{ "best_metric": 0.2419528067111969, "best_model_checkpoint": "final_models/lait_bur_mpt/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": 126.80790230283516, "learning_rate": 1.5e-07, "loss": 22.75, "step": 1 }, { "epoch": 0.0009411764705882353, "grad_norm": 141.89163747819543, "learning_rate": 3e-07, "loss": 25.375, "step": 2 }, { "epoch": 0.001411764705882353, "grad_norm": 120.38634946502475, "learning_rate": 4.5e-07, "loss": 21.5, "step": 3 }, { "epoch": 0.0018823529411764706, "grad_norm": 127.84874857486952, "learning_rate": 6e-07, "loss": 23.0, "step": 4 }, { "epoch": 0.002352941176470588, "grad_norm": 140.02816693447738, "learning_rate": 7.499999999999999e-07, "loss": 24.875, "step": 5 }, { "epoch": 0.002823529411764706, "grad_norm": 164.07975189273387, "learning_rate": 9e-07, "loss": 26.875, "step": 6 }, { "epoch": 0.0032941176470588237, "grad_norm": 121.11149714320504, "learning_rate": 1.05e-06, "loss": 20.375, "step": 7 }, { "epoch": 0.0037647058823529413, "grad_norm": 181.00933411594866, "learning_rate": 1.2e-06, "loss": 26.5, "step": 8 }, { "epoch": 0.0042352941176470585, "grad_norm": 147.42436823756182, "learning_rate": 1.3499999999999998e-06, "loss": 20.75, "step": 9 }, { "epoch": 0.004705882352941176, "grad_norm": 196.0964237819939, "learning_rate": 1.4999999999999998e-06, "loss": 23.75, "step": 10 }, { "epoch": 0.0051764705882352945, "grad_norm": 148.50175051006903, "learning_rate": 1.6499999999999997e-06, "loss": 18.625, "step": 11 }, { "epoch": 0.005647058823529412, "grad_norm": 157.1663152697669, "learning_rate": 1.8e-06, "loss": 13.9375, "step": 12 }, { "epoch": 0.00611764705882353, "grad_norm": 101.43305543623924, "learning_rate": 1.9499999999999995e-06, "loss": 9.0, "step": 13 }, { "epoch": 0.006588235294117647, "grad_norm": 93.8500045217412, "learning_rate": 2.1e-06, "loss": 8.125, "step": 14 }, { "epoch": 0.007058823529411765, "grad_norm": 94.08067682549492, "learning_rate": 2.2499999999999996e-06, "loss": 7.125, "step": 15 }, { "epoch": 0.0075294117647058826, "grad_norm": 64.58797694106946, "learning_rate": 2.4e-06, "loss": 8.6875, "step": 16 }, { "epoch": 0.008, "grad_norm": 114.7266994404903, "learning_rate": 2.55e-06, "loss": 13.625, "step": 17 }, { "epoch": 0.008470588235294117, "grad_norm": 59.80630922855903, "learning_rate": 2.6999999999999996e-06, "loss": 9.375, "step": 18 }, { "epoch": 0.008941176470588235, "grad_norm": 27.071192639455848, "learning_rate": 2.85e-06, "loss": 6.4688, "step": 19 }, { "epoch": 0.009411764705882352, "grad_norm": 38.90216894391841, "learning_rate": 2.9999999999999997e-06, "loss": 7.3125, "step": 20 }, { "epoch": 0.00988235294117647, "grad_norm": 61.48770059268234, "learning_rate": 3.15e-06, "loss": 8.5, "step": 21 }, { "epoch": 0.010352941176470589, "grad_norm": 36.96923274205829, "learning_rate": 3.2999999999999993e-06, "loss": 6.6875, "step": 22 }, { "epoch": 0.010823529411764706, "grad_norm": 61.74256195284704, "learning_rate": 3.4499999999999996e-06, "loss": 8.0625, "step": 23 }, { "epoch": 0.011294117647058824, "grad_norm": 36.07402096673762, "learning_rate": 3.6e-06, "loss": 6.625, "step": 24 }, { "epoch": 0.011764705882352941, "grad_norm": 155.5883129466431, "learning_rate": 3.7499999999999997e-06, "loss": 7.5, "step": 25 }, { "epoch": 0.01223529411764706, "grad_norm": 31.792624719625362, "learning_rate": 3.899999999999999e-06, "loss": 6.125, "step": 26 }, { "epoch": 0.012705882352941176, "grad_norm": 33.22905775480454, "learning_rate": 4.049999999999999e-06, "loss": 5.875, "step": 27 }, { "epoch": 0.013176470588235295, "grad_norm": 80.84677624569694, "learning_rate": 4.2e-06, "loss": 8.8125, "step": 28 }, { "epoch": 0.013647058823529411, "grad_norm": 31.54871496327248, "learning_rate": 4.35e-06, "loss": 6.125, "step": 29 }, { "epoch": 0.01411764705882353, "grad_norm": 38.79347161393182, "learning_rate": 4.499999999999999e-06, "loss": 6.5, "step": 30 }, { "epoch": 0.014588235294117647, "grad_norm": 40.276638372858216, "learning_rate": 4.6499999999999995e-06, "loss": 5.9375, "step": 31 }, { "epoch": 0.015058823529411765, "grad_norm": 48.421224332812315, "learning_rate": 4.8e-06, "loss": 5.5, "step": 32 }, { "epoch": 0.015529411764705882, "grad_norm": 99.94288841967486, "learning_rate": 4.95e-06, "loss": 6.1562, "step": 33 }, { "epoch": 0.016, "grad_norm": 61.19448411307994, "learning_rate": 5.1e-06, "loss": 5.375, "step": 34 }, { "epoch": 0.01647058823529412, "grad_norm": 30.586351291964363, "learning_rate": 5.25e-06, "loss": 5.375, "step": 35 }, { "epoch": 0.016941176470588234, "grad_norm": 48.950121593455485, "learning_rate": 5.399999999999999e-06, "loss": 5.25, "step": 36 }, { "epoch": 0.017411764705882352, "grad_norm": 51.507089663191245, "learning_rate": 5.549999999999999e-06, "loss": 5.3438, "step": 37 }, { "epoch": 0.01788235294117647, "grad_norm": 26.846820909485864, "learning_rate": 5.7e-06, "loss": 5.0625, "step": 38 }, { "epoch": 0.01835294117647059, "grad_norm": 33.442394073640294, "learning_rate": 5.85e-06, "loss": 5.3438, "step": 39 }, { "epoch": 0.018823529411764704, "grad_norm": 30.256690367315894, "learning_rate": 5.999999999999999e-06, "loss": 4.8438, "step": 40 }, { "epoch": 0.019294117647058823, "grad_norm": 24.664419126548932, "learning_rate": 6.1499999999999996e-06, "loss": 4.5, "step": 41 }, { "epoch": 0.01976470588235294, "grad_norm": 36.307067882535186, "learning_rate": 6.3e-06, "loss": 5.0, "step": 42 }, { "epoch": 0.02023529411764706, "grad_norm": 23.477943277402286, "learning_rate": 6.449999999999999e-06, "loss": 3.8438, "step": 43 }, { "epoch": 0.020705882352941178, "grad_norm": 26.58889754005323, "learning_rate": 6.599999999999999e-06, "loss": 4.8125, "step": 44 }, { "epoch": 0.021176470588235293, "grad_norm": 25.817071998453613, "learning_rate": 6.749999999999999e-06, "loss": 4.5312, "step": 45 }, { "epoch": 0.02164705882352941, "grad_norm": 37.14744567517535, "learning_rate": 6.899999999999999e-06, "loss": 3.9531, "step": 46 }, { "epoch": 0.02211764705882353, "grad_norm": 21.922458893823382, "learning_rate": 7.049999999999999e-06, "loss": 3.8281, "step": 47 }, { "epoch": 0.02258823529411765, "grad_norm": 26.74639707180004, "learning_rate": 7.2e-06, "loss": 4.5625, "step": 48 }, { "epoch": 0.023058823529411764, "grad_norm": 33.145080512238444, "learning_rate": 7.35e-06, "loss": 4.9688, "step": 49 }, { "epoch": 0.023529411764705882, "grad_norm": 27.50822057091598, "learning_rate": 7.499999999999999e-06, "loss": 4.625, "step": 50 }, { "epoch": 0.024, "grad_norm": 24.528894969315363, "learning_rate": 7.65e-06, "loss": 4.3125, "step": 51 }, { "epoch": 0.02447058823529412, "grad_norm": 33.38626537200227, "learning_rate": 7.799999999999998e-06, "loss": 4.75, "step": 52 }, { "epoch": 0.024941176470588234, "grad_norm": 21.323006718316684, "learning_rate": 7.949999999999998e-06, "loss": 3.8281, "step": 53 }, { "epoch": 0.025411764705882352, "grad_norm": 36.1102175432494, "learning_rate": 8.099999999999999e-06, "loss": 5.2188, "step": 54 }, { "epoch": 0.02588235294117647, "grad_norm": 35.35865998029587, "learning_rate": 8.249999999999999e-06, "loss": 5.0312, "step": 55 }, { "epoch": 0.02635294117647059, "grad_norm": 30.342783164041112, "learning_rate": 8.4e-06, "loss": 4.5625, "step": 56 }, { "epoch": 0.026823529411764704, "grad_norm": 22.780986222782506, "learning_rate": 8.55e-06, "loss": 3.8281, "step": 57 }, { "epoch": 0.027294117647058823, "grad_norm": 36.44762412767495, "learning_rate": 8.7e-06, "loss": 4.7188, "step": 58 }, { "epoch": 0.02776470588235294, "grad_norm": 38.96395593510465, "learning_rate": 8.849999999999998e-06, "loss": 4.875, "step": 59 }, { "epoch": 0.02823529411764706, "grad_norm": 29.21683727301997, "learning_rate": 8.999999999999999e-06, "loss": 4.3438, "step": 60 }, { "epoch": 0.028705882352941175, "grad_norm": 28.09093451278362, "learning_rate": 9.149999999999999e-06, "loss": 4.0, "step": 61 }, { "epoch": 0.029176470588235293, "grad_norm": 29.226592643555474, "learning_rate": 9.299999999999999e-06, "loss": 3.8594, "step": 62 }, { "epoch": 0.029647058823529412, "grad_norm": 33.785248252856064, "learning_rate": 9.45e-06, "loss": 4.2188, "step": 63 }, { "epoch": 0.03011764705882353, "grad_norm": 35.5707239530583, "learning_rate": 9.6e-06, "loss": 4.4688, "step": 64 }, { "epoch": 0.03058823529411765, "grad_norm": 24.850706379053925, "learning_rate": 9.75e-06, "loss": 3.6094, "step": 65 }, { "epoch": 0.031058823529411764, "grad_norm": 36.28235989283603, "learning_rate": 9.9e-06, "loss": 4.3125, "step": 66 }, { "epoch": 0.03152941176470588, "grad_norm": 25.800232471790114, "learning_rate": 1.005e-05, "loss": 4.0, "step": 67 }, { "epoch": 0.032, "grad_norm": 37.39883069979816, "learning_rate": 1.02e-05, "loss": 4.1875, "step": 68 }, { "epoch": 0.03247058823529412, "grad_norm": 29.628651404759903, "learning_rate": 1.035e-05, "loss": 4.2188, "step": 69 }, { "epoch": 0.03294117647058824, "grad_norm": 31.522508477513295, "learning_rate": 1.05e-05, "loss": 4.2188, "step": 70 }, { "epoch": 0.033411764705882356, "grad_norm": 37.937293713546026, "learning_rate": 1.0649999999999998e-05, "loss": 4.0312, "step": 71 }, { "epoch": 0.03388235294117647, "grad_norm": 35.479951891926945, "learning_rate": 1.0799999999999998e-05, "loss": 4.0312, "step": 72 }, { "epoch": 0.034352941176470586, "grad_norm": 34.972402510477615, "learning_rate": 1.0949999999999998e-05, "loss": 3.7188, "step": 73 }, { "epoch": 0.034823529411764705, "grad_norm": 28.684363670794294, "learning_rate": 1.1099999999999999e-05, "loss": 3.5312, "step": 74 }, { "epoch": 0.03529411764705882, "grad_norm": 26.887409269309636, "learning_rate": 1.1249999999999999e-05, "loss": 3.125, "step": 75 }, { "epoch": 0.03576470588235294, "grad_norm": 27.099553865669122, "learning_rate": 1.14e-05, "loss": 3.4531, "step": 76 }, { "epoch": 0.03623529411764706, "grad_norm": 27.076371392296856, "learning_rate": 1.155e-05, "loss": 3.3125, "step": 77 }, { "epoch": 0.03670588235294118, "grad_norm": 22.9146585235841, "learning_rate": 1.17e-05, "loss": 3.4844, "step": 78 }, { "epoch": 0.0371764705882353, "grad_norm": 38.926453464196385, "learning_rate": 1.1849999999999998e-05, "loss": 3.625, "step": 79 }, { "epoch": 0.03764705882352941, "grad_norm": 33.26198406616535, "learning_rate": 1.1999999999999999e-05, "loss": 4.0312, "step": 80 }, { "epoch": 0.03811764705882353, "grad_norm": 31.045779932027436, "learning_rate": 1.2149999999999999e-05, "loss": 3.4688, "step": 81 }, { "epoch": 0.038588235294117645, "grad_norm": 22.29688835265743, "learning_rate": 1.2299999999999999e-05, "loss": 2.5156, "step": 82 }, { "epoch": 0.039058823529411764, "grad_norm": 26.77322799965209, "learning_rate": 1.245e-05, "loss": 3.375, "step": 83 }, { "epoch": 0.03952941176470588, "grad_norm": 27.28915746404361, "learning_rate": 1.26e-05, "loss": 2.6719, "step": 84 }, { "epoch": 0.04, "grad_norm": 26.39929722024932, "learning_rate": 1.275e-05, "loss": 3.2188, "step": 85 }, { "epoch": 0.04047058823529412, "grad_norm": 20.58382350475664, "learning_rate": 1.2899999999999998e-05, "loss": 2.3125, "step": 86 }, { "epoch": 0.04094117647058824, "grad_norm": 41.968880597886645, "learning_rate": 1.3049999999999999e-05, "loss": 3.25, "step": 87 }, { "epoch": 0.041411764705882356, "grad_norm": 25.972797749481746, "learning_rate": 1.3199999999999997e-05, "loss": 3.0312, "step": 88 }, { "epoch": 0.04188235294117647, "grad_norm": 39.86413933844889, "learning_rate": 1.3349999999999998e-05, "loss": 3.3281, "step": 89 }, { "epoch": 0.042352941176470586, "grad_norm": 29.594133331889935, "learning_rate": 1.3499999999999998e-05, "loss": 3.1562, "step": 90 }, { "epoch": 0.042823529411764705, "grad_norm": 40.91476735929124, "learning_rate": 1.3649999999999998e-05, "loss": 3.1406, "step": 91 }, { "epoch": 0.04329411764705882, "grad_norm": 38.74025194344423, "learning_rate": 1.3799999999999998e-05, "loss": 2.8438, "step": 92 }, { "epoch": 0.04376470588235294, "grad_norm": 26.631631341454383, "learning_rate": 1.3949999999999999e-05, "loss": 2.75, "step": 93 }, { "epoch": 0.04423529411764706, "grad_norm": 29.41126027935915, "learning_rate": 1.4099999999999999e-05, "loss": 2.7812, "step": 94 }, { "epoch": 0.04470588235294118, "grad_norm": 21.964583101964713, "learning_rate": 1.4249999999999999e-05, "loss": 2.0156, "step": 95 }, { "epoch": 0.0451764705882353, "grad_norm": 30.842232847521704, "learning_rate": 1.44e-05, "loss": 2.7188, "step": 96 }, { "epoch": 0.04564705882352941, "grad_norm": 29.96867558968851, "learning_rate": 1.455e-05, "loss": 2.4844, "step": 97 }, { "epoch": 0.04611764705882353, "grad_norm": 30.577424049428558, "learning_rate": 1.47e-05, "loss": 2.7031, "step": 98 }, { "epoch": 0.046588235294117646, "grad_norm": 38.579875896337946, "learning_rate": 1.485e-05, "loss": 2.7188, "step": 99 }, { "epoch": 0.047058823529411764, "grad_norm": 30.46230334616613, "learning_rate": 1.4999999999999999e-05, "loss": 2.4844, "step": 100 }, { "epoch": 0.04752941176470588, "grad_norm": 34.33236923143771, "learning_rate": 1.5149999999999999e-05, "loss": 2.0469, "step": 101 }, { "epoch": 0.048, "grad_norm": 36.304834657917056, "learning_rate": 1.53e-05, "loss": 2.0312, "step": 102 }, { "epoch": 0.04847058823529412, "grad_norm": 35.8742991068029, "learning_rate": 1.545e-05, "loss": 1.8906, "step": 103 }, { "epoch": 0.04894117647058824, "grad_norm": 23.908339217963434, "learning_rate": 1.5599999999999996e-05, "loss": 1.7109, "step": 104 }, { "epoch": 0.04941176470588235, "grad_norm": 34.07617474364691, "learning_rate": 1.5749999999999997e-05, "loss": 1.8281, "step": 105 }, { "epoch": 0.04988235294117647, "grad_norm": 34.77087527429948, "learning_rate": 1.5899999999999997e-05, "loss": 1.6875, "step": 106 }, { "epoch": 0.050352941176470586, "grad_norm": 24.59122707514179, "learning_rate": 1.6049999999999997e-05, "loss": 2.1406, "step": 107 }, { "epoch": 0.050823529411764705, "grad_norm": 23.400944947094004, "learning_rate": 1.6199999999999997e-05, "loss": 1.7109, "step": 108 }, { "epoch": 0.05129411764705882, "grad_norm": 25.85120402344135, "learning_rate": 1.6349999999999998e-05, "loss": 1.5391, "step": 109 }, { "epoch": 0.05176470588235294, "grad_norm": 31.88003089033638, "learning_rate": 1.6499999999999998e-05, "loss": 1.6094, "step": 110 }, { "epoch": 0.05223529411764706, "grad_norm": 29.95000744806244, "learning_rate": 1.6649999999999998e-05, "loss": 1.3438, "step": 111 }, { "epoch": 0.05270588235294118, "grad_norm": 30.936988272458894, "learning_rate": 1.68e-05, "loss": 1.4453, "step": 112 }, { "epoch": 0.0531764705882353, "grad_norm": 24.62053828047722, "learning_rate": 1.695e-05, "loss": 1.3906, "step": 113 }, { "epoch": 0.05364705882352941, "grad_norm": 27.875136972034788, "learning_rate": 1.71e-05, "loss": 1.4844, "step": 114 }, { "epoch": 0.05411764705882353, "grad_norm": 27.810411313613436, "learning_rate": 1.725e-05, "loss": 1.2812, "step": 115 }, { "epoch": 0.054588235294117646, "grad_norm": 23.145075367687852, "learning_rate": 1.74e-05, "loss": 1.375, "step": 116 }, { "epoch": 0.055058823529411764, "grad_norm": 18.58828380610928, "learning_rate": 1.755e-05, "loss": 1.2891, "step": 117 }, { "epoch": 0.05552941176470588, "grad_norm": 20.68244164352036, "learning_rate": 1.7699999999999997e-05, "loss": 1.1016, "step": 118 }, { "epoch": 0.056, "grad_norm": 27.67075830075517, "learning_rate": 1.7849999999999997e-05, "loss": 1.1016, "step": 119 }, { "epoch": 0.05647058823529412, "grad_norm": 22.74986884447893, "learning_rate": 1.7999999999999997e-05, "loss": 1.7109, "step": 120 }, { "epoch": 0.05694117647058824, "grad_norm": 24.045327378210736, "learning_rate": 1.8149999999999997e-05, "loss": 0.9727, "step": 121 }, { "epoch": 0.05741176470588235, "grad_norm": 21.38133283386783, "learning_rate": 1.8299999999999998e-05, "loss": 1.0312, "step": 122 }, { "epoch": 0.05788235294117647, "grad_norm": 24.071272892012896, "learning_rate": 1.8449999999999998e-05, "loss": 1.0625, "step": 123 }, { "epoch": 0.05835294117647059, "grad_norm": 21.14689156206203, "learning_rate": 1.8599999999999998e-05, "loss": 0.9141, "step": 124 }, { "epoch": 0.058823529411764705, "grad_norm": 20.385135229600486, "learning_rate": 1.875e-05, "loss": 0.8086, "step": 125 }, { "epoch": 0.059294117647058824, "grad_norm": 14.980303793228291, "learning_rate": 1.89e-05, "loss": 0.875, "step": 126 }, { "epoch": 0.05976470588235294, "grad_norm": 24.556770356374706, "learning_rate": 1.905e-05, "loss": 1.2422, "step": 127 }, { "epoch": 0.06023529411764706, "grad_norm": 19.818624508024612, "learning_rate": 1.92e-05, "loss": 1.1875, "step": 128 }, { "epoch": 0.06070588235294118, "grad_norm": 19.15324056937183, "learning_rate": 1.935e-05, "loss": 0.9336, "step": 129 }, { "epoch": 0.0611764705882353, "grad_norm": 16.11363111477877, "learning_rate": 1.95e-05, "loss": 0.9023, "step": 130 }, { "epoch": 0.06164705882352941, "grad_norm": 11.829453085355334, "learning_rate": 1.965e-05, "loss": 0.6992, "step": 131 }, { "epoch": 0.06211764705882353, "grad_norm": 13.831116620052702, "learning_rate": 1.98e-05, "loss": 0.7109, "step": 132 }, { "epoch": 0.06258823529411765, "grad_norm": 8.665437582401621, "learning_rate": 1.995e-05, "loss": 0.8594, "step": 133 }, { "epoch": 0.06305882352941176, "grad_norm": 12.331892212811855, "learning_rate": 2.01e-05, "loss": 0.543, "step": 134 }, { "epoch": 0.06352941176470588, "grad_norm": 9.70589192709707, "learning_rate": 2.025e-05, "loss": 0.75, "step": 135 }, { "epoch": 0.064, "grad_norm": 9.684673338933775, "learning_rate": 2.04e-05, "loss": 0.7539, "step": 136 }, { "epoch": 0.06447058823529411, "grad_norm": 11.396350041194111, "learning_rate": 2.055e-05, "loss": 0.4141, "step": 137 }, { "epoch": 0.06494117647058824, "grad_norm": 8.8440309900305, "learning_rate": 2.07e-05, "loss": 0.6602, "step": 138 }, { "epoch": 0.06541176470588235, "grad_norm": 8.717020825533307, "learning_rate": 2.085e-05, "loss": 0.4766, "step": 139 }, { "epoch": 0.06588235294117648, "grad_norm": 8.614885057278169, "learning_rate": 2.1e-05, "loss": 0.332, "step": 140 }, { "epoch": 0.06635294117647059, "grad_norm": 7.783198278707622, "learning_rate": 2.1149999999999996e-05, "loss": 0.6328, "step": 141 }, { "epoch": 0.06682352941176471, "grad_norm": 9.46337197777407, "learning_rate": 2.1299999999999996e-05, "loss": 0.2334, "step": 142 }, { "epoch": 0.06729411764705882, "grad_norm": 7.1471850476906855, "learning_rate": 2.1449999999999996e-05, "loss": 0.4531, "step": 143 }, { "epoch": 0.06776470588235294, "grad_norm": 6.395882887736069, "learning_rate": 2.1599999999999996e-05, "loss": 0.4414, "step": 144 }, { "epoch": 0.06823529411764706, "grad_norm": 7.784029986550863, "learning_rate": 2.1749999999999997e-05, "loss": 0.2207, "step": 145 }, { "epoch": 0.06870588235294117, "grad_norm": 7.193544573701644, "learning_rate": 2.1899999999999997e-05, "loss": 0.2041, "step": 146 }, { "epoch": 0.0691764705882353, "grad_norm": 73.67192476853974, "learning_rate": 2.2049999999999997e-05, "loss": 0.8828, "step": 147 }, { "epoch": 0.06964705882352941, "grad_norm": 6.771312334327687, "learning_rate": 2.2199999999999998e-05, "loss": 0.4082, "step": 148 }, { "epoch": 0.07011764705882353, "grad_norm": 128.85702363978734, "learning_rate": 2.2349999999999998e-05, "loss": 2.875, "step": 149 }, { "epoch": 0.07058823529411765, "grad_norm": 5.805726238737508, "learning_rate": 2.2499999999999998e-05, "loss": 0.3359, "step": 150 }, { "epoch": 0.07105882352941177, "grad_norm": 25.000954681086423, "learning_rate": 2.2649999999999998e-05, "loss": 0.918, "step": 151 }, { "epoch": 0.07152941176470588, "grad_norm": 34.69605033685843, "learning_rate": 2.28e-05, "loss": 1.2656, "step": 152 }, { "epoch": 0.072, "grad_norm": 12.891973849935882, "learning_rate": 2.295e-05, "loss": 1.1406, "step": 153 }, { "epoch": 0.07247058823529412, "grad_norm": 9.276864456457687, "learning_rate": 2.31e-05, "loss": 0.582, "step": 154 }, { "epoch": 0.07294117647058823, "grad_norm": 10.247145923506967, "learning_rate": 2.325e-05, "loss": 0.8594, "step": 155 }, { "epoch": 0.07341176470588236, "grad_norm": 53.696271270621985, "learning_rate": 2.34e-05, "loss": 1.5391, "step": 156 }, { "epoch": 0.07388235294117647, "grad_norm": 15.538819617700973, "learning_rate": 2.3549999999999996e-05, "loss": 0.8242, "step": 157 }, { "epoch": 0.0743529411764706, "grad_norm": 12.635659754845358, "learning_rate": 2.3699999999999997e-05, "loss": 0.8125, "step": 158 }, { "epoch": 0.0748235294117647, "grad_norm": 8.041435527852633, "learning_rate": 2.3849999999999997e-05, "loss": 0.9414, "step": 159 }, { "epoch": 0.07529411764705882, "grad_norm": 9.277157987856228, "learning_rate": 2.3999999999999997e-05, "loss": 0.4785, "step": 160 }, { "epoch": 0.07576470588235294, "grad_norm": 7.818579085835733, "learning_rate": 2.4149999999999997e-05, "loss": 0.75, "step": 161 }, { "epoch": 0.07623529411764705, "grad_norm": 10.51881708421973, "learning_rate": 2.4299999999999998e-05, "loss": 0.5742, "step": 162 }, { "epoch": 0.07670588235294118, "grad_norm": 8.545929540208599, "learning_rate": 2.4449999999999998e-05, "loss": 0.6641, "step": 163 }, { "epoch": 0.07717647058823529, "grad_norm": 6.812044414092388, "learning_rate": 2.4599999999999998e-05, "loss": 0.5938, "step": 164 }, { "epoch": 0.07764705882352942, "grad_norm": 6.28817129168233, "learning_rate": 2.475e-05, "loss": 0.6406, "step": 165 }, { "epoch": 0.07811764705882353, "grad_norm": 6.5719427052375625, "learning_rate": 2.49e-05, "loss": 0.3965, "step": 166 }, { "epoch": 0.07858823529411765, "grad_norm": 4.340535241889796, "learning_rate": 2.505e-05, "loss": 0.5625, "step": 167 }, { "epoch": 0.07905882352941176, "grad_norm": 3.8711719706309466, "learning_rate": 2.52e-05, "loss": 0.6797, "step": 168 }, { "epoch": 0.07952941176470588, "grad_norm": 3.7610167575751263, "learning_rate": 2.535e-05, "loss": 0.1445, "step": 169 }, { "epoch": 0.08, "grad_norm": 4.591109494353126, "learning_rate": 2.55e-05, "loss": 0.5664, "step": 170 }, { "epoch": 0.08047058823529411, "grad_norm": 3.179225130823475, "learning_rate": 2.565e-05, "loss": 0.1289, "step": 171 }, { "epoch": 0.08094117647058824, "grad_norm": 2.9443611395213676, "learning_rate": 2.5799999999999997e-05, "loss": 0.4375, "step": 172 }, { "epoch": 0.08141176470588235, "grad_norm": 2.487882859232921, "learning_rate": 2.5949999999999997e-05, "loss": 0.2031, "step": 173 }, { "epoch": 0.08188235294117648, "grad_norm": 2.2300922744693668, "learning_rate": 2.6099999999999997e-05, "loss": 0.3438, "step": 174 }, { "epoch": 0.08235294117647059, "grad_norm": 1.9737670482952636, "learning_rate": 2.6249999999999998e-05, "loss": 0.2598, "step": 175 }, { "epoch": 0.08282352941176471, "grad_norm": 1.9429082288625046, "learning_rate": 2.6399999999999995e-05, "loss": 0.2656, "step": 176 }, { "epoch": 0.08329411764705882, "grad_norm": 2.236804743808334, "learning_rate": 2.6549999999999995e-05, "loss": 0.4316, "step": 177 }, { "epoch": 0.08376470588235294, "grad_norm": 1.8088382768474316, "learning_rate": 2.6699999999999995e-05, "loss": 0.3613, "step": 178 }, { "epoch": 0.08423529411764706, "grad_norm": 1.4794526566974708, "learning_rate": 2.6849999999999995e-05, "loss": 0.0664, "step": 179 }, { "epoch": 0.08470588235294117, "grad_norm": 2.812003127068479, "learning_rate": 2.6999999999999996e-05, "loss": 0.6562, "step": 180 }, { "epoch": 0.0851764705882353, "grad_norm": 1.568641545639836, "learning_rate": 2.7149999999999996e-05, "loss": 0.3438, "step": 181 }, { "epoch": 0.08564705882352941, "grad_norm": 1.6016991379145489, "learning_rate": 2.7299999999999996e-05, "loss": 0.2812, "step": 182 }, { "epoch": 0.08611764705882353, "grad_norm": 1.7488238440848805, "learning_rate": 2.7449999999999996e-05, "loss": 0.3633, "step": 183 }, { "epoch": 0.08658823529411765, "grad_norm": 1.7652322862556513, "learning_rate": 2.7599999999999997e-05, "loss": 0.4121, "step": 184 }, { "epoch": 0.08705882352941176, "grad_norm": 1.3088886916747102, "learning_rate": 2.7749999999999997e-05, "loss": 0.1768, "step": 185 }, { "epoch": 0.08752941176470588, "grad_norm": 1.320421443688735, "learning_rate": 2.7899999999999997e-05, "loss": 0.165, "step": 186 }, { "epoch": 0.088, "grad_norm": 3.4284840018725413, "learning_rate": 2.8049999999999997e-05, "loss": 0.3848, "step": 187 }, { "epoch": 0.08847058823529412, "grad_norm": 1.4113661171710425, "learning_rate": 2.8199999999999998e-05, "loss": 0.4004, "step": 188 }, { "epoch": 0.08894117647058823, "grad_norm": 3.9786940247439606, "learning_rate": 2.8349999999999998e-05, "loss": 0.7344, "step": 189 }, { "epoch": 0.08941176470588236, "grad_norm": 2.0415627408046273, "learning_rate": 2.8499999999999998e-05, "loss": 0.4375, "step": 190 }, { "epoch": 0.08988235294117647, "grad_norm": 1.9050954467372154, "learning_rate": 2.865e-05, "loss": 0.2676, "step": 191 }, { "epoch": 0.0903529411764706, "grad_norm": 1.4750503794390664, "learning_rate": 2.88e-05, "loss": 0.2871, "step": 192 }, { "epoch": 0.0908235294117647, "grad_norm": 1.264504043977634, "learning_rate": 2.895e-05, "loss": 0.2637, "step": 193 }, { "epoch": 0.09129411764705882, "grad_norm": 1.1990584983378303, "learning_rate": 2.91e-05, "loss": 0.2715, "step": 194 }, { "epoch": 0.09176470588235294, "grad_norm": 1.3414544070220438, "learning_rate": 2.925e-05, "loss": 0.4023, "step": 195 }, { "epoch": 0.09223529411764705, "grad_norm": 1.3135057154772607, "learning_rate": 2.94e-05, "loss": 0.2832, "step": 196 }, { "epoch": 0.09270588235294118, "grad_norm": 1.383484833155175, "learning_rate": 2.955e-05, "loss": 0.3008, "step": 197 }, { "epoch": 0.09317647058823529, "grad_norm": 0.8790851276504076, "learning_rate": 2.97e-05, "loss": 0.1465, "step": 198 }, { "epoch": 0.09364705882352942, "grad_norm": 1.80395261005332, "learning_rate": 2.985e-05, "loss": 0.4883, "step": 199 }, { "epoch": 0.09411764705882353, "grad_norm": 0.7472704787735531, "learning_rate": 2.9999999999999997e-05, "loss": 0.0605, "step": 200 }, { "epoch": 0.09458823529411765, "grad_norm": 0.7472481057406283, "learning_rate": 3.0149999999999998e-05, "loss": 0.0635, "step": 201 }, { "epoch": 0.09505882352941177, "grad_norm": 2.6557036171232524, "learning_rate": 3.0299999999999998e-05, "loss": 0.5703, "step": 202 }, { "epoch": 0.09552941176470588, "grad_norm": 1.0617336273339346, "learning_rate": 3.0449999999999998e-05, "loss": 0.1895, "step": 203 }, { "epoch": 0.096, "grad_norm": 0.6866803608513821, "learning_rate": 3.06e-05, "loss": 0.0586, "step": 204 }, { "epoch": 0.09647058823529411, "grad_norm": 1.062520397468133, "learning_rate": 3.0749999999999995e-05, "loss": 0.2754, "step": 205 }, { "epoch": 0.09694117647058824, "grad_norm": 1.1142598358754312, "learning_rate": 3.09e-05, "loss": 0.2139, "step": 206 }, { "epoch": 0.09741176470588235, "grad_norm": 1.8054286919682772, "learning_rate": 3.1049999999999996e-05, "loss": 0.459, "step": 207 }, { "epoch": 0.09788235294117648, "grad_norm": 0.8254896763542591, "learning_rate": 3.119999999999999e-05, "loss": 0.1021, "step": 208 }, { "epoch": 0.09835294117647059, "grad_norm": 1.5686205949895398, "learning_rate": 3.1349999999999996e-05, "loss": 0.293, "step": 209 }, { "epoch": 0.0988235294117647, "grad_norm": 0.6706589252725473, "learning_rate": 3.149999999999999e-05, "loss": 0.0737, "step": 210 }, { "epoch": 0.09929411764705882, "grad_norm": 0.9312633480593814, "learning_rate": 3.165e-05, "loss": 0.1797, "step": 211 }, { "epoch": 0.09976470588235294, "grad_norm": 0.7750864716295603, "learning_rate": 3.1799999999999994e-05, "loss": 0.2021, "step": 212 }, { "epoch": 0.10023529411764706, "grad_norm": 1.0983343123854903, "learning_rate": 3.195e-05, "loss": 0.3203, "step": 213 }, { "epoch": 0.10070588235294117, "grad_norm": 0.8054111425424797, "learning_rate": 3.2099999999999994e-05, "loss": 0.167, "step": 214 }, { "epoch": 0.1011764705882353, "grad_norm": 0.6704444508180226, "learning_rate": 3.225e-05, "loss": 0.0913, "step": 215 }, { "epoch": 0.10164705882352941, "grad_norm": 0.7895650632691829, "learning_rate": 3.2399999999999995e-05, "loss": 0.2061, "step": 216 }, { "epoch": 0.10211764705882354, "grad_norm": 1.9609702696143563, "learning_rate": 3.255e-05, "loss": 0.582, "step": 217 }, { "epoch": 0.10258823529411765, "grad_norm": 0.6163651450389701, "learning_rate": 3.2699999999999995e-05, "loss": 0.0698, "step": 218 }, { "epoch": 0.10305882352941176, "grad_norm": 0.4464838569943109, "learning_rate": 3.285e-05, "loss": 0.0211, "step": 219 }, { "epoch": 0.10352941176470588, "grad_norm": 0.917860778695507, "learning_rate": 3.2999999999999996e-05, "loss": 0.1904, "step": 220 }, { "epoch": 0.104, "grad_norm": 0.734314329193583, "learning_rate": 3.315e-05, "loss": 0.1187, "step": 221 }, { "epoch": 0.10447058823529412, "grad_norm": 1.5621951861774843, "learning_rate": 3.3299999999999996e-05, "loss": 0.3613, "step": 222 }, { "epoch": 0.10494117647058823, "grad_norm": 1.608364930808748, "learning_rate": 3.345e-05, "loss": 0.5117, "step": 223 }, { "epoch": 0.10541176470588236, "grad_norm": 1.7189615838161418, "learning_rate": 3.36e-05, "loss": 0.4434, "step": 224 }, { "epoch": 0.10588235294117647, "grad_norm": 1.418278375966527, "learning_rate": 3.375e-05, "loss": 0.3027, "step": 225 }, { "epoch": 0.1063529411764706, "grad_norm": 0.6834634839138635, "learning_rate": 3.39e-05, "loss": 0.104, "step": 226 }, { "epoch": 0.1068235294117647, "grad_norm": 1.4264577438446433, "learning_rate": 3.405e-05, "loss": 0.3711, "step": 227 }, { "epoch": 0.10729411764705882, "grad_norm": 1.2201197804600774, "learning_rate": 3.42e-05, "loss": 0.3691, "step": 228 }, { "epoch": 0.10776470588235294, "grad_norm": 0.9529252189878412, "learning_rate": 3.435e-05, "loss": 0.2949, "step": 229 }, { "epoch": 0.10823529411764705, "grad_norm": 0.9332926522548369, "learning_rate": 3.45e-05, "loss": 0.2051, "step": 230 }, { "epoch": 0.10870588235294118, "grad_norm": 2.9248918281842275, "learning_rate": 3.465e-05, "loss": 0.6758, "step": 231 }, { "epoch": 0.10917647058823529, "grad_norm": 0.8683055220670575, "learning_rate": 3.48e-05, "loss": 0.208, "step": 232 }, { "epoch": 0.10964705882352942, "grad_norm": 3.6432680187975666, "learning_rate": 3.4949999999999996e-05, "loss": 0.5703, "step": 233 }, { "epoch": 0.11011764705882353, "grad_norm": 2.1753805483480457, "learning_rate": 3.51e-05, "loss": 0.5898, "step": 234 }, { "epoch": 0.11058823529411765, "grad_norm": 2.78143344362244, "learning_rate": 3.5249999999999996e-05, "loss": 0.6758, "step": 235 }, { "epoch": 0.11105882352941177, "grad_norm": 2.316883658617287, "learning_rate": 3.539999999999999e-05, "loss": 0.5117, "step": 236 }, { "epoch": 0.11152941176470588, "grad_norm": 0.899239772082746, "learning_rate": 3.555e-05, "loss": 0.1289, "step": 237 }, { "epoch": 0.112, "grad_norm": 1.4055094376517179, "learning_rate": 3.5699999999999994e-05, "loss": 0.2598, "step": 238 }, { "epoch": 0.11247058823529411, "grad_norm": 1.0789189296915482, "learning_rate": 3.585e-05, "loss": 0.103, "step": 239 }, { "epoch": 0.11294117647058824, "grad_norm": 1.1869218184930037, "learning_rate": 3.5999999999999994e-05, "loss": 0.2773, "step": 240 }, { "epoch": 0.11341176470588235, "grad_norm": 2.4886239784927966, "learning_rate": 3.615e-05, "loss": 0.3262, "step": 241 }, { "epoch": 0.11388235294117648, "grad_norm": 1.5487557112552695, "learning_rate": 3.6299999999999995e-05, "loss": 0.3262, "step": 242 }, { "epoch": 0.11435294117647059, "grad_norm": 1.0988170891751576, "learning_rate": 3.645e-05, "loss": 0.2207, "step": 243 }, { "epoch": 0.1148235294117647, "grad_norm": 2.485563921096238, "learning_rate": 3.6599999999999995e-05, "loss": 0.3926, "step": 244 }, { "epoch": 0.11529411764705882, "grad_norm": 2.639999438554555, "learning_rate": 3.675e-05, "loss": 0.5234, "step": 245 }, { "epoch": 0.11576470588235294, "grad_norm": 1.162980122708776, "learning_rate": 3.6899999999999996e-05, "loss": 0.2002, "step": 246 }, { "epoch": 0.11623529411764706, "grad_norm": 4.357256068964088, "learning_rate": 3.705e-05, "loss": 0.4531, "step": 247 }, { "epoch": 0.11670588235294117, "grad_norm": 3.411384430022824, "learning_rate": 3.7199999999999996e-05, "loss": 0.7461, "step": 248 }, { "epoch": 0.1171764705882353, "grad_norm": 1.3641650660710323, "learning_rate": 3.735e-05, "loss": 0.293, "step": 249 }, { "epoch": 0.11764705882352941, "grad_norm": 2.5934868293417357, "learning_rate": 3.75e-05, "loss": 0.5938, "step": 250 }, { "epoch": 0.11811764705882354, "grad_norm": 2.1841217514124454, "learning_rate": 3.7649999999999994e-05, "loss": 0.6602, "step": 251 }, { "epoch": 0.11858823529411765, "grad_norm": 1.039535406491042, "learning_rate": 3.78e-05, "loss": 0.3027, "step": 252 }, { "epoch": 0.11905882352941176, "grad_norm": 0.6856808168081825, "learning_rate": 3.7949999999999994e-05, "loss": 0.1108, "step": 253 }, { "epoch": 0.11952941176470588, "grad_norm": 0.806854992716898, "learning_rate": 3.81e-05, "loss": 0.1069, "step": 254 }, { "epoch": 0.12, "grad_norm": 5.022424949418036, "learning_rate": 3.8249999999999995e-05, "loss": 0.6914, "step": 255 }, { "epoch": 0.12047058823529412, "grad_norm": 1.0438173046690482, "learning_rate": 3.84e-05, "loss": 0.0197, "step": 256 }, { "epoch": 0.12094117647058823, "grad_norm": 0.820749760487373, "learning_rate": 3.8549999999999995e-05, "loss": 0.1777, "step": 257 }, { "epoch": 0.12141176470588236, "grad_norm": 2.531655408265771, "learning_rate": 3.87e-05, "loss": 0.668, "step": 258 }, { "epoch": 0.12188235294117647, "grad_norm": 1.029720240671538, "learning_rate": 3.8849999999999996e-05, "loss": 0.2715, "step": 259 }, { "epoch": 0.1223529411764706, "grad_norm": 0.23609914011099625, "learning_rate": 3.9e-05, "loss": 0.0102, "step": 260 }, { "epoch": 0.1228235294117647, "grad_norm": 2.4012511353305057, "learning_rate": 3.9149999999999996e-05, "loss": 0.5117, "step": 261 }, { "epoch": 0.12329411764705882, "grad_norm": 2.0859617825258097, "learning_rate": 3.93e-05, "loss": 0.6367, "step": 262 }, { "epoch": 0.12376470588235294, "grad_norm": 0.614002053969302, "learning_rate": 3.945e-05, "loss": 0.1523, "step": 263 }, { "epoch": 0.12423529411764705, "grad_norm": 1.9041283279185184, "learning_rate": 3.96e-05, "loss": 0.334, "step": 264 }, { "epoch": 0.12470588235294118, "grad_norm": 0.7927695782511892, "learning_rate": 3.975e-05, "loss": 0.1611, "step": 265 }, { "epoch": 0.1251764705882353, "grad_norm": 0.9868614680594472, "learning_rate": 3.99e-05, "loss": 0.2637, "step": 266 }, { "epoch": 0.12564705882352942, "grad_norm": 1.2257341041732663, "learning_rate": 4.005e-05, "loss": 0.3066, "step": 267 }, { "epoch": 0.12611764705882353, "grad_norm": 0.7473535103447487, "learning_rate": 4.02e-05, "loss": 0.208, "step": 268 }, { "epoch": 0.12658823529411764, "grad_norm": 1.2180529266999345, "learning_rate": 4.035e-05, "loss": 0.3496, "step": 269 }, { "epoch": 0.12705882352941175, "grad_norm": 1.0180285085845597, "learning_rate": 4.05e-05, "loss": 0.2617, "step": 270 }, { "epoch": 0.1275294117647059, "grad_norm": 1.7760523851931367, "learning_rate": 4.065e-05, "loss": 0.5469, "step": 271 }, { "epoch": 0.128, "grad_norm": 0.703364228960797, "learning_rate": 4.08e-05, "loss": 0.248, "step": 272 }, { "epoch": 0.12847058823529411, "grad_norm": 1.4380413592590233, "learning_rate": 4.095e-05, "loss": 0.5586, "step": 273 }, { "epoch": 0.12894117647058823, "grad_norm": 0.7225081734895676, "learning_rate": 4.11e-05, "loss": 0.2393, "step": 274 }, { "epoch": 0.12941176470588237, "grad_norm": 0.3264294954422337, "learning_rate": 4.125e-05, "loss": 0.0515, "step": 275 }, { "epoch": 0.12988235294117648, "grad_norm": 0.7808191931455151, "learning_rate": 4.14e-05, "loss": 0.2236, "step": 276 }, { "epoch": 0.1303529411764706, "grad_norm": 0.6798882744297382, "learning_rate": 4.155e-05, "loss": 0.1484, "step": 277 }, { "epoch": 0.1308235294117647, "grad_norm": 1.2134699699286127, "learning_rate": 4.17e-05, "loss": 0.373, "step": 278 }, { "epoch": 0.1312941176470588, "grad_norm": 0.2815614452975243, "learning_rate": 4.185e-05, "loss": 0.0437, "step": 279 }, { "epoch": 0.13176470588235295, "grad_norm": 0.6762187068587895, "learning_rate": 4.2e-05, "loss": 0.1836, "step": 280 }, { "epoch": 0.13223529411764706, "grad_norm": 0.6563150173398641, "learning_rate": 4.215e-05, "loss": 0.1592, "step": 281 }, { "epoch": 0.13270588235294117, "grad_norm": 0.782736387915129, "learning_rate": 4.229999999999999e-05, "loss": 0.2852, "step": 282 }, { "epoch": 0.13317647058823529, "grad_norm": 1.0181988381284752, "learning_rate": 4.2449999999999995e-05, "loss": 0.4941, "step": 283 }, { "epoch": 0.13364705882352942, "grad_norm": 0.8841286864860903, "learning_rate": 4.259999999999999e-05, "loss": 0.3398, "step": 284 }, { "epoch": 0.13411764705882354, "grad_norm": 0.7049999201846086, "learning_rate": 4.2749999999999996e-05, "loss": 0.2578, "step": 285 }, { "epoch": 0.13458823529411765, "grad_norm": 1.6051814892961798, "learning_rate": 4.289999999999999e-05, "loss": 0.5391, "step": 286 }, { "epoch": 0.13505882352941176, "grad_norm": 0.733342324800991, "learning_rate": 4.3049999999999996e-05, "loss": 0.3008, "step": 287 }, { "epoch": 0.13552941176470587, "grad_norm": 1.2210182050422451, "learning_rate": 4.319999999999999e-05, "loss": 0.5312, "step": 288 }, { "epoch": 0.136, "grad_norm": 0.8132979590459677, "learning_rate": 4.334999999999999e-05, "loss": 0.2441, "step": 289 }, { "epoch": 0.13647058823529412, "grad_norm": 1.1137549636910253, "learning_rate": 4.3499999999999993e-05, "loss": 0.3613, "step": 290 }, { "epoch": 0.13694117647058823, "grad_norm": 0.9739550923701275, "learning_rate": 4.364999999999999e-05, "loss": 0.3125, "step": 291 }, { "epoch": 0.13741176470588234, "grad_norm": 0.6481200350425838, "learning_rate": 4.3799999999999994e-05, "loss": 0.2168, "step": 292 }, { "epoch": 0.13788235294117648, "grad_norm": 0.5391540822394856, "learning_rate": 4.394999999999999e-05, "loss": 0.1641, "step": 293 }, { "epoch": 0.1383529411764706, "grad_norm": 1.1524345800357936, "learning_rate": 4.4099999999999995e-05, "loss": 0.4277, "step": 294 }, { "epoch": 0.1388235294117647, "grad_norm": 0.4145107180260662, "learning_rate": 4.424999999999999e-05, "loss": 0.1172, "step": 295 }, { "epoch": 0.13929411764705882, "grad_norm": 0.35669358313324306, "learning_rate": 4.4399999999999995e-05, "loss": 0.0674, "step": 296 }, { "epoch": 0.13976470588235293, "grad_norm": 0.563179808655922, "learning_rate": 4.454999999999999e-05, "loss": 0.0928, "step": 297 }, { "epoch": 0.14023529411764707, "grad_norm": 0.4305892445008516, "learning_rate": 4.4699999999999996e-05, "loss": 0.0771, "step": 298 }, { "epoch": 0.14070588235294118, "grad_norm": 1.326966246247397, "learning_rate": 4.484999999999999e-05, "loss": 0.3184, "step": 299 }, { "epoch": 0.1411764705882353, "grad_norm": 0.6687933748985586, "learning_rate": 4.4999999999999996e-05, "loss": 0.3047, "step": 300 }, { "epoch": 0.1416470588235294, "grad_norm": 0.9812272074085899, "learning_rate": 4.514999999999999e-05, "loss": 0.3008, "step": 301 }, { "epoch": 0.14211764705882354, "grad_norm": 2.031557504695254, "learning_rate": 4.5299999999999997e-05, "loss": 0.668, "step": 302 }, { "epoch": 0.14258823529411765, "grad_norm": 0.40513478676645, "learning_rate": 4.5449999999999993e-05, "loss": 0.0513, "step": 303 }, { "epoch": 0.14305882352941177, "grad_norm": 3.6878363271256234, "learning_rate": 4.56e-05, "loss": 0.6719, "step": 304 }, { "epoch": 0.14352941176470588, "grad_norm": 0.8793237658201585, "learning_rate": 4.5749999999999994e-05, "loss": 0.2275, "step": 305 }, { "epoch": 0.144, "grad_norm": 1.6139537340377474, "learning_rate": 4.59e-05, "loss": 0.375, "step": 306 }, { "epoch": 0.14447058823529413, "grad_norm": 1.3097631619064187, "learning_rate": 4.6049999999999994e-05, "loss": 0.418, "step": 307 }, { "epoch": 0.14494117647058824, "grad_norm": 0.9388893123258139, "learning_rate": 4.62e-05, "loss": 0.4062, "step": 308 }, { "epoch": 0.14541176470588235, "grad_norm": 0.5548712604488774, "learning_rate": 4.6349999999999995e-05, "loss": 0.168, "step": 309 }, { "epoch": 0.14588235294117646, "grad_norm": 0.2677545277666903, "learning_rate": 4.65e-05, "loss": 0.0361, "step": 310 }, { "epoch": 0.14635294117647057, "grad_norm": 0.5168059117968169, "learning_rate": 4.6649999999999996e-05, "loss": 0.0698, "step": 311 }, { "epoch": 0.1468235294117647, "grad_norm": 2.128580032838602, "learning_rate": 4.68e-05, "loss": 0.3691, "step": 312 }, { "epoch": 0.14729411764705883, "grad_norm": 0.9100332876723771, "learning_rate": 4.6949999999999996e-05, "loss": 0.3027, "step": 313 }, { "epoch": 0.14776470588235294, "grad_norm": 0.7546426353780566, "learning_rate": 4.709999999999999e-05, "loss": 0.2354, "step": 314 }, { "epoch": 0.14823529411764705, "grad_norm": 0.2811459522863485, "learning_rate": 4.7249999999999997e-05, "loss": 0.0544, "step": 315 }, { "epoch": 0.1487058823529412, "grad_norm": 0.8505656266438859, "learning_rate": 4.7399999999999993e-05, "loss": 0.1445, "step": 316 }, { "epoch": 0.1491764705882353, "grad_norm": 1.227155661944277, "learning_rate": 4.755e-05, "loss": 0.2471, "step": 317 }, { "epoch": 0.1496470588235294, "grad_norm": 0.4308929283540831, "learning_rate": 4.7699999999999994e-05, "loss": 0.0713, "step": 318 }, { "epoch": 0.15011764705882352, "grad_norm": 1.7530274071372416, "learning_rate": 4.785e-05, "loss": 0.5547, "step": 319 }, { "epoch": 0.15058823529411763, "grad_norm": 1.1120647523994263, "learning_rate": 4.7999999999999994e-05, "loss": 0.4629, "step": 320 }, { "epoch": 0.15105882352941177, "grad_norm": 0.8080399255387463, "learning_rate": 4.815e-05, "loss": 0.2637, "step": 321 }, { "epoch": 0.15152941176470588, "grad_norm": 1.5087426556303478, "learning_rate": 4.8299999999999995e-05, "loss": 0.3359, "step": 322 }, { "epoch": 0.152, "grad_norm": 0.7446279641584979, "learning_rate": 4.845e-05, "loss": 0.1631, "step": 323 }, { "epoch": 0.1524705882352941, "grad_norm": 0.7812858645632035, "learning_rate": 4.8599999999999995e-05, "loss": 0.3008, "step": 324 }, { "epoch": 0.15294117647058825, "grad_norm": 0.3650776855137046, "learning_rate": 4.875e-05, "loss": 0.0938, "step": 325 }, { "epoch": 0.15341176470588236, "grad_norm": 0.7210291187932456, "learning_rate": 4.8899999999999996e-05, "loss": 0.1074, "step": 326 }, { "epoch": 0.15388235294117647, "grad_norm": 0.7995833123031412, "learning_rate": 4.905e-05, "loss": 0.1992, "step": 327 }, { "epoch": 0.15435294117647058, "grad_norm": 0.8426529460702371, "learning_rate": 4.9199999999999997e-05, "loss": 0.3359, "step": 328 }, { "epoch": 0.1548235294117647, "grad_norm": 0.7060700164767392, "learning_rate": 4.935e-05, "loss": 0.2676, "step": 329 }, { "epoch": 0.15529411764705883, "grad_norm": 1.00982958998498, "learning_rate": 4.95e-05, "loss": 0.4531, "step": 330 }, { "epoch": 0.15576470588235294, "grad_norm": 0.46462265838628436, "learning_rate": 4.965e-05, "loss": 0.0991, "step": 331 }, { "epoch": 0.15623529411764706, "grad_norm": 1.3104341272344626, "learning_rate": 4.98e-05, "loss": 0.4297, "step": 332 }, { "epoch": 0.15670588235294117, "grad_norm": 0.6317950797364377, "learning_rate": 4.995e-05, "loss": 0.2061, "step": 333 }, { "epoch": 0.1571764705882353, "grad_norm": 0.3735775365089679, "learning_rate": 5.01e-05, "loss": 0.1328, "step": 334 }, { "epoch": 0.15764705882352942, "grad_norm": 0.19578042416571886, "learning_rate": 5.025e-05, "loss": 0.0225, "step": 335 }, { "epoch": 0.15811764705882353, "grad_norm": 1.0129311177684392, "learning_rate": 5.04e-05, "loss": 0.3574, "step": 336 }, { "epoch": 0.15858823529411764, "grad_norm": 0.8216872920761591, "learning_rate": 5.055e-05, "loss": 0.291, "step": 337 }, { "epoch": 0.15905882352941175, "grad_norm": 0.7766025902132522, "learning_rate": 5.07e-05, "loss": 0.3242, "step": 338 }, { "epoch": 0.1595294117647059, "grad_norm": 0.8782366546087371, "learning_rate": 5.0849999999999996e-05, "loss": 0.208, "step": 339 }, { "epoch": 0.16, "grad_norm": 1.3039201321842566, "learning_rate": 5.1e-05, "loss": 0.4434, "step": 340 }, { "epoch": 0.16047058823529411, "grad_norm": 0.5872420844249224, "learning_rate": 5.1149999999999996e-05, "loss": 0.2637, "step": 341 }, { "epoch": 0.16094117647058823, "grad_norm": 1.2212968876690322, "learning_rate": 5.13e-05, "loss": 0.3125, "step": 342 }, { "epoch": 0.16141176470588237, "grad_norm": 1.3413506567900586, "learning_rate": 5.145e-05, "loss": 0.5156, "step": 343 }, { "epoch": 0.16188235294117648, "grad_norm": 0.7712952435865623, "learning_rate": 5.1599999999999994e-05, "loss": 0.3711, "step": 344 }, { "epoch": 0.1623529411764706, "grad_norm": 0.8107797338360077, "learning_rate": 5.174999999999999e-05, "loss": 0.0923, "step": 345 }, { "epoch": 0.1628235294117647, "grad_norm": 1.1225161154543688, "learning_rate": 5.1899999999999994e-05, "loss": 0.2949, "step": 346 }, { "epoch": 0.1632941176470588, "grad_norm": 0.6107122134633587, "learning_rate": 5.204999999999999e-05, "loss": 0.1826, "step": 347 }, { "epoch": 0.16376470588235295, "grad_norm": 0.22954251409908097, "learning_rate": 5.2199999999999995e-05, "loss": 0.0092, "step": 348 }, { "epoch": 0.16423529411764706, "grad_norm": 0.6274110867520188, "learning_rate": 5.234999999999999e-05, "loss": 0.2598, "step": 349 }, { "epoch": 0.16470588235294117, "grad_norm": 0.7000252720218264, "learning_rate": 5.2499999999999995e-05, "loss": 0.2266, "step": 350 }, { "epoch": 0.16517647058823529, "grad_norm": 0.6213698357175527, "learning_rate": 5.264999999999999e-05, "loss": 0.25, "step": 351 }, { "epoch": 0.16564705882352943, "grad_norm": 0.8909972457830243, "learning_rate": 5.279999999999999e-05, "loss": 0.3711, "step": 352 }, { "epoch": 0.16611764705882354, "grad_norm": 0.4936553098874544, "learning_rate": 5.294999999999999e-05, "loss": 0.1455, "step": 353 }, { "epoch": 0.16658823529411765, "grad_norm": 0.3537652685315274, "learning_rate": 5.309999999999999e-05, "loss": 0.0732, "step": 354 }, { "epoch": 0.16705882352941176, "grad_norm": 0.4119115106879398, "learning_rate": 5.324999999999999e-05, "loss": 0.0591, "step": 355 }, { "epoch": 0.16752941176470587, "grad_norm": 0.4982765027433932, "learning_rate": 5.339999999999999e-05, "loss": 0.0957, "step": 356 }, { "epoch": 0.168, "grad_norm": 1.000007497537447, "learning_rate": 5.3549999999999994e-05, "loss": 0.2246, "step": 357 }, { "epoch": 0.16847058823529412, "grad_norm": 0.4508520437366611, "learning_rate": 5.369999999999999e-05, "loss": 0.1504, "step": 358 }, { "epoch": 0.16894117647058823, "grad_norm": 0.5893178600122496, "learning_rate": 5.3849999999999994e-05, "loss": 0.1543, "step": 359 }, { "epoch": 0.16941176470588235, "grad_norm": 0.24277319607285167, "learning_rate": 5.399999999999999e-05, "loss": 0.0457, "step": 360 }, { "epoch": 0.16988235294117648, "grad_norm": 0.6573092638152923, "learning_rate": 5.4149999999999995e-05, "loss": 0.2539, "step": 361 }, { "epoch": 0.1703529411764706, "grad_norm": 0.4919912485735255, "learning_rate": 5.429999999999999e-05, "loss": 0.124, "step": 362 }, { "epoch": 0.1708235294117647, "grad_norm": 0.6058591604077519, "learning_rate": 5.4449999999999995e-05, "loss": 0.1621, "step": 363 }, { "epoch": 0.17129411764705882, "grad_norm": 0.8928929212617583, "learning_rate": 5.459999999999999e-05, "loss": 0.3184, "step": 364 }, { "epoch": 0.17176470588235293, "grad_norm": 0.4428589560452053, "learning_rate": 5.4749999999999996e-05, "loss": 0.0928, "step": 365 }, { "epoch": 0.17223529411764707, "grad_norm": 0.28073746829684465, "learning_rate": 5.489999999999999e-05, "loss": 0.0393, "step": 366 }, { "epoch": 0.17270588235294118, "grad_norm": 1.4742716984243056, "learning_rate": 5.5049999999999996e-05, "loss": 0.3027, "step": 367 }, { "epoch": 0.1731764705882353, "grad_norm": 0.9148221260404318, "learning_rate": 5.519999999999999e-05, "loss": 0.2383, "step": 368 }, { "epoch": 0.1736470588235294, "grad_norm": 0.845325956551309, "learning_rate": 5.535e-05, "loss": 0.3691, "step": 369 }, { "epoch": 0.17411764705882352, "grad_norm": 0.9892830942438218, "learning_rate": 5.5499999999999994e-05, "loss": 0.3008, "step": 370 }, { "epoch": 0.17458823529411766, "grad_norm": 0.7548578862086637, "learning_rate": 5.565e-05, "loss": 0.2471, "step": 371 }, { "epoch": 0.17505882352941177, "grad_norm": 0.36580253066671575, "learning_rate": 5.5799999999999994e-05, "loss": 0.0938, "step": 372 }, { "epoch": 0.17552941176470588, "grad_norm": 0.49316503739156975, "learning_rate": 5.595e-05, "loss": 0.0376, "step": 373 }, { "epoch": 0.176, "grad_norm": 0.7156799910787572, "learning_rate": 5.6099999999999995e-05, "loss": 0.2246, "step": 374 }, { "epoch": 0.17647058823529413, "grad_norm": 0.8847083019501687, "learning_rate": 5.625e-05, "loss": 0.4141, "step": 375 }, { "epoch": 0.17694117647058824, "grad_norm": 1.1198844056253916, "learning_rate": 5.6399999999999995e-05, "loss": 0.6602, "step": 376 }, { "epoch": 0.17741176470588235, "grad_norm": 0.6119151995686666, "learning_rate": 5.654999999999999e-05, "loss": 0.2715, "step": 377 }, { "epoch": 0.17788235294117646, "grad_norm": 0.39510575694094247, "learning_rate": 5.6699999999999996e-05, "loss": 0.0796, "step": 378 }, { "epoch": 0.17835294117647058, "grad_norm": 0.3214425214660216, "learning_rate": 5.684999999999999e-05, "loss": 0.0742, "step": 379 }, { "epoch": 0.17882352941176471, "grad_norm": 0.9086914922866958, "learning_rate": 5.6999999999999996e-05, "loss": 0.3203, "step": 380 }, { "epoch": 0.17929411764705883, "grad_norm": 0.5522092819602247, "learning_rate": 5.714999999999999e-05, "loss": 0.1914, "step": 381 }, { "epoch": 0.17976470588235294, "grad_norm": 0.6613782775513914, "learning_rate": 5.73e-05, "loss": 0.2578, "step": 382 }, { "epoch": 0.18023529411764705, "grad_norm": 0.1294373515193004, "learning_rate": 5.7449999999999994e-05, "loss": 0.0217, "step": 383 }, { "epoch": 0.1807058823529412, "grad_norm": 1.0484255093831976, "learning_rate": 5.76e-05, "loss": 0.3945, "step": 384 }, { "epoch": 0.1811764705882353, "grad_norm": 0.6986041206497703, "learning_rate": 5.7749999999999994e-05, "loss": 0.3145, "step": 385 }, { "epoch": 0.1816470588235294, "grad_norm": 0.784932608342283, "learning_rate": 5.79e-05, "loss": 0.293, "step": 386 }, { "epoch": 0.18211764705882352, "grad_norm": 0.33365401429267033, "learning_rate": 5.8049999999999995e-05, "loss": 0.0476, "step": 387 }, { "epoch": 0.18258823529411763, "grad_norm": 0.878382417482022, "learning_rate": 5.82e-05, "loss": 0.3848, "step": 388 }, { "epoch": 0.18305882352941177, "grad_norm": 0.3611674675367117, "learning_rate": 5.8349999999999995e-05, "loss": 0.1641, "step": 389 }, { "epoch": 0.18352941176470589, "grad_norm": 0.62342578501463, "learning_rate": 5.85e-05, "loss": 0.2832, "step": 390 }, { "epoch": 0.184, "grad_norm": 0.8597020247910956, "learning_rate": 5.8649999999999996e-05, "loss": 0.3359, "step": 391 }, { "epoch": 0.1844705882352941, "grad_norm": 0.2795931374667671, "learning_rate": 5.88e-05, "loss": 0.0874, "step": 392 }, { "epoch": 0.18494117647058825, "grad_norm": 0.8897741151449923, "learning_rate": 5.8949999999999996e-05, "loss": 0.4375, "step": 393 }, { "epoch": 0.18541176470588236, "grad_norm": 0.5338206601639112, "learning_rate": 5.91e-05, "loss": 0.125, "step": 394 }, { "epoch": 0.18588235294117647, "grad_norm": 0.6074791286328972, "learning_rate": 5.925e-05, "loss": 0.1484, "step": 395 }, { "epoch": 0.18635294117647058, "grad_norm": 1.0495820905817197, "learning_rate": 5.94e-05, "loss": 0.5039, "step": 396 }, { "epoch": 0.1868235294117647, "grad_norm": 0.4101257288299696, "learning_rate": 5.955e-05, "loss": 0.1387, "step": 397 }, { "epoch": 0.18729411764705883, "grad_norm": 0.21056130908533172, "learning_rate": 5.97e-05, "loss": 0.041, "step": 398 }, { "epoch": 0.18776470588235294, "grad_norm": 0.7260693110410155, "learning_rate": 5.985e-05, "loss": 0.3027, "step": 399 }, { "epoch": 0.18823529411764706, "grad_norm": 0.3256384173182311, "learning_rate": 5.9999999999999995e-05, "loss": 0.0593, "step": 400 }, { "epoch": 0.18870588235294117, "grad_norm": 0.3161799216859178, "learning_rate": 6.015e-05, "loss": 0.0845, "step": 401 }, { "epoch": 0.1891764705882353, "grad_norm": 0.5875948134362763, "learning_rate": 6.0299999999999995e-05, "loss": 0.105, "step": 402 }, { "epoch": 0.18964705882352942, "grad_norm": 0.8583041383070734, "learning_rate": 6.045e-05, "loss": 0.377, "step": 403 }, { "epoch": 0.19011764705882353, "grad_norm": 0.8762188750877999, "learning_rate": 6.0599999999999996e-05, "loss": 0.5469, "step": 404 }, { "epoch": 0.19058823529411764, "grad_norm": 0.14963508344794363, "learning_rate": 6.075e-05, "loss": 0.0193, "step": 405 }, { "epoch": 0.19105882352941175, "grad_norm": 1.0416974399734202, "learning_rate": 6.0899999999999996e-05, "loss": 0.5859, "step": 406 }, { "epoch": 0.1915294117647059, "grad_norm": 0.607716537866937, "learning_rate": 6.104999999999999e-05, "loss": 0.2344, "step": 407 }, { "epoch": 0.192, "grad_norm": 0.2593713201505706, "learning_rate": 6.12e-05, "loss": 0.0435, "step": 408 }, { "epoch": 0.19247058823529412, "grad_norm": 0.6123104919557523, "learning_rate": 6.134999999999999e-05, "loss": 0.2256, "step": 409 }, { "epoch": 0.19294117647058823, "grad_norm": 0.8954554595003723, "learning_rate": 6.149999999999999e-05, "loss": 0.3809, "step": 410 }, { "epoch": 0.19341176470588237, "grad_norm": 0.5954497499517709, "learning_rate": 6.165e-05, "loss": 0.2969, "step": 411 }, { "epoch": 0.19388235294117648, "grad_norm": 0.3620595153552274, "learning_rate": 6.18e-05, "loss": 0.0903, "step": 412 }, { "epoch": 0.1943529411764706, "grad_norm": 0.5347798840888999, "learning_rate": 6.194999999999999e-05, "loss": 0.0854, "step": 413 }, { "epoch": 0.1948235294117647, "grad_norm": 0.3784450157758678, "learning_rate": 6.209999999999999e-05, "loss": 0.0583, "step": 414 }, { "epoch": 0.1952941176470588, "grad_norm": 0.31760143748447656, "learning_rate": 6.225e-05, "loss": 0.0396, "step": 415 }, { "epoch": 0.19576470588235295, "grad_norm": 2.0079605714091993, "learning_rate": 6.239999999999999e-05, "loss": 0.6328, "step": 416 }, { "epoch": 0.19623529411764706, "grad_norm": 0.639836193413352, "learning_rate": 6.254999999999999e-05, "loss": 0.3242, "step": 417 }, { "epoch": 0.19670588235294117, "grad_norm": 0.4364868463200541, "learning_rate": 6.269999999999999e-05, "loss": 0.1196, "step": 418 }, { "epoch": 0.1971764705882353, "grad_norm": 1.3305030156842392, "learning_rate": 6.285e-05, "loss": 0.3633, "step": 419 }, { "epoch": 0.1976470588235294, "grad_norm": 0.32578023484070545, "learning_rate": 6.299999999999999e-05, "loss": 0.0698, "step": 420 }, { "epoch": 0.19811764705882354, "grad_norm": 0.49727999294153447, "learning_rate": 6.314999999999999e-05, "loss": 0.1621, "step": 421 }, { "epoch": 0.19858823529411765, "grad_norm": 1.666758416301894, "learning_rate": 6.33e-05, "loss": 0.4336, "step": 422 }, { "epoch": 0.19905882352941176, "grad_norm": 0.46698452561837406, "learning_rate": 6.345e-05, "loss": 0.0342, "step": 423 }, { "epoch": 0.19952941176470587, "grad_norm": 0.7558905902204056, "learning_rate": 6.359999999999999e-05, "loss": 0.3457, "step": 424 }, { "epoch": 0.2, "grad_norm": 0.3406085306108928, "learning_rate": 6.374999999999999e-05, "loss": 0.0928, "step": 425 }, { "epoch": 0.20047058823529412, "grad_norm": 1.0897278046875087, "learning_rate": 6.39e-05, "loss": 0.3887, "step": 426 }, { "epoch": 0.20094117647058823, "grad_norm": 0.6148236227756396, "learning_rate": 6.405e-05, "loss": 0.2734, "step": 427 }, { "epoch": 0.20141176470588235, "grad_norm": 0.51612077959547, "learning_rate": 6.419999999999999e-05, "loss": 0.2793, "step": 428 }, { "epoch": 0.20188235294117646, "grad_norm": 0.34031574680878823, "learning_rate": 6.434999999999999e-05, "loss": 0.1641, "step": 429 }, { "epoch": 0.2023529411764706, "grad_norm": 0.5110638296438663, "learning_rate": 6.45e-05, "loss": 0.2227, "step": 430 }, { "epoch": 0.2028235294117647, "grad_norm": 0.39288346443883754, "learning_rate": 6.465e-05, "loss": 0.0713, "step": 431 }, { "epoch": 0.20329411764705882, "grad_norm": 0.5588364850565736, "learning_rate": 6.479999999999999e-05, "loss": 0.2285, "step": 432 }, { "epoch": 0.20376470588235293, "grad_norm": 0.15318611386600406, "learning_rate": 6.494999999999999e-05, "loss": 0.0236, "step": 433 }, { "epoch": 0.20423529411764707, "grad_norm": 0.37436211248501616, "learning_rate": 6.51e-05, "loss": 0.1416, "step": 434 }, { "epoch": 0.20470588235294118, "grad_norm": 0.6049966740901396, "learning_rate": 6.525e-05, "loss": 0.2793, "step": 435 }, { "epoch": 0.2051764705882353, "grad_norm": 0.3223346081975797, "learning_rate": 6.539999999999999e-05, "loss": 0.0698, "step": 436 }, { "epoch": 0.2056470588235294, "grad_norm": 0.6006374825129116, "learning_rate": 6.555e-05, "loss": 0.3184, "step": 437 }, { "epoch": 0.20611764705882352, "grad_norm": 0.4389315970406382, "learning_rate": 6.57e-05, "loss": 0.165, "step": 438 }, { "epoch": 0.20658823529411766, "grad_norm": 0.43173895779624955, "learning_rate": 6.584999999999999e-05, "loss": 0.2168, "step": 439 }, { "epoch": 0.20705882352941177, "grad_norm": 0.49792277263344475, "learning_rate": 6.599999999999999e-05, "loss": 0.2314, "step": 440 }, { "epoch": 0.20752941176470588, "grad_norm": 0.41898619920986585, "learning_rate": 6.615e-05, "loss": 0.1689, "step": 441 }, { "epoch": 0.208, "grad_norm": 1.2584914487497825, "learning_rate": 6.63e-05, "loss": 0.3301, "step": 442 }, { "epoch": 0.20847058823529413, "grad_norm": 0.5121239039402783, "learning_rate": 6.644999999999999e-05, "loss": 0.2256, "step": 443 }, { "epoch": 0.20894117647058824, "grad_norm": 0.33778109071021467, "learning_rate": 6.659999999999999e-05, "loss": 0.1035, "step": 444 }, { "epoch": 0.20941176470588235, "grad_norm": 0.7997395626604156, "learning_rate": 6.675e-05, "loss": 0.3945, "step": 445 }, { "epoch": 0.20988235294117646, "grad_norm": 0.4462568564530529, "learning_rate": 6.69e-05, "loss": 0.0894, "step": 446 }, { "epoch": 0.21035294117647058, "grad_norm": 0.41070857676590655, "learning_rate": 6.704999999999999e-05, "loss": 0.1387, "step": 447 }, { "epoch": 0.21082352941176472, "grad_norm": 0.496804410891264, "learning_rate": 6.72e-05, "loss": 0.1553, "step": 448 }, { "epoch": 0.21129411764705883, "grad_norm": 0.8025027750203263, "learning_rate": 6.735e-05, "loss": 0.3047, "step": 449 }, { "epoch": 0.21176470588235294, "grad_norm": 0.7207483410988521, "learning_rate": 6.75e-05, "loss": 0.3945, "step": 450 }, { "epoch": 0.21223529411764705, "grad_norm": 0.6304435676349105, "learning_rate": 6.764999999999999e-05, "loss": 0.2051, "step": 451 }, { "epoch": 0.2127058823529412, "grad_norm": 0.8516451554679207, "learning_rate": 6.78e-05, "loss": 0.3398, "step": 452 }, { "epoch": 0.2131764705882353, "grad_norm": 0.23894992960862882, "learning_rate": 6.795e-05, "loss": 0.0388, "step": 453 }, { "epoch": 0.2136470588235294, "grad_norm": 0.5871485969892457, "learning_rate": 6.81e-05, "loss": 0.2871, "step": 454 }, { "epoch": 0.21411764705882352, "grad_norm": 0.29564096105264387, "learning_rate": 6.824999999999999e-05, "loss": 0.0977, "step": 455 }, { "epoch": 0.21458823529411764, "grad_norm": 0.7252065572068596, "learning_rate": 6.84e-05, "loss": 0.3398, "step": 456 }, { "epoch": 0.21505882352941177, "grad_norm": 0.45325877997989816, "learning_rate": 6.855e-05, "loss": 0.166, "step": 457 }, { "epoch": 0.21552941176470589, "grad_norm": 0.43495290834454664, "learning_rate": 6.87e-05, "loss": 0.165, "step": 458 }, { "epoch": 0.216, "grad_norm": 0.9273215989586128, "learning_rate": 6.884999999999999e-05, "loss": 0.291, "step": 459 }, { "epoch": 0.2164705882352941, "grad_norm": 0.8752895148265798, "learning_rate": 6.9e-05, "loss": 0.3164, "step": 460 }, { "epoch": 0.21694117647058825, "grad_norm": 0.6988075314960966, "learning_rate": 6.915e-05, "loss": 0.3535, "step": 461 }, { "epoch": 0.21741176470588236, "grad_norm": 0.693355398458505, "learning_rate": 6.93e-05, "loss": 0.1523, "step": 462 }, { "epoch": 0.21788235294117647, "grad_norm": 0.32773893193378284, "learning_rate": 6.945e-05, "loss": 0.0825, "step": 463 }, { "epoch": 0.21835294117647058, "grad_norm": 0.7898509824725533, "learning_rate": 6.96e-05, "loss": 0.2344, "step": 464 }, { "epoch": 0.2188235294117647, "grad_norm": 0.12608295250122234, "learning_rate": 6.975e-05, "loss": 0.0265, "step": 465 }, { "epoch": 0.21929411764705883, "grad_norm": 0.4912714019912382, "learning_rate": 6.989999999999999e-05, "loss": 0.1206, "step": 466 }, { "epoch": 0.21976470588235295, "grad_norm": 1.2432490697807095, "learning_rate": 7.005e-05, "loss": 0.4688, "step": 467 }, { "epoch": 0.22023529411764706, "grad_norm": 0.5400097515399174, "learning_rate": 7.02e-05, "loss": 0.1543, "step": 468 }, { "epoch": 0.22070588235294117, "grad_norm": 1.0129023757677407, "learning_rate": 7.034999999999999e-05, "loss": 0.1299, "step": 469 }, { "epoch": 0.2211764705882353, "grad_norm": 0.532562549368051, "learning_rate": 7.049999999999999e-05, "loss": 0.1826, "step": 470 }, { "epoch": 0.22164705882352942, "grad_norm": 0.6662667167218821, "learning_rate": 7.065e-05, "loss": 0.3184, "step": 471 }, { "epoch": 0.22211764705882353, "grad_norm": 0.563618677082448, "learning_rate": 7.079999999999999e-05, "loss": 0.2754, "step": 472 }, { "epoch": 0.22258823529411764, "grad_norm": 0.6459057439378746, "learning_rate": 7.094999999999999e-05, "loss": 0.2207, "step": 473 }, { "epoch": 0.22305882352941175, "grad_norm": 0.17337711917943557, "learning_rate": 7.11e-05, "loss": 0.0469, "step": 474 }, { "epoch": 0.2235294117647059, "grad_norm": 1.134205922539958, "learning_rate": 7.125e-05, "loss": 0.3652, "step": 475 }, { "epoch": 0.224, "grad_norm": 0.6684568400872262, "learning_rate": 7.139999999999999e-05, "loss": 0.3691, "step": 476 }, { "epoch": 0.22447058823529412, "grad_norm": 0.9416921232423248, "learning_rate": 7.154999999999999e-05, "loss": 0.4531, "step": 477 }, { "epoch": 0.22494117647058823, "grad_norm": 0.5251379724266426, "learning_rate": 7.17e-05, "loss": 0.127, "step": 478 }, { "epoch": 0.22541176470588234, "grad_norm": 0.31626866470954956, "learning_rate": 7.184999999999998e-05, "loss": 0.0659, "step": 479 }, { "epoch": 0.22588235294117648, "grad_norm": 0.367494260674843, "learning_rate": 7.199999999999999e-05, "loss": 0.1309, "step": 480 }, { "epoch": 0.2263529411764706, "grad_norm": 1.0215051683912817, "learning_rate": 7.214999999999999e-05, "loss": 0.5273, "step": 481 }, { "epoch": 0.2268235294117647, "grad_norm": 0.5593100053612886, "learning_rate": 7.23e-05, "loss": 0.2471, "step": 482 }, { "epoch": 0.2272941176470588, "grad_norm": 0.7930743399689717, "learning_rate": 7.244999999999999e-05, "loss": 0.2852, "step": 483 }, { "epoch": 0.22776470588235295, "grad_norm": 0.6308836568989419, "learning_rate": 7.259999999999999e-05, "loss": 0.3047, "step": 484 }, { "epoch": 0.22823529411764706, "grad_norm": 0.4187886907713778, "learning_rate": 7.274999999999999e-05, "loss": 0.1177, "step": 485 }, { "epoch": 0.22870588235294118, "grad_norm": 0.4709005863985298, "learning_rate": 7.29e-05, "loss": 0.1768, "step": 486 }, { "epoch": 0.2291764705882353, "grad_norm": 0.2907112586568621, "learning_rate": 7.304999999999999e-05, "loss": 0.0825, "step": 487 }, { "epoch": 0.2296470588235294, "grad_norm": 0.2266152899291524, "learning_rate": 7.319999999999999e-05, "loss": 0.0659, "step": 488 }, { "epoch": 0.23011764705882354, "grad_norm": 0.35338820223565276, "learning_rate": 7.335e-05, "loss": 0.1172, "step": 489 }, { "epoch": 0.23058823529411765, "grad_norm": 0.5599941458269876, "learning_rate": 7.35e-05, "loss": 0.2041, "step": 490 }, { "epoch": 0.23105882352941176, "grad_norm": 0.15835780189893633, "learning_rate": 7.364999999999999e-05, "loss": 0.0432, "step": 491 }, { "epoch": 0.23152941176470587, "grad_norm": 0.5264218836801889, "learning_rate": 7.379999999999999e-05, "loss": 0.3516, "step": 492 }, { "epoch": 0.232, "grad_norm": 0.7107715480155977, "learning_rate": 7.395e-05, "loss": 0.4219, "step": 493 }, { "epoch": 0.23247058823529412, "grad_norm": 0.25573175517541363, "learning_rate": 7.41e-05, "loss": 0.1138, "step": 494 }, { "epoch": 0.23294117647058823, "grad_norm": 0.6795786656586539, "learning_rate": 7.424999999999999e-05, "loss": 0.4082, "step": 495 }, { "epoch": 0.23341176470588235, "grad_norm": 0.48194253972786166, "learning_rate": 7.439999999999999e-05, "loss": 0.2598, "step": 496 }, { "epoch": 0.23388235294117646, "grad_norm": 0.4490334695880054, "learning_rate": 7.455e-05, "loss": 0.3262, "step": 497 }, { "epoch": 0.2343529411764706, "grad_norm": 0.3631633424912013, "learning_rate": 7.47e-05, "loss": 0.1641, "step": 498 }, { "epoch": 0.2348235294117647, "grad_norm": 0.5811951539771774, "learning_rate": 7.484999999999999e-05, "loss": 0.2598, "step": 499 }, { "epoch": 0.23529411764705882, "grad_norm": 0.32029200475331054, "learning_rate": 7.5e-05, "loss": 0.0898, "step": 500 }, { "epoch": 0.23576470588235293, "grad_norm": 0.46793957176816253, "learning_rate": 7.515e-05, "loss": 0.1611, "step": 501 }, { "epoch": 0.23623529411764707, "grad_norm": 0.5243470932629672, "learning_rate": 7.529999999999999e-05, "loss": 0.1768, "step": 502 }, { "epoch": 0.23670588235294118, "grad_norm": 0.32848328456312254, "learning_rate": 7.544999999999999e-05, "loss": 0.1484, "step": 503 }, { "epoch": 0.2371764705882353, "grad_norm": 0.33432019852142564, "learning_rate": 7.56e-05, "loss": 0.1377, "step": 504 }, { "epoch": 0.2376470588235294, "grad_norm": 1.3214215138991232, "learning_rate": 7.575e-05, "loss": 0.5547, "step": 505 }, { "epoch": 0.23811764705882352, "grad_norm": 0.3033218806702392, "learning_rate": 7.589999999999999e-05, "loss": 0.0913, "step": 506 }, { "epoch": 0.23858823529411766, "grad_norm": 0.553332337411801, "learning_rate": 7.604999999999999e-05, "loss": 0.2256, "step": 507 }, { "epoch": 0.23905882352941177, "grad_norm": 0.1811864169625102, "learning_rate": 7.62e-05, "loss": 0.0454, "step": 508 }, { "epoch": 0.23952941176470588, "grad_norm": 0.3063672485132768, "learning_rate": 7.635e-05, "loss": 0.0898, "step": 509 }, { "epoch": 0.24, "grad_norm": 0.2745689795254573, "learning_rate": 7.649999999999999e-05, "loss": 0.0542, "step": 510 }, { "epoch": 0.24047058823529413, "grad_norm": 0.7810839812006951, "learning_rate": 7.664999999999999e-05, "loss": 0.3398, "step": 511 }, { "epoch": 0.24094117647058824, "grad_norm": 0.28500470399757855, "learning_rate": 7.68e-05, "loss": 0.0737, "step": 512 }, { "epoch": 0.24141176470588235, "grad_norm": 0.33259789393122047, "learning_rate": 7.695e-05, "loss": 0.1025, "step": 513 }, { "epoch": 0.24188235294117647, "grad_norm": 0.28102158487600865, "learning_rate": 7.709999999999999e-05, "loss": 0.0757, "step": 514 }, { "epoch": 0.24235294117647058, "grad_norm": 0.508766350605634, "learning_rate": 7.725e-05, "loss": 0.2393, "step": 515 }, { "epoch": 0.24282352941176472, "grad_norm": 0.48638967428610286, "learning_rate": 7.74e-05, "loss": 0.0942, "step": 516 }, { "epoch": 0.24329411764705883, "grad_norm": 0.4538404478212116, "learning_rate": 7.755e-05, "loss": 0.1504, "step": 517 }, { "epoch": 0.24376470588235294, "grad_norm": 0.29651954071734377, "learning_rate": 7.769999999999999e-05, "loss": 0.0811, "step": 518 }, { "epoch": 0.24423529411764705, "grad_norm": 0.6179704299617457, "learning_rate": 7.785e-05, "loss": 0.2363, "step": 519 }, { "epoch": 0.2447058823529412, "grad_norm": 0.36578394437715933, "learning_rate": 7.8e-05, "loss": 0.0996, "step": 520 }, { "epoch": 0.2451764705882353, "grad_norm": 0.5595840324770631, "learning_rate": 7.815e-05, "loss": 0.1992, "step": 521 }, { "epoch": 0.2456470588235294, "grad_norm": 0.23060167931235076, "learning_rate": 7.829999999999999e-05, "loss": 0.084, "step": 522 }, { "epoch": 0.24611764705882352, "grad_norm": 0.39245004215107543, "learning_rate": 7.845e-05, "loss": 0.1963, "step": 523 }, { "epoch": 0.24658823529411764, "grad_norm": 0.6266412660633984, "learning_rate": 7.86e-05, "loss": 0.248, "step": 524 }, { "epoch": 0.24705882352941178, "grad_norm": 0.3791345014595914, "learning_rate": 7.874999999999999e-05, "loss": 0.0464, "step": 525 }, { "epoch": 0.2475294117647059, "grad_norm": 0.6115959769652561, "learning_rate": 7.89e-05, "loss": 0.2754, "step": 526 }, { "epoch": 0.248, "grad_norm": 0.23762978638071394, "learning_rate": 7.905e-05, "loss": 0.0596, "step": 527 }, { "epoch": 0.2484705882352941, "grad_norm": 0.662265029810628, "learning_rate": 7.92e-05, "loss": 0.4023, "step": 528 }, { "epoch": 0.24894117647058825, "grad_norm": 0.36998120770290055, "learning_rate": 7.934999999999999e-05, "loss": 0.1138, "step": 529 }, { "epoch": 0.24941176470588236, "grad_norm": 0.9381967595223902, "learning_rate": 7.95e-05, "loss": 0.5938, "step": 530 }, { "epoch": 0.24988235294117647, "grad_norm": 0.33251490329196254, "learning_rate": 7.965e-05, "loss": 0.167, "step": 531 }, { "epoch": 0.2503529411764706, "grad_norm": 0.1175707743539181, "learning_rate": 7.98e-05, "loss": 0.0084, "step": 532 }, { "epoch": 0.2508235294117647, "grad_norm": 0.773406913703576, "learning_rate": 7.994999999999999e-05, "loss": 0.334, "step": 533 }, { "epoch": 0.25129411764705883, "grad_norm": 0.4488770294916928, "learning_rate": 8.01e-05, "loss": 0.1543, "step": 534 }, { "epoch": 0.25176470588235295, "grad_norm": 0.612213186814733, "learning_rate": 8.025e-05, "loss": 0.208, "step": 535 }, { "epoch": 0.25223529411764706, "grad_norm": 0.42647786039846386, "learning_rate": 8.04e-05, "loss": 0.1514, "step": 536 }, { "epoch": 0.25270588235294117, "grad_norm": 0.450043417178882, "learning_rate": 8.054999999999999e-05, "loss": 0.2031, "step": 537 }, { "epoch": 0.2531764705882353, "grad_norm": 0.37608450151506745, "learning_rate": 8.07e-05, "loss": 0.2012, "step": 538 }, { "epoch": 0.2536470588235294, "grad_norm": 0.15052176513542637, "learning_rate": 8.085e-05, "loss": 0.0083, "step": 539 }, { "epoch": 0.2541176470588235, "grad_norm": 0.7407634937306354, "learning_rate": 8.1e-05, "loss": 0.3145, "step": 540 }, { "epoch": 0.25458823529411767, "grad_norm": 0.3750865634040353, "learning_rate": 8.115e-05, "loss": 0.2139, "step": 541 }, { "epoch": 0.2550588235294118, "grad_norm": 0.26981213477881105, "learning_rate": 8.13e-05, "loss": 0.0952, "step": 542 }, { "epoch": 0.2555294117647059, "grad_norm": 0.3994448210569468, "learning_rate": 8.145e-05, "loss": 0.1641, "step": 543 }, { "epoch": 0.256, "grad_norm": 0.5345534633588487, "learning_rate": 8.16e-05, "loss": 0.2148, "step": 544 }, { "epoch": 0.2564705882352941, "grad_norm": 1.2216490203923538, "learning_rate": 8.175e-05, "loss": 0.4004, "step": 545 }, { "epoch": 0.25694117647058823, "grad_norm": 0.5668454560197818, "learning_rate": 8.19e-05, "loss": 0.2715, "step": 546 }, { "epoch": 0.25741176470588234, "grad_norm": 1.5055054026410901, "learning_rate": 8.205e-05, "loss": 0.3887, "step": 547 }, { "epoch": 0.25788235294117645, "grad_norm": 0.7124678565313054, "learning_rate": 8.22e-05, "loss": 0.4258, "step": 548 }, { "epoch": 0.25835294117647056, "grad_norm": 0.12633767227246923, "learning_rate": 8.235e-05, "loss": 0.0366, "step": 549 }, { "epoch": 0.25882352941176473, "grad_norm": 0.8900625832301083, "learning_rate": 8.25e-05, "loss": 0.2305, "step": 550 }, { "epoch": 0.25929411764705884, "grad_norm": 0.7881587148748513, "learning_rate": 8.265e-05, "loss": 0.2041, "step": 551 }, { "epoch": 0.25976470588235295, "grad_norm": 0.35306096327742487, "learning_rate": 8.28e-05, "loss": 0.1035, "step": 552 }, { "epoch": 0.26023529411764706, "grad_norm": 1.0652716529346682, "learning_rate": 8.295e-05, "loss": 0.2676, "step": 553 }, { "epoch": 0.2607058823529412, "grad_norm": 0.985807353415178, "learning_rate": 8.31e-05, "loss": 0.4355, "step": 554 }, { "epoch": 0.2611764705882353, "grad_norm": 1.4340133302555131, "learning_rate": 8.325e-05, "loss": 0.5781, "step": 555 }, { "epoch": 0.2616470588235294, "grad_norm": 0.5229914227778542, "learning_rate": 8.34e-05, "loss": 0.2969, "step": 556 }, { "epoch": 0.2621176470588235, "grad_norm": 0.6467929437113247, "learning_rate": 8.355e-05, "loss": 0.293, "step": 557 }, { "epoch": 0.2625882352941176, "grad_norm": 0.501456820047123, "learning_rate": 8.37e-05, "loss": 0.2139, "step": 558 }, { "epoch": 0.2630588235294118, "grad_norm": 0.573035517309768, "learning_rate": 8.385e-05, "loss": 0.2988, "step": 559 }, { "epoch": 0.2635294117647059, "grad_norm": 0.38565125479667395, "learning_rate": 8.4e-05, "loss": 0.2207, "step": 560 }, { "epoch": 0.264, "grad_norm": 0.4463722728923688, "learning_rate": 8.415e-05, "loss": 0.1611, "step": 561 }, { "epoch": 0.2644705882352941, "grad_norm": 0.21765139960720392, "learning_rate": 8.43e-05, "loss": 0.0444, "step": 562 }, { "epoch": 0.26494117647058824, "grad_norm": 0.20307346908742074, "learning_rate": 8.444999999999998e-05, "loss": 0.0723, "step": 563 }, { "epoch": 0.26541176470588235, "grad_norm": 0.6607736869410401, "learning_rate": 8.459999999999998e-05, "loss": 0.2324, "step": 564 }, { "epoch": 0.26588235294117646, "grad_norm": 0.5800158274712031, "learning_rate": 8.474999999999999e-05, "loss": 0.3047, "step": 565 }, { "epoch": 0.26635294117647057, "grad_norm": 0.9539533498061589, "learning_rate": 8.489999999999999e-05, "loss": 0.459, "step": 566 }, { "epoch": 0.2668235294117647, "grad_norm": 0.7273806182887096, "learning_rate": 8.504999999999998e-05, "loss": 0.4082, "step": 567 }, { "epoch": 0.26729411764705885, "grad_norm": 0.2863483545425986, "learning_rate": 8.519999999999998e-05, "loss": 0.0752, "step": 568 }, { "epoch": 0.26776470588235296, "grad_norm": 0.27434520408190927, "learning_rate": 8.534999999999999e-05, "loss": 0.0845, "step": 569 }, { "epoch": 0.26823529411764707, "grad_norm": 1.1597649291549932, "learning_rate": 8.549999999999999e-05, "loss": 0.3945, "step": 570 }, { "epoch": 0.2687058823529412, "grad_norm": 0.5366826090077484, "learning_rate": 8.564999999999998e-05, "loss": 0.1611, "step": 571 }, { "epoch": 0.2691764705882353, "grad_norm": 1.2228498326035842, "learning_rate": 8.579999999999998e-05, "loss": 0.4492, "step": 572 }, { "epoch": 0.2696470588235294, "grad_norm": 0.44739009857410317, "learning_rate": 8.594999999999999e-05, "loss": 0.1152, "step": 573 }, { "epoch": 0.2701176470588235, "grad_norm": 0.5558862395017841, "learning_rate": 8.609999999999999e-05, "loss": 0.2197, "step": 574 }, { "epoch": 0.27058823529411763, "grad_norm": 0.35220130603907746, "learning_rate": 8.624999999999998e-05, "loss": 0.1436, "step": 575 }, { "epoch": 0.27105882352941174, "grad_norm": 0.6055521104257731, "learning_rate": 8.639999999999999e-05, "loss": 0.3184, "step": 576 }, { "epoch": 0.2715294117647059, "grad_norm": 0.5626876793736684, "learning_rate": 8.654999999999999e-05, "loss": 0.3398, "step": 577 }, { "epoch": 0.272, "grad_norm": 0.3968796598412928, "learning_rate": 8.669999999999998e-05, "loss": 0.2266, "step": 578 }, { "epoch": 0.27247058823529413, "grad_norm": 0.4993394951909001, "learning_rate": 8.684999999999998e-05, "loss": 0.2031, "step": 579 }, { "epoch": 0.27294117647058824, "grad_norm": 0.5492088280666209, "learning_rate": 8.699999999999999e-05, "loss": 0.2559, "step": 580 }, { "epoch": 0.27341176470588235, "grad_norm": 0.22916907892956437, "learning_rate": 8.714999999999999e-05, "loss": 0.0669, "step": 581 }, { "epoch": 0.27388235294117647, "grad_norm": 0.37737992598369624, "learning_rate": 8.729999999999998e-05, "loss": 0.1152, "step": 582 }, { "epoch": 0.2743529411764706, "grad_norm": 0.6683743744635253, "learning_rate": 8.744999999999998e-05, "loss": 0.293, "step": 583 }, { "epoch": 0.2748235294117647, "grad_norm": 1.2987993146087722, "learning_rate": 8.759999999999999e-05, "loss": 0.4043, "step": 584 }, { "epoch": 0.2752941176470588, "grad_norm": 0.7755148073956045, "learning_rate": 8.774999999999999e-05, "loss": 0.21, "step": 585 }, { "epoch": 0.27576470588235297, "grad_norm": 0.2819082838326636, "learning_rate": 8.789999999999998e-05, "loss": 0.0466, "step": 586 }, { "epoch": 0.2762352941176471, "grad_norm": 0.9255256654449154, "learning_rate": 8.804999999999999e-05, "loss": 0.5, "step": 587 }, { "epoch": 0.2767058823529412, "grad_norm": 0.5654152760274705, "learning_rate": 8.819999999999999e-05, "loss": 0.0562, "step": 588 }, { "epoch": 0.2771764705882353, "grad_norm": 0.8936800480795688, "learning_rate": 8.834999999999999e-05, "loss": 0.2383, "step": 589 }, { "epoch": 0.2776470588235294, "grad_norm": 0.5235777576426814, "learning_rate": 8.849999999999998e-05, "loss": 0.3262, "step": 590 }, { "epoch": 0.2781176470588235, "grad_norm": 0.7618695035709355, "learning_rate": 8.864999999999999e-05, "loss": 0.2432, "step": 591 }, { "epoch": 0.27858823529411764, "grad_norm": 0.14035584817353378, "learning_rate": 8.879999999999999e-05, "loss": 0.0354, "step": 592 }, { "epoch": 0.27905882352941175, "grad_norm": 0.1199383768646397, "learning_rate": 8.895e-05, "loss": 0.0122, "step": 593 }, { "epoch": 0.27952941176470586, "grad_norm": 0.6229840317983422, "learning_rate": 8.909999999999998e-05, "loss": 0.2266, "step": 594 }, { "epoch": 0.28, "grad_norm": 0.9166530357219094, "learning_rate": 8.924999999999999e-05, "loss": 0.4004, "step": 595 }, { "epoch": 0.28047058823529414, "grad_norm": 0.38566332451561625, "learning_rate": 8.939999999999999e-05, "loss": 0.1875, "step": 596 }, { "epoch": 0.28094117647058825, "grad_norm": 0.5949879488679902, "learning_rate": 8.955e-05, "loss": 0.3711, "step": 597 }, { "epoch": 0.28141176470588236, "grad_norm": 0.43020782812509256, "learning_rate": 8.969999999999998e-05, "loss": 0.1973, "step": 598 }, { "epoch": 0.2818823529411765, "grad_norm": 0.6870291619030582, "learning_rate": 8.984999999999999e-05, "loss": 0.3789, "step": 599 }, { "epoch": 0.2823529411764706, "grad_norm": 0.35772534034172726, "learning_rate": 8.999999999999999e-05, "loss": 0.1152, "step": 600 }, { "epoch": 0.2828235294117647, "grad_norm": 0.40380464069155, "learning_rate": 9.014999999999998e-05, "loss": 0.1367, "step": 601 }, { "epoch": 0.2832941176470588, "grad_norm": 0.20682170570307243, "learning_rate": 9.029999999999999e-05, "loss": 0.0466, "step": 602 }, { "epoch": 0.2837647058823529, "grad_norm": 0.4921353899061832, "learning_rate": 9.044999999999999e-05, "loss": 0.1865, "step": 603 }, { "epoch": 0.2842352941176471, "grad_norm": 0.4019016478338475, "learning_rate": 9.059999999999999e-05, "loss": 0.165, "step": 604 }, { "epoch": 0.2847058823529412, "grad_norm": 0.39233835199866973, "learning_rate": 9.074999999999998e-05, "loss": 0.2559, "step": 605 }, { "epoch": 0.2851764705882353, "grad_norm": 0.21009506560118327, "learning_rate": 9.089999999999999e-05, "loss": 0.0806, "step": 606 }, { "epoch": 0.2856470588235294, "grad_norm": 1.1996666056179672, "learning_rate": 9.104999999999999e-05, "loss": 0.4785, "step": 607 }, { "epoch": 0.28611764705882353, "grad_norm": 0.511936659152366, "learning_rate": 9.12e-05, "loss": 0.0688, "step": 608 }, { "epoch": 0.28658823529411764, "grad_norm": 0.4438232279220451, "learning_rate": 9.134999999999998e-05, "loss": 0.2354, "step": 609 }, { "epoch": 0.28705882352941176, "grad_norm": 0.2180128013868957, "learning_rate": 9.149999999999999e-05, "loss": 0.0493, "step": 610 }, { "epoch": 0.28752941176470587, "grad_norm": 1.0250608555142793, "learning_rate": 9.164999999999999e-05, "loss": 0.4004, "step": 611 }, { "epoch": 0.288, "grad_norm": 0.3536270699584222, "learning_rate": 9.18e-05, "loss": 0.1118, "step": 612 }, { "epoch": 0.28847058823529415, "grad_norm": 0.32569127769505213, "learning_rate": 9.194999999999999e-05, "loss": 0.127, "step": 613 }, { "epoch": 0.28894117647058826, "grad_norm": 0.33461733085655504, "learning_rate": 9.209999999999999e-05, "loss": 0.0488, "step": 614 }, { "epoch": 0.28941176470588237, "grad_norm": 0.48537702975667785, "learning_rate": 9.224999999999999e-05, "loss": 0.2129, "step": 615 }, { "epoch": 0.2898823529411765, "grad_norm": 0.6843005690047845, "learning_rate": 9.24e-05, "loss": 0.2832, "step": 616 }, { "epoch": 0.2903529411764706, "grad_norm": 0.27840178010711836, "learning_rate": 9.254999999999999e-05, "loss": 0.0693, "step": 617 }, { "epoch": 0.2908235294117647, "grad_norm": 0.3558505486428398, "learning_rate": 9.269999999999999e-05, "loss": 0.1377, "step": 618 }, { "epoch": 0.2912941176470588, "grad_norm": 0.46768967177689824, "learning_rate": 9.285e-05, "loss": 0.2188, "step": 619 }, { "epoch": 0.2917647058823529, "grad_norm": 0.5241655397072673, "learning_rate": 9.3e-05, "loss": 0.2754, "step": 620 }, { "epoch": 0.29223529411764704, "grad_norm": 0.3517192227471972, "learning_rate": 9.314999999999999e-05, "loss": 0.1592, "step": 621 }, { "epoch": 0.29270588235294115, "grad_norm": 1.5883781520857974, "learning_rate": 9.329999999999999e-05, "loss": 0.3027, "step": 622 }, { "epoch": 0.2931764705882353, "grad_norm": 0.28756549369649115, "learning_rate": 9.345e-05, "loss": 0.0869, "step": 623 }, { "epoch": 0.2936470588235294, "grad_norm": 0.654970331468186, "learning_rate": 9.36e-05, "loss": 0.4531, "step": 624 }, { "epoch": 0.29411764705882354, "grad_norm": 0.3784166057978448, "learning_rate": 9.374999999999999e-05, "loss": 0.1367, "step": 625 }, { "epoch": 0.29458823529411765, "grad_norm": 0.3804131970194466, "learning_rate": 9.389999999999999e-05, "loss": 0.252, "step": 626 }, { "epoch": 0.29505882352941176, "grad_norm": 0.2617103503652112, "learning_rate": 9.405e-05, "loss": 0.0947, "step": 627 }, { "epoch": 0.2955294117647059, "grad_norm": 0.4241841621365012, "learning_rate": 9.419999999999999e-05, "loss": 0.2393, "step": 628 }, { "epoch": 0.296, "grad_norm": 0.5744508201633514, "learning_rate": 9.434999999999999e-05, "loss": 0.2314, "step": 629 }, { "epoch": 0.2964705882352941, "grad_norm": 0.4089714232705533, "learning_rate": 9.449999999999999e-05, "loss": 0.209, "step": 630 }, { "epoch": 0.2969411764705882, "grad_norm": 0.5753366938929578, "learning_rate": 9.465e-05, "loss": 0.3027, "step": 631 }, { "epoch": 0.2974117647058824, "grad_norm": 0.6903276214809534, "learning_rate": 9.479999999999999e-05, "loss": 0.249, "step": 632 }, { "epoch": 0.2978823529411765, "grad_norm": 0.18687233730869354, "learning_rate": 9.494999999999999e-05, "loss": 0.0332, "step": 633 }, { "epoch": 0.2983529411764706, "grad_norm": 0.3342936685722516, "learning_rate": 9.51e-05, "loss": 0.1108, "step": 634 }, { "epoch": 0.2988235294117647, "grad_norm": 0.25673635336057166, "learning_rate": 9.525e-05, "loss": 0.126, "step": 635 }, { "epoch": 0.2992941176470588, "grad_norm": 0.4798663867608047, "learning_rate": 9.539999999999999e-05, "loss": 0.2002, "step": 636 }, { "epoch": 0.29976470588235293, "grad_norm": 0.9587545975688218, "learning_rate": 9.554999999999999e-05, "loss": 0.6211, "step": 637 }, { "epoch": 0.30023529411764704, "grad_norm": 0.33834616489556013, "learning_rate": 9.57e-05, "loss": 0.1514, "step": 638 }, { "epoch": 0.30070588235294116, "grad_norm": 1.6974426278425727, "learning_rate": 9.585e-05, "loss": 0.4629, "step": 639 }, { "epoch": 0.30117647058823527, "grad_norm": 0.3032635652821546, "learning_rate": 9.599999999999999e-05, "loss": 0.1152, "step": 640 }, { "epoch": 0.30164705882352943, "grad_norm": 1.4650129395090468, "learning_rate": 9.614999999999999e-05, "loss": 0.4727, "step": 641 }, { "epoch": 0.30211764705882355, "grad_norm": 0.5007027483998651, "learning_rate": 9.63e-05, "loss": 0.2871, "step": 642 }, { "epoch": 0.30258823529411766, "grad_norm": 0.3074075330336198, "learning_rate": 9.645e-05, "loss": 0.1436, "step": 643 }, { "epoch": 0.30305882352941177, "grad_norm": 0.5382694541191785, "learning_rate": 9.659999999999999e-05, "loss": 0.1787, "step": 644 }, { "epoch": 0.3035294117647059, "grad_norm": 1.2405420455815717, "learning_rate": 9.675e-05, "loss": 0.6016, "step": 645 }, { "epoch": 0.304, "grad_norm": 0.29757712260760655, "learning_rate": 9.69e-05, "loss": 0.1123, "step": 646 }, { "epoch": 0.3044705882352941, "grad_norm": 0.3647424268480989, "learning_rate": 9.705e-05, "loss": 0.1094, "step": 647 }, { "epoch": 0.3049411764705882, "grad_norm": 1.2855288129622389, "learning_rate": 9.719999999999999e-05, "loss": 0.4902, "step": 648 }, { "epoch": 0.3054117647058823, "grad_norm": 0.49991856469854623, "learning_rate": 9.735e-05, "loss": 0.2471, "step": 649 }, { "epoch": 0.3058823529411765, "grad_norm": 0.5060315848393442, "learning_rate": 9.75e-05, "loss": 0.2305, "step": 650 }, { "epoch": 0.3063529411764706, "grad_norm": 0.39569333672560236, "learning_rate": 9.764999999999999e-05, "loss": 0.1279, "step": 651 }, { "epoch": 0.3068235294117647, "grad_norm": 0.5560834580024201, "learning_rate": 9.779999999999999e-05, "loss": 0.2002, "step": 652 }, { "epoch": 0.30729411764705883, "grad_norm": 0.321634602438879, "learning_rate": 9.795e-05, "loss": 0.1069, "step": 653 }, { "epoch": 0.30776470588235294, "grad_norm": 0.3217851194309024, "learning_rate": 9.81e-05, "loss": 0.1191, "step": 654 }, { "epoch": 0.30823529411764705, "grad_norm": 0.8530874456102066, "learning_rate": 9.824999999999999e-05, "loss": 0.3555, "step": 655 }, { "epoch": 0.30870588235294116, "grad_norm": 0.32097319553586584, "learning_rate": 9.839999999999999e-05, "loss": 0.1289, "step": 656 }, { "epoch": 0.3091764705882353, "grad_norm": 0.613084483819731, "learning_rate": 9.855e-05, "loss": 0.3242, "step": 657 }, { "epoch": 0.3096470588235294, "grad_norm": 0.6998296720018863, "learning_rate": 9.87e-05, "loss": 0.1562, "step": 658 }, { "epoch": 0.31011764705882355, "grad_norm": 0.393779844733004, "learning_rate": 9.884999999999999e-05, "loss": 0.168, "step": 659 }, { "epoch": 0.31058823529411766, "grad_norm": 0.16407635570805554, "learning_rate": 9.9e-05, "loss": 0.0245, "step": 660 }, { "epoch": 0.3110588235294118, "grad_norm": 0.7839514768080947, "learning_rate": 9.915e-05, "loss": 0.418, "step": 661 }, { "epoch": 0.3115294117647059, "grad_norm": 0.3809999606784037, "learning_rate": 9.93e-05, "loss": 0.1289, "step": 662 }, { "epoch": 0.312, "grad_norm": 0.17870545789822523, "learning_rate": 9.944999999999999e-05, "loss": 0.042, "step": 663 }, { "epoch": 0.3124705882352941, "grad_norm": 0.39064063866338505, "learning_rate": 9.96e-05, "loss": 0.1963, "step": 664 }, { "epoch": 0.3129411764705882, "grad_norm": 0.2614773357440662, "learning_rate": 9.975e-05, "loss": 0.0605, "step": 665 }, { "epoch": 0.31341176470588233, "grad_norm": 0.6426607171522486, "learning_rate": 9.99e-05, "loss": 0.3145, "step": 666 }, { "epoch": 0.31388235294117645, "grad_norm": 0.3975308353478714, "learning_rate": 0.00010004999999999999, "loss": 0.1709, "step": 667 }, { "epoch": 0.3143529411764706, "grad_norm": 0.9253632831908153, "learning_rate": 0.0001002, "loss": 0.4062, "step": 668 }, { "epoch": 0.3148235294117647, "grad_norm": 0.4997746414923037, "learning_rate": 0.00010035, "loss": 0.1079, "step": 669 }, { "epoch": 0.31529411764705884, "grad_norm": 0.40904504403264325, "learning_rate": 0.0001005, "loss": 0.1104, "step": 670 }, { "epoch": 0.31576470588235295, "grad_norm": 0.6532967266168576, "learning_rate": 0.00010065, "loss": 0.25, "step": 671 }, { "epoch": 0.31623529411764706, "grad_norm": 0.6986705912163381, "learning_rate": 0.0001008, "loss": 0.377, "step": 672 }, { "epoch": 0.31670588235294117, "grad_norm": 0.616730430007992, "learning_rate": 0.00010095, "loss": 0.3301, "step": 673 }, { "epoch": 0.3171764705882353, "grad_norm": 0.7631606898501914, "learning_rate": 0.0001011, "loss": 0.4629, "step": 674 }, { "epoch": 0.3176470588235294, "grad_norm": 0.8272344948133583, "learning_rate": 0.00010125, "loss": 0.2578, "step": 675 }, { "epoch": 0.3181176470588235, "grad_norm": 0.21113227241376983, "learning_rate": 0.0001014, "loss": 0.0466, "step": 676 }, { "epoch": 0.31858823529411767, "grad_norm": 0.7496373150486257, "learning_rate": 0.00010155, "loss": 0.3086, "step": 677 }, { "epoch": 0.3190588235294118, "grad_norm": 1.162433799688981, "learning_rate": 0.00010169999999999999, "loss": 0.5859, "step": 678 }, { "epoch": 0.3195294117647059, "grad_norm": 0.4235712697731388, "learning_rate": 0.00010185, "loss": 0.2539, "step": 679 }, { "epoch": 0.32, "grad_norm": 0.5115208376843903, "learning_rate": 0.000102, "loss": 0.3477, "step": 680 }, { "epoch": 0.3204705882352941, "grad_norm": 0.6785297732960747, "learning_rate": 0.00010215, "loss": 0.2695, "step": 681 }, { "epoch": 0.32094117647058823, "grad_norm": 0.5287180565868146, "learning_rate": 0.00010229999999999999, "loss": 0.2832, "step": 682 }, { "epoch": 0.32141176470588234, "grad_norm": 0.9055580539402787, "learning_rate": 0.00010245, "loss": 0.2734, "step": 683 }, { "epoch": 0.32188235294117645, "grad_norm": 0.5076416649674839, "learning_rate": 0.0001026, "loss": 0.2178, "step": 684 }, { "epoch": 0.32235294117647056, "grad_norm": 0.21525158180940768, "learning_rate": 0.00010275, "loss": 0.0403, "step": 685 }, { "epoch": 0.32282352941176473, "grad_norm": 0.2046358775917218, "learning_rate": 0.0001029, "loss": 0.053, "step": 686 }, { "epoch": 0.32329411764705884, "grad_norm": 0.6018279035432887, "learning_rate": 0.00010305, "loss": 0.2422, "step": 687 }, { "epoch": 0.32376470588235295, "grad_norm": 0.37674580858508344, "learning_rate": 0.00010319999999999999, "loss": 0.167, "step": 688 }, { "epoch": 0.32423529411764707, "grad_norm": 0.6803378880617801, "learning_rate": 0.00010334999999999998, "loss": 0.3105, "step": 689 }, { "epoch": 0.3247058823529412, "grad_norm": 0.6653581152028256, "learning_rate": 0.00010349999999999998, "loss": 0.373, "step": 690 }, { "epoch": 0.3251764705882353, "grad_norm": 0.4807947686896532, "learning_rate": 0.00010364999999999999, "loss": 0.2275, "step": 691 }, { "epoch": 0.3256470588235294, "grad_norm": 4.476288615900905, "learning_rate": 0.00010379999999999999, "loss": 0.373, "step": 692 }, { "epoch": 0.3261176470588235, "grad_norm": 3.4460093603578357, "learning_rate": 0.00010394999999999998, "loss": 0.2432, "step": 693 }, { "epoch": 0.3265882352941176, "grad_norm": 2.510212014656726, "learning_rate": 0.00010409999999999998, "loss": 0.4023, "step": 694 }, { "epoch": 0.3270588235294118, "grad_norm": 0.7439915418254579, "learning_rate": 0.00010424999999999999, "loss": 0.1992, "step": 695 }, { "epoch": 0.3275294117647059, "grad_norm": 0.6886238510528011, "learning_rate": 0.00010439999999999999, "loss": 0.1553, "step": 696 }, { "epoch": 0.328, "grad_norm": 1.4596933301936774, "learning_rate": 0.00010454999999999998, "loss": 0.6328, "step": 697 }, { "epoch": 0.3284705882352941, "grad_norm": 0.3541867286776129, "learning_rate": 0.00010469999999999998, "loss": 0.1582, "step": 698 }, { "epoch": 0.32894117647058824, "grad_norm": 0.4008239847220212, "learning_rate": 0.00010484999999999999, "loss": 0.1104, "step": 699 }, { "epoch": 0.32941176470588235, "grad_norm": 1.6234652101075382, "learning_rate": 0.00010499999999999999, "loss": 0.5156, "step": 700 }, { "epoch": 0.32988235294117646, "grad_norm": 0.9369396505697087, "learning_rate": 0.00010514999999999998, "loss": 0.4277, "step": 701 }, { "epoch": 0.33035294117647057, "grad_norm": 0.8703415696909208, "learning_rate": 0.00010529999999999998, "loss": 0.3672, "step": 702 }, { "epoch": 0.3308235294117647, "grad_norm": 11.395236399602009, "learning_rate": 0.00010544999999999999, "loss": 4.1562, "step": 703 }, { "epoch": 0.33129411764705885, "grad_norm": 62.81841278746279, "learning_rate": 0.00010559999999999998, "loss": 23.625, "step": 704 }, { "epoch": 0.33176470588235296, "grad_norm": 0.4994999498165253, "learning_rate": 0.00010574999999999998, "loss": 0.106, "step": 705 }, { "epoch": 0.3322352941176471, "grad_norm": 23.443408510237916, "learning_rate": 0.00010589999999999999, "loss": 1.9141, "step": 706 }, { "epoch": 0.3327058823529412, "grad_norm": 2.069182250074985, "learning_rate": 0.00010604999999999999, "loss": 0.2773, "step": 707 }, { "epoch": 0.3331764705882353, "grad_norm": 18.46327004067902, "learning_rate": 0.00010619999999999998, "loss": 2.0938, "step": 708 }, { "epoch": 0.3336470588235294, "grad_norm": 8.757400385718602, "learning_rate": 0.00010634999999999998, "loss": 0.9648, "step": 709 }, { "epoch": 0.3341176470588235, "grad_norm": 2.535620403133512, "learning_rate": 0.00010649999999999999, "loss": 0.8594, "step": 710 }, { "epoch": 0.33458823529411763, "grad_norm": 2.5037759197223712, "learning_rate": 0.00010664999999999999, "loss": 0.2891, "step": 711 }, { "epoch": 0.33505882352941174, "grad_norm": 1.4952973378705774, "learning_rate": 0.00010679999999999998, "loss": 0.1318, "step": 712 }, { "epoch": 0.3355294117647059, "grad_norm": 2.667023890760785, "learning_rate": 0.00010694999999999998, "loss": 0.4492, "step": 713 }, { "epoch": 0.336, "grad_norm": 0.14294853971951924, "learning_rate": 0.00010709999999999999, "loss": 0.0596, "step": 714 }, { "epoch": 0.33647058823529413, "grad_norm": 35.85595544750816, "learning_rate": 0.00010724999999999999, "loss": 0.8828, "step": 715 }, { "epoch": 0.33694117647058824, "grad_norm": 11.238010263149732, "learning_rate": 0.00010739999999999998, "loss": 3.1562, "step": 716 }, { "epoch": 0.33741176470588236, "grad_norm": 32.635600449142416, "learning_rate": 0.00010754999999999999, "loss": 1.1016, "step": 717 }, { "epoch": 0.33788235294117647, "grad_norm": 2.302824370722843, "learning_rate": 0.00010769999999999999, "loss": 1.2266, "step": 718 }, { "epoch": 0.3383529411764706, "grad_norm": 4.644358013526273, "learning_rate": 0.00010784999999999999, "loss": 0.7734, "step": 719 }, { "epoch": 0.3388235294117647, "grad_norm": 6.639081237080364, "learning_rate": 0.00010799999999999998, "loss": 0.6289, "step": 720 }, { "epoch": 0.3392941176470588, "grad_norm": 0.21377398540636852, "learning_rate": 0.00010814999999999999, "loss": 0.103, "step": 721 }, { "epoch": 0.33976470588235297, "grad_norm": 0.7750891805148357, "learning_rate": 0.00010829999999999999, "loss": 0.5117, "step": 722 }, { "epoch": 0.3402352941176471, "grad_norm": 0.6341446714589778, "learning_rate": 0.00010845, "loss": 0.377, "step": 723 }, { "epoch": 0.3407058823529412, "grad_norm": 1.252287579998299, "learning_rate": 0.00010859999999999998, "loss": 0.8164, "step": 724 }, { "epoch": 0.3411764705882353, "grad_norm": 0.6042033245211691, "learning_rate": 0.00010874999999999999, "loss": 0.3281, "step": 725 }, { "epoch": 0.3416470588235294, "grad_norm": 2.398572364386284, "learning_rate": 0.00010889999999999999, "loss": 1.0547, "step": 726 }, { "epoch": 0.3421176470588235, "grad_norm": 1.6087766310850853, "learning_rate": 0.00010904999999999998, "loss": 0.4141, "step": 727 }, { "epoch": 0.34258823529411764, "grad_norm": 45.469245753892906, "learning_rate": 0.00010919999999999998, "loss": 5.5938, "step": 728 }, { "epoch": 0.34305882352941175, "grad_norm": 35.46281469409353, "learning_rate": 0.00010934999999999999, "loss": 5.5938, "step": 729 }, { "epoch": 0.34352941176470586, "grad_norm": 12.41339906269861, "learning_rate": 0.00010949999999999999, "loss": 1.2031, "step": 730 }, { "epoch": 0.344, "grad_norm": 4.57476780340973, "learning_rate": 0.00010964999999999998, "loss": 1.3516, "step": 731 }, { "epoch": 0.34447058823529414, "grad_norm": 3.422562034874079, "learning_rate": 0.00010979999999999999, "loss": 1.1172, "step": 732 }, { "epoch": 0.34494117647058825, "grad_norm": 0.9779367232169753, "learning_rate": 0.00010994999999999999, "loss": 0.6172, "step": 733 }, { "epoch": 0.34541176470588236, "grad_norm": 3.7057756063783094, "learning_rate": 0.00011009999999999999, "loss": 0.5352, "step": 734 }, { "epoch": 0.3458823529411765, "grad_norm": 11.721459315339997, "learning_rate": 0.00011024999999999998, "loss": 0.9805, "step": 735 }, { "epoch": 0.3463529411764706, "grad_norm": 2.261752041645444, "learning_rate": 0.00011039999999999999, "loss": 0.7148, "step": 736 }, { "epoch": 0.3468235294117647, "grad_norm": 0.6293546756282684, "learning_rate": 0.00011054999999999999, "loss": 0.2451, "step": 737 }, { "epoch": 0.3472941176470588, "grad_norm": 10.040747544238222, "learning_rate": 0.0001107, "loss": 2.7188, "step": 738 }, { "epoch": 0.3477647058823529, "grad_norm": 1.6132181507444872, "learning_rate": 0.00011084999999999998, "loss": 1.4766, "step": 739 }, { "epoch": 0.34823529411764703, "grad_norm": 2.7666506512748374, "learning_rate": 0.00011099999999999999, "loss": 0.8516, "step": 740 }, { "epoch": 0.3487058823529412, "grad_norm": 1.5584106195596537, "learning_rate": 0.00011114999999999999, "loss": 0.4902, "step": 741 }, { "epoch": 0.3491764705882353, "grad_norm": 6.118610410950182, "learning_rate": 0.0001113, "loss": 0.6445, "step": 742 }, { "epoch": 0.3496470588235294, "grad_norm": 1.5634385972397047, "learning_rate": 0.00011144999999999998, "loss": 0.5586, "step": 743 }, { "epoch": 0.35011764705882353, "grad_norm": 1.952412877768451, "learning_rate": 0.00011159999999999999, "loss": 0.9414, "step": 744 }, { "epoch": 0.35058823529411764, "grad_norm": 5.804978877105826, "learning_rate": 0.00011174999999999999, "loss": 2.6719, "step": 745 }, { "epoch": 0.35105882352941176, "grad_norm": 2.3917710929360685, "learning_rate": 0.0001119, "loss": 0.75, "step": 746 }, { "epoch": 0.35152941176470587, "grad_norm": 1.5637425015298705, "learning_rate": 0.00011204999999999999, "loss": 0.332, "step": 747 }, { "epoch": 0.352, "grad_norm": 0.22952527134237738, "learning_rate": 0.00011219999999999999, "loss": 0.0679, "step": 748 }, { "epoch": 0.3524705882352941, "grad_norm": 5.554775358580264, "learning_rate": 0.00011235, "loss": 1.3906, "step": 749 }, { "epoch": 0.35294117647058826, "grad_norm": 1.837757108914798, "learning_rate": 0.0001125, "loss": 0.5898, "step": 750 }, { "epoch": 0.35341176470588237, "grad_norm": 2.076841441802926, "learning_rate": 0.00011264999999999999, "loss": 0.6758, "step": 751 }, { "epoch": 0.3538823529411765, "grad_norm": 0.5502959154627333, "learning_rate": 0.00011279999999999999, "loss": 0.1504, "step": 752 }, { "epoch": 0.3543529411764706, "grad_norm": 2.2394040662103842, "learning_rate": 0.00011295, "loss": 0.6133, "step": 753 }, { "epoch": 0.3548235294117647, "grad_norm": 0.3302409657280562, "learning_rate": 0.00011309999999999998, "loss": 0.1787, "step": 754 }, { "epoch": 0.3552941176470588, "grad_norm": 0.5136217032709538, "learning_rate": 0.00011324999999999999, "loss": 0.3203, "step": 755 }, { "epoch": 0.3557647058823529, "grad_norm": 0.9870232118520796, "learning_rate": 0.00011339999999999999, "loss": 0.4375, "step": 756 }, { "epoch": 0.35623529411764704, "grad_norm": 1.5451032563921072, "learning_rate": 0.00011355, "loss": 0.5117, "step": 757 }, { "epoch": 0.35670588235294115, "grad_norm": 1.044860252870971, "learning_rate": 0.00011369999999999999, "loss": 0.3379, "step": 758 }, { "epoch": 0.3571764705882353, "grad_norm": 0.47814680324663894, "learning_rate": 0.00011384999999999999, "loss": 0.1396, "step": 759 }, { "epoch": 0.35764705882352943, "grad_norm": 0.6112011771771941, "learning_rate": 0.00011399999999999999, "loss": 0.1953, "step": 760 }, { "epoch": 0.35811764705882354, "grad_norm": 0.9754867533568051, "learning_rate": 0.00011415, "loss": 0.4297, "step": 761 }, { "epoch": 0.35858823529411765, "grad_norm": 0.8907322967109768, "learning_rate": 0.00011429999999999999, "loss": 0.4375, "step": 762 }, { "epoch": 0.35905882352941176, "grad_norm": 0.7735770351325458, "learning_rate": 0.00011444999999999999, "loss": 0.4277, "step": 763 }, { "epoch": 0.3595294117647059, "grad_norm": 0.39900647628912705, "learning_rate": 0.0001146, "loss": 0.1289, "step": 764 }, { "epoch": 0.36, "grad_norm": 0.7542251132231567, "learning_rate": 0.00011475, "loss": 0.3594, "step": 765 }, { "epoch": 0.3604705882352941, "grad_norm": 2.113811768397646, "learning_rate": 0.00011489999999999999, "loss": 0.5469, "step": 766 }, { "epoch": 0.3609411764705882, "grad_norm": 0.6645829904993801, "learning_rate": 0.00011504999999999999, "loss": 0.5508, "step": 767 }, { "epoch": 0.3614117647058824, "grad_norm": 0.45592483670748174, "learning_rate": 0.0001152, "loss": 0.2598, "step": 768 }, { "epoch": 0.3618823529411765, "grad_norm": 1.566738336971126, "learning_rate": 0.00011535, "loss": 0.7344, "step": 769 }, { "epoch": 0.3623529411764706, "grad_norm": 1.0946767388718226, "learning_rate": 0.00011549999999999999, "loss": 0.707, "step": 770 }, { "epoch": 0.3628235294117647, "grad_norm": 0.8635207511659837, "learning_rate": 0.00011564999999999999, "loss": 0.3184, "step": 771 }, { "epoch": 0.3632941176470588, "grad_norm": 1.389396467082158, "learning_rate": 0.0001158, "loss": 0.4375, "step": 772 }, { "epoch": 0.36376470588235293, "grad_norm": 1.3839451208232412, "learning_rate": 0.00011595, "loss": 0.5273, "step": 773 }, { "epoch": 0.36423529411764705, "grad_norm": 2.148033789867013, "learning_rate": 0.00011609999999999999, "loss": 0.9805, "step": 774 }, { "epoch": 0.36470588235294116, "grad_norm": 1.3783209931877305, "learning_rate": 0.00011624999999999999, "loss": 0.8477, "step": 775 }, { "epoch": 0.36517647058823527, "grad_norm": 0.6839169951104881, "learning_rate": 0.0001164, "loss": 0.2812, "step": 776 }, { "epoch": 0.36564705882352944, "grad_norm": 1.3288499246242353, "learning_rate": 0.00011654999999999999, "loss": 0.4473, "step": 777 }, { "epoch": 0.36611764705882355, "grad_norm": 1.4947806651113587, "learning_rate": 0.00011669999999999999, "loss": 0.418, "step": 778 }, { "epoch": 0.36658823529411766, "grad_norm": 0.7511946599247681, "learning_rate": 0.00011685, "loss": 0.3066, "step": 779 }, { "epoch": 0.36705882352941177, "grad_norm": 0.8704286596039543, "learning_rate": 0.000117, "loss": 0.6094, "step": 780 }, { "epoch": 0.3675294117647059, "grad_norm": 0.34264781298285163, "learning_rate": 0.00011714999999999999, "loss": 0.2598, "step": 781 }, { "epoch": 0.368, "grad_norm": 0.12457098282907672, "learning_rate": 0.00011729999999999999, "loss": 0.0654, "step": 782 }, { "epoch": 0.3684705882352941, "grad_norm": 1.0710171650265568, "learning_rate": 0.00011745, "loss": 0.3965, "step": 783 }, { "epoch": 0.3689411764705882, "grad_norm": 0.6489455421712819, "learning_rate": 0.0001176, "loss": 0.2754, "step": 784 }, { "epoch": 0.36941176470588233, "grad_norm": 0.40210311908925245, "learning_rate": 0.00011774999999999999, "loss": 0.1235, "step": 785 }, { "epoch": 0.3698823529411765, "grad_norm": 0.4587710694396334, "learning_rate": 0.00011789999999999999, "loss": 0.1533, "step": 786 }, { "epoch": 0.3703529411764706, "grad_norm": 0.676950800402359, "learning_rate": 0.00011805, "loss": 0.3965, "step": 787 }, { "epoch": 0.3708235294117647, "grad_norm": 1.0741430459060748, "learning_rate": 0.0001182, "loss": 0.9609, "step": 788 }, { "epoch": 0.37129411764705883, "grad_norm": 1.4261045955341234, "learning_rate": 0.00011834999999999999, "loss": 0.793, "step": 789 }, { "epoch": 0.37176470588235294, "grad_norm": 0.21927272998291186, "learning_rate": 0.0001185, "loss": 0.0972, "step": 790 }, { "epoch": 0.37223529411764705, "grad_norm": 1.802102865073379, "learning_rate": 0.00011865, "loss": 0.5039, "step": 791 }, { "epoch": 0.37270588235294116, "grad_norm": 1.2999417948678365, "learning_rate": 0.0001188, "loss": 0.5703, "step": 792 }, { "epoch": 0.3731764705882353, "grad_norm": 0.6485062909077706, "learning_rate": 0.00011894999999999999, "loss": 0.4902, "step": 793 }, { "epoch": 0.3736470588235294, "grad_norm": 0.41163049533696083, "learning_rate": 0.0001191, "loss": 0.3477, "step": 794 }, { "epoch": 0.37411764705882355, "grad_norm": 0.9604990009342059, "learning_rate": 0.00011925, "loss": 0.5117, "step": 795 }, { "epoch": 0.37458823529411767, "grad_norm": 0.7330368191272117, "learning_rate": 0.0001194, "loss": 0.543, "step": 796 }, { "epoch": 0.3750588235294118, "grad_norm": 0.33572463001707703, "learning_rate": 0.00011954999999999999, "loss": 0.2393, "step": 797 }, { "epoch": 0.3755294117647059, "grad_norm": 0.12640884264548793, "learning_rate": 0.0001197, "loss": 0.0854, "step": 798 }, { "epoch": 0.376, "grad_norm": 0.8645980682938456, "learning_rate": 0.00011985, "loss": 0.6133, "step": 799 }, { "epoch": 0.3764705882352941, "grad_norm": 0.7102713001059413, "learning_rate": 0.00011999999999999999, "loss": 0.8047, "step": 800 }, { "epoch": 0.3769411764705882, "grad_norm": 0.2735206376460154, "learning_rate": 0.00012014999999999999, "loss": 0.1147, "step": 801 }, { "epoch": 0.37741176470588234, "grad_norm": 0.2559733486195856, "learning_rate": 0.0001203, "loss": 0.1445, "step": 802 }, { "epoch": 0.37788235294117645, "grad_norm": 0.21207722485348895, "learning_rate": 0.00012045, "loss": 0.1172, "step": 803 }, { "epoch": 0.3783529411764706, "grad_norm": 0.441293776133396, "learning_rate": 0.00012059999999999999, "loss": 0.252, "step": 804 }, { "epoch": 0.3788235294117647, "grad_norm": 0.5100914814209183, "learning_rate": 0.00012075, "loss": 0.2695, "step": 805 }, { "epoch": 0.37929411764705884, "grad_norm": 0.5135308801524721, "learning_rate": 0.0001209, "loss": 0.4961, "step": 806 }, { "epoch": 0.37976470588235295, "grad_norm": 0.24876193645104863, "learning_rate": 0.00012105, "loss": 0.1021, "step": 807 }, { "epoch": 0.38023529411764706, "grad_norm": 1.1859051862128176, "learning_rate": 0.00012119999999999999, "loss": 0.543, "step": 808 }, { "epoch": 0.38070588235294117, "grad_norm": 0.6587010330770888, "learning_rate": 0.00012135, "loss": 0.4082, "step": 809 }, { "epoch": 0.3811764705882353, "grad_norm": 0.23326979247197133, "learning_rate": 0.0001215, "loss": 0.1445, "step": 810 }, { "epoch": 0.3816470588235294, "grad_norm": 0.6248470362527676, "learning_rate": 0.00012165, "loss": 0.3594, "step": 811 }, { "epoch": 0.3821176470588235, "grad_norm": 0.3055763289148157, "learning_rate": 0.00012179999999999999, "loss": 0.1191, "step": 812 }, { "epoch": 0.3825882352941177, "grad_norm": 0.19038763998206623, "learning_rate": 0.00012194999999999998, "loss": 0.0593, "step": 813 }, { "epoch": 0.3830588235294118, "grad_norm": 0.38567210744847946, "learning_rate": 0.00012209999999999999, "loss": 0.2236, "step": 814 }, { "epoch": 0.3835294117647059, "grad_norm": 0.4689322477380041, "learning_rate": 0.00012225, "loss": 0.1934, "step": 815 }, { "epoch": 0.384, "grad_norm": 0.17441278287223216, "learning_rate": 0.0001224, "loss": 0.0713, "step": 816 }, { "epoch": 0.3844705882352941, "grad_norm": 0.9207882182309842, "learning_rate": 0.00012254999999999997, "loss": 0.3398, "step": 817 }, { "epoch": 0.38494117647058823, "grad_norm": 0.29114036268202326, "learning_rate": 0.00012269999999999997, "loss": 0.1719, "step": 818 }, { "epoch": 0.38541176470588234, "grad_norm": 0.4268931764456148, "learning_rate": 0.00012284999999999998, "loss": 0.2314, "step": 819 }, { "epoch": 0.38588235294117645, "grad_norm": 0.5835604534947156, "learning_rate": 0.00012299999999999998, "loss": 0.3066, "step": 820 }, { "epoch": 0.38635294117647057, "grad_norm": 0.7165240968508761, "learning_rate": 0.00012314999999999998, "loss": 0.5625, "step": 821 }, { "epoch": 0.38682352941176473, "grad_norm": 0.44434752765798036, "learning_rate": 0.0001233, "loss": 0.3555, "step": 822 }, { "epoch": 0.38729411764705884, "grad_norm": 0.04493024583835492, "learning_rate": 0.00012345, "loss": 0.0143, "step": 823 }, { "epoch": 0.38776470588235296, "grad_norm": 0.5824412089769833, "learning_rate": 0.0001236, "loss": 0.4062, "step": 824 }, { "epoch": 0.38823529411764707, "grad_norm": 0.7273785099911753, "learning_rate": 0.00012374999999999997, "loss": 0.3867, "step": 825 }, { "epoch": 0.3887058823529412, "grad_norm": 0.3892526418788562, "learning_rate": 0.00012389999999999998, "loss": 0.2021, "step": 826 }, { "epoch": 0.3891764705882353, "grad_norm": 0.6632704549744459, "learning_rate": 0.00012404999999999998, "loss": 0.3984, "step": 827 }, { "epoch": 0.3896470588235294, "grad_norm": 0.6646774687676704, "learning_rate": 0.00012419999999999998, "loss": 0.4629, "step": 828 }, { "epoch": 0.3901176470588235, "grad_norm": 0.8058621861709193, "learning_rate": 0.00012435, "loss": 0.3711, "step": 829 }, { "epoch": 0.3905882352941176, "grad_norm": 0.6403817498369757, "learning_rate": 0.0001245, "loss": 0.5117, "step": 830 }, { "epoch": 0.3910588235294118, "grad_norm": 0.3344074867723219, "learning_rate": 0.00012465, "loss": 0.2539, "step": 831 }, { "epoch": 0.3915294117647059, "grad_norm": 0.7523896348789751, "learning_rate": 0.00012479999999999997, "loss": 0.4082, "step": 832 }, { "epoch": 0.392, "grad_norm": 0.6594190626613722, "learning_rate": 0.00012494999999999997, "loss": 0.4551, "step": 833 }, { "epoch": 0.3924705882352941, "grad_norm": 0.4760180044809535, "learning_rate": 0.00012509999999999998, "loss": 0.4395, "step": 834 }, { "epoch": 0.39294117647058824, "grad_norm": 0.4743355862088893, "learning_rate": 0.00012524999999999998, "loss": 0.2773, "step": 835 }, { "epoch": 0.39341176470588235, "grad_norm": 0.5155164856102634, "learning_rate": 0.00012539999999999999, "loss": 0.2139, "step": 836 }, { "epoch": 0.39388235294117646, "grad_norm": 0.26079070326822607, "learning_rate": 0.00012555, "loss": 0.0947, "step": 837 }, { "epoch": 0.3943529411764706, "grad_norm": 0.5781319309990642, "learning_rate": 0.0001257, "loss": 0.4453, "step": 838 }, { "epoch": 