, { "epoch": 5.4701176470588235, "grad_norm": 0.38746681521094584, "learning_rate": 8.048193930883933e-06, "loss": 0.1493, "step": 11624 }, { "epoch": 5.470588235294118, "grad_norm": 0.47142602297604574, "learning_rate": 8.034034026358587e-06, "loss": 0.2453, "step": 11625 }, { "epoch": 5.471058823529412, "grad_norm": 0.4854031008233041, "learning_rate": 8.019886246423107e-06, "loss": 0.2591, "step": 11626 }, { "epoch": 5.471529411764706, "grad_norm": 0.48568191416366324, "learning_rate": 8.005750592285725e-06, "loss": 0.2773, "step": 11627 }, { "epoch": 5.4719999999999995, "grad_norm": 0.5023842500670381, "learning_rate": 7.991627065153727e-06, "loss": 0.2666, "step": 11628 }, { "epoch": 5.472470588235294, "grad_norm": 0.3007902685427286, "learning_rate": 7.977515666233347e-06, "loss": 0.1006, "step": 11629 }, { "epoch": 5.472941176470588, "grad_norm": 0.5246629870876479, "learning_rate": 7.963416396729722e-06, "loss": 0.2429, "step": 11630 }, { "epoch": 5.473411764705882, "grad_norm": 0.47927440718429726, "learning_rate": 7.94932925784702e-06, "loss": 0.2034, "step": 11631 }, { "epoch": 5.473882352941176, "grad_norm": 0.27912613640939926, "learning_rate": 7.935254250788364e-06, "loss": 0.0532, "step": 11632 }, { "epoch": 5.474352941176471, "grad_norm": 0.5623789059545502, "learning_rate": 7.921191376755808e-06, "loss": 0.3253, "step": 11633 }, { "epoch": 5.474823529411765, "grad_norm": 0.2686991662145402, "learning_rate": 7.90714063695041e-06, "loss": 0.078, "step": 11634 }, { "epoch": 5.475294117647059, "grad_norm": 0.29079022280785605, "learning_rate": 7.893102032572174e-06, "loss": 0.0856, "step": 11635 }, { "epoch": 5.475764705882353, "grad_norm": 0.3891175131653843, "learning_rate": 7.879075564820025e-06, "loss": 0.1613, "step": 11636 }, { "epoch": 5.476235294117647, "grad_norm": 0.09794837836939513, "learning_rate": 7.86506123489194e-06, "loss": 0.0197, "step": 11637 }, { "epoch": 5.476705882352941, "grad_norm": 0.3860842930935366, "learning_rate": 7.85105904398481e-06, "loss": 0.1797, "step": 11638 }, { "epoch": 5.477176470588235, "grad_norm": 0.41847096924873384, "learning_rate": 7.837068993294443e-06, "loss": 0.1371, "step": 11639 }, { "epoch": 5.477647058823529, "grad_norm": 0.34047448382806605, "learning_rate": 7.823091084015687e-06, "loss": 0.1034, "step": 11640 }, { "epoch": 5.4781176470588235, "grad_norm": 0.41493715172247214, "learning_rate": 7.809125317342369e-06, "loss": 0.2628, "step": 11641 }, { "epoch": 5.478588235294118, "grad_norm": 0.5865594831504289, "learning_rate": 7.79517169446715e-06, "loss": 0.2814, "step": 11642 }, { "epoch": 5.479058823529412, "grad_norm": 0.27487137198186995, "learning_rate": 7.781230216581779e-06, "loss": 0.0553, "step": 11643 }, { "epoch": 5.479529411764706, "grad_norm": 0.29574947091450743, "learning_rate": 7.767300884876971e-06, "loss": 0.0589, "step": 11644 }, { "epoch": 5.48, "grad_norm": 0.1983684715535635, "learning_rate": 7.753383700542271e-06, "loss": 0.0317, "step": 11645 }, { "epoch": 5.480470588235294, "grad_norm": 0.21806867880989872, "learning_rate": 7.739478664766347e-06, "loss": 0.0521, "step": 11646 }, { "epoch": 5.480941176470588, "grad_norm": 0.25501646455007776, "learning_rate": 7.725585778736698e-06, "loss": 0.0558, "step": 11647 }, { "epoch": 5.481411764705882, "grad_norm": 0.43936553438333115, "learning_rate": 7.711705043639893e-06, "loss": 0.2779, "step": 11648 }, { "epoch": 5.481882352941176, "grad_norm": 0.4484560349266533, "learning_rate": 7.697836460661399e-06, "loss": 0.2785, "step": 11649 }, { "epoch": 5.482352941176471, "grad_norm": 0.3601315749460284, "learning_rate": 7.683980030985654e-06, "loss": 0.1503, "step": 11650 }, { "epoch": 5.482823529411765, "grad_norm": 0.344722830690345, "learning_rate": 7.670135755796074e-06, "loss": 0.1035, "step": 11651 }, { "epoch": 5.483294117647059, "grad_norm": 0.2491193742639546, "learning_rate": 7.656303636275036e-06, "loss": 0.0636, "step": 11652 }, { "epoch": 5.483764705882353, "grad_norm": 0.2089440076704305, "learning_rate": 7.64248367360384e-06, "loss": 0.038, "step": 11653 }, { "epoch": 5.484235294117647, "grad_norm": 0.5710930649830581, "learning_rate": 7.6286758689627915e-06, "loss": 0.3125, "step": 11654 }, { "epoch": 5.484705882352941, "grad_norm": 0.338563681117998, "learning_rate": 7.614880223531183e-06, "loss": 0.0783, "step": 11655 }, { "epoch": 5.485176470588235, "grad_norm": 0.543683973314488, "learning_rate": 7.6010967384871725e-06, "loss": 0.3023, "step": 11656 }, { "epoch": 5.485647058823529, "grad_norm": 0.24535349038944673, "learning_rate": 7.587325415007967e-06, "loss": 0.0794, "step": 11657 }, { "epoch": 5.4861176470588235, "grad_norm": 0.3765628943518027, "learning_rate": 7.573566254269708e-06, "loss": 0.1686, "step": 11658 }, { "epoch": 5.486588235294118, "grad_norm": 0.515427542369375, "learning_rate": 7.559819257447492e-06, "loss": 0.3093, "step": 11659 }, { "epoch": 5.487058823529412, "grad_norm": 0.4695290548876871, "learning_rate": 7.546084425715359e-06, "loss": 0.1985, "step": 11660 }, { "epoch": 5.487529411764706, "grad_norm": 0.3377188148188675, "learning_rate": 7.532361760246375e-06, "loss": 0.0937, "step": 11661 }, { "epoch": 5.4879999999999995, "grad_norm": 0.20890019042842142, "learning_rate": 7.5186512622124495e-06, "loss": 0.0471, "step": 11662 }, { "epoch": 5.488470588235294, "grad_norm": 0.15294980114844245, "learning_rate": 7.5049529327846135e-06, "loss": 0.0255, "step": 11663 }, { "epoch": 5.488941176470588, "grad_norm": 0.41065942638396824, "learning_rate": 7.491266773132715e-06, "loss": 0.2019, "step": 11664 }, { "epoch": 5.489411764705882, "grad_norm": 0.44270997046466687, "learning_rate": 7.477592784425634e-06, "loss": 0.134, "step": 11665 }, { "epoch": 5.489882352941176, "grad_norm": 0.30370070669947113, "learning_rate": 7.463930967831222e-06, "loss": 0.0823, "step": 11666 }, { "epoch": 5.490352941176471, "grad_norm": 0.20744155901149175, "learning_rate": 7.450281324516228e-06, "loss": 0.0743, "step": 11667 }, { "epoch": 5.490823529411765, "grad_norm": 0.3171890761297109, "learning_rate": 7.436643855646402e-06, "loss": 0.11, "step": 11668 }, { "epoch": 5.491294117647059, "grad_norm": 0.33351047952479995, "learning_rate": 7.423018562386479e-06, "loss": 0.0942, "step": 11669 }, { "epoch": 5.491764705882353, "grad_norm": 0.41898850883730626, "learning_rate": 7.40940544590008e-06, "loss": 0.1436, "step": 11670 }, { "epoch": 5.492235294117647, "grad_norm": 0.4296581316777872, "learning_rate": 7.395804507349873e-06, "loss": 0.1702, "step": 11671 }, { "epoch": 5.492705882352941, "grad_norm": 0.27790429267867733, "learning_rate": 7.3822157478974465e-06, "loss": 0.0738, "step": 11672 }, { "epoch": 5.493176470588235, "grad_norm": 0.4323997081199505, "learning_rate": 7.368639168703306e-06, "loss": 0.1909, "step": 11673 }, { "epoch": 5.493647058823529, "grad_norm": 0.652875291100111, "learning_rate": 7.3550747709269885e-06, "loss": 0.1519, "step": 11674 }, { "epoch": 5.4941176470588236, "grad_norm": 0.5847114832557503, "learning_rate": 7.34152255572697e-06, "loss": 0.3013, "step": 11675 }, { "epoch": 5.494588235294118, "grad_norm": 0.3032939569095036, "learning_rate": 7.327982524260639e-06, "loss": 0.1148, "step": 11676 }, { "epoch": 5.495058823529412, "grad_norm": 0.20932675335184686, "learning_rate": 7.314454677684439e-06, "loss": 0.0507, "step": 11677 }, { "epoch": 5.495529411764706, "grad_norm": 0.5322645385599301, "learning_rate": 7.300939017153645e-06, "loss": 0.3061, "step": 11678 }, { "epoch": 5.496, "grad_norm": 0.5006406661761742, "learning_rate": 7.287435543822584e-06, "loss": 0.2479, "step": 11679 }, { "epoch": 5.496470588235294, "grad_norm": 0.5261568061878747, "learning_rate": 7.2739442588445666e-06, "loss": 0.2553, "step": 11680 }, { "epoch": 5.496941176470588, "grad_norm": 0.36374510897233575, "learning_rate": 7.260465163371754e-06, "loss": 0.1014, "step": 11681 }, { "epoch": 5.497411764705882, "grad_norm": 0.5614214599374924, "learning_rate": 7.246998258555342e-06, "loss": 0.4052, "step": 11682 }, { "epoch": 5.497882352941176, "grad_norm": 0.7402028001984986, "learning_rate": 7.233543545545511e-06, "loss": 0.3179, "step": 11683 }, { "epoch": 5.498352941176471, "grad_norm": 0.5508474196902837, "learning_rate": 7.220101025491293e-06, "loss": 0.1785, "step": 11684 }, { "epoch": 5.498823529411765, "grad_norm": 1.8044385270614258, "learning_rate": 7.2066706995407845e-06, "loss": 0.3842, "step": 11685 }, { "epoch": 5.499294117647059, "grad_norm": 0.20380650007106518, "learning_rate": 7.193252568841018e-06, "loss": 0.0601, "step": 11686 }, { "epoch": 5.499764705882353, "grad_norm": 0.8162755356523704, "learning_rate": 7.1798466345379115e-06, "loss": 0.1751, "step": 11687 }, { "epoch": 5.5002352941176476, "grad_norm": 0.45044243987889104, "learning_rate": 7.166452897776448e-06, "loss": 0.2796, "step": 11688 }, { "epoch": 5.500705882352941, "grad_norm": 0.7872650890909677, "learning_rate": 7.153071359700513e-06, "loss": 0.2757, "step": 11689 }, { "epoch": 5.501176470588235, "grad_norm": 0.4310645039789697, "learning_rate": 7.1397020214529255e-06, "loss": 0.1966, "step": 11690 }, { "epoch": 5.501647058823529, "grad_norm": 0.5206831625697994, "learning_rate": 7.126344884175522e-06, "loss": 0.1746, "step": 11691 }, { "epoch": 5.502117647058824, "grad_norm": 0.39011916269200875, "learning_rate": 7.1129999490090735e-06, "loss": 0.1267, "step": 11692 }, { "epoch": 5.502588235294118, "grad_norm": 0.41641068208166987, "learning_rate": 7.099667217093252e-06, "loss": 0.2339, "step": 11693 }, { "epoch": 5.503058823529412, "grad_norm": 0.3706089465976132, "learning_rate": 7.08634668956678e-06, "loss": 0.1429, "step": 11694 }, { "epoch": 5.503529411764706, "grad_norm": 0.3561513624711835, "learning_rate": 7.073038367567313e-06, "loss": 0.1708, "step": 11695 }, { "epoch": 5.504, "grad_norm": 0.6385589917892722, "learning_rate": 7.059742252231393e-06, "loss": 0.3247, "step": 11696 }, { "epoch": 5.504470588235294, "grad_norm": 0.36708962525947, "learning_rate": 7.046458344694628e-06, "loss": 0.1723, "step": 11697 }, { "epoch": 5.504941176470588, "grad_norm": 0.6357333897154973, "learning_rate": 7.033186646091477e-06, "loss": 0.256, "step": 11698 }, { "epoch": 5.505411764705882, "grad_norm": 0.3898755500574676, "learning_rate": 7.019927157555449e-06, "loss": 0.1932, "step": 11699 }, { "epoch": 5.5058823529411764, "grad_norm": 0.7610133025571476, "learning_rate": 7.006679880218974e-06, "loss": 0.381, "step": 11700 }, { "epoch": 5.506352941176471, "grad_norm": 0.38468735419412964, "learning_rate": 6.993444815213378e-06, "loss": 0.0812, "step": 11701 }, { "epoch": 5.506823529411765, "grad_norm": 0.3584885852640074, "learning_rate": 6.9802219636690585e-06, "loss": 0.2028, "step": 11702 }, { "epoch": 5.507294117647059, "grad_norm": 0.7508466359146181, "learning_rate": 6.967011326715294e-06, "loss": 0.1535, "step": 11703 }, { "epoch": 5.507764705882353, "grad_norm": 0.37965447610724795, "learning_rate": 6.953812905480316e-06, "loss": 0.1425, "step": 11704 }, { "epoch": 5.508235294117647, "grad_norm": 0.2308165951089318, "learning_rate": 6.940626701091356e-06, "loss": 0.0537, "step": 11705 }, { "epoch": 5.508705882352941, "grad_norm": 0.5676205924884794, "learning_rate": 6.927452714674614e-06, "loss": 0.259, "step": 11706 }, { "epoch": 5.509176470588235, "grad_norm": 0.3747468376826811, "learning_rate": 6.914290947355122e-06, "loss": 0.1038, "step": 11707 }, { "epoch": 5.509647058823529, "grad_norm": 0.47064044985451664, "learning_rate": 6.901141400257048e-06, "loss": 0.2342, "step": 11708 }, { "epoch": 5.510117647058824, "grad_norm": 0.366599164277286, "learning_rate": 6.888004074503395e-06, "loss": 0.2069, "step": 11709 }, { "epoch": 5.510588235294118, "grad_norm": 0.309215701986677, "learning_rate": 6.87487897121613e-06, "loss": 0.1096, "step": 11710 }, { "epoch": 5.511058823529412, "grad_norm": 0.3944298525071958, "learning_rate": 6.861766091516274e-06, "loss": 0.0918, "step": 11711 }, { "epoch": 5.511529411764706, "grad_norm": 0.2632512036382365, "learning_rate": 6.848665436523631e-06, "loss": 0.0589, "step": 11712 }, { "epoch": 5.5120000000000005, "grad_norm": 0.3399426827609103, "learning_rate": 6.83557700735714e-06, "loss": 0.1151, "step": 11713 }, { "epoch": 5.512470588235294, "grad_norm": 0.458813529469794, "learning_rate": 6.82250080513459e-06, "loss": 0.2638, "step": 11714 }, { "epoch": 5.512941176470588, "grad_norm": 0.44365474878484523, "learning_rate": 6.809436830972753e-06, "loss": 0.1686, "step": 11715 }, { "epoch": 5.513411764705882, "grad_norm": 0.3372326726880487, "learning_rate": 6.796385085987338e-06, "loss": 0.056, "step": 11716 }, { "epoch": 5.5138823529411765, "grad_norm": 0.42633940502163115, "learning_rate": 6.783345571293069e-06, "loss": 0.1969, "step": 11717 }, { "epoch": 5.514352941176471, "grad_norm": 0.8493310589648145, "learning_rate": 6.770318288003556e-06, "loss": 0.3269, "step": 11718 }, { "epoch": 5.514823529411765, "grad_norm": 0.48501712257872304, "learning_rate": 6.757303237231393e-06, "loss": 0.2036, "step": 11719 }, { "epoch": 5.515294117647059, "grad_norm": 0.6039942323327708, "learning_rate": 6.744300420088156e-06, "loss": 0.3786, "step": 11720 }, { "epoch": 5.5157647058823525, "grad_norm": 0.5113228483537563, "learning_rate": 6.73130983768429e-06, "loss": 0.2527, "step": 11721 }, { "epoch": 5.516235294117647, "grad_norm": 0.20026069800310065, "learning_rate": 6.718331491129309e-06, "loss": 0.0499, "step": 11722 }, { "epoch": 5.516705882352941, "grad_norm": 0.333795490829832, "learning_rate": 6.705365381531641e-06, "loss": 0.1304, "step": 11723 }, { "epoch": 5.517176470588235, "grad_norm": 0.4002195315877828, "learning_rate": 6.6924115099985856e-06, "loss": 0.1159, "step": 11724 }, { "epoch": 5.517647058823529, "grad_norm": 0.23443083051795685, "learning_rate": 6.679469877636523e-06, "loss": 0.046, "step": 11725 }, { "epoch": 5.518117647058824, "grad_norm": 0.2132783377230856, "learning_rate": 6.6665404855507195e-06, "loss": 0.0616, "step": 11726 }, { "epoch": 5.518588235294118, "grad_norm": 0.7013190980913002, "learning_rate": 6.6536233348453925e-06, "loss": 0.2402, "step": 11727 }, { "epoch": 5.519058823529412, "grad_norm": 0.3366370885737166, "learning_rate": 6.640718426623759e-06, "loss": 0.0941, "step": 11728 }, { "epoch": 5.519529411764706, "grad_norm": 0.5059791902580918, "learning_rate": 6.627825761987937e-06, "loss": 0.2249, "step": 11729 }, { "epoch": 5.52, "grad_norm": 0.4004639252145831, "learning_rate": 6.61494534203903e-06, "loss": 0.1848, "step": 11730 }, { "epoch": 5.520470588235294, "grad_norm": 0.3992196072268141, "learning_rate": 6.6020771678771066e-06, "loss": 0.1007, "step": 11731 }, { "epoch": 5.520941176470588, "grad_norm": 0.39807519776125033, "learning_rate": 6.589221240601123e-06, "loss": 0.1161, "step": 11732 }, { "epoch": 5.521411764705882, "grad_norm": 0.5155283520969245, "learning_rate": 6.576377561309082e-06, "loss": 0.2241, "step": 11733 }, { "epoch": 5.5218823529411765, "grad_norm": 0.10136876719406757, "learning_rate": 6.563546131097907e-06, "loss": 0.012, "step": 11734 }, { "epoch": 5.522352941176471, "grad_norm": 0.5576744583403056, "learning_rate": 6.550726951063423e-06, "loss": 0.2814, "step": 11735 }, { "epoch": 5.522823529411765, "grad_norm": 0.32916986210450067, "learning_rate": 6.5379200223004855e-06, "loss": 0.1568, "step": 11736 }, { "epoch": 5.523294117647059, "grad_norm": 0.37091532686642464, "learning_rate": 6.5251253459028545e-06, "loss": 0.122, "step": 11737 }, { "epoch": 5.523764705882353, "grad_norm": 0.3697323325265708, "learning_rate": 6.512342922963254e-06, "loss": 0.1692, "step": 11738 }, { "epoch": 5.524235294117647, "grad_norm": 0.4929353215675222, "learning_rate": 6.499572754573362e-06, "loss": 0.1878, "step": 11739 }, { "epoch": 5.524705882352941, "grad_norm": 0.29788716508604224, "learning_rate": 6.48681484182384e-06, "loss": 0.1174, "step": 11740 }, { "epoch": 5.525176470588235, "grad_norm": 0.577846599909599, "learning_rate": 6.474069185804248e-06, "loss": 0.2544, "step": 11741 }, { "epoch": 5.525647058823529, "grad_norm": 0.32060669555727167, "learning_rate": 6.461335787603134e-06, "loss": 0.0901, "step": 11742 }, { "epoch": 5.526117647058824, "grad_norm": 0.5249909657407713, "learning_rate": 6.448614648308009e-06, "loss": 0.2506, "step": 11743 }, { "epoch": 5.526588235294118, "grad_norm": 0.5599371937980736, "learning_rate": 6.435905769005306e-06, "loss": 0.3709, "step": 11744 }, { "epoch": 5.527058823529412, "grad_norm": 0.4633351308749633, "learning_rate": 6.423209150780423e-06, "loss": 0.1518, "step": 11745 }, { "epoch": 5.527529411764706, "grad_norm": 0.28789378382103403, "learning_rate": 6.4105247947177256e-06, "loss": 0.0711, "step": 11746 }, { "epoch": 5.5280000000000005, "grad_norm": 0.2786682127577309, "learning_rate": 6.3978527019004964e-06, "loss": 0.0593, "step": 11747 }, { "epoch": 5.528470588235294, "grad_norm": 0.22550106387888225, "learning_rate": 6.385192873411038e-06, "loss": 0.0572, "step": 11748 }, { "epoch": 5.528941176470588, "grad_norm": 0.2721968915860003, "learning_rate": 6.372545310330518e-06, "loss": 0.0502, "step": 11749 }, { "epoch": 5.529411764705882, "grad_norm": 0.2753924228514459, "learning_rate": 6.359910013739122e-06, "loss": 0.1191, "step": 11750 }, { "epoch": 5.5298823529411765, "grad_norm": 0.4057659842827627, "learning_rate": 6.347286984715938e-06, "loss": 0.1134, "step": 11751 }, { "epoch": 5.530352941176471, "grad_norm": 0.3667677194451269, "learning_rate": 6.334676224339086e-06, "loss": 0.1773, "step": 11752 }, { "epoch": 5.530823529411765, "grad_norm": 0.3460806984059296, "learning_rate": 6.322077733685538e-06, "loss": 0.1112, "step": 11753 }, { "epoch": 5.531294117647059, "grad_norm": 0.5497867775653249, "learning_rate": 6.309491513831317e-06, "loss": 0.3203, "step": 11754 }, { "epoch": 5.5317647058823525, "grad_norm": 0.5898260748582497, "learning_rate": 6.296917565851295e-06, "loss": 0.3754, "step": 11755 }, { "epoch": 5.532235294117647, "grad_norm": 0.18185027270209114, "learning_rate": 6.284355890819398e-06, "loss": 0.0441, "step": 11756 }, { "epoch": 5.532705882352941, "grad_norm": 0.5303230625159979, "learning_rate": 6.271806489808401e-06, "loss": 0.3145, "step": 11757 }, { "epoch": 5.533176470588235, "grad_norm": 0.4217243638997182, "learning_rate": 6.2592693638901125e-06, "loss": 0.246, "step": 11758 }, { "epoch": 5.533647058823529, "grad_norm": 0.21796674469447946, "learning_rate": 6.246744514135293e-06, "loss": 0.0468, "step": 11759 }, { "epoch": 5.534117647058824, "grad_norm": 0.28431999504294714, "learning_rate": 6.234231941613571e-06, "loss": 0.0835, "step": 11760 }, { "epoch": 5.534588235294118, "grad_norm": 0.44864918743295573, "learning_rate": 6.221731647393607e-06, "loss": 0.2297, "step": 11761 }, { "epoch": 5.535058823529412, "grad_norm": 0.10685588528552646, "learning_rate": 6.209243632543015e-06, "loss": 0.015, "step": 11762 }, { "epoch": 5.535529411764706, "grad_norm": 1.021380433586945, "learning_rate": 6.196767898128291e-06, "loss": 0.3751, "step": 11763 }, { "epoch": 5.536, "grad_norm": 0.5299920928641174, "learning_rate": 6.184304445214932e-06, "loss": 0.207, "step": 11764 }, { "epoch": 5.536470588235294, "grad_norm": 