0.3948235294117647, "grad_norm": 1.0071583300515972, "learning_rate": 0.00012585, "loss": 0.8633, "step": 839 }, { "epoch": 0.3952941176470588, "grad_norm": 0.9252205962454083, "learning_rate": 0.00012599999999999997, "loss": 0.7812, "step": 840 }, { "epoch": 0.39576470588235296, "grad_norm": 0.26417216552935496, "learning_rate": 0.00012614999999999998, "loss": 0.1328, "step": 841 }, { "epoch": 0.3962352941176471, "grad_norm": 0.9334877917026346, "learning_rate": 0.00012629999999999998, "loss": 0.5742, "step": 842 }, { "epoch": 0.3967058823529412, "grad_norm": 0.5561380116689559, "learning_rate": 0.00012644999999999998, "loss": 0.3359, "step": 843 }, { "epoch": 0.3971764705882353, "grad_norm": 0.44867589108072825, "learning_rate": 0.0001266, "loss": 0.1973, "step": 844 }, { "epoch": 0.3976470588235294, "grad_norm": 0.7413356661074205, "learning_rate": 0.00012675, "loss": 0.6523, "step": 845 }, { "epoch": 0.3981176470588235, "grad_norm": 0.17194946928554697, "learning_rate": 0.0001269, "loss": 0.0688, "step": 846 }, { "epoch": 0.39858823529411763, "grad_norm": 0.4370001973650655, "learning_rate": 0.00012705, "loss": 0.2256, "step": 847 }, { "epoch": 0.39905882352941174, "grad_norm": 0.1197139844989413, "learning_rate": 0.00012719999999999997, "loss": 0.0352, "step": 848 }, { "epoch": 0.39952941176470586, "grad_norm": 0.43161167130919437, "learning_rate": 0.00012734999999999998, "loss": 0.1875, "step": 849 }, { "epoch": 0.4, "grad_norm": 1.104536173162915, "learning_rate": 0.00012749999999999998, "loss": 0.5586, "step": 850 }, { "epoch": 0.40047058823529413, "grad_norm": 0.5253924686903438, "learning_rate": 0.00012764999999999999, "loss": 0.2969, "step": 851 }, { "epoch": 0.40094117647058825, "grad_norm": 0.5375445030252899, "learning_rate": 0.0001278, "loss": 0.3008, "step": 852 }, { "epoch": 0.40141176470588236, "grad_norm": 0.07133565776214476, "learning_rate": 0.00012795, "loss": 0.0206, "step": 853 }, { "epoch": 0.40188235294117647, "grad_norm": 1.1852526997860668, "learning_rate": 0.0001281, "loss": 0.418, "step": 854 }, { "epoch": 0.4023529411764706, "grad_norm": 0.591747091914598, "learning_rate": 0.00012824999999999997, "loss": 0.3242, "step": 855 }, { "epoch": 0.4028235294117647, "grad_norm": 0.6418454367875942, "learning_rate": 0.00012839999999999998, "loss": 0.332, "step": 856 }, { "epoch": 0.4032941176470588, "grad_norm": 1.1054440272257946, "learning_rate": 0.00012854999999999998, "loss": 0.5195, "step": 857 }, { "epoch": 0.4037647058823529, "grad_norm": 1.0486020497997566, "learning_rate": 0.00012869999999999998, "loss": 0.75, "step": 858 }, { "epoch": 0.4042352941176471, "grad_norm": 0.3441282327492822, "learning_rate": 0.00012885, "loss": 0.2119, "step": 859 }, { "epoch": 0.4047058823529412, "grad_norm": 0.9212646784324253, "learning_rate": 0.000129, "loss": 0.5117, "step": 860 }, { "epoch": 0.4051764705882353, "grad_norm": 1.1908694245139193, "learning_rate": 0.00012915, "loss": 0.4688, "step": 861 }, { "epoch": 0.4056470588235294, "grad_norm": 0.5621380742383486, "learning_rate": 0.0001293, "loss": 0.4238, "step": 862 }, { "epoch": 0.4061176470588235, "grad_norm": 0.4638889445816288, "learning_rate": 0.00012944999999999998, "loss": 0.3125, "step": 863 }, { "epoch": 0.40658823529411764, "grad_norm": 1.101512278632625, "learning_rate": 0.00012959999999999998, "loss": 0.4941, "step": 864 }, { "epoch": 0.40705882352941175, "grad_norm": 0.5173437301117318, "learning_rate": 0.00012974999999999998, "loss": 0.209, "step": 865 }, { "epoch": 0.40752941176470586, "grad_norm": 0.8999273936782953, "learning_rate": 0.00012989999999999999, "loss": 0.5273, "step": 866 }, { "epoch": 0.408, "grad_norm": 0.31955363240370277, "learning_rate": 0.00013005, "loss": 0.1338, "step": 867 }, { "epoch": 0.40847058823529414, "grad_norm": 0.31286825396448537, "learning_rate": 0.0001302, "loss": 0.1387, "step": 868 }, { "epoch": 0.40894117647058825, "grad_norm": 0.7402916869883935, "learning_rate": 0.00013035, "loss": 0.3418, "step": 869 }, { "epoch": 0.40941176470588236, "grad_norm": 1.9002808641640418, "learning_rate": 0.0001305, "loss": 0.6719, "step": 870 }, { "epoch": 0.4098823529411765, "grad_norm": 0.6102542120611387, "learning_rate": 0.00013064999999999998, "loss": 0.5938, "step": 871 }, { "epoch": 0.4103529411764706, "grad_norm": 0.3252429367507606, "learning_rate": 0.00013079999999999998, "loss": 0.1904, "step": 872 }, { "epoch": 0.4108235294117647, "grad_norm": 0.9600501951844836, "learning_rate": 0.00013094999999999998, "loss": 0.416, "step": 873 }, { "epoch": 0.4112941176470588, "grad_norm": 1.5785862485791862, "learning_rate": 0.0001311, "loss": 0.7148, "step": 874 }, { "epoch": 0.4117647058823529, "grad_norm": 0.5640816042807212, "learning_rate": 0.00013125, "loss": 0.3301, "step": 875 }, { "epoch": 0.41223529411764703, "grad_norm": 0.4178420777984672, "learning_rate": 0.0001314, "loss": 0.166, "step": 876 }, { "epoch": 0.4127058823529412, "grad_norm": 0.38372709909856356, "learning_rate": 0.00013155, "loss": 0.1309, "step": 877 }, { "epoch": 0.4131764705882353, "grad_norm": 0.8346541845915121, "learning_rate": 0.00013169999999999998, "loss": 0.2695, "step": 878 }, { "epoch": 0.4136470588235294, "grad_norm": 0.7733267010122818, "learning_rate": 0.00013184999999999998, "loss": 0.3301, "step": 879 }, { "epoch": 0.41411764705882353, "grad_norm": 0.4512562444891831, "learning_rate": 0.00013199999999999998, "loss": 0.1533, "step": 880 }, { "epoch": 0.41458823529411765, "grad_norm": 0.6478752769726613, "learning_rate": 0.00013215, "loss": 0.373, "step": 881 }, { "epoch": 0.41505882352941176, "grad_norm": 0.43241831316881674, "learning_rate": 0.0001323, "loss": 0.2168, "step": 882 }, { "epoch": 0.41552941176470587, "grad_norm": 1.1949161449042693, "learning_rate": 0.00013245, "loss": 0.582, "step": 883 }, { "epoch": 0.416, "grad_norm": 0.2323468958573826, "learning_rate": 0.0001326, "loss": 0.0752, "step": 884 }, { "epoch": 0.4164705882352941, "grad_norm": 0.6475431323145501, "learning_rate": 0.00013275, "loss": 0.4102, "step": 885 }, { "epoch": 0.41694117647058826, "grad_norm": 0.5471458105310245, "learning_rate": 0.00013289999999999998, "loss": 0.25, "step": 886 }, { "epoch": 0.41741176470588237, "grad_norm": 1.0137546093741379, "learning_rate": 0.00013304999999999998, "loss": 0.3789, "step": 887 }, { "epoch": 0.4178823529411765, "grad_norm": 0.8069382419794647, "learning_rate": 0.00013319999999999999, "loss": 0.4297, "step": 888 }, { "epoch": 0.4183529411764706, "grad_norm": 0.49259346663922693, "learning_rate": 0.00013335, "loss": 0.3223, "step": 889 }, { "epoch": 0.4188235294117647, "grad_norm": 0.1487295071996931, "learning_rate": 0.0001335, "loss": 0.0659, "step": 890 }, { "epoch": 0.4192941176470588, "grad_norm": 0.9935373585398379, "learning_rate": 0.00013365, "loss": 0.3887, "step": 891 }, { "epoch": 0.41976470588235293, "grad_norm": 0.20940451565877446, "learning_rate": 0.0001338, "loss": 0.1113, "step": 892 }, { "epoch": 0.42023529411764704, "grad_norm": 0.5345617645753284, "learning_rate": 0.00013395, "loss": 0.2344, "step": 893 }, { "epoch": 0.42070588235294115, "grad_norm": 0.33012767642306057, "learning_rate": 0.00013409999999999998, "loss": 0.209, "step": 894 }, { "epoch": 0.4211764705882353, "grad_norm": 0.5356733460970422, "learning_rate": 0.00013424999999999998, "loss": 0.3516, "step": 895 }, { "epoch": 0.42164705882352943, "grad_norm": 0.43532711303597715, "learning_rate": 0.0001344, "loss": 0.3203, "step": 896 }, { "epoch": 0.42211764705882354, "grad_norm": 0.27263520602137303, "learning_rate": 0.00013455, "loss": 0.167, "step": 897 }, { "epoch": 0.42258823529411765, "grad_norm": 0.48371236010862445, "learning_rate": 0.0001347, "loss": 0.2773, "step": 898 }, { "epoch": 0.42305882352941176, "grad_norm": 1.12327589145164, "learning_rate": 0.00013485, "loss": 0.6641, "step": 899 }, { "epoch": 0.4235294117647059, "grad_norm": 0.3568671131556459, "learning_rate": 0.000135, "loss": 0.2188, "step": 900 }, { "epoch": 0.424, "grad_norm": 0.3223878390694201, "learning_rate": 0.00013514999999999998, "loss": 0.1895, "step": 901 }, { "epoch": 0.4244705882352941, "grad_norm": 0.699152313812253, "learning_rate": 0.00013529999999999998, "loss": 0.2852, "step": 902 }, { "epoch": 0.4249411764705882, "grad_norm": 0.7618164639677238, "learning_rate": 0.00013544999999999999, "loss": 0.3848, "step": 903 }, { "epoch": 0.4254117647058824, "grad_norm": 0.3603658364666382, "learning_rate": 0.0001356, "loss": 0.0869, "step": 904 }, { "epoch": 0.4258823529411765, "grad_norm": 0.1247440327384394, "learning_rate": 0.00013575, "loss": 0.0439, "step": 905 }, { "epoch": 0.4263529411764706, "grad_norm": 1.6972350384297803, "learning_rate": 0.0001359, "loss": 0.668, "step": 906 }, { "epoch": 0.4268235294117647, "grad_norm": 0.4463973650606885, "learning_rate": 0.00013605, "loss": 0.2871, "step": 907 }, { "epoch": 0.4272941176470588, "grad_norm": 0.5915404568397832, "learning_rate": 0.0001362, "loss": 0.3281, "step": 908 }, { "epoch": 0.42776470588235294, "grad_norm": 0.2543179143534107, "learning_rate": 0.00013634999999999998, "loss": 0.1138, "step": 909 }, { "epoch": 0.42823529411764705, "grad_norm": 1.006408654660531, "learning_rate": 0.00013649999999999998, "loss": 0.5312, "step": 910 }, { "epoch": 0.42870588235294116, "grad_norm": 0.7477594525724014, "learning_rate": 0.00013665, "loss": 0.4375, "step": 911 }, { "epoch": 0.42917647058823527, "grad_norm": 0.7720444521751664, "learning_rate": 0.0001368, "loss": 0.4629, "step": 912 }, { "epoch": 0.42964705882352944, "grad_norm": 0.20661106635632423, "learning_rate": 0.00013695, "loss": 0.0491, "step": 913 }, { "epoch": 0.43011764705882355, "grad_norm": 0.7451853417739873, "learning_rate": 0.0001371, "loss": 0.2754, "step": 914 }, { "epoch": 0.43058823529411766, "grad_norm": 0.4673712337286548, "learning_rate": 0.00013725, "loss": 0.1777, "step": 915 }, { "epoch": 0.43105882352941177, "grad_norm": 0.6705853843993065, "learning_rate": 0.0001374, "loss": 0.3965, "step": 916 }, { "epoch": 0.4315294117647059, "grad_norm": 0.3859183660035513, "learning_rate": 0.00013754999999999998, "loss": 0.2305, "step": 917 }, { "epoch": 0.432, "grad_norm": 1.074007115105099, "learning_rate": 0.00013769999999999999, "loss": 0.7578, "step": 918 }, { "epoch": 0.4324705882352941, "grad_norm": 0.47544282001849186, "learning_rate": 0.00013785, "loss": 0.377, "step": 919 }, { "epoch": 0.4329411764705882, "grad_norm": 0.47833680092517594, "learning_rate": 0.000138, "loss": 0.4512, "step": 920 }, { "epoch": 0.43341176470588233, "grad_norm": 0.17127333336204123, "learning_rate": 0.00013815, "loss": 0.0444, "step": 921 }, { "epoch": 0.4338823529411765, "grad_norm": 0.580901694304692, "learning_rate": 0.0001383, "loss": 0.248, "step": 922 }, { "epoch": 0.4343529411764706, "grad_norm": 0.3171081629347101, "learning_rate": 0.00013845, "loss": 0.126, "step": 923 }, { "epoch": 0.4348235294117647, "grad_norm": 0.25590176425309125, "learning_rate": 0.0001386, "loss": 0.125, "step": 924 }, { "epoch": 0.43529411764705883, "grad_norm": 0.6412140701368668, "learning_rate": 0.00013874999999999998, "loss": 0.4062, "step": 925 }, { "epoch": 0.43576470588235294, "grad_norm": 0.5974558467293047, "learning_rate": 0.0001389, "loss": 0.334, "step": 926 }, { "epoch": 0.43623529411764705, "grad_norm": 0.565299352162957, "learning_rate": 0.00013905, "loss": 0.418, "step": 927 }, { "epoch": 0.43670588235294117, "grad_norm": 0.3396658028986845, "learning_rate": 0.0001392, "loss": 0.2656, "step": 928 }, { "epoch": 0.4371764705882353, "grad_norm": 0.5301232640039317, "learning_rate": 0.00013935, "loss": 0.3809, "step": 929 }, { "epoch": 0.4376470588235294, "grad_norm": 0.26774324555190027, "learning_rate": 0.0001395, "loss": 0.1279, "step": 930 }, { "epoch": 0.43811764705882356, "grad_norm": 0.8849661735400011, "learning_rate": 0.00013965, "loss": 0.4844, "step": 931 }, { "epoch": 0.43858823529411767, "grad_norm": 0.4834591055514268, "learning_rate": 0.00013979999999999998, "loss": 0.4805, "step": 932 }, { "epoch": 0.4390588235294118, "grad_norm": 0.6914648810818953, "learning_rate": 0.00013995, "loss": 0.5195, "step": 933 }, { "epoch": 0.4395294117647059, "grad_norm": 0.309089730374881, "learning_rate": 0.0001401, "loss": 0.3125, "step": 934 }, { "epoch": 0.44, "grad_norm": 0.5107850603362507, "learning_rate": 0.00014025, "loss": 0.3574, "step": 935 }, { "epoch": 0.4404705882352941, "grad_norm": 0.5400177153911806, "learning_rate": 0.0001404, "loss": 0.5195, "step": 936 }, { "epoch": 0.4409411764705882, "grad_norm": 130.17727964061726, "learning_rate": 0.00014055, "loss": 0.9961, "step": 937 }, { "epoch": 0.44141176470588234, "grad_norm": 11.149901284994415, "learning_rate": 0.00014069999999999998, "loss": 1.4531, "step": 938 }, { "epoch": 0.44188235294117645, "grad_norm": 7.62040367677822, "learning_rate": 0.00014084999999999998, "loss": 0.9961, "step": 939 }, { "epoch": 0.4423529411764706, "grad_norm": 2.153478140913247, "learning_rate": 0.00014099999999999998, "loss": 0.1514, "step": 940 }, { "epoch": 0.4428235294117647, "grad_norm": 0.6710066539532712, "learning_rate": 0.00014115, "loss": 0.3457, "step": 941 }, { "epoch": 0.44329411764705884, "grad_norm": 0.5634310041243139, "learning_rate": 0.0001413, "loss": 0.2207, "step": 942 }, { "epoch": 0.44376470588235295, "grad_norm": 0.3749494521841048, "learning_rate": 0.00014144999999999997, "loss": 0.1318, "step": 943 }, { "epoch": 0.44423529411764706, "grad_norm": 2.9309350272984855, "learning_rate": 0.00014159999999999997, "loss": 0.5547, "step": 944 }, { "epoch": 0.4447058823529412, "grad_norm": 0.8103991013889708, "learning_rate": 0.00014174999999999998, "loss": 0.4727, "step": 945 }, { "epoch": 0.4451764705882353, "grad_norm": 1.051618885417863, "learning_rate": 0.00014189999999999998, "loss": 0.5273, "step": 946 }, { "epoch": 0.4456470588235294, "grad_norm": 1.1106137448454652, "learning_rate": 0.00014204999999999998, "loss": 0.6406, "step": 947 }, { "epoch": 0.4461176470588235, "grad_norm": 1.5325320872741002, "learning_rate": 0.0001422, "loss": 0.3203, "step": 948 }, { "epoch": 0.4465882352941177, "grad_norm": 7.738079450020742, "learning_rate": 0.00014235, "loss": 1.8047, "step": 949 }, { "epoch": 0.4470588235294118, "grad_norm": 1.073168382556028, "learning_rate": 0.0001425, "loss": 0.4316, "step": 950 }, { "epoch": 0.4475294117647059, "grad_norm": 14.7348382860641, "learning_rate": 0.00014264999999999997, "loss": 1.4609, "step": 951 }, { "epoch": 0.448, "grad_norm": 2.067562253301923, "learning_rate": 0.00014279999999999997, "loss": 0.9062, "step": 952 }, { "epoch": 0.4484705882352941, "grad_norm": 6.27921576287008, "learning_rate": 0.00014294999999999998, "loss": 2.5625, "step": 953 }, { "epoch": 0.44894117647058823, "grad_norm": 1.9611474415747037, "learning_rate": 0.00014309999999999998, "loss": 0.7461, "step": 954 }, { "epoch": 0.44941176470588234, "grad_norm": 5.709048026965382, "learning_rate": 0.00014324999999999999, "loss": 0.8086, "step": 955 }, { "epoch": 0.44988235294117646, "grad_norm": 1.4939269492238372, "learning_rate": 0.0001434, "loss": 0.377, "step": 956 }, { "epoch": 0.45035294117647057, "grad_norm": 1.3720541381074287, "learning_rate": 0.00014355, "loss": 0.3145, "step": 957 }, { "epoch": 0.4508235294117647, "grad_norm": 0.7574448207015058, "learning_rate": 0.00014369999999999997, "loss": 0.291, "step": 958 }, { "epoch": 0.45129411764705885, "grad_norm": 0.4802099245114083, "learning_rate": 0.00014384999999999997, "loss": 0.1318, "step": 959 }, { "epoch": 0.45176470588235296, "grad_norm": 1.3020746476909402, "learning_rate": 0.00014399999999999998, "loss": 0.4922, "step": 960 }, { "epoch": 0.45223529411764707, "grad_norm": 0.4575126639614981, "learning_rate": 0.00014414999999999998, "loss": 0.3027, "step": 961 }, { "epoch": 0.4527058823529412, "grad_norm": 0.6259000856229613, "learning_rate": 0.00014429999999999998, "loss": 0.334, "step": 962 }, { "epoch": 0.4531764705882353, "grad_norm": 0.8065692492808693, "learning_rate": 0.00014445, "loss": 0.3242, "step": 963 }, { "epoch": 0.4536470588235294, "grad_norm": 0.7102892883533377, "learning_rate": 0.0001446, "loss": 0.1885, "step": 964 }, { "epoch": 0.4541176470588235, "grad_norm": 2.214293486611632, "learning_rate": 0.00014475, "loss": 0.7734, "step": 965 }, { "epoch": 0.4545882352941176, "grad_norm": 0.5848697516565188, "learning_rate": 0.00014489999999999997, "loss": 0.4004, "step": 966 }, { "epoch": 0.45505882352941174, "grad_norm": 0.620912577378254, "learning_rate": 0.00014504999999999997, "loss": 0.1992, "step": 967 }, { "epoch": 0.4555294117647059, "grad_norm": 2.0661415067613342, "learning_rate": 0.00014519999999999998, "loss": 0.4863, "step": 968 }, { "epoch": 0.456, "grad_norm": 2.0267144759109854, "learning_rate": 0.00014534999999999998, "loss": 0.4766, "step": 969 }, { "epoch": 0.45647058823529413, "grad_norm": 1.0456748051094453, "learning_rate": 0.00014549999999999999, "loss": 0.7188, "step": 970 }, { "epoch": 0.45694117647058824, "grad_norm": 1.1271303909845993, "learning_rate": 0.00014565, "loss": 0.6367, "step": 971 }, { "epoch": 0.45741176470588235, "grad_norm": 0.7073774914073668, "learning_rate": 0.0001458, "loss": 0.252, "step": 972 }, { "epoch": 0.45788235294117646, "grad_norm": 1.8099488109635762, "learning_rate": 0.00014595, "loss": 0.4883, "step": 973 }, { "epoch": 0.4583529411764706, "grad_norm": 1.7171995335048804, "learning_rate": 0.00014609999999999997, "loss": 0.5195, "step": 974 }, { "epoch": 0.4588235294117647, "grad_norm": 0.25356444957911045, "learning_rate": 0.00014624999999999998, "loss": 0.0938, "step": 975 }, { "epoch": 0.4592941176470588, "grad_norm": 0.764201887242091, "learning_rate": 0.00014639999999999998, "loss": 0.5156, "step": 976 }, { "epoch": 0.45976470588235296, "grad_norm": 0.5200283370908744, "learning_rate": 0.00014654999999999998, "loss": 0.2812, "step": 977 }, { "epoch": 0.4602352941176471, "grad_norm": 1.24589549332828, "learning_rate": 0.0001467, "loss": 0.373, "step": 978 }, { "epoch": 0.4607058823529412, "grad_norm": 1.02197846526516, "learning_rate": 0.00014685, "loss": 0.3223, "step": 979 }, { "epoch": 0.4611764705882353, "grad_norm": 0.38513442569321665, "learning_rate": 0.000147, "loss": 0.1289, "step": 980 }, { "epoch": 0.4616470588235294, "grad_norm": 0.8998791790048354, "learning_rate": 0.00014714999999999997, "loss": 0.4375, "step": 981 }, { "epoch": 0.4621176470588235, "grad_norm": 0.37587637375378047, "learning_rate": 0.00014729999999999998, "loss": 0.2129, "step": 982 }, { "epoch": 0.46258823529411763, "grad_norm": 0.45489794430585806, "learning_rate": 0.00014744999999999998, "loss": 0.3613, "step": 983 }, { "epoch": 0.46305882352941174, "grad_norm": 0.7618701371124983, "learning_rate": 0.00014759999999999998, "loss": 0.3184, "step": 984 }, { "epoch": 0.46352941176470586, "grad_norm": 1.5070352886002691, "learning_rate": 0.00014774999999999999, "loss": 0.5938, "step": 985 }, { "epoch": 0.464, "grad_norm": 0.5437309484874063, "learning_rate": 0.0001479, "loss": 0.3418, "step": 986 }, { "epoch": 0.46447058823529414, "grad_norm": 0.5634587270998968, "learning_rate": 0.00014805, "loss": 0.3418, "step": 987 }, { "epoch": 0.46494117647058825, "grad_norm": 0.4259915453220922, "learning_rate": 0.0001482, "loss": 0.1904, "step": 988 }, { "epoch": 0.46541176470588236, "grad_norm": 0.44135286944586183, "learning_rate": 0.00014834999999999997, "loss": 0.2002, "step": 989 }, { "epoch": 0.46588235294117647, "grad_norm": 0.8937848541702058, "learning_rate": 0.00014849999999999998, "loss": 0.5078, "step": 990 }, { "epoch": 0.4663529411764706, "grad_norm": 0.594195716473444, "learning_rate": 0.00014864999999999998, "loss": 0.4082, "step": 991 }, { "epoch": 0.4668235294117647, "grad_norm": 0.22652189655478774, "learning_rate": 0.00014879999999999998, "loss": 0.1367, "step": 992 }, { "epoch": 0.4672941176470588, "grad_norm": 0.5192840667396642, "learning_rate": 0.00014895, "loss": 0.2656, "step": 993 }, { "epoch": 0.4677647058823529, "grad_norm": 0.5722078484899096, "learning_rate": 0.0001491, "loss": 0.2598, "step": 994 }, { "epoch": 0.4682352941176471, "grad_norm": 0.21349146607078218, "learning_rate": 0.00014925, "loss": 0.1426, "step": 995 }, { "epoch": 0.4687058823529412, "grad_norm": 0.4466233991421943, "learning_rate": 0.0001494, "loss": 0.1992, "step": 996 }, { "epoch": 0.4691764705882353, "grad_norm": 0.46472969878260245, "learning_rate": 0.00014954999999999998, "loss": 0.332, "step": 997 }, { "epoch": 0.4696470588235294, "grad_norm": 0.7907071353676258, "learning_rate": 0.00014969999999999998, "loss": 0.8047, "step": 998 }, { "epoch": 0.47011764705882353, "grad_norm": 0.6229493718728893, "learning_rate": 0.00014984999999999998, "loss": 0.3652, "step": 999 }, { "epoch": 0.47058823529411764, "grad_norm": 0.4332538063714341, "learning_rate": 0.00015, "loss": 0.4512, "step": 1000 }, { "epoch": 0.47105882352941175, "grad_norm": 0.12807719899979783, "learning_rate": 0.00015014999999999996, "loss": 0.0618, "step": 1001 }, { "epoch": 0.47152941176470586, "grad_norm": 0.4532007753022373, "learning_rate": 0.0001503, "loss": 0.3203, "step": 1002 }, { "epoch": 0.472, "grad_norm": 0.44800786755260275, "learning_rate": 0.00015044999999999997, "loss": 0.1592, "step": 1003 }, { "epoch": 0.47247058823529414, "grad_norm": 0.5548174613616775, "learning_rate": 0.00015059999999999997, "loss": 0.3418, "step": 1004 }, { "epoch": 0.47294117647058825, "grad_norm": 0.29348923010121103, "learning_rate": 0.00015074999999999998, "loss": 0.1768, "step": 1005 }, { "epoch": 0.47341176470588237, "grad_norm": 0.4082443914205907, "learning_rate": 0.00015089999999999998, "loss": 0.1357, "step": 1006 }, { "epoch": 0.4738823529411765, "grad_norm": 1.0660939217098717, "learning_rate": 0.00015104999999999996, "loss": 0.5977, "step": 1007 }, { "epoch": 0.4743529411764706, "grad_norm": 0.18061788901443507, "learning_rate": 0.0001512, "loss": 0.0625, "step": 1008 }, { "epoch": 0.4748235294117647, "grad_norm": 0.4918310756256643, "learning_rate": 0.00015134999999999997, "loss": 0.4004, "step": 1009 }, { "epoch": 0.4752941176470588, "grad_norm": 0.5718895033479666, "learning_rate": 0.0001515, "loss": 0.3555, "step": 1010 }, { "epoch": 0.4757647058823529, "grad_norm": 0.4962152991228443, "learning_rate": 0.00015164999999999997, "loss": 0.3711, "step": 1011 }, { "epoch": 0.47623529411764703, "grad_norm": 0.5110505670744595, "learning_rate": 0.00015179999999999998, "loss": 0.3691, "step": 1012 }, { "epoch": 0.4767058823529412, "grad_norm": 0.5562760113810503, "learning_rate": 0.00015194999999999998, "loss": 0.4238, "step": 1013 }, { "epoch": 0.4771764705882353, "grad_norm": 0.26707530809488705, "learning_rate": 0.00015209999999999998, "loss": 0.0845, "step": 1014 }, { "epoch": 0.4776470588235294, "grad_norm": 0.342376070573134, "learning_rate": 0.00015224999999999996, "loss": 0.2383, "step": 1015 }, { "epoch": 0.47811764705882354, "grad_norm": 0.38482457586058816, "learning_rate": 0.0001524, "loss": 0.334, "step": 1016 }, { "epoch": 0.47858823529411765, "grad_norm": 0.5010378074039405, "learning_rate": 0.00015254999999999997, "loss": 0.2363, "step": 1017 }, { "epoch": 0.47905882352941176, "grad_norm": 0.5541810667677003, "learning_rate": 0.0001527, "loss": 0.4551, "step": 1018 }, { "epoch": 0.47952941176470587, "grad_norm": 0.3081419158540009, "learning_rate": 0.00015284999999999997, "loss": 0.1895, "step": 1019 }, { "epoch": 0.48, "grad_norm": 0.30808647126583905, "learning_rate": 0.00015299999999999998, "loss": 0.1455, "step": 1020 }, { "epoch": 0.4804705882352941, "grad_norm": 1.0537490937435874, "learning_rate": 0.00015314999999999998, "loss": 0.5078, "step": 1021 }, { "epoch": 0.48094117647058826, "grad_norm": 0.4668256547728154, "learning_rate": 0.00015329999999999999, "loss": 0.3574, "step": 1022 }, { "epoch": 0.4814117647058824, "grad_norm": 0.35183286685135096, "learning_rate": 0.00015344999999999996, "loss": 0.2676, "step": 1023 }, { "epoch": 0.4818823529411765, "grad_norm": 0.44407863653827107, "learning_rate": 0.0001536, "loss": 0.25, "step": 1024 }, { "epoch": 0.4823529411764706, "grad_norm": 0.21485808233795675, "learning_rate": 0.00015374999999999997, "loss": 0.126, "step": 1025 }, { "epoch": 0.4828235294117647, "grad_norm": 0.6073529025389884, "learning_rate": 0.0001539, "loss": 0.3633, "step": 1026 }, { "epoch": 0.4832941176470588, "grad_norm": 0.24256860071111075, "learning_rate": 0.00015404999999999998, "loss": 0.1299, "step": 1027 }, { "epoch": 0.48376470588235293, "grad_norm": 0.22413806847008247, "learning_rate": 0.00015419999999999998, "loss": 0.1377, "step": 1028 }, { "epoch": 0.48423529411764704, "grad_norm": 0.05052480738086909, "learning_rate": 0.00015434999999999998, "loss": 0.0114, "step": 1029 }, { "epoch": 0.48470588235294115, "grad_norm": 0.5816583670467674, "learning_rate": 0.0001545, "loss": 0.4395, "step": 1030 }, { "epoch": 0.4851764705882353, "grad_norm": 0.5564192908445479, "learning_rate": 0.00015464999999999996, "loss": 0.3965, "step": 1031 }, { "epoch": 0.48564705882352943, "grad_norm": 130.36383900166956, "learning_rate": 0.0001548, "loss": 0.793, "step": 1032 }, { "epoch": 0.48611764705882354, "grad_norm": 10.887046127156134, "learning_rate": 0.00015494999999999997, "loss": 0.9609, "step": 1033 }, { "epoch": 0.48658823529411765, "grad_norm": 1.101321709442698, "learning_rate": 0.0001551, "loss": 0.5898, "step": 1034 }, { "epoch": 0.48705882352941177, "grad_norm": 0.2820166081737122, "learning_rate": 0.00015524999999999998, "loss": 0.1201, "step": 1035 }, { "epoch": 0.4875294117647059, "grad_norm": 0.5975434010502195, "learning_rate": 0.00015539999999999998, "loss": 0.3809, "step": 1036 }, { "epoch": 0.488, "grad_norm": 0.24021145516370684, "learning_rate": 0.00015554999999999999, "loss": 0.1025, "step": 1037 }, { "epoch": 0.4884705882352941, "grad_norm": 1.4402900462842971, "learning_rate": 0.0001557, "loss": 0.6914, "step": 1038 }, { "epoch": 0.4889411764705882, "grad_norm": 0.33538616948657785, "learning_rate": 0.00015584999999999997, "loss": 0.1572, "step": 1039 }, { "epoch": 0.4894117647058824, "grad_norm": 1.3672954459345001, "learning_rate": 0.000156, "loss": 0.4473, "step": 1040 }, { "epoch": 0.4898823529411765, "grad_norm": 0.6936719125365621, "learning_rate": 0.00015614999999999997, "loss": 0.3516, "step": 1041 }, { "epoch": 0.4903529411764706, "grad_norm": 0.524652832816304, "learning_rate": 0.0001563, "loss": 0.2949, "step": 1042 }, { "epoch": 0.4908235294117647, "grad_norm": 0.8088601391774861, "learning_rate": 0.00015644999999999998, "loss": 0.4082, "step": 1043 }, { "epoch": 0.4912941176470588, "grad_norm": 0.8133514797721847, "learning_rate": 0.00015659999999999998, "loss": 0.7031, "step": 1044 }, { "epoch": 0.49176470588235294, "grad_norm": 0.8715808371599304, "learning_rate": 0.00015675, "loss": 0.8164, "step": 1045 }, { "epoch": 0.49223529411764705, "grad_norm": 0.6764377689088178, "learning_rate": 0.0001569, "loss": 0.5156, "step": 1046 }, { "epoch": 0.49270588235294116, "grad_norm": 0.6018609716143039, "learning_rate": 0.00015704999999999997, "loss": 0.2832, "step": 1047 }, { "epoch": 0.49317647058823527, "grad_norm": 1.3022570862722698, "learning_rate": 0.0001572, "loss": 0.8086, "step": 1048 }, { "epoch": 0.49364705882352944, "grad_norm": 0.5523986242078239, "learning_rate": 0.00015734999999999998, "loss": 0.3887, "step": 1049 }, { "epoch": 0.49411764705882355, "grad_norm": 0.27288001583793725, "learning_rate": 0.00015749999999999998, "loss": 0.1299, "step": 1050 }, { "epoch": 0.49458823529411766, "grad_norm": 1.0925443783253115, "learning_rate": 0.00015764999999999998, "loss": 0.4023, "step": 1051 }, { "epoch": 0.4950588235294118, "grad_norm": 0.9902890145792097, "learning_rate": 0.0001578, "loss": 0.4805, "step": 1052 }, { "epoch": 0.4955294117647059, "grad_norm": 0.2707468959959759, "learning_rate": 0.00015794999999999996, "loss": 0.1846, "step": 1053 }, { "epoch": 0.496, "grad_norm": 1.0962020012758753, "learning_rate": 0.0001581, "loss": 0.5312, "step": 1054 }, { "epoch": 0.4964705882352941, "grad_norm": 0.592992321402021, "learning_rate": 0.00015824999999999997, "loss": 0.2676, "step": 1055 }, { "epoch": 0.4969411764705882, "grad_norm": 0.8336383921616137, "learning_rate": 0.0001584, "loss": 0.6641, "step": 1056 }, { "epoch": 0.49741176470588233, "grad_norm": 0.9026143375198125, "learning_rate": 0.00015854999999999998, "loss": 0.5703, "step": 1057 }, { "epoch": 0.4978823529411765, "grad_norm": 0.8117138371430479, "learning_rate": 0.00015869999999999998, "loss": 0.6719, "step": 1058 }, { "epoch": 0.4983529411764706, "grad_norm": 0.2996201979149486, "learning_rate": 0.00015884999999999999, "loss": 0.1475, "step": 1059 }, { "epoch": 0.4988235294117647, "grad_norm": 0.19040380712233396, "learning_rate": 0.000159, "loss": 0.0654, "step": 1060 }, { "epoch": 0.49929411764705883, "grad_norm": 0.6767345432394254, "learning_rate": 0.00015914999999999997, "loss": 0.168, "step": 1061 }, { "epoch": 0.49976470588235294, "grad_norm": 0.80483016317124, "learning_rate": 0.0001593, "loss": 0.2676, "step": 1062 }, { "epoch": 0.5002352941176471, "grad_norm": 0.3909471712109635, "learning_rate": 0.00015944999999999997, "loss": 0.1533, "step": 1063 }, { "epoch": 0.5007058823529412, "grad_norm": 0.5509135897968765, "learning_rate": 0.0001596, "loss": 0.2793, "step": 1064 }, { "epoch": 0.5011764705882353, "grad_norm": 7.621971610532651, "learning_rate": 0.00015974999999999998, "loss": 0.3965, "step": 1065 }, { "epoch": 0.5016470588235294, "grad_norm": 14.497614242064484, "learning_rate": 0.00015989999999999998, "loss": 1.4219, "step": 1066 }, { "epoch": 0.5021176470588236, "grad_norm": 31.04438483717043, "learning_rate": 0.00016005, "loss": 2.9219, "step": 1067 }, { "epoch": 0.5025882352941177, "grad_norm": 0.9507505079689481, "learning_rate": 0.0001602, "loss": 0.6328, "step": 1068 }, { "epoch": 0.5030588235294118, "grad_norm": 2.562485366640766, "learning_rate": 0.00016034999999999997, "loss": 0.1846, "step": 1069 }, { "epoch": 0.5035294117647059, "grad_norm": 1.3331384822660648, "learning_rate": 0.0001605, "loss": 0.1797, "step": 1070 }, { "epoch": 0.504, "grad_norm": 7.444012601366254, "learning_rate": 0.00016064999999999997, "loss": 1.1172, "step": 1071 }, { "epoch": 0.5044705882352941, "grad_norm": 0.45362605794592775, "learning_rate": 0.0001608, "loss": 0.1582, "step": 1072 }, { "epoch": 0.5049411764705882, "grad_norm": 1.3790475132355446, "learning_rate": 0.00016094999999999998, "loss": 0.5781, "step": 1073 }, { "epoch": 0.5054117647058823, "grad_norm": 0.9130988162421337, "learning_rate": 0.00016109999999999999, "loss": 0.6289, "step": 1074 }, { "epoch": 0.5058823529411764, "grad_norm": 0.554624986485218, "learning_rate": 0.00016125, "loss": 0.416, "step": 1075 }, { "epoch": 0.5063529411764706, "grad_norm": 0.9172378003571626, "learning_rate": 0.0001614, "loss": 0.4062, "step": 1076 }, { "epoch": 0.5068235294117647, "grad_norm": 0.16586056603121577, "learning_rate": 0.00016154999999999997, "loss": 0.0559, "step": 1077 }, { "epoch": 0.5072941176470588, "grad_norm": 0.4144962003272826, "learning_rate": 0.0001617, "loss": 0.1094, "step": 1078 }, { "epoch": 0.5077647058823529, "grad_norm": 0.8596561533127368, "learning_rate": 0.00016184999999999998, "loss": 0.4316, "step": 1079 }, { "epoch": 0.508235294117647, "grad_norm": 0.5206510783325673, "learning_rate": 0.000162, "loss": 0.2227, "step": 1080 }, { "epoch": 0.5087058823529412, "grad_norm": 1.287918184549901, "learning_rate": 0.00016214999999999998, "loss": 0.5664, "step": 1081 }, { "epoch": 0.5091764705882353, "grad_norm": 0.3957977849106666, "learning_rate": 0.0001623, "loss": 0.3184, "step": 1082 }, { "epoch": 0.5096470588235295, "grad_norm": 0.43930513947804306, "learning_rate": 0.00016245, "loss": 0.3027, "step": 1083 }, { "epoch": 0.5101176470588236, "grad_norm": 0.6844852305229466, "learning_rate": 0.0001626, "loss": 0.6406, "step": 1084 }, { "epoch": 0.5105882352941177, "grad_norm": 0.3457634774529634, "learning_rate": 0.00016274999999999997, "loss": 0.2217, "step": 1085 }, { "epoch": 0.5110588235294118, "grad_norm": 0.43353142952289675, "learning_rate": 0.0001629, "loss": 0.3066, "step": 1086 }, { "epoch": 0.5115294117647059, "grad_norm": 0.4109303597137811, "learning_rate": 0.00016304999999999998, "loss": 0.2754, "step": 1087 }, { "epoch": 0.512, "grad_norm": 0.3182848140371119, "learning_rate": 0.0001632, "loss": 0.1235, "step": 1088 }, { "epoch": 0.5124705882352941, "grad_norm": 1.158758053237921, "learning_rate": 0.00016334999999999999, "loss": 0.7656, "step": 1089 }, { "epoch": 0.5129411764705882, "grad_norm": 0.7003879104366687, "learning_rate": 0.0001635, "loss": 0.3574, "step": 1090 }, { "epoch": 0.5134117647058823, "grad_norm": 0.9002260109306522, "learning_rate": 0.00016365, "loss": 0.4414, "step": 1091 }, { "epoch": 0.5138823529411765, "grad_norm": 0.2636348442089857, "learning_rate": 0.0001638, "loss": 0.1553, "step": 1092 }, { "epoch": 0.5143529411764706, "grad_norm": 0.9398155526988595, "learning_rate": 0.00016394999999999997, "loss": 0.4102, "step": 1093 }, { "epoch": 0.5148235294117647, "grad_norm": 0.87071504318831, "learning_rate": 0.0001641, "loss": 0.4102, "step": 1094 }, { "epoch": 0.5152941176470588, "grad_norm": 0.7702865248803222, "learning_rate": 0.00016424999999999998, "loss": 0.4531, "step": 1095 }, { "epoch": 0.5157647058823529, "grad_norm": 0.4980913518794899, "learning_rate": 0.0001644, "loss": 0.165, "step": 1096 }, { "epoch": 0.516235294117647, "grad_norm": 1.3072065448269046, "learning_rate": 0.00016455, "loss": 0.4512, "step": 1097 }, { "epoch": 0.5167058823529411, "grad_norm": 0.4104350327010381, "learning_rate": 0.0001647, "loss": 0.2061, "step": 1098 }, { "epoch": 0.5171764705882353, "grad_norm": 0.72410857714883, "learning_rate": 0.00016485, "loss": 0.4434, "step": 1099 }, { "epoch": 0.5176470588235295, "grad_norm": 0.5188643618085057, "learning_rate": 0.000165, "loss": 0.2168, "step": 1100 }, { "epoch": 0.5181176470588236, "grad_norm": 0.4218343026049592, "learning_rate": 0.00016514999999999998, "loss": 0.2598, "step": 1101 }, { "epoch": 0.5185882352941177, "grad_norm": 0.6907733135700641, "learning_rate": 0.0001653, "loss": 0.4395, "step": 1102 }, { "epoch": 0.5190588235294118, "grad_norm": 0.7820837416661196, "learning_rate": 0.00016544999999999998, "loss": 0.5156, "step": 1103 }, { "epoch": 0.5195294117647059, "grad_norm": 0.7459866995586093, "learning_rate": 0.0001656, "loss": 0.543, "step": 1104 }, { "epoch": 0.52, "grad_norm": 0.5631044698980698, "learning_rate": 0.00016575, "loss": 0.334, "step": 1105 }, { "epoch": 0.5204705882352941, "grad_norm": 0.22319921167007595, "learning_rate": 0.0001659, "loss": 0.0996, "step": 1106 }, { "epoch": 0.5209411764705882, "grad_norm": 0.2270054989658639, "learning_rate": 0.00016604999999999997, "loss": 0.1328, "step": 1107 }, { "epoch": 0.5214117647058824, "grad_norm": 0.42987130913483845, "learning_rate": 0.0001662, "loss": 0.1836, "step": 1108 }, { "epoch": 0.5218823529411765, "grad_norm": 1.1942828485598984, "learning_rate": 0.00016634999999999998, "loss": 0.6094, "step": 1109 }, { "epoch": 0.5223529411764706, "grad_norm": 0.5203818144501925, "learning_rate": 0.0001665, "loss": 0.4336, "step": 1110 }, { "epoch": 0.5228235294117647, "grad_norm": 0.529460050408733, "learning_rate": 0.00016664999999999998, "loss": 0.2334, "step": 1111 }, { "epoch": 0.5232941176470588, "grad_norm": 0.7275054861202723, "learning_rate": 0.0001668, "loss": 0.377, "step": 1112 }, { "epoch": 0.5237647058823529, "grad_norm": 0.5538674422227177, "learning_rate": 0.00016695, "loss": 0.373, "step": 1113 }, { "epoch": 0.524235294117647, "grad_norm": 0.436226637191822, "learning_rate": 0.0001671, "loss": 0.3809, "step": 1114 }, { "epoch": 0.5247058823529411, "grad_norm": 0.9088185243275496, "learning_rate": 0.00016724999999999997, "loss": 0.416, "step": 1115 }, { "epoch": 0.5251764705882352, "grad_norm": 0.615981987442095, "learning_rate": 0.0001674, "loss": 0.3301, "step": 1116 }, { "epoch": 0.5256470588235295, "grad_norm": 0.39440408083240835, "learning_rate": 0.00016754999999999998, "loss": 0.2676, "step": 1117 }, { "epoch": 0.5261176470588236, "grad_norm": 1.1498369004884696, "learning_rate": 0.0001677, "loss": 0.6211, "step": 1118 }, { "epoch": 0.5265882352941177, "grad_norm": 0.45936482840322757, "learning_rate": 0.00016785, "loss": 0.3184, "step": 1119 }, { "epoch": 0.5270588235294118, "grad_norm": 0.6273165559172755, "learning_rate": 0.000168, "loss": 0.4961, "step": 1120 }, { "epoch": 0.5275294117647059, "grad_norm": 0.4927038324120246, "learning_rate": 0.00016815, "loss": 0.3906, "step": 1121 }, { "epoch": 0.528, "grad_norm": 0.2710386308830355, "learning_rate": 0.0001683, "loss": 0.1611, "step": 1122 }, { "epoch": 0.5284705882352941, "grad_norm": 0.43150941996264935, "learning_rate": 0.00016844999999999997, "loss": 0.2197, "step": 1123 }, { "epoch": 0.5289411764705882, "grad_norm": 0.8218392672923381, "learning_rate": 0.0001686, "loss": 0.5273, "step": 1124 }, { "epoch": 0.5294117647058824, "grad_norm": 0.669381105447083, "learning_rate": 0.00016874999999999998, "loss": 0.4551, "step": 1125 }, { "epoch": 0.5298823529411765, "grad_norm": 0.3349158179702733, "learning_rate": 0.00016889999999999996, "loss": 0.1475, "step": 1126 }, { "epoch": 0.5303529411764706, "grad_norm": 0.5648660588252027, "learning_rate": 0.00016905, "loss": 0.2197, "step": 1127 }, { "epoch": 0.5308235294117647, "grad_norm": 0.6686739918857232, "learning_rate": 0.00016919999999999997, "loss": 0.3945, "step": 1128 }, { "epoch": 0.5312941176470588, "grad_norm": 0.27165387179779277, "learning_rate": 0.00016935, "loss": 0.2129, "step": 1129 }, { "epoch": 0.5317647058823529, "grad_norm": 0.6042434358779378, "learning_rate": 0.00016949999999999997, "loss": 0.377, "step": 1130 }, { "epoch": 0.532235294117647, "grad_norm": 0.5968182425149446, "learning_rate": 0.00016964999999999998, "loss": 0.3125, "step": 1131 }, { "epoch": 0.5327058823529411, "grad_norm": 0.5109222459530585, "learning_rate": 0.00016979999999999998, "loss": 0.2432, "step": 1132 }, { "epoch": 0.5331764705882353, "grad_norm": 0.7310528967042255, "learning_rate": 0.00016994999999999998, "loss": 0.4629, "step": 1133 }, { "epoch": 0.5336470588235294, "grad_norm": 0.14788174572514964, "learning_rate": 0.00017009999999999996, "loss": 0.084, "step": 1134 }, { "epoch": 0.5341176470588235, "grad_norm": 0.48298783826063485, "learning_rate": 0.00017025, "loss": 0.2754, "step": 1135 }, { "epoch": 0.5345882352941177, "grad_norm": 0.7894997255983062, "learning_rate": 0.00017039999999999997, "loss": 0.4004, "step": 1136 }, { "epoch": 0.5350588235294118, "grad_norm": 0.47112810834410207, "learning_rate": 0.00017055, "loss": 0.2559, "step": 1137 }, { "epoch": 0.5355294117647059, "grad_norm": 0.8985681153613565, "learning_rate": 0.00017069999999999998, "loss": 0.2891, "step": 1138 }, { "epoch": 0.536, "grad_norm": 0.4318152372452802, "learning_rate": 0.00017084999999999998, "loss": 0.2158, "step": 1139 }, { "epoch": 0.5364705882352941, "grad_norm": 0.23278971415238212, "learning_rate": 0.00017099999999999998, "loss": 0.0693, "step": 1140 }, { "epoch": 0.5369411764705883, "grad_norm": 0.8358493309350203, "learning_rate": 0.00017114999999999999, "loss": 0.543, "step": 1141 }, { "epoch": 0.5374117647058824, "grad_norm": 0.8187930739768438, "learning_rate": 0.00017129999999999996, "loss": 0.4707, "step": 1142 }, { "epoch": 0.5378823529411765, "grad_norm": 1.29112237186011, "learning_rate": 0.00017145, "loss": 0.6602, "step": 1143 }, { "epoch": 0.5383529411764706, "grad_norm": 0.313532928642508, "learning_rate": 0.00017159999999999997, "loss": 0.1904, "step": 1144 }, { "epoch": 0.5388235294117647, "grad_norm": 0.6357197626505234, "learning_rate": 0.00017175, "loss": 0.373, "step": 1145 }, { "epoch": 0.5392941176470588, "grad_norm": 1.2594614523907277, "learning_rate": 0.00017189999999999998, "loss": 0.6328, "step": 1146 }, { "epoch": 0.5397647058823529, "grad_norm": 0.12846818386709824, "learning_rate": 0.00017204999999999998, "loss": 0.0598, "step": 1147 }, { "epoch": 0.540235294117647, "grad_norm": 0.35344729149814674, "learning_rate": 0.00017219999999999998, "loss": 0.2031, "step": 1148 }, { "epoch": 0.5407058823529411, "grad_norm": 1.0056971293957233, "learning_rate": 0.00017235, "loss": 0.3926, "step": 1149 }, { "epoch": 0.5411764705882353, "grad_norm": 0.47184067501756555, "learning_rate": 0.00017249999999999996, "loss": 0.1826, "step": 1150 }, { "epoch": 0.5416470588235294, "grad_norm": 0.19504669158171506, "learning_rate": 0.00017265, "loss": 0.1069, "step": 1151 }, { "epoch": 0.5421176470588235, "grad_norm": 0.2355682494398515, "learning_rate": 0.00017279999999999997, "loss": 0.1157, "step": 1152 }, { "epoch": 0.5425882352941176, "grad_norm": 0.36700156504220793, "learning_rate": 0.00017294999999999998, "loss": 0.2461, "step": 1153 }, { "epoch": 0.5430588235294118, "grad_norm": 0.7684337058534506, "learning_rate": 0.00017309999999999998, "loss": 0.3809, "step": 1154 }, { "epoch": 0.5435294117647059, "grad_norm": 0.17672658376946357, "learning_rate": 0.00017324999999999998, "loss": 0.0908, "step": 1155 }, { "epoch": 0.544, "grad_norm": 0.2670982620047788, "learning_rate": 0.00017339999999999996, "loss": 0.2129, "step": 1156 }, { "epoch": 0.5444705882352942, "grad_norm": 0.19161027285707122, "learning_rate": 0.00017355, "loss": 0.1079, "step": 1157 }, { "epoch": 0.5449411764705883, "grad_norm": 0.3542981231000798, "learning_rate": 0.00017369999999999997, "loss": 0.3223, "step": 1158 }, { "epoch": 0.5454117647058824, "grad_norm": 0.606442255392123, "learning_rate": 0.00017385, "loss": 0.3496, "step": 1159 }, { "epoch": 0.5458823529411765, "grad_norm": 0.48162007944649093, "learning_rate": 0.00017399999999999997, "loss": 0.3477, "step": 1160 }, { "epoch": 0.5463529411764706, "grad_norm": 0.43483880792508, "learning_rate": 0.00017414999999999998, "loss": 0.2656, "step": 1161 }, { "epoch": 0.5468235294117647, "grad_norm": 0.33595205912219744, "learning_rate": 0.00017429999999999998, "loss": 0.1807, "step": 1162 }, { "epoch": 0.5472941176470588, "grad_norm": 0.46168691318745403, "learning_rate": 0.00017444999999999998, "loss": 0.2988, "step": 1163 }, { "epoch": 0.5477647058823529, "grad_norm": 0.26798822204069195, "learning_rate": 0.00017459999999999996, "loss": 0.1406, "step": 1164 }, { "epoch": 0.548235294117647, "grad_norm": 0.3950519709607946, "learning_rate": 0.00017475, "loss": 0.1196, "step": 1165 }, { "epoch": 0.5487058823529412, "grad_norm": 0.45417260000119486, "learning_rate": 0.00017489999999999997, "loss": 0.2852, "step": 1166 }, { "epoch": 0.5491764705882353, "grad_norm": 0.4609525984825212, "learning_rate": 0.00017505, "loss": 0.2676, "step": 1167 }, { "epoch": 0.5496470588235294, "grad_norm": 0.2978110485493122, "learning_rate": 0.00017519999999999998, "loss": 0.1406, "step": 1168 }, { "epoch": 0.5501176470588235, "grad_norm": 0.334102878379339, "learning_rate": 0.00017534999999999998, "loss": 0.1992, "step": 1169 }, { "epoch": 0.5505882352941176, "grad_norm": 0.06559836535589754, "learning_rate": 0.00017549999999999998, "loss": 0.0126, "step": 1170 }, { "epoch": 0.5510588235294117, "grad_norm": 0.40595239480006046, "learning_rate": 0.00017565, "loss": 0.1367, "step": 1171 }, { "epoch": 0.5515294117647059, "grad_norm": 0.3551210630287856, "learning_rate": 0.00017579999999999996, "loss": 0.1914, "step": 1172 }, { "epoch": 0.552, "grad_norm": 0.59201459557432, "learning_rate": 0.00017595, "loss": 0.3477, "step": 1173 }, { "epoch": 0.5524705882352942, "grad_norm": 0.21148736490469888, "learning_rate": 0.00017609999999999997, "loss": 0.0776, "step": 1174 }, { "epoch": 0.5529411764705883, "grad_norm": 0.5642803214884526, "learning_rate": 0.00017625, "loss": 0.3105, "step": 1175 }, { "epoch": 0.5534117647058824, "grad_norm": 0.43735541694456503, "learning_rate": 0.00017639999999999998, "loss": 0.1826, "step": 1176 }, { "epoch": 0.5538823529411765, "grad_norm": 0.5097457642937747, "learning_rate": 0.00017654999999999998, "loss": 0.209, "step": 1177 }, { "epoch": 0.5543529411764706, "grad_norm": 0.28166781243980665, "learning_rate": 0.00017669999999999999, "loss": 0.2412, "step": 1178 }, { "epoch": 0.5548235294117647, "grad_norm": 0.269729630207667, "learning_rate": 0.00017685, "loss": 0.1377, "step": 1179 }, { "epoch": 0.5552941176470588, "grad_norm": 0.2613693491678193, "learning_rate": 0.00017699999999999997, "loss": 0.1445, "step": 1180 }, { "epoch": 0.5557647058823529, "grad_norm": 0.7047858615563941, "learning_rate": 0.00017715, "loss": 0.3867, "step": 1181 }, { "epoch": 0.556235294117647, "grad_norm": 0.4632279957909421, "learning_rate": 0.00017729999999999997, "loss": 0.3184, "step": 1182 }, { "epoch": 0.5567058823529412, "grad_norm": 0.5134713085470163, "learning_rate": 0.00017745, "loss": 0.2441, "step": 1183 }, { "epoch": 0.5571764705882353, "grad_norm": 0.2971168719738786, "learning_rate": 0.00017759999999999998, "loss": 0.2188, "step": 1184 }, { "epoch": 0.5576470588235294, "grad_norm": 0.588083001733109, "learning_rate": 0.00017774999999999998, "loss": 0.4023, "step": 1185 }, { "epoch": 0.5581176470588235, "grad_norm": 0.30044402628584416, "learning_rate": 0.0001779, "loss": 0.0728, "step": 1186 }, { "epoch": 0.5585882352941176, "grad_norm": 0.5560675492221114, "learning_rate": 0.00017805, "loss": 0.4277, "step": 1187 }, { "epoch": 0.5590588235294117, "grad_norm": 0.39352299271871377, "learning_rate": 0.00017819999999999997, "loss": 0.334, "step": 1188 }, { "epoch": 0.5595294117647058, "grad_norm": 0.37831929588394964, "learning_rate": 0.00017835, "loss": 0.1855, "step": 1189 }, { "epoch": 0.56, "grad_norm": 0.4747975149180563, "learning_rate": 0.00017849999999999997, "loss": 0.252, "step": 1190 }, { "epoch": 0.5604705882352942, "grad_norm": 0.3218938885080376, "learning_rate": 0.00017865, "loss": 0.1748, "step": 1191 }, { "epoch": 0.5609411764705883, "grad_norm": 0.48062905984727605, "learning_rate": 0.00017879999999999998, "loss": 0.3184, "step": 1192 }, { "epoch": 0.5614117647058824, "grad_norm": 0.22338024378256896, "learning_rate": 0.00017894999999999999, "loss": 0.1338, "step": 1193 }, { "epoch": 0.5618823529411765, "grad_norm": 0.16542968614763015, "learning_rate": 0.0001791, "loss": 0.0752, "step": 1194 }, { "epoch": 0.5623529411764706, "grad_norm": 0.27391765527969697, "learning_rate": 0.00017925, "loss": 0.1099, "step": 1195 }, { "epoch": 0.5628235294117647, "grad_norm": 0.7811223585042465, "learning_rate": 0.00017939999999999997, "loss": 0.3262, "step": 1196 }, { "epoch": 0.5632941176470588, "grad_norm": 0.30801002752799234, "learning_rate": 0.00017955, "loss": 0.2197, "step": 1197 }, { "epoch": 0.563764705882353, "grad_norm": 0.4492241500436995, "learning_rate": 0.00017969999999999998, "loss": 0.4102, "step": 1198 }, { "epoch": 0.5642352941176471, "grad_norm": 0.4996849622423084, "learning_rate": 0.00017984999999999998, "loss": 0.4629, "step": 1199 }, { "epoch": 0.5647058823529412, "grad_norm": 1.745363667800955, "learning_rate": 0.00017999999999999998, "loss": 0.3125, "step": 1200 }, { "epoch": 0.5651764705882353, "grad_norm": 0.4734023676234772, "learning_rate": 0.00018015, "loss": 0.4004, "step": 1201 }, { "epoch": 0.5656470588235294, "grad_norm": 0.29723399360126634, "learning_rate": 0.00018029999999999996, "loss": 0.1934, "step": 1202 }, { "epoch": 0.5661176470588235, "grad_norm": 0.3201405271912412, "learning_rate": 0.00018045, "loss": 0.1445, "step": 1203 }, { "epoch": 0.5665882352941176, "grad_norm": 0.3194660016270721, "learning_rate": 0.00018059999999999997, "loss": 0.1533, "step": 1204 }, { "epoch": 0.5670588235294117, "grad_norm": 1.3207439975559794, "learning_rate": 0.00018075, "loss": 0.75, "step": 1205 }, { "epoch": 0.5675294117647058, "grad_norm": 0.42044854637637313, "learning_rate": 0.00018089999999999998, "loss": 0.377, "step": 1206 }, { "epoch": 0.568, "grad_norm": 0.43325562615174806, "learning_rate": 0.00018104999999999998, "loss": 0.293, "step": 1207 }, { "epoch": 0.5684705882352942, "grad_norm": 0.33247244666386333, "learning_rate": 0.00018119999999999999, "loss": 0.1504, "step": 1208 }, { "epoch": 0.5689411764705883, "grad_norm": 1.3215588921080286, "learning_rate": 0.00018135, "loss": 0.3906, "step": 1209 }, { "epoch": 0.5694117647058824, "grad_norm": 0.6654696155079777, "learning_rate": 0.00018149999999999997, "loss": 0.3965, "step": 1210 }, { "epoch": 0.5698823529411765, "grad_norm": 0.7795394500954109, "learning_rate": 0.00018165, "loss": 0.3477, "step": 1211 }, { "epoch": 0.5703529411764706, "grad_norm": 0.8890747131536718, "learning_rate": 0.00018179999999999997, "loss": 0.3867, "step": 1212 }, { "epoch": 0.5708235294117647, "grad_norm": 0.9170651968640205, "learning_rate": 0.00018195, "loss": 0.4883, "step": 1213 }, { "epoch": 0.5712941176470588, "grad_norm": 1.0721965147512325, "learning_rate": 0.00018209999999999998, "loss": 0.5742, "step": 1214 }, { "epoch": 0.571764705882353, "grad_norm": 0.6377100097140628, "learning_rate": 0.00018224999999999998, "loss": 0.291, "step": 1215 }, { "epoch": 0.5722352941176471, "grad_norm": 0.9129319777562448, "learning_rate": 0.0001824, "loss": 0.291, "step": 1216 }, { "epoch": 0.5727058823529412, "grad_norm": 0.7241774011424386, "learning_rate": 0.00018255, "loss": 0.4473, "step": 1217 }, { "epoch": 0.5731764705882353, "grad_norm": 0.6191618701919391, "learning_rate": 0.00018269999999999997, "loss": 0.3301, "step": 1218 }, { "epoch": 0.5736470588235294, "grad_norm": 0.7376465413461164, "learning_rate": 0.00018285, "loss": 0.5508, "step": 1219 }, { "epoch": 0.5741176470588235, "grad_norm": 0.6233800598910109, "learning_rate": 0.00018299999999999998, "loss": 0.3809, "step": 1220 }, { "epoch": 0.5745882352941176, "grad_norm": 0.8968588620802962, "learning_rate": 0.00018315, "loss": 0.5586, "step": 1221 }, { "epoch": 0.5750588235294117, "grad_norm": 0.43130737001879427, "learning_rate": 0.00018329999999999998, "loss": 0.4473, "step": 1222 }, { "epoch": 0.5755294117647058, "grad_norm": 0.5459026720849964, "learning_rate": 0.00018345, "loss": 0.3027, "step": 1223 }, { "epoch": 0.576, "grad_norm": 0.7065162714104544, "learning_rate": 0.0001836, "loss": 0.4648, "step": 1224 }, { "epoch": 0.5764705882352941, "grad_norm": 0.29379081298055815, "learning_rate": 0.00018375, "loss": 0.1865, "step": 1225 }, { "epoch": 0.5769411764705883, "grad_norm": 0.5239764523426201, "learning_rate": 0.00018389999999999997, "loss": 0.4199, "step": 1226 }, { "epoch": 0.5774117647058824, "grad_norm": 0.4750127000686836, "learning_rate": 0.00018405, "loss": 0.3477, "step": 1227 }, { "epoch": 0.5778823529411765, "grad_norm": 0.4554816922090711, "learning_rate": 0.00018419999999999998, "loss": 0.2021, "step": 1228 }, { "epoch": 0.5783529411764706, "grad_norm": 0.4265091692275325, "learning_rate": 0.00018435, "loss": 0.334, "step": 1229 }, { "epoch": 0.5788235294117647, "grad_norm": 0.3135862783863988, "learning_rate": 0.00018449999999999999, "loss": 0.2012, "step": 1230 }, { "epoch": 0.5792941176470588, "grad_norm": 0.3014438805183728, "learning_rate": 0.00018465, "loss": 0.104, "step": 1231 }, { "epoch": 0.579764705882353, "grad_norm": 0.0838336410073949, "learning_rate": 0.0001848, "loss": 0.0199, "step": 1232 }, { "epoch": 0.5802352941176471, "grad_norm": 0.5130702883359033, "learning_rate": 0.00018495, "loss": 0.249, "step": 1233 }, { "epoch": 0.5807058823529412, "grad_norm": 0.9142097344797621, "learning_rate": 0.00018509999999999997, "loss": 0.3457, "step": 1234 }, { "epoch": 0.5811764705882353, "grad_norm": 0.29137775881610023, "learning_rate": 0.00018525, "loss": 0.0952, "step": 1235 }, { "epoch": 0.5816470588235294, "grad_norm": 0.4131213183701821, "learning_rate": 0.00018539999999999998, "loss": 0.2441, "step": 1236 }, { "epoch": 0.5821176470588235, "grad_norm": 0.3962145050069321, "learning_rate": 0.00018555, "loss": 0.2275, "step": 1237 }, { "epoch": 0.5825882352941176, "grad_norm": 0.3930691069732539, "learning_rate": 0.0001857, "loss": 0.2656, "step": 1238 }, { "epoch": 0.5830588235294117, "grad_norm": 0.9913589015634764, "learning_rate": 0.00018585, "loss": 0.6953, "step": 1239 }, { "epoch": 0.5835294117647059, "grad_norm": 0.1805098181643691, "learning_rate": 0.000186, "loss": 0.0884, "step": 1240 }, { "epoch": 0.584, "grad_norm": 0.43708993942179286, "learning_rate": 0.00018615, "loss": 0.332, "step": 1241 }, { "epoch": 0.5844705882352941, "grad_norm": 0.1017556511869099, "learning_rate": 0.00018629999999999997, "loss": 0.0259, "step": 1242 }, { "epoch": 0.5849411764705882, "grad_norm": 0.7802313409660265, "learning_rate": 0.00018645, "loss": 0.2793, "step": 1243 }, { "epoch": 0.5854117647058823, "grad_norm": 0.7123972986751385, "learning_rate": 0.00018659999999999998, "loss": 0.3594, "step": 1244 }, { "epoch": 0.5858823529411765, "grad_norm": 0.7190632257413204, "learning_rate": 0.00018675, "loss": 0.416, "step": 1245 }, { "epoch": 0.5863529411764706, "grad_norm": 0.03750538656503901, "learning_rate": 0.0001869, "loss": 0.0067, "step": 1246 }, { "epoch": 0.5868235294117647, "grad_norm": 0.41188565339116356, "learning_rate": 0.00018705, "loss": 0.2354, "step": 1247 }, { "epoch": 0.5872941176470589, "grad_norm": 0.6614101136539572, "learning_rate": 0.0001872, "loss": 0.3809, "step": 1248 }, { "epoch": 0.587764705882353, "grad_norm": 0.7215611776507429, "learning_rate": 0.00018735, "loss": 0.4648, "step": 1249 }, { "epoch": 0.5882352941176471, "grad_norm": 0.45833711561484103, "learning_rate": 0.00018749999999999998, "loss": 0.3477, "step": 1250 }, { "epoch": 0.5887058823529412, "grad_norm": 0.21849226341268052, "learning_rate": 0.00018764999999999998, "loss": 0.1406, "step": 1251 }, { "epoch": 0.5891764705882353, "grad_norm": 0.8795154519151053, "learning_rate": 0.00018779999999999998, "loss": 0.3301, "step": 1252 }, { "epoch": 0.5896470588235294, "grad_norm": 0.8837011052811973, "learning_rate": 0.00018794999999999996, "loss": 0.3457, "step": 1253 }, { "epoch": 0.5901176470588235, "grad_norm": 0.08888720116113552, "learning_rate": 0.0001881, "loss": 0.0225, "step": 1254 }, { "epoch": 0.5905882352941176, "grad_norm": 0.1846484812654719, "learning_rate": 0.00018824999999999997, "loss": 0.0574, "step": 1255 }, { "epoch": 0.5910588235294117, "grad_norm": 0.3718152365828226, "learning_rate": 0.00018839999999999997, "loss": 0.2188, "step": 1256 }, { "epoch": 0.5915294117647059, "grad_norm": 0.29105641718147074, "learning_rate": 0.00018854999999999998, "loss": 0.1758, "step": 1257 }, { "epoch": 0.592, "grad_norm": 0.47244577373352264, "learning_rate": 0.00018869999999999998, "loss": 0.3867, "step": 1258 }, { "epoch": 0.5924705882352941, "grad_norm": 0.8134323115433131, "learning_rate": 0.00018884999999999996, "loss": 0.5938, "step": 1259 }, { "epoch": 0.5929411764705882, "grad_norm": 0.29972315049160136, "learning_rate": 0.00018899999999999999, "loss": 0.1963, "step": 1260 }, { "epoch": 0.5934117647058823, "grad_norm": 0.34868557038550085, "learning_rate": 0.00018914999999999996, "loss": 0.2676, "step": 1261 }, { "epoch": 0.5938823529411764, "grad_norm": 0.44739076890796964, "learning_rate": 0.0001893, "loss": 0.2637, "step": 1262 }, { "epoch": 0.5943529411764706, "grad_norm": 0.5415435696942915, "learning_rate": 0.00018944999999999997, "loss": 0.3145, "step": 1263 }, { "epoch": 0.5948235294117648, "grad_norm": 0.38315817224468357, "learning_rate": 0.00018959999999999997, "loss": 0.2031, "step": 1264 }, { "epoch": 0.5952941176470589, "grad_norm": 0.27156588290654354, "learning_rate": 0.00018974999999999998, "loss": 0.2314, "step": 1265 }, { "epoch": 0.595764705882353, "grad_norm": 0.31032310329814017, "learning_rate": 0.00018989999999999998, "loss": 0.1592, "step": 1266 }, { "epoch": 0.5962352941176471, "grad_norm": 0.12777369234926833, "learning_rate": 0.00019004999999999996, "loss": 0.0762, "step": 1267 }, { "epoch": 0.5967058823529412, "grad_norm": 0.6530819027842283, "learning_rate": 0.0001902, "loss": 0.3711, "step": 1268 }, { "epoch": 0.5971764705882353, "grad_norm": 0.32982212308904496, "learning_rate": 0.00019034999999999996, "loss": 0.25, "step": 1269 }, { "epoch": 0.5976470588235294, "grad_norm": 0.35101097400202624, "learning_rate": 0.0001905, "loss": 0.25, "step": 1270 }, { "epoch": 0.5981176470588235, "grad_norm": 0.45348633309817216, "learning_rate": 0.00019064999999999997, "loss": 0.373, "step": 1271 }, { "epoch": 0.5985882352941176, "grad_norm": 0.39133099961123147, "learning_rate": 0.00019079999999999998, "loss": 0.3262, "step": 1272 }, { "epoch": 0.5990588235294118, "grad_norm": 0.2642267253373357, "learning_rate": 0.00019094999999999998, "loss": 0.1455, "step": 1273 }, { "epoch": 0.5995294117647059, "grad_norm": 0.20687632107962559, "learning_rate": 0.00019109999999999998, "loss": 0.1855, "step": 1274 }, { "epoch": 0.6, "grad_norm": 0.06549905669070386, "learning_rate": 0.00019124999999999996, "loss": 0.0186, "step": 1275 }, { "epoch": 0.6004705882352941, "grad_norm": 0.564776907893941, "learning_rate": 0.0001914, "loss": 0.3086, "step": 1276 }, { "epoch": 0.6009411764705882, "grad_norm": 0.7499842055727834, "learning_rate": 0.00019154999999999997, "loss": 0.5352, "step": 1277 }, { "epoch": 0.6014117647058823, "grad_norm": 0.44073758844217265, "learning_rate": 0.0001917, "loss": 0.625, "step": 1278 }, { "epoch": 0.6018823529411764, "grad_norm": 0.4184020594037526, "learning_rate": 0.00019184999999999997, "loss": 0.3672, "step": 1279 }, { "epoch": 0.6023529411764705, "grad_norm": 0.04292379224165461, "learning_rate": 0.00019199999999999998, "loss": 0.0106, "step": 1280 }, { "epoch": 0.6028235294117648, "grad_norm": 0.662280138220446, "learning_rate": 0.00019214999999999998, "loss": 0.3809, "step": 1281 }, { "epoch": 0.6032941176470589, "grad_norm": 1.0773997521302827, "learning_rate": 0.00019229999999999999, "loss": 0.6445, "step": 1282 }, { "epoch": 0.603764705882353, "grad_norm": 0.9222320924510616, "learning_rate": 0.00019244999999999996, "loss": 0.5742, "step": 1283 }, { "epoch": 0.6042352941176471, "grad_norm": 0.15993095499729082, "learning_rate": 0.0001926, "loss": 0.1006, "step": 1284 }, { "epoch": 0.6047058823529412, "grad_norm": 0.4797516284086731, "learning_rate": 0.00019274999999999997, "loss": 0.2432, "step": 1285 }, { "epoch": 0.6051764705882353, "grad_norm": 0.1400703342026864, "learning_rate": 0.0001929, "loss": 0.0488, "step": 1286 }, { "epoch": 0.6056470588235294, "grad_norm": 0.9765273768811992, "learning_rate": 0.00019304999999999998, "loss": 0.5391, "step": 1287 }, { "epoch": 0.6061176470588235, "grad_norm": 0.7607782145859179, "learning_rate": 0.00019319999999999998, "loss": 0.5664, "step": 1288 }, { "epoch": 0.6065882352941177, "grad_norm": 0.6447488275776124, "learning_rate": 0.00019334999999999998, "loss": 0.3984, "step": 1289 }, { "epoch": 0.6070588235294118, "grad_norm": 0.8088472844794387, "learning_rate": 0.0001935, "loss": 0.8828, "step": 1290 }, { "epoch": 0.6075294117647059, "grad_norm": 0.07170586951771434, "learning_rate": 0.00019364999999999996, "loss": 0.0325, "step": 1291 }, { "epoch": 0.608, "grad_norm": 0.48298185059166254, "learning_rate": 0.0001938, "loss": 0.3047, "step": 1292 }, { "epoch": 0.6084705882352941, "grad_norm": 0.349743274820686, "learning_rate": 0.00019394999999999997, "loss": 0.2188, "step": 1293 }, { "epoch": 0.6089411764705882, "grad_norm": 0.2820250683754545, "learning_rate": 0.0001941, "loss": 0.1387, "step": 1294 }, { "epoch": 0.6094117647058823, "grad_norm": 0.5265138966044554, "learning_rate": 0.00019424999999999998, "loss": 0.3438, "step": 1295 }, { "epoch": 0.6098823529411764, "grad_norm": 0.20282412184591347, "learning_rate": 0.00019439999999999998, "loss": 0.0991, "step": 1296 }, { "epoch": 0.6103529411764705, "grad_norm": 0.23790388676967164, "learning_rate": 0.00019454999999999999, "loss": 0.1611, "step": 1297 }, { "epoch": 0.6108235294117647, "grad_norm": 0.3481049153284812, "learning_rate": 0.0001947, "loss": 0.2363, "step": 1298 }, { "epoch": 0.6112941176470589, "grad_norm": 0.6624481540432229, "learning_rate": 0.00019484999999999997, "loss": 0.4199, "step": 1299 }, { "epoch": 0.611764705882353, "grad_norm": 0.26130377151276907, "learning_rate": 0.000195, "loss": 0.2012, "step": 1300 }, { "epoch": 0.6122352941176471, "grad_norm": 0.5130751597005535, "learning_rate": 0.00019514999999999997, "loss": 0.4141, "step": 1301 }, { "epoch": 0.6127058823529412, "grad_norm": 0.3000730098547914, "learning_rate": 0.00019529999999999998, "loss": 0.1514, "step": 1302 }, { "epoch": 0.6131764705882353, "grad_norm": 0.5043455542997886, "learning_rate": 0.00019544999999999998, "loss": 0.4434, "step": 1303 }, { "epoch": 0.6136470588235294, "grad_norm": 0.31627901702261346, "learning_rate": 0.00019559999999999998, "loss": 0.1797, "step": 1304 }, { "epoch": 0.6141176470588235, "grad_norm": 0.49696720998627786, "learning_rate": 0.00019574999999999996, "loss": 0.3555, "step": 1305 }, { "epoch": 0.6145882352941177, "grad_norm": 0.6065170871875628, "learning_rate": 0.0001959, "loss": 0.377, "step": 1306 }, { "epoch": 0.6150588235294118, "grad_norm": 0.31649758173222076, "learning_rate": 0.00019604999999999997, "loss": 0.168, "step": 1307 }, { "epoch": 0.6155294117647059, "grad_norm": 0.5811382880478959, "learning_rate": 0.0001962, "loss": 0.7891, "step": 1308 }, { "epoch": 0.616, "grad_norm": 0.08494047408478438, "learning_rate": 0.00019634999999999998, "loss": 0.019, "step": 1309 }, { "epoch": 0.6164705882352941, "grad_norm": 0.4575658428953346, "learning_rate": 0.00019649999999999998, "loss": 0.3965, "step": 1310 }, { "epoch": 0.6169411764705882, "grad_norm": 0.612256453766549, "learning_rate": 0.00019664999999999998, "loss": 0.3984, "step": 1311 }, { "epoch": 0.6174117647058823, "grad_norm": 0.19905294082363764, "learning_rate": 0.00019679999999999999, "loss": 0.0654, "step": 1312 }, { "epoch": 0.6178823529411764, "grad_norm": 0.29096201984441616, "learning_rate": 0.00019694999999999996, "loss": 0.1406, "step": 1313 }, { "epoch": 0.6183529411764705, "grad_norm": 0.6314519585436147, "learning_rate": 0.0001971, "loss": 0.7891, "step": 1314 }, { "epoch": 0.6188235294117647, "grad_norm": 0.4293635313499881, "learning_rate": 0.00019724999999999997, "loss": 0.4473, "step": 1315 }, { "epoch": 0.6192941176470588, "grad_norm": 0.6704754709170613, "learning_rate": 0.0001974, "loss": 0.4141, "step": 1316 }, { "epoch": 0.619764705882353, "grad_norm": 0.4212709553270446, "learning_rate": 0.00019754999999999998, "loss": 0.2852, "step": 1317 }, { "epoch": 0.6202352941176471, "grad_norm": 0.4217144326718646, "learning_rate": 0.00019769999999999998, "loss": 0.3672, "step": 1318 }, { "epoch": 0.6207058823529412, "grad_norm": 0.35139718411251497, "learning_rate": 0.00019784999999999998, "loss": 0.3223, "step": 1319 }, { "epoch": 0.6211764705882353, "grad_norm": 0.26484951784071453, "learning_rate": 0.000198, "loss": 0.2148, "step": 1320 }, { "epoch": 0.6216470588235294, "grad_norm": 0.22523786082091138, "learning_rate": 0.00019814999999999996, "loss": 0.1045, "step": 1321 }, { "epoch": 0.6221176470588236, "grad_norm": 0.27954723098559375, "learning_rate": 0.0001983, "loss": 0.1118, "step": 1322 }, { "epoch": 0.6225882352941177, "grad_norm": 0.31888406411386383, "learning_rate": 0.00019844999999999997, "loss": 0.1387, "step": 1323 }, { "epoch": 0.6230588235294118, "grad_norm": 0.15285975379671865, "learning_rate": 0.0001986, "loss": 0.1079, "step": 1324 }, { "epoch": 0.6235294117647059, "grad_norm": 0.2716829664621721, "learning_rate": 0.00019874999999999998, "loss": 0.1777, "step": 1325 }, { "epoch": 0.624, "grad_norm": 0.40251634994703417, "learning_rate": 0.00019889999999999998, "loss": 0.3008, "step": 1326 }, { "epoch": 0.6244705882352941, "grad_norm": 1.0970754413113506, "learning_rate": 0.00019905, "loss": 1.2656, "step": 1327 }, { "epoch": 0.6249411764705882, "grad_norm": 0.12392048401151869, "learning_rate": 0.0001992, "loss": 0.04, "step": 1328 }, { "epoch": 0.6254117647058823, "grad_norm": 0.33019006408187773, "learning_rate": 0.00019934999999999997, "loss": 0.1836, "step": 1329 }, { "epoch": 0.6258823529411764, "grad_norm": 0.34143532444449237, "learning_rate": 0.0001995, "loss": 0.1279, "step": 1330 }, { "epoch": 0.6263529411764706, "grad_norm": 0.5471859738568038, "learning_rate": 0.00019964999999999997, "loss": 0.249, "step": 1331 }, { "epoch": 0.6268235294117647, "grad_norm": 0.16716977779205172, "learning_rate": 0.0001998, "loss": 0.1055, "step": 1332 }, { "epoch": 0.6272941176470588, "grad_norm": 0.41954846163651505, "learning_rate": 0.00019994999999999998, "loss": 0.2988, "step": 1333 }, { "epoch": 0.6277647058823529, "grad_norm": 0.9736538196479534, "learning_rate": 0.00020009999999999998, "loss": 0.5508, "step": 1334 }, { "epoch": 0.6282352941176471, "grad_norm": 0.20346601471100464, "learning_rate": 0.00020025, "loss": 0.1445, "step": 1335 }, { "epoch": 0.6287058823529412, "grad_norm": 0.7121282666808582, "learning_rate": 0.0002004, "loss": 0.334, "step": 1336 }, { "epoch": 0.6291764705882353, "grad_norm": 7.126889799730739, "learning_rate": 0.00020054999999999997, "loss": 0.6445, "step": 1337 }, { "epoch": 0.6296470588235294, "grad_norm": 6.6876995985061285, "learning_rate": 0.0002007, "loss": 0.7461, "step": 1338 }, { "epoch": 0.6301176470588236, "grad_norm": 0.5175264655332582, "learning_rate": 0.00020084999999999998, "loss": 0.2754, "step": 1339 }, { "epoch": 0.6305882352941177, "grad_norm": 0.3200292424425301, "learning_rate": 0.000201, "loss": 0.1709, "step": 1340 }, { "epoch": 0.6310588235294118, "grad_norm": 0.9955667967418849, "learning_rate": 0.00020114999999999998, "loss": 0.3008, "step": 1341 }, { "epoch": 0.6315294117647059, "grad_norm": 0.5501315942166686, "learning_rate": 0.0002013, "loss": 0.2148, "step": 1342 }, { "epoch": 0.632, "grad_norm": 0.7474186678471962, "learning_rate": 0.00020145, "loss": 0.3418, "step": 1343 }, { "epoch": 0.6324705882352941, "grad_norm": 43.10914571860515, "learning_rate": 0.0002016, "loss": 0.1836, "step": 1344 }, { "epoch": 0.6329411764705882, "grad_norm": 82.00502190268998, "learning_rate": 0.00020174999999999997, "loss": 6.0, "step": 1345 }, { "epoch": 0.6334117647058823, "grad_norm": 13.100635782913484, "learning_rate": 0.0002019, "loss": 9.375, "step": 1346 }, { "epoch": 0.6338823529411765, "grad_norm": 25.451704369773207, "learning_rate": 0.00020204999999999998, "loss": 30.0, "step": 1347 }, { "epoch": 0.6343529411764706, "grad_norm": 42.3037474884192, "learning_rate": 0.0002022, "loss": 39.75, "step": 1348 }, { "epoch": 0.6348235294117647, "grad_norm": 34.95298801691623, "learning_rate": 0.00020234999999999999, "loss": 17.0, "step": 1349 }, { "epoch": 0.6352941176470588, "grad_norm": 36.21601108480889, "learning_rate": 0.0002025, "loss": 14.8125, "step": 1350 }, { "epoch": 0.6357647058823529, "grad_norm": 32.96373930309115, "learning_rate": 0.00020264999999999997, "loss": 14.0625, "step": 1351 }, { "epoch": 0.636235294117647, "grad_norm": 39.547161836197496, "learning_rate": 0.0002028, "loss": 4.625, "step": 1352 }, { "epoch": 0.6367058823529411, "grad_norm": 7.424337119724559, "learning_rate": 0.00020294999999999997, "loss": 7.3438, "step": 1353 }, { "epoch": 0.6371764705882353, "grad_norm": 20.61234329581765, "learning_rate": 0.0002031, "loss": 2.5, "step": 1354 }, { "epoch": 0.6376470588235295, "grad_norm": 5.452547525349031, "learning_rate": 0.00020324999999999998, "loss": 2.4531, "step": 1355 }, { "epoch": 0.6381176470588236, "grad_norm": 22.92388110458262, "learning_rate": 0.00020339999999999998, "loss": 2.2656, "step": 1356 }, { "epoch": 0.6385882352941177, "grad_norm": 42.26833571302864, "learning_rate": 0.00020355, "loss": 3.4844, "step": 1357 }, { "epoch": 0.6390588235294118, "grad_norm": 9.321998055371223, "learning_rate": 0.0002037, "loss": 2.1094, "step": 1358 }, { "epoch": 0.6395294117647059, "grad_norm": 8.075313487464898, "learning_rate": 0.00020384999999999997, "loss": 1.5547, "step": 1359 }, { "epoch": 0.64, "grad_norm": 11.574273344605505, "learning_rate": 0.000204, "loss": 0.9492, "step": 1360 }, { "epoch": 0.6404705882352941, "grad_norm": 7.9023684779460694, "learning_rate": 0.00020414999999999997, "loss": 3.9062, "step": 1361 }, { "epoch": 0.6409411764705882, "grad_norm": 4.372233435759768, "learning_rate": 0.0002043, "loss": 1.3359, "step": 1362 }, { "epoch": 0.6414117647058823, "grad_norm": 8.826984445876825, "learning_rate": 0.00020444999999999998, "loss": 2.5156, "step": 1363 }, { "epoch": 0.6418823529411765, "grad_norm": 8.214429801776024, "learning_rate": 0.00020459999999999999, "loss": 2.5156, "step": 1364 }, { "epoch": 0.6423529411764706, "grad_norm": 8.219088023452167, "learning_rate": 0.00020475, "loss": 2.5938, "step": 1365 }, { "epoch": 0.6428235294117647, "grad_norm": 15.879187137612112, "learning_rate": 0.0002049, "loss": 2.4219, "step": 1366 }, { "epoch": 0.6432941176470588, "grad_norm": 2.278063194282542, "learning_rate": 0.00020504999999999997, "loss": 0.8711, "step": 1367 }, { "epoch": 0.6437647058823529, "grad_norm": 13.200729196568528, "learning_rate": 0.0002052, "loss": 2.8281, "step": 1368 }, { "epoch": 0.644235294117647, "grad_norm": 7.924187587120109, "learning_rate": 0.00020534999999999998, "loss": 4.0938, "step": 1369 }, { "epoch": 0.6447058823529411, "grad_norm": 10.983052120242826, "learning_rate": 0.0002055, "loss": 2.3125, "step": 1370 }, { "epoch": 0.6451764705882352, "grad_norm": 15.66919749282821, "learning_rate": 0.00020564999999999998, "loss": 1.5078, "step": 1371 }, { "epoch": 0.6456470588235295, "grad_norm": 19.114035068966256, "learning_rate": 0.0002058, "loss": 1.4531, "step": 1372 }, { "epoch": 0.6461176470588236, "grad_norm": 8.405883068806304, "learning_rate": 0.00020595, "loss": 1.8672, "step": 1373 }, { "epoch": 0.6465882352941177, "grad_norm": 8.356345060709378, "learning_rate": 0.0002061, "loss": 1.75, "step": 1374 }, { "epoch": 0.6470588235294118, "grad_norm": 5.092550116348749, "learning_rate": 0.00020624999999999997, "loss": 1.1406, "step": 1375 }, { "epoch": 0.6475294117647059, "grad_norm": 15.39239328932158, "learning_rate": 0.00020639999999999998, "loss": 3.0469, "step": 1376 }, { "epoch": 0.648, "grad_norm": 6.612409749034143, "learning_rate": 0.00020654999999999998, "loss": 2.5, "step": 1377 }, { "epoch": 0.6484705882352941, "grad_norm": 41.22834245121602, "learning_rate": 0.00020669999999999996, "loss": 2.5469, "step": 1378 }, { "epoch": 0.6489411764705882, "grad_norm": 14.045831536971543, "learning_rate": 0.00020684999999999999, "loss": 0.9023, "step": 1379 }, { "epoch": 0.6494117647058824, "grad_norm": 15.125447225779837, "learning_rate": 0.00020699999999999996, "loss": 3.1562, "step": 1380 }, { "epoch": 0.6498823529411765, "grad_norm": 12.606462272278717, "learning_rate": 0.00020715, "loss": 2.7656, "step": 1381 }, { "epoch": 0.6503529411764706, "grad_norm": 4.743223596898047, "learning_rate": 0.00020729999999999997, "loss": 1.0234, "step": 1382 }, { "epoch": 0.6508235294117647, "grad_norm": 6.662152233078737, "learning_rate": 0.00020744999999999997, "loss": 1.3828, "step": 1383 }, { "epoch": 0.6512941176470588, "grad_norm": 11.302909604486846, "learning_rate": 0.00020759999999999998, "loss": 2.3594, "step": 1384 }, { "epoch": 0.6517647058823529, "grad_norm": 9.184770018283492, "learning_rate": 0.00020774999999999998, "loss": 0.7266, "step": 1385 }, { "epoch": 0.652235294117647, "grad_norm": 21.835638354329273, "learning_rate": 0.00020789999999999996, "loss": 2.8125, "step": 1386 }, { "epoch": 0.6527058823529411, "grad_norm": 5.661353813716896, "learning_rate": 0.00020805, "loss": 0.2871, "step": 1387 }, { "epoch": 0.6531764705882352, "grad_norm": 19.129168872518928, "learning_rate": 0.00020819999999999996, "loss": 3.4219, "step": 1388 }, { "epoch": 0.6536470588235294, "grad_norm": 13.627631102720422, "learning_rate": 0.00020835, "loss": 2.6562, "step": 1389 }, { "epoch": 0.6541176470588236, "grad_norm": 4.331199191218265, "learning_rate": 0.00020849999999999997, "loss": 1.1172, "step": 1390 }, { "epoch": 0.6545882352941177, "grad_norm": 8.125456196539671, "learning_rate": 0.00020864999999999998, "loss": 1.8047, "step": 1391 }, { "epoch": 0.6550588235294118, "grad_norm": 24.87723420576104, "learning_rate": 0.00020879999999999998, "loss": 1.8281, "step": 1392 }, { "epoch": 0.6555294117647059, "grad_norm": 7.739430729072587, "learning_rate": 0.00020894999999999998, "loss": 0.4766, "step": 1393 }, { "epoch": 0.656, "grad_norm": 10.43487877087986, "learning_rate": 0.00020909999999999996, "loss": 1.75, "step": 1394 }, { "epoch": 0.6564705882352941, "grad_norm": 9.687269426810905, "learning_rate": 0.00020925, "loss": 2.1562, "step": 1395 }, { "epoch": 0.6569411764705883, "grad_norm": 22.414496710587315, "learning_rate": 0.00020939999999999997, "loss": 4.7188, "step": 1396 }, { "epoch": 0.6574117647058824, "grad_norm": 10.825621711221826, "learning_rate": 0.00020955, "loss": 1.9297, "step": 1397 }, { "epoch": 0.6578823529411765, "grad_norm": 21.468875044707715, "learning_rate": 0.00020969999999999997, "loss": 1.9453, "step": 1398 }, { "epoch": 0.6583529411764706, "grad_norm": 40.12203967325378, "learning_rate": 0.00020984999999999998, "loss": 1.8359, "step": 1399 }, { "epoch": 0.6588235294117647, "grad_norm": 5.270029629976205, "learning_rate": 0.00020999999999999998, "loss": 1.2891, "step": 1400 }, { "epoch": 0.6592941176470588, "grad_norm": 14.31125368030509, "learning_rate": 0.00021014999999999999, "loss": 2.8125, "step": 1401 }, { "epoch": 0.6597647058823529, "grad_norm": 15.574996454065094, "learning_rate": 0.00021029999999999996, "loss": 3.25, "step": 1402 }, { "epoch": 0.660235294117647, "grad_norm": 7.438463457680565, "learning_rate": 0.00021045, "loss": 1.4141, "step": 1403 }, { "epoch": 0.6607058823529411, "grad_norm": 4.533077878744562, "learning_rate": 0.00021059999999999997, "loss": 2.125, "step": 1404 }, { "epoch": 0.6611764705882353, "grad_norm": 12.356421893470163, "learning_rate": 0.00021074999999999997, "loss": 2.3281, "step": 1405 }, { "epoch": 0.6616470588235294, "grad_norm": 7.178462479049681, "learning_rate": 0.00021089999999999998, "loss": 1.5234, "step": 1406 }, { "epoch": 0.6621176470588235, "grad_norm": 34.14996214623403, "learning_rate": 0.00021104999999999998, "loss": 1.9609, "step": 1407 }, { "epoch": 0.6625882352941177, "grad_norm": 6.299923940487153, "learning_rate": 0.00021119999999999996, "loss": 1.375, "step": 1408 }, { "epoch": 0.6630588235294118, "grad_norm": 0.2656552737096778, "learning_rate": 0.00021135, "loss": 0.2129, "step": 1409 }, { "epoch": 0.6635294117647059, "grad_norm": 13.749957882063212, "learning_rate": 0.00021149999999999996, "loss": 2.9531, "step": 1410 }, { "epoch": 0.664, "grad_norm": 10.510648636570577, "learning_rate": 0.00021165, "loss": 1.75, "step": 1411 }, { "epoch": 0.6644705882352941, "grad_norm": 4.715012091018497, "learning_rate": 0.00021179999999999997, "loss": 0.8398, "step": 1412 }, { "epoch": 0.6649411764705883, "grad_norm": 12.04925376748768, "learning_rate": 0.00021194999999999997, "loss": 1.7656, "step": 1413 }, { "epoch": 0.6654117647058824, "grad_norm": 10.108324071430902, "learning_rate": 0.00021209999999999998, "loss": 1.7578, "step": 1414 }, { "epoch": 0.6658823529411765, "grad_norm": 24.837855800591527, "learning_rate": 0.00021224999999999998, "loss": 2.0156, "step": 1415 }, { "epoch": 0.6663529411764706, "grad_norm": 31.218751369824044, "learning_rate": 0.00021239999999999996, "loss": 1.1562, "step": 1416 }, { "epoch": 0.6668235294117647, "grad_norm": 1.3352886504784927, "learning_rate": 0.00021255, "loss": 1.0625, "step": 1417 }, { "epoch": 0.6672941176470588, "grad_norm": 12.853032766448415, "learning_rate": 0.00021269999999999997, "loss": 1.6484, "step": 1418 }, { "epoch": 0.6677647058823529, "grad_norm": 2.461292528782212, "learning_rate": 0.00021285, "loss": 0.4414, "step": 1419 }, { "epoch": 0.668235294117647, "grad_norm": 6.901824475071449, "learning_rate": 0.00021299999999999997, "loss": 0.8672, "step": 1420 }, { "epoch": 0.6687058823529411, "grad_norm": 4.339001637137442, "learning_rate": 0.00021314999999999998, "loss": 0.8203, "step": 1421 }, { "epoch": 0.6691764705882353, "grad_norm": 2.9721678590957388, "learning_rate": 0.00021329999999999998, "loss": 1.6719, "step": 1422 }, { "epoch": 0.6696470588235294, "grad_norm": 3.1249287914358566, "learning_rate": 0.00021344999999999998, "loss": 0.6289, "step": 1423 }, { "epoch": 0.6701176470588235, "grad_norm": 1.6961190810850328, "learning_rate": 0.00021359999999999996, "loss": 1.4062, "step": 1424 }, { "epoch": 0.6705882352941176, "grad_norm": 0.6490130015064681, "learning_rate": 0.00021375, "loss": 1.1562, "step": 1425 }, { "epoch": 0.6710588235294118, "grad_norm": 0.8849542098332686, "learning_rate": 0.00021389999999999997, "loss": 0.8203, "step": 1426 }, { "epoch": 0.6715294117647059, "grad_norm": 1.7689897417532967, "learning_rate": 0.00021405, "loss": 1.2344, "step": 1427 }, { "epoch": 0.672, "grad_norm": 1.4813454896524119, "learning_rate": 0.00021419999999999998, "loss": 1.4844, "step": 1428 }, { "epoch": 0.6724705882352942, "grad_norm": 0.9100796301448788, "learning_rate": 0.00021434999999999998, "loss": 1.4922, "step": 1429 }, { "epoch": 0.6729411764705883, "grad_norm": 2.2270407790206157, "learning_rate": 0.00021449999999999998, "loss": 0.8086, "step": 1430 }, { "epoch": 0.6734117647058824, "grad_norm": 0.5148123789770275, "learning_rate": 0.00021464999999999999, "loss": 0.416, "step": 1431 }, { "epoch": 0.6738823529411765, "grad_norm": 1.3299363536368851, "learning_rate": 0.00021479999999999996, "loss": 1.0, "step": 1432 }, { "epoch": 0.6743529411764706, "grad_norm": 1.451025325920524, "learning_rate": 0.00021495, "loss": 0.9883, "step": 1433 }, { "epoch": 0.6748235294117647, "grad_norm": 0.6139694396784833, "learning_rate": 0.00021509999999999997, "loss": 0.4258, "step": 1434 }, { "epoch": 0.6752941176470588, "grad_norm": 0.8541096478317249, "learning_rate": 0.00021525, "loss": 0.6914, "step": 1435 }, { "epoch": 0.6757647058823529, "grad_norm": 1.1681689648851223, "learning_rate": 0.00021539999999999998, "loss": 1.4688, "step": 1436 }, { "epoch": 0.676235294117647, "grad_norm": 1.5841710948122811, "learning_rate": 0.00021554999999999998, "loss": 1.3047, "step": 1437 }, { "epoch": 0.6767058823529412, "grad_norm": 0.3897976206289636, "learning_rate": 0.00021569999999999998, "loss": 0.0869, "step": 1438 }, { "epoch": 0.6771764705882353, "grad_norm": 0.8344836160470648, "learning_rate": 0.00021585, "loss": 1.0312, "step": 1439 }, { "epoch": 0.6776470588235294, "grad_norm": 1.2474742282031905, "learning_rate": 0.00021599999999999996, "loss": 1.4375, "step": 1440 }, { "epoch": 0.6781176470588235, "grad_norm": 0.8863294307937158, "learning_rate": 0.00021615, "loss": 0.8359, "step": 1441 }, { "epoch": 0.6785882352941176, "grad_norm": 1.1984894162812867, "learning_rate": 0.00021629999999999997, "loss": 0.957, "step": 1442 }, { "epoch": 0.6790588235294117, "grad_norm": 1.67005935103557, "learning_rate": 0.00021645, "loss": 1.2812, "step": 1443 }, { "epoch": 0.6795294117647059, "grad_norm": 0.2986655764204079, "learning_rate": 0.00021659999999999998, "loss": 0.3066, "step": 1444 }, { "epoch": 0.68, "grad_norm": 1.2413663911302026, "learning_rate": 0.00021674999999999998, "loss": 2.9062, "step": 1445 }, { "epoch": 0.6804705882352942, "grad_norm": 0.5032345294381938, "learning_rate": 0.0002169, "loss": 0.543, "step": 1446 }, { "epoch": 0.6809411764705883, "grad_norm": 0.4439725525596708, "learning_rate": 0.00021705, "loss": 0.3184, "step": 1447 }, { "epoch": 0.6814117647058824, "grad_norm": 0.6975366058160798, "learning_rate": 0.00021719999999999997, "loss": 0.7148, "step": 1448 }, { "epoch": 0.6818823529411765, "grad_norm": 0.5363618237998186, "learning_rate": 0.00021735, "loss": 0.6211, "step": 1449 }, { "epoch": 0.6823529411764706, "grad_norm": 0.355201378732584, "learning_rate": 0.00021749999999999997, "loss": 0.4648, "step": 1450 }, { "epoch": 0.6828235294117647, "grad_norm": 0.5246496029905264, "learning_rate": 0.00021764999999999998, "loss": 0.7344, "step": 1451 }, { "epoch": 0.6832941176470588, "grad_norm": 1.2524112507430671, "learning_rate": 0.00021779999999999998, "loss": 0.7617, "step": 1452 }, { "epoch": 0.6837647058823529, "grad_norm": 0.45603512147135716, "learning_rate": 0.00021794999999999999, "loss": 0.3379, "step": 1453 }, { "epoch": 0.684235294117647, "grad_norm": 0.6685114519004622, "learning_rate": 0.00021809999999999996, "loss": 0.8711, "step": 1454 }, { "epoch": 0.6847058823529412, "grad_norm": 0.6758653683661042, "learning_rate": 0.00021825, "loss": 0.9766, "step": 1455 }, { "epoch": 0.6851764705882353, "grad_norm": 0.939312565281696, "learning_rate": 0.00021839999999999997, "loss": 1.4609, "step": 1456 }, { "epoch": 0.6856470588235294, "grad_norm": 0.9104641914080901, "learning_rate": 0.00021855, "loss": 1.0312, "step": 1457 }, { "epoch": 0.6861176470588235, "grad_norm": 0.5125324784888075, "learning_rate": 0.00021869999999999998, "loss": 0.1543, "step": 1458 }, { "epoch": 0.6865882352941176, "grad_norm": 0.2886433920958697, "learning_rate": 0.00021884999999999998, "loss": 0.1924, "step": 1459 }, { "epoch": 0.6870588235294117, "grad_norm": 0.28445965404021395, "learning_rate": 0.00021899999999999998, "loss": 0.2852, "step": 1460 }, { "epoch": 0.6875294117647058, "grad_norm": 0.3930389426529099, "learning_rate": 0.00021915, "loss": 0.4453, "step": 1461 }, { "epoch": 0.688, "grad_norm": 1.0324489455281982, "learning_rate": 0.00021929999999999996, "loss": 0.3008, "step": 1462 }, { "epoch": 0.6884705882352942, "grad_norm": 1.2399122768835715, "learning_rate": 0.00021945, "loss": 1.4219, "step": 1463 }, { "epoch": 0.6889411764705883, "grad_norm": 0.6855905358760419, "learning_rate": 0.00021959999999999997, "loss": 0.6602, "step": 1464 }, { "epoch": 0.6894117647058824, "grad_norm": 1.2583793732845265, "learning_rate": 0.00021975, "loss": 0.4922, "step": 1465 }, { "epoch": 0.6898823529411765, "grad_norm": 3.062299331161181, "learning_rate": 0.00021989999999999998, "loss": 0.3652, "step": 1466 }, { "epoch": 0.6903529411764706, "grad_norm": 0.1340182512802856, "learning_rate": 0.00022004999999999998, "loss": 0.0669, "step": 1467 }, { "epoch": 0.6908235294117647, "grad_norm": 0.4336462865070905, "learning_rate": 0.00022019999999999999, "loss": 0.3691, "step": 1468 }, { "epoch": 0.6912941176470588, "grad_norm": 0.5419374134736205, "learning_rate": 0.00022035, "loss": 0.5234, "step": 1469 }, { "epoch": 0.691764705882353, "grad_norm": 0.5438094469518234, "learning_rate": 0.00022049999999999997, "loss": 0.7383, "step": 1470 }, { "epoch": 0.6922352941176471, "grad_norm": 0.8786148260460748, "learning_rate": 0.00022065, "loss": 0.6523, "step": 1471 }, { "epoch": 0.6927058823529412, "grad_norm": 0.9374668689499195, "learning_rate": 0.00022079999999999997, "loss": 0.5078, "step": 1472 }, { "epoch": 0.6931764705882353, "grad_norm": 0.6657167714551305, "learning_rate": 0.00022095, "loss": 0.8164, "step": 1473 }, { "epoch": 0.6936470588235294, "grad_norm": 1.2747282483422417, "learning_rate": 0.00022109999999999998, "loss": 0.6289, "step": 1474 }, { "epoch": 0.6941176470588235, "grad_norm": 0.7243029453332123, "learning_rate": 0.00022124999999999998, "loss": 1.2734, "step": 1475 }, { "epoch": 0.6945882352941176, "grad_norm": 1.1132091732915435, "learning_rate": 0.0002214, "loss": 0.6328, "step": 1476 }, { "epoch": 0.6950588235294117, "grad_norm": 0.7734133044765937, "learning_rate": 0.00022155, "loss": 0.5703, "step": 1477 }, { "epoch": 0.6955294117647058, "grad_norm": 0.5429893328181674, "learning_rate": 0.00022169999999999997, "loss": 0.4355, "step": 1478 }, { "epoch": 0.696, "grad_norm": 0.5457564203004177, "learning_rate": 0.00022185, "loss": 0.5742, "step": 1479 }, { "epoch": 0.6964705882352941, "grad_norm": 1.0436174331711712, "learning_rate": 0.00022199999999999998, "loss": 0.5781, "step": 1480 }, { "epoch": 0.6969411764705883, "grad_norm": 0.7302006812312426, "learning_rate": 0.00022215, "loss": 0.6016, "step": 1481 }, { "epoch": 0.6974117647058824, "grad_norm": 0.31515859209130465, "learning_rate": 0.00022229999999999998, "loss": 0.1826, "step": 1482 }, { "epoch": 0.6978823529411765, "grad_norm": 0.2673469852160898, "learning_rate": 0.00022244999999999999, "loss": 0.1875, "step": 1483 }, { "epoch": 0.6983529411764706, "grad_norm": 0.6092174557832118, "learning_rate": 0.0002226, "loss": 0.4473, "step": 1484 }, { "epoch": 0.6988235294117647, "grad_norm": 1.410290652894044, "learning_rate": 0.00022275, "loss": 0.9492, "step": 1485 }, { "epoch": 0.6992941176470588, "grad_norm": 15.290923389307455, "learning_rate": 0.00022289999999999997, "loss": 0.5469, "step": 1486 }, { "epoch": 0.699764705882353, "grad_norm": 0.39998252443432786, "learning_rate": 0.00022305, "loss": 0.1777, "step": 1487 }, { "epoch": 0.7002352941176471, "grad_norm": 83.99252139687047, "learning_rate": 0.00022319999999999998, "loss": 9.75, "step": 1488 }, { "epoch": 0.7007058823529412, "grad_norm": 15.858491424206683, "learning_rate": 0.00022335, "loss": 2.5312, "step": 1489 }, { "epoch": 0.7011764705882353, "grad_norm": 10.464685111377275, "learning_rate": 0.00022349999999999998, "loss": 2.5625, "step": 1490 }, { "epoch": 0.7016470588235294, "grad_norm": 5.89095625703063, "learning_rate": 0.00022365, "loss": 1.0625, "step": 1491 }, { "epoch": 0.7021176470588235, "grad_norm": 2.7498283391336185, "learning_rate": 0.0002238, "loss": 0.4492, "step": 1492 }, { "epoch": 0.7025882352941176, "grad_norm": 28.84939104128909, "learning_rate": 0.00022395, "loss": 1.1016, "step": 1493 }, { "epoch": 0.7030588235294117, "grad_norm": 5.493660920616416, "learning_rate": 0.00022409999999999997, "loss": 0.5859, "step": 1494 }, { "epoch": 0.7035294117647058, "grad_norm": 0.6002092704571573, "learning_rate": 0.00022425, "loss": 0.2178, "step": 1495 }, { "epoch": 0.704, "grad_norm": 2.9124156160791212, "learning_rate": 0.00022439999999999998, "loss": 1.1016, "step": 1496 }, { "epoch": 0.7044705882352941, "grad_norm": 4.057643097869166, "learning_rate": 0.00022455, "loss": 0.4688, "step": 1497 }, { "epoch": 0.7049411764705882, "grad_norm": 1.9967250522228046, "learning_rate": 0.0002247, "loss": 0.4297, "step": 1498 }, { "epoch": 0.7054117647058824, "grad_norm": 9.514226392497935, "learning_rate": 0.00022485, "loss": 1.3828, "step": 1499 }, { "epoch": 0.7058823529411765, "grad_norm": 6.972244849179647, "learning_rate": 0.000225, "loss": 1.4844, "step": 1500 }, { "epoch": 0.7063529411764706, "grad_norm": 4.295813571370161, "learning_rate": 0.00022514999999999997, "loss": 1.0859, "step": 1501 }, { "epoch": 0.7068235294117647, "grad_norm": 2.9740554334363623, "learning_rate": 0.00022529999999999997, "loss": 0.7578, "step": 1502 }, { "epoch": 0.7072941176470589, "grad_norm": 1.2309616969062167, "learning_rate": 0.00022544999999999995, "loss": 0.8398, "step": 1503 }, { "epoch": 0.707764705882353, "grad_norm": 3.4710835833406097, "learning_rate": 0.00022559999999999998, "loss": 1.0469, "step": 1504 }, { "epoch": 0.7082352941176471, "grad_norm": 1.0447151228478282, "learning_rate": 0.00022574999999999996, "loss": 0.3789, "step": 1505 }, { "epoch": 0.7087058823529412, "grad_norm": 2.3392788622795226, "learning_rate": 0.0002259, "loss": 0.6055, "step": 1506 }, { "epoch": 0.7091764705882353, "grad_norm": 2.0176856830194523, "learning_rate": 0.00022604999999999997, "loss": 0.1182, "step": 1507 }, { "epoch": 0.7096470588235294, "grad_norm": 2.4278201534675765, "learning_rate": 0.00022619999999999997, "loss": 0.3477, "step": 1508 }, { "epoch": 0.7101176470588235, "grad_norm": 1.1208615184292465, "learning_rate": 0.00022634999999999997, "loss": 0.1582, "step": 1509 }, { "epoch": 0.7105882352941176, "grad_norm": 2.936124588636442, "learning_rate": 0.00022649999999999998, "loss": 0.6484, "step": 1510 }, { "epoch": 0.7110588235294117, "grad_norm": 6.1063424890546205, "learning_rate": 0.00022664999999999995, "loss": 1.2812, "step": 1511 }, { "epoch": 0.7115294117647059, "grad_norm": 1.9460615207290934, "learning_rate": 0.00022679999999999998, "loss": 0.543, "step": 1512 }, { "epoch": 0.712, "grad_norm": 1.92426234371515, "learning_rate": 0.00022694999999999996, "loss": 1.0859, "step": 1513 }, { "epoch": 0.7124705882352941, "grad_norm": 1.8111423709548853, "learning_rate": 0.0002271, "loss": 0.9414, "step": 1514 }, { "epoch": 0.7129411764705882, "grad_norm": 0.7047006409198756, "learning_rate": 0.00022724999999999997, "loss": 0.3027, "step": 1515 }, { "epoch": 0.7134117647058823, "grad_norm": 0.3903527183758082, "learning_rate": 0.00022739999999999997, "loss": 0.1201, "step": 1516 }, { "epoch": 0.7138823529411765, "grad_norm": 2.3252097230737667, "learning_rate": 0.00022754999999999997, "loss": 0.6289, "step": 1517 }, { "epoch": 0.7143529411764706, "grad_norm": 1.6487314263864021, "learning_rate": 0.00022769999999999998, "loss": 0.6875, "step": 1518 }, { "epoch": 0.7148235294117647, "grad_norm": 1.744643116069344, "learning_rate": 0.00022784999999999995, "loss": 0.9688, "step": 1519 }, { "epoch": 0.7152941176470589, "grad_norm": 1.019682359305576, "learning_rate": 0.00022799999999999999, "loss": 0.3691, "step": 1520 }, { "epoch": 0.715764705882353, "grad_norm": 1.3815080101060855, "learning_rate": 0.00022814999999999996, "loss": 0.4023, "step": 1521 }, { "epoch": 0.7162352941176471, "grad_norm": 3.742310163666799, "learning_rate": 0.0002283, "loss": 1.0312, "step": 1522 }, { "epoch": 0.7167058823529412, "grad_norm": 2.899972835589401, "learning_rate": 0.00022844999999999997, "loss": 0.8438, "step": 1523 }, { "epoch": 0.7171764705882353, "grad_norm": 1.6955739639313931, "learning_rate": 0.00022859999999999997, "loss": 0.832, "step": 1524 }, { "epoch": 0.7176470588235294, "grad_norm": 0.8023875956966636, "learning_rate": 0.00022874999999999998, "loss": 0.8008, "step": 1525 }, { "epoch": 0.7181176470588235, "grad_norm": 0.8727447025000873, "learning_rate": 0.00022889999999999998, "loss": 0.6875, "step": 1526 }, { "epoch": 0.7185882352941176, "grad_norm": 0.7851451162187459, "learning_rate": 0.00022904999999999996, "loss": 0.4141, "step": 1527 }, { "epoch": 0.7190588235294118, "grad_norm": 0.9048457557953513, "learning_rate": 0.0002292, "loss": 0.334, "step": 1528 }, { "epoch": 0.7195294117647059, "grad_norm": 0.6221912437196012, "learning_rate": 0.00022934999999999996, "loss": 0.2402, "step": 1529 }, { "epoch": 0.72, "grad_norm": 1.0390395027964907, "learning_rate": 0.0002295, "loss": 0.334, "step": 1530 }, { "epoch": 0.7204705882352941, "grad_norm": 2.311817351505413, "learning_rate": 0.00022964999999999997, "loss": 0.7891, "step": 1531 }, { "epoch": 0.7209411764705882, "grad_norm": 1.1794315757420804, "learning_rate": 0.00022979999999999997, "loss": 0.4277, "step": 1532 }, { "epoch": 0.7214117647058823, "grad_norm": 1.673545000422853, "learning_rate": 0.00022994999999999998, "loss": 0.8672, "step": 1533 }, { "epoch": 0.7218823529411764, "grad_norm": 0.5248801813012701, "learning_rate": 0.00023009999999999998, "loss": 0.4043, "step": 1534 }, { "epoch": 0.7223529411764706, "grad_norm": 0.6628696276715341, "learning_rate": 0.00023024999999999996, "loss": 0.6719, "step": 1535 }, { "epoch": 0.7228235294117648, "grad_norm": 1.0799641308369212, "learning_rate": 0.0002304, "loss": 0.832, "step": 1536 }, { "epoch": 0.7232941176470589, "grad_norm": 1.2551956200429188, "learning_rate": 0.00023054999999999997, "loss": 0.8672, "step": 1537 }, { "epoch": 0.723764705882353, "grad_norm": 0.7920333336998496, "learning_rate": 0.0002307, "loss": 0.7422, "step": 1538 }, { "epoch": 0.7242352941176471, "grad_norm": 0.14755247052875414, "learning_rate": 0.00023084999999999997, "loss": 0.0952, "step": 1539 }, { "epoch": 0.7247058823529412, "grad_norm": 0.819565002072272, "learning_rate": 0.00023099999999999998, "loss": 0.6562, "step": 1540 }, { "epoch": 0.7251764705882353, "grad_norm": 0.2255958760436914, "learning_rate": 0.00023114999999999998, "loss": 0.063, "step": 1541 }, { "epoch": 0.7256470588235294, "grad_norm": 1.2108041216842307, "learning_rate": 0.00023129999999999998, "loss": 0.3691, "step": 1542 }, { "epoch": 0.7261176470588235, "grad_norm": 1.8417886479397818, "learning_rate": 0.00023144999999999996, "loss": 0.6016, "step": 1543 }, { "epoch": 0.7265882352941176, "grad_norm": 0.15777565667977922, "learning_rate": 0.0002316, "loss": 0.0688, "step": 1544 }, { "epoch": 0.7270588235294118, "grad_norm": 1.5089225977000271, "learning_rate": 0.00023174999999999997, "loss": 0.8477, "step": 1545 }, { "epoch": 0.7275294117647059, "grad_norm": 0.48031713815268107, "learning_rate": 0.0002319, "loss": 0.4434, "step": 1546 }, { "epoch": 0.728, "grad_norm": 0.4237243790078257, "learning_rate": 0.00023204999999999998, "loss": 0.4102, "step": 1547 }, { "epoch": 0.7284705882352941, "grad_norm": 0.1734048744312187, "learning_rate": 0.00023219999999999998, "loss": 0.0845, "step": 1548 }, { "epoch": 0.7289411764705882, "grad_norm": 0.2893980903308708, "learning_rate": 0.00023234999999999998, "loss": 0.2188, "step": 1549 }, { "epoch": 0.7294117647058823, "grad_norm": 0.8432018171387861, "learning_rate": 0.00023249999999999999, "loss": 0.6992, "step": 1550 }, { "epoch": 0.7298823529411764, "grad_norm": 1.2800696992369578, "learning_rate": 0.00023264999999999996, "loss": 1.1484, "step": 1551 }, { "epoch": 0.7303529411764705, "grad_norm": 1.3999479139816042, "learning_rate": 0.0002328, "loss": 1.3125, "step": 1552 }, { "epoch": 0.7308235294117648, "grad_norm": 0.7611109043055622, "learning_rate": 0.00023294999999999997, "loss": 0.6094, "step": 1553 }, { "epoch": 0.7312941176470589, "grad_norm": 0.7471013911279517, "learning_rate": 0.00023309999999999997, "loss": 0.4629, "step": 1554 }, { "epoch": 0.731764705882353, "grad_norm": 1.2154880857918553, "learning_rate": 0.00023324999999999998, "loss": 0.7422, "step": 1555 }, { "epoch": 0.7322352941176471, "grad_norm": 0.5972946270102102, "learning_rate": 0.00023339999999999998, "loss": 0.3613, "step": 1556 }, { "epoch": 0.7327058823529412, "grad_norm": 0.448888702682091, "learning_rate": 0.00023354999999999996, "loss": 0.6445, "step": 1557 }, { "epoch": 0.7331764705882353, "grad_norm": 0.9666308626045933, "learning_rate": 0.0002337, "loss": 0.7227, "step": 1558 }, { "epoch": 0.7336470588235294, "grad_norm": 0.4971971761851123, "learning_rate": 0.00023384999999999997, "loss": 0.2559, "step": 1559 }, { "epoch": 0.7341176470588235, "grad_norm": 0.17785067680120928, "learning_rate": 0.000234, "loss": 0.0928, "step": 1560 }, { "epoch": 0.7345882352941177, "grad_norm": 0.14845154451386486, "learning_rate": 0.00023414999999999997, "loss": 0.0845, "step": 1561 }, { "epoch": 0.7350588235294118, "grad_norm": 0.4655521054929147, "learning_rate": 0.00023429999999999998, "loss": 0.2412, "step": 1562 }, { "epoch": 0.7355294117647059, "grad_norm": 0.4415832374303001, "learning_rate": 0.00023444999999999998, "loss": 0.3633, "step": 1563 }, { "epoch": 0.736, "grad_norm": 0.8310544235113982, "learning_rate": 0.00023459999999999998, "loss": 0.7383, "step": 1564 }, { "epoch": 0.7364705882352941, "grad_norm": 0.285720731199029, "learning_rate": 0.00023474999999999996, "loss": 0.2451, "step": 1565 }, { "epoch": 0.7369411764705882, "grad_norm": 0.6134035921478573, "learning_rate": 0.0002349, "loss": 0.5195, "step": 1566 }, { "epoch": 0.7374117647058823, "grad_norm": 0.43106928988460974, "learning_rate": 0.00023504999999999997, "loss": 0.4805, "step": 1567 }, { "epoch": 0.7378823529411764, "grad_norm": 0.7970972855900171, "learning_rate": 0.0002352, "loss": 0.4746, "step": 1568 }, { "epoch": 0.7383529411764705, "grad_norm": 1.0619604221796353, "learning_rate": 0.00023534999999999997, "loss": 0.2676, "step": 1569 }, { "epoch": 0.7388235294117647, "grad_norm": 0.935192959117847, "learning_rate": 0.00023549999999999998, "loss": 0.75, "step": 1570 }, { "epoch": 0.7392941176470588, "grad_norm": 0.9168626034647783, "learning_rate": 0.00023564999999999998, "loss": 0.6367, "step": 1571 }, { "epoch": 0.739764705882353, "grad_norm": 0.4836692950225115, "learning_rate": 0.00023579999999999999, "loss": 0.2871, "step": 1572 }, { "epoch": 0.7402352941176471, "grad_norm": 1.29327113737219, "learning_rate": 0.00023594999999999996, "loss": 1.1562, "step": 1573 }, { "epoch": 0.7407058823529412, "grad_norm": 0.7664447019353549, "learning_rate": 0.0002361, "loss": 0.5664, "step": 1574 }, { "epoch": 0.7411764705882353, "grad_norm": 0.736805322699185, "learning_rate": 0.00023624999999999997, "loss": 0.8086, "step": 1575 }, { "epoch": 0.7416470588235294, "grad_norm": 0.694000576644725, "learning_rate": 0.0002364, "loss": 0.6914, "step": 1576 }, { "epoch": 0.7421176470588235, "grad_norm": 0.206186115045359, "learning_rate": 0.00023654999999999998, "loss": 0.0972, "step": 1577 }, { "epoch": 0.7425882352941177, "grad_norm": 0.5392360953301502, "learning_rate": 0.00023669999999999998, "loss": 0.543, "step": 1578 }, { "epoch": 0.7430588235294118, "grad_norm": 1.4515535862298796, "learning_rate": 0.00023684999999999998, "loss": 1.7656, "step": 1579 }, { "epoch": 0.7435294117647059, "grad_norm": 0.3904614543551327, "learning_rate": 0.000237, "loss": 0.2051, "step": 1580 }, { "epoch": 0.744, "grad_norm": 0.1872594194824693, "learning_rate": 0.00023714999999999996, "loss": 0.0317, "step": 1581 }, { "epoch": 0.7444705882352941, "grad_norm": 2.5201054681906117, "learning_rate": 0.0002373, "loss": 0.5117, "step": 1582 }, { "epoch": 0.7449411764705882, "grad_norm": 1.3830025042393248, "learning_rate": 0.00023744999999999997, "loss": 0.8594, "step": 1583 }, { "epoch": 0.7454117647058823, "grad_norm": 0.4947579149471535, "learning_rate": 0.0002376, "loss": 0.293, "step": 1584 }, { "epoch": 0.7458823529411764, "grad_norm": 1.3450201599606146, "learning_rate": 0.00023774999999999998, "loss": 1.0312, "step": 1585 }, { "epoch": 0.7463529411764706, "grad_norm": 0.9639939097835344, "learning_rate": 0.00023789999999999998, "loss": 0.7227, "step": 1586 }, { "epoch": 0.7468235294117647, "grad_norm": 0.692973596105788, "learning_rate": 0.00023804999999999999, "loss": 0.582, "step": 1587 }, { "epoch": 0.7472941176470588, "grad_norm": 0.4561063149380402, "learning_rate": 0.0002382, "loss": 0.3418, "step": 1588 }, { "epoch": 0.7477647058823529, "grad_norm": 0.24517722331227576, "learning_rate": 0.00023834999999999997, "loss": 0.0703, "step": 1589 }, { "epoch": 0.7482352941176471, "grad_norm": 0.6035068418513635, "learning_rate": 0.0002385, "loss": 0.373, "step": 1590 }, { "epoch": 0.7487058823529412, "grad_norm": 0.9244699747287953, "learning_rate": 0.00023864999999999997, "loss": 0.6641, "step": 1591 }, { "epoch": 0.7491764705882353, "grad_norm": 1.1934462921255757, "learning_rate": 0.0002388, "loss": 0.8008, "step": 1592 }, { "epoch": 0.7496470588235294, "grad_norm": 0.6831697917403781, "learning_rate": 0.00023894999999999998, "loss": 0.459, "step": 1593 }, { "epoch": 0.7501176470588236, "grad_norm": 0.20952292910216713, "learning_rate": 0.00023909999999999998, "loss": 0.1465, "step": 1594 }, { "epoch": 0.7505882352941177, "grad_norm": 0.9972619498906786, "learning_rate": 0.00023925, "loss": 0.6211, "step": 1595 }, { "epoch": 0.7510588235294118, "grad_norm": 0.6488394254044758, "learning_rate": 0.0002394, "loss": 0.377, "step": 1596 }, { "epoch": 0.7515294117647059, "grad_norm": 1.1124465014324454, "learning_rate": 0.00023954999999999997, "loss": 0.3691, "step": 1597 }, { "epoch": 0.752, "grad_norm": 1.476595498917147, "learning_rate": 0.0002397, "loss": 0.7148, "step": 1598 }, { "epoch": 0.7524705882352941, "grad_norm": 0.671942816964095, "learning_rate": 0.00023984999999999998, "loss": 0.4746, "step": 1599 }, { "epoch": 0.7529411764705882, "grad_norm": 0.761852443820748, "learning_rate": 0.00023999999999999998, "loss": 0.582, "step": 1600 }, { "epoch": 0.7534117647058823, "grad_norm": 1.1191481535146905, "learning_rate": 0.00024014999999999998, "loss": 1.0156, "step": 1601 }, { "epoch": 0.7538823529411764, "grad_norm": 0.7069922891648587, "learning_rate": 0.00024029999999999999, "loss": 0.7734, "step": 1602 }, { "epoch": 0.7543529411764706, "grad_norm": 0.5631929257440909, "learning_rate": 0.00024044999999999996, "loss": 0.1914, "step": 1603 }, { "epoch": 0.7548235294117647, "grad_norm": 0.9620214057327504, "learning_rate": 0.0002406, "loss": 0.9219, "step": 1604 }, { "epoch": 0.7552941176470588, "grad_norm": 1.7086193922241837, "learning_rate": 0.00024074999999999997, "loss": 0.6094, "step": 1605 }, { "epoch": 0.7557647058823529, "grad_norm": 1.2450649243855718, "learning_rate": 0.0002409, "loss": 0.7383, "step": 1606 }, { "epoch": 0.756235294117647, "grad_norm": 0.44639313198148894, "learning_rate": 0.00024104999999999998, "loss": 0.2891, "step": 1607 }, { "epoch": 0.7567058823529412, "grad_norm": 0.41396498549064653, "learning_rate": 0.00024119999999999998, "loss": 0.2432, "step": 1608 }, { "epoch": 0.7571764705882353, "grad_norm": 1.0789344768855669, "learning_rate": 0.00024134999999999998, "loss": 1.0234, "step": 1609 }, { "epoch": 0.7576470588235295, "grad_norm": 1.2876461239643782, "learning_rate": 0.0002415, "loss": 1.0938, "step": 1610 }, { "epoch": 0.7581176470588236, "grad_norm": 0.8132739401156546, "learning_rate": 0.00024164999999999996, "loss": 0.5547, "step": 1611 }, { "epoch": 0.7585882352941177, "grad_norm": 0.7001550146377488, "learning_rate": 0.0002418, "loss": 0.582, "step": 1612 }, { "epoch": 0.7590588235294118, "grad_norm": 0.310044295322753, "learning_rate": 0.00024194999999999997, "loss": 0.1816, "step": 1613 }, { "epoch": 0.7595294117647059, "grad_norm": 1.8022537157865453, "learning_rate": 0.0002421, "loss": 0.8516, "step": 1614 }, { "epoch": 0.76, "grad_norm": 0.8058173183778412, "learning_rate": 0.00024224999999999998, "loss": 0.7344, "step": 1615 }, { "epoch": 0.7604705882352941, "grad_norm": 0.5807090707247538, "learning_rate": 0.00024239999999999998, "loss": 0.5273, "step": 1616 }, { "epoch": 0.7609411764705882, "grad_norm": 0.4655774904018283, "learning_rate": 0.00024255, "loss": 0.4316, "step": 1617 }, { "epoch": 0.7614117647058823, "grad_norm": 0.20553851636832937, "learning_rate": 0.0002427, "loss": 0.0947, "step": 1618 }, { "epoch": 0.7618823529411765, "grad_norm": 0.6288498765341047, "learning_rate": 0.00024284999999999997, "loss": 0.3574, "step": 1619 }, { "epoch": 0.7623529411764706, "grad_norm": 0.7193899505904776, "learning_rate": 0.000243, "loss": 0.5273, "step": 1620 }, { "epoch": 0.7628235294117647, "grad_norm": 0.9980508595234882, "learning_rate": 0.00024314999999999997, "loss": 0.5508, "step": 1621 }, { "epoch": 0.7632941176470588, "grad_norm": 0.3265150218470262, "learning_rate": 0.0002433, "loss": 0.2129, "step": 1622 }, { "epoch": 0.7637647058823529, "grad_norm": 1.1135541900457773, "learning_rate": 0.00024344999999999998, "loss": 1.25, "step": 1623 }, { "epoch": 0.764235294117647, "grad_norm": 0.852084041666733, "learning_rate": 0.00024359999999999999, "loss": 0.6523, "step": 1624 }, { "epoch": 0.7647058823529411, "grad_norm": 0.6704702002422857, "learning_rate": 0.00024375, "loss": 0.3555, "step": 1625 }, { "epoch": 0.7651764705882353, "grad_norm": 0.520236515642979, "learning_rate": 0.00024389999999999997, "loss": 0.3184, "step": 1626 }, { "epoch": 0.7656470588235295, "grad_norm": 0.22375637670117599, "learning_rate": 0.00024404999999999997, "loss": 0.1118, "step": 1627 }, { "epoch": 0.7661176470588236, "grad_norm": 0.8204167824517862, "learning_rate": 0.00024419999999999997, "loss": 0.5234, "step": 1628 }, { "epoch": 0.7665882352941177, "grad_norm": 0.5323589631422363, "learning_rate": 0.00024435, "loss": 0.4551, "step": 1629 }, { "epoch": 0.7670588235294118, "grad_norm": 1.2746711858144715, "learning_rate": 0.0002445, "loss": 0.8945, "step": 1630 }, { "epoch": 0.7675294117647059, "grad_norm": 1.2419997600874704, "learning_rate": 0.00024464999999999996, "loss": 1.1406, "step": 1631 }, { "epoch": 0.768, "grad_norm": 0.8180698430388603, "learning_rate": 0.0002448, "loss": 0.6602, "step": 1632 }, { "epoch": 0.7684705882352941, "grad_norm": 0.5696513379390673, "learning_rate": 0.00024494999999999996, "loss": 0.6484, "step": 1633 }, { "epoch": 0.7689411764705882, "grad_norm": 0.3055911121159137, "learning_rate": 0.00024509999999999994, "loss": 0.2148, "step": 1634 }, { "epoch": 0.7694117647058824, "grad_norm": 0.20274066920174402, "learning_rate": 0.00024524999999999997, "loss": 0.0928, "step": 1635 }, { "epoch": 0.7698823529411765, "grad_norm": 0.6208905979044019, "learning_rate": 0.00024539999999999995, "loss": 0.2207, "step": 1636 }, { "epoch": 0.7703529411764706, "grad_norm": 1.7216493294189914, "learning_rate": 0.00024555, "loss": 0.8867, "step": 1637 }, { "epoch": 0.7708235294117647, "grad_norm": 0.4974807122879204, "learning_rate": 0.00024569999999999995, "loss": 0.3926, "step": 1638 }, { "epoch": 0.7712941176470588, "grad_norm": 1.2539978043236406, "learning_rate": 0.00024585, "loss": 0.9609, "step": 1639 }, { "epoch": 0.7717647058823529, "grad_norm": 0.3903544197438216, "learning_rate": 0.00024599999999999996, "loss": 0.416, "step": 1640 }, { "epoch": 0.772235294117647, "grad_norm": 0.43184644503764624, "learning_rate": 0.00024615, "loss": 0.3047, "step": 1641 }, { "epoch": 0.7727058823529411, "grad_norm": 0.3292915495470157, "learning_rate": 0.00024629999999999997, "loss": 0.2148, "step": 1642 }, { "epoch": 0.7731764705882352, "grad_norm": 0.9058561137441741, "learning_rate": 0.00024645, "loss": 0.707, "step": 1643 }, { "epoch": 0.7736470588235295, "grad_norm": 0.47871402257536444, "learning_rate": 0.0002466, "loss": 0.2373, "step": 1644 }, { "epoch": 0.7741176470588236, "grad_norm": 1.349857040677693, "learning_rate": 0.00024675, "loss": 0.8945, "step": 1645 }, { "epoch": 0.7745882352941177, "grad_norm": 0.3102114033569848, "learning_rate": 0.0002469, "loss": 0.1475, "step": 1646 }, { "epoch": 0.7750588235294118, "grad_norm": 1.0911004674021807, "learning_rate": 0.00024704999999999996, "loss": 0.6797, "step": 1647 }, { "epoch": 0.7755294117647059, "grad_norm": 0.29909312659122084, "learning_rate": 0.0002472, "loss": 0.2324, "step": 1648 }, { "epoch": 0.776, "grad_norm": 1.2700272788981055, "learning_rate": 0.00024734999999999997, "loss": 1.0312, "step": 1649 }, { "epoch": 0.7764705882352941, "grad_norm": 0.27300408411358384, "learning_rate": 0.00024749999999999994, "loss": 0.1436, "step": 1650 }, { "epoch": 0.7769411764705882, "grad_norm": 0.6912220339702603, "learning_rate": 0.00024765, "loss": 0.2773, "step": 1651 }, { "epoch": 0.7774117647058824, "grad_norm": 1.9907815726728488, "learning_rate": 0.00024779999999999995, "loss": 1.0312, "step": 1652 }, { "epoch": 0.7778823529411765, "grad_norm": 0.2701525451642164, "learning_rate": 0.00024795, "loss": 0.3281, "step": 1653 }, { "epoch": 0.7783529411764706, "grad_norm": 0.7767907649077823, "learning_rate": 0.00024809999999999996, "loss": 0.5156, "step": 1654 }, { "epoch": 0.7788235294117647, "grad_norm": 2.2044794737209803, "learning_rate": 0.00024825, "loss": 1.2344, "step": 1655 }, { "epoch": 0.7792941176470588, "grad_norm": 0.8120598637679295, "learning_rate": 0.00024839999999999997, "loss": 0.5742, "step": 1656 }, { "epoch": 0.7797647058823529, "grad_norm": 1.3385437823558124, "learning_rate": 0.00024855, "loss": 0.707, "step": 1657 }, { "epoch": 0.780235294117647, "grad_norm": 2.7741467918767047, "learning_rate": 0.0002487, "loss": 1.2656, "step": 1658 }, { "epoch": 0.7807058823529411, "grad_norm": 1.2197966391503687, "learning_rate": 0.00024885, "loss": 0.5312, "step": 1659 }, { "epoch": 0.7811764705882352, "grad_norm": 1.2113351030952866, "learning_rate": 0.000249, "loss": 0.7188, "step": 1660 }, { "epoch": 0.7816470588235294, "grad_norm": 0.9289748414887277, "learning_rate": 0.00024914999999999996, "loss": 0.6445, "step": 1661 }, { "epoch": 0.7821176470588236, "grad_norm": 0.8476838074601832, "learning_rate": 0.0002493, "loss": 0.5977, "step": 1662 }, { "epoch": 0.7825882352941177, "grad_norm": 0.3079661528058667, "learning_rate": 0.00024944999999999996, "loss": 0.2031, "step": 1663 }, { "epoch": 0.7830588235294118, "grad_norm": 0.3967962101081234, "learning_rate": 0.00024959999999999994, "loss": 0.168, "step": 1664 }, { "epoch": 0.7835294117647059, "grad_norm": 1.035315343770387, "learning_rate": 0.00024974999999999997, "loss": 0.668, "step": 1665 }, { "epoch": 0.784, "grad_norm": 0.34416233373889527, "learning_rate": 0.00024989999999999995, "loss": 0.1709, "step": 1666 }, { "epoch": 0.7844705882352941, "grad_norm": 0.05291348138536339, "learning_rate": 0.00025005, "loss": 0.009, "step": 1667 }, { "epoch": 0.7849411764705883, "grad_norm": 0.38201704957487315, "learning_rate": 0.00025019999999999996, "loss": 0.1875, "step": 1668 }, { "epoch": 0.7854117647058824, "grad_norm": 0.3201867137525833, "learning_rate": 0.00025035, "loss": 0.2139, "step": 1669 }, { "epoch": 0.7858823529411765, "grad_norm": 0.33164308337785486, "learning_rate": 0.00025049999999999996, "loss": 0.2432, "step": 1670 }, { "epoch": 0.7863529411764706, "grad_norm": 0.4764174524587233, "learning_rate": 0.00025065, "loss": 0.2773, "step": 1671 }, { "epoch": 0.7868235294117647, "grad_norm": 0.716466845401015, "learning_rate": 0.00025079999999999997, "loss": 0.377, "step": 1672 }, { "epoch": 0.7872941176470588, "grad_norm": 0.5757375246112131, "learning_rate": 0.00025095, "loss": 0.2354, "step": 1673 }, { "epoch": 0.7877647058823529, "grad_norm": 0.6823841816909341, "learning_rate": 0.0002511, "loss": 0.582, "step": 1674 }, { "epoch": 0.788235294117647, "grad_norm": 0.08113893006398998, "learning_rate": 0.00025125, "loss": 0.0147, "step": 1675 }, { "epoch": 0.7887058823529411, "grad_norm": 1.2484872836011762, "learning_rate": 0.0002514, "loss": 1.0312, "step": 1676 }, { "epoch": 0.7891764705882353, "grad_norm": 0.9904881255444773, "learning_rate": 0.00025154999999999996, "loss": 0.7109, "step": 1677 }, { "epoch": 0.7896470588235294, "grad_norm": 0.6969579372698647, "learning_rate": 0.0002517, "loss": 0.4922, "step": 1678 }, { "epoch": 0.7901176470588235, "grad_norm": 0.8397363000864592, "learning_rate": 0.00025184999999999997, "loss": 0.5742, "step": 1679 }, { "epoch": 0.7905882352941176, "grad_norm": 0.1567590120323918, "learning_rate": 0.00025199999999999995, "loss": 0.0586, "step": 1680 }, { "epoch": 0.7910588235294118, "grad_norm": 1.9953688288845415, "learning_rate": 0.00025215, "loss": 0.6758, "step": 1681 }, { "epoch": 0.7915294117647059, "grad_norm": 1.057981786505867, "learning_rate": 0.00025229999999999995, "loss": 1.0781, "step": 1682 }, { "epoch": 0.792, "grad_norm": 0.30922692702420995, "learning_rate": 0.00025245, "loss": 0.1602, "step": 1683 }, { "epoch": 0.7924705882352941, "grad_norm": 0.8675762518771549, "learning_rate": 0.00025259999999999996, "loss": 0.5508, "step": 1684 }, { "epoch": 0.7929411764705883, "grad_norm": 1.1474779086230331, "learning_rate": 0.00025275, "loss": 0.6289, "step": 1685 }, { "epoch": 0.7934117647058824, "grad_norm": 1.187478934985439, "learning_rate": 0.00025289999999999997, "loss": 0.668, "step": 1686 }, { "epoch": 0.7938823529411765, "grad_norm": 0.6946356253516658, "learning_rate": 0.00025305, "loss": 0.6133, "step": 1687 }, { "epoch": 0.7943529411764706, "grad_norm": 186.3412451284468, "learning_rate": 0.0002532, "loss": 0.6289, "step": 1688 }, { "epoch": 0.7948235294117647, "grad_norm": 0.3983865282620059, "learning_rate": 0.00025335, "loss": 0.1602, "step": 1689 }, { "epoch": 0.7952941176470588, "grad_norm": 1.4967442061798062, "learning_rate": 0.0002535, "loss": 0.4395, "step": 1690 }, { "epoch": 0.7957647058823529, "grad_norm": 1.8733392256301058, "learning_rate": 0.00025365, "loss": 0.8281, "step": 1691 }, { "epoch": 0.796235294117647, "grad_norm": 1.1992095123260624, "learning_rate": 0.0002538, "loss": 1.0625, "step": 1692 }, { "epoch": 0.7967058823529412, "grad_norm": 0.3910107200650718, "learning_rate": 0.00025394999999999997, "loss": 0.1699, "step": 1693 }, { "epoch": 0.7971764705882353, "grad_norm": 0.57261701916655, "learning_rate": 0.0002541, "loss": 0.4043, "step": 1694 }, { "epoch": 0.7976470588235294, "grad_norm": 1.0288600760021558, "learning_rate": 0.00025425, "loss": 0.5273, "step": 1695 }, { "epoch": 0.7981176470588235, "grad_norm": 0.868127346004536, "learning_rate": 0.00025439999999999995, "loss": 0.7656, "step": 1696 }, { "epoch": 0.7985882352941176, "grad_norm": 0.4413650231145831, "learning_rate": 0.00025455, "loss": 0.2158, "step": 1697 }, { "epoch": 0.7990588235294117, "grad_norm": 1.0397104986594359, "learning_rate": 0.00025469999999999996, "loss": 0.4707, "step": 1698 }, { "epoch": 0.7995294117647059, "grad_norm": 1.4267153353367295, "learning_rate": 0.00025485, "loss": 1.0938, "step": 1699 }, { "epoch": 0.8, "grad_norm": 0.387027870856493, "learning_rate": 0.00025499999999999996, "loss": 0.3457, "step": 1700 }, { "epoch": 0.8004705882352942, "grad_norm": 0.7405648058610118, "learning_rate": 0.00025515, "loss": 0.2461, "step": 1701 }, { "epoch": 0.8009411764705883, "grad_norm": 4.281486273908819, "learning_rate": 0.00025529999999999997, "loss": 1.1562, "step": 1702 }, { "epoch": 0.8014117647058824, "grad_norm": 1.2929540674309885, "learning_rate": 0.00025545, "loss": 0.875, "step": 1703 }, { "epoch": 0.8018823529411765, "grad_norm": 1.7337135854285934, "learning_rate": 0.0002556, "loss": 0.6602, "step": 1704 }, { "epoch": 0.8023529411764706, "grad_norm": 1.769752962731271, "learning_rate": 0.00025575, "loss": 0.8711, "step": 1705 }, { "epoch": 0.8028235294117647, "grad_norm": 0.5560683440663149, "learning_rate": 0.0002559, "loss": 0.1855, "step": 1706 }, { "epoch": 0.8032941176470588, "grad_norm": 2.093376774809631, "learning_rate": 0.00025604999999999996, "loss": 0.2891, "step": 1707 }, { "epoch": 0.8037647058823529, "grad_norm": 7.826432241446053, "learning_rate": 0.0002562, "loss": 0.8633, "step": 1708 }, { "epoch": 0.804235294117647, "grad_norm": 0.8169399325318537, "learning_rate": 0.00025634999999999997, "loss": 0.6719, "step": 1709 }, { "epoch": 0.8047058823529412, "grad_norm": 0.5613033950811338, "learning_rate": 0.00025649999999999995, "loss": 0.2852, "step": 1710 }, { "epoch": 0.8051764705882353, "grad_norm": 0.6361746717894778, "learning_rate": 0.00025665, "loss": 0.2969, "step": 1711 }, { "epoch": 0.8056470588235294, "grad_norm": 1.3410858918731317, "learning_rate": 0.00025679999999999995, "loss": 0.5625, "step": 1712 }, { "epoch": 0.8061176470588235, "grad_norm": 2.21492971352735, "learning_rate": 0.00025695, "loss": 1.1094, "step": 1713 }, { "epoch": 0.8065882352941176, "grad_norm": 0.6112465406603899, "learning_rate": 0.00025709999999999996, "loss": 0.5117, "step": 1714 }, { "epoch": 0.8070588235294117, "grad_norm": 1.0938125703884711, "learning_rate": 0.00025725, "loss": 0.5703, "step": 1715 }, { "epoch": 0.8075294117647058, "grad_norm": 1.1799728937199754, "learning_rate": 0.00025739999999999997, "loss": 0.3965, "step": 1716 }, { "epoch": 0.808, "grad_norm": 2.805380252112087, "learning_rate": 0.00025755, "loss": 1.1953, "step": 1717 }, { "epoch": 0.8084705882352942, "grad_norm": 0.35362038801230117, "learning_rate": 0.0002577, "loss": 0.2236, "step": 1718 }, { "epoch": 0.8089411764705883, "grad_norm": 0.751614153109755, "learning_rate": 0.00025785, "loss": 0.2812, "step": 1719 }, { "epoch": 0.8094117647058824, "grad_norm": 0.31239720207587335, "learning_rate": 0.000258, "loss": 0.0549, "step": 1720 }, { "epoch": 0.8098823529411765, "grad_norm": 8.828444804042459, "learning_rate": 0.00025815, "loss": 1.3516, "step": 1721 }, { "epoch": 0.8103529411764706, "grad_norm": 0.6319756384634746, "learning_rate": 0.0002583, "loss": 0.375, "step": 1722 }, { "epoch": 0.8108235294117647, "grad_norm": 0.42712108014560307, "learning_rate": 0.00025844999999999997, "loss": 0.1611, "step": 1723 }, { "epoch": 0.8112941176470588, "grad_norm": 3.302118235497819, "learning_rate": 0.0002586, "loss": 1.3516, "step": 1724 }, { "epoch": 0.8117647058823529, "grad_norm": 0.7342205065185035, "learning_rate": 0.00025875, "loss": 0.1348, "step": 1725 }, { "epoch": 0.812235294117647, "grad_norm": 0.7509531533334682, "learning_rate": 0.00025889999999999995, "loss": 0.5, "step": 1726 }, { "epoch": 0.8127058823529412, "grad_norm": 1.6242038984280827, "learning_rate": 0.00025905, "loss": 0.4805, "step": 1727 }, { "epoch": 0.8131764705882353, "grad_norm": 0.969642537580155, "learning_rate": 0.00025919999999999996, "loss": 0.8203, "step": 1728 }, { "epoch": 0.8136470588235294, "grad_norm": 0.6758016703460737, "learning_rate": 0.00025935, "loss": 0.3711, "step": 1729 }, { "epoch": 0.8141176470588235, "grad_norm": 0.22685885005993595, "learning_rate": 0.00025949999999999997, "loss": 0.0938, "step": 1730 }, { "epoch": 0.8145882352941176, "grad_norm": 1.3056239718795652, "learning_rate": 0.00025965, "loss": 0.7734, "step": 1731 }, { "epoch": 0.8150588235294117, "grad_norm": 0.482766103764615, "learning_rate": 0.00025979999999999997, "loss": 0.2949, "step": 1732 }, { "epoch": 0.8155294117647058, "grad_norm": 1.4685957559870733, "learning_rate": 0.00025995, "loss": 0.8125, "step": 1733 }, { "epoch": 0.816, "grad_norm": 0.33554324296637394, "learning_rate": 0.0002601, "loss": 0.1797, "step": 1734 }, { "epoch": 0.8164705882352942, "grad_norm": 1.0796176601772443, "learning_rate": 0.00026025, "loss": 1.0547, "step": 1735 }, { "epoch": 0.8169411764705883, "grad_norm": 0.5940454775677864, "learning_rate": 0.0002604, "loss": 0.4883, "step": 1736 }, { "epoch": 0.8174117647058824, "grad_norm": 0.614795098006916, "learning_rate": 0.00026055, "loss": 0.5508, "step": 1737 }, { "epoch": 0.8178823529411765, "grad_norm": 0.5519304374659372, "learning_rate": 0.0002607, "loss": 0.5078, "step": 1738 }, { "epoch": 0.8183529411764706, "grad_norm": 0.3559998188111999, "learning_rate": 0.00026084999999999997, "loss": 0.1787, "step": 1739 }, { "epoch": 0.8188235294117647, "grad_norm": 1.2254841381154877, "learning_rate": 0.000261, "loss": 0.8867, "step": 1740 }, { "epoch": 0.8192941176470588, "grad_norm": 0.3751437949347406, "learning_rate": 0.00026115, "loss": 0.165, "step": 1741 }, { "epoch": 0.819764705882353, "grad_norm": 1.0112496963766513, "learning_rate": 0.00026129999999999995, "loss": 0.1924, "step": 1742 }, { "epoch": 0.8202352941176471, "grad_norm": 0.22370754699850026, "learning_rate": 0.00026145, "loss": 0.1187, "step": 1743 }, { "epoch": 0.8207058823529412, "grad_norm": 1.692231601861776, "learning_rate": 0.00026159999999999996, "loss": 0.8047, "step": 1744 }, { "epoch": 0.8211764705882353, "grad_norm": 4.561578780621376, "learning_rate": 0.00026175, "loss": 1.375, "step": 1745 }, { "epoch": 0.8216470588235294, "grad_norm": 1.640053579357598, "learning_rate": 0.00026189999999999997, "loss": 1.2109, "step": 1746 }, { "epoch": 0.8221176470588235, "grad_norm": 1.088016565069868, "learning_rate": 0.00026205, "loss": 1.2891, "step": 1747 }, { "epoch": 0.8225882352941176, "grad_norm": 0.7984023055979929, "learning_rate": 0.0002622, "loss": 0.5273, "step": 1748 }, { "epoch": 0.8230588235294117, "grad_norm": 0.8429717876544, "learning_rate": 0.00026235, "loss": 0.7109, "step": 1749 }, { "epoch": 0.8235294117647058, "grad_norm": 1.1822161780261968, "learning_rate": 0.0002625, "loss": 1.0312, "step": 1750 }, { "epoch": 0.824, "grad_norm": 1.1309913768379594, "learning_rate": 0.00026264999999999996, "loss": 0.9375, "step": 1751 }, { "epoch": 0.8244705882352941, "grad_norm": 2.3240539904173336, "learning_rate": 0.0002628, "loss": 0.9805, "step": 1752 }, { "epoch": 0.8249411764705883, "grad_norm": 3.087425371135697, "learning_rate": 0.00026294999999999997, "loss": 0.8125, "step": 1753 }, { "epoch": 0.8254117647058824, "grad_norm": 1.0487562591672608, "learning_rate": 0.0002631, "loss": 0.7695, "step": 1754 }, { "epoch": 0.8258823529411765, "grad_norm": 0.6074143119748796, "learning_rate": 0.00026325, "loss": 0.4395, "step": 1755 }, { "epoch": 0.8263529411764706, "grad_norm": 0.6755744634200169, "learning_rate": 0.00026339999999999995, "loss": 0.5703, "step": 1756 }, { "epoch": 0.8268235294117647, "grad_norm": 1.19146780775608, "learning_rate": 0.00026355, "loss": 1.1328, "step": 1757 }, { "epoch": 0.8272941176470588, "grad_norm": 0.7142321479674486, "learning_rate": 0.00026369999999999996, "loss": 0.582, "step": 1758 }, { "epoch": 0.827764705882353, "grad_norm": 0.5672914697326136, "learning_rate": 0.00026384999999999994, "loss": 0.4883, "step": 1759 }, { "epoch": 0.8282352941176471, "grad_norm": 0.682697436157297, "learning_rate": 0.00026399999999999997, "loss": 0.5977, "step": 1760 }, { "epoch": 0.8287058823529412, "grad_norm": 0.6543285923456857, "learning_rate": 0.00026414999999999994, "loss": 0.1074, "step": 1761 }, { "epoch": 0.8291764705882353, "grad_norm": 14.64664523414107, "learning_rate": 0.0002643, "loss": 0.418, "step": 1762 }, { "epoch": 0.8296470588235294, "grad_norm": 0.9638882296884215, "learning_rate": 0.00026444999999999995, "loss": 0.5547, "step": 1763 }, { "epoch": 0.8301176470588235, "grad_norm": 0.5832891477543776, "learning_rate": 0.0002646, "loss": 0.4219, "step": 1764 }, { "epoch": 0.8305882352941176, "grad_norm": 0.8395998426176483, "learning_rate": 0.00026474999999999996, "loss": 0.7227, "step": 1765 }, { "epoch": 0.8310588235294117, "grad_norm": 0.5417936191874949, "learning_rate": 0.0002649, "loss": 0.5352, "step": 1766 }, { "epoch": 0.8315294117647059, "grad_norm": 1.022810116602172, "learning_rate": 0.00026504999999999996, "loss": 0.2598, "step": 1767 }, { "epoch": 0.832, "grad_norm": 0.3851979725799651, "learning_rate": 0.0002652, "loss": 0.1387, "step": 1768 }, { "epoch": 0.8324705882352941, "grad_norm": 1.0820063075110884, "learning_rate": 0.00026534999999999997, "loss": 0.6914, "step": 1769 }, { "epoch": 0.8329411764705882, "grad_norm": 0.30391992636383114, "learning_rate": 0.0002655, "loss": 0.2637, "step": 1770 }, { "epoch": 0.8334117647058824, "grad_norm": 2.11842463092649, "learning_rate": 0.00026565, "loss": 1.1328, "step": 1771 }, { "epoch": 0.8338823529411765, "grad_norm": 0.4310064055023741, "learning_rate": 0.00026579999999999996, "loss": 0.2969, "step": 1772 }, { "epoch": 0.8343529411764706, "grad_norm": 1.869478597334778, "learning_rate": 0.00026595, "loss": 1.1172, "step": 1773 }, { "epoch": 0.8348235294117647, "grad_norm": 0.6594821548033462, "learning_rate": 0.00026609999999999996, "loss": 0.2812, "step": 1774 }, { "epoch": 0.8352941176470589, "grad_norm": 1.9339584680197313, "learning_rate": 0.00026624999999999994, "loss": 1.375, "step": 1775 }, { "epoch": 0.835764705882353, "grad_norm": 3.189963636087494, "learning_rate": 0.00026639999999999997, "loss": 1.5078, "step": 1776 }, { "epoch": 0.8362352941176471, "grad_norm": 0.42237145831556705, "learning_rate": 0.00026654999999999995, "loss": 0.3281, "step": 1777 }, { "epoch": 0.8367058823529412, "grad_norm": 4.92965865078828, "learning_rate": 0.0002667, "loss": 1.1875, "step": 1778 }, { "epoch": 0.8371764705882353, "grad_norm": 1.0559490361158421, "learning_rate": 0.00026684999999999995, "loss": 0.1504, "step": 1779 }, { "epoch": 0.8376470588235294, "grad_norm": 8.599035203756598, "learning_rate": 0.000267, "loss": 1.9062, "step": 1780 }, { "epoch": 0.8381176470588235, "grad_norm": 0.7252685058463623, "learning_rate": 0.00026714999999999996, "loss": 0.1699, "step": 1781 }, { "epoch": 0.8385882352941176, "grad_norm": 3.1246499542469723, "learning_rate": 0.0002673, "loss": 1.2969, "step": 1782 }, { "epoch": 0.8390588235294117, "grad_norm": 6.143490170013402, "learning_rate": 0.00026744999999999997, "loss": 1.125, "step": 1783 }, { "epoch": 0.8395294117647059, "grad_norm": 0.7454246133782326, "learning_rate": 0.0002676, "loss": 0.4512, "step": 1784 }, { "epoch": 0.84, "grad_norm": 0.7106325996518073, "learning_rate": 0.00026775, "loss": 0.4883, "step": 1785 }, { "epoch": 0.8404705882352941, "grad_norm": 1.1982139719353364, "learning_rate": 0.0002679, "loss": 0.918, "step": 1786 }, { "epoch": 0.8409411764705882, "grad_norm": 0.2490229269152589, "learning_rate": 0.00026805, "loss": 0.1797, "step": 1787 }, { "epoch": 0.8414117647058823, "grad_norm": 2.018104921050214, "learning_rate": 0.00026819999999999996, "loss": 0.8789, "step": 1788 }, { "epoch": 0.8418823529411765, "grad_norm": 1.7648974121367895, "learning_rate": 0.00026835, "loss": 0.875, "step": 1789 }, { "epoch": 0.8423529411764706, "grad_norm": 0.4161250657572892, "learning_rate": 0.00026849999999999997, "loss": 0.2354, "step": 1790 }, { "epoch": 0.8428235294117647, "grad_norm": 0.5878801752634752, "learning_rate": 0.00026864999999999994, "loss": 0.4219, "step": 1791 }, { "epoch": 0.8432941176470589, "grad_norm": 1.0611752819722304, "learning_rate": 0.0002688, "loss": 0.6914, "step": 1792 }, { "epoch": 0.843764705882353, "grad_norm": 1.2073466875747187, "learning_rate": 0.00026894999999999995, "loss": 0.8945, "step": 1793 }, { "epoch": 0.8442352941176471, "grad_norm": 1.3331499201051968, "learning_rate": 0.0002691, "loss": 0.9766, "step": 1794 }, { "epoch": 0.8447058823529412, "grad_norm": 0.6086115185109574, "learning_rate": 0.00026924999999999996, "loss": 0.793, "step": 1795 }, { "epoch": 0.8451764705882353, "grad_norm": 0.819472772184276, "learning_rate": 0.0002694, "loss": 0.5391, "step": 1796 }, { "epoch": 0.8456470588235294, "grad_norm": 2.2848721799998555, "learning_rate": 0.00026954999999999997, "loss": 1.0078, "step": 1797 }, { "epoch": 0.8461176470588235, "grad_norm": 0.5681817395444358, "learning_rate": 0.0002697, "loss": 0.3008, "step": 1798 }, { "epoch": 0.8465882352941176, "grad_norm": 8.04753569072354, "learning_rate": 0.00026984999999999997, "loss": 1.6641, "step": 1799 }, { "epoch": 0.8470588235294118, "grad_norm": 0.7331220413534095, "learning_rate": 0.00027, "loss": 0.2754, "step": 1800 }, { "epoch": 0.8475294117647059, "grad_norm": 4.578407269581459, "learning_rate": 0.00027015, "loss": 1.0078, "step": 1801 }, { "epoch": 0.848, "grad_norm": 2.7258895176416584, "learning_rate": 0.00027029999999999996, "loss": 0.2793, "step": 1802 }, { "epoch": 0.8484705882352941, "grad_norm": 7.243356087802088, "learning_rate": 0.00027045, "loss": 1.1094, "step": 1803 }, { "epoch": 0.8489411764705882, "grad_norm": 0.6609311582587903, "learning_rate": 0.00027059999999999996, "loss": 0.0854, "step": 1804 }, { "epoch": 0.8494117647058823, "grad_norm": 4.515442479961137, "learning_rate": 0.00027074999999999994, "loss": 0.3438, "step": 1805 }, { "epoch": 0.8498823529411764, "grad_norm": 0.9056707932734825, "learning_rate": 0.00027089999999999997, "loss": 0.6016, "step": 1806 }, { "epoch": 0.8503529411764705, "grad_norm": 0.7433409844461097, "learning_rate": 0.00027104999999999995, "loss": 0.6094, "step": 1807 }, { "epoch": 0.8508235294117648, "grad_norm": 0.728354193619272, "learning_rate": 0.0002712, "loss": 0.6211, "step": 1808 }, { "epoch": 0.8512941176470589, "grad_norm": 0.4358004472503555, "learning_rate": 0.00027134999999999995, "loss": 0.3047, "step": 1809 }, { "epoch": 0.851764705882353, "grad_norm": 1.972607723723633, "learning_rate": 0.0002715, "loss": 0.6875, "step": 1810 }, { "epoch": 0.8522352941176471, "grad_norm": 2.6299826425787716, "learning_rate": 0.00027164999999999996, "loss": 0.9883, "step": 1811 }, { "epoch": 0.8527058823529412, "grad_norm": 0.42072744928215033, "learning_rate": 0.0002718, "loss": 0.2734, "step": 1812 }, { "epoch": 0.8531764705882353, "grad_norm": 1.3499024183548385, "learning_rate": 0.00027194999999999997, "loss": 1.1562, "step": 1813 }, { "epoch": 0.8536470588235294, "grad_norm": 1.116700358875539, "learning_rate": 0.0002721, "loss": 0.793, "step": 1814 }, { "epoch": 0.8541176470588235, "grad_norm": 0.5583066795108653, "learning_rate": 0.00027225, "loss": 0.2852, "step": 1815 }, { "epoch": 0.8545882352941176, "grad_norm": 1.261446366665725, "learning_rate": 0.0002724, "loss": 0.5234, "step": 1816 }, { "epoch": 0.8550588235294118, "grad_norm": 1.2995254751215888, "learning_rate": 0.00027255, "loss": 0.416, "step": 1817 }, { "epoch": 0.8555294117647059, "grad_norm": 0.6027995859449442, "learning_rate": 0.00027269999999999996, "loss": 0.3926, "step": 1818 }, { "epoch": 0.856, "grad_norm": 163.116823460591, "learning_rate": 0.00027285, "loss": 0.6211, "step": 1819 }, { "epoch": 0.8564705882352941, "grad_norm": 35.87903267120068, "learning_rate": 0.00027299999999999997, "loss": 0.4805, "step": 1820 }, { "epoch": 0.8569411764705882, "grad_norm": 9.34788053438399, "learning_rate": 0.00027314999999999994, "loss": 1.0625, "step": 1821 }, { "epoch": 0.8574117647058823, "grad_norm": 3.030286419237993, "learning_rate": 0.0002733, "loss": 0.4004, "step": 1822 }, { "epoch": 0.8578823529411764, "grad_norm": 15.795761603272226, "learning_rate": 0.00027344999999999995, "loss": 0.7617, "step": 1823 }, { "epoch": 0.8583529411764705, "grad_norm": 0.15335410723401965, "learning_rate": 0.0002736, "loss": 0.0435, "step": 1824 }, { "epoch": 0.8588235294117647, "grad_norm": 9.751441207759118, "learning_rate": 0.00027374999999999996, "loss": 0.4531, "step": 1825 }, { "epoch": 0.8592941176470589, "grad_norm": 1.224515259753223, "learning_rate": 0.0002739, "loss": 0.4082, "step": 1826 }, { "epoch": 0.859764705882353, "grad_norm": 2.9096600736454747, "learning_rate": 0.00027404999999999997, "loss": 0.5938, "step": 1827 }, { "epoch": 0.8602352941176471, "grad_norm": 1268.0424415924601, "learning_rate": 0.0002742, "loss": 1.7266, "step": 1828 }, { "epoch": 0.8607058823529412, "grad_norm": 3.5433601473308327, "learning_rate": 0.00027435, "loss": 0.6133, "step": 1829 }, { "epoch": 0.8611764705882353, "grad_norm": 0.4998933926907155, "learning_rate": 0.0002745, "loss": 0.5195, "step": 1830 }, { "epoch": 0.8616470588235294, "grad_norm": 0.9115180381898776, "learning_rate": 0.00027465, "loss": 0.6953, "step": 1831 }, { "epoch": 0.8621176470588235, "grad_norm": 2.8979182811512887, "learning_rate": 0.0002748, "loss": 0.9336, "step": 1832 }, { "epoch": 0.8625882352941177, "grad_norm": 1.7345506619702489, "learning_rate": 0.00027495, "loss": 0.3379, "step": 1833 }, { "epoch": 0.8630588235294118, "grad_norm": 0.8259267951459591, "learning_rate": 0.00027509999999999996, "loss": 0.209, "step": 1834 }, { "epoch": 0.8635294117647059, "grad_norm": 1.2751958352619799, "learning_rate": 0.00027525, "loss": 1.0, "step": 1835 }, { "epoch": 0.864, "grad_norm": 0.771797900473269, "learning_rate": 0.00027539999999999997, "loss": 0.4727, "step": 1836 }, { "epoch": 0.8644705882352941, "grad_norm": 0.7154266926028713, "learning_rate": 0.00027554999999999995, "loss": 0.4297, "step": 1837 }, { "epoch": 0.8649411764705882, "grad_norm": 1.3340087762811903, "learning_rate": 0.0002757, "loss": 0.4805, "step": 1838 }, { "epoch": 0.8654117647058823, "grad_norm": 2.603736806451102, "learning_rate": 0.00027584999999999996, "loss": 0.7031, "step": 1839 }, { "epoch": 0.8658823529411764, "grad_norm": 0.6079663819564365, "learning_rate": 0.000276, "loss": 0.2578, "step": 1840 }, { "epoch": 0.8663529411764705, "grad_norm": 0.6368915839229126, "learning_rate": 0.00027614999999999996, "loss": 0.3418, "step": 1841 }, { "epoch": 0.8668235294117647, "grad_norm": 2.4687839547903385, "learning_rate": 0.0002763, "loss": 1.3047, "step": 1842 }, { "epoch": 0.8672941176470588, "grad_norm": 0.634543284721965, "learning_rate": 0.00027644999999999997, "loss": 0.3926, "step": 1843 }, { "epoch": 0.867764705882353, "grad_norm": 0.2959463475458897, "learning_rate": 0.0002766, "loss": 0.1021, "step": 1844 }, { "epoch": 0.8682352941176471, "grad_norm": 0.6738947257617132, "learning_rate": 0.00027675, "loss": 0.5039, "step": 1845 }, { "epoch": 0.8687058823529412, "grad_norm": 0.7855091228424183, "learning_rate": 0.0002769, "loss": 0.5234, "step": 1846 }, { "epoch": 0.8691764705882353, "grad_norm": 0.7133879568723548, "learning_rate": 0.00027705, "loss": 0.2295, "step": 1847 }, { "epoch": 0.8696470588235294, "grad_norm": 0.49141625986971776, "learning_rate": 0.0002772, "loss": 0.3008, "step": 1848 }, { "epoch": 0.8701176470588236, "grad_norm": 1.2886034897141623, "learning_rate": 0.00027735, "loss": 0.6602, "step": 1849 }, { "epoch": 0.8705882352941177, "grad_norm": 2.5597804258525954, "learning_rate": 0.00027749999999999997, "loss": 1.0, "step": 1850 }, { "epoch": 0.8710588235294118, "grad_norm": 1.8213842129846909, "learning_rate": 0.00027764999999999995, "loss": 0.7578, "step": 1851 }, { "epoch": 0.8715294117647059, "grad_norm": 0.2711396545219131, "learning_rate": 0.0002778, "loss": 0.1484, "step": 1852 }, { "epoch": 0.872, "grad_norm": 1.3145731668449412, "learning_rate": 0.00027794999999999995, "loss": 0.3379, "step": 1853 }, { "epoch": 0.8724705882352941, "grad_norm": 5.530710941373088, "learning_rate": 0.0002781, "loss": 1.3047, "step": 1854 }, { "epoch": 0.8729411764705882, "grad_norm": 1.0779274828574648, "learning_rate": 0.00027824999999999996, "loss": 0.7891, "step": 1855 }, { "epoch": 0.8734117647058823, "grad_norm": 0.37723331061005155, "learning_rate": 0.0002784, "loss": 0.1787, "step": 1856 }, { "epoch": 0.8738823529411764, "grad_norm": 0.853045526470145, "learning_rate": 0.00027854999999999997, "loss": 0.7266, "step": 1857 }, { "epoch": 0.8743529411764706, "grad_norm": 0.542702268459568, "learning_rate": 0.0002787, "loss": 0.3691, "step": 1858 }, { "epoch": 0.8748235294117647, "grad_norm": 0.6783664772790605, "learning_rate": 0.00027885, "loss": 0.6445, "step": 1859 }, { "epoch": 0.8752941176470588, "grad_norm": 0.7712354685875629, "learning_rate": 0.000279, "loss": 0.6328, "step": 1860 }, { "epoch": 0.8757647058823529, "grad_norm": 0.7540294027255894, "learning_rate": 0.00027915, "loss": 0.6602, "step": 1861 }, { "epoch": 0.8762352941176471, "grad_norm": 2.9193482737840077, "learning_rate": 0.0002793, "loss": 1.5156, "step": 1862 }, { "epoch": 0.8767058823529412, "grad_norm": 0.6121042946902189, "learning_rate": 0.00027945, "loss": 0.3457, "step": 1863 }, { "epoch": 0.8771764705882353, "grad_norm": 2.531194166218884, "learning_rate": 0.00027959999999999997, "loss": 0.5352, "step": 1864 }, { "epoch": 0.8776470588235294, "grad_norm": 0.6483251323900107, "learning_rate": 0.00027975, "loss": 0.2695, "step": 1865 }, { "epoch": 0.8781176470588236, "grad_norm": 0.6150185569296112, "learning_rate": 0.0002799, "loss": 0.3379, "step": 1866 }, { "epoch": 0.8785882352941177, "grad_norm": 1.0740282496210194, "learning_rate": 0.00028004999999999995, "loss": 0.5508, "step": 1867 }, { "epoch": 0.8790588235294118, "grad_norm": 1.8832961326515643, "learning_rate": 0.0002802, "loss": 0.4766, "step": 1868 }, { "epoch": 0.8795294117647059, "grad_norm": 19.89388342762319, "learning_rate": 0.00028034999999999996, "loss": 0.5156, "step": 1869 }, { "epoch": 0.88, "grad_norm": 0.6574458518874003, "learning_rate": 0.0002805, "loss": 0.4395, "step": 1870 }, { "epoch": 0.8804705882352941, "grad_norm": 0.7471944048224677, "learning_rate": 0.00028064999999999996, "loss": 0.5078, "step": 1871 }, { "epoch": 0.8809411764705882, "grad_norm": 0.5525677057412882, "learning_rate": 0.0002808, "loss": 0.3691, "step": 1872 }, { "epoch": 0.8814117647058823, "grad_norm": 1.1824302340432575, "learning_rate": 0.00028094999999999997, "loss": 0.5156, "step": 1873 }, { "epoch": 0.8818823529411765, "grad_norm": 0.5270083984580488, "learning_rate": 0.0002811, "loss": 0.2559, "step": 1874 }, { "epoch": 0.8823529411764706, "grad_norm": 0.8566964360048256, "learning_rate": 0.00028125, "loss": 0.5781, "step": 1875 }, { "epoch": 0.8828235294117647, "grad_norm": 2.422684073339713, "learning_rate": 0.00028139999999999996, "loss": 1.1172, "step": 1876 }, { "epoch": 0.8832941176470588, "grad_norm": 0.42809440251889264, "learning_rate": 0.00028155, "loss": 0.3438, "step": 1877 }, { "epoch": 0.8837647058823529, "grad_norm": 2.0033548396582868, "learning_rate": 0.00028169999999999996, "loss": 0.9258, "step": 1878 }, { "epoch": 0.884235294117647, "grad_norm": 0.6946020012838222, "learning_rate": 0.00028185, "loss": 0.5312, "step": 1879 }, { "epoch": 0.8847058823529412, "grad_norm": 0.6910206763830444, "learning_rate": 0.00028199999999999997, "loss": 0.3516, "step": 1880 }, { "epoch": 0.8851764705882353, "grad_norm": 1.407452888364625, "learning_rate": 0.00028215, "loss": 0.4414, "step": 1881 }, { "epoch": 0.8856470588235295, "grad_norm": 1.2901371107600303, "learning_rate": 0.0002823, "loss": 0.5508, "step": 1882 }, { "epoch": 0.8861176470588236, "grad_norm": 1.2973640114144305, "learning_rate": 0.00028244999999999995, "loss": 0.6367, "step": 1883 }, { "epoch": 0.8865882352941177, "grad_norm": 0.5250064736991586, "learning_rate": 0.0002826, "loss": 0.3574, "step": 1884 }, { "epoch": 0.8870588235294118, "grad_norm": 0.8990297732506728, "learning_rate": 0.00028274999999999996, "loss": 0.3984, "step": 1885 }, { "epoch": 0.8875294117647059, "grad_norm": 0.6265038857021148, "learning_rate": 0.00028289999999999994, "loss": 0.3398, "step": 1886 }, { "epoch": 0.888, "grad_norm": 2.256320797114715, "learning_rate": 0.00028304999999999997, "loss": 0.6484, "step": 1887 }, { "epoch": 0.8884705882352941, "grad_norm": 0.7061231522328684, "learning_rate": 0.00028319999999999994, "loss": 0.543, "step": 1888 }, { "epoch": 0.8889411764705882, "grad_norm": 0.8096315339171369, "learning_rate": 0.00028335, "loss": 0.625, "step": 1889 }, { "epoch": 0.8894117647058823, "grad_norm": 0.1364486322005271, "learning_rate": 0.00028349999999999995, "loss": 0.0659, "step": 1890 }, { "epoch": 0.8898823529411765, "grad_norm": 12.678087145491448, "learning_rate": 0.00028365, "loss": 0.5, "step": 1891 }, { "epoch": 0.8903529411764706, "grad_norm": 0.3860045764334932, "learning_rate": 0.00028379999999999996, "loss": 0.1318, "step": 1892 }, { "epoch": 0.8908235294117647, "grad_norm": 1.1886949272881602, "learning_rate": 0.00028395, "loss": 0.4043, "step": 1893 }, { "epoch": 0.8912941176470588, "grad_norm": 1.7967275107873397, "learning_rate": 0.00028409999999999997, "loss": 0.3184, "step": 1894 }, { "epoch": 0.8917647058823529, "grad_norm": 2.615547126908021, "learning_rate": 0.00028425, "loss": 0.4297, "step": 1895 }, { "epoch": 0.892235294117647, "grad_norm": 0.9945803075829914, "learning_rate": 0.0002844, "loss": 0.5195, "step": 1896 }, { "epoch": 0.8927058823529411, "grad_norm": 0.9703466676353008, "learning_rate": 0.00028455, "loss": 0.2832, "step": 1897 }, { "epoch": 0.8931764705882353, "grad_norm": 2.228645482164307, "learning_rate": 0.0002847, "loss": 0.5977, "step": 1898 }, { "epoch": 0.8936470588235295, "grad_norm": 0.5359043335203522, "learning_rate": 0.00028484999999999996, "loss": 0.2871, "step": 1899 }, { "epoch": 0.8941176470588236, "grad_norm": 1.1438426756876052, "learning_rate": 0.000285, "loss": 0.8398, "step": 1900 }, { "epoch": 0.8945882352941177, "grad_norm": 0.4435712358771744, "learning_rate": 0.00028514999999999997, "loss": 0.3633, "step": 1901 }, { "epoch": 0.8950588235294118, "grad_norm": 0.6679259486204332, "learning_rate": 0.00028529999999999994, "loss": 0.5938, "step": 1902 }, { "epoch": 0.8955294117647059, "grad_norm": 0.690178931914832, "learning_rate": 0.00028544999999999997, "loss": 0.4883, "step": 1903 }, { "epoch": 0.896, "grad_norm": 0.17685279810416246, "learning_rate": 0.00028559999999999995, "loss": 0.062, "step": 1904 }, { "epoch": 0.8964705882352941, "grad_norm": 1.3500445349890846, "learning_rate": 0.00028575, "loss": 0.7773, "step": 1905 }, { "epoch": 0.8969411764705882, "grad_norm": 1.3160067672610978, "learning_rate": 0.00028589999999999996, "loss": 1.0234, "step": 1906 }, { "epoch": 0.8974117647058824, "grad_norm": 0.5001616267686528, "learning_rate": 0.00028605, "loss": 0.2891, "step": 1907 }, { "epoch": 0.8978823529411765, "grad_norm": 0.6325572867018808, "learning_rate": 0.00028619999999999996, "loss": 0.6914, "step": 1908 }, { "epoch": 0.8983529411764706, "grad_norm": 0.41829076477371985, "learning_rate": 0.00028635, "loss": 0.124, "step": 1909 }, { "epoch": 0.8988235294117647, "grad_norm": 0.33371888475261297, "learning_rate": 0.00028649999999999997, "loss": 0.1719, "step": 1910 }, { "epoch": 0.8992941176470588, "grad_norm": 1.7284874366083154, "learning_rate": 0.00028665, "loss": 0.9102, "step": 1911 }, { "epoch": 0.8997647058823529, "grad_norm": 0.4239841647406126, "learning_rate": 0.0002868, "loss": 0.2031, "step": 1912 }, { "epoch": 0.900235294117647, "grad_norm": 0.7660448866164389, "learning_rate": 0.00028694999999999995, "loss": 0.2295, "step": 1913 }, { "epoch": 0.9007058823529411, "grad_norm": 0.18166741998244437, "learning_rate": 0.0002871, "loss": 0.0457, "step": 1914 }, { "epoch": 0.9011764705882352, "grad_norm": 0.6493036031302608, "learning_rate": 0.00028724999999999996, "loss": 0.1865, "step": 1915 }, { "epoch": 0.9016470588235294, "grad_norm": 1.6194078455388108, "learning_rate": 0.00028739999999999994, "loss": 0.9336, "step": 1916 }, { "epoch": 0.9021176470588236, "grad_norm": 1.0307936536069973, "learning_rate": 0.00028754999999999997, "loss": 0.5898, "step": 1917 }, { "epoch": 0.9025882352941177, "grad_norm": 0.0663768960217112, "learning_rate": 0.00028769999999999995, "loss": 0.0143, "step": 1918 }, { "epoch": 0.9030588235294118, "grad_norm": 0.5478313792458261, "learning_rate": 0.00028785, "loss": 0.4629, "step": 1919 }, { "epoch": 0.9035294117647059, "grad_norm": 0.43610609115169907, "learning_rate": 0.00028799999999999995, "loss": 0.3262, "step": 1920 }, { "epoch": 0.904, "grad_norm": 2.818249691671347, "learning_rate": 0.00028815, "loss": 0.1875, "step": 1921 }, { "epoch": 0.9044705882352941, "grad_norm": 1.1307708578030062, "learning_rate": 0.00028829999999999996, "loss": 0.6094, "step": 1922 }, { "epoch": 0.9049411764705882, "grad_norm": 0.6847682223877308, "learning_rate": 0.00028845, "loss": 0.418, "step": 1923 }, { "epoch": 0.9054117647058824, "grad_norm": 2.9570397616511586, "learning_rate": 0.00028859999999999997, "loss": 0.2871, "step": 1924 }, { "epoch": 0.9058823529411765, "grad_norm": 1.3021519397897587, "learning_rate": 0.00028875, "loss": 0.7891, "step": 1925 }, { "epoch": 0.9063529411764706, "grad_norm": 0.6609000170640847, "learning_rate": 0.0002889, "loss": 0.3926, "step": 1926 }, { "epoch": 0.9068235294117647, "grad_norm": 0.7748880596815708, "learning_rate": 0.00028905, "loss": 0.6953, "step": 1927 }, { "epoch": 0.9072941176470588, "grad_norm": 0.9763490564588232, "learning_rate": 0.0002892, "loss": 0.7031, "step": 1928 }, { "epoch": 0.9077647058823529, "grad_norm": 1.0003827747298377, "learning_rate": 0.00028934999999999996, "loss": 0.9297, "step": 1929 }, { "epoch": 0.908235294117647, "grad_norm": 0.9128993686990414, "learning_rate": 0.0002895, "loss": 0.6523, "step": 1930 }, { "epoch": 0.9087058823529411, "grad_norm": 0.6538744039840173, "learning_rate": 0.00028964999999999997, "loss": 0.4434, "step": 1931 }, { "epoch": 0.9091764705882353, "grad_norm": 1.1447602641313595, "learning_rate": 0.00028979999999999994, "loss": 0.707, "step": 1932 }, { "epoch": 0.9096470588235294, "grad_norm": 0.647955288537468, "learning_rate": 0.00028995, "loss": 0.3672, "step": 1933 }, { "epoch": 0.9101176470588235, "grad_norm": 1.2070772180956826, "learning_rate": 0.00029009999999999995, "loss": 0.625, "step": 1934 }, { "epoch": 0.9105882352941177, "grad_norm": 0.9473091377548979, "learning_rate": 0.00029025, "loss": 0.3516, "step": 1935 }, { "epoch": 0.9110588235294118, "grad_norm": 0.6133207314821894, "learning_rate": 0.00029039999999999996, "loss": 0.3047, "step": 1936 }, { "epoch": 0.9115294117647059, "grad_norm": 0.44543850655588907, "learning_rate": 0.00029055, "loss": 0.1982, "step": 1937 }, { "epoch": 0.912, "grad_norm": 0.9088560214592156, "learning_rate": 0.00029069999999999996, "loss": 0.4023, "step": 1938 }, { "epoch": 0.9124705882352941, "grad_norm": 0.7917600925163306, "learning_rate": 0.00029085, "loss": 0.3926, "step": 1939 }, { "epoch": 0.9129411764705883, "grad_norm": 3.1544588080324076, "learning_rate": 0.00029099999999999997, "loss": 0.9414, "step": 1940 }, { "epoch": 0.9134117647058824, "grad_norm": 1.1196477980633504, "learning_rate": 0.00029115, "loss": 1.0938, "step": 1941 }, { "epoch": 0.9138823529411765, "grad_norm": 0.3386924686372468, "learning_rate": 0.0002913, "loss": 0.1914, "step": 1942 }, { "epoch": 0.9143529411764706, "grad_norm": 0.619002644957469, "learning_rate": 0.00029145, "loss": 0.5078, "step": 1943 }, { "epoch": 0.9148235294117647, "grad_norm": 0.5146358254669068, "learning_rate": 0.0002916, "loss": 0.3379, "step": 1944 }, { "epoch": 0.9152941176470588, "grad_norm": 1.2670806338302243, "learning_rate": 0.00029174999999999996, "loss": 0.7578, "step": 1945 }, { "epoch": 0.9157647058823529, "grad_norm": 0.5084646505820173, "learning_rate": 0.0002919, "loss": 0.4824, "step": 1946 }, { "epoch": 0.916235294117647, "grad_norm": 0.9316861001592689, "learning_rate": 0.00029204999999999997, "loss": 0.5039, "step": 1947 }, { "epoch": 0.9167058823529411, "grad_norm": 0.6216180063752146, "learning_rate": 0.00029219999999999995, "loss": 0.582, "step": 1948 }, { "epoch": 0.9171764705882353, "grad_norm": 1.3039962689699127, "learning_rate": 0.00029235, "loss": 0.4258, "step": 1949 }, { "epoch": 0.9176470588235294, "grad_norm": 0.5702193010439696, "learning_rate": 0.00029249999999999995, "loss": 0.3789, "step": 1950 }, { "epoch": 0.9181176470588235, "grad_norm": 0.5186334346367387, "learning_rate": 0.00029265, "loss": 0.2227, "step": 1951 }, { "epoch": 0.9185882352941176, "grad_norm": 0.4199476297622594, "learning_rate": 0.00029279999999999996, "loss": 0.1797, "step": 1952 }, { "epoch": 0.9190588235294118, "grad_norm": 0.6942672442085024, "learning_rate": 0.00029295, "loss": 0.3203, "step": 1953 }, { "epoch": 0.9195294117647059, "grad_norm": 0.9425310472099631, "learning_rate": 0.00029309999999999997, "loss": 0.6211, "step": 1954 }, { "epoch": 0.92, "grad_norm": 0.5238695195042602, "learning_rate": 0.00029325, "loss": 0.332, "step": 1955 }, { "epoch": 0.9204705882352942, "grad_norm": 0.33053882523291556, "learning_rate": 0.0002934, "loss": 0.1924, "step": 1956 }, { "epoch": 0.9209411764705883, "grad_norm": 0.4903340578030495, "learning_rate": 0.00029355, "loss": 0.2158, "step": 1957 }, { "epoch": 0.9214117647058824, "grad_norm": 0.3227351416083805, "learning_rate": 0.0002937, "loss": 0.1797, "step": 1958 }, { "epoch": 0.9218823529411765, "grad_norm": 0.8813084071668971, "learning_rate": 0.00029384999999999996, "loss": 0.5391, "step": 1959 }, { "epoch": 0.9223529411764706, "grad_norm": 0.8528665882691603, "learning_rate": 0.000294, "loss": 0.5938, "step": 1960 }, { "epoch": 0.9228235294117647, "grad_norm": 0.5764429067409618, "learning_rate": 0.00029414999999999997, "loss": 0.2969, "step": 1961 }, { "epoch": 0.9232941176470588, "grad_norm": 6.238535386856139, "learning_rate": 0.00029429999999999994, "loss": 1.0234, "step": 1962 }, { "epoch": 0.9237647058823529, "grad_norm": 0.5634579564944823, "learning_rate": 0.00029445, "loss": 0.416, "step": 1963 }, { "epoch": 0.924235294117647, "grad_norm": 0.41759861954839606, "learning_rate": 0.00029459999999999995, "loss": 0.2256, "step": 1964 }, { "epoch": 0.9247058823529412, "grad_norm": 1.8560742782119914, "learning_rate": 0.00029475, "loss": 0.7656, "step": 1965 }, { "epoch": 0.9251764705882353, "grad_norm": 3.2187958251369677, "learning_rate": 0.00029489999999999996, "loss": 0.5625, "step": 1966 }, { "epoch": 0.9256470588235294, "grad_norm": 1.186766524411532, "learning_rate": 0.00029505, "loss": 0.2891, "step": 1967 }, { "epoch": 0.9261176470588235, "grad_norm": 1.5082046953244581, "learning_rate": 0.00029519999999999997, "loss": 0.2197, "step": 1968 }, { "epoch": 0.9265882352941176, "grad_norm": 1.2878901215998726, "learning_rate": 0.00029535, "loss": 0.2275, "step": 1969 }, { "epoch": 0.9270588235294117, "grad_norm": 3.471748814074499, "learning_rate": 0.00029549999999999997, "loss": 0.6484, "step": 1970 }, { "epoch": 0.9275294117647059, "grad_norm": 0.7158601270440363, "learning_rate": 0.00029565, "loss": 0.5352, "step": 1971 }, { "epoch": 0.928, "grad_norm": 0.5582404092172346, "learning_rate": 0.0002958, "loss": 0.2891, "step": 1972 }, { "epoch": 0.9284705882352942, "grad_norm": 1.2955536858255987, "learning_rate": 0.00029595, "loss": 0.6094, "step": 1973 }, { "epoch": 0.9289411764705883, "grad_norm": 1.0774536340157639, "learning_rate": 0.0002961, "loss": 0.5859, "step": 1974 }, { "epoch": 0.9294117647058824, "grad_norm": 0.33335906595654385, "learning_rate": 0.00029624999999999996, "loss": 0.2227, "step": 1975 }, { "epoch": 0.9298823529411765, "grad_norm": 0.7222999091249296, "learning_rate": 0.0002964, "loss": 0.4199, "step": 1976 }, { "epoch": 0.9303529411764706, "grad_norm": 0.960941360583951, "learning_rate": 0.00029654999999999997, "loss": 0.4082, "step": 1977 }, { "epoch": 0.9308235294117647, "grad_norm": 0.982387974085573, "learning_rate": 0.00029669999999999995, "loss": 0.5703, "step": 1978 }, { "epoch": 0.9312941176470588, "grad_norm": 0.5285451459620838, "learning_rate": 0.00029685, "loss": 0.1982, "step": 1979 }, { "epoch": 0.9317647058823529, "grad_norm": 0.5114367988120364, "learning_rate": 0.00029699999999999996, "loss": 0.2773, "step": 1980 }, { "epoch": 0.932235294117647, "grad_norm": 1.4086136325998355, "learning_rate": 0.00029715, "loss": 0.7891, "step": 1981 }, { "epoch": 0.9327058823529412, "grad_norm": 0.7122305205195403, "learning_rate": 0.00029729999999999996, "loss": 0.2578, "step": 1982 }, { "epoch": 0.9331764705882353, "grad_norm": 4.618044496236377, "learning_rate": 0.00029745, "loss": 1.2422, "step": 1983 }, { "epoch": 0.9336470588235294, "grad_norm": 0.727071627574374, "learning_rate": 0.00029759999999999997, "loss": 0.3965, "step": 1984 }, { "epoch": 0.9341176470588235, "grad_norm": 1.1866920619867174, "learning_rate": 0.00029775, "loss": 0.5977, "step": 1985 }, { "epoch": 0.9345882352941176, "grad_norm": 0.5731421792714855, "learning_rate": 0.0002979, "loss": 0.2227, "step": 1986 }, { "epoch": 0.9350588235294117, "grad_norm": 1.2853790951647153, "learning_rate": 0.00029805, "loss": 0.6406, "step": 1987 }, { "epoch": 0.9355294117647058, "grad_norm": 1.020857272946937, "learning_rate": 0.0002982, "loss": 0.4121, "step": 1988 }, { "epoch": 0.936, "grad_norm": 2.778305815917275, "learning_rate": 0.00029835, "loss": 0.6953, "step": 1989 }, { "epoch": 0.9364705882352942, "grad_norm": 6.491799945766335, "learning_rate": 0.0002985, "loss": 0.6328, "step": 1990 }, { "epoch": 0.9369411764705883, "grad_norm": 0.8685380544227272, "learning_rate": 0.00029864999999999997, "loss": 0.4238, "step": 1991 }, { "epoch": 0.9374117647058824, "grad_norm": 0.612793335557577, "learning_rate": 0.0002988, "loss": 0.2754, "step": 1992 }, { "epoch": 0.9378823529411765, "grad_norm": 0.46076033844471426, "learning_rate": 0.00029895, "loss": 0.1758, "step": 1993 }, { "epoch": 0.9383529411764706, "grad_norm": 0.8029739858779902, "learning_rate": 0.00029909999999999995, "loss": 0.3945, "step": 1994 }, { "epoch": 0.9388235294117647, "grad_norm": 0.5672599028256854, "learning_rate": 0.00029925, "loss": 0.2402, "step": 1995 }, { "epoch": 0.9392941176470588, "grad_norm": 0.8799321018092006, "learning_rate": 0.00029939999999999996, "loss": 0.5391, "step": 1996 }, { "epoch": 0.939764705882353, "grad_norm": 0.9778644621549533, "learning_rate": 0.00029955, "loss": 0.6719, "step": 1997 }, { "epoch": 0.9402352941176471, "grad_norm": 0.3643624851167493, "learning_rate": 0.00029969999999999997, "loss": 0.1216, "step": 1998 }, { "epoch": 0.9407058823529412, "grad_norm": 1.0908913110059226, "learning_rate": 0.00029985, "loss": 0.6562, "step": 1999 }, { "epoch": 0.9411764705882353, "grad_norm": 1.4256905838908749, "learning_rate": 0.0003, "loss": 1.0234, "step": 2000 }, { "epoch": 0.9416470588235294, "grad_norm": 6.176491329133198, "learning_rate": 0.0002999999935946321, "loss": 1.0469, "step": 2001 }, { "epoch": 0.9421176470588235, "grad_norm": 2.0343716636401017, "learning_rate": 0.00029999997437852905, "loss": 0.6055, "step": 2002 }, { "epoch": 0.9425882352941176, "grad_norm": 0.6813020860736175, "learning_rate": 0.0002999999423516924, "loss": 0.4219, "step": 2003 }, { "epoch": 0.9430588235294117, "grad_norm": 0.37176642140164773, "learning_rate": 0.00029999989751412494, "loss": 0.1104, "step": 2004 }, { "epoch": 0.9435294117647058, "grad_norm": 1.180287241541776, "learning_rate": 0.00029999983986583054, "loss": 0.5547, "step": 2005 }, { "epoch": 0.944, "grad_norm": 0.17022118464252417, "learning_rate": 0.000299999769406814, "loss": 0.0674, "step": 2006 }, { "epoch": 0.9444705882352942, "grad_norm": 4.972987045636348, "learning_rate": 0.0002999996861370814, "loss": 0.8594, "step": 2007 }, { "epoch": 0.9449411764705883, "grad_norm": 0.6502731218138986, "learning_rate": 0.0002999995900566399, "loss": 0.4062, "step": 2008 }, { "epoch": 0.9454117647058824, "grad_norm": 1.2222935058183297, "learning_rate": 0.0002999994811654977, "loss": 0.3711, "step": 2009 }, { "epoch": 0.9458823529411765, "grad_norm": 6.545927084795194, "learning_rate": 0.000299999359463664, "loss": 0.8633, "step": 2010 }, { "epoch": 0.9463529411764706, "grad_norm": 0.35049973146771635, "learning_rate": 0.0002999992249511493, "loss": 0.0649, "step": 2011 }, { "epoch": 0.9468235294117647, "grad_norm": 1.3890711840001146, "learning_rate": 0.00029999907762796505, "loss": 0.7461, "step": 2012 }, { "epoch": 0.9472941176470588, "grad_norm": 0.8165927129615663, "learning_rate": 0.00029999891749412375, "loss": 0.3926, "step": 2013 }, { "epoch": 0.947764705882353, "grad_norm": 0.7012991020213204, "learning_rate": 0.0002999987445496392, "loss": 0.1758, "step": 2014 }, { "epoch": 0.9482352941176471, "grad_norm": 1.2880655258395075, "learning_rate": 0.00029999855879452616, "loss": 0.7031, "step": 2015 }, { "epoch": 0.9487058823529412, "grad_norm": 0.7639820119374141, "learning_rate": 0.00029999836022880045, "loss": 0.5312, "step": 2016 }, { "epoch": 0.9491764705882353, "grad_norm": 18.52711225942869, "learning_rate": 0.00029999814885247906, "loss": 1.0, "step": 2017 }, { "epoch": 0.9496470588235294, "grad_norm": 1.6256840117134443, "learning_rate": 0.00029999792466558, "loss": 0.8594, "step": 2018 }, { "epoch": 0.9501176470588235, "grad_norm": 3.1204055968997806, "learning_rate": 0.0002999976876681224, "loss": 0.5977, "step": 2019 }, { "epoch": 0.9505882352941176, "grad_norm": 0.7597843588280848, "learning_rate": 0.00029999743786012656, "loss": 0.1592, "step": 2020 }, { "epoch": 0.9510588235294117, "grad_norm": 488.1626024537324, "learning_rate": 0.0002999971752416138, "loss": 1.9766, "step": 2021 }, { "epoch": 0.9515294117647058, "grad_norm": 2.9955490529546647, "learning_rate": 0.00029999689981260655, "loss": 0.8906, "step": 2022 }, { "epoch": 0.952, "grad_norm": 0.3080543894738393, "learning_rate": 0.0002999966115731283, "loss": 0.126, "step": 2023 }, { "epoch": 0.9524705882352941, "grad_norm": 0.8909780898825914, "learning_rate": 0.00029999631052320374, "loss": 0.5469, "step": 2024 }, { "epoch": 0.9529411764705882, "grad_norm": 0.7457438776787849, "learning_rate": 0.0002999959966628585, "loss": 0.2295, "step": 2025 }, { "epoch": 0.9534117647058824, "grad_norm": 1.099629804641442, "learning_rate": 0.0002999956699921194, "loss": 0.5273, "step": 2026 }, { "epoch": 0.9538823529411765, "grad_norm": 0.8388375332828327, "learning_rate": 0.0002999953305110144, "loss": 0.4277, "step": 2027 }, { "epoch": 0.9543529411764706, "grad_norm": 1.7500287957295801, "learning_rate": 0.0002999949782195724, "loss": 0.75, "step": 2028 }, { "epoch": 0.9548235294117647, "grad_norm": 3.5605447812956834, "learning_rate": 0.0002999946131178236, "loss": 0.918, "step": 2029 }, { "epoch": 0.9552941176470588, "grad_norm": 3.801951629494748, "learning_rate": 0.0002999942352057991, "loss": 1.0078, "step": 2030 }, { "epoch": 0.955764705882353, "grad_norm": 3.428626146194582, "learning_rate": 0.00029999384448353114, "loss": 0.4961, "step": 2031 }, { "epoch": 0.9562352941176471, "grad_norm": 19.615578378036165, "learning_rate": 0.00029999344095105326, "loss": 0.6055, "step": 2032 }, { "epoch": 0.9567058823529412, "grad_norm": 3.2062430117471163, "learning_rate": 0.0002999930246083997, "loss": 1.3438, "step": 2033 }, { "epoch": 0.9571764705882353, "grad_norm": 68.00022189244498, "learning_rate": 0.0002999925954556062, "loss": 0.9453, "step": 2034 }, { "epoch": 0.9576470588235294, "grad_norm": 20.402948721037493, "learning_rate": 0.0002999921534927093, "loss": 0.6719, "step": 2035 }, { "epoch": 0.9581176470588235, "grad_norm": 63.02393529846124, "learning_rate": 0.00029999169871974677, "loss": 0.7656, "step": 2036 }, { "epoch": 0.9585882352941176, "grad_norm": 25.342520534858416, "learning_rate": 0.0002999912311367575, "loss": 0.5352, "step": 2037 }, { "epoch": 0.9590588235294117, "grad_norm": 138.769570424601, "learning_rate": 0.00029999075074378136, "loss": 0.8281, "step": 2038 }, { "epoch": 0.9595294117647059, "grad_norm": 2.6291753633050488, "learning_rate": 0.00029999025754085944, "loss": 0.8867, "step": 2039 }, { "epoch": 0.96, "grad_norm": 2.0202591378066033, "learning_rate": 0.00029998975152803383, "loss": 0.4121, "step": 2040 }, { "epoch": 0.9604705882352941, "grad_norm": 4.914931732260481, "learning_rate": 0.0002999892327053477, "loss": 0.6562, "step": 2041 }, { "epoch": 0.9609411764705882, "grad_norm": 144.30375532875166, "learning_rate": 0.0002999887010728454, "loss": 1.0938, "step": 2042 }, { "epoch": 0.9614117647058823, "grad_norm": 14.736974908546241, "learning_rate": 0.0002999881566305724, "loss": 0.8555, "step": 2043 }, { "epoch": 0.9618823529411765, "grad_norm": 2.592276683499716, "learning_rate": 0.0002999875993785751, "loss": 1.3828, "step": 2044 }, { "epoch": 0.9623529411764706, "grad_norm": 6.415814924537517, "learning_rate": 0.0002999870293169011, "loss": 0.9492, "step": 2045 }, { "epoch": 0.9628235294117647, "grad_norm": 13.91790730946338, "learning_rate": 0.00029998644644559916, "loss": 0.8828, "step": 2046 }, { "epoch": 0.9632941176470589, "grad_norm": 7.173601191751188, "learning_rate": 0.000299985850764719, "loss": 0.9727, "step": 2047 }, { "epoch": 0.963764705882353, "grad_norm": 22.82384591123687, "learning_rate": 0.00029998524227431146, "loss": 1.0469, "step": 2048 }, { "epoch": 0.9642352941176471, "grad_norm": 18.529125473436682, "learning_rate": 0.00029998462097442866, "loss": 0.459, "step": 2049 }, { "epoch": 0.9647058823529412, "grad_norm": 19.072359571418556, "learning_rate": 0.0002999839868651235, "loss": 0.6914, "step": 2050 }, { "epoch": 0.9651764705882353, "grad_norm": 1.2521350109345826, "learning_rate": 0.0002999833399464502, "loss": 0.5664, "step": 2051 }, { "epoch": 0.9656470588235294, "grad_norm": 3.418921079214178, "learning_rate": 0.000299982680218464, "loss": 0.3691, "step": 2052 }, { "epoch": 0.9661176470588235, "grad_norm": 5.429779567882132, "learning_rate": 0.00029998200768122123, "loss": 0.8789, "step": 2053 }, { "epoch": 0.9665882352941176, "grad_norm": 26.717811200267466, "learning_rate": 0.00029998132233477936, "loss": 0.8594, "step": 2054 }, { "epoch": 0.9670588235294117, "grad_norm": 1.783666023594037, "learning_rate": 0.0002999806241791969, "loss": 0.8516, "step": 2055 }, { "epoch": 0.9675294117647059, "grad_norm": 1.784939912214696, "learning_rate": 0.0002999799132145335, "loss": 0.7305, "step": 2056 }, { "epoch": 0.968, "grad_norm": 1.222854058554226, "learning_rate": 0.00029997918944084986, "loss": 0.793, "step": 2057 }, { "epoch": 0.9684705882352941, "grad_norm": 0.7489440717682019, "learning_rate": 0.0002999784528582078, "loss": 0.625, "step": 2058 }, { "epoch": 0.9689411764705882, "grad_norm": 1.0846055009840208, "learning_rate": 0.0002999777034666702, "loss": 0.4219, "step": 2059 }, { "epoch": 0.9694117647058823, "grad_norm": 1.3331391717644518, "learning_rate": 0.00029997694126630115, "loss": 0.6016, "step": 2060 }, { "epoch": 0.9698823529411764, "grad_norm": 0.673964162864541, "learning_rate": 0.00029997616625716567, "loss": 0.0605, "step": 2061 }, { "epoch": 0.9703529411764706, "grad_norm": 0.8483682174352695, "learning_rate": 0.0002999753784393299, "loss": 0.3086, "step": 2062 }, { "epoch": 0.9708235294117648, "grad_norm": 1.0866962358200138, "learning_rate": 0.00029997457781286125, "loss": 0.375, "step": 2063 }, { "epoch": 0.9712941176470589, "grad_norm": 1.833481592384574, "learning_rate": 0.000299973764377828, "loss": 0.6172, "step": 2064 }, { "epoch": 0.971764705882353, "grad_norm": 0.3837193275680536, "learning_rate": 0.00029997293813429965, "loss": 0.2139, "step": 2065 }, { "epoch": 0.9722352941176471, "grad_norm": 0.49443450036778763, "learning_rate": 0.0002999720990823468, "loss": 0.2754, "step": 2066 }, { "epoch": 0.9727058823529412, "grad_norm": 0.4053053171332955, "learning_rate": 0.0002999712472220411, "loss": 0.2139, "step": 2067 }, { "epoch": 0.9731764705882353, "grad_norm": 0.7823341397777712, "learning_rate": 0.0002999703825534552, "loss": 0.6055, "step": 2068 }, { "epoch": 0.9736470588235294, "grad_norm": 0.6107322713873271, "learning_rate": 0.0002999695050766631, "loss": 0.498, "step": 2069 }, { "epoch": 0.9741176470588235, "grad_norm": 0.6090096588146482, "learning_rate": 0.0002999686147917396, "loss": 0.5742, "step": 2070 }, { "epoch": 0.9745882352941176, "grad_norm": 0.8042256388011494, "learning_rate": 0.0002999677116987609, "loss": 0.6914, "step": 2071 }, { "epoch": 0.9750588235294118, "grad_norm": 0.8641895100434002, "learning_rate": 0.00029996679579780394, "loss": 0.625, "step": 2072 }, { "epoch": 0.9755294117647059, "grad_norm": 1.0431613215523872, "learning_rate": 0.0002999658670889471, "loss": 0.6602, "step": 2073 }, { "epoch": 0.976, "grad_norm": 0.9925470137504194, "learning_rate": 0.0002999649255722696, "loss": 0.5312, "step": 2074 }, { "epoch": 0.9764705882352941, "grad_norm": 0.6274692937099692, "learning_rate": 0.0002999639712478519, "loss": 0.4102, "step": 2075 }, { "epoch": 0.9769411764705882, "grad_norm": 0.6775421266476722, "learning_rate": 0.00029996300411577547, "loss": 0.582, "step": 2076 }, { "epoch": 0.9774117647058823, "grad_norm": 0.5163352851448774, "learning_rate": 0.00029996202417612294, "loss": 0.4824, "step": 2077 }, { "epoch": 0.9778823529411764, "grad_norm": 0.4333028280571672, "learning_rate": 0.000299961031428978, "loss": 0.1768, "step": 2078 }, { "epoch": 0.9783529411764705, "grad_norm": 0.5215459100226125, "learning_rate": 0.00029996002587442537, "loss": 0.3711, "step": 2079 }, { "epoch": 0.9788235294117648, "grad_norm": 0.7866026691922703, "learning_rate": 0.000299959007512551, "loss": 0.5195, "step": 2080 }, { "epoch": 0.9792941176470589, "grad_norm": 0.4732777263341195, "learning_rate": 0.00029995797634344186, "loss": 0.2129, "step": 2081 }, { "epoch": 0.979764705882353, "grad_norm": 1.4037769240110463, "learning_rate": 0.000299956932367186, "loss": 0.668, "step": 2082 }, { "epoch": 0.9802352941176471, "grad_norm": 0.7602060073546716, "learning_rate": 0.0002999558755838726, "loss": 0.5664, "step": 2083 }, { "epoch": 0.9807058823529412, "grad_norm": 0.3691593898154028, "learning_rate": 0.00029995480599359186, "loss": 0.165, "step": 2084 }, { "epoch": 0.9811764705882353, "grad_norm": 0.5682735711354729, "learning_rate": 0.00029995372359643515, "loss": 0.3887, "step": 2085 }, { "epoch": 0.9816470588235294, "grad_norm": 0.7100127333929447, "learning_rate": 0.00029995262839249497, "loss": 0.5898, "step": 2086 }, { "epoch": 0.9821176470588235, "grad_norm": 0.7603694912317529, "learning_rate": 0.0002999515203818648, "loss": 0.7109, "step": 2087 }, { "epoch": 0.9825882352941177, "grad_norm": 0.9889884017902557, "learning_rate": 0.0002999503995646393, "loss": 0.5078, "step": 2088 }, { "epoch": 0.9830588235294118, "grad_norm": 0.4436307256026286, "learning_rate": 0.00029994926594091413, "loss": 0.1592, "step": 2089 }, { "epoch": 0.9835294117647059, "grad_norm": 0.8596694371838055, "learning_rate": 0.00029994811951078616, "loss": 0.6133, "step": 2090 }, { "epoch": 0.984, "grad_norm": 0.36351231691081864, "learning_rate": 0.0002999469602743533, "loss": 0.334, "step": 2091 }, { "epoch": 0.9844705882352941, "grad_norm": 0.3738552006432788, "learning_rate": 0.0002999457882317146, "loss": 0.2754, "step": 2092 }, { "epoch": 0.9849411764705882, "grad_norm": 0.4168012525187384, "learning_rate": 0.00029994460338297, "loss": 0.4199, "step": 2093 }, { "epoch": 0.9854117647058823, "grad_norm": 0.21162491265561553, "learning_rate": 0.0002999434057282209, "loss": 0.1406, "step": 2094 }, { "epoch": 0.9858823529411764, "grad_norm": 0.590876342150037, "learning_rate": 0.0002999421952675694, "loss": 0.3301, "step": 2095 }, { "epoch": 0.9863529411764705, "grad_norm": 0.4661820179352778, "learning_rate": 0.00029994097200111903, "loss": 0.2246, "step": 2096 }, { "epoch": 0.9868235294117647, "grad_norm": 0.3429451949029688, "learning_rate": 0.0002999397359289741, "loss": 0.1787, "step": 2097 }, { "epoch": 0.9872941176470589, "grad_norm": 1.4882838401868934, "learning_rate": 0.00029993848705124035, "loss": 1.0312, "step": 2098 }, { "epoch": 0.987764705882353, "grad_norm": 0.4994530292860533, "learning_rate": 0.0002999372253680244, "loss": 0.6445, "step": 2099 }, { "epoch": 0.9882352941176471, "grad_norm": 0.30740915649622147, "learning_rate": 0.0002999359508794339, "loss": 0.2051, "step": 2100 }, { "epoch": 0.9887058823529412, "grad_norm": 0.34019818098241417, "learning_rate": 0.0002999346635855778, "loss": 0.1836, "step": 2101 }, { "epoch": 0.9891764705882353, "grad_norm": 1.0306174317703332, "learning_rate": 0.00029993336348656594, "loss": 0.8008, "step": 2102 }, { "epoch": 0.9896470588235294, "grad_norm": 0.17071795419322494, "learning_rate": 0.00029993205058250947, "loss": 0.0649, "step": 2103 }, { "epoch": 0.9901176470588235, "grad_norm": 0.6882843972864152, "learning_rate": 0.00029993072487352043, "loss": 0.3691, "step": 2104 }, { "epoch": 0.9905882352941177, "grad_norm": 1.095002124778347, "learning_rate": 0.0002999293863597121, "loss": 0.4766, "step": 2105 }, { "epoch": 0.9910588235294118, "grad_norm": 1.1140078166457148, "learning_rate": 0.00029992803504119876, "loss": 0.5508, "step": 2106 }, { "epoch": 0.9915294117647059, "grad_norm": 0.38639281437621614, "learning_rate": 0.0002999266709180959, "loss": 0.2656, "step": 2107 }, { "epoch": 0.992, "grad_norm": 0.3024992195988211, "learning_rate": 0.0002999252939905199, "loss": 0.1475, "step": 2108 }, { "epoch": 0.9924705882352941, "grad_norm": 0.44599734198815905, "learning_rate": 0.0002999239042585884, "loss": 0.3203, "step": 2109 }, { "epoch": 0.9929411764705882, "grad_norm": 0.6349402121830598, "learning_rate": 0.0002999225017224201, "loss": 0.4941, "step": 2110 }, { "epoch": 0.9934117647058823, "grad_norm": 1.5872603556503992, "learning_rate": 0.0002999210863821349, "loss": 0.293, "step": 2111 }, { "epoch": 0.9938823529411764, "grad_norm": 0.8840887542645859, "learning_rate": 0.0002999196582378535, "loss": 0.7227, "step": 2112 }, { "epoch": 0.9943529411764706, "grad_norm": 0.7727838295073933, "learning_rate": 0.0002999182172896979, "loss": 0.7031, "step": 2113 }, { "epoch": 0.9948235294117647, "grad_norm": 0.2525000849299281, "learning_rate": 0.0002999167635377912, "loss": 0.1069, "step": 2114 }, { "epoch": 0.9952941176470588, "grad_norm": 0.250716154447261, "learning_rate": 0.00029991529698225757, "loss": 0.1592, "step": 2115 }, { "epoch": 0.995764705882353, "grad_norm": 0.8264836865129309, "learning_rate": 0.0002999138176232223, "loss": 0.6562, "step": 2116 }, { "epoch": 0.9962352941176471, "grad_norm": 0.7663781830226135, "learning_rate": 0.00029991232546081156, "loss": 0.4258, "step": 2117 }, { "epoch": 0.9967058823529412, "grad_norm": 0.9093919816261683, "learning_rate": 0.000299910820495153, "loss": 0.5078, "step": 2118 }, { "epoch": 0.9971764705882353, "grad_norm": 0.5203257945187729, "learning_rate": 0.00029990930272637506, "loss": 0.3438, "step": 2119 }, { "epoch": 0.9976470588235294, "grad_norm": 1.323622243146971, "learning_rate": 0.00029990777215460737, "loss": 1.1328, "step": 2120 }, { "epoch": 0.9981176470588236, "grad_norm": 0.8132984304172503, "learning_rate": 0.0002999062287799806, "loss": 0.6719, "step": 2121 }, { "epoch": 0.9985882352941177, "grad_norm": 0.4473889605893884, "learning_rate": 0.0002999046726026266, "loss": 0.416, "step": 2122 }, { "epoch": 0.9990588235294118, "grad_norm": 0.9283392386273461, "learning_rate": 0.00029990310362267833, "loss": 0.6562, "step": 2123 }, { "epoch": 0.9995294117647059, "grad_norm": 0.6461156499244688, "learning_rate": 0.0002999015218402697, "loss": 0.4258, "step": 2124 }, { "epoch": 1.0, "grad_norm": 1.002367603470952, "learning_rate": 0.0002998999272555359, "loss": 0.7891, "step": 2125 }, { "epoch": 1.0, "eval_loss": 0.4773147702217102, "eval_runtime": 61.6519, "eval_samples_per_second": 12.165, "eval_steps_per_second": 6.083, "step": 2125 }, { "epoch": 1.0004705882352942, "grad_norm": 0.3082483854327443, "learning_rate": 0.000299898319868613, "loss": 0.1914, "step": 2126 }, { "epoch": 1.0009411764705882, "grad_norm": 0.5335204750174635, "learning_rate": 0.00029989669967963835, "loss": 0.5469, "step": 2127 }, { "epoch": 1.0014117647058824, "grad_norm": 0.5497151849387469, "learning_rate": 0.00029989506668875033, "loss": 0.3203, "step": 2128 }, { "epoch": 1.0018823529411764, "grad_norm": 0.39173121433208546, "learning_rate": 0.00029989342089608835, "loss": 0.3125, "step": 2129 }, { "epoch": 1.0023529411764707, "grad_norm": 0.45345104133229325, "learning_rate": 0.000299891762301793, "loss": 0.208, "step": 2130 }, { "epoch": 1.0028235294117647, "grad_norm": 1.0856527137445398, "learning_rate": 0.00029989009090600595, "loss": 0.5859, "step": 2131 }, { "epoch": 1.003294117647059, "grad_norm": 0.7408302978586546, "learning_rate": 0.00029988840670887, "loss": 0.5508, "step": 2132 }, { "epoch": 1.003764705882353, "grad_norm": 0.5134448855919611, "learning_rate": 0.0002998867097105288, "loss": 0.7383, "step": 2133 }, { "epoch": 1.0042352941176471, "grad_norm": 0.644948402331713, "learning_rate": 0.00029988499991112747, "loss": 0.5625, "step": 2134 }, { "epoch": 1.0047058823529411, "grad_norm": 0.7889912644180597, "learning_rate": 0.000299883277310812, "loss": 0.6133, "step": 2135 }, { "epoch": 1.0051764705882353, "grad_norm": 0.642081452557094, "learning_rate": 0.0002998815419097294, "loss": 0.1021, "step": 2136 }, { "epoch": 1.0056470588235293, "grad_norm": 0.7846235968624644, "learning_rate": 0.0002998797937080279, "loss": 0.625, "step": 2137 }, { "epoch": 1.0061176470588236, "grad_norm": 0.4779421458288335, "learning_rate": 0.000299878032705857, "loss": 0.4121, "step": 2138 }, { "epoch": 1.0065882352941176, "grad_norm": 0.5012458842879002, "learning_rate": 0.0002998762589033669, "loss": 0.332, "step": 2139 }, { "epoch": 1.0070588235294118, "grad_norm": 0.9258806169188526, "learning_rate": 0.00029987447230070917, "loss": 0.6797, "step": 2140 }, { "epoch": 1.0075294117647058, "grad_norm": 1.2686336192723202, "learning_rate": 0.00029987267289803635, "loss": 0.9336, "step": 2141 }, { "epoch": 1.008, "grad_norm": 0.43352431248549544, "learning_rate": 0.00029987086069550216, "loss": 0.3887, "step": 2142 }, { "epoch": 1.008470588235294, "grad_norm": 0.3739906404234125, "learning_rate": 0.0002998690356932613, "loss": 0.3047, "step": 2143 }, { "epoch": 1.0089411764705882, "grad_norm": 0.7593762711781534, "learning_rate": 0.0002998671978914698, "loss": 0.5703, "step": 2144 }, { "epoch": 1.0094117647058825, "grad_norm": 0.15351379751920438, "learning_rate": 0.00029986534729028444, "loss": 0.0542, "step": 2145 }, { "epoch": 1.0098823529411765, "grad_norm": 0.7700729318016318, "learning_rate": 0.00029986348388986333, "loss": 0.6523, "step": 2146 }, { "epoch": 1.0103529411764707, "grad_norm": 0.7878182512147147, "learning_rate": 0.00029986160769036565, "loss": 0.5508, "step": 2147 }, { "epoch": 1.0108235294117647, "grad_norm": 0.6202104497727321, "learning_rate": 0.0002998597186919515, "loss": 0.5508, "step": 2148 }, { "epoch": 1.011294117647059, "grad_norm": 0.1385848572429925, "learning_rate": 0.0002998578168947824, "loss": 0.0688, "step": 2149 }, { "epoch": 1.011764705882353, "grad_norm": 0.19062496830240022, "learning_rate": 0.00029985590229902073, "loss": 0.0708, "step": 2150 }, { "epoch": 1.0122352941176471, "grad_norm": 0.74685583908396, "learning_rate": 0.0002998539749048299, "loss": 0.6875, "step": 2151 }, { "epoch": 1.0127058823529411, "grad_norm": 0.4547724872210965, "learning_rate": 0.0002998520347123746, "loss": 0.3594, "step": 2152 }, { "epoch": 1.0131764705882353, "grad_norm": 0.26091837965467146, "learning_rate": 0.00029985008172182055, "loss": 0.1523, "step": 2153 }, { "epoch": 1.0136470588235293, "grad_norm": 2.3789975281874334, "learning_rate": 0.00029984811593333446, "loss": 0.8789, "step": 2154 }, { "epoch": 1.0141176470588236, "grad_norm": 1.0045574301687348, "learning_rate": 0.00029984613734708427, "loss": 0.7422, "step": 2155 }, { "epoch": 1.0145882352941176, "grad_norm": 0.5008529722744754, "learning_rate": 0.00029984414596323896, "loss": 0.2793, "step": 2156 }, { "epoch": 1.0150588235294118, "grad_norm": 1.5983412594719995, "learning_rate": 0.00029984214178196863, "loss": 0.6562, "step": 2157 }, { "epoch": 1.0155294117647058, "grad_norm": 0.5890563605824383, "learning_rate": 0.00029984012480344444, "loss": 0.2285, "step": 2158 }, { "epoch": 1.016, "grad_norm": 0.7949052039760869, "learning_rate": 0.0002998380950278386, "loss": 0.4316, "step": 2159 }, { "epoch": 1.016470588235294, "grad_norm": 0.3846571822770932, "learning_rate": 0.00029983605245532453, "loss": 0.2949, "step": 2160 }, { "epoch": 1.0169411764705882, "grad_norm": 0.3277148903981383, "learning_rate": 0.0002998339970860766, "loss": 0.2148, "step": 2161 }, { "epoch": 1.0174117647058825, "grad_norm": 0.9249357419529531, "learning_rate": 0.00029983192892027037, "loss": 0.7695, "step": 2162 }, { "epoch": 1.0178823529411765, "grad_norm": 0.6660536484535448, "learning_rate": 0.0002998298479580825, "loss": 0.3242, "step": 2163 }, { "epoch": 1.0183529411764707, "grad_norm": 1.2671297033247573, "learning_rate": 0.0002998277541996907, "loss": 1.125, "step": 2164 }, { "epoch": 1.0188235294117647, "grad_norm": 0.6095411785050607, "learning_rate": 0.0002998256476452738, "loss": 0.7031, "step": 2165 }, { "epoch": 1.019294117647059, "grad_norm": 0.3904082680043325, "learning_rate": 0.0002998235282950117, "loss": 0.2363, "step": 2166 }, { "epoch": 1.019764705882353, "grad_norm": 1.620579947839787, "learning_rate": 0.0002998213961490854, "loss": 0.5547, "step": 2167 }, { "epoch": 1.0202352941176471, "grad_norm": 0.2664400199205829, "learning_rate": 0.00029981925120767705, "loss": 0.0933, "step": 2168 }, { "epoch": 1.0207058823529411, "grad_norm": 0.9266314979285506, "learning_rate": 0.00029981709347096975, "loss": 0.4082, "step": 2169 }, { "epoch": 1.0211764705882354, "grad_norm": 0.8938352857539457, "learning_rate": 0.00029981492293914777, "loss": 0.4316, "step": 2170 }, { "epoch": 1.0216470588235294, "grad_norm": 0.8125641788125866, "learning_rate": 0.0002998127396123966, "loss": 0.5234, "step": 2171 }, { "epoch": 1.0221176470588236, "grad_norm": 0.8669920538381192, "learning_rate": 0.00029981054349090264, "loss": 0.6641, "step": 2172 }, { "epoch": 1.0225882352941176, "grad_norm": 1.2282131983902662, "learning_rate": 0.0002998083345748534, "loss": 0.7227, "step": 2173 }, { "epoch": 1.0230588235294118, "grad_norm": 0.8219307258929561, "learning_rate": 0.0002998061128644376, "loss": 0.5898, "step": 2174 }, { "epoch": 1.0235294117647058, "grad_norm": 0.7402216977068184, "learning_rate": 0.00029980387835984494, "loss": 0.1406, "step": 2175 }, { "epoch": 1.024, "grad_norm": 0.18046092898695568, "learning_rate": 0.0002998016310612663, "loss": 0.0491, "step": 2176 }, { "epoch": 1.024470588235294, "grad_norm": 0.6617538114659017, "learning_rate": 0.0002997993709688937, "loss": 0.2354, "step": 2177 }, { "epoch": 1.0249411764705882, "grad_norm": 2.1864405873475565, "learning_rate": 0.0002997970980829199, "loss": 0.5703, "step": 2178 }, { "epoch": 1.0254117647058822, "grad_norm": 0.8507863466082145, "learning_rate": 0.0002997948124035393, "loss": 0.3945, "step": 2179 }, { "epoch": 1.0258823529411765, "grad_norm": 0.4327713291357767, "learning_rate": 0.00029979251393094685, "loss": 0.2676, "step": 2180 }, { "epoch": 1.0263529411764707, "grad_norm": 0.3281995067962772, "learning_rate": 0.00029979020266533913, "loss": 0.2217, "step": 2181 }, { "epoch": 1.0268235294117647, "grad_norm": 0.33336468129353974, "learning_rate": 0.0002997878786069133, "loss": 0.2539, "step": 2182 }, { "epoch": 1.027294117647059, "grad_norm": 1.6701325217526124, "learning_rate": 0.0002997855417558679, "loss": 0.8047, "step": 2183 }, { "epoch": 1.027764705882353, "grad_norm": 0.5950981248535397, "learning_rate": 0.0002997831921124026, "loss": 0.3516, "step": 2184 }, { "epoch": 1.0282352941176471, "grad_norm": 0.7081287160507042, "learning_rate": 0.000299780829676718, "loss": 0.3965, "step": 2185 }, { "epoch": 1.0287058823529411, "grad_norm": 0.21442141573702503, "learning_rate": 0.00029977845444901586, "loss": 0.1611, "step": 2186 }, { "epoch": 1.0291764705882354, "grad_norm": 1.2782218081476642, "learning_rate": 0.0002997760664294991, "loss": 0.7656, "step": 2187 }, { "epoch": 1.0296470588235294, "grad_norm": 0.6225184869149138, "learning_rate": 0.00029977366561837154, "loss": 0.4766, "step": 2188 }, { "epoch": 1.0301176470588236, "grad_norm": 0.42126887002782887, "learning_rate": 0.0002997712520158383, "loss": 0.1934, "step": 2189 }, { "epoch": 1.0305882352941176, "grad_norm": 1.1093218986780617, "learning_rate": 0.00029976882562210563, "loss": 1.125, "step": 2190 }, { "epoch": 1.0310588235294118, "grad_norm": 0.5595143828388118, "learning_rate": 0.00029976638643738055, "loss": 0.3203, "step": 2191 }, { "epoch": 1.0315294117647058, "grad_norm": 0.7096293168856254, "learning_rate": 0.0002997639344618714, "loss": 0.4336, "step": 2192 }, { "epoch": 1.032, "grad_norm": 0.5263018000759135, "learning_rate": 0.0002997614696957878, "loss": 0.4023, "step": 2193 }, { "epoch": 1.032470588235294, "grad_norm": 0.48776791524907015, "learning_rate": 0.00029975899213934004, "loss": 0.167, "step": 2194 }, { "epoch": 1.0329411764705883, "grad_norm": 0.7719353432656099, "learning_rate": 0.0002997565017927398, "loss": 0.4844, "step": 2195 }, { "epoch": 1.0334117647058823, "grad_norm": 0.8541807739250561, "learning_rate": 0.00029975399865619977, "loss": 0.4961, "step": 2196 }, { "epoch": 1.0338823529411765, "grad_norm": 0.41783433233041634, "learning_rate": 0.00029975148272993373, "loss": 0.2871, "step": 2197 }, { "epoch": 1.0343529411764707, "grad_norm": 0.47698071293093214, "learning_rate": 0.0002997489540141565, "loss": 0.2334, "step": 2198 }, { "epoch": 1.0348235294117647, "grad_norm": 0.5426671292690118, "learning_rate": 0.00029974641250908406, "loss": 0.6992, "step": 2199 }, { "epoch": 1.035294117647059, "grad_norm": 0.9908349088416134, "learning_rate": 0.0002997438582149335, "loss": 0.9023, "step": 2200 }, { "epoch": 1.035764705882353, "grad_norm": 0.4810520643048888, "learning_rate": 0.000299741291131923, "loss": 0.3984, "step": 2201 }, { "epoch": 1.0362352941176471, "grad_norm": 0.7823929527389105, "learning_rate": 0.00029973871126027173, "loss": 0.6562, "step": 2202 }, { "epoch": 1.0367058823529411, "grad_norm": 2.6892021938589568, "learning_rate": 0.00029973611860020013, "loss": 0.1924, "step": 2203 }, { "epoch": 1.0371764705882354, "grad_norm": 1.1027063547156997, "learning_rate": 0.0002997335131519294, "loss": 0.8555, "step": 2204 }, { "epoch": 1.0376470588235294, "grad_norm": 0.8337245994677625, "learning_rate": 0.0002997308949156823, "loss": 0.3359, "step": 2205 }, { "epoch": 1.0381176470588236, "grad_norm": 0.5223271492502598, "learning_rate": 0.00029972826389168235, "loss": 0.3535, "step": 2206 }, { "epoch": 1.0385882352941176, "grad_norm": 0.24723734348370244, "learning_rate": 0.00029972562008015425, "loss": 0.0522, "step": 2207 }, { "epoch": 1.0390588235294118, "grad_norm": 0.5418777576807218, "learning_rate": 0.0002997229634813238, "loss": 0.4141, "step": 2208 }, { "epoch": 1.0395294117647058, "grad_norm": 0.31773099440888525, "learning_rate": 0.0002997202940954179, "loss": 0.1846, "step": 2209 }, { "epoch": 1.04, "grad_norm": 0.45746892404869927, "learning_rate": 0.00029971761192266447, "loss": 0.2734, "step": 2210 }, { "epoch": 1.040470588235294, "grad_norm": 0.5718111411136858, "learning_rate": 0.0002997149169632926, "loss": 0.2676, "step": 2211 }, { "epoch": 1.0409411764705883, "grad_norm": 0.6257624484540292, "learning_rate": 0.0002997122092175325, "loss": 0.0591, "step": 2212 }, { "epoch": 1.0414117647058823, "grad_norm": 0.7873588159920986, "learning_rate": 0.00029970948868561547, "loss": 0.4707, "step": 2213 }, { "epoch": 1.0418823529411765, "grad_norm": 0.5903894390421355, "learning_rate": 0.00029970675536777365, "loss": 0.5664, "step": 2214 }, { "epoch": 1.0423529411764705, "grad_norm": 0.5418525705333153, "learning_rate": 0.0002997040092642407, "loss": 0.6289, "step": 2215 }, { "epoch": 1.0428235294117647, "grad_norm": 0.5244342535805936, "learning_rate": 0.00029970125037525106, "loss": 0.4531, "step": 2216 }, { "epoch": 1.043294117647059, "grad_norm": 0.36631662487312844, "learning_rate": 0.00029969847870104033, "loss": 0.1963, "step": 2217 }, { "epoch": 1.043764705882353, "grad_norm": 0.47217787810542966, "learning_rate": 0.00029969569424184527, "loss": 0.3301, "step": 2218 }, { "epoch": 1.0442352941176472, "grad_norm": 0.804294613570468, "learning_rate": 0.0002996928969979037, "loss": 0.582, "step": 2219 }, { "epoch": 1.0447058823529412, "grad_norm": 0.8212773052920217, "learning_rate": 0.00029969008696945445, "loss": 0.6562, "step": 2220 }, { "epoch": 1.0451764705882354, "grad_norm": 1.9036615574975395, "learning_rate": 0.00029968726415673753, "loss": 1.7188, "step": 2221 }, { "epoch": 1.0456470588235294, "grad_norm": 0.607619743099934, "learning_rate": 0.00029968442855999406, "loss": 0.4121, "step": 2222 }, { "epoch": 1.0461176470588236, "grad_norm": 0.6876157147953116, "learning_rate": 0.0002996815801794662, "loss": 0.4902, "step": 2223 }, { "epoch": 1.0465882352941176, "grad_norm": 0.9079558709771741, "learning_rate": 0.0002996787190153972, "loss": 0.6523, "step": 2224 }, { "epoch": 1.0470588235294118, "grad_norm": 0.7494813211383777, "learning_rate": 0.0002996758450680314, "loss": 0.4883, "step": 2225 }, { "epoch": 1.0475294117647058, "grad_norm": 0.39942695828761526, "learning_rate": 0.0002996729583376143, "loss": 0.1777, "step": 2226 }, { "epoch": 1.048, "grad_norm": 0.4138342600498021, "learning_rate": 0.00029967005882439244, "loss": 0.3027, "step": 2227 }, { "epoch": 1.048470588235294, "grad_norm": 0.7146707765896653, "learning_rate": 0.0002996671465286134, "loss": 0.4004, "step": 2228 }, { "epoch": 1.0489411764705883, "grad_norm": 0.1095241988265912, "learning_rate": 0.00029966422145052594, "loss": 0.0283, "step": 2229 }, { "epoch": 1.0494117647058823, "grad_norm": 0.9019004827027765, "learning_rate": 0.00029966128359037986, "loss": 0.6602, "step": 2230 }, { "epoch": 1.0498823529411765, "grad_norm": 0.6633273040749156, "learning_rate": 0.0002996583329484261, "loss": 0.543, "step": 2231 }, { "epoch": 1.0503529411764705, "grad_norm": 0.9439047360944087, "learning_rate": 0.00029965536952491665, "loss": 0.4434, "step": 2232 }, { "epoch": 1.0508235294117647, "grad_norm": 1.6619008379315172, "learning_rate": 0.00029965239332010454, "loss": 0.8828, "step": 2233 }, { "epoch": 1.0512941176470587, "grad_norm": 0.2681102669589502, "learning_rate": 0.000299649404334244, "loss": 0.1191, "step": 2234 }, { "epoch": 1.051764705882353, "grad_norm": 0.7681567627112421, "learning_rate": 0.0002996464025675903, "loss": 0.6016, "step": 2235 }, { "epoch": 1.0522352941176472, "grad_norm": 0.34530565661175694, "learning_rate": 0.00029964338802039987, "loss": 0.1768, "step": 2236 }, { "epoch": 1.0527058823529412, "grad_norm": 0.8649267018360983, "learning_rate": 0.00029964036069293006, "loss": 0.3027, "step": 2237 }, { "epoch": 1.0531764705882354, "grad_norm": 0.15259162103839982, "learning_rate": 0.0002996373205854395, "loss": 0.0437, "step": 2238 }, { "epoch": 1.0536470588235294, "grad_norm": 0.6973096905854785, "learning_rate": 0.00029963426769818774, "loss": 0.2637, "step": 2239 }, { "epoch": 1.0541176470588236, "grad_norm": 1.23185671155823, "learning_rate": 0.0002996312020314356, "loss": 0.7617, "step": 2240 }, { "epoch": 1.0545882352941176, "grad_norm": 1.1478325574949058, "learning_rate": 0.00029962812358544485, "loss": 0.8789, "step": 2241 }, { "epoch": 1.0550588235294118, "grad_norm": 1.3693228133481832, "learning_rate": 0.0002996250323604785, "loss": 1.0781, "step": 2242 }, { "epoch": 1.0555294117647058, "grad_norm": 3.8676863594272374e-05, "learning_rate": 0.00029962192835680037, "loss": 0.0, "step": 2243 }, { "epoch": 1.056, "grad_norm": 1.0728778105072077, "learning_rate": 0.0002996188115746757, "loss": 0.582, "step": 2244 }, { "epoch": 1.056470588235294, "grad_norm": 1.1074818043318793, "learning_rate": 0.0002996156820143707, "loss": 0.5977, "step": 2245 }, { "epoch": 1.0569411764705883, "grad_norm": 0.39093988167977656, "learning_rate": 0.00029961253967615255, "loss": 0.2051, "step": 2246 }, { "epoch": 1.0574117647058823, "grad_norm": 1.1192986928143918, "learning_rate": 0.0002996093845602897, "loss": 0.5742, "step": 2247 }, { "epoch": 1.0578823529411765, "grad_norm": 1.0630557813126562, "learning_rate": 0.00029960621666705155, "loss": 0.5273, "step": 2248 }, { "epoch": 1.0583529411764705, "grad_norm": 0.3162515440476433, "learning_rate": 0.00029960303599670866, "loss": 0.168, "step": 2249 }, { "epoch": 1.0588235294117647, "grad_norm": 0.7687292664263685, "learning_rate": 0.0002995998425495327, "loss": 0.4453, "step": 2250 }, { "epoch": 1.0592941176470587, "grad_norm": 0.3568281152996, "learning_rate": 0.0002995966363257964, "loss": 0.1299, "step": 2251 }, { "epoch": 1.059764705882353, "grad_norm": 1.1583457306872151, "learning_rate": 0.0002995934173257736, "loss": 0.8047, "step": 2252 }, { "epoch": 1.0602352941176472, "grad_norm": 1.0753882978657308, "learning_rate": 0.0002995901855497393, "loss": 0.957, "step": 2253 }, { "epoch": 1.0607058823529412, "grad_norm": 1.1889334795857482, "learning_rate": 0.00029958694099796934, "loss": 0.8242, "step": 2254 }, { "epoch": 1.0611764705882354, "grad_norm": 0.8190329099078439, "learning_rate": 0.0002995836836707409, "loss": 0.5039, "step": 2255 }, { "epoch": 1.0616470588235294, "grad_norm": 0.4219868745400458, "learning_rate": 0.00029958041356833217, "loss": 0.3457, "step": 2256 }, { "epoch": 1.0621176470588236, "grad_norm": 0.6924869115142891, "learning_rate": 0.00029957713069102244, "loss": 0.1426, "step": 2257 }, { "epoch": 1.0625882352941176, "grad_norm": 0.4897389791545995, "learning_rate": 0.00029957383503909204, "loss": 0.3398, "step": 2258 }, { "epoch": 1.0630588235294118, "grad_norm": 0.5887977060671058, "learning_rate": 0.00029957052661282256, "loss": 0.2852, "step": 2259 }, { "epoch": 1.0635294117647058, "grad_norm": 0.677738333268181, "learning_rate": 0.0002995672054124964, "loss": 0.3691, "step": 2260 }, { "epoch": 1.064, "grad_norm": 0.4601556089136804, "learning_rate": 0.0002995638714383973, "loss": 0.291, "step": 2261 }, { "epoch": 1.064470588235294, "grad_norm": 0.6694253239423346, "learning_rate": 0.00029956052469080994, "loss": 0.4629, "step": 2262 }, { "epoch": 1.0649411764705883, "grad_norm": 0.9934579530853513, "learning_rate": 0.0002995571651700202, "loss": 0.6953, "step": 2263 }, { "epoch": 1.0654117647058823, "grad_norm": 0.6377833238210402, "learning_rate": 0.000299553792876315, "loss": 0.4004, "step": 2264 }, { "epoch": 1.0658823529411765, "grad_norm": 0.09617063580510633, "learning_rate": 0.0002995504078099823, "loss": 0.0151, "step": 2265 }, { "epoch": 1.0663529411764705, "grad_norm": 0.4306801687900688, "learning_rate": 0.00029954700997131126, "loss": 0.2412, "step": 2266 }, { "epoch": 1.0668235294117647, "grad_norm": 0.5358250193078988, "learning_rate": 0.000299543599360592, "loss": 0.3008, "step": 2267 }, { "epoch": 1.0672941176470587, "grad_norm": 0.5323333029340063, "learning_rate": 0.0002995401759781159, "loss": 0.2695, "step": 2268 }, { "epoch": 1.067764705882353, "grad_norm": 0.8130328105326683, "learning_rate": 0.00029953673982417527, "loss": 0.582, "step": 2269 }, { "epoch": 1.0682352941176472, "grad_norm": 1.061369227522017, "learning_rate": 0.00029953329089906357, "loss": 0.3477, "step": 2270 }, { "epoch": 1.0687058823529412, "grad_norm": 0.7133722043173718, "learning_rate": 0.0002995298292030754, "loss": 0.4785, "step": 2271 }, { "epoch": 1.0691764705882354, "grad_norm": 0.38300912690564215, "learning_rate": 0.00029952635473650633, "loss": 0.2695, "step": 2272 }, { "epoch": 1.0696470588235294, "grad_norm": 0.6666165370739969, "learning_rate": 0.00029952286749965315, "loss": 0.5469, "step": 2273 }, { "epoch": 1.0701176470588236, "grad_norm": 0.31293757746819684, "learning_rate": 0.0002995193674928137, "loss": 0.252, "step": 2274 }, { "epoch": 1.0705882352941176, "grad_norm": 0.7470034288837749, "learning_rate": 0.0002995158547162869, "loss": 0.6758, "step": 2275 }, { "epoch": 1.0710588235294118, "grad_norm": 1.083018726257417, "learning_rate": 0.00029951232917037264, "loss": 0.8789, "step": 2276 }, { "epoch": 1.0715294117647058, "grad_norm": 0.2748544700119028, "learning_rate": 0.0002995087908553722, "loss": 0.1553, "step": 2277 }, { "epoch": 1.072, "grad_norm": 0.4258699522491939, "learning_rate": 0.00029950523977158766, "loss": 0.0674, "step": 2278 }, { "epoch": 1.072470588235294, "grad_norm": 0.9280940374073156, "learning_rate": 0.00029950167591932233, "loss": 0.8008, "step": 2279 }, { "epoch": 1.0729411764705883, "grad_norm": 0.652628291316283, "learning_rate": 0.0002994980992988806, "loss": 0.6094, "step": 2280 }, { "epoch": 1.0734117647058823, "grad_norm": 0.5482725721106498, "learning_rate": 0.0002994945099105679, "loss": 0.3184, "step": 2281 }, { "epoch": 1.0738823529411765, "grad_norm": 0.3926825510196246, "learning_rate": 0.00029949090775469077, "loss": 0.2109, "step": 2282 }, { "epoch": 1.0743529411764705, "grad_norm": 0.8420067545734012, "learning_rate": 0.00029948729283155683, "loss": 0.6641, "step": 2283 }, { "epoch": 1.0748235294117647, "grad_norm": 0.4337272457888629, "learning_rate": 0.0002994836651414749, "loss": 0.291, "step": 2284 }, { "epoch": 1.0752941176470587, "grad_norm": 0.9793800749549587, "learning_rate": 0.00029948002468475475, "loss": 0.9102, "step": 2285 }, { "epoch": 1.075764705882353, "grad_norm": 0.5965423104773878, "learning_rate": 0.0002994763714617073, "loss": 0.625, "step": 2286 }, { "epoch": 1.076235294117647, "grad_norm": 0.4594982928129791, "learning_rate": 0.00029947270547264456, "loss": 0.2314, "step": 2287 }, { "epoch": 1.0767058823529412, "grad_norm": 0.40477086393664496, "learning_rate": 0.00029946902671787955, "loss": 0.2432, "step": 2288 }, { "epoch": 1.0771764705882352, "grad_norm": 10.840159212797637, "learning_rate": 0.00029946533519772657, "loss": 0.7461, "step": 2289 }, { "epoch": 1.0776470588235294, "grad_norm": 0.6302491211202365, "learning_rate": 0.0002994616309125008, "loss": 0.3438, "step": 2290 }, { "epoch": 1.0781176470588236, "grad_norm": 0.9954590261097361, "learning_rate": 0.0002994579138625187, "loss": 0.3789, "step": 2291 }, { "epoch": 1.0785882352941176, "grad_norm": 1.4906812405206666, "learning_rate": 0.00029945418404809767, "loss": 0.4668, "step": 2292 }, { "epoch": 1.0790588235294118, "grad_norm": 1.5527143184491625, "learning_rate": 0.0002994504414695562, "loss": 0.7969, "step": 2293 }, { "epoch": 1.0795294117647058, "grad_norm": 0.7440130029725314, "learning_rate": 0.000299446686127214, "loss": 0.5078, "step": 2294 }, { "epoch": 1.08, "grad_norm": 0.38959135015769414, "learning_rate": 0.00029944291802139184, "loss": 0.2461, "step": 2295 }, { "epoch": 1.080470588235294, "grad_norm": 0.9325181947548646, "learning_rate": 0.00029943913715241137, "loss": 0.5234, "step": 2296 }, { "epoch": 1.0809411764705883, "grad_norm": 1.5843030878007829, "learning_rate": 0.0002994353435205956, "loss": 1.0078, "step": 2297 }, { "epoch": 1.0814117647058823, "grad_norm": 0.28736717824113284, "learning_rate": 0.00029943153712626855, "loss": 0.1504, "step": 2298 }, { "epoch": 1.0818823529411765, "grad_norm": 0.5728349421085469, "learning_rate": 0.0002994277179697553, "loss": 0.2676, "step": 2299 }, { "epoch": 1.0823529411764705, "grad_norm": 0.2547769488486275, "learning_rate": 0.000299423886051382, "loss": 0.1367, "step": 2300 }, { "epoch": 1.0828235294117647, "grad_norm": 3.029171672658658, "learning_rate": 0.00029942004137147586, "loss": 0.8594, "step": 2301 }, { "epoch": 1.0832941176470587, "grad_norm": 0.5790323547344932, "learning_rate": 0.0002994161839303653, "loss": 0.2451, "step": 2302 }, { "epoch": 1.083764705882353, "grad_norm": 1.292636141348619, "learning_rate": 0.00029941231372837974, "loss": 0.5977, "step": 2303 }, { "epoch": 1.084235294117647, "grad_norm": 0.19387889420086823, "learning_rate": 0.0002994084307658497, "loss": 0.0479, "step": 2304 }, { "epoch": 1.0847058823529412, "grad_norm": 0.8622764454912082, "learning_rate": 0.0002994045350431069, "loss": 0.4219, "step": 2305 }, { "epoch": 1.0851764705882352, "grad_norm": 1.522380234102741, "learning_rate": 0.00029940062656048396, "loss": 0.5312, "step": 2306 }, { "epoch": 1.0856470588235294, "grad_norm": 0.2707881544068233, "learning_rate": 0.0002993967053183147, "loss": 0.1357, "step": 2307 }, { "epoch": 1.0861176470588236, "grad_norm": 0.6044387538881097, "learning_rate": 0.000299392771316934, "loss": 0.4531, "step": 2308 }, { "epoch": 1.0865882352941176, "grad_norm": 0.9253284503818813, "learning_rate": 0.00029938882455667786, "loss": 0.582, "step": 2309 }, { "epoch": 1.0870588235294119, "grad_norm": 0.45738436501163054, "learning_rate": 0.00029938486503788337, "loss": 0.3164, "step": 2310 }, { "epoch": 1.0875294117647059, "grad_norm": 0.8504663272470254, "learning_rate": 0.0002993808927608887, "loss": 0.543, "step": 2311 }, { "epoch": 1.088, "grad_norm": 0.8844402780159984, "learning_rate": 0.000299376907726033, "loss": 0.6172, "step": 2312 }, { "epoch": 1.088470588235294, "grad_norm": 0.5547660623837287, "learning_rate": 0.0002993729099336568, "loss": 0.4434, "step": 2313 }, { "epoch": 1.0889411764705883, "grad_norm": 0.33076750922266723, "learning_rate": 0.00029936889938410137, "loss": 0.1436, "step": 2314 }, { "epoch": 1.0894117647058823, "grad_norm": 0.6851068129595647, "learning_rate": 0.0002993648760777092, "loss": 0.5742, "step": 2315 }, { "epoch": 1.0898823529411765, "grad_norm": 1.2153335344644611, "learning_rate": 0.0002993608400148241, "loss": 0.8789, "step": 2316 }, { "epoch": 1.0903529411764705, "grad_norm": 0.7543448961716512, "learning_rate": 0.00029935679119579066, "loss": 0.5, "step": 2317 }, { "epoch": 1.0908235294117647, "grad_norm": 0.46601955760150754, "learning_rate": 0.0002993527296209546, "loss": 0.2139, "step": 2318 }, { "epoch": 1.0912941176470587, "grad_norm": 0.2691284482621596, "learning_rate": 0.0002993486552906629, "loss": 0.1846, "step": 2319 }, { "epoch": 1.091764705882353, "grad_norm": 0.2518828505255079, "learning_rate": 0.00029934456820526344, "loss": 0.1069, "step": 2320 }, { "epoch": 1.092235294117647, "grad_norm": 0.6416776447707003, "learning_rate": 0.00029934046836510537, "loss": 0.3965, "step": 2321 }, { "epoch": 1.0927058823529412, "grad_norm": 0.8136788597804331, "learning_rate": 0.00029933635577053877, "loss": 0.2656, "step": 2322 }, { "epoch": 1.0931764705882352, "grad_norm": 0.834195608040877, "learning_rate": 0.00029933223042191487, "loss": 1.0156, "step": 2323 }, { "epoch": 1.0936470588235294, "grad_norm": 1.0543477267898937, "learning_rate": 0.000299328092319586, "loss": 0.8242, "step": 2324 }, { "epoch": 1.0941176470588236, "grad_norm": 0.6343401578061785, "learning_rate": 0.0002993239414639057, "loss": 0.4512, "step": 2325 }, { "epoch": 1.0945882352941176, "grad_norm": 0.8777914312158983, "learning_rate": 0.0002993197778552283, "loss": 0.5859, "step": 2326 }, { "epoch": 1.0950588235294119, "grad_norm": 1.2723218350805883, "learning_rate": 0.0002993156014939095, "loss": 0.9375, "step": 2327 }, { "epoch": 1.0955294117647059, "grad_norm": 0.3311866115499116, "learning_rate": 0.00029931141238030587, "loss": 0.252, "step": 2328 }, { "epoch": 1.096, "grad_norm": 0.16042464114363406, "learning_rate": 0.00029930721051477533, "loss": 0.0791, "step": 2329 }, { "epoch": 1.096470588235294, "grad_norm": 2.297198679422061, "learning_rate": 0.0002993029958976766, "loss": 1.1484, "step": 2330 }, { "epoch": 1.0969411764705883, "grad_norm": 0.5691823561322112, "learning_rate": 0.00029929876852936973, "loss": 0.1553, "step": 2331 }, { "epoch": 1.0974117647058823, "grad_norm": 1.3943066009962621, "learning_rate": 0.00029929452841021573, "loss": 0.6445, "step": 2332 }, { "epoch": 1.0978823529411765, "grad_norm": 0.9145850221240114, "learning_rate": 0.00029929027554057675, "loss": 0.6484, "step": 2333 }, { "epoch": 1.0983529411764705, "grad_norm": 1.7868645023591108, "learning_rate": 0.0002992860099208159, "loss": 0.625, "step": 2334 }, { "epoch": 1.0988235294117648, "grad_norm": 0.7690639971605028, "learning_rate": 0.0002992817315512976, "loss": 0.3867, "step": 2335 }, { "epoch": 1.0992941176470588, "grad_norm": 1.2517741136121638, "learning_rate": 0.0002992774404323872, "loss": 0.5547, "step": 2336 }, { "epoch": 1.099764705882353, "grad_norm": 0.7078675285612387, "learning_rate": 0.0002992731365644512, "loss": 0.4902, "step": 2337 }, { "epoch": 1.100235294117647, "grad_norm": 0.5609112381135855, "learning_rate": 0.0002992688199478571, "loss": 0.3887, "step": 2338 }, { "epoch": 1.1007058823529412, "grad_norm": 0.7760940497696047, "learning_rate": 0.00029926449058297364, "loss": 0.4102, "step": 2339 }, { "epoch": 1.1011764705882352, "grad_norm": 0.7022481824965607, "learning_rate": 0.0002992601484701706, "loss": 0.2754, "step": 2340 }, { "epoch": 1.1016470588235294, "grad_norm": 0.23750571280866162, "learning_rate": 0.0002992557936098187, "loss": 0.1245, "step": 2341 }, { "epoch": 1.1021176470588236, "grad_norm": 1.899808709237349, "learning_rate": 0.00029925142600229, "loss": 1.0, "step": 2342 }, { "epoch": 1.1025882352941176, "grad_norm": 0.735319399754126, "learning_rate": 0.00029924704564795735, "loss": 0.4844, "step": 2343 }, { "epoch": 1.1030588235294116, "grad_norm": 0.2944309051886778, "learning_rate": 0.000299242652547195, "loss": 0.1128, "step": 2344 }, { "epoch": 1.1035294117647059, "grad_norm": 0.6774747763130411, "learning_rate": 0.0002992382467003781, "loss": 0.2852, "step": 2345 }, { "epoch": 1.104, "grad_norm": 1.9569910663168564, "learning_rate": 0.00029923382810788287, "loss": 0.8633, "step": 2346 }, { "epoch": 1.104470588235294, "grad_norm": 1.2865228483942694, "learning_rate": 0.0002992293967700868, "loss": 0.7031, "step": 2347 }, { "epoch": 1.1049411764705883, "grad_norm": 0.782852444163298, "learning_rate": 0.00029922495268736824, "loss": 0.8984, "step": 2348 }, { "epoch": 1.1054117647058823, "grad_norm": 1.159160227562001, "learning_rate": 0.0002992204958601068, "loss": 0.6211, "step": 2349 }, { "epoch": 1.1058823529411765, "grad_norm": 0.6179473245326553, "learning_rate": 0.0002992160262886831, "loss": 0.2334, "step": 2350 }, { "epoch": 1.1063529411764705, "grad_norm": 0.3156179031714266, "learning_rate": 0.0002992115439734788, "loss": 0.1279, "step": 2351 }, { "epoch": 1.1068235294117648, "grad_norm": 0.9216834720786508, "learning_rate": 0.0002992070489148768, "loss": 0.5781, "step": 2352 }, { "epoch": 1.1072941176470588, "grad_norm": 0.7295827528945122, "learning_rate": 0.000299202541113261, "loss": 0.4453, "step": 2353 }, { "epoch": 1.107764705882353, "grad_norm": 0.490516004460376, "learning_rate": 0.0002991980205690163, "loss": 0.2207, "step": 2354 }, { "epoch": 1.108235294117647, "grad_norm": 0.8733208225629261, "learning_rate": 0.0002991934872825289, "loss": 0.5469, "step": 2355 }, { "epoch": 1.1087058823529412, "grad_norm": 0.75638918391718, "learning_rate": 0.0002991889412541858, "loss": 0.5039, "step": 2356 }, { "epoch": 1.1091764705882352, "grad_norm": 0.9352445211112369, "learning_rate": 0.0002991843824843754, "loss": 0.6719, "step": 2357 }, { "epoch": 1.1096470588235294, "grad_norm": 0.45906246398047706, "learning_rate": 0.000299179810973487, "loss": 0.1914, "step": 2358 }, { "epoch": 1.1101176470588234, "grad_norm": 0.6420819460499354, "learning_rate": 0.000299175226721911, "loss": 0.5508, "step": 2359 }, { "epoch": 1.1105882352941177, "grad_norm": 0.9508187728242627, "learning_rate": 0.00029917062973003897, "loss": 0.6523, "step": 2360 }, { "epoch": 1.1110588235294117, "grad_norm": 0.6103589922814426, "learning_rate": 0.0002991660199982634, "loss": 0.4805, "step": 2361 }, { "epoch": 1.1115294117647059, "grad_norm": 0.9978719514969167, "learning_rate": 0.0002991613975269782, "loss": 1.0, "step": 2362 }, { "epoch": 1.112, "grad_norm": 0.8203502536213675, "learning_rate": 0.0002991567623165779, "loss": 0.8359, "step": 2363 }, { "epoch": 1.112470588235294, "grad_norm": 0.6215742782751649, "learning_rate": 0.00029915211436745855, "loss": 0.625, "step": 2364 }, { "epoch": 1.1129411764705883, "grad_norm": 0.6504345713578128, "learning_rate": 0.00029914745368001704, "loss": 0.3672, "step": 2365 }, { "epoch": 1.1134117647058823, "grad_norm": 0.9060427359438034, "learning_rate": 0.0002991427802546514, "loss": 0.5273, "step": 2366 }, { "epoch": 1.1138823529411765, "grad_norm": 0.48882230313720926, "learning_rate": 0.0002991380940917608, "loss": 0.2236, "step": 2367 }, { "epoch": 1.1143529411764705, "grad_norm": 0.6122676124968307, "learning_rate": 0.0002991333951917454, "loss": 0.5352, "step": 2368 }, { "epoch": 1.1148235294117648, "grad_norm": 0.3851709995316498, "learning_rate": 0.00029912868355500666, "loss": 0.1855, "step": 2369 }, { "epoch": 1.1152941176470588, "grad_norm": 0.5271265402215474, "learning_rate": 0.0002991239591819468, "loss": 0.2383, "step": 2370 }, { "epoch": 1.115764705882353, "grad_norm": 0.5200088160491695, "learning_rate": 0.00029911922207296937, "loss": 0.2139, "step": 2371 }, { "epoch": 1.116235294117647, "grad_norm": 0.4900575599117993, "learning_rate": 0.00029911447222847893, "loss": 0.3613, "step": 2372 }, { "epoch": 1.1167058823529412, "grad_norm": 0.5755375697010797, "learning_rate": 0.00029910970964888126, "loss": 0.4883, "step": 2373 }, { "epoch": 1.1171764705882352, "grad_norm": 0.7372676187324112, "learning_rate": 0.0002991049343345829, "loss": 0.5508, "step": 2374 }, { "epoch": 1.1176470588235294, "grad_norm": 0.2629104798255747, "learning_rate": 0.00029910014628599184, "loss": 0.1455, "step": 2375 }, { "epoch": 1.1181176470588234, "grad_norm": 0.7566910053781494, "learning_rate": 0.00029909534550351695, "loss": 0.5977, "step": 2376 }, { "epoch": 1.1185882352941177, "grad_norm": 0.7366063477592191, "learning_rate": 0.00029909053198756827, "loss": 0.6016, "step": 2377 }, { "epoch": 1.1190588235294117, "grad_norm": 0.7833368721324311, "learning_rate": 0.00029908570573855683, "loss": 0.6016, "step": 2378 }, { "epoch": 1.1195294117647059, "grad_norm": 0.37897318006936165, "learning_rate": 0.0002990808667568949, "loss": 0.2715, "step": 2379 }, { "epoch": 1.12, "grad_norm": 1.0313212533524416, "learning_rate": 0.0002990760150429957, "loss": 0.7617, "step": 2380 }, { "epoch": 1.120470588235294, "grad_norm": 0.4844683306382725, "learning_rate": 0.0002990711505972736, "loss": 0.2793, "step": 2381 }, { "epoch": 1.1209411764705883, "grad_norm": 0.4000087839301186, "learning_rate": 0.00029906627342014407, "loss": 0.3359, "step": 2382 }, { "epoch": 1.1214117647058823, "grad_norm": 0.6957953036076342, "learning_rate": 0.00029906138351202356, "loss": 0.5312, "step": 2383 }, { "epoch": 1.1218823529411766, "grad_norm": 0.6609176350080997, "learning_rate": 0.0002990564808733298, "loss": 0.4453, "step": 2384 }, { "epoch": 1.1223529411764706, "grad_norm": 0.83047218895649, "learning_rate": 0.00029905156550448145, "loss": 0.8164, "step": 2385 }, { "epoch": 1.1228235294117648, "grad_norm": 0.7016355096822686, "learning_rate": 0.00029904663740589827, "loss": 0.4844, "step": 2386 }, { "epoch": 1.1232941176470588, "grad_norm": 0.6268462057933534, "learning_rate": 0.0002990416965780012, "loss": 0.3789, "step": 2387 }, { "epoch": 1.123764705882353, "grad_norm": 0.5152672897129131, "learning_rate": 0.0002990367430212123, "loss": 0.1797, "step": 2388 }, { "epoch": 1.124235294117647, "grad_norm": 0.5094492465345267, "learning_rate": 0.0002990317767359544, "loss": 0.2393, "step": 2389 }, { "epoch": 1.1247058823529412, "grad_norm": 0.5455385202945487, "learning_rate": 0.0002990267977226518, "loss": 0.2158, "step": 2390 }, { "epoch": 1.1251764705882352, "grad_norm": 0.317842466819763, "learning_rate": 0.00029902180598172977, "loss": 0.1826, "step": 2391 }, { "epoch": 1.1256470588235294, "grad_norm": 1.048502028398976, "learning_rate": 0.00029901680151361447, "loss": 0.6328, "step": 2392 }, { "epoch": 1.1261176470588234, "grad_norm": 0.7159603470716766, "learning_rate": 0.00029901178431873346, "loss": 0.5469, "step": 2393 }, { "epoch": 1.1265882352941177, "grad_norm": 0.7254013053266671, "learning_rate": 0.0002990067543975151, "loss": 0.4688, "step": 2394 }, { "epoch": 1.1270588235294117, "grad_norm": 1.3447625318422918, "learning_rate": 0.0002990017117503891, "loss": 0.7812, "step": 2395 }, { "epoch": 1.127529411764706, "grad_norm": 1.366783435789541, "learning_rate": 0.0002989966563777861, "loss": 1.0547, "step": 2396 }, { "epoch": 1.1280000000000001, "grad_norm": 1.4870901080325236, "learning_rate": 0.00029899158828013774, "loss": 0.7773, "step": 2397 }, { "epoch": 1.1284705882352941, "grad_norm": 0.7581030356779755, "learning_rate": 0.000298986507457877, "loss": 0.5039, "step": 2398 }, { "epoch": 1.1289411764705881, "grad_norm": 0.38688310713278506, "learning_rate": 0.0002989814139114377, "loss": 0.168, "step": 2399 }, { "epoch": 1.1294117647058823, "grad_norm": 0.2784728144330515, "learning_rate": 0.0002989763076412549, "loss": 0.0869, "step": 2400 }, { "epoch": 1.1298823529411766, "grad_norm": 0.5650539158541238, "learning_rate": 0.00029897118864776476, "loss": 0.3809, "step": 2401 }, { "epoch": 1.1303529411764706, "grad_norm": 1.5534731679287923, "learning_rate": 0.00029896605693140437, "loss": 0.9844, "step": 2402 }, { "epoch": 1.1308235294117648, "grad_norm": 0.2284542311325825, "learning_rate": 0.0002989609124926121, "loss": 0.1338, "step": 2403 }, { "epoch": 1.1312941176470588, "grad_norm": 1.5913774590891863, "learning_rate": 0.0002989557553318272, "loss": 0.8125, "step": 2404 }, { "epoch": 1.131764705882353, "grad_norm": 0.5211090767123966, "learning_rate": 0.0002989505854494902, "loss": 0.3379, "step": 2405 }, { "epoch": 1.132235294117647, "grad_norm": 0.580876660006031, "learning_rate": 0.00029894540284604255, "loss": 0.3926, "step": 2406 }, { "epoch": 1.1327058823529412, "grad_norm": 1.1714946546635243, "learning_rate": 0.000298940207521927, "loss": 0.7461, "step": 2407 }, { "epoch": 1.1331764705882352, "grad_norm": 0.977049080153735, "learning_rate": 0.00029893499947758706, "loss": 0.6406, "step": 2408 }, { "epoch": 1.1336470588235295, "grad_norm": 0.3731786458609919, "learning_rate": 0.00029892977871346776, "loss": 0.2314, "step": 2409 }, { "epoch": 1.1341176470588235, "grad_norm": 0.43443390746471344, "learning_rate": 0.00029892454523001484, "loss": 0.3359, "step": 2410 }, { "epoch": 1.1345882352941177, "grad_norm": 0.2138286586051741, "learning_rate": 0.00029891929902767526, "loss": 0.0952, "step": 2411 }, { "epoch": 1.1350588235294117, "grad_norm": 0.579300079810714, "learning_rate": 0.0002989140401068971, "loss": 0.416, "step": 2412 }, { "epoch": 1.135529411764706, "grad_norm": 0.6078823093922304, "learning_rate": 0.0002989087684681295, "loss": 0.4082, "step": 2413 }, { "epoch": 1.1360000000000001, "grad_norm": 0.5614640413072002, "learning_rate": 0.00029890348411182267, "loss": 0.3965, "step": 2414 }, { "epoch": 1.1364705882352941, "grad_norm": 0.7729312289056457, "learning_rate": 0.00029889818703842796, "loss": 0.373, "step": 2415 }, { "epoch": 1.1369411764705881, "grad_norm": 1.1993869771699024, "learning_rate": 0.0002988928772483977, "loss": 0.8516, "step": 2416 }, { "epoch": 1.1374117647058823, "grad_norm": 0.5981602156492171, "learning_rate": 0.0002988875547421854, "loss": 0.4434, "step": 2417 }, { "epoch": 1.1378823529411766, "grad_norm": 1.020963437421633, "learning_rate": 0.00029888221952024567, "loss": 0.5547, "step": 2418 }, { "epoch": 1.1383529411764706, "grad_norm": 0.68669937834542, "learning_rate": 0.00029887687158303416, "loss": 0.2539, "step": 2419 }, { "epoch": 1.1388235294117648, "grad_norm": 0.8469695025859382, "learning_rate": 0.00029887151093100755, "loss": 0.2598, "step": 2420 }, { "epoch": 1.1392941176470588, "grad_norm": 0.5385590358573453, "learning_rate": 0.00029886613756462365, "loss": 0.3457, "step": 2421 }, { "epoch": 1.139764705882353, "grad_norm": 0.4108224957513745, "learning_rate": 0.00029886075148434146, "loss": 0.3691, "step": 2422 }, { "epoch": 1.140235294117647, "grad_norm": 0.6310196055116304, "learning_rate": 0.0002988553526906209, "loss": 0.4629, "step": 2423 }, { "epoch": 1.1407058823529412, "grad_norm": 1.1739143298801102, "learning_rate": 0.0002988499411839232, "loss": 0.6133, "step": 2424 }, { "epoch": 1.1411764705882352, "grad_norm": 0.4120792930759276, "learning_rate": 0.0002988445169647103, "loss": 0.2393, "step": 2425 }, { "epoch": 1.1416470588235295, "grad_norm": 0.6127250220593682, "learning_rate": 0.00029883908003344554, "loss": 0.4492, "step": 2426 }, { "epoch": 1.1421176470588235, "grad_norm": 0.607042317192988, "learning_rate": 0.00029883363039059343, "loss": 0.3145, "step": 2427 }, { "epoch": 1.1425882352941177, "grad_norm": 0.3398279419763395, "learning_rate": 0.0002988281680366192, "loss": 0.1689, "step": 2428 }, { "epoch": 1.1430588235294117, "grad_norm": 0.722172166977404, "learning_rate": 0.0002988226929719894, "loss": 0.3945, "step": 2429 }, { "epoch": 1.143529411764706, "grad_norm": 0.7182509471054064, "learning_rate": 0.0002988172051971717, "loss": 0.3906, "step": 2430 }, { "epoch": 1.144, "grad_norm": 0.44317860646763124, "learning_rate": 0.0002988117047126346, "loss": 0.2275, "step": 2431 }, { "epoch": 1.1444705882352941, "grad_norm": 0.7635362847320891, "learning_rate": 0.00029880619151884813, "loss": 0.4336, "step": 2432 }, { "epoch": 1.1449411764705881, "grad_norm": 0.405168560625751, "learning_rate": 0.000298800665616283, "loss": 0.2598, "step": 2433 }, { "epoch": 1.1454117647058824, "grad_norm": 0.6973125928327072, "learning_rate": 0.00029879512700541115, "loss": 0.6875, "step": 2434 }, { "epoch": 1.1458823529411766, "grad_norm": 0.9377509867502763, "learning_rate": 0.0002987895756867056, "loss": 0.5547, "step": 2435 }, { "epoch": 1.1463529411764706, "grad_norm": 0.6246041192530115, "learning_rate": 0.00029878401166064046, "loss": 0.2617, "step": 2436 }, { "epoch": 1.1468235294117648, "grad_norm": 2.190668823555416, "learning_rate": 0.000298778434927691, "loss": 0.9414, "step": 2437 }, { "epoch": 1.1472941176470588, "grad_norm": 0.7168509676020413, "learning_rate": 0.0002987728454883334, "loss": 0.6289, "step": 2438 }, { "epoch": 1.147764705882353, "grad_norm": 0.38682757827171266, "learning_rate": 0.00029876724334304506, "loss": 0.2158, "step": 2439 }, { "epoch": 1.148235294117647, "grad_norm": 0.4363529239536492, "learning_rate": 0.00029876162849230446, "loss": 0.2559, "step": 2440 }, { "epoch": 1.1487058823529412, "grad_norm": 2.455842008074943, "learning_rate": 0.00029875600093659115, "loss": 0.5898, "step": 2441 }, { "epoch": 1.1491764705882352, "grad_norm": 1.255402425297146, "learning_rate": 0.00029875036067638565, "loss": 0.4043, "step": 2442 }, { "epoch": 1.1496470588235295, "grad_norm": 1.2351187684299532, "learning_rate": 0.00029874470771216983, "loss": 0.5547, "step": 2443 }, { "epoch": 1.1501176470588235, "grad_norm": 0.5439907811000767, "learning_rate": 0.0002987390420444262, "loss": 0.25, "step": 2444 }, { "epoch": 1.1505882352941177, "grad_norm": 1.3528801805717021, "learning_rate": 0.00029873336367363903, "loss": 0.457, "step": 2445 }, { "epoch": 1.1510588235294117, "grad_norm": 0.9427046436254591, "learning_rate": 0.00029872767260029294, "loss": 0.4395, "step": 2446 }, { "epoch": 1.151529411764706, "grad_norm": 1.0937756534216658, "learning_rate": 0.00029872196882487417, "loss": 0.457, "step": 2447 }, { "epoch": 1.152, "grad_norm": 0.880336376096104, "learning_rate": 0.00029871625234786974, "loss": 0.2988, "step": 2448 }, { "epoch": 1.1524705882352941, "grad_norm": 0.6314695683819336, "learning_rate": 0.0002987105231697679, "loss": 0.377, "step": 2449 }, { "epoch": 1.1529411764705881, "grad_norm": 0.29680866126149835, "learning_rate": 0.000298704781291058, "loss": 0.208, "step": 2450 }, { "epoch": 1.1534117647058824, "grad_norm": 0.7633092327705396, "learning_rate": 0.0002986990267122304, "loss": 0.3535, "step": 2451 }, { "epoch": 1.1538823529411766, "grad_norm": 1.6543423780202273, "learning_rate": 0.0002986932594337765, "loss": 0.6914, "step": 2452 }, { "epoch": 1.1543529411764706, "grad_norm": 0.5559355576631164, "learning_rate": 0.0002986874794561889, "loss": 0.3105, "step": 2453 }, { "epoch": 1.1548235294117646, "grad_norm": 1.100417828984083, "learning_rate": 0.0002986816867799613, "loss": 0.7461, "step": 2454 }, { "epoch": 1.1552941176470588, "grad_norm": 0.37758294652401964, "learning_rate": 0.0002986758814055883, "loss": 0.3867, "step": 2455 }, { "epoch": 1.155764705882353, "grad_norm": 0.6953551724862149, "learning_rate": 0.0002986700633335659, "loss": 0.5352, "step": 2456 }, { "epoch": 1.156235294117647, "grad_norm": 1.1391354423171363, "learning_rate": 0.0002986642325643908, "loss": 0.5664, "step": 2457 }, { "epoch": 1.1567058823529413, "grad_norm": 0.3853179334998502, "learning_rate": 0.00029865838909856103, "loss": 0.2334, "step": 2458 }, { "epoch": 1.1571764705882353, "grad_norm": 0.43804022381754326, "learning_rate": 0.0002986525329365757, "loss": 0.3223, "step": 2459 }, { "epoch": 1.1576470588235295, "grad_norm": 1.0248411198747276, "learning_rate": 0.00029864666407893487, "loss": 0.8555, "step": 2460 }, { "epoch": 1.1581176470588235, "grad_norm": 0.7448906432423426, "learning_rate": 0.00029864078252613983, "loss": 0.6875, "step": 2461 }, { "epoch": 1.1585882352941177, "grad_norm": 0.44662525463760566, "learning_rate": 0.0002986348882786929, "loss": 0.3984, "step": 2462 }, { "epoch": 1.1590588235294117, "grad_norm": 0.31242692725260046, "learning_rate": 0.00029862898133709747, "loss": 0.1885, "step": 2463 }, { "epoch": 1.159529411764706, "grad_norm": 0.7088619592256338, "learning_rate": 0.000298623061701858, "loss": 0.2031, "step": 2464 }, { "epoch": 1.16, "grad_norm": 0.7115456152114484, "learning_rate": 0.00029861712937348007, "loss": 0.5117, "step": 2465 }, { "epoch": 1.1604705882352941, "grad_norm": 1.1280569813001848, "learning_rate": 0.0002986111843524703, "loss": 0.7969, "step": 2466 }, { "epoch": 1.1609411764705881, "grad_norm": 0.6337351304263517, "learning_rate": 0.00029860522663933655, "loss": 0.4941, "step": 2467 }, { "epoch": 1.1614117647058824, "grad_norm": 0.791356866629777, "learning_rate": 0.00029859925623458745, "loss": 0.2734, "step": 2468 }, { "epoch": 1.1618823529411766, "grad_norm": 1.1416830260915845, "learning_rate": 0.000298593273138733, "loss": 0.6367, "step": 2469 }, { "epoch": 1.1623529411764706, "grad_norm": 0.43850953211802346, "learning_rate": 0.00029858727735228416, "loss": 0.2793, "step": 2470 }, { "epoch": 1.1628235294117646, "grad_norm": 0.4278416313101789, "learning_rate": 0.00029858126887575306, "loss": 0.291, "step": 2471 }, { "epoch": 1.1632941176470588, "grad_norm": 0.8631295484422393, "learning_rate": 0.00029857524770965285, "loss": 0.543, "step": 2472 }, { "epoch": 1.163764705882353, "grad_norm": 0.4128286359552405, "learning_rate": 0.00029856921385449763, "loss": 0.3125, "step": 2473 }, { "epoch": 1.164235294117647, "grad_norm": 0.9675626959084614, "learning_rate": 0.0002985631673108029, "loss": 0.6445, "step": 2474 }, { "epoch": 1.1647058823529413, "grad_norm": 0.6746168313819753, "learning_rate": 0.00029855710807908493, "loss": 0.3926, "step": 2475 }, { "epoch": 1.1651764705882353, "grad_norm": 0.6812536793018557, "learning_rate": 0.00029855103615986133, "loss": 0.334, "step": 2476 }, { "epoch": 1.1656470588235295, "grad_norm": 0.9712947527695285, "learning_rate": 0.0002985449515536505, "loss": 0.8516, "step": 2477 }, { "epoch": 1.1661176470588235, "grad_norm": 0.6630080805762254, "learning_rate": 0.0002985388542609723, "loss": 0.4199, "step": 2478 }, { "epoch": 1.1665882352941177, "grad_norm": 0.9205491671356729, "learning_rate": 0.0002985327442823473, "loss": 0.5234, "step": 2479 }, { "epoch": 1.1670588235294117, "grad_norm": 0.47535985918092294, "learning_rate": 0.00029852662161829744, "loss": 0.1816, "step": 2480 }, { "epoch": 1.167529411764706, "grad_norm": 1.5532485903057573, "learning_rate": 0.0002985204862693455, "loss": 0.918, "step": 2481 }, { "epoch": 1.168, "grad_norm": 0.5294738297628215, "learning_rate": 0.00029851433823601563, "loss": 0.2871, "step": 2482 }, { "epoch": 1.1684705882352941, "grad_norm": 0.533177842137827, "learning_rate": 0.00029850817751883276, "loss": 0.4141, "step": 2483 }, { "epoch": 1.1689411764705882, "grad_norm": 0.8944204367189023, "learning_rate": 