0.30100814422523975, "learning_rate": 6.171853274867406e-06, "loss": 0.0715, "step": 11765 }, { "epoch": 5.536941176470588, "grad_norm": 0.3248844769223006, "learning_rate": 6.159414388149059e-06, "loss": 0.0611, "step": 11766 }, { "epoch": 5.537411764705882, "grad_norm": 0.41484188837327096, "learning_rate": 6.146987786122259e-06, "loss": 0.1217, "step": 11767 }, { "epoch": 5.5378823529411765, "grad_norm": 0.3510506591421233, "learning_rate": 6.134573469848325e-06, "loss": 0.0661, "step": 11768 }, { "epoch": 5.538352941176471, "grad_norm": 0.9328508517831882, "learning_rate": 6.1221714403874415e-06, "loss": 0.3015, "step": 11769 }, { "epoch": 5.538823529411765, "grad_norm": 0.35486594283015305, "learning_rate": 6.109781698798843e-06, "loss": 0.0802, "step": 11770 }, { "epoch": 5.539294117647059, "grad_norm": 0.6644582038344785, "learning_rate": 6.097404246140669e-06, "loss": 0.1989, "step": 11771 }, { "epoch": 5.539764705882353, "grad_norm": 0.4638781927497331, "learning_rate": 6.085039083470006e-06, "loss": 0.2393, "step": 11772 }, { "epoch": 5.540235294117647, "grad_norm": 0.3241881174194506, "learning_rate": 6.072686211842909e-06, "loss": 0.0856, "step": 11773 }, { "epoch": 5.540705882352941, "grad_norm": 0.32438330364812745, "learning_rate": 6.060345632314384e-06, "loss": 0.1085, "step": 11774 }, { "epoch": 5.541176470588235, "grad_norm": 0.34672018975802044, "learning_rate": 6.048017345938339e-06, "loss": 0.0728, "step": 11775 }, { "epoch": 5.541647058823529, "grad_norm": 0.27668629537260625, "learning_rate": 6.035701353767713e-06, "loss": 0.0787, "step": 11776 }, { "epoch": 5.542117647058824, "grad_norm": 0.2743297487441436, "learning_rate": 6.023397656854301e-06, "loss": 0.0676, "step": 11777 }, { "epoch": 5.542588235294118, "grad_norm": 0.8599455174246087, "learning_rate": 6.011106256248943e-06, "loss": 0.2599, "step": 11778 }, { "epoch": 5.543058823529412, "grad_norm": 0.3809906771869555, "learning_rate": 5.998827153001401e-06, "loss": 0.1243, "step": 11779 }, { "epoch": 5.543529411764706, "grad_norm": 0.7069784027301036, "learning_rate": 5.986560348160302e-06, "loss": 0.2969, "step": 11780 }, { "epoch": 5.5440000000000005, "grad_norm": 0.39254203262948834, "learning_rate": 5.9743058427733414e-06, "loss": 0.112, "step": 11781 }, { "epoch": 5.544470588235294, "grad_norm": 0.35330224136116944, "learning_rate": 5.9620636378871325e-06, "loss": 0.099, "step": 11782 }, { "epoch": 5.544941176470588, "grad_norm": 0.6284662683989387, "learning_rate": 5.9498337345471544e-06, "loss": 0.2536, "step": 11783 }, { "epoch": 5.545411764705882, "grad_norm": 0.6013063822193266, "learning_rate": 5.93761613379794e-06, "loss": 0.3024, "step": 11784 }, { "epoch": 5.5458823529411765, "grad_norm": 0.1398132948314497, "learning_rate": 5.925410836682953e-06, "loss": 0.0169, "step": 11785 }, { "epoch": 5.546352941176471, "grad_norm": 0.42832517176551815, "learning_rate": 5.913217844244545e-06, "loss": 0.1059, "step": 11786 }, { "epoch": 5.546823529411765, "grad_norm": 0.2883722099377356, "learning_rate": 5.90103715752408e-06, "loss": 0.0957, "step": 11787 }, { "epoch": 5.547294117647059, "grad_norm": 0.2863081022070237, "learning_rate": 5.888868777561862e-06, "loss": 0.0652, "step": 11788 }, { "epoch": 5.5477647058823525, "grad_norm": 0.5438619988850377, "learning_rate": 5.8767127053970745e-06, "loss": 0.2957, "step": 11789 }, { "epoch": 5.548235294117647, "grad_norm": 0.36711191263472914, "learning_rate": 5.864568942067954e-06, "loss": 0.1724, "step": 11790 }, { "epoch": 5.548705882352941, "grad_norm": 0.3826588061975344, "learning_rate": 5.852437488611656e-06, "loss": 0.1294, "step": 11791 }, { "epoch": 5.549176470588235, "grad_norm": 0.4719705026414908, "learning_rate": 5.840318346064199e-06, "loss": 0.2836, "step": 11792 }, { "epoch": 5.549647058823529, "grad_norm": 0.4368220678687495, "learning_rate": 5.828211515460673e-06, "loss": 0.2599, "step": 11793 }, { "epoch": 5.550117647058824, "grad_norm": 0.6601116960750829, "learning_rate": 5.816116997835052e-06, "loss": 0.1662, "step": 11794 }, { "epoch": 5.550588235294118, "grad_norm": 0.30240964638967543, "learning_rate": 5.804034794220208e-06, "loss": 0.1305, "step": 11795 }, { "epoch": 5.551058823529412, "grad_norm": 0.5419769936871109, "learning_rate": 5.791964905648116e-06, "loss": 0.1224, "step": 11796 }, { "epoch": 5.551529411764706, "grad_norm": 0.3981516313481435, "learning_rate": 5.779907333149553e-06, "loss": 0.2263, "step": 11797 }, { "epoch": 5.552, "grad_norm": 0.2967660867591144, "learning_rate": 5.767862077754276e-06, "loss": 0.0535, "step": 11798 }, { "epoch": 5.552470588235294, "grad_norm": 0.1726592080447743, "learning_rate": 5.755829140491064e-06, "loss": 0.0337, "step": 11799 }, { "epoch": 5.552941176470588, "grad_norm": 0.27824010991823667, "learning_rate": 5.743808522387544e-06, "loss": 0.095, "step": 11800 }, { "epoch": 5.553411764705882, "grad_norm": 0.22790877422234257, "learning_rate": 5.7318002244703285e-06, "loss": 0.06, "step": 11801 }, { "epoch": 5.5538823529411765, "grad_norm": 0.8751312543255948, "learning_rate": 5.7198042477650475e-06, "loss": 0.423, "step": 11802 }, { "epoch": 5.554352941176471, "grad_norm": 0.4909817029700525, "learning_rate": 5.7078205932961476e-06, "loss": 0.2404, "step": 11803 }, { "epoch": 5.554823529411765, "grad_norm": 0.4844238339496939, "learning_rate": 5.69584926208711e-06, "loss": 0.2698, "step": 11804 }, { "epoch": 5.555294117647059, "grad_norm": 0.4690303674282952, "learning_rate": 5.683890255160367e-06, "loss": 0.1831, "step": 11805 }, { "epoch": 5.555764705882353, "grad_norm": 0.42967221418774115, "learning_rate": 5.671943573537269e-06, "loss": 0.2302, "step": 11806 }, { "epoch": 5.556235294117647, "grad_norm": 0.1962182490950896, "learning_rate": 5.6600092182381156e-06, "loss": 0.0339, "step": 11807 }, { "epoch": 5.556705882352941, "grad_norm": 0.16516008926934192, "learning_rate": 5.648087190282141e-06, "loss": 0.0196, "step": 11808 }, { "epoch": 5.557176470588235, "grad_norm": 0.34654442806919067, "learning_rate": 5.636177490687565e-06, "loss": 0.1483, "step": 11809 }, { "epoch": 5.557647058823529, "grad_norm": 0.23887498611898023, "learning_rate": 5.624280120471542e-06, "loss": 0.0564, "step": 11810 }, { "epoch": 5.558117647058824, "grad_norm": 0.4086102356888619, "learning_rate": 5.612395080650156e-06, "loss": 0.28, "step": 11811 }, { "epoch": 5.558588235294118, "grad_norm": 0.3127640382126154, "learning_rate": 5.600522372238431e-06, "loss": 0.1229, "step": 11812 }, { "epoch": 5.559058823529412, "grad_norm": 0.24258665594144965, "learning_rate": 5.588661996250387e-06, "loss": 0.0705, "step": 11813 }, { "epoch": 5.559529411764705, "grad_norm": 0.29658230534238156, "learning_rate": 5.576813953698933e-06, "loss": 0.1512, "step": 11814 }, { "epoch": 5.5600000000000005, "grad_norm": 0.8318029568873816, "learning_rate": 5.564978245595958e-06, "loss": 0.2672, "step": 11815 }, { "epoch": 5.560470588235294, "grad_norm": 0.2403999286714593, "learning_rate": 5.553154872952303e-06, "loss": 0.0689, "step": 11816 }, { "epoch": 5.560941176470588, "grad_norm": 0.47671708362643384, "learning_rate": 5.541343836777729e-06, "loss": 0.159, "step": 11817 }, { "epoch": 5.561411764705882, "grad_norm": 0.5725132443937174, "learning_rate": 5.529545138080943e-06, "loss": 0.2789, "step": 11818 }, { "epoch": 5.5618823529411765, "grad_norm": 0.0626395508320418, "learning_rate": 5.517758777869641e-06, "loss": 0.0062, "step": 11819 }, { "epoch": 5.562352941176471, "grad_norm": 0.42771523130378103, "learning_rate": 5.5059847571504e-06, "loss": 0.1422, "step": 11820 }, { "epoch": 5.562823529411765, "grad_norm": 0.25603304167647367, "learning_rate": 5.494223076928817e-06, "loss": 0.0864, "step": 11821 }, { "epoch": 5.563294117647059, "grad_norm": 0.3684873083170584, "learning_rate": 5.482473738209386e-06, "loss": 0.0987, "step": 11822 }, { "epoch": 5.5637647058823525, "grad_norm": 0.4499856918858861, "learning_rate": 5.470736741995541e-06, "loss": 0.1739, "step": 11823 }, { "epoch": 5.564235294117647, "grad_norm": 0.3454442132318533, "learning_rate": 5.4590120892896775e-06, "loss": 0.1219, "step": 11824 }, { "epoch": 5.564705882352941, "grad_norm": 0.31860028172464233, "learning_rate": 5.447299781093178e-06, "loss": 0.0898, "step": 11825 }, { "epoch": 5.565176470588235, "grad_norm": 0.2895134433156408, "learning_rate": 5.435599818406278e-06, "loss": 0.0771, "step": 11826 }, { "epoch": 5.565647058823529, "grad_norm": 0.350260832193626, "learning_rate": 5.423912202228275e-06, "loss": 0.0846, "step": 11827 }, { "epoch": 5.566117647058824, "grad_norm": 0.42192960010944375, "learning_rate": 5.412236933557274e-06, "loss": 0.1902, "step": 11828 }, { "epoch": 5.566588235294118, "grad_norm": 0.623848647092177, "learning_rate": 5.400574013390441e-06, "loss": 0.278, "step": 11829 }, { "epoch": 5.567058823529412, "grad_norm": 0.4746571768320563, "learning_rate": 5.388923442723864e-06, "loss": 0.258, "step": 11830 }, { "epoch": 5.567529411764706, "grad_norm": 0.5547158516153804, "learning_rate": 5.37728522255253e-06, "loss": 0.3392, "step": 11831 }, { "epoch": 5.568, "grad_norm": 0.2758365361875383, "learning_rate": 5.365659353870394e-06, "loss": 0.097, "step": 11832 }, { "epoch": 5.568470588235294, "grad_norm": 0.2866796055132264, "learning_rate": 5.354045837670412e-06, "loss": 0.1349, "step": 11833 }, { "epoch": 5.568941176470588, "grad_norm": 0.3397144444555645, "learning_rate": 5.342444674944374e-06, "loss": 0.1466, "step": 11834 }, { "epoch": 5.569411764705882, "grad_norm": 0.23699823148309615, "learning_rate": 5.330855866683087e-06, "loss": 0.089, "step": 11835 }, { "epoch": 5.5698823529411765, "grad_norm": 0.4950946971450544, "learning_rate": 5.319279413876343e-06, "loss": 0.2379, "step": 11836 }, { "epoch": 5.570352941176471, "grad_norm": 0.41239351672599917, "learning_rate": 5.307715317512767e-06, "loss": 0.1664, "step": 11837 }, { "epoch": 5.570823529411765, "grad_norm": 0.542224661430773, "learning_rate": 5.29616357858002e-06, "loss": 0.1865, "step": 11838 }, { "epoch": 5.571294117647059, "grad_norm": 0.47481235471024164, "learning_rate": 5.2846241980646795e-06, "loss": 0.1934, "step": 11839 }, { "epoch": 5.571764705882353, "grad_norm": 0.473957586073674, "learning_rate": 5.273097176952224e-06, "loss": 0.222, "step": 11840 }, { "epoch": 5.572235294117647, "grad_norm": 0.44052447523005983, "learning_rate": 5.261582516227197e-06, "loss": 0.1916, "step": 11841 }, { "epoch": 5.572705882352941, "grad_norm": 0.37611010521152216, "learning_rate": 5.250080216872932e-06, "loss": 0.1586, "step": 11842 }, { "epoch": 5.573176470588235, "grad_norm": 0.4248083559100502, "learning_rate": 5.238590279871824e-06, "loss": 0.2014, "step": 11843 }, { "epoch": 5.573647058823529, "grad_norm": 0.6627079509783114, "learning_rate": 5.227112706205155e-06, "loss": 0.3442, "step": 11844 }, { "epoch": 5.574117647058824, "grad_norm": 0.4810485778178097, "learning_rate": 5.2156474968531595e-06, "loss": 0.2571, "step": 11845 }, { "epoch": 5.574588235294118, "grad_norm": 0.266010909377068, "learning_rate": 5.204194652795035e-06, "loss": 0.0912, "step": 11846 }, { "epoch": 5.575058823529412, "grad_norm": 0.38413110675333995, "learning_rate": 5.1927541750089175e-06, "loss": 0.1811, "step": 11847 }, { "epoch": 5.575529411764705, "grad_norm": 0.4099085586802709, "learning_rate": 5.181326064471858e-06, "loss": 0.2236, "step": 11848 }, { "epoch": 5.576, "grad_norm": 0.432465159802487, "learning_rate": 5.169910322159876e-06, "loss": 0.2624, "step": 11849 }, { "epoch": 5.576470588235294, "grad_norm": 0.4404333524750053, "learning_rate": 5.158506949047975e-06, "loss": 0.1856, "step": 11850 }, { "epoch": 5.576941176470588, "grad_norm": 0.48302527350264995, "learning_rate": 5.147115946109992e-06, "loss": 0.2187, "step": 11851 }, { "epoch": 5.577411764705882, "grad_norm": 0.41109218701801375, "learning_rate": 5.1357373143188e-06, "loss": 0.2319, "step": 11852 }, { "epoch": 5.5778823529411765, "grad_norm": 0.37985743506505304, "learning_rate": 5.124371054646237e-06, "loss": 0.1522, "step": 11853 }, { "epoch": 5.578352941176471, "grad_norm": 0.4863353270370313, "learning_rate": 5.11301716806296e-06, "loss": 0.2612, "step": 11854 }, { "epoch": 5.578823529411765, "grad_norm": 0.26641916471454147, "learning_rate": 5.101675655538695e-06, "loss": 0.0606, "step": 11855 }, { "epoch": 5.579294117647059, "grad_norm": 0.293537329554613, "learning_rate": 5.090346518042049e-06, "loss": 0.0702, "step": 11856 }, { "epoch": 5.5797647058823525, "grad_norm": 0.9218846397335487, "learning_rate": 5.079029756540598e-06, "loss": 0.1269, "step": 11857 }, { "epoch": 5.580235294117647, "grad_norm": 0.3744556308848323, "learning_rate": 5.067725372000836e-06, "loss": 0.1851, "step": 11858 }, { "epoch": 5.580705882352941, "grad_norm": 0.5697608065593661, "learning_rate": 5.056433365388207e-06, "loss": 0.3571, "step": 11859 }, { "epoch": 5.581176470588235, "grad_norm": 0.37421213930116776, "learning_rate": 5.045153737667107e-06, "loss": 0.1463, "step": 11860 }, { "epoch": 5.581647058823529, "grad_norm": 0.3027012864268716, "learning_rate": 5.033886489800898e-06, "loss": 0.1227, "step": 11861 }, { "epoch": 5.582117647058824, "grad_norm": 0.30049530902022276, "learning_rate": 5.022631622751827e-06, "loss": 0.1059, "step": 11862 }, { "epoch": 5.582588235294118, "grad_norm": 0.8404650423073906, "learning_rate": 5.011389137481109e-06, "loss": 0.3829, "step": 11863 }, { "epoch": 5.583058823529412, "grad_norm": 0.5102693579640882, "learning_rate": 5.000159034948958e-06, "loss": 0.354, "step": 11864 }, { "epoch": 5.583529411764706, "grad_norm": 0.09021619266285173, "learning_rate": 4.988941316114425e-06, "loss": 0.0149, "step": 11865 }, { "epoch": 5.584, "grad_norm": 0.3939129215886215, "learning_rate": 4.9777359819355765e-06, "loss": 0.2102, "step": 11866 }, { "epoch": 5.584470588235294, "grad_norm": 0.4994296429054334, "learning_rate": 4.9665430333694135e-06, "loss": 0.2377, "step": 11867 }, { "epoch": 5.584941176470588, "grad_norm": 0.5092422208567806, "learning_rate": 4.955362471371854e-06, "loss": 0.1992, "step": 11868 }, { "epoch": 5.585411764705882, "grad_norm": 0.2951815144417374, "learning_rate": 4.944194296897785e-06, "loss": 0.1067, "step": 11869 }, { "epoch": 5.5858823529411765, "grad_norm": 0.44741089900339714, "learning_rate": 4.933038510901027e-06, "loss": 0.1435, "step": 11870 }, { "epoch": 5.586352941176471, "grad_norm": 0.2777865664013191, "learning_rate": 4.921895114334301e-06, "loss": 0.1241, "step": 11871 }, { "epoch": 5.586823529411765, "grad_norm": 0.44132581898812995, "learning_rate": 4.910764108149346e-06, "loss": 0.1485, "step": 11872 }, { "epoch": 5.587294117647059, "grad_norm": 0.6994126344329107, "learning_rate": 4.899645493296817e-06, "loss": 0.4384, "step": 11873 }, { "epoch": 5.587764705882353, "grad_norm": 0.5052734185187049, "learning_rate": 4.888539270726255e-06, "loss": 0.1876, "step": 11874 }, { "epoch": 5.588235294117647, "grad_norm": 0.3840497834763075, "learning_rate": 4.877445441386218e-06, "loss": 0.1061, "step": 11875 }, { "epoch": 5.588705882352941, "grad_norm": 0.4987650861897492, "learning_rate": 4.866364006224166e-06, "loss": 0.2677, "step": 11876 }, { "epoch": 5.589176470588235, "grad_norm": 0.38816773002116267, "learning_rate": 4.85529496618649e-06, "loss": 0.1691, "step": 11877 }, { "epoch": 5.589647058823529, "grad_norm": 0.5201100050009914, "learning_rate": 4.844238322218568e-06, "loss": 0.293, "step": 11878 }, { "epoch": 5.590117647058824, "grad_norm": 0.353294668057429, "learning_rate": 4.833194075264663e-06, "loss": 0.0862, "step": 11879 }, { "epoch": 5.590588235294118, "grad_norm": 0.3059025526567546, "learning_rate": 4.8221622262680336e-06, "loss": 0.1218, "step": 11880 }, { "epoch": 5.591058823529412, "grad_norm": 0.6439329182860452, "learning_rate": 4.811142776170845e-06, "loss": 0.4315, "step": 11881 }, { "epoch": 5.591529411764705, "grad_norm": 0.30199625974293026, "learning_rate": 4.800135725914195e-06, "loss": 0.0765, "step": 11882 }, { "epoch": 5.592, "grad_norm": 0.47956795201734664, "learning_rate": 4.789141076438163e-06, "loss": 0.1986, "step": 11883 }, { "epoch": 5.592470588235294, "grad_norm": 0.401395014497302, "learning_rate": 4.778158828681733e-06, "loss": 0.2084, "step": 11884 }, { "epoch": 5.592941176470588, "grad_norm": 0.24734429062645555, "learning_rate": 4.767188983582854e-06, "loss": 0.0625, "step": 11885 }, { "epoch": 5.593411764705882, "grad_norm": 0.5475419828183155, "learning_rate": 4.75623154207841e-06, "loss": 0.1877, "step": 11886 }, { "epoch": 5.5938823529411765, "grad_norm": 0.21520438211464346, "learning_rate": 4.745286505104201e-06, "loss": 0.0228, "step": 11887 }, { "epoch": 5.594352941176471, "grad_norm": 0.27549386394411374, "learning_rate": 4.734353873594998e-06, "loss": 0.1101, "step": 11888 }, { "epoch": 5.594823529411765, "grad_norm": 0.3167515878664871, "learning_rate": 4.723433648484503e-06, "loss": 0.0669, "step": 11889 }, { "epoch": 5.595294117647059, "grad_norm": 0.36070586110885483, "learning_rate": 4.712525830705338e-06, "loss": 0.1585, "step": 11890 }, { "epoch": 5.5957647058823525, "grad_norm": 0.5337138512260707, "learning_rate": 4.70163042118909e-06, "loss": 0.2295, "step": 11891 }, { "epoch": 5.596235294117647, "grad_norm": 0.38352354064034083, "learning_rate": 4.690747420866314e-06, "loss": 0.1305, "step": 11892 }, { "epoch": 5.596705882352941, "grad_norm": 0.4785719905290515, "learning_rate": 4.679876830666434e-06, "loss": 0.3069, "step": 11893 }, { "epoch": 5.597176470588235, "grad_norm": 0.6923063341559793, "learning_rate": 4.669018651517859e-06, "loss": 0.3063, "step": 11894 }, { "epoch": 5.597647058823529, "grad_norm": 0.5265492472858376, "learning_rate": 4.658172884347943e-06, "loss": 0.1339, "step": 11895 }, { "epoch": 5.598117647058824, "grad_norm": 0.315682223927009, "learning_rate": 4.6473395300829495e-06, "loss": 0.12, "step": 11896 }, { "epoch": 5.598588235294118, "grad_norm": 0.6378162009401648, "learning_rate": 4.636518589648119e-06, "loss": 0.3848, "step": 11897 }, { "epoch": 5.599058823529412, "grad_norm": 0.3818207129183248, "learning_rate": 4.625710063967613e-06, "loss": 0.1933, "step": 11898 }, { "epoch": 5.599529411764706, "grad_norm": 0.38340482432324097, "learning_rate": 4.614913953964511e-06, "loss": 0.1805, "step": 11899 }, { "epoch": 5.6, "grad_norm": 0.42778008633513803, "learning_rate": 4.604130260560873e-06, "loss": 0.1879, "step": 11900 }, { "epoch": 5.600470588235294, "grad_norm": 0.5060562103958107, "learning_rate": 4.593358984677681e-06, "loss": 0.1464, "step": 11901 }, { "epoch": 5.600941176470588, "grad_norm": 0.4325014273358173, "learning_rate": 4.5826001272348315e-06, "loss": 0.2032, "step": 11902 }, { "epoch": 5.601411764705882, "grad_norm": 0.42358534877090576, "learning_rate": 4.571853689151206e-06, "loss": 0.1628, "step": 11903 }, { "epoch": 5.6018823529411765, "grad_norm": 0.347080057502976, "learning_rate": 4.5611196713446045e-06, "loss": 0.1437, "step": 11904 }, { "epoch": 5.602352941176471, "grad_norm": 0.3988694502072107, "learning_rate": 4.5503980747317425e-06, "loss": 0.1447, "step": 11905 }, { "epoch": 5.602823529411765, "grad_norm": 0.3515063511734542, "learning_rate": 4.539688900228322e-06, "loss": 0.1768, "step": 11906 }, { "epoch": 5.603294117647059, "grad_norm": 0.3353485303822607, "learning_rate": 4.528992148748944e-06, "loss": 0.1634, "step": 11907 }, { "epoch": 5.603764705882353, "grad_norm": 0.3684873793818719, "learning_rate": 4.518307821207162e-06, "loss": 0.075, "step": 11908 }, { "epoch": 5.604235294117647, "grad_norm": 0.3544304999819855, "learning_rate": 4.507635918515496e-06, "loss": 0.2053, "step": 11909 }, { "epoch": 5.604705882352941, "grad_norm": 0.45685960255485697, "learning_rate": 4.496976441585348e-06, "loss": 0.1781, "step": 11910 }, { "epoch": 5.605176470588235, "grad_norm": 0.38329338453420936, "learning_rate": 4.486329391327093e-06, "loss": 0.2056, "step": 11911 }, { "epoch": 5.605647058823529, "grad_norm": 0.18653257591625236, "learning_rate": 4.475694768650051e-06, "loss": 0.0617, "step": 11912 }, { "epoch": 5.606117647058824, "grad_norm": 0.3724666156190349, "learning_rate": 4.4650725744624635e-06, "loss": 0.2165, "step": 11913 }, { "epoch": 5.606588235294118, "grad_norm": 0.41551992188735787, "learning_rate": 4.454462809671538e-06, "loss": 0.2322, "step": 11914 }, { "epoch": 5.607058823529412, "grad_norm": 0.318478371946494, "learning_rate": 4.443865475183384e-06, "loss": 0.1107, "step": 11915 }, { "epoch": 5.607529411764705, "grad_norm": 0.38838771113592363, "learning_rate": 4.4332805719030586e-06, "loss": 0.1582, "step": 11916 }, { "epoch": 5.608, "grad_norm": 0.46099076675307055, "learning_rate": 4.4227081007345565e-06, "loss": 0.1997, "step": 11917 }, { "epoch": 5.608470588235294, "grad_norm": 0.32759258219775467, "learning_rate": 4.412148062580873e-06, "loss": 0.1368, "step": 11918 }, { "epoch": 5.608941176470588, "grad_norm": 0.4816925446831323, "learning_rate": 4.401600458343818e-06, "loss": 0.1323, "step": 11919 }, { "epoch": 5.609411764705882, "grad_norm": 0.38733715474362956, "learning_rate": 4.39106528892424e-06, "loss": 0.1663, "step": 11920 }, { "epoch": 5.6098823529411765, "grad_norm": 0.43257962841966174, "learning_rate": 4.380542555221917e-06, "loss": 0.1976, "step": 11921 }, { "epoch": 5.610352941176471, "grad_norm": 0.6137182477728531, "learning_rate": 4.370032258135514e-06, "loss": 0.2677, "step": 11922 }, { "epoch": 5.610823529411765, "grad_norm": 0.3121997911816059, "learning_rate": 4.3595343985626476e-06, "loss": 0.0768, "step": 11923 }, { "epoch": 5.611294117647059, "grad_norm": 0.2413280971354655, "learning_rate": 4.349048977399933e-06, "loss": 0.0561, "step": 11924 }, { "epoch": 5.6117647058823525, "grad_norm": 0.45468793263091145, "learning_rate": 4.33857599554282e-06, "loss": 0.2534, "step": 11925 }, { "epoch": 5.612235294117647, "grad_norm": 0.46250099599321126, "learning_rate": 4.328115453885794e-06, "loss": 0.1349, "step": 11926 }, { "epoch": 5.612705882352941, "grad_norm": 0.3268189077528746, "learning_rate": 4.317667353322241e-06, "loss": 0.1687, "step": 11927 }, { "epoch": 5.613176470588235, "grad_norm": 0.41193538883283665, "learning_rate": 4.307231694744445e-06, "loss": 0.1171, "step": 11928 }, { "epoch": 5.613647058823529, "grad_norm": 0.5614935973225681, "learning_rate": 4.296808479043678e-06, "loss": 0.1802, "step": 11929 }, { "epoch": 5.614117647058824, "grad_norm": 0.2814467150825151, "learning_rate": 4.286397707110145e-06, "loss": 0.0677, "step": 11930 }, { "epoch": 5.614588235294118, "grad_norm": 0.39779664391494124, "learning_rate": 4.275999379832967e-06, "loss": 0.1831, "step": 11931 }, { "epoch": 5.615058823529412, "grad_norm": 0.18241099175316217, "learning_rate": 4.265613498100234e-06, "loss": 0.0221, "step": 11932 }, { "epoch": 5.615529411764706, "grad_norm": 0.45860611825815384, "learning_rate": 4.255240062798904e-06, "loss": 0.288, "step": 11933 }, { "epoch": 5.616, "grad_norm": 0.43577197053597266, "learning_rate": 4.24487907481495e-06, "loss": 0.2413, "step": 11934 }, { "epoch": 5.616470588235294, "grad_norm": 0.21975569498398811, "learning_rate": 4.234530535033265e-06, "loss": 0.0197, "step": 11935 }, { "epoch": 5.616941176470588, "grad_norm": 0.4061736920088683, "learning_rate": 4.224194444337642e-06, "loss": 0.1972, "step": 11936 }, { "epoch": 5.617411764705882, "grad_norm": 0.5333139883854424, "learning_rate": 4.213870803610842e-06, "loss": 0.1447, "step": 11937 }, { "epoch": 5.6178823529411765, "grad_norm": 0.32437082561044583, "learning_rate": 4.203559613734542e-06, "loss": 0.0987, "step": 11938 }, { "epoch": 5.618352941176471, "grad_norm": 0.42317340219099686, "learning_rate": 4.193260875589371e-06, "loss": 0.1652, "step": 11939 }, { "epoch": 5.618823529411765, "grad_norm": 0.3397657223183693, "learning_rate": 4.1829745900549275e-06, "loss": 0.1356, "step": 11940 }, { "epoch": 5.619294117647058, "grad_norm": 0.29663669447173907, "learning_rate": 4.1727007580096565e-06, "loss": 0.1388, "step": 11941 }, { "epoch": 5.619764705882353, "grad_norm": 0.4657293413505383, "learning_rate": 4.162439380331023e-06, "loss": 0.0988, "step": 11942 }, { "epoch": 5.620235294117647, "grad_norm": 0.5411295792363046, "learning_rate": 4.15219045789541e-06, "loss": 0.4073, "step": 11943 }, { "epoch": 5.620705882352941, "grad_norm": 0.39829202776442874, "learning_rate": 4.141953991578101e-06, "loss": 0.2079, "step": 11944 }, { "epoch": 5.621176470588235, "grad_norm": 0.34059399237943916, "learning_rate": 4.131729982253345e-06, "loss": 0.1732, "step": 11945 }, { "epoch": 5.621647058823529, "grad_norm": 0.3691665237804986, "learning_rate": 4.121518430794346e-06, "loss": 0.1024, "step": 11946 }, { "epoch": 5.622117647058824, "grad_norm": 0.36742319168835785, "learning_rate": 4.111319338073188e-06, "loss": 0.217, "step": 11947 }, { "epoch": 5.622588235294118, "grad_norm": 0.4339115487012016, "learning_rate": 4.101132704960941e-06, "loss": 0.2971, "step": 11948 }, { "epoch": 5.623058823529412, "grad_norm": 0.4269723035637699, "learning_rate": 4.09095853232761e-06, "loss": 0.2604, "step": 11949 }, { "epoch": 5.623529411764705, "grad_norm": 0.3230180477774646, "learning_rate": 4.080796821042082e-06, "loss": 0.1246, "step": 11950 }, { "epoch": 5.624, "grad_norm": 0.5030001868124448, "learning_rate": 4.07064757197223e-06, "loss": 0.2761, "step": 11951 }, { "epoch": 5.624470588235294, "grad_norm": 0.2395316839630274, "learning_rate": 4.0605107859848604e-06, "loss": 0.078, "step": 11952 }, { "epoch": 5.624941176470588, "grad_norm": 0.6311550679511057, "learning_rate": 4.050386463945699e-06, "loss": 0.4696, "step": 11953 }, { "epoch": 5.625411764705882, "grad_norm": 0.35608447494407186, "learning_rate": 4.040274606719418e-06, "loss": 0.0753, "step": 11954 }, { "epoch": 5.625882352941177, "grad_norm": 0.39433649187624215, "learning_rate": 4.030175215169612e-06, "loss": 0.1542, "step": 11955 }, { "epoch": 5.626352941176471, "grad_norm": 0.5195243300531664, "learning_rate": 4.020088290158807e-06, "loss": 0.307, "step": 11956 }, { "epoch": 5.626823529411765, "grad_norm": 0.3202203000112282, "learning_rate": 4.010013832548514e-06, "loss": 0.1176, "step": 11957 }, { "epoch": 5.627294117647059, "grad_norm": 0.46762356910266095, "learning_rate": 3.999951843199095e-06, "loss": 0.1998, "step": 11958 }, { "epoch": 5.627764705882353, "grad_norm": 0.42438176189045224, "learning_rate": 3.989902322969929e-06, "loss": 0.0837, "step": 11959 }, { "epoch": 5.628235294117647, "grad_norm": 0.3537700459303499, "learning_rate": 3.979865272719279e-06, "loss": 0.1692, "step": 11960 }, { "epoch": 5.628705882352941, "grad_norm": 0.1839694586074556, "learning_rate": 3.969840693304344e-06, "loss": 0.0411, "step": 11961 }, { "epoch": 5.629176470588235, "grad_norm": 0.5232367212280938, "learning_rate": 3.959828585581287e-06, "loss": 0.3691, "step": 11962 }, { "epoch": 5.6296470588235294, "grad_norm": 0.36383815558278, "learning_rate": 3.949828950405226e-06, "loss": 0.1099, "step": 11963 }, { "epoch": 5.630117647058824, "grad_norm": 0.4950885728382521, "learning_rate": 3.939841788630111e-06, "loss": 0.1051, "step": 11964 }, { "epoch": 5.630588235294118, "grad_norm": 0.38725611529599074, "learning_rate": 3.9298671011089245e-06, "loss": 0.219, "step": 11965 }, { "epoch": 5.631058823529412, "grad_norm": 0.3820955363153786, "learning_rate": 3.9199048886935865e-06, "loss": 0.1726, "step": 11966 }, { "epoch": 5.631529411764706, "grad_norm": 0.2803340259792611, "learning_rate": 3.909955152234867e-06, "loss": 0.0784, "step": 11967 }, { "epoch": 5.632, "grad_norm": 0.2504705808756861, "learning_rate": 3.900017892582552e-06, "loss": 0.1284, "step": 11968 }, { "epoch": 5.632470588235294, "grad_norm": 0.3113842710883873, "learning_rate": 3.8900931105853294e-06, "loss": 0.1004, "step": 11969 }, { "epoch": 5.632941176470588, "grad_norm": 0.33916757754892096, "learning_rate": 3.8801808070908215e-06, "loss": 0.1077, "step": 11970 }, { "epoch": 5.633411764705882, "grad_norm": 0.7154159319154135, "learning_rate": 3.870280982945567e-06, "loss": 0.3202, "step": 11971 }, { "epoch": 5.633882352941177, "grad_norm": 0.5518900205273918, "learning_rate": 3.860393638995107e-06, "loss": 0.3447, "step": 11972 }, { "epoch": 5.634352941176471, "grad_norm": 0.33939038079143635, "learning_rate": 3.850518776083817e-06, "loss": 0.1284, "step": 11973 }, { "epoch": 5.634823529411765, "grad_norm": 0.2607693280798234, "learning_rate": 3.840656395055103e-06, "loss": 0.0771, "step": 11974 }, { "epoch": 5.635294117647058, "grad_norm": 0.3228365106044445, "learning_rate": 3.8308064967512095e-06, "loss": 0.1393, "step": 11975 }, { "epoch": 5.635764705882353, "grad_norm": 0.2918331278654302, "learning_rate": 3.820969082013414e-06, "loss": 0.0862, "step": 11976 }, { "epoch": 5.636235294117647, "grad_norm": 0.3094332606878871, "learning_rate": 3.8111441516818776e-06, "loss": 0.08, "step": 11977 }, { "epoch": 5.636705882352941, "grad_norm": 0.3813822931285919, "learning_rate": 3.8013317065956784e-06, "loss": 0.1274, "step": 11978 }, { "epoch": 5.637176470588235, "grad_norm": 0.37261004550508814, "learning_rate": 3.7915317475928296e-06, "loss": 0.207, "step": 11979 }, { "epoch": 5.6376470588235295, "grad_norm": 0.5182984665095642, "learning_rate": 3.7817442755103457e-06, "loss": 0.1784, "step": 11980 }, { "epoch": 5.638117647058824, "grad_norm": 1.2528215116588324, "learning_rate": 3.7719692911840905e-06, "loss": 0.541, "step": 11981 }, { "epoch": 5.638588235294118, "grad_norm": 0.36361277461004965, "learning_rate": 3.76220679544888e-06, "loss": 0.1453, "step": 11982 }, { "epoch": 5.639058823529412, "grad_norm": 0.27426616118863706, "learning_rate": 3.7524567891385305e-06, "loss": 0.0882, "step": 11983 }, { "epoch": 5.6395294117647055, "grad_norm": 0.3306099494003304, "learning_rate": 3.7427192730856927e-06, "loss": 0.147, "step": 11984 }, { "epoch": 5.64, "grad_norm": 0.49336328546469166, "learning_rate": 3.7329942481220176e-06, "loss": 0.1836, "step": 11985 }, { "epoch": 5.640470588235294, "grad_norm": 0.36603355297344775, "learning_rate": 3.723281715078091e-06, "loss": 0.103, "step": 11986 }, { "epoch": 5.640941176470588, "grad_norm": 0.6383609028447473, "learning_rate": 3.7135816747833667e-06, "loss": 0.1866, "step": 11987 }, { "epoch": 5.641411764705882, "grad_norm": 0.41555637168147, "learning_rate": 3.7038941280663147e-06, "loss": 0.1679, "step": 11988 }, { "epoch": 5.641882352941177, "grad_norm": 0.35707419629963044, "learning_rate": 3.6942190757542735e-06, "loss": 0.1324, "step": 11989 }, { "epoch": 5.642352941176471, "grad_norm": 0.5225678809701626, "learning_rate": 3.684556518673548e-06, "loss": 0.1792, "step": 11990 }, { "epoch": 5.642823529411765, "grad_norm": 0.7036974912483117, "learning_rate": 3.6749064576493793e-06, "loss": 0.2811, "step": 11991 }, { "epoch": 5.643294117647059, "grad_norm": 0.6799838125341455, "learning_rate": 3.6652688935058904e-06, "loss": 0.3713, "step": 11992 }, { "epoch": 5.643764705882353, "grad_norm": 0.7960669948280389, "learning_rate": 3.6556438270662236e-06, "loss": 0.4815, "step": 11993 }, { "epoch": 5.644235294117647, "grad_norm": 0.4257416013004998, "learning_rate": 3.6460312591524044e-06, "loss": 0.1824, "step": 11994 }, { "epoch": 5.644705882352941, "grad_norm": 0.37784492702142386, "learning_rate": 3.6364311905853594e-06, "loss": 0.1564, "step": 11995 }, { "epoch": 5.645176470588235, "grad_norm": 0.31691598808763327, "learning_rate": 3.6268436221849994e-06, "loss": 0.1345, "step": 11996 }, { "epoch": 5.6456470588235295, "grad_norm": 0.4166029922081381, "learning_rate": 3.6172685547701695e-06, "loss": 0.2018, "step": 11997 }, { "epoch": 5.646117647058824, "grad_norm": 1.378577473533113, "learning_rate": 3.6077059891585826e-06, "loss": 0.4208, "step": 11998 }, { "epoch": 5.646588235294118, "grad_norm": 0.5782010977637754, "learning_rate": 3.598155926166968e-06, "loss": 0.3482, "step": 11999 }, { "epoch": 5.647058823529412, "grad_norm": 0.3743934832292039, "learning_rate": 3.5886183666109405e-06, "loss": 0.1064, "step": 12000 }, { "epoch": 5.647529411764706, "grad_norm": 0.45461199434577465, "learning_rate": 3.579093311305031e-06, "loss": 0.138, "step": 12001 }, { "epoch": 5.648, "grad_norm": 0.5225553671938022, "learning_rate": 3.5695807610627402e-06, "loss": 0.1224, "step": 12002 }, { "epoch": 5.648470588235294, "grad_norm": 0.27322478388676835, "learning_rate": 3.5600807166965173e-06, "loss": 0.0854, "step": 12003 }, { "epoch": 5.648941176470588, "grad_norm": 0.40319037233494326, "learning_rate": 3.5505931790176635e-06, "loss": 0.1275, "step": 12004 }, { "epoch": 5.649411764705882, "grad_norm": 0.4062350558136103, "learning_rate": 3.5411181488364802e-06, "loss": 0.1471, "step": 12005 }, { "epoch": 5.649882352941177, "grad_norm": 0.3724126471718209, "learning_rate": 3.531655626962204e-06, "loss": 0.1891, "step": 12006 }, { "epoch": 5.650352941176471, "grad_norm": 0.4004427322458092, "learning_rate": 3.5222056142029387e-06, "loss": 0.1944, "step": 12007 }, { "epoch": 5.650823529411765, "grad_norm": 0.2994323984689297, "learning_rate": 3.5127681113658044e-06, "loss": 0.1273, "step": 12008 }, { "epoch": 5.651294117647058, "grad_norm": 0.5123665232600527, "learning_rate": 3.503343119256774e-06, "loss": 0.3252, "step": 12009 }, { "epoch": 5.651764705882353, "grad_norm": 0.45264844955823874, "learning_rate": 3.493930638680803e-06, "loss": 0.174, "step": 12010 }, { "epoch": 5.652235294117647, "grad_norm": 0.3738280532943441, "learning_rate": 3.484530670441782e-06, "loss": 0.0974, "step": 12011 }, { "epoch": 5.652705882352941, "grad_norm": 0.16246368379545503, "learning_rate": 3.4751432153424687e-06, "loss": 0.0412, "step": 12012 }, { "epoch": 5.653176470588235, "grad_norm": 0.4196002668720498, "learning_rate": 3.465768274184638e-06, "loss": 0.2383, "step": 12013 }, { "epoch": 5.6536470588235295, "grad_norm": 0.34783644141470943, "learning_rate": 3.4564058477689503e-06, "loss": 0.0866, "step": 12014 }, { "epoch": 5.654117647058824, "grad_norm": 0.5346772222026979, "learning_rate": 3.4470559368949825e-06, "loss": 0.3053, "step": 12015 }, { "epoch": 5.654588235294118, "grad_norm": 0.32251369485891757, "learning_rate": 3.4377185423612785e-06, "loss": 0.1693, "step": 12016 }, { "epoch": 5.655058823529412, "grad_norm": 0.7489404788599114, "learning_rate": 3.428393664965301e-06, "loss": 0.225, "step": 12017 }, { "epoch": 5.6555294117647055, "grad_norm": 0.5028109216862187, "learning_rate": 3.4190813055034127e-06, "loss": 0.0962, "step": 12018 }, { "epoch": 5.656, "grad_norm": 0.3531291923020026, "learning_rate": 3.4097814647709775e-06, "loss": 0.0928, "step": 12019 }, { "epoch": 5.656470588235294, "grad_norm": 0.5061898641989244, "learning_rate": 3.400494143562227e-06, "loss": 0.2319, "step": 12020 }, { "epoch": 5.656941176470588, "grad_norm": 0.7650603988664763, "learning_rate": 3.391219342670326e-06, "loss": 0.1861, "step": 12021 }, { "epoch": 5.657411764705882, "grad_norm": 0.2662567267423737, "learning_rate": 3.3819570628874414e-06, "loss": 0.0671, "step": 12022 }, { "epoch": 5.657882352941177, "grad_norm": 0.5509335624132994, "learning_rate": 3.3727073050045406e-06, "loss": 0.2718, "step": 12023 }, { "epoch": 5.658352941176471, "grad_norm": 0.39322470431465106, "learning_rate": 3.3634700698116577e-06, "loss": 0.1139, "step": 12024 }, { "epoch": 5.658823529411765, "grad_norm": 0.42189254400120246, "learning_rate": 3.354245358097679e-06, "loss": 0.1351, "step": 12025 }, { "epoch": 5.659294117647059, "grad_norm": 0.4467192518137178, "learning_rate": 3.345033170650441e-06, "loss": 0.2071, "step": 12026 }, { "epoch": 5.659764705882353, "grad_norm": 0.48981830509587615, "learning_rate": 3.3358335082566966e-06, "loss": 0.1401, "step": 12027 }, { "epoch": 5.660235294117647, "grad_norm": 0.4831274167792141, "learning_rate": 3.326646371702185e-06, "loss": 0.1933, "step": 12028 }, { "epoch": 5.660705882352941, "grad_norm": 0.3756224112183197, "learning_rate": 3.317471761771462e-06, "loss": 0.1499, "step": 12029 }, { "epoch": 5.661176470588235, "grad_norm": 0.25780475289856525, "learning_rate": 3.30830967924815e-06, "loss": 0.0682, "step": 12030 }, { "epoch": 5.6616470588235295, "grad_norm": 0.29489653441750774, "learning_rate": 3.299160124914707e-06, "loss": 0.1669, "step": 12031 }, { "epoch": 5.662117647058824, "grad_norm": 0.8838040364579014, "learning_rate": 3.290023099552541e-06, "loss": 0.4123, "step": 12032 }, { "epoch": 5.662588235294118, "grad_norm": 0.4075230952441548, "learning_rate": 3.2808986039420115e-06, "loss": 0.2035, "step": 12033 }, { "epoch": 5.663058823529412, "grad_norm": 0.5481696885323531, "learning_rate": 3.2717866388624114e-06, "loss": 0.1945, "step": 12034 }, { "epoch": 5.663529411764706, "grad_norm": 0.41415298274126616, "learning_rate": 3.2626872050919185e-06, "loss": 0.2743, "step": 12035 }, { "epoch": 5.664, "grad_norm": 0.2620939530936169, "learning_rate": 3.2536003034076774e-06, "loss": 0.0571, "step": 12036 }, { "epoch": 5.664470588235294, "grad_norm": 0.44353320116930645, "learning_rate": 3.2445259345857676e-06, "loss": 0.1881, "step": 12037 }, { "epoch": 5.664941176470588, "grad_norm": 0.463359483229457, "learning_rate": 3.2354640994011693e-06, "loss": 0.1484, "step": 12038 }, { "epoch": 5.665411764705882, "grad_norm": 0.608042458862914, "learning_rate": 3.22641479862783e-06, "loss": 0.3232, "step": 12039 }, { "epoch": 5.665882352941177, "grad_norm": 0.5957909987140839, "learning_rate": 3.2173780330385647e-06, "loss": 0.2986, "step": 12040 }, { "epoch": 5.666352941176471, "grad_norm": 0.34206113724485837, "learning_rate": 3.2083538034051726e-06, "loss": 0.1617, "step": 12041 }, { "epoch": 5.666823529411765, "grad_norm": 0.2835394414321824, "learning_rate": 3.199342110498404e-06, "loss": 0.114, "step": 12042 }, { "epoch": 5.667294117647058, "grad_norm": 0.3203677238381262, "learning_rate": 3.1903429550878435e-06, "loss": 0.0764, "step": 12043 }, { "epoch": 5.667764705882353, "grad_norm": 0.38349120561984773, "learning_rate": 3.181356337942109e-06, "loss": 0.1546, "step": 12044 }, { "epoch": 5.668235294117647, "grad_norm": 0.3430328984307246, "learning_rate": 3.1723822598286875e-06, "loss": 0.1189, "step": 12045 }, { "epoch": 5.668705882352941, "grad_norm": 0.33673864768388234, "learning_rate": 3.163420721513982e-06, "loss": 0.09, "step": 12046 }, { "epoch": 5.669176470588235, "grad_norm": 0.4276136832218465, "learning_rate": 3.1544717237633965e-06, "loss": 0.1736, "step": 12047 }, { "epoch": 