0.0002985020041183231, "loss": 0.6445, "step": 2484 }, { "epoch": 1.1694117647058824, "grad_norm": 0.19381547593014445, "learning_rate": 0.00029849581803501393, "loss": 0.1641, "step": 2485 }, { "epoch": 1.1698823529411766, "grad_norm": 0.41424464737150507, "learning_rate": 0.0002984896192694335, "loss": 0.252, "step": 2486 }, { "epoch": 1.1703529411764706, "grad_norm": 0.2502656775329353, "learning_rate": 0.00029848340782211125, "loss": 0.1172, "step": 2487 }, { "epoch": 1.1708235294117646, "grad_norm": 1.2065034193878823, "learning_rate": 0.0002984771836935777, "loss": 0.543, "step": 2488 }, { "epoch": 1.1712941176470588, "grad_norm": 1.4915350206825875, "learning_rate": 0.0002984709468843644, "loss": 0.4961, "step": 2489 }, { "epoch": 1.171764705882353, "grad_norm": 0.6481091242891962, "learning_rate": 0.0002984646973950039, "loss": 0.5234, "step": 2490 }, { "epoch": 1.172235294117647, "grad_norm": 0.7411198224760289, "learning_rate": 0.00029845843522603006, "loss": 0.2539, "step": 2491 }, { "epoch": 1.1727058823529413, "grad_norm": 0.3126815689280448, "learning_rate": 0.0002984521603779777, "loss": 0.0708, "step": 2492 }, { "epoch": 1.1731764705882353, "grad_norm": 0.7774889945389237, "learning_rate": 0.0002984458728513827, "loss": 0.75, "step": 2493 }, { "epoch": 1.1736470588235295, "grad_norm": 1.1066288226558159, "learning_rate": 0.000298439572646782, "loss": 0.6484, "step": 2494 }, { "epoch": 1.1741176470588235, "grad_norm": 0.9174823689875835, "learning_rate": 0.0002984332597647137, "loss": 0.5156, "step": 2495 }, { "epoch": 1.1745882352941177, "grad_norm": 0.6590030329087833, "learning_rate": 0.000298426934205717, "loss": 0.4316, "step": 2496 }, { "epoch": 1.1750588235294117, "grad_norm": 0.9087770736522945, "learning_rate": 0.00029842059597033204, "loss": 0.5898, "step": 2497 }, { "epoch": 1.175529411764706, "grad_norm": 1.4302066552621568, "learning_rate": 0.0002984142450591002, "loss": 0.8203, "step": 2498 }, { "epoch": 1.176, "grad_norm": 0.706502955573733, "learning_rate": 0.0002984078814725639, "loss": 0.5117, "step": 2499 }, { "epoch": 1.1764705882352942, "grad_norm": 0.8158176086782409, "learning_rate": 0.0002984015052112665, "loss": 0.543, "step": 2500 }, { "epoch": 1.1769411764705882, "grad_norm": 0.22703390371537752, "learning_rate": 0.0002983951162757527, "loss": 0.0854, "step": 2501 }, { "epoch": 1.1774117647058824, "grad_norm": 1.5871182320815276, "learning_rate": 0.0002983887146665681, "loss": 0.7383, "step": 2502 }, { "epoch": 1.1778823529411764, "grad_norm": 0.6614031554002686, "learning_rate": 0.00029838230038425944, "loss": 0.459, "step": 2503 }, { "epoch": 1.1783529411764706, "grad_norm": 0.25223370362948794, "learning_rate": 0.0002983758734293745, "loss": 0.0981, "step": 2504 }, { "epoch": 1.1788235294117646, "grad_norm": 1.2423280912715027, "learning_rate": 0.00029836943380246215, "loss": 0.5625, "step": 2505 }, { "epoch": 1.1792941176470588, "grad_norm": 1.5322413653468328, "learning_rate": 0.00029836298150407245, "loss": 0.8867, "step": 2506 }, { "epoch": 1.179764705882353, "grad_norm": 0.5565476353106314, "learning_rate": 0.00029835651653475643, "loss": 0.3047, "step": 2507 }, { "epoch": 1.180235294117647, "grad_norm": 0.31177151678968246, "learning_rate": 0.00029835003889506616, "loss": 0.1113, "step": 2508 }, { "epoch": 1.1807058823529413, "grad_norm": 0.802187432559577, "learning_rate": 0.000298343548585555, "loss": 0.4434, "step": 2509 }, { "epoch": 1.1811764705882353, "grad_norm": 0.4896368407407459, "learning_rate": 0.0002983370456067771, "loss": 0.1885, "step": 2510 }, { "epoch": 1.1816470588235295, "grad_norm": 0.6585776236103731, "learning_rate": 0.0002983305299592879, "loss": 0.3516, "step": 2511 }, { "epoch": 1.1821176470588235, "grad_norm": 1.7052230061239466, "learning_rate": 0.0002983240016436439, "loss": 0.9961, "step": 2512 }, { "epoch": 1.1825882352941177, "grad_norm": 0.48301768307215626, "learning_rate": 0.0002983174606604027, "loss": 0.2695, "step": 2513 }, { "epoch": 1.1830588235294117, "grad_norm": 0.7264183878638125, "learning_rate": 0.0002983109070101228, "loss": 0.6484, "step": 2514 }, { "epoch": 1.183529411764706, "grad_norm": 1.2123483202626806, "learning_rate": 0.00029830434069336403, "loss": 0.8203, "step": 2515 }, { "epoch": 1.184, "grad_norm": 0.36814809297524226, "learning_rate": 0.0002982977617106871, "loss": 0.1475, "step": 2516 }, { "epoch": 1.1844705882352942, "grad_norm": 0.8374498347173281, "learning_rate": 0.0002982911700626539, "loss": 0.5742, "step": 2517 }, { "epoch": 1.1849411764705882, "grad_norm": 0.6876904803958768, "learning_rate": 0.0002982845657498274, "loss": 0.3555, "step": 2518 }, { "epoch": 1.1854117647058824, "grad_norm": 0.6770046250676408, "learning_rate": 0.0002982779487727718, "loss": 0.5352, "step": 2519 }, { "epoch": 1.1858823529411764, "grad_norm": 0.4878792801358265, "learning_rate": 0.0002982713191320519, "loss": 0.3066, "step": 2520 }, { "epoch": 1.1863529411764706, "grad_norm": 0.678472356200618, "learning_rate": 0.00029826467682823415, "loss": 0.5391, "step": 2521 }, { "epoch": 1.1868235294117646, "grad_norm": 0.33452582121601065, "learning_rate": 0.0002982580218618858, "loss": 0.1196, "step": 2522 }, { "epoch": 1.1872941176470588, "grad_norm": 1.111939587776276, "learning_rate": 0.0002982513542335751, "loss": 0.5352, "step": 2523 }, { "epoch": 1.187764705882353, "grad_norm": 0.504554997545113, "learning_rate": 0.00029824467394387166, "loss": 0.2988, "step": 2524 }, { "epoch": 1.188235294117647, "grad_norm": 0.5644154626342992, "learning_rate": 0.0002982379809933459, "loss": 0.3574, "step": 2525 }, { "epoch": 1.188705882352941, "grad_norm": 0.5667464005828541, "learning_rate": 0.00029823127538256943, "loss": 0.3164, "step": 2526 }, { "epoch": 1.1891764705882353, "grad_norm": 0.898291704505829, "learning_rate": 0.000298224557112115, "loss": 0.6445, "step": 2527 }, { "epoch": 1.1896470588235295, "grad_norm": 0.4462940980086925, "learning_rate": 0.0002982178261825563, "loss": 0.3242, "step": 2528 }, { "epoch": 1.1901176470588235, "grad_norm": 0.3904932971899054, "learning_rate": 0.0002982110825944683, "loss": 0.3125, "step": 2529 }, { "epoch": 1.1905882352941177, "grad_norm": 0.3344792874328435, "learning_rate": 0.0002982043263484269, "loss": 0.1494, "step": 2530 }, { "epoch": 1.1910588235294117, "grad_norm": 0.29655387179443804, "learning_rate": 0.00029819755744500904, "loss": 0.1172, "step": 2531 }, { "epoch": 1.191529411764706, "grad_norm": 0.9179679810918975, "learning_rate": 0.0002981907758847929, "loss": 0.543, "step": 2532 }, { "epoch": 1.192, "grad_norm": 0.5528624681090901, "learning_rate": 0.0002981839816683576, "loss": 0.4863, "step": 2533 }, { "epoch": 1.1924705882352942, "grad_norm": 0.03497695004587073, "learning_rate": 0.00029817717479628346, "loss": 0.0075, "step": 2534 }, { "epoch": 1.1929411764705882, "grad_norm": 0.38625618240963255, "learning_rate": 0.00029817035526915177, "loss": 0.1514, "step": 2535 }, { "epoch": 1.1934117647058824, "grad_norm": 0.41406655964866257, "learning_rate": 0.00029816352308754496, "loss": 0.2266, "step": 2536 }, { "epoch": 1.1938823529411764, "grad_norm": 0.6159798173770084, "learning_rate": 0.00029815667825204654, "loss": 0.4512, "step": 2537 }, { "epoch": 1.1943529411764706, "grad_norm": 0.8951566593819397, "learning_rate": 0.0002981498207632411, "loss": 0.6055, "step": 2538 }, { "epoch": 1.1948235294117646, "grad_norm": 0.7948873854972002, "learning_rate": 0.0002981429506217143, "loss": 0.5508, "step": 2539 }, { "epoch": 1.1952941176470588, "grad_norm": 0.357347236928761, "learning_rate": 0.0002981360678280529, "loss": 0.2656, "step": 2540 }, { "epoch": 1.195764705882353, "grad_norm": 0.3973128147984877, "learning_rate": 0.00029812917238284466, "loss": 0.1504, "step": 2541 }, { "epoch": 1.196235294117647, "grad_norm": 0.43359132646870097, "learning_rate": 0.00029812226428667854, "loss": 0.0986, "step": 2542 }, { "epoch": 1.196705882352941, "grad_norm": 1.8610015686164703, "learning_rate": 0.0002981153435401445, "loss": 0.6094, "step": 2543 }, { "epoch": 1.1971764705882353, "grad_norm": 1.17595505844382, "learning_rate": 0.00029810841014383366, "loss": 0.6641, "step": 2544 }, { "epoch": 1.1976470588235295, "grad_norm": 0.27776598082261483, "learning_rate": 0.00029810146409833815, "loss": 0.123, "step": 2545 }, { "epoch": 1.1981176470588235, "grad_norm": 1.567502136163231, "learning_rate": 0.0002980945054042511, "loss": 0.6562, "step": 2546 }, { "epoch": 1.1985882352941177, "grad_norm": 0.5634015938052396, "learning_rate": 0.00029808753406216693, "loss": 0.2168, "step": 2547 }, { "epoch": 1.1990588235294117, "grad_norm": 0.132811203693956, "learning_rate": 0.000298080550072681, "loss": 0.0396, "step": 2548 }, { "epoch": 1.199529411764706, "grad_norm": 0.7448548222179526, "learning_rate": 0.0002980735534363897, "loss": 0.3184, "step": 2549 }, { "epoch": 1.2, "grad_norm": 0.8016722983790723, "learning_rate": 0.0002980665441538907, "loss": 0.2676, "step": 2550 }, { "epoch": 1.2004705882352942, "grad_norm": 0.3088438545192593, "learning_rate": 0.00029805952222578257, "loss": 0.0664, "step": 2551 }, { "epoch": 1.2009411764705882, "grad_norm": 0.5842095111228159, "learning_rate": 0.000298052487652665, "loss": 0.3008, "step": 2552 }, { "epoch": 1.2014117647058824, "grad_norm": 0.6029152808693562, "learning_rate": 0.0002980454404351388, "loss": 0.2246, "step": 2553 }, { "epoch": 1.2018823529411764, "grad_norm": 1.3978207238056897, "learning_rate": 0.0002980383805738058, "loss": 0.5469, "step": 2554 }, { "epoch": 1.2023529411764706, "grad_norm": 0.6626488556944777, "learning_rate": 0.00029803130806926897, "loss": 0.2305, "step": 2555 }, { "epoch": 1.2028235294117646, "grad_norm": 0.3805553062325879, "learning_rate": 0.0002980242229221324, "loss": 0.166, "step": 2556 }, { "epoch": 1.2032941176470588, "grad_norm": 0.7109129199629735, "learning_rate": 0.00029801712513300105, "loss": 0.4902, "step": 2557 }, { "epoch": 1.2037647058823528, "grad_norm": 0.5551150752558437, "learning_rate": 0.00029801001470248123, "loss": 0.332, "step": 2558 }, { "epoch": 1.204235294117647, "grad_norm": 0.7436696663353932, "learning_rate": 0.00029800289163118014, "loss": 0.3359, "step": 2559 }, { "epoch": 1.204705882352941, "grad_norm": 0.8398809124165089, "learning_rate": 0.00029799575591970623, "loss": 0.4941, "step": 2560 }, { "epoch": 1.2051764705882353, "grad_norm": 0.5810511773852425, "learning_rate": 0.00029798860756866883, "loss": 0.1162, "step": 2561 }, { "epoch": 1.2056470588235295, "grad_norm": 0.5742303907062304, "learning_rate": 0.00029798144657867845, "loss": 0.3438, "step": 2562 }, { "epoch": 1.2061176470588235, "grad_norm": 1.0506168069544926, "learning_rate": 0.0002979742729503467, "loss": 0.9102, "step": 2563 }, { "epoch": 1.2065882352941177, "grad_norm": 0.3611655664013743, "learning_rate": 0.0002979670866842862, "loss": 0.2559, "step": 2564 }, { "epoch": 1.2070588235294117, "grad_norm": 0.3760744313789838, "learning_rate": 0.0002979598877811107, "loss": 0.332, "step": 2565 }, { "epoch": 1.207529411764706, "grad_norm": 1.0319933163516155, "learning_rate": 0.0002979526762414351, "loss": 0.4023, "step": 2566 }, { "epoch": 1.208, "grad_norm": 1.3551309540748557, "learning_rate": 0.0002979454520658752, "loss": 0.7109, "step": 2567 }, { "epoch": 1.2084705882352942, "grad_norm": 0.48322877093653677, "learning_rate": 0.0002979382152550481, "loss": 0.2354, "step": 2568 }, { "epoch": 1.2089411764705882, "grad_norm": 0.9917139868016243, "learning_rate": 0.0002979309658095717, "loss": 0.7461, "step": 2569 }, { "epoch": 1.2094117647058824, "grad_norm": 0.5723903500913126, "learning_rate": 0.0002979237037300653, "loss": 0.4766, "step": 2570 }, { "epoch": 1.2098823529411764, "grad_norm": 0.9083645822562818, "learning_rate": 0.00029791642901714897, "loss": 0.6875, "step": 2571 }, { "epoch": 1.2103529411764706, "grad_norm": 0.419522054773501, "learning_rate": 0.0002979091416714441, "loss": 0.2217, "step": 2572 }, { "epoch": 1.2108235294117646, "grad_norm": 0.6924096126844205, "learning_rate": 0.00029790184169357307, "loss": 0.416, "step": 2573 }, { "epoch": 1.2112941176470589, "grad_norm": 1.0461917004881498, "learning_rate": 0.00029789452908415923, "loss": 0.4746, "step": 2574 }, { "epoch": 1.2117647058823529, "grad_norm": 0.6083711311499007, "learning_rate": 0.0002978872038438273, "loss": 0.4043, "step": 2575 }, { "epoch": 1.212235294117647, "grad_norm": 0.8311002404687304, "learning_rate": 0.0002978798659732027, "loss": 0.3926, "step": 2576 }, { "epoch": 1.212705882352941, "grad_norm": 0.5512836596097301, "learning_rate": 0.00029787251547291215, "loss": 0.3418, "step": 2577 }, { "epoch": 1.2131764705882353, "grad_norm": 0.44257741406243584, "learning_rate": 0.00029786515234358357, "loss": 0.2852, "step": 2578 }, { "epoch": 1.2136470588235295, "grad_norm": 0.5976677464853372, "learning_rate": 0.0002978577765858457, "loss": 0.3984, "step": 2579 }, { "epoch": 1.2141176470588235, "grad_norm": 0.3520722203894363, "learning_rate": 0.00029785038820032844, "loss": 0.1963, "step": 2580 }, { "epoch": 1.2145882352941175, "grad_norm": 0.46257660468468603, "learning_rate": 0.0002978429871876628, "loss": 0.2969, "step": 2581 }, { "epoch": 1.2150588235294117, "grad_norm": 0.3358451335490024, "learning_rate": 0.00029783557354848094, "loss": 0.1748, "step": 2582 }, { "epoch": 1.215529411764706, "grad_norm": 0.31668089886251327, "learning_rate": 0.0002978281472834159, "loss": 0.1445, "step": 2583 }, { "epoch": 1.216, "grad_norm": 0.6820487999577355, "learning_rate": 0.00029782070839310206, "loss": 0.4141, "step": 2584 }, { "epoch": 1.2164705882352942, "grad_norm": 0.4688019642397194, "learning_rate": 0.0002978132568781746, "loss": 0.3086, "step": 2585 }, { "epoch": 1.2169411764705882, "grad_norm": 0.5094125539735361, "learning_rate": 0.00029780579273927006, "loss": 0.3945, "step": 2586 }, { "epoch": 1.2174117647058824, "grad_norm": 0.7169496938550357, "learning_rate": 0.0002977983159770258, "loss": 0.457, "step": 2587 }, { "epoch": 1.2178823529411764, "grad_norm": 0.7769555500746157, "learning_rate": 0.0002977908265920804, "loss": 0.4141, "step": 2588 }, { "epoch": 1.2183529411764706, "grad_norm": 0.690973921667228, "learning_rate": 0.00029778332458507353, "loss": 0.459, "step": 2589 }, { "epoch": 1.2188235294117646, "grad_norm": 0.9359778178647983, "learning_rate": 0.0002977758099566458, "loss": 0.8008, "step": 2590 }, { "epoch": 1.2192941176470589, "grad_norm": 0.22730534417389267, "learning_rate": 0.0002977682827074391, "loss": 0.1064, "step": 2591 }, { "epoch": 1.2197647058823529, "grad_norm": 0.3357202290318851, "learning_rate": 0.0002977607428380963, "loss": 0.1475, "step": 2592 }, { "epoch": 1.220235294117647, "grad_norm": 0.7144740073375032, "learning_rate": 0.00029775319034926125, "loss": 0.5547, "step": 2593 }, { "epoch": 1.220705882352941, "grad_norm": 1.0018099694840839, "learning_rate": 0.000297745625241579, "loss": 0.5859, "step": 2594 }, { "epoch": 1.2211764705882353, "grad_norm": 0.6814796143190006, "learning_rate": 0.0002977380475156958, "loss": 0.5273, "step": 2595 }, { "epoch": 1.2216470588235295, "grad_norm": 1.1683085516299927, "learning_rate": 0.00029773045717225856, "loss": 0.6758, "step": 2596 }, { "epoch": 1.2221176470588235, "grad_norm": 1.0932724716067337, "learning_rate": 0.0002977228542119157, "loss": 0.6602, "step": 2597 }, { "epoch": 1.2225882352941175, "grad_norm": 0.11672841131637796, "learning_rate": 0.0002977152386353165, "loss": 0.0217, "step": 2598 }, { "epoch": 1.2230588235294118, "grad_norm": 0.3886576067734427, "learning_rate": 0.0002977076104431114, "loss": 0.1943, "step": 2599 }, { "epoch": 1.223529411764706, "grad_norm": 0.5582598876178959, "learning_rate": 0.00029769996963595184, "loss": 0.2285, "step": 2600 }, { "epoch": 1.224, "grad_norm": 0.3964878068227268, "learning_rate": 0.0002976923162144904, "loss": 0.3008, "step": 2601 }, { "epoch": 1.2244705882352942, "grad_norm": 1.0589545476958147, "learning_rate": 0.0002976846501793808, "loss": 0.625, "step": 2602 }, { "epoch": 1.2249411764705882, "grad_norm": 0.3931602131652867, "learning_rate": 0.00029767697153127766, "loss": 0.2051, "step": 2603 }, { "epoch": 1.2254117647058824, "grad_norm": 0.8716710064156584, "learning_rate": 0.0002976692802708368, "loss": 0.5156, "step": 2604 }, { "epoch": 1.2258823529411764, "grad_norm": 0.6140809066525067, "learning_rate": 0.000297661576398715, "loss": 0.2676, "step": 2605 }, { "epoch": 1.2263529411764706, "grad_norm": 0.4389064410289762, "learning_rate": 0.0002976538599155704, "loss": 0.2002, "step": 2606 }, { "epoch": 1.2268235294117646, "grad_norm": 0.08134237985670083, "learning_rate": 0.00029764613082206183, "loss": 0.0459, "step": 2607 }, { "epoch": 1.2272941176470589, "grad_norm": 0.09208152270507533, "learning_rate": 0.0002976383891188496, "loss": 0.0222, "step": 2608 }, { "epoch": 1.2277647058823529, "grad_norm": 0.9559002885287028, "learning_rate": 0.0002976306348065947, "loss": 0.8125, "step": 2609 }, { "epoch": 1.228235294117647, "grad_norm": 0.8873710013967337, "learning_rate": 0.00029762286788595954, "loss": 0.6328, "step": 2610 }, { "epoch": 1.228705882352941, "grad_norm": 0.3528485897406086, "learning_rate": 0.0002976150883576073, "loss": 0.0732, "step": 2611 }, { "epoch": 1.2291764705882353, "grad_norm": 0.6994956229285968, "learning_rate": 0.0002976072962222025, "loss": 0.4199, "step": 2612 }, { "epoch": 1.2296470588235293, "grad_norm": 0.3595949875746125, "learning_rate": 0.0002975994914804106, "loss": 0.2598, "step": 2613 }, { "epoch": 1.2301176470588235, "grad_norm": 0.3460523758981697, "learning_rate": 0.00029759167413289815, "loss": 0.1514, "step": 2614 }, { "epoch": 1.2305882352941175, "grad_norm": 1.017864208748841, "learning_rate": 0.00029758384418033276, "loss": 0.5547, "step": 2615 }, { "epoch": 1.2310588235294118, "grad_norm": 0.08193019788281472, "learning_rate": 0.00029757600162338326, "loss": 0.0157, "step": 2616 }, { "epoch": 1.231529411764706, "grad_norm": 0.6128854336132394, "learning_rate": 0.0002975681464627193, "loss": 0.3105, "step": 2617 }, { "epoch": 1.232, "grad_norm": 0.6646184772645541, "learning_rate": 0.00029756027869901183, "loss": 0.3672, "step": 2618 }, { "epoch": 1.2324705882352942, "grad_norm": 0.8649737395476247, "learning_rate": 0.0002975523983329328, "loss": 0.4668, "step": 2619 }, { "epoch": 1.2329411764705882, "grad_norm": 0.5951501801298859, "learning_rate": 0.00029754450536515517, "loss": 0.2656, "step": 2620 }, { "epoch": 1.2334117647058824, "grad_norm": 0.4406569599375948, "learning_rate": 0.00029753659979635315, "loss": 0.1934, "step": 2621 }, { "epoch": 1.2338823529411764, "grad_norm": 0.6049943423179337, "learning_rate": 0.00029752868162720176, "loss": 0.2109, "step": 2622 }, { "epoch": 1.2343529411764707, "grad_norm": 0.4017157409930296, "learning_rate": 0.0002975207508583774, "loss": 0.1465, "step": 2623 }, { "epoch": 1.2348235294117647, "grad_norm": 1.430369981537313, "learning_rate": 0.00029751280749055727, "loss": 0.8945, "step": 2624 }, { "epoch": 1.2352941176470589, "grad_norm": 0.2255516222000614, "learning_rate": 0.0002975048515244199, "loss": 0.0767, "step": 2625 }, { "epoch": 1.2357647058823529, "grad_norm": 0.8124046478778999, "learning_rate": 0.0002974968829606447, "loss": 0.4863, "step": 2626 }, { "epoch": 1.236235294117647, "grad_norm": 1.0251880324671803, "learning_rate": 0.0002974889017999122, "loss": 0.4434, "step": 2627 }, { "epoch": 1.236705882352941, "grad_norm": 0.8718892258549767, "learning_rate": 0.00029748090804290403, "loss": 0.4727, "step": 2628 }, { "epoch": 1.2371764705882353, "grad_norm": 0.5685253987708282, "learning_rate": 0.00029747290169030295, "loss": 0.3613, "step": 2629 }, { "epoch": 1.2376470588235293, "grad_norm": 0.6728971432190224, "learning_rate": 0.0002974648827427927, "loss": 0.4746, "step": 2630 }, { "epoch": 1.2381176470588235, "grad_norm": 0.41123316888771655, "learning_rate": 0.0002974568512010582, "loss": 0.4238, "step": 2631 }, { "epoch": 1.2385882352941175, "grad_norm": 0.2994197338178247, "learning_rate": 0.00029744880706578526, "loss": 0.1611, "step": 2632 }, { "epoch": 1.2390588235294118, "grad_norm": 0.5201927071583586, "learning_rate": 0.00029744075033766103, "loss": 0.3262, "step": 2633 }, { "epoch": 1.239529411764706, "grad_norm": 0.6017057771865146, "learning_rate": 0.0002974326810173735, "loss": 0.5586, "step": 2634 }, { "epoch": 1.24, "grad_norm": 0.1706668673950123, "learning_rate": 0.00029742459910561187, "loss": 0.0381, "step": 2635 }, { "epoch": 1.2404705882352942, "grad_norm": 0.8623603936162891, "learning_rate": 0.00029741650460306636, "loss": 0.5273, "step": 2636 }, { "epoch": 1.2409411764705882, "grad_norm": 0.22626359854875794, "learning_rate": 0.0002974083975104283, "loss": 0.0679, "step": 2637 }, { "epoch": 1.2414117647058824, "grad_norm": 0.500515044772063, "learning_rate": 0.00029740027782839003, "loss": 0.2012, "step": 2638 }, { "epoch": 1.2418823529411764, "grad_norm": 0.8608447513520148, "learning_rate": 0.00029739214555764504, "loss": 0.416, "step": 2639 }, { "epoch": 1.2423529411764707, "grad_norm": 0.3224481829780719, "learning_rate": 0.00029738400069888787, "loss": 0.126, "step": 2640 }, { "epoch": 1.2428235294117647, "grad_norm": 0.11710786795481377, "learning_rate": 0.00029737584325281415, "loss": 0.0525, "step": 2641 }, { "epoch": 1.2432941176470589, "grad_norm": 0.9873284350614592, "learning_rate": 0.0002973676732201205, "loss": 0.5117, "step": 2642 }, { "epoch": 1.2437647058823529, "grad_norm": 0.8135723206569472, "learning_rate": 0.00029735949060150476, "loss": 0.3672, "step": 2643 }, { "epoch": 1.244235294117647, "grad_norm": 0.7894413245140134, "learning_rate": 0.0002973512953976657, "loss": 0.4766, "step": 2644 }, { "epoch": 1.244705882352941, "grad_norm": 0.39738628113984514, "learning_rate": 0.0002973430876093033, "loss": 0.1865, "step": 2645 }, { "epoch": 1.2451764705882353, "grad_norm": 0.6802501891287442, "learning_rate": 0.00029733486723711845, "loss": 0.3906, "step": 2646 }, { "epoch": 1.2456470588235293, "grad_norm": 1.0488655282635098, "learning_rate": 0.0002973266342818133, "loss": 0.5117, "step": 2647 }, { "epoch": 1.2461176470588236, "grad_norm": 0.8130920857064506, "learning_rate": 0.00029731838874409093, "loss": 0.6094, "step": 2648 }, { "epoch": 1.2465882352941176, "grad_norm": 1.8225807206352815, "learning_rate": 0.0002973101306246556, "loss": 0.8281, "step": 2649 }, { "epoch": 1.2470588235294118, "grad_norm": 0.18269944661667045, "learning_rate": 0.0002973018599242125, "loss": 0.0454, "step": 2650 }, { "epoch": 1.247529411764706, "grad_norm": 1.365472474536242, "learning_rate": 0.0002972935766434681, "loss": 0.4355, "step": 2651 }, { "epoch": 1.248, "grad_norm": 1.6467570228449542, "learning_rate": 0.00029728528078312976, "loss": 0.8867, "step": 2652 }, { "epoch": 1.248470588235294, "grad_norm": 1.5771005605310593, "learning_rate": 0.00029727697234390604, "loss": 0.7617, "step": 2653 }, { "epoch": 1.2489411764705882, "grad_norm": 0.9007864880160934, "learning_rate": 0.0002972686513265065, "loss": 0.4629, "step": 2654 }, { "epoch": 1.2494117647058824, "grad_norm": 0.9605375823524221, "learning_rate": 0.00029726031773164176, "loss": 0.4199, "step": 2655 }, { "epoch": 1.2498823529411764, "grad_norm": 0.32027069958626647, "learning_rate": 0.00029725197156002357, "loss": 0.1426, "step": 2656 }, { "epoch": 1.2503529411764707, "grad_norm": 1.5640201037277366, "learning_rate": 0.00029724361281236475, "loss": 0.5781, "step": 2657 }, { "epoch": 1.2508235294117647, "grad_norm": 0.6563283752247898, "learning_rate": 0.0002972352414893792, "loss": 0.3438, "step": 2658 }, { "epoch": 1.251294117647059, "grad_norm": 1.0562174908735715, "learning_rate": 0.00029722685759178187, "loss": 0.6055, "step": 2659 }, { "epoch": 1.251764705882353, "grad_norm": 0.41828067772441063, "learning_rate": 0.0002972184611202886, "loss": 0.3984, "step": 2660 }, { "epoch": 1.2522352941176471, "grad_norm": 0.7601381249200325, "learning_rate": 0.00029721005207561683, "loss": 0.2637, "step": 2661 }, { "epoch": 1.2527058823529411, "grad_norm": 1.3031042452540227, "learning_rate": 0.00029720163045848444, "loss": 0.4746, "step": 2662 }, { "epoch": 1.2531764705882353, "grad_norm": 0.7918951301873076, "learning_rate": 0.0002971931962696109, "loss": 0.5508, "step": 2663 }, { "epoch": 1.2536470588235293, "grad_norm": 0.560132775037991, "learning_rate": 0.00029718474950971626, "loss": 0.377, "step": 2664 }, { "epoch": 1.2541176470588236, "grad_norm": 0.7062668416422698, "learning_rate": 0.0002971762901795222, "loss": 0.3848, "step": 2665 }, { "epoch": 1.2545882352941176, "grad_norm": 0.7458885067148028, "learning_rate": 0.000297167818279751, "loss": 0.5469, "step": 2666 }, { "epoch": 1.2550588235294118, "grad_norm": 0.8973235390959523, "learning_rate": 0.0002971593338111263, "loss": 0.4824, "step": 2667 }, { "epoch": 1.255529411764706, "grad_norm": 0.6181082412386886, "learning_rate": 0.0002971508367743727, "loss": 0.4082, "step": 2668 }, { "epoch": 1.256, "grad_norm": 0.5824287368841624, "learning_rate": 0.0002971423271702158, "loss": 0.2969, "step": 2669 }, { "epoch": 1.256470588235294, "grad_norm": 0.18921676749908908, "learning_rate": 0.00029713380499938246, "loss": 0.0767, "step": 2670 }, { "epoch": 1.2569411764705882, "grad_norm": 1.9649841084655533, "learning_rate": 0.00029712527026260047, "loss": 1.75, "step": 2671 }, { "epoch": 1.2574117647058825, "grad_norm": 0.3802764915611933, "learning_rate": 0.0002971167229605988, "loss": 0.2197, "step": 2672 }, { "epoch": 1.2578823529411765, "grad_norm": 0.6065043515467189, "learning_rate": 0.00029710816309410737, "loss": 0.2344, "step": 2673 }, { "epoch": 1.2583529411764705, "grad_norm": 0.49398899210576613, "learning_rate": 0.00029709959066385723, "loss": 0.1621, "step": 2674 }, { "epoch": 1.2588235294117647, "grad_norm": 0.5112113268272244, "learning_rate": 0.0002970910056705806, "loss": 0.3926, "step": 2675 }, { "epoch": 1.259294117647059, "grad_norm": 1.1185732774988597, "learning_rate": 0.0002970824081150105, "loss": 0.9531, "step": 2676 }, { "epoch": 1.259764705882353, "grad_norm": 0.378651601457219, "learning_rate": 0.00029707379799788134, "loss": 0.2812, "step": 2677 }, { "epoch": 1.2602352941176471, "grad_norm": 0.06552266798089532, "learning_rate": 0.0002970651753199285, "loss": 0.0119, "step": 2678 }, { "epoch": 1.2607058823529411, "grad_norm": 0.5280436505483518, "learning_rate": 0.0002970565400818883, "loss": 0.1475, "step": 2679 }, { "epoch": 1.2611764705882353, "grad_norm": 2.0962951553582165, "learning_rate": 0.0002970478922844983, "loss": 0.9414, "step": 2680 }, { "epoch": 1.2616470588235293, "grad_norm": 0.8469555334038726, "learning_rate": 0.00029703923192849694, "loss": 0.5234, "step": 2681 }, { "epoch": 1.2621176470588236, "grad_norm": 0.5215362282601556, "learning_rate": 0.000297030559014624, "loss": 0.2734, "step": 2682 }, { "epoch": 1.2625882352941176, "grad_norm": 0.6821303711516671, "learning_rate": 0.00029702187354362016, "loss": 0.332, "step": 2683 }, { "epoch": 1.2630588235294118, "grad_norm": 0.4625755783022477, "learning_rate": 0.0002970131755162272, "loss": 0.2168, "step": 2684 }, { "epoch": 1.263529411764706, "grad_norm": 0.9415050156366147, "learning_rate": 0.0002970044649331879, "loss": 0.4727, "step": 2685 }, { "epoch": 1.264, "grad_norm": 0.4185774909029987, "learning_rate": 0.0002969957417952463, "loss": 0.3105, "step": 2686 }, { "epoch": 1.264470588235294, "grad_norm": 2.0018469846052236, "learning_rate": 0.00029698700610314725, "loss": 0.9961, "step": 2687 }, { "epoch": 1.2649411764705882, "grad_norm": 0.48764415435577474, "learning_rate": 0.000296978257857637, "loss": 0.167, "step": 2688 }, { "epoch": 1.2654117647058825, "grad_norm": 0.8492293083318698, "learning_rate": 0.00029696949705946255, "loss": 0.5039, "step": 2689 }, { "epoch": 1.2658823529411765, "grad_norm": 0.4930862975433324, "learning_rate": 0.0002969607237093722, "loss": 0.3691, "step": 2690 }, { "epoch": 1.2663529411764705, "grad_norm": 0.7598826360824054, "learning_rate": 0.00029695193780811523, "loss": 0.8867, "step": 2691 }, { "epoch": 1.2668235294117647, "grad_norm": 0.9361769141671799, "learning_rate": 0.0002969431393564419, "loss": 0.4277, "step": 2692 }, { "epoch": 1.267294117647059, "grad_norm": 1.1690888355320852, "learning_rate": 0.0002969343283551038, "loss": 0.3965, "step": 2693 }, { "epoch": 1.267764705882353, "grad_norm": 2.246016828635721, "learning_rate": 0.0002969255048048533, "loss": 0.9727, "step": 2694 }, { "epoch": 1.2682352941176471, "grad_norm": 0.5107443414148816, "learning_rate": 0.00029691666870644405, "loss": 0.3008, "step": 2695 }, { "epoch": 1.2687058823529411, "grad_norm": 0.7815372649651733, "learning_rate": 0.0002969078200606307, "loss": 0.3086, "step": 2696 }, { "epoch": 1.2691764705882354, "grad_norm": 1.3107294410032766, "learning_rate": 0.0002968989588681688, "loss": 0.4648, "step": 2697 }, { "epoch": 1.2696470588235294, "grad_norm": 0.18410243512449234, "learning_rate": 0.0002968900851298154, "loss": 0.062, "step": 2698 }, { "epoch": 1.2701176470588236, "grad_norm": 0.4179805382814441, "learning_rate": 0.00029688119884632827, "loss": 0.1836, "step": 2699 }, { "epoch": 1.2705882352941176, "grad_norm": 1.0286299942711516, "learning_rate": 0.0002968723000184662, "loss": 0.4609, "step": 2700 }, { "epoch": 1.2710588235294118, "grad_norm": 1.2356935193284484, "learning_rate": 0.0002968633886469894, "loss": 0.5742, "step": 2701 }, { "epoch": 1.271529411764706, "grad_norm": 0.8538526841393902, "learning_rate": 0.00029685446473265876, "loss": 0.543, "step": 2702 }, { "epoch": 1.272, "grad_norm": 1.1919660526336888, "learning_rate": 0.00029684552827623657, "loss": 0.8203, "step": 2703 }, { "epoch": 1.272470588235294, "grad_norm": 0.2435549542753988, "learning_rate": 0.000296836579278486, "loss": 0.1104, "step": 2704 }, { "epoch": 1.2729411764705882, "grad_norm": 0.5100363909349487, "learning_rate": 0.0002968276177401713, "loss": 0.1738, "step": 2705 }, { "epoch": 1.2734117647058825, "grad_norm": 0.6310453040817836, "learning_rate": 0.00029681864366205786, "loss": 0.249, "step": 2706 }, { "epoch": 1.2738823529411765, "grad_norm": 0.5142092137787913, "learning_rate": 0.0002968096570449121, "loss": 0.1611, "step": 2707 }, { "epoch": 1.2743529411764705, "grad_norm": 1.1466852757722998, "learning_rate": 0.0002968006578895016, "loss": 0.3672, "step": 2708 }, { "epoch": 1.2748235294117647, "grad_norm": 0.23018072347954316, "learning_rate": 0.0002967916461965948, "loss": 0.0459, "step": 2709 }, { "epoch": 1.275294117647059, "grad_norm": 1.1286883045375715, "learning_rate": 0.00029678262196696143, "loss": 0.4707, "step": 2710 }, { "epoch": 1.275764705882353, "grad_norm": 0.3520216152652662, "learning_rate": 0.0002967735852013722, "loss": 0.1475, "step": 2711 }, { "epoch": 1.2762352941176471, "grad_norm": 0.448711911336247, "learning_rate": 0.0002967645359005988, "loss": 0.2559, "step": 2712 }, { "epoch": 1.2767058823529411, "grad_norm": 0.5612095846485782, "learning_rate": 0.0002967554740654142, "loss": 0.2451, "step": 2713 }, { "epoch": 1.2771764705882354, "grad_norm": 1.208920184978549, "learning_rate": 0.0002967463996965923, "loss": 0.832, "step": 2714 }, { "epoch": 1.2776470588235294, "grad_norm": 0.711321497818644, "learning_rate": 0.0002967373127949081, "loss": 0.5508, "step": 2715 }, { "epoch": 1.2781176470588236, "grad_norm": 0.27773130420794123, "learning_rate": 0.0002967282133611375, "loss": 0.127, "step": 2716 }, { "epoch": 1.2785882352941176, "grad_norm": 0.43509929186008695, "learning_rate": 0.0002967191013960579, "loss": 0.1426, "step": 2717 }, { "epoch": 1.2790588235294118, "grad_norm": 1.7104987701559629, "learning_rate": 0.00029670997690044745, "loss": 0.8789, "step": 2718 }, { "epoch": 1.2795294117647058, "grad_norm": 0.9533531612193377, "learning_rate": 0.0002967008398750853, "loss": 0.4434, "step": 2719 }, { "epoch": 1.28, "grad_norm": 0.7814714024117533, "learning_rate": 0.00029669169032075186, "loss": 0.4023, "step": 2720 }, { "epoch": 1.280470588235294, "grad_norm": 0.7327336450921323, "learning_rate": 0.0002966825282382285, "loss": 0.4863, "step": 2721 }, { "epoch": 1.2809411764705882, "grad_norm": 0.951929542817249, "learning_rate": 0.0002966733536282978, "loss": 0.3262, "step": 2722 }, { "epoch": 1.2814117647058825, "grad_norm": 0.09401306680497497, "learning_rate": 0.0002966641664917433, "loss": 0.0119, "step": 2723 }, { "epoch": 1.2818823529411765, "grad_norm": 1.2876231270793927, "learning_rate": 0.0002966549668293495, "loss": 0.4492, "step": 2724 }, { "epoch": 1.2823529411764705, "grad_norm": 0.9956625810447196, "learning_rate": 0.0002966457546419023, "loss": 0.5117, "step": 2725 }, { "epoch": 1.2828235294117647, "grad_norm": 0.6263579278218235, "learning_rate": 0.0002966365299301883, "loss": 0.5, "step": 2726 }, { "epoch": 1.283294117647059, "grad_norm": 0.583106106117435, "learning_rate": 0.00029662729269499546, "loss": 0.2695, "step": 2727 }, { "epoch": 1.283764705882353, "grad_norm": 0.6021033588802701, "learning_rate": 0.00029661804293711254, "loss": 0.2539, "step": 2728 }, { "epoch": 1.2842352941176471, "grad_norm": 1.0751897986046595, "learning_rate": 0.00029660878065732965, "loss": 0.332, "step": 2729 }, { "epoch": 1.2847058823529411, "grad_norm": 1.0872287111115755, "learning_rate": 0.0002965995058564377, "loss": 0.4258, "step": 2730 }, { "epoch": 1.2851764705882354, "grad_norm": 1.3421472787359723, "learning_rate": 0.000296590218535229, "loss": 0.7109, "step": 2731 }, { "epoch": 1.2856470588235294, "grad_norm": 0.32115508916283686, "learning_rate": 0.00029658091869449657, "loss": 0.1152, "step": 2732 }, { "epoch": 1.2861176470588236, "grad_norm": 0.46164476109327357, "learning_rate": 0.0002965716063350347, "loss": 0.1631, "step": 2733 }, { "epoch": 1.2865882352941176, "grad_norm": 0.37950542455506436, "learning_rate": 0.0002965622814576387, "loss": 0.0732, "step": 2734 }, { "epoch": 1.2870588235294118, "grad_norm": 0.6255035675804157, "learning_rate": 0.000296552944063105, "loss": 0.1621, "step": 2735 }, { "epoch": 1.2875294117647058, "grad_norm": 1.435716439135893, "learning_rate": 0.0002965435941522311, "loss": 0.6055, "step": 2736 }, { "epoch": 1.288, "grad_norm": 0.3971176304892575, "learning_rate": 0.00029653423172581534, "loss": 0.1982, "step": 2737 }, { "epoch": 1.288470588235294, "grad_norm": 1.0127555154445238, "learning_rate": 0.0002965248567846575, "loss": 0.5938, "step": 2738 }, { "epoch": 1.2889411764705883, "grad_norm": 0.6379649657008292, "learning_rate": 0.0002965154693295582, "loss": 0.377, "step": 2739 }, { "epoch": 1.2894117647058825, "grad_norm": 0.8155048180894552, "learning_rate": 0.00029650606936131913, "loss": 0.5742, "step": 2740 }, { "epoch": 1.2898823529411765, "grad_norm": 0.8371042668902209, "learning_rate": 0.0002964966568807432, "loss": 0.416, "step": 2741 }, { "epoch": 1.2903529411764705, "grad_norm": 1.519413393663649, "learning_rate": 0.00029648723188863417, "loss": 0.8555, "step": 2742 }, { "epoch": 1.2908235294117647, "grad_norm": 0.27201381884343817, "learning_rate": 0.000296477794385797, "loss": 0.0845, "step": 2743 }, { "epoch": 1.291294117647059, "grad_norm": 0.3992719130815345, "learning_rate": 0.00029646834437303777, "loss": 0.2061, "step": 2744 }, { "epoch": 1.291764705882353, "grad_norm": 0.36300699841909556, "learning_rate": 0.00029645888185116353, "loss": 0.0835, "step": 2745 }, { "epoch": 1.292235294117647, "grad_norm": 1.297399662302259, "learning_rate": 0.00029644940682098234, "loss": 0.625, "step": 2746 }, { "epoch": 1.2927058823529411, "grad_norm": 0.5238122033816949, "learning_rate": 0.00029643991928330346, "loss": 0.1167, "step": 2747 }, { "epoch": 1.2931764705882354, "grad_norm": 0.4892073960545997, "learning_rate": 0.0002964304192389372, "loss": 0.2891, "step": 2748 }, { "epoch": 1.2936470588235294, "grad_norm": 0.27228493476681975, "learning_rate": 0.000296420906688695, "loss": 0.1074, "step": 2749 }, { "epoch": 1.2941176470588236, "grad_norm": 0.6238805975003487, "learning_rate": 0.00029641138163338907, "loss": 0.2383, "step": 2750 }, { "epoch": 1.2945882352941176, "grad_norm": 1.5389871554186705, "learning_rate": 0.00029640184407383303, "loss": 0.5938, "step": 2751 }, { "epoch": 1.2950588235294118, "grad_norm": 0.8802663050089031, "learning_rate": 0.00029639229401084137, "loss": 0.6289, "step": 2752 }, { "epoch": 1.2955294117647058, "grad_norm": 0.23917847475941356, "learning_rate": 0.00029638273144522983, "loss": 0.0469, "step": 2753 }, { "epoch": 1.296, "grad_norm": 1.4281458667698625, "learning_rate": 0.00029637315637781494, "loss": 0.6289, "step": 2754 }, { "epoch": 1.296470588235294, "grad_norm": 0.7259625619338094, "learning_rate": 0.0002963635688094146, "loss": 0.416, "step": 2755 }, { "epoch": 1.2969411764705883, "grad_norm": 0.36956324764177445, "learning_rate": 0.0002963539687408476, "loss": 0.1553, "step": 2756 }, { "epoch": 1.2974117647058825, "grad_norm": 0.9472359461662344, "learning_rate": 0.00029634435617293374, "loss": 0.4492, "step": 2757 }, { "epoch": 1.2978823529411765, "grad_norm": 0.4976165917675633, "learning_rate": 0.00029633473110649407, "loss": 0.2598, "step": 2758 }, { "epoch": 1.2983529411764705, "grad_norm": 0.7509439839415394, "learning_rate": 0.00029632509354235063, "loss": 0.3945, "step": 2759 }, { "epoch": 1.2988235294117647, "grad_norm": 0.8918635463158838, "learning_rate": 0.00029631544348132643, "loss": 0.5391, "step": 2760 }, { "epoch": 1.299294117647059, "grad_norm": 0.5125930971720142, "learning_rate": 0.0002963057809242457, "loss": 0.332, "step": 2761 }, { "epoch": 1.299764705882353, "grad_norm": 0.2940954870605505, "learning_rate": 0.00029629610587193365, "loss": 0.1621, "step": 2762 }, { "epoch": 1.300235294117647, "grad_norm": 0.9574580843265589, "learning_rate": 0.00029628641832521663, "loss": 0.543, "step": 2763 }, { "epoch": 1.3007058823529412, "grad_norm": 1.3255013703728018, "learning_rate": 0.0002962767182849219, "loss": 0.6758, "step": 2764 }, { "epoch": 1.3011764705882354, "grad_norm": 0.711876507111717, "learning_rate": 0.00029626700575187794, "loss": 0.5117, "step": 2765 }, { "epoch": 1.3016470588235294, "grad_norm": 0.6096330238029697, "learning_rate": 0.0002962572807269143, "loss": 0.5547, "step": 2766 }, { "epoch": 1.3021176470588236, "grad_norm": 0.5932548732597558, "learning_rate": 0.00029624754321086146, "loss": 0.2852, "step": 2767 }, { "epoch": 1.3025882352941176, "grad_norm": 1.0166619831352324, "learning_rate": 0.00029623779320455106, "loss": 0.4531, "step": 2768 }, { "epoch": 1.3030588235294118, "grad_norm": 1.2196030655543952, "learning_rate": 0.00029622803070881586, "loss": 0.4883, "step": 2769 }, { "epoch": 1.3035294117647058, "grad_norm": 0.5904448310143794, "learning_rate": 0.00029621825572448965, "loss": 0.3633, "step": 2770 }, { "epoch": 1.304, "grad_norm": 0.6100005518867367, "learning_rate": 0.0002962084682524071, "loss": 0.6172, "step": 2771 }, { "epoch": 1.304470588235294, "grad_norm": 1.0627463585762937, "learning_rate": 0.0002961986682934043, "loss": 0.5156, "step": 2772 }, { "epoch": 1.3049411764705883, "grad_norm": 0.6566467808814681, "learning_rate": 0.0002961888558483181, "loss": 0.2559, "step": 2773 }, { "epoch": 1.3054117647058823, "grad_norm": 1.1731226649793878, "learning_rate": 0.0002961790309179866, "loss": 0.6797, "step": 2774 }, { "epoch": 1.3058823529411765, "grad_norm": 0.09889751453102923, "learning_rate": 0.00029616919350324873, "loss": 0.0227, "step": 2775 }, { "epoch": 1.3063529411764705, "grad_norm": 0.7111518302910435, "learning_rate": 0.0002961593436049449, "loss": 0.5039, "step": 2776 }, { "epoch": 1.3068235294117647, "grad_norm": 0.4468106814150315, "learning_rate": 0.0002961494812239162, "loss": 0.1797, "step": 2777 }, { "epoch": 1.307294117647059, "grad_norm": 0.1855575389195614, "learning_rate": 0.00029613960636100487, "loss": 0.0659, "step": 2778 }, { "epoch": 1.307764705882353, "grad_norm": 0.441693225462945, "learning_rate": 0.0002961297190170544, "loss": 0.1338, "step": 2779 }, { "epoch": 1.308235294117647, "grad_norm": 0.8174783870278101, "learning_rate": 0.00029611981919290916, "loss": 0.3613, "step": 2780 }, { "epoch": 1.3087058823529412, "grad_norm": 2.010384630813015, "learning_rate": 0.0002961099068894146, "loss": 0.8594, "step": 2781 }, { "epoch": 1.3091764705882354, "grad_norm": 0.63988364822848, "learning_rate": 0.00029609998210741744, "loss": 0.4785, "step": 2782 }, { "epoch": 1.3096470588235294, "grad_norm": 0.5142942884697997, "learning_rate": 0.0002960900448477651, "loss": 0.1895, "step": 2783 }, { "epoch": 1.3101176470588236, "grad_norm": 2.905332786920547, "learning_rate": 0.0002960800951113064, "loss": 0.9297, "step": 2784 }, { "epoch": 1.3105882352941176, "grad_norm": 1.7598588826914252, "learning_rate": 0.00029607013289889106, "loss": 0.8594, "step": 2785 }, { "epoch": 1.3110588235294118, "grad_norm": 0.7501014308716617, "learning_rate": 0.0002960601582113699, "loss": 0.3398, "step": 2786 }, { "epoch": 1.3115294117647058, "grad_norm": 1.4986293596418048, "learning_rate": 0.00029605017104959475, "loss": 0.7656, "step": 2787 }, { "epoch": 1.312, "grad_norm": 0.6931918529678324, "learning_rate": 0.00029604017141441865, "loss": 0.1777, "step": 2788 }, { "epoch": 1.312470588235294, "grad_norm": 0.4710942987546686, "learning_rate": 0.0002960301593066956, "loss": 0.1309, "step": 2789 }, { "epoch": 1.3129411764705883, "grad_norm": 1.701846009411546, "learning_rate": 0.0002960201347272807, "loss": 0.4062, "step": 2790 }, { "epoch": 1.3134117647058823, "grad_norm": 1.2810894489149, "learning_rate": 0.00029601009767703007, "loss": 0.4238, "step": 2791 }, { "epoch": 1.3138823529411765, "grad_norm": 0.8329007404204406, "learning_rate": 0.0002960000481568009, "loss": 0.4824, "step": 2792 }, { "epoch": 1.3143529411764705, "grad_norm": 0.3706452123992428, "learning_rate": 0.0002959899861674515, "loss": 0.3184, "step": 2793 }, { "epoch": 1.3148235294117647, "grad_norm": 0.49153270539620064, "learning_rate": 0.0002959799117098412, "loss": 0.1855, "step": 2794 }, { "epoch": 1.315294117647059, "grad_norm": 0.9562578703571376, "learning_rate": 0.00029596982478483037, "loss": 0.3457, "step": 2795 }, { "epoch": 1.315764705882353, "grad_norm": 2.2211048627061465, "learning_rate": 0.00029595972539328053, "loss": 0.7109, "step": 2796 }, { "epoch": 1.316235294117647, "grad_norm": 2.0478275860932924, "learning_rate": 0.00029594961353605426, "loss": 0.6562, "step": 2797 }, { "epoch": 1.3167058823529412, "grad_norm": 0.8461753703262666, "learning_rate": 0.0002959394892140151, "loss": 0.4336, "step": 2798 }, { "epoch": 1.3171764705882354, "grad_norm": 0.9756567941827132, "learning_rate": 0.00029592935242802774, "loss": 0.5859, "step": 2799 }, { "epoch": 1.3176470588235294, "grad_norm": 0.2991748520456577, "learning_rate": 0.0002959192031789579, "loss": 0.1182, "step": 2800 }, { "epoch": 1.3181176470588234, "grad_norm": 1.951752760605327, "learning_rate": 0.00029590904146767236, "loss": 0.5352, "step": 2801 }, { "epoch": 1.3185882352941176, "grad_norm": 0.7604925137815194, "learning_rate": 0.000295898867295039, "loss": 0.1582, "step": 2802 }, { "epoch": 1.3190588235294118, "grad_norm": 1.700266333547303, "learning_rate": 0.0002958886806619268, "loss": 0.6172, "step": 2803 }, { "epoch": 1.3195294117647058, "grad_norm": 2.6275764391363383, "learning_rate": 0.00029587848156920564, "loss": 0.8281, "step": 2804 }, { "epoch": 1.32, "grad_norm": 0.9274964408192514, "learning_rate": 0.0002958682700177466, "loss": 0.5586, "step": 2805 }, { "epoch": 1.320470588235294, "grad_norm": 0.6828581303599337, "learning_rate": 0.00029585804600842187, "loss": 0.4629, "step": 2806 }, { "epoch": 1.3209411764705883, "grad_norm": 0.7098704745430147, "learning_rate": 0.00029584780954210454, "loss": 0.334, "step": 2807 }, { "epoch": 1.3214117647058823, "grad_norm": 0.7543659302892135, "learning_rate": 0.0002958375606196689, "loss": 0.7656, "step": 2808 }, { "epoch": 1.3218823529411765, "grad_norm": 2.0515765021491346, "learning_rate": 0.0002958272992419903, "loss": 0.4844, "step": 2809 }, { "epoch": 1.3223529411764705, "grad_norm": 2.2211924148304822, "learning_rate": 0.000295817025409945, "loss": 0.4961, "step": 2810 }, { "epoch": 1.3228235294117647, "grad_norm": 1.6059895255585361, "learning_rate": 0.00029580673912441057, "loss": 0.334, "step": 2811 }, { "epoch": 1.323294117647059, "grad_norm": 0.7555464962185174, "learning_rate": 0.00029579644038626546, "loss": 0.2246, "step": 2812 }, { "epoch": 1.323764705882353, "grad_norm": 0.5482802170813846, "learning_rate": 0.00029578612919638913, "loss": 0.3926, "step": 2813 }, { "epoch": 1.324235294117647, "grad_norm": 0.9192661973497169, "learning_rate": 0.0002957758055556623, "loss": 0.3848, "step": 2814 }, { "epoch": 1.3247058823529412, "grad_norm": 0.7895255957474493, "learning_rate": 0.0002957654694649667, "loss": 0.4922, "step": 2815 }, { "epoch": 1.3251764705882354, "grad_norm": 0.8408924919304067, "learning_rate": 0.000295755120925185, "loss": 0.208, "step": 2816 }, { "epoch": 1.3256470588235294, "grad_norm": 1.8856350967845763, "learning_rate": 0.0002957447599372011, "loss": 0.7617, "step": 2817 }, { "epoch": 1.3261176470588234, "grad_norm": 0.6151155707164783, "learning_rate": 0.00029573438650189974, "loss": 0.3359, "step": 2818 }, { "epoch": 1.3265882352941176, "grad_norm": 2.4030100515217474, "learning_rate": 0.000295724000620167, "loss": 1.0156, "step": 2819 }, { "epoch": 1.3270588235294118, "grad_norm": 0.9629509133619298, "learning_rate": 0.00029571360229288984, "loss": 0.5859, "step": 2820 }, { "epoch": 1.3275294117647058, "grad_norm": 0.40915040863061647, "learning_rate": 0.0002957031915209563, "loss": 0.2207, "step": 2821 }, { "epoch": 1.328, "grad_norm": 1.7789452673460082, "learning_rate": 0.00029569276830525553, "loss": 0.498, "step": 2822 }, { "epoch": 1.328470588235294, "grad_norm": 0.2383774378164193, "learning_rate": 0.00029568233264667775, "loss": 0.0549, "step": 2823 }, { "epoch": 1.3289411764705883, "grad_norm": 0.7685544693992786, "learning_rate": 0.00029567188454611415, "loss": 0.2295, "step": 2824 }, { "epoch": 1.3294117647058823, "grad_norm": 0.35300782403465697, "learning_rate": 0.00029566142400445716, "loss": 0.1118, "step": 2825 }, { "epoch": 1.3298823529411765, "grad_norm": 1.5420484822480267, "learning_rate": 0.0002956509510226001, "loss": 0.4629, "step": 2826 }, { "epoch": 1.3303529411764705, "grad_norm": 0.9224442109256791, "learning_rate": 0.00029564046560143735, "loss": 0.498, "step": 2827 }, { "epoch": 1.3308235294117647, "grad_norm": 1.0417281060382173, "learning_rate": 0.0002956299677418645, "loss": 0.4668, "step": 2828 }, { "epoch": 1.331294117647059, "grad_norm": 0.49265005561058833, "learning_rate": 0.000295619457444778, "loss": 0.2139, "step": 2829 }, { "epoch": 1.331764705882353, "grad_norm": 0.8827324645239742, "learning_rate": 0.00029560893471107573, "loss": 0.4805, "step": 2830 }, { "epoch": 1.332235294117647, "grad_norm": 0.4411014635675758, "learning_rate": 0.00029559839954165616, "loss": 0.3125, "step": 2831 }, { "epoch": 1.3327058823529412, "grad_norm": 0.4596978523154006, "learning_rate": 0.00029558785193741914, "loss": 0.2061, "step": 2832 }, { "epoch": 1.3331764705882354, "grad_norm": 0.28163095250142983, "learning_rate": 0.0002955772918992654, "loss": 0.0869, "step": 2833 }, { "epoch": 1.3336470588235294, "grad_norm": 0.8224462834283784, "learning_rate": 0.00029556671942809694, "loss": 0.2812, "step": 2834 }, { "epoch": 1.3341176470588234, "grad_norm": 0.1471367633478047, "learning_rate": 0.00029555613452481657, "loss": 0.0303, "step": 2835 }, { "epoch": 1.3345882352941176, "grad_norm": 2.677695277571631, "learning_rate": 0.00029554553719032844, "loss": 0.8203, "step": 2836 }, { "epoch": 1.3350588235294119, "grad_norm": 0.3129745929925292, "learning_rate": 0.00029553492742553754, "loss": 0.1768, "step": 2837 }, { "epoch": 1.3355294117647059, "grad_norm": 0.36619559299998333, "learning_rate": 0.00029552430523134993, "loss": 0.2002, "step": 2838 }, { "epoch": 1.336, "grad_norm": 0.4286828934211613, "learning_rate": 0.0002955136706086729, "loss": 0.2715, "step": 2839 }, { "epoch": 1.336470588235294, "grad_norm": 0.8362651906327431, "learning_rate": 0.0002955030235584146, "loss": 0.5, "step": 2840 }, { "epoch": 1.3369411764705883, "grad_norm": 0.3155503799125853, "learning_rate": 0.00029549236408148453, "loss": 0.1826, "step": 2841 }, { "epoch": 1.3374117647058823, "grad_norm": 1.2130356247840777, "learning_rate": 0.00029548169217879285, "loss": 0.6094, "step": 2842 }, { "epoch": 1.3378823529411765, "grad_norm": 7.605167598055386, "learning_rate": 0.000295471007851251, "loss": 0.3887, "step": 2843 }, { "epoch": 1.3383529411764705, "grad_norm": 0.573978771966205, "learning_rate": 0.00029546031109977164, "loss": 0.2656, "step": 2844 }, { "epoch": 1.3388235294117647, "grad_norm": 1.2747379266148593, "learning_rate": 0.00029544960192526824, "loss": 0.4844, "step": 2845 }, { "epoch": 1.3392941176470587, "grad_norm": 0.8243944362188917, "learning_rate": 0.0002954388803286554, "loss": 0.4746, "step": 2846 }, { "epoch": 1.339764705882353, "grad_norm": 0.8206121002147876, "learning_rate": 0.0002954281463108488, "loss": 0.4727, "step": 2847 }, { "epoch": 1.340235294117647, "grad_norm": 2.706901754690993, "learning_rate": 0.0002954173998727652, "loss": 0.8633, "step": 2848 }, { "epoch": 1.3407058823529412, "grad_norm": 0.45982759783749155, "learning_rate": 0.00029540664101532227, "loss": 0.165, "step": 2849 }, { "epoch": 1.3411764705882354, "grad_norm": 0.5720305081571759, "learning_rate": 0.0002953958697394391, "loss": 0.3125, "step": 2850 }, { "epoch": 1.3416470588235294, "grad_norm": 0.5952492606783187, "learning_rate": 0.0002953850860460355, "loss": 0.3594, "step": 2851 }, { "epoch": 1.3421176470588234, "grad_norm": 0.2972262583332857, "learning_rate": 0.00029537428993603237, "loss": 0.1787, "step": 2852 }, { "epoch": 1.3425882352941176, "grad_norm": 0.9677112927364723, "learning_rate": 0.00029536348141035183, "loss": 0.3672, "step": 2853 }, { "epoch": 1.3430588235294119, "grad_norm": 0.9785813136466791, "learning_rate": 0.000295352660469917, "loss": 0.6562, "step": 2854 }, { "epoch": 1.3435294117647059, "grad_norm": 0.9504036427128452, "learning_rate": 0.00029534182711565203, "loss": 0.4883, "step": 2855 }, { "epoch": 1.3439999999999999, "grad_norm": 0.7123636105764138, "learning_rate": 0.0002953309813484821, "loss": 0.6133, "step": 2856 }, { "epoch": 1.344470588235294, "grad_norm": 0.7425267258773184, "learning_rate": 0.0002953201231693336, "loss": 0.4609, "step": 2857 }, { "epoch": 1.3449411764705883, "grad_norm": 1.2840457512576893, "learning_rate": 0.00029530925257913366, "loss": 0.7461, "step": 2858 }, { "epoch": 1.3454117647058823, "grad_norm": 0.6399383809796425, "learning_rate": 0.0002952983695788109, "loss": 0.4355, "step": 2859 }, { "epoch": 1.3458823529411765, "grad_norm": 0.8026315527345066, "learning_rate": 0.00029528747416929463, "loss": 0.4336, "step": 2860 }, { "epoch": 1.3463529411764705, "grad_norm": 0.6031020777388744, "learning_rate": 0.0002952765663515155, "loss": 0.375, "step": 2861 }, { "epoch": 1.3468235294117648, "grad_norm": 0.4705380666883062, "learning_rate": 0.000295265646126405, "loss": 0.167, "step": 2862 }, { "epoch": 1.3472941176470588, "grad_norm": 0.159317698066715, "learning_rate": 0.00029525471349489576, "loss": 0.0527, "step": 2863 }, { "epoch": 1.347764705882353, "grad_norm": 1.4299756131459356, "learning_rate": 0.00029524376845792153, "loss": 0.4824, "step": 2864 }, { "epoch": 1.348235294117647, "grad_norm": 0.1253323965928411, "learning_rate": 0.0002952328110164171, "loss": 0.0479, "step": 2865 }, { "epoch": 1.3487058823529412, "grad_norm": 0.5940540707802817, "learning_rate": 0.0002952218411713182, "loss": 0.3086, "step": 2866 }, { "epoch": 1.3491764705882354, "grad_norm": 0.8669482686096048, "learning_rate": 0.0002952108589235618, "loss": 0.4082, "step": 2867 }, { "epoch": 1.3496470588235294, "grad_norm": 0.5169538938860797, "learning_rate": 0.0002951998642740858, "loss": 0.3438, "step": 2868 }, { "epoch": 1.3501176470588234, "grad_norm": 0.9136581469253631, "learning_rate": 0.00029518885722382913, "loss": 0.4453, "step": 2869 }, { "epoch": 1.3505882352941176, "grad_norm": 0.8572106305580024, "learning_rate": 0.00029517783777373194, "loss": 0.3301, "step": 2870 }, { "epoch": 1.3510588235294119, "grad_norm": 0.6593260809997873, "learning_rate": 0.0002951668059247353, "loss": 0.5, "step": 2871 }, { "epoch": 1.3515294117647059, "grad_norm": 0.3527876718956774, "learning_rate": 0.0002951557616777814, "loss": 0.1904, "step": 2872 }, { "epoch": 1.3519999999999999, "grad_norm": 0.18687028670108388, "learning_rate": 0.0002951447050338135, "loss": 0.0684, "step": 2873 }, { "epoch": 1.352470588235294, "grad_norm": 1.2829359192225092, "learning_rate": 0.00029513363599377587, "loss": 0.7344, "step": 2874 }, { "epoch": 1.3529411764705883, "grad_norm": 0.7226019390900488, "learning_rate": 0.00029512255455861375, "loss": 0.5703, "step": 2875 }, { "epoch": 1.3534117647058823, "grad_norm": 0.7780952414872212, "learning_rate": 0.0002951114607292737, "loss": 0.5195, "step": 2876 }, { "epoch": 1.3538823529411765, "grad_norm": 0.29873186484680436, "learning_rate": 0.0002951003545067032, "loss": 0.127, "step": 2877 }, { "epoch": 1.3543529411764705, "grad_norm": 1.1733125864121223, "learning_rate": 0.0002950892358918506, "loss": 0.707, "step": 2878 }, { "epoch": 1.3548235294117648, "grad_norm": 0.6899763647470281, "learning_rate": 0.00029507810488566563, "loss": 0.4902, "step": 2879 }, { "epoch": 1.3552941176470588, "grad_norm": 1.1557364816963718, "learning_rate": 0.000295066961489099, "loss": 0.5742, "step": 2880 }, { "epoch": 1.355764705882353, "grad_norm": 0.3776398260691145, "learning_rate": 0.0002950558057031022, "loss": 0.2188, "step": 2881 }, { "epoch": 1.356235294117647, "grad_norm": 1.0178020002030095, "learning_rate": 0.0002950446375286281, "loss": 0.4004, "step": 2882 }, { "epoch": 1.3567058823529412, "grad_norm": 0.8742010785026686, "learning_rate": 0.00029503345696663054, "loss": 0.6289, "step": 2883 }, { "epoch": 1.3571764705882354, "grad_norm": 0.6658297791414713, "learning_rate": 0.0002950222640180644, "loss": 0.457, "step": 2884 }, { "epoch": 1.3576470588235294, "grad_norm": 0.6400074855420779, "learning_rate": 0.0002950110586838855, "loss": 0.5117, "step": 2885 }, { "epoch": 1.3581176470588234, "grad_norm": 0.7901590888797486, "learning_rate": 0.00029499984096505107, "loss": 0.6875, "step": 2886 }, { "epoch": 1.3585882352941177, "grad_norm": 0.5672317334148411, "learning_rate": 0.00029498861086251887, "loss": 0.3652, "step": 2887 }, { "epoch": 1.3590588235294119, "grad_norm": 0.8144954787409819, "learning_rate": 0.00029497736837724817, "loss": 0.4551, "step": 2888 }, { "epoch": 1.3595294117647059, "grad_norm": 0.9997350503086946, "learning_rate": 0.0002949661135101991, "loss": 0.6914, "step": 2889 }, { "epoch": 1.3599999999999999, "grad_norm": 0.5123790923998042, "learning_rate": 0.0002949548462623329, "loss": 0.4668, "step": 2890 }, { "epoch": 1.360470588235294, "grad_norm": 0.7028758717293614, "learning_rate": 0.0002949435666346118, "loss": 0.4824, "step": 2891 }, { "epoch": 1.3609411764705883, "grad_norm": 0.6658146017313548, "learning_rate": 0.00029493227462799914, "loss": 0.4043, "step": 2892 }, { "epoch": 1.3614117647058823, "grad_norm": 0.5592927718223956, "learning_rate": 0.00029492097024345936, "loss": 0.1699, "step": 2893 }, { "epoch": 1.3618823529411765, "grad_norm": 0.30227438983208493, "learning_rate": 0.00029490965348195794, "loss": 0.1211, "step": 2894 }, { "epoch": 1.3623529411764705, "grad_norm": 0.9201558746870834, "learning_rate": 0.00029489832434446126, "loss": 0.543, "step": 2895 }, { "epoch": 1.3628235294117648, "grad_norm": 0.7271020402718875, "learning_rate": 0.000294886982831937, "loss": 0.5039, "step": 2896 }, { "epoch": 1.3632941176470588, "grad_norm": 0.07269339967904642, "learning_rate": 0.00029487562894535375, "loss": 0.0117, "step": 2897 }, { "epoch": 1.363764705882353, "grad_norm": 1.0622123205863032, "learning_rate": 0.0002948642626856812, "loss": 0.5156, "step": 2898 }, { "epoch": 1.364235294117647, "grad_norm": 0.05747626688379106, "learning_rate": 0.00029485288405389, "loss": 0.0127, "step": 2899 }, { "epoch": 1.3647058823529412, "grad_norm": 0.2434525107633764, "learning_rate": 0.000294841493050952, "loss": 0.0957, "step": 2900 }, { "epoch": 1.3651764705882352, "grad_norm": 0.38634327065360274, "learning_rate": 0.00029483008967784007, "loss": 0.1592, "step": 2901 }, { "epoch": 1.3656470588235294, "grad_norm": 0.8284016963172915, "learning_rate": 0.0002948186739355281, "loss": 0.3672, "step": 2902 }, { "epoch": 1.3661176470588234, "grad_norm": 0.5323849054254881, "learning_rate": 0.00029480724582499107, "loss": 0.3301, "step": 2903 }, { "epoch": 1.3665882352941177, "grad_norm": 1.1665921794551295, "learning_rate": 0.0002947958053472049, "loss": 0.6367, "step": 2904 }, { "epoch": 1.3670588235294119, "grad_norm": 0.3660602939361605, "learning_rate": 0.0002947843525031468, "loss": 0.1787, "step": 2905 }, { "epoch": 1.3675294117647059, "grad_norm": 0.25182986785262035, "learning_rate": 0.0002947728872937948, "loss": 0.0564, "step": 2906 }, { "epoch": 1.3679999999999999, "grad_norm": 0.6068242793704626, "learning_rate": 0.00029476140972012815, "loss": 0.2715, "step": 2907 }, { "epoch": 1.368470588235294, "grad_norm": 0.3832636206936227, "learning_rate": 0.000294749919783127, "loss": 0.1875, "step": 2908 }, { "epoch": 1.3689411764705883, "grad_norm": 0.6581506787351928, "learning_rate": 0.00029473841748377277, "loss": 0.5664, "step": 2909 }, { "epoch": 1.3694117647058823, "grad_norm": 0.5446011878018309, "learning_rate": 0.0002947269028230477, "loss": 0.3594, "step": 2910 }, { "epoch": 1.3698823529411766, "grad_norm": 0.6096779898904713, "learning_rate": 0.0002947153758019353, "loss": 0.3965, "step": 2911 }, { "epoch": 1.3703529411764706, "grad_norm": 0.6119051253774141, "learning_rate": 0.00029470383642142, "loss": 0.2559, "step": 2912 }, { "epoch": 1.3708235294117648, "grad_norm": 0.7171456047841113, "learning_rate": 0.0002946922846824872, "loss": 0.5625, "step": 2913 }, { "epoch": 1.3712941176470588, "grad_norm": 0.7059320547842534, "learning_rate": 0.00029468072058612367, "loss": 0.6562, "step": 2914 }, { "epoch": 1.371764705882353, "grad_norm": 0.6214868684824649, "learning_rate": 0.0002946691441333169, "loss": 0.3633, "step": 2915 }, { "epoch": 1.372235294117647, "grad_norm": 0.6938195982247195, "learning_rate": 0.0002946575553250556, "loss": 0.4922, "step": 2916 }, { "epoch": 1.3727058823529412, "grad_norm": 0.39136561053116503, "learning_rate": 0.00029464595416232956, "loss": 0.1289, "step": 2917 }, { "epoch": 1.3731764705882352, "grad_norm": 0.7405968699016022, "learning_rate": 0.0002946343406461296, "loss": 0.4102, "step": 2918 }, { "epoch": 1.3736470588235294, "grad_norm": 0.5919062854200144, "learning_rate": 0.0002946227147774475, "loss": 0.373, "step": 2919 }, { "epoch": 1.3741176470588234, "grad_norm": 0.8920995014864062, "learning_rate": 0.0002946110765572761, "loss": 0.4746, "step": 2920 }, { "epoch": 1.3745882352941177, "grad_norm": 0.9928724874648257, "learning_rate": 0.00029459942598660953, "loss": 0.707, "step": 2921 }, { "epoch": 1.3750588235294119, "grad_norm": 0.4992157601352473, "learning_rate": 0.00029458776306644273, "loss": 0.3984, "step": 2922 }, { "epoch": 1.375529411764706, "grad_norm": 0.7097305984292055, "learning_rate": 0.0002945760877977717, "loss": 0.459, "step": 2923 }, { "epoch": 1.376, "grad_norm": 0.7600874053253832, "learning_rate": 0.0002945644001815937, "loss": 0.293, "step": 2924 }, { "epoch": 1.3764705882352941, "grad_norm": 0.6533996918385554, "learning_rate": 0.0002945527002189068, "loss": 0.5195, "step": 2925 }, { "epoch": 1.3769411764705883, "grad_norm": 1.2339288980651588, "learning_rate": 0.0002945409879107103, "loss": 1.1641, "step": 2926 }, { "epoch": 1.3774117647058823, "grad_norm": 0.30143773741565294, "learning_rate": 0.00029452926325800445, "loss": 0.1533, "step": 2927 }, { "epoch": 1.3778823529411763, "grad_norm": 0.5918579720013963, "learning_rate": 0.0002945175262617906, "loss": 0.1846, "step": 2928 }, { "epoch": 1.3783529411764706, "grad_norm": 0.7132919072588351, "learning_rate": 0.00029450577692307115, "loss": 0.1895, "step": 2929 }, { "epoch": 1.3788235294117648, "grad_norm": 0.5564843835019143, "learning_rate": 0.0002944940152428496, "loss": 0.1099, "step": 2930 }, { "epoch": 1.3792941176470588, "grad_norm": 0.8002513769223962, "learning_rate": 0.0002944822412221303, "loss": 0.2754, "step": 2931 }, { "epoch": 1.379764705882353, "grad_norm": 0.8161945842822336, "learning_rate": 0.00029447045486191904, "loss": 0.3301, "step": 2932 }, { "epoch": 1.380235294117647, "grad_norm": 0.6332930372254167, "learning_rate": 0.0002944586561632223, "loss": 0.334, "step": 2933 }, { "epoch": 1.3807058823529412, "grad_norm": 0.6793561958935377, "learning_rate": 0.0002944468451270477, "loss": 0.3906, "step": 2934 }, { "epoch": 1.3811764705882352, "grad_norm": 0.37745534752927956, "learning_rate": 0.000294435021754404, "loss": 0.1611, "step": 2935 }, { "epoch": 1.3816470588235295, "grad_norm": 1.899214786911377, "learning_rate": 0.000294423186046301, "loss": 0.5508, "step": 2936 }, { "epoch": 1.3821176470588235, "grad_norm": 1.0939405704059748, "learning_rate": 0.00029441133800374957, "loss": 0.3906, "step": 2937 }, { "epoch": 1.3825882352941177, "grad_norm": 0.6121068808883147, "learning_rate": 0.00029439947762776154, "loss": 0.2695, "step": 2938 }, { "epoch": 1.383058823529412, "grad_norm": 0.4105174072866703, "learning_rate": 0.00029438760491934987, "loss": 0.2012, "step": 2939 }, { "epoch": 1.383529411764706, "grad_norm": 1.718855484631404, "learning_rate": 0.0002943757198795284, "loss": 0.5703, "step": 2940 }, { "epoch": 1.384, "grad_norm": 0.23631915032966874, "learning_rate": 0.0002943638225093124, "loss": 0.0498, "step": 2941 }, { "epoch": 1.3844705882352941, "grad_norm": 0.8633164489407618, "learning_rate": 0.00029435191280971783, "loss": 0.3828, "step": 2942 }, { "epoch": 1.3849411764705883, "grad_norm": 0.31450757539624324, "learning_rate": 0.0002943399907817619, "loss": 0.1099, "step": 2943 }, { "epoch": 1.3854117647058823, "grad_norm": 0.45163626282907143, "learning_rate": 0.0002943280564264627, "loss": 0.2041, "step": 2944 }, { "epoch": 1.3858823529411763, "grad_norm": 0.44736212476220405, "learning_rate": 0.0002943161097448396, "loss": 0.2188, "step": 2945 }, { "epoch": 1.3863529411764706, "grad_norm": 0.8282842940581313, "learning_rate": 0.0002943041507379128, "loss": 0.4316, "step": 2946 }, { "epoch": 1.3868235294117648, "grad_norm": 0.6759173430190617, "learning_rate": 0.00029429217940670383, "loss": 0.3145, "step": 2947 }, { "epoch": 1.3872941176470588, "grad_norm": 0.47997726691333265, "learning_rate": 0.0002942801957522349, "loss": 0.2734, "step": 2948 }, { "epoch": 1.387764705882353, "grad_norm": 1.0333901914475037, "learning_rate": 0.00029426819977552965, "loss": 0.4863, "step": 2949 }, { "epoch": 1.388235294117647, "grad_norm": 0.2741815464962453, "learning_rate": 0.0002942561914776124, "loss": 0.0903, "step": 2950 }, { "epoch": 1.3887058823529412, "grad_norm": 0.6036486501633591, "learning_rate": 0.0002942441708595089, "loss": 0.3516, "step": 2951 }, { "epoch": 1.3891764705882352, "grad_norm": 0.7875991616626988, "learning_rate": 0.0002942321379222457, "loss": 0.3965, "step": 2952 }, { "epoch": 1.3896470588235295, "grad_norm": 1.2485906460188043, "learning_rate": 0.0002942200926668504, "loss": 0.6875, "step": 2953 }, { "epoch": 1.3901176470588235, "grad_norm": 0.5161872848533656, "learning_rate": 0.00029420803509435186, "loss": 0.2236, "step": 2954 }, { "epoch": 1.3905882352941177, "grad_norm": 0.4771376105119258, "learning_rate": 0.00029419596520577975, "loss": 0.1865, "step": 2955 }, { "epoch": 1.391058823529412, "grad_norm": 0.3295490977363582, "learning_rate": 0.00029418388300216493, "loss": 0.1084, "step": 2956 }, { "epoch": 1.391529411764706, "grad_norm": 1.3381630647671052, "learning_rate": 0.0002941717884845393, "loss": 0.4629, "step": 2957 }, { "epoch": 1.392, "grad_norm": 0.4475850778477822, "learning_rate": 0.00029415968165393577, "loss": 0.2227, "step": 2958 }, { "epoch": 1.3924705882352941, "grad_norm": 0.39383580401573953, "learning_rate": 0.00029414756251138835, "loss": 0.1201, "step": 2959 }, { "epoch": 1.3929411764705883, "grad_norm": 0.5587680398529477, "learning_rate": 0.000294135431057932, "loss": 0.2354, "step": 2960 }, { "epoch": 1.3934117647058823, "grad_norm": 0.8989620010075672, "learning_rate": 0.0002941232872946029, "loss": 0.6445, "step": 2961 }, { "epoch": 1.3938823529411764, "grad_norm": 0.9528237280017864, "learning_rate": 0.00029411113122243814, "loss": 0.5664, "step": 2962 }, { "epoch": 1.3943529411764706, "grad_norm": 0.41539767948692796, "learning_rate": 0.0002940989628424759, "loss": 0.2314, "step": 2963 }, { "epoch": 1.3948235294117648, "grad_norm": 0.7033097608060835, "learning_rate": 0.0002940867821557554, "loss": 0.4805, "step": 2964 }, { "epoch": 1.3952941176470588, "grad_norm": 1.1957291752427621, "learning_rate": 0.00029407458916331705, "loss": 0.6289, "step": 2965 }, { "epoch": 1.395764705882353, "grad_norm": 0.8016428557425166, "learning_rate": 0.00029406238386620205, "loss": 0.5, "step": 2966 }, { "epoch": 1.396235294117647, "grad_norm": 0.20886227098042212, "learning_rate": 0.0002940501662654528, "loss": 0.0579, "step": 2967 }, { "epoch": 1.3967058823529412, "grad_norm": 1.3851116470325875, "learning_rate": 0.0002940379363621129, "loss": 0.3613, "step": 2968 }, { "epoch": 1.3971764705882352, "grad_norm": 1.1838162163681256, "learning_rate": 0.00029402569415722666, "loss": 0.498, "step": 2969 }, { "epoch": 1.3976470588235295, "grad_norm": 0.5638196191000622, "learning_rate": 0.00029401343965183965, "loss": 0.2891, "step": 2970 }, { "epoch": 1.3981176470588235, "grad_norm": 1.244579209924009, "learning_rate": 0.0002940011728469986, "loss": 0.3926, "step": 2971 }, { "epoch": 1.3985882352941177, "grad_norm": 1.7563445888077576, "learning_rate": 0.000293988893743751, "loss": 0.6094, "step": 2972 }, { "epoch": 1.3990588235294117, "grad_norm": 1.3066461019156361, "learning_rate": 0.0002939766023431457, "loss": 0.5312, "step": 2973 }, { "epoch": 1.399529411764706, "grad_norm": 0.7668502039329009, "learning_rate": 0.0002939642986462323, "loss": 0.3027, "step": 2974 }, { "epoch": 1.4, "grad_norm": 1.779888919766425, "learning_rate": 0.0002939519826540616, "loss": 0.5078, "step": 2975 }, { "epoch": 1.4004705882352941, "grad_norm": 1.0051988196251789, "learning_rate": 0.0002939396543676856, "loss": 0.3027, "step": 2976 }, { "epoch": 1.4009411764705884, "grad_norm": 0.908488989653713, "learning_rate": 0.00029392731378815706, "loss": 0.4902, "step": 2977 }, { "epoch": 1.4014117647058824, "grad_norm": 0.4655666028998395, "learning_rate": 0.00029391496091652993, "loss": 0.1387, "step": 2978 }, { "epoch": 1.4018823529411764, "grad_norm": 0.44984030341193654, "learning_rate": 0.0002939025957538593, "loss": 0.1543, "step": 2979 }, { "epoch": 1.4023529411764706, "grad_norm": 1.1328870881707198, "learning_rate": 0.0002938902183012011, "loss": 0.3457, "step": 2980 }, { "epoch": 1.4028235294117648, "grad_norm": 0.5068414965504481, "learning_rate": 0.00029387782855961253, "loss": 0.2314, "step": 2981 }, { "epoch": 1.4032941176470588, "grad_norm": 0.6929801603707006, "learning_rate": 0.00029386542653015167, "loss": 0.5039, "step": 2982 }, { "epoch": 1.4037647058823528, "grad_norm": 0.6569941936683029, "learning_rate": 0.0002938530122138777, "loss": 0.3105, "step": 2983 }, { "epoch": 1.404235294117647, "grad_norm": 0.8862226325153073, "learning_rate": 0.0002938405856118509, "loss": 0.4395, "step": 2984 }, { "epoch": 1.4047058823529412, "grad_norm": 0.7245994493678537, "learning_rate": 0.0002938281467251326, "loss": 0.3379, "step": 2985 }, { "epoch": 1.4051764705882352, "grad_norm": 0.29484011636736135, "learning_rate": 0.00029381569555478505, "loss": 0.1562, "step": 2986 }, { "epoch": 1.4056470588235295, "grad_norm": 0.7339427596859406, "learning_rate": 0.00029380323210187166, "loss": 0.3418, "step": 2987 }, { "epoch": 1.4061176470588235, "grad_norm": 0.08682626197769823, "learning_rate": 0.00029379075636745694, "loss": 0.0184, "step": 2988 }, { "epoch": 1.4065882352941177, "grad_norm": 0.7018463354286016, "learning_rate": 0.0002937782683526063, "loss": 0.3926, "step": 2989 }, { "epoch": 1.4070588235294117, "grad_norm": 0.4952243511393426, "learning_rate": 0.0002937657680583864, "loss": 0.1699, "step": 2990 }, { "epoch": 1.407529411764706, "grad_norm": 0.6155630320586918, "learning_rate": 0.0002937532554858647, "loss": 0.332, "step": 2991 }, { "epoch": 1.408, "grad_norm": 0.7234036935693355, "learning_rate": 0.0002937407306361099, "loss": 0.4492, "step": 2992 }, { "epoch": 1.4084705882352941, "grad_norm": 0.9782306476745619, "learning_rate": 0.0002937281935101916, "loss": 0.5664, "step": 2993 }, { "epoch": 1.4089411764705884, "grad_norm": 0.7186996540904889, "learning_rate": 0.0002937156441091806, "loss": 0.5117, "step": 2994 }, { "epoch": 1.4094117647058824, "grad_norm": 0.3817247414802484, "learning_rate": 0.0002937030824341487, "loss": 0.25, "step": 2995 }, { "epoch": 1.4098823529411764, "grad_norm": 0.6235663590793403, "learning_rate": 0.00029369050848616863, "loss": 0.2969, "step": 2996 }, { "epoch": 1.4103529411764706, "grad_norm": 1.0066386522975064, "learning_rate": 0.0002936779222663144, "loss": 0.4199, "step": 2997 }, { "epoch": 1.4108235294117648, "grad_norm": 0.2673758719157291, "learning_rate": 0.0002936653237756609, "loss": 0.0879, "step": 2998 }, { "epoch": 1.4112941176470588, "grad_norm": 1.0324052829736958, "learning_rate": 0.00029365271301528406, "loss": 0.5977, "step": 2999 }, { "epoch": 1.4117647058823528, "grad_norm": 0.5555661452070343, "learning_rate": 0.00029364008998626086, "loss": 0.4961, "step": 3000 }, { "epoch": 1.412235294117647, "grad_norm": 0.9996786650396697, "learning_rate": 0.00029362745468966946, "loss": 0.416, "step": 3001 }, { "epoch": 1.4127058823529413, "grad_norm": 0.12080654872334692, "learning_rate": 0.0002936148071265889, "loss": 0.0322, "step": 3002 }, { "epoch": 1.4131764705882353, "grad_norm": 0.9473514230854797, "learning_rate": 0.00029360214729809946, "loss": 0.4492, "step": 3003 }, { "epoch": 1.4136470588235295, "grad_norm": 0.7458435352097934, "learning_rate": 0.00029358947520528227, "loss": 0.4531, "step": 3004 }, { "epoch": 1.4141176470588235, "grad_norm": 0.9063122200157733, "learning_rate": 0.0002935767908492196, "loss": 0.5469, "step": 3005 }, { "epoch": 1.4145882352941177, "grad_norm": 0.5660199243467389, "learning_rate": 0.0002935640942309947, "loss": 0.5781, "step": 3006 }, { "epoch": 1.4150588235294117, "grad_norm": 0.36285422845750415, "learning_rate": 0.000293551385351692, "loss": 0.1816, "step": 3007 }, { "epoch": 1.415529411764706, "grad_norm": 0.5500125037685311, "learning_rate": 0.0002935386642123968, "loss": 0.2773, "step": 3008 }, { "epoch": 1.416, "grad_norm": 0.4764091742332767, "learning_rate": 0.00029352593081419573, "loss": 0.1709, "step": 3009 }, { "epoch": 1.4164705882352941, "grad_norm": 1.1007293893960972, "learning_rate": 0.0002935131851581762, "loss": 0.6484, "step": 3010 }, { "epoch": 1.4169411764705884, "grad_norm": 1.5970130079415463, "learning_rate": 0.00029350042724542657, "loss": 0.7383, "step": 3011 }, { "epoch": 1.4174117647058824, "grad_norm": 0.7703646382695025, "learning_rate": 0.0002934876570770367, "loss": 0.625, "step": 3012 }, { "epoch": 1.4178823529411764, "grad_norm": 0.22529810899909467, "learning_rate": 0.00029347487465409714, "loss": 0.0381, "step": 3013 }, { "epoch": 1.4183529411764706, "grad_norm": 0.2729039233636633, "learning_rate": 0.0002934620799776995, "loss": 0.0767, "step": 3014 }, { "epoch": 1.4188235294117648, "grad_norm": 1.4531435793307697, "learning_rate": 0.00029344927304893653, "loss": 0.793, "step": 3015 }, { "epoch": 1.4192941176470588, "grad_norm": 1.1077855602934397, "learning_rate": 0.00029343645386890204, "loss": 0.4902, "step": 3016 }, { "epoch": 1.4197647058823528, "grad_norm": 1.0318481716169552, "learning_rate": 0.00029342362243869093, "loss": 0.4648, "step": 3017 }, { "epoch": 1.420235294117647, "grad_norm": 0.724577474014082, "learning_rate": 0.00029341077875939884, "loss": 0.2793, "step": 3018 }, { "epoch": 1.4207058823529413, "grad_norm": 0.6855949893612144, "learning_rate": 0.0002933979228321229, "loss": 0.2949, "step": 3019 }, { "epoch": 1.4211764705882353, "grad_norm": 0.4988905091236628, "learning_rate": 0.00029338505465796095, "loss": 0.1396, "step": 3020 }, { "epoch": 1.4216470588235295, "grad_norm": 2.2728277307287192, "learning_rate": 0.00029337217423801206, "loss": 0.4707, "step": 3021 }, { "epoch": 1.4221176470588235, "grad_norm": 0.7677734340052971, "learning_rate": 0.00029335928157337625, "loss": 0.2539, "step": 3022 }, { "epoch": 1.4225882352941177, "grad_norm": 0.9491195733487032, "learning_rate": 0.00029334637666515457, "loss": 0.5469, "step": 3023 }, { "epoch": 1.4230588235294117, "grad_norm": 0.7184363956565454, "learning_rate": 0.00029333345951444923, "loss": 0.3262, "step": 3024 }, { "epoch": 1.423529411764706, "grad_norm": 0.8079245314374938, "learning_rate": 0.00029332053012236345, "loss": 0.2246, "step": 3025 }, { "epoch": 1.424, "grad_norm": 0.7585473798804008, "learning_rate": 0.0002933075884900014, "loss": 0.2559, "step": 3026 }, { "epoch": 1.4244705882352942, "grad_norm": 0.5725340513509566, "learning_rate": 0.00029329463461846833, "loss": 0.1738, "step": 3027 }, { "epoch": 1.4249411764705882, "grad_norm": 1.1946568900776504, "learning_rate": 0.00029328166850887066, "loss": 0.543, "step": 3028 }, { "epoch": 1.4254117647058824, "grad_norm": 0.544496849592075, "learning_rate": 0.00029326869016231566, "loss": 0.2617, "step": 3029 }, { "epoch": 1.4258823529411764, "grad_norm": 1.5549013368096671, "learning_rate": 0.0002932556995799118, "loss": 0.7461, "step": 3030 }, { "epoch": 1.4263529411764706, "grad_norm": 0.8056162368184797, "learning_rate": 0.00029324269676276856, "loss": 0.5312, "step": 3031 }, { "epoch": 1.4268235294117648, "grad_norm": 0.9157060026545634, "learning_rate": 0.0002932296817119964, "loss": 0.4277, "step": 3032 }, { "epoch": 1.4272941176470588, "grad_norm": 1.6925309192750986, "learning_rate": 0.0002932166544287069, "loss": 0.6641, "step": 3033 }, { "epoch": 1.4277647058823528, "grad_norm": 1.3063275824128306, "learning_rate": 0.00029320361491401263, "loss": 0.4043, "step": 3034 }, { "epoch": 1.428235294117647, "grad_norm": 0.35108460449580386, "learning_rate": 0.00029319056316902723, "loss": 0.167, "step": 3035 }, { "epoch": 1.4287058823529413, "grad_norm": 0.18968684191046525, "learning_rate": 0.0002931774991948654, "loss": 0.0349, "step": 3036 }, { "epoch": 1.4291764705882353, "grad_norm": 0.3575179918857989, "learning_rate": 0.0002931644229926428, "loss": 0.1157, "step": 3037 }, { "epoch": 1.4296470588235295, "grad_norm": 1.500588929000258, "learning_rate": 0.00029315133456347633, "loss": 0.5352, "step": 3038 }, { "epoch": 1.4301176470588235, "grad_norm": 1.6821331214370783, "learning_rate": 0.0002931382339084837, "loss": 0.7891, "step": 3039 }, { "epoch": 1.4305882352941177, "grad_norm": 0.6009006223371941, "learning_rate": 0.0002931251210287838, "loss": 0.3164, "step": 3040 }, { "epoch": 1.4310588235294117, "grad_norm": 0.3513978605940606, "learning_rate": 0.00029311199592549656, "loss": 0.2236, "step": 3041 }, { "epoch": 1.431529411764706, "grad_norm": 0.38139351979790886, "learning_rate": 0.00029309885859974294, "loss": 0.1377, "step": 3042 }, { "epoch": 1.432, "grad_norm": 1.608989116483698, "learning_rate": 0.00029308570905264484, "loss": 0.8672, "step": 3043 }, { "epoch": 1.4324705882352942, "grad_norm": 0.928433671618733, "learning_rate": 0.00029307254728532536, "loss": 0.8008, "step": 3044 }, { "epoch": 1.4329411764705882, "grad_norm": 0.2372437104153451, "learning_rate": 0.00029305937329890863, "loss": 0.0952, "step": 3045 }, { "epoch": 1.4334117647058824, "grad_norm": 1.0478307439350167, "learning_rate": 0.00029304618709451966, "loss": 0.8672, "step": 3046 }, { "epoch": 1.4338823529411764, "grad_norm": 0.4194585630170711, "learning_rate": 0.0002930329886732847, "loss": 0.2002, "step": 3047 }, { "epoch": 1.4343529411764706, "grad_norm": 0.4717108532368173, "learning_rate": 0.00029301977803633094, "loss": 0.1875, "step": 3048 }, { "epoch": 1.4348235294117648, "grad_norm": 1.3198641356748904, "learning_rate": 0.0002930065551847866, "loss": 0.4766, "step": 3049 }, { "epoch": 1.4352941176470588, "grad_norm": 0.5621481203960071, "learning_rate": 0.00029299332011978107, "loss": 0.21, "step": 3050 }, { "epoch": 1.4357647058823528, "grad_norm": 0.299661844751439, "learning_rate": 0.00029298007284244456, "loss": 0.1885, "step": 3051 }, { "epoch": 1.436235294117647, "grad_norm": 0.5460558088997721, "learning_rate": 0.0002929668133539085, "loss": 0.4277, "step": 3052 }, { "epoch": 1.4367058823529413, "grad_norm": 0.8785610485773877, "learning_rate": 0.0002929535416553054, "loss": 0.2461, "step": 3053 }, { "epoch": 1.4371764705882353, "grad_norm": 0.8553229345957273, "learning_rate": 0.0002929402577477686, "loss": 0.3828, "step": 3054 }, { "epoch": 1.4376470588235293, "grad_norm": 0.7757526052898732, "learning_rate": 0.0002929269616324327, "loss": 0.377, "step": 3055 }, { "epoch": 1.4381176470588235, "grad_norm": 0.4247839327442265, "learning_rate": 0.0002929136533104332, "loss": 0.2559, "step": 3056 }, { "epoch": 1.4385882352941177, "grad_norm": 0.5462414622071737, "learning_rate": 0.0002929003327829067, "loss": 0.3379, "step": 3057 }, { "epoch": 1.4390588235294117, "grad_norm": 0.3444692031341709, "learning_rate": 0.0002928870000509909, "loss": 0.1514, "step": 3058 }, { "epoch": 1.439529411764706, "grad_norm": 0.8989466794049074, "learning_rate": 0.0002928736551158244, "loss": 0.5156, "step": 3059 }, { "epoch": 1.44, "grad_norm": 0.2930561418972236, "learning_rate": 0.00029286029797854704, "loss": 0.0952, "step": 3060 }, { "epoch": 1.4404705882352942, "grad_norm": 0.7854424945848034, "learning_rate": 0.00029284692864029946, "loss": 0.7539, "step": 3061 }, { "epoch": 1.4409411764705882, "grad_norm": 0.1609377877817217, "learning_rate": 0.0002928335471022235, "loss": 0.041, "step": 3062 }, { "epoch": 1.4414117647058824, "grad_norm": 1.7028790496280084, "learning_rate": 0.00029282015336546205, "loss": 0.8125, "step": 3063 }, { "epoch": 1.4418823529411764, "grad_norm": 1.201238445741116, "learning_rate": 0.00029280674743115896, "loss": 0.7422, "step": 3064 }, { "epoch": 1.4423529411764706, "grad_norm": 0.9236525031790386, "learning_rate": 0.0002927933293004592, "loss": 0.5352, "step": 3065 }, { "epoch": 1.4428235294117648, "grad_norm": 1.4492315230179769, "learning_rate": 0.00029277989897450867, "loss": 0.7617, "step": 3066 }, { "epoch": 1.4432941176470588, "grad_norm": 0.7177740530970874, "learning_rate": 0.0002927664564544545, "loss": 0.4141, "step": 3067 }, { "epoch": 1.4437647058823528, "grad_norm": 0.8088563577542524, "learning_rate": 0.00029275300174144464, "loss": 0.4844, "step": 3068 }, { "epoch": 1.444235294117647, "grad_norm": 0.6073966220606444, "learning_rate": 0.0002927395348366282, "loss": 0.4336, "step": 3069 }, { "epoch": 1.4447058823529413, "grad_norm": 0.6508009856862789, "learning_rate": 0.00029272605574115544, "loss": 0.3984, "step": 3070 }, { "epoch": 1.4451764705882353, "grad_norm": 0.45124768204495713, "learning_rate": 0.00029271256445617737, "loss": 0.1875, "step": 3071 }, { "epoch": 1.4456470588235293, "grad_norm": 0.5219027661432298, "learning_rate": 0.0002926990609828463, "loss": 0.1562, "step": 3072 }, { "epoch": 1.4461176470588235, "grad_norm": 0.7180406042862771, "learning_rate": 0.00029268554532231556, "loss": 0.3047, "step": 3073 }, { "epoch": 1.4465882352941177, "grad_norm": 0.4997008879912999, "learning_rate": 0.00029267201747573934, "loss": 0.166, "step": 3074 }, { "epoch": 1.4470588235294117, "grad_norm": 0.2907939481413796, "learning_rate": 0.00029265847744427303, "loss": 0.1387, "step": 3075 }, { "epoch": 1.447529411764706, "grad_norm": 0.9452734151450879, "learning_rate": 0.000292644925229073, "loss": 0.4199, "step": 3076 }, { "epoch": 1.448, "grad_norm": 0.8061991238530709, "learning_rate": 0.00029263136083129666, "loss": 0.3633, "step": 3077 }, { "epoch": 1.4484705882352942, "grad_norm": 0.7701190497572092, "learning_rate": 0.00029261778425210256, "loss": 0.4395, "step": 3078 }, { "epoch": 1.4489411764705882, "grad_norm": 0.841472410294005, "learning_rate": 0.0002926041954926501, "loss": 0.4922, "step": 3079 }, { "epoch": 1.4494117647058824, "grad_norm": 0.5062728661990734, "learning_rate": 0.00029259059455409987, "loss": 0.4316, "step": 3080 }, { "epoch": 1.4498823529411764, "grad_norm": 0.6949676297237439, "learning_rate": 0.00029257698143761354, "loss": 0.3438, "step": 3081 }, { "epoch": 1.4503529411764706, "grad_norm": 0.5107267605296302, "learning_rate": 0.00029256335614435356, "loss": 0.2344, "step": 3082 }, { "epoch": 1.4508235294117646, "grad_norm": 0.7886603127478929, "learning_rate": 0.00029254971867548373, "loss": 0.4141, "step": 3083 }, { "epoch": 1.4512941176470588, "grad_norm": 0.4562564432304691, "learning_rate": 0.00029253606903216874, "loss": 0.1855, "step": 3084 }, { "epoch": 1.4517647058823528, "grad_norm": 0.8494967900672381, "learning_rate": 0.00029252240721557433, "loss": 0.5195, "step": 3085 }, { "epoch": 1.452235294117647, "grad_norm": 0.3685613205866255, "learning_rate": 0.00029250873322686727, "loss": 0.1631, "step": 3086 }, { "epoch": 1.4527058823529413, "grad_norm": 0.1950326022708509, "learning_rate": 0.00029249504706721534, "loss": 0.0674, "step": 3087 }, { "epoch": 1.4531764705882353, "grad_norm": 0.6095681040927862, "learning_rate": 0.00029248134873778754, "loss": 0.2637, "step": 3088 }, { "epoch": 1.4536470588235293, "grad_norm": 0.28580839230971744, "learning_rate": 0.0002924676382397536, "loss": 0.1074, "step": 3089 }, { "epoch": 1.4541176470588235, "grad_norm": 0.7968027202797223, "learning_rate": 0.0002924539155742846, "loss": 0.4648, "step": 3090 }, { "epoch": 1.4545882352941177, "grad_norm": 0.6159570613328179, "learning_rate": 0.0002924401807425525, "loss": 0.2852, "step": 3091 }, { "epoch": 1.4550588235294117, "grad_norm": 0.3951855590670128, "learning_rate": 0.0002924264337457303, "loss": 0.1064, "step": 3092 }, { "epoch": 1.455529411764706, "grad_norm": 0.552988204151373, "learning_rate": 0.000292412674584992, "loss": 0.4102, "step": 3093 }, { "epoch": 1.456, "grad_norm": 0.6195942558922274, "learning_rate": 0.0002923989032615128, "loss": 0.4805, "step": 3094 }, { "epoch": 1.4564705882352942, "grad_norm": 0.4488556889918597, "learning_rate": 0.0002923851197764688, "loss": 0.2344, "step": 3095 }, { "epoch": 1.4569411764705882, "grad_norm": 0.8155998203938263, "learning_rate": 0.00029237132413103715, "loss": 0.6602, "step": 3096 }, { "epoch": 1.4574117647058824, "grad_norm": 0.5772717650011892, "learning_rate": 0.00029235751632639617, "loss": 0.3438, "step": 3097 }, { "epoch": 1.4578823529411764, "grad_norm": 0.7037903727701418, "learning_rate": 0.00029234369636372496, "loss": 0.5352, "step": 3098 }, { "epoch": 1.4583529411764706, "grad_norm": 0.6710154596725083, "learning_rate": 0.0002923298642442039, "loss": 0.3711, "step": 3099 }, { "epoch": 1.4588235294117646, "grad_norm": 0.7464997608836487, "learning_rate": 0.00029231601996901433, "loss": 0.4043, "step": 3100 }, { "epoch": 1.4592941176470589, "grad_norm": 0.42157537819815455, "learning_rate": 0.0002923021635393386, "loss": 0.2949, "step": 3101 }, { "epoch": 1.4597647058823529, "grad_norm": 0.4008121120178334, "learning_rate": 0.0002922882949563601, "loss": 0.2061, "step": 3102 }, { "epoch": 1.460235294117647, "grad_norm": 0.21673792231096545, "learning_rate": 0.0002922744142212633, "loss": 0.0723, "step": 3103 }, { "epoch": 1.4607058823529413, "grad_norm": 0.8725156020838979, "learning_rate": 0.00029226052133523363, "loss": 0.3711, "step": 3104 }, { "epoch": 1.4611764705882353, "grad_norm": 0.7237622825942133, "learning_rate": 0.0002922466162994577, "loss": 0.4473, "step": 3105 }, { "epoch": 1.4616470588235293, "grad_norm": 0.8505589214821299, "learning_rate": 0.000292232699115123, "loss": 0.375, "step": 3106 }, { "epoch": 1.4621176470588235, "grad_norm": 0.4751437181584391, "learning_rate": 0.0002922187697834182, "loss": 0.1621, "step": 3107 }, { "epoch": 1.4625882352941177, "grad_norm": 0.6987834731409167, "learning_rate": 0.00029220482830553284, "loss": 0.3457, "step": 3108 }, { "epoch": 1.4630588235294117, "grad_norm": 0.2969046079098138, "learning_rate": 0.0002921908746826576, "loss": 0.165, "step": 3109 }, { "epoch": 1.4635294117647057, "grad_norm": 0.40321648978858415, "learning_rate": 0.0002921769089159843, "loss": 0.208, "step": 3110 }, { "epoch": 1.464, "grad_norm": 1.2891461688851416, "learning_rate": 0.00029216293100670554, "loss": 0.5625, "step": 3111 }, { "epoch": 1.4644705882352942, "grad_norm": 0.48482644774252226, "learning_rate": 0.0002921489409560152, "loss": 0.2715, "step": 3112 }, { "epoch": 1.4649411764705882, "grad_norm": 0.40900535022701534, "learning_rate": 0.00029213493876510807, "loss": 0.2461, "step": 3113 }, { "epoch": 1.4654117647058824, "grad_norm": 0.33112914574235014, "learning_rate": 0.0002921209244351799, "loss": 0.1924, "step": 3114 }, { "epoch": 1.4658823529411764, "grad_norm": 0.5061448067692157, "learning_rate": 0.00029210689796742783, "loss": 0.2734, "step": 3115 }, { "epoch": 1.4663529411764706, "grad_norm": 0.8070816904254846, "learning_rate": 0.00029209285936304956, "loss": 0.4375, "step": 3116 }, { "epoch": 1.4668235294117646, "grad_norm": 0.7302937067436797, "learning_rate": 0.0002920788086232442, "loss": 0.3594, "step": 3117 }, { "epoch": 1.4672941176470589, "grad_norm": 0.7779040054065939, "learning_rate": 0.00029206474574921165, "loss": 0.5508, "step": 3118 }, { "epoch": 1.4677647058823529, "grad_norm": 0.3351912009022298, "learning_rate": 0.0002920506707421529, "loss": 0.0864, "step": 3119 }, { "epoch": 1.468235294117647, "grad_norm": 0.5599547984663217, "learning_rate": 0.0002920365836032703, "loss": 0.2275, "step": 3120 }, { "epoch": 1.4687058823529413, "grad_norm": 0.5841147395095765, "learning_rate": 0.00029202248433376663, "loss": 0.2383, "step": 3121 }, { "epoch": 1.4691764705882353, "grad_norm": 0.5688754906327314, "learning_rate": 0.00029200837293484627, "loss": 0.2471, "step": 3122 }, { "epoch": 1.4696470588235293, "grad_norm": 1.1776709650792483, "learning_rate": 0.0002919942494077142, "loss": 0.582, "step": 3123 }, { "epoch": 1.4701176470588235, "grad_norm": 0.4956054210653614, "learning_rate": 0.0002919801137535769, "loss": 0.2402, "step": 3124 }, { "epoch": 1.4705882352941178, "grad_norm": 0.3059344198312254, "learning_rate": 0.0002919659659736414, "loss": 0.1226, "step": 3125 }, { "epoch": 1.4710588235294118, "grad_norm": 0.5188847319708784, "learning_rate": 0.00029195180606911606, "loss": 0.1875, "step": 3126 }, { "epoch": 1.4715294117647058, "grad_norm": 0.6937722855680643, "learning_rate": 0.0002919376340412102, "loss": 0.457, "step": 3127 }, { "epoch": 1.472, "grad_norm": 0.8618784074264217, "learning_rate": 0.00029192344989113425, "loss": 0.4785, "step": 3128 }, { "epoch": 1.4724705882352942, "grad_norm": 0.8801950169261521, "learning_rate": 0.0002919092536200995, "loss": 0.375, "step": 3129 }, { "epoch": 1.4729411764705882, "grad_norm": 0.5871805958519569, "learning_rate": 0.00029189504522931847, "loss": 0.3672, "step": 3130 }, { "epoch": 1.4734117647058824, "grad_norm": 1.0256911400491313, "learning_rate": 0.00029188082472000453, "loss": 0.6406, "step": 3131 }, { "epoch": 1.4738823529411764, "grad_norm": 0.8363312722899093, "learning_rate": 0.0002918665920933722, "loss": 0.375, "step": 3132 }, { "epoch": 1.4743529411764706, "grad_norm": 0.36554560412562476, "learning_rate": 0.0002918523473506372, "loss": 0.2207, "step": 3133 }, { "epoch": 1.4748235294117646, "grad_norm": 0.8389442789749372, "learning_rate": 0.00029183809049301585, "loss": 0.4531, "step": 3134 }, { "epoch": 1.4752941176470589, "grad_norm": 1.000151431436795, "learning_rate": 0.0002918238215217259, "loss": 0.5391, "step": 3135 }, { "epoch": 1.4757647058823529, "grad_norm": 0.5296387211997113, "learning_rate": 0.00029180954043798595, "loss": 0.3379, "step": 3136 }, { "epoch": 1.476235294117647, "grad_norm": 0.4049050174390671, "learning_rate": 0.00029179524724301563, "loss": 0.2207, "step": 3137 }, { "epoch": 1.4767058823529413, "grad_norm": 0.44296545759798767, "learning_rate": 0.00029178094193803576, "loss": 0.1167, "step": 3138 }, { "epoch": 1.4771764705882353, "grad_norm": 0.3086051127189462, "learning_rate": 0.00029176662452426796, "loss": 0.0986, "step": 3139 }, { "epoch": 1.4776470588235293, "grad_norm": 0.6763662375873982, "learning_rate": 0.0002917522950029351, "loss": 0.3145, "step": 3140 }, { "epoch": 1.4781176470588235, "grad_norm": 0.44682838057116636, "learning_rate": 0.00029173795337526093, "loss": 0.1279, "step": 3141 }, { "epoch": 1.4785882352941178, "grad_norm": 1.0212775159763734, "learning_rate": 0.0002917235996424703, "loss": 0.5039, "step": 3142 }, { "epoch": 1.4790588235294118, "grad_norm": 0.8809647849474782, "learning_rate": 0.0002917092338057892, "loss": 0.3984, "step": 3143 }, { "epoch": 1.4795294117647058, "grad_norm": 0.4062929054669361, "learning_rate": 0.00029169485586644434, "loss": 0.2129, "step": 3144 }, { "epoch": 1.48, "grad_norm": 0.985559629304731, "learning_rate": 0.00029168046582566385, "loss": 0.5156, "step": 3145 }, { "epoch": 1.4804705882352942, "grad_norm": 0.9476333789201534, "learning_rate": 0.00029166606368467667, "loss": 0.6992, "step": 3146 }, { "epoch": 1.4809411764705882, "grad_norm": 0.4139095550004189, "learning_rate": 0.0002916516494447127, "loss": 0.3438, "step": 3147 }, { "epoch": 1.4814117647058824, "grad_norm": 1.2718820516721514, "learning_rate": 0.0002916372231070031, "loss": 0.8203, "step": 3148 }, { "epoch": 1.4818823529411764, "grad_norm": 0.8545195230433796, "learning_rate": 0.00029162278467277994, "loss": 0.4941, "step": 3149 }, { "epoch": 1.4823529411764707, "grad_norm": 0.730722262161757, "learning_rate": 0.0002916083341432763, "loss": 0.5547, "step": 3150 }, { "epoch": 1.4828235294117647, "grad_norm": 0.3970793541405011, "learning_rate": 0.0002915938715197264, "loss": 0.2275, "step": 3151 }, { "epoch": 1.4832941176470589, "grad_norm": 0.48589647867732566, "learning_rate": 0.00029157939680336527, "loss": 0.2393, "step": 3152 }, { "epoch": 1.4837647058823529, "grad_norm": 0.6025667142755579, "learning_rate": 0.00029156490999542923, "loss": 0.3594, "step": 3153 }, { "epoch": 1.484235294117647, "grad_norm": 0.6067351820761697, "learning_rate": 0.0002915504110971555, "loss": 0.3438, "step": 3154 }, { "epoch": 1.484705882352941, "grad_norm": 0.5968602198174439, "learning_rate": 0.0002915359001097824, "loss": 0.4375, "step": 3155 }, { "epoch": 1.4851764705882353, "grad_norm": 0.800122514410122, "learning_rate": 0.0002915213770345492, "loss": 0.3027, "step": 3156 }, { "epoch": 1.4856470588235293, "grad_norm": 0.5416932924978439, "learning_rate": 0.0002915068418726962, "loss": 0.3125, "step": 3157 }, { "epoch": 1.4861176470588235, "grad_norm": 0.4702242717424918, "learning_rate": 0.00029149229462546483, "loss": 0.0947, "step": 3158 }, { "epoch": 1.4865882352941178, "grad_norm": 0.910667655085887, "learning_rate": 0.0002914777352940975, "loss": 0.4355, "step": 3159 }, { "epoch": 1.4870588235294118, "grad_norm": 0.34736132933325964, "learning_rate": 0.0002914631638798376, "loss": 0.2256, "step": 3160 }, { "epoch": 1.4875294117647058, "grad_norm": 0.2719494892242855, "learning_rate": 0.00029144858038392964, "loss": 0.1191, "step": 3161 }, { "epoch": 1.488, "grad_norm": 0.6192855804843529, "learning_rate": 0.00029143398480761915, "loss": 0.252, "step": 3162 }, { "epoch": 1.4884705882352942, "grad_norm": 0.1603833518386672, "learning_rate": 0.00029141937715215256, "loss": 0.0461, "step": 3163 }, { "epoch": 1.4889411764705882, "grad_norm": 0.19414192548082113, "learning_rate": 0.0002914047574187776, "loss": 0.1069, "step": 3164 }, { "epoch": 1.4894117647058824, "grad_norm": 0.10733326929577801, "learning_rate": 0.0002913901256087427, "loss": 0.0245, "step": 3165 }, { "epoch": 1.4898823529411764, "grad_norm": 0.318607087351056, "learning_rate": 0.00029137548172329756, "loss": 0.1123, "step": 3166 }, { "epoch": 1.4903529411764707, "grad_norm": 0.33035853773915475, "learning_rate": 0.0002913608257636929, "loss": 0.1562, "step": 3167 }, { "epoch": 1.4908235294117647, "grad_norm": 0.04674087527268697, "learning_rate": 0.0002913461577311803, "loss": 0.0079, "step": 3168 }, { "epoch": 1.4912941176470589, "grad_norm": 0.9213577633574086, "learning_rate": 0.00029133147762701246, "loss": 0.3945, "step": 3169 }, { "epoch": 1.4917647058823529, "grad_norm": 1.1509025727970648, "learning_rate": 0.00029131678545244333, "loss": 0.6289, "step": 3170 }, { "epoch": 1.492235294117647, "grad_norm": 0.6906313738241581, "learning_rate": 0.00029130208120872747, "loss": 0.4629, "step": 3171 }, { "epoch": 1.492705882352941, "grad_norm": 0.565104065024753, "learning_rate": 0.00029128736489712087, "loss": 0.4473, "step": 3172 }, { "epoch": 1.4931764705882353, "grad_norm": 0.7225622869385315, "learning_rate": 0.0002912726365188803, "loss": 0.3613, "step": 3173 }, { "epoch": 1.4936470588235293, "grad_norm": 0.7723724378813056, "learning_rate": 0.0002912578960752636, "loss": 0.4375, "step": 3174 }, { "epoch": 1.4941176470588236, "grad_norm": 0.9807304275639461, "learning_rate": 0.00029124314356752967, "loss": 0.5273, "step": 3175 }, { "epoch": 1.4945882352941178, "grad_norm": 0.35279538270553723, "learning_rate": 0.00029122837899693847, "loss": 0.105, "step": 3176 }, { "epoch": 1.4950588235294118, "grad_norm": 0.47197008492281367, "learning_rate": 0.0002912136023647511, "loss": 0.1875, "step": 3177 }, { "epoch": 1.4955294117647058, "grad_norm": 0.4823699497903927, "learning_rate": 0.0002911988136722293, "loss": 0.1992, "step": 3178 }, { "epoch": 1.496, "grad_norm": 0.6152766753884527, "learning_rate": 0.00029118401292063634, "loss": 0.2432, "step": 3179 }, { "epoch": 1.4964705882352942, "grad_norm": 0.49610040448845094, "learning_rate": 0.00029116920011123615, "loss": 0.2305, "step": 3180 }, { "epoch": 1.4969411764705882, "grad_norm": 0.6237223434336241, "learning_rate": 0.00029115437524529383, "loss": 0.2617, "step": 3181 }, { "epoch": 1.4974117647058822, "grad_norm": 0.6407677995138333, "learning_rate": 0.0002911395383240755, "loss": 0.3379, "step": 3182 }, { "epoch": 1.4978823529411764, "grad_norm": 0.4758341740340116, "learning_rate": 0.0002911246893488483, "loss": 0.1885, "step": 3183 }, { "epoch": 1.4983529411764707, "grad_norm": 1.1813399452355395, "learning_rate": 0.0002911098283208805, "loss": 0.6953, "step": 3184 }, { "epoch": 1.4988235294117647, "grad_norm": 0.5001415635981636, "learning_rate": 0.0002910949552414411, "loss": 0.2188, "step": 3185 }, { "epoch": 1.4992941176470589, "grad_norm": 0.424881387810134, "learning_rate": 0.0002910800701118005, "loss": 0.2031, "step": 3186 }, { "epoch": 1.4997647058823529, "grad_norm": 1.0974792446629305, "learning_rate": 0.00029106517293323, "loss": 0.6758, "step": 3187 }, { "epoch": 1.5002352941176471, "grad_norm": 0.25005556966630876, "learning_rate": 0.0002910502637070017, "loss": 0.0698, "step": 3188 }, { "epoch": 1.5007058823529413, "grad_norm": 0.25846879549272567, "learning_rate": 0.0002910353424343891, "loss": 0.0967, "step": 3189 }, { "epoch": 1.5011764705882353, "grad_norm": 0.28593034575815696, "learning_rate": 0.00029102040911666645, "loss": 0.1113, "step": 3190 }, { "epoch": 1.5016470588235293, "grad_norm": 0.8883969271994847, "learning_rate": 0.0002910054637551092, "loss": 0.3926, "step": 3191 }, { "epoch": 1.5021176470588236, "grad_norm": 0.4601472576456537, "learning_rate": 0.0002909905063509936, "loss": 0.248, "step": 3192 }, { "epoch": 1.5025882352941178, "grad_norm": 1.1946131457865448, "learning_rate": 0.0002909755369055973, "loss": 0.7461, "step": 3193 }, { "epoch": 1.5030588235294118, "grad_norm": 0.5910610483757786, "learning_rate": 0.00029096055542019865, "loss": 0.4551, "step": 3194 }, { "epoch": 1.5035294117647058, "grad_norm": 0.6654547202435543, "learning_rate": 0.0002909455618960772, "loss": 0.4395, "step": 3195 }, { "epoch": 1.504, "grad_norm": 0.4162088768544188, "learning_rate": 0.0002909305563345134, "loss": 0.1108, "step": 3196 }, { "epoch": 1.5044705882352942, "grad_norm": 0.5770905006014709, "learning_rate": 0.0002909155387367888, "loss": 0.4277, "step": 3197 }, { "epoch": 1.5049411764705882, "grad_norm": 0.7072108708629942, "learning_rate": 0.0002909005091041861, "loss": 0.3438, "step": 3198 }, { "epoch": 1.5054117647058822, "grad_norm": 0.7043373997642437, "learning_rate": 0.00029088546743798866, "loss": 0.5, "step": 3199 }, { "epoch": 1.5058823529411764, "grad_norm": 0.9350628955176378, "learning_rate": 0.00029087041373948135, "loss": 0.4883, "step": 3200 }, { "epoch": 1.5063529411764707, "grad_norm": 0.2930644452309073, "learning_rate": 0.00029085534800994976, "loss": 0.1064, "step": 3201 }, { "epoch": 1.5068235294117647, "grad_norm": 0.3387759647054437, "learning_rate": 0.0002908402702506805, "loss": 0.1689, "step": 3202 }, { "epoch": 1.5072941176470587, "grad_norm": 0.4860216645743824, "learning_rate": 0.0002908251804629614, "loss": 0.4238, "step": 3203 }, { "epoch": 1.507764705882353, "grad_norm": 1.2154000509989538, "learning_rate": 0.0002908100786480811, "loss": 0.7773, "step": 3204 }, { "epoch": 1.5082352941176471, "grad_norm": 1.0024591060053152, "learning_rate": 0.0002907949648073294, "loss": 0.5703, "step": 3205 }, { "epoch": 1.5087058823529413, "grad_norm": 0.3451465591689003, "learning_rate": 0.00029077983894199716, "loss": 0.1367, "step": 3206 }, { "epoch": 1.5091764705882353, "grad_norm": 0.9792906908682882, "learning_rate": 0.00029076470105337615, "loss": 0.4746, "step": 3207 }, { "epoch": 1.5096470588235293, "grad_norm": 1.4964373162425788, "learning_rate": 0.0002907495511427592, "loss": 0.668, "step": 3208 }, { "epoch": 1.5101176470588236, "grad_norm": 0.302894471174187, "learning_rate": 0.0002907343892114402, "loss": 0.1226, "step": 3209 }, { "epoch": 1.5105882352941178, "grad_norm": 0.5937259703996237, "learning_rate": 0.0002907192152607141, "loss": 0.2578, "step": 3210 }, { "epoch": 1.5110588235294118, "grad_norm": 0.3862567169523613, "learning_rate": 0.00029070402929187673, "loss": 0.1064, "step": 3211 }, { "epoch": 1.5115294117647058, "grad_norm": 1.5175293547980173, "learning_rate": 0.0002906888313062252, "loss": 0.5742, "step": 3212 }, { "epoch": 1.512, "grad_norm": 0.7169238199621281, "learning_rate": 0.00029067362130505734, "loss": 0.2344, "step": 3213 }, { "epoch": 1.5124705882352942, "grad_norm": 0.46729404291012516, "learning_rate": 0.00029065839928967224, "loss": 0.2793, "step": 3214 }, { "epoch": 1.5129411764705882, "grad_norm": 0.597483875179376, "learning_rate": 0.0002906431652613699, "loss": 0.2578, "step": 3215 }, { "epoch": 1.5134117647058822, "grad_norm": 0.8095084589548126, "learning_rate": 0.0002906279192214514, "loss": 0.2598, "step": 3216 }, { "epoch": 1.5138823529411765, "grad_norm": 0.06806238761045512, "learning_rate": 0.0002906126611712188, "loss": 0.0103, "step": 3217 }, { "epoch": 1.5143529411764707, "grad_norm": 1.0516501852888183, "learning_rate": 0.00029059739111197524, "loss": 0.4512, "step": 3218 }, { "epoch": 1.5148235294117647, "grad_norm": 1.444816784023697, "learning_rate": 0.0002905821090450249, "loss": 0.6328, "step": 3219 }, { "epoch": 1.5152941176470587, "grad_norm": 0.7219584505831064, "learning_rate": 0.00029056681497167285, "loss": 0.3203, "step": 3220 }, { "epoch": 1.515764705882353, "grad_norm": 0.5302510803549249, "learning_rate": 0.00029055150889322536, "loss": 0.2949, "step": 3221 }, { "epoch": 1.5162352941176471, "grad_norm": 1.316439740809237, "learning_rate": 0.0002905361908109896, "loss": 0.5352, "step": 3222 }, { "epoch": 1.5167058823529411, "grad_norm": 0.8073940569158071, "learning_rate": 0.0002905208607262738, "loss": 0.2891, "step": 3223 }, { "epoch": 1.5171764705882353, "grad_norm": 0.9208903296447348, "learning_rate": 0.0002905055186403873, "loss": 0.4238, "step": 3224 }, { "epoch": 1.5176470588235293, "grad_norm": 0.6300451216186169, "learning_rate": 0.0002904901645546403, "loss": 0.3438, "step": 3225 }, { "epoch": 1.5181176470588236, "grad_norm": 0.6604942502691994, "learning_rate": 0.0002904747984703441, "loss": 0.4824, "step": 3226 }, { "epoch": 1.5185882352941178, "grad_norm": 0.7512346247804114, "learning_rate": 0.00029045942038881113, "loss": 0.3906, "step": 3227 }, { "epoch": 1.5190588235294118, "grad_norm": 1.165295089057958, "learning_rate": 0.00029044403031135466, "loss": 0.5625, "step": 3228 }, { "epoch": 1.5195294117647058, "grad_norm": 0.7531086454358745, "learning_rate": 0.0002904286282392892, "loss": 0.3008, "step": 3229 }, { "epoch": 1.52, "grad_norm": 0.9048136872756994, "learning_rate": 0.00029041321417393006, "loss": 0.3047, "step": 3230 }, { "epoch": 1.5204705882352942, "grad_norm": 2.7785065103156565, "learning_rate": 0.0002903977881165937, "loss": 1.8672, "step": 3231 }, { "epoch": 1.5209411764705882, "grad_norm": 0.7335320003045418, "learning_rate": 0.0002903823500685976, "loss": 0.375, "step": 3232 }, { "epoch": 1.5214117647058822, "grad_norm": 1.0030334424345342, "learning_rate": 0.0002903669000312602, "loss": 0.3574, "step": 3233 }, { "epoch": 1.5218823529411765, "grad_norm": 1.5679728599072225, "learning_rate": 0.0002903514380059011, "loss": 0.2539, "step": 3234 }, { "epoch": 1.5223529411764707, "grad_norm": 0.7961036664994136, "learning_rate": 0.0002903359639938407, "loss": 0.4531, "step": 3235 }, { "epoch": 1.5228235294117647, "grad_norm": 0.283361423126367, "learning_rate": 0.0002903204779964007, "loss": 0.1084, "step": 3236 }, { "epoch": 1.5232941176470587, "grad_norm": 0.9736095574244049, "learning_rate": 0.00029030498001490356, "loss": 0.75, "step": 3237 }, { "epoch": 1.523764705882353, "grad_norm": 1.2206557873696773, "learning_rate": 0.00029028947005067297, "loss": 0.7344, "step": 3238 }, { "epoch": 1.5242352941176471, "grad_norm": 0.3718394144934464, "learning_rate": 0.00029027394810503346, "loss": 0.207, "step": 3239 }, { "epoch": 1.5247058823529411, "grad_norm": 1.0037269636579527, "learning_rate": 0.0002902584141793108, "loss": 0.6523, "step": 3240 }, { "epoch": 1.5251764705882351, "grad_norm": 0.33469470542717294, "learning_rate": 0.0002902428682748316, "loss": 0.2051, "step": 3241 }, { "epoch": 1.5256470588235294, "grad_norm": 0.8276970414038041, "learning_rate": 0.0002902273103929236, "loss": 0.4434, "step": 3242 }, { "epoch": 1.5261176470588236, "grad_norm": 0.8473892195346581, "learning_rate": 0.00029021174053491547, "loss": 0.6367, "step": 3243 }, { "epoch": 1.5265882352941178, "grad_norm": 0.5743080763016544, "learning_rate": 0.0002901961587021369, "loss": 0.2305, "step": 3244 }, { "epoch": 1.5270588235294118, "grad_norm": 0.9287278370289362, "learning_rate": 0.00029018056489591874, "loss": 0.4043, "step": 3245 }, { "epoch": 1.5275294117647058, "grad_norm": 0.9830275827115587, "learning_rate": 0.0002901649591175928, "loss": 0.416, "step": 3246 }, { "epoch": 1.528, "grad_norm": 0.5961883327676067, "learning_rate": 0.0002901493413684918, "loss": 0.3105, "step": 3247 }, { "epoch": 1.5284705882352942, "grad_norm": 0.37561459152289634, "learning_rate": 0.0002901337116499496, "loss": 0.2197, "step": 3248 }, { "epoch": 1.5289411764705882, "grad_norm": 0.4813200707872986, "learning_rate": 0.0002901180699633011, "loss": 0.2129, "step": 3249 }, { "epoch": 1.5294117647058822, "grad_norm": 1.6597381861586202, "learning_rate": 0.00029010241630988217, "loss": 0.7812, "step": 3250 }, { "epoch": 1.5298823529411765, "grad_norm": 0.4412653376937733, "learning_rate": 0.0002900867506910297, "loss": 0.3535, "step": 3251 }, { "epoch": 1.5303529411764707, "grad_norm": 0.7069960969887371, "learning_rate": 0.00029007107310808153, "loss": 0.4023, "step": 3252 }, { "epoch": 1.5308235294117647, "grad_norm": 0.6535034193118592, "learning_rate": 0.00029005538356237674, "loss": 0.4082, "step": 3253 }, { "epoch": 1.5312941176470587, "grad_norm": 0.8681713270898427, "learning_rate": 0.00029003968205525523, "loss": 0.498, "step": 3254 }, { "epoch": 1.531764705882353, "grad_norm": 0.8292216818778146, "learning_rate": 0.00029002396858805794, "loss": 0.4395, "step": 3255 }, { "epoch": 1.5322352941176471, "grad_norm": 0.4444698184860006, "learning_rate": 0.0002900082431621269, "loss": 0.1797, "step": 3256 }, { "epoch": 1.5327058823529411, "grad_norm": 0.6115055457150091, "learning_rate": 0.0002899925057788052, "loss": 0.3965, "step": 3257 }, { "epoch": 1.5331764705882351, "grad_norm": 0.3561611900665171, "learning_rate": 0.00028997675643943686, "loss": 0.1709, "step": 3258 }, { "epoch": 1.5336470588235294, "grad_norm": 0.8381939305709574, "learning_rate": 0.0002899609951453669, "loss": 0.4883, "step": 3259 }, { "epoch": 1.5341176470588236, "grad_norm": 0.7717582496441046, "learning_rate": 0.0002899452218979415, "loss": 0.3691, "step": 3260 }, { "epoch": 1.5345882352941178, "grad_norm": 0.2628887262837708, "learning_rate": 0.0002899294366985076, "loss": 0.1001, "step": 3261 }, { "epoch": 1.5350588235294118, "grad_norm": 0.6403486503016241, "learning_rate": 0.0002899136395484136, "loss": 0.4766, "step": 3262 }, { "epoch": 1.5355294117647058, "grad_norm": 0.499894922753514, "learning_rate": 0.00028989783044900845, "loss": 0.2832, "step": 3263 }, { "epoch": 1.536, "grad_norm": 0.48457956018040793, "learning_rate": 0.0002898820094016424, "loss": 0.2197, "step": 3264 }, { "epoch": 1.5364705882352943, "grad_norm": 0.6606943297171065, "learning_rate": 0.0002898661764076666, "loss": 0.3262, "step": 3265 }, { "epoch": 1.5369411764705883, "grad_norm": 0.30939428812011266, "learning_rate": 0.00028985033146843335, "loss": 0.1182, "step": 3266 }, { "epoch": 1.5374117647058823, "grad_norm": 0.6293345704081018, "learning_rate": 0.00028983447458529575, "loss": 0.4102, "step": 3267 }, { "epoch": 1.5378823529411765, "grad_norm": 0.46394496037832406, "learning_rate": 0.0002898186057596082, "loss": 0.3574, "step": 3268 }, { "epoch": 1.5383529411764707, "grad_norm": 0.21134177105965543, "learning_rate": 0.0002898027249927259, "loss": 0.123, "step": 3269 }, { "epoch": 1.5388235294117647, "grad_norm": 0.39538855190339167, "learning_rate": 0.00028978683228600514, "loss": 0.1914, "step": 3270 }, { "epoch": 1.5392941176470587, "grad_norm": 0.7134567839094356, "learning_rate": 0.0002897709276408032, "loss": 0.418, "step": 3271 }, { "epoch": 1.539764705882353, "grad_norm": 0.5863435342465918, "learning_rate": 0.00028975501105847856, "loss": 0.293, "step": 3272 }, { "epoch": 1.5402352941176471, "grad_norm": 0.2589620295906423, "learning_rate": 0.00028973908254039044, "loss": 0.0903, "step": 3273 }, { "epoch": 1.5407058823529411, "grad_norm": 0.5495561905184214, "learning_rate": 0.00028972314208789926, "loss": 0.4473, "step": 3274 }, { "epoch": 1.5411764705882351, "grad_norm": 0.723343234359232, "learning_rate": 0.00028970718970236644, "loss": 0.4023, "step": 3275 }, { "epoch": 1.5416470588235294, "grad_norm": 0.4884920989887348, "learning_rate": 0.00028969122538515435, "loss": 0.3496, "step": 3276 }, { "epoch": 1.5421176470588236, "grad_norm": 0.8807266566542935, "learning_rate": 0.00028967524913762636, "loss": 0.5156, "step": 3277 }, { "epoch": 1.5425882352941176, "grad_norm": 0.22480029280925717, "learning_rate": 0.00028965926096114706, "loss": 0.0723, "step": 3278 }, { "epoch": 1.5430588235294118, "grad_norm": 0.5022860429984638, "learning_rate": 0.00028964326085708183, "loss": 0.2676, "step": 3279 }, { "epoch": 1.5435294117647058, "grad_norm": 1.2156742150397388, "learning_rate": 0.00028962724882679715, "loss": 0.4922, "step": 3280 }, { "epoch": 1.544, "grad_norm": 0.30442011546655806, "learning_rate": 0.00028961122487166063, "loss": 0.106, "step": 3281 }, { "epoch": 1.5444705882352943, "grad_norm": 1.1326935166589125, "learning_rate": 0.00028959518899304064, "loss": 0.7969, "step": 3282 }, { "epoch": 1.5449411764705883, "grad_norm": 0.6730238763088898, "learning_rate": 0.00028957914119230686, "loss": 0.3633, "step": 3283 }, { "epoch": 1.5454117647058823, "grad_norm": 0.6962632378100578, "learning_rate": 0.0002895630814708297, "loss": 