5.6696470588235295, "grad_norm": 0.23270518157200765, "learning_rate": 3.145535267341204e-06, "loss": 0.0569, "step": 12048 }, { "epoch": 5.670117647058824, "grad_norm": 0.5733573871412626, "learning_rate": 3.1366113530106097e-06, "loss": 0.1965, "step": 12049 }, { "epoch": 5.670588235294118, "grad_norm": 0.49657443667767287, "learning_rate": 3.1276999815337544e-06, "loss": 0.261, "step": 12050 }, { "epoch": 5.671058823529412, "grad_norm": 0.6021823083489858, "learning_rate": 3.118801153671746e-06, "loss": 0.4639, "step": 12051 }, { "epoch": 5.6715294117647055, "grad_norm": 0.7023796382845751, "learning_rate": 3.1099148701845435e-06, "loss": 0.2517, "step": 12052 }, { "epoch": 5.672, "grad_norm": 0.2304563903604225, "learning_rate": 3.1010411318311226e-06, "loss": 0.0723, "step": 12053 }, { "epoch": 5.672470588235294, "grad_norm": 0.22325273740675056, "learning_rate": 3.092179939369327e-06, "loss": 0.047, "step": 12054 }, { "epoch": 5.672941176470588, "grad_norm": 0.4359156464453798, "learning_rate": 3.083331293555935e-06, "loss": 0.1834, "step": 12055 }, { "epoch": 5.673411764705882, "grad_norm": 0.3819063865817147, "learning_rate": 3.074495195146659e-06, "loss": 0.2159, "step": 12056 }, { "epoch": 5.673882352941177, "grad_norm": 0.31067544395202673, "learning_rate": 3.0656716448961784e-06, "loss": 0.0845, "step": 12057 }, { "epoch": 5.674352941176471, "grad_norm": 0.43355230983929827, "learning_rate": 3.05686064355804e-06, "loss": 0.0853, "step": 12058 }, { "epoch": 5.674823529411765, "grad_norm": 0.37702371333988566, "learning_rate": 3.048062191884759e-06, "loss": 0.2312, "step": 12059 }, { "epoch": 5.675294117647059, "grad_norm": 0.2756445668846196, "learning_rate": 3.039276290627768e-06, "loss": 0.101, "step": 12060 }, { "epoch": 5.675764705882353, "grad_norm": 0.4012867151974477, "learning_rate": 3.0305029405373993e-06, "loss": 0.1493, "step": 12061 }, { "epoch": 5.676235294117647, "grad_norm": 0.4757440932147367, "learning_rate": 3.0217421423629706e-06, "loss": 0.2995, "step": 12062 }, { "epoch": 5.676705882352941, "grad_norm": 0.3238066288029223, "learning_rate": 3.0129938968526836e-06, "loss": 0.1799, "step": 12063 }, { "epoch": 5.677176470588235, "grad_norm": 0.2873102386995056, "learning_rate": 3.00425820475369e-06, "loss": 0.1086, "step": 12064 }, { "epoch": 5.6776470588235295, "grad_norm": 0.3742737569167978, "learning_rate": 2.9955350668120605e-06, "loss": 0.1231, "step": 12065 }, { "epoch": 5.678117647058824, "grad_norm": 0.5726307544333289, "learning_rate": 2.9868244837727815e-06, "loss": 0.3134, "step": 12066 }, { "epoch": 5.678588235294118, "grad_norm": 0.2361997922703122, "learning_rate": 2.978126456379809e-06, "loss": 0.0529, "step": 12067 }, { "epoch": 5.679058823529411, "grad_norm": 0.1291296203518325, "learning_rate": 2.969440985375948e-06, "loss": 0.0171, "step": 12068 }, { "epoch": 5.679529411764706, "grad_norm": 0.5761222690088845, "learning_rate": 2.960768071503006e-06, "loss": 0.4365, "step": 12069 }, { "epoch": 5.68, "grad_norm": 0.7654388296500148, "learning_rate": 2.9521077155017224e-06, "loss": 0.3118, "step": 12070 }, { "epoch": 5.680470588235294, "grad_norm": 0.3933857125017221, "learning_rate": 2.9434599181116736e-06, "loss": 0.1529, "step": 12071 }, { "epoch": 5.680941176470588, "grad_norm": 0.7291280623541259, "learning_rate": 2.934824680071468e-06, "loss": 0.4496, "step": 12072 }, { "epoch": 5.681411764705882, "grad_norm": 0.6072065796893752, "learning_rate": 2.9262020021185993e-06, "loss": 0.4292, "step": 12073 }, { "epoch": 5.681882352941177, "grad_norm": 0.22516333013810536, "learning_rate": 2.917591884989462e-06, "loss": 0.041, "step": 12074 }, { "epoch": 5.682352941176471, "grad_norm": 0.46338688275413425, "learning_rate": 2.908994329419401e-06, "loss": 0.2097, "step": 12075 }, { "epoch": 5.682823529411765, "grad_norm": 0.38453977473495593, "learning_rate": 2.9004093361427282e-06, "loss": 0.1663, "step": 12076 }, { "epoch": 5.683294117647058, "grad_norm": 0.19672556113128875, "learning_rate": 2.891836905892592e-06, "loss": 0.0544, "step": 12077 }, { "epoch": 5.683764705882353, "grad_norm": 0.5460713527252795, "learning_rate": 2.883277039401155e-06, "loss": 0.2985, "step": 12078 }, { "epoch": 5.684235294117647, "grad_norm": 0.38822792870076034, "learning_rate": 2.874729737399467e-06, "loss": 0.2507, "step": 12079 }, { "epoch": 5.684705882352941, "grad_norm": 0.24923454197812647, "learning_rate": 2.866195000617494e-06, "loss": 0.0835, "step": 12080 }, { "epoch": 5.685176470588235, "grad_norm": 0.46936037468386294, "learning_rate": 2.857672829784169e-06, "loss": 0.2743, "step": 12081 }, { "epoch": 5.6856470588235295, "grad_norm": 0.35573065741008564, "learning_rate": 2.84916322562731e-06, "loss": 0.1931, "step": 12082 }, { "epoch": 5.686117647058824, "grad_norm": 0.5146215661916121, "learning_rate": 2.840666188873669e-06, "loss": 0.2599, "step": 12083 }, { "epoch": 5.686588235294118, "grad_norm": 0.22751888232998657, "learning_rate": 2.8321817202489493e-06, "loss": 0.0439, "step": 12084 }, { "epoch": 5.687058823529412, "grad_norm": 0.4573589207982352, "learning_rate": 2.8237098204777874e-06, "loss": 0.1944, "step": 12085 }, { "epoch": 5.6875294117647055, "grad_norm": 0.2996713185841234, "learning_rate": 2.815250490283688e-06, "loss": 0.1003, "step": 12086 }, { "epoch": 5.688, "grad_norm": 0.5020717955569836, "learning_rate": 2.8068037303891233e-06, "loss": 0.2579, "step": 12087 }, { "epoch": 5.688470588235294, "grad_norm": 0.46568941417241827, "learning_rate": 2.7983695415155162e-06, "loss": 0.0891, "step": 12088 }, { "epoch": 5.688941176470588, "grad_norm": 0.1984743276153381, "learning_rate": 2.789947924383157e-06, "loss": 0.0624, "step": 12089 }, { "epoch": 5.689411764705882, "grad_norm": 0.46937467135908506, "learning_rate": 2.78153887971132e-06, "loss": 0.1703, "step": 12090 }, { "epoch": 5.689882352941177, "grad_norm": 0.3161952817422803, "learning_rate": 2.773142408218165e-06, "loss": 0.1178, "step": 12091 }, { "epoch": 5.690352941176471, "grad_norm": 0.4097171040971054, "learning_rate": 2.7647585106207682e-06, "loss": 0.2076, "step": 12092 }, { "epoch": 5.690823529411765, "grad_norm": 0.5166799052864396, "learning_rate": 2.7563871876352063e-06, "loss": 0.2387, "step": 12093 }, { "epoch": 5.691294117647059, "grad_norm": 0.4284088432798029, "learning_rate": 2.7480284399763907e-06, "loss": 0.2283, "step": 12094 }, { "epoch": 5.691764705882353, "grad_norm": 0.329654844731472, "learning_rate": 2.7396822683582177e-06, "loss": 0.1838, "step": 12095 }, { "epoch": 5.692235294117647, "grad_norm": 0.33815033464889255, "learning_rate": 2.731348673493483e-06, "loss": 0.1423, "step": 12096 }, { "epoch": 5.692705882352941, "grad_norm": 0.29459787699173534, "learning_rate": 2.723027656093918e-06, "loss": 0.0648, "step": 12097 }, { "epoch": 5.693176470588235, "grad_norm": 0.5508544099636309, "learning_rate": 2.714719216870187e-06, "loss": 0.2576, "step": 12098 }, { "epoch": 5.6936470588235295, "grad_norm": 0.40438252308949957, "learning_rate": 2.706423356531873e-06, "loss": 0.1882, "step": 12099 }, { "epoch": 5.694117647058824, "grad_norm": 0.36811099137320064, "learning_rate": 2.6981400757874584e-06, "loss": 0.0603, "step": 12100 }, { "epoch": 5.694588235294118, "grad_norm": 0.4886880946903176, "learning_rate": 2.6898693753443945e-06, "loss": 0.2297, "step": 12101 }, { "epoch": 5.695058823529411, "grad_norm": 0.7260460538458205, "learning_rate": 2.68161125590905e-06, "loss": 0.2737, "step": 12102 }, { "epoch": 5.6955294117647055, "grad_norm": 0.15323243196778188, "learning_rate": 2.673365718186693e-06, "loss": 0.0191, "step": 12103 }, { "epoch": 5.696, "grad_norm": 0.19319887912575837, "learning_rate": 2.6651327628815277e-06, "loss": 0.0437, "step": 12104 }, { "epoch": 5.696470588235294, "grad_norm": 0.5841039882174119, "learning_rate": 2.656912390696708e-06, "loss": 0.1996, "step": 12105 }, { "epoch": 5.696941176470588, "grad_norm": 0.32432249679504166, "learning_rate": 2.6487046023342563e-06, "loss": 0.0764, "step": 12106 }, { "epoch": 5.697411764705882, "grad_norm": 0.18378492507635857, "learning_rate": 2.6405093984952275e-06, "loss": 0.04, "step": 12107 }, { "epoch": 5.697882352941177, "grad_norm": 0.7595543839517066, "learning_rate": 2.6323267798794625e-06, "loss": 0.1604, "step": 12108 }, { "epoch": 5.698352941176471, "grad_norm": 0.7089731802065445, "learning_rate": 2.624156747185835e-06, "loss": 0.3046, "step": 12109 }, { "epoch": 5.698823529411765, "grad_norm": 0.31848149206039966, "learning_rate": 2.615999301112104e-06, "loss": 0.0589, "step": 12110 }, { "epoch": 5.699294117647058, "grad_norm": 0.4032318508985775, "learning_rate": 2.6078544423549286e-06, "loss": 0.1536, "step": 12111 }, { "epoch": 5.699764705882353, "grad_norm": 0.3787563009267203, "learning_rate": 2.599722171609936e-06, "loss": 0.211, "step": 12112 }, { "epoch": 5.700235294117647, "grad_norm": 0.4020100657760978, "learning_rate": 2.591602489571687e-06, "loss": 0.2064, "step": 12113 }, { "epoch": 5.700705882352941, "grad_norm": 0.16078848685286726, "learning_rate": 2.5834953969336105e-06, "loss": 0.0245, "step": 12114 }, { "epoch": 5.701176470588235, "grad_norm": 0.6928214108581957, "learning_rate": 2.5754008943880855e-06, "loss": 0.3758, "step": 12115 }, { "epoch": 5.7016470588235295, "grad_norm": 0.472714300397954, "learning_rate": 2.5673189826264596e-06, "loss": 0.1586, "step": 12116 }, { "epoch": 5.702117647058824, "grad_norm": 0.5577710751548521, "learning_rate": 2.559249662338947e-06, "loss": 0.1882, "step": 12117 }, { "epoch": 5.702588235294118, "grad_norm": 0.31958383073417573, "learning_rate": 2.5511929342146964e-06, "loss": 0.0855, "step": 12118 }, { "epoch": 5.703058823529412, "grad_norm": 0.20223744195975416, "learning_rate": 2.543148798941791e-06, "loss": 0.0489, "step": 12119 }, { "epoch": 5.7035294117647055, "grad_norm": 0.355815014048281, "learning_rate": 2.535117257207264e-06, "loss": 0.2303, "step": 12120 }, { "epoch": 5.704, "grad_norm": 0.5480785811013056, "learning_rate": 2.5270983096970344e-06, "loss": 0.2552, "step": 12121 }, { "epoch": 5.704470588235294, "grad_norm": 0.5052679246198154, "learning_rate": 2.5190919570959533e-06, "loss": 0.3452, "step": 12122 }, { "epoch": 5.704941176470588, "grad_norm": 0.6768150465782806, "learning_rate": 2.511098200087808e-06, "loss": 0.359, "step": 12123 }, { "epoch": 5.705411764705882, "grad_norm": 0.27889160591717976, "learning_rate": 2.503117039355318e-06, "loss": 0.1121, "step": 12124 }, { "epoch": 5.705882352941177, "grad_norm": 0.6125107748757913, "learning_rate": 2.4951484755800886e-06, "loss": 0.3676, "step": 12125 }, { "epoch": 5.706352941176471, "grad_norm": 0.27672348436377314, "learning_rate": 2.487192509442676e-06, "loss": 0.0932, "step": 12126 }, { "epoch": 5.706823529411765, "grad_norm": 0.24211053205662317, "learning_rate": 2.4792491416226024e-06, "loss": 0.0717, "step": 12127 }, { "epoch": 5.707294117647059, "grad_norm": 0.2990082406379024, "learning_rate": 2.471318372798209e-06, "loss": 0.0613, "step": 12128 }, { "epoch": 5.707764705882353, "grad_norm": 0.34166005781111863, "learning_rate": 2.4634002036468536e-06, "loss": 0.1024, "step": 12129 }, { "epoch": 5.708235294117647, "grad_norm": 0.34100091175399666, "learning_rate": 2.4554946348447958e-06, "loss": 0.1185, "step": 12130 }, { "epoch": 5.708705882352941, "grad_norm": 0.3662968139379496, "learning_rate": 2.447601667067195e-06, "loss": 0.0842, "step": 12131 }, { "epoch": 5.709176470588235, "grad_norm": 0.27822124901938133, "learning_rate": 2.4397213009881456e-06, "loss": 0.0585, "step": 12132 }, { "epoch": 5.7096470588235295, "grad_norm": 0.5513706992088774, "learning_rate": 2.4318535372806757e-06, "loss": 0.2584, "step": 12133 }, { "epoch": 5.710117647058824, "grad_norm": 0.3411217368713796, "learning_rate": 2.4239983766167315e-06, "loss": 0.1305, "step": 12134 }, { "epoch": 5.710588235294118, "grad_norm": 0.44196918138124247, "learning_rate": 2.416155819667193e-06, "loss": 0.1277, "step": 12135 }, { "epoch": 5.711058823529411, "grad_norm": 1.0939398577690742, "learning_rate": 2.4083258671018403e-06, "loss": 0.566, "step": 12136 }, { "epoch": 5.7115294117647055, "grad_norm": 0.24037346379788435, "learning_rate": 2.4005085195893725e-06, "loss": 0.0616, "step": 12137 }, { "epoch": 5.712, "grad_norm": 0.22481118855372847, "learning_rate": 2.3927037777974722e-06, "loss": 0.0452, "step": 12138 }, { "epoch": 5.712470588235294, "grad_norm": 0.45251355083341405, "learning_rate": 2.384911642392673e-06, "loss": 0.1827, "step": 12139 }, { "epoch": 5.712941176470588, "grad_norm": 0.39318911316273997, "learning_rate": 2.377132114040459e-06, "loss": 0.1392, "step": 12140 }, { "epoch": 5.713411764705882, "grad_norm": 0.4002292907139873, "learning_rate": 2.369365193405248e-06, "loss": 0.2304, "step": 12141 }, { "epoch": 5.713882352941177, "grad_norm": 0.4425000560646051, "learning_rate": 2.3616108811503775e-06, "loss": 0.1846, "step": 12142 }, { "epoch": 5.714352941176471, "grad_norm": 0.4321966161754647, "learning_rate": 2.3538691779381003e-06, "loss": 0.2261, "step": 12143 }, { "epoch": 5.714823529411765, "grad_norm": 0.5286826281758431, "learning_rate": 2.346140084429604e-06, "loss": 0.1328, "step": 12144 }, { "epoch": 5.715294117647058, "grad_norm": 0.15815943101828084, "learning_rate": 2.3384236012849778e-06, "loss": 0.0374, "step": 12145 }, { "epoch": 5.715764705882353, "grad_norm": 0.2010123572063485, "learning_rate": 2.330719729163244e-06, "loss": 0.0611, "step": 12146 }, { "epoch": 5.716235294117647, "grad_norm": 0.3950346912779776, "learning_rate": 2.32302846872236e-06, "loss": 0.1207, "step": 12147 }, { "epoch": 5.716705882352941, "grad_norm": 0.47226705059697466, "learning_rate": 2.3153498206192e-06, "loss": 0.397, "step": 12148 }, { "epoch": 5.717176470588235, "grad_norm": 0.21092873230686224, "learning_rate": 2.3076837855095563e-06, "loss": 0.0434, "step": 12149 }, { "epoch": 5.7176470588235295, "grad_norm": 1.0227019221561984, "learning_rate": 2.3000303640481386e-06, "loss": 0.4674, "step": 12150 }, { "epoch": 5.718117647058824, "grad_norm": 0.23099945759522647, "learning_rate": 2.292389556888591e-06, "loss": 0.0637, "step": 12151 }, { "epoch": 5.718588235294118, "grad_norm": 0.5072363285119175, "learning_rate": 2.2847613646834905e-06, "loss": 0.3351, "step": 12152 }, { "epoch": 5.719058823529412, "grad_norm": 0.30576369945017295, "learning_rate": 2.2771457880842834e-06, "loss": 0.1606, "step": 12153 }, { "epoch": 5.7195294117647055, "grad_norm": 0.2997455808245731, "learning_rate": 2.2695428277414165e-06, "loss": 0.0932, "step": 12154 }, { "epoch": 5.72, "grad_norm": 0.5156949423127966, "learning_rate": 2.261952484304219e-06, "loss": 0.2523, "step": 12155 }, { "epoch": 5.720470588235294, "grad_norm": 0.2190191984631647, "learning_rate": 2.254374758420924e-06, "loss": 0.0925, "step": 12156 }, { "epoch": 5.720941176470588, "grad_norm": 0.34808974501483336, "learning_rate": 2.2468096507386956e-06, "loss": 0.1072, "step": 12157 }, { "epoch": 5.721411764705882, "grad_norm": 0.498970821068712, "learning_rate": 2.239257161903668e-06, "loss": 0.1478, "step": 12158 }, { "epoch": 5.721882352941177, "grad_norm": 0.3442952919928759, "learning_rate": 2.231717292560825e-06, "loss": 0.114, "step": 12159 }, { "epoch": 5.722352941176471, "grad_norm": 0.381357235650271, "learning_rate": 2.224190043354135e-06, "loss": 0.2039, "step": 12160 }, { "epoch": 5.722823529411765, "grad_norm": 0.04299840983190299, "learning_rate": 2.2166754149264667e-06, "loss": 0.0024, "step": 12161 }, { "epoch": 5.723294117647059, "grad_norm": 0.6317467182583631, "learning_rate": 2.2091734079195565e-06, "loss": 0.3761, "step": 12162 }, { "epoch": 5.723764705882353, "grad_norm": 0.11455075882732489, "learning_rate": 2.2016840229741593e-06, "loss": 0.0237, "step": 12163 }, { "epoch": 5.724235294117647, "grad_norm": 0.2221052162069878, "learning_rate": 2.1942072607299135e-06, "loss": 0.0601, "step": 12164 }, { "epoch": 5.724705882352941, "grad_norm": 0.4611319042272, "learning_rate": 2.1867431218253406e-06, "loss": 0.2009, "step": 12165 }, { "epoch": 5.725176470588235, "grad_norm": 0.39936287033222956, "learning_rate": 2.179291606897915e-06, "loss": 0.2376, "step": 12166 }, { "epoch": 5.7256470588235295, "grad_norm": 0.3678034346142254, "learning_rate": 2.171852716584044e-06, "loss": 0.1867, "step": 12167 }, { "epoch": 5.726117647058824, "grad_norm": 0.4497135811727665, "learning_rate": 2.1644264515190524e-06, "loss": 0.2246, "step": 12168 }, { "epoch": 5.726588235294118, "grad_norm": 0.5310231598776098, "learning_rate": 2.1570128123371667e-06, "loss": 0.1886, "step": 12169 }, { "epoch": 5.727058823529411, "grad_norm": 0.37471924869693174, "learning_rate": 2.1496117996715636e-06, "loss": 0.2562, "step": 12170 }, { "epoch": 5.7275294117647055, "grad_norm": 0.4548567680464673, "learning_rate": 2.142223414154287e-06, "loss": 0.2097, "step": 12171 }, { "epoch": 5.728, "grad_norm": 0.4157461981728918, "learning_rate": 2.1348476564163997e-06, "loss": 0.2352, "step": 12172 }, { "epoch": 5.728470588235294, "grad_norm": 0.40256441834341944, "learning_rate": 2.1274845270877804e-06, "loss": 0.2249, "step": 12173 }, { "epoch": 5.728941176470588, "grad_norm": 0.4199026647992453, "learning_rate": 2.1201340267972933e-06, "loss": 0.1954, "step": 12174 }, { "epoch": 5.729411764705882, "grad_norm": 0.3405979244551161, "learning_rate": 2.1127961561727193e-06, "loss": 0.0978, "step": 12175 }, { "epoch": 5.729882352941177, "grad_norm": 0.2612391144610253, "learning_rate": 2.1054709158407235e-06, "loss": 0.0661, "step": 12176 }, { "epoch": 5.730352941176471, "grad_norm": 0.5237320891910938, "learning_rate": 2.0981583064269226e-06, "loss": 0.3398, "step": 12177 }, { "epoch": 5.730823529411765, "grad_norm": 0.4310431622326833, "learning_rate": 2.0908583285558834e-06, "loss": 0.1667, "step": 12178 }, { "epoch": 5.731294117647058, "grad_norm": 0.5323190992907549, "learning_rate": 2.083570982851007e-06, "loss": 0.3272, "step": 12179 }, { "epoch": 5.731764705882353, "grad_norm": 0.4780214460274693, "learning_rate": 2.0762962699346965e-06, "loss": 0.2403, "step": 12180 }, { "epoch": 5.732235294117647, "grad_norm": 0.47371076368895226, "learning_rate": 2.06903419042827e-06, "loss": 0.1807, "step": 12181 }, { "epoch": 5.732705882352941, "grad_norm": 0.25070714123887333, "learning_rate": 2.0617847449519e-06, "loss": 0.0555, "step": 12182 }, { "epoch": 5.733176470588235, "grad_norm": 0.39803204450661944, "learning_rate": 2.054547934124756e-06, "loss": 0.0851, "step": 12183 }, { "epoch": 5.7336470588235295, "grad_norm": 0.7894903916979802, "learning_rate": 2.047323758564878e-06, "loss": 0.1381, "step": 12184 }, { "epoch": 5.734117647058824, "grad_norm": 0.42567757462693007, "learning_rate": 2.040112218889256e-06, "loss": 0.1992, "step": 12185 }, { "epoch": 5.734588235294118, "grad_norm": 0.3390490671373284, "learning_rate": 2.0329133157137965e-06, "loss": 0.1253, "step": 12186 }, { "epoch": 5.735058823529412, "grad_norm": 0.33083481163956746, "learning_rate": 2.0257270496533086e-06, "loss": 0.1152, "step": 12187 }, { "epoch": 5.7355294117647055, "grad_norm": 0.40163783206714393, "learning_rate": 2.0185534213215506e-06, "loss": 0.1884, "step": 12188 }, { "epoch": 5.736, "grad_norm": 