0.3496, "step": 3284 }, { "epoch": 1.5458823529411765, "grad_norm": 0.9630860560125789, "learning_rate": 0.00028954700982998093, "loss": 0.4043, "step": 3285 }, { "epoch": 1.5463529411764707, "grad_norm": 0.43031561415626585, "learning_rate": 0.000289530926271133, "loss": 0.1309, "step": 3286 }, { "epoch": 1.5468235294117647, "grad_norm": 0.26954059229510785, "learning_rate": 0.0002895148307956596, "loss": 0.1172, "step": 3287 }, { "epoch": 1.5472941176470587, "grad_norm": 0.5448087285408556, "learning_rate": 0.00028949872340493525, "loss": 0.1904, "step": 3288 }, { "epoch": 1.547764705882353, "grad_norm": 0.47518406232092564, "learning_rate": 0.00028948260410033576, "loss": 0.1416, "step": 3289 }, { "epoch": 1.5482352941176472, "grad_norm": 0.6814463294309605, "learning_rate": 0.00028946647288323766, "loss": 0.3945, "step": 3290 }, { "epoch": 1.5487058823529412, "grad_norm": 0.6751572674888878, "learning_rate": 0.00028945032975501877, "loss": 0.6016, "step": 3291 }, { "epoch": 1.5491764705882352, "grad_norm": 0.31705268274275517, "learning_rate": 0.0002894341747170577, "loss": 0.1484, "step": 3292 }, { "epoch": 1.5496470588235294, "grad_norm": 0.2746014093062837, "learning_rate": 0.00028941800777073414, "loss": 0.167, "step": 3293 }, { "epoch": 1.5501176470588236, "grad_norm": 1.0388545384639416, "learning_rate": 0.0002894018289174289, "loss": 0.4141, "step": 3294 }, { "epoch": 1.5505882352941176, "grad_norm": 0.7815512659693453, "learning_rate": 0.0002893856381585237, "loss": 0.4297, "step": 3295 }, { "epoch": 1.5510588235294116, "grad_norm": 0.5333901097540472, "learning_rate": 0.0002893694354954014, "loss": 0.1973, "step": 3296 }, { "epoch": 1.5515294117647058, "grad_norm": 0.864707672113009, "learning_rate": 0.00028935322092944564, "loss": 0.4512, "step": 3297 }, { "epoch": 1.552, "grad_norm": 0.5748059699734884, "learning_rate": 0.00028933699446204126, "loss": 0.2734, "step": 3298 }, { "epoch": 1.5524705882352943, "grad_norm": 0.6551897622617444, "learning_rate": 0.0002893207560945742, "loss": 0.5898, "step": 3299 }, { "epoch": 1.5529411764705883, "grad_norm": 0.5717589681653132, "learning_rate": 0.0002893045058284311, "loss": 0.3066, "step": 3300 }, { "epoch": 1.5534117647058823, "grad_norm": 0.3888443690607038, "learning_rate": 0.000289288243665, "loss": 0.2129, "step": 3301 }, { "epoch": 1.5538823529411765, "grad_norm": 0.9498512735935507, "learning_rate": 0.00028927196960566964, "loss": 0.3984, "step": 3302 }, { "epoch": 1.5543529411764707, "grad_norm": 0.5227650602001622, "learning_rate": 0.00028925568365182995, "loss": 0.416, "step": 3303 }, { "epoch": 1.5548235294117647, "grad_norm": 0.637147829895383, "learning_rate": 0.00028923938580487183, "loss": 0.4062, "step": 3304 }, { "epoch": 1.5552941176470587, "grad_norm": 0.5886777640267865, "learning_rate": 0.0002892230760661872, "loss": 0.3906, "step": 3305 }, { "epoch": 1.555764705882353, "grad_norm": 0.2192182094955465, "learning_rate": 0.000289206754437169, "loss": 0.0801, "step": 3306 }, { "epoch": 1.5562352941176472, "grad_norm": 0.2983364441899268, "learning_rate": 0.0002891904209192112, "loss": 0.0737, "step": 3307 }, { "epoch": 1.5567058823529412, "grad_norm": 0.7052224877735163, "learning_rate": 0.00028917407551370863, "loss": 0.4727, "step": 3308 }, { "epoch": 1.5571764705882352, "grad_norm": 0.5897432495372956, "learning_rate": 0.00028915771822205747, "loss": 0.3691, "step": 3309 }, { "epoch": 1.5576470588235294, "grad_norm": 0.23668722417350704, "learning_rate": 0.00028914134904565455, "loss": 0.0566, "step": 3310 }, { "epoch": 1.5581176470588236, "grad_norm": 0.164122033093889, "learning_rate": 0.00028912496798589796, "loss": 0.0376, "step": 3311 }, { "epoch": 1.5585882352941176, "grad_norm": 0.3135910826138231, "learning_rate": 0.0002891085750441867, "loss": 0.0801, "step": 3312 }, { "epoch": 1.5590588235294116, "grad_norm": 0.85378351442452, "learning_rate": 0.0002890921702219208, "loss": 0.3438, "step": 3313 }, { "epoch": 1.5595294117647058, "grad_norm": 0.7893335399528032, "learning_rate": 0.00028907575352050133, "loss": 0.5078, "step": 3314 }, { "epoch": 1.56, "grad_norm": 0.5301127224275124, "learning_rate": 0.00028905932494133034, "loss": 0.2969, "step": 3315 }, { "epoch": 1.5604705882352943, "grad_norm": 0.493803098250413, "learning_rate": 0.0002890428844858109, "loss": 0.2266, "step": 3316 }, { "epoch": 1.5609411764705883, "grad_norm": 0.3691127580524037, "learning_rate": 0.0002890264321553471, "loss": 0.1836, "step": 3317 }, { "epoch": 1.5614117647058823, "grad_norm": 0.49827691634788246, "learning_rate": 0.00028900996795134414, "loss": 0.2676, "step": 3318 }, { "epoch": 1.5618823529411765, "grad_norm": 0.45146148441686795, "learning_rate": 0.000288993491875208, "loss": 0.3086, "step": 3319 }, { "epoch": 1.5623529411764707, "grad_norm": 0.5658504751864312, "learning_rate": 0.000288977003928346, "loss": 0.3281, "step": 3320 }, { "epoch": 1.5628235294117647, "grad_norm": 0.8436013024544687, "learning_rate": 0.0002889605041121661, "loss": 0.668, "step": 3321 }, { "epoch": 1.5632941176470587, "grad_norm": 0.8042231574677605, "learning_rate": 0.0002889439924280776, "loss": 0.4414, "step": 3322 }, { "epoch": 1.563764705882353, "grad_norm": 0.6923537477772361, "learning_rate": 0.0002889274688774906, "loss": 0.5195, "step": 3323 }, { "epoch": 1.5642352941176472, "grad_norm": 0.6721486139728803, "learning_rate": 0.00028891093346181633, "loss": 0.4434, "step": 3324 }, { "epoch": 1.5647058823529412, "grad_norm": 0.29545256792047575, "learning_rate": 0.000288894386182467, "loss": 0.1387, "step": 3325 }, { "epoch": 1.5651764705882352, "grad_norm": 0.47787010002095437, "learning_rate": 0.0002888778270408558, "loss": 0.2637, "step": 3326 }, { "epoch": 1.5656470588235294, "grad_norm": 0.34497486517172304, "learning_rate": 0.000288861256038397, "loss": 0.1592, "step": 3327 }, { "epoch": 1.5661176470588236, "grad_norm": 0.9469728083441037, "learning_rate": 0.0002888446731765058, "loss": 0.4434, "step": 3328 }, { "epoch": 1.5665882352941176, "grad_norm": 0.5190051701589937, "learning_rate": 0.00028882807845659846, "loss": 0.2373, "step": 3329 }, { "epoch": 1.5670588235294116, "grad_norm": 0.8905092467961419, "learning_rate": 0.00028881147188009233, "loss": 0.5938, "step": 3330 }, { "epoch": 1.5675294117647058, "grad_norm": 0.25921666067467186, "learning_rate": 0.00028879485344840563, "loss": 0.1025, "step": 3331 }, { "epoch": 1.568, "grad_norm": 0.7305552532108316, "learning_rate": 0.00028877822316295763, "loss": 0.459, "step": 3332 }, { "epoch": 1.5684705882352943, "grad_norm": 0.9838577734743685, "learning_rate": 0.0002887615810251687, "loss": 0.5391, "step": 3333 }, { "epoch": 1.5689411764705883, "grad_norm": 0.3018100256459629, "learning_rate": 0.0002887449270364601, "loss": 0.1348, "step": 3334 }, { "epoch": 1.5694117647058823, "grad_norm": 0.43033777600565953, "learning_rate": 0.0002887282611982542, "loss": 0.1758, "step": 3335 }, { "epoch": 1.5698823529411765, "grad_norm": 0.7713558143993403, "learning_rate": 0.0002887115835119743, "loss": 0.4219, "step": 3336 }, { "epoch": 1.5703529411764707, "grad_norm": 0.8520930761796794, "learning_rate": 0.00028869489397904486, "loss": 0.4004, "step": 3337 }, { "epoch": 1.5708235294117647, "grad_norm": 0.8401225587941268, "learning_rate": 0.00028867819260089113, "loss": 0.6992, "step": 3338 }, { "epoch": 1.5712941176470587, "grad_norm": 0.7825573469228737, "learning_rate": 0.0002886614793789396, "loss": 0.4004, "step": 3339 }, { "epoch": 1.571764705882353, "grad_norm": 1.039974057396725, "learning_rate": 0.0002886447543146175, "loss": 0.4277, "step": 3340 }, { "epoch": 1.5722352941176472, "grad_norm": 0.45784302518256226, "learning_rate": 0.0002886280174093534, "loss": 0.2754, "step": 3341 }, { "epoch": 1.5727058823529412, "grad_norm": 0.3645950111968548, "learning_rate": 0.0002886112686645766, "loss": 0.1719, "step": 3342 }, { "epoch": 1.5731764705882352, "grad_norm": 1.2605563049084585, "learning_rate": 0.00028859450808171763, "loss": 0.6367, "step": 3343 }, { "epoch": 1.5736470588235294, "grad_norm": 0.8038680689358217, "learning_rate": 0.00028857773566220784, "loss": 0.377, "step": 3344 }, { "epoch": 1.5741176470588236, "grad_norm": 0.3610777905309743, "learning_rate": 0.0002885609514074797, "loss": 0.1973, "step": 3345 }, { "epoch": 1.5745882352941176, "grad_norm": 1.2016647428268203, "learning_rate": 0.0002885441553189667, "loss": 0.6484, "step": 3346 }, { "epoch": 1.5750588235294116, "grad_norm": 0.797941327347, "learning_rate": 0.00028852734739810324, "loss": 0.4668, "step": 3347 }, { "epoch": 1.5755294117647058, "grad_norm": 0.481033335416695, "learning_rate": 0.00028851052764632487, "loss": 0.2754, "step": 3348 }, { "epoch": 1.576, "grad_norm": 0.7331501072489436, "learning_rate": 0.00028849369606506807, "loss": 0.4434, "step": 3349 }, { "epoch": 1.576470588235294, "grad_norm": 0.5052341910581787, "learning_rate": 0.0002884768526557703, "loss": 0.3457, "step": 3350 }, { "epoch": 1.5769411764705883, "grad_norm": 1.3207496039709568, "learning_rate": 0.00028845999741987003, "loss": 0.7812, "step": 3351 }, { "epoch": 1.5774117647058823, "grad_norm": 0.4636958169485171, "learning_rate": 0.0002884431303588069, "loss": 0.3555, "step": 3352 }, { "epoch": 1.5778823529411765, "grad_norm": 0.581553110504934, "learning_rate": 0.00028842625147402134, "loss": 0.3438, "step": 3353 }, { "epoch": 1.5783529411764707, "grad_norm": 0.5776267325251178, "learning_rate": 0.00028840936076695493, "loss": 0.3516, "step": 3354 }, { "epoch": 1.5788235294117647, "grad_norm": 0.7659287444298065, "learning_rate": 0.0002883924582390503, "loss": 0.4727, "step": 3355 }, { "epoch": 1.5792941176470587, "grad_norm": 0.7502957896780829, "learning_rate": 0.0002883755438917508, "loss": 0.5039, "step": 3356 }, { "epoch": 1.579764705882353, "grad_norm": 0.20088575916394055, "learning_rate": 0.0002883586177265012, "loss": 0.0378, "step": 3357 }, { "epoch": 1.5802352941176472, "grad_norm": 0.2852712619275631, "learning_rate": 0.000288341679744747, "loss": 0.0664, "step": 3358 }, { "epoch": 1.5807058823529412, "grad_norm": 0.04182320349158939, "learning_rate": 0.0002883247299479348, "loss": 0.0128, "step": 3359 }, { "epoch": 1.5811764705882352, "grad_norm": 0.2617430669152892, "learning_rate": 0.0002883077683375121, "loss": 0.0854, "step": 3360 }, { "epoch": 1.5816470588235294, "grad_norm": 0.9790460071458565, "learning_rate": 0.0002882907949149277, "loss": 0.3496, "step": 3361 }, { "epoch": 1.5821176470588236, "grad_norm": 0.5138345185338906, "learning_rate": 0.000288273809681631, "loss": 0.209, "step": 3362 }, { "epoch": 1.5825882352941176, "grad_norm": 0.8271651434201458, "learning_rate": 0.00028825681263907276, "loss": 0.4395, "step": 3363 }, { "epoch": 1.5830588235294116, "grad_norm": 0.2744249110943665, "learning_rate": 0.00028823980378870457, "loss": 0.1079, "step": 3364 }, { "epoch": 1.5835294117647059, "grad_norm": 0.85138097088943, "learning_rate": 0.0002882227831319791, "loss": 0.3789, "step": 3365 }, { "epoch": 1.584, "grad_norm": 0.5084966201248217, "learning_rate": 0.00028820575067035, "loss": 0.2949, "step": 3366 }, { "epoch": 1.584470588235294, "grad_norm": 0.8607906367727092, "learning_rate": 0.00028818870640527183, "loss": 0.4707, "step": 3367 }, { "epoch": 1.584941176470588, "grad_norm": 0.9405297763820358, "learning_rate": 0.0002881716503382004, "loss": 0.5742, "step": 3368 }, { "epoch": 1.5854117647058823, "grad_norm": 0.7266105431447538, "learning_rate": 0.0002881545824705922, "loss": 0.5039, "step": 3369 }, { "epoch": 1.5858823529411765, "grad_norm": 0.408551986121882, "learning_rate": 0.00028813750280390506, "loss": 0.1484, "step": 3370 }, { "epoch": 1.5863529411764707, "grad_norm": 0.6007010841155924, "learning_rate": 0.00028812041133959764, "loss": 0.375, "step": 3371 }, { "epoch": 1.5868235294117647, "grad_norm": 0.8286488678750633, "learning_rate": 0.0002881033080791296, "loss": 0.4863, "step": 3372 }, { "epoch": 1.5872941176470587, "grad_norm": 0.5873550795854943, "learning_rate": 0.0002880861930239616, "loss": 0.3828, "step": 3373 }, { "epoch": 1.587764705882353, "grad_norm": 0.5647963551499091, "learning_rate": 0.0002880690661755555, "loss": 0.3242, "step": 3374 }, { "epoch": 1.5882352941176472, "grad_norm": 0.3135304631976477, "learning_rate": 0.00028805192753537386, "loss": 0.1201, "step": 3375 }, { "epoch": 1.5887058823529412, "grad_norm": 0.4225008254589228, "learning_rate": 0.00028803477710488055, "loss": 0.1709, "step": 3376 }, { "epoch": 1.5891764705882352, "grad_norm": 0.8297310940303364, "learning_rate": 0.00028801761488554017, "loss": 0.4258, "step": 3377 }, { "epoch": 1.5896470588235294, "grad_norm": 0.8860463217918532, "learning_rate": 0.0002880004408788185, "loss": 0.4004, "step": 3378 }, { "epoch": 1.5901176470588236, "grad_norm": 0.4592493901245264, "learning_rate": 0.0002879832550861823, "loss": 0.1943, "step": 3379 }, { "epoch": 1.5905882352941176, "grad_norm": 0.3387853011550775, "learning_rate": 0.0002879660575090993, "loss": 0.167, "step": 3380 }, { "epoch": 1.5910588235294116, "grad_norm": 0.8228418290666584, "learning_rate": 0.0002879488481490382, "loss": 0.4102, "step": 3381 }, { "epoch": 1.5915294117647059, "grad_norm": 0.7692634786680473, "learning_rate": 0.0002879316270074689, "loss": 0.0928, "step": 3382 }, { "epoch": 1.592, "grad_norm": 0.26235839773301645, "learning_rate": 0.00028791439408586215, "loss": 0.0942, "step": 3383 }, { "epoch": 1.592470588235294, "grad_norm": 0.5122369299416855, "learning_rate": 0.00028789714938568965, "loss": 0.2412, "step": 3384 }, { "epoch": 1.592941176470588, "grad_norm": 0.8400545251312498, "learning_rate": 0.0002878798929084242, "loss": 0.5625, "step": 3385 }, { "epoch": 1.5934117647058823, "grad_norm": 0.4984773125449095, "learning_rate": 0.00028786262465553954, "loss": 0.3828, "step": 3386 }, { "epoch": 1.5938823529411765, "grad_norm": 0.6512133900639157, "learning_rate": 0.00028784534462851063, "loss": 0.332, "step": 3387 }, { "epoch": 1.5943529411764708, "grad_norm": 0.9485922384952272, "learning_rate": 0.0002878280528288131, "loss": 0.6719, "step": 3388 }, { "epoch": 1.5948235294117648, "grad_norm": 0.5856841387191777, "learning_rate": 0.00028781074925792377, "loss": 0.4277, "step": 3389 }, { "epoch": 1.5952941176470588, "grad_norm": 0.765675429066403, "learning_rate": 0.00028779343391732053, "loss": 0.4414, "step": 3390 }, { "epoch": 1.595764705882353, "grad_norm": 1.0372351080894264, "learning_rate": 0.0002877761068084822, "loss": 0.6406, "step": 3391 }, { "epoch": 1.5962352941176472, "grad_norm": 0.29897441168694877, "learning_rate": 0.00028775876793288855, "loss": 0.1533, "step": 3392 }, { "epoch": 1.5967058823529412, "grad_norm": 0.2898915899194574, "learning_rate": 0.0002877414172920204, "loss": 0.0972, "step": 3393 }, { "epoch": 1.5971764705882352, "grad_norm": 0.6313794659631421, "learning_rate": 0.00028772405488735963, "loss": 0.373, "step": 3394 }, { "epoch": 1.5976470588235294, "grad_norm": 0.551248253716249, "learning_rate": 0.000287706680720389, "loss": 0.1729, "step": 3395 }, { "epoch": 1.5981176470588236, "grad_norm": 0.8511173474516102, "learning_rate": 0.0002876892947925924, "loss": 0.5234, "step": 3396 }, { "epoch": 1.5985882352941176, "grad_norm": 0.530760326533525, "learning_rate": 0.0002876718971054547, "loss": 0.3574, "step": 3397 }, { "epoch": 1.5990588235294116, "grad_norm": 0.7719155853096871, "learning_rate": 0.00028765448766046163, "loss": 0.4238, "step": 3398 }, { "epoch": 1.5995294117647059, "grad_norm": 0.7754093626002447, "learning_rate": 0.00028763706645910014, "loss": 0.4102, "step": 3399 }, { "epoch": 1.6, "grad_norm": 0.4300153388128627, "learning_rate": 0.0002876196335028581, "loss": 0.124, "step": 3400 }, { "epoch": 1.600470588235294, "grad_norm": 0.38213289944909107, "learning_rate": 0.00028760218879322436, "loss": 0.2949, "step": 3401 }, { "epoch": 1.600941176470588, "grad_norm": 0.36029258289970717, "learning_rate": 0.00028758473233168874, "loss": 0.1689, "step": 3402 }, { "epoch": 1.6014117647058823, "grad_norm": 1.0522654263338878, "learning_rate": 0.0002875672641197421, "loss": 0.5859, "step": 3403 }, { "epoch": 1.6018823529411765, "grad_norm": 0.38930067135264124, "learning_rate": 0.00028754978415887635, "loss": 0.1943, "step": 3404 }, { "epoch": 1.6023529411764705, "grad_norm": 0.5253036372904321, "learning_rate": 0.0002875322924505844, "loss": 0.375, "step": 3405 }, { "epoch": 1.6028235294117648, "grad_norm": 0.21354244932009503, "learning_rate": 0.00028751478899636, "loss": 0.0562, "step": 3406 }, { "epoch": 1.6032941176470588, "grad_norm": 1.292442408925704, "learning_rate": 0.00028749727379769817, "loss": 0.6641, "step": 3407 }, { "epoch": 1.603764705882353, "grad_norm": 0.46005024927222027, "learning_rate": 0.0002874797468560948, "loss": 0.2656, "step": 3408 }, { "epoch": 1.6042352941176472, "grad_norm": 1.0508197295840613, "learning_rate": 0.0002874622081730466, "loss": 0.8672, "step": 3409 }, { "epoch": 1.6047058823529412, "grad_norm": 0.8463172806847586, "learning_rate": 0.0002874446577500516, "loss": 0.625, "step": 3410 }, { "epoch": 1.6051764705882352, "grad_norm": 0.7522926489126089, "learning_rate": 0.0002874270955886087, "loss": 0.4004, "step": 3411 }, { "epoch": 1.6056470588235294, "grad_norm": 0.3966101956056107, "learning_rate": 0.0002874095216902178, "loss": 0.1699, "step": 3412 }, { "epoch": 1.6061176470588236, "grad_norm": 0.37769581768739224, "learning_rate": 0.00028739193605637967, "loss": 0.1719, "step": 3413 }, { "epoch": 1.6065882352941177, "grad_norm": 0.5863688963422279, "learning_rate": 0.00028737433868859634, "loss": 0.1514, "step": 3414 }, { "epoch": 1.6070588235294117, "grad_norm": 0.5509192448890498, "learning_rate": 0.0002873567295883707, "loss": 0.248, "step": 3415 }, { "epoch": 1.6075294117647059, "grad_norm": 0.806592223209973, "learning_rate": 0.00028733910875720654, "loss": 0.5781, "step": 3416 }, { "epoch": 1.608, "grad_norm": 0.38944827425767364, "learning_rate": 0.00028732147619660883, "loss": 0.1475, "step": 3417 }, { "epoch": 1.608470588235294, "grad_norm": 0.7650065870232743, "learning_rate": 0.00028730383190808364, "loss": 0.3359, "step": 3418 }, { "epoch": 1.608941176470588, "grad_norm": 0.7160044800215675, "learning_rate": 0.0002872861758931376, "loss": 0.3516, "step": 3419 }, { "epoch": 1.6094117647058823, "grad_norm": 1.127969636005743, "learning_rate": 0.0002872685081532788, "loss": 0.7422, "step": 3420 }, { "epoch": 1.6098823529411765, "grad_norm": 0.609561279037776, "learning_rate": 0.00028725082869001615, "loss": 0.3711, "step": 3421 }, { "epoch": 1.6103529411764705, "grad_norm": 0.07089399574476477, "learning_rate": 0.00028723313750485946, "loss": 0.0138, "step": 3422 }, { "epoch": 1.6108235294117645, "grad_norm": 0.3270753227488268, "learning_rate": 0.0002872154345993197, "loss": 0.1147, "step": 3423 }, { "epoch": 1.6112941176470588, "grad_norm": 0.24404870826703057, "learning_rate": 0.0002871977199749088, "loss": 0.0713, "step": 3424 }, { "epoch": 1.611764705882353, "grad_norm": 1.8935321254781954, "learning_rate": 0.00028717999363313967, "loss": 0.7891, "step": 3425 }, { "epoch": 1.6122352941176472, "grad_norm": 0.5431863419105253, "learning_rate": 0.0002871622555755262, "loss": 0.2197, "step": 3426 }, { "epoch": 1.6127058823529412, "grad_norm": 1.2491205731129842, "learning_rate": 0.00028714450580358334, "loss": 0.7656, "step": 3427 }, { "epoch": 1.6131764705882352, "grad_norm": 0.9280694662768819, "learning_rate": 0.000287126744318827, "loss": 0.7578, "step": 3428 }, { "epoch": 1.6136470588235294, "grad_norm": 0.49863955040639363, "learning_rate": 0.0002871089711227741, "loss": 0.2441, "step": 3429 }, { "epoch": 1.6141176470588237, "grad_norm": 0.8815489574267806, "learning_rate": 0.0002870911862169426, "loss": 0.4141, "step": 3430 }, { "epoch": 1.6145882352941177, "grad_norm": 0.7935103029611038, "learning_rate": 0.0002870733896028513, "loss": 0.3926, "step": 3431 }, { "epoch": 1.6150588235294117, "grad_norm": 0.22154801987501957, "learning_rate": 0.00028705558128202017, "loss": 0.0869, "step": 3432 }, { "epoch": 1.6155294117647059, "grad_norm": 0.6266633801543559, "learning_rate": 0.0002870377612559702, "loss": 0.5312, "step": 3433 }, { "epoch": 1.616, "grad_norm": 0.862137707936665, "learning_rate": 0.0002870199295262232, "loss": 0.4141, "step": 3434 }, { "epoch": 1.616470588235294, "grad_norm": 0.7097385471252584, "learning_rate": 0.00028700208609430217, "loss": 0.3516, "step": 3435 }, { "epoch": 1.616941176470588, "grad_norm": 0.5794981945392625, "learning_rate": 0.00028698423096173097, "loss": 0.2969, "step": 3436 }, { "epoch": 1.6174117647058823, "grad_norm": 0.5596266201221978, "learning_rate": 0.0002869663641300346, "loss": 0.4824, "step": 3437 }, { "epoch": 1.6178823529411765, "grad_norm": 0.8983678150982479, "learning_rate": 0.00028694848560073885, "loss": 0.5547, "step": 3438 }, { "epoch": 1.6183529411764705, "grad_norm": 0.6823194251244922, "learning_rate": 0.00028693059537537077, "loss": 0.4023, "step": 3439 }, { "epoch": 1.6188235294117646, "grad_norm": 0.7998735631760006, "learning_rate": 0.00028691269345545816, "loss": 0.3945, "step": 3440 }, { "epoch": 1.6192941176470588, "grad_norm": 0.9372143680174381, "learning_rate": 0.00028689477984252996, "loss": 0.5508, "step": 3441 }, { "epoch": 1.619764705882353, "grad_norm": 0.7134890218671077, "learning_rate": 0.00028687685453811614, "loss": 0.1865, "step": 3442 }, { "epoch": 1.6202352941176472, "grad_norm": 0.43211869322543317, "learning_rate": 0.00028685891754374755, "loss": 0.1553, "step": 3443 }, { "epoch": 1.6207058823529412, "grad_norm": 0.5716155268110942, "learning_rate": 0.0002868409688609561, "loss": 0.2891, "step": 3444 }, { "epoch": 1.6211764705882352, "grad_norm": 0.4240158414405704, "learning_rate": 0.00028682300849127465, "loss": 0.1748, "step": 3445 }, { "epoch": 1.6216470588235294, "grad_norm": 0.9751182739334988, "learning_rate": 0.0002868050364362373, "loss": 0.3203, "step": 3446 }, { "epoch": 1.6221176470588237, "grad_norm": 0.6125562866517958, "learning_rate": 0.00028678705269737873, "loss": 0.2314, "step": 3447 }, { "epoch": 1.6225882352941177, "grad_norm": 0.7712903716389199, "learning_rate": 0.0002867690572762349, "loss": 0.3125, "step": 3448 }, { "epoch": 1.6230588235294117, "grad_norm": 0.5889682690500896, "learning_rate": 0.0002867510501743428, "loss": 0.2637, "step": 3449 }, { "epoch": 1.6235294117647059, "grad_norm": 0.5983193540558828, "learning_rate": 0.0002867330313932402, "loss": 0.4238, "step": 3450 }, { "epoch": 1.624, "grad_norm": 0.6156922935758209, "learning_rate": 0.00028671500093446604, "loss": 0.3672, "step": 3451 }, { "epoch": 1.624470588235294, "grad_norm": 0.0928199072326897, "learning_rate": 0.0002866969587995602, "loss": 0.0215, "step": 3452 }, { "epoch": 1.6249411764705881, "grad_norm": 0.6197473007831555, "learning_rate": 0.00028667890499006364, "loss": 0.2402, "step": 3453 }, { "epoch": 1.6254117647058823, "grad_norm": 1.1141285334040416, "learning_rate": 0.00028666083950751824, "loss": 0.4375, "step": 3454 }, { "epoch": 1.6258823529411766, "grad_norm": 0.3362948144204534, "learning_rate": 0.0002866427623534667, "loss": 0.1738, "step": 3455 }, { "epoch": 1.6263529411764706, "grad_norm": 0.8630194767423445, "learning_rate": 0.0002866246735294531, "loss": 0.4219, "step": 3456 }, { "epoch": 1.6268235294117646, "grad_norm": 0.3977805378507253, "learning_rate": 0.00028660657303702225, "loss": 0.1875, "step": 3457 }, { "epoch": 1.6272941176470588, "grad_norm": 0.6718950997076049, "learning_rate": 0.00028658846087771994, "loss": 0.1934, "step": 3458 }, { "epoch": 1.627764705882353, "grad_norm": 0.7687368959017442, "learning_rate": 0.0002865703370530932, "loss": 0.3711, "step": 3459 }, { "epoch": 1.6282352941176472, "grad_norm": 1.5977185587631875, "learning_rate": 0.0002865522015646897, "loss": 0.5859, "step": 3460 }, { "epoch": 1.6287058823529412, "grad_norm": 0.27934016750744317, "learning_rate": 0.0002865340544140585, "loss": 0.0859, "step": 3461 }, { "epoch": 1.6291764705882352, "grad_norm": 0.4325976149474358, "learning_rate": 0.00028651589560274934, "loss": 0.1748, "step": 3462 }, { "epoch": 1.6296470588235294, "grad_norm": 1.1232840203681778, "learning_rate": 0.00028649772513231306, "loss": 0.4102, "step": 3463 }, { "epoch": 1.6301176470588237, "grad_norm": 1.619198512210644, "learning_rate": 0.00028647954300430154, "loss": 0.5664, "step": 3464 }, { "epoch": 1.6305882352941177, "grad_norm": 0.45698692721044626, "learning_rate": 0.0002864613492202676, "loss": 0.3262, "step": 3465 }, { "epoch": 1.6310588235294117, "grad_norm": 0.909213670750775, "learning_rate": 0.00028644314378176513, "loss": 0.5781, "step": 3466 }, { "epoch": 1.631529411764706, "grad_norm": 0.4540306502331186, "learning_rate": 0.00028642492669034894, "loss": 0.1514, "step": 3467 }, { "epoch": 1.6320000000000001, "grad_norm": 1.5655117113574628, "learning_rate": 0.0002864066979475748, "loss": 0.3906, "step": 3468 }, { "epoch": 1.6324705882352941, "grad_norm": 0.8340453187644473, "learning_rate": 0.00028638845755499964, "loss": 0.3125, "step": 3469 }, { "epoch": 1.6329411764705881, "grad_norm": 1.0393405103411542, "learning_rate": 0.00028637020551418123, "loss": 0.416, "step": 3470 }, { "epoch": 1.6334117647058823, "grad_norm": 1.0275306643340154, "learning_rate": 0.00028635194182667834, "loss": 0.7148, "step": 3471 }, { "epoch": 1.6338823529411766, "grad_norm": 0.6033240868666676, "learning_rate": 0.0002863336664940508, "loss": 0.3418, "step": 3472 }, { "epoch": 1.6343529411764706, "grad_norm": 0.3634139591908357, "learning_rate": 0.00028631537951785945, "loss": 0.1699, "step": 3473 }, { "epoch": 1.6348235294117646, "grad_norm": 0.8003754469658498, "learning_rate": 0.000286297080899666, "loss": 0.2451, "step": 3474 }, { "epoch": 1.6352941176470588, "grad_norm": 0.8172217151169006, "learning_rate": 0.0002862787706410334, "loss": 0.252, "step": 3475 }, { "epoch": 1.635764705882353, "grad_norm": 1.2922757237910016, "learning_rate": 0.00028626044874352534, "loss": 0.4766, "step": 3476 }, { "epoch": 1.636235294117647, "grad_norm": 0.28605149147601144, "learning_rate": 0.0002862421152087066, "loss": 0.1113, "step": 3477 }, { "epoch": 1.636705882352941, "grad_norm": 0.8292428558486432, "learning_rate": 0.00028622377003814295, "loss": 0.375, "step": 3478 }, { "epoch": 1.6371764705882352, "grad_norm": 0.5280597927792031, "learning_rate": 0.00028620541323340115, "loss": 0.3086, "step": 3479 }, { "epoch": 1.6376470588235295, "grad_norm": 0.5218480713198957, "learning_rate": 0.00028618704479604895, "loss": 0.3008, "step": 3480 }, { "epoch": 1.6381176470588237, "grad_norm": 0.6248127147326854, "learning_rate": 0.00028616866472765514, "loss": 0.3477, "step": 3481 }, { "epoch": 1.6385882352941177, "grad_norm": 0.814887730824187, "learning_rate": 0.00028615027302978947, "loss": 0.1855, "step": 3482 }, { "epoch": 1.6390588235294117, "grad_norm": 1.985242814949125, "learning_rate": 0.00028613186970402274, "loss": 0.6016, "step": 3483 }, { "epoch": 1.639529411764706, "grad_norm": 0.4768636834127736, "learning_rate": 0.0002861134547519265, "loss": 0.1973, "step": 3484 }, { "epoch": 1.6400000000000001, "grad_norm": 0.7346499254607937, "learning_rate": 0.0002860950281750737, "loss": 0.6562, "step": 3485 }, { "epoch": 1.6404705882352941, "grad_norm": 0.9427316159372734, "learning_rate": 0.00028607658997503786, "loss": 0.5977, "step": 3486 }, { "epoch": 1.6409411764705881, "grad_norm": 0.5865639403141694, "learning_rate": 0.00028605814015339383, "loss": 0.2637, "step": 3487 }, { "epoch": 1.6414117647058823, "grad_norm": 0.3971465719929169, "learning_rate": 0.0002860396787117173, "loss": 0.0972, "step": 3488 }, { "epoch": 1.6418823529411766, "grad_norm": 1.1770261784618887, "learning_rate": 0.00028602120565158486, "loss": 0.5586, "step": 3489 }, { "epoch": 1.6423529411764706, "grad_norm": 0.5805096268342019, "learning_rate": 0.0002860027209745743, "loss": 0.3516, "step": 3490 }, { "epoch": 1.6428235294117646, "grad_norm": 0.9132920023122486, "learning_rate": 0.00028598422468226425, "loss": 0.5352, "step": 3491 }, { "epoch": 1.6432941176470588, "grad_norm": 0.4919369025832289, "learning_rate": 0.0002859657167762345, "loss": 0.375, "step": 3492 }, { "epoch": 1.643764705882353, "grad_norm": 0.8802916006653778, "learning_rate": 0.00028594719725806556, "loss": 0.3555, "step": 3493 }, { "epoch": 1.644235294117647, "grad_norm": 0.7964796044509754, "learning_rate": 0.0002859286661293392, "loss": 0.3516, "step": 3494 }, { "epoch": 1.644705882352941, "grad_norm": 0.8080572573540259, "learning_rate": 0.00028591012339163806, "loss": 0.3496, "step": 3495 }, { "epoch": 1.6451764705882352, "grad_norm": 0.5620655172166396, "learning_rate": 0.00028589156904654565, "loss": 0.2422, "step": 3496 }, { "epoch": 1.6456470588235295, "grad_norm": 0.40638968723501157, "learning_rate": 0.00028587300309564675, "loss": 0.126, "step": 3497 }, { "epoch": 1.6461176470588237, "grad_norm": 0.35605086632859884, "learning_rate": 0.0002858544255405269, "loss": 0.1406, "step": 3498 }, { "epoch": 1.6465882352941177, "grad_norm": 0.7337739403821542, "learning_rate": 0.0002858358363827728, "loss": 0.5586, "step": 3499 }, { "epoch": 1.6470588235294117, "grad_norm": 0.5862875221095291, "learning_rate": 0.000285817235623972, "loss": 0.2812, "step": 3500 }, { "epoch": 1.647529411764706, "grad_norm": 1.0675784920031546, "learning_rate": 0.00028579862326571305, "loss": 0.8008, "step": 3501 }, { "epoch": 1.6480000000000001, "grad_norm": 0.6703736258410206, "learning_rate": 0.0002857799993095856, "loss": 0.4121, "step": 3502 }, { "epoch": 1.6484705882352941, "grad_norm": 0.6595728561052334, "learning_rate": 0.0002857613637571802, "loss": 0.4258, "step": 3503 }, { "epoch": 1.6489411764705881, "grad_norm": 0.7819129003120052, "learning_rate": 0.00028574271661008845, "loss": 0.416, "step": 3504 }, { "epoch": 1.6494117647058824, "grad_norm": 0.9325531245694538, "learning_rate": 0.00028572405786990294, "loss": 0.707, "step": 3505 }, { "epoch": 1.6498823529411766, "grad_norm": 0.5645883795372104, "learning_rate": 0.00028570538753821704, "loss": 0.543, "step": 3506 }, { "epoch": 1.6503529411764706, "grad_norm": 0.5829271095437708, "learning_rate": 0.0002856867056166255, "loss": 0.4551, "step": 3507 }, { "epoch": 1.6508235294117646, "grad_norm": 0.3566359635463925, "learning_rate": 0.00028566801210672376, "loss": 0.1377, "step": 3508 }, { "epoch": 1.6512941176470588, "grad_norm": 0.7798259015430598, "learning_rate": 0.0002856493070101084, "loss": 0.4473, "step": 3509 }, { "epoch": 1.651764705882353, "grad_norm": 0.7011050292411276, "learning_rate": 0.0002856305903283767, "loss": 0.3594, "step": 3510 }, { "epoch": 1.652235294117647, "grad_norm": 1.4604287283175355, "learning_rate": 0.00028561186206312744, "loss": 0.8008, "step": 3511 }, { "epoch": 1.652705882352941, "grad_norm": 0.8764208154699447, "learning_rate": 0.00028559312221595996, "loss": 0.4336, "step": 3512 }, { "epoch": 1.6531764705882352, "grad_norm": 0.5382166248277216, "learning_rate": 0.00028557437078847477, "loss": 0.293, "step": 3513 }, { "epoch": 1.6536470588235295, "grad_norm": 0.49712857832532165, "learning_rate": 0.0002855556077822733, "loss": 0.3164, "step": 3514 }, { "epoch": 1.6541176470588237, "grad_norm": 0.5435994056009108, "learning_rate": 0.00028553683319895804, "loss": 0.2715, "step": 3515 }, { "epoch": 1.6545882352941177, "grad_norm": 0.25780153101525827, "learning_rate": 0.00028551804704013244, "loss": 0.1807, "step": 3516 }, { "epoch": 1.6550588235294117, "grad_norm": 0.27216910787902093, "learning_rate": 0.0002854992493074009, "loss": 0.0718, "step": 3517 }, { "epoch": 1.655529411764706, "grad_norm": 0.22412001919915428, "learning_rate": 0.00028548044000236886, "loss": 0.0659, "step": 3518 }, { "epoch": 1.6560000000000001, "grad_norm": 0.5298203315282076, "learning_rate": 0.0002854616191266427, "loss": 0.1895, "step": 3519 }, { "epoch": 1.6564705882352941, "grad_norm": 1.334684681973561, "learning_rate": 0.0002854427866818298, "loss": 0.6133, "step": 3520 }, { "epoch": 1.6569411764705881, "grad_norm": 0.44958299865845003, "learning_rate": 0.00028542394266953865, "loss": 0.1875, "step": 3521 }, { "epoch": 1.6574117647058824, "grad_norm": 0.9364385496056143, "learning_rate": 0.00028540508709137846, "loss": 0.4727, "step": 3522 }, { "epoch": 1.6578823529411766, "grad_norm": 0.5911583017584983, "learning_rate": 0.00028538621994895976, "loss": 0.4258, "step": 3523 }, { "epoch": 1.6583529411764706, "grad_norm": 0.5857500126891103, "learning_rate": 0.00028536734124389373, "loss": 0.2891, "step": 3524 }, { "epoch": 1.6588235294117646, "grad_norm": 0.528902535581868, "learning_rate": 0.0002853484509777928, "loss": 0.291, "step": 3525 }, { "epoch": 1.6592941176470588, "grad_norm": 0.9441338599277899, "learning_rate": 0.00028532954915227033, "loss": 0.4219, "step": 3526 }, { "epoch": 1.659764705882353, "grad_norm": 0.18858090629532637, "learning_rate": 0.0002853106357689405, "loss": 0.0591, "step": 3527 }, { "epoch": 1.660235294117647, "grad_norm": 0.7684128936557731, "learning_rate": 0.0002852917108294187, "loss": 0.5391, "step": 3528 }, { "epoch": 1.660705882352941, "grad_norm": 0.5476601788374061, "learning_rate": 0.0002852727743353213, "loss": 0.2617, "step": 3529 }, { "epoch": 1.6611764705882353, "grad_norm": 1.0013497710344277, "learning_rate": 0.00028525382628826536, "loss": 0.7461, "step": 3530 }, { "epoch": 1.6616470588235295, "grad_norm": 0.7777850145699229, "learning_rate": 0.0002852348666898692, "loss": 0.3926, "step": 3531 }, { "epoch": 1.6621176470588235, "grad_norm": 0.687685111508236, "learning_rate": 0.0002852158955417522, "loss": 0.4531, "step": 3532 }, { "epoch": 1.6625882352941177, "grad_norm": 0.5737090090464597, "learning_rate": 0.0002851969128455345, "loss": 0.3359, "step": 3533 }, { "epoch": 1.6630588235294117, "grad_norm": 0.5644561928144999, "learning_rate": 0.0002851779186028372, "loss": 0.4766, "step": 3534 }, { "epoch": 1.663529411764706, "grad_norm": 0.4478431913174095, "learning_rate": 0.0002851589128152827, "loss": 0.1206, "step": 3535 }, { "epoch": 1.6640000000000001, "grad_norm": 0.28854880783397935, "learning_rate": 0.000285139895484494, "loss": 0.2012, "step": 3536 }, { "epoch": 1.6644705882352941, "grad_norm": 0.14777644604824244, "learning_rate": 0.0002851208666120955, "loss": 0.0518, "step": 3537 }, { "epoch": 1.6649411764705881, "grad_norm": 0.5512623183788512, "learning_rate": 0.0002851018261997121, "loss": 0.3984, "step": 3538 }, { "epoch": 1.6654117647058824, "grad_norm": 0.5812246183314734, "learning_rate": 0.00028508277424897015, "loss": 0.4121, "step": 3539 }, { "epoch": 1.6658823529411766, "grad_norm": 1.1740232395133146, "learning_rate": 0.00028506371076149665, "loss": 0.5156, "step": 3540 }, { "epoch": 1.6663529411764706, "grad_norm": 0.5807468285408109, "learning_rate": 0.0002850446357389198, "loss": 0.2471, "step": 3541 }, { "epoch": 1.6668235294117646, "grad_norm": 0.7320880408502146, "learning_rate": 0.00028502554918286867, "loss": 0.4707, "step": 3542 }, { "epoch": 1.6672941176470588, "grad_norm": 0.9948234684507037, "learning_rate": 0.0002850064510949733, "loss": 0.4297, "step": 3543 }, { "epoch": 1.667764705882353, "grad_norm": 0.625621601750781, "learning_rate": 0.00028498734147686485, "loss": 0.4297, "step": 3544 }, { "epoch": 1.668235294117647, "grad_norm": 0.7648552545512166, "learning_rate": 0.0002849682203301753, "loss": 0.3262, "step": 3545 }, { "epoch": 1.668705882352941, "grad_norm": 0.1687913637027432, "learning_rate": 0.0002849490876565378, "loss": 0.0327, "step": 3546 }, { "epoch": 1.6691764705882353, "grad_norm": 0.5701281763638483, "learning_rate": 0.0002849299434575862, "loss": 0.3301, "step": 3547 }, { "epoch": 1.6696470588235295, "grad_norm": 0.781220234969054, "learning_rate": 0.00028491078773495564, "loss": 0.3633, "step": 3548 }, { "epoch": 1.6701176470588235, "grad_norm": 0.9408148721763459, "learning_rate": 0.00028489162049028205, "loss": 0.459, "step": 3549 }, { "epoch": 1.6705882352941175, "grad_norm": 1.109886745777379, "learning_rate": 0.00028487244172520246, "loss": 0.5469, "step": 3550 }, { "epoch": 1.6710588235294117, "grad_norm": 0.16514198427937696, "learning_rate": 0.0002848532514413548, "loss": 0.0471, "step": 3551 }, { "epoch": 1.671529411764706, "grad_norm": 0.3050360734147803, "learning_rate": 0.00028483404964037794, "loss": 0.1494, "step": 3552 }, { "epoch": 1.6720000000000002, "grad_norm": 1.2368427231193828, "learning_rate": 0.000284814836323912, "loss": 0.5195, "step": 3553 }, { "epoch": 1.6724705882352942, "grad_norm": 0.3857700688522147, "learning_rate": 0.0002847956114935977, "loss": 0.1504, "step": 3554 }, { "epoch": 1.6729411764705882, "grad_norm": 0.8552737267435145, "learning_rate": 0.000284776375151077, "loss": 0.3809, "step": 3555 }, { "epoch": 1.6734117647058824, "grad_norm": 1.0735716864140143, "learning_rate": 0.00028475712729799285, "loss": 0.5898, "step": 3556 }, { "epoch": 1.6738823529411766, "grad_norm": 0.8168301844541009, "learning_rate": 0.000284737867935989, "loss": 0.4121, "step": 3557 }, { "epoch": 1.6743529411764706, "grad_norm": 0.6090567833014102, "learning_rate": 0.0002847185970667104, "loss": 0.3926, "step": 3558 }, { "epoch": 1.6748235294117646, "grad_norm": 0.5299749706626915, "learning_rate": 0.00028469931469180275, "loss": 0.2266, "step": 3559 }, { "epoch": 1.6752941176470588, "grad_norm": 1.0506288764253957, "learning_rate": 0.00028468002081291296, "loss": 0.5234, "step": 3560 }, { "epoch": 1.675764705882353, "grad_norm": 0.8695169702713037, "learning_rate": 0.0002846607154316888, "loss": 0.4277, "step": 3561 }, { "epoch": 1.676235294117647, "grad_norm": 0.6993711250942359, "learning_rate": 0.000284641398549779, "loss": 0.3867, "step": 3562 }, { "epoch": 1.676705882352941, "grad_norm": 1.2103686790570318, "learning_rate": 0.00028462207016883336, "loss": 0.8555, "step": 3563 }, { "epoch": 1.6771764705882353, "grad_norm": 0.7821500591236633, "learning_rate": 0.0002846027302905026, "loss": 0.4492, "step": 3564 }, { "epoch": 1.6776470588235295, "grad_norm": 0.728483449724052, "learning_rate": 0.0002845833789164385, "loss": 0.6875, "step": 3565 }, { "epoch": 1.6781176470588235, "grad_norm": 0.5489532486591993, "learning_rate": 0.00028456401604829367, "loss": 0.2773, "step": 3566 }, { "epoch": 1.6785882352941175, "grad_norm": 0.3032464930181781, "learning_rate": 0.00028454464168772186, "loss": 0.0947, "step": 3567 }, { "epoch": 1.6790588235294117, "grad_norm": 0.6510824484166666, "learning_rate": 0.00028452525583637774, "loss": 0.3047, "step": 3568 }, { "epoch": 1.679529411764706, "grad_norm": 0.842904181894912, "learning_rate": 0.0002845058584959169, "loss": 0.416, "step": 3569 }, { "epoch": 1.6800000000000002, "grad_norm": 0.4589921059114043, "learning_rate": 0.00028448644966799595, "loss": 0.1758, "step": 3570 }, { "epoch": 1.6804705882352942, "grad_norm": 0.8014642565263207, "learning_rate": 0.0002844670293542726, "loss": 0.4668, "step": 3571 }, { "epoch": 1.6809411764705882, "grad_norm": 0.35769500915342806, "learning_rate": 0.00028444759755640535, "loss": 0.1729, "step": 3572 }, { "epoch": 1.6814117647058824, "grad_norm": 0.25898602309046836, "learning_rate": 0.0002844281542760539, "loss": 0.106, "step": 3573 }, { "epoch": 1.6818823529411766, "grad_norm": 0.3737569325881282, "learning_rate": 0.0002844086995148786, "loss": 0.1475, "step": 3574 }, { "epoch": 1.6823529411764706, "grad_norm": 0.6828102696805299, "learning_rate": 0.00028438923327454114, "loss": 0.1079, "step": 3575 }, { "epoch": 1.6828235294117646, "grad_norm": 0.3037858074794296, "learning_rate": 0.000284369755556704, "loss": 0.0967, "step": 3576 }, { "epoch": 1.6832941176470588, "grad_norm": 1.1003091582471634, "learning_rate": 0.0002843502663630306, "loss": 0.4766, "step": 3577 }, { "epoch": 1.683764705882353, "grad_norm": 0.9222373793861636, "learning_rate": 0.0002843307656951855, "loss": 0.4395, "step": 3578 }, { "epoch": 1.684235294117647, "grad_norm": 0.6279358415503371, "learning_rate": 0.0002843112535548341, "loss": 0.3887, "step": 3579 }, { "epoch": 1.684705882352941, "grad_norm": 0.9793501733689143, "learning_rate": 0.0002842917299436429, "loss": 0.6328, "step": 3580 }, { "epoch": 1.6851764705882353, "grad_norm": 0.9459577974511543, "learning_rate": 0.00028427219486327923, "loss": 0.6836, "step": 3581 }, { "epoch": 1.6856470588235295, "grad_norm": 0.711713550428957, "learning_rate": 0.00028425264831541153, "loss": 0.3438, "step": 3582 }, { "epoch": 1.6861176470588235, "grad_norm": 0.30231773587286725, "learning_rate": 0.0002842330903017091, "loss": 0.123, "step": 3583 }, { "epoch": 1.6865882352941175, "grad_norm": 0.8103950448101285, "learning_rate": 0.00028421352082384245, "loss": 0.4082, "step": 3584 }, { "epoch": 1.6870588235294117, "grad_norm": 0.3285209900215518, "learning_rate": 0.00028419393988348275, "loss": 0.1777, "step": 3585 }, { "epoch": 1.687529411764706, "grad_norm": 1.0420956533910979, "learning_rate": 0.00028417434748230237, "loss": 0.4473, "step": 3586 }, { "epoch": 1.688, "grad_norm": 0.8202707061086957, "learning_rate": 0.0002841547436219746, "loss": 0.4102, "step": 3587 }, { "epoch": 1.6884705882352942, "grad_norm": 0.1327558986924051, "learning_rate": 0.0002841351283041737, "loss": 0.0461, "step": 3588 }, { "epoch": 1.6889411764705882, "grad_norm": 0.22903767218092452, "learning_rate": 0.0002841155015305749, "loss": 0.0864, "step": 3589 }, { "epoch": 1.6894117647058824, "grad_norm": 0.6123761605793643, "learning_rate": 0.00028409586330285444, "loss": 0.2656, "step": 3590 }, { "epoch": 1.6898823529411766, "grad_norm": 0.9288112232158269, "learning_rate": 0.00028407621362268955, "loss": 0.5547, "step": 3591 }, { "epoch": 1.6903529411764706, "grad_norm": 1.057045437615541, "learning_rate": 0.0002840565524917583, "loss": 0.5156, "step": 3592 }, { "epoch": 1.6908235294117646, "grad_norm": 0.6683353648339626, "learning_rate": 0.00028403687991174003, "loss": 0.4258, "step": 3593 }, { "epoch": 1.6912941176470588, "grad_norm": 1.0333011201193765, "learning_rate": 0.0002840171958843147, "loss": 0.4297, "step": 3594 }, { "epoch": 1.691764705882353, "grad_norm": 7.765675831501892, "learning_rate": 0.0002839975004111635, "loss": 0.4609, "step": 3595 }, { "epoch": 1.692235294117647, "grad_norm": 45.98685808389334, "learning_rate": 0.00028397779349396854, "loss": 0.4277, "step": 3596 }, { "epoch": 1.692705882352941, "grad_norm": 33.279206132313774, "learning_rate": 0.0002839580751344128, "loss": 0.7656, "step": 3597 }, { "epoch": 1.6931764705882353, "grad_norm": 327.9253408513855, "learning_rate": 0.00028393834533418043, "loss": 0.7891, "step": 3598 }, { "epoch": 1.6936470588235295, "grad_norm": 1.793649157908063, "learning_rate": 0.0002839186040949564, "loss": 0.2715, "step": 3599 }, { "epoch": 1.6941176470588235, "grad_norm": 0.9630475061635608, "learning_rate": 0.0002838988514184267, "loss": 0.3965, "step": 3600 }, { "epoch": 1.6945882352941175, "grad_norm": 0.5035099432947172, "learning_rate": 0.0002838790873062784, "loss": 0.4043, "step": 3601 }, { "epoch": 1.6950588235294117, "grad_norm": 0.605047184491221, "learning_rate": 0.0002838593117601992, "loss": 0.4746, "step": 3602 }, { "epoch": 1.695529411764706, "grad_norm": 0.8939656836887572, "learning_rate": 0.0002838395247818783, "loss": 0.3086, "step": 3603 }, { "epoch": 1.696, "grad_norm": 0.7951285837096982, "learning_rate": 0.0002838197263730055, "loss": 0.3086, "step": 3604 }, { "epoch": 1.696470588235294, "grad_norm": 0.783017928820617, "learning_rate": 0.0002837999165352717, "loss": 0.4141, "step": 3605 }, { "epoch": 1.6969411764705882, "grad_norm": 0.8255543306833171, "learning_rate": 0.0002837800952703687, "loss": 0.4316, "step": 3606 }, { "epoch": 1.6974117647058824, "grad_norm": 0.48730169029989034, "learning_rate": 0.00028376026257998947, "loss": 0.1416, "step": 3607 }, { "epoch": 1.6978823529411766, "grad_norm": 0.796329108488359, "learning_rate": 0.0002837404184658276, "loss": 0.0713, "step": 3608 }, { "epoch": 1.6983529411764706, "grad_norm": 0.7628634230594373, "learning_rate": 0.00028372056292957806, "loss": 0.3438, "step": 3609 }, { "epoch": 1.6988235294117646, "grad_norm": 0.6900383955136066, "learning_rate": 0.0002837006959729366, "loss": 0.2676, "step": 3610 }, { "epoch": 1.6992941176470588, "grad_norm": 0.7020691424201109, "learning_rate": 0.00028368081759759983, "loss": 0.5859, "step": 3611 }, { "epoch": 1.699764705882353, "grad_norm": 0.5724244388759112, "learning_rate": 0.00028366092780526554, "loss": 0.293, "step": 3612 }, { "epoch": 1.700235294117647, "grad_norm": 0.3239789912592156, "learning_rate": 0.00028364102659763247, "loss": 0.0972, "step": 3613 }, { "epoch": 1.700705882352941, "grad_norm": 0.5434759394198182, "learning_rate": 0.00028362111397640016, "loss": 0.3789, "step": 3614 }, { "epoch": 1.7011764705882353, "grad_norm": 0.4353350283175255, "learning_rate": 0.00028360118994326936, "loss": 0.2559, "step": 3615 }, { "epoch": 1.7016470588235295, "grad_norm": 0.47881422962386405, "learning_rate": 0.00028358125449994156, "loss": 0.2148, "step": 3616 }, { "epoch": 1.7021176470588235, "grad_norm": 1.1074501103000352, "learning_rate": 0.0002835613076481195, "loss": 0.5195, "step": 3617 }, { "epoch": 1.7025882352941175, "grad_norm": 0.6126762445113554, "learning_rate": 0.0002835413493895066, "loss": 0.375, "step": 3618 }, { "epoch": 1.7030588235294117, "grad_norm": 0.4812072396579831, "learning_rate": 0.00028352137972580745, "loss": 0.2676, "step": 3619 }, { "epoch": 1.703529411764706, "grad_norm": 0.5915998674115696, "learning_rate": 0.0002835013986587276, "loss": 0.4531, "step": 3620 }, { "epoch": 1.704, "grad_norm": 0.5766512090290205, "learning_rate": 0.0002834814061899734, "loss": 0.5078, "step": 3621 }, { "epoch": 1.704470588235294, "grad_norm": 0.6610485228442287, "learning_rate": 0.0002834614023212525, "loss": 0.4492, "step": 3622 }, { "epoch": 1.7049411764705882, "grad_norm": 0.33252758290042755, "learning_rate": 0.0002834413870542731, "loss": 0.1436, "step": 3623 }, { "epoch": 1.7054117647058824, "grad_norm": 0.4994278622211525, "learning_rate": 0.00028342136039074476, "loss": 0.1895, "step": 3624 }, { "epoch": 1.7058823529411766, "grad_norm": 0.16834523073448457, "learning_rate": 0.0002834013223323778, "loss": 0.05, "step": 3625 }, { "epoch": 1.7063529411764706, "grad_norm": 0.3623626294321344, "learning_rate": 0.0002833812728808836, "loss": 0.1484, "step": 3626 }, { "epoch": 1.7068235294117646, "grad_norm": 0.1283326563327976, "learning_rate": 0.00028336121203797446, "loss": 0.0197, "step": 3627 }, { "epoch": 1.7072941176470589, "grad_norm": 0.2791440309100756, "learning_rate": 0.0002833411398053637, "loss": 0.1011, "step": 3628 }, { "epoch": 1.707764705882353, "grad_norm": 0.8073604340705169, "learning_rate": 0.00028332105618476556, "loss": 0.3789, "step": 3629 }, { "epoch": 1.708235294117647, "grad_norm": 0.5314077011134644, "learning_rate": 0.0002833009611778953, "loss": 0.2334, "step": 3630 }, { "epoch": 1.708705882352941, "grad_norm": 0.7311471876017961, "learning_rate": 0.0002832808547864691, "loss": 0.4648, "step": 3631 }, { "epoch": 1.7091764705882353, "grad_norm": 0.7417738879378096, "learning_rate": 0.00028326073701220416, "loss": 0.4648, "step": 3632 }, { "epoch": 1.7096470588235295, "grad_norm": 0.5751208237856914, "learning_rate": 0.0002832406078568186, "loss": 0.2793, "step": 3633 }, { "epoch": 1.7101176470588235, "grad_norm": 0.7465254551928308, "learning_rate": 0.00028322046732203165, "loss": 0.5664, "step": 3634 }, { "epoch": 1.7105882352941175, "grad_norm": 0.4996752522166758, "learning_rate": 0.0002832003154095633, "loss": 0.2402, "step": 3635 }, { "epoch": 1.7110588235294117, "grad_norm": 0.7713562898153319, "learning_rate": 0.0002831801521211347, "loss": 0.4043, "step": 3636 }, { "epoch": 1.711529411764706, "grad_norm": 0.5273745158488228, "learning_rate": 0.0002831599774584679, "loss": 0.2637, "step": 3637 }, { "epoch": 1.712, "grad_norm": 0.3236764382084382, "learning_rate": 0.0002831397914232858, "loss": 0.1406, "step": 3638 }, { "epoch": 1.712470588235294, "grad_norm": 0.7421634340152626, "learning_rate": 0.0002831195940173125, "loss": 0.2891, "step": 3639 }, { "epoch": 1.7129411764705882, "grad_norm": 0.5205478150821575, "learning_rate": 0.00028309938524227296, "loss": 0.3105, "step": 3640 }, { "epoch": 1.7134117647058824, "grad_norm": 0.5050835451970671, "learning_rate": 0.000283079165099893, "loss": 0.1738, "step": 3641 }, { "epoch": 1.7138823529411766, "grad_norm": 0.45753247255615037, "learning_rate": 0.0002830589335918997, "loss": 0.168, "step": 3642 }, { "epoch": 1.7143529411764706, "grad_norm": 0.4313170800591543, "learning_rate": 0.0002830386907200207, "loss": 0.1514, "step": 3643 }, { "epoch": 1.7148235294117646, "grad_norm": 0.9830260016127061, "learning_rate": 0.000283018436485985, "loss": 0.5117, "step": 3644 }, { "epoch": 1.7152941176470589, "grad_norm": 0.35037370435657916, "learning_rate": 0.0002829981708915224, "loss": 0.1348, "step": 3645 }, { "epoch": 1.715764705882353, "grad_norm": 0.3558862211309021, "learning_rate": 0.0002829778939383636, "loss": 0.1777, "step": 3646 }, { "epoch": 1.716235294117647, "grad_norm": 0.46707920845177814, "learning_rate": 0.0002829576056282405, "loss": 0.2363, "step": 3647 }, { "epoch": 1.716705882352941, "grad_norm": 0.7310042864836375, "learning_rate": 0.00028293730596288565, "loss": 0.373, "step": 3648 }, { "epoch": 1.7171764705882353, "grad_norm": 0.6990890448525916, "learning_rate": 0.00028291699494403284, "loss": 0.3516, "step": 3649 }, { "epoch": 1.7176470588235295, "grad_norm": 0.39802708582689084, "learning_rate": 0.0002828966725734167, "loss": 0.1826, "step": 3650 }, { "epoch": 1.7181176470588235, "grad_norm": 0.43649306906544144, "learning_rate": 0.0002828763388527729, "loss": 0.2656, "step": 3651 }, { "epoch": 1.7185882352941175, "grad_norm": 0.8273684300678723, "learning_rate": 0.000282855993783838, "loss": 0.2988, "step": 3652 }, { "epoch": 1.7190588235294118, "grad_norm": 0.7614457585165283, "learning_rate": 0.0002828356373683495, "loss": 0.4336, "step": 3653 }, { "epoch": 1.719529411764706, "grad_norm": 0.698163489148244, "learning_rate": 0.0002828152696080461, "loss": 0.3594, "step": 3654 }, { "epoch": 1.72, "grad_norm": 0.29774823511042786, "learning_rate": 0.0002827948905046672, "loss": 0.1387, "step": 3655 }, { "epoch": 1.720470588235294, "grad_norm": 0.7990664993106855, "learning_rate": 0.00028277450005995326, "loss": 0.4395, "step": 3656 }, { "epoch": 1.7209411764705882, "grad_norm": 0.4280092889217418, "learning_rate": 0.0002827540982756458, "loss": 0.2168, "step": 3657 }, { "epoch": 1.7214117647058824, "grad_norm": 0.7660301874061176, "learning_rate": 0.00028273368515348713, "loss": 0.4551, "step": 3658 }, { "epoch": 1.7218823529411764, "grad_norm": 0.7598325935503237, "learning_rate": 0.00028271326069522076, "loss": 0.4434, "step": 3659 }, { "epoch": 1.7223529411764706, "grad_norm": 0.7425331231670033, "learning_rate": 0.0002826928249025909, "loss": 0.3906, "step": 3660 }, { "epoch": 1.7228235294117646, "grad_norm": 0.33669750675466154, "learning_rate": 0.000282672377777343, "loss": 0.1475, "step": 3661 }, { "epoch": 1.7232941176470589, "grad_norm": 0.5970877690226029, "learning_rate": 0.0002826519193212233, "loss": 0.3262, "step": 3662 }, { "epoch": 1.723764705882353, "grad_norm": 0.30312876517185844, "learning_rate": 0.000282631449535979, "loss": 0.1167, "step": 3663 }, { "epoch": 1.724235294117647, "grad_norm": 0.667839755681149, "learning_rate": 0.00028261096842335836, "loss": 0.5312, "step": 3664 }, { "epoch": 1.724705882352941, "grad_norm": 0.5955151622382114, "learning_rate": 0.00028259047598511053, "loss": 0.3125, "step": 3665 }, { "epoch": 1.7251764705882353, "grad_norm": 0.2825281537563074, "learning_rate": 0.0002825699722229858, "loss": 0.0693, "step": 3666 }, { "epoch": 1.7256470588235295, "grad_norm": 0.7098101392899883, "learning_rate": 0.0002825494571387351, "loss": 0.4395, "step": 3667 }, { "epoch": 1.7261176470588235, "grad_norm": 0.80088625480772, "learning_rate": 0.00028252893073411065, "loss": 0.4062, "step": 3668 }, { "epoch": 1.7265882352941175, "grad_norm": 0.7498166564626751, "learning_rate": 0.00028250839301086544, "loss": 0.4609, "step": 3669 }, { "epoch": 1.7270588235294118, "grad_norm": 0.3866062635386392, "learning_rate": 0.0002824878439707535, "loss": 0.1631, "step": 3670 }, { "epoch": 1.727529411764706, "grad_norm": 0.4824693669318062, "learning_rate": 0.0002824672836155299, "loss": 0.3086, "step": 3671 }, { "epoch": 1.728, "grad_norm": 0.6799010121342822, "learning_rate": 0.0002824467119469505, "loss": 0.3691, "step": 3672 }, { "epoch": 1.728470588235294, "grad_norm": 0.5707740201208708, "learning_rate": 0.0002824261289667722, "loss": 0.3594, "step": 3673 }, { "epoch": 1.7289411764705882, "grad_norm": 0.5725906263099982, "learning_rate": 0.000282405534676753, "loss": 0.5547, "step": 3674 }, { "epoch": 1.7294117647058824, "grad_norm": 0.6404409740407215, "learning_rate": 0.0002823849290786517, "loss": 0.2891, "step": 3675 }, { "epoch": 1.7298823529411764, "grad_norm": 0.5089366021715698, "learning_rate": 0.0002823643121742281, "loss": 0.252, "step": 3676 }, { "epoch": 1.7303529411764704, "grad_norm": 0.3833625205459714, "learning_rate": 0.000282343683965243, "loss": 0.1904, "step": 3677 }, { "epoch": 1.7308235294117646, "grad_norm": 0.34827595098934644, "learning_rate": 0.0002823230444534581, "loss": 0.1807, "step": 3678 }, { "epoch": 1.7312941176470589, "grad_norm": 0.10519999103000217, "learning_rate": 0.0002823023936406362, "loss": 0.0342, "step": 3679 }, { "epoch": 1.731764705882353, "grad_norm": 1.1842082522831838, "learning_rate": 0.0002822817315285409, "loss": 0.6797, "step": 3680 }, { "epoch": 1.732235294117647, "grad_norm": 0.6474490261231644, "learning_rate": 0.000282261058118937, "loss": 0.3535, "step": 3681 }, { "epoch": 1.732705882352941, "grad_norm": 0.4916722096145163, "learning_rate": 0.00028224037341358993, "loss": 0.1348, "step": 3682 }, { "epoch": 1.7331764705882353, "grad_norm": 0.5859496799640609, "learning_rate": 0.0002822196774142663, "loss": 0.3984, "step": 3683 }, { "epoch": 1.7336470588235295, "grad_norm": 0.7735728524665025, "learning_rate": 0.0002821989701227337, "loss": 0.5469, "step": 3684 }, { "epoch": 1.7341176470588235, "grad_norm": 0.6356960341101053, "learning_rate": 0.00028217825154076067, "loss": 0.3359, "step": 3685 }, { "epoch": 1.7345882352941175, "grad_norm": 0.36807070575076417, "learning_rate": 0.0002821575216701166, "loss": 0.1592, "step": 3686 }, { "epoch": 1.7350588235294118, "grad_norm": 0.2451000305678796, "learning_rate": 0.00028213678051257194, "loss": 0.0752, "step": 3687 }, { "epoch": 1.735529411764706, "grad_norm": 0.8530247271945357, "learning_rate": 0.00028211602806989814, "loss": 0.4492, "step": 3688 }, { "epoch": 1.736, "grad_norm": 0.8272854199355689, "learning_rate": 0.00028209526434386744, "loss": 0.4824, "step": 3689 }, { "epoch": 1.736470588235294, "grad_norm": 0.7562826044146214, "learning_rate": 0.0002820744893362533, "loss": 0.3906, "step": 3690 }, { "epoch": 1.7369411764705882, "grad_norm": 0.6244983221424016, "learning_rate": 0.0002820537030488299, "loss": 0.418, "step": 3691 }, { "epoch": 1.7374117647058824, "grad_norm": 0.4388614154477667, "learning_rate": 0.00028203290548337257, "loss": 0.1992, "step": 3692 }, { "epoch": 1.7378823529411764, "grad_norm": 0.3924718902689277, "learning_rate": 0.0002820120966416575, "loss": 0.127, "step": 3693 }, { "epoch": 1.7383529411764704, "grad_norm": 0.5449552475762766, "learning_rate": 0.0002819912765254619, "loss": 0.1699, "step": 3694 }, { "epoch": 1.7388235294117647, "grad_norm": 0.5880995113501354, "learning_rate": 0.0002819704451365638, "loss": 0.334, "step": 3695 }, { "epoch": 1.7392941176470589, "grad_norm": 0.9543992705492474, "learning_rate": 0.0002819496024767424, "loss": 0.2988, "step": 3696 }, { "epoch": 1.739764705882353, "grad_norm": 0.8206994334565839, "learning_rate": 0.0002819287485477778, "loss": 0.3984, "step": 3697 }, { "epoch": 1.740235294117647, "grad_norm": 0.542267982998744, "learning_rate": 0.0002819078833514509, "loss": 0.2734, "step": 3698 }, { "epoch": 1.740705882352941, "grad_norm": 0.5053372752994848, "learning_rate": 0.0002818870068895437, "loss": 0.1846, "step": 3699 }, { "epoch": 1.7411764705882353, "grad_norm": 0.8353209186074757, "learning_rate": 0.0002818661191638393, "loss": 0.5312, "step": 3700 }, { "epoch": 1.7416470588235295, "grad_norm": 1.1156747785506953, "learning_rate": 0.0002818452201761215, "loss": 0.6328, "step": 3701 }, { "epoch": 1.7421176470588235, "grad_norm": 0.7206577946348878, "learning_rate": 0.0002818243099281752, "loss": 0.5977, "step": 3702 }, { "epoch": 1.7425882352941175, "grad_norm": 0.3874016477225195, "learning_rate": 0.00028180338842178623, "loss": 0.1953, "step": 3703 }, { "epoch": 1.7430588235294118, "grad_norm": 0.8845062652624719, "learning_rate": 0.0002817824556587414, "loss": 0.4688, "step": 3704 }, { "epoch": 1.743529411764706, "grad_norm": 0.9209327892101524, "learning_rate": 0.0002817615116408285, "loss": 0.4785, "step": 3705 }, { "epoch": 1.744, "grad_norm": 0.6851552195813684, "learning_rate": 0.0002817405563698362, "loss": 0.291, "step": 3706 }, { "epoch": 1.744470588235294, "grad_norm": 1.2133838200412124, "learning_rate": 0.0002817195898475542, "loss": 0.4336, "step": 3707 }, { "epoch": 1.7449411764705882, "grad_norm": 0.6152416978535032, "learning_rate": 0.00028169861207577313, "loss": 0.4121, "step": 3708 }, { "epoch": 1.7454117647058824, "grad_norm": 0.29970928629824867, "learning_rate": 0.0002816776230562846, "loss": 0.1055, "step": 3709 }, { "epoch": 1.7458823529411764, "grad_norm": 0.584276586471385, "learning_rate": 0.0002816566227908812, "loss": 0.2373, "step": 3710 }, { "epoch": 1.7463529411764704, "grad_norm": 1.5999798651430077, "learning_rate": 0.00028163561128135645, "loss": 0.7695, "step": 3711 }, { "epoch": 1.7468235294117647, "grad_norm": 0.8030722065231918, "learning_rate": 0.0002816145885295048, "loss": 0.5312, "step": 3712 }, { "epoch": 1.7472941176470589, "grad_norm": 0.561977967048474, "learning_rate": 0.0002815935545371218, "loss": 0.3281, "step": 3713 }, { "epoch": 1.7477647058823529, "grad_norm": 0.4115768710974495, "learning_rate": 0.00028157250930600374, "loss": 0.2461, "step": 3714 }, { "epoch": 1.748235294117647, "grad_norm": 0.7050330864052464, "learning_rate": 0.00028155145283794797, "loss": 0.3223, "step": 3715 }, { "epoch": 1.748705882352941, "grad_norm": 0.939708508839632, "learning_rate": 0.0002815303851347529, "loss": 0.2295, "step": 3716 }, { "epoch": 1.7491764705882353, "grad_norm": 0.8518068565248567, "learning_rate": 0.0002815093061982178, "loss": 0.3418, "step": 3717 }, { "epoch": 1.7496470588235296, "grad_norm": 1.373920681417691, "learning_rate": 0.0002814882160301429, "loss": 0.7852, "step": 3718 }, { "epoch": 1.7501176470588236, "grad_norm": 0.7868976312816243, "learning_rate": 0.00028146711463232943, "loss": 0.4473, "step": 3719 }, { "epoch": 1.7505882352941176, "grad_norm": 0.3850252713636653, "learning_rate": 0.0002814460020065795, "loss": 0.1177, "step": 3720 }, { "epoch": 1.7510588235294118, "grad_norm": 0.49534131725987934, "learning_rate": 0.00028142487815469627, "loss": 0.2129, "step": 3721 }, { "epoch": 1.751529411764706, "grad_norm": 2.0007698984716225, "learning_rate": 0.00028140374307848384, "loss": 0.9258, "step": 3722 }, { "epoch": 1.752, "grad_norm": 0.6784368260354362, "learning_rate": 0.00028138259677974714, "loss": 0.3262, "step": 3723 }, { "epoch": 1.752470588235294, "grad_norm": 0.5969341217727377, "learning_rate": 0.0002813614392602923, "loss": 0.3516, "step": 3724 }, { "epoch": 1.7529411764705882, "grad_norm": 0.3510223414596249, "learning_rate": 0.00028134027052192624, "loss": 0.1621, "step": 3725 }, { "epoch": 1.7534117647058824, "grad_norm": 1.1895312335977695, "learning_rate": 0.0002813190905664568, "loss": 0.5391, "step": 3726 }, { "epoch": 1.7538823529411764, "grad_norm": 0.461450568289904, "learning_rate": 0.0002812978993956929, "loss": 0.1309, "step": 3727 }, { "epoch": 1.7543529411764704, "grad_norm": 0.8620329321452963, "learning_rate": 0.0002812766970114445, "loss": 0.2402, "step": 3728 }, { "epoch": 1.7548235294117647, "grad_norm": 0.8448374346477764, "learning_rate": 0.00028125548341552213, "loss": 0.4316, "step": 3729 }, { "epoch": 1.755294117647059, "grad_norm": 0.39423074396657193, "learning_rate": 0.00028123425860973773, "loss": 0.2217, "step": 3730 }, { "epoch": 1.755764705882353, "grad_norm": 0.5099307563936175, "learning_rate": 0.000281213022595904, "loss": 0.3789, "step": 3731 }, { "epoch": 1.756235294117647, "grad_norm": 1.2995099138152366, "learning_rate": 0.0002811917753758345, "loss": 0.8945, "step": 3732 }, { "epoch": 1.7567058823529411, "grad_norm": 0.6519454154168242, "learning_rate": 0.00028117051695134384, "loss": 0.4668, "step": 3733 }, { "epoch": 1.7571764705882353, "grad_norm": 0.6448373306110493, "learning_rate": 0.0002811492473242477, "loss": 0.2832, "step": 3734 }, { "epoch": 1.7576470588235296, "grad_norm": 0.6469192514348066, "learning_rate": 0.00028112796649636254, "loss": 0.459, "step": 3735 }, { "epoch": 1.7581176470588236, "grad_norm": 0.34110276188292366, "learning_rate": 0.0002811066744695059, "loss": 0.1118, "step": 3736 }, { "epoch": 1.7585882352941176, "grad_norm": 0.9862541332823129, "learning_rate": 0.0002810853712454961, "loss": 0.4062, "step": 3737 }, { "epoch": 1.7590588235294118, "grad_norm": 0.4626564287936353, "learning_rate": 0.00028106405682615265, "loss": 0.2773, "step": 3738 }, { "epoch": 1.759529411764706, "grad_norm": 0.7869755982620595, "learning_rate": 0.0002810427312132959, "loss": 0.4141, "step": 3739 }, { "epoch": 1.76, "grad_norm": 0.5825038451434952, "learning_rate": 0.00028102139440874715, "loss": 0.3477, "step": 3740 }, { "epoch": 1.760470588235294, "grad_norm": 0.375075754663659, "learning_rate": 0.0002810000464143287, "loss": 0.2754, "step": 3741 }, { "epoch": 1.7609411764705882, "grad_norm": 0.9419357126853739, "learning_rate": 0.0002809786872318637, "loss": 0.3555, "step": 3742 }, { "epoch": 1.7614117647058825, "grad_norm": 0.6342115596013616, "learning_rate": 0.00028095731686317635, "loss": 0.3594, "step": 3743 }, { "epoch": 1.7618823529411765, "grad_norm": 0.258659468361624, "learning_rate": 0.0002809359353100918, "loss": 0.0869, "step": 3744 }, { "epoch": 1.7623529411764705, "grad_norm": 0.2066399727138459, "learning_rate": 0.0002809145425744361, "loss": 0.1016, "step": 3745 }, { "epoch": 1.7628235294117647, "grad_norm": 1.1130282151441648, "learning_rate": 0.00028089313865803644, "loss": 0.5195, "step": 3746 }, { "epoch": 1.763294117647059, "grad_norm": 0.792810929391764, "learning_rate": 0.0002808717235627206, "loss": 0.4844, "step": 3747 }, { "epoch": 1.763764705882353, "grad_norm": 0.8299795963901603, "learning_rate": 0.0002808502972903177, "loss": 0.4609, "step": 3748 }, { "epoch": 1.764235294117647, "grad_norm": 0.6482650488614158, "learning_rate": 0.0002808288598426576, "loss": 0.3496, "step": 3749 }, { "epoch": 1.7647058823529411, "grad_norm": 0.4230220619277034, "learning_rate": 0.0002808074112215711, "loss": 0.2471, "step": 3750 }, { "epoch": 1.7651764705882353, "grad_norm": 1.3599583016561867, "learning_rate": 0.00028078595142889013, "loss": 0.625, "step": 3751 }, { "epoch": 1.7656470588235296, "grad_norm": 0.6703669392851356, "learning_rate": 0.0002807644804664474, "loss": 0.25, "step": 3752 }, { "epoch": 1.7661176470588236, "grad_norm": 0.5052287984840137, "learning_rate": 0.0002807429983360766, "loss": 0.2812, "step": 3753 }, { "epoch": 1.7665882352941176, "grad_norm": 0.4242097382664636, "learning_rate": 0.0002807215050396125, "loss": 0.248, "step": 3754 }, { "epoch": 1.7670588235294118, "grad_norm": 0.48960070143126894, "learning_rate": 0.0002807000005788907, "loss": 0.3516, "step": 3755 }, { "epoch": 1.767529411764706, "grad_norm": 0.8341019374284204, "learning_rate": 0.00028067848495574773, "loss": 0.6328, "step": 3756 }, { "epoch": 1.768, "grad_norm": 0.8276728556166796, "learning_rate": 0.0002806569581720212, "loss": 0.4785, "step": 3757 }, { "epoch": 1.768470588235294, "grad_norm": 1.060789200913506, "learning_rate": 0.00028063542022954956, "loss": 0.668, "step": 3758 }, { "epoch": 1.7689411764705882, "grad_norm": 0.6817602141314077, "learning_rate": 0.00028061387113017226, "loss": 0.2891, "step": 3759 }, { "epoch": 1.7694117647058825, "grad_norm": 0.570817393528066, "learning_rate": 0.0002805923108757297, "loss": 0.3906, "step": 3760 }, { "epoch": 1.7698823529411765, "grad_norm": 0.2902548216803911, "learning_rate": 0.00028057073946806323, "loss": 0.1084, "step": 3761 }, { "epoch": 1.7703529411764705, "grad_norm": 0.9227551894351123, "learning_rate": 0.00028054915690901524, "loss": 0.2891, "step": 3762 }, { "epoch": 1.7708235294117647, "grad_norm": 0.3908195929691962, "learning_rate": 0.0002805275632004288, "loss": 0.0918, "step": 3763 }, { "epoch": 1.771294117647059, "grad_norm": 0.5984366424326687, "learning_rate": 0.00028050595834414835, "loss": 0.3418, "step": 3764 }, { "epoch": 1.771764705882353, "grad_norm": 0.35709547262772806, "learning_rate": 0.00028048434234201883, "loss": 0.106, "step": 3765 }, { "epoch": 1.772235294117647, "grad_norm": 0.2971844188192754, "learning_rate": 0.00028046271519588644, "loss": 0.1138, "step": 3766 }, { "epoch": 1.7727058823529411, "grad_norm": 0.7332544818544294, "learning_rate": 0.0002804410769075984, "loss": 0.4473, "step": 3767 }, { "epoch": 1.7731764705882354, "grad_norm": 0.2510192851295027, "learning_rate": 0.0002804194274790024, "loss": 0.082, "step": 3768 }, { "epoch": 1.7736470588235296, "grad_norm": 0.5141611277491169, "learning_rate": 0.00028039776691194773, "loss": 0.2754, "step": 3769 }, { "epoch": 1.7741176470588236, "grad_norm": 0.36935793777244835, "learning_rate": 0.0002803760952082841, "loss": 0.2422, "step": 3770 }, { "epoch": 1.7745882352941176, "grad_norm": 0.4716903544093982, "learning_rate": 0.0002803544123698624, "loss": 0.4102, "step": 3771 }, { "epoch": 1.7750588235294118, "grad_norm": 0.47237004526910326, "learning_rate": 0.0002803327183985346, "loss": 0.1846, "step": 3772 }, { "epoch": 1.775529411764706, "grad_norm": 0.4681015624532315, "learning_rate": 0.00028031101329615334, "loss": 0.1533, "step": 3773 }, { "epoch": 1.776, "grad_norm": 0.5902998698516534, "learning_rate": 0.00028028929706457235, "loss": 0.3555, "step": 3774 }, { "epoch": 1.776470588235294, "grad_norm": 0.2831267224919553, "learning_rate": 0.0002802675697056463, "loss": 0.0698, "step": 3775 }, { "epoch": 1.7769411764705882, "grad_norm": 0.4804262367145264, "learning_rate": 0.00028024583122123094, "loss": 0.2383, "step": 3776 }, { "epoch": 1.7774117647058825, "grad_norm": 0.9524363479261132, "learning_rate": 0.00028022408161318266, "loss": 0.4863, "step": 3777 }, { "epoch": 1.7778823529411765, "grad_norm": 0.5352327321497511, "learning_rate": 0.00028020232088335907, "loss": 0.3184, "step": 3778 }, { "epoch": 1.7783529411764705, "grad_norm": 1.0829575617989042, "learning_rate": 0.00028018054903361864, "loss": 0.6406, "step": 3779 }, { "epoch": 1.7788235294117647, "grad_norm": 0.30714526824048094, "learning_rate": 0.00028015876606582083, "loss": 0.1826, "step": 3780 }, { "epoch": 1.779294117647059, "grad_norm": 0.9273670961051127, "learning_rate": 0.0002801369719818259, "loss": 0.4844, "step": 3781 }, { "epoch": 1.779764705882353, "grad_norm": 0.10419715455087332, "learning_rate": 0.0002801151667834953, "loss": 0.0194, "step": 3782 }, { "epoch": 1.780235294117647, "grad_norm": 0.48191431994901396, "learning_rate": 0.00028009335047269126, "loss": 0.2197, "step": 3783 }, { "epoch": 1.7807058823529411, "grad_norm": 0.5346547752949506, "learning_rate": 0.00028007152305127694, "loss": 0.209, "step": 3784 }, { "epoch": 1.7811764705882354, "grad_norm": 0.2978978097551027, "learning_rate": 0.0002800496845211166, "loss": 0.1157, "step": 3785 }, { "epoch": 1.7816470588235294, "grad_norm": 1.0077886976505643, "learning_rate": 0.0002800278348840753, "loss": 0.5586, "step": 3786 }, { "epoch": 1.7821176470588236, "grad_norm": 0.3606762798948609, "learning_rate": 0.00028000597414201906, "loss": 0.1611, "step": 3787 }, { "epoch": 1.7825882352941176, "grad_norm": 1.3126427413748611, "learning_rate": 0.000279984102296815, "loss": 0.3984, "step": 3788 }, { "epoch": 1.7830588235294118, "grad_norm": 0.9970885455599817, "learning_rate": 0.000279962219350331, "loss": 0.3789, "step": 3789 }, { "epoch": 1.783529411764706, "grad_norm": 1.4854100274656492, "learning_rate": 0.00027994032530443606, "loss": 0.5352, "step": 3790 }, { "epoch": 1.784, "grad_norm": 0.5265899100957072, "learning_rate": 0.00027991842016099994, "loss": 0.2471, "step": 3791 }, { "epoch": 1.784470588235294, "grad_norm": 0.29404079832454805, "learning_rate": 0.0002798965039218936, "loss": 0.126, "step": 3792 }, { "epoch": 1.7849411764705883, "grad_norm": 0.5771235389476416, "learning_rate": 0.0002798745765889886, "loss": 0.2207, "step": 3793 }, { "epoch": 1.7854117647058825, "grad_norm": 0.39951086431737026, "learning_rate": 0.00027985263816415767, "loss": 0.168, "step": 3794 }, { "epoch": 1.7858823529411765, "grad_norm": 0.08210848589985972, "learning_rate": 0.0002798306886492746, "loss": 0.0109, "step": 3795 }, { "epoch": 1.7863529411764705, "grad_norm": 0.9452916506372964, "learning_rate": 0.00027980872804621385, "loss": 0.6641, "step": 3796 }, { "epoch": 1.7868235294117647, "grad_norm": 0.6657350668306088, "learning_rate": 0.0002797867563568511, "loss": 0.2246, "step": 3797 }, { "epoch": 1.787294117647059, "grad_norm": 0.9582360222364475, "learning_rate": 0.0002797647735830627, "loss": 0.6602, "step": 3798 }, { "epoch": 1.787764705882353, "grad_norm": 0.1875830222338957, "learning_rate": 0.00027974277972672614, "loss": 0.0579, "step": 3799 }, { "epoch": 1.788235294117647, "grad_norm": 0.6578865420460747, "learning_rate": 0.0002797207747897198, "loss": 0.3809, "step": 3800 }, { "epoch": 1.7887058823529411, "grad_norm": 0.28660386497576246, "learning_rate": 0.000279698758773923, "loss": 0.0791, "step": 3801 }, { "epoch": 1.7891764705882354, "grad_norm": 0.8107139812430298, "learning_rate": 0.0002796767316812161, "loss": 0.3438, "step": 3802 }, { "epoch": 1.7896470588235294, "grad_norm": 0.5140894027618798, "learning_rate": 0.0002796546935134802, "loss": 0.1543, "step": 3803 }, { "epoch": 1.7901176470588234, "grad_norm": 0.6173093329477303, "learning_rate": 0.00027963264427259755, "loss": 0.3223, "step": 3804 }, { "epoch": 1.7905882352941176, "grad_norm": 0.7800050650529952, "learning_rate": 0.00027961058396045124, "loss": 0.3828, "step": 3805 }, { "epoch": 1.7910588235294118, "grad_norm": 0.2796481006457584, "learning_rate": 0.00027958851257892526, "loss": 0.085, "step": 3806 }, { "epoch": 1.791529411764706, "grad_norm": 0.2099224971059292, "learning_rate": 0.0002795664301299047, "loss": 0.0713, "step": 3807 }, { "epoch": 1.792, "grad_norm": 0.6103416590317726, "learning_rate": 0.00027954433661527554, "loss": 0.3145, "step": 3808 }, { "epoch": 1.792470588235294, "grad_norm": 0.942398330759289, "learning_rate": 0.0002795222320369246, "loss": 0.5078, "step": 3809 }, { "epoch": 1.7929411764705883, "grad_norm": 0.3043911990979957, "learning_rate": 0.00027950011639673976, "loss": 0.0742, "step": 3810 }, { "epoch": 1.7934117647058825, "grad_norm": 0.18635418515980973, "learning_rate": 0.00027947798969660975, "loss": 0.0623, "step": 3811 }, { "epoch": 1.7938823529411765, "grad_norm": 0.7558854336711831, "learning_rate": 0.00027945585193842433, "loss": 0.4531, "step": 3812 }, { "epoch": 1.7943529411764705, "grad_norm": 0.41228237546848995, "learning_rate": 0.00027943370312407415, "loss": 0.0908, "step": 3813 }, { "epoch": 1.7948235294117647, "grad_norm": 0.5623084362873914, "learning_rate": 0.00027941154325545096, "loss": 0.4082, "step": 3814 }, { "epoch": 1.795294117647059, "grad_norm": 0.8546537122741202, "learning_rate": 0.00027938937233444713, "loss": 0.4707, "step": 3815 }, { "epoch": 1.795764705882353, "grad_norm": 0.4116203681423941, "learning_rate": 0.00027936719036295623, "loss": 0.1943, "step": 3816 }, { "epoch": 1.796235294117647, "grad_norm": 0.7451663326622401, "learning_rate": 0.0002793449973428728, "loss": 0.3809, "step": 3817 }, { "epoch": 1.7967058823529412, "grad_norm": 0.18681499224537382, "learning_rate": 0.0002793227932760921, "loss": 0.0327, "step": 3818 }, { "epoch": 1.7971764705882354, "grad_norm": 0.6375603841301485, "learning_rate": 0.0002793005781645106, "loss": 0.3867, "step": 3819 }, { "epoch": 1.7976470588235294, "grad_norm": 0.6724668716328301, "learning_rate": 0.00027927835201002545, "loss": 0.3613, "step": 3820 }, { "epoch": 1.7981176470588234, "grad_norm": 0.973642975211751, "learning_rate": 0.00027925611481453497, "loss": 0.5273, "step": 3821 }, { "epoch": 1.7985882352941176, "grad_norm": 0.9732881670467017, "learning_rate": 0.0002792338665799383, "loss": 0.5742, "step": 3822 }, { "epoch": 1.7990588235294118, "grad_norm": 0.4642238452478658, "learning_rate": 0.00027921160730813544, "loss": 0.1484, "step": 3823 }, { "epoch": 1.799529411764706, "grad_norm": 1.263900920396467, "learning_rate": 0.0002791893370010276, "loss": 0.7734, "step": 3824 }, { "epoch": 1.8, "grad_norm": 0.9767915291890125, "learning_rate": 0.00027916705566051676, "loss": 0.4141, "step": 3825 }, { "epoch": 1.800470588235294, "grad_norm": 0.3826163379876425, "learning_rate": 0.00027914476328850575, "loss": 0.2256, "step": 3826 }, { "epoch": 1.8009411764705883, "grad_norm": 0.5483122924766313, "learning_rate": 0.0002791224598868985, "loss": 0.1924, "step": 3827 }, { "epoch": 1.8014117647058825, "grad_norm": 1.1812899788144802, "learning_rate": 0.00027910014545759986, "loss": 0.707, "step": 3828 }, { "epoch": 1.8018823529411765, "grad_norm": 0.8081902446062275, "learning_rate": 0.00027907782000251557, "loss": 0.4902, "step": 3829 }, { "epoch": 1.8023529411764705, "grad_norm": 0.28828393673465547, "learning_rate": 0.00027905548352355226, "loss": 0.1328, "step": 3830 }, { "epoch": 1.8028235294117647, "grad_norm": 0.5350272849719232, "learning_rate": 0.00027903313602261774, "loss": 0.2539, "step": 3831 }, { "epoch": 1.803294117647059, "grad_norm": 1.780707421144444, "learning_rate": 0.00027901077750162043, "loss": 0.668, "step": 3832 }, { "epoch": 1.803764705882353, "grad_norm": 0.863855163274048, "learning_rate": 0.00027898840796246994, "loss": 0.3926, "step": 3833 }, { "epoch": 1.804235294117647, "grad_norm": 0.39802921918128736, "learning_rate": 0.00027896602740707673, "loss": 0.1855, "step": 3834 }, { "epoch": 1.8047058823529412, "grad_norm": 0.6472915208229849, "learning_rate": 0.0002789436358373522, "loss": 0.4199, "step": 3835 }, { "epoch": 1.8051764705882354, "grad_norm": 0.408016899896795, "learning_rate": 0.0002789212332552088, "loss": 0.1729, "step": 3836 }, { "epoch": 1.8056470588235294, "grad_norm": 1.0726630896117626, "learning_rate": 0.0002788988196625596, "loss": 0.5078, "step": 3837 }, { "epoch": 1.8061176470588234, "grad_norm": 1.337148284179441, "learning_rate": 0.000278876395061319, "loss": 0.4941, "step": 3838 }, { "epoch": 1.8065882352941176, "grad_norm": 1.0646145473453266, "learning_rate": 0.0002788539594534021, "loss": 0.5469, "step": 3839 }, { "epoch": 1.8070588235294118, "grad_norm": 0.315675122832031, "learning_rate": 0.0002788315128407251, "loss": 0.1064, "step": 3840 }, { "epoch": 1.8075294117647058, "grad_norm": 1.146558498525613, "learning_rate": 0.0002788090552252049, "loss": 0.6562, "step": 3841 }, { "epoch": 1.808, "grad_norm": 0.9062567019381595, "learning_rate": 0.00027878658660875967, "loss": 0.4648, "step": 3842 }, { "epoch": 1.808470588235294, "grad_norm": 1.2051619448188045, "learning_rate": 0.0002787641069933082, "loss": 0.5469, "step": 3843 }, { "epoch": 1.8089411764705883, "grad_norm": 1.2415609771749416, "learning_rate": 0.0002787416163807704, "loss": 0.6641, "step": 3844 }, { "epoch": 1.8094117647058825, "grad_norm": 0.565947518263322, "learning_rate": 0.0002787191147730671, "loss": 0.4277, "step": 3845 }, { "epoch": 1.8098823529411765, "grad_norm": 0.47164186804827446, "learning_rate": 0.00027869660217212005, "loss": 0.1348, "step": 3846 }, { "epoch": 1.8103529411764705, "grad_norm": 0.4876851604359634, "learning_rate": 0.0002786740785798519, "loss": 0.2275, "step": 3847 }, { "epoch": 1.8108235294117647, "grad_norm": 0.47243496588010725, "learning_rate": 0.00027865154399818627, "loss": 0.1436, "step": 3848 }, { "epoch": 1.811294117647059, "grad_norm": 1.1580826305580372, "learning_rate": 0.00027862899842904783, "loss": 0.3828, "step": 3849 }, { "epoch": 1.811764705882353, "grad_norm": 1.3910992293785842, "learning_rate": 0.00027860644187436195, "loss": 0.4297, "step": 3850 }, { "epoch": 1.812235294117647, "grad_norm": 0.9243904337748972, "learning_rate": 0.0002785838743360551, "loss": 0.4082, "step": 3851 }, { "epoch": 1.8127058823529412, "grad_norm": 0.30059581753687653, "learning_rate": 0.00027856129581605474, "loss": 0.1143, "step": 3852 }, { "epoch": 1.8131764705882354, "grad_norm": 0.42859173654692206, "learning_rate": 0.0002785387063162891, "loss": 0.1807, "step": 3853 }, { "epoch": 1.8136470588235294, "grad_norm": 0.3232968503109475, "learning_rate": 0.00027851610583868745, "loss": 0.1494, "step": 3854 }, { "epoch": 1.8141176470588234, "grad_norm": 1.2142756682274969, "learning_rate": 0.00027849349438518003, "loss": 0.3984, "step": 3855 }, { "epoch": 1.8145882352941176, "grad_norm": 0.7012954027152409, "learning_rate": 0.0002784708719576979, "loss": 0.2119, "step": 3856 }, { "epoch": 1.8150588235294118, "grad_norm": 1.0608114826127804, "learning_rate": 0.00027844823855817315, "loss": 0.4941, "step": 3857 }, { "epoch": 1.8155294117647058, "grad_norm": 0.14268916529298908, "learning_rate": 0.0002784255941885388, "loss": 0.0718, "step": 3858 }, { "epoch": 1.8159999999999998, "grad_norm": 1.0642526204134113, "learning_rate": 0.0002784029388507288, "loss": 0.6211, "step": 3859 }, { "epoch": 1.816470588235294, "grad_norm": 0.5425379747374066, "learning_rate": 0.000278380272546678, "loss": 0.3262, "step": 3860 }, { "epoch": 1.8169411764705883, "grad_norm": 0.8120126117324393, "learning_rate": 0.00027835759527832226, "loss": 0.4941, "step": 3861 }, { "epoch": 1.8174117647058825, "grad_norm": 0.27635391951280186, "learning_rate": 0.0002783349070475982, "loss": 0.1279, "step": 3862 }, { "epoch": 1.8178823529411765, "grad_norm": 0.8414721343107857, "learning_rate": 0.00027831220785644366, "loss": 0.3652, "step": 3863 }, { "epoch": 1.8183529411764705, "grad_norm": 0.37975819262138466, "learning_rate": 0.0002782894977067972, "loss": 0.1514, "step": 3864 }, { "epoch": 1.8188235294117647, "grad_norm": 0.44359730239726886, "learning_rate": 0.0002782667766005984, "loss": 0.1406, "step": 3865 }, { "epoch": 1.819294117647059, "grad_norm": 0.7950024731693178, "learning_rate": 0.0002782440445397877, "loss": 0.5742, "step": 3866 }, { "epoch": 1.819764705882353, "grad_norm": 1.1578194618946773, "learning_rate": 0.00027822130152630657, "loss": 0.5508, "step": 3867 }, { "epoch": 1.820235294117647, "grad_norm": 0.5550369384727578, "learning_rate": 0.00027819854756209734, "loss": 0.2812, "step": 3868 }, { "epoch": 1.8207058823529412, "grad_norm": 0.4268442941131041, "learning_rate": 0.00027817578264910343, "loss": 0.2344, "step": 3869 }, { "epoch": 1.8211764705882354, "grad_norm": 0.4340440718107654, "learning_rate": 0.00027815300678926886, "loss": 0.2354, "step": 3870 }, { "epoch": 1.8216470588235294, "grad_norm": 0.3911587781576124, "learning_rate": 0.000278130219984539, "loss": 0.2793, "step": 3871 }, { "epoch": 1.8221176470588234, "grad_norm": 1.1270188382911863, "learning_rate": 0.00027810742223685983, "loss": 0.5977, "step": 3872 }, { "epoch": 1.8225882352941176, "grad_norm": 0.42816928973850554, "learning_rate": 0.00027808461354817847, "loss": 0.207, "step": 3873 }, { "epoch": 1.8230588235294118, "grad_norm": 0.31322198031657794, "learning_rate": 0.00027806179392044286, "loss": 0.1221, "step": 3874 }, { "epoch": 1.8235294117647058, "grad_norm": 0.1515126221296573, "learning_rate": 0.0002780389633556019, "loss": 0.0325, "step": 3875 }, { "epoch": 1.8239999999999998, "grad_norm": 0.7663652585382263, "learning_rate": 0.00027801612185560554, "loss": 0.4414, "step": 3876 }, { "epoch": 1.824470588235294, "grad_norm": 0.6757405865636325, "learning_rate": 0.00027799326942240434, "loss": 0.3418, "step": 3877 }, { "epoch": 1.8249411764705883, "grad_norm": 0.46473628086985885, "learning_rate": 0.0002779704060579501, "loss": 0.2578, "step": 3878 }, { "epoch": 1.8254117647058825, "grad_norm": 0.5986711502170833, "learning_rate": 0.00027794753176419557, "loss": 0.3125, "step": 3879 }, { "epoch": 1.8258823529411765, "grad_norm": 0.7679115609480975, "learning_rate": 0.00027792464654309427, "loss": 0.707, "step": 3880 }, { "epoch": 1.8263529411764705, "grad_norm": 0.38059942655977486, "learning_rate": 0.0002779017503966006, "loss": 0.2539, "step": 3881 }, { "epoch": 1.8268235294117647, "grad_norm": 0.27559999754376696, "learning_rate": 0.0002778788433266702, "loss": 0.1191, "step": 3882 }, { "epoch": 1.827294117647059, "grad_norm": 0.39891970545763844, "learning_rate": 0.00027785592533525924, "loss": 0.1982, "step": 3883 }, { "epoch": 1.827764705882353, "grad_norm": 1.2310783076848177, "learning_rate": 0.00027783299642432515, "loss": 0.707, "step": 3884 }, { "epoch": 1.828235294117647, "grad_norm": 0.3507621808411768, "learning_rate": 0.0002778100565958262, "loss": 0.1328, "step": 3885 }, { "epoch": 1.8287058823529412, "grad_norm": 0.5700348514272032, "learning_rate": 0.0002777871058517215, "loss": 0.3965, "step": 3886 }, { "epoch": 1.8291764705882354, "grad_norm": 0.29562752931115344, "learning_rate": 0.00027776414419397106, "loss": 0.1279, "step": 3887 }, { "epoch": 1.8296470588235294, "grad_norm": 0.6557209628415375, "learning_rate": 0.00027774117162453617, "loss": 0.5898, "step": 3888 }, { "epoch": 1.8301176470588234, "grad_norm": 0.690936400888537, "learning_rate": 0.00027771818814537855, "loss": 0.3359, "step": 3889 }, { "epoch": 1.8305882352941176, "grad_norm": 0.16743877109748556, "learning_rate": 0.00027769519375846126, "loss": 0.0776, "step": 3890 }, { "epoch": 1.8310588235294118, "grad_norm": 0.5720619739405142, "learning_rate": 0.0002776721884657481, "loss": 0.3789, "step": 3891 }, { "epoch": 1.8315294117647059, "grad_norm": 0.8124176231144937, "learning_rate": 0.00027764917226920377, "loss": 0.4551, "step": 3892 }, { "epoch": 1.8319999999999999, "grad_norm": 0.5860239571326551, "learning_rate": 0.000277626145170794, "loss": 0.3555, "step": 3893 }, { "epoch": 1.832470588235294, "grad_norm": 0.31519841070035803, "learning_rate": 0.0002776031071724855, "loss": 0.1128, "step": 3894 }, { "epoch": 1.8329411764705883, "grad_norm": 1.118607397347264, "learning_rate": 0.0002775800582762457, "loss": 0.625, "step": 3895 }, { "epoch": 1.8334117647058825, "grad_norm": 4.949043724392093e-06, "learning_rate": 0.00027755699848404315, "loss": 0.0, "step": 3896 }, { "epoch": 1.8338823529411765, "grad_norm": 0.8522537444271392, "learning_rate": 0.0002775339277978473, "loss": 0.4434, "step": 3897 }, { "epoch": 1.8343529411764705, "grad_norm": 0.7495366085170441, "learning_rate": 0.0002775108462196284, "loss": 0.3867, "step": 3898 }, { "epoch": 1.8348235294117647, "grad_norm": 0.3383640439612085, "learning_rate": 0.0002774877537513579, "loss": 0.0967, "step": 3899 }, { "epoch": 1.835294117647059, "grad_norm": 0.21584614290258927, "learning_rate": 0.0002774646503950078, "loss": 0.0723, "step": 3900 }, { "epoch": 1.835764705882353, "grad_norm": 0.4959078239820497, "learning_rate": 0.0002774415361525514, "loss": 0.252, "step": 3901 }, { "epoch": 1.836235294117647, "grad_norm": 0.8900952049641199, "learning_rate": 0.00027741841102596276, "loss": 0.3398, "step": 3902 }, { "epoch": 1.8367058823529412, "grad_norm": 1.583755502432744, "learning_rate": 0.0002773952750172168, "loss": 0.7305, "step": 3903 }, { "epoch": 1.8371764705882354, "grad_norm": 0.8398385716699854, "learning_rate": 0.0002773721281282894, "loss": 0.4609, "step": 3904 }, { "epoch": 1.8376470588235294, "grad_norm": 0.8653296150152197, "learning_rate": 0.0002773489703611576, "loss": 0.3359, "step": 3905 }, { "epoch": 1.8381176470588234, "grad_norm": 1.1287066295912782, "learning_rate": 0.00027732580171779906, "loss": 0.4375, "step": 3906 }, { "epoch": 1.8385882352941176, "grad_norm": 1.0852053547257674, "learning_rate": 0.0002773026222001925, "loss": 0.3633, "step": 3907 }, { "epoch": 1.8390588235294119, "grad_norm": 0.6232627457532051, "learning_rate": 0.00027727943181031766, "loss": 0.459, "step": 3908 }, { "epoch": 1.8395294117647059, "grad_norm": 0.8660864288788546, "learning_rate": 0.00027725623055015495, "loss": 0.4863, "step": 3909 }, { "epoch": 1.8399999999999999, "grad_norm": 0.3256486514448379, "learning_rate": 0.000277233018421686, "loss": 0.1689, "step": 3910 }, { "epoch": 1.840470588235294, "grad_norm": 0.7640262106178614, "learning_rate": 0.00027720979542689324, "loss": 0.25, "step": 3911 }, { "epoch": 1.8409411764705883, "grad_norm": 0.7023536552536782, "learning_rate": 0.0002771865615677599, "loss": 0.3789, "step": 3912 }, { "epoch": 1.8414117647058823, "grad_norm": 1.0214619031823382, "learning_rate": 0.00027716331684627044, "loss": 0.4961, "step": 3913 }, { "epoch": 1.8418823529411765, "grad_norm": 0.3651440109432169, "learning_rate": 0.00027714006126441, "loss": 0.104, "step": 3914 }, { "epoch": 1.8423529411764705, "grad_norm": 0.8612389414898728, "learning_rate": 0.00027711679482416465, "loss": 0.5195, "step": 3915 }, { "epoch": 1.8428235294117647, "grad_norm": 0.2844917558164128, "learning_rate": 0.0002770935175275215, "loss": 0.1533, "step": 3916 }, { "epoch": 1.843294117647059, "grad_norm": 0.7834870406147976, "learning_rate": 0.0002770702293764686, "loss": 0.373, "step": 3917 }, { "epoch": 1.843764705882353, "grad_norm": 0.4491162784244002, "learning_rate": 0.0002770469303729949, "loss": 0.1699, "step": 3918 }, { "epoch": 1.844235294117647, "grad_norm": 0.7431910230914255, "learning_rate": 0.0002770236205190901, "loss": 0.4297, "step": 3919 }, { "epoch": 1.8447058823529412, "grad_norm": 0.6143871903263908, "learning_rate": 0.00027700029981674506, "loss": 0.3301, "step": 3920 }, { "epoch": 1.8451764705882354, "grad_norm": 0.5935344385899067, "learning_rate": 0.0002769769682679515, "loss": 0.3691, "step": 3921 }, { "epoch": 1.8456470588235294, "grad_norm": 0.3040746967507643, "learning_rate": 0.00027695362587470203, "loss": 0.1543, "step": 3922 }, { "epoch": 1.8461176470588234, "grad_norm": 0.7285716335513348, "learning_rate": 0.0002769302726389902, "loss": 0.4473, "step": 3923 }, { "epoch": 1.8465882352941176, "grad_norm": 0.45035418277755523, "learning_rate": 0.00027690690856281054, "loss": 0.3711, "step": 3924 }, { "epoch": 1.8470588235294119, "grad_norm": 0.524462416178943, "learning_rate": 0.00027688353364815834, "loss": 0.3809, "step": 3925 }, { "epoch": 1.8475294117647059, "grad_norm": 0.9411082887036368, "learning_rate": 0.0002768601478970301, "loss": 0.8008, "step": 3926 }, { "epoch": 1.8479999999999999, "grad_norm": 0.8661813985168554, "learning_rate": 0.00027683675131142294, "loss": 0.5078, "step": 3927 }, { "epoch": 1.848470588235294, "grad_norm": 0.4355892750271831, "learning_rate": 0.000276813343893335, "loss": 0.1357, "step": 3928 }, { "epoch": 1.8489411764705883, "grad_norm": 0.6815835600934493, "learning_rate": 0.0002767899256447656, "loss": 0.3574, "step": 3929 }, { "epoch": 1.8494117647058823, "grad_norm": 1.1940000545705531, "learning_rate": 0.0002767664965677146, "loss": 0.7656, "step": 3930 }, { "epoch": 1.8498823529411763, "grad_norm": 0.5895409025889713, "learning_rate": 0.000276743056664183, "loss": 0.3828, "step": 3931 }, { "epoch": 1.8503529411764705, "grad_norm": 0.824423117288264, "learning_rate": 0.0002767196059361727, "loss": 0.3926, "step": 3932 }, { "epoch": 1.8508235294117648, "grad_norm": 1.0991894281169263, "learning_rate": 0.0002766961443856865, "loss": 0.3164, "step": 3933 }, { "epoch": 1.851294117647059, "grad_norm": 0.3550648628797242, "learning_rate": 0.00027667267201472815, "loss": 0.1357, "step": 3934 }, { "epoch": 1.851764705882353, "grad_norm": 0.20063338405800085, "learning_rate": 0.00027664918882530225, "loss": 0.0884, "step": 3935 }, { "epoch": 1.852235294117647, "grad_norm": 0.7766923230200695, "learning_rate": 0.0002766256948194144, "loss": 0.4766, "step": 3936 }, { "epoch": 1.8527058823529412, "grad_norm": 0.6876931108996382, "learning_rate": 0.0002766021899990712, "loss": 0.4629, "step": 3937 }, { "epoch": 1.8531764705882354, "grad_norm": 0.8157229687834556, "learning_rate": 0.0002765786743662799, "loss": 0.5742, "step": 3938 }, { "epoch": 1.8536470588235294, "grad_norm": 0.6511930234838821, "learning_rate": 0.0002765551479230489, "loss": 0.3848, "step": 3939 }, { "epoch": 1.8541176470588234, "grad_norm": 0.5928261519847694, "learning_rate": 0.00027653161067138764, "loss": 0.4199, "step": 3940 }, { "epoch": 1.8545882352941176, "grad_norm": 0.44410582414662475, "learning_rate": 0.00027650806261330616, "loss": 0.1738, "step": 3941 }, { "epoch": 1.8550588235294119, "grad_norm": 0.33247772629916905, "learning_rate": 0.0002764845037508156, "loss": 0.127, "step": 3942 }, { "epoch": 1.8555294117647059, "grad_norm": 0.6984146750528241, "learning_rate": 0.00027646093408592804, "loss": 0.3105, "step": 3943 }, { "epoch": 1.8559999999999999, "grad_norm": 0.598247868955683, "learning_rate": 0.0002764373536206564, "loss": 0.2598, "step": 3944 }, { "epoch": 1.856470588235294, "grad_norm": 0.5240400247451853, "learning_rate": 0.0002764137623570146, "loss": 0.4766, "step": 3945 }, { "epoch": 1.8569411764705883, "grad_norm": 0.8816165073098859, "learning_rate": 0.0002763901602970174, "loss": 0.7695, "step": 3946 }, { "epoch": 1.8574117647058823, "grad_norm": 0.5386349390770168, "learning_rate": 0.0002763665474426806, "loss": 0.3418, "step": 3947 }, { "epoch": 1.8578823529411763, "grad_norm": 0.6426416836164873, "learning_rate": 0.00027634292379602086, "loss": 0.3574, "step": 3948 }, { "epoch": 1.8583529411764705, "grad_norm": 0.8484251872360186, "learning_rate": 0.00027631928935905567, "loss": 0.5625, "step": 3949 }, { "epoch": 1.8588235294117648, "grad_norm": 0.6548795049385904, "learning_rate": 0.0002762956441338036, "loss": 0.3223, "step": 3950 }, { "epoch": 1.859294117647059, "grad_norm": 0.4487677084091063, "learning_rate": 0.00027627198812228405, "loss": 0.2197, "step": 3951 }, { "epoch": 1.859764705882353, "grad_norm": 0.6429498152149756, "learning_rate": 0.0002762483213265173, "loss": 0.418, "step": 3952 }, { "epoch": 1.860235294117647, "grad_norm": 0.6199898906599667, "learning_rate": 0.0002762246437485247, "loss": 0.2373, "step": 3953 }, { "epoch": 1.8607058823529412, "grad_norm": 0.2715622952634902, "learning_rate": 0.0002762009553903284, "loss": 0.1069, "step": 3954 }, { "epoch": 1.8611764705882354, "grad_norm": 0.3785385717794578, "learning_rate": 0.0002761772562539515, "loss": 0.1377, "step": 3955 }, { "epoch": 1.8616470588235294, "grad_norm": 1.3841070977643757, "learning_rate": 0.0002761535463414181, "loss": 0.5039, "step": 3956 }, { "epoch": 1.8621176470588234, "grad_norm": 0.7479756193001401, "learning_rate": 0.0002761298256547529, "loss": 0.3594, "step": 3957 }, { "epoch": 1.8625882352941177, "grad_norm": 0.44142943561430975, "learning_rate": 0.00027610609419598204, "loss": 0.2168, "step": 3958 }, { "epoch": 1.8630588235294119, "grad_norm": 0.8453208474839619, "learning_rate": 0.0002760823519671321, "loss": 0.4473, "step": 3959 }, { "epoch": 1.8635294117647059, "grad_norm": 0.6081006064313307, "learning_rate": 0.0002760585989702309, "loss": 0.3398, "step": 3960 }, { "epoch": 1.8639999999999999, "grad_norm": 0.6370541764133888, "learning_rate": 0.0002760348352073071, "loss": 0.3184, "step": 3961 }, { "epoch": 1.864470588235294, "grad_norm": 0.5899506223260625, "learning_rate": 0.0002760110606803901, "loss": 0.25, "step": 3962 }, { "epoch": 1.8649411764705883, "grad_norm": 0.16374058873243416, "learning_rate": 0.00027598727539151054, "loss": 0.0635, "step": 3963 }, { "epoch": 1.8654117647058823, "grad_norm": 0.7127964072464803, "learning_rate": 0.0002759634793426996, "loss": 0.4355, "step": 3964 }, { "epoch": 1.8658823529411763, "grad_norm": 0.34231074802251216, "learning_rate": 0.00027593967253598974, "loss": 0.1602, "step": 3965 }, { "epoch": 1.8663529411764705, "grad_norm": 0.8534682532062593, "learning_rate": 0.00027591585497341406, "loss": 0.416, "step": 3966 }, { "epoch": 1.8668235294117648, "grad_norm": 0.6039346941022213, "learning_rate": 0.00027589202665700676, "loss": 0.1865, "step": 3967 }, { "epoch": 1.8672941176470588, "grad_norm": 1.0290732165941632, "learning_rate": 0.0002758681875888029, "loss": 0.5586, "step": 3968 }, { "epoch": 1.867764705882353, "grad_norm": 0.5925037274610244, "learning_rate": 0.00027584433777083845, "loss": 0.2148, "step": 3969 }, { "epoch": 1.868235294117647, "grad_norm": 0.9765632461483865, "learning_rate": 0.0002758204772051503, "loss": 0.5391, "step": 3970 }, { "epoch": 1.8687058823529412, "grad_norm": 0.3251248011320681, "learning_rate": 0.00027579660589377626, "loss": 0.1099, "step": 3971 }, { "epoch": 1.8691764705882354, "grad_norm": 0.9015339656534083, "learning_rate": 0.00027577272383875497, "loss": 0.4707, "step": 3972 }, { "epoch": 1.8696470588235294, "grad_norm": 0.5734628853121092, "learning_rate": 0.0002757488310421262, "loss": 0.2656, "step": 3973 }, { "epoch": 1.8701176470588234, "grad_norm": 0.3509613668282453, "learning_rate": 0.0002757249275059305, "loss": 0.168, "step": 3974 }, { "epoch": 1.8705882352941177, "grad_norm": 0.6446428821050263, "learning_rate": 0.00027570101323220927, "loss": 0.3652, "step": 3975 }, { "epoch": 1.8710588235294119, "grad_norm": 0.8487072776691496, "learning_rate": 0.00027567708822300496, "loss": 0.5547, "step": 3976 }, { "epoch": 1.8715294117647059, "grad_norm": 0.6275848392620613, "learning_rate": 0.0002756531524803609, "loss": 0.4141, "step": 3977 }, { "epoch": 1.8719999999999999, "grad_norm": 0.4409254396880624, "learning_rate": 0.00027562920600632127, "loss": 0.2266, "step": 3978 }, { "epoch": 1.872470588235294, "grad_norm": 0.4571490820803724, "learning_rate": 0.0002756052488029312, "loss": 0.1855, "step": 3979 }, { "epoch": 1.8729411764705883, "grad_norm": 0.939658841544043, "learning_rate": 0.0002755812808722368, "loss": 0.668, "step": 3980 }, { "epoch": 1.8734117647058823, "grad_norm": 0.6532970720535978, "learning_rate": 0.0002755573022162851, "loss": 0.3945, "step": 3981 }, { "epoch": 1.8738823529411763, "grad_norm": 0.5221342835480686, "learning_rate": 0.0002755333128371239, "loss": 0.3027, "step": 3982 }, { "epoch": 1.8743529411764706, "grad_norm": 0.5373120425047474, "learning_rate": 0.00027550931273680207, "loss": 0.2256, "step": 3983 }, { "epoch": 1.8748235294117648, "grad_norm": 0.5478283013658282, "learning_rate": 0.00027548530191736926, "loss": 0.3926, "step": 3984 }, { "epoch": 1.8752941176470588, "grad_norm": 0.9088182134027826, "learning_rate": 0.0002754612803808762, "loss": 0.5586, "step": 3985 }, { "epoch": 1.8757647058823528, "grad_norm": 0.5640533686767626, "learning_rate": 0.00027543724812937443, "loss": 0.4277, "step": 3986 }, { "epoch": 1.876235294117647, "grad_norm": 0.6433681174161633, "learning_rate": 0.0002754132051649164, "loss": 0.4473, "step": 3987 }, { "epoch": 1.8767058823529412, "grad_norm": 0.3928262558880872, "learning_rate": 0.00027538915148955545, "loss": 0.1006, "step": 3988 }, { "epoch": 1.8771764705882354, "grad_norm": 0.7879055504117866, "learning_rate": 0.000275365087105346, "loss": 0.4609, "step": 3989 }, { "epoch": 1.8776470588235294, "grad_norm": 1.1137548352999385, "learning_rate": 0.00027534101201434315, "loss": 0.7539, "step": 3990 }, { "epoch": 1.8781176470588234, "grad_norm": 0.7658402691246536, "learning_rate": 0.00027531692621860307, "loss": 0.377, "step": 3991 }, { "epoch": 1.8785882352941177, "grad_norm": 0.5911930814274785, "learning_rate": 0.00027529282972018287, "loss": 0.3242, "step": 3992 }, { "epoch": 1.879058823529412, "grad_norm": 0.6957575402338313, "learning_rate": 0.0002752687225211404, "loss": 0.3008, "step": 3993 }, { "epoch": 1.879529411764706, "grad_norm": 0.17363291446197626, "learning_rate": 0.0002752446046235346, "loss": 0.0664, "step": 3994 }, { "epoch": 1.88, "grad_norm": 0.4646712156066344, "learning_rate": 0.00027522047602942525, "loss": 0.1797, "step": 3995 }, { "epoch": 1.8804705882352941, "grad_norm": 0.5432746880393022, "learning_rate": 0.00027519633674087305, "loss": 0.2373, "step": 3996 }, { "epoch": 1.8809411764705883, "grad_norm": 1.2896288148684805, "learning_rate": 0.0002751721867599396, "loss": 0.668, "step": 3997 }, { "epoch": 1.8814117647058823, "grad_norm": 0.22887452755913854, "learning_rate": 0.00027514802608868743, "loss": 0.0549, "step": 3998 }, { "epoch": 1.8818823529411763, "grad_norm": 0.6493253112167112, "learning_rate": 0.00027512385472918, "loss": 0.4258, "step": 3999 }, { "epoch": 1.8823529411764706, "grad_norm": 1.05954316173421, "learning_rate": 0.0002750996726834817, "loss": 0.7344, "step": 4000 }, { "epoch": 1.8828235294117648, "grad_norm": 0.33926591989806587, "learning_rate": 0.00027507547995365766, "loss": 0.1387, "step": 4001 }, { "epoch": 1.8832941176470588, "grad_norm": 0.6778590051936878, "learning_rate": 0.00027505127654177413, "loss": 0.3906, "step": 4002 }, { "epoch": 1.8837647058823528, "grad_norm": 1.1822636831874764, "learning_rate": 0.0002750270624498983, "loss": 0.6484, "step": 4003 }, { "epoch": 1.884235294117647, "grad_norm": 0.6659095161929643, "learning_rate": 0.000275002837680098, "loss": 0.3086, "step": 4004 }, { "epoch": 1.8847058823529412, "grad_norm": 0.8941904863221932, "learning_rate": 0.0002749786022344422, "loss": 0.4004, "step": 4005 }, { "epoch": 1.8851764705882355, "grad_norm": 0.7753990304654802, "learning_rate": 0.0002749543561150009, "loss": 0.4434, "step": 4006 }, { "epoch": 1.8856470588235295, "grad_norm": 0.9660006195026851, "learning_rate": 0.00027493009932384456, "loss": 0.4395, "step": 4007 }, { "epoch": 1.8861176470588235, "grad_norm": 0.47257294472639744, "learning_rate": 0.000274905831863045, "loss": 0.2715, "step": 4008 }, { "epoch": 1.8865882352941177, "grad_norm": 0.6507010704656147, "learning_rate": 0.00027488155373467474, "loss": 0.2852, "step": 4009 }, { "epoch": 1.887058823529412, "grad_norm": 0.9871929242551599, "learning_rate": 0.00027485726494080724, "loss": 0.5586, "step": 4010 }, { "epoch": 1.887529411764706, "grad_norm": 0.7464012221499577, "learning_rate": 0.0002748329654835169, "loss": 0.4023, "step": 4011 }, { "epoch": 1.888, "grad_norm": 0.7685768101266142, "learning_rate": 0.000274808655364879, "loss": 0.4062, "step": 4012 }, { "epoch": 1.8884705882352941, "grad_norm": 1.518603638156052, "learning_rate": 0.00027478433458696976, "loss": 0.7891, "step": 4013 }, { "epoch": 1.8889411764705883, "grad_norm": 0.5049012871770346, "learning_rate": 0.0002747600031518663, "loss": 0.1406, "step": 4014 }, { "epoch": 1.8894117647058823, "grad_norm": 0.7591536759205634, "learning_rate": 0.0002747356610616466, "loss": 0.3535, "step": 4015 }, { "epoch": 1.8898823529411763, "grad_norm": 0.4439618049804798, "learning_rate": 0.0002747113083183896, "loss": 0.252, "step": 4016 }, { "epoch": 1.8903529411764706, "grad_norm": 0.562164246459679, "learning_rate": 0.00027468694492417515, "loss": 0.2773, "step": 4017 }, { "epoch": 1.8908235294117648, "grad_norm": 1.2878784189953323, "learning_rate": 0.0002746625708810841, "loss": 0.5547, "step": 4018 }, { "epoch": 1.8912941176470588, "grad_norm": 0.40935084582441383, "learning_rate": 0.00027463818619119794, "loss": 0.1123, "step": 4019 }, { "epoch": 1.8917647058823528, "grad_norm": 1.849743380989478, "learning_rate": 0.0002746137908565994, "loss": 0.8125, "step": 4020 }, { "epoch": 1.892235294117647, "grad_norm": 0.7203976637812926, "learning_rate": 0.00027458938487937186, "loss": 0.4062, "step": 4021 }, { "epoch": 1.8927058823529412, "grad_norm": 1.0568313725263117, "learning_rate": 0.0002745649682615998, "loss": 0.3652, "step": 4022 }, { "epoch": 1.8931764705882355, "grad_norm": 0.5795483636353984, "learning_rate": 0.00027454054100536836, "loss": 0.2217, "step": 4023 }, { "epoch": 1.8936470588235295, "grad_norm": 0.48834581007744776, "learning_rate": 0.00027451610311276397, "loss": 0.2578, "step": 4024 }, { "epoch": 1.8941176470588235, "grad_norm": 0.45107616677016443, "learning_rate": 0.00027449165458587354, "loss": 0.291, "step": 4025 }, { "epoch": 1.8945882352941177, "grad_norm": 0.3893767442362545, "learning_rate": 0.0002744671954267852, "loss": 0.2207, "step": 4026 }, { "epoch": 1.895058823529412, "grad_norm": 0.5311460693259638, "learning_rate": 0.00027444272563758784, "loss": 0.21, "step": 4027 }, { "epoch": 1.895529411764706, "grad_norm": 0.5384001759165298, "learning_rate": 0.00027441824522037137, "loss": 0.2432, "step": 4028 }, { "epoch": 1.896, "grad_norm": 0.3459885650350138, "learning_rate": 0.0002743937541772265, "loss": 0.1494, "step": 4029 }, { "epoch": 1.8964705882352941, "grad_norm": 0.33667076916312755, "learning_rate": 0.0002743692525102448, "loss": 0.2197, "step": 4030 }, { "epoch": 1.8969411764705884, "grad_norm": 0.5118641015628198, "learning_rate": 0.000274344740221519, "loss": 0.1455, "step": 4031 }, { "epoch": 1.8974117647058824, "grad_norm": 0.6764776361448467, "learning_rate": 0.0002743202173131424, "loss": 0.4531, "step": 4032 }, { "epoch": 1.8978823529411764, "grad_norm": 0.7095046068914053, "learning_rate": 0.00027429568378720945, "loss": 0.4551, "step": 4033 }, { "epoch": 1.8983529411764706, "grad_norm": 0.3536550267374807, "learning_rate": 0.0002742711396458155, "loss": 0.1387, "step": 4034 }, { "epoch": 1.8988235294117648, "grad_norm": 0.7351338807231597, "learning_rate": 0.00027424658489105667, "loss": 0.3633, "step": 4035 }, { "epoch": 1.8992941176470588, "grad_norm": 0.5103203074546883, "learning_rate": 0.00027422201952503005, "loss": 0.3379, "step": 4036 }, { "epoch": 1.8997647058823528, "grad_norm": 0.684977882543098, "learning_rate": 0.00027419744354983364, "loss": 0.375, "step": 4037 }, { "epoch": 1.900235294117647, "grad_norm": 0.8210586379558052, "learning_rate": 0.00027417285696756634, "loss": 0.6836, "step": 4038 }, { "epoch": 1.9007058823529412, "grad_norm": 0.8317491131772318, "learning_rate": 0.00027414825978032805, "loss": 0.4395, "step": 4039 }, { "epoch": 1.9011764705882352, "grad_norm": 0.243050109088622, "learning_rate": 0.0002741236519902194, "loss": 0.1338, "step": 4040 }, { "epoch": 1.9016470588235292, "grad_norm": 0.5087913836702378, "learning_rate": 0.0002740990335993421, "loss": 0.3379, "step": 4041 }, { "epoch": 1.9021176470588235, "grad_norm": 0.681109034884753, "learning_rate": 0.00027407440460979854, "loss": 0.3516, "step": 4042 }, { "epoch": 1.9025882352941177, "grad_norm": 1.3998471543307973, "learning_rate": 0.0002740497650236923, "loss": 1.1328, "step": 4043 }, { "epoch": 1.903058823529412, "grad_norm": 0.5774329476320269, "learning_rate": 0.00027402511484312764, "loss": 0.2539, "step": 4044 }, { "epoch": 1.903529411764706, "grad_norm": 0.4256831514402977, "learning_rate": 0.00027400045407020986, "loss": 0.252, "step": 4045 }, { "epoch": 1.904, "grad_norm": 0.47984251913419484, "learning_rate": 0.00027397578270704504, "loss": 0.2051, "step": 4046 }, { "epoch": 1.9044705882352941, "grad_norm": 0.5371283921005668, "learning_rate": 0.0002739511007557403, "loss": 0.1787, "step": 4047 }, { "epoch": 1.9049411764705884, "grad_norm": 0.22539308951128828, "learning_rate": 0.00027392640821840365, "loss": 0.0608, "step": 4048 }, { "epoch": 1.9054117647058824, "grad_norm": 0.7063013501610246, "learning_rate": 0.0002739017050971438, "loss": 0.3379, "step": 4049 }, { "epoch": 1.9058823529411764, "grad_norm": 0.3161031774481649, "learning_rate": 0.0002738769913940706, "loss": 0.1338, "step": 4050 }, { "epoch": 1.9063529411764706, "grad_norm": 0.8445067445471733, "learning_rate": 0.00027385226711129477, "loss": 0.6094, "step": 4051 }, { "epoch": 1.9068235294117648, "grad_norm": 0.5196903483338, "learning_rate": 0.00027382753225092777, "loss": 0.2324, "step": 4052 }, { "epoch": 1.9072941176470588, "grad_norm": 0.6741934196074226, "learning_rate": 0.0002738027868150822, "loss": 0.459, "step": 4053 }, { "epoch": 1.9077647058823528, "grad_norm": 1.0207440905925593, "learning_rate": 0.0002737780308058714, "loss": 0.5156, "step": 4054 }, { "epoch": 1.908235294117647, "grad_norm": 0.4099995136773267, "learning_rate": 0.00027375326422540957, "loss": 0.1865, "step": 4055 }, { "epoch": 1.9087058823529413, "grad_norm": 0.6554015461768926, "learning_rate": 0.000273728487075812, "loss": 0.3906, "step": 4056 }, { "epoch": 1.9091764705882353, "grad_norm": 0.4492382044838128, "learning_rate": 0.0002737036993591948, "loss": 0.3516, "step": 4057 }, { "epoch": 1.9096470588235293, "grad_norm": 0.39048811515065507, "learning_rate": 0.0002736789010776749, "loss": 0.1494, "step": 4058 }, { "epoch": 1.9101176470588235, "grad_norm": 0.3016475066040567, "learning_rate": 0.00027365409223337016, "loss": 0.105, "step": 4059 }, { "epoch": 1.9105882352941177, "grad_norm": 0.4962591673981564, "learning_rate": 0.0002736292728283994, "loss": 0.2773, "step": 4060 }, { "epoch": 1.911058823529412, "grad_norm": 0.5158104328099821, "learning_rate": 0.0002736044428648824, "loss": 0.3047, "step": 4061 }, { "epoch": 1.911529411764706, "grad_norm": 0.5119802398247325, "learning_rate": 0.0002735796023449397, "loss": 0.293, "step": 4062 }, { "epoch": 1.912, "grad_norm": 0.6395180657837585, "learning_rate": 0.00027355475127069275, "loss": 0.4043, "step": 4063 }, { "epoch": 1.9124705882352941, "grad_norm": 0.3775400516690439, "learning_rate": 0.0002735298896442641, "loss": 0.1523, "step": 4064 }, { "epoch": 1.9129411764705884, "grad_norm": 0.8379024630612059, "learning_rate": 0.0002735050174677769, "loss": 0.4297, "step": 4065 }, { "epoch": 1.9134117647058824, "grad_norm": 0.6403489985152722, "learning_rate": 0.00027348013474335546, "loss": 0.3145, "step": 4066 }, { "epoch": 1.9138823529411764, "grad_norm": 0.5544422809300148, "learning_rate": 0.0002734552414731248, "loss": 0.4688, "step": 4067 }, { "epoch": 1.9143529411764706, "grad_norm": 0.48164694839438577, "learning_rate": 0.000273430337659211, "loss": 0.3398, "step": 4068 }, { "epoch": 1.9148235294117648, "grad_norm": 0.5103773703475097, "learning_rate": 0.0002734054233037409, "loss": 0.3535, "step": 4069 }, { "epoch": 1.9152941176470588, "grad_norm": 0.6900642547563317, "learning_rate": 0.00027338049840884245, "loss": 0.3242, "step": 4070 }, { "epoch": 1.9157647058823528, "grad_norm": 0.6572643104305829, "learning_rate": 0.00027335556297664417, "loss": 0.4473, "step": 4071 }, { "epoch": 1.916235294117647, "grad_norm": 0.7166919132132807, "learning_rate": 0.00027333061700927577, "loss": 0.4531, "step": 4072 }, { "epoch": 1.9167058823529413, "grad_norm": 0.11204037921492618, "learning_rate": 0.00027330566050886776, "loss": 0.0253, "step": 4073 }, { "epoch": 1.9171764705882353, "grad_norm": 0.6248270564048417, "learning_rate": 0.0002732806934775515, "loss": 0.3711, "step": 4074 }, { "epoch": 1.9176470588235293, "grad_norm": 0.7604235839495506, "learning_rate": 0.0002732557159174594, "loss": 0.4395, "step": 4075 }, { "epoch": 1.9181176470588235, "grad_norm": 0.236032883101031, "learning_rate": 0.0002732307278307246, "loss": 0.0874, "step": 4076 }, { "epoch": 1.9185882352941177, "grad_norm": 0.4632000523983454, "learning_rate": 0.0002732057292194811, "loss": 0.2139, "step": 4077 }, { "epoch": 1.919058823529412, "grad_norm": 0.7958651749445348, "learning_rate": 0.000273180720085864, "loss": 0.2676, "step": 4078 }, { "epoch": 1.919529411764706, "grad_norm": 0.3726697101128571, "learning_rate": 0.0002731557004320093, "loss": 0.2969, "step": 4079 }, { "epoch": 1.92, "grad_norm": 0.7956009041216338, "learning_rate": 0.0002731306702600537, "loss": 0.3496, "step": 4080 }, { "epoch": 1.9204705882352942, "grad_norm": 0.5594063046553679, "learning_rate": 0.00027310562957213487, "loss": 0.3965, "step": 4081 }, { "epoch": 1.9209411764705884, "grad_norm": 0.6694167784827791, "learning_rate": 0.00027308057837039144, "loss": 0.4121, "step": 4082 }, { "epoch": 1.9214117647058824, "grad_norm": 0.2950404280097077, "learning_rate": 0.0002730555166569629, "loss": 0.1357, "step": 4083 }, { "epoch": 