0.2963888288676682, "learning_rate": 2.011392431331166e-06, "loss": 0.1033, "step": 12189 }, { "epoch": 5.736470588235294, "grad_norm": 0.09143463398007748, "learning_rate": 2.0042440802937485e-06, "loss": 0.0118, "step": 12190 }, { "epoch": 5.736941176470588, "grad_norm": 0.25907009412811244, "learning_rate": 1.9971083688197943e-06, "loss": 0.0674, "step": 12191 }, { "epoch": 5.737411764705882, "grad_norm": 0.22145470584964688, "learning_rate": 1.989985297518748e-06, "loss": 0.0333, "step": 12192 }, { "epoch": 5.737882352941177, "grad_norm": 0.3910960364508917, "learning_rate": 1.9828748669989405e-06, "loss": 0.107, "step": 12193 }, { "epoch": 5.738352941176471, "grad_norm": 0.3154675080391808, "learning_rate": 1.9757770778676183e-06, "loss": 0.0896, "step": 12194 }, { "epoch": 5.738823529411764, "grad_norm": 0.1858643894702649, "learning_rate": 1.9686919307310145e-06, "loss": 0.0427, "step": 12195 }, { "epoch": 5.739294117647058, "grad_norm": 0.547520484622163, "learning_rate": 1.961619426194194e-06, "loss": 0.2888, "step": 12196 }, { "epoch": 5.739764705882353, "grad_norm": 0.4355778010461193, "learning_rate": 1.9545595648612078e-06, "loss": 0.112, "step": 12197 }, { "epoch": 5.740235294117647, "grad_norm": 0.34537217348294363, "learning_rate": 1.9475123473349897e-06, "loss": 0.1098, "step": 12198 }, { "epoch": 5.740705882352941, "grad_norm": 0.6267839617566252, "learning_rate": 1.9404777742174083e-06, "loss": 0.3418, "step": 12199 }, { "epoch": 5.741176470588235, "grad_norm": 0.46761926521258573, "learning_rate": 1.9334558461092663e-06, "loss": 0.1682, "step": 12200 }, { "epoch": 5.7416470588235295, "grad_norm": 0.5470545039920593, "learning_rate": 1.9264465636102342e-06, "loss": 0.258, "step": 12201 }, { "epoch": 5.742117647058824, "grad_norm": 0.49085651584573853, "learning_rate": 1.9194499273189834e-06, "loss": 0.1592, "step": 12202 }, { "epoch": 5.742588235294118, "grad_norm": 0.16756710422228335, "learning_rate": 1.912465937833035e-06, "loss": 0.0249, "step": 12203 }, { "epoch": 5.743058823529411, "grad_norm": 0.5100024300988586, "learning_rate": 1.9054945957488466e-06, "loss": 0.2561, "step": 12204 }, { "epoch": 5.7435294117647056, "grad_norm": 0.3877347927266225, "learning_rate": 1.898535901661824e-06, "loss": 0.1598, "step": 12205 }, { "epoch": 5.744, "grad_norm": 0.3167341128407802, "learning_rate": 1.8915898561662923e-06, "loss": 0.0458, "step": 12206 }, { "epoch": 5.744470588235294, "grad_norm": 0.28824167967122005, "learning_rate": 1.884656459855427e-06, "loss": 0.0841, "step": 12207 }, { "epoch": 5.744941176470588, "grad_norm": 0.4106110091599334, "learning_rate": 1.8777357133214044e-06, "loss": 0.2674, "step": 12208 }, { "epoch": 5.745411764705882, "grad_norm": 0.44876970157989965, "learning_rate": 1.8708276171553017e-06, "loss": 0.2542, "step": 12209 }, { "epoch": 5.745882352941177, "grad_norm": 0.38685804102323446, "learning_rate": 1.86393217194708e-06, "loss": 0.1195, "step": 12210 }, { "epoch": 5.746352941176471, "grad_norm": 0.3947345049828282, "learning_rate": 1.857049378285652e-06, "loss": 0.1754, "step": 12211 }, { "epoch": 5.746823529411765, "grad_norm": 0.30061685093245205, "learning_rate": 1.8501792367588642e-06, "loss": 0.0885, "step": 12212 }, { "epoch": 5.747294117647058, "grad_norm": 0.5102759470471817, "learning_rate": 1.84332174795343e-06, "loss": 0.2878, "step": 12213 }, { "epoch": 5.747764705882353, "grad_norm": 0.54393177201954, "learning_rate": 1.8364769124550149e-06, "loss": 0.3211, "step": 12214 }, { "epoch": 5.748235294117647, "grad_norm": 0.46002354192608136, "learning_rate": 1.8296447308482176e-06, "loss": 0.1541, "step": 12215 }, { "epoch": 5.748705882352941, "grad_norm": 0.6654389157473971, "learning_rate": 1.8228252037165215e-06, "loss": 0.2442, "step": 12216 }, { "epoch": 5.749176470588235, "grad_norm": 0.48030227875822773, "learning_rate": 1.8160183316423605e-06, "loss": 0.1666, "step": 12217 }, { "epoch": 5.7496470588235296, "grad_norm": 0.5255385487093384, "learning_rate": 1.8092241152070696e-06, "loss": 0.4132, "step": 12218 }, { "epoch": 5.750117647058824, "grad_norm": 0.4990297223842684, "learning_rate": 1.8024425549909183e-06, "loss": 0.2627, "step": 12219 }, { "epoch": 5.750588235294118, "grad_norm": 0.30228721073372494, "learning_rate": 1.795673651573093e-06, "loss": 0.1318, "step": 12220 }, { "epoch": 5.751058823529412, "grad_norm": 0.3963506374588191, "learning_rate": 1.788917405531648e-06, "loss": 0.2255, "step": 12221 }, { "epoch": 5.751529411764706, "grad_norm": 0.5066016852848748, "learning_rate": 1.7821738174436384e-06, "loss": 0.3191, "step": 12222 }, { "epoch": 5.752, "grad_norm": 0.29302807514104023, "learning_rate": 1.7754428878849868e-06, "loss": 0.0752, "step": 12223 }, { "epoch": 5.752470588235294, "grad_norm": 0.3082059906308489, "learning_rate": 1.7687246174305502e-06, "loss": 0.1361, "step": 12224 }, { "epoch": 5.752941176470588, "grad_norm": 0.27497719671984333, "learning_rate": 1.762019006654103e-06, "loss": 0.1082, "step": 12225 }, { "epoch": 5.7534117647058824, "grad_norm": 0.528551225778353, "learning_rate": 1.7553260561283533e-06, "loss": 0.1641, "step": 12226 }, { "epoch": 5.753882352941177, "grad_norm": 0.20079562844514015, "learning_rate": 1.748645766424861e-06, "loss": 0.0338, "step": 12227 }, { "epoch": 5.754352941176471, "grad_norm": 0.3743547176970686, "learning_rate": 1.7419781381142028e-06, "loss": 0.2048, "step": 12228 }, { "epoch": 5.754823529411764, "grad_norm": 0.39671544379998747, "learning_rate": 1.7353231717658233e-06, "loss": 0.1802, "step": 12229 }, { "epoch": 5.7552941176470584, "grad_norm": 0.2945938487147145, "learning_rate": 1.728680867948068e-06, "loss": 0.1294, "step": 12230 }, { "epoch": 5.755764705882353, "grad_norm": 0.23753489281526127, "learning_rate": 1.7220512272282328e-06, "loss": 0.1202, "step": 12231 }, { "epoch": 5.756235294117647, "grad_norm": 0.5105384114676466, "learning_rate": 1.7154342501725316e-06, "loss": 0.3166, "step": 12232 }, { "epoch": 5.756705882352941, "grad_norm": 0.5469437240578532, "learning_rate": 1.708829937346079e-06, "loss": 0.0995, "step": 12233 }, { "epoch": 5.757176470588235, "grad_norm": 0.6602247529374244, "learning_rate": 1.7022382893129072e-06, "loss": 0.4761, "step": 12234 }, { "epoch": 5.75764705882353, "grad_norm": 0.6897297655836678, "learning_rate": 1.6956593066359825e-06, "loss": 0.2704, "step": 12235 }, { "epoch": 5.758117647058824, "grad_norm": 0.46267179176522116, "learning_rate": 1.6890929898771722e-06, "loss": 0.2073, "step": 12236 }, { "epoch": 5.758588235294118, "grad_norm": 0.4166635730922703, "learning_rate": 1.6825393395972942e-06, "loss": 0.1979, "step": 12237 }, { "epoch": 5.759058823529411, "grad_norm": 0.33309097711314084, "learning_rate": 1.6759983563560508e-06, "loss": 0.102, "step": 12238 }, { "epoch": 5.759529411764706, "grad_norm": 0.39871627531126863, "learning_rate": 1.669470040712062e-06, "loss": 0.1714, "step": 12239 }, { "epoch": 5.76, "grad_norm": 0.3179166884214627, "learning_rate": 1.6629543932228817e-06, "loss": 0.0949, "step": 12240 }, { "epoch": 5.760470588235294, "grad_norm": 0.4158216824978592, "learning_rate": 1.6564514144449982e-06, "loss": 0.1528, "step": 12241 }, { "epoch": 5.760941176470588, "grad_norm": 0.39807994356425613, "learning_rate": 1.6499611049337835e-06, "loss": 0.2213, "step": 12242 }, { "epoch": 5.7614117647058825, "grad_norm": 0.4602792623901965, "learning_rate": 1.6434834652435447e-06, "loss": 0.2656, "step": 12243 }, { "epoch": 5.761882352941177, "grad_norm": 0.3828425597444992, "learning_rate": 1.6370184959275058e-06, "loss": 0.1614, "step": 12244 }, { "epoch": 5.762352941176471, "grad_norm": 0.11322340798770739, "learning_rate": 1.6305661975377916e-06, "loss": 0.0199, "step": 12245 }, { "epoch": 5.762823529411765, "grad_norm": 0.34859448698845713, "learning_rate": 1.6241265706254947e-06, "loss": 0.1446, "step": 12246 }, { "epoch": 5.7632941176470585, "grad_norm": 0.25431712749634466, "learning_rate": 1.617699615740542e-06, "loss": 0.0478, "step": 12247 }, { "epoch": 5.763764705882353, "grad_norm": 0.32793865047394205, "learning_rate": 1.6112853334318775e-06, "loss": 0.1171, "step": 12248 }, { "epoch": 5.764235294117647, "grad_norm": 0.12726041748566724, "learning_rate": 1.6048837242472634e-06, "loss": 0.0243, "step": 12249 }, { "epoch": 5.764705882352941, "grad_norm": 0.30663086606871665, "learning_rate": 1.598494788733462e-06, "loss": 0.1261, "step": 12250 }, { "epoch": 5.765176470588235, "grad_norm": 0.29812892112226774, "learning_rate": 1.5921185274361203e-06, "loss": 0.0584, "step": 12251 }, { "epoch": 5.76564705882353, "grad_norm": 0.4368144277417818, "learning_rate": 1.585754940899786e-06, "loss": 0.167, "step": 12252 }, { "epoch": 5.766117647058824, "grad_norm": 0.22422733712673104, "learning_rate": 1.5794040296679411e-06, "loss": 0.0726, "step": 12253 }, { "epoch": 5.766588235294118, "grad_norm": 0.4465628587697366, "learning_rate": 1.5730657942830183e-06, "loss": 0.3207, "step": 12254 }, { "epoch": 5.767058823529412, "grad_norm": 0.30813313484754157, "learning_rate": 1.5667402352862846e-06, "loss": 0.1045, "step": 12255 }, { "epoch": 5.767529411764706, "grad_norm": 0.5196795387776865, "learning_rate": 1.5604273532179913e-06, "loss": 0.3546, "step": 12256 }, { "epoch": 5.768, "grad_norm": 0.6340120555508291, "learning_rate": 1.554127148617307e-06, "loss": 0.377, "step": 12257 }, { "epoch": 5.768470588235294, "grad_norm": 0.55524915123121, "learning_rate": 1.5478396220222678e-06, "loss": 0.2345, "step": 12258 }, { "epoch": 5.768941176470588, "grad_norm": 0.3980788478313158, "learning_rate": 1.5415647739698944e-06, "loss": 0.1823, "step": 12259 }, { "epoch": 5.7694117647058825, "grad_norm": 0.45930902349782604, "learning_rate": 1.5353026049960748e-06, "loss": 0.2732, "step": 12260 }, { "epoch": 5.769882352941177, "grad_norm": 0.5883604352140864, "learning_rate": 1.5290531156356146e-06, "loss": 0.2995, "step": 12261 }, { "epoch": 5.770352941176471, "grad_norm": 0.36078791601571425, "learning_rate": 1.5228163064222698e-06, "loss": 0.1362, "step": 12262 }, { "epoch": 5.770823529411764, "grad_norm": 0.5605720515511645, "learning_rate": 1.516592177888698e-06, "loss": 0.354, "step": 12263 }, { "epoch": 5.7712941176470585, "grad_norm": 0.4249309007396193, "learning_rate": 1.5103807305664573e-06, "loss": 0.2617, "step": 12264 }, { "epoch": 5.771764705882353, "grad_norm": 0.5659654834501456, "learning_rate": 1.5041819649860399e-06, "loss": 0.2334, "step": 12265 }, { "epoch": 5.772235294117647, "grad_norm": 0.3572523703092294, "learning_rate": 1.4979958816768556e-06, "loss": 0.1373, "step": 12266 }, { "epoch": 5.772705882352941, "grad_norm": 0.3440265718111436, "learning_rate": 1.4918224811672153e-06, "loss": 0.1625, "step": 12267 }, { "epoch": 5.773176470588235, "grad_norm": 0.42059089610327555, "learning_rate": 1.485661763984347e-06, "loss": 0.2433, "step": 12268 }, { "epoch": 5.77364705882353, "grad_norm": 0.3042643633472815, "learning_rate": 1.479513730654447e-06, "loss": 0.0648, "step": 12269 }, { "epoch": 5.774117647058824, "grad_norm": 0.34531528924492455, "learning_rate": 1.4733783817025613e-06, "loss": 0.111, "step": 12270 }, { "epoch": 5.774588235294118, "grad_norm": 0.435339888270501, "learning_rate": 1.4672557176526712e-06, "loss": 0.2187, "step": 12271 }, { "epoch": 5.775058823529411, "grad_norm": 0.4067130029272295, "learning_rate": 1.4611457390276915e-06, "loss": 0.2435, "step": 12272 }, { "epoch": 5.775529411764706, "grad_norm": 0.33509915027773857, "learning_rate": 1.4550484463494384e-06, "loss": 0.1549, "step": 12273 }, { "epoch": 5.776, "grad_norm": 0.2885151610424307, "learning_rate": 1.4489638401386783e-06, "loss": 0.067, "step": 12274 }, { "epoch": 5.776470588235294, "grad_norm": 0.5038273445739218, "learning_rate": 1.442891920915029e-06, "loss": 0.2644, "step": 12275 }, { "epoch": 5.776941176470588, "grad_norm": 0.5676821324631598, "learning_rate": 1.436832689197076e-06, "loss": 0.3028, "step": 12276 }, { "epoch": 5.7774117647058825, "grad_norm": 0.32535003305240257, "learning_rate": 1.4307861455023218e-06, "loss": 0.1012, "step": 12277 }, { "epoch": 5.777882352941177, "grad_norm": 0.44057188655163704, "learning_rate": 1.4247522903471531e-06, "loss": 0.1162, "step": 12278 }, { "epoch": 5.778352941176471, "grad_norm": 0.4946867472271462, "learning_rate": 1.4187311242468914e-06, "loss": 0.3426, "step": 12279 }, { "epoch": 5.778823529411765, "grad_norm": 0.4642819142551971, "learning_rate": 1.4127226477157915e-06, "loss": 0.2675, "step": 12280 }, { "epoch": 5.7792941176470585, "grad_norm": 0.5030754497933887, "learning_rate": 1.4067268612669769e-06, "loss": 0.2654, "step": 12281 }, { "epoch": 5.779764705882353, "grad_norm": 0.45818813177363493, "learning_rate": 1.4007437654125375e-06, "loss": 0.1588, "step": 12282 }, { "epoch": 5.780235294117647, "grad_norm": 0.4837292049515313, "learning_rate": 1.3947733606634647e-06, "loss": 0.2473, "step": 12283 }, { "epoch": 5.780705882352941, "grad_norm": 0.6652360816566028, "learning_rate": 1.388815647529634e-06, "loss": 0.2106, "step": 12284 }, { "epoch": 5.781176470588235, "grad_norm": 0.3110120201747281, "learning_rate": 1.3828706265199051e-06, "loss": 0.0917, "step": 12285 }, { "epoch": 5.78164705882353, "grad_norm": 0.4642128641717683, "learning_rate": 1.376938298141972e-06, "loss": 0.1758, "step": 12286 }, { "epoch": 5.782117647058824, "grad_norm": 0.41138818899562385, "learning_rate": 1.3710186629025121e-06, "loss": 0.0652, "step": 12287 }, { "epoch": 5.782588235294118, "grad_norm": 0.5879763375352142, "learning_rate": 1.3651117213070716e-06, "loss": 0.2398, "step": 12288 }, { "epoch": 5.783058823529412, "grad_norm": 0.16260798859727776, "learning_rate": 1.3592174738601302e-06, "loss": 0.0426, "step": 12289 }, { "epoch": 5.783529411764706, "grad_norm": 0.285457436101118, "learning_rate": 1.3533359210651184e-06, "loss": 0.0833, "step": 12290 }, { "epoch": 5.784, "grad_norm": 0.4251487734488235, "learning_rate": 1.3474670634243178e-06, "loss": 0.1405, "step": 12291 }, { "epoch": 5.784470588235294, "grad_norm": 0.4831901724378473, "learning_rate": 1.341610901438961e-06, "loss": 0.268, "step": 12292 }, { "epoch": 5.784941176470588, "grad_norm": 0.3384811874171769, "learning_rate": 1.3357674356091974e-06, "loss": 0.1816, "step": 12293 }, { "epoch": 5.7854117647058825, "grad_norm": 0.30231950584236844, "learning_rate": 1.3299366664340948e-06, "loss": 0.0907, "step": 12294 }, { "epoch": 5.785882352941177, "grad_norm": 0.41160970933782587, "learning_rate": 1.3241185944116217e-06, "loss": 0.2051, "step": 12295 }, { "epoch": 5.786352941176471, "grad_norm": 0.28251407474620277, "learning_rate": 1.3183132200386637e-06, "loss": 0.1004, "step": 12296 }, { "epoch": 5.786823529411764, "grad_norm": 0.573811939351539, "learning_rate": 1.3125205438110408e-06, "loss": 0.27, "step": 12297 }, { "epoch": 5.7872941176470585, "grad_norm": 0.34171566320252594, "learning_rate": 1.3067405662234743e-06, "loss": 0.0816, "step": 12298 }, { "epoch": 5.787764705882353, "grad_norm": 0.5908081058697748, "learning_rate": 1.3009732877696022e-06, "loss": 0.3795, "step": 12299 }, { "epoch": 5.788235294117647, "grad_norm": 0.5054959131019396, "learning_rate": 1.2952187089419642e-06, "loss": 0.3088, "step": 12300 }, { "epoch": 5.788705882352941, "grad_norm": 0.30324762460206894, "learning_rate": 1.28947683023205e-06, "loss": 0.118, "step": 12301 }, { "epoch": 5.789176470588235, "grad_norm": 0.3662748557741967, "learning_rate": 1.2837476521302514e-06, "loss": 0.1894, "step": 12302 }, { "epoch": 5.78964705882353, "grad_norm": 0.3851874597208231, "learning_rate": 1.2780311751258265e-06, "loss": 0.1768, "step": 12303 }, { "epoch": 5.790117647058824, "grad_norm": 0.3256765590944982, "learning_rate": 1.2723273997070182e-06, "loss": 0.1563, "step": 12304 }, { "epoch": 5.790588235294118, "grad_norm": 0.8240936239405204, "learning_rate": 1.2666363263609702e-06, "loss": 0.3405, "step": 12305 }, { "epoch": 5.791058823529411, "grad_norm": 0.435134296955456, "learning_rate": 1.2609579555737104e-06, "loss": 0.182, "step": 12306 }, { "epoch": 5.791529411764706, "grad_norm": 0.5529388836466014, "learning_rate": 1.2552922878301841e-06, "loss": 0.2591, "step": 12307 }, { "epoch": 5.792, "grad_norm": 0.33275588523924315, "learning_rate": 1.2496393236143044e-06, "loss": 0.0926, "step": 12308 }, { "epoch": 5.792470588235294, "grad_norm": 0.3439159676794316, "learning_rate": 1.2439990634088348e-06, "loss": 0.1749, "step": 12309 }, { "epoch": 5.792941176470588, "grad_norm": 0.24949464702080656, "learning_rate": 1.2383715076954902e-06, "loss": 0.0823, "step": 12310 }, { "epoch": 5.7934117647058825, "grad_norm": 0.47750250583989357, "learning_rate": 1.232756656954903e-06, "loss": 0.2209, "step": 12311 }, { "epoch": 5.793882352941177, "grad_norm": 0.12823514648358342, "learning_rate": 1.2271545116665727e-06, "loss": 0.0265, "step": 12312 }, { "epoch": 5.794352941176471, "grad_norm": 0.4110620435763575, "learning_rate": 1.2215650723089832e-06, "loss": 0.1286, "step": 12313 }, { "epoch": 5.794823529411765, "grad_norm": 0.47237445970225217, "learning_rate": 1.2159883393595026e-06, "loss": 0.233, "step": 12314 }, { "epoch": 5.7952941176470585, "grad_norm": 0.45722281998632136, "learning_rate": 1.2104243132943836e-06, "loss": 0.1417, "step": 12315 }, { "epoch": 5.795764705882353, "grad_norm": 0.4025493968118455, "learning_rate": 1.2048729945888458e-06, "loss": 0.1667, "step": 12316 }, { "epoch": 5.796235294117647, "grad_norm": 0.5450897161109743, "learning_rate": 1.1993343837169933e-06, "loss": 0.2284, "step": 12317 }, { "epoch": 5.796705882352941, "grad_norm": 0.33171707821170765, "learning_rate": 1.1938084811518311e-06, "loss": 0.1894, "step": 12318 }, { "epoch": 5.797176470588235, "grad_norm": 0.3297977721886466, "learning_rate": 1.1882952873653317e-06, "loss": 0.1768, "step": 12319 }, { "epoch": 5.79764705882353, "grad_norm": 0.39940128296569555, "learning_rate": 1.1827948028283352e-06, "loss": 0.1837, "step": 12320 }, { "epoch": 5.798117647058824, "grad_norm": 0.4177750408275451, "learning_rate": 1.1773070280105822e-06, "loss": 0.2345, "step": 12321 }, { "epoch": 5.798588235294117, "grad_norm": 0.3542187826548442, "learning_rate": 1.171831963380815e-06, "loss": 0.1344, "step": 12322 }, { "epoch": 5.799058823529411, "grad_norm": 0.4328367075605091, "learning_rate": 1.1663696094065755e-06, "loss": 0.2131, "step": 12323 }, { "epoch": 5.799529411764706, "grad_norm": 0.4095954515348485, "learning_rate": 1.160919966554391e-06, "loss": 0.1306, "step": 12324 }, { "epoch": 5.8, "grad_norm": 0.3560976257495609, "learning_rate": 1.155483035289706e-06, "loss": 0.1544, "step": 12325 }, { "epoch": 5.800470588235294, "grad_norm": 0.3811830912389316, "learning_rate": 1.150058816076832e-06, "loss": 0.1883, "step": 12326 }, { "epoch": 5.800941176470588, "grad_norm": 0.46637977896716076, "learning_rate": 1.1446473093790487e-06, "loss": 0.3052, "step": 12327 }, { "epoch": 5.8014117647058825, "grad_norm": 0.9271367532753095, "learning_rate": 1.1392485156585196e-06, "loss": 0.4072, "step": 12328 }, { "epoch": 5.801882352941177, "grad_norm": 0.2312461834627152, "learning_rate": 1.1338624353763093e-06, "loss": 0.0564, "step": 12329 }, { "epoch": 