1.9218823529411764, "grad_norm": 0.6345874379490247, "learning_rate": 0.00027303044443398967, "loss": 0.3027, "step": 4084 }, { "epoch": 1.9223529411764706, "grad_norm": 0.18421825688985105, "learning_rate": 0.00027300536170361307, "loss": 0.0474, "step": 4085 }, { "epoch": 1.9228235294117648, "grad_norm": 0.9749295141706882, "learning_rate": 0.00027298026846797515, "loss": 0.5469, "step": 4086 }, { "epoch": 1.9232941176470588, "grad_norm": 0.6560961991204883, "learning_rate": 0.0002729551647292191, "loss": 0.4121, "step": 4087 }, { "epoch": 1.9237647058823528, "grad_norm": 0.8122890270238483, "learning_rate": 0.000272930050489489, "loss": 0.5859, "step": 4088 }, { "epoch": 1.924235294117647, "grad_norm": 0.36461895035503705, "learning_rate": 0.0002729049257509295, "loss": 0.0996, "step": 4089 }, { "epoch": 1.9247058823529413, "grad_norm": 0.04345451053489233, "learning_rate": 0.0002728797905156866, "loss": 0.0112, "step": 4090 }, { "epoch": 1.9251764705882353, "grad_norm": 0.43332978599657096, "learning_rate": 0.0002728546447859068, "loss": 0.3457, "step": 4091 }, { "epoch": 1.9256470588235293, "grad_norm": 0.5599814170811723, "learning_rate": 0.00027282948856373774, "loss": 0.2832, "step": 4092 }, { "epoch": 1.9261176470588235, "grad_norm": 0.6527517893097506, "learning_rate": 0.0002728043218513279, "loss": 0.2988, "step": 4093 }, { "epoch": 1.9265882352941177, "grad_norm": 0.3522716110269936, "learning_rate": 0.0002727791446508266, "loss": 0.0679, "step": 4094 }, { "epoch": 1.9270588235294117, "grad_norm": 0.6393769613325891, "learning_rate": 0.00027275395696438413, "loss": 0.3145, "step": 4095 }, { "epoch": 1.927529411764706, "grad_norm": 0.6537188515288065, "learning_rate": 0.0002727287587941516, "loss": 0.3633, "step": 4096 }, { "epoch": 1.928, "grad_norm": 0.6002132889166114, "learning_rate": 0.00027270355014228113, "loss": 0.4062, "step": 4097 }, { "epoch": 1.9284705882352942, "grad_norm": 0.18375912956319965, "learning_rate": 0.00027267833101092564, "loss": 0.0952, "step": 4098 }, { "epoch": 1.9289411764705884, "grad_norm": 0.9872243720865852, "learning_rate": 0.0002726531014022389, "loss": 0.4434, "step": 4099 }, { "epoch": 1.9294117647058824, "grad_norm": 0.280271390998025, "learning_rate": 0.00027262786131837573, "loss": 0.127, "step": 4100 }, { "epoch": 1.9298823529411764, "grad_norm": 0.49804286869414544, "learning_rate": 0.00027260261076149166, "loss": 0.377, "step": 4101 }, { "epoch": 1.9303529411764706, "grad_norm": 0.2214929429569479, "learning_rate": 0.00027257734973374335, "loss": 0.0728, "step": 4102 }, { "epoch": 1.9308235294117648, "grad_norm": 0.620838195781694, "learning_rate": 0.0002725520782372881, "loss": 0.2197, "step": 4103 }, { "epoch": 1.9312941176470588, "grad_norm": 0.7289204238153438, "learning_rate": 0.00027252679627428423, "loss": 0.3203, "step": 4104 }, { "epoch": 1.9317647058823528, "grad_norm": 0.3051356345243907, "learning_rate": 0.00027250150384689094, "loss": 0.127, "step": 4105 }, { "epoch": 1.932235294117647, "grad_norm": 0.3686876309964001, "learning_rate": 0.0002724762009572684, "loss": 0.1758, "step": 4106 }, { "epoch": 1.9327058823529413, "grad_norm": 0.4781573146618943, "learning_rate": 0.0002724508876075776, "loss": 0.293, "step": 4107 }, { "epoch": 1.9331764705882353, "grad_norm": 0.6246072711003641, "learning_rate": 0.0002724255637999803, "loss": 0.373, "step": 4108 }, { "epoch": 1.9336470588235293, "grad_norm": 0.38049426173946815, "learning_rate": 0.0002724002295366394, "loss": 0.2559, "step": 4109 }, { "epoch": 1.9341176470588235, "grad_norm": 0.49753146460186304, "learning_rate": 0.0002723748848197185, "loss": 0.1426, "step": 4110 }, { "epoch": 1.9345882352941177, "grad_norm": 0.5086992637383695, "learning_rate": 0.00027234952965138223, "loss": 0.2598, "step": 4111 }, { "epoch": 1.9350588235294117, "grad_norm": 0.43873162948476446, "learning_rate": 0.000272324164033796, "loss": 0.1846, "step": 4112 }, { "epoch": 1.9355294117647057, "grad_norm": 1.0635470024068236, "learning_rate": 0.0002722987879691261, "loss": 0.7344, "step": 4113 }, { "epoch": 1.936, "grad_norm": 0.637384451049744, "learning_rate": 0.00027227340145953994, "loss": 0.2832, "step": 4114 }, { "epoch": 1.9364705882352942, "grad_norm": 0.8193597393472482, "learning_rate": 0.0002722480045072055, "loss": 0.5156, "step": 4115 }, { "epoch": 1.9369411764705884, "grad_norm": 0.3069019588345523, "learning_rate": 0.0002722225971142918, "loss": 0.1206, "step": 4116 }, { "epoch": 1.9374117647058824, "grad_norm": 0.3595623154831395, "learning_rate": 0.00027219717928296885, "loss": 0.1729, "step": 4117 }, { "epoch": 1.9378823529411764, "grad_norm": 0.4827859473832828, "learning_rate": 0.0002721717510154074, "loss": 0.2598, "step": 4118 }, { "epoch": 1.9383529411764706, "grad_norm": 0.16161296952808615, "learning_rate": 0.0002721463123137792, "loss": 0.0339, "step": 4119 }, { "epoch": 1.9388235294117648, "grad_norm": 0.46351079303647297, "learning_rate": 0.0002721208631802568, "loss": 0.2158, "step": 4120 }, { "epoch": 1.9392941176470588, "grad_norm": 0.3378745762035209, "learning_rate": 0.00027209540361701365, "loss": 0.1299, "step": 4121 }, { "epoch": 1.9397647058823528, "grad_norm": 0.3655902499106526, "learning_rate": 0.0002720699336262242, "loss": 0.106, "step": 4122 }, { "epoch": 1.940235294117647, "grad_norm": 0.5894745156552506, "learning_rate": 0.00027204445321006365, "loss": 0.2988, "step": 4123 }, { "epoch": 1.9407058823529413, "grad_norm": 0.3771441771909052, "learning_rate": 0.00027201896237070815, "loss": 0.1973, "step": 4124 }, { "epoch": 1.9411764705882353, "grad_norm": 0.49892621542266036, "learning_rate": 0.0002719934611103348, "loss": 0.2695, "step": 4125 }, { "epoch": 1.9416470588235293, "grad_norm": 0.4407724749210434, "learning_rate": 0.00027196794943112147, "loss": 0.1943, "step": 4126 }, { "epoch": 1.9421176470588235, "grad_norm": 0.5997367453953644, "learning_rate": 0.000271942427335247, "loss": 0.2754, "step": 4127 }, { "epoch": 1.9425882352941177, "grad_norm": 0.2659883466162374, "learning_rate": 0.00027191689482489114, "loss": 0.1182, "step": 4128 }, { "epoch": 1.9430588235294117, "grad_norm": 0.4164144813855835, "learning_rate": 0.0002718913519022344, "loss": 0.1953, "step": 4129 }, { "epoch": 1.9435294117647057, "grad_norm": 0.8549607080110786, "learning_rate": 0.00027186579856945843, "loss": 0.4492, "step": 4130 }, { "epoch": 1.944, "grad_norm": 0.4352811440004546, "learning_rate": 0.00027184023482874545, "loss": 0.2715, "step": 4131 }, { "epoch": 1.9444705882352942, "grad_norm": 0.44470517187774555, "learning_rate": 0.00027181466068227885, "loss": 0.1758, "step": 4132 }, { "epoch": 1.9449411764705884, "grad_norm": 0.7879922969425378, "learning_rate": 0.00027178907613224264, "loss": 0.4258, "step": 4133 }, { "epoch": 1.9454117647058824, "grad_norm": 0.49188390778227403, "learning_rate": 0.00027176348118082203, "loss": 0.2461, "step": 4134 }, { "epoch": 1.9458823529411764, "grad_norm": 0.9500719296617347, "learning_rate": 0.0002717378758302029, "loss": 0.6406, "step": 4135 }, { "epoch": 1.9463529411764706, "grad_norm": 1.0396611497807666, "learning_rate": 0.000271712260082572, "loss": 0.5742, "step": 4136 }, { "epoch": 1.9468235294117648, "grad_norm": 0.6119365813032608, "learning_rate": 0.0002716866339401171, "loss": 0.3105, "step": 4137 }, { "epoch": 1.9472941176470588, "grad_norm": 0.3459919078844705, "learning_rate": 0.0002716609974050268, "loss": 0.2461, "step": 4138 }, { "epoch": 1.9477647058823528, "grad_norm": 0.7324101068327012, "learning_rate": 0.0002716353504794906, "loss": 0.6445, "step": 4139 }, { "epoch": 1.948235294117647, "grad_norm": 0.39104781626011875, "learning_rate": 0.0002716096931656988, "loss": 0.1611, "step": 4140 }, { "epoch": 1.9487058823529413, "grad_norm": 1.030846790731753, "learning_rate": 0.0002715840254658427, "loss": 0.5312, "step": 4141 }, { "epoch": 1.9491764705882353, "grad_norm": 0.8714042870065627, "learning_rate": 0.0002715583473821145, "loss": 0.4043, "step": 4142 }, { "epoch": 1.9496470588235293, "grad_norm": 0.504546898570171, "learning_rate": 0.0002715326589167072, "loss": 0.2314, "step": 4143 }, { "epoch": 1.9501176470588235, "grad_norm": 0.8858449538905794, "learning_rate": 0.0002715069600718147, "loss": 0.3223, "step": 4144 }, { "epoch": 1.9505882352941177, "grad_norm": 0.9005729853928928, "learning_rate": 0.0002714812508496319, "loss": 0.4473, "step": 4145 }, { "epoch": 1.9510588235294117, "grad_norm": 0.8279670855262014, "learning_rate": 0.0002714555312523543, "loss": 0.3613, "step": 4146 }, { "epoch": 1.9515294117647057, "grad_norm": 1.2454607215498918, "learning_rate": 0.00027142980128217864, "loss": 0.6094, "step": 4147 }, { "epoch": 1.952, "grad_norm": 0.6184115290791009, "learning_rate": 0.00027140406094130234, "loss": 0.3105, "step": 4148 }, { "epoch": 1.9524705882352942, "grad_norm": 0.6441601353697481, "learning_rate": 0.0002713783102319237, "loss": 0.2949, "step": 4149 }, { "epoch": 1.9529411764705882, "grad_norm": 0.9405680180774264, "learning_rate": 0.0002713525491562421, "loss": 0.3281, "step": 4150 }, { "epoch": 1.9534117647058824, "grad_norm": 0.6077833388144943, "learning_rate": 0.0002713267777164575, "loss": 0.2695, "step": 4151 }, { "epoch": 1.9538823529411764, "grad_norm": 0.5252850388717462, "learning_rate": 0.00027130099591477104, "loss": 0.1973, "step": 4152 }, { "epoch": 1.9543529411764706, "grad_norm": 1.0795687459236774, "learning_rate": 0.0002712752037533845, "loss": 0.6055, "step": 4153 }, { "epoch": 1.9548235294117648, "grad_norm": 0.9466378527596505, "learning_rate": 0.00027124940123450075, "loss": 0.6562, "step": 4154 }, { "epoch": 1.9552941176470588, "grad_norm": 1.0935158775119775, "learning_rate": 0.00027122358836032337, "loss": 0.6094, "step": 4155 }, { "epoch": 1.9557647058823528, "grad_norm": 0.22642347648811287, "learning_rate": 0.000271197765133057, "loss": 0.0596, "step": 4156 }, { "epoch": 1.956235294117647, "grad_norm": 0.6506790695322383, "learning_rate": 0.00027117193155490695, "loss": 0.1699, "step": 4157 }, { "epoch": 1.9567058823529413, "grad_norm": 1.5744891443167013, "learning_rate": 0.00027114608762807965, "loss": 0.459, "step": 4158 }, { "epoch": 1.9571764705882353, "grad_norm": 1.2335293241609273, "learning_rate": 0.0002711202333547822, "loss": 0.5938, "step": 4159 }, { "epoch": 1.9576470588235293, "grad_norm": 0.5531599261757548, "learning_rate": 0.00027109436873722277, "loss": 0.3652, "step": 4160 }, { "epoch": 1.9581176470588235, "grad_norm": 0.36641210361509385, "learning_rate": 0.00027106849377761023, "loss": 0.0918, "step": 4161 }, { "epoch": 1.9585882352941177, "grad_norm": 0.9011299573636279, "learning_rate": 0.00027104260847815457, "loss": 0.2852, "step": 4162 }, { "epoch": 1.9590588235294117, "grad_norm": 0.9692788901037857, "learning_rate": 0.0002710167128410664, "loss": 0.3945, "step": 4163 }, { "epoch": 1.9595294117647057, "grad_norm": 1.4998447363260754, "learning_rate": 0.00027099080686855735, "loss": 0.4434, "step": 4164 }, { "epoch": 1.96, "grad_norm": 1.1353632636810185, "learning_rate": 0.00027096489056283995, "loss": 0.5312, "step": 4165 }, { "epoch": 1.9604705882352942, "grad_norm": 0.4011101694854458, "learning_rate": 0.00027093896392612753, "loss": 0.2715, "step": 4166 }, { "epoch": 1.9609411764705882, "grad_norm": 0.8120085526050767, "learning_rate": 0.0002709130269606345, "loss": 0.5039, "step": 4167 }, { "epoch": 1.9614117647058822, "grad_norm": 1.157469452390679, "learning_rate": 0.00027088707966857576, "loss": 0.4766, "step": 4168 }, { "epoch": 1.9618823529411764, "grad_norm": 0.8154533181242408, "learning_rate": 0.0002708611220521675, "loss": 0.291, "step": 4169 }, { "epoch": 1.9623529411764706, "grad_norm": 0.7745355366983023, "learning_rate": 0.0002708351541136267, "loss": 0.3203, "step": 4170 }, { "epoch": 1.9628235294117649, "grad_norm": 0.9942364236516972, "learning_rate": 0.0002708091758551709, "loss": 0.4453, "step": 4171 }, { "epoch": 1.9632941176470589, "grad_norm": 0.4118159902842822, "learning_rate": 0.000270783187279019, "loss": 0.1738, "step": 4172 }, { "epoch": 1.9637647058823529, "grad_norm": 1.190763848045446, "learning_rate": 0.00027075718838739043, "loss": 0.6211, "step": 4173 }, { "epoch": 1.964235294117647, "grad_norm": 0.14827056310636008, "learning_rate": 0.0002707311791825057, "loss": 0.0381, "step": 4174 }, { "epoch": 1.9647058823529413, "grad_norm": 1.1904353449363039, "learning_rate": 0.00027070515966658604, "loss": 0.5977, "step": 4175 }, { "epoch": 1.9651764705882353, "grad_norm": 0.46064632343728534, "learning_rate": 0.00027067912984185376, "loss": 0.104, "step": 4176 }, { "epoch": 1.9656470588235293, "grad_norm": 0.3900546702202347, "learning_rate": 0.0002706530897105318, "loss": 0.1152, "step": 4177 }, { "epoch": 1.9661176470588235, "grad_norm": 1.068331633683494, "learning_rate": 0.0002706270392748442, "loss": 0.4688, "step": 4178 }, { "epoch": 1.9665882352941177, "grad_norm": 0.7810541899417911, "learning_rate": 0.00027060097853701576, "loss": 0.3867, "step": 4179 }, { "epoch": 1.9670588235294117, "grad_norm": 0.5029425132982974, "learning_rate": 0.0002705749074992722, "loss": 0.2373, "step": 4180 }, { "epoch": 1.9675294117647057, "grad_norm": 0.1722611633187372, "learning_rate": 0.00027054882616384015, "loss": 0.0654, "step": 4181 }, { "epoch": 1.968, "grad_norm": 0.6661538411650157, "learning_rate": 0.0002705227345329471, "loss": 0.4824, "step": 4182 }, { "epoch": 1.9684705882352942, "grad_norm": 0.5110097860347951, "learning_rate": 0.00027049663260882124, "loss": 0.2393, "step": 4183 }, { "epoch": 1.9689411764705882, "grad_norm": 0.5920761793403551, "learning_rate": 0.00027047052039369204, "loss": 0.3105, "step": 4184 }, { "epoch": 1.9694117647058822, "grad_norm": 0.4435352396789925, "learning_rate": 0.0002704443978897894, "loss": 0.2578, "step": 4185 }, { "epoch": 1.9698823529411764, "grad_norm": 0.3756081415182603, "learning_rate": 0.0002704182650993445, "loss": 0.1533, "step": 4186 }, { "epoch": 1.9703529411764706, "grad_norm": 0.8967084399995298, "learning_rate": 0.00027039212202458905, "loss": 0.457, "step": 4187 }, { "epoch": 1.9708235294117649, "grad_norm": 0.39627514352083903, "learning_rate": 0.0002703659686677559, "loss": 0.1089, "step": 4188 }, { "epoch": 1.9712941176470589, "grad_norm": 0.3572662515846285, "learning_rate": 0.00027033980503107856, "loss": 0.1729, "step": 4189 }, { "epoch": 1.9717647058823529, "grad_norm": 1.234974194554183, "learning_rate": 0.00027031363111679164, "loss": 0.6133, "step": 4190 }, { "epoch": 1.972235294117647, "grad_norm": 0.3801333245143472, "learning_rate": 0.0002702874469271305, "loss": 0.2461, "step": 4191 }, { "epoch": 1.9727058823529413, "grad_norm": 0.4928638622442806, "learning_rate": 0.00027026125246433144, "loss": 0.3145, "step": 4192 }, { "epoch": 1.9731764705882353, "grad_norm": 1.464527196378793, "learning_rate": 0.0002702350477306315, "loss": 0.6367, "step": 4193 }, { "epoch": 1.9736470588235293, "grad_norm": 0.42160066517503764, "learning_rate": 0.0002702088327282687, "loss": 0.1592, "step": 4194 }, { "epoch": 1.9741176470588235, "grad_norm": 0.43157775838570217, "learning_rate": 0.00027018260745948196, "loss": 0.1885, "step": 4195 }, { "epoch": 1.9745882352941178, "grad_norm": 0.26553305112451014, "learning_rate": 0.0002701563719265111, "loss": 0.0977, "step": 4196 }, { "epoch": 1.9750588235294118, "grad_norm": 0.5664100626649606, "learning_rate": 0.0002701301261315967, "loss": 0.3906, "step": 4197 }, { "epoch": 1.9755294117647058, "grad_norm": 0.5432216019207542, "learning_rate": 0.0002701038700769803, "loss": 0.2676, "step": 4198 }, { "epoch": 1.976, "grad_norm": 0.6145401554050415, "learning_rate": 0.00027007760376490425, "loss": 0.4395, "step": 4199 }, { "epoch": 1.9764705882352942, "grad_norm": 1.0428005967802063, "learning_rate": 0.0002700513271976119, "loss": 0.6328, "step": 4200 }, { "epoch": 1.9769411764705882, "grad_norm": 0.39749292825289495, "learning_rate": 0.0002700250403773473, "loss": 0.1904, "step": 4201 }, { "epoch": 1.9774117647058822, "grad_norm": 1.0984407683799726, "learning_rate": 0.0002699987433063556, "loss": 0.4785, "step": 4202 }, { "epoch": 1.9778823529411764, "grad_norm": 0.5419716742012964, "learning_rate": 0.0002699724359868826, "loss": 0.4219, "step": 4203 }, { "epoch": 1.9783529411764706, "grad_norm": 0.46547044926958764, "learning_rate": 0.0002699461184211751, "loss": 0.3086, "step": 4204 }, { "epoch": 1.9788235294117649, "grad_norm": 0.7875838520883358, "learning_rate": 0.00026991979061148077, "loss": 0.377, "step": 4205 }, { "epoch": 1.9792941176470589, "grad_norm": 0.27188594694403057, "learning_rate": 0.00026989345256004814, "loss": 0.1099, "step": 4206 }, { "epoch": 1.9797647058823529, "grad_norm": 0.8562653830019358, "learning_rate": 0.00026986710426912656, "loss": 0.293, "step": 4207 }, { "epoch": 1.980235294117647, "grad_norm": 0.4408314047832049, "learning_rate": 0.0002698407457409663, "loss": 0.1729, "step": 4208 }, { "epoch": 1.9807058823529413, "grad_norm": 0.5191847394315839, "learning_rate": 0.00026981437697781854, "loss": 0.1719, "step": 4209 }, { "epoch": 1.9811764705882353, "grad_norm": 0.8504773499111443, "learning_rate": 0.0002697879979819354, "loss": 0.4395, "step": 4210 }, { "epoch": 1.9816470588235293, "grad_norm": 0.3568932117344741, "learning_rate": 0.0002697616087555696, "loss": 0.1719, "step": 4211 }, { "epoch": 1.9821176470588235, "grad_norm": 0.7503759585046258, "learning_rate": 0.00026973520930097494, "loss": 0.4316, "step": 4212 }, { "epoch": 1.9825882352941178, "grad_norm": 1.0209085423092348, "learning_rate": 0.0002697087996204062, "loss": 0.6094, "step": 4213 }, { "epoch": 1.9830588235294118, "grad_norm": 0.30867686535929845, "learning_rate": 0.0002696823797161187, "loss": 0.1143, "step": 4214 }, { "epoch": 1.9835294117647058, "grad_norm": 0.33187114729573697, "learning_rate": 0.00026965594959036906, "loss": 0.2852, "step": 4215 }, { "epoch": 1.984, "grad_norm": 0.5963433928007793, "learning_rate": 0.00026962950924541435, "loss": 0.1553, "step": 4216 }, { "epoch": 1.9844705882352942, "grad_norm": 0.8815886790907018, "learning_rate": 0.0002696030586835128, "loss": 0.4414, "step": 4217 }, { "epoch": 1.9849411764705882, "grad_norm": 1.208774595599944, "learning_rate": 0.00026957659790692335, "loss": 0.4688, "step": 4218 }, { "epoch": 1.9854117647058822, "grad_norm": 0.21698609877864627, "learning_rate": 0.0002695501269179059, "loss": 0.0728, "step": 4219 }, { "epoch": 1.9858823529411764, "grad_norm": 0.30941439245423835, "learning_rate": 0.00026952364571872125, "loss": 0.1172, "step": 4220 }, { "epoch": 1.9863529411764707, "grad_norm": 0.5310141896431626, "learning_rate": 0.000269497154311631, "loss": 0.2393, "step": 4221 }, { "epoch": 1.9868235294117647, "grad_norm": 0.6027312876933395, "learning_rate": 0.00026947065269889765, "loss": 0.3984, "step": 4222 }, { "epoch": 1.9872941176470589, "grad_norm": 0.7980271850411194, "learning_rate": 0.00026944414088278453, "loss": 0.2617, "step": 4223 }, { "epoch": 1.9877647058823529, "grad_norm": 0.8073537475228729, "learning_rate": 0.00026941761886555595, "loss": 0.4492, "step": 4224 }, { "epoch": 1.988235294117647, "grad_norm": 0.7049885124288462, "learning_rate": 0.00026939108664947694, "loss": 0.4961, "step": 4225 }, { "epoch": 1.9887058823529413, "grad_norm": 0.5136331634479008, "learning_rate": 0.0002693645442368135, "loss": 0.332, "step": 4226 }, { "epoch": 1.9891764705882353, "grad_norm": 0.5162341274758, "learning_rate": 0.0002693379916298326, "loss": 0.3086, "step": 4227 }, { "epoch": 1.9896470588235293, "grad_norm": 0.31144165209124924, "learning_rate": 0.00026931142883080176, "loss": 0.1035, "step": 4228 }, { "epoch": 1.9901176470588235, "grad_norm": 0.5573153832379506, "learning_rate": 0.00026928485584198973, "loss": 0.1924, "step": 4229 }, { "epoch": 1.9905882352941178, "grad_norm": 0.5819979349234548, "learning_rate": 0.00026925827266566595, "loss": 0.2812, "step": 4230 }, { "epoch": 1.9910588235294118, "grad_norm": 0.6162112643214661, "learning_rate": 0.0002692316793041007, "loss": 0.2676, "step": 4231 }, { "epoch": 1.9915294117647058, "grad_norm": 0.6848450087380703, "learning_rate": 0.0002692050757595652, "loss": 0.4297, "step": 4232 }, { "epoch": 1.992, "grad_norm": 0.6632921774575664, "learning_rate": 0.00026917846203433155, "loss": 0.3828, "step": 4233 }, { "epoch": 1.9924705882352942, "grad_norm": 0.6633402301216967, "learning_rate": 0.00026915183813067264, "loss": 0.5547, "step": 4234 }, { "epoch": 1.9929411764705882, "grad_norm": 0.48144220346445177, "learning_rate": 0.0002691252040508623, "loss": 0.2393, "step": 4235 }, { "epoch": 1.9934117647058822, "grad_norm": 0.6504739795425377, "learning_rate": 0.0002690985597971753, "loss": 0.5664, "step": 4236 }, { "epoch": 1.9938823529411764, "grad_norm": 0.8497264915583647, "learning_rate": 0.00026907190537188707, "loss": 0.4668, "step": 4237 }, { "epoch": 1.9943529411764707, "grad_norm": 0.4465780405873704, "learning_rate": 0.0002690452407772741, "loss": 0.2715, "step": 4238 }, { "epoch": 1.9948235294117647, "grad_norm": 0.17658267603356584, "learning_rate": 0.0002690185660156136, "loss": 0.0598, "step": 4239 }, { "epoch": 1.9952941176470587, "grad_norm": 0.8000735599355846, "learning_rate": 0.00026899188108918387, "loss": 0.418, "step": 4240 }, { "epoch": 1.9957647058823529, "grad_norm": 0.25904040907766157, "learning_rate": 0.00026896518600026377, "loss": 0.0898, "step": 4241 }, { "epoch": 1.996235294117647, "grad_norm": 0.6030852165317331, "learning_rate": 0.00026893848075113327, "loss": 0.2969, "step": 4242 }, { "epoch": 1.9967058823529413, "grad_norm": 0.6223711678031595, "learning_rate": 0.00026891176534407315, "loss": 0.3652, "step": 4243 }, { "epoch": 1.9971764705882353, "grad_norm": 1.5156128155044253, "learning_rate": 0.00026888503978136506, "loss": 0.5273, "step": 4244 }, { "epoch": 1.9976470588235293, "grad_norm": 0.05779289375159494, "learning_rate": 0.00026885830406529136, "loss": 0.0072, "step": 4245 }, { "epoch": 1.9981176470588236, "grad_norm": 0.9653130814283235, "learning_rate": 0.0002688315581981355, "loss": 0.4297, "step": 4246 }, { "epoch": 1.9985882352941178, "grad_norm": 0.354313280962286, "learning_rate": 0.0002688048021821817, "loss": 0.1348, "step": 4247 }, { "epoch": 1.9990588235294118, "grad_norm": 0.43709298010379993, "learning_rate": 0.00026877803601971516, "loss": 0.126, "step": 4248 }, { "epoch": 1.9995294117647058, "grad_norm": 0.4979781435612921, "learning_rate": 0.00026875125971302167, "loss": 0.1777, "step": 4249 }, { "epoch": 2.0, "grad_norm": 1.8427142505522496, "learning_rate": 0.0002687244732643881, "loss": 0.5234, "step": 4250 }, { "epoch": 2.0, "eval_loss": 0.3175669014453888, "eval_runtime": 61.679, "eval_samples_per_second": 12.16, "eval_steps_per_second": 6.08, "step": 4250 }, { "epoch": 2.0004705882352942, "grad_norm": 0.8320275929037299, "learning_rate": 0.0002686976766761022, "loss": 0.3828, "step": 4251 }, { "epoch": 2.0009411764705884, "grad_norm": 0.3106304355385026, "learning_rate": 0.0002686708699504525, "loss": 0.1064, "step": 4252 }, { "epoch": 2.0014117647058822, "grad_norm": 0.6586067868584167, "learning_rate": 0.0002686440530897284, "loss": 0.3633, "step": 4253 }, { "epoch": 2.0018823529411764, "grad_norm": 0.7303442653999923, "learning_rate": 0.00026861722609622027, "loss": 0.3184, "step": 4254 }, { "epoch": 2.0023529411764707, "grad_norm": 0.8513347343871455, "learning_rate": 0.0002685903889722192, "loss": 0.4473, "step": 4255 }, { "epoch": 2.002823529411765, "grad_norm": 0.365983783858796, "learning_rate": 0.00026856354172001724, "loss": 0.1553, "step": 4256 }, { "epoch": 2.0032941176470587, "grad_norm": 0.9789902302838233, "learning_rate": 0.0002685366843419073, "loss": 0.3262, "step": 4257 }, { "epoch": 2.003764705882353, "grad_norm": 0.8704939110787284, "learning_rate": 0.000268509816840183, "loss": 0.625, "step": 4258 }, { "epoch": 2.004235294117647, "grad_norm": 0.4659218887655477, "learning_rate": 0.00026848293921713913, "loss": 0.2256, "step": 4259 }, { "epoch": 2.0047058823529413, "grad_norm": 0.7377833618091157, "learning_rate": 0.0002684560514750711, "loss": 0.4062, "step": 4260 }, { "epoch": 2.005176470588235, "grad_norm": 0.5098204018993794, "learning_rate": 0.00026842915361627523, "loss": 0.2891, "step": 4261 }, { "epoch": 2.0056470588235293, "grad_norm": 0.6199885450406663, "learning_rate": 0.0002684022456430487, "loss": 0.3496, "step": 4262 }, { "epoch": 2.0061176470588236, "grad_norm": 0.291609570609489, "learning_rate": 0.0002683753275576897, "loss": 0.1582, "step": 4263 }, { "epoch": 2.006588235294118, "grad_norm": 0.5352753803023074, "learning_rate": 0.0002683483993624971, "loss": 0.3789, "step": 4264 }, { "epoch": 2.0070588235294116, "grad_norm": 0.3357470326079738, "learning_rate": 0.0002683214610597706, "loss": 0.063, "step": 4265 }, { "epoch": 2.007529411764706, "grad_norm": 0.48860828007436724, "learning_rate": 0.0002682945126518111, "loss": 0.1699, "step": 4266 }, { "epoch": 2.008, "grad_norm": 0.5658344618496152, "learning_rate": 0.00026826755414091993, "loss": 0.4688, "step": 4267 }, { "epoch": 2.0084705882352942, "grad_norm": 0.776937790705901, "learning_rate": 0.00026824058552939954, "loss": 0.377, "step": 4268 }, { "epoch": 2.0089411764705885, "grad_norm": 0.8436845291055466, "learning_rate": 0.00026821360681955316, "loss": 0.5234, "step": 4269 }, { "epoch": 2.0094117647058822, "grad_norm": 0.5807263353348123, "learning_rate": 0.00026818661801368494, "loss": 0.3125, "step": 4270 }, { "epoch": 2.0098823529411765, "grad_norm": 0.20496991601113074, "learning_rate": 0.0002681596191140998, "loss": 0.0757, "step": 4271 }, { "epoch": 2.0103529411764707, "grad_norm": 0.6001469371101925, "learning_rate": 0.00026813261012310365, "loss": 0.2676, "step": 4272 }, { "epoch": 2.010823529411765, "grad_norm": 1.0327064749312496, "learning_rate": 0.00026810559104300315, "loss": 0.3809, "step": 4273 }, { "epoch": 2.0112941176470587, "grad_norm": 0.25709095083910394, "learning_rate": 0.0002680785618761059, "loss": 0.0688, "step": 4274 }, { "epoch": 2.011764705882353, "grad_norm": 0.24262840768922234, "learning_rate": 0.00026805152262472025, "loss": 0.1201, "step": 4275 }, { "epoch": 2.012235294117647, "grad_norm": 0.5389431899077448, "learning_rate": 0.00026802447329115557, "loss": 0.375, "step": 4276 }, { "epoch": 2.0127058823529413, "grad_norm": 0.19405012668411606, "learning_rate": 0.00026799741387772193, "loss": 0.0942, "step": 4277 }, { "epoch": 2.013176470588235, "grad_norm": 0.9953168183495427, "learning_rate": 0.0002679703443867304, "loss": 0.4062, "step": 4278 }, { "epoch": 2.0136470588235293, "grad_norm": 0.4561872653689488, "learning_rate": 0.0002679432648204928, "loss": 0.2461, "step": 4279 }, { "epoch": 2.0141176470588236, "grad_norm": 0.6820852681875939, "learning_rate": 0.00026791617518132186, "loss": 0.3281, "step": 4280 }, { "epoch": 2.014588235294118, "grad_norm": 0.6302162240232356, "learning_rate": 0.00026788907547153126, "loss": 0.21, "step": 4281 }, { "epoch": 2.0150588235294116, "grad_norm": 0.5228841518026071, "learning_rate": 0.00026786196569343525, "loss": 0.2852, "step": 4282 }, { "epoch": 2.015529411764706, "grad_norm": 0.2227670555939001, "learning_rate": 0.0002678348458493494, "loss": 0.0796, "step": 4283 }, { "epoch": 2.016, "grad_norm": 0.5760142935078132, "learning_rate": 0.00026780771594158966, "loss": 0.252, "step": 4284 }, { "epoch": 2.0164705882352942, "grad_norm": 0.22258282988680708, "learning_rate": 0.0002677805759724731, "loss": 0.0791, "step": 4285 }, { "epoch": 2.016941176470588, "grad_norm": 0.642333772965735, "learning_rate": 0.0002677534259443177, "loss": 0.2969, "step": 4286 }, { "epoch": 2.0174117647058822, "grad_norm": 0.7792179096160876, "learning_rate": 0.00026772626585944213, "loss": 0.4434, "step": 4287 }, { "epoch": 2.0178823529411765, "grad_norm": 0.4552380113331904, "learning_rate": 0.00026769909572016605, "loss": 0.2051, "step": 4288 }, { "epoch": 2.0183529411764707, "grad_norm": 0.2578189495323732, "learning_rate": 0.0002676719155288098, "loss": 0.0859, "step": 4289 }, { "epoch": 2.018823529411765, "grad_norm": 0.5124193679843935, "learning_rate": 0.00026764472528769485, "loss": 0.2422, "step": 4290 }, { "epoch": 2.0192941176470587, "grad_norm": 0.6427644311862184, "learning_rate": 0.0002676175249991433, "loss": 0.3262, "step": 4291 }, { "epoch": 2.019764705882353, "grad_norm": 0.27267074216527526, "learning_rate": 0.0002675903146654782, "loss": 0.1172, "step": 4292 }, { "epoch": 2.020235294117647, "grad_norm": 0.5305317552894295, "learning_rate": 0.00026756309428902343, "loss": 0.1465, "step": 4293 }, { "epoch": 2.0207058823529414, "grad_norm": 1.0577191176260703, "learning_rate": 0.0002675358638721038, "loss": 0.4316, "step": 4294 }, { "epoch": 2.021176470588235, "grad_norm": 1.928631005078379, "learning_rate": 0.0002675086234170449, "loss": 0.2344, "step": 4295 }, { "epoch": 2.0216470588235294, "grad_norm": 0.4883294480194424, "learning_rate": 0.0002674813729261731, "loss": 0.1572, "step": 4296 }, { "epoch": 2.0221176470588236, "grad_norm": 0.4472668012637688, "learning_rate": 0.0002674541124018159, "loss": 0.2305, "step": 4297 }, { "epoch": 2.022588235294118, "grad_norm": 0.2784737769102859, "learning_rate": 0.00026742684184630135, "loss": 0.1377, "step": 4298 }, { "epoch": 2.0230588235294116, "grad_norm": 0.2253185407415786, "learning_rate": 0.00026739956126195855, "loss": 0.0645, "step": 4299 }, { "epoch": 2.023529411764706, "grad_norm": 0.5881847548464965, "learning_rate": 0.0002673722706511174, "loss": 0.126, "step": 4300 }, { "epoch": 2.024, "grad_norm": 1.5493967072334338, "learning_rate": 0.00026734497001610854, "loss": 0.5898, "step": 4301 }, { "epoch": 2.0244705882352942, "grad_norm": 0.972640912402038, "learning_rate": 0.00026731765935926374, "loss": 0.4512, "step": 4302 }, { "epoch": 2.024941176470588, "grad_norm": 0.3071190140206751, "learning_rate": 0.00026729033868291537, "loss": 0.1963, "step": 4303 }, { "epoch": 2.0254117647058822, "grad_norm": 0.38816020802570383, "learning_rate": 0.00026726300798939675, "loss": 0.2119, "step": 4304 }, { "epoch": 2.0258823529411765, "grad_norm": 0.7707282179713657, "learning_rate": 0.0002672356672810421, "loss": 0.2031, "step": 4305 }, { "epoch": 2.0263529411764707, "grad_norm": 0.6852738358925002, "learning_rate": 0.00026720831656018644, "loss": 0.3457, "step": 4306 }, { "epoch": 2.026823529411765, "grad_norm": 0.34226671791354857, "learning_rate": 0.0002671809558291656, "loss": 0.0522, "step": 4307 }, { "epoch": 2.0272941176470587, "grad_norm": 0.22487469892778766, "learning_rate": 0.00026715358509031643, "loss": 0.0588, "step": 4308 }, { "epoch": 2.027764705882353, "grad_norm": 0.5492464589643016, "learning_rate": 0.00026712620434597636, "loss": 0.2451, "step": 4309 }, { "epoch": 2.028235294117647, "grad_norm": 0.7626578292876429, "learning_rate": 0.00026709881359848406, "loss": 0.3516, "step": 4310 }, { "epoch": 2.0287058823529414, "grad_norm": 0.8629875049149519, "learning_rate": 0.0002670714128501786, "loss": 0.4883, "step": 4311 }, { "epoch": 2.029176470588235, "grad_norm": 0.5791914235950862, "learning_rate": 0.0002670440021034003, "loss": 0.3516, "step": 4312 }, { "epoch": 2.0296470588235294, "grad_norm": 0.31896129587864896, "learning_rate": 0.00026701658136049015, "loss": 0.1914, "step": 4313 }, { "epoch": 2.0301176470588236, "grad_norm": 0.6423098460811862, "learning_rate": 0.00026698915062378994, "loss": 0.3359, "step": 4314 }, { "epoch": 2.030588235294118, "grad_norm": 1.649212494909248, "learning_rate": 0.00026696170989564245, "loss": 1.0312, "step": 4315 }, { "epoch": 2.0310588235294116, "grad_norm": 0.5741985006803473, "learning_rate": 0.0002669342591783913, "loss": 0.293, "step": 4316 }, { "epoch": 2.031529411764706, "grad_norm": 0.6953514633704793, "learning_rate": 0.00026690679847438077, "loss": 0.3281, "step": 4317 }, { "epoch": 2.032, "grad_norm": 0.2725549180550329, "learning_rate": 0.0002668793277859563, "loss": 0.0889, "step": 4318 }, { "epoch": 2.0324705882352943, "grad_norm": 0.7096479302535291, "learning_rate": 0.0002668518471154639, "loss": 0.2793, "step": 4319 }, { "epoch": 2.032941176470588, "grad_norm": 1.5949327347762916, "learning_rate": 0.0002668243564652506, "loss": 0.6328, "step": 4320 }, { "epoch": 2.0334117647058823, "grad_norm": 0.7143645621566991, "learning_rate": 0.0002667968558376642, "loss": 0.4395, "step": 4321 }, { "epoch": 2.0338823529411765, "grad_norm": 3.1407268757310463, "learning_rate": 0.0002667693452350535, "loss": 0.6484, "step": 4322 }, { "epoch": 2.0343529411764707, "grad_norm": 0.7868992525179156, "learning_rate": 0.0002667418246597679, "loss": 0.3926, "step": 4323 }, { "epoch": 2.034823529411765, "grad_norm": 0.20232401164684793, "learning_rate": 0.0002667142941141579, "loss": 0.0425, "step": 4324 }, { "epoch": 2.0352941176470587, "grad_norm": 0.7944908548701184, "learning_rate": 0.0002666867536005747, "loss": 0.416, "step": 4325 }, { "epoch": 2.035764705882353, "grad_norm": 0.4645955723141125, "learning_rate": 0.00026665920312137045, "loss": 0.1826, "step": 4326 }, { "epoch": 2.036235294117647, "grad_norm": 1.0221209687771253, "learning_rate": 0.000266631642678898, "loss": 0.3965, "step": 4327 }, { "epoch": 2.0367058823529414, "grad_norm": 0.7671317107299694, "learning_rate": 0.00026660407227551113, "loss": 0.3477, "step": 4328 }, { "epoch": 2.037176470588235, "grad_norm": 0.4279775111760865, "learning_rate": 0.0002665764919135646, "loss": 0.3008, "step": 4329 }, { "epoch": 2.0376470588235294, "grad_norm": 0.35933529128904756, "learning_rate": 0.00026654890159541386, "loss": 0.1982, "step": 4330 }, { "epoch": 2.0381176470588236, "grad_norm": 1.6607987196314162, "learning_rate": 0.0002665213013234153, "loss": 0.7891, "step": 4331 }, { "epoch": 2.038588235294118, "grad_norm": 0.6111250473401995, "learning_rate": 0.00026649369109992596, "loss": 0.208, "step": 4332 }, { "epoch": 2.0390588235294116, "grad_norm": 0.2894217559317421, "learning_rate": 0.0002664660709273041, "loss": 0.0845, "step": 4333 }, { "epoch": 2.039529411764706, "grad_norm": 0.242094090686814, "learning_rate": 0.0002664384408079085, "loss": 0.0591, "step": 4334 }, { "epoch": 2.04, "grad_norm": 0.4024195561427554, "learning_rate": 0.0002664108007440989, "loss": 0.106, "step": 4335 }, { "epoch": 2.0404705882352943, "grad_norm": 0.4604606220816121, "learning_rate": 0.00026638315073823594, "loss": 0.125, "step": 4336 }, { "epoch": 2.040941176470588, "grad_norm": 0.3480245268902986, "learning_rate": 0.00026635549079268104, "loss": 0.167, "step": 4337 }, { "epoch": 2.0414117647058823, "grad_norm": 0.8460826088954516, "learning_rate": 0.00026632782090979653, "loss": 0.3574, "step": 4338 }, { "epoch": 2.0418823529411765, "grad_norm": 1.3058099357189208, "learning_rate": 0.0002663001410919455, "loss": 0.3047, "step": 4339 }, { "epoch": 2.0423529411764707, "grad_norm": 0.39861443968809485, "learning_rate": 0.000266272451341492, "loss": 0.126, "step": 4340 }, { "epoch": 2.0428235294117645, "grad_norm": 0.6060213436251883, "learning_rate": 0.0002662447516608008, "loss": 0.1973, "step": 4341 }, { "epoch": 2.0432941176470587, "grad_norm": 0.8908622656191993, "learning_rate": 0.0002662170420522377, "loss": 0.4629, "step": 4342 }, { "epoch": 2.043764705882353, "grad_norm": 0.4711882924345044, "learning_rate": 0.0002661893225181691, "loss": 0.209, "step": 4343 }, { "epoch": 2.044235294117647, "grad_norm": 0.7724310712408982, "learning_rate": 0.0002661615930609625, "loss": 0.3574, "step": 4344 }, { "epoch": 2.0447058823529414, "grad_norm": 0.5891705400186736, "learning_rate": 0.000266133853682986, "loss": 0.2832, "step": 4345 }, { "epoch": 2.045176470588235, "grad_norm": 0.6850549891531619, "learning_rate": 0.0002661061043866089, "loss": 0.2773, "step": 4346 }, { "epoch": 2.0456470588235294, "grad_norm": 0.897311252994932, "learning_rate": 0.00026607834517420085, "loss": 0.3105, "step": 4347 }, { "epoch": 2.0461176470588236, "grad_norm": 0.6109065003119837, "learning_rate": 0.00026605057604813285, "loss": 0.1953, "step": 4348 }, { "epoch": 2.046588235294118, "grad_norm": 0.4904714895860333, "learning_rate": 0.00026602279701077637, "loss": 0.1738, "step": 4349 }, { "epoch": 2.0470588235294116, "grad_norm": 0.9139762239640978, "learning_rate": 0.000265995008064504, "loss": 0.4336, "step": 4350 }, { "epoch": 2.047529411764706, "grad_norm": 0.6805753795289322, "learning_rate": 0.00026596720921168897, "loss": 0.2041, "step": 4351 }, { "epoch": 2.048, "grad_norm": 0.8737740102109165, "learning_rate": 0.00026593940045470545, "loss": 0.625, "step": 4352 }, { "epoch": 2.0484705882352943, "grad_norm": 0.5595856668184119, "learning_rate": 0.00026591158179592846, "loss": 0.2402, "step": 4353 }, { "epoch": 2.048941176470588, "grad_norm": 0.4246320299919359, "learning_rate": 0.0002658837532377338, "loss": 0.1582, "step": 4354 }, { "epoch": 2.0494117647058823, "grad_norm": 0.5170933566420168, "learning_rate": 0.00026585591478249824, "loss": 0.1572, "step": 4355 }, { "epoch": 2.0498823529411765, "grad_norm": 1.4546715926554477, "learning_rate": 0.0002658280664325993, "loss": 0.6719, "step": 4356 }, { "epoch": 2.0503529411764707, "grad_norm": 0.7716614261034735, "learning_rate": 0.00026580020819041533, "loss": 0.459, "step": 4357 }, { "epoch": 2.0508235294117645, "grad_norm": 0.3795781889390534, "learning_rate": 0.0002657723400583256, "loss": 0.1895, "step": 4358 }, { "epoch": 2.0512941176470587, "grad_norm": 0.2526074609515635, "learning_rate": 0.00026574446203871013, "loss": 0.0435, "step": 4359 }, { "epoch": 2.051764705882353, "grad_norm": 1.1133230727323666, "learning_rate": 0.0002657165741339499, "loss": 0.4824, "step": 4360 }, { "epoch": 2.052235294117647, "grad_norm": 2.4588122399896695, "learning_rate": 0.00026568867634642665, "loss": 0.7734, "step": 4361 }, { "epoch": 2.0527058823529414, "grad_norm": 0.6393528411165551, "learning_rate": 0.00026566076867852304, "loss": 0.248, "step": 4362 }, { "epoch": 2.053176470588235, "grad_norm": 0.047008974721623104, "learning_rate": 0.0002656328511326224, "loss": 0.0062, "step": 4363 }, { "epoch": 2.0536470588235294, "grad_norm": 1.0302681795075446, "learning_rate": 0.00026560492371110913, "loss": 0.3066, "step": 4364 }, { "epoch": 2.0541176470588236, "grad_norm": 0.04424327485323875, "learning_rate": 0.00026557698641636835, "loss": 0.0051, "step": 4365 }, { "epoch": 2.054588235294118, "grad_norm": 0.7831391568757388, "learning_rate": 0.000265549039250786, "loss": 0.3105, "step": 4366 }, { "epoch": 2.0550588235294116, "grad_norm": 0.9029179459311518, "learning_rate": 0.000265521082216749, "loss": 0.1953, "step": 4367 }, { "epoch": 2.055529411764706, "grad_norm": 1.7866640916190197, "learning_rate": 0.0002654931153166449, "loss": 0.4512, "step": 4368 }, { "epoch": 2.056, "grad_norm": 0.1395850807345972, "learning_rate": 0.0002654651385528622, "loss": 0.0417, "step": 4369 }, { "epoch": 2.0564705882352943, "grad_norm": 0.4414290309834292, "learning_rate": 0.00026543715192779043, "loss": 0.2402, "step": 4370 }, { "epoch": 2.056941176470588, "grad_norm": 0.7484377434434545, "learning_rate": 0.0002654091554438196, "loss": 0.457, "step": 4371 }, { "epoch": 2.0574117647058823, "grad_norm": 0.8004865062049494, "learning_rate": 0.0002653811491033408, "loss": 0.6055, "step": 4372 }, { "epoch": 2.0578823529411765, "grad_norm": 0.3198840777879755, "learning_rate": 0.00026535313290874593, "loss": 0.0898, "step": 4373 }, { "epoch": 2.0583529411764707, "grad_norm": 0.7777836768536661, "learning_rate": 0.0002653251068624278, "loss": 0.2891, "step": 4374 }, { "epoch": 2.0588235294117645, "grad_norm": 1.6214128600769542, "learning_rate": 0.00026529707096677977, "loss": 0.5859, "step": 4375 }, { "epoch": 2.0592941176470587, "grad_norm": 0.4153612761854799, "learning_rate": 0.00026526902522419637, "loss": 0.2129, "step": 4376 }, { "epoch": 2.059764705882353, "grad_norm": 0.5351217210000311, "learning_rate": 0.0002652409696370729, "loss": 0.1895, "step": 4377 }, { "epoch": 2.060235294117647, "grad_norm": 0.43006155485202374, "learning_rate": 0.0002652129042078053, "loss": 0.1826, "step": 4378 }, { "epoch": 2.0607058823529414, "grad_norm": 0.9988860909831699, "learning_rate": 0.00026518482893879056, "loss": 0.5703, "step": 4379 }, { "epoch": 2.061176470588235, "grad_norm": 0.37169251510034435, "learning_rate": 0.0002651567438324265, "loss": 0.1514, "step": 4380 }, { "epoch": 2.0616470588235294, "grad_norm": 1.9858760777048776, "learning_rate": 0.00026512864889111163, "loss": 0.7422, "step": 4381 }, { "epoch": 2.0621176470588236, "grad_norm": 0.8019414061535947, "learning_rate": 0.00026510054411724545, "loss": 0.4434, "step": 4382 }, { "epoch": 2.062588235294118, "grad_norm": 0.7425242522380046, "learning_rate": 0.00026507242951322825, "loss": 0.4023, "step": 4383 }, { "epoch": 2.0630588235294116, "grad_norm": 0.6350952199126365, "learning_rate": 0.00026504430508146116, "loss": 0.3516, "step": 4384 }, { "epoch": 2.063529411764706, "grad_norm": 0.6713090936407171, "learning_rate": 0.00026501617082434607, "loss": 0.2871, "step": 4385 }, { "epoch": 2.064, "grad_norm": 1.1099177633553872, "learning_rate": 0.0002649880267442859, "loss": 0.3574, "step": 4386 }, { "epoch": 2.0644705882352943, "grad_norm": 0.9235670858225079, "learning_rate": 0.00026495987284368415, "loss": 0.4023, "step": 4387 }, { "epoch": 2.064941176470588, "grad_norm": 0.7596962590416075, "learning_rate": 0.0002649317091249455, "loss": 0.4043, "step": 4388 }, { "epoch": 2.0654117647058823, "grad_norm": 0.5474734118953172, "learning_rate": 0.00026490353559047505, "loss": 0.416, "step": 4389 }, { "epoch": 2.0658823529411765, "grad_norm": 0.6519287250316945, "learning_rate": 0.0002648753522426791, "loss": 0.2168, "step": 4390 }, { "epoch": 2.0663529411764707, "grad_norm": 0.6306228869573557, "learning_rate": 0.0002648471590839646, "loss": 0.4141, "step": 4391 }, { "epoch": 2.0668235294117645, "grad_norm": 0.134838583828882, "learning_rate": 0.00026481895611673944, "loss": 0.0271, "step": 4392 }, { "epoch": 2.0672941176470587, "grad_norm": 1.2406736990943656, "learning_rate": 0.0002647907433434122, "loss": 0.8398, "step": 4393 }, { "epoch": 2.067764705882353, "grad_norm": 0.49943324053011307, "learning_rate": 0.00026476252076639245, "loss": 0.1797, "step": 4394 }, { "epoch": 2.068235294117647, "grad_norm": 0.23529829031782162, "learning_rate": 0.0002647342883880905, "loss": 0.053, "step": 4395 }, { "epoch": 2.0687058823529414, "grad_norm": 0.6652469372934053, "learning_rate": 0.0002647060462109176, "loss": 0.3711, "step": 4396 }, { "epoch": 2.069176470588235, "grad_norm": 1.323454584570122, "learning_rate": 0.00026467779423728564, "loss": 0.5742, "step": 4397 }, { "epoch": 2.0696470588235294, "grad_norm": 0.6370905936624864, "learning_rate": 0.0002646495324696076, "loss": 0.3633, "step": 4398 }, { "epoch": 2.0701176470588236, "grad_norm": 0.38178841095476246, "learning_rate": 0.0002646212609102972, "loss": 0.1729, "step": 4399 }, { "epoch": 2.070588235294118, "grad_norm": 0.3668848253328038, "learning_rate": 0.00026459297956176885, "loss": 0.207, "step": 4400 }, { "epoch": 2.0710588235294116, "grad_norm": 0.8762296346516614, "learning_rate": 0.000264564688426438, "loss": 0.2393, "step": 4401 }, { "epoch": 2.071529411764706, "grad_norm": 0.3548065185330906, "learning_rate": 0.0002645363875067208, "loss": 0.0889, "step": 4402 }, { "epoch": 2.072, "grad_norm": 0.7988029857336995, "learning_rate": 0.0002645080768050342, "loss": 0.3047, "step": 4403 }, { "epoch": 2.0724705882352943, "grad_norm": 1.0424678245531038, "learning_rate": 0.00026447975632379637, "loss": 0.5117, "step": 4404 }, { "epoch": 2.072941176470588, "grad_norm": 0.39590496496928107, "learning_rate": 0.0002644514260654257, "loss": 0.1206, "step": 4405 }, { "epoch": 2.0734117647058823, "grad_norm": 0.5462700725143714, "learning_rate": 0.0002644230860323419, "loss": 0.543, "step": 4406 }, { "epoch": 2.0738823529411765, "grad_norm": 0.46613892260982354, "learning_rate": 0.00026439473622696534, "loss": 0.1768, "step": 4407 }, { "epoch": 2.0743529411764707, "grad_norm": 0.5851007473361696, "learning_rate": 0.0002643663766517172, "loss": 0.1787, "step": 4408 }, { "epoch": 2.0748235294117645, "grad_norm": 0.44858779306588736, "learning_rate": 0.00026433800730901947, "loss": 0.1709, "step": 4409 }, { "epoch": 2.0752941176470587, "grad_norm": 0.6576041249432074, "learning_rate": 0.00026430962820129513, "loss": 0.3105, "step": 4410 }, { "epoch": 2.075764705882353, "grad_norm": 0.888434980831216, "learning_rate": 0.00026428123933096786, "loss": 0.4492, "step": 4411 }, { "epoch": 2.076235294117647, "grad_norm": 0.6997687189941643, "learning_rate": 0.0002642528407004622, "loss": 0.3047, "step": 4412 }, { "epoch": 2.0767058823529414, "grad_norm": 0.4170611938113522, "learning_rate": 0.00026422443231220353, "loss": 0.168, "step": 4413 }, { "epoch": 2.077176470588235, "grad_norm": 0.9317057137804267, "learning_rate": 0.0002641960141686181, "loss": 0.6484, "step": 4414 }, { "epoch": 2.0776470588235294, "grad_norm": 1.903906561624605, "learning_rate": 0.0002641675862721329, "loss": 0.6992, "step": 4415 }, { "epoch": 2.0781176470588236, "grad_norm": 1.0057208248054856, "learning_rate": 0.00026413914862517585, "loss": 0.7344, "step": 4416 }, { "epoch": 2.078588235294118, "grad_norm": 0.08877775462348769, "learning_rate": 0.00026411070123017567, "loss": 0.0179, "step": 4417 }, { "epoch": 2.0790588235294116, "grad_norm": 0.3700053012633482, "learning_rate": 0.0002640822440895619, "loss": 0.1885, "step": 4418 }, { "epoch": 2.079529411764706, "grad_norm": 0.655407987669269, "learning_rate": 0.0002640537772057649, "loss": 0.3477, "step": 4419 }, { "epoch": 2.08, "grad_norm": 0.8359684568574518, "learning_rate": 0.0002640253005812159, "loss": 0.3574, "step": 4420 }, { "epoch": 2.0804705882352943, "grad_norm": 1.35454929268078, "learning_rate": 0.00026399681421834694, "loss": 0.75, "step": 4421 }, { "epoch": 2.080941176470588, "grad_norm": 0.52983296461903, "learning_rate": 0.0002639683181195909, "loss": 0.2344, "step": 4422 }, { "epoch": 2.0814117647058823, "grad_norm": 0.7947429600581367, "learning_rate": 0.00026393981228738144, "loss": 0.4102, "step": 4423 }, { "epoch": 2.0818823529411765, "grad_norm": 0.48334046050628743, "learning_rate": 0.0002639112967241532, "loss": 0.2363, "step": 4424 }, { "epoch": 2.0823529411764707, "grad_norm": 0.44829636647289983, "learning_rate": 0.00026388277143234146, "loss": 0.1143, "step": 4425 }, { "epoch": 2.0828235294117645, "grad_norm": 0.5347579368369387, "learning_rate": 0.0002638542364143825, "loss": 0.2295, "step": 4426 }, { "epoch": 2.0832941176470587, "grad_norm": 0.555091189454891, "learning_rate": 0.00026382569167271317, "loss": 0.4023, "step": 4427 }, { "epoch": 2.083764705882353, "grad_norm": 0.6041740305237255, "learning_rate": 0.00026379713720977156, "loss": 0.2139, "step": 4428 }, { "epoch": 2.084235294117647, "grad_norm": 0.3101482721725143, "learning_rate": 0.0002637685730279962, "loss": 0.1289, "step": 4429 }, { "epoch": 2.084705882352941, "grad_norm": 0.6614702975433389, "learning_rate": 0.00026373999912982676, "loss": 0.3945, "step": 4430 }, { "epoch": 2.085176470588235, "grad_norm": 0.32369468479401625, "learning_rate": 0.00026371141551770344, "loss": 0.1836, "step": 4431 }, { "epoch": 2.0856470588235294, "grad_norm": 0.3110450884970438, "learning_rate": 0.0002636828221940675, "loss": 0.1113, "step": 4432 }, { "epoch": 2.0861176470588236, "grad_norm": 0.7144967708244496, "learning_rate": 0.00026365421916136093, "loss": 0.4941, "step": 4433 }, { "epoch": 2.086588235294118, "grad_norm": 0.9510885671603797, "learning_rate": 0.00026362560642202655, "loss": 0.4199, "step": 4434 }, { "epoch": 2.0870588235294116, "grad_norm": 0.5272763736574855, "learning_rate": 0.0002635969839785081, "loss": 0.3301, "step": 4435 }, { "epoch": 2.087529411764706, "grad_norm": 0.7348850533664014, "learning_rate": 0.00026356835183325, "loss": 0.4785, "step": 4436 }, { "epoch": 2.088, "grad_norm": 0.31459260765984376, "learning_rate": 0.00026353970998869763, "loss": 0.1387, "step": 4437 }, { "epoch": 2.0884705882352943, "grad_norm": 0.8581644725604585, "learning_rate": 0.0002635110584472971, "loss": 0.5039, "step": 4438 }, { "epoch": 2.088941176470588, "grad_norm": 0.2120316012817219, "learning_rate": 0.00026348239721149535, "loss": 0.0427, "step": 4439 }, { "epoch": 2.0894117647058823, "grad_norm": 0.5796915825453435, "learning_rate": 0.0002634537262837403, "loss": 0.3027, "step": 4440 }, { "epoch": 2.0898823529411765, "grad_norm": 0.29812707357030577, "learning_rate": 0.00026342504566648055, "loss": 0.1079, "step": 4441 }, { "epoch": 2.0903529411764707, "grad_norm": 0.5856969972052861, "learning_rate": 0.0002633963553621655, "loss": 0.3066, "step": 4442 }, { "epoch": 2.0908235294117645, "grad_norm": 0.43856121429743533, "learning_rate": 0.0002633676553732456, "loss": 0.1543, "step": 4443 }, { "epoch": 2.0912941176470587, "grad_norm": 0.502085038617022, "learning_rate": 0.0002633389457021718, "loss": 0.2109, "step": 4444 }, { "epoch": 2.091764705882353, "grad_norm": 0.40108710778567025, "learning_rate": 0.0002633102263513961, "loss": 0.4238, "step": 4445 }, { "epoch": 2.092235294117647, "grad_norm": 1.053257203030967, "learning_rate": 0.0002632814973233713, "loss": 0.7344, "step": 4446 }, { "epoch": 2.092705882352941, "grad_norm": 0.7858368226635274, "learning_rate": 0.00026325275862055104, "loss": 0.4512, "step": 4447 }, { "epoch": 2.093176470588235, "grad_norm": 0.38301188645800205, "learning_rate": 0.0002632240102453896, "loss": 0.104, "step": 4448 }, { "epoch": 2.0936470588235294, "grad_norm": 0.662781689372486, "learning_rate": 0.0002631952522003424, "loss": 0.3926, "step": 4449 }, { "epoch": 2.0941176470588236, "grad_norm": 0.9197915760879822, "learning_rate": 0.00026316648448786536, "loss": 0.4043, "step": 4450 }, { "epoch": 2.094588235294118, "grad_norm": 0.5436324535567783, "learning_rate": 0.00026313770711041556, "loss": 0.2734, "step": 4451 }, { "epoch": 2.0950588235294116, "grad_norm": 0.2085483115625084, "learning_rate": 0.0002631089200704506, "loss": 0.0898, "step": 4452 }, { "epoch": 2.095529411764706, "grad_norm": 0.7410575935433751, "learning_rate": 0.00026308012337042906, "loss": 0.3652, "step": 4453 }, { "epoch": 2.096, "grad_norm": 0.8267711167566011, "learning_rate": 0.0002630513170128103, "loss": 0.3613, "step": 4454 }, { "epoch": 2.0964705882352943, "grad_norm": 0.46279895017051537, "learning_rate": 0.0002630225010000546, "loss": 0.1885, "step": 4455 }, { "epoch": 2.096941176470588, "grad_norm": 0.3676316442262274, "learning_rate": 0.0002629936753346229, "loss": 0.1367, "step": 4456 }, { "epoch": 2.0974117647058823, "grad_norm": 0.7640088732293066, "learning_rate": 0.00026296484001897715, "loss": 0.3652, "step": 4457 }, { "epoch": 2.0978823529411765, "grad_norm": 0.514627863122915, "learning_rate": 0.00026293599505558, "loss": 0.2637, "step": 4458 }, { "epoch": 2.0983529411764708, "grad_norm": 0.7156265347372303, "learning_rate": 0.00026290714044689483, "loss": 0.4883, "step": 4459 }, { "epoch": 2.0988235294117645, "grad_norm": 0.2130313567937843, "learning_rate": 0.0002628782761953861, "loss": 0.0579, "step": 4460 }, { "epoch": 2.0992941176470588, "grad_norm": 1.0709514573678116, "learning_rate": 0.0002628494023035189, "loss": 0.5703, "step": 4461 }, { "epoch": 2.099764705882353, "grad_norm": 0.12274876081289735, "learning_rate": 0.00026282051877375923, "loss": 0.041, "step": 4462 }, { "epoch": 2.100235294117647, "grad_norm": 0.4918450930799125, "learning_rate": 0.0002627916256085739, "loss": 0.2324, "step": 4463 }, { "epoch": 2.100705882352941, "grad_norm": 0.36312619798225976, "learning_rate": 0.0002627627228104305, "loss": 0.1299, "step": 4464 }, { "epoch": 2.101176470588235, "grad_norm": 0.38347857666588775, "learning_rate": 0.00026273381038179744, "loss": 0.1138, "step": 4465 }, { "epoch": 2.1016470588235294, "grad_norm": 1.1620855294499848, "learning_rate": 0.000262704888325144, "loss": 0.543, "step": 4466 }, { "epoch": 2.1021176470588236, "grad_norm": 0.8963499921633558, "learning_rate": 0.0002626759566429404, "loss": 0.3945, "step": 4467 }, { "epoch": 2.1025882352941174, "grad_norm": 0.3967995931382431, "learning_rate": 0.0002626470153376573, "loss": 0.1138, "step": 4468 }, { "epoch": 2.1030588235294116, "grad_norm": 0.5570354201227695, "learning_rate": 0.00026261806441176666, "loss": 0.2871, "step": 4469 }, { "epoch": 2.103529411764706, "grad_norm": 0.750629965209041, "learning_rate": 0.0002625891038677408, "loss": 0.707, "step": 4470 }, { "epoch": 2.104, "grad_norm": 0.8728546358317452, "learning_rate": 0.0002625601337080533, "loss": 0.3828, "step": 4471 }, { "epoch": 2.1044705882352943, "grad_norm": 0.18919691233098215, "learning_rate": 0.00026253115393517834, "loss": 0.033, "step": 4472 }, { "epoch": 2.104941176470588, "grad_norm": 0.4514368078394034, "learning_rate": 0.0002625021645515908, "loss": 0.1963, "step": 4473 }, { "epoch": 2.1054117647058823, "grad_norm": 0.4872535954695474, "learning_rate": 0.00026247316555976657, "loss": 0.3086, "step": 4474 }, { "epoch": 2.1058823529411765, "grad_norm": 0.7262217690456692, "learning_rate": 0.00026244415696218237, "loss": 0.2754, "step": 4475 }, { "epoch": 2.1063529411764708, "grad_norm": 0.3028568204667117, "learning_rate": 0.00026241513876131564, "loss": 0.0977, "step": 4476 }, { "epoch": 2.1068235294117645, "grad_norm": 0.4239409047397617, "learning_rate": 0.00026238611095964466, "loss": 0.2314, "step": 4477 }, { "epoch": 2.1072941176470588, "grad_norm": 0.3633026022296357, "learning_rate": 0.0002623570735596485, "loss": 0.3008, "step": 4478 }, { "epoch": 2.107764705882353, "grad_norm": 0.6216497388613067, "learning_rate": 0.0002623280265638072, "loss": 0.3633, "step": 4479 }, { "epoch": 2.108235294117647, "grad_norm": 0.668142084825107, "learning_rate": 0.0002622989699746014, "loss": 0.2949, "step": 4480 }, { "epoch": 2.108705882352941, "grad_norm": 0.31724695547785353, "learning_rate": 0.00026226990379451284, "loss": 0.1611, "step": 4481 }, { "epoch": 2.109176470588235, "grad_norm": 0.8556880577000465, "learning_rate": 0.00026224082802602373, "loss": 0.3867, "step": 4482 }, { "epoch": 2.1096470588235294, "grad_norm": 0.3769277674166264, "learning_rate": 0.00026221174267161744, "loss": 0.1289, "step": 4483 }, { "epoch": 2.1101176470588237, "grad_norm": 0.28499170963886766, "learning_rate": 0.0002621826477337779, "loss": 0.0747, "step": 4484 }, { "epoch": 2.1105882352941174, "grad_norm": 0.5541558132468691, "learning_rate": 0.00026215354321498997, "loss": 0.2695, "step": 4485 }, { "epoch": 2.1110588235294117, "grad_norm": 0.31123365556761795, "learning_rate": 0.00026212442911773936, "loss": 0.207, "step": 4486 }, { "epoch": 2.111529411764706, "grad_norm": 0.7650942676675905, "learning_rate": 0.0002620953054445125, "loss": 0.4824, "step": 4487 }, { "epoch": 2.112, "grad_norm": 0.5432219190822737, "learning_rate": 0.0002620661721977968, "loss": 0.3164, "step": 4488 }, { "epoch": 2.1124705882352943, "grad_norm": 0.7080534731353697, "learning_rate": 0.00026203702938008025, "loss": 0.418, "step": 4489 }, { "epoch": 2.112941176470588, "grad_norm": 0.5786314561030036, "learning_rate": 0.00026200787699385186, "loss": 0.2812, "step": 4490 }, { "epoch": 2.1134117647058823, "grad_norm": 0.5221569878183993, "learning_rate": 0.0002619787150416014, "loss": 0.3789, "step": 4491 }, { "epoch": 2.1138823529411765, "grad_norm": 0.7760262243992951, "learning_rate": 0.00026194954352581944, "loss": 0.4453, "step": 4492 }, { "epoch": 2.1143529411764708, "grad_norm": 0.18186712013173192, "learning_rate": 0.0002619203624489973, "loss": 0.0786, "step": 4493 }, { "epoch": 2.1148235294117645, "grad_norm": 0.7180182456686001, "learning_rate": 0.00026189117181362733, "loss": 0.3945, "step": 4494 }, { "epoch": 2.1152941176470588, "grad_norm": 0.609525838014791, "learning_rate": 0.0002618619716222024, "loss": 0.3945, "step": 4495 }, { "epoch": 2.115764705882353, "grad_norm": 0.8568863225563251, "learning_rate": 0.00026183276187721635, "loss": 0.5938, "step": 4496 }, { "epoch": 2.116235294117647, "grad_norm": 0.5218066554018781, "learning_rate": 0.000261803542581164, "loss": 0.3047, "step": 4497 }, { "epoch": 2.116705882352941, "grad_norm": 0.6974465820615533, "learning_rate": 0.0002617743137365407, "loss": 0.4258, "step": 4498 }, { "epoch": 2.117176470588235, "grad_norm": 0.6864793595840449, "learning_rate": 0.0002617450753458427, "loss": 0.3418, "step": 4499 }, { "epoch": 2.1176470588235294, "grad_norm": 0.7233805545590319, "learning_rate": 0.00026171582741156725, "loss": 0.3164, "step": 4500 }, { "epoch": 2.1181176470588237, "grad_norm": 1.0686732393350733, "learning_rate": 0.0002616865699362121, "loss": 0.5391, "step": 4501 }, { "epoch": 2.1185882352941174, "grad_norm": 0.579871597381889, "learning_rate": 0.0002616573029222761, "loss": 0.3242, "step": 4502 }, { "epoch": 2.1190588235294117, "grad_norm": 0.230952549937822, "learning_rate": 0.0002616280263722588, "loss": 0.0845, "step": 4503 }, { "epoch": 2.119529411764706, "grad_norm": 0.7040627506293542, "learning_rate": 0.00026159874028866044, "loss": 0.3789, "step": 4504 }, { "epoch": 2.12, "grad_norm": 1.0277676373359623, "learning_rate": 0.0002615694446739823, "loss": 0.3926, "step": 4505 }, { "epoch": 2.1204705882352943, "grad_norm": 0.36954059636423214, "learning_rate": 0.0002615401395307263, "loss": 0.125, "step": 4506 }, { "epoch": 2.120941176470588, "grad_norm": 0.3132372944581515, "learning_rate": 0.00026151082486139534, "loss": 0.1157, "step": 4507 }, { "epoch": 2.1214117647058823, "grad_norm": 0.6873164450546774, "learning_rate": 0.0002614815006684929, "loss": 0.3262, "step": 4508 }, { "epoch": 2.1218823529411766, "grad_norm": 0.4923856728403684, "learning_rate": 0.0002614521669545236, "loss": 0.2314, "step": 4509 }, { "epoch": 2.1223529411764708, "grad_norm": 0.1655936819207218, "learning_rate": 0.0002614228237219925, "loss": 0.0498, "step": 4510 }, { "epoch": 2.1228235294117646, "grad_norm": 0.2392800814379388, "learning_rate": 0.00026139347097340574, "loss": 0.0991, "step": 4511 }, { "epoch": 2.1232941176470588, "grad_norm": 0.7290430103950397, "learning_rate": 0.00026136410871127017, "loss": 0.3594, "step": 4512 }, { "epoch": 2.123764705882353, "grad_norm": 0.5365293403207232, "learning_rate": 0.0002613347369380935, "loss": 0.3125, "step": 4513 }, { "epoch": 2.124235294117647, "grad_norm": 0.23708199563686916, "learning_rate": 0.00026130535565638415, "loss": 0.0898, "step": 4514 }, { "epoch": 2.124705882352941, "grad_norm": 0.920640237014037, "learning_rate": 0.00026127596486865155, "loss": 0.6602, "step": 4515 }, { "epoch": 2.1251764705882352, "grad_norm": 1.084266472203102, "learning_rate": 0.00026124656457740574, "loss": 0.3926, "step": 4516 }, { "epoch": 2.1256470588235294, "grad_norm": 0.47385405228973293, "learning_rate": 0.00026121715478515757, "loss": 0.3828, "step": 4517 }, { "epoch": 2.1261176470588237, "grad_norm": 0.697641396140033, "learning_rate": 0.0002611877354944189, "loss": 0.4336, "step": 4518 }, { "epoch": 2.1265882352941174, "grad_norm": 0.5330489333833021, "learning_rate": 0.00026115830670770225, "loss": 0.3438, "step": 4519 }, { "epoch": 2.1270588235294117, "grad_norm": 0.5493690626845187, "learning_rate": 0.00026112886842752104, "loss": 0.4316, "step": 4520 }, { "epoch": 2.127529411764706, "grad_norm": 0.43519034867546263, "learning_rate": 0.0002610994206563893, "loss": 0.2539, "step": 4521 }, { "epoch": 2.128, "grad_norm": 1.1915026549759162, "learning_rate": 0.0002610699633968222, "loss": 0.668, "step": 4522 }, { "epoch": 2.1284705882352943, "grad_norm": 0.13460106342257325, "learning_rate": 0.00026104049665133533, "loss": 0.063, "step": 4523 }, { "epoch": 2.128941176470588, "grad_norm": 0.3454172007025011, "learning_rate": 0.00026101102042244543, "loss": 0.1758, "step": 4524 }, { "epoch": 2.1294117647058823, "grad_norm": 0.5028015823836306, "learning_rate": 0.0002609815347126699, "loss": 0.1787, "step": 4525 }, { "epoch": 2.1298823529411766, "grad_norm": 0.44677290924185653, "learning_rate": 0.00026095203952452694, "loss": 0.1611, "step": 4526 }, { "epoch": 2.130352941176471, "grad_norm": 0.7189885844446572, "learning_rate": 0.00026092253486053555, "loss": 0.4141, "step": 4527 }, { "epoch": 2.1308235294117646, "grad_norm": 0.8640153342788428, "learning_rate": 0.0002608930207232156, "loss": 0.4531, "step": 4528 }, { "epoch": 2.131294117647059, "grad_norm": 0.3896566147772299, "learning_rate": 0.0002608634971150878, "loss": 0.1211, "step": 4529 }, { "epoch": 2.131764705882353, "grad_norm": 1.033378884773599, "learning_rate": 0.0002608339640386735, "loss": 0.334, "step": 4530 }, { "epoch": 2.1322352941176472, "grad_norm": 0.49914195497713126, "learning_rate": 0.00026080442149649516, "loss": 0.2344, "step": 4531 }, { "epoch": 2.132705882352941, "grad_norm": 0.29636022311116705, "learning_rate": 0.0002607748694910756, "loss": 0.0757, "step": 4532 }, { "epoch": 2.1331764705882352, "grad_norm": 0.6865785688402909, "learning_rate": 0.0002607453080249389, "loss": 0.2988, "step": 4533 }, { "epoch": 2.1336470588235295, "grad_norm": 0.7150954364961197, "learning_rate": 0.0002607157371006097, "loss": 0.3789, "step": 4534 }, { "epoch": 2.1341176470588237, "grad_norm": 0.8254068156121949, "learning_rate": 0.00026068615672061353, "loss": 0.4336, "step": 4535 }, { "epoch": 2.1345882352941175, "grad_norm": 0.7625357689440792, "learning_rate": 0.0002606565668874766, "loss": 0.4062, "step": 4536 }, { "epoch": 2.1350588235294117, "grad_norm": 0.2347794086033765, "learning_rate": 0.00026062696760372606, "loss": 0.1108, "step": 4537 }, { "epoch": 2.135529411764706, "grad_norm": 0.7129170527532083, "learning_rate": 0.0002605973588718899, "loss": 0.4258, "step": 4538 }, { "epoch": 2.136, "grad_norm": 0.3146218682330918, "learning_rate": 0.00026056774069449685, "loss": 0.252, "step": 4539 }, { "epoch": 2.1364705882352943, "grad_norm": 0.5579014075905012, "learning_rate": 0.00026053811307407635, "loss": 0.2373, "step": 4540 }, { "epoch": 2.136941176470588, "grad_norm": 0.33753572831315887, "learning_rate": 0.00026050847601315886, "loss": 0.1177, "step": 4541 }, { "epoch": 2.1374117647058823, "grad_norm": 0.725405335567999, "learning_rate": 0.00026047882951427543, "loss": 0.459, "step": 4542 }, { "epoch": 2.1378823529411766, "grad_norm": 0.5320091693458014, "learning_rate": 0.00026044917357995813, "loss": 0.3887, "step": 4543 }, { "epoch": 2.138352941176471, "grad_norm": 0.232312224625645, "learning_rate": 0.00026041950821273965, "loss": 0.1221, "step": 4544 }, { "epoch": 2.1388235294117646, "grad_norm": 0.9151678961496816, "learning_rate": 0.0002603898334151535, "loss": 0.5898, "step": 4545 }, { "epoch": 2.139294117647059, "grad_norm": 0.725565548455918, "learning_rate": 0.00026036014918973414, "loss": 0.4355, "step": 4546 }, { "epoch": 2.139764705882353, "grad_norm": 0.6446850788449624, "learning_rate": 0.00026033045553901674, "loss": 0.4336, "step": 4547 }, { "epoch": 2.1402352941176472, "grad_norm": 0.25518778142443754, "learning_rate": 0.0002603007524655373, "loss": 0.0723, "step": 4548 }, { "epoch": 2.140705882352941, "grad_norm": 0.6875715746545299, "learning_rate": 0.0002602710399718325, "loss": 0.3164, "step": 4549 }, { "epoch": 2.1411764705882352, "grad_norm": 0.8972144322867595, "learning_rate": 0.0002602413180604401, "loss": 0.668, "step": 4550 }, { "epoch": 2.1416470588235295, "grad_norm": 0.21650277197259113, "learning_rate": 0.0002602115867338983, "loss": 0.0884, "step": 4551 }, { "epoch": 2.1421176470588237, "grad_norm": 0.3922269301974963, "learning_rate": 0.0002601818459947465, "loss": 0.209, "step": 4552 }, { "epoch": 2.1425882352941175, "grad_norm": 0.7339127081944927, "learning_rate": 0.0002601520958455246, "loss": 0.3613, "step": 4553 }, { "epoch": 2.1430588235294117, "grad_norm": 0.6701944681705325, "learning_rate": 0.00026012233628877336, "loss": 0.3242, "step": 4554 }, { "epoch": 2.143529411764706, "grad_norm": 0.5918399387287264, "learning_rate": 0.0002600925673270345, "loss": 0.3848, "step": 4555 }, { "epoch": 2.144, "grad_norm": 0.8245983978262789, "learning_rate": 0.0002600627889628504, "loss": 0.4668, "step": 4556 }, { "epoch": 2.1444705882352944, "grad_norm": 0.5501476903552112, "learning_rate": 0.00026003300119876424, "loss": 0.3008, "step": 4557 }, { "epoch": 2.144941176470588, "grad_norm": 0.26364331178456063, "learning_rate": 0.0002600032040373201, "loss": 0.1426, "step": 4558 }, { "epoch": 2.1454117647058824, "grad_norm": 0.37043638576554655, "learning_rate": 0.00025997339748106275, "loss": 0.2598, "step": 4559 }, { "epoch": 2.1458823529411766, "grad_norm": 0.5796352279635016, "learning_rate": 0.00025994358153253784, "loss": 0.1562, "step": 4560 }, { "epoch": 2.1463529411764704, "grad_norm": 1.0118660364529473, "learning_rate": 0.0002599137561942918, "loss": 0.6094, "step": 4561 }, { "epoch": 2.1468235294117646, "grad_norm": 0.47181302712514, "learning_rate": 0.00025988392146887183, "loss": 0.1973, "step": 4562 }, { "epoch": 2.147294117647059, "grad_norm": 1.0284356036309292, "learning_rate": 0.00025985407735882606, "loss": 0.6484, "step": 4563 }, { "epoch": 2.147764705882353, "grad_norm": 0.4766271205993331, "learning_rate": 0.0002598242238667032, "loss": 0.2432, "step": 4564 }, { "epoch": 2.1482352941176472, "grad_norm": 0.4821148318732412, "learning_rate": 0.00025979436099505296, "loss": 0.1177, "step": 4565 }, { "epoch": 2.148705882352941, "grad_norm": 0.3984891787011641, "learning_rate": 0.0002597644887464257, "loss": 0.1729, "step": 4566 }, { "epoch": 2.1491764705882352, "grad_norm": 0.5758003410474078, "learning_rate": 0.00025973460712337275, "loss": 0.2129, "step": 4567 }, { "epoch": 2.1496470588235295, "grad_norm": 1.0729361027698578, "learning_rate": 0.0002597047161284461, "loss": 0.8086, "step": 4568 }, { "epoch": 2.1501176470588237, "grad_norm": 0.5882579283959066, "learning_rate": 0.00025967481576419864, "loss": 0.3672, "step": 4569 }, { "epoch": 2.1505882352941175, "grad_norm": 0.8149571306236061, "learning_rate": 0.0002596449060331839, "loss": 0.6133, "step": 4570 }, { "epoch": 2.1510588235294117, "grad_norm": 0.6394962728629808, "learning_rate": 0.0002596149869379564, "loss": 0.3008, "step": 4571 }, { "epoch": 2.151529411764706, "grad_norm": 0.16693230579408141, "learning_rate": 0.00025958505848107133, "loss": 0.0874, "step": 4572 }, { "epoch": 2.152, "grad_norm": 0.6802448501812529, "learning_rate": 0.00025955512066508477, "loss": 0.291, "step": 4573 }, { "epoch": 2.152470588235294, "grad_norm": 0.9361203683025668, "learning_rate": 0.0002595251734925536, "loss": 0.4512, "step": 4574 }, { "epoch": 2.152941176470588, "grad_norm": 0.40524888025599715, "learning_rate": 0.00025949521696603536, "loss": 0.1963, "step": 4575 }, { "epoch": 2.1534117647058824, "grad_norm": 0.3574413886850256, "learning_rate": 0.0002594652510880885, "loss": 0.1426, "step": 4576 }, { "epoch": 2.1538823529411766, "grad_norm": 0.2939174748609174, "learning_rate": 0.0002594352758612723, "loss": 0.0718, "step": 4577 }, { "epoch": 2.1543529411764704, "grad_norm": 0.6702666274290352, "learning_rate": 0.0002594052912881467, "loss": 0.457, "step": 4578 }, { "epoch": 2.1548235294117646, "grad_norm": 0.32799668871238186, "learning_rate": 0.00025937529737127263, "loss": 0.1465, "step": 4579 }, { "epoch": 2.155294117647059, "grad_norm": 0.26563449211089274, "learning_rate": 0.0002593452941132117, "loss": 0.1426, "step": 4580 }, { "epoch": 2.155764705882353, "grad_norm": 0.8468936101239343, "learning_rate": 0.0002593152815165263, "loss": 0.3848, "step": 4581 }, { "epoch": 2.1562352941176472, "grad_norm": 0.46471576979833235, "learning_rate": 0.0002592852595837797, "loss": 0.2051, "step": 4582 }, { "epoch": 2.156705882352941, "grad_norm": 0.6516199238163795, "learning_rate": 0.00025925522831753584, "loss": 0.3262, "step": 4583 }, { "epoch": 2.1571764705882353, "grad_norm": 0.828278310087784, "learning_rate": 0.00025922518772035963, "loss": 0.5625, "step": 4584 }, { "epoch": 2.1576470588235295, "grad_norm": 1.0555983451781088, "learning_rate": 0.0002591951377948166, "loss": 0.7148, "step": 4585 }, { "epoch": 2.1581176470588237, "grad_norm": 0.7378525901569598, "learning_rate": 0.00025916507854347327, "loss": 0.3301, "step": 4586 }, { "epoch": 2.1585882352941175, "grad_norm": 0.4238194367264434, "learning_rate": 0.00025913500996889677, "loss": 0.1631, "step": 4587 }, { "epoch": 2.1590588235294117, "grad_norm": 0.42514026339929284, "learning_rate": 0.00025910493207365504, "loss": 0.1201, "step": 4588 }, { "epoch": 2.159529411764706, "grad_norm": 0.4141265068947832, "learning_rate": 0.000259074844860317, "loss": 0.1475, "step": 4589 }, { "epoch": 2.16, "grad_norm": 0.5392114372693098, "learning_rate": 0.00025904474833145225, "loss": 0.2441, "step": 4590 }, { "epoch": 2.160470588235294, "grad_norm": 0.5073307646745818, "learning_rate": 0.00025901464248963106, "loss": 0.2295, "step": 4591 }, { "epoch": 2.160941176470588, "grad_norm": 0.3857924962353249, "learning_rate": 0.00025898452733742473, "loss": 0.1504, "step": 4592 }, { "epoch": 2.1614117647058824, "grad_norm": 0.3997776446784195, "learning_rate": 0.00025895440287740514, "loss": 0.1748, "step": 4593 }, { "epoch": 2.1618823529411766, "grad_norm": 0.48502540367881863, "learning_rate": 0.00025892426911214517, "loss": 0.1719, "step": 4594 }, { "epoch": 2.1623529411764704, "grad_norm": 0.12231682737993084, "learning_rate": 0.0002588941260442183, "loss": 0.0187, "step": 4595 }, { "epoch": 2.1628235294117646, "grad_norm": 0.7281437952189962, "learning_rate": 0.00025886397367619904, "loss": 0.3691, "step": 4596 }, { "epoch": 2.163294117647059, "grad_norm": 0.47579771690405914, "learning_rate": 0.00025883381201066236, "loss": 0.2412, "step": 4597 }, { "epoch": 2.163764705882353, "grad_norm": 0.40916278506888326, "learning_rate": 0.0002588036410501844, "loss": 0.1699, "step": 4598 }, { "epoch": 2.1642352941176473, "grad_norm": 0.4197835480719002, "learning_rate": 0.00025877346079734175, "loss": 0.1875, "step": 4599 }, { "epoch": 2.164705882352941, "grad_norm": 0.8883272280919292, "learning_rate": 0.000258743271254712, "loss": 0.4297, "step": 4600 }, { "epoch": 2.1651764705882353, "grad_norm": 0.8719212338744408, "learning_rate": 0.00025871307242487354, "loss": 0.5469, "step": 4601 }, { "epoch": 2.1656470588235295, "grad_norm": 0.548846040280004, "learning_rate": 0.00025868286431040546, "loss": 0.4531, "step": 4602 }, { "epoch": 2.1661176470588237, "grad_norm": 0.32289998791198826, "learning_rate": 0.0002586526469138877, "loss": 0.1279, "step": 4603 }, { "epoch": 2.1665882352941175, "grad_norm": 0.3914232481085348, "learning_rate": 0.0002586224202379009, "loss": 0.2148, "step": 4604 }, { "epoch": 2.1670588235294117, "grad_norm": 0.3320573958673442, "learning_rate": 0.00025859218428502667, "loss": 0.1426, "step": 4605 }, { "epoch": 2.167529411764706, "grad_norm": 0.9676650750545013, "learning_rate": 0.00025856193905784725, "loss": 0.3359, "step": 4606 }, { "epoch": 2.168, "grad_norm": 0.6197760448471998, "learning_rate": 0.00025853168455894566, "loss": 0.5234, "step": 4607 }, { "epoch": 2.168470588235294, "grad_norm": 0.90080661933843, "learning_rate": 0.000258501420790906, "loss": 0.6094, "step": 4608 }, { "epoch": 2.168941176470588, "grad_norm": 0.6520467497324546, "learning_rate": 0.0002584711477563127, "loss": 0.4238, "step": 4609 }, { "epoch": 2.1694117647058824, "grad_norm": 0.341751919988087, "learning_rate": 0.0002584408654577514, "loss": 0.2119, "step": 4610 }, { "epoch": 2.1698823529411766, "grad_norm": 0.2515765358211208, "learning_rate": 0.0002584105738978083, "loss": 0.0659, "step": 4611 }, { "epoch": 2.1703529411764704, "grad_norm": 0.6336296215733933, "learning_rate": 0.00025838027307907037, "loss": 0.4043, "step": 4612 }, { "epoch": 2.1708235294117646, "grad_norm": 0.5587968257451592, "learning_rate": 0.00025834996300412556, "loss": 0.2656, "step": 4613 }, { "epoch": 2.171294117647059, "grad_norm": 0.6885245024596995, "learning_rate": 0.0002583196436755624, "loss": 0.3906, "step": 4614 }, { "epoch": 2.171764705882353, "grad_norm": 0.4100528059438338, "learning_rate": 0.00025828931509597045, "loss": 0.1885, "step": 4615 }, { "epoch": 2.1722352941176473, "grad_norm": 1.0532072820904443, "learning_rate": 0.0002582589772679398, "loss": 0.5898, "step": 4616 }, { "epoch": 2.172705882352941, "grad_norm": 0.3989065085556223, "learning_rate": 0.0002582286301940615, "loss": 0.4277, "step": 4617 }, { "epoch": 2.1731764705882353, "grad_norm": 0.2122692701436995, "learning_rate": 0.00025819827387692733, "loss": 0.0669, "step": 4618 }, { "epoch": 2.1736470588235295, "grad_norm": 0.4897928637271234, "learning_rate": 0.0002581679083191298, "loss": 0.291, "step": 4619 }, { "epoch": 2.1741176470588237, "grad_norm": 0.35573628289113074, "learning_rate": 0.0002581375335232624, "loss": 0.084, "step": 4620 }, { "epoch": 2.1745882352941175, "grad_norm": 1.445405733113211, "learning_rate": 0.00025810714949191916, "loss": 0.7422, "step": 4621 }, { "epoch": 2.1750588235294117, "grad_norm": 0.7754139486418665, "learning_rate": 0.00025807675622769514, "loss": 0.4355, "step": 4622 }, { "epoch": 2.175529411764706, "grad_norm": 0.7064034783563272, "learning_rate": 0.00025804635373318603, "loss": 0.4395, "step": 4623 }, { "epoch": 2.176, "grad_norm": 0.7651116680770333, "learning_rate": 0.0002580159420109883, "loss": 0.6992, "step": 4624 }, { "epoch": 2.176470588235294, "grad_norm": 0.6634600253171034, "learning_rate": 0.00025798552106369937, "loss": 0.4844, "step": 4625 }, { "epoch": 2.176941176470588, "grad_norm": 0.5749991068276878, "learning_rate": 0.00025795509089391726, "loss": 0.2334, "step": 4626 }, { "epoch": 2.1774117647058824, "grad_norm": 0.8208430058511523, "learning_rate": 0.00025792465150424083, "loss": 0.5312, "step": 4627 }, { "epoch": 2.1778823529411766, "grad_norm": 1.1231836210681807, "learning_rate": 0.00025789420289726983, "loss": 0.6719, "step": 4628 }, { "epoch": 2.1783529411764704, "grad_norm": 1.015727498974389, "learning_rate": 0.00025786374507560467, "loss": 0.4531, "step": 4629 }, { "epoch": 2.1788235294117646, "grad_norm": 0.6182031435844523, "learning_rate": 0.0002578332780418466, "loss": 0.3633, "step": 4630 }, { "epoch": 2.179294117647059, "grad_norm": 0.47195735469230915, "learning_rate": 0.0002578028017985977, "loss": 0.1035, "step": 4631 }, { "epoch": 2.179764705882353, "grad_norm": 0.9542980202985503, "learning_rate": 0.00025777231634846074, "loss": 0.6523, "step": 4632 }, { "epoch": 2.1802352941176473, "grad_norm": 1.3513961961074799, "learning_rate": 0.00025774182169403933, "loss": 0.5703, "step": 4633 }, { "epoch": 2.180705882352941, "grad_norm": 0.36503335872034853, "learning_rate": 0.00025771131783793787, "loss": 0.1216, "step": 4634 }, { "epoch": 2.1811764705882353, "grad_norm": 0.20632405084777902, "learning_rate": 0.00025768080478276163, "loss": 0.0454, "step": 4635 }, { "epoch": 2.1816470588235295, "grad_norm": 1.2069544549092412, "learning_rate": 0.0002576502825311164, "loss": 0.6875, "step": 4636 }, { "epoch": 2.1821176470588237, "grad_norm": 0.17847685062328583, "learning_rate": 0.000257619751085609, "loss": 0.0588, "step": 4637 }, { "epoch": 2.1825882352941175, "grad_norm": 0.7373183250689943, "learning_rate": 0.00025758921044884706, "loss": 0.3477, "step": 4638 }, { "epoch": 2.1830588235294117, "grad_norm": 0.7079646806432442, "learning_rate": 0.0002575586606234388, "loss": 0.4434, "step": 4639 }, { "epoch": 2.183529411764706, "grad_norm": 0.6173393005337582, "learning_rate": 0.0002575281016119933, "loss": 0.2793, "step": 4640 }, { "epoch": 2.184, "grad_norm": 0.6610279231815804, "learning_rate": 0.00025749753341712056, "loss": 0.4023, "step": 4641 }, { "epoch": 2.184470588235294, "grad_norm": 0.31710610992536725, "learning_rate": 0.0002574669560414312, "loss": 0.1895, "step": 4642 }, { "epoch": 2.184941176470588, "grad_norm": 0.258471407166285, "learning_rate": 0.0002574363694875367, "loss": 0.0776, "step": 4643 }, { "epoch": 2.1854117647058824, "grad_norm": 0.9452771869261776, "learning_rate": 0.0002574057737580492, "loss": 0.5586, "step": 4644 }, { "epoch": 2.1858823529411766, "grad_norm": 0.6761274327728912, "learning_rate": 0.0002573751688555818, "loss": 0.3457, "step": 4645 }, { "epoch": 2.1863529411764704, "grad_norm": 0.6394853788488253, "learning_rate": 0.0002573445547827484, "loss": 0.4121, "step": 4646 }, { "epoch": 2.1868235294117646, "grad_norm": 0.641155448394473, "learning_rate": 0.0002573139315421634, "loss": 0.3535, "step": 4647 }, { "epoch": 2.187294117647059, "grad_norm": 0.39341069239782966, "learning_rate": 0.0002572832991364423, "loss": 0.1196, "step": 4648 }, { "epoch": 2.187764705882353, "grad_norm": 0.938209530008667, "learning_rate": 0.00025725265756820125, "loss": 0.4258, "step": 4649 }, { "epoch": 2.1882352941176473, "grad_norm": 0.5494609790098176, "learning_rate": 0.00025722200684005715, "loss": 0.3086, "step": 4650 }, { "epoch": 2.188705882352941, "grad_norm": 0.375184261510577, "learning_rate": 0.0002571913469546277, "loss": 0.1025, "step": 4651 }, { "epoch": 2.1891764705882353, "grad_norm": 1.3051366344076838, "learning_rate": 0.00025716067791453157, "loss": 0.5547, "step": 4652 }, { "epoch": 2.1896470588235295, "grad_norm": 0.8828573406277952, "learning_rate": 0.0002571299997223878, "loss": 0.5547, "step": 4653 }, { "epoch": 2.1901176470588237, "grad_norm": 0.6846264831169487, "learning_rate": 0.0002570993123808167, "loss": 0.3867, "step": 4654 }, { "epoch": 2.1905882352941175, "grad_norm": 1.0134503395359487, "learning_rate": 0.00025706861589243894, "loss": 0.6055, "step": 4655 }, { "epoch": 2.1910588235294117, "grad_norm": 0.5443894178908287, "learning_rate": 0.0002570379102598762, "loss": 0.3281, "step": 4656 }, { "epoch": 2.191529411764706, "grad_norm": 0.717939364448048, "learning_rate": 0.000257007195485751, "loss": 0.5352, "step": 4657 }, { "epoch": 2.192, "grad_norm": 0.4553823018453367, "learning_rate": 0.00025697647157268635, "loss": 0.2578, "step": 4658 }, { "epoch": 2.192470588235294, "grad_norm": 0.5055574321348052, "learning_rate": 0.0002569457385233063, "loss": 0.377, "step": 4659 }, { "epoch": 2.192941176470588, "grad_norm": 0.4755647878645639, "learning_rate": 0.00025691499634023564, "loss": 0.3027, "step": 4660 }, { "epoch": 2.1934117647058824, "grad_norm": 0.7584625571267166, "learning_rate": 0.0002568842450260999, "loss": 0.5508, "step": 4661 }, { "epoch": 2.1938823529411766, "grad_norm": 0.22477588599731027, "learning_rate": 0.0002568534845835253, "loss": 0.0449, "step": 4662 }, { "epoch": 2.1943529411764704, "grad_norm": 0.4476820993382767, "learning_rate": 0.0002568227150151391, "loss": 0.1426, "step": 4663 }, { "epoch": 2.1948235294117646, "grad_norm": 0.46353653371339965, "learning_rate": 0.000256791936323569, "loss": 0.2559, "step": 4664 }, { "epoch": 2.195294117647059, "grad_norm": 0.5251373625710158, "learning_rate": 0.00025676114851144375, "loss": 0.3438, "step": 4665 }, { "epoch": 2.195764705882353, "grad_norm": 0.392476574677241, "learning_rate": 0.0002567303515813928, "loss": 0.1689, "step": 4666 }, { "epoch": 2.1962352941176473, "grad_norm": 0.514201775889668, "learning_rate": 0.0002566995455360463, "loss": 0.332, "step": 4667 }, { "epoch": 2.196705882352941, "grad_norm": 0.42689597097069026, "learning_rate": 0.00025666873037803517, "loss": 0.1416, "step": 4668 }, { "epoch": 2.1971764705882353, "grad_norm": 0.471736933657457, "learning_rate": 0.00025663790610999134, "loss": 0.2158, "step": 4669 }, { "epoch": 2.1976470588235295, "grad_norm": 0.17600518592849304, "learning_rate": 0.0002566070727345473, "loss": 0.0674, "step": 4670 }, { "epoch": 2.1981176470588237, "grad_norm": 0.3625741284361011, "learning_rate": 0.00025657623025433627, "loss": 0.1143, "step": 4671 }, { "epoch": 2.1985882352941175, "grad_norm": 0.5607324663676265, "learning_rate": 0.0002565453786719924, "loss": 0.3281, "step": 4672 }, { "epoch": 2.1990588235294117, "grad_norm": 0.4193374951856342, "learning_rate": 0.00025651451799015067, "loss": 0.2324, "step": 4673 }, { "epoch": 2.199529411764706, "grad_norm": 0.3564096448643347, "learning_rate": 0.0002564836482114466, "loss": 0.1611, "step": 4674 }, { "epoch": 2.2, "grad_norm": 0.9965847523781665, "learning_rate": 0.00025645276933851667, "loss": 0.6133, "step": 4675 }, { "epoch": 2.200470588235294, "grad_norm": 0.4952234382244605, "learning_rate": 0.00025642188137399815, "loss": 0.2178, "step": 4676 }, { "epoch": 2.200941176470588, "grad_norm": 0.5082672228787917, "learning_rate": 0.00025639098432052895, "loss": 0.1855, "step": 4677 }, { "epoch": 2.2014117647058824, "grad_norm": 0.7163071872325057, "learning_rate": 0.0002563600781807478, "loss": 0.3281, "step": 4678 }, { "epoch": 2.2018823529411766, "grad_norm": 0.8876945542615994, "learning_rate": 0.0002563291629572943, "loss": 0.3555, "step": 4679 }, { "epoch": 2.2023529411764704, "grad_norm": 0.724712110639052, "learning_rate": 0.00025629823865280874, "loss": 0.334, "step": 4680 }, { "epoch": 2.2028235294117646, "grad_norm": 1.1321144432523655, "learning_rate": 0.00025626730526993223, "loss": 0.4668, "step": 4681 }, { "epoch": 2.203294117647059, "grad_norm": 0.8839836301107834, "learning_rate": 0.00025623636281130657, "loss": 0.5234, "step": 4682 }, { "epoch": 2.203764705882353, "grad_norm": 0.6963687161703425, "learning_rate": 0.0002562054112795745, "loss": 0.4062, "step": 4683 }, { "epoch": 2.2042352941176473, "grad_norm": 0.6893986316552888, "learning_rate": 0.00025617445067737933, "loss": 0.4277, "step": 4684 }, { "epoch": 2.204705882352941, "grad_norm": 0.2879195930384086, "learning_rate": 0.0002561434810073653, "loss": 0.1055, "step": 4685 }, { "epoch": 2.2051764705882353, "grad_norm": 0.9844420859285459, "learning_rate": 0.00025611250227217737, "loss": 0.3789, "step": 4686 }, { "epoch": 2.2056470588235295, "grad_norm": 0.7569178683751882, "learning_rate": 0.0002560815144744612, "loss": 0.3301, "step": 4687 }, { "epoch": 2.2061176470588233, "grad_norm": 0.35842773593833493, "learning_rate": 0.00025605051761686345, "loss": 0.1406, "step": 4688 }, { "epoch": 2.2065882352941175, "grad_norm": 0.4707335096386357, "learning_rate": 0.0002560195117020313, "loss": 0.3555, "step": 4689 }, { "epoch": 2.2070588235294117, "grad_norm": 0.6857391469355768, "learning_rate": 0.0002559884967326128, "loss": 0.2412, "step": 4690 }, { "epoch": 2.207529411764706, "grad_norm": 1.1248471154339197, "learning_rate": 0.0002559574727112569, "loss": 0.3965, "step": 4691 }, { "epoch": 2.208, "grad_norm": 0.46926607101323986, "learning_rate": 0.000255926439640613, "loss": 0.1719, "step": 4692 }, { "epoch": 2.208470588235294, "grad_norm": 0.8858475006978717, "learning_rate": 0.00025589539752333165, "loss": 0.418, "step": 4693 }, { "epoch": 2.208941176470588, "grad_norm": 0.4206551792709373, "learning_rate": 0.0002558643463620639, "loss": 0.248, "step": 4694 }, { "epoch": 2.2094117647058824, "grad_norm": 0.7589176986228539, "learning_rate": 0.0002558332861594617, "loss": 0.4375, "step": 4695 }, { "epoch": 2.2098823529411766, "grad_norm": 0.3315278146246091, "learning_rate": 0.0002558022169181778, "loss": 0.0981, "step": 4696 }, { "epoch": 2.2103529411764704, "grad_norm": 0.6981564630982651, "learning_rate": 0.00025577113864086557, "loss": 0.2949, "step": 4697 }, { "epoch": 2.2108235294117646, "grad_norm": 0.48482035733521406, "learning_rate": 0.0002557400513301794, "loss": 0.1416, "step": 4698 }, { "epoch": 2.211294117647059, "grad_norm": 0.4588551442220821, "learning_rate": 0.00025570895498877404, "loss": 0.248, "step": 4699 }, { "epoch": 2.211764705882353, "grad_norm": 0.576796688126738, "learning_rate": 0.00025567784961930546, "loss": 0.373, "step": 4700 }, { "epoch": 2.212235294117647, "grad_norm": 0.36528321930513813, "learning_rate": 0.0002556467352244302, "loss": 0.1807, "step": 4701 }, { "epoch": 2.212705882352941, "grad_norm": 0.7436881494693952, "learning_rate": 0.00025561561180680563, "loss": 0.4004, "step": 4702 }, { "epoch": 2.2131764705882353, "grad_norm": 0.2730625355640298, "learning_rate": 0.0002555844793690897, "loss": 0.166, "step": 4703 }, { "epoch": 2.2136470588235295, "grad_norm": 0.8688240005723895, "learning_rate": 0.0002555533379139414, "loss": 0.6289, "step": 4704 }, { "epoch": 2.2141176470588233, "grad_norm": 0.5014897918238692, "learning_rate": 0.00025552218744402026, "loss": 0.2559, "step": 4705 }, { "epoch": 2.2145882352941175, "grad_norm": 0.5639400698147855, "learning_rate": 0.00025549102796198673, "loss": 0.2734, "step": 4706 }, { "epoch": 2.2150588235294117, "grad_norm": 0.8673423022700468, "learning_rate": 0.00025545985947050197, "loss": 0.4766, "step": 4707 }, { "epoch": 2.215529411764706, "grad_norm": 0.5496694428779668, "learning_rate": 0.00025542868197222797, "loss": 0.3262, "step": 4708 }, { "epoch": 2.216, "grad_norm": 0.4092568790125873, "learning_rate": 0.00025539749546982736, "loss": 0.208, "step": 4709 }, { "epoch": 2.216470588235294, "grad_norm": 0.6581894011508607, "learning_rate": 0.00025536629996596367, "loss": 0.2617, "step": 4710 }, { "epoch": 2.216941176470588, "grad_norm": 0.9032135199744293, "learning_rate": 0.0002553350954633012, "loss": 0.7031, "step": 4711 }, { "epoch": 2.2174117647058824, "grad_norm": 0.6878093100028401, "learning_rate": 0.00025530388196450487, "loss": 0.3555, "step": 4712 }, { "epoch": 2.2178823529411766, "grad_norm": 0.7640103926899467, "learning_rate": 0.0002552726594722405, "loss": 0.6406, "step": 4713 }, { "epoch": 2.2183529411764704, "grad_norm": 0.6166727089686072, "learning_rate": 0.00025524142798917466, "loss": 0.416, "step": 4714 }, { "epoch": 2.2188235294117646, "grad_norm": 0.25881125271498406, "learning_rate": 0.00025521018751797464, "loss": 0.1367, "step": 4715 }, { "epoch": 2.219294117647059, "grad_norm": 0.34757853006128553, "learning_rate": 0.0002551789380613086, "loss": 0.1416, "step": 4716 }, { "epoch": 2.219764705882353, "grad_norm": 0.6312581396846493, "learning_rate": 0.00025514767962184535, "loss": 0.249, "step": 4717 }, { "epoch": 2.220235294117647, "grad_norm": 0.32868044435282584, "learning_rate": 0.0002551164122022545, "loss": 0.0835, "step": 4718 }, { "epoch": 2.220705882352941, "grad_norm": 0.6761456454952035, "learning_rate": 0.00025508513580520644, "loss": 0.4414, "step": 4719 }, { "epoch": 2.2211764705882353, "grad_norm": 0.7169011894704393, "learning_rate": 0.00025505385043337236, "loss": 0.3984, "step": 4720 }, { "epoch": 2.2216470588235295, "grad_norm": 0.7704716274554653, "learning_rate": 0.0002550225560894241, "loss": 0.3281, "step": 4721 }, { "epoch": 2.2221176470588233, "grad_norm": 0.7915841741203175, "learning_rate": 0.00025499125277603446, "loss": 0.4043, "step": 4722 }, { "epoch": 2.2225882352941175, "grad_norm": 0.4112816104918387, "learning_rate": 0.0002549599404958769, "loss": 0.1475, "step": 4723 }, { "epoch": 2.2230588235294118, "grad_norm": 0.7917717568070624, "learning_rate": 0.0002549286192516256, "loss": 0.2715, "step": 4724 }, { "epoch": 2.223529411764706, "grad_norm": 0.48709496275197406, "learning_rate": 0.0002548972890459555, "loss": 0.1592, "step": 4725 }, { "epoch": 2.224, "grad_norm": 0.4767054898227113, "learning_rate": 0.00025486594988154235, "loss": 0.1445, "step": 4726 }, { "epoch": 2.224470588235294, "grad_norm": 0.5874674577246577, "learning_rate": 0.0002548346017610628, "loss": 0.1475, "step": 4727 }, { "epoch": 2.224941176470588, "grad_norm": 0.20839354181349815, "learning_rate": 0.000254803244687194, "loss": 0.1387, "step": 4728 }, { "epoch": 2.2254117647058824, "grad_norm": 0.6537241873077314, "learning_rate": 0.0002547718786626141, "loss": 0.3594, "step": 4729 }, { "epoch": 2.2258823529411766, "grad_norm": 0.7131058350015165, "learning_rate": 0.0002547405036900018, "loss": 0.3398, "step": 4730 }, { "epoch": 2.2263529411764704, "grad_norm": 0.837232091599251, "learning_rate": 0.0002547091197720368, "loss": 0.418, "step": 4731 }, { "epoch": 2.2268235294117646, "grad_norm": 0.6749584588465367, "learning_rate": 0.0002546777269113993, "loss": 0.2119, "step": 4732 }, { "epoch": 2.227294117647059, "grad_norm": 0.672224652809802, "learning_rate": 0.0002546463251107705, "loss": 0.3535, "step": 4733 }, { "epoch": 2.227764705882353, "grad_norm": 0.4570195519289637, "learning_rate": 0.0002546149143728323, "loss": 0.1943, "step": 4734 }, { "epoch": 2.228235294117647, "grad_norm": 0.7770501932626676, "learning_rate": 0.0002545834947002672, "loss": 0.3457, "step": 4735 }, { "epoch": 2.228705882352941, "grad_norm": 0.7960270798254538, "learning_rate": 0.00025455206609575876, "loss": 0.3906, "step": 4736 }, { "epoch": 2.2291764705882353, "grad_norm": 0.5054042073356894, "learning_rate": 0.000254520628561991, "loss": 0.1865, "step": 4737 }, { "epoch": 2.2296470588235295, "grad_norm": 0.10489187639082252, "learning_rate": 0.00025448918210164886, "loss": 0.0205, "step": 4738 }, { "epoch": 2.2301176470588233, "grad_norm": 0.9154600197271495, "learning_rate": 0.00025445772671741816, "loss": 0.3984, "step": 4739 }, { "epoch": 2.2305882352941175, "grad_norm": 1.0449783194896034, "learning_rate": 0.00025442626241198513, "loss": 0.5156, "step": 4740 }, { "epoch": 2.2310588235294118, "grad_norm": 0.7457630024790696, "learning_rate": 0.00025439478918803715, "loss": 0.4199, "step": 4741 }, { "epoch": 2.231529411764706, "grad_norm": 0.47658042852579224, "learning_rate": 0.00025436330704826205, "loss": 0.4082, "step": 4742 }, { "epoch": 2.232, "grad_norm": 0.47672273523562414, "learning_rate": 0.0002543318159953487, "loss": 0.2344, "step": 4743 }, { "epoch": 2.232470588235294, "grad_norm": 0.513254850862088, "learning_rate": 0.00025430031603198644, "loss": 0.2754, "step": 4744 }, { "epoch": 2.232941176470588, "grad_norm": 0.25350698221616935, "learning_rate": 0.0002542688071608657, "loss": 0.0493, "step": 4745 }, { "epoch": 2.2334117647058824, "grad_norm": 0.5472036889335512, "learning_rate": 0.0002542372893846773, "loss": 0.2227, "step": 4746 }, { "epoch": 2.2338823529411767, "grad_norm": 1.182763995339556, "learning_rate": 0.00025420576270611313, "loss": 0.6289, "step": 4747 }, { "epoch": 2.2343529411764704, "grad_norm": 0.6857468273760879, "learning_rate": 0.00025417422712786575, "loss": 0.3672, "step": 4748 }, { "epoch": 2.2348235294117647, "grad_norm": 1.1126416952471392, "learning_rate": 0.00025414268265262833, "loss": 0.6211, "step": 4749 }, { "epoch": 2.235294117647059, "grad_norm": 0.6735221008371276, "learning_rate": 0.0002541111292830951, "loss": 0.3594, "step": 4750 }, { "epoch": 2.235764705882353, "grad_norm": 0.9701080116552132, "learning_rate": 0.0002540795670219606, "loss": 0.3984, "step": 4751 }, { "epoch": 2.236235294117647, "grad_norm": 0.41348503212023874, "learning_rate": 0.0002540479958719207, "loss": 0.1553, "step": 4752 }, { "epoch": 2.236705882352941, "grad_norm": 0.46552003566968003, "learning_rate": 0.0002540164158356716, "loss": 0.1484, "step": 4753 }, { "epoch": 2.2371764705882353, "grad_norm": 1.0897870746764677, "learning_rate": 0.0002539848269159104, "loss": 0.5938, "step": 4754 }, { "epoch": 2.2376470588235295, "grad_norm": 0.633918264318149, "learning_rate": 0.00025395322911533484, "loss": 0.2988, "step": 4755 }, { "epoch": 2.2381176470588233, "grad_norm": 0.4353145881567354, "learning_rate": 0.0002539216224366437, "loss": 0.1631, "step": 4756 }, { "epoch": 2.2385882352941175, "grad_norm": 0.4545647643815696, "learning_rate": 0.00025389000688253626, "loss": 0.2422, "step": 4757 }, { "epoch": 2.2390588235294118, "grad_norm": 0.8117026108552673, "learning_rate": 0.00025385838245571264, "loss": 0.4922, "step": 4758 }, { "epoch": 2.239529411764706, "grad_norm": 0.7951356102023124, "learning_rate": 0.00025382674915887374, "loss": 0.3789, "step": 4759 }, { "epoch": 2.24, "grad_norm": 0.8452122698556983, "learning_rate": 0.0002537951069947212, "loss": 0.543, "step": 4760 }, { "epoch": 2.240470588235294, "grad_norm": 0.3647680111107824, "learning_rate": 0.00025376345596595743, "loss": 0.1504, "step": 4761 }, { "epoch": 2.240941176470588, "grad_norm": 0.6334417087763149, "learning_rate": 0.0002537317960752855, "loss": 0.3008, "step": 4762 }, { "epoch": 2.2414117647058824, "grad_norm": 0.4181319319079749, "learning_rate": 0.0002537001273254094, "loss": 0.1973, "step": 4763 }, { "epoch": 2.2418823529411767, "grad_norm": 0.4057771104035387, "learning_rate": 0.00025366844971903385, "loss": 0.1455, "step": 4764 }, { "epoch": 2.2423529411764704, "grad_norm": 0.7898717506630499, "learning_rate": 0.00025363676325886416, "loss": 0.3086, "step": 4765 }, { "epoch": 2.2428235294117647, "grad_norm": 0.8694835732643444, "learning_rate": 0.00025360506794760653, "loss": 0.4473, "step": 4766 }, { "epoch": 2.243294117647059, "grad_norm": 0.7373160567428463, "learning_rate": 0.00025357336378796795, "loss": 0.4648, "step": 4767 }, { "epoch": 2.243764705882353, "grad_norm": 0.4562867994395678, "learning_rate": 0.0002535416507826561, "loss": 0.1914, "step": 4768 }, { "epoch": 2.244235294117647, "grad_norm": 0.7323250027162094, "learning_rate": 0.0002535099289343793, "loss": 0.2344, "step": 4769 }, { "epoch": 2.244705882352941, "grad_norm": 0.4460557158537172, "learning_rate": 0.00025347819824584695, "loss": 0.1807, "step": 4770 }, { "epoch": 2.2451764705882353, "grad_norm": 0.19779140871977477, "learning_rate": 0.0002534464587197689, "loss": 0.0461, "step": 4771 }, { "epoch": 2.2456470588235296, "grad_norm": 1.2245480178302384, "learning_rate": 0.00025341471035885587, "loss": 0.3457, "step": 4772 }, { "epoch": 2.2461176470588233, "grad_norm": 1.915502290502909, "learning_rate": 0.0002533829531658193, "loss": 0.7578, "step": 4773 }, { "epoch": 2.2465882352941176, "grad_norm": 0.5513686458851862, "learning_rate": 0.00025335118714337146, "loss": 0.4434, "step": 4774 }, { "epoch": 2.2470588235294118, "grad_norm": 1.408218614343185, "learning_rate": 0.0002533194122942253, "loss": 0.5938, "step": 4775 }, { "epoch": 2.247529411764706, "grad_norm": 0.719388828122023, "learning_rate": 0.00025328762862109457, "loss": 0.2598, "step": 4776 }, { "epoch": 2.248, "grad_norm": 0.4886512217383577, "learning_rate": 0.00025325583612669373, "loss": 0.1279, "step": 4777 }, { "epoch": 2.248470588235294, "grad_norm": 0.378470537540145, "learning_rate": 0.00025322403481373797, "loss": 0.168, "step": 4778 }, { "epoch": 2.248941176470588, "grad_norm": 0.8791457167235265, "learning_rate": 0.00025319222468494335, "loss": 0.3672, "step": 4779 }, { "epoch": 2.2494117647058824, "grad_norm": 0.9917151789280976, "learning_rate": 0.00025316040574302663, "loss": 0.3301, "step": 4780 }, { "epoch": 2.2498823529411767, "grad_norm": 0.43899921422934496, "learning_rate": 0.00025312857799070524, "loss": 0.1377, "step": 4781 }, { "epoch": 2.2503529411764704, "grad_norm": 0.5085953281413705, "learning_rate": 0.0002530967414306974, "loss": 0.2236, "step": 4782 }, { "epoch": 2.2508235294117647, "grad_norm": 0.5346413702938485, "learning_rate": 0.0002530648960657222, "loss": 0.1426, "step": 4783 }, { "epoch": 2.251294117647059, "grad_norm": 0.39193652200517565, "learning_rate": 0.0002530330418984993, "loss": 0.1025, "step": 4784 }, { "epoch": 2.251764705882353, "grad_norm": 1.7434919308623729, "learning_rate": 0.00025300117893174934, "loss": 0.6406, "step": 4785 }, { "epoch": 2.252235294117647, "grad_norm": 0.5056205257310283, "learning_rate": 0.0002529693071681934, "loss": 0.1738, "step": 4786 }, { "epoch": 2.252705882352941, "grad_norm": 0.36798041348874094, "learning_rate": 0.00025293742661055364, "loss": 0.1001, "step": 4787 }, { "epoch": 2.2531764705882353, "grad_norm": 0.29663254602905786, "learning_rate": 0.0002529055372615527, "loss": 0.1611, "step": 4788 }, { "epoch": 2.2536470588235296, "grad_norm": 1.375833082047387, "learning_rate": 0.00025287363912391403, "loss": 0.6953, "step": 4789 }, { "epoch": 2.2541176470588233, "grad_norm": 0.7631774660290596, "learning_rate": 0.00025284173220036215, "loss": 0.2891, "step": 4790 }, { "epoch": 2.2545882352941176, "grad_norm": 0.8340702142861858, "learning_rate": 0.00025280981649362184, "loss": 0.459, "step": 4791 }, { "epoch": 2.255058823529412, "grad_norm": 0.490790153798792, "learning_rate": 0.0002527778920064189, "loss": 0.1543, "step": 4792 }, { "epoch": 2.255529411764706, "grad_norm": 0.5346431611422944, "learning_rate": 0.00025274595874147984, "loss": 0.2715, "step": 4793 }, { "epoch": 2.2560000000000002, "grad_norm": 0.4178450740385689, "learning_rate": 0.000252714016701532, "loss": 0.165, "step": 4794 }, { "epoch": 2.256470588235294, "grad_norm": 0.6686324378030953, "learning_rate": 0.0002526820658893033, "loss": 0.4082, "step": 4795 }, { "epoch": 2.2569411764705882, "grad_norm": 0.39824280852219596, "learning_rate": 0.0002526501063075225, "loss": 0.1719, "step": 4796 }, { "epoch": 2.2574117647058825, "grad_norm": 0.6884138819388149, "learning_rate": 0.00025261813795891916, "loss": 0.334, "step": 4797 }, { "epoch": 2.2578823529411762, "grad_norm": 0.6332114015555256, "learning_rate": 0.00025258616084622345, "loss": 0.2734, "step": 4798 }, { "epoch": 2.2583529411764705, "grad_norm": 0.6090851153685339, "learning_rate": 0.00025255417497216647, "loss": 0.3672, "step": 4799 }, { "epoch": 2.2588235294117647, "grad_norm": 0.8948945274790322, "learning_rate": 0.00025252218033947993, "loss": 0.4805, "step": 4800 }, { "epoch": 2.259294117647059, "grad_norm": 0.6135701554352305, "learning_rate": 0.0002524901769508963, "loss": 0.377, "step": 4801 }, { "epoch": 2.259764705882353, "grad_norm": 1.0129987330458798, "learning_rate": 0.0002524581648091488, "loss": 0.6367, "step": 4802 }, { "epoch": 2.260235294117647, "grad_norm": 0.8068474107411387, "learning_rate": 0.00025242614391697153, "loss": 0.4395, "step": 4803 }, { "epoch": 2.260705882352941, "grad_norm": 1.3508586007744492, "learning_rate": 0.00025239411427709915, "loss": 0.6172, "step": 4804 }, { "epoch": 2.2611764705882353, "grad_norm": 0.4133009524349282, "learning_rate": 0.0002523620758922672, "loss": 0.1445, "step": 4805 }, { "epoch": 2.2616470588235296, "grad_norm": 0.13656581626756717, "learning_rate": 0.00025233002876521183, "loss": 0.0311, "step": 4806 }, { "epoch": 2.2621176470588233, "grad_norm": 0.41422850681658263, "learning_rate": 0.0002522979728986701, "loss": 0.0972, "step": 4807 }, { "epoch": 2.2625882352941176, "grad_norm": 0.44468806211872114, "learning_rate": 0.00025226590829537977, "loss": 0.2256, "step": 4808 }, { "epoch": 2.263058823529412, "grad_norm": 1.1419171360851315, "learning_rate": 0.0002522338349580792, "loss": 0.4062, "step": 4809 }, { "epoch": 2.263529411764706, "grad_norm": 0.7919883952366881, "learning_rate": 0.00025220175288950765, "loss": 0.2559, "step": 4810 }, { "epoch": 2.2640000000000002, "grad_norm": 0.8600842271763787, "learning_rate": 0.00025216966209240516, "loss": 0.3848, "step": 4811 }, { "epoch": 2.264470588235294, "grad_norm": 0.35638892864770183, "learning_rate": 0.00025213756256951234, "loss": 0.1289, "step": 4812 }, { "epoch": 2.2649411764705882, "grad_norm": 0.9067358488147585, "learning_rate": 0.00025210545432357076, "loss": 0.5586, "step": 4813 }, { "epoch": 2.2654117647058825, "grad_norm": 0.4164569631630565, "learning_rate": 0.0002520733373573225, "loss": 0.2266, "step": 4814 }, { "epoch": 2.2658823529411762, "grad_norm": 0.2303591157720705, "learning_rate": 0.0002520412116735106, "loss": 0.084, "step": 4815 }, { "epoch": 2.2663529411764705, "grad_norm": 0.42999197354885227, "learning_rate": 0.0002520090772748787, "loss": 0.1455, "step": 4816 }, { "epoch": 2.2668235294117647, "grad_norm": 0.6074612810874049, "learning_rate": 0.00025197693416417126, "loss": 0.375, "step": 4817 }, { "epoch": 2.267294117647059, "grad_norm": 0.35559724045224045, "learning_rate": 0.0002519447823441334, "loss": 0.1807, "step": 4818 }, { "epoch": 2.267764705882353, "grad_norm": 0.9729230992018347, "learning_rate": 0.00025191262181751113, "loss": 0.3301, "step": 4819 }, { "epoch": 2.268235294117647, "grad_norm": 1.00052024467795, "learning_rate": 0.00025188045258705107, "loss": 0.5625, "step": 4820 }, { "epoch": 2.268705882352941, "grad_norm": 0.5763001248819799, "learning_rate": 0.0002518482746555007, "loss": 0.3242, "step": 4821 }, { "epoch": 2.2691764705882354, "grad_norm": 0.27789248028386965, "learning_rate": 0.000251816088025608, "loss": 0.127, "step": 4822 }, { "epoch": 2.2696470588235296, "grad_norm": 0.8153102614349669, "learning_rate": 0.00025178389270012205, "loss": 0.5, "step": 4823 }, { "epoch": 2.2701176470588234, "grad_norm": 0.4297591030147104, "learning_rate": 0.0002517516886817924, "loss": 0.1152, "step": 4824 }, { "epoch": 2.2705882352941176, "grad_norm": 0.9532102144512983, "learning_rate": 0.00025171947597336946, "loss": 0.2891, "step": 4825 }, { "epoch": 2.271058823529412, "grad_norm": 0.6128613730711236, "learning_rate": 0.00025168725457760434, "loss": 0.2441, "step": 4826 }, { "epoch": 2.271529411764706, "grad_norm": 0.20429436926300876, "learning_rate": 0.0002516550244972489, "loss": 0.0967, "step": 4827 }, { "epoch": 2.2720000000000002, "grad_norm": 0.6838447151737324, "learning_rate": 0.00025162278573505575, "loss": 0.3945, "step": 4828 }, { "epoch": 2.272470588235294, "grad_norm": 0.26119743645576515, "learning_rate": 0.0002515905382937782, "loss": 0.1089, "step": 4829 }, { "epoch": 2.2729411764705882, "grad_norm": 1.302689306748011, "learning_rate": 0.0002515582821761704, "loss": 0.5469, "step": 4830 }, { "epoch": 2.2734117647058825, "grad_norm": 0.44982268166806133, "learning_rate": 0.0002515260173849872, "loss": 0.1543, "step": 4831 }, { "epoch": 2.2738823529411762, "grad_norm": 0.7431725356950288, "learning_rate": 0.00025149374392298407, "loss": 0.3652, "step": 4832 }, { "epoch": 2.2743529411764705, "grad_norm": 0.2622938512008574, "learning_rate": 0.00025146146179291746, "loss": 0.1123, "step": 4833 }, { "epoch": 2.2748235294117647, "grad_norm": 1.2129017894277068, "learning_rate": 0.0002514291709975443, "loss": 0.6328, "step": 4834 }, { "epoch": 2.275294117647059, "grad_norm": 0.9226024192686515, "learning_rate": 0.00025139687153962246, "loss": 0.3145, "step": 4835 }, { "epoch": 2.275764705882353, "grad_norm": 0.5095490951955288, "learning_rate": 0.00025136456342191043, "loss": 0.2578, "step": 4836 }, { "epoch": 2.276235294117647, "grad_norm": 0.40290380212384624, "learning_rate": 0.0002513322466471675, "loss": 0.1318, "step": 4837 }, { "epoch": 2.276705882352941, "grad_norm": 0.55285215835112, "learning_rate": 0.00025129992121815365, "loss": 0.3047, "step": 4838 }, { "epoch": 2.2771764705882354, "grad_norm": 0.6638953718351187, "learning_rate": 0.00025126758713762964, "loss": 0.4004, "step": 4839 }, { "epoch": 2.2776470588235296, "grad_norm": 0.7474803334399567, "learning_rate": 0.000251235244408357, "loss": 0.3613, "step": 4840 }, { "epoch": 2.2781176470588234, "grad_norm": 0.7667567757700962, "learning_rate": 0.00025120289303309795, "loss": 0.3789, "step": 4841 }, { "epoch": 2.2785882352941176, "grad_norm": 0.1732937172751923, "learning_rate": 0.00025117053301461546, "loss": 0.0894, "step": 4842 }, { "epoch": 2.279058823529412, "grad_norm": 0.9910713499425761, "learning_rate": 0.00025113816435567315, "loss": 0.4297, "step": 4843 }, { "epoch": 2.279529411764706, "grad_norm": 0.47753231340468233, "learning_rate": 0.00025110578705903557, "loss": 0.2051, "step": 4844 }, { "epoch": 2.2800000000000002, "grad_norm": 1.1221451125468227, "learning_rate": 0.0002510734011274678, "loss": 0.6328, "step": 4845 }, { "epoch": 2.280470588235294, "grad_norm": 0.5939393361721864, "learning_rate": 0.00025104100656373585, "loss": 0.168, "step": 4846 }, { "epoch": 2.2809411764705882, "grad_norm": 0.41032598291471634, "learning_rate": 0.0002510086033706063, "loss": 0.1787, "step": 4847 }, { "epoch": 2.2814117647058825, "grad_norm": 0.3516606111435239, "learning_rate": 0.0002509761915508466, "loss": 0.1074, "step": 4848 }, { "epoch": 2.2818823529411763, "grad_norm": 0.16527743502710932, "learning_rate": 0.0002509437711072249, "loss": 0.0449, "step": 4849 }, { "epoch": 2.2823529411764705, "grad_norm": 1.8489551340737047, "learning_rate": 0.00025091134204250997, "loss": 0.7461, "step": 4850 }, { "epoch": 2.2828235294117647, "grad_norm": 0.6843446606856772, "learning_rate": 0.0002508789043594715, "loss": 0.2949, "step": 4851 }, { "epoch": 2.283294117647059, "grad_norm": 0.5426544997931458, "learning_rate": 0.00025084645806087974, "loss": 0.2793, "step": 4852 }, { "epoch": 2.283764705882353, "grad_norm": 0.14001435042715815, "learning_rate": 0.00025081400314950586, "loss": 0.0216, "step": 4853 }, { "epoch": 2.284235294117647, "grad_norm": 0.4146192228323712, "learning_rate": 0.0002507815396281216, "loss": 0.1445, "step": 4854 }, { "epoch": 2.284705882352941, "grad_norm": 0.781993859282193, "learning_rate": 0.0002507490674994995, "loss": 0.1475, "step": 4855 }, { "epoch": 2.2851764705882354, "grad_norm": 0.28811727428403106, "learning_rate": 0.0002507165867664129, "loss": 0.1367, "step": 4856 }, { "epoch": 2.2856470588235296, "grad_norm": 1.0879318541687981, "learning_rate": 0.0002506840974316357, "loss": 0.3867, "step": 4857 }, { "epoch": 2.2861176470588234, "grad_norm": 0.21849907115026795, "learning_rate": 0.00025065159949794283, "loss": 0.0747, "step": 4858 }, { "epoch": 2.2865882352941176, "grad_norm": 0.643804346639359, "learning_rate": 0.00025061909296810966, "loss": 0.3398, "step": 4859 }, { "epoch": 2.287058823529412, "grad_norm": 0.34656679220485714, "learning_rate": 0.0002505865778449124, "loss": 0.1445, "step": 4860 }, { "epoch": 2.287529411764706, "grad_norm": 0.6402798446447823, "learning_rate": 0.000250554054131128, "loss": 0.416, "step": 4861 }, { "epoch": 2.288, "grad_norm": 0.786406284957311, "learning_rate": 0.00025052152182953414, "loss": 0.5078, "step": 4862 }, { "epoch": 2.288470588235294, "grad_norm": 0.07843496904188749, "learning_rate": 0.0002504889809429093, "loss": 0.0415, "step": 4863 }, { "epoch": 2.2889411764705883, "grad_norm": 0.4640857931404265, "learning_rate": 0.0002504564314740326, "loss": 0.167, "step": 4864 }, { "epoch": 2.2894117647058825, "grad_norm": 0.5399259076809724, "learning_rate": 0.0002504238734256839, "loss": 0.1865, "step": 4865 }, { "epoch": 2.2898823529411763, "grad_norm": 0.18571129926585062, "learning_rate": 0.0002503913068006438, "loss": 0.085, "step": 4866 }, { "epoch": 2.2903529411764705, "grad_norm": 0.461143632830966, "learning_rate": 0.00025035873160169373, "loss": 0.1846, "step": 4867 }, { "epoch": 2.2908235294117647, "grad_norm": 0.9196119164565877, "learning_rate": 0.0002503261478316157, "loss": 0.5312, "step": 4868 }, { "epoch": 2.291294117647059, "grad_norm": 0.4873505777288735, "learning_rate": 0.0002502935554931926, "loss": 0.1963, "step": 4869 }, { "epoch": 2.291764705882353, "grad_norm": 0.6960147027814763, "learning_rate": 0.0002502609545892079, "loss": 0.2695, "step": 4870 }, { "epoch": 2.292235294117647, "grad_norm": 0.6017806669727108, "learning_rate": 0.00025022834512244583, "loss": 0.293, "step": 4871 }, { "epoch": 2.292705882352941, "grad_norm": 0.6190296402646038, "learning_rate": 0.00025019572709569153, "loss": 0.3457, "step": 4872 }, { "epoch": 2.2931764705882354, "grad_norm": 0.9694632696591824, "learning_rate": 0.0002501631005117307, "loss": 0.3926, "step": 4873 }, { "epoch": 2.2936470588235296, "grad_norm": 1.2110684514709598, "learning_rate": 0.0002501304653733497, "loss": 0.5547, "step": 4874 }, { "epoch": 2.2941176470588234, "grad_norm": 0.5787845780658014, "learning_rate": 0.0002500978216833359, "loss": 0.2852, "step": 4875 }, { "epoch": 2.2945882352941176, "grad_norm": 0.19451837768388644, "learning_rate": 0.0002500651694444771, "loss": 0.1167, "step": 4876 }, { "epoch": 2.295058823529412, "grad_norm": 0.5212224079292161, "learning_rate": 0.00025003250865956203, "loss": 0.2285, "step": 4877 }, { "epoch": 2.295529411764706, "grad_norm": 1.0810595670206153, "learning_rate": 0.0002499998393313801, "loss": 0.4805, "step": 4878 }, { "epoch": 2.296, "grad_norm": 0.710008537945075, "learning_rate": 0.00024996716146272137, "loss": 0.3242, "step": 4879 }, { "epoch": 2.296470588235294, "grad_norm": 0.648459343980186, "learning_rate": 0.0002499344750563767, "loss": 0.3047, "step": 4880 }, { "epoch": 2.2969411764705883, "grad_norm": 0.4303688262092395, "learning_rate": 0.0002499017801151377, "loss": 0.2119, "step": 4881 }, { "epoch": 2.2974117647058825, "grad_norm": 0.5011312766192015, "learning_rate": 0.0002498690766417967, "loss": 0.2695, "step": 4882 }, { "epoch": 2.2978823529411763, "grad_norm": 0.4066942519406138, "learning_rate": 