5.802352941176471, "grad_norm": 1.136644811811799, "learning_rate": 1.1284890689924497e-06, "loss": 0.2318, "step": 12330 }, { "epoch": 5.802823529411764, "grad_norm": 0.3073715144015624, "learning_rate": 1.1231284169658239e-06, "loss": 0.1533, "step": 12331 }, { "epoch": 5.8032941176470585, "grad_norm": 1.004471516893751, "learning_rate": 1.1177804797542821e-06, "loss": 0.0984, "step": 12332 }, { "epoch": 5.803764705882353, "grad_norm": 0.47019671974622973, "learning_rate": 1.1124452578145425e-06, "loss": 0.2167, "step": 12333 }, { "epoch": 5.804235294117647, "grad_norm": 0.5911294818721665, "learning_rate": 1.1071227516022574e-06, "loss": 0.2677, "step": 12334 }, { "epoch": 5.804705882352941, "grad_norm": 0.08370841621366282, "learning_rate": 1.1018129615720129e-06, "loss": 0.0093, "step": 12335 }, { "epoch": 5.805176470588235, "grad_norm": 0.31964937979525015, "learning_rate": 1.0965158881772961e-06, "loss": 0.0727, "step": 12336 }, { "epoch": 5.80564705882353, "grad_norm": 0.46674799786838367, "learning_rate": 1.091231531870479e-06, "loss": 0.1732, "step": 12337 }, { "epoch": 5.806117647058824, "grad_norm": 0.20533689483374448, "learning_rate": 1.0859598931028668e-06, "loss": 0.0436, "step": 12338 }, { "epoch": 5.806588235294118, "grad_norm": 0.32694673872569996, "learning_rate": 1.0807009723247328e-06, "loss": 0.1601, "step": 12339 }, { "epoch": 5.807058823529411, "grad_norm": 0.3428868932935534, "learning_rate": 1.0754547699851513e-06, "loss": 0.0877, "step": 12340 }, { "epoch": 5.807529411764706, "grad_norm": 0.5965767243063327, "learning_rate": 1.0702212865322135e-06, "loss": 0.258, "step": 12341 }, { "epoch": 5.808, "grad_norm": 1.2205492746493818, "learning_rate": 1.0650005224128788e-06, "loss": 0.3766, "step": 12342 }, { "epoch": 5.808470588235294, "grad_norm": 0.3665666053702097, "learning_rate": 1.059792478073007e-06, "loss": 0.1755, "step": 12343 }, { "epoch": 5.808941176470588, "grad_norm": 0.37493534964376296, "learning_rate": 1.0545971539574094e-06, "loss": 0.1146, "step": 12344 }, { "epoch": 5.8094117647058825, "grad_norm": 0.42899783991956714, "learning_rate": 1.0494145505097973e-06, "loss": 0.2559, "step": 12345 }, { "epoch": 5.809882352941177, "grad_norm": 0.3239500235170833, "learning_rate": 1.0442446681727834e-06, "loss": 0.1406, "step": 12346 }, { "epoch": 5.810352941176471, "grad_norm": 0.1356980497509674, "learning_rate": 1.0390875073878812e-06, "loss": 0.0298, "step": 12347 }, { "epoch": 5.810823529411765, "grad_norm": 0.1622366524733969, "learning_rate": 1.0339430685955719e-06, "loss": 0.0218, "step": 12348 }, { "epoch": 5.8112941176470585, "grad_norm": 0.340234579349746, "learning_rate": 1.0288113522352038e-06, "loss": 0.1229, "step": 12349 }, { "epoch": 5.811764705882353, "grad_norm": 0.05440532939641402, "learning_rate": 1.0236923587450263e-06, "loss": 0.0057, "step": 12350 }, { "epoch": 5.812235294117647, "grad_norm": 0.2979174311776874, "learning_rate": 1.0185860885622565e-06, "loss": 0.0847, "step": 12351 }, { "epoch": 5.812705882352941, "grad_norm": 0.7135168531997668, "learning_rate": 1.013492542122979e-06, "loss": 0.2565, "step": 12352 }, { "epoch": 5.813176470588235, "grad_norm": 0.4509462998049467, "learning_rate": 1.0084117198622287e-06, "loss": 0.1768, "step": 12353 }, { "epoch": 5.81364705882353, "grad_norm": 0.5953865521306603, "learning_rate": 1.0033436222138924e-06, "loss": 0.3284, "step": 12354 }, { "epoch": 5.814117647058824, "grad_norm": 0.317501692344246, "learning_rate": 9.98288249610857e-07, "loss": 0.1052, "step": 12355 }, { "epoch": 5.814588235294117, "grad_norm": 0.4377928666811095, "learning_rate": 9.932456024848435e-07, "loss": 0.2203, "step": 12356 }, { "epoch": 5.815058823529411, "grad_norm": 0.33890863704930485, "learning_rate": 9.882156812665244e-07, "loss": 0.1274, "step": 12357 }, { "epoch": 5.815529411764706, "grad_norm": 0.5944377189822255, "learning_rate": 9.831984863854891e-07, "loss": 0.3138, "step": 12358 }, { "epoch": 5.816, "grad_norm": 0.5145492449491817, "learning_rate": 9.781940182702451e-07, "loss": 0.1791, "step": 12359 }, { "epoch": 5.816470588235294, "grad_norm": 0.20429344222865725, "learning_rate": 9.732022773481506e-07, "loss": 0.065, "step": 12360 }, { "epoch": 5.816941176470588, "grad_norm": 0.6737789162435037, "learning_rate": 9.682232640455644e-07, "loss": 0.2077, "step": 12361 }, { "epoch": 5.8174117647058825, "grad_norm": 0.5373777998708084, "learning_rate": 9.632569787877298e-07, "loss": 0.2227, "step": 12362 }, { "epoch": 5.817882352941177, "grad_norm": 0.34389440513600694, "learning_rate": 9.583034219987406e-07, "loss": 0.1269, "step": 12363 }, { "epoch": 5.818352941176471, "grad_norm": 0.36041550613934314, "learning_rate": 9.53362594101692e-07, "loss": 0.1978, "step": 12364 }, { "epoch": 5.818823529411764, "grad_norm": 0.3614267144148638, "learning_rate": 9.48434495518563e-07, "loss": 0.2064, "step": 12365 }, { "epoch": 5.8192941176470585, "grad_norm": 0.16679579726159718, "learning_rate": 9.435191266702003e-07, "loss": 0.0285, "step": 12366 }, { "epoch": 5.819764705882353, "grad_norm": 0.39331317893891304, "learning_rate": 9.386164879764346e-07, "loss": 0.0741, "step": 12367 }, { "epoch": 5.820235294117647, "grad_norm": 0.3840675962900643, "learning_rate": 9.337265798559645e-07, "loss": 0.1942, "step": 12368 }, { "epoch": 5.820705882352941, "grad_norm": 0.64183767148517, "learning_rate": 9.28849402726406e-07, "loss": 0.1975, "step": 12369 }, { "epoch": 5.821176470588235, "grad_norm": 0.4719331914676973, "learning_rate": 9.239849570042924e-07, "loss": 0.2678, "step": 12370 }, { "epoch": 5.82164705882353, "grad_norm": 0.5872780987824416, "learning_rate": 9.191332431050747e-07, "loss": 0.281, "step": 12371 }, { "epoch": 5.822117647058824, "grad_norm": 0.5094834874775562, "learning_rate": 9.142942614431215e-07, "loss": 0.125, "step": 12372 }, { "epoch": 5.822588235294118, "grad_norm": 0.5982247760165464, "learning_rate": 9.094680124317022e-07, "loss": 0.3087, "step": 12373 }, { "epoch": 5.823058823529411, "grad_norm": 0.39447426555455084, "learning_rate": 9.046544964830038e-07, "loss": 0.1593, "step": 12374 }, { "epoch": 5.823529411764706, "grad_norm": 0.4680009267317047, "learning_rate": 8.998537140081141e-07, "loss": 0.1816, "step": 12375 }, { "epoch": 5.824, "grad_norm": 0.43751752572717306, "learning_rate": 8.950656654170552e-07, "loss": 0.2152, "step": 12376 }, { "epoch": 5.824470588235294, "grad_norm": 0.4049253649643877, "learning_rate": 8.902903511187332e-07, "loss": 0.2177, "step": 12377 }, { "epoch": 5.824941176470588, "grad_norm": 0.39622707286397524, "learning_rate": 8.855277715210052e-07, "loss": 0.2327, "step": 12378 }, { "epoch": 5.8254117647058825, "grad_norm": 0.054180309385957824, "learning_rate": 8.807779270305959e-07, "loss": 0.0037, "step": 12379 }, { "epoch": 5.825882352941177, "grad_norm": 0.6673624585640222, "learning_rate": 8.760408180531809e-07, "loss": 0.2974, "step": 12380 }, { "epoch": 5.826352941176471, "grad_norm": 0.4768627376811324, "learning_rate": 8.713164449933364e-07, "loss": 0.2764, "step": 12381 }, { "epoch": 5.826823529411765, "grad_norm": 0.36166371031712735, "learning_rate": 8.666048082545396e-07, "loss": 0.0998, "step": 12382 }, { "epoch": 5.8272941176470585, "grad_norm": 0.3665475149616975, "learning_rate": 8.619059082391689e-07, "loss": 0.1814, "step": 12383 }, { "epoch": 5.827764705882353, "grad_norm": 0.42025306523210954, "learning_rate": 8.572197453485696e-07, "loss": 0.2445, "step": 12384 }, { "epoch": 5.828235294117647, "grad_norm": 0.5623534951994106, "learning_rate": 8.525463199829385e-07, "loss": 0.2946, "step": 12385 }, { "epoch": 5.828705882352941, "grad_norm": 0.33688789858131546, "learning_rate": 8.47885632541423e-07, "loss": 0.0998, "step": 12386 }, { "epoch": 5.829176470588235, "grad_norm": 0.4835024143114279, "learning_rate": 8.432376834220544e-07, "loss": 0.1992, "step": 12387 }, { "epoch": 5.82964705882353, "grad_norm": 0.2484951019577272, "learning_rate": 8.386024730217988e-07, "loss": 0.0488, "step": 12388 }, { "epoch": 5.830117647058824, "grad_norm": 0.2756554437662093, "learning_rate": 8.339800017365228e-07, "loss": 0.0956, "step": 12389 }, { "epoch": 5.830588235294117, "grad_norm": 0.534164495743422, "learning_rate": 8.293702699610272e-07, "loss": 0.3044, "step": 12390 }, { "epoch": 5.831058823529411, "grad_norm": 0.755863439193739, "learning_rate": 8.247732780889637e-07, "loss": 0.4261, "step": 12391 }, { "epoch": 5.831529411764706, "grad_norm": 0.43815558557670636, "learning_rate": 8.201890265129684e-07, "loss": 0.2545, "step": 12392 }, { "epoch": 5.832, "grad_norm": 0.349652514796142, "learning_rate": 8.156175156245614e-07, "loss": 0.1605, "step": 12393 }, { "epoch": 5.832470588235294, "grad_norm": 0.3589802017936206, "learning_rate": 8.110587458141637e-07, "loss": 0.1468, "step": 12394 }, { "epoch": 5.832941176470588, "grad_norm": 0.506560178465911, "learning_rate": 8.065127174711139e-07, "loss": 0.2369, "step": 12395 }, { "epoch": 5.8334117647058825, "grad_norm": 0.40950508242087275, "learning_rate": 8.01979430983668e-07, "loss": 0.1643, "step": 12396 }, { "epoch": 5.833882352941177, "grad_norm": 0.2516238899352039, "learning_rate": 7.97458886738983e-07, "loss": 0.0778, "step": 12397 }, { "epoch": 5.834352941176471, "grad_norm": 0.5008447661339626, "learning_rate": 7.929510851231502e-07, "loss": 0.1271, "step": 12398 }, { "epoch": 5.834823529411764, "grad_norm": 0.8436889759409725, "learning_rate": 7.884560265211615e-07, "loss": 0.2925, "step": 12399 }, { "epoch": 5.8352941176470585, "grad_norm": 0.24991497833896967, "learning_rate": 7.839737113168931e-07, "loss": 0.057, "step": 12400 }, { "epoch": 5.835764705882353, "grad_norm": 0.16895124558542604, "learning_rate": 7.795041398931723e-07, "loss": 0.0469, "step": 12401 }, { "epoch": 5.836235294117647, "grad_norm": 0.2987007347901206, "learning_rate": 7.750473126317269e-07, "loss": 0.0859, "step": 12402 }, { "epoch": 5.836705882352941, "grad_norm": 0.3961732628641651, "learning_rate": 7.706032299131692e-07, "loss": 0.2445, "step": 12403 }, { "epoch": 5.837176470588235, "grad_norm": 0.4287124537971291, "learning_rate": 7.66171892117079e-07, "loss": 0.0945, "step": 12404 }, { "epoch": 5.83764705882353, "grad_norm": 0.28497982081232565, "learning_rate": 7.617532996218867e-07, "loss": 0.1118, "step": 12405 }, { "epoch": 5.838117647058824, "grad_norm": 0.35569276386750975, "learning_rate": 7.573474528049739e-07, "loss": 0.1696, "step": 12406 }, { "epoch": 5.838588235294118, "grad_norm": 0.5373092048282134, "learning_rate": 7.529543520426228e-07, "loss": 0.2353, "step": 12407 }, { "epoch": 5.839058823529411, "grad_norm": 0.5510971366569588, "learning_rate": 7.485739977100336e-07, "loss": 0.2755, "step": 12408 }, { "epoch": 5.839529411764706, "grad_norm": 0.3042809490746334, "learning_rate": 7.442063901812734e-07, "loss": 0.1189, "step": 12409 }, { "epoch": 5.84, "grad_norm": 0.3133803938438597, "learning_rate": 7.398515298294106e-07, "loss": 0.1025, "step": 12410 }, { "epoch": 5.840470588235294, "grad_norm": 0.4631215907086382, "learning_rate": 7.355094170263143e-07, "loss": 0.1377, "step": 12411 }, { "epoch": 5.840941176470588, "grad_norm": 0.2969218577786666, "learning_rate": 7.311800521428712e-07, "loss": 0.0547, "step": 12412 }, { "epoch": 5.8414117647058825, "grad_norm": 0.3621375270866556, "learning_rate": 7.26863435548819e-07, "loss": 0.2338, "step": 12413 }, { "epoch": 5.841882352941177, "grad_norm": 0.5278414774829868, "learning_rate": 7.22559567612796e-07, "loss": 0.3047, "step": 12414 }, { "epoch": 5.842352941176471, "grad_norm": 0.42133551943312686, "learning_rate": 7.182684487023748e-07, "loss": 0.186, "step": 12415 }, { "epoch": 5.842823529411765, "grad_norm": 0.4228703100586571, "learning_rate": 7.139900791840792e-07, "loss": 0.177, "step": 12416 }, { "epoch": 5.8432941176470585, "grad_norm": 0.23831344335620616, "learning_rate": 7.097244594232498e-07, "loss": 0.0354, "step": 12417 }, { "epoch": 5.843764705882353, "grad_norm": 0.41007462540144246, "learning_rate": 7.054715897842289e-07, "loss": 0.1135, "step": 12418 }, { "epoch": 5.844235294117647, "grad_norm": 0.26413463610620935, "learning_rate": 7.012314706302257e-07, "loss": 0.1091, "step": 12419 }, { "epoch": 5.844705882352941, "grad_norm": 0.5482650961607559, "learning_rate": 6.970041023233508e-07, "loss": 0.2626, "step": 12420 }, { "epoch": 5.845176470588235, "grad_norm": 0.7825065420728761, "learning_rate": 6.927894852246485e-07, "loss": 0.2266, "step": 12421 }, { "epoch": 5.84564705882353, "grad_norm": 0.4388625803591499, "learning_rate": 6.88587619694081e-07, "loss": 0.2163, "step": 12422 }, { "epoch": 5.846117647058824, "grad_norm": 0.4492144479737783, "learning_rate": 6.843985060904944e-07, "loss": 0.2539, "step": 12423 }, { "epoch": 5.846588235294117, "grad_norm": 0.44634260559038275, "learning_rate": 6.802221447716694e-07, "loss": 0.2294, "step": 12424 }, { "epoch": 5.847058823529411, "grad_norm": 0.28052883179508464, "learning_rate": 6.760585360942872e-07, "loss": 0.0737, "step": 12425 }, { "epoch": 5.847529411764706, "grad_norm": 0.32041205969869446, "learning_rate": 6.719076804139134e-07, "loss": 0.1383, "step": 12426 }, { "epoch": 5.848, "grad_norm": 0.2810452667304543, "learning_rate": 6.677695780850978e-07, "loss": 0.1028, "step": 12427 }, { "epoch": 5.848470588235294, "grad_norm": 0.4856525824299952, "learning_rate": 6.636442294612243e-07, "loss": 0.2027, "step": 12428 }, { "epoch": 5.848941176470588, "grad_norm": 0.4578828086439143, "learning_rate": 6.595316348946278e-07, "loss": 0.1228, "step": 12429 }, { "epoch": 5.8494117647058825, "grad_norm": 0.35464864156996573, "learning_rate": 6.554317947365273e-07, "loss": 0.1603, "step": 12430 }, { "epoch": 5.849882352941177, "grad_norm": 0.7980402366011341, "learning_rate": 6.513447093370927e-07, "loss": 0.5533, "step": 12431 }, { "epoch": 5.850352941176471, "grad_norm": 0.3497727994198488, "learning_rate": 6.472703790453781e-07, "loss": 0.0934, "step": 12432 }, { "epoch": 5.850823529411764, "grad_norm": 0.48008905584529044, "learning_rate": 6.43208804209322e-07, "loss": 0.1308, "step": 12433 }, { "epoch": 5.8512941176470585, "grad_norm": 0.4533932897630135, "learning_rate": 6.391599851758466e-07, "loss": 0.2319, "step": 12434 }, { "epoch": 5.851764705882353, "grad_norm": 0.5136087707271776, "learning_rate": 6.351239222907089e-07, "loss": 0.2175, "step": 12435 }, { "epoch": 5.852235294117647, "grad_norm": 0.6693924337254467, "learning_rate": 6.311006158986165e-07, "loss": 0.378, "step": 12436 }, { "epoch": 5.852705882352941, "grad_norm": 0.670166349447949, "learning_rate": 6.270900663431777e-07, "loss": 0.2323, "step": 12437 }, { "epoch": 5.853176470588235, "grad_norm": 0.3048578617801998, "learning_rate": 6.230922739669353e-07, "loss": 0.1391, "step": 12438 }, { "epoch": 5.85364705882353, "grad_norm": 0.6475593718619959, "learning_rate": 6.191072391112828e-07, "loss": 0.3604, "step": 12439 }, { "epoch": 5.854117647058824, "grad_norm": 0.4661882988220373, "learning_rate": 6.151349621165979e-07, "loss": 0.2527, "step": 12440 }, { "epoch": 5.854588235294118, "grad_norm": 0.5338864605008747, "learning_rate": 6.111754433221094e-07, "loss": 0.2169, "step": 12441 }, { "epoch": 5.855058823529411, "grad_norm": 0.4339112823025102, "learning_rate": 6.0722868306598e-07, "loss": 0.1141, "step": 12442 }, { "epoch": 5.855529411764706, "grad_norm": 0.4696136268964804, "learning_rate": 6.032946816852902e-07, "loss": 0.2238, "step": 12443 }, { "epoch": 5.856, "grad_norm": 0.9145131292500703, "learning_rate": 5.99373439516021e-07, "loss": 0.5513, "step": 12444 }, { "epoch": 5.856470588235294, "grad_norm": 0.4081106361707383, "learning_rate": 5.954649568930714e-07, "loss": 0.1743, "step": 12445 }, { "epoch": 5.856941176470588, "grad_norm": 0.5315231812977067, "learning_rate": 5.91569234150241e-07, "loss": 0.3203, "step": 12446 }, { "epoch": 5.8574117647058825, "grad_norm": 0.35150070261288285, "learning_rate": 5.876862716202302e-07, "loss": 0.1155, "step": 12447 }, { "epoch": 5.857882352941177, "grad_norm": 0.502237798229033, "learning_rate": 5.838160696346906e-07, "loss": 0.2503, "step": 12448 }, { "epoch": 5.85835294117647, "grad_norm": 0.5236140834416976, "learning_rate": 5.799586285241242e-07, "loss": 0.3354, "step": 12449 }, { "epoch": 5.858823529411764, "grad_norm": 0.5320419568981516, "learning_rate": 5.761139486180178e-07, "loss": 0.277, "step": 12450 }, { "epoch": 5.8592941176470585, "grad_norm": 0.5078388320536629, "learning_rate": 5.722820302446751e-07, "loss": 0.2206, "step": 12451 }, { "epoch": 5.859764705882353, "grad_norm": 0.538514945730946, "learning_rate": 5.684628737314012e-07, "loss": 0.2458, "step": 12452 }, { "epoch": 5.860235294117647, "grad_norm": 0.3142375819112392, "learning_rate": 5.646564794043518e-07, "loss": 0.1364, "step": 12453 }, { "epoch": 5.860705882352941, "grad_norm": 0.32291834860592816, "learning_rate": 5.608628475886001e-07, "loss": 0.144, "step": 12454 }, { "epoch": 5.861176470588235, "grad_norm": 0.4734174547257439, "learning_rate": 5.570819786081704e-07, "loss": 0.2329, "step": 12455 }, { "epoch": 5.86164705882353, "grad_norm": 0.3570799956030496, "learning_rate": 5.533138727859543e-07, "loss": 0.1012, "step": 12456 }, { "epoch": 5.862117647058824, "grad_norm": 0.7470478488987372, "learning_rate": 5.495585304437611e-07, "loss": 0.2326, "step": 12457 }, { "epoch": 5.862588235294117, "grad_norm": 0.38424580788909274, "learning_rate": 5.458159519023175e-07, "loss": 0.2138, "step": 12458 }, { "epoch": 5.863058823529411, "grad_norm": 0.4072666049689284, "learning_rate": 5.42086137481268e-07, "loss": 0.1844, "step": 12459 }, { "epoch": 5.863529411764706, "grad_norm": 0.4050286052015678, "learning_rate": 5.383690874991408e-07, "loss": 0.1372, "step": 12460 }, { "epoch": 5.864, "grad_norm": 0.5508480638968949, "learning_rate": 5.34664802273399e-07, "loss": 0.2063, "step": 12461 }, { "epoch": 5.864470588235294, "grad_norm": 0.7770240552815074, "learning_rate": 5.30973282120406e-07, "loss": 0.2708, "step": 12462 }, { "epoch": 5.864941176470588, "grad_norm": 0.1830329226229569, "learning_rate": 5.272945273554429e-07, "loss": 0.0356, "step": 12463 }, { "epoch": 5.8654117647058825, "grad_norm": 0.323321238815173, "learning_rate": 5.236285382926752e-07, "loss": 0.0612, "step": 12464 }, { "epoch": 5.865882352941177, "grad_norm": 0.2834817864097676, "learning_rate": 5.19975315245219e-07, "loss": 0.0995, "step": 12465 }, { "epoch": 5.866352941176471, "grad_norm": 0.2945004881292919, "learning_rate": 5.163348585250748e-07, "loss": 0.1064, "step": 12466 }, { "epoch": 5.866823529411764, "grad_norm": 0.4494236050355192, "learning_rate": 5.127071684431272e-07, "loss": 0.2059, "step": 12467 }, { "epoch": 5.8672941176470585, "grad_norm": 0.5390747925598715, "learning_rate": 5.090922453092283e-07, "loss": 0.3503, "step": 12468 }, { "epoch": 5.867764705882353, "grad_norm": 0.6001782868347395, "learning_rate": 5.054900894320979e-07, "loss": 0.3253, "step": 12469 }, { "epoch": 5.868235294117647, "grad_norm": 0.09020303908662548, "learning_rate": 5.019007011193732e-07, "loss": 0.0147, "step": 12470 }, { "epoch": 5.868705882352941, "grad_norm": 0.8271273469718815, "learning_rate": 4.983240806776257e-07, "loss": 0.106, "step": 12471 }, { "epoch": 5.869176470588235, "grad_norm": 0.27348333028463045, "learning_rate": 4.947602284122942e-07, "loss": 0.0586, "step": 12472 }, { "epoch": 5.86964705882353, "grad_norm": 0.48666878795871726, "learning_rate": 4.912091446277688e-07, "loss": 0.2643, "step": 12473 }, { "epoch": 5.870117647058824, "grad_norm": 0.3950088202985532, "learning_rate": 4.876708296273068e-07, "loss": 0.1752, "step": 12474 }, { "epoch": 5.870588235294118, "grad_norm": 0.4743350952353156, "learning_rate": 4.841452837131166e-07, "loss": 0.1722, "step": 12475 }, { "epoch": 5.871058823529411, "grad_norm": 0.1832952811212393, "learning_rate": 4.806325071862905e-07, "loss": 0.0519, "step": 12476 }, { "epoch": 5.871529411764706, "grad_norm": 0.3936273767640116, "learning_rate": 4.771325003468385e-07, "loss": 0.1458, "step": 12477 }, { "epoch": 5.872, "grad_norm": 0.4762031211210788, "learning_rate": 4.736452634936716e-07, "loss": 0.2537, "step": 12478 }, { "epoch": 5.872470588235294, "grad_norm": 0.39013718743773557, "learning_rate": 4.701707969246182e-07, "loss": 0.218, "step": 12479 }, { "epoch": 5.872941176470588, "grad_norm": 0.3835834993269191, "learning_rate": 4.6670910093642435e-07, "loss": 0.1685, "step": 12480 }, { "epoch": 5.8734117647058826, "grad_norm": 0.4876877310175919, "learning_rate": 4.6326017582473674e-07, "loss": 0.2754, "step": 12481 }, { "epoch": 5.873882352941177, "grad_norm": 0.39361868553209894, "learning_rate": 4.5982402188408653e-07, "loss": 0.1479, "step": 12482 }, { "epoch": 5.87435294117647, "grad_norm": 0.704994031141779, "learning_rate": 4.5640063940797223e-07, "loss": 0.1354, "step": 12483 }, { "epoch": 5.874823529411764, "grad_norm": 0.20283284730564077, "learning_rate": 4.529900286887433e-07, "loss": 0.0264, "step": 12484 }, { "epoch": 5.875294117647059, "grad_norm": 0.20736664829578763, "learning_rate": 4.4959219001768355e-07, "loss": 0.0672, "step": 12485 }, { "epoch": 5.875764705882353, "grad_norm": 0.5318653988269917, "learning_rate": 4.4620712368501065e-07, "loss": 0.1354, "step": 12486 }, { "epoch": 5.876235294117647, "grad_norm": 0.3076256993836966, "learning_rate": 4.428348299797768e-07, "loss": 0.0735, "step": 12487 }, { "epoch": 5.876705882352941, "grad_norm": 0.5605864461461738, "learning_rate": 4.394753091900516e-07, "loss": 0.4117, "step": 12488 }, { "epoch": 5.8771764705882354, "grad_norm": 0.4559810147849918, "learning_rate": 4.361285616027055e-07, "loss": 0.2255, "step": 12489 }, { "epoch": 5.87764705882353, "grad_norm": 0.4257971471012315, "learning_rate": 4.327945875035932e-07, "loss": 0.2232, "step": 12490 }, { "epoch": 5.878117647058824, "grad_norm": 0.34507867807574694, "learning_rate": 4.2947338717743694e-07, "loss": 0.0947, "step": 12491 }, { "epoch": 5.878588235294117, "grad_norm": 0.40301102750597295, "learning_rate": 4.2616496090790983e-07, "loss": 0.2019, "step": 12492 }, { "epoch": 5.8790588235294114, "grad_norm": 0.29179392917896885, "learning_rate": 4.228693089775526e-07, "loss": 0.0563, "step": 12493 }, { "epoch": 5.879529411764706, "grad_norm": 0.49140681194475583, "learning_rate": 4.195864316678066e-07, "loss": 0.2428, "step": 12494 }, { "epoch": 5.88, "grad_norm": 0.4534115378351574, "learning_rate": 4.16316329259081e-07, "loss": 0.2254, "step": 12495 }, { "epoch": 5.880470588235294, "grad_norm": 0.48092931124931876, "learning_rate": 4.130590020306523e-07, "loss": 0.1509, "step": 12496 }, { "epoch": 5.880941176470588, "grad_norm": 0.5273909783606819, "learning_rate": 4.0981445026068147e-07, "loss": 0.3675, "step": 12497 }, { "epoch": 5.881411764705883, "grad_norm": 0.4022791214357403, "learning_rate": 4.0658267422631343e-07, "loss": 0.204, "step": 12498 }, { "epoch": 5.881882352941177, "grad_norm": 0.20636803599045989, "learning_rate": 4.033636742035273e-07, "loss": 0.0512, "step": 12499 }, { "epoch": 5.882352941176471, "grad_norm": 0.366311860818471, "learning_rate": 4.0015745046725336e-07, "loss": 0.1529, "step": 12500 }, { "epoch": 5.882823529411764, "grad_norm": 0.3704023047722576, "learning_rate": 3.9696400329132236e-07, "loss": 0.1491, "step": 12501 }, { "epoch": 5.883294117647059, "grad_norm": 0.32443604997642, "learning_rate": 3.9378333294844967e-07, "loss": 0.1327, "step": 12502 }, { "epoch": 5.883764705882353, "grad_norm": 0.3675925192798372, "learning_rate": 3.906154397103012e-07, "loss": 0.1237, "step": 12503 }, { "epoch": 5.884235294117647, "grad_norm": 0.5336483517526178, "learning_rate": 3.874603238474272e-07, "loss": 0.2717, "step": 12504 }, { "epoch": 5.884705882352941, "grad_norm": 0.5876279765669866, "learning_rate": 3.8431798562927887e-07, "loss": 0.343, "step": 12505 }, { "epoch": 5.8851764705882355, "grad_norm": 0.5912388212770834, "learning_rate": 3.8118842532424145e-07, "loss": 0.1526, "step": 12506 }, { "epoch": 5.88564705882353, "grad_norm": 0.30504814904185396, "learning_rate": 3.780716431995845e-07, "loss": 0.0974, "step": 12507 }, { "epoch": 5.886117647058824, "grad_norm": 0.1685203979117645, "learning_rate": 3.7496763952151177e-07, "loss": 0.0342, "step": 12508 }, { "epoch": 5.886588235294118, "grad_norm": 0.4566077198746201, "learning_rate": 3.7187641455509455e-07, "loss": 0.1677, "step": 12509 }, { "epoch": 5.8870588235294115, "grad_norm": 6.915949669704572e-05, "learning_rate": 3.687979685643549e-07, "loss": 0.0, "step": 12510 }, { "epoch": 5.887529411764706, "grad_norm": 0.2708867313073871, "learning_rate": 3.6573230181221604e-07, "loss": 0.0773, "step": 12511 }, { "epoch": 5.888, "grad_norm": 0.5581888480135343, "learning_rate": 3.62679414560485e-07, "loss": 0.2523, "step": 12512 }, { "epoch": 5.888470588235294, "grad_norm": 0.44708898681253045, "learning_rate": 3.596393070699033e-07, "loss": 0.1827, "step": 12513 }, { "epoch": 5.888941176470588, "grad_norm": 0.25622238489596033, "learning_rate": 3.566119796000966e-07, "loss": 0.0845, "step": 12514 }, { "epoch": 5.889411764705883, "grad_norm": 0.44405479296329875, "learning_rate": 3.535974324096413e-07, "loss": 0.2058, "step": 12515 }, { "epoch": 5.889882352941177, "grad_norm": 0.43009280562549146, "learning_rate": 3.5059566575596497e-07, "loss": 0.137, "step": 12516 }, { "epoch": 5.89035294117647, "grad_norm": 0.3731583927482378, "learning_rate": 3.4760667989544575e-07, "loss": 0.1204, "step": 12517 }, { "epoch": 5.890823529411764, "grad_norm": 0.4279994633383302, "learning_rate": 3.446304750833462e-07, "loss": 0.2077, "step": 12518 }, { "epoch": 5.891294117647059, "grad_norm": 0.6633268152847723, "learning_rate": 3.416670515738629e-07, "loss": 0.3044, "step": 12519 }, { "epoch": 5.891764705882353, "grad_norm": 1.1015138726774203, "learning_rate": 3.387164096200934e-07, "loss": 0.4164, "step": 12520 }, { "epoch": 5.892235294117647, "grad_norm": 0.5449540818191345, "learning_rate": 3.3577854947403615e-07, "loss": 0.229, "step": 12521 }, { "epoch": 5.892705882352941, "grad_norm": 0.38579692483993716, "learning_rate": 3.3285347138657383e-07, "loss": 0.2238, "step": 12522 }, { "epoch": 5.8931764705882355, "grad_norm": 0.3348189052762599, "learning_rate": 3.2994117560753986e-07, "loss": 0.1678, "step": 12523 }, { "epoch": 5.89364705882353, "grad_norm": 0.35510264036001815, "learning_rate": 3.2704166238565197e-07, "loss": 0.1939, "step": 12524 }, { "epoch": 5.894117647058824, "grad_norm": 0.42383543175761274, "learning_rate": 3.241549319685621e-07, "loss": 0.2245, "step": 12525 }, { "epoch": 5.894588235294117, "grad_norm": 0.27915709098706976, "learning_rate": 3.212809846027897e-07, "loss": 0.0772, "step": 12526 }, { "epoch": 5.8950588235294115, "grad_norm": 0.4380312238379397, "learning_rate": 3.1841982053378846e-07, "loss": 0.1561, "step": 12527 }, { "epoch": 5.895529411764706, "grad_norm": 0.46894244745761365, "learning_rate": 3.155714400059128e-07, "loss": 0.1199, "step": 12528 }, { "epoch": 5.896, "grad_norm": 0.325344702784228, "learning_rate": 3.127358432624516e-07, "loss": 0.078, "step": 12529 }, { "epoch": 5.896470588235294, "grad_norm": 0.4975927362338577, "learning_rate": 3.099130305455444e-07, "loss": 0.1958, "step": 12530 }, { "epoch": 5.896941176470588, "grad_norm": 0.287406075522886, "learning_rate": 3.0710300209629836e-07, "loss": 0.095, "step": 12531 }, { "epoch": 5.897411764705883, "grad_norm": 0.44395130599452104, "learning_rate": 3.043057581547048e-07, "loss": 0.1784, "step": 12532 }, { "epoch": 5.897882352941177, "grad_norm": 0.42217225899498334, "learning_rate": 3.0152129895963937e-07, "loss": 0.129, "step": 12533 }, { "epoch": 5.898352941176471, "grad_norm": 0.4691510431899866, "learning_rate": 2.987496247489285e-07, "loss": 0.2673, "step": 12534 }, { "epoch": 5.898823529411764, "grad_norm": 0.8578012474367509, "learning_rate": 2.959907357592661e-07, "loss": 0.4639, "step": 12535 }, { "epoch": 5.899294117647059, "grad_norm": 0.28078272297578594, "learning_rate": 2.9324463222629715e-07, "loss": 0.1313, "step": 12536 }, { "epoch": 5.899764705882353, "grad_norm": 0.18766885883800868, "learning_rate": 2.9051131438455055e-07, "loss": 0.0518, "step": 12537 }, { "epoch": 5.900235294117647, "grad_norm": 0.6164798249612946, "learning_rate": 2.8779078246745635e-07, "loss": 0.3871, "step": 12538 }, { "epoch": 5.900705882352941, "grad_norm": 0.4570606207493512, "learning_rate": 2.85083036707362e-07, "loss": 0.292, "step": 12539 }, { "epoch": 5.9011764705882355, "grad_norm": 0.5297467493090203, "learning_rate": 2.8238807733551583e-07, "loss": 0.2532, "step": 12540 }, { "epoch": 5.90164705882353, "grad_norm": 0.39556271608671567, "learning_rate": 2.797059045821004e-07, "loss": 0.1765, "step": 12541 }, { "epoch": 5.902117647058824, "grad_norm": 0.3612406748423263, "learning_rate": 2.7703651867616583e-07, "loss": 0.1552, "step": 12542 }, { "epoch": 5.902588235294118, "grad_norm": 0.3071591479498154, "learning_rate": 2.743799198457131e-07, "loss": 0.081, "step": 12543 }, { "epoch": 5.9030588235294115, "grad_norm": 0.4781673701402833, "learning_rate": 2.71736108317594e-07, "loss": 0.1245, "step": 12544 }, { "epoch": 5.903529411764706, "grad_norm": 0.41424464993659016, "learning_rate": 2.6910508431764457e-07, "loss": 0.2079, "step": 12545 }, { "epoch": 5.904, "grad_norm": 2.562141721879202, "learning_rate": 2.6648684807053513e-07, "loss": 1.0521, "step": 12546 }, { "epoch": 5.904470588235294, "grad_norm": 0.5119671782620504, "learning_rate": 2.6388139979988677e-07, "loss": 0.246, "step": 12547 }, { "epoch": 5.904941176470588, "grad_norm": 0.3631297555316738, "learning_rate": 2.612887397282215e-07, "loss": 0.159, "step": 12548 }, { "epoch": 5.905411764705883, "grad_norm": 0.30332381162609, "learning_rate": 2.5870886807696225e-07, "loss": 0.0817, "step": 12549 }, { "epoch": 5.905882352941177, "grad_norm": 0.6035749841733056, "learning_rate": 2.5614178506644934e-07, "loss": 0.1942, "step": 12550 }, { "epoch": 5.90635294117647, "grad_norm": 0.5698094206521125, "learning_rate": 2.5358749091590746e-07, "loss": 0.4219, "step": 12551 }, { "epoch": 5.906823529411764, "grad_norm": 0.35904280455889553, "learning_rate": 2.5104598584349543e-07, "loss": 0.1675, "step": 12552 }, { "epoch": 5.907294117647059, "grad_norm": 0.34012660218173696, "learning_rate": 2.4851727006627297e-07, "loss": 0.1447, "step": 12553 }, { "epoch": 5.907764705882353, "grad_norm": 0.4953460768910073, "learning_rate": 2.460013438002173e-07, "loss": 0.1884, "step": 12554 }, { "epoch": 5.908235294117647, "grad_norm": 0.4423746406118466, "learning_rate": 2.434982072601732e-07, "loss": 0.0704, "step": 12555 }, { "epoch": 5.908705882352941, "grad_norm": 0.44324790665312636, "learning_rate": 2.4100786065991975e-07, "loss": 0.2418, "step": 12556 }, { "epoch": 5.9091764705882355, "grad_norm": 0.3956435717162078, "learning_rate": 2.3853030421218667e-07, "loss": 0.2495, "step": 12557 }, { "epoch": 5.90964705882353, "grad_norm": 0.4444051881190328, "learning_rate": 2.3606553812852146e-07, "loss": 0.141, "step": 12558 }, { "epoch": 5.910117647058824, "grad_norm": 0.4017737000904737, "learning_rate": 2.3361356261945575e-07, "loss": 0.2539, "step": 12559 }, { "epoch": 5.910588235294117, "grad_norm": 0.3712146696349644, "learning_rate": 2.3117437789438865e-07, "loss": 0.1048, "step": 12560 }, { "epoch": 5.9110588235294115, "grad_norm": 0.5975346162047074, "learning_rate": 2.2874798416163686e-07, "loss": 0.2015, "step": 12561 }, { "epoch": 5.911529411764706, "grad_norm": 0.2789102947731265, "learning_rate": 2.2633438162843465e-07, "loss": 0.0589, "step": 12562 }, { "epoch": 5.912, "grad_norm": 0.4696698055844325, "learning_rate": 2.239335705009171e-07, "loss": 0.2671, "step": 12563 }, { "epoch": 5.912470588235294, "grad_norm": 0.690590311443185, "learning_rate": 2.215455509841202e-07, "loss": 0.3711, "step": 12564 }, { "epoch": 5.912941176470588, "grad_norm": 0.49588141190220936, "learning_rate": 2.1917032328199747e-07, "loss": 0.3084, "step": 12565 }, { "epoch": 5.913411764705883, "grad_norm": 0.21711123325663137, "learning_rate": 2.1680788759738664e-07, "loss": 0.063, "step": 12566 }, { "epoch": 5.913882352941177, "grad_norm": 2.9618067352198483, "learning_rate": 2.144582441320597e-07, "loss": 1.0918, "step": 12567 }, { "epoch": 5.914352941176471, "grad_norm": 0.3687150624910495, "learning_rate": 2.1212139308670607e-07, "loss": 0.1572, "step": 12568 }, { "epoch": 5.914823529411764, "grad_norm": 0.5709648799031557, "learning_rate": 2.0979733466088277e-07, "loss": 0.2938, "step": 12569 }, { "epoch": 5.915294117647059, "grad_norm": 0.23416371733622585, "learning_rate": 2.0748606905309774e-07, "loss": 0.0671, "step": 12570 }, { "epoch": 5.915764705882353, "grad_norm": 0.26374559460727354, "learning_rate": 2.051875964607097e-07, "loss": 0.0522, "step": 12571 }, { "epoch": 5.916235294117647, "grad_norm": 0.19886279560281062, "learning_rate": 2.0290191708006165e-07, "loss": 0.0388, "step": 12572 }, { "epoch": 5.916705882352941, "grad_norm": 0.42067568878922973, "learning_rate": 2.0062903110633077e-07, "loss": 0.1567, "step": 12573 }, { "epoch": 5.9171764705882355, "grad_norm": 0.3985863016747597, "learning_rate": 1.9836893873364513e-07, "loss": 0.1804, "step": 12574 }, { "epoch": 5.91764705882353, "grad_norm": 0.3126949941795773, "learning_rate": 1.9612164015501696e-07, "loss": 0.1608, "step": 12575 }, { "epoch": 5.918117647058823, "grad_norm": 0.30786680177883324, "learning_rate": 1.9388713556239277e-07, "loss": 0.0733, "step": 12576 }, { "epoch": 5.918588235294117, "grad_norm": 0.2970290295661651, "learning_rate": 1.9166542514658656e-07, "loss": 0.0834, "step": 12577 }, { "epoch": 5.9190588235294115, "grad_norm": 0.4359907265931007, "learning_rate": 1.8945650909737985e-07, "loss": 0.1829, "step": 12578 }, { "epoch": 5.919529411764706, "grad_norm": 0.31044884182138727, "learning_rate": 1.8726038760338847e-07, "loss": 0.1356, "step": 12579 }, { "epoch": 5.92, "grad_norm": 0.5413189595149878, "learning_rate": 1.8507706085219566e-07, "loss": 0.1274, "step": 12580 }, { "epoch": 5.920470588235294, "grad_norm": 0.3100294732932705, "learning_rate": 1.8290652903025228e-07, "loss": 0.1741, "step": 12581 }, { "epoch": 5.920941176470588, "grad_norm": 0.6016711938580879, "learning_rate": 1.8074879232296002e-07, "loss": 0.2391, "step": 12582 }, { "epoch": 5.921411764705883, "grad_norm": 0.41086561606490096, "learning_rate": 1.7860385091455488e-07, "loss": 0.1808, "step": 12583 }, { "epoch": 5.921882352941177, "grad_norm": 0.41585099753837695, "learning_rate": 1.764717049882569e-07, "loss": 0.1847, "step": 12584 }, { "epoch": 5.92235294117647, "grad_norm": 0.5148272493575093, "learning_rate": 1.743523547261705e-07, "loss": 0.23, "step": 12585 }, { "epoch": 5.922823529411764, "grad_norm": 0.4317847333095888, "learning_rate": 1.722458003092675e-07, "loss": 0.2305, "step": 12586 }, { "epoch": 5.923294117647059, "grad_norm": 0.3891756786228416, "learning_rate": 1.701520419174873e-07, "loss": 0.1124, "step": 12587 }, { "epoch": 5.923764705882353, "grad_norm": 0.34558955408333253, "learning_rate": 1.680710797296203e-07, "loss": 0.0919, "step": 12588 }, { "epoch": 5.924235294117647, "grad_norm": 0.33076363829733757, "learning_rate": 1.660029139234076e-07, "loss": 0.1519, "step": 12589 }, { "epoch": 5.924705882352941, "grad_norm": 0.23760481848398513, "learning_rate": 1.639475446754912e-07, "loss": 0.0494, "step": 12590 }, { "epoch": 5.9251764705882355, "grad_norm": 0.4002283697908901, "learning_rate": 1.6190497216138076e-07, "loss": 0.1848, "step": 12591 }, { "epoch": 5.92564705882353, "grad_norm": 0.5767596248358092, "learning_rate": 1.598751965555367e-07, "loss": 0.3138, "step": 12592 }, { "epoch": 5.926117647058824, "grad_norm": 0.8184414736758399, "learning_rate": 1.5785821803132037e-07, "loss": 0.245, "step": 12593 }, { "epoch": 5.926588235294117, "grad_norm": 0.26852065554227034, "learning_rate": 1.558540367609773e-07, "loss": 0.0472, "step": 12594 }, { "epoch": 5.9270588235294115, "grad_norm": 0.30651926069517055, "learning_rate": 1.5386265291568723e-07, "loss": 0.0677, "step": 12595 }, { "epoch": 5.927529411764706, "grad_norm": 0.3863738703647947, "learning_rate": 1.5188406666551412e-07, "loss": 0.1026, "step": 12596 }, { "epoch": 5.928, "grad_norm": 0.43843814321463387, "learning_rate": 1.4991827817943947e-07, "loss": 0.2428, "step": 12597 }, { "epoch": 5.928470588235294, "grad_norm": 0.23136456001171776, "learning_rate": 1.479652876253623e-07, "loss": 0.0388, "step": 12598 }, { "epoch": 5.928941176470588, "grad_norm": 0.5028322470663013, "learning_rate": 1.4602509517006588e-07, "loss": 0.1926, "step": 12599 }, { "epoch": 5.929411764705883, "grad_norm": 0.2696054494422805, "learning_rate": 1.4409770097926765e-07, "loss": 0.0867, "step": 12600 }, { "epoch": 5.929882352941177, "grad_norm": 0.3117572064858016, "learning_rate": 1.421831052175526e-07, "loss": 0.1916, "step": 12601 }, { "epoch": 5.930352941176471, "grad_norm": 0.5681303157312482, "learning_rate": 1.4028130804843995e-07, "loss": 0.2454, "step": 12602 }, { "epoch": 5.930823529411764, "grad_norm": 0.4591862543126826, "learning_rate": 1.3839230963436642e-07, "loss": 0.202, "step": 12603 }, { "epoch": 5.931294117647059, "grad_norm": 0.5008449816719717, "learning_rate": 1.3651611013666964e-07, "loss": 0.281, "step": 12604 }, { "epoch": 5.931764705882353, "grad_norm": 0.2605633558232753, "learning_rate": 1.346527097155714e-07, "loss": 0.0555, "step": 12605 }, { "epoch": 5.932235294117647, "grad_norm": 0.19716918089520424, "learning_rate": 1.3280210853019447e-07, "loss": 0.0485, "step": 12606 }, { "epoch": 5.932705882352941, "grad_norm": 0.2649776423446085, "learning_rate": 1.3096430673862901e-07, "loss": 0.0639, "step": 12607 }, { "epoch": 5.9331764705882355, "grad_norm": 0.3916891027534406, "learning_rate": 1.2913930449781619e-07, "loss": 0.1502, "step": 12608 }, { "epoch": 5.93364705882353, "grad_norm": 0.06878998891681354, "learning_rate": 1.2732710196361463e-07, "loss": 0.0109, "step": 12609 }, { "epoch": 5.934117647058823, "grad_norm": 0.4207914228628074, "learning_rate": 1.2552769929081717e-07, "loss": 0.1573, "step": 12610 }, { "epoch": 5.934588235294117, "grad_norm": 0.47773270554877617, "learning_rate": 1.2374109663306763e-07, "loss": 0.2345, "step": 12611 }, { "epoch": 5.9350588235294115, "grad_norm": 0.32922201904096043, "learning_rate": 1.2196729414296058e-07, "loss": 0.1303, "step": 12612 }, { "epoch": 