0.00024983636463914665, "loss": 0.209, "step": 4883 }, { "epoch": 2.2983529411764705, "grad_norm": 0.7877819121919463, "learning_rate": 0.00024980364410998133, "loss": 0.4102, "step": 4884 }, { "epoch": 2.2988235294117647, "grad_norm": 0.7762017973420764, "learning_rate": 0.0002497709150570953, "loss": 0.3379, "step": 4885 }, { "epoch": 2.299294117647059, "grad_norm": 0.3702757893160591, "learning_rate": 0.0002497381774832838, "loss": 0.1406, "step": 4886 }, { "epoch": 2.299764705882353, "grad_norm": 0.4842965462810677, "learning_rate": 0.00024970543139134275, "loss": 0.2773, "step": 4887 }, { "epoch": 2.300235294117647, "grad_norm": 0.32535425076503016, "learning_rate": 0.0002496726767840687, "loss": 0.1001, "step": 4888 }, { "epoch": 2.300705882352941, "grad_norm": 0.43491490886273965, "learning_rate": 0.0002496399136642592, "loss": 0.3027, "step": 4889 }, { "epoch": 2.3011764705882354, "grad_norm": 0.7665507886322744, "learning_rate": 0.0002496071420347124, "loss": 0.3652, "step": 4890 }, { "epoch": 2.3016470588235296, "grad_norm": 0.5901613100582624, "learning_rate": 0.00024957436189822706, "loss": 0.2891, "step": 4891 }, { "epoch": 2.3021176470588234, "grad_norm": 0.5426400110603145, "learning_rate": 0.0002495415732576028, "loss": 0.2695, "step": 4892 }, { "epoch": 2.3025882352941176, "grad_norm": 0.7996571648943143, "learning_rate": 0.00024950877611564, "loss": 0.5039, "step": 4893 }, { "epoch": 2.303058823529412, "grad_norm": 0.5272054781444143, "learning_rate": 0.0002494759704751395, "loss": 0.252, "step": 4894 }, { "epoch": 2.303529411764706, "grad_norm": 0.5074560730565847, "learning_rate": 0.00024944315633890327, "loss": 0.3047, "step": 4895 }, { "epoch": 2.304, "grad_norm": 0.44481985832940457, "learning_rate": 0.0002494103337097337, "loss": 0.1338, "step": 4896 }, { "epoch": 2.304470588235294, "grad_norm": 0.6116735055971195, "learning_rate": 0.00024937750259043407, "loss": 0.4258, "step": 4897 }, { "epoch": 2.3049411764705883, "grad_norm": 0.5764357579116152, "learning_rate": 0.00024934466298380824, "loss": 0.2412, "step": 4898 }, { "epoch": 2.3054117647058825, "grad_norm": 0.6322483346030578, "learning_rate": 0.0002493118148926609, "loss": 0.2734, "step": 4899 }, { "epoch": 2.3058823529411763, "grad_norm": 0.9672791235951634, "learning_rate": 0.00024927895831979745, "loss": 0.2441, "step": 4900 }, { "epoch": 2.3063529411764705, "grad_norm": 0.8650166935218045, "learning_rate": 0.000249246093268024, "loss": 0.4277, "step": 4901 }, { "epoch": 2.3068235294117647, "grad_norm": 0.7043344299423698, "learning_rate": 0.0002492132197401474, "loss": 0.2363, "step": 4902 }, { "epoch": 2.307294117647059, "grad_norm": 0.669017184673712, "learning_rate": 0.0002491803377389752, "loss": 0.4004, "step": 4903 }, { "epoch": 2.307764705882353, "grad_norm": 0.8951919216623455, "learning_rate": 0.0002491474472673157, "loss": 0.5273, "step": 4904 }, { "epoch": 2.308235294117647, "grad_norm": 1.0021276155256174, "learning_rate": 0.00024911454832797784, "loss": 0.6133, "step": 4905 }, { "epoch": 2.308705882352941, "grad_norm": 0.3498956316521572, "learning_rate": 0.0002490816409237714, "loss": 0.0889, "step": 4906 }, { "epoch": 2.3091764705882354, "grad_norm": 0.629083604861006, "learning_rate": 0.0002490487250575069, "loss": 0.4082, "step": 4907 }, { "epoch": 2.309647058823529, "grad_norm": 0.09746228486353385, "learning_rate": 0.00024901580073199536, "loss": 0.0503, "step": 4908 }, { "epoch": 2.3101176470588234, "grad_norm": 0.5697312541662328, "learning_rate": 0.0002489828679500488, "loss": 0.25, "step": 4909 }, { "epoch": 2.3105882352941176, "grad_norm": 0.4721926237543816, "learning_rate": 0.00024894992671447983, "loss": 0.2061, "step": 4910 }, { "epoch": 2.311058823529412, "grad_norm": 0.6636142140371338, "learning_rate": 0.0002489169770281018, "loss": 0.3145, "step": 4911 }, { "epoch": 2.311529411764706, "grad_norm": 0.7922000714086018, "learning_rate": 0.0002488840188937287, "loss": 0.3789, "step": 4912 }, { "epoch": 2.312, "grad_norm": 0.5673154084963069, "learning_rate": 0.00024885105231417536, "loss": 0.4531, "step": 4913 }, { "epoch": 2.312470588235294, "grad_norm": 0.5608118131703475, "learning_rate": 0.00024881807729225727, "loss": 0.2695, "step": 4914 }, { "epoch": 2.3129411764705883, "grad_norm": 0.5366372650202932, "learning_rate": 0.00024878509383079075, "loss": 0.2354, "step": 4915 }, { "epoch": 2.3134117647058825, "grad_norm": 0.5580203236892662, "learning_rate": 0.0002487521019325926, "loss": 0.2354, "step": 4916 }, { "epoch": 2.3138823529411763, "grad_norm": 0.5651776423204833, "learning_rate": 0.00024871910160048063, "loss": 0.2324, "step": 4917 }, { "epoch": 2.3143529411764705, "grad_norm": 0.5656631884623008, "learning_rate": 0.00024868609283727316, "loss": 0.2871, "step": 4918 }, { "epoch": 2.3148235294117647, "grad_norm": 0.4107954533373266, "learning_rate": 0.00024865307564578927, "loss": 0.2119, "step": 4919 }, { "epoch": 2.315294117647059, "grad_norm": 0.295610738873652, "learning_rate": 0.00024862005002884886, "loss": 0.0815, "step": 4920 }, { "epoch": 2.315764705882353, "grad_norm": 0.47186665438009195, "learning_rate": 0.0002485870159892724, "loss": 0.2031, "step": 4921 }, { "epoch": 2.316235294117647, "grad_norm": 0.8161548096168095, "learning_rate": 0.00024855397352988127, "loss": 0.4551, "step": 4922 }, { "epoch": 2.316705882352941, "grad_norm": 0.19847048814207144, "learning_rate": 0.0002485209226534974, "loss": 0.0481, "step": 4923 }, { "epoch": 2.3171764705882354, "grad_norm": 0.5752985919085581, "learning_rate": 0.00024848786336294346, "loss": 0.1797, "step": 4924 }, { "epoch": 2.317647058823529, "grad_norm": 0.7007606360931382, "learning_rate": 0.0002484547956610429, "loss": 0.3145, "step": 4925 }, { "epoch": 2.3181176470588234, "grad_norm": 0.9457046143457716, "learning_rate": 0.0002484217195506199, "loss": 0.5547, "step": 4926 }, { "epoch": 2.3185882352941176, "grad_norm": 0.5626037506287279, "learning_rate": 0.00024838863503449926, "loss": 0.3047, "step": 4927 }, { "epoch": 2.319058823529412, "grad_norm": 0.9850186337343574, "learning_rate": 0.00024835554211550664, "loss": 0.4902, "step": 4928 }, { "epoch": 2.319529411764706, "grad_norm": 0.7343041234650929, "learning_rate": 0.00024832244079646827, "loss": 0.2461, "step": 4929 }, { "epoch": 2.32, "grad_norm": 0.6869557790652765, "learning_rate": 0.00024828933108021116, "loss": 0.373, "step": 4930 }, { "epoch": 2.320470588235294, "grad_norm": 0.31039915934774653, "learning_rate": 0.00024825621296956313, "loss": 0.0938, "step": 4931 }, { "epoch": 2.3209411764705883, "grad_norm": 0.31156904797734936, "learning_rate": 0.00024822308646735254, "loss": 0.1055, "step": 4932 }, { "epoch": 2.3214117647058825, "grad_norm": 0.7201209215995821, "learning_rate": 0.00024818995157640855, "loss": 0.2168, "step": 4933 }, { "epoch": 2.3218823529411763, "grad_norm": 0.7397000128543164, "learning_rate": 0.00024815680829956115, "loss": 0.2539, "step": 4934 }, { "epoch": 2.3223529411764705, "grad_norm": 1.1255252662002617, "learning_rate": 0.0002481236566396408, "loss": 0.5703, "step": 4935 }, { "epoch": 2.3228235294117647, "grad_norm": 0.7667396794219558, "learning_rate": 0.0002480904965994789, "loss": 0.4805, "step": 4936 }, { "epoch": 2.323294117647059, "grad_norm": 0.4150304507552023, "learning_rate": 0.00024805732818190747, "loss": 0.1348, "step": 4937 }, { "epoch": 2.323764705882353, "grad_norm": 0.8537960543405843, "learning_rate": 0.00024802415138975924, "loss": 0.416, "step": 4938 }, { "epoch": 2.324235294117647, "grad_norm": 1.2131278666779917, "learning_rate": 0.00024799096622586767, "loss": 0.3613, "step": 4939 }, { "epoch": 2.324705882352941, "grad_norm": 0.7707403531336795, "learning_rate": 0.00024795777269306696, "loss": 0.3066, "step": 4940 }, { "epoch": 2.3251764705882354, "grad_norm": 0.7870222670667114, "learning_rate": 0.0002479245707941919, "loss": 0.4902, "step": 4941 }, { "epoch": 2.325647058823529, "grad_norm": 0.3551906560656289, "learning_rate": 0.0002478913605320783, "loss": 0.1572, "step": 4942 }, { "epoch": 2.3261176470588234, "grad_norm": 0.7330743843550657, "learning_rate": 0.0002478581419095623, "loss": 0.5742, "step": 4943 }, { "epoch": 2.3265882352941176, "grad_norm": 0.8007161096291721, "learning_rate": 0.0002478249149294809, "loss": 0.375, "step": 4944 }, { "epoch": 2.327058823529412, "grad_norm": 1.3588222697631513, "learning_rate": 0.00024779167959467207, "loss": 0.5781, "step": 4945 }, { "epoch": 2.327529411764706, "grad_norm": 0.3676737774556191, "learning_rate": 0.0002477584359079741, "loss": 0.1299, "step": 4946 }, { "epoch": 2.328, "grad_norm": 0.4793881432094604, "learning_rate": 0.00024772518387222616, "loss": 0.1562, "step": 4947 }, { "epoch": 2.328470588235294, "grad_norm": 0.4586525417043881, "learning_rate": 0.0002476919234902682, "loss": 0.1279, "step": 4948 }, { "epoch": 2.3289411764705883, "grad_norm": 0.8134719163161219, "learning_rate": 0.0002476586547649408, "loss": 0.373, "step": 4949 }, { "epoch": 2.3294117647058825, "grad_norm": 0.5896009670310772, "learning_rate": 0.00024762537769908535, "loss": 0.167, "step": 4950 }, { "epoch": 2.3298823529411763, "grad_norm": 0.3824913950666755, "learning_rate": 0.0002475920922955437, "loss": 0.2402, "step": 4951 }, { "epoch": 2.3303529411764705, "grad_norm": 0.46957539355645994, "learning_rate": 0.0002475587985571587, "loss": 0.1758, "step": 4952 }, { "epoch": 2.3308235294117647, "grad_norm": 0.7079543281712112, "learning_rate": 0.0002475254964867738, "loss": 0.1748, "step": 4953 }, { "epoch": 2.331294117647059, "grad_norm": 0.7815260428276413, "learning_rate": 0.0002474921860872331, "loss": 0.3691, "step": 4954 }, { "epoch": 2.331764705882353, "grad_norm": 0.9501176722382467, "learning_rate": 0.00024745886736138154, "loss": 0.4746, "step": 4955 }, { "epoch": 2.332235294117647, "grad_norm": 0.6132443208196687, "learning_rate": 0.0002474255403120647, "loss": 0.2793, "step": 4956 }, { "epoch": 2.332705882352941, "grad_norm": 0.3991118182734481, "learning_rate": 0.0002473922049421288, "loss": 0.2363, "step": 4957 }, { "epoch": 2.3331764705882354, "grad_norm": 0.24605883625924224, "learning_rate": 0.0002473588612544209, "loss": 0.0737, "step": 4958 }, { "epoch": 2.333647058823529, "grad_norm": 0.07670495003559365, "learning_rate": 0.0002473255092517887, "loss": 0.011, "step": 4959 }, { "epoch": 2.3341176470588234, "grad_norm": 0.6157157907103984, "learning_rate": 0.0002472921489370807, "loss": 0.1826, "step": 4960 }, { "epoch": 2.3345882352941176, "grad_norm": 0.4937919610285864, "learning_rate": 0.0002472587803131459, "loss": 0.1865, "step": 4961 }, { "epoch": 2.335058823529412, "grad_norm": 0.46309447829626776, "learning_rate": 0.0002472254033828342, "loss": 0.1445, "step": 4962 }, { "epoch": 2.335529411764706, "grad_norm": 0.7442682254919171, "learning_rate": 0.00024719201814899617, "loss": 0.4023, "step": 4963 }, { "epoch": 2.336, "grad_norm": 0.8800201155987945, "learning_rate": 0.0002471586246144831, "loss": 0.4336, "step": 4964 }, { "epoch": 2.336470588235294, "grad_norm": 0.996691036940441, "learning_rate": 0.0002471252227821469, "loss": 0.5664, "step": 4965 }, { "epoch": 2.3369411764705883, "grad_norm": 0.610301226132299, "learning_rate": 0.0002470918126548403, "loss": 0.2852, "step": 4966 }, { "epoch": 2.3374117647058825, "grad_norm": 0.7554530470437539, "learning_rate": 0.00024705839423541664, "loss": 0.3574, "step": 4967 }, { "epoch": 2.3378823529411763, "grad_norm": 0.5189457791906937, "learning_rate": 0.00024702496752673005, "loss": 0.2734, "step": 4968 }, { "epoch": 2.3383529411764705, "grad_norm": 0.9427210911857005, "learning_rate": 0.00024699153253163533, "loss": 0.4727, "step": 4969 }, { "epoch": 2.3388235294117647, "grad_norm": 0.6903514847344661, "learning_rate": 0.000246958089252988, "loss": 0.4082, "step": 4970 }, { "epoch": 2.339294117647059, "grad_norm": 0.42220878563741243, "learning_rate": 0.0002469246376936443, "loss": 0.1797, "step": 4971 }, { "epoch": 2.339764705882353, "grad_norm": 0.10463330856069322, "learning_rate": 0.00024689117785646106, "loss": 0.0137, "step": 4972 }, { "epoch": 2.340235294117647, "grad_norm": 0.4940935964349975, "learning_rate": 0.000246857709744296, "loss": 0.1777, "step": 4973 }, { "epoch": 2.340705882352941, "grad_norm": 0.6541226675965064, "learning_rate": 0.00024682423336000746, "loss": 0.2051, "step": 4974 }, { "epoch": 2.3411764705882354, "grad_norm": 0.4100809006196458, "learning_rate": 0.00024679074870645446, "loss": 0.126, "step": 4975 }, { "epoch": 2.341647058823529, "grad_norm": 0.36200477725751895, "learning_rate": 0.0002467572557864968, "loss": 0.1367, "step": 4976 }, { "epoch": 2.3421176470588234, "grad_norm": 0.538155863649543, "learning_rate": 0.0002467237546029949, "loss": 0.2432, "step": 4977 }, { "epoch": 2.3425882352941176, "grad_norm": 0.9211634563981537, "learning_rate": 0.0002466902451588099, "loss": 0.3867, "step": 4978 }, { "epoch": 2.343058823529412, "grad_norm": 0.686363679220836, "learning_rate": 0.0002466567274568037, "loss": 0.2812, "step": 4979 }, { "epoch": 2.343529411764706, "grad_norm": 0.4869176423026318, "learning_rate": 0.0002466232014998389, "loss": 0.2324, "step": 4980 }, { "epoch": 2.344, "grad_norm": 0.5728629209438983, "learning_rate": 0.00024658966729077875, "loss": 0.334, "step": 4981 }, { "epoch": 2.344470588235294, "grad_norm": 0.742737204130242, "learning_rate": 0.00024655612483248727, "loss": 0.3379, "step": 4982 }, { "epoch": 2.3449411764705883, "grad_norm": 0.40833987782376585, "learning_rate": 0.00024652257412782904, "loss": 0.21, "step": 4983 }, { "epoch": 2.3454117647058825, "grad_norm": 0.6245137571762436, "learning_rate": 0.0002464890151796696, "loss": 0.4668, "step": 4984 }, { "epoch": 2.3458823529411763, "grad_norm": 0.3157018438293412, "learning_rate": 0.000246455447990875, "loss": 0.1064, "step": 4985 }, { "epoch": 2.3463529411764705, "grad_norm": 0.7407260793688563, "learning_rate": 0.000246421872564312, "loss": 0.4258, "step": 4986 }, { "epoch": 2.3468235294117648, "grad_norm": 0.6167931676195839, "learning_rate": 0.0002463882889028481, "loss": 0.3086, "step": 4987 }, { "epoch": 2.347294117647059, "grad_norm": 0.6665804118754938, "learning_rate": 0.0002463546970093516, "loss": 0.4336, "step": 4988 }, { "epoch": 2.3477647058823528, "grad_norm": 0.47910788088724937, "learning_rate": 0.00024632109688669135, "loss": 0.3496, "step": 4989 }, { "epoch": 2.348235294117647, "grad_norm": 0.8713590884418393, "learning_rate": 0.000246287488537737, "loss": 0.4043, "step": 4990 }, { "epoch": 2.348705882352941, "grad_norm": 0.5096427034634736, "learning_rate": 0.00024625387196535885, "loss": 0.248, "step": 4991 }, { "epoch": 2.3491764705882354, "grad_norm": 1.3205881995565136, "learning_rate": 0.0002462202471724278, "loss": 0.793, "step": 4992 }, { "epoch": 2.349647058823529, "grad_norm": 1.0179581031253282, "learning_rate": 0.0002461866141618158, "loss": 0.6367, "step": 4993 }, { "epoch": 2.3501176470588234, "grad_norm": 0.641619542045715, "learning_rate": 0.0002461529729363951, "loss": 0.3965, "step": 4994 }, { "epoch": 2.3505882352941176, "grad_norm": 0.4454467518088805, "learning_rate": 0.0002461193234990389, "loss": 0.1807, "step": 4995 }, { "epoch": 2.351058823529412, "grad_norm": 0.840413081430781, "learning_rate": 0.00024608566585262103, "loss": 0.377, "step": 4996 }, { "epoch": 2.351529411764706, "grad_norm": 0.8499857604590335, "learning_rate": 0.000246052000000016, "loss": 0.4355, "step": 4997 }, { "epoch": 2.352, "grad_norm": 0.9823400892389513, "learning_rate": 0.00024601832594409905, "loss": 0.3984, "step": 4998 }, { "epoch": 2.352470588235294, "grad_norm": 0.5564535728602248, "learning_rate": 0.00024598464368774604, "loss": 0.3281, "step": 4999 }, { "epoch": 2.3529411764705883, "grad_norm": 0.5876505415946125, "learning_rate": 0.00024595095323383365, "loss": 0.3262, "step": 5000 }, { "epoch": 2.3534117647058825, "grad_norm": 0.6813695679487884, "learning_rate": 0.0002459172545852393, "loss": 0.5234, "step": 5001 }, { "epoch": 2.3538823529411763, "grad_norm": 0.5799505769686545, "learning_rate": 0.0002458835477448408, "loss": 0.2119, "step": 5002 }, { "epoch": 2.3543529411764705, "grad_norm": 0.7322259159547697, "learning_rate": 0.00024584983271551717, "loss": 0.5703, "step": 5003 }, { "epoch": 2.3548235294117648, "grad_norm": 0.890099556341197, "learning_rate": 0.0002458161095001476, "loss": 0.457, "step": 5004 }, { "epoch": 2.355294117647059, "grad_norm": 0.7570049502053894, "learning_rate": 0.0002457823781016123, "loss": 0.2559, "step": 5005 }, { "epoch": 2.3557647058823528, "grad_norm": 0.2798244697318081, "learning_rate": 0.0002457486385227922, "loss": 0.0786, "step": 5006 }, { "epoch": 2.356235294117647, "grad_norm": 0.31264867177805494, "learning_rate": 0.00024571489076656865, "loss": 0.0781, "step": 5007 }, { "epoch": 2.356705882352941, "grad_norm": 0.28070623460942484, "learning_rate": 0.0002456811348358239, "loss": 0.0859, "step": 5008 }, { "epoch": 2.3571764705882354, "grad_norm": 0.7123583294352264, "learning_rate": 0.000245647370733441, "loss": 0.3457, "step": 5009 }, { "epoch": 2.357647058823529, "grad_norm": 0.6233218256017379, "learning_rate": 0.0002456135984623034, "loss": 0.1748, "step": 5010 }, { "epoch": 2.3581176470588234, "grad_norm": 0.576403788211275, "learning_rate": 0.0002455798180252956, "loss": 0.2188, "step": 5011 }, { "epoch": 2.3585882352941177, "grad_norm": 0.8620771071397318, "learning_rate": 0.0002455460294253025, "loss": 0.4219, "step": 5012 }, { "epoch": 2.359058823529412, "grad_norm": 0.580708154019551, "learning_rate": 0.00024551223266520983, "loss": 0.3711, "step": 5013 }, { "epoch": 2.359529411764706, "grad_norm": 0.3077911754126724, "learning_rate": 0.00024547842774790403, "loss": 0.0801, "step": 5014 }, { "epoch": 2.36, "grad_norm": 0.4253806267064142, "learning_rate": 0.0002454446146762721, "loss": 0.0874, "step": 5015 }, { "epoch": 2.360470588235294, "grad_norm": 0.33218882580464604, "learning_rate": 0.000245410793453202, "loss": 0.1172, "step": 5016 }, { "epoch": 2.3609411764705883, "grad_norm": 0.7273749602982771, "learning_rate": 0.00024537696408158215, "loss": 0.1641, "step": 5017 }, { "epoch": 2.3614117647058825, "grad_norm": 0.2260396684157586, "learning_rate": 0.0002453431265643017, "loss": 0.0571, "step": 5018 }, { "epoch": 2.3618823529411763, "grad_norm": 0.9500918493057616, "learning_rate": 0.0002453092809042507, "loss": 0.3945, "step": 5019 }, { "epoch": 2.3623529411764705, "grad_norm": 0.7610159039683726, "learning_rate": 0.0002452754271043195, "loss": 0.3496, "step": 5020 }, { "epoch": 2.3628235294117648, "grad_norm": 0.5259641433720941, "learning_rate": 0.0002452415651673995, "loss": 0.2188, "step": 5021 }, { "epoch": 2.363294117647059, "grad_norm": 0.9084452573970885, "learning_rate": 0.00024520769509638274, "loss": 0.582, "step": 5022 }, { "epoch": 2.3637647058823528, "grad_norm": 0.760306807543429, "learning_rate": 0.0002451738168941618, "loss": 0.459, "step": 5023 }, { "epoch": 2.364235294117647, "grad_norm": 0.9613334398959105, "learning_rate": 0.0002451399305636301, "loss": 0.2812, "step": 5024 }, { "epoch": 2.364705882352941, "grad_norm": 0.5784848986558229, "learning_rate": 0.00024510603610768166, "loss": 0.3906, "step": 5025 }, { "epoch": 2.3651764705882354, "grad_norm": 0.8654692336773477, "learning_rate": 0.0002450721335292112, "loss": 0.5156, "step": 5026 }, { "epoch": 2.365647058823529, "grad_norm": 0.8847621511272054, "learning_rate": 0.00024503822283111423, "loss": 0.4395, "step": 5027 }, { "epoch": 2.3661176470588234, "grad_norm": 0.3772938350907761, "learning_rate": 0.0002450043040162869, "loss": 0.1445, "step": 5028 }, { "epoch": 2.3665882352941177, "grad_norm": 0.563389110357283, "learning_rate": 0.000244970377087626, "loss": 0.2285, "step": 5029 }, { "epoch": 2.367058823529412, "grad_norm": 0.381470288117856, "learning_rate": 0.00024493644204802906, "loss": 0.2041, "step": 5030 }, { "epoch": 2.367529411764706, "grad_norm": 0.2018156140595798, "learning_rate": 0.0002449024989003943, "loss": 0.0422, "step": 5031 }, { "epoch": 2.368, "grad_norm": 0.5304968438676998, "learning_rate": 0.0002448685476476207, "loss": 0.2559, "step": 5032 }, { "epoch": 2.368470588235294, "grad_norm": 0.07358690546335822, "learning_rate": 0.0002448345882926077, "loss": 0.0189, "step": 5033 }, { "epoch": 2.3689411764705883, "grad_norm": 0.666929988909018, "learning_rate": 0.0002448006208382558, "loss": 0.2188, "step": 5034 }, { "epoch": 2.369411764705882, "grad_norm": 0.7910202528269651, "learning_rate": 0.0002447666452874658, "loss": 0.2559, "step": 5035 }, { "epoch": 2.3698823529411763, "grad_norm": 0.8700151705791009, "learning_rate": 0.00024473266164313953, "loss": 0.4219, "step": 5036 }, { "epoch": 2.3703529411764706, "grad_norm": 0.6465387579337807, "learning_rate": 0.00024469866990817923, "loss": 0.3613, "step": 5037 }, { "epoch": 2.3708235294117648, "grad_norm": 2.0676931388394575, "learning_rate": 0.00024466467008548814, "loss": 1.5391, "step": 5038 }, { "epoch": 2.371294117647059, "grad_norm": 0.9614046470424926, "learning_rate": 0.0002446306621779698, "loss": 0.4824, "step": 5039 }, { "epoch": 2.3717647058823528, "grad_norm": 0.5022501587315182, "learning_rate": 0.00024459664618852877, "loss": 0.3652, "step": 5040 }, { "epoch": 2.372235294117647, "grad_norm": 0.5794487337911851, "learning_rate": 0.0002445626221200702, "loss": 0.3359, "step": 5041 }, { "epoch": 2.372705882352941, "grad_norm": 0.5415927467143092, "learning_rate": 0.0002445285899754998, "loss": 0.1289, "step": 5042 }, { "epoch": 2.3731764705882354, "grad_norm": 0.9022375415958175, "learning_rate": 0.0002444945497577242, "loss": 0.459, "step": 5043 }, { "epoch": 2.373647058823529, "grad_norm": 0.8322994023280142, "learning_rate": 0.0002444605014696506, "loss": 0.5234, "step": 5044 }, { "epoch": 2.3741176470588234, "grad_norm": 0.25945242089591064, "learning_rate": 0.0002444264451141868, "loss": 0.1191, "step": 5045 }, { "epoch": 2.3745882352941177, "grad_norm": 0.47727551998217993, "learning_rate": 0.0002443923806942414, "loss": 0.3047, "step": 5046 }, { "epoch": 2.375058823529412, "grad_norm": 0.7683693689742849, "learning_rate": 0.0002443583082127237, "loss": 0.4473, "step": 5047 }, { "epoch": 2.375529411764706, "grad_norm": 1.102270174009831, "learning_rate": 0.0002443242276725437, "loss": 0.7109, "step": 5048 }, { "epoch": 2.376, "grad_norm": 0.7948758471018498, "learning_rate": 0.00024429013907661197, "loss": 0.3652, "step": 5049 }, { "epoch": 2.376470588235294, "grad_norm": 0.5938305503364824, "learning_rate": 0.0002442560424278399, "loss": 0.3418, "step": 5050 }, { "epoch": 2.3769411764705883, "grad_norm": 0.28793948975904776, "learning_rate": 0.0002442219377291395, "loss": 0.0752, "step": 5051 }, { "epoch": 2.377411764705882, "grad_norm": 0.32116066900690776, "learning_rate": 0.00024418782498342336, "loss": 0.1226, "step": 5052 }, { "epoch": 2.3778823529411763, "grad_norm": 0.2716042430654707, "learning_rate": 0.00024415370419360506, "loss": 0.085, "step": 5053 }, { "epoch": 2.3783529411764706, "grad_norm": 0.47823194660356494, "learning_rate": 0.00024411957536259857, "loss": 0.1348, "step": 5054 }, { "epoch": 2.378823529411765, "grad_norm": 0.5866248104635597, "learning_rate": 0.0002440854384933187, "loss": 0.2949, "step": 5055 }, { "epoch": 2.379294117647059, "grad_norm": 0.5208184745016499, "learning_rate": 0.0002440512935886809, "loss": 0.21, "step": 5056 }, { "epoch": 2.379764705882353, "grad_norm": 0.24216227651897837, "learning_rate": 0.0002440171406516013, "loss": 0.0723, "step": 5057 }, { "epoch": 2.380235294117647, "grad_norm": 0.7033936957298588, "learning_rate": 0.00024398297968499677, "loss": 0.3691, "step": 5058 }, { "epoch": 2.3807058823529412, "grad_norm": 0.5773541721480262, "learning_rate": 0.00024394881069178477, "loss": 0.3887, "step": 5059 }, { "epoch": 2.3811764705882354, "grad_norm": 0.2964119957542581, "learning_rate": 0.00024391463367488348, "loss": 0.0669, "step": 5060 }, { "epoch": 2.3816470588235292, "grad_norm": 1.0628854631309639, "learning_rate": 0.00024388044863721188, "loss": 0.6602, "step": 5061 }, { "epoch": 2.3821176470588235, "grad_norm": 0.5962397324154697, "learning_rate": 0.00024384625558168948, "loss": 0.3438, "step": 5062 }, { "epoch": 2.3825882352941177, "grad_norm": 0.41857669242805223, "learning_rate": 0.0002438120545112365, "loss": 0.1196, "step": 5063 }, { "epoch": 2.383058823529412, "grad_norm": 0.608555903314444, "learning_rate": 0.000243777845428774, "loss": 0.332, "step": 5064 }, { "epoch": 2.383529411764706, "grad_norm": 0.7797774653969497, "learning_rate": 0.00024374362833722344, "loss": 0.3711, "step": 5065 }, { "epoch": 2.384, "grad_norm": 0.8600007629871061, "learning_rate": 0.00024370940323950724, "loss": 0.5547, "step": 5066 }, { "epoch": 2.384470588235294, "grad_norm": 0.8798867648365197, "learning_rate": 0.0002436751701385484, "loss": 0.6016, "step": 5067 }, { "epoch": 2.3849411764705883, "grad_norm": 0.6264430536391349, "learning_rate": 0.0002436409290372705, "loss": 0.3926, "step": 5068 }, { "epoch": 2.385411764705882, "grad_norm": 0.8617436661439033, "learning_rate": 0.000243606679938598, "loss": 0.4531, "step": 5069 }, { "epoch": 2.3858823529411763, "grad_norm": 0.4314905042854918, "learning_rate": 0.00024357242284545587, "loss": 0.2354, "step": 5070 }, { "epoch": 2.3863529411764706, "grad_norm": 0.3954280479306051, "learning_rate": 0.00024353815776076985, "loss": 0.167, "step": 5071 }, { "epoch": 2.386823529411765, "grad_norm": 0.658801691470938, "learning_rate": 0.00024350388468746635, "loss": 0.4707, "step": 5072 }, { "epoch": 2.387294117647059, "grad_norm": 0.7382049717179208, "learning_rate": 0.00024346960362847247, "loss": 0.5, "step": 5073 }, { "epoch": 2.387764705882353, "grad_norm": 0.6892978981554783, "learning_rate": 0.000243435314586716, "loss": 0.3926, "step": 5074 }, { "epoch": 2.388235294117647, "grad_norm": 0.6675964876331153, "learning_rate": 0.00024340101756512532, "loss": 0.3496, "step": 5075 }, { "epoch": 2.3887058823529412, "grad_norm": 0.49080908021117947, "learning_rate": 0.00024336671256662963, "loss": 0.4629, "step": 5076 }, { "epoch": 2.3891764705882355, "grad_norm": 1.084536656341079, "learning_rate": 0.00024333239959415873, "loss": 0.4629, "step": 5077 }, { "epoch": 2.3896470588235292, "grad_norm": 0.30471068530337, "learning_rate": 0.0002432980786506431, "loss": 0.1113, "step": 5078 }, { "epoch": 2.3901176470588235, "grad_norm": 0.6401942668644001, "learning_rate": 0.00024326374973901395, "loss": 0.3809, "step": 5079 }, { "epoch": 2.3905882352941177, "grad_norm": 0.4140713995093678, "learning_rate": 0.0002432294128622031, "loss": 0.2324, "step": 5080 }, { "epoch": 2.391058823529412, "grad_norm": 0.35411502763609237, "learning_rate": 0.0002431950680231431, "loss": 0.0928, "step": 5081 }, { "epoch": 2.391529411764706, "grad_norm": 0.326668582070955, "learning_rate": 0.00024316071522476718, "loss": 0.1201, "step": 5082 }, { "epoch": 2.392, "grad_norm": 0.7496644493792496, "learning_rate": 0.0002431263544700093, "loss": 0.6445, "step": 5083 }, { "epoch": 2.392470588235294, "grad_norm": 0.5366512268028542, "learning_rate": 0.00024309198576180388, "loss": 0.252, "step": 5084 }, { "epoch": 2.3929411764705883, "grad_norm": 0.8463672045082099, "learning_rate": 0.00024305760910308627, "loss": 0.4414, "step": 5085 }, { "epoch": 2.393411764705882, "grad_norm": 1.0186501070664233, "learning_rate": 0.00024302322449679245, "loss": 0.6797, "step": 5086 }, { "epoch": 2.3938823529411764, "grad_norm": 0.43873542437306734, "learning_rate": 0.0002429888319458589, "loss": 0.1758, "step": 5087 }, { "epoch": 2.3943529411764706, "grad_norm": 0.8132881484896033, "learning_rate": 0.00024295443145322303, "loss": 0.3418, "step": 5088 }, { "epoch": 2.394823529411765, "grad_norm": 0.9824907838180164, "learning_rate": 0.0002429200230218228, "loss": 0.5039, "step": 5089 }, { "epoch": 2.395294117647059, "grad_norm": 0.6733135083930363, "learning_rate": 0.00024288560665459679, "loss": 0.2656, "step": 5090 }, { "epoch": 2.395764705882353, "grad_norm": 1.3008993327662732, "learning_rate": 0.0002428511823544844, "loss": 0.4062, "step": 5091 }, { "epoch": 2.396235294117647, "grad_norm": 0.6585066450661475, "learning_rate": 0.00024281675012442563, "loss": 0.2451, "step": 5092 }, { "epoch": 2.3967058823529412, "grad_norm": 0.7798392857959129, "learning_rate": 0.00024278230996736109, "loss": 0.2891, "step": 5093 }, { "epoch": 2.3971764705882355, "grad_norm": 0.8130196470611057, "learning_rate": 0.0002427478618862322, "loss": 0.4023, "step": 5094 }, { "epoch": 2.3976470588235292, "grad_norm": 0.4153800802095795, "learning_rate": 0.00024271340588398097, "loss": 0.2021, "step": 5095 }, { "epoch": 2.3981176470588235, "grad_norm": 0.47192819367696254, "learning_rate": 0.00024267894196355015, "loss": 0.207, "step": 5096 }, { "epoch": 2.3985882352941177, "grad_norm": 0.4954959506724215, "learning_rate": 0.00024264447012788308, "loss": 0.1094, "step": 5097 }, { "epoch": 2.399058823529412, "grad_norm": 0.8074851699331167, "learning_rate": 0.00024260999037992384, "loss": 0.459, "step": 5098 }, { "epoch": 2.3995294117647057, "grad_norm": 0.6402961139978278, "learning_rate": 0.00024257550272261714, "loss": 0.248, "step": 5099 }, { "epoch": 2.4, "grad_norm": 0.5764831834156797, "learning_rate": 0.00024254100715890846, "loss": 0.3535, "step": 5100 }, { "epoch": 2.400470588235294, "grad_norm": 0.6172747718613393, "learning_rate": 0.00024250650369174386, "loss": 0.3984, "step": 5101 }, { "epoch": 2.4009411764705884, "grad_norm": 0.3166840542374814, "learning_rate": 0.0002424719923240701, "loss": 0.1934, "step": 5102 }, { "epoch": 2.401411764705882, "grad_norm": 1.0518906643213544, "learning_rate": 0.00024243747305883461, "loss": 0.4004, "step": 5103 }, { "epoch": 2.4018823529411764, "grad_norm": 0.7337255343805572, "learning_rate": 0.00024240294589898554, "loss": 0.2754, "step": 5104 }, { "epoch": 2.4023529411764706, "grad_norm": 0.9455058227100106, "learning_rate": 0.00024236841084747165, "loss": 0.4922, "step": 5105 }, { "epoch": 2.402823529411765, "grad_norm": 0.4527361783877507, "learning_rate": 0.00024233386790724242, "loss": 0.25, "step": 5106 }, { "epoch": 2.403294117647059, "grad_norm": 0.5000544408822555, "learning_rate": 0.000242299317081248, "loss": 0.3535, "step": 5107 }, { "epoch": 2.403764705882353, "grad_norm": 0.5523041378595738, "learning_rate": 0.00024226475837243908, "loss": 0.377, "step": 5108 }, { "epoch": 2.404235294117647, "grad_norm": 0.4071028677962844, "learning_rate": 0.00024223019178376732, "loss": 0.1079, "step": 5109 }, { "epoch": 2.4047058823529412, "grad_norm": 0.31911349738720873, "learning_rate": 0.00024219561731818475, "loss": 0.1226, "step": 5110 }, { "epoch": 2.4051764705882355, "grad_norm": 0.6655566967001002, "learning_rate": 0.00024216103497864426, "loss": 0.373, "step": 5111 }, { "epoch": 2.4056470588235292, "grad_norm": 0.1890692906165805, "learning_rate": 0.00024212644476809934, "loss": 0.1279, "step": 5112 }, { "epoch": 2.4061176470588235, "grad_norm": 0.5387867781232091, "learning_rate": 0.0002420918466895042, "loss": 0.2949, "step": 5113 }, { "epoch": 2.4065882352941177, "grad_norm": 0.6303272586262267, "learning_rate": 0.0002420572407458136, "loss": 0.3086, "step": 5114 }, { "epoch": 2.407058823529412, "grad_norm": 0.5701147084343036, "learning_rate": 0.0002420226269399831, "loss": 0.3242, "step": 5115 }, { "epoch": 2.4075294117647057, "grad_norm": 0.4122816901413045, "learning_rate": 0.00024198800527496889, "loss": 0.1914, "step": 5116 }, { "epoch": 2.408, "grad_norm": 0.41779043919172054, "learning_rate": 0.00024195337575372788, "loss": 0.2373, "step": 5117 }, { "epoch": 2.408470588235294, "grad_norm": 0.6173297644844439, "learning_rate": 0.00024191873837921748, "loss": 0.3223, "step": 5118 }, { "epoch": 2.4089411764705884, "grad_norm": 0.5534930876540773, "learning_rate": 0.000241884093154396, "loss": 0.291, "step": 5119 }, { "epoch": 2.409411764705882, "grad_norm": 0.42437016423516205, "learning_rate": 0.00024184944008222227, "loss": 0.2002, "step": 5120 }, { "epoch": 2.4098823529411764, "grad_norm": 0.38682158057082316, "learning_rate": 0.00024181477916565582, "loss": 0.291, "step": 5121 }, { "epoch": 2.4103529411764706, "grad_norm": 0.3176827979991266, "learning_rate": 0.0002417801104076569, "loss": 0.085, "step": 5122 }, { "epoch": 2.410823529411765, "grad_norm": 0.7587744886790557, "learning_rate": 0.00024174543381118638, "loss": 0.3516, "step": 5123 }, { "epoch": 2.411294117647059, "grad_norm": 0.4702573243893701, "learning_rate": 0.00024171074937920579, "loss": 0.2295, "step": 5124 }, { "epoch": 2.411764705882353, "grad_norm": 0.5140736472547476, "learning_rate": 0.00024167605711467738, "loss": 0.3086, "step": 5125 }, { "epoch": 2.412235294117647, "grad_norm": 0.47815210792178087, "learning_rate": 0.000241641357020564, "loss": 0.1953, "step": 5126 }, { "epoch": 2.4127058823529413, "grad_norm": 0.5722224275049738, "learning_rate": 0.00024160664909982928, "loss": 0.293, "step": 5127 }, { "epoch": 2.4131764705882355, "grad_norm": 1.157831428995274, "learning_rate": 0.00024157193335543734, "loss": 0.5625, "step": 5128 }, { "epoch": 2.4136470588235293, "grad_norm": 0.5098379499269692, "learning_rate": 0.00024153720979035317, "loss": 0.3945, "step": 5129 }, { "epoch": 2.4141176470588235, "grad_norm": 0.47346661963275466, "learning_rate": 0.00024150247840754228, "loss": 0.1191, "step": 5130 }, { "epoch": 2.4145882352941177, "grad_norm": 0.34561674662267616, "learning_rate": 0.00024146773920997091, "loss": 0.127, "step": 5131 }, { "epoch": 2.415058823529412, "grad_norm": 0.27403408971425586, "learning_rate": 0.000241432992200606, "loss": 0.0898, "step": 5132 }, { "epoch": 2.4155294117647057, "grad_norm": 0.7005539496098562, "learning_rate": 0.000241398237382415, "loss": 0.2715, "step": 5133 }, { "epoch": 2.416, "grad_norm": 0.46194763481339357, "learning_rate": 0.00024136347475836633, "loss": 0.1816, "step": 5134 }, { "epoch": 2.416470588235294, "grad_norm": 0.3477691132164582, "learning_rate": 0.0002413287043314287, "loss": 0.1055, "step": 5135 }, { "epoch": 2.4169411764705884, "grad_norm": 0.46904865635498955, "learning_rate": 0.00024129392610457176, "loss": 0.1875, "step": 5136 }, { "epoch": 2.417411764705882, "grad_norm": 0.787930295668137, "learning_rate": 0.0002412591400807658, "loss": 0.3262, "step": 5137 }, { "epoch": 2.4178823529411764, "grad_norm": 0.4740921120955274, "learning_rate": 0.0002412243462629816, "loss": 0.2041, "step": 5138 }, { "epoch": 2.4183529411764706, "grad_norm": 0.4953446758370144, "learning_rate": 0.0002411895446541908, "loss": 0.2041, "step": 5139 }, { "epoch": 2.418823529411765, "grad_norm": 0.8277345669935008, "learning_rate": 0.00024115473525736561, "loss": 0.3691, "step": 5140 }, { "epoch": 2.419294117647059, "grad_norm": 1.0258862738094312, "learning_rate": 0.0002411199180754789, "loss": 0.6211, "step": 5141 }, { "epoch": 2.419764705882353, "grad_norm": 0.6199952873761098, "learning_rate": 0.00024108509311150422, "loss": 0.2812, "step": 5142 }, { "epoch": 2.420235294117647, "grad_norm": 1.2850458859162461, "learning_rate": 0.00024105026036841582, "loss": 0.6172, "step": 5143 }, { "epoch": 2.4207058823529413, "grad_norm": 1.0026229719908923, "learning_rate": 0.00024101541984918864, "loss": 0.582, "step": 5144 }, { "epoch": 2.4211764705882355, "grad_norm": 0.6545415214517007, "learning_rate": 0.00024098057155679813, "loss": 0.2969, "step": 5145 }, { "epoch": 2.4216470588235293, "grad_norm": 0.42123160468200166, "learning_rate": 0.0002409457154942206, "loss": 0.2217, "step": 5146 }, { "epoch": 2.4221176470588235, "grad_norm": 0.3313306267622709, "learning_rate": 0.00024091085166443285, "loss": 0.126, "step": 5147 }, { "epoch": 2.4225882352941177, "grad_norm": 0.31543023691254046, "learning_rate": 0.00024087598007041243, "loss": 0.1416, "step": 5148 }, { "epoch": 2.423058823529412, "grad_norm": 0.8750654281756439, "learning_rate": 0.0002408411007151376, "loss": 0.3125, "step": 5149 }, { "epoch": 2.4235294117647057, "grad_norm": 0.6955897341594057, "learning_rate": 0.00024080621360158717, "loss": 0.3145, "step": 5150 }, { "epoch": 2.424, "grad_norm": 0.37134740939545186, "learning_rate": 0.00024077131873274072, "loss": 0.123, "step": 5151 }, { "epoch": 2.424470588235294, "grad_norm": 0.627177041510824, "learning_rate": 0.0002407364161115784, "loss": 0.2773, "step": 5152 }, { "epoch": 2.4249411764705884, "grad_norm": 0.5025360099691598, "learning_rate": 0.00024070150574108113, "loss": 0.3027, "step": 5153 }, { "epoch": 2.425411764705882, "grad_norm": 0.672114178327289, "learning_rate": 0.00024066658762423034, "loss": 0.3906, "step": 5154 }, { "epoch": 2.4258823529411764, "grad_norm": 0.571790634405378, "learning_rate": 0.00024063166176400825, "loss": 0.2178, "step": 5155 }, { "epoch": 2.4263529411764706, "grad_norm": 0.3956491115589719, "learning_rate": 0.0002405967281633977, "loss": 0.1094, "step": 5156 }, { "epoch": 2.426823529411765, "grad_norm": 0.5080551776321752, "learning_rate": 0.00024056178682538222, "loss": 0.3457, "step": 5157 }, { "epoch": 2.427294117647059, "grad_norm": 0.9422207305866093, "learning_rate": 0.00024052683775294592, "loss": 0.457, "step": 5158 }, { "epoch": 2.427764705882353, "grad_norm": 0.571053974628086, "learning_rate": 0.00024049188094907363, "loss": 0.2852, "step": 5159 }, { "epoch": 2.428235294117647, "grad_norm": 0.33576807319622193, "learning_rate": 0.00024045691641675086, "loss": 0.1328, "step": 5160 }, { "epoch": 2.4287058823529413, "grad_norm": 0.724589129012906, "learning_rate": 0.00024042194415896376, "loss": 0.4316, "step": 5161 }, { "epoch": 2.429176470588235, "grad_norm": 0.669386226014204, "learning_rate": 0.00024038696417869908, "loss": 0.5039, "step": 5162 }, { "epoch": 2.4296470588235293, "grad_norm": 0.6751809208608537, "learning_rate": 0.00024035197647894434, "loss": 0.4297, "step": 5163 }, { "epoch": 2.4301176470588235, "grad_norm": 0.31430989499701556, "learning_rate": 0.00024031698106268763, "loss": 0.1001, "step": 5164 }, { "epoch": 2.4305882352941177, "grad_norm": 0.49550774434921385, "learning_rate": 0.0002402819779329177, "loss": 0.3594, "step": 5165 }, { "epoch": 2.431058823529412, "grad_norm": 0.4749231224519177, "learning_rate": 0.00024024696709262406, "loss": 0.1895, "step": 5166 }, { "epoch": 2.4315294117647057, "grad_norm": 0.48059025408286116, "learning_rate": 0.00024021194854479674, "loss": 0.2461, "step": 5167 }, { "epoch": 2.432, "grad_norm": 1.033198429326603, "learning_rate": 0.00024017692229242655, "loss": 0.5938, "step": 5168 }, { "epoch": 2.432470588235294, "grad_norm": 1.0710057452075068, "learning_rate": 0.00024014188833850486, "loss": 0.6406, "step": 5169 }, { "epoch": 2.4329411764705884, "grad_norm": 0.6043779370690753, "learning_rate": 0.0002401068466860238, "loss": 0.2812, "step": 5170 }, { "epoch": 2.433411764705882, "grad_norm": 0.7509867071988019, "learning_rate": 0.00024007179733797604, "loss": 0.3066, "step": 5171 }, { "epoch": 2.4338823529411764, "grad_norm": 1.0229758212074451, "learning_rate": 0.00024003674029735497, "loss": 0.3379, "step": 5172 }, { "epoch": 2.4343529411764706, "grad_norm": 1.0814448479048155, "learning_rate": 0.00024000167556715468, "loss": 0.4512, "step": 5173 }, { "epoch": 2.434823529411765, "grad_norm": 0.5167239627764915, "learning_rate": 0.00023996660315036986, "loss": 0.2412, "step": 5174 }, { "epoch": 2.435294117647059, "grad_norm": 0.3910810575757353, "learning_rate": 0.00023993152304999582, "loss": 0.1572, "step": 5175 }, { "epoch": 2.435764705882353, "grad_norm": 1.1198684383891342, "learning_rate": 0.0002398964352690286, "loss": 0.416, "step": 5176 }, { "epoch": 2.436235294117647, "grad_norm": 1.6543971965749753, "learning_rate": 0.00023986133981046485, "loss": 0.6602, "step": 5177 }, { "epoch": 2.4367058823529413, "grad_norm": 0.6315596884582475, "learning_rate": 0.00023982623667730195, "loss": 0.3496, "step": 5178 }, { "epoch": 2.437176470588235, "grad_norm": 0.31908877498242266, "learning_rate": 0.0002397911258725378, "loss": 0.0947, "step": 5179 }, { "epoch": 2.4376470588235293, "grad_norm": 0.5264896175792634, "learning_rate": 0.00023975600739917113, "loss": 0.3867, "step": 5180 }, { "epoch": 2.4381176470588235, "grad_norm": 1.054030662212989, "learning_rate": 0.00023972088126020114, "loss": 0.5117, "step": 5181 }, { "epoch": 2.4385882352941177, "grad_norm": 0.6575891857694875, "learning_rate": 0.00023968574745862782, "loss": 0.2891, "step": 5182 }, { "epoch": 2.439058823529412, "grad_norm": 0.9506495962153193, "learning_rate": 0.0002396506059974518, "loss": 0.5312, "step": 5183 }, { "epoch": 2.4395294117647057, "grad_norm": 0.1322030602256953, "learning_rate": 0.00023961545687967422, "loss": 0.0242, "step": 5184 }, { "epoch": 2.44, "grad_norm": 0.5866553142623478, "learning_rate": 0.0002395803001082971, "loss": 0.1953, "step": 5185 }, { "epoch": 2.440470588235294, "grad_norm": 0.4835058775245684, "learning_rate": 0.00023954513568632296, "loss": 0.2236, "step": 5186 }, { "epoch": 2.4409411764705884, "grad_norm": 0.6885723505109197, "learning_rate": 0.000239509963616755, "loss": 0.2969, "step": 5187 }, { "epoch": 2.441411764705882, "grad_norm": 0.4458673509932739, "learning_rate": 0.00023947478390259712, "loss": 0.25, "step": 5188 }, { "epoch": 2.4418823529411764, "grad_norm": 0.45610289556609657, "learning_rate": 0.00023943959654685383, "loss": 0.2715, "step": 5189 }, { "epoch": 2.4423529411764706, "grad_norm": 0.14889978955947342, "learning_rate": 0.00023940440155253024, "loss": 0.025, "step": 5190 }, { "epoch": 2.442823529411765, "grad_norm": 0.20218157479308702, "learning_rate": 0.0002393691989226323, "loss": 0.0781, "step": 5191 }, { "epoch": 2.443294117647059, "grad_norm": 0.6741453396120217, "learning_rate": 0.00023933398866016642, "loss": 0.6953, "step": 5192 }, { "epoch": 2.443764705882353, "grad_norm": 0.821267750423313, "learning_rate": 0.0002392987707681397, "loss": 0.6328, "step": 5193 }, { "epoch": 2.444235294117647, "grad_norm": 0.6526554260602797, "learning_rate": 0.00023926354524955994, "loss": 0.3594, "step": 5194 }, { "epoch": 2.4447058823529413, "grad_norm": 0.5933399721577161, "learning_rate": 0.0002392283121074356, "loss": 0.418, "step": 5195 }, { "epoch": 2.445176470588235, "grad_norm": 0.31846925913005697, "learning_rate": 0.00023919307134477572, "loss": 0.1689, "step": 5196 }, { "epoch": 2.4456470588235293, "grad_norm": 0.4002161514390015, "learning_rate": 0.00023915782296459007, "loss": 0.0933, "step": 5197 }, { "epoch": 2.4461176470588235, "grad_norm": 0.8236107485674216, "learning_rate": 0.00023912256696988904, "loss": 0.3672, "step": 5198 }, { "epoch": 2.4465882352941177, "grad_norm": 0.5595764734046778, "learning_rate": 0.00023908730336368364, "loss": 0.3867, "step": 5199 }, { "epoch": 2.447058823529412, "grad_norm": 0.5695026959657875, "learning_rate": 0.00023905203214898558, "loss": 0.2656, "step": 5200 }, { "epoch": 2.4475294117647057, "grad_norm": 0.3085644323054856, "learning_rate": 0.00023901675332880715, "loss": 0.1719, "step": 5201 }, { "epoch": 2.448, "grad_norm": 0.41602064793980603, "learning_rate": 0.00023898146690616134, "loss": 0.2412, "step": 5202 }, { "epoch": 2.448470588235294, "grad_norm": 0.6684957605232907, "learning_rate": 0.0002389461728840618, "loss": 0.4043, "step": 5203 }, { "epoch": 2.4489411764705884, "grad_norm": 0.5422446889700338, "learning_rate": 0.00023891087126552292, "loss": 0.2539, "step": 5204 }, { "epoch": 2.449411764705882, "grad_norm": 0.32060321078370113, "learning_rate": 0.00023887556205355945, "loss": 0.0908, "step": 5205 }, { "epoch": 2.4498823529411764, "grad_norm": 0.477071781002468, "learning_rate": 0.00023884024525118702, "loss": 0.3457, "step": 5206 }, { "epoch": 2.4503529411764706, "grad_norm": 0.5913933251842917, "learning_rate": 0.00023880492086142193, "loss": 0.3672, "step": 5207 }, { "epoch": 2.450823529411765, "grad_norm": 0.4012967622577294, "learning_rate": 0.00023876958888728098, "loss": 0.21, "step": 5208 }, { "epoch": 2.451294117647059, "grad_norm": 0.39234700501855535, "learning_rate": 0.00023873424933178178, "loss": 0.166, "step": 5209 }, { "epoch": 2.451764705882353, "grad_norm": 0.4467298174388728, "learning_rate": 0.00023869890219794238, "loss": 0.1514, "step": 5210 }, { "epoch": 2.452235294117647, "grad_norm": 0.32287821086185037, "learning_rate": 0.0002386635474887817, "loss": 0.1406, "step": 5211 }, { "epoch": 2.4527058823529413, "grad_norm": 0.998545812393593, "learning_rate": 0.0002386281852073192, "loss": 0.6406, "step": 5212 }, { "epoch": 2.453176470588235, "grad_norm": 0.2743305145912476, "learning_rate": 0.00023859281535657496, "loss": 0.0452, "step": 5213 }, { "epoch": 2.4536470588235293, "grad_norm": 0.22305119714223623, "learning_rate": 0.00023855743793956967, "loss": 0.0747, "step": 5214 }, { "epoch": 2.4541176470588235, "grad_norm": 0.6136287447740382, "learning_rate": 0.0002385220529593249, "loss": 0.3926, "step": 5215 }, { "epoch": 2.4545882352941177, "grad_norm": 0.7791278030143096, "learning_rate": 0.00023848666041886258, "loss": 0.332, "step": 5216 }, { "epoch": 2.455058823529412, "grad_norm": 0.3804474821089739, "learning_rate": 0.00023845126032120539, "loss": 0.2002, "step": 5217 }, { "epoch": 2.4555294117647057, "grad_norm": 0.26943989398479085, "learning_rate": 0.00023841585266937677, "loss": 0.0845, "step": 5218 }, { "epoch": 2.456, "grad_norm": 1.106351964160502, "learning_rate": 0.00023838043746640065, "loss": 0.5, "step": 5219 }, { "epoch": 2.456470588235294, "grad_norm": 0.8909002130712631, "learning_rate": 0.00023834501471530165, "loss": 0.3555, "step": 5220 }, { "epoch": 2.4569411764705884, "grad_norm": 0.8886106470615616, "learning_rate": 0.0002383095844191051, "loss": 0.5469, "step": 5221 }, { "epoch": 2.457411764705882, "grad_norm": 0.4674864144533376, "learning_rate": 0.0002382741465808368, "loss": 0.2969, "step": 5222 }, { "epoch": 2.4578823529411764, "grad_norm": 0.4615172387377678, "learning_rate": 0.0002382387012035235, "loss": 0.2461, "step": 5223 }, { "epoch": 2.4583529411764706, "grad_norm": 0.8600109038179683, "learning_rate": 0.0002382032482901922, "loss": 0.3457, "step": 5224 }, { "epoch": 2.458823529411765, "grad_norm": 0.24544229024425301, "learning_rate": 0.00023816778784387094, "loss": 0.0532, "step": 5225 }, { "epoch": 2.4592941176470586, "grad_norm": 0.8378820944570707, "learning_rate": 0.00023813231986758815, "loss": 0.4004, "step": 5226 }, { "epoch": 2.459764705882353, "grad_norm": 0.27497001296538226, "learning_rate": 0.00023809684436437294, "loss": 0.1001, "step": 5227 }, { "epoch": 2.460235294117647, "grad_norm": 0.90279182280152, "learning_rate": 0.00023806136133725506, "loss": 0.5078, "step": 5228 }, { "epoch": 2.4607058823529413, "grad_norm": 0.4657759466551752, "learning_rate": 0.00023802587078926505, "loss": 0.2412, "step": 5229 }, { "epoch": 2.461176470588235, "grad_norm": 0.8792806047988263, "learning_rate": 0.00023799037272343387, "loss": 0.543, "step": 5230 }, { "epoch": 2.4616470588235293, "grad_norm": 0.4457449253021764, "learning_rate": 0.00023795486714279327, "loss": 0.1621, "step": 5231 }, { "epoch": 2.4621176470588235, "grad_norm": 0.48517682560123815, "learning_rate": 0.0002379193540503756, "loss": 0.1377, "step": 5232 }, { "epoch": 2.4625882352941177, "grad_norm": 0.9646301378883598, "learning_rate": 0.00023788383344921384, "loss": 0.5234, "step": 5233 }, { "epoch": 2.463058823529412, "grad_norm": 0.6578541923842577, "learning_rate": 0.0002378483053423417, "loss": 0.4121, "step": 5234 }, { "epoch": 2.4635294117647057, "grad_norm": 0.5450784642597967, "learning_rate": 0.00023781276973279335, "loss": 0.4531, "step": 5235 }, { "epoch": 2.464, "grad_norm": 0.5583612238097911, "learning_rate": 0.0002377772266236037, "loss": 0.3359, "step": 5236 }, { "epoch": 2.464470588235294, "grad_norm": 0.22915253472475147, "learning_rate": 0.0002377416760178084, "loss": 0.0713, "step": 5237 }, { "epoch": 2.4649411764705884, "grad_norm": 0.6701883772773813, "learning_rate": 0.00023770611791844364, "loss": 0.2949, "step": 5238 }, { "epoch": 2.465411764705882, "grad_norm": 0.557107819150734, "learning_rate": 0.00023767055232854615, "loss": 0.2578, "step": 5239 }, { "epoch": 2.4658823529411764, "grad_norm": 1.1123009788435871, "learning_rate": 0.00023763497925115347, "loss": 0.7539, "step": 5240 }, { "epoch": 2.4663529411764706, "grad_norm": 0.6738398468348324, "learning_rate": 0.00023759939868930373, "loss": 0.3457, "step": 5241 }, { "epoch": 2.466823529411765, "grad_norm": 0.9322664202069794, "learning_rate": 0.00023756381064603564, "loss": 0.6992, "step": 5242 }, { "epoch": 2.4672941176470586, "grad_norm": 0.5300429549245725, "learning_rate": 0.0002375282151243887, "loss": 0.3418, "step": 5243 }, { "epoch": 2.467764705882353, "grad_norm": 0.6696414479233899, "learning_rate": 0.00023749261212740278, "loss": 0.416, "step": 5244 }, { "epoch": 2.468235294117647, "grad_norm": 0.4201091403011071, "learning_rate": 0.00023745700165811868, "loss": 0.1465, "step": 5245 }, { "epoch": 2.4687058823529413, "grad_norm": 0.8355767753018798, "learning_rate": 0.00023742138371957765, "loss": 0.3359, "step": 5246 }, { "epoch": 2.469176470588235, "grad_norm": 1.138083065426251, "learning_rate": 0.00023738575831482164, "loss": 0.6562, "step": 5247 }, { "epoch": 2.4696470588235293, "grad_norm": 0.7293463599225983, "learning_rate": 0.0002373501254468933, "loss": 0.4473, "step": 5248 }, { "epoch": 2.4701176470588235, "grad_norm": 0.3043216155468721, "learning_rate": 0.00023731448511883575, "loss": 0.1387, "step": 5249 }, { "epoch": 2.4705882352941178, "grad_norm": 0.3371748579623761, "learning_rate": 0.00023727883733369292, "loss": 0.1221, "step": 5250 }, { "epoch": 2.471058823529412, "grad_norm": 0.5535709029851912, "learning_rate": 0.00023724318209450925, "loss": 0.3008, "step": 5251 }, { "epoch": 2.4715294117647058, "grad_norm": 0.29074979988778765, "learning_rate": 0.00023720751940432992, "loss": 0.1108, "step": 5252 }, { "epoch": 2.472, "grad_norm": 1.0468950591922788, "learning_rate": 0.0002371718492662007, "loss": 0.3594, "step": 5253 }, { "epoch": 2.472470588235294, "grad_norm": 0.6265305296334751, "learning_rate": 0.00023713617168316797, "loss": 0.2676, "step": 5254 }, { "epoch": 2.4729411764705884, "grad_norm": 1.1809059328984481, "learning_rate": 0.00023710048665827872, "loss": 0.5898, "step": 5255 }, { "epoch": 2.473411764705882, "grad_norm": 0.3574819373448321, "learning_rate": 0.00023706479419458075, "loss": 0.2236, "step": 5256 }, { "epoch": 2.4738823529411764, "grad_norm": 0.27766843386515105, "learning_rate": 0.00023702909429512228, "loss": 0.0903, "step": 5257 }, { "epoch": 2.4743529411764706, "grad_norm": 0.7362892848094721, "learning_rate": 0.00023699338696295228, "loss": 0.5312, "step": 5258 }, { "epoch": 2.474823529411765, "grad_norm": 0.3861998449956519, "learning_rate": 0.00023695767220112033, "loss": 0.1543, "step": 5259 }, { "epoch": 2.4752941176470586, "grad_norm": 0.6295373226888556, "learning_rate": 0.0002369219500126767, "loss": 0.2393, "step": 5260 }, { "epoch": 2.475764705882353, "grad_norm": 0.9028830932744525, "learning_rate": 0.00023688622040067217, "loss": 0.3984, "step": 5261 }, { "epoch": 2.476235294117647, "grad_norm": 0.693339960419591, "learning_rate": 0.0002368504833681582, "loss": 0.3066, "step": 5262 }, { "epoch": 2.4767058823529413, "grad_norm": 0.8312424611147522, "learning_rate": 0.00023681473891818699, "loss": 0.4102, "step": 5263 }, { "epoch": 2.477176470588235, "grad_norm": 0.9678926011167207, "learning_rate": 0.00023677898705381123, "loss": 0.4961, "step": 5264 }, { "epoch": 2.4776470588235293, "grad_norm": 0.5225347468168254, "learning_rate": 0.00023674322777808432, "loss": 0.1885, "step": 5265 }, { "epoch": 2.4781176470588235, "grad_norm": 1.0629927385545512, "learning_rate": 0.0002367074610940603, "loss": 0.3867, "step": 5266 }, { "epoch": 2.4785882352941178, "grad_norm": 0.5176131821697195, "learning_rate": 0.00023667168700479383, "loss": 0.291, "step": 5267 }, { "epoch": 2.479058823529412, "grad_norm": 0.570220645787642, "learning_rate": 0.0002366359055133401, "loss": 0.332, "step": 5268 }, { "epoch": 2.4795294117647058, "grad_norm": 0.5064097626905196, "learning_rate": 0.00023660011662275516, "loss": 0.1377, "step": 5269 }, { "epoch": 2.48, "grad_norm": 0.26545512412674466, "learning_rate": 0.00023656432033609544, "loss": 0.0269, "step": 5270 }, { "epoch": 2.480470588235294, "grad_norm": 0.861890496372941, "learning_rate": 0.0002365285166564182, "loss": 0.3516, "step": 5271 }, { "epoch": 2.4809411764705884, "grad_norm": 1.0706253142768145, "learning_rate": 0.00023649270558678123, "loss": 0.6172, "step": 5272 }, { "epoch": 2.481411764705882, "grad_norm": 0.338228423292168, "learning_rate": 0.00023645688713024295, "loss": 0.1787, "step": 5273 }, { "epoch": 2.4818823529411764, "grad_norm": 0.6250723178877179, "learning_rate": 0.0002364210612898624, "loss": 0.7188, "step": 5274 }, { "epoch": 2.4823529411764707, "grad_norm": 0.6367209729549962, "learning_rate": 0.00023638522806869935, "loss": 0.3301, "step": 5275 }, { "epoch": 2.482823529411765, "grad_norm": 1.0079298552481084, "learning_rate": 0.0002363493874698141, "loss": 0.5078, "step": 5276 }, { "epoch": 2.4832941176470587, "grad_norm": 1.0197577076711766, "learning_rate": 0.0002363135394962676, "loss": 0.6484, "step": 5277 }, { "epoch": 2.483764705882353, "grad_norm": 0.5805550839674827, "learning_rate": 0.00023627768415112154, "loss": 0.4297, "step": 5278 }, { "epoch": 2.484235294117647, "grad_norm": 0.5600114651835364, "learning_rate": 0.00023624182143743797, "loss": 0.3164, "step": 5279 }, { "epoch": 2.4847058823529413, "grad_norm": 0.7365552116180569, "learning_rate": 0.00023620595135827985, "loss": 0.3535, "step": 5280 }, { "epoch": 2.485176470588235, "grad_norm": 1.2044523714028872, "learning_rate": 0.0002361700739167107, "loss": 0.5312, "step": 5281 }, { "epoch": 2.4856470588235293, "grad_norm": 0.7509884044542737, "learning_rate": 0.00023613418911579454, "loss": 0.3379, "step": 5282 }, { "epoch": 2.4861176470588235, "grad_norm": 0.3958062019538459, "learning_rate": 0.00023609829695859618, "loss": 0.1787, "step": 5283 }, { "epoch": 2.4865882352941178, "grad_norm": 0.68998758863302, "learning_rate": 0.00023606239744818094, "loss": 0.3008, "step": 5284 }, { "epoch": 2.487058823529412, "grad_norm": 0.5271086484695544, "learning_rate": 0.0002360264905876148, "loss": 0.2119, "step": 5285 }, { "epoch": 2.4875294117647058, "grad_norm": 0.08250735997851609, "learning_rate": 0.0002359905763799644, "loss": 0.0557, "step": 5286 }, { "epoch": 2.488, "grad_norm": 0.4106028918417294, "learning_rate": 0.00023595465482829697, "loss": 0.1494, "step": 5287 }, { "epoch": 2.488470588235294, "grad_norm": 0.709901284139537, "learning_rate": 0.0002359187259356805, "loss": 0.293, "step": 5288 }, { "epoch": 2.488941176470588, "grad_norm": 0.5808341038754148, "learning_rate": 0.00023588278970518333, "loss": 0.2373, "step": 5289 }, { "epoch": 2.489411764705882, "grad_norm": 0.7169399809465391, "learning_rate": 0.00023584684613987472, "loss": 0.4102, "step": 5290 }, { "epoch": 2.4898823529411764, "grad_norm": 0.44163443284011156, "learning_rate": 0.00023581089524282432, "loss": 0.1748, "step": 5291 }, { "epoch": 2.4903529411764707, "grad_norm": 0.36181183638537634, "learning_rate": 0.0002357749370171026, "loss": 0.1455, "step": 5292 }, { "epoch": 2.490823529411765, "grad_norm": 0.36821992129502135, "learning_rate": 0.00023573897146578053, "loss": 0.2578, "step": 5293 }, { "epoch": 2.4912941176470587, "grad_norm": 0.3316492108698467, "learning_rate": 0.00023570299859192976, "loss": 0.0742, "step": 5294 }, { "epoch": 2.491764705882353, "grad_norm": 0.1822664890232378, "learning_rate": 0.00023566701839862253, "loss": 0.0596, "step": 5295 }, { "epoch": 2.492235294117647, "grad_norm": 1.3962784293423016, "learning_rate": 0.00023563103088893172, "loss": 0.668, "step": 5296 }, { "epoch": 2.4927058823529413, "grad_norm": 0.8774595433816683, "learning_rate": 0.00023559503606593088, "loss": 0.5469, "step": 5297 }, { "epoch": 2.493176470588235, "grad_norm": 0.1416959299003023, "learning_rate": 0.0002355590339326941, "loss": 0.0464, "step": 5298 }, { "epoch": 2.4936470588235293, "grad_norm": 0.6806978911699028, "learning_rate": 0.00023552302449229612, "loss": 0.4199, "step": 5299 }, { "epoch": 2.4941176470588236, "grad_norm": 0.42525694726248936, "learning_rate": 0.00023548700774781242, "loss": 0.1147, "step": 5300 }, { "epoch": 2.4945882352941178, "grad_norm": 0.9164167032554784, "learning_rate": 0.00023545098370231893, "loss": 0.543, "step": 5301 }, { "epoch": 2.495058823529412, "grad_norm": 0.644323021458261, "learning_rate": 0.00023541495235889227, "loss": 0.2656, "step": 5302 }, { "epoch": 2.4955294117647058, "grad_norm": 0.2888128887484852, "learning_rate": 0.0002353789137206098, "loss": 0.063, "step": 5303 }, { "epoch": 2.496, "grad_norm": 0.7437768802115137, "learning_rate": 0.00023534286779054923, "loss": 0.5391, "step": 5304 }, { "epoch": 2.496470588235294, "grad_norm": 0.47474854709720915, "learning_rate": 0.00023530681457178916, "loss": 0.252, "step": 5305 }, { "epoch": 2.496941176470588, "grad_norm": 0.5841480887574718, "learning_rate": 0.00023527075406740876, "loss": 0.3848, "step": 5306 }, { "epoch": 2.497411764705882, "grad_norm": 0.3748713772635085, "learning_rate": 0.00023523468628048764, "loss": 0.1787, "step": 5307 }, { "epoch": 2.4978823529411764, "grad_norm": 0.38694777054014495, "learning_rate": 0.0002351986112141063, "loss": 0.25, "step": 5308 }, { "epoch": 2.4983529411764707, "grad_norm": 0.5898256145616778, "learning_rate": 0.00023516252887134558, "loss": 0.3184, "step": 5309 }, { "epoch": 2.498823529411765, "grad_norm": 0.18168574173097485, "learning_rate": 0.00023512643925528725, "loss": 0.0439, "step": 5310 }, { "epoch": 2.4992941176470587, "grad_norm": 0.3191284277274752, "learning_rate": 0.00023509034236901348, "loss": 0.1436, "step": 5311 }, { "epoch": 2.499764705882353, "grad_norm": 0.9839878772379211, "learning_rate": 0.00023505423821560705, "loss": 0.4297, "step": 5312 }, { "epoch": 2.500235294117647, "grad_norm": 0.702004982235106, "learning_rate": 0.00023501812679815152, "loss": 0.3262, "step": 5313 }, { "epoch": 2.5007058823529413, "grad_norm": 0.7066114551517132, "learning_rate": 0.00023498200811973092, "loss": 0.375, "step": 5314 }, { "epoch": 2.501176470588235, "grad_norm": 0.3720636166105665, "learning_rate": 0.00023494588218343003, "loss": 0.1719, "step": 5315 }, { "epoch": 2.5016470588235293, "grad_norm": 0.29917273108786474, "learning_rate": 0.0002349097489923341, "loss": 0.1011, "step": 5316 }, { "epoch": 2.5021176470588236, "grad_norm": 0.29444425246808836, "learning_rate": 0.00023487360854952915, "loss": 0.0645, "step": 5317 }, { "epoch": 2.502588235294118, "grad_norm": 0.40121252192026746, "learning_rate": 0.00023483746085810176, "loss": 0.1025, "step": 5318 }, { "epoch": 2.503058823529412, "grad_norm": 0.6778865216438478, "learning_rate": 0.00023480130592113904, "loss": 0.3691, "step": 5319 }, { "epoch": 2.503529411764706, "grad_norm": 0.7664569111256563, "learning_rate": 0.00023476514374172887, "loss": 0.416, "step": 5320 }, { "epoch": 2.504, "grad_norm": 0.40767868024554615, "learning_rate": 0.0002347289743229596, "loss": 0.1338, "step": 5321 }, { "epoch": 2.5044705882352942, "grad_norm": 0.6243017991529776, "learning_rate": 0.00023469279766792037, "loss": 0.3281, "step": 5322 }, { "epoch": 2.504941176470588, "grad_norm": 0.798945161187189, "learning_rate": 0.0002346566137797008, "loss": 0.4082, "step": 5323 }, { "epoch": 2.5054117647058822, "grad_norm": 0.31320650838973185, "learning_rate": 0.0002346204226613912, "loss": 0.1152, "step": 5324 }, { "epoch": 2.5058823529411764, "grad_norm": 0.6972800645417779, "learning_rate": 0.0002345842243160824, "loss": 0.2637, "step": 5325 }, { "epoch": 2.5063529411764707, "grad_norm": 0.4525842703124199, "learning_rate": 0.00023454801874686596, "loss": 0.2451, "step": 5326 }, { "epoch": 2.506823529411765, "grad_norm": 0.8937513452251545, "learning_rate": 0.00023451180595683409, "loss": 0.3809, "step": 5327 }, { "epoch": 2.5072941176470587, "grad_norm": 0.4700460789875531, "learning_rate": 0.0002344755859490794, "loss": 0.2246, "step": 5328 }, { "epoch": 2.507764705882353, "grad_norm": 0.6145388899753846, "learning_rate": 0.00023443935872669534, "loss": 0.1875, "step": 5329 }, { "epoch": 2.508235294117647, "grad_norm": 0.6509536840671437, "learning_rate": 0.00023440312429277583, "loss": 0.3105, "step": 5330 }, { "epoch": 2.5087058823529413, "grad_norm": 0.40054674597328727, "learning_rate": 0.00023436688265041555, "loss": 0.2314, "step": 5331 }, { "epoch": 2.509176470588235, "grad_norm": 0.1304200429266923, "learning_rate": 0.0002343306338027097, "loss": 0.0294, "step": 5332 }, { "epoch": 2.5096470588235293, "grad_norm": 0.34567660566944497, "learning_rate": 0.00023429437775275406, "loss": 0.1226, "step": 5333 }, { "epoch": 2.5101176470588236, "grad_norm": 0.6648457237288172, "learning_rate": 0.00023425811450364508, "loss": 0.2715, "step": 5334 }, { "epoch": 2.510588235294118, "grad_norm": 0.48132367061645126, "learning_rate": 0.00023422184405847985, "loss": 0.167, "step": 5335 }, { "epoch": 2.511058823529412, "grad_norm": 0.23840843256836314, "learning_rate": 0.00023418556642035602, "loss": 0.1069, "step": 5336 }, { "epoch": 2.511529411764706, "grad_norm": 0.6349079438685585, "learning_rate": 0.0002341492815923719, "loss": 0.3496, "step": 5337 }, { "epoch": 2.512, "grad_norm": 0.47600934399966216, "learning_rate": 0.00023411298957762636, "loss": 0.1709, "step": 5338 }, { "epoch": 2.5124705882352942, "grad_norm": 0.6667215100550932, "learning_rate": 0.00023407669037921902, "loss": 0.5273, "step": 5339 }, { "epoch": 2.512941176470588, "grad_norm": 0.3036495298678612, "learning_rate": 0.00023404038400024984, "loss": 0.127, "step": 5340 }, { "epoch": 2.5134117647058822, "grad_norm": 1.0882710812936693, "learning_rate": 0.00023400407044381965, "loss": 0.3301, "step": 5341 }, { "epoch": 2.5138823529411765, "grad_norm": 0.42424218909308675, "learning_rate": 0.00023396774971302985, "loss": 0.165, "step": 5342 }, { "epoch": 2.5143529411764707, "grad_norm": 0.9663470363326188, "learning_rate": 0.00023393142181098235, "loss": 0.5977, "step": 5343 }, { "epoch": 2.514823529411765, "grad_norm": 0.2825570450588366, "learning_rate": 0.00023389508674077973, "loss": 0.0693, "step": 5344 }, { "epoch": 2.5152941176470587, "grad_norm": 0.9350611805644287, "learning_rate": 0.00023385874450552522, "loss": 0.5078, "step": 5345 }, { "epoch": 2.515764705882353, "grad_norm": 0.6384809389485296, "learning_rate": 0.00023382239510832262, "loss": 0.3008, "step": 5346 }, { "epoch": 2.516235294117647, "grad_norm": 0.39299878989613796, "learning_rate": 0.0002337860385522763, "loss": 0.1846, "step": 5347 }, { "epoch": 2.516705882352941, "grad_norm": 0.37383816802574793, "learning_rate": 0.00023374967484049132, "loss": 0.1089, "step": 5348 }, { "epoch": 2.517176470588235, "grad_norm": 0.2633985388647448, "learning_rate": 0.00023371330397607335, "loss": 0.0977, "step": 5349 }, { "epoch": 2.5176470588235293, "grad_norm": 0.748440865400015, "learning_rate": 0.00023367692596212858, "loss": 0.3711, "step": 5350 }, { "epoch": 2.5181176470588236, "grad_norm": 0.7953760445439904, "learning_rate": 0.0002336405408017639, "loss": 0.375, "step": 5351 }, { "epoch": 2.518588235294118, "grad_norm": 0.6860041230269978, "learning_rate": 0.0002336041484980868, "loss": 0.3789, "step": 5352 }, { "epoch": 2.519058823529412, "grad_norm": 0.6823322894854287, "learning_rate": 0.0002335677490542053, "loss": 0.3535, "step": 5353 }, { "epoch": 2.519529411764706, "grad_norm": 0.581954914703267, "learning_rate": 0.0002335313424732282, "loss": 0.3008, "step": 5354 }, { "epoch": 2.52, "grad_norm": 0.48058582835269104, "learning_rate": 0.0002334949287582647, "loss": 0.1011, "step": 5355 }, { "epoch": 2.5204705882352942, "grad_norm": 1.8285398832650164, "learning_rate": 0.00023345850791242477, "loss": 0.252, "step": 5356 }, { "epoch": 2.520941176470588, "grad_norm": 0.35310977818058004, "learning_rate": 0.00023342207993881883, "loss": 0.1475, "step": 5357 }, { "epoch": 2.5214117647058822, "grad_norm": 0.35668173940362047, "learning_rate": 0.0002333856448405581, "loss": 0.1885, "step": 5358 }, { "epoch": 2.5218823529411765, "grad_norm": 0.3055163493422738, "learning_rate": 0.00023334920262075437, "loss": 0.1089, "step": 5359 }, { "epoch": 2.5223529411764707, "grad_norm": 0.2311669179619939, "learning_rate": 0.00023331275328251983, "loss": 0.0723, "step": 5360 }, { "epoch": 2.522823529411765, "grad_norm": 0.7341372018209597, "learning_rate": 0.00023327629682896758, "loss": 0.4336, "step": 5361 }, { "epoch": 2.5232941176470587, "grad_norm": 0.8266235321917347, "learning_rate": 0.00023323983326321106, "loss": 0.3613, "step": 5362 }, { "epoch": 2.523764705882353, "grad_norm": 0.6514261026568585, "learning_rate": 0.0002332033625883645, "loss": 0.3359, "step": 5363 }, { "epoch": 2.524235294117647, "grad_norm": 1.1342905650918773, "learning_rate": 0.00023316688480754267, "loss": 0.4355, "step": 5364 }, { "epoch": 2.524705882352941, "grad_norm": 0.3641303611066924, "learning_rate": 0.00023313039992386094, "loss": 0.126, "step": 5365 }, { "epoch": 2.525176470588235, "grad_norm": 0.440703708415517, "learning_rate": 0.00023309390794043532, "loss": 0.1895, "step": 5366 }, { "epoch": 2.5256470588235294, "grad_norm": 0.5223648951222887, "learning_rate": 0.00023305740886038235, "loss": 0.2676, "step": 5367 }, { "epoch": 2.5261176470588236, "grad_norm": 0.28578564860467454, "learning_rate": 0.00023302090268681932, "loss": 0.0977, "step": 5368 }, { "epoch": 2.526588235294118, "grad_norm": 0.5996500526563102, "learning_rate": 0.00023298438942286395, "loss": 0.2656, "step": 5369 }, { "epoch": 2.527058823529412, "grad_norm": 0.8871507278309158, "learning_rate": 0.00023294786907163473, "loss": 0.3359, "step": 5370 }, { "epoch": 2.527529411764706, "grad_norm": 0.6157648041778638, "learning_rate": 0.0002329113416362506, "loss": 0.3047, "step": 5371 }, { "epoch": 2.528, "grad_norm": 0.5295443609054393, "learning_rate": 0.00023287480711983125, "loss": 0.4121, "step": 5372 }, { "epoch": 2.5284705882352942, "grad_norm": 0.3294687581677163, "learning_rate": 0.00023283826552549682, "loss": 0.2695, "step": 5373 }, { "epoch": 2.528941176470588, "grad_norm": 0.5245666266183101, "learning_rate": 0.00023280171685636822, "loss": 0.2178, "step": 5374 }, { "epoch": 2.5294117647058822, "grad_norm": 0.7131463655686878, "learning_rate": 0.0002327651611155669, "loss": 0.4453, "step": 5375 }, { "epoch": 2.5298823529411765, "grad_norm": 0.43468149099936265, "learning_rate": 0.0002327285983062148, "loss": 0.1475, "step": 5376 }, { "epoch": 2.5303529411764707, "grad_norm": 0.5680683911749898, "learning_rate": 0.00023269202843143462, "loss": 0.3574, "step": 5377 }, { "epoch": 2.530823529411765, "grad_norm": 0.3425965424562437, "learning_rate": 0.00023265545149434968, "loss": 0.1279, "step": 5378 }, { "epoch": 2.5312941176470587, "grad_norm": 0.6717798556102157, "learning_rate": 0.0002326188674980837, "loss": 0.3223, "step": 5379 }, { "epoch": 2.531764705882353, "grad_norm": 1.3769268753861206, "learning_rate": 0.0002325822764457612, "loss": 0.6289, "step": 5380 }, { "epoch": 2.532235294117647, "grad_norm": 0.5338779558906656, "learning_rate": 0.0002325456783405072, "loss": 0.582, "step": 5381 }, { "epoch": 2.532705882352941, "grad_norm": 0.88825859867802, "learning_rate": 0.00023250907318544744, "loss": 0.5547, "step": 5382 }, { "epoch": 2.533176470588235, "grad_norm": 0.6610345083481953, "learning_rate": 0.0002324724609837081, "loss": 0.3496, "step": 5383 }, { "epoch": 2.5336470588235294, "grad_norm": 0.579495255081084, "learning_rate": 0.0002324358417384161, "loss": 0.4512, "step": 5384 }, { "epoch": 2.5341176470588236, "grad_norm": 0.6940482024804978, "learning_rate": 0.00023239921545269884, "loss": 0.3242, "step": 5385 }, { "epoch": 2.534588235294118, "grad_norm": 0.5478731898724907, "learning_rate": 0.00023236258212968445, "loss": 0.3125, "step": 5386 }, { "epoch": 2.535058823529412, "grad_norm": 0.3700230426588584, "learning_rate": 0.00023232594177250148, "loss": 0.1738, "step": 5387 }, { "epoch": 2.535529411764706, "grad_norm": 0.586825359407587, "learning_rate": 0.00023228929438427932, "loss": 0.373, "step": 5388 }, { "epoch": 2.536, "grad_norm": 0.43445956463331553, "learning_rate": 0.0002322526399681478, "loss": 0.2354, "step": 5389 }, { "epoch": 2.5364705882352943, "grad_norm": 0.4101924910369805, "learning_rate": 0.0002322159785272374, "loss": 0.1621, "step": 5390 }, { "epoch": 2.536941176470588, "grad_norm": 0.8138907842446805, "learning_rate": 0.00023217931006467912, "loss": 0.375, "step": 5391 }, { "epoch": 2.5374117647058823, "grad_norm": 0.565972925013291, "learning_rate": 0.00023214263458360467, "loss": 0.3105, "step": 5392 }, { "epoch": 2.5378823529411765, "grad_norm": 0.7986943329798215, "learning_rate": 0.00023210595208714635, "loss": 0.5, "step": 5393 }, { "epoch": 2.5383529411764707, "grad_norm": 0.1087316515498479, "learning_rate": 0.00023206926257843702, "loss": 0.0189, "step": 5394 }, { "epoch": 2.538823529411765, "grad_norm": 0.5991598084811915, "learning_rate": 0.00023203256606061008, "loss": 0.2871, "step": 5395 }, { "epoch": 2.5392941176470587, "grad_norm": 0.5386790176765304, "learning_rate": 0.00023199586253679964, "loss": 0.2354, "step": 5396 }, { "epoch": 2.539764705882353, "grad_norm": 0.9113089708142086, "learning_rate": 0.00023195915201014035, "loss": 0.4824, "step": 5397 }, { "epoch": 2.540235294117647, "grad_norm": 0.3079963191749291, "learning_rate": 0.00023192243448376745, "loss": 0.1108, "step": 5398 }, { "epoch": 2.540705882352941, "grad_norm": 0.6219611358734066, "learning_rate": 0.00023188570996081687, "loss": 0.3516, "step": 5399 }, { "epoch": 2.541176470588235, "grad_norm": 0.7481384149054705, "learning_rate": 0.00023184897844442495, "loss": 0.3789, "step": 5400 }, { "epoch": 2.5416470588235294, "grad_norm": 0.6813446736210256, "learning_rate": 0.00023181223993772886, "loss": 0.4961, "step": 5401 }, { "epoch": 2.5421176470588236, "grad_norm": 0.7875022877777751, "learning_rate": 0.00023177549444386618, "loss": 0.4395, "step": 5402 }, { "epoch": 2.542588235294118, "grad_norm": 0.617257789831907, "learning_rate": 0.00023173874196597515, "loss": 0.3145, "step": 5403 }, { "epoch": 2.543058823529412, "grad_norm": 0.5921492056106311, "learning_rate": 0.00023170198250719464, "loss": 0.3555, "step": 5404 }, { "epoch": 2.543529411764706, "grad_norm": 0.9856430799645806, "learning_rate": 0.00023166521607066409, "loss": 0.6172, "step": 5405 }, { "epoch": 2.544, "grad_norm": 0.6139270675898811, "learning_rate": 0.00023162844265952348, "loss": 0.2441, "step": 5406 }, { "epoch": 2.5444705882352943, "grad_norm": 1.0900044028612252, "learning_rate": 0.0002315916622769135, "loss": 0.7422, "step": 5407 }, { "epoch": 2.544941176470588, "grad_norm": 0.44195038755256877, "learning_rate": 0.00023155487492597536, "loss": 0.2266, "step": 5408 }, { "epoch": 2.5454117647058823, "grad_norm": 0.5542446826581118, "learning_rate": 0.0002315180806098509, "loss": 0.3359, "step": 5409 }, { "epoch": 2.5458823529411765, "grad_norm": 0.5337231880377593, "learning_rate": 0.00023148127933168245, "loss": 0.2578, "step": 5410 }, { "epoch": 2.5463529411764707, "grad_norm": 0.5894057353612641, "learning_rate": 0.00023144447109461314, "loss": 0.3574, "step": 5411 }, { "epoch": 2.546823529411765, "grad_norm": 0.5871227787647295, "learning_rate": 0.0002314076559017865, "loss": 0.377, "step": 5412 }, { "epoch": 2.5472941176470587, "grad_norm": 0.8312472189284773, "learning_rate": 0.00023137083375634676, "loss": 0.4629, "step": 5413 }, { "epoch": 2.547764705882353, "grad_norm": 0.8149872973926496, "learning_rate": 0.00023133400466143867, "loss": 0.6094, "step": 5414 }, { "epoch": 2.548235294117647, "grad_norm": 0.15812360152622434, "learning_rate": 0.00023129716862020766, "loss": 0.0654, "step": 5415 }, { "epoch": 2.548705882352941, "grad_norm": 0.4707558347828575, "learning_rate": 0.0002312603256357997, "loss": 0.1797, "step": 5416 }, { "epoch": 2.549176470588235, "grad_norm": 0.8612873231185479, "learning_rate": 0.00023122347571136127, "loss": 0.4648, "step": 5417 }, { "epoch": 2.5496470588235294, "grad_norm": 1.05889228842311, "learning_rate": 0.00023118661885003969, "loss": 0.6602, "step": 5418 }, { "epoch": 2.5501176470588236, "grad_norm": 0.5970963691973102, "learning_rate": 0.00023114975505498264, "loss": 0.4043, "step": 5419 }, { "epoch": 2.550588235294118, "grad_norm": 0.8512387898370558, "learning_rate": 0.00023111288432933847, "loss": 0.582, "step": 5420 }, { "epoch": 2.5510588235294116, "grad_norm": 0.38820558887505946, "learning_rate": 0.0002310760066762561, "loss": 0.1484, "step": 5421 }, { "epoch": 2.551529411764706, "grad_norm": 0.7054931186631825, "learning_rate": 0.00023103912209888512, "loss": 0.3047, "step": 5422 }, { "epoch": 2.552, "grad_norm": 0.5848948101005098, "learning_rate": 0.00023100223060037558, "loss": 0.2539, "step": 5423 }, { "epoch": 2.5524705882352943, "grad_norm": 0.6425191160461174, "learning_rate": 0.00023096533218387827, "loss": 0.5234, "step": 5424 }, { "epoch": 2.552941176470588, "grad_norm": 0.48725496247146943, "learning_rate": 0.00023092842685254442, "loss": 0.2188, "step": 5425 }, { "epoch": 2.5534117647058823, "grad_norm": 0.9077043039357833, "learning_rate": 0.000230891514609526, "loss": 0.3633, "step": 5426 }, { "epoch": 2.5538823529411765, "grad_norm": 0.7432669842214747, "learning_rate": 0.00023085459545797542, "loss": 0.4902, "step": 5427 }, { "epoch": 2.5543529411764707, "grad_norm": 0.6951972343107778, "learning_rate": 0.00023081766940104585, "loss": 0.3926, "step": 5428 }, { "epoch": 2.554823529411765, "grad_norm": 0.34854766001264, "learning_rate": 0.00023078073644189085, "loss": 0.0737, "step": 5429 }, { "epoch": 2.5552941176470587, "grad_norm": 0.8400817393222086, "learning_rate": 0.00023074379658366478, "loss": 0.3027, "step": 5430 }, { "epoch": 2.555764705882353, "grad_norm": 1.6174087415863747, "learning_rate": 0.0002307068498295224, "loss": 0.5977, "step": 5431 }, { "epoch": 2.556235294117647, "grad_norm": 0.7018386321898995, "learning_rate": 0.00023066989618261919, "loss": 0.2852, "step": 5432 }, { "epoch": 2.556705882352941, "grad_norm": 0.38365826201368747, "learning_rate": 0.00023063293564611114, "loss": 0.1768, "step": 5433 }, { "epoch": 2.557176470588235, "grad_norm": 0.69939248029183, "learning_rate": 0.00023059596822315491, "loss": 0.4238, "step": 5434 }, { "epoch": 2.5576470588235294, "grad_norm": 0.6552285809721593, "learning_rate": 0.00023055899391690766, "loss": 0.3164, "step": 5435 }, { "epoch": 2.5581176470588236, "grad_norm": 0.4110846884989259, "learning_rate": 0.00023052201273052723, "loss": 0.126, "step": 5436 }, { "epoch": 2.558588235294118, "grad_norm": 0.41194015480502494, "learning_rate": 0.00023048502466717192, "loss": 0.1387, "step": 5437 }, { "epoch": 2.5590588235294116, "grad_norm": 1.2634489811942147, "learning_rate": 0.00023044802973000073, "loss": 0.3613, "step": 5438 }, { "epoch": 2.559529411764706, "grad_norm": 0.5035059434121655, "learning_rate": 0.00023041102792217323, "loss": 0.2256, "step": 5439 }, { "epoch": 2.56, "grad_norm": 0.5660540317992611, "learning_rate": 0.00023037401924684946, "loss": 0.1768, "step": 5440 }, { "epoch": 2.5604705882352943, "grad_norm": 0.4326172294815317, "learning_rate": 0.00023033700370719026, "loss": 0.3848, "step": 5441 }, { "epoch": 2.560941176470588, "grad_norm": 0.5363965647407598, "learning_rate": 0.00023029998130635692, "loss": 0.3125, "step": 5442 }, { "epoch": 2.5614117647058823, "grad_norm": 0.9584213897864251, "learning_rate": 0.00023026295204751126, "loss": 0.6406, "step": 5443 }, { "epoch": 2.5618823529411765, "grad_norm": 0.49595353184432833, "learning_rate": 0.00023022591593381582, "loss": 0.1982, "step": 5444 }, { "epoch": 2.5623529411764707, "grad_norm": 0.34702437052888696, "learning_rate": 0.0002301888729684337, "loss": 0.1562, "step": 5445 }, { "epoch": 2.562823529411765, "grad_norm": 0.8673896928032009, "learning_rate": 0.00023015182315452843, "loss": 0.3105, "step": 5446 }, { "epoch": 2.5632941176470587, "grad_norm": 0.38276877435818035, "learning_rate": 0.00023011476649526438, "loss": 0.1641, "step": 5447 }, { "epoch": 2.563764705882353, "grad_norm": 0.49710172622572213, "learning_rate": 0.0002300777029938063, "loss": 0.1279, "step": 5448 }, { "epoch": 2.564235294117647, "grad_norm": 0.7798761978255146, "learning_rate": 0.0002300406326533196, "loss": 0.3887, "step": 5449 }, { "epoch": 2.564705882352941, "grad_norm": 0.38794178115143363, "learning_rate": 0.00023000355547697027, "loss": 0.1816, "step": 5450 }, { "epoch": 2.565176470588235, "grad_norm": 0.8532111401389505, "learning_rate": 0.00022996647146792488, "loss": 0.5312, "step": 5451 }, { "epoch": 2.5656470588235294, "grad_norm": 0.45344311878744054, "learning_rate": 0.0002299293806293506, "loss": 0.168, "step": 5452 }, { "epoch": 2.5661176470588236, "grad_norm": 0.6776993369095108, "learning_rate": 0.00022989228296441515, "loss": 0.3262, "step": 5453 }, { "epoch": 2.566588235294118, "grad_norm": 0.6370228826889177, "learning_rate": 0.00022985517847628686, "loss": 0.4434, "step": 5454 }, { "epoch": 2.5670588235294116, "grad_norm": 0.5636412557033355, "learning_rate": 0.0002298180671681347, "loss": 0.3809, "step": 5455 }, { "epoch": 2.567529411764706, "grad_norm": 0.41048221621779274, "learning_rate": 0.00022978094904312803, "loss": 0.2109, "step": 5456 }, { "epoch": 2.568, "grad_norm": 0.698987884974562, "learning_rate": 0.000229743824104437, "loss": 0.3477, "step": 5457 }, { "epoch": 2.5684705882352943, "grad_norm": 0.2654895202376332, "learning_rate": 0.00022970669235523224, "loss": 0.0674, "step": 5458 }, { "epoch": 2.568941176470588, "grad_norm": 0.09497809744493142, "learning_rate": 0.00022966955379868502, "loss": 0.0159, "step": 5459 }, { "epoch": 2.5694117647058823, "grad_norm": 0.5800005513552452, "learning_rate": 0.00022963240843796714, "loss": 0.3125, "step": 5460 }, { "epoch": 2.5698823529411765, "grad_norm": 0.5489975870714889, "learning_rate": 0.000229595256276251, "loss": 0.3105, "step": 5461 }, { "epoch": 2.5703529411764707, "grad_norm": 0.24163315668339502, "learning_rate": 0.00022955809731670955, "loss": 0.0752, "step": 5462 }, { "epoch": 2.570823529411765, "grad_norm": 0.6244718360157333, "learning_rate": 0.00022952093156251635, "loss": 0.3398, "step": 5463 }, { "epoch": 2.5712941176470587, "grad_norm": 0.47672209773276325, "learning_rate": 0.0002294837590168455, "loss": 0.2041, "step": 5464 }, { "epoch": 2.571764705882353, "grad_norm": 0.8299952194158938, "learning_rate": 0.00022944657968287184, "loss": 0.3945, "step": 5465 }, { "epoch": 2.572235294117647, "grad_norm": 0.6802588221887002, "learning_rate": 0.00022940939356377052, "loss": 0.3633, "step": 5466 }, { "epoch": 2.572705882352941, "grad_norm": 0.4772594908714841, "learning_rate": 0.00022937220066271752, "loss": 0.2773, "step": 5467 }, { "epoch": 2.573176470588235, "grad_norm": 0.40553598436881666, "learning_rate": 0.0002293350009828893, "loss": 0.2051, "step": 5468 }, { "epoch": 2.5736470588235294, "grad_norm": 0.6131462814274042, "learning_rate": 0.00022929779452746283, "loss": 0.332, "step": 5469 }, { "epoch": 2.5741176470588236, "grad_norm": 0.461096357092062, "learning_rate": 0.0002292605812996157, "loss": 0.2617, "step": 5470 }, { "epoch": 2.574588235294118, "grad_norm": 0.3298253676257666, "learning_rate": 0.00022922336130252622, "loss": 0.084, "step": 5471 }, { "epoch": 2.5750588235294116, "grad_norm": 1.0711290118250958, "learning_rate": 0.00022918613453937314, "loss": 0.707, "step": 5472 }, { "epoch": 2.575529411764706, "grad_norm": 0.2776411181351887, "learning_rate": 0.00022914890101333573, "loss": 0.1553, "step": 5473 }, { "epoch": 2.576, "grad_norm": 0.6933065449440994, "learning_rate": 0.00022911166072759386, "loss": 0.2441, "step": 5474 }, { "epoch": 2.576470588235294, "grad_norm": 0.6799733967714799, "learning_rate": 0.00022907441368532822, "loss": 0.2598, "step": 5475 }, { "epoch": 2.576941176470588, "grad_norm": 0.4351007378646998, "learning_rate": 0.00022903715988971973, "loss": 0.1602, "step": 5476 }, { "epoch": 2.5774117647058823, "grad_norm": 0.5151373552383063, "learning_rate": 0.0002289998993439502, "loss": 0.207, "step": 5477 }, { "epoch": 2.5778823529411765, "grad_norm": 0.2684782796002512, "learning_rate": 0.00022896263205120167, "loss": 0.1108, "step": 5478 }, { "epoch": 2.5783529411764707, "grad_norm": 0.47657194063821395, "learning_rate": 0.0002289253580146571, "loss": 0.2949, "step": 5479 }, { "epoch": 2.578823529411765, "grad_norm": 0.6403648323881513, "learning_rate": 0.00022888807723749988, "loss": 0.3066, "step": 5480 }, { "epoch": 2.5792941176470587, "grad_norm": 0.8632469480149915, "learning_rate": 0.0002288507897229138, "loss": 0.2041, "step": 5481 }, { "epoch": 2.579764705882353, "grad_norm": 0.31828222711612686, "learning_rate": 0.00022881349547408363, "loss": 0.1328, "step": 5482 }, { "epoch": 2.580235294117647, "grad_norm": 0.8654196419350753, "learning_rate": 0.00022877619449419436, "loss": 0.543, "step": 5483 }, { "epoch": 2.580705882352941, "grad_norm": 0.31629076653609034, "learning_rate": 0.00022873888678643164, "loss": 0.1279, "step": 5484 }, { "epoch": 2.581176470588235, "grad_norm": 0.6415090331126613, "learning_rate": 0.0002287015723539818, "loss": 0.3613, "step": 5485 }, { "epoch": 2.5816470588235294, "grad_norm": 0.1898147309214671, "learning_rate": 0.00022866425120003165, "loss": 0.1289, "step": 5486 }, { "epoch": 2.5821176470588236, "grad_norm": 0.38581047982660915, "learning_rate": 0.0002286269233277686, "loss": 0.1309, "step": 5487 }, { "epoch": 2.582588235294118, "grad_norm": 0.8497142488134545, "learning_rate": 0.0002285895887403806, "loss": 0.3906, "step": 5488 }, { "epoch": 2.5830588235294116, "grad_norm": 0.5663731329301769, "learning_rate": 0.00022855224744105625, "loss": 0.2871, "step": 5489 }, { "epoch": 2.583529411764706, "grad_norm": 0.5582748650785885, "learning_rate": 0.00022851489943298472, "loss": 0.2393, "step": 5490 }, { "epoch": 2.584, "grad_norm": 0.3165626451191853, "learning_rate": 0.0002284775447193556, "loss": 0.1177, "step": 5491 }, { "epoch": 2.584470588235294, "grad_norm": 0.33751350634531957, "learning_rate": 0.00022844018330335928, "loss": 0.1504, "step": 5492 }, { "epoch": 2.584941176470588, "grad_norm": 0.6972313459558673, "learning_rate": 0.00022840281518818657, "loss": 0.6836, "step": 5493 }, { "epoch": 2.5854117647058823, "grad_norm": 0.5729875855397045, "learning_rate": 0.00022836544037702883, "loss": 0.3477, "step": 5494 }, { "epoch": 2.5858823529411765, "grad_norm": 0.5926954559057458, "learning_rate": 0.00022832805887307815, "loss": 0.3926, "step": 5495 }, { "epoch": 2.5863529411764707, "grad_norm": 0.6287409732566493, "learning_rate": 0.00022829067067952702, "loss": 0.2578, "step": 5496 }, { "epoch": 2.586823529411765, "grad_norm": 1.1652043453690597, "learning_rate": 0.00022825327579956863, "loss": 0.707, "step": 5497 }, { "epoch": 2.5872941176470587, "grad_norm": 0.4788245852183584, "learning_rate": 0.0002282158742363966, "loss": 0.1826, "step": 5498 }, { "epoch": 2.587764705882353, "grad_norm": 0.8242207519360928, "learning_rate": 0.00022817846599320528, "loss": 0.3809, "step": 5499 }, { "epoch": 2.588235294117647, "grad_norm": 0.7885732272377171, "learning_rate": 0.00022814105107318952, "loss": 0.3789, "step": 5500 }, { "epoch": 2.588705882352941, "grad_norm": 0.46524647598084395, "learning_rate": 0.00022810362947954475, "loss": 0.1953, "step": 5501 }, { "epoch": 2.589176470588235, "grad_norm": 0.4051183080342506, "learning_rate": 0.00022806620121546684, "loss": 0.1338, "step": 5502 }, { "epoch": 2.5896470588235294, "grad_norm": 0.4416297919213443, "learning_rate": 0.0002280287662841525, "loss": 0.2598, "step": 5503 }, { "epoch": 2.5901176470588236, "grad_norm": 0.4068389385883523, "learning_rate": 0.00022799132468879878, "loss": 0.1719, "step": 5504 }, { "epoch": 2.590588235294118, "grad_norm": 0.6462074038082136, "learning_rate": 0.00022795387643260336, "loss": 0.3477, "step": 5505 }, { "epoch": 2.5910588235294116, "grad_norm": 0.7455443025566505, "learning_rate": 0.00022791642151876455, "loss": 0.377, "step": 5506 }, { "epoch": 2.591529411764706, "grad_norm": 0.3577590736090264, "learning_rate": 0.00022787895995048118, "loss": 0.1904, "step": 5507 }, { "epoch": 2.592, "grad_norm": 0.7413760953919808, "learning_rate": 0.00022784149173095262, "loss": 0.3457, "step": 5508 }, { "epoch": 2.592470588235294, "grad_norm": 0.3059074303471914, "learning_rate": 0.00022780401686337885, "loss": 0.1328, "step": 5509 }, { "epoch": 2.592941176470588, "grad_norm": 0.43964792802290303, "learning_rate": 0.00022776653535096044, "loss": 0.2227, "step": 5510 }, { "epoch": 2.5934117647058823, "grad_norm": 0.49208907164658827, "learning_rate": 0.00022772904719689844, "loss": 0.2734, "step": 5511 }, { "epoch": 2.5938823529411765, "grad_norm": 0.43418927023905035, "learning_rate": 0.00022769155240439455, "loss": 0.2734, "step": 5512 }, { "epoch": 2.5943529411764708, "grad_norm": 0.39642211141015005, "learning_rate": 0.00022765405097665103, "loss": 0.1641, "step": 5513 }, { "epoch": 2.594823529411765, "grad_norm": 0.6075772788282886, "learning_rate": 0.00022761654291687065, "loss": 0.252, "step": 5514 }, { "epoch": 2.5952941176470588, "grad_norm": 0.49128660647490907, "learning_rate": 0.00022757902822825683, "loss": 0.2812, "step": 5515 }, { "epoch": 2.595764705882353, "grad_norm": 0.7223001284306708, "learning_rate": 0.00022754150691401345, "loss": 0.3789, "step": 5516 }, { "epoch": 2.596235294117647, "grad_norm": 0.9440583344750372, "learning_rate": 0.00022750397897734504, "loss": 0.4082, "step": 5517 }, { "epoch": 2.596705882352941, "grad_norm": 0.521945439384442, "learning_rate": 0.00022746644442145665, "loss": 0.1182, "step": 5518 }, { "epoch": 2.597176470588235, "grad_norm": 0.6910499059546841, "learning_rate": 0.00022742890324955396, "loss": 0.3945, "step": 5519 }, { "epoch": 2.5976470588235294, "grad_norm": 0.37140084211999475, "learning_rate": 0.00022739135546484315, "loss": 0.1367, "step": 5520 }, { "epoch": 2.5981176470588236, "grad_norm": 0.5891542959758188, "learning_rate": 0.00022735380107053094, "loss": 0.3477, "step": 5521 }, { "epoch": 2.598588235294118, "grad_norm": 0.4604500646264872, "learning_rate": 0.00022731624006982473, "loss": 0.1953, "step": 5522 }, { "epoch": 2.5990588235294116, "grad_norm": 0.5675610556617754, "learning_rate": 0.0002272786724659324, "loss": 0.2305, "step": 5523 }, { "epoch": 2.599529411764706, "grad_norm": 0.5425827580529484, "learning_rate": 0.00022724109826206235, "loss": 0.25, "step": 5524 }, { "epoch": 2.6, "grad_norm": 0.17309763452450191, "learning_rate": 0.00022720351746142366, "loss": 0.0898, "step": 5525 }, { "epoch": 2.600470588235294, "grad_norm": 0.45027876398945305, "learning_rate": 0.00022716593006722594, "loss": 0.1729, "step": 5526 }, { "epoch": 2.600941176470588, "grad_norm": 1.223685050202344, "learning_rate": 0.0002271283360826793, "loss": 0.8906, "step": 5527 }, { "epoch": 2.6014117647058823, "grad_norm": 0.5040236673049538, "learning_rate": 0.0002270907355109944, "loss": 0.1953, "step": 5528 }, { "epoch": 2.6018823529411765, "grad_norm": 1.1337299102159284, "learning_rate": 0.00022705312835538257, "loss": 0.7461, "step": 5529 }, { "epoch": 2.6023529411764708, "grad_norm": 1.3736689409765428, "learning_rate": 0.00022701551461905563, "loss": 1.6172, "step": 5530 }, { "epoch": 2.602823529411765, "grad_norm": 0.5430902921455337, "learning_rate": 0.00022697789430522602, "loss": 0.1768, "step": 5531 }, { "epoch": 2.6032941176470588, "grad_norm": 0.34970543927637393, "learning_rate": 0.00022694026741710657, "loss": 0.123, "step": 5532 }, { "epoch": 2.603764705882353, "grad_norm": 0.20917412322787357, "learning_rate": 0.00022690263395791095, "loss": 0.0688, "step": 5533 }, { "epoch": 2.604235294117647, "grad_norm": 0.4188138513786484, "learning_rate": 0.00022686499393085322, "loss": 0.1885, "step": 5534 }, { "epoch": 2.604705882352941, "grad_norm": 0.5708225876164785, "learning_rate": 0.00022682734733914792, "loss": 0.293, "step": 5535 }, { "epoch": 2.605176470588235, "grad_norm": 0.8796341131670267, "learning_rate": 0.00022678969418601038, "loss": 0.5195, "step": 5536 }, { "epoch": 2.6056470588235294, "grad_norm": 0.5657376592222856, "learning_rate": 0.0002267520344746563, "loss": 0.2891, "step": 5537 }, { "epoch": 2.6061176470588236, "grad_norm": 0.19891215617036406, "learning_rate": 0.000226714368208302, "loss": 0.0569, "step": 5538 }, { "epoch": 2.606588235294118, "grad_norm": 0.2738024407742498, "learning_rate": 0.00022667669539016435, "loss": 0.1885, "step": 5539 }, { "epoch": 2.6070588235294117, "grad_norm": 1.0457307820516315, "learning_rate": 0.00022663901602346082, "loss": 0.6641, "step": 5540 }, { "epoch": 2.607529411764706, "grad_norm": 0.35357430084650854, "learning_rate": 0.00022660133011140942, "loss": 0.1514, "step": 5541 }, { "epoch": 2.608, "grad_norm": 0.9463061639851986, "learning_rate": 0.00022656363765722872, "loss": 0.7266, "step": 5542 }, { "epoch": 2.608470588235294, "grad_norm": 0.6485120138409064, "learning_rate": 0.00022652593866413775, "loss": 0.3613, "step": 5543 }, { "epoch": 2.608941176470588, "grad_norm": 0.5803102471014759, "learning_rate": 0.00022648823313535635, "loss": 0.1875, "step": 5544 }, { "epoch": 2.6094117647058823, "grad_norm": 0.31901714008272686, "learning_rate": 0.00022645052107410463, "loss": 0.1055, "step": 5545 }, { "epoch": 2.6098823529411765, "grad_norm": 0.22344040278926247, "learning_rate": 0.0002264128024836034, "loss": 0.0498, "step": 5546 }, { "epoch": 2.6103529411764708, "grad_norm": 0.7048959447368551, "learning_rate": 0.00022637507736707408, "loss": 0.3105, "step": 5547 }, { "epoch": 2.6108235294117645, "grad_norm": 0.6547595111014132, "learning_rate": 0.00022633734572773852, "loss": 0.3496, "step": 5548 }, { "epoch": 2.6112941176470588, "grad_norm": 1.1093783078067347, "learning_rate": 0.00022629960756881916, "loss": 0.5586, "step": 5549 }, { "epoch": 2.611764705882353, "grad_norm": 0.31292893195735566, "learning_rate": 0.00022626186289353913, "loss": 0.1001, "step": 5550 }, { "epoch": 2.612235294117647, "grad_norm": 0.6568728977840544, "learning_rate": 0.00022622411170512193, "loss": 0.3223, "step": 5551 }, { "epoch": 2.612705882352941, "grad_norm": 0.5415382455654462, "learning_rate": 0.0002261863540067917, "loss": 0.2383, "step": 5552 }, { "epoch": 2.613176470588235, "grad_norm": 0.5303019717397162, "learning_rate": 0.00022614858980177314, "loss": 0.1504, "step": 5553 }, { "epoch": 2.6136470588235294, "grad_norm": 0.24388623633335157, "learning_rate": 0.00022611081909329153, "loss": 0.0601, "step": 5554 }, { "epoch": 2.6141176470588237, "grad_norm": 0.964010383431064, "learning_rate": 0.00022607304188457264, "loss": 0.9492, "step": 5555 }, { "epoch": 2.614588235294118, "grad_norm": 0.8038318784405831, "learning_rate": 0.0002260352581788428, "loss": 0.3594, "step": 5556 }, { "epoch": 2.6150588235294117, "grad_norm": 0.29734401611477346, "learning_rate": 0.000225997467979329, "loss": 0.1025, "step": 5557 }, { "epoch": 2.615529411764706, "grad_norm": 0.40332605850488673, "learning_rate": 0.00022595967128925866, "loss": 0.1235, "step": 5558 }, { "epoch": 2.616, "grad_norm": 0.37262985040642926, "learning_rate": 0.0002259218681118598, "loss": 0.082, "step": 5559 }, { "epoch": 2.616470588235294, "grad_norm": 0.5101408604520494, "learning_rate": 0.000225884058450361, "loss": 0.2412, "step": 5560 }, { "epoch": 2.616941176470588, "grad_norm": 0.49982865500169604, "learning_rate": 0.00022584624230799144, "loss": 0.293, "step": 5561 }, { "epoch": 2.6174117647058823, "grad_norm": 0.8544704487279419, "learning_rate": 0.00022580841968798072, "loss": 0.3047, "step": 5562 }, { "epoch": 2.6178823529411765, "grad_norm": 0.4481168701076181, "learning_rate": 0.00022577059059355914, "loss": 0.2373, "step": 5563 }, { "epoch": 2.6183529411764708, "grad_norm": 1.057033680460755, "learning_rate": 0.00022573275502795747, "loss": 0.7344, "step": 5564 }, { "epoch": 2.6188235294117646, "grad_norm": 0.4955906021126774, "learning_rate": 0.00022569491299440704, "loss": 0.249, "step": 5565 }, { "epoch": 2.6192941176470588, "grad_norm": 0.6768171198494805, "learning_rate": 0.00022565706449613978, "loss": 0.3398, "step": 5566 }, { "epoch": 2.619764705882353, "grad_norm": 0.4235924558597225, "learning_rate": 0.0002256192095363881, "loss": 0.1953, "step": 5567 }, { "epoch": 2.620235294117647, "grad_norm": 0.7415094211879192, "learning_rate": 0.00022558134811838504, "loss": 0.1738, "step": 5568 }, { "epoch": 2.620705882352941, "grad_norm": 0.6158226801337059, "learning_rate": 0.00022554348024536413, "loss": 0.209, "step": 5569 }, { "epoch": 2.621176470588235, "grad_norm": 0.9097995216291047, "learning_rate": 0.00022550560592055942, "loss": 0.3887, "step": 5570 }, { "epoch": 2.6216470588235294, "grad_norm": 0.5955169410403827, "learning_rate": 0.00022546772514720566, "loss": 0.2412, "step": 5571 }, { "epoch": 2.6221176470588237, "grad_norm": 0.3093852442367186, "learning_rate": 0.00022542983792853797, "loss": 0.2012, "step": 5572 }, { "epoch": 2.622588235294118, "grad_norm": 0.45874061898190793, "learning_rate": 0.00022539194426779216, "loss": 0.2129, "step": 5573 }, { "epoch": 2.6230588235294117, "grad_norm": 0.23992192100597926, "learning_rate": 0.00022535404416820453, "loss": 0.0574, "step": 5574 }, { "epoch": 2.623529411764706, "grad_norm": 0.685872579578883, "learning_rate": 0.0002253161376330119, "loss": 0.3594, "step": 5575 }, { "epoch": 2.624, "grad_norm": 0.5603948733263727, "learning_rate": 0.00022527822466545164, "loss": 0.2256, "step": 5576 }, { "epoch": 2.624470588235294, "grad_norm": 0.6209437541184025, "learning_rate": 0.00022524030526876183, "loss": 0.2871, "step": 5577 }, { "epoch": 2.624941176470588, "grad_norm": 0.7129110304283741, "learning_rate": 0.0002252023794461809, "loss": 0.4414, "step": 5578 }, { "epoch": 2.6254117647058823, "grad_norm": 0.7483243580803434, "learning_rate": 0.00022516444720094785, "loss": 0.6055, "step": 5579 }, { "epoch": 2.6258823529411766, "grad_norm": 0.583592060525961, "learning_rate": 0.00022512650853630234, "loss": 0.377, "step": 5580 }, { "epoch": 2.626352941176471, "grad_norm": 0.4075861218623849, "learning_rate": 0.00022508856345548455, "loss": 0.1982, "step": 5581 }, { "epoch": 2.6268235294117646, "grad_norm": 0.41305353140640266, "learning_rate": 0.00022505061196173513, "loss": 0.1934, "step": 5582 }, { "epoch": 2.627294117647059, "grad_norm": 0.49902316396201385, "learning_rate": 0.00022501265405829526, "loss": 0.2012, "step": 5583 }, { "epoch": 2.627764705882353, "grad_norm": 0.7127737771620065, "learning_rate": 0.00022497468974840687, "loss": 0.377, "step": 5584 }, { "epoch": 2.6282352941176472, "grad_norm": 0.719991397205954, "learning_rate": 0.00022493671903531218, "loss": 0.3633, "step": 5585 }, { "epoch": 2.628705882352941, "grad_norm": 1.2592957281365502, "learning_rate": 0.00022489874192225414, "loss": 0.6172, "step": 5586 }, { "epoch": 2.6291764705882352, "grad_norm": 0.6847479619574183, "learning_rate": 0.00022486075841247617, "loss": 0.3887, "step": 5587 }, { "epoch": 2.6296470588235294, "grad_norm": 0.3562034059484368, "learning_rate": 0.0002248227685092222, "loss": 0.1826, "step": 5588 }, { "epoch": 2.6301176470588237, "grad_norm": 0.8089287528573087, "learning_rate": 0.00022478477221573687, "loss": 0.4902, "step": 5589 }, { "epoch": 2.630588235294118, "grad_norm": 0.3860549113107808, "learning_rate": 0.00022474676953526508, "loss": 0.1348, "step": 5590 }, { "epoch": 2.6310588235294117, "grad_norm": 0.529942375366553, "learning_rate": 0.00022470876047105262, "loss": 0.3262, "step": 5591 }, { "epoch": 2.631529411764706, "grad_norm": 0.6085851864598586, "learning_rate": 0.00022467074502634558, "loss": 0.2256, "step": 5592 }, { "epoch": 2.632, "grad_norm": 0.8247694200860833, "learning_rate": 0.00022463272320439063, "loss": 0.334, "step": 5593 }, { "epoch": 2.632470588235294, "grad_norm": 0.7393502782526121, "learning_rate": 0.00022459469500843502, "loss": 0.4004, "step": 5594 }, { "epoch": 2.632941176470588, "grad_norm": 0.039530787367874506, "learning_rate": 0.0002245566604417266, "loss": 0.0042, "step": 5595 }, { "epoch": 2.6334117647058823, "grad_norm": 1.4116172168648635, "learning_rate": 0.0002245186195075137, "loss": 1.0156, "step": 5596 }, { "epoch": 2.6338823529411766, "grad_norm": 0.3446686281815196, "learning_rate": 0.0002244805722090451, "loss": 0.1592, "step": 5597 }, { "epoch": 2.634352941176471, "grad_norm": 0.41962761066767706, "learning_rate": 0.0002244425185495704, "loss": 0.1387, "step": 5598 }, { "epoch": 2.6348235294117646, "grad_norm": 0.6005711206254032, "learning_rate": 0.00022440445853233945, "loss": 0.2461, "step": 5599 }, { "epoch": 2.635294117647059, "grad_norm": 0.3523000099926978, "learning_rate": 0.00022436639216060275, "loss": 0.2158, "step": 5600 }, { "epoch": 2.635764705882353, "grad_norm": 0.7009534704845541, "learning_rate": 0.00022432831943761146, "loss": 0.3535, "step": 5601 }, { "epoch": 2.636235294117647, "grad_norm": 0.37781065454892526, "learning_rate": 0.00022429024036661708, "loss": 0.2021, "step": 5602 }, { "epoch": 2.636705882352941, "grad_norm": 0.6715494607190132, "learning_rate": 0.00022425215495087177, "loss": 0.4199, "step": 5603 }, { "epoch": 2.6371764705882352, "grad_norm": 0.448385869278421, "learning_rate": 0.00022421406319362823, "loss": 0.2314, "step": 5604 }, { "epoch": 2.6376470588235295, "grad_norm": 0.2291092215200947, "learning_rate": 0.00022417596509813968, "loss": 0.0684, "step": 5605 }, { "epoch": 2.6381176470588237, "grad_norm": 0.04851853518512525, "learning_rate": 0.0002241378606676599, "loss": 0.006, "step": 5606 }, { "epoch": 2.638588235294118, "grad_norm": 0.4329785907678265, "learning_rate": 0.00022409974990544312, "loss": 0.1826, "step": 5607 }, { "epoch": 2.6390588235294117, "grad_norm": 0.5918409889499947, "learning_rate": 0.0002240616328147443, "loss": 0.4082, "step": 5608 }, { "epoch": 2.639529411764706, "grad_norm": 1.0627310207657905, "learning_rate": 0.00022402350939881873, "loss": 0.5625, "step": 5609 }, { "epoch": 2.64, "grad_norm": 0.7485411518067263, "learning_rate": 0.00022398537966092236, "loss": 0.4082, "step": 5610 }, { "epoch": 2.640470588235294, "grad_norm": 0.5206652859533079, "learning_rate": 0.00022394724360431167, "loss": 0.3828, "step": 5611 }, { "epoch": 2.640941176470588, "grad_norm": 0.3847964065631854, "learning_rate": 0.00022390910123224373, "loss": 0.127, "step": 5612 }, { "epoch": 2.6414117647058823, "grad_norm": 0.567762658651356, "learning_rate": 0.00022387095254797597, "loss": 0.3594, "step": 5613 }, { "epoch": 2.6418823529411766, "grad_norm": 0.30187026613564516, "learning_rate": 0.00022383279755476651, "loss": 0.1035, "step": 5614 }, { "epoch": 2.642352941176471, "grad_norm": 0.8009111255446881, "learning_rate": 0.00022379463625587406, "loss": 0.4277, "step": 5615 }, { "epoch": 2.6428235294117646, "grad_norm": 0.9119102433170289, "learning_rate": 0.00022375646865455766, "loss": 0.543, "step": 5616 }, { "epoch": 2.643294117647059, "grad_norm": 0.06893636059146437, "learning_rate": 0.00022371829475407713, "loss": 0.0091, "step": 5617 }, { "epoch": 2.643764705882353, "grad_norm": 0.8558903289076308, "learning_rate": 0.00022368011455769251, "loss": 0.4277, "step": 5618 }, { "epoch": 2.644235294117647, "grad_norm": 0.7374659632649702, "learning_rate": 0.00022364192806866481, "loss": 0.3984, "step": 5619 }, { "epoch": 2.644705882352941, "grad_norm": 0.6309842074716011, "learning_rate": 0.00022360373529025523, "loss": 0.3477, "step": 5620 }, { "epoch": 2.6451764705882352, "grad_norm": 0.7581726839861296, "learning_rate": 0.00022356553622572564, "loss": 0.4395, "step": 5621 }, { "epoch": 2.6456470588235295, "grad_norm": 0.6585810572285379, "learning_rate": 0.00022352733087833838, "loss": 0.3594, "step": 5622 }, { "epoch": 2.6461176470588237, "grad_norm": 0.6832852014217388, "learning_rate": 0.00022348911925135646, "loss": 0.3887, "step": 5623 }, { "epoch": 2.646588235294118, "grad_norm": 0.5538821999815674, "learning_rate": 0.00022345090134804325, "loss": 0.3184, "step": 5624 }, { "epoch": 2.6470588235294117, "grad_norm": 0.25675560187331176, "learning_rate": 0.0002234126771716628, "loss": 0.1299, "step": 5625 }, { "epoch": 2.647529411764706, "grad_norm": 0.719983878544125, "learning_rate": 0.00022337444672547967, "loss": 0.3848, "step": 5626 }, { "epoch": 2.648, "grad_norm": 0.8194094036467646, "learning_rate": 0.00022333621001275884, "loss": 0.4121, "step": 5627 }, { "epoch": 2.648470588235294, "grad_norm": 0.43020969183263175, "learning_rate": 0.00022329796703676597, "loss": 0.2236, "step": 5628 }, { "epoch": 2.648941176470588, "grad_norm": 0.6629514420216228, "learning_rate": 0.00022325971780076721, "loss": 0.3691, "step": 5629 }, { "epoch": 2.6494117647058824, "grad_norm": 0.6663819453699507, "learning_rate": 0.00022322146230802924, "loss": 0.3574, "step": 5630 }, { "epoch": 2.6498823529411766, "grad_norm": 0.7190192664935228, "learning_rate": 0.0002231832005618192, "loss": 0.373, "step": 5631 }, { "epoch": 2.650352941176471, "grad_norm": 0.45670595786849366, "learning_rate": 0.00022314493256540485, "loss": 0.2266, "step": 5632 }, { "epoch": 2.6508235294117646, "grad_norm": 0.5703694224388972, "learning_rate": 0.0002231066583220545, "loss": 0.3105, "step": 5633 }, { "epoch": 2.651294117647059, "grad_norm": 0.586244879594808, "learning_rate": 0.00022306837783503698, "loss": 0.3145, "step": 5634 }, { "epoch": 2.651764705882353, "grad_norm": 0.5995595810279016, "learning_rate": 0.00022303009110762154, "loss": 0.3379, "step": 5635 }, { "epoch": 2.652235294117647, "grad_norm": 0.7913989917994299, "learning_rate": 0.00022299179814307818, "loss": 0.3496, "step": 5636 }, { "epoch": 2.652705882352941, "grad_norm": 0.5751268542150385, "learning_rate": 0.0002229534989446772, "loss": 0.2578, "step": 5637 }, { "epoch": 2.6531764705882352, "grad_norm": 0.8592041930506349, "learning_rate": 0.00022291519351568955, "loss": 0.4551, "step": 5638 }, { "epoch": 2.6536470588235295, "grad_norm": 0.496129832168406, "learning_rate": 0.00022287688185938673, "loss": 0.2598, "step": 5639 }, { "epoch": 2.6541176470588237, "grad_norm": 0.6926235552097688, "learning_rate": 0.00022283856397904077, "loss": 0.4727, "step": 5640 }, { "epoch": 2.654588235294118, "grad_norm": 0.3494913682769114, "learning_rate": 0.00022280023987792416, "loss": 0.1797, "step": 5641 }, { "epoch": 2.6550588235294117, "grad_norm": 0.44676229239280263, "learning_rate": 0.00022276190955930998, "loss": 0.2617, "step": 5642 }, { "epoch": 2.655529411764706, "grad_norm": 0.7230574387412679, "learning_rate": 0.00022272357302647187, "loss": 0.4961, "step": 5643 }, { "epoch": 2.656, "grad_norm": 0.5585006060983765, "learning_rate": 0.00022268523028268385, "loss": 0.3105, "step": 5644 }, { "epoch": 2.656470588235294, "grad_norm": 0.4240814481328337, "learning_rate": 0.00022264688133122069, "loss": 0.1689, "step": 5645 }, { "epoch": 2.656941176470588, "grad_norm": 0.4234906025748563, "learning_rate": 0.00022260852617535752, "loss": 0.252, "step": 5646 }, { "epoch": 2.6574117647058824, "grad_norm": 0.5711636881912378, "learning_rate": 0.00022257016481837006, "loss": 0.3105, "step": 5647 }, { "epoch": 2.6578823529411766, "grad_norm": 0.7028917579455725, "learning_rate": 0.00022253179726353458, "loss": 0.3223, "step": 5648 }, { "epoch": 2.658352941176471, "grad_norm": 0.5356281246087629, "learning_rate": 0.00022249342351412783, "loss": 0.2324, "step": 5649 }, { "epoch": 2.6588235294117646, "grad_norm": 0.524339267824856, "learning_rate": 0.00022245504357342716, "loss": 0.1748, "step": 5650 }, { "epoch": 2.659294117647059, "grad_norm": 0.65666478736624, "learning_rate": 0.00022241665744471037, "loss": 0.3262, "step": 5651 }, { "epoch": 2.659764705882353, "grad_norm": 0.22225151789157488, "learning_rate": 0.00022237826513125584, "loss": 0.0732, "step": 5652 }, { "epoch": 2.660235294117647, "grad_norm": 0.6567360370040344, "learning_rate": 0.00022233986663634241, "loss": 0.3242, "step": 5653 }, { "epoch": 2.660705882352941, "grad_norm": 0.5435885079337702, "learning_rate": 0.0002223014619632496, "loss": 0.3184, "step": 5654 }, { "epoch": 2.6611764705882353, "grad_norm": 0.03965171335747086, "learning_rate": 0.00022226305111525726, "loss": 0.0042, "step": 5655 }, { "epoch": 2.6616470588235295, "grad_norm": 0.7494341091199428, "learning_rate": 0.00022222463409564595, "loss": 0.3887, "step": 5656 }, { "epoch": 2.6621176470588237, "grad_norm": 0.7184313034437767, "learning_rate": 0.0002221862109076966, "loss": 0.4316, "step": 5657 }, { "epoch": 2.662588235294118, "grad_norm": 0.16442350111120652, "learning_rate": 0.00022214778155469077, "loss": 0.0299, "step": 5658 }, { "epoch": 2.6630588235294117, "grad_norm": 0.23772699201491526, "learning_rate": 0.0002221093460399105, "loss": 0.0957, "step": 5659 }, { "epoch": 2.663529411764706, "grad_norm": 0.6809802572479673, "learning_rate": 0.0002220709043666384, "loss": 0.3398, "step": 5660 }, { "epoch": 2.664, "grad_norm": 0.6049842789355948, "learning_rate": 0.0002220324565381576, "loss": 0.1147, "step": 5661 }, { "epoch": 2.664470588235294, "grad_norm": 0.7263486450297874, "learning_rate": 0.0002219940025577516, "loss": 0.459, "step": 5662 }, { "epoch": 2.664941176470588, "grad_norm": 0.27141791869233545, "learning_rate": 0.00022195554242870473, "loss": 0.061, "step": 5663 }, { "epoch": 2.6654117647058824, "grad_norm": 1.0832919106538823, "learning_rate": 0.0002219170761543016, "loss": 0.4395, "step": 5664 }, { "epoch": 2.6658823529411766, "grad_norm": 0.5212976776357282, "learning_rate": 0.00022187860373782737, "loss": 0.2734, "step": 5665 }, { "epoch": 2.666352941176471, "grad_norm": 0.7653515873480496, "learning_rate": 0.00022184012518256785, "loss": 0.4805, "step": 5666 }, { "epoch": 2.6668235294117646, "grad_norm": 0.4786237674204526, "learning_rate": 0.00022180164049180926, "loss": 0.3359, "step": 5667 }, { "epoch": 2.667294117647059, "grad_norm": 0.4751110414465731, "learning_rate": 0.00022176314966883837, "loss": 0.1992, "step": 5668 }, { "epoch": 2.667764705882353, "grad_norm": 0.30432170615769194, "learning_rate": 0.0002217246527169425, "loss": 0.1006, "step": 5669 }, { "epoch": 2.668235294117647, "grad_norm": 0.5790737658729866, "learning_rate": 0.00022168614963940952, "loss": 0.3301, "step": 5670 }, { "epoch": 2.668705882352941, "grad_norm": 0.44595104066296615, "learning_rate": 0.00022164764043952772, "loss": 0.2354, "step": 5671 }, { "epoch": 2.6691764705882353, "grad_norm": 0.8931277253342417, "learning_rate": 0.00022160912512058598, "loss": 0.5273, "step": 5672 }, { "epoch": 2.6696470588235295, "grad_norm": 0.3780533057871416, "learning_rate": 0.0002215706036858737, "loss": 0.0781, "step": 5673 }, { "epoch": 2.6701176470588237, "grad_norm": 0.2906265906057969, "learning_rate": 0.00022153207613868087, "loss": 0.2002, "step": 5674 }, { "epoch": 2.6705882352941175, "grad_norm": 0.1381483689309586, "learning_rate": 0.00022149354248229784, "loss": 0.0364, "step": 5675 }, { "epoch": 2.6710588235294117, "grad_norm": 0.2900825789124446, "learning_rate": 0.00022145500272001558, "loss": 0.0669, "step": 5676 }, { "epoch": 2.671529411764706, "grad_norm": 0.46419540190707664, "learning_rate": 0.00022141645685512564, "loss": 0.1455, "step": 5677 }, { "epoch": 2.672, "grad_norm": 0.662895349832821, "learning_rate": 0.00022137790489091999, "loss": 0.3828, "step": 5678 }, { "epoch": 2.672470588235294, "grad_norm": 0.45692155275875246, "learning_rate": 0.00022133934683069115, "loss": 0.2559, "step": 5679 }, { "epoch": 2.672941176470588, "grad_norm": 0.5246189926609092, "learning_rate": 0.00022130078267773218, "loss": 0.2236, "step": 5680 }, { "epoch": 2.6734117647058824, "grad_norm": 0.7671590063935863, "learning_rate": 0.00022126221243533664, "loss": 0.4082, "step": 5681 }, { "epoch": 2.6738823529411766, "grad_norm": 0.40219643325115934, "learning_rate": 0.00022122363610679863, "loss": 0.1953, "step": 5682 }, { "epoch": 2.674352941176471, "grad_norm": 0.5066931996193308, "learning_rate": 0.0002211850536954127, "loss": 0.1953, "step": 5683 }, { "epoch": 2.6748235294117646, "grad_norm": 