5.935529411764706, "grad_norm": 0.6609178125447679, "learning_rate": 1.2020629197200814e-07, "loss": 0.3975, "step": 12613 }, { "epoch": 5.936, "grad_norm": 0.3744390348544421, "learning_rate": 1.1845809027059006e-07, "loss": 0.2139, "step": 12614 }, { "epoch": 5.936470588235294, "grad_norm": 0.6341280181871066, "learning_rate": 1.1672268918800353e-07, "loss": 0.3323, "step": 12615 }, { "epoch": 5.936941176470588, "grad_norm": 0.7870556372742298, "learning_rate": 1.1500008887249668e-07, "loss": 0.4517, "step": 12616 }, { "epoch": 5.937411764705883, "grad_norm": 0.21901892634816403, "learning_rate": 1.1329028947115181e-07, "loss": 0.0472, "step": 12617 }, { "epoch": 5.937882352941177, "grad_norm": 0.4683660378694691, "learning_rate": 1.1159329113000215e-07, "loss": 0.2267, "step": 12618 }, { "epoch": 5.93835294117647, "grad_norm": 0.17341461588892718, "learning_rate": 1.0990909399398173e-07, "loss": 0.052, "step": 12619 }, { "epoch": 5.938823529411764, "grad_norm": 0.1634011257132534, "learning_rate": 1.0823769820694218e-07, "loss": 0.0436, "step": 12620 }, { "epoch": 5.939294117647059, "grad_norm": 0.47200480761491087, "learning_rate": 1.0657910391161928e-07, "loss": 0.2852, "step": 12621 }, { "epoch": 5.939764705882353, "grad_norm": 0.5262796613665418, "learning_rate": 1.0493331124964976e-07, "loss": 0.3354, "step": 12622 }, { "epoch": 5.940235294117647, "grad_norm": 0.41170261829034294, "learning_rate": 1.0330032036162117e-07, "loss": 0.1817, "step": 12623 }, { "epoch": 5.940705882352941, "grad_norm": 0.17819884244295492, "learning_rate": 1.0168013138697196e-07, "loss": 0.0205, "step": 12624 }, { "epoch": 5.9411764705882355, "grad_norm": 0.3034291550625384, "learning_rate": 1.0007274446409141e-07, "loss": 0.1417, "step": 12625 }, { "epoch": 5.94164705882353, "grad_norm": 0.400370913416935, "learning_rate": 9.847815973025308e-08, "loss": 0.1332, "step": 12626 }, { "epoch": 5.942117647058824, "grad_norm": 0.33135927832763606, "learning_rate": 9.689637732164802e-08, "loss": 0.1214, "step": 12627 }, { "epoch": 5.942588235294117, "grad_norm": 0.4142724692619003, "learning_rate": 9.532739737336814e-08, "loss": 0.2264, "step": 12628 }, { "epoch": 5.9430588235294115, "grad_norm": 0.5282289314426467, "learning_rate": 9.377122001938963e-08, "loss": 0.2402, "step": 12629 }, { "epoch": 5.943529411764706, "grad_norm": 0.4950463654379639, "learning_rate": 9.222784539263949e-08, "loss": 0.1363, "step": 12630 }, { "epoch": 5.944, "grad_norm": 0.3602078737919224, "learning_rate": 9.069727362492896e-08, "loss": 0.1687, "step": 12631 }, { "epoch": 5.944470588235294, "grad_norm": 0.3566697336249507, "learning_rate": 8.917950484697012e-08, "loss": 0.1162, "step": 12632 }, { "epoch": 5.944941176470588, "grad_norm": 0.30213340552335755, "learning_rate": 8.767453918839263e-08, "loss": 0.0892, "step": 12633 }, { "epoch": 5.945411764705883, "grad_norm": 0.2876586330341623, "learning_rate": 8.618237677772699e-08, "loss": 0.061, "step": 12634 }, { "epoch": 5.945882352941177, "grad_norm": 0.5075783712804034, "learning_rate": 8.470301774240463e-08, "loss": 0.3397, "step": 12635 }, { "epoch": 5.946352941176471, "grad_norm": 0.25851958819177606, "learning_rate": 8.323646220877444e-08, "loss": 0.0784, "step": 12636 }, { "epoch": 5.946823529411764, "grad_norm": 0.5247758109358625, "learning_rate": 8.178271030210292e-08, "loss": 0.2616, "step": 12637 }, { "epoch": 5.947294117647059, "grad_norm": 0.1943718142338147, "learning_rate": 8.034176214650745e-08, "loss": 0.0455, "step": 12638 }, { "epoch": 5.947764705882353, "grad_norm": 0.28742794395092186, "learning_rate": 7.891361786508953e-08, "loss": 0.0646, "step": 12639 }, { "epoch": 5.948235294117647, "grad_norm": 0.5267872624429979, "learning_rate": 7.749827757981831e-08, "loss": 0.3116, "step": 12640 }, { "epoch": 5.948705882352941, "grad_norm": 0.33514251249330324, "learning_rate": 7.609574141154706e-08, "loss": 0.1102, "step": 12641 }, { "epoch": 5.9491764705882355, "grad_norm": 0.28696961334844756, "learning_rate": 7.470600948006333e-08, "loss": 0.1358, "step": 12642 }, { "epoch": 5.94964705882353, "grad_norm": 0.6396465794944032, "learning_rate": 7.332908190408882e-08, "loss": 0.2296, "step": 12643 }, { "epoch": 5.950117647058823, "grad_norm": 0.443236365248776, "learning_rate": 7.196495880117947e-08, "loss": 0.1481, "step": 12644 }, { "epoch": 5.950588235294117, "grad_norm": 0.4169520009275197, "learning_rate": 7.061364028785876e-08, "loss": 0.1748, "step": 12645 }, { "epoch": 5.9510588235294115, "grad_norm": 0.44056000959017283, "learning_rate": 6.927512647953437e-08, "loss": 0.2451, "step": 12646 }, { "epoch": 5.951529411764706, "grad_norm": 1.111235597483251, "learning_rate": 6.794941749053151e-08, "loss": 0.3407, "step": 12647 }, { "epoch": 5.952, "grad_norm": 0.4402155388359179, "learning_rate": 6.663651343404297e-08, "loss": 0.2697, "step": 12648 }, { "epoch": 5.952470588235294, "grad_norm": 0.1713327172325056, "learning_rate": 6.533641442222904e-08, "loss": 0.0604, "step": 12649 }, { "epoch": 5.952941176470588, "grad_norm": 0.37472563532826125, "learning_rate": 6.40491205661009e-08, "loss": 0.1845, "step": 12650 }, { "epoch": 5.953411764705883, "grad_norm": 0.5155192837749001, "learning_rate": 6.277463197560395e-08, "loss": 0.2929, "step": 12651 }, { "epoch": 5.953882352941177, "grad_norm": 0.37943670983128064, "learning_rate": 6.151294875960111e-08, "loss": 0.1889, "step": 12652 }, { "epoch": 5.95435294117647, "grad_norm": 0.24513349171161328, "learning_rate": 6.026407102583952e-08, "loss": 0.0422, "step": 12653 }, { "epoch": 5.954823529411764, "grad_norm": 0.4449891188580757, "learning_rate": 5.902799888096721e-08, "loss": 0.1731, "step": 12654 }, { "epoch": 5.955294117647059, "grad_norm": 0.5456905709014903, "learning_rate": 5.780473243056638e-08, "loss": 0.1967, "step": 12655 }, { "epoch": 5.955764705882353, "grad_norm": 0.4809071988115002, "learning_rate": 5.659427177910347e-08, "loss": 0.2712, "step": 12656 }, { "epoch": 5.956235294117647, "grad_norm": 0.3652928148224184, "learning_rate": 5.539661702996245e-08, "loss": 0.1022, "step": 12657 }, { "epoch": 5.956705882352941, "grad_norm": 0.16657932009687842, "learning_rate": 5.421176828541152e-08, "loss": 0.0326, "step": 12658 }, { "epoch": 5.9571764705882355, "grad_norm": 0.27138251661918295, "learning_rate": 5.30397256466697e-08, "loss": 0.0608, "step": 12659 }, { "epoch": 5.95764705882353, "grad_norm": 0.47226350749096085, "learning_rate": 5.1880489213806944e-08, "loss": 0.2707, "step": 12660 }, { "epoch": 5.958117647058824, "grad_norm": 0.43244621325446286, "learning_rate": 5.0734059085860704e-08, "loss": 0.131, "step": 12661 }, { "epoch": 5.958588235294117, "grad_norm": 0.5213267780214141, "learning_rate": 4.9600435360702685e-08, "loss": 0.3252, "step": 12662 }, { "epoch": 5.9590588235294115, "grad_norm": 0.4541094904023527, "learning_rate": 4.847961813517209e-08, "loss": 0.1818, "step": 12663 }, { "epoch": 5.959529411764706, "grad_norm": 0.4531796776373933, "learning_rate": 4.737160750500901e-08, "loss": 0.1442, "step": 12664 }, { "epoch": 5.96, "grad_norm": 0.24709793898199986, "learning_rate": 4.6276403564804444e-08, "loss": 0.0933, "step": 12665 }, { "epoch": 5.960470588235294, "grad_norm": 0.5783793538509567, "learning_rate": 4.519400640811688e-08, "loss": 0.191, "step": 12666 }, { "epoch": 5.960941176470588, "grad_norm": 0.6907486914257231, "learning_rate": 4.412441612738904e-08, "loss": 0.1992, "step": 12667 }, { "epoch": 5.961411764705883, "grad_norm": 0.36668276978602743, "learning_rate": 4.3067632813964524e-08, "loss": 0.1213, "step": 12668 }, { "epoch": 5.961882352941177, "grad_norm": 0.5156826859182707, "learning_rate": 4.2023656558104466e-08, "loss": 0.245, "step": 12669 }, { "epoch": 5.962352941176471, "grad_norm": 0.5053477619156591, "learning_rate": 4.099248744897088e-08, "loss": 0.1747, "step": 12670 }, { "epoch": 5.962823529411764, "grad_norm": 0.4190326227381024, "learning_rate": 3.9974125574609994e-08, "loss": 0.1904, "step": 12671 }, { "epoch": 5.963294117647059, "grad_norm": 0.6364323614080616, "learning_rate": 3.896857102200224e-08, "loss": 0.2785, "step": 12672 }, { "epoch": 5.963764705882353, "grad_norm": 0.43156529112077, "learning_rate": 3.797582387704556e-08, "loss": 0.2341, "step": 12673 }, { "epoch": 5.964235294117647, "grad_norm": 0.42197002483499724, "learning_rate": 3.69958842245055e-08, "loss": 0.2371, "step": 12674 }, { "epoch": 5.964705882352941, "grad_norm": 0.35418700971905503, "learning_rate": 3.6028752148081766e-08, "loss": 0.1572, "step": 12675 }, { "epoch": 5.9651764705882355, "grad_norm": 0.22861222982782817, "learning_rate": 3.5074427730374945e-08, "loss": 0.0756, "step": 12676 }, { "epoch": 5.96564705882353, "grad_norm": 0.40080863646244175, "learning_rate": 3.4132911052886514e-08, "loss": 0.0815, "step": 12677 }, { "epoch": 5.966117647058823, "grad_norm": 0.6135188462699558, "learning_rate": 3.320420219603548e-08, "loss": 0.2858, "step": 12678 }, { "epoch": 5.966588235294117, "grad_norm": 0.2779108700690283, "learning_rate": 3.228830123910842e-08, "loss": 0.0789, "step": 12679 }, { "epoch": 5.9670588235294115, "grad_norm": 0.6905057068839608, "learning_rate": 3.13852082603594e-08, "loss": 0.1223, "step": 12680 }, { "epoch": 5.967529411764706, "grad_norm": 0.6057176494641224, "learning_rate": 3.049492333691006e-08, "loss": 0.372, "step": 12681 }, { "epoch": 5.968, "grad_norm": 0.3888165810642977, "learning_rate": 2.961744654478293e-08, "loss": 0.2351, "step": 12682 }, { "epoch": 5.968470588235294, "grad_norm": 0.4139811080634488, "learning_rate": 2.875277795891806e-08, "loss": 0.2361, "step": 12683 }, { "epoch": 5.968941176470588, "grad_norm": 0.32866266860113347, "learning_rate": 2.7900917653189692e-08, "loss": 0.1101, "step": 12684 }, { "epoch": 5.969411764705883, "grad_norm": 0.20118005079556645, "learning_rate": 2.706186570032298e-08, "loss": 0.0444, "step": 12685 }, { "epoch": 5.969882352941177, "grad_norm": 0.3522714669152879, "learning_rate": 2.6235622171977278e-08, "loss": 0.1612, "step": 12686 }, { "epoch": 5.97035294117647, "grad_norm": 0.38911337263947915, "learning_rate": 2.542218713874611e-08, "loss": 0.1941, "step": 12687 }, { "epoch": 5.970823529411764, "grad_norm": 0.5702975183885534, "learning_rate": 2.462156067005727e-08, "loss": 0.3842, "step": 12688 }, { "epoch": 5.971294117647059, "grad_norm": 0.5368720463742964, "learning_rate": 2.3833742834339365e-08, "loss": 0.234, "step": 12689 }, { "epoch": 5.971764705882353, "grad_norm": 0.3607368321480323, "learning_rate": 2.305873369882194e-08, "loss": 0.1936, "step": 12690 }, { "epoch": 5.972235294117647, "grad_norm": 0.41278687404367576, "learning_rate": 2.2296533329735356e-08, "loss": 0.1899, "step": 12691 }, { "epoch": 5.972705882352941, "grad_norm": 0.34698027784352214, "learning_rate": 2.1547141792160882e-08, "loss": 0.0878, "step": 12692 }, { "epoch": 5.9731764705882355, "grad_norm": 0.8184521982229178, "learning_rate": 2.0810559150097327e-08, "loss": 0.4492, "step": 12693 }, { "epoch": 5.97364705882353, "grad_norm": 0.49229142172128143, "learning_rate": 2.008678546646103e-08, "loss": 0.2169, "step": 12694 }, { "epoch": 5.974117647058824, "grad_norm": 0.384096499727702, "learning_rate": 1.9375820803052555e-08, "loss": 0.1557, "step": 12695 }, { "epoch": 5.974588235294117, "grad_norm": 0.4477410323851928, "learning_rate": 1.8677665220606653e-08, "loss": 0.2158, "step": 12696 }, { "epoch": 5.9750588235294115, "grad_norm": 0.5137025394405624, "learning_rate": 1.7992318778742298e-08, "loss": 0.2694, "step": 12697 }, { "epoch": 5.975529411764706, "grad_norm": 0.37483942461537667, "learning_rate": 1.7319781535996002e-08, "loss": 0.089, "step": 12698 }, { "epoch": 5.976, "grad_norm": 0.4433622561818996, "learning_rate": 1.6660053549788498e-08, "loss": 0.2847, "step": 12699 }, { "epoch": 5.976470588235294, "grad_norm": 0.2904126821094812, "learning_rate": 1.6013134876491362e-08, "loss": 0.0871, "step": 12700 }, { "epoch": 5.976941176470588, "grad_norm": 0.4397004330647622, "learning_rate": 1.5379025571327088e-08, "loss": 0.1856, "step": 12701 }, { "epoch": 5.977411764705883, "grad_norm": 0.42576678829475473, "learning_rate": 1.4757725688469002e-08, "loss": 0.1848, "step": 12702 }, { "epoch": 5.977882352941176, "grad_norm": 0.7042440003185594, "learning_rate": 1.414923528097467e-08, "loss": 0.2001, "step": 12703 }, { "epoch": 5.97835294117647, "grad_norm": 0.3321750458853403, "learning_rate": 1.3553554400802524e-08, "loss": 0.119, "step": 12704 }, { "epoch": 5.978823529411764, "grad_norm": 0.322182401087817, "learning_rate": 1.2970683098845192e-08, "loss": 0.1442, "step": 12705 }, { "epoch": 5.979294117647059, "grad_norm": 0.2629340929197358, "learning_rate": 1.240062142487952e-08, "loss": 0.0864, "step": 12706 }, { "epoch": 5.979764705882353, "grad_norm": 0.3607963154777243, "learning_rate": 1.1843369427583238e-08, "loss": 0.0751, "step": 12707 }, { "epoch": 5.980235294117647, "grad_norm": 0.43468542903379326, "learning_rate": 1.1298927154551608e-08, "loss": 0.2912, "step": 12708 }, { "epoch": 5.980705882352941, "grad_norm": 0.1113914654058489, "learning_rate": 1.0767294652280767e-08, "loss": 0.0109, "step": 12709 }, { "epoch": 5.9811764705882355, "grad_norm": 0.16237231006654376, "learning_rate": 1.0248471966167737e-08, "loss": 0.0119, "step": 12710 }, { "epoch": 5.98164705882353, "grad_norm": 0.45293113349127767, "learning_rate": 9.74245914052707e-09, "loss": 0.1968, "step": 12711 }, { "epoch": 5.982117647058823, "grad_norm": 0.4764801193622346, "learning_rate": 9.24925621859085e-09, "loss": 0.228, "step": 12712 }, { "epoch": 5.982588235294117, "grad_norm": 0.4114600302617416, "learning_rate": 8.768863242475389e-09, "loss": 0.1233, "step": 12713 }, { "epoch": 5.9830588235294115, "grad_norm": 0.36654758848170865, "learning_rate": 8.30128025318122e-09, "loss": 0.1027, "step": 12714 }, { "epoch": 5.983529411764706, "grad_norm": 0.8927495881221981, "learning_rate": 7.846507290676373e-09, "loss": 0.4863, "step": 12715 }, { "epoch": 5.984, "grad_norm": 0.33912000272010917, "learning_rate": 7.404544393779799e-09, "loss": 0.1241, "step": 12716 }, { "epoch": 5.984470588235294, "grad_norm": 0.5485942055106412, "learning_rate": 6.975391600261282e-09, "loss": 0.1915, "step": 12717 }, { "epoch": 5.984941176470588, "grad_norm": 0.2660543768555559, "learning_rate": 6.55904894674153e-09, "loss": 0.0789, "step": 12718 }, { "epoch": 5.985411764705883, "grad_norm": 0.3594875613301142, "learning_rate": 6.155516468792088e-09, "loss": 0.0628, "step": 12719 }, { "epoch": 5.985882352941177, "grad_norm": 0.20208032184175406, "learning_rate": 5.764794200885381e-09, "loss": 0.0511, "step": 12720 }, { "epoch": 5.98635294117647, "grad_norm": 0.32905212756267244, "learning_rate": 5.38688217637806e-09, "loss": 0.0719, "step": 12721 }, { "epoch": 5.986823529411764, "grad_norm": 0.33560850061263164, "learning_rate": 5.0217804275443084e-09, "loss": 0.1772, "step": 12722 }, { "epoch": 5.987294117647059, "grad_norm": 0.44527560500311547, "learning_rate": 4.669488985575842e-09, "loss": 0.2236, "step": 12723 }, { "epoch": 5.987764705882353, "grad_norm": 0.22452868814195726, "learning_rate": 4.3300078805652564e-09, "loss": 0.0492, "step": 12724 }, { "epoch": 5.988235294117647, "grad_norm": 0.24587447937378573, "learning_rate": 4.003337141489371e-09, "loss": 0.0603, "step": 12725 }, { "epoch": 5.988705882352941, "grad_norm": 0.29982105401964326, "learning_rate": 3.689476796242541e-09, "loss": 0.1195, "step": 12726 }, { "epoch": 5.9891764705882355, "grad_norm": 0.34401951697245636, "learning_rate": 3.388426871653305e-09, "loss": 0.1366, "step": 12727 }, { "epoch": 5.98964705882353, "grad_norm": 0.5811299561548067, "learning_rate": 3.100187393434428e-09, "loss": 0.3197, "step": 12728 }, { "epoch": 5.990117647058824, "grad_norm": 0.403837624451558, "learning_rate": 2.8247583861662484e-09, "loss": 0.1477, "step": 12729 }, { "epoch": 5.990588235294117, "grad_norm": 0.3580975228351413, "learning_rate": 2.562139873413249e-09, "loss": 0.0897, "step": 12730 }, { "epoch": 5.9910588235294115, "grad_norm": 0.35135712625733573, "learning_rate": 2.3123318775741805e-09, "loss": 0.08, "step": 12731 }, { "epoch": 5.991529411764706, "grad_norm": 0.4250888064287909, "learning_rate": 2.075334419998631e-09, "loss": 0.1563, "step": 12732 }, { "epoch": 5.992, "grad_norm": 0.43745528997897254, "learning_rate": 1.8511475209370685e-09, "loss": 0.2051, "step": 12733 }, { "epoch": 5.992470588235294, "grad_norm": 0.8338123248690035, "learning_rate": 1.6397711995241869e-09, "loss": 0.3566, "step": 12734 }, { "epoch": 5.992941176470588, "grad_norm": 0.252200326081689, "learning_rate": 1.441205473795559e-09, "loss": 0.0381, "step": 12735 }, { "epoch": 5.993411764705883, "grad_norm": 0.3302885240553069, "learning_rate": 1.2554503607375976e-09, "loss": 0.1718, "step": 12736 }, { "epoch": 5.993882352941176, "grad_norm": 0.2826752315954214, "learning_rate": 1.082505876204287e-09, "loss": 0.0764, "step": 12737 }, { "epoch": 5.99435294117647, "grad_norm": 0.25482916274553286, "learning_rate": 9.223720349671449e-10, "loss": 0.0863, "step": 12738 }, { "epoch": 5.994823529411764, "grad_norm": 0.40172755933100207, "learning_rate": 7.750488506985674e-10, "loss": 0.1722, "step": 12739 }, { "epoch": 5.995294117647059, "grad_norm": 0.7602670438779824, "learning_rate": 6.405363359718308e-10, "loss": 0.5341, "step": 12740 }, { "epoch": 5.995764705882353, "grad_norm": 0.41655758882247157, "learning_rate": 5.188345022943963e-10, "loss": 0.198, "step": 12741 }, { "epoch": 5.996235294117647, "grad_norm": 0.12809915506589803, "learning_rate": 4.099433600579516e-10, "loss": 0.0236, "step": 12742 }, { "epoch": 5.996705882352941, "grad_norm": 0.13187919343701868, "learning_rate": 3.138629185550634e-10, "loss": 0.0289, "step": 12743 }, { "epoch": 5.9971764705882356, "grad_norm": 0.2675820775051434, "learning_rate": 2.3059318597917763e-10, "loss": 0.0743, "step": 12744 }, { "epoch": 5.99764705882353, "grad_norm": 0.4004927120181915, "learning_rate": 1.6013416947457946e-10, "loss": 0.1263, "step": 12745 }, { "epoch": 5.998117647058823, "grad_norm": 0.4537899425381827, "learning_rate": 1.0248587501981986e-10, "loss": 0.2919, "step": 12746 }, { "epoch": 5.998588235294117, "grad_norm": 0.3497367621403661, "learning_rate": 5.7648307577595756e-11, "loss": 0.0801, "step": 12747 }, { "epoch": 5.9990588235294116, "grad_norm": 0.3455610206433515, "learning_rate": 2.5621470944869883e-11, "loss": 0.1001, "step": 12748 }, { "epoch": 5.999529411764706, "grad_norm": 0.585156075467105, "learning_rate": 6.4053678694442335e-12, "loss": 0.2674, "step": 12749 }, { "epoch": 6.0, "grad_norm": 0.3350017022001738, "learning_rate": 0.0, "loss": 0.0839, "step": 12750 }, { "epoch": 6.0, "eval_loss": 0.19756343960762024, "eval_runtime": 65.4027, "eval_samples_per_second": 11.467, "eval_steps_per_second": 5.734, "step": 12750 }, { "epoch": 6.0, "step": 12750, "total_flos": 35859537592320.0, "train_loss": 0.3054850370849501, "train_runtime": 107922.1025, "train_samples_per_second": 0.236, "train_steps_per_second": 0.118 } ], "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": 35859537592320.0, "train_batch_size": 2, "trial_name": null, "trial_params": null }