0.6293659705163136, "learning_rate": 0.00022114646520447406, "loss": 0.377, "step": 5684 }, { "epoch": 2.675294117647059, "grad_norm": 0.238933475746968, "learning_rate": 0.00022110787063727832, "loss": 0.1079, "step": 5685 }, { "epoch": 2.675764705882353, "grad_norm": 0.8713924648806143, "learning_rate": 0.00022106926999712165, "loss": 0.5469, "step": 5686 }, { "epoch": 2.676235294117647, "grad_norm": 0.5897399009724148, "learning_rate": 0.0002210306632873007, "loss": 0.4277, "step": 5687 }, { "epoch": 2.676705882352941, "grad_norm": 0.25421183581127005, "learning_rate": 0.00022099205051111272, "loss": 0.053, "step": 5688 }, { "epoch": 2.6771764705882353, "grad_norm": 0.5704023026597387, "learning_rate": 0.0002209534316718554, "loss": 0.2217, "step": 5689 }, { "epoch": 2.6776470588235295, "grad_norm": 0.3672859072301418, "learning_rate": 0.000220914806772827, "loss": 0.1133, "step": 5690 }, { "epoch": 2.6781176470588237, "grad_norm": 0.402693708203335, "learning_rate": 0.0002208761758173263, "loss": 0.1357, "step": 5691 }, { "epoch": 2.6785882352941175, "grad_norm": 0.697486750512712, "learning_rate": 0.0002208375388086525, "loss": 0.6484, "step": 5692 }, { "epoch": 2.6790588235294117, "grad_norm": 0.545847265198256, "learning_rate": 0.00022079889575010543, "loss": 0.126, "step": 5693 }, { "epoch": 2.679529411764706, "grad_norm": 0.588468366656754, "learning_rate": 0.0002207602466449854, "loss": 0.3203, "step": 5694 }, { "epoch": 2.68, "grad_norm": 0.5787047440715446, "learning_rate": 0.00022072159149659327, "loss": 0.2676, "step": 5695 }, { "epoch": 2.680470588235294, "grad_norm": 0.3762144121108437, "learning_rate": 0.00022068293030823026, "loss": 0.165, "step": 5696 }, { "epoch": 2.680941176470588, "grad_norm": 0.3712083772840556, "learning_rate": 0.00022064426308319838, "loss": 0.1406, "step": 5697 }, { "epoch": 2.6814117647058824, "grad_norm": 0.6142567954042727, "learning_rate": 0.0002206055898247999, "loss": 0.2695, "step": 5698 }, { "epoch": 2.6818823529411766, "grad_norm": 0.5296936040118292, "learning_rate": 0.0002205669105363377, "loss": 0.2129, "step": 5699 }, { "epoch": 2.682352941176471, "grad_norm": 0.3695930264182106, "learning_rate": 0.00022052822522111522, "loss": 0.167, "step": 5700 }, { "epoch": 2.6828235294117646, "grad_norm": 0.3376298557751779, "learning_rate": 0.0002204895338824364, "loss": 0.207, "step": 5701 }, { "epoch": 2.683294117647059, "grad_norm": 0.8924015645752833, "learning_rate": 0.0002204508365236056, "loss": 0.4512, "step": 5702 }, { "epoch": 2.683764705882353, "grad_norm": 0.39811006193442716, "learning_rate": 0.0002204121331479278, "loss": 0.123, "step": 5703 }, { "epoch": 2.684235294117647, "grad_norm": 0.4340518548125632, "learning_rate": 0.0002203734237587085, "loss": 0.3945, "step": 5704 }, { "epoch": 2.684705882352941, "grad_norm": 0.5230108461334062, "learning_rate": 0.00022033470835925361, "loss": 0.2637, "step": 5705 }, { "epoch": 2.6851764705882353, "grad_norm": 0.4692377404661241, "learning_rate": 0.00022029598695286963, "loss": 0.1816, "step": 5706 }, { "epoch": 2.6856470588235295, "grad_norm": 0.2544928269738162, "learning_rate": 0.00022025725954286356, "loss": 0.0889, "step": 5707 }, { "epoch": 2.6861176470588237, "grad_norm": 1.1518131054423812, "learning_rate": 0.00022021852613254297, "loss": 0.6328, "step": 5708 }, { "epoch": 2.6865882352941175, "grad_norm": 0.6567722490403293, "learning_rate": 0.00022017978672521577, "loss": 0.3809, "step": 5709 }, { "epoch": 2.6870588235294117, "grad_norm": 0.9325320102196359, "learning_rate": 0.00022014104132419055, "loss": 0.75, "step": 5710 }, { "epoch": 2.687529411764706, "grad_norm": 0.6814552432556048, "learning_rate": 0.0002201022899327764, "loss": 0.3828, "step": 5711 }, { "epoch": 2.6879999999999997, "grad_norm": 0.6853852867950758, "learning_rate": 0.00022006353255428281, "loss": 0.2617, "step": 5712 }, { "epoch": 2.688470588235294, "grad_norm": 0.5378140506515986, "learning_rate": 0.00022002476919201987, "loss": 0.3105, "step": 5713 }, { "epoch": 2.688941176470588, "grad_norm": 0.2216199677068858, "learning_rate": 0.00021998599984929818, "loss": 0.1138, "step": 5714 }, { "epoch": 2.6894117647058824, "grad_norm": 0.7266113163155263, "learning_rate": 0.00021994722452942886, "loss": 0.332, "step": 5715 }, { "epoch": 2.6898823529411766, "grad_norm": 0.6407671334117209, "learning_rate": 0.00021990844323572346, "loss": 0.3906, "step": 5716 }, { "epoch": 2.690352941176471, "grad_norm": 0.34227635522298056, "learning_rate": 0.0002198696559714941, "loss": 0.084, "step": 5717 }, { "epoch": 2.6908235294117646, "grad_norm": 0.6252115338065254, "learning_rate": 0.0002198308627400534, "loss": 0.2754, "step": 5718 }, { "epoch": 2.691294117647059, "grad_norm": 0.4362780084237329, "learning_rate": 0.00021979206354471453, "loss": 0.1748, "step": 5719 }, { "epoch": 2.691764705882353, "grad_norm": 0.9345607731993836, "learning_rate": 0.00021975325838879107, "loss": 0.5312, "step": 5720 }, { "epoch": 2.692235294117647, "grad_norm": 0.6019298162305381, "learning_rate": 0.0002197144472755972, "loss": 0.3398, "step": 5721 }, { "epoch": 2.692705882352941, "grad_norm": 0.3351726121323779, "learning_rate": 0.00021967563020844762, "loss": 0.1396, "step": 5722 }, { "epoch": 2.6931764705882353, "grad_norm": 0.3452037530409717, "learning_rate": 0.00021963680719065746, "loss": 0.1138, "step": 5723 }, { "epoch": 2.6936470588235295, "grad_norm": 0.44084586332449227, "learning_rate": 0.0002195979782255424, "loss": 0.332, "step": 5724 }, { "epoch": 2.6941176470588237, "grad_norm": 0.6213184936798757, "learning_rate": 0.0002195591433164186, "loss": 0.2832, "step": 5725 }, { "epoch": 2.6945882352941175, "grad_norm": 0.6355614876630944, "learning_rate": 0.00021952030246660283, "loss": 0.248, "step": 5726 }, { "epoch": 2.6950588235294117, "grad_norm": 0.9013067701462507, "learning_rate": 0.00021948145567941217, "loss": 0.334, "step": 5727 }, { "epoch": 2.695529411764706, "grad_norm": 0.5422327555017841, "learning_rate": 0.0002194426029581644, "loss": 0.2393, "step": 5728 }, { "epoch": 2.6959999999999997, "grad_norm": 0.9173806727362426, "learning_rate": 0.00021940374430617777, "loss": 0.7344, "step": 5729 }, { "epoch": 2.696470588235294, "grad_norm": 0.3463904469580823, "learning_rate": 0.0002193648797267709, "loss": 0.1201, "step": 5730 }, { "epoch": 2.696941176470588, "grad_norm": 0.38703126052627235, "learning_rate": 0.0002193260092232631, "loss": 0.209, "step": 5731 }, { "epoch": 2.6974117647058824, "grad_norm": 0.5844285606398748, "learning_rate": 0.00021928713279897408, "loss": 0.1904, "step": 5732 }, { "epoch": 2.6978823529411766, "grad_norm": 1.0516739167027085, "learning_rate": 0.00021924825045722404, "loss": 0.5625, "step": 5733 }, { "epoch": 2.698352941176471, "grad_norm": 0.5734000376276852, "learning_rate": 0.00021920936220133373, "loss": 0.418, "step": 5734 }, { "epoch": 2.6988235294117646, "grad_norm": 0.42409255788468486, "learning_rate": 0.0002191704680346245, "loss": 0.1318, "step": 5735 }, { "epoch": 2.699294117647059, "grad_norm": 0.4692572776465915, "learning_rate": 0.00021913156796041794, "loss": 0.2188, "step": 5736 }, { "epoch": 2.699764705882353, "grad_norm": 0.824958453067471, "learning_rate": 0.0002190926619820364, "loss": 0.3242, "step": 5737 }, { "epoch": 2.700235294117647, "grad_norm": 0.6448292050567147, "learning_rate": 0.00021905375010280266, "loss": 0.2061, "step": 5738 }, { "epoch": 2.700705882352941, "grad_norm": 0.45970327263632327, "learning_rate": 0.00021901483232603995, "loss": 0.3008, "step": 5739 }, { "epoch": 2.7011764705882353, "grad_norm": 0.3805846974599225, "learning_rate": 0.000218975908655072, "loss": 0.1738, "step": 5740 }, { "epoch": 2.7016470588235295, "grad_norm": 0.732816669065521, "learning_rate": 0.0002189369790932232, "loss": 0.3535, "step": 5741 }, { "epoch": 2.7021176470588237, "grad_norm": 0.7675450513670273, "learning_rate": 0.0002188980436438182, "loss": 0.3066, "step": 5742 }, { "epoch": 2.7025882352941175, "grad_norm": 0.7221321529853295, "learning_rate": 0.0002188591023101823, "loss": 0.2754, "step": 5743 }, { "epoch": 2.7030588235294117, "grad_norm": 0.4222566922161189, "learning_rate": 0.00021882015509564132, "loss": 0.2207, "step": 5744 }, { "epoch": 2.703529411764706, "grad_norm": 0.4353999933769465, "learning_rate": 0.00021878120200352158, "loss": 0.1846, "step": 5745 }, { "epoch": 2.7039999999999997, "grad_norm": 0.30249273106746216, "learning_rate": 0.00021874224303714983, "loss": 0.1416, "step": 5746 }, { "epoch": 2.704470588235294, "grad_norm": 0.861918342005828, "learning_rate": 0.00021870327819985331, "loss": 0.5977, "step": 5747 }, { "epoch": 2.704941176470588, "grad_norm": 0.6065734685619708, "learning_rate": 0.00021866430749495987, "loss": 0.3555, "step": 5748 }, { "epoch": 2.7054117647058824, "grad_norm": 0.4625601396523521, "learning_rate": 0.00021862533092579774, "loss": 0.2158, "step": 5749 }, { "epoch": 2.7058823529411766, "grad_norm": 0.7366646204091252, "learning_rate": 0.00021858634849569576, "loss": 0.582, "step": 5750 }, { "epoch": 2.706352941176471, "grad_norm": 0.9032804831274797, "learning_rate": 0.00021854736020798319, "loss": 0.6406, "step": 5751 }, { "epoch": 2.7068235294117646, "grad_norm": 0.7504390047326555, "learning_rate": 0.00021850836606598987, "loss": 0.5039, "step": 5752 }, { "epoch": 2.707294117647059, "grad_norm": 0.3313749993317656, "learning_rate": 0.00021846936607304603, "loss": 0.1484, "step": 5753 }, { "epoch": 2.707764705882353, "grad_norm": 0.2880049606168152, "learning_rate": 0.00021843036023248247, "loss": 0.084, "step": 5754 }, { "epoch": 2.708235294117647, "grad_norm": 0.5879351299029894, "learning_rate": 0.00021839134854763055, "loss": 0.3496, "step": 5755 }, { "epoch": 2.708705882352941, "grad_norm": 0.45431168097367164, "learning_rate": 0.00021835233102182199, "loss": 0.2363, "step": 5756 }, { "epoch": 2.7091764705882353, "grad_norm": 0.22022636085613895, "learning_rate": 0.00021831330765838903, "loss": 0.0276, "step": 5757 }, { "epoch": 2.7096470588235295, "grad_norm": 0.5739720268047102, "learning_rate": 0.00021827427846066462, "loss": 0.3828, "step": 5758 }, { "epoch": 2.7101176470588237, "grad_norm": 0.4183567410519087, "learning_rate": 0.0002182352434319819, "loss": 0.1709, "step": 5759 }, { "epoch": 2.7105882352941175, "grad_norm": 0.5075696478667784, "learning_rate": 0.0002181962025756747, "loss": 0.1904, "step": 5760 }, { "epoch": 2.7110588235294117, "grad_norm": 0.3613644848486076, "learning_rate": 0.00021815715589507728, "loss": 0.209, "step": 5761 }, { "epoch": 2.711529411764706, "grad_norm": 0.6729611184454214, "learning_rate": 0.00021811810339352446, "loss": 0.2598, "step": 5762 }, { "epoch": 2.7119999999999997, "grad_norm": 0.7334241556930056, "learning_rate": 0.0002180790450743515, "loss": 0.2852, "step": 5763 }, { "epoch": 2.712470588235294, "grad_norm": 0.26088259893031573, "learning_rate": 0.0002180399809408941, "loss": 0.1035, "step": 5764 }, { "epoch": 2.712941176470588, "grad_norm": 0.21539961493556128, "learning_rate": 0.00021800091099648865, "loss": 0.0581, "step": 5765 }, { "epoch": 2.7134117647058824, "grad_norm": 0.6256556581434194, "learning_rate": 0.0002179618352444719, "loss": 0.3945, "step": 5766 }, { "epoch": 2.7138823529411766, "grad_norm": 1.035656570051487, "learning_rate": 0.00021792275368818102, "loss": 0.6406, "step": 5767 }, { "epoch": 2.714352941176471, "grad_norm": 0.2945959231969235, "learning_rate": 0.00021788366633095375, "loss": 0.1069, "step": 5768 }, { "epoch": 2.7148235294117646, "grad_norm": 0.6560833152511912, "learning_rate": 0.00021784457317612852, "loss": 0.3828, "step": 5769 }, { "epoch": 2.715294117647059, "grad_norm": 0.3224015628590368, "learning_rate": 0.0002178054742270439, "loss": 0.1035, "step": 5770 }, { "epoch": 2.715764705882353, "grad_norm": 1.0508507080427614, "learning_rate": 0.0002177663694870392, "loss": 0.6016, "step": 5771 }, { "epoch": 2.716235294117647, "grad_norm": 0.4173684867741772, "learning_rate": 0.00021772725895945422, "loss": 0.2021, "step": 5772 }, { "epoch": 2.716705882352941, "grad_norm": 0.6867062510076279, "learning_rate": 0.0002176881426476291, "loss": 0.5039, "step": 5773 }, { "epoch": 2.7171764705882353, "grad_norm": 0.2808782833372979, "learning_rate": 0.00021764902055490456, "loss": 0.1289, "step": 5774 }, { "epoch": 2.7176470588235295, "grad_norm": 0.3635207677233803, "learning_rate": 0.00021760989268462182, "loss": 0.1465, "step": 5775 }, { "epoch": 2.7181176470588237, "grad_norm": 0.25807366078253324, "learning_rate": 0.00021757075904012266, "loss": 0.0869, "step": 5776 }, { "epoch": 2.7185882352941175, "grad_norm": 0.40867815919080047, "learning_rate": 0.00021753161962474924, "loss": 0.1445, "step": 5777 }, { "epoch": 2.7190588235294118, "grad_norm": 0.7941231454337039, "learning_rate": 0.00021749247444184422, "loss": 0.3184, "step": 5778 }, { "epoch": 2.719529411764706, "grad_norm": 0.9547618704244681, "learning_rate": 0.00021745332349475092, "loss": 0.3848, "step": 5779 }, { "epoch": 2.7199999999999998, "grad_norm": 0.8500541031964702, "learning_rate": 0.0002174141667868129, "loss": 0.4668, "step": 5780 }, { "epoch": 2.720470588235294, "grad_norm": 0.3344397393560259, "learning_rate": 0.00021737500432137435, "loss": 0.1426, "step": 5781 }, { "epoch": 2.720941176470588, "grad_norm": 0.8393653753080613, "learning_rate": 0.00021733583610177997, "loss": 0.4961, "step": 5782 }, { "epoch": 2.7214117647058824, "grad_norm": 1.0232881528357936, "learning_rate": 0.00021729666213137494, "loss": 0.4355, "step": 5783 }, { "epoch": 2.7218823529411766, "grad_norm": 0.26637702429216975, "learning_rate": 0.00021725748241350486, "loss": 0.103, "step": 5784 }, { "epoch": 2.722352941176471, "grad_norm": 0.1334517085990278, "learning_rate": 0.00021721829695151585, "loss": 0.0253, "step": 5785 }, { "epoch": 2.7228235294117646, "grad_norm": 0.8161558620420155, "learning_rate": 0.00021717910574875458, "loss": 0.5391, "step": 5786 }, { "epoch": 2.723294117647059, "grad_norm": 0.2713264188193682, "learning_rate": 0.00021713990880856818, "loss": 0.1113, "step": 5787 }, { "epoch": 2.723764705882353, "grad_norm": 0.8622061019504709, "learning_rate": 0.0002171007061343043, "loss": 0.4414, "step": 5788 }, { "epoch": 2.724235294117647, "grad_norm": 0.9456332338985828, "learning_rate": 0.0002170614977293109, "loss": 0.4238, "step": 5789 }, { "epoch": 2.724705882352941, "grad_norm": 0.8771742723574395, "learning_rate": 0.00021702228359693672, "loss": 0.3906, "step": 5790 }, { "epoch": 2.7251764705882353, "grad_norm": 0.31639596452181745, "learning_rate": 0.00021698306374053075, "loss": 0.1099, "step": 5791 }, { "epoch": 2.7256470588235295, "grad_norm": 0.6522207483396305, "learning_rate": 0.00021694383816344258, "loss": 0.2422, "step": 5792 }, { "epoch": 2.7261176470588238, "grad_norm": 0.41263385106768724, "learning_rate": 0.0002169046068690223, "loss": 0.2793, "step": 5793 }, { "epoch": 2.7265882352941175, "grad_norm": 0.2398931929322939, "learning_rate": 0.00021686536986062043, "loss": 0.0615, "step": 5794 }, { "epoch": 2.7270588235294118, "grad_norm": 0.6668323658895895, "learning_rate": 0.000216826127141588, "loss": 0.3555, "step": 5795 }, { "epoch": 2.727529411764706, "grad_norm": 0.770968328217191, "learning_rate": 0.0002167868787152765, "loss": 0.4453, "step": 5796 }, { "epoch": 2.7279999999999998, "grad_norm": 0.46660621497323634, "learning_rate": 0.000216747624585038, "loss": 0.2676, "step": 5797 }, { "epoch": 2.728470588235294, "grad_norm": 0.7373628629089988, "learning_rate": 0.00021670836475422498, "loss": 0.3965, "step": 5798 }, { "epoch": 2.728941176470588, "grad_norm": 0.7781170743105642, "learning_rate": 0.00021666909922619037, "loss": 0.3145, "step": 5799 }, { "epoch": 2.7294117647058824, "grad_norm": 0.5378833409388679, "learning_rate": 0.0002166298280042877, "loss": 0.3164, "step": 5800 }, { "epoch": 2.7298823529411766, "grad_norm": 0.5060844946242495, "learning_rate": 0.00021659055109187088, "loss": 0.2256, "step": 5801 }, { "epoch": 2.7303529411764704, "grad_norm": 0.8358967651455761, "learning_rate": 0.0002165512684922944, "loss": 0.5391, "step": 5802 }, { "epoch": 2.7308235294117646, "grad_norm": 0.40518707413450716, "learning_rate": 0.0002165119802089131, "loss": 0.1816, "step": 5803 }, { "epoch": 2.731294117647059, "grad_norm": 0.7019615581140011, "learning_rate": 0.00021647268624508254, "loss": 0.4688, "step": 5804 }, { "epoch": 2.731764705882353, "grad_norm": 0.7094199513723102, "learning_rate": 0.00021643338660415844, "loss": 0.3926, "step": 5805 }, { "epoch": 2.732235294117647, "grad_norm": 0.6961257590324894, "learning_rate": 0.0002163940812894973, "loss": 0.4629, "step": 5806 }, { "epoch": 2.732705882352941, "grad_norm": 0.6608889167413357, "learning_rate": 0.00021635477030445597, "loss": 0.3359, "step": 5807 }, { "epoch": 2.7331764705882353, "grad_norm": 0.49546336668332486, "learning_rate": 0.0002163154536523918, "loss": 0.2471, "step": 5808 }, { "epoch": 2.7336470588235295, "grad_norm": 0.3753330556915324, "learning_rate": 0.0002162761313366626, "loss": 0.1973, "step": 5809 }, { "epoch": 2.7341176470588238, "grad_norm": 0.04148487787385711, "learning_rate": 0.0002162368033606267, "loss": 0.012, "step": 5810 }, { "epoch": 2.7345882352941175, "grad_norm": 0.30727794700956346, "learning_rate": 0.00021619746972764288, "loss": 0.1279, "step": 5811 }, { "epoch": 2.7350588235294118, "grad_norm": 1.2902137069129334, "learning_rate": 0.00021615813044107045, "loss": 0.5078, "step": 5812 }, { "epoch": 2.735529411764706, "grad_norm": 0.5679445564569252, "learning_rate": 0.00021611878550426922, "loss": 0.2734, "step": 5813 }, { "epoch": 2.7359999999999998, "grad_norm": 0.2942919130150801, "learning_rate": 0.00021607943492059936, "loss": 0.1006, "step": 5814 }, { "epoch": 2.736470588235294, "grad_norm": 0.06907875407417392, "learning_rate": 0.00021604007869342168, "loss": 0.0125, "step": 5815 }, { "epoch": 2.736941176470588, "grad_norm": 0.47197967719431916, "learning_rate": 0.00021600071682609733, "loss": 0.1582, "step": 5816 }, { "epoch": 2.7374117647058824, "grad_norm": 0.4514308307942719, "learning_rate": 0.00021596134932198803, "loss": 0.1318, "step": 5817 }, { "epoch": 2.7378823529411767, "grad_norm": 1.2389339367907548, "learning_rate": 0.00021592197618445596, "loss": 0.5898, "step": 5818 }, { "epoch": 2.7383529411764704, "grad_norm": 0.40280036175453876, "learning_rate": 0.00021588259741686373, "loss": 0.165, "step": 5819 }, { "epoch": 2.7388235294117647, "grad_norm": 0.9430323540311641, "learning_rate": 0.00021584321302257464, "loss": 0.4004, "step": 5820 }, { "epoch": 2.739294117647059, "grad_norm": 0.42414010789169576, "learning_rate": 0.00021580382300495215, "loss": 0.2031, "step": 5821 }, { "epoch": 2.739764705882353, "grad_norm": 0.6315760084540638, "learning_rate": 0.00021576442736736032, "loss": 0.3242, "step": 5822 }, { "epoch": 2.740235294117647, "grad_norm": 0.26120007953093843, "learning_rate": 0.00021572502611316393, "loss": 0.0928, "step": 5823 }, { "epoch": 2.740705882352941, "grad_norm": 0.3423891587917207, "learning_rate": 0.0002156856192457279, "loss": 0.0908, "step": 5824 }, { "epoch": 2.7411764705882353, "grad_norm": 0.8236069105220113, "learning_rate": 0.00021564620676841782, "loss": 0.3301, "step": 5825 }, { "epoch": 2.7416470588235295, "grad_norm": 0.31598366253294047, "learning_rate": 0.0002156067886845996, "loss": 0.0957, "step": 5826 }, { "epoch": 2.7421176470588238, "grad_norm": 1.16732532093957, "learning_rate": 0.00021556736499763992, "loss": 0.4004, "step": 5827 }, { "epoch": 2.7425882352941175, "grad_norm": 1.1150886911171225, "learning_rate": 0.00021552793571090568, "loss": 0.5391, "step": 5828 }, { "epoch": 2.7430588235294118, "grad_norm": 0.7527953824458513, "learning_rate": 0.0002154885008277642, "loss": 0.3945, "step": 5829 }, { "epoch": 2.743529411764706, "grad_norm": 0.16338139134628332, "learning_rate": 0.0002154490603515836, "loss": 0.0635, "step": 5830 }, { "epoch": 2.7439999999999998, "grad_norm": 0.283614308019049, "learning_rate": 0.0002154096142857323, "loss": 0.0981, "step": 5831 }, { "epoch": 2.744470588235294, "grad_norm": 0.2520345571069311, "learning_rate": 0.000215370162633579, "loss": 0.082, "step": 5832 }, { "epoch": 2.744941176470588, "grad_norm": 0.7562892571886374, "learning_rate": 0.00021533070539849316, "loss": 0.2119, "step": 5833 }, { "epoch": 2.7454117647058824, "grad_norm": 0.5025635968772647, "learning_rate": 0.00021529124258384466, "loss": 0.0977, "step": 5834 }, { "epoch": 2.7458823529411767, "grad_norm": 0.8389704756366402, "learning_rate": 0.00021525177419300383, "loss": 0.3398, "step": 5835 }, { "epoch": 2.7463529411764704, "grad_norm": 0.4176648802410199, "learning_rate": 0.0002152123002293414, "loss": 0.1187, "step": 5836 }, { "epoch": 2.7468235294117647, "grad_norm": 0.9517567511104448, "learning_rate": 0.0002151728206962287, "loss": 0.4238, "step": 5837 }, { "epoch": 2.747294117647059, "grad_norm": 0.6102091509545075, "learning_rate": 0.00021513333559703738, "loss": 0.3828, "step": 5838 }, { "epoch": 2.7477647058823527, "grad_norm": 1.3053740004243524, "learning_rate": 0.00021509384493513978, "loss": 0.6094, "step": 5839 }, { "epoch": 2.748235294117647, "grad_norm": 0.3241327378815687, "learning_rate": 0.00021505434871390848, "loss": 0.1357, "step": 5840 }, { "epoch": 2.748705882352941, "grad_norm": 0.39479143777754994, "learning_rate": 0.00021501484693671677, "loss": 0.1504, "step": 5841 }, { "epoch": 2.7491764705882353, "grad_norm": 0.7620690772808107, "learning_rate": 0.00021497533960693823, "loss": 0.3125, "step": 5842 }, { "epoch": 2.7496470588235296, "grad_norm": 0.20245628041532376, "learning_rate": 0.00021493582672794693, "loss": 0.0781, "step": 5843 }, { "epoch": 2.7501176470588238, "grad_norm": 0.33575076252720654, "learning_rate": 0.00021489630830311758, "loss": 0.1099, "step": 5844 }, { "epoch": 2.7505882352941176, "grad_norm": 1.3816097944743102, "learning_rate": 0.00021485678433582516, "loss": 0.5742, "step": 5845 }, { "epoch": 2.751058823529412, "grad_norm": 0.7848636716616955, "learning_rate": 0.00021481725482944518, "loss": 0.3809, "step": 5846 }, { "epoch": 2.751529411764706, "grad_norm": 0.3504902018441477, "learning_rate": 0.0002147777197873538, "loss": 0.1089, "step": 5847 }, { "epoch": 2.752, "grad_norm": 1.0187430061299791, "learning_rate": 0.0002147381792129274, "loss": 0.6094, "step": 5848 }, { "epoch": 2.752470588235294, "grad_norm": 1.0170871987326393, "learning_rate": 0.00021469863310954296, "loss": 0.498, "step": 5849 }, { "epoch": 2.7529411764705882, "grad_norm": 0.6154625040739933, "learning_rate": 0.00021465908148057787, "loss": 0.3945, "step": 5850 }, { "epoch": 2.7534117647058824, "grad_norm": 0.7651552721101357, "learning_rate": 0.00021461952432941006, "loss": 0.3574, "step": 5851 }, { "epoch": 2.7538823529411767, "grad_norm": 1.2489662444466194, "learning_rate": 0.00021457996165941798, "loss": 0.875, "step": 5852 }, { "epoch": 2.7543529411764704, "grad_norm": 0.6338350919013422, "learning_rate": 0.00021454039347398033, "loss": 0.459, "step": 5853 }, { "epoch": 2.7548235294117647, "grad_norm": 0.4951966829786431, "learning_rate": 0.00021450081977647658, "loss": 0.1387, "step": 5854 }, { "epoch": 2.755294117647059, "grad_norm": 0.6982256319370556, "learning_rate": 0.00021446124057028642, "loss": 0.4023, "step": 5855 }, { "epoch": 2.7557647058823527, "grad_norm": 0.7394875172646475, "learning_rate": 0.00021442165585879014, "loss": 0.4746, "step": 5856 }, { "epoch": 2.756235294117647, "grad_norm": 0.24298211973329004, "learning_rate": 0.00021438206564536845, "loss": 0.1562, "step": 5857 }, { "epoch": 2.756705882352941, "grad_norm": 0.22811502844763182, "learning_rate": 0.00021434246993340256, "loss": 0.064, "step": 5858 }, { "epoch": 2.7571764705882353, "grad_norm": 0.30048096821007003, "learning_rate": 0.00021430286872627417, "loss": 0.1191, "step": 5859 }, { "epoch": 2.7576470588235296, "grad_norm": 0.6757670455974006, "learning_rate": 0.00021426326202736532, "loss": 0.2812, "step": 5860 }, { "epoch": 2.758117647058824, "grad_norm": 0.42500169746388883, "learning_rate": 0.00021422364984005876, "loss": 0.291, "step": 5861 }, { "epoch": 2.7585882352941176, "grad_norm": 0.20754755056432136, "learning_rate": 0.00021418403216773746, "loss": 0.0601, "step": 5862 }, { "epoch": 2.759058823529412, "grad_norm": 0.5987939020061627, "learning_rate": 0.000214144409013785, "loss": 0.3125, "step": 5863 }, { "epoch": 2.759529411764706, "grad_norm": 0.6567305134352406, "learning_rate": 0.00021410478038158537, "loss": 0.3086, "step": 5864 }, { "epoch": 2.76, "grad_norm": 0.6048229313509995, "learning_rate": 0.00021406514627452306, "loss": 0.2285, "step": 5865 }, { "epoch": 2.760470588235294, "grad_norm": 0.3407429165387324, "learning_rate": 0.00021402550669598303, "loss": 0.0962, "step": 5866 }, { "epoch": 2.7609411764705882, "grad_norm": 0.35631834397508416, "learning_rate": 0.00021398586164935067, "loss": 0.1416, "step": 5867 }, { "epoch": 2.7614117647058825, "grad_norm": 0.5394957710606311, "learning_rate": 0.0002139462111380119, "loss": 0.459, "step": 5868 }, { "epoch": 2.7618823529411767, "grad_norm": 0.5332440463868264, "learning_rate": 0.00021390655516535304, "loss": 0.3398, "step": 5869 }, { "epoch": 2.7623529411764705, "grad_norm": 0.4979249041000976, "learning_rate": 0.00021386689373476087, "loss": 0.2197, "step": 5870 }, { "epoch": 2.7628235294117647, "grad_norm": 0.1652523820978458, "learning_rate": 0.00021382722684962276, "loss": 0.031, "step": 5871 }, { "epoch": 2.763294117647059, "grad_norm": 0.307398686729234, "learning_rate": 0.00021378755451332637, "loss": 0.1099, "step": 5872 }, { "epoch": 2.7637647058823527, "grad_norm": 0.8246597163351636, "learning_rate": 0.00021374787672926, "loss": 0.5312, "step": 5873 }, { "epoch": 2.764235294117647, "grad_norm": 0.5848667139457773, "learning_rate": 0.00021370819350081222, "loss": 0.3789, "step": 5874 }, { "epoch": 2.764705882352941, "grad_norm": 0.34336795856783664, "learning_rate": 0.00021366850483137226, "loss": 0.2227, "step": 5875 }, { "epoch": 2.7651764705882353, "grad_norm": 0.21660812776377966, "learning_rate": 0.00021362881072432965, "loss": 0.0693, "step": 5876 }, { "epoch": 2.7656470588235296, "grad_norm": 1.2692432042629076, "learning_rate": 0.00021358911118307455, "loss": 0.707, "step": 5877 }, { "epoch": 2.766117647058824, "grad_norm": 0.4006756517503151, "learning_rate": 0.00021354940621099745, "loss": 0.2031, "step": 5878 }, { "epoch": 2.7665882352941176, "grad_norm": 0.3580318702884004, "learning_rate": 0.00021350969581148934, "loss": 0.127, "step": 5879 }, { "epoch": 2.767058823529412, "grad_norm": 0.5174471128600763, "learning_rate": 0.00021346997998794172, "loss": 0.2715, "step": 5880 }, { "epoch": 2.767529411764706, "grad_norm": 0.5771557697025739, "learning_rate": 0.00021343025874374645, "loss": 0.3281, "step": 5881 }, { "epoch": 2.768, "grad_norm": 0.5135169770218664, "learning_rate": 0.000213390532082296, "loss": 0.3125, "step": 5882 }, { "epoch": 2.768470588235294, "grad_norm": 0.8205100503010385, "learning_rate": 0.00021335080000698317, "loss": 0.3555, "step": 5883 }, { "epoch": 2.7689411764705882, "grad_norm": 0.4275847528908253, "learning_rate": 0.00021331106252120126, "loss": 0.1113, "step": 5884 }, { "epoch": 2.7694117647058825, "grad_norm": 0.35307414797756914, "learning_rate": 0.00021327131962834407, "loss": 0.1196, "step": 5885 }, { "epoch": 2.7698823529411767, "grad_norm": 0.6566682117668556, "learning_rate": 0.00021323157133180587, "loss": 0.3164, "step": 5886 }, { "epoch": 2.7703529411764705, "grad_norm": 0.49110063944979215, "learning_rate": 0.0002131918176349813, "loss": 0.332, "step": 5887 }, { "epoch": 2.7708235294117647, "grad_norm": 0.4498426983679977, "learning_rate": 0.0002131520585412656, "loss": 0.2295, "step": 5888 }, { "epoch": 2.771294117647059, "grad_norm": 0.07639601562847208, "learning_rate": 0.0002131122940540543, "loss": 0.0123, "step": 5889 }, { "epoch": 2.7717647058823527, "grad_norm": 0.47026758872386015, "learning_rate": 0.00021307252417674356, "loss": 0.2363, "step": 5890 }, { "epoch": 2.772235294117647, "grad_norm": 1.1887567501912404, "learning_rate": 0.0002130327489127298, "loss": 0.7266, "step": 5891 }, { "epoch": 2.772705882352941, "grad_norm": 0.8768455844200989, "learning_rate": 0.0002129929682654102, "loss": 0.4961, "step": 5892 }, { "epoch": 2.7731764705882354, "grad_norm": 0.44651786940308963, "learning_rate": 0.0002129531822381821, "loss": 0.3066, "step": 5893 }, { "epoch": 2.7736470588235296, "grad_norm": 0.3265262967141683, "learning_rate": 0.00021291339083444344, "loss": 0.1426, "step": 5894 }, { "epoch": 2.774117647058824, "grad_norm": 0.42828726642857345, "learning_rate": 0.00021287359405759265, "loss": 0.1748, "step": 5895 }, { "epoch": 2.7745882352941176, "grad_norm": 0.7258147795840901, "learning_rate": 0.0002128337919110285, "loss": 0.2598, "step": 5896 }, { "epoch": 2.775058823529412, "grad_norm": 1.0879779186265168, "learning_rate": 0.00021279398439815034, "loss": 0.3945, "step": 5897 }, { "epoch": 2.775529411764706, "grad_norm": 0.697802010837887, "learning_rate": 0.00021275417152235787, "loss": 0.3145, "step": 5898 }, { "epoch": 2.776, "grad_norm": 0.9378789212562793, "learning_rate": 0.00021271435328705138, "loss": 0.6602, "step": 5899 }, { "epoch": 2.776470588235294, "grad_norm": 0.6987890920615151, "learning_rate": 0.00021267452969563153, "loss": 0.3457, "step": 5900 }, { "epoch": 2.7769411764705882, "grad_norm": 0.8831418065482001, "learning_rate": 0.00021263470075149937, "loss": 0.2637, "step": 5901 }, { "epoch": 2.7774117647058825, "grad_norm": 0.2731980336795791, "learning_rate": 0.00021259486645805662, "loss": 0.124, "step": 5902 }, { "epoch": 2.7778823529411767, "grad_norm": 0.3426750631173772, "learning_rate": 0.00021255502681870518, "loss": 0.1494, "step": 5903 }, { "epoch": 2.7783529411764705, "grad_norm": 0.5767625867655659, "learning_rate": 0.00021251518183684767, "loss": 0.2695, "step": 5904 }, { "epoch": 2.7788235294117647, "grad_norm": 0.8958971993168812, "learning_rate": 0.00021247533151588693, "loss": 0.3984, "step": 5905 }, { "epoch": 2.779294117647059, "grad_norm": 0.33087023064792864, "learning_rate": 0.0002124354758592265, "loss": 0.1021, "step": 5906 }, { "epoch": 2.7797647058823527, "grad_norm": 1.018151391972336, "learning_rate": 0.00021239561487027017, "loss": 0.5469, "step": 5907 }, { "epoch": 2.780235294117647, "grad_norm": 0.544790220192589, "learning_rate": 0.00021235574855242226, "loss": 0.375, "step": 5908 }, { "epoch": 2.780705882352941, "grad_norm": 1.0705281371592603, "learning_rate": 0.00021231587690908758, "loss": 0.6836, "step": 5909 }, { "epoch": 2.7811764705882354, "grad_norm": 0.6320814570560858, "learning_rate": 0.00021227599994367137, "loss": 0.2891, "step": 5910 }, { "epoch": 2.7816470588235296, "grad_norm": 0.544953585621215, "learning_rate": 0.0002122361176595793, "loss": 0.2373, "step": 5911 }, { "epoch": 2.782117647058824, "grad_norm": 0.4639285195435328, "learning_rate": 0.0002121962300602175, "loss": 0.1504, "step": 5912 }, { "epoch": 2.7825882352941176, "grad_norm": 0.8046004832350726, "learning_rate": 0.00021215633714899264, "loss": 0.5273, "step": 5913 }, { "epoch": 2.783058823529412, "grad_norm": 0.8860051559117776, "learning_rate": 0.00021211643892931167, "loss": 0.5703, "step": 5914 }, { "epoch": 2.783529411764706, "grad_norm": 0.9401122226183337, "learning_rate": 0.00021207653540458209, "loss": 0.5625, "step": 5915 }, { "epoch": 2.784, "grad_norm": 0.3099402924668749, "learning_rate": 0.00021203662657821198, "loss": 0.1953, "step": 5916 }, { "epoch": 2.784470588235294, "grad_norm": 0.9986765044607971, "learning_rate": 0.00021199671245360964, "loss": 0.6562, "step": 5917 }, { "epoch": 2.7849411764705883, "grad_norm": 0.39682066382348163, "learning_rate": 0.0002119567930341839, "loss": 0.1216, "step": 5918 }, { "epoch": 2.7854117647058825, "grad_norm": 0.4988183302102686, "learning_rate": 0.00021191686832334423, "loss": 0.2256, "step": 5919 }, { "epoch": 2.7858823529411767, "grad_norm": 1.0815354051552735, "learning_rate": 0.0002118769383245003, "loss": 0.4746, "step": 5920 }, { "epoch": 2.7863529411764705, "grad_norm": 0.4804896431731503, "learning_rate": 0.0002118370030410623, "loss": 0.3887, "step": 5921 }, { "epoch": 2.7868235294117647, "grad_norm": 0.4157431118487654, "learning_rate": 0.0002117970624764409, "loss": 0.2129, "step": 5922 }, { "epoch": 2.787294117647059, "grad_norm": 0.20332666893446208, "learning_rate": 0.00021175711663404728, "loss": 0.0811, "step": 5923 }, { "epoch": 2.7877647058823527, "grad_norm": 0.950591156168723, "learning_rate": 0.000211717165517293, "loss": 0.5664, "step": 5924 }, { "epoch": 2.788235294117647, "grad_norm": 0.6092640052880577, "learning_rate": 0.00021167720912959004, "loss": 0.3887, "step": 5925 }, { "epoch": 2.788705882352941, "grad_norm": 0.5125197113433573, "learning_rate": 0.00021163724747435091, "loss": 0.3047, "step": 5926 }, { "epoch": 2.7891764705882354, "grad_norm": 0.18009319208231153, "learning_rate": 0.00021159728055498856, "loss": 0.0508, "step": 5927 }, { "epoch": 2.7896470588235296, "grad_norm": 0.6254967092941872, "learning_rate": 0.00021155730837491625, "loss": 0.373, "step": 5928 }, { "epoch": 2.7901176470588234, "grad_norm": 0.14620136314305565, "learning_rate": 0.0002115173309375479, "loss": 0.0491, "step": 5929 }, { "epoch": 2.7905882352941176, "grad_norm": 0.7151411405642288, "learning_rate": 0.00021147734824629772, "loss": 0.3223, "step": 5930 }, { "epoch": 2.791058823529412, "grad_norm": 0.5064469104732341, "learning_rate": 0.0002114373603045805, "loss": 0.252, "step": 5931 }, { "epoch": 2.791529411764706, "grad_norm": 14.99666170992909, "learning_rate": 0.00021139736711581132, "loss": 0.2598, "step": 5932 }, { "epoch": 2.792, "grad_norm": 0.6713391503853732, "learning_rate": 0.00021135736868340587, "loss": 0.3789, "step": 5933 }, { "epoch": 2.792470588235294, "grad_norm": 0.7265005506191626, "learning_rate": 0.00021131736501078018, "loss": 0.7031, "step": 5934 }, { "epoch": 2.7929411764705883, "grad_norm": 0.6161094174159844, "learning_rate": 0.0002112773561013507, "loss": 0.3594, "step": 5935 }, { "epoch": 2.7934117647058825, "grad_norm": 0.44616732582665586, "learning_rate": 0.0002112373419585345, "loss": 0.2773, "step": 5936 }, { "epoch": 2.7938823529411767, "grad_norm": 2.5187287851324704, "learning_rate": 0.00021119732258574893, "loss": 0.293, "step": 5937 }, { "epoch": 2.7943529411764705, "grad_norm": 0.9271095902054705, "learning_rate": 0.0002111572979864118, "loss": 0.3672, "step": 5938 }, { "epoch": 2.7948235294117647, "grad_norm": 1.1482256938589037, "learning_rate": 0.00021111726816394147, "loss": 0.3789, "step": 5939 }, { "epoch": 2.795294117647059, "grad_norm": 1.5774141428515562, "learning_rate": 0.00021107723312175665, "loss": 0.1006, "step": 5940 }, { "epoch": 2.7957647058823527, "grad_norm": 0.16501600713470202, "learning_rate": 0.00021103719286327658, "loss": 0.0398, "step": 5941 }, { "epoch": 2.796235294117647, "grad_norm": 4.570576915441877, "learning_rate": 0.0002109971473919208, "loss": 0.3516, "step": 5942 }, { "epoch": 2.796705882352941, "grad_norm": 2.684329345690095, "learning_rate": 0.00021095709671110942, "loss": 0.2734, "step": 5943 }, { "epoch": 2.7971764705882354, "grad_norm": 0.5527479028875477, "learning_rate": 0.00021091704082426308, "loss": 0.1504, "step": 5944 }, { "epoch": 2.7976470588235296, "grad_norm": 0.6348407635316596, "learning_rate": 0.0002108769797348026, "loss": 0.4121, "step": 5945 }, { "epoch": 2.7981176470588234, "grad_norm": 0.4071934949973202, "learning_rate": 0.00021083691344614947, "loss": 0.0933, "step": 5946 }, { "epoch": 2.7985882352941176, "grad_norm": 0.7304081803476722, "learning_rate": 0.00021079684196172554, "loss": 0.3691, "step": 5947 }, { "epoch": 2.799058823529412, "grad_norm": 0.5815794478423291, "learning_rate": 0.00021075676528495307, "loss": 0.3008, "step": 5948 }, { "epoch": 2.799529411764706, "grad_norm": 0.37709621886982597, "learning_rate": 0.00021071668341925482, "loss": 0.1504, "step": 5949 }, { "epoch": 2.8, "grad_norm": 0.14485964872346427, "learning_rate": 0.00021067659636805403, "loss": 0.026, "step": 5950 }, { "epoch": 2.800470588235294, "grad_norm": 0.251320635699071, "learning_rate": 0.00021063650413477428, "loss": 0.1084, "step": 5951 }, { "epoch": 2.8009411764705883, "grad_norm": 0.708807234104542, "learning_rate": 0.00021059640672283964, "loss": 0.2676, "step": 5952 }, { "epoch": 2.8014117647058825, "grad_norm": 0.49334416218862837, "learning_rate": 0.0002105563041356747, "loss": 0.3223, "step": 5953 }, { "epoch": 2.8018823529411767, "grad_norm": 0.4752786427467906, "learning_rate": 0.00021051619637670431, "loss": 0.3164, "step": 5954 }, { "epoch": 2.8023529411764705, "grad_norm": 0.3043273438675844, "learning_rate": 0.00021047608344935395, "loss": 0.1338, "step": 5955 }, { "epoch": 2.8028235294117647, "grad_norm": 0.4115563004068518, "learning_rate": 0.0002104359653570494, "loss": 0.2812, "step": 5956 }, { "epoch": 2.803294117647059, "grad_norm": 0.7573584472384687, "learning_rate": 0.00021039584210321699, "loss": 0.5781, "step": 5957 }, { "epoch": 2.8037647058823527, "grad_norm": 0.7236864911312088, "learning_rate": 0.00021035571369128341, "loss": 0.3672, "step": 5958 }, { "epoch": 2.804235294117647, "grad_norm": 1.1046687256251262, "learning_rate": 0.00021031558012467582, "loss": 0.6992, "step": 5959 }, { "epoch": 2.804705882352941, "grad_norm": 0.6549518715093314, "learning_rate": 0.00021027544140682188, "loss": 0.3223, "step": 5960 }, { "epoch": 2.8051764705882354, "grad_norm": 0.6616401748991358, "learning_rate": 0.00021023529754114956, "loss": 0.2793, "step": 5961 }, { "epoch": 2.8056470588235296, "grad_norm": 0.3469227780711986, "learning_rate": 0.0002101951485310874, "loss": 0.123, "step": 5962 }, { "epoch": 2.8061176470588234, "grad_norm": 0.7726634136329202, "learning_rate": 0.0002101549943800643, "loss": 0.3301, "step": 5963 }, { "epoch": 2.8065882352941176, "grad_norm": 1.0054450022804087, "learning_rate": 0.00021011483509150963, "loss": 0.4473, "step": 5964 }, { "epoch": 2.807058823529412, "grad_norm": 0.7184554115744978, "learning_rate": 0.00021007467066885314, "loss": 0.5, "step": 5965 }, { "epoch": 2.8075294117647056, "grad_norm": 0.47103275032596575, "learning_rate": 0.00021003450111552514, "loss": 0.2852, "step": 5966 }, { "epoch": 2.808, "grad_norm": 0.33677918323680983, "learning_rate": 0.00020999432643495626, "loss": 0.1328, "step": 5967 }, { "epoch": 2.808470588235294, "grad_norm": 0.3278058706378958, "learning_rate": 0.00020995414663057768, "loss": 0.1216, "step": 5968 }, { "epoch": 2.8089411764705883, "grad_norm": 1.2691838385973784, "learning_rate": 0.00020991396170582088, "loss": 0.4727, "step": 5969 }, { "epoch": 2.8094117647058825, "grad_norm": 0.7033713846690505, "learning_rate": 0.00020987377166411785, "loss": 0.4395, "step": 5970 }, { "epoch": 2.8098823529411767, "grad_norm": 0.7620711717860276, "learning_rate": 0.00020983357650890106, "loss": 0.3867, "step": 5971 }, { "epoch": 2.8103529411764705, "grad_norm": 0.9024228795401736, "learning_rate": 0.00020979337624360336, "loss": 0.5547, "step": 5972 }, { "epoch": 2.8108235294117647, "grad_norm": 0.6304140724328223, "learning_rate": 0.000209753170871658, "loss": 0.3711, "step": 5973 }, { "epoch": 2.811294117647059, "grad_norm": 0.1585866929769258, "learning_rate": 0.00020971296039649882, "loss": 0.0554, "step": 5974 }, { "epoch": 2.8117647058823527, "grad_norm": 0.49327390131266224, "learning_rate": 0.00020967274482155992, "loss": 0.291, "step": 5975 }, { "epoch": 2.812235294117647, "grad_norm": 0.4574652987293924, "learning_rate": 0.00020963252415027588, "loss": 0.1641, "step": 5976 }, { "epoch": 2.812705882352941, "grad_norm": 0.8823649516807085, "learning_rate": 0.0002095922983860818, "loss": 0.3613, "step": 5977 }, { "epoch": 2.8131764705882354, "grad_norm": 0.5975316323851587, "learning_rate": 0.00020955206753241315, "loss": 0.2471, "step": 5978 }, { "epoch": 2.8136470588235296, "grad_norm": 0.6646673884465418, "learning_rate": 0.00020951183159270586, "loss": 0.3105, "step": 5979 }, { "epoch": 2.8141176470588234, "grad_norm": 0.6286629237393629, "learning_rate": 0.0002094715905703962, "loss": 0.4453, "step": 5980 }, { "epoch": 2.8145882352941176, "grad_norm": 0.49159440952570016, "learning_rate": 0.00020943134446892102, "loss": 0.2734, "step": 5981 }, { "epoch": 2.815058823529412, "grad_norm": 0.5315169053796337, "learning_rate": 0.00020939109329171752, "loss": 0.2676, "step": 5982 }, { "epoch": 2.8155294117647056, "grad_norm": 0.30872696269470756, "learning_rate": 0.00020935083704222331, "loss": 0.126, "step": 5983 }, { "epoch": 2.816, "grad_norm": 0.45964555028009063, "learning_rate": 0.00020931057572387655, "loss": 0.125, "step": 5984 }, { "epoch": 2.816470588235294, "grad_norm": 0.7311503733257041, "learning_rate": 0.00020927030934011568, "loss": 0.2734, "step": 5985 }, { "epoch": 2.8169411764705883, "grad_norm": 0.20797549178633992, "learning_rate": 0.00020923003789437967, "loss": 0.0693, "step": 5986 }, { "epoch": 2.8174117647058825, "grad_norm": 1.1379436924277413, "learning_rate": 0.00020918976139010784, "loss": 0.4844, "step": 5987 }, { "epoch": 2.8178823529411767, "grad_norm": 0.5043996662617862, "learning_rate": 0.00020914947983074016, "loss": 0.3496, "step": 5988 }, { "epoch": 2.8183529411764705, "grad_norm": 0.9306656646110929, "learning_rate": 0.00020910919321971675, "loss": 0.6055, "step": 5989 }, { "epoch": 2.8188235294117647, "grad_norm": 0.4947854266705662, "learning_rate": 0.00020906890156047826, "loss": 0.2695, "step": 5990 }, { "epoch": 2.819294117647059, "grad_norm": 0.533859801470082, "learning_rate": 0.0002090286048564659, "loss": 0.2559, "step": 5991 }, { "epoch": 2.8197647058823527, "grad_norm": 0.37393146145550227, "learning_rate": 0.00020898830311112118, "loss": 0.2109, "step": 5992 }, { "epoch": 2.820235294117647, "grad_norm": 0.5820491989529576, "learning_rate": 0.00020894799632788596, "loss": 0.25, "step": 5993 }, { "epoch": 2.820705882352941, "grad_norm": 0.6953657607125616, "learning_rate": 0.0002089076845102028, "loss": 0.4941, "step": 5994 }, { "epoch": 2.8211764705882354, "grad_norm": 0.45935397079479057, "learning_rate": 0.00020886736766151438, "loss": 0.2119, "step": 5995 }, { "epoch": 2.8216470588235296, "grad_norm": 0.6596781184672418, "learning_rate": 0.00020882704578526406, "loss": 0.3184, "step": 5996 }, { "epoch": 2.8221176470588234, "grad_norm": 1.4263189864531955, "learning_rate": 0.0002087867188848954, "loss": 0.5352, "step": 5997 }, { "epoch": 2.8225882352941176, "grad_norm": 0.6657447149701318, "learning_rate": 0.0002087463869638527, "loss": 0.3438, "step": 5998 }, { "epoch": 2.823058823529412, "grad_norm": 0.3311725793181296, "learning_rate": 0.00020870605002558038, "loss": 0.1226, "step": 5999 }, { "epoch": 2.8235294117647056, "grad_norm": 0.571075371673141, "learning_rate": 0.00020866570807352337, "loss": 0.3594, "step": 6000 }, { "epoch": 2.824, "grad_norm": 0.501795851924751, "learning_rate": 0.0002086253611111272, "loss": 0.375, "step": 6001 }, { "epoch": 2.824470588235294, "grad_norm": 0.3608864620427188, "learning_rate": 0.00020858500914183763, "loss": 0.1436, "step": 6002 }, { "epoch": 2.8249411764705883, "grad_norm": 0.4513501995928211, "learning_rate": 0.0002085446521691009, "loss": 0.1816, "step": 6003 }, { "epoch": 2.8254117647058825, "grad_norm": 0.5558178718769067, "learning_rate": 0.00020850429019636372, "loss": 0.2891, "step": 6004 }, { "epoch": 2.8258823529411767, "grad_norm": 0.3486892255835896, "learning_rate": 0.00020846392322707323, "loss": 0.1367, "step": 6005 }, { "epoch": 2.8263529411764705, "grad_norm": 0.3933512663148448, "learning_rate": 0.00020842355126467693, "loss": 0.1924, "step": 6006 }, { "epoch": 2.8268235294117647, "grad_norm": 0.4774002599800635, "learning_rate": 0.00020838317431262274, "loss": 0.2354, "step": 6007 }, { "epoch": 2.827294117647059, "grad_norm": 0.7616320453125097, "learning_rate": 0.0002083427923743591, "loss": 0.6562, "step": 6008 }, { "epoch": 2.8277647058823527, "grad_norm": 0.42013463699615783, "learning_rate": 0.00020830240545333489, "loss": 0.2031, "step": 6009 }, { "epoch": 2.828235294117647, "grad_norm": 0.5514852109216106, "learning_rate": 0.00020826201355299918, "loss": 0.2412, "step": 6010 }, { "epoch": 2.828705882352941, "grad_norm": 0.49080136309471006, "learning_rate": 0.0002082216166768018, "loss": 0.1982, "step": 6011 }, { "epoch": 2.8291764705882354, "grad_norm": 0.6810326314406893, "learning_rate": 0.00020818121482819277, "loss": 0.291, "step": 6012 }, { "epoch": 2.8296470588235296, "grad_norm": 0.33552976593826583, "learning_rate": 0.00020814080801062263, "loss": 0.1143, "step": 6013 }, { "epoch": 2.8301176470588234, "grad_norm": 0.6324982572201376, "learning_rate": 0.00020810039622754228, "loss": 0.418, "step": 6014 }, { "epoch": 2.8305882352941176, "grad_norm": 0.46919280082974585, "learning_rate": 0.00020805997948240314, "loss": 0.2236, "step": 6015 }, { "epoch": 2.831058823529412, "grad_norm": 0.5502382411991118, "learning_rate": 0.00020801955777865694, "loss": 0.3184, "step": 6016 }, { "epoch": 2.8315294117647056, "grad_norm": 1.113880534037828, "learning_rate": 0.00020797913111975593, "loss": 0.6719, "step": 6017 }, { "epoch": 2.832, "grad_norm": 0.4135086546166559, "learning_rate": 0.00020793869950915276, "loss": 0.1602, "step": 6018 }, { "epoch": 2.832470588235294, "grad_norm": 0.5088114707012835, "learning_rate": 0.00020789826295030046, "loss": 0.2988, "step": 6019 }, { "epoch": 2.8329411764705883, "grad_norm": 0.3237767387091123, "learning_rate": 0.0002078578214466525, "loss": 0.1934, "step": 6020 }, { "epoch": 2.8334117647058825, "grad_norm": 0.3579367714539943, "learning_rate": 0.00020781737500166279, "loss": 0.1924, "step": 6021 }, { "epoch": 2.8338823529411767, "grad_norm": 0.3733328487805513, "learning_rate": 0.00020777692361878569, "loss": 0.1367, "step": 6022 }, { "epoch": 2.8343529411764705, "grad_norm": 0.5532321844030693, "learning_rate": 0.0002077364673014759, "loss": 0.3516, "step": 6023 }, { "epoch": 2.8348235294117647, "grad_norm": 0.7444549891540458, "learning_rate": 0.00020769600605318857, "loss": 0.5859, "step": 6024 }, { "epoch": 2.835294117647059, "grad_norm": 0.42518437928035535, "learning_rate": 0.0002076555398773794, "loss": 0.1797, "step": 6025 }, { "epoch": 2.8357647058823527, "grad_norm": 0.3671164433392334, "learning_rate": 0.00020761506877750428, "loss": 0.1465, "step": 6026 }, { "epoch": 2.836235294117647, "grad_norm": 0.40907838093471227, "learning_rate": 0.00020757459275701966, "loss": 0.2119, "step": 6027 }, { "epoch": 2.836705882352941, "grad_norm": 0.4477339643158027, "learning_rate": 0.0002075341118193825, "loss": 0.1182, "step": 6028 }, { "epoch": 2.8371764705882354, "grad_norm": 0.6292675589513401, "learning_rate": 0.00020749362596804994, "loss": 0.4844, "step": 6029 }, { "epoch": 2.8376470588235296, "grad_norm": 0.46825416064454745, "learning_rate": 0.00020745313520647971, "loss": 0.1914, "step": 6030 }, { "epoch": 2.8381176470588234, "grad_norm": 0.3707207549473386, "learning_rate": 0.00020741263953812992, "loss": 0.1406, "step": 6031 }, { "epoch": 2.8385882352941176, "grad_norm": 0.47647994953604556, "learning_rate": 0.0002073721389664591, "loss": 0.3184, "step": 6032 }, { "epoch": 2.839058823529412, "grad_norm": 0.7983461725820109, "learning_rate": 0.00020733163349492625, "loss": 0.5664, "step": 6033 }, { "epoch": 2.8395294117647056, "grad_norm": 0.373816841056974, "learning_rate": 0.00020729112312699066, "loss": 0.3535, "step": 6034 }, { "epoch": 2.84, "grad_norm": 0.7316675029004335, "learning_rate": 0.00020725060786611212, "loss": 0.4961, "step": 6035 }, { "epoch": 2.840470588235294, "grad_norm": 0.7111027235650856, "learning_rate": 0.0002072100877157509, "loss": 0.5547, "step": 6036 }, { "epoch": 2.8409411764705883, "grad_norm": 0.6076062173511734, "learning_rate": 0.00020716956267936759, "loss": 0.3105, "step": 6037 }, { "epoch": 2.8414117647058825, "grad_norm": 0.5876868976233681, "learning_rate": 0.0002071290327604231, "loss": 0.4102, "step": 6038 }, { "epoch": 2.8418823529411767, "grad_norm": 0.1674719620897819, "learning_rate": 0.00020708849796237914, "loss": 0.0845, "step": 6039 }, { "epoch": 2.8423529411764705, "grad_norm": 0.23179581853418285, "learning_rate": 0.00020704795828869736, "loss": 0.0752, "step": 6040 }, { "epoch": 2.8428235294117647, "grad_norm": 0.8127628702261183, "learning_rate": 0.00020700741374284014, "loss": 0.375, "step": 6041 }, { "epoch": 2.843294117647059, "grad_norm": 0.6152069722871729, "learning_rate": 0.00020696686432827017, "loss": 0.2266, "step": 6042 }, { "epoch": 2.8437647058823528, "grad_norm": 0.30060253340416293, "learning_rate": 0.00020692631004845055, "loss": 0.0811, "step": 6043 }, { "epoch": 2.844235294117647, "grad_norm": 0.40031734937589886, "learning_rate": 0.00020688575090684484, "loss": 0.1094, "step": 6044 }, { "epoch": 2.844705882352941, "grad_norm": 0.5179325652533137, "learning_rate": 0.00020684518690691696, "loss": 0.2852, "step": 6045 }, { "epoch": 2.8451764705882354, "grad_norm": 0.5809572263848658, "learning_rate": 0.00020680461805213133, "loss": 0.2793, "step": 6046 }, { "epoch": 2.8456470588235296, "grad_norm": 0.5629921656125699, "learning_rate": 0.00020676404434595265, "loss": 0.3535, "step": 6047 }, { "epoch": 2.8461176470588234, "grad_norm": 0.7604620061303897, "learning_rate": 0.00020672346579184617, "loss": 0.3164, "step": 6048 }, { "epoch": 2.8465882352941176, "grad_norm": 0.5304431334409291, "learning_rate": 0.00020668288239327753, "loss": 0.2773, "step": 6049 }, { "epoch": 2.847058823529412, "grad_norm": 1.08113678027774, "learning_rate": 0.00020664229415371266, "loss": 0.625, "step": 6050 }, { "epoch": 2.8475294117647056, "grad_norm": 0.5407818453514592, "learning_rate": 0.00020660170107661803, "loss": 0.2695, "step": 6051 }, { "epoch": 2.848, "grad_norm": 0.7210844633193844, "learning_rate": 0.00020656110316546054, "loss": 0.5547, "step": 6052 }, { "epoch": 2.848470588235294, "grad_norm": 0.24962831906815647, "learning_rate": 0.00020652050042370743, "loss": 0.0854, "step": 6053 }, { "epoch": 2.8489411764705883, "grad_norm": 0.3169850111977464, "learning_rate": 0.0002064798928548263, "loss": 0.0874, "step": 6054 }, { "epoch": 2.8494117647058825, "grad_norm": 0.9436311853939425, "learning_rate": 0.00020643928046228525, "loss": 0.3262, "step": 6055 }, { "epoch": 2.8498823529411763, "grad_norm": 0.9197344478559081, "learning_rate": 0.00020639866324955285, "loss": 0.2188, "step": 6056 }, { "epoch": 2.8503529411764705, "grad_norm": 0.5108874523343165, "learning_rate": 0.00020635804122009802, "loss": 0.2598, "step": 6057 }, { "epoch": 2.8508235294117648, "grad_norm": 0.3907546741146878, "learning_rate": 0.00020631741437738998, "loss": 0.207, "step": 6058 }, { "epoch": 2.851294117647059, "grad_norm": 0.742061302449235, "learning_rate": 0.00020627678272489854, "loss": 0.4395, "step": 6059 }, { "epoch": 2.8517647058823528, "grad_norm": 0.5383037986105963, "learning_rate": 0.00020623614626609385, "loss": 0.2969, "step": 6060 }, { "epoch": 2.852235294117647, "grad_norm": 0.07093966838226524, "learning_rate": 0.0002061955050044464, "loss": 0.0114, "step": 6061 }, { "epoch": 2.852705882352941, "grad_norm": 0.33495469513448284, "learning_rate": 0.00020615485894342716, "loss": 0.1172, "step": 6062 }, { "epoch": 2.8531764705882354, "grad_norm": 0.45973121852072424, "learning_rate": 0.00020611420808650758, "loss": 0.1406, "step": 6063 }, { "epoch": 2.8536470588235296, "grad_norm": 1.2698242724125728, "learning_rate": 0.00020607355243715936, "loss": 0.5156, "step": 6064 }, { "epoch": 2.8541176470588234, "grad_norm": 0.7616392415376196, "learning_rate": 0.00020603289199885473, "loss": 0.3594, "step": 6065 }, { "epoch": 2.8545882352941176, "grad_norm": 0.5316607103527404, "learning_rate": 0.00020599222677506628, "loss": 0.2227, "step": 6066 }, { "epoch": 2.855058823529412, "grad_norm": 0.4487875033874322, "learning_rate": 0.00020595155676926704, "loss": 0.3301, "step": 6067 }, { "epoch": 2.8555294117647056, "grad_norm": 0.2381946995426419, "learning_rate": 0.0002059108819849304, "loss": 0.0562, "step": 6068 }, { "epoch": 2.856, "grad_norm": 0.6914662418790918, "learning_rate": 0.0002058702024255302, "loss": 0.25, "step": 6069 }, { "epoch": 2.856470588235294, "grad_norm": 0.9153406089482861, "learning_rate": 0.00020582951809454067, "loss": 0.4648, "step": 6070 }, { "epoch": 2.8569411764705883, "grad_norm": 0.37750521840747286, "learning_rate": 0.00020578882899543644, "loss": 0.1719, "step": 6071 }, { "epoch": 2.8574117647058825, "grad_norm": 0.7588315986728014, "learning_rate": 0.0002057481351316926, "loss": 0.3613, "step": 6072 }, { "epoch": 2.8578823529411763, "grad_norm": 0.5303046716186106, "learning_rate": 0.00020570743650678454, "loss": 0.2715, "step": 6073 }, { "epoch": 2.8583529411764705, "grad_norm": 0.35020108923366017, "learning_rate": 0.00020566673312418817, "loss": 0.1328, "step": 6074 }, { "epoch": 2.8588235294117648, "grad_norm": 0.9560704974986298, "learning_rate": 0.00020562602498737975, "loss": 0.5586, "step": 6075 }, { "epoch": 2.859294117647059, "grad_norm": 0.6942883920367368, "learning_rate": 0.00020558531209983593, "loss": 0.3145, "step": 6076 }, { "epoch": 2.8597647058823528, "grad_norm": 0.7333887465190565, "learning_rate": 0.00020554459446503386, "loss": 0.3418, "step": 6077 }, { "epoch": 2.860235294117647, "grad_norm": 0.25260019313014015, "learning_rate": 0.00020550387208645093, "loss": 0.1069, "step": 6078 }, { "epoch": 2.860705882352941, "grad_norm": 0.38938616967319023, "learning_rate": 0.0002054631449675651, "loss": 0.1729, "step": 6079 }, { "epoch": 2.8611764705882354, "grad_norm": 0.42510630823867146, "learning_rate": 0.00020542241311185467, "loss": 0.2119, "step": 6080 }, { "epoch": 2.8616470588235297, "grad_norm": 0.6783570274393126, "learning_rate": 0.0002053816765227983, "loss": 0.2773, "step": 6081 }, { "epoch": 2.8621176470588234, "grad_norm": 0.8517878611713666, "learning_rate": 0.0002053409352038751, "loss": 0.4922, "step": 6082 }, { "epoch": 2.8625882352941177, "grad_norm": 1.002931428580105, "learning_rate": 0.00020530018915856457, "loss": 0.3613, "step": 6083 }, { "epoch": 2.863058823529412, "grad_norm": 0.33637103996886764, "learning_rate": 0.00020525943839034668, "loss": 0.0884, "step": 6084 }, { "epoch": 2.8635294117647057, "grad_norm": 0.9228546326758682, "learning_rate": 0.0002052186829027017, "loss": 0.4414, "step": 6085 }, { "epoch": 2.864, "grad_norm": 0.4256289361172423, "learning_rate": 0.00020517792269911033, "loss": 0.1553, "step": 6086 }, { "epoch": 2.864470588235294, "grad_norm": 0.5762721558786431, "learning_rate": 0.0002051371577830537, "loss": 0.4824, "step": 6087 }, { "epoch": 2.8649411764705883, "grad_norm": 0.4368684848928292, "learning_rate": 0.00020509638815801342, "loss": 0.3457, "step": 6088 }, { "epoch": 2.8654117647058825, "grad_norm": 0.9692770138316953, "learning_rate": 0.00020505561382747127, "loss": 0.668, "step": 6089 }, { "epoch": 2.8658823529411763, "grad_norm": 0.34595349001649756, "learning_rate": 0.00020501483479490968, "loss": 0.0781, "step": 6090 }, { "epoch": 2.8663529411764705, "grad_norm": 1.2452502170051607, "learning_rate": 0.00020497405106381138, "loss": 0.5664, "step": 6091 }, { "epoch": 2.8668235294117648, "grad_norm": 0.3124660183324574, "learning_rate": 0.0002049332626376594, "loss": 0.1094, "step": 6092 }, { "epoch": 2.8672941176470585, "grad_norm": 0.36241966492367766, "learning_rate": 0.00020489246951993742, "loss": 0.105, "step": 6093 }, { "epoch": 2.8677647058823528, "grad_norm": 0.7400859028828258, "learning_rate": 0.00020485167171412926, "loss": 0.4336, "step": 6094 }, { "epoch": 2.868235294117647, "grad_norm": 0.1222673487186654, "learning_rate": 0.00020481086922371935, "loss": 0.0172, "step": 6095 }, { "epoch": 2.868705882352941, "grad_norm": 0.7511992053175859, "learning_rate": 0.00020477006205219228, "loss": 0.3223, "step": 6096 }, { "epoch": 2.8691764705882354, "grad_norm": 0.3213379457156073, "learning_rate": 0.0002047292502030333, "loss": 0.103, "step": 6097 }, { "epoch": 2.8696470588235297, "grad_norm": 1.2075156383687806, "learning_rate": 0.00020468843367972786, "loss": 0.4531, "step": 6098 }, { "epoch": 2.8701176470588234, "grad_norm": 0.6464436889442068, "learning_rate": 0.00020464761248576203, "loss": 0.3301, "step": 6099 }, { "epoch": 2.8705882352941177, "grad_norm": 0.43947856184344164, "learning_rate": 0.00020460678662462194, "loss": 0.2852, "step": 6100 }, { "epoch": 2.871058823529412, "grad_norm": 0.6963569468493432, "learning_rate": 0.0002045659560997945, "loss": 0.3105, "step": 6101 }, { "epoch": 2.8715294117647057, "grad_norm": 1.0498743607728462, "learning_rate": 0.00020452512091476678, "loss": 0.4004, "step": 6102 }, { "epoch": 2.872, "grad_norm": 0.8394888040778876, "learning_rate": 0.0002044842810730262, "loss": 0.4727, "step": 6103 }, { "epoch": 2.872470588235294, "grad_norm": 0.6773459390059396, "learning_rate": 0.00020444343657806087, "loss": 0.3828, "step": 6104 }, { "epoch": 2.8729411764705883, "grad_norm": 0.6277113118169757, "learning_rate": 0.000204402587433359, "loss": 0.3301, "step": 6105 }, { "epoch": 2.8734117647058826, "grad_norm": 0.49700012374871183, "learning_rate": 0.0002043617336424093, "loss": 0.2598, "step": 6106 }, { "epoch": 2.8738823529411763, "grad_norm": 0.5875221585277721, "learning_rate": 0.00020432087520870086, "loss": 0.1426, "step": 6107 }, { "epoch": 2.8743529411764706, "grad_norm": 0.885929188016757, "learning_rate": 0.0002042800121357233, "loss": 0.3145, "step": 6108 }, { "epoch": 2.8748235294117648, "grad_norm": 0.41925901469777616, "learning_rate": 0.00020423914442696642, "loss": 0.0742, "step": 6109 }, { "epoch": 2.8752941176470586, "grad_norm": 0.7403693032394938, "learning_rate": 0.00020419827208592058, "loss": 0.5508, "step": 6110 }, { "epoch": 2.875764705882353, "grad_norm": 0.7669509873360305, "learning_rate": 0.00020415739511607653, "loss": 0.3164, "step": 6111 }, { "epoch": 2.876235294117647, "grad_norm": 0.9860034153648598, "learning_rate": 0.0002041165135209252, "loss": 0.4824, "step": 6112 }, { "epoch": 2.8767058823529412, "grad_norm": 0.6244349607836001, "learning_rate": 0.0002040756273039582, "loss": 0.3242, "step": 6113 }, { "epoch": 2.8771764705882354, "grad_norm": 0.24715242005847873, "learning_rate": 0.0002040347364686674, "loss": 0.103, "step": 6114 }, { "epoch": 2.8776470588235297, "grad_norm": 0.8186822635587649, "learning_rate": 0.00020399384101854506, "loss": 0.373, "step": 6115 }, { "epoch": 2.8781176470588234, "grad_norm": 0.9911767540293148, "learning_rate": 0.00020395294095708387, "loss": 0.6602, "step": 6116 }, { "epoch": 2.8785882352941177, "grad_norm": 0.4244738602292612, "learning_rate": 0.00020391203628777687, "loss": 0.1631, "step": 6117 }, { "epoch": 2.879058823529412, "grad_norm": 0.4908928060208077, "learning_rate": 0.00020387112701411754, "loss": 0.1719, "step": 6118 }, { "epoch": 2.8795294117647057, "grad_norm": 0.298208749910716, "learning_rate": 0.00020383021313959972, "loss": 0.1387, "step": 6119 }, { "epoch": 2.88, "grad_norm": 0.2131803346601466, "learning_rate": 0.00020378929466771767, "loss": 0.1104, "step": 6120 }, { "epoch": 2.880470588235294, "grad_norm": 0.5268053995335494, "learning_rate": 0.000203748371601966, "loss": 0.2656, "step": 6121 }, { "epoch": 2.8809411764705883, "grad_norm": 0.2309764382909473, "learning_rate": 0.0002037074439458398, "loss": 0.0649, "step": 6122 }, { "epoch": 2.8814117647058826, "grad_norm": 1.3057644572772908, "learning_rate": 0.0002036665117028344, "loss": 0.334, "step": 6123 }, { "epoch": 2.8818823529411763, "grad_norm": 0.34831208078841197, "learning_rate": 0.0002036255748764457, "loss": 0.0898, "step": 6124 }, { "epoch": 2.8823529411764706, "grad_norm": 0.43448134331324384, "learning_rate": 0.00020358463347016988, "loss": 0.1689, "step": 6125 }, { "epoch": 2.882823529411765, "grad_norm": 0.6021185269849992, "learning_rate": 0.00020354368748750357, "loss": 0.1177, "step": 6126 }, { "epoch": 2.8832941176470586, "grad_norm": 0.8673084429839151, "learning_rate": 0.00020350273693194363, "loss": 0.5078, "step": 6127 }, { "epoch": 2.883764705882353, "grad_norm": 0.30513318861367206, "learning_rate": 0.00020346178180698758, "loss": 0.1113, "step": 6128 }, { "epoch": 2.884235294117647, "grad_norm": 0.580087165443106, "learning_rate": 0.00020342082211613312, "loss": 0.3105, "step": 6129 }, { "epoch": 2.8847058823529412, "grad_norm": 0.7049666163394611, "learning_rate": 0.00020337985786287842, "loss": 0.5352, "step": 6130 }, { "epoch": 2.8851764705882355, "grad_norm": 0.4185635144701722, "learning_rate": 0.00020333888905072208, "loss": 0.1797, "step": 6131 }, { "epoch": 2.8856470588235297, "grad_norm": 0.3994157192338907, "learning_rate": 0.00020329791568316294, "loss": 0.1963, "step": 6132 }, { "epoch": 2.8861176470588235, "grad_norm": 0.7434850897150878, "learning_rate": 0.00020325693776370037, "loss": 0.3379, "step": 6133 }, { "epoch": 2.8865882352941177, "grad_norm": 0.9278356605075562, "learning_rate": 0.0002032159552958341, "loss": 0.4863, "step": 6134 }, { "epoch": 2.887058823529412, "grad_norm": 0.708194618354491, "learning_rate": 0.00020317496828306427, "loss": 0.334, "step": 6135 }, { "epoch": 2.8875294117647057, "grad_norm": 0.5487847984618209, "learning_rate": 0.00020313397672889132, "loss": 0.3008, "step": 6136 }, { "epoch": 2.888, "grad_norm": 0.7741535150769263, "learning_rate": 0.00020309298063681607, "loss": 0.3828, "step": 6137 }, { "epoch": 2.888470588235294, "grad_norm": 0.7767398802986298, "learning_rate": 0.00020305198001033988, "loss": 0.3105, "step": 6138 }, { "epoch": 2.8889411764705883, "grad_norm": 0.5364527794442937, "learning_rate": 0.00020301097485296443, "loss": 0.5469, "step": 6139 }, { "epoch": 2.8894117647058826, "grad_norm": 0.27444304481200654, "learning_rate": 0.00020296996516819163, "loss": 0.1318, "step": 6140 }, { "epoch": 2.8898823529411763, "grad_norm": 0.31540465002042506, "learning_rate": 0.00020292895095952402, "loss": 0.1299, "step": 6141 }, { "epoch": 2.8903529411764706, "grad_norm": 0.35411282970181257, "learning_rate": 0.0002028879322304644, "loss": 0.1641, "step": 6142 }, { "epoch": 2.890823529411765, "grad_norm": 0.7115644140310925, "learning_rate": 0.00020284690898451593, "loss": 0.3574, "step": 6143 }, { "epoch": 2.8912941176470586, "grad_norm": 0.9366383095166059, "learning_rate": 0.00020280588122518223, "loss": 0.5898, "step": 6144 }, { "epoch": 2.891764705882353, "grad_norm": 0.47336570666954414, "learning_rate": 0.00020276484895596726, "loss": 0.3398, "step": 6145 }, { "epoch": 2.892235294117647, "grad_norm": 0.3396469033093128, "learning_rate": 0.00020272381218037542, "loss": 0.1592, "step": 6146 }, { "epoch": 2.8927058823529412, "grad_norm": 0.32689594497226315, "learning_rate": 0.00020268277090191133, "loss": 0.0698, "step": 6147 }, { "epoch": 2.8931764705882355, "grad_norm": 0.471737278796329, "learning_rate": 0.00020264172512408026, "loss": 0.1963, "step": 6148 }, { "epoch": 2.8936470588235297, "grad_norm": 0.857735607087419, "learning_rate": 0.00020260067485038767, "loss": 0.459, "step": 6149 }, { "epoch": 2.8941176470588235, "grad_norm": 0.6397758969739435, "learning_rate": 0.0002025596200843394, "loss": 0.3223, "step": 6150 }, { "epoch": 2.8945882352941177, "grad_norm": 0.9305038083552174, "learning_rate": 0.00020251856082944178, "loss": 0.3555, "step": 6151 }, { "epoch": 2.895058823529412, "grad_norm": 0.5045990415342232, "learning_rate": 0.00020247749708920146, "loss": 0.2676, "step": 6152 }, { "epoch": 2.8955294117647057, "grad_norm": 0.7702146906284236, "learning_rate": 0.00020243642886712553, "loss": 0.2598, "step": 6153 }, { "epoch": 2.896, "grad_norm": 0.5446766305289977, "learning_rate": 0.0002023953561667213, "loss": 0.2852, "step": 6154 }, { "epoch": 2.896470588235294, "grad_norm": 0.43229918217348534, "learning_rate": 0.0002023542789914967, "loss": 0.2617, "step": 6155 }, { "epoch": 2.8969411764705884, "grad_norm": 0.6965903665691267, "learning_rate": 0.0002023131973449599, "loss": 0.4766, "step": 6156 }, { "epoch": 2.8974117647058826, "grad_norm": 0.19374696986064607, "learning_rate": 0.00020227211123061942, "loss": 0.0552, "step": 6157 }, { "epoch": 2.8978823529411764, "grad_norm": 0.31299300612353215, "learning_rate": 0.00020223102065198426, "loss": 0.1445, "step": 6158 }, { "epoch": 2.8983529411764706, "grad_norm": 1.2394217254621387, "learning_rate": 0.00020218992561256375, "loss": 0.6094, "step": 6159 }, { "epoch": 2.898823529411765, "grad_norm": 0.48744022380097096, "learning_rate": 0.0002021488261158676, "loss": 0.2812, "step": 6160 }, { "epoch": 2.8992941176470586, "grad_norm": 0.5067701453931774, "learning_rate": 0.0002021077221654059, "loss": 0.168, "step": 6161 }, { "epoch": 2.899764705882353, "grad_norm": 0.30450241722571475, "learning_rate": 0.00020206661376468912, "loss": 0.0942, "step": 6162 }, { "epoch": 2.900235294117647, "grad_norm": 0.4404135269859224, "learning_rate": 0.00020202550091722818, "loss": 0.2363, "step": 6163 }, { "epoch": 2.9007058823529412, "grad_norm": 0.5396623718569982, "learning_rate": 0.00020198438362653426, "loss": 0.2617, "step": 6164 }, { "epoch": 2.9011764705882355, "grad_norm": 0.4500903593566814, "learning_rate": 0.00020194326189611897, "loss": 0.1641, "step": 6165 }, { "epoch": 2.9016470588235292, "grad_norm": 0.29226536136047576, "learning_rate": 0.00020190213572949435, "loss": 0.1035, "step": 6166 }, { "epoch": 2.9021176470588235, "grad_norm": 0.5006861795823025, "learning_rate": 0.0002018610051301728, "loss": 0.2832, "step": 6167 }, { "epoch": 2.9025882352941177, "grad_norm": 0.5746949491477054, "learning_rate": 0.00020181987010166699, "loss": 0.2793, "step": 6168 }, { "epoch": 2.903058823529412, "grad_norm": 0.69531871094651, "learning_rate": 0.0002017787306474901, "loss": 0.4512, "step": 6169 }, { "epoch": 2.9035294117647057, "grad_norm": 0.3881809490544036, "learning_rate": 0.00020173758677115568, "loss": 0.2139, "step": 6170 }, { "epoch": 2.904, "grad_norm": 0.8901788871132248, "learning_rate": 0.00020169643847617753, "loss": 0.4219, "step": 6171 }, { "epoch": 2.904470588235294, "grad_norm": 0.24606223447810874, "learning_rate": 0.00020165528576607, "loss": 0.1484, "step": 6172 }, { "epoch": 2.9049411764705884, "grad_norm": 0.22810965524817137, "learning_rate": 0.0002016141286443477, "loss": 0.0649, "step": 6173 }, { "epoch": 2.9054117647058826, "grad_norm": 0.6562322645952045, "learning_rate": 0.0002015729671145256, "loss": 0.4746, "step": 6174 }, { "epoch": 2.9058823529411764, "grad_norm": 0.3414006369449746, "learning_rate": 0.0002015318011801192, "loss": 0.165, "step": 6175 }, { "epoch": 2.9063529411764706, "grad_norm": 0.320538220551639, "learning_rate": 0.0002014906308446442, "loss": 0.1641, "step": 6176 }, { "epoch": 2.906823529411765, "grad_norm": 0.35534476952436395, "learning_rate": 0.0002014494561116168, "loss": 0.2002, "step": 6177 }, { "epoch": 2.9072941176470586, "grad_norm": 0.547990640379465, "learning_rate": 0.00020140827698455338, "loss": 0.3359, "step": 6178 }, { "epoch": 2.907764705882353, "grad_norm": 0.2371587431561166, "learning_rate": 0.00020136709346697107, "loss": 0.1123, "step": 6179 }, { "epoch": 2.908235294117647, "grad_norm": 0.6781700914305144, "learning_rate": 0.00020132590556238696, "loss": 0.334, "step": 6180 }, { "epoch": 2.9087058823529413, "grad_norm": 0.556758972580293, "learning_rate": 0.00020128471327431877, "loss": 0.2832, "step": 6181 }, { "epoch": 2.9091764705882355, "grad_norm": 0.367087736943273, "learning_rate": 0.0002012435166062845, "loss": 0.1973, "step": 6182 }, { "epoch": 2.9096470588235293, "grad_norm": 0.7536747990724059, "learning_rate": 0.00020120231556180262, "loss": 0.5234, "step": 6183 }, { "epoch": 2.9101176470588235, "grad_norm": 0.5116047189010988, "learning_rate": 0.00020116111014439178, "loss": 0.3223, "step": 6184 }, { "epoch": 2.9105882352941177, "grad_norm": 0.09982614591952663, "learning_rate": 0.00020111990035757125, "loss": 0.0236, "step": 6185 }, { "epoch": 2.911058823529412, "grad_norm": 0.06868814478974687, "learning_rate": 0.00020107868620486047, "loss": 0.0133, "step": 6186 }, { "epoch": 2.9115294117647057, "grad_norm": 0.6020979096181659, "learning_rate": 0.0002010374676897793, "loss": 0.168, "step": 6187 }, { "epoch": 2.912, "grad_norm": 1.2181743348706482, "learning_rate": 0.00020099624481584807, "loss": 0.5078, "step": 6188 }, { "epoch": 2.912470588235294, "grad_norm": 0.6944849496898902, "learning_rate": 0.0002009550175865874, "loss": 0.2871, "step": 6189 }, { "epoch": 2.9129411764705884, "grad_norm": 0.6755803158715302, "learning_rate": 0.00020091378600551834, "loss": 0.6211, "step": 6190 }, { "epoch": 2.9134117647058826, "grad_norm": 0.6828039590411406, "learning_rate": 0.00020087255007616222, "loss": 0.4004, "step": 6191 }, { "epoch": 2.9138823529411764, "grad_norm": 0.6898933584277098, "learning_rate": 0.00020083130980204074, "loss": 0.2891, "step": 6192 }, { "epoch": 2.9143529411764706, "grad_norm": 1.1784627622395678, "learning_rate": 0.0002007900651866761, "loss": 0.3887, "step": 6193 }, { "epoch": 2.914823529411765, "grad_norm": 1.2684309439735528, "learning_rate": 0.00020074881623359083, "loss": 0.8711, "step": 6194 }, { "epoch": 2.9152941176470586, "grad_norm": 0.3992519627452765, "learning_rate": 0.00020070756294630765, "loss": 0.1553, "step": 6195 }, { "epoch": 2.915764705882353, "grad_norm": 0.649407901714305, "learning_rate": 0.0002006663053283499, "loss": 0.5039, "step": 6196 }, { "epoch": 2.916235294117647, "grad_norm": 0.749981738110135, "learning_rate": 0.00020062504338324116, "loss": 0.3691, "step": 6197 }, { "epoch": 2.9167058823529413, "grad_norm": 0.5850294295396545, "learning_rate": 0.0002005837771145054, "loss": 0.25, "step": 6198 }, { "epoch": 2.9171764705882355, "grad_norm": 0.6341996465170444, "learning_rate": 0.00020054250652566696, "loss": 0.4961, "step": 6199 }, { "epoch": 2.9176470588235293, "grad_norm": 0.6687675737784627, "learning_rate": 0.0002005012316202506, "loss": 0.3008, "step": 6200 }, { "epoch": 2.9181176470588235, "grad_norm": 0.6798381736057191, "learning_rate": 0.00020045995240178133, "loss": 0.3086, "step": 6201 }, { "epoch": 2.9185882352941177, "grad_norm": 0.4883089305305859, "learning_rate": 0.00020041866887378457, "loss": 0.1973, "step": 6202 }, { "epoch": 2.919058823529412, "grad_norm": 0.7504365144502415, "learning_rate": 0.0002003773810397862, "loss": 0.5195, "step": 6203 }, { "epoch": 2.9195294117647057, "grad_norm": 0.6197171013741635, "learning_rate": 0.00020033608890331246, "loss": 0.4238, "step": 6204 }, { "epoch": 2.92, "grad_norm": 0.9000296031639302, "learning_rate": 0.00020029479246788973, "loss": 0.5078, "step": 6205 }, { "epoch": 2.920470588235294, "grad_norm": 0.65975664261136, "learning_rate": 0.00020025349173704508, "loss": 0.3711, "step": 6206 }, { "epoch": 2.9209411764705884, "grad_norm": 0.23516996280808264, "learning_rate": 0.00020021218671430573, "loss": 0.0728, "step": 6207 }, { "epoch": 2.9214117647058826, "grad_norm": 0.6094703415225406, "learning_rate": 0.00020017087740319936, "loss": 0.2793, "step": 6208 }, { "epoch": 2.9218823529411764, "grad_norm": 0.17231677955270938, "learning_rate": 0.00020012956380725393, "loss": 0.0493, "step": 6209 }, { "epoch": 2.9223529411764706, "grad_norm": 1.4781134775882223, "learning_rate": 0.00020008824592999795, "loss": 0.4785, "step": 6210 }, { "epoch": 2.922823529411765, "grad_norm": 0.5451207161509343, "learning_rate": 0.00020004692377496004, "loss": 0.2832, "step": 6211 }, { "epoch": 2.9232941176470586, "grad_norm": 0.549882261989153, "learning_rate": 0.00020000559734566933, "loss": 0.3398, "step": 6212 }, { "epoch": 2.923764705882353, "grad_norm": 0.6162755850087082, "learning_rate": 0.00019996426664565538, "loss": 0.2168, "step": 6213 }, { "epoch": 2.924235294117647, "grad_norm": 0.5471357168659542, "learning_rate": 0.000199922931678448, "loss": 0.2891, "step": 6214 }, { "epoch": 2.9247058823529413, "grad_norm": 0.5894203538750471, "learning_rate": 0.00019988159244757736, "loss": 0.2197, "step": 6215 }, { "epoch": 2.9251764705882355, "grad_norm": 0.703540891877273, "learning_rate": 0.000199840248956574, "loss": 0.3574, "step": 6216 }, { "epoch": 2.9256470588235293, "grad_norm": 0.3711522249402425, "learning_rate": 0.000199798901208969, "loss": 0.1934, "step": 6217 }, { "epoch": 2.9261176470588235, "grad_norm": 0.6009805827165962, "learning_rate": 0.0001997575492082935, "loss": 0.1982, "step": 6218 }, { "epoch": 2.9265882352941177, "grad_norm": 1.0165486548922313, "learning_rate": 0.0001997161929580793, "loss": 0.4727, "step": 6219 }, { "epoch": 2.9270588235294115, "grad_norm": 0.39910225316010284, "learning_rate": 0.00019967483246185834, "loss": 0.1465, "step": 6220 }, { "epoch": 2.9275294117647057, "grad_norm": 0.5952609969414533, "learning_rate": 0.00019963346772316307, "loss": 0.4062, "step": 6221 }, { "epoch": 2.928, "grad_norm": 0.5748570137372201, "learning_rate": 0.00019959209874552612, "loss": 0.3535, "step": 6222 }, { "epoch": 2.928470588235294, "grad_norm": 0.9049686509041697, "learning_rate": 0.00019955072553248075, "loss": 0.6367, "step": 6223 }, { "epoch": 2.9289411764705884, "grad_norm": 0.2627434405795739, "learning_rate": 0.0001995093480875604, "loss": 0.082, "step": 6224 }, { "epoch": 2.9294117647058826, "grad_norm": 0.4312258931303729, "learning_rate": 0.00019946796641429884, "loss": 0.0884, "step": 6225 }, { "epoch": 2.9298823529411764, "grad_norm": 0.4071069679873724, "learning_rate": 0.0001994265805162303, "loss": 0.1553, "step": 6226 }, { "epoch": 2.9303529411764706, "grad_norm": 0.7940380976638789, "learning_rate": 0.00019938519039688937, "loss": 0.6953, "step": 6227 }, { "epoch": 2.930823529411765, "grad_norm": 0.8839595722691558, "learning_rate": 0.00019934379605981097, "loss": 0.3086, "step": 6228 }, { "epoch": 2.9312941176470586, "grad_norm": 0.9012655039769791, "learning_rate": 0.0001993023975085303, "loss": 0.4551, "step": 6229 }, { "epoch": 2.931764705882353, "grad_norm": 0.3348257540092137, "learning_rate": 0.00019926099474658305, "loss": 0.1465, "step": 6230 }, { "epoch": 2.932235294117647, "grad_norm": 0.46307492333923506, "learning_rate": 0.0001992195877775053, "loss": 0.2891, "step": 6231 }, { "epoch": 2.9327058823529413, "grad_norm": 1.0028129498489873, "learning_rate": 0.00019917817660483327, "loss": 0.6445, "step": 6232 }, { "epoch": 2.9331764705882355, "grad_norm": 0.439873755593142, "learning_rate": 0.00019913676123210373, "loss": 0.1592, "step": 6233 }, { "epoch": 2.9336470588235293, "grad_norm": 0.6304407181721868, "learning_rate": 0.00019909534166285382, "loss": 0.3105, "step": 6234 }, { "epoch": 2.9341176470588235, "grad_norm": 0.5843416820348565, "learning_rate": 0.00019905391790062092, "loss": 0.5078, "step": 6235 }, { "epoch": 2.9345882352941177, "grad_norm": 0.6649840650684066, "learning_rate": 0.00019901248994894276, "loss": 0.4238, "step": 6236 }, { "epoch": 2.9350588235294115, "grad_norm": 0.6382156614913462, "learning_rate": 0.0001989710578113576, "loss": 0.332, "step": 6237 }, { "epoch": 2.9355294117647057, "grad_norm": 0.3240829720289809, "learning_rate": 0.00019892962149140392, "loss": 0.0977, "step": 6238 }, { "epoch": 2.936, "grad_norm": 0.39863524490354996, "learning_rate": 0.0001988881809926205, "loss": 0.1904, "step": 6239 }, { "epoch": 2.936470588235294, "grad_norm": 0.6389943633119379, "learning_rate": 0.00019884673631854672, "loss": 0.2656, "step": 6240 }, { "epoch": 2.9369411764705884, "grad_norm": 0.7862822043895529, "learning_rate": 0.000198805287472722, "loss": 0.252, "step": 6241 }, { "epoch": 2.9374117647058826, "grad_norm": 0.22021154300786824, "learning_rate": 0.0001987638344586864, "loss": 0.0366, "step": 6242 }, { "epoch": 2.9378823529411764, "grad_norm": 0.6243728714626801, "learning_rate": 0.00019872237727998004, "loss": 0.2871, "step": 6243 }, { "epoch": 2.9383529411764706, "grad_norm": 0.5749751980030593, "learning_rate": 0.0001986809159401438, "loss": 0.2373, "step": 6244 }, { "epoch": 2.938823529411765, "grad_norm": 0.2637104637581865, "learning_rate": 0.00019863945044271846, "loss": 0.0879, "step": 6245 }, { "epoch": 2.9392941176470586, "grad_norm": 0.7428843411788777, "learning_rate": 0.00019859798079124554, "loss": 0.3809, "step": 6246 }, { "epoch": 2.939764705882353, "grad_norm": 0.358473285080412, "learning_rate": 0.00019855650698926666, "loss": 0.2031, "step": 6247 }, { "epoch": 2.940235294117647, "grad_norm": 0.42224834492175023, "learning_rate": 0.00019851502904032396, "loss": 0.2871, "step": 6248 }, { "epoch": 2.9407058823529413, "grad_norm": 0.5094855729990716, "learning_rate": 0.00019847354694795978, "loss": 0.2256, "step": 6249 }, { "epoch": 2.9411764705882355, "grad_norm": 0.631121806363893, "learning_rate": 0.00019843206071571692, "loss": 0.418, "step": 6250 }, { "epoch": 2.9416470588235293, "grad_norm": 0.041264858665411686, "learning_rate": 0.00019839057034713858, "loss": 0.0037, "step": 6251 }, { "epoch": 2.9421176470588235, "grad_norm": 1.4741303473939245, "learning_rate": 0.00019834907584576812, "loss": 1.1484, "step": 6252 }, { "epoch": 2.9425882352941177, "grad_norm": 0.9293475903733931, "learning_rate": 0.00019830757721514943, "loss": 0.3945, "step": 6253 }, { "epoch": 2.9430588235294115, "grad_norm": 0.5341371003618399, "learning_rate": 0.00019826607445882667, "loss": 0.2656, "step": 6254 }, { "epoch": 2.9435294117647057, "grad_norm": 0.399515037577877, "learning_rate": 0.00019822456758034445, "loss": 0.2148, "step": 6255 }, { "epoch": 2.944, "grad_norm": 0.5991903124391074, "learning_rate": 0.0001981830565832476, "loss": 0.2832, "step": 6256 }, { "epoch": 2.944470588235294, "grad_norm": 0.21112422815724843, "learning_rate": 0.0001981415414710814, "loss": 0.0552, "step": 6257 }, { "epoch": 2.9449411764705884, "grad_norm": 0.18725238427979685, "learning_rate": 0.00019810002224739142, "loss": 0.0527, "step": 6258 }, { "epoch": 2.9454117647058826, "grad_norm": 0.6814733068543498, "learning_rate": 0.00019805849891572363, "loss": 0.2256, "step": 6259 }, { "epoch": 2.9458823529411764, "grad_norm": 0.49098394454383215, "learning_rate": 0.00019801697147962428, "loss": 0.2373, "step": 6260 }, { "epoch": 2.9463529411764706, "grad_norm": 0.9482909531875718, "learning_rate": 0.00019797543994264004, "loss": 0.4473, "step": 6261 }, { "epoch": 2.946823529411765, "grad_norm": 0.6258832863429501, "learning_rate": 0.00019793390430831794, "loss": 0.3633, "step": 6262 }, { "epoch": 2.9472941176470586, "grad_norm": 0.7592038200713381, "learning_rate": 0.00019789236458020527, "loss": 0.4102, "step": 6263 }, { "epoch": 2.947764705882353, "grad_norm": 0.7542233442359475, "learning_rate": 0.00019785082076184977, "loss": 0.4766, "step": 6264 }, { "epoch": 2.948235294117647, "grad_norm": 0.41085139248246744, "learning_rate": 0.00019780927285679945, "loss": 0.1426, "step": 6265 }, { "epoch": 2.9487058823529413, "grad_norm": 0.3450045543576048, "learning_rate": 0.00019776772086860277, "loss": 0.1631, "step": 6266 }, { "epoch": 2.9491764705882355, "grad_norm": 0.37918233847408983, "learning_rate": 0.0001977261648008084, "loss": 0.2373, "step": 6267 }, { "epoch": 2.9496470588235293, "grad_norm": 0.24515316503214932, "learning_rate": 0.00019768460465696545, "loss": 0.0518, "step": 6268 }, { "epoch": 2.9501176470588235, "grad_norm": 0.45232144401201546, "learning_rate": 0.0001976430404406234, "loss": 0.2852, "step": 6269 }, { "epoch": 2.9505882352941177, "grad_norm": 0.9372420186317776, "learning_rate": 0.00019760147215533197, "loss": 0.5859, "step": 6270 }, { "epoch": 2.9510588235294115, "grad_norm": 0.4992124743137893, "learning_rate": 0.00019755989980464137, "loss": 0.2949, "step": 6271 }, { "epoch": 2.9515294117647057, "grad_norm": 0.5525487029227147, "learning_rate": 0.00019751832339210204, "loss": 0.2676, "step": 6272 }, { "epoch": 2.952, "grad_norm": 0.3314742814271528, "learning_rate": 0.00019747674292126477, "loss": 0.1021, "step": 6273 }, { "epoch": 2.952470588235294, "grad_norm": 0.6613592650141941, "learning_rate": 0.00019743515839568083, "loss": 0.4492, "step": 6274 }, { "epoch": 2.9529411764705884, "grad_norm": 0.4644653152600244, "learning_rate": 0.00019739356981890167, "loss": 0.1133, "step": 6275 }, { "epoch": 2.9534117647058826, "grad_norm": 0.47587418458608294, "learning_rate": 0.00019735197719447916, "loss": 0.3965, "step": 6276 }, { "epoch": 2.9538823529411764, "grad_norm": 0.7708186276853086, "learning_rate": 0.00019731038052596558, "loss": 0.4434, "step": 6277 }, { "epoch": 2.9543529411764706, "grad_norm": 0.3104089829601269, "learning_rate": 0.00019726877981691347, "loss": 0.1133, "step": 6278 }, { "epoch": 2.954823529411765, "grad_norm": 0.8642489314863485, "learning_rate": 0.0001972271750708757, "loss": 0.4102, "step": 6279 }, { "epoch": 2.9552941176470586, "grad_norm": 0.5905462971891802, "learning_rate": 0.0001971855662914055, "loss": 0.293, "step": 6280 }, { "epoch": 2.955764705882353, "grad_norm": 0.5554080828698138, "learning_rate": 0.00019714395348205649, "loss": 0.3281, "step": 6281 }, { "epoch": 2.956235294117647, "grad_norm": 0.8769737837849125, "learning_rate": 0.00019710233664638264, "loss": 0.4414, "step": 6282 }, { "epoch": 2.9567058823529413, "grad_norm": 0.7011333392236605, "learning_rate": 0.00019706071578793819, "loss": 0.3535, "step": 6283 }, { "epoch": 2.9571764705882355, "grad_norm": 0.7903557464574158, "learning_rate": 0.00019701909091027778, "loss": 0.5664, "step": 6284 }, { "epoch": 2.9576470588235293, "grad_norm": 0.8821080264634521, "learning_rate": 0.0001969774620169564, "loss": 0.4531, "step": 6285 }, { "epoch": 2.9581176470588235, "grad_norm": 0.37474278348821, "learning_rate": 0.00019693582911152935, "loss": 0.2637, "step": 6286 }, { "epoch": 2.9585882352941177, "grad_norm": 0.616242222964146, "learning_rate": 0.0001968941921975522, "loss": 0.3223, "step": 6287 }, { "epoch": 2.9590588235294115, "grad_norm": 0.8484831076836118, "learning_rate": 0.00019685255127858107, "loss": 0.3984, "step": 6288 }, { "epoch": 2.9595294117647057, "grad_norm": 0.5226178553454515, "learning_rate": 0.00019681090635817228, "loss": 0.2471, "step": 6289 }, { "epoch": 2.96, "grad_norm": 0.4164084081070091, "learning_rate": 0.00019676925743988244, "loss": 0.127, "step": 6290 }, { "epoch": 2.960470588235294, "grad_norm": 0.34299832232519, "learning_rate": 0.00019672760452726863, "loss": 0.1865, "step": 6291 }, { "epoch": 2.9609411764705884, "grad_norm": 0.2810599876825006, "learning_rate": 0.0001966859476238882, "loss": 0.168, "step": 6292 }, { "epoch": 2.961411764705882, "grad_norm": 0.6953844372985415, "learning_rate": 0.00019664428673329883, "loss": 0.3906, "step": 6293 }, { "epoch": 2.9618823529411764, "grad_norm": 0.361578964578794, "learning_rate": 0.00019660262185905856, "loss": 0.1108, "step": 6294 }, { "epoch": 2.9623529411764706, "grad_norm": 1.0242061590165394, "learning_rate": 0.0001965609530047259, "loss": 0.4238, "step": 6295 }, { "epoch": 2.962823529411765, "grad_norm": 1.0287025355351807, "learning_rate": 0.0001965192801738594, "loss": 0.5078, "step": 6296 }, { "epoch": 2.9632941176470586, "grad_norm": 0.37756152170766627, "learning_rate": 0.0001964776033700182, "loss": 0.0767, "step": 6297 }, { "epoch": 2.963764705882353, "grad_norm": 0.45458376749315893, "learning_rate": 0.00019643592259676164, "loss": 0.1973, "step": 6298 }, { "epoch": 2.964235294117647, "grad_norm": 0.15172675361679866, "learning_rate": 0.0001963942378576496, "loss": 0.0466, "step": 6299 }, { "epoch": 2.9647058823529413, "grad_norm": 1.1878902605644597, "learning_rate": 0.0001963525491562421, "loss": 0.543, "step": 6300 }, { "epoch": 2.9651764705882355, "grad_norm": 0.3406195205873971, "learning_rate": 0.0001963108564960995, "loss": 0.1699, "step": 6301 }, { "epoch": 2.9656470588235293, "grad_norm": 0.450623950168524, "learning_rate": 0.00019626915988078263, "loss": 0.3281, "step": 6302 }, { "epoch": 2.9661176470588235, "grad_norm": 0.8109868677100465, "learning_rate": 0.00019622745931385254, "loss": 0.2578, "step": 6303 }, { "epoch": 2.9665882352941177, "grad_norm": 0.6481216703586661, "learning_rate": 0.00019618575479887066, "loss": 0.2598, "step": 6304 }, { "epoch": 2.9670588235294115, "grad_norm": 0.2894707595377978, "learning_rate": 0.00019614404633939886, "loss": 0.1035, "step": 6305 }, { "epoch": 2.9675294117647057, "grad_norm": 0.2656348298486315, "learning_rate": 0.0001961023339389991, "loss": 0.1543, "step": 6306 }, { "epoch": 2.968, "grad_norm": 0.65509830642269, "learning_rate": 0.00019606061760123392, "loss": 0.3379, "step": 6307 }, { "epoch": 2.968470588235294, "grad_norm": 0.7390309348972192, "learning_rate": 0.00019601889732966604, "loss": 0.3008, "step": 6308 }, { "epoch": 2.9689411764705884, "grad_norm": 0.30668115027299686, "learning_rate": 0.0001959771731278586, "loss": 0.1118, "step": 6309 }, { "epoch": 2.969411764705882, "grad_norm": 0.26893775609588566, "learning_rate": 0.0001959354449993751, "loss": 0.0962, "step": 6310 }, { "epoch": 2.9698823529411764, "grad_norm": 0.3084365133317706, "learning_rate": 0.00019589371294777925, "loss": 0.0884, "step": 6311 }, { "epoch": 2.9703529411764706, "grad_norm": 0.7137563928334816, "learning_rate": 0.00019585197697663526, "loss": 0.3105, "step": 6312 }, { "epoch": 2.970823529411765, "grad_norm": 0.5555504590645702, "learning_rate": 0.0001958102370895075, "loss": 0.2539, "step": 6313 }, { "epoch": 2.9712941176470586, "grad_norm": 0.641695891238703, "learning_rate": 0.00019576849328996075, "loss": 0.3379, "step": 6314 }, { "epoch": 2.971764705882353, "grad_norm": 0.5025737984879556, "learning_rate": 0.00019572674558156024, "loss": 0.2441, "step": 6315 }, { "epoch": 2.972235294117647, "grad_norm": 0.5111309630352316, "learning_rate": 0.00019568499396787134, "loss": 0.293, "step": 6316 }, { "epoch": 2.9727058823529413, "grad_norm": 0.9743213270245913, "learning_rate": 0.0001956432384524599, "loss": 0.6484, "step": 6317 }, { "epoch": 2.9731764705882355, "grad_norm": 0.5193967778220546, "learning_rate": 0.000195601479038892, "loss": 0.1807, "step": 6318 }, { "epoch": 2.9736470588235293, "grad_norm": 0.3570143758923615, "learning_rate": 0.0001955597157307341, "loss": 0.1768, "step": 6319 }, { "epoch": 2.9741176470588235, "grad_norm": 0.5052797192227093, "learning_rate": 0.000195517948531553, "loss": 0.3457, "step": 6320 }, { "epoch": 2.9745882352941178, "grad_norm": 0.6857929773473299, "learning_rate": 0.00019547617744491587, "loss": 0.6172, "step": 6321 }, { "epoch": 2.9750588235294115, "grad_norm": 0.3443671042751411, "learning_rate": 0.00019543440247439008, "loss": 0.1816, "step": 6322 }, { "epoch": 2.9755294117647058, "grad_norm": 0.17738901265288987, "learning_rate": 0.00019539262362354353, "loss": 0.061, "step": 6323 }, { "epoch": 2.976, "grad_norm": 0.6754186345856364, "learning_rate": 0.00019535084089594422, "loss": 0.5703, "step": 6324 }, { "epoch": 2.976470588235294, "grad_norm": 0.4976053956112988, "learning_rate": 0.00019530905429516066, "loss": 0.3301, "step": 6325 }, { "epoch": 2.9769411764705884, "grad_norm": 0.34751258603697815, "learning_rate": 0.00019526726382476165, "loss": 0.1318, "step": 6326 }, { "epoch": 2.977411764705882, "grad_norm": 0.1438195381250165, "learning_rate": 0.00019522546948831627, "loss": 0.0957, "step": 6327 }, { "epoch": 2.9778823529411764, "grad_norm": 0.8584312432788573, "learning_rate": 0.00019518367128939392, "loss": 0.4727, "step": 6328 }, { "epoch": 2.9783529411764706, "grad_norm": 0.4497798484890942, "learning_rate": 0.0001951418692315645, "loss": 0.2734, "step": 6329 }, { "epoch": 2.978823529411765, "grad_norm": 0.44211606731758124, "learning_rate": 0.00019510006331839794, "loss": 0.2109, "step": 6330 }, { "epoch": 2.9792941176470586, "grad_norm": 0.3727746071700409, "learning_rate": 0.00019505825355346483, "loss": 0.0845, "step": 6331 }, { "epoch": 2.979764705882353, "grad_norm": 0.8436633469206926, "learning_rate": 0.0001950164399403358, "loss": 0.5195, "step": 6332 }, { "epoch": 2.980235294117647, "grad_norm": 0.7680581843685419, "learning_rate": 0.00019497462248258202, "loss": 0.6484, "step": 6333 }, { "epoch": 2.9807058823529413, "grad_norm": 0.3967704570805833, "learning_rate": 0.0001949328011837749, "loss": 0.2227, "step": 6334 }, { "epoch": 2.9811764705882355, "grad_norm": 0.6704307402362367, "learning_rate": 0.00019489097604748613, "loss": 0.3613, "step": 6335 }, { "epoch": 2.9816470588235293, "grad_norm": 0.502796499557877, "learning_rate": 0.00019484914707728785, "loss": 0.252, "step": 6336 }, { "epoch": 2.9821176470588235, "grad_norm": 0.5948353522806974, "learning_rate": 0.00019480731427675238, "loss": 0.3418, "step": 6337 }, { "epoch": 2.9825882352941178, "grad_norm": 0.43166392313964075, "learning_rate": 0.0001947654776494525, "loss": 0.1973, "step": 6338 }, { "epoch": 2.9830588235294115, "grad_norm": 0.2012672252748564, "learning_rate": 0.0001947236371989613, "loss": 0.063, "step": 6339 }, { "epoch": 2.9835294117647058, "grad_norm": 0.7857979065764017, "learning_rate": 0.0001946817929288521, "loss": 0.4668, "step": 6340 }, { "epoch": 2.984, "grad_norm": 0.7261265678576789, "learning_rate": 0.0001946399448426986, "loss": 0.4316, "step": 6341 }, { "epoch": 2.984470588235294, "grad_norm": 0.6538385363975371, "learning_rate": 0.00019459809294407486, "loss": 0.3535, "step": 6342 }, { "epoch": 2.9849411764705884, "grad_norm": 0.34154524490377514, "learning_rate": 0.00019455623723655521, "loss": 0.1406, "step": 6343 }, { "epoch": 2.985411764705882, "grad_norm": 0.5427698042804658, "learning_rate": 0.00019451437772371438, "loss": 0.4766, "step": 6344 }, { "epoch": 2.9858823529411764, "grad_norm": 0.7838514315059024, "learning_rate": 0.00019447251440912736, "loss": 0.5586, "step": 6345 }, { "epoch": 2.9863529411764707, "grad_norm": 0.48481695072425035, "learning_rate": 0.00019443064729636947, "loss": 0.2207, "step": 6346 }, { "epoch": 2.9868235294117644, "grad_norm": 0.4940203494790976, "learning_rate": 0.00019438877638901635, "loss": 0.1943, "step": 6347 }, { "epoch": 2.9872941176470587, "grad_norm": 0.4773202713940455, "learning_rate": 0.00019434690169064402, "loss": 0.1973, "step": 6348 }, { "epoch": 2.987764705882353, "grad_norm": 0.36691557610106823, "learning_rate": 0.00019430502320482871, "loss": 0.1982, "step": 6349 }, { "epoch": 2.988235294117647, "grad_norm": 0.5693479038549359, "learning_rate": 0.00019426314093514717, "loss": 0.3379, "step": 6350 }, { "epoch": 2.9887058823529413, "grad_norm": 0.4702237956285457, "learning_rate": 0.0001942212548851763, "loss": 0.1846, "step": 6351 }, { "epoch": 2.9891764705882355, "grad_norm": 0.6361512117449258, "learning_rate": 0.00019417936505849328, "loss": 0.3789, "step": 6352 }, { "epoch": 2.9896470588235293, "grad_norm": 0.8778587549337787, "learning_rate": 0.00019413747145867583, "loss": 0.4062, "step": 6353 }, { "epoch": 2.9901176470588235, "grad_norm": 1.0713906494627774, "learning_rate": 0.00019409557408930184, "loss": 0.4375, "step": 6354 }, { "epoch": 2.9905882352941178, "grad_norm": 0.5910382187453344, "learning_rate": 0.0001940536729539495, "loss": 0.2988, "step": 6355 }, { "epoch": 2.9910588235294115, "grad_norm": 0.321085280472087, "learning_rate": 0.00019401176805619746, "loss": 0.1118, "step": 6356 }, { "epoch": 2.9915294117647058, "grad_norm": 0.4241526804342609, "learning_rate": 0.0001939698593996245, "loss": 0.1553, "step": 6357 }, { "epoch": 2.992, "grad_norm": 0.24542095625917, "learning_rate": 0.00019392794698780988, "loss": 0.0752, "step": 6358 }, { "epoch": 2.992470588235294, "grad_norm": 0.6986268025304806, "learning_rate": 0.00019388603082433315, "loss": 0.5703, "step": 6359 }, { "epoch": 2.9929411764705884, "grad_norm": 0.1350926408420689, "learning_rate": 0.00019384411091277412, "loss": 0.0527, "step": 6360 }, { "epoch": 2.993411764705882, "grad_norm": 0.4081779486108775, "learning_rate": 0.00019380218725671298, "loss": 0.1875, "step": 6361 }, { "epoch": 2.9938823529411764, "grad_norm": 0.6380729207227459, "learning_rate": 0.00019376025985973013, "loss": 0.2695, "step": 6362 }, { "epoch": 2.9943529411764707, "grad_norm": 0.1943858858112964, "learning_rate": 0.00019371832872540652, "loss": 0.0864, "step": 6363 }, { "epoch": 2.9948235294117644, "grad_norm": 0.46174773705713457, "learning_rate": 0.00019367639385732316, "loss": 0.2178, "step": 6364 }, { "epoch": 2.9952941176470587, "grad_norm": 0.50756634238749, "learning_rate": 0.00019363445525906152, "loss": 0.249, "step": 6365 }, { "epoch": 2.995764705882353, "grad_norm": 0.7638066222081874, "learning_rate": 0.00019359251293420336, "loss": 0.3906, "step": 6366 }, { "epoch": 2.996235294117647, "grad_norm": 0.7523183176621423, "learning_rate": 0.00019355056688633086, "loss": 0.5, "step": 6367 }, { "epoch": 2.9967058823529413, "grad_norm": 0.06496331000028763, "learning_rate": 0.0001935086171190263, "loss": 0.0205, "step": 6368 }, { "epoch": 2.9971764705882356, "grad_norm": 0.5024785773081517, "learning_rate": 0.00019346666363587238, "loss": 0.3477, "step": 6369 }, { "epoch": 2.9976470588235293, "grad_norm": 0.9913649812580256, "learning_rate": 0.00019342470644045223, "loss": 0.6094, "step": 6370 }, { "epoch": 2.9981176470588236, "grad_norm": 0.35479743861166324, "learning_rate": 0.00019338274553634913, "loss": 0.1875, "step": 6371 }, { "epoch": 2.9985882352941178, "grad_norm": 0.3823864330437947, "learning_rate": 0.00019334078092714683, "loss": 0.1089, "step": 6372 }, { "epoch": 2.9990588235294116, "grad_norm": 0.17329420773323023, "learning_rate": 0.00019329881261642918, "loss": 0.0327, "step": 6373 }, { "epoch": 2.9995294117647058, "grad_norm": 0.3023358553900344, "learning_rate": 0.00019325684060778056, "loss": 0.1582, "step": 6374 }, { "epoch": 3.0, "grad_norm": 0.36156407844506483, "learning_rate": 0.00019321486490478563, "loss": 0.1592, "step": 6375 }, { "epoch": 3.0, "eval_loss": 0.2771148979663849, "eval_runtime": 61.6096, "eval_samples_per_second": 12.173, "eval_steps_per_second": 6.087, "step": 6375 }, { "epoch": 3.0004705882352942, "grad_norm": 0.7796569737266739, "learning_rate": 0.0001931728855110292, "loss": 0.4102, "step": 6376 }, { "epoch": 3.0009411764705884, "grad_norm": 0.4223675262181774, "learning_rate": 0.00019313090243009665, "loss": 0.2314, "step": 6377 }, { "epoch": 3.0014117647058822, "grad_norm": 0.5849164486684518, "learning_rate": 0.00019308891566557345, "loss": 0.2695, "step": 6378 }, { "epoch": 3.0018823529411764, "grad_norm": 0.6568205190396736, "learning_rate": 0.0001930469252210455, "loss": 0.3594, "step": 6379 }, { "epoch": 3.0023529411764707, "grad_norm": 0.3939123231598081, "learning_rate": 0.000193004931100099, "loss": 0.1318, "step": 6380 }, { "epoch": 3.002823529411765, "grad_norm": 0.8581930252328285, "learning_rate": 0.00019296293330632047, "loss": 0.6875, "step": 6381 }, { "epoch": 3.0032941176470587, "grad_norm": 0.36771605259321877, "learning_rate": 0.0001929209318432967, "loss": 0.1973, "step": 6382 }, { "epoch": 3.003764705882353, "grad_norm": 0.539719911179031, "learning_rate": 0.00019287892671461475, "loss": 0.3496, "step": 6383 }, { "epoch": 3.004235294117647, "grad_norm": 0.5337403377040445, "learning_rate": 0.0001928369179238622, "loss": 0.3711, "step": 6384 }, { "epoch": 3.0047058823529413, "grad_norm": 0.555438142883609, "learning_rate": 0.00019279490547462674, "loss": 0.3496, "step": 6385 }, { "epoch": 3.005176470588235, "grad_norm": 0.5093533949313972, "learning_rate": 0.0001927528893704964, "loss": 0.3008, "step": 6386 }, { "epoch": 3.0056470588235293, "grad_norm": 0.31975237616921126, "learning_rate": 0.00019271086961505964, "loss": 0.1279, "step": 6387 }, { "epoch": 3.0061176470588236, "grad_norm": 0.620313661890212, "learning_rate": 0.0001926688462119051, "loss": 0.3691, "step": 6388 }, { "epoch": 3.006588235294118, "grad_norm": 0.9182768146640328, "learning_rate": 0.00019262681916462187, "loss": 0.6094, "step": 6389 }, { "epoch": 3.0070588235294116, "grad_norm": 0.25814360465943265, "learning_rate": 0.00019258478847679914, "loss": 0.1465, "step": 6390 }, { "epoch": 3.007529411764706, "grad_norm": 0.3162685772420085, "learning_rate": 0.0001925427541520266, "loss": 0.1562, "step": 6391 }, { "epoch": 3.008, "grad_norm": 0.5670063667952792, "learning_rate": 0.00019250071619389422, "loss": 0.3105, "step": 6392 }, { "epoch": 3.0084705882352942, "grad_norm": 0.5190093655489778, "learning_rate": 0.00019245867460599217, "loss": 0.252, "step": 6393 }, { "epoch": 3.0089411764705885, "grad_norm": 0.6185993508182276, "learning_rate": 0.00019241662939191106, "loss": 0.3281, "step": 6394 }, { "epoch": 3.0094117647058822, "grad_norm": 0.4105369609239397, "learning_rate": 0.00019237458055524178, "loss": 0.1338, "step": 6395 }, { "epoch": 3.0098823529411765, "grad_norm": 0.6864169565338775, "learning_rate": 0.00019233252809957546, "loss": 0.5078, "step": 6396 }, { "epoch": 3.0103529411764707, "grad_norm": 0.29597744465189646, "learning_rate": 0.00019229047202850357, "loss": 0.1064, "step": 6397 }, { "epoch": 3.010823529411765, "grad_norm": 0.7057813093569419, "learning_rate": 0.00019224841234561797, "loss": 0.3164, "step": 6398 }, { "epoch": 3.0112941176470587, "grad_norm": 1.3164899725712043, "learning_rate": 0.00019220634905451074, "loss": 0.6562, "step": 6399 }, { "epoch": 3.011764705882353, "grad_norm": 0.44601382218225194, "learning_rate": 0.00019216428215877425, "loss": 0.1689, "step": 6400 }, { "epoch": 3.012235294117647, "grad_norm": 0.7924784557260869, "learning_rate": 0.00019212221166200124, "loss": 0.498, "step": 6401 }, { "epoch": 3.0127058823529413, "grad_norm": 0.464723528570025, "learning_rate": 0.00019208013756778476, "loss": 0.0889, "step": 6402 }, { "epoch": 3.013176470588235, "grad_norm": 0.34077248921323483, "learning_rate": 0.0001920380598797181, "loss": 0.0864, "step": 6403 }, { "epoch": 3.0136470588235293, "grad_norm": 0.21183297647257135, "learning_rate": 0.000191995978601395, "loss": 0.041, "step": 6404 }, { "epoch": 3.0141176470588236, "grad_norm": 0.964549091941062, "learning_rate": 0.0001919538937364093, "loss": 0.3477, "step": 6405 }, { "epoch": 3.014588235294118, "grad_norm": 0.7472109241508773, "learning_rate": 0.00019191180528835527, "loss": 0.3809, "step": 6406 }, { "epoch": 3.0150588235294116, "grad_norm": 0.6112671941885142, "learning_rate": 0.00019186971326082752, "loss": 0.3535, "step": 6407 }, { "epoch": 3.015529411764706, "grad_norm": 0.7322767464084692, "learning_rate": 0.0001918276176574209, "loss": 0.3359, "step": 6408 }, { "epoch": 3.016, "grad_norm": 0.8530069110439841, "learning_rate": 0.00019178551848173054, "loss": 0.5195, "step": 6409 }, { "epoch": 3.0164705882352942, "grad_norm": 0.5778581739842896, "learning_rate": 0.00019174341573735196, "loss": 0.3223, "step": 6410 }, { "epoch": 3.016941176470588, "grad_norm": 0.48217523782680677, "learning_rate": 0.0001917013094278809, "loss": 0.3008, "step": 6411 }, { "epoch": 3.0174117647058822, "grad_norm": 0.934227176744375, "learning_rate": 0.0001916591995569135, "loss": 0.6172, "step": 6412 }, { "epoch": 3.0178823529411765, "grad_norm": 0.6172803306704264, "learning_rate": 0.00019161708612804615, "loss": 0.3633, "step": 6413 }, { "epoch": 3.0183529411764707, "grad_norm": 0.6731152670110475, "learning_rate": 0.00019157496914487545, "loss": 0.4746, "step": 6414 }, { "epoch": 3.018823529411765, "grad_norm": 0.9270517240039355, "learning_rate": 0.00019153284861099848, "loss": 0.5156, "step": 6415 }, { "epoch": 3.0192941176470587, "grad_norm": 0.2954897735059412, "learning_rate": 0.00019149072453001254, "loss": 0.1348, "step": 6416 }, { "epoch": 3.019764705882353, "grad_norm": 0.85474466273413, "learning_rate": 0.00019144859690551518, "loss": 0.3965, "step": 6417 }, { "epoch": 3.020235294117647, "grad_norm": 0.5496862354899181, "learning_rate": 0.00019140646574110438, "loss": 0.2197, "step": 6418 }, { "epoch": 3.0207058823529414, "grad_norm": 0.48600876676131494, "learning_rate": 0.0001913643310403783, "loss": 0.1709, "step": 6419 }, { "epoch": 3.021176470588235, "grad_norm": 0.6005055970850678, "learning_rate": 0.00019132219280693544, "loss": 0.1318, "step": 6420 }, { "epoch": 3.0216470588235294, "grad_norm": 0.3460361531603845, "learning_rate": 0.00019128005104437462, "loss": 0.1177, "step": 6421 }, { "epoch": 3.0221176470588236, "grad_norm": 0.7265716795781588, "learning_rate": 0.00019123790575629503, "loss": 0.4043, "step": 6422 }, { "epoch": 3.022588235294118, "grad_norm": 0.24960272712765547, "learning_rate": 0.00019119575694629594, "loss": 0.0559, "step": 6423 }, { "epoch": 3.0230588235294116, "grad_norm": 0.8861115886999735, "learning_rate": 0.00019115360461797716, "loss": 0.3164, "step": 6424 }, { "epoch": 3.023529411764706, "grad_norm": 0.8145449090575372, "learning_rate": 0.00019111144877493873, "loss": 0.416, "step": 6425 }, { "epoch": 3.024, "grad_norm": 0.8589772949650406, "learning_rate": 0.0001910692894207809, "loss": 0.4258, "step": 6426 }, { "epoch": 3.0244705882352942, "grad_norm": 0.2813535197343175, "learning_rate": 0.00019102712655910426, "loss": 0.0608, "step": 6427 }, { "epoch": 3.024941176470588, "grad_norm": 0.17136471118935162, "learning_rate": 0.00019098496019350985, "loss": 0.0708, "step": 6428 }, { "epoch": 3.0254117647058822, "grad_norm": 0.7676055152433899, "learning_rate": 0.0001909427903275988, "loss": 0.3438, "step": 6429 }, { "epoch": 3.0258823529411765, "grad_norm": 0.8825288950972447, "learning_rate": 0.00019090061696497258, "loss": 0.3926, "step": 6430 }, { "epoch": 3.0263529411764707, "grad_norm": 0.8235243848713525, "learning_rate": 0.0001908584401092331, "loss": 0.5078, "step": 6431 }, { "epoch": 3.026823529411765, "grad_norm": 0.3542032982066136, "learning_rate": 0.0001908162597639824, "loss": 0.2402, "step": 6432 }, { "epoch": 3.0272941176470587, "grad_norm": 0.44748431695109764, "learning_rate": 0.00019077407593282294, "loss": 0.2891, "step": 6433 }, { "epoch": 3.027764705882353, "grad_norm": 0.3020395313460797, "learning_rate": 0.00019073188861935735, "loss": 0.1982, "step": 6434 }, { "epoch": 3.028235294117647, "grad_norm": 0.42923555808794367, "learning_rate": 0.00019068969782718874, "loss": 0.2168, "step": 6435 }, { "epoch": 3.0287058823529414, "grad_norm": 0.7963387038582185, "learning_rate": 0.00019064750355992036, "loss": 0.5195, "step": 6436 }, { "epoch": 3.029176470588235, "grad_norm": 0.5410892860641043, "learning_rate": 0.00019060530582115578, "loss": 0.3223, "step": 6437 }, { "epoch": 3.0296470588235294, "grad_norm": 0.37693705827776897, "learning_rate": 0.00019056310461449885, "loss": 0.2617, "step": 6438 }, { "epoch": 3.0301176470588236, "grad_norm": 0.5982894421886157, "learning_rate": 0.00019052089994355394, "loss": 0.2969, "step": 6439 }, { "epoch": 3.030588235294118, "grad_norm": 0.6573620818848988, "learning_rate": 0.00019047869181192535, "loss": 0.4375, "step": 6440 }, { "epoch": 3.0310588235294116, "grad_norm": 0.5316936063153891, "learning_rate": 0.00019043648022321793, "loss": 0.1836, "step": 6441 }, { "epoch": 3.031529411764706, "grad_norm": 0.21841557812771467, "learning_rate": 0.00019039426518103678, "loss": 0.0767, "step": 6442 }, { "epoch": 3.032, "grad_norm": 0.7287585858050226, "learning_rate": 0.00019035204668898725, "loss": 0.5039, "step": 6443 }, { "epoch": 3.0324705882352943, "grad_norm": 0.5125325421541603, "learning_rate": 0.00019030982475067498, "loss": 0.3184, "step": 6444 }, { "epoch": 3.032941176470588, "grad_norm": 0.19013038283529762, "learning_rate": 0.00019026759936970598, "loss": 0.0957, "step": 6445 }, { "epoch": 3.0334117647058823, "grad_norm": 0.7927862543612101, "learning_rate": 0.0001902253705496865, "loss": 0.3945, "step": 6446 }, { "epoch": 3.0338823529411765, "grad_norm": 0.6142272283020553, "learning_rate": 0.00019018313829422307, "loss": 0.4414, "step": 6447 }, { "epoch": 3.0343529411764707, "grad_norm": 0.30438574709425326, "learning_rate": 0.0001901409026069225, "loss": 0.1514, "step": 6448 }, { "epoch": 3.034823529411765, "grad_norm": 0.6164707857892407, "learning_rate": 0.000190098663491392, "loss": 0.375, "step": 6449 }, { "epoch": 3.0352941176470587, "grad_norm": 0.8836219274177567, "learning_rate": 0.00019005642095123895, "loss": 0.418, "step": 6450 }, { "epoch": 3.035764705882353, "grad_norm": 0.36289220411112877, "learning_rate": 0.00019001417499007103, "loss": 0.1514, "step": 6451 }, { "epoch": 3.036235294117647, "grad_norm": 0.784610661498285, "learning_rate": 0.0001899719256114963, "loss": 0.3887, "step": 6452 }, { "epoch": 3.0367058823529414, "grad_norm": 0.6169213721123376, "learning_rate": 0.00018992967281912312, "loss": 0.2305, "step": 6453 }, { "epoch": 3.037176470588235, "grad_norm": 0.6480763542684604, "learning_rate": 0.00018988741661655996, "loss": 0.3242, "step": 6454 }, { "epoch": 3.0376470588235294, "grad_norm": 0.7141874631522788, "learning_rate": 0.00018984515700741579, "loss": 0.5703, "step": 6455 }, { "epoch": 3.0381176470588236, "grad_norm": 0.4263154794186237, "learning_rate": 0.0001898028939952998, "loss": 0.2852, "step": 6456 }, { "epoch": 3.038588235294118, "grad_norm": 0.7908003665805735, "learning_rate": 0.00018976062758382144, "loss": 0.4062, "step": 6457 }, { "epoch": 3.0390588235294116, "grad_norm": 0.4203889230243344, "learning_rate": 0.0001897183577765904, "loss": 0.3008, "step": 6458 }, { "epoch": 3.039529411764706, "grad_norm": 0.4191848473077498, "learning_rate": 0.00018967608457721684, "loss": 0.1611, "step": 6459 }, { "epoch": 3.04, "grad_norm": 0.3341677817630966, "learning_rate": 0.00018963380798931102, "loss": 0.1216, "step": 6460 }, { "epoch": 3.0404705882352943, "grad_norm": 0.47214618962002874, "learning_rate": 0.00018959152801648364, "loss": 0.1206, "step": 6461 }, { "epoch": 3.040941176470588, "grad_norm": 0.5974337083132218, "learning_rate": 0.00018954924466234554, "loss": 0.2754, "step": 6462 }, { "epoch": 3.0414117647058823, "grad_norm": 0.3311826900641352, "learning_rate": 0.00018950695793050796, "loss": 0.2031, "step": 6463 }, { "epoch": 3.0418823529411765, "grad_norm": 0.5270875963106073, "learning_rate": 0.0001894646678245824, "loss": 0.1621, "step": 6464 }, { "epoch": 3.0423529411764707, "grad_norm": 0.7891700854645634, "learning_rate": 0.00018942237434818062, "loss": 0.3828, "step": 6465 }, { "epoch": 3.0428235294117645, "grad_norm": 0.45044775404753346, "learning_rate": 0.00018938007750491468, "loss": 0.2363, "step": 6466 }, { "epoch": 3.0432941176470587, "grad_norm": 0.5155932613178419, "learning_rate": 0.00018933777729839701, "loss": 0.2617, "step": 6467 }, { "epoch": 3.043764705882353, "grad_norm": 0.7816889857406982, "learning_rate": 0.0001892954737322402, "loss": 0.3418, "step": 6468 }, { "epoch": 3.044235294117647, "grad_norm": 0.6261620866152721, "learning_rate": 0.00018925316681005718, "loss": 0.4316, "step": 6469 }, { "epoch": 3.0447058823529414, "grad_norm": 0.6405708630275181, "learning_rate": 0.0001892108565354612, "loss": 0.3516, "step": 6470 }, { "epoch": 3.045176470588235, "grad_norm": 0.5571168866365444, "learning_rate": 0.0001891685429120657, "loss": 0.3242, "step": 6471 }, { "epoch": 3.0456470588235294, "grad_norm": 0.8468694123885491, "learning_rate": 0.00018912622594348454, "loss": 0.6328, "step": 6472 }, { "epoch": 3.0461176470588236, "grad_norm": 0.9565038950400039, "learning_rate": 0.00018908390563333175, "loss": 0.6211, "step": 6473 }, { "epoch": 3.046588235294118, "grad_norm": 0.6680916155926221, "learning_rate": 0.00018904158198522177, "loss": 0.4473, "step": 6474 }, { "epoch": 3.0470588235294116, "grad_norm": 0.3786947805589229, "learning_rate": 0.00018899925500276912, "loss": 0.2227, "step": 6475 }, { "epoch": 3.047529411764706, "grad_norm": 0.5705500530999708, "learning_rate": 0.00018895692468958884, "loss": 0.2354, "step": 6476 }, { "epoch": 3.048, "grad_norm": 0.520037265597708, "learning_rate": 0.00018891459104929608, "loss": 0.1787, "step": 6477 }, { "epoch": 3.0484705882352943, "grad_norm": 0.5048538659276309, "learning_rate": 0.0001888722540855064, "loss": 0.2148, "step": 6478 }, { "epoch": 3.048941176470588, "grad_norm": 0.31116228649968036, "learning_rate": 0.00018882991380183556, "loss": 0.1748, "step": 6479 }, { "epoch": 3.0494117647058823, "grad_norm": 0.8294919301947921, "learning_rate": 0.0001887875702018996, "loss": 0.416, "step": 6480 }, { "epoch": 3.0498823529411765, "grad_norm": 0.8845589677739797, "learning_rate": 0.00018874522328931492, "loss": 0.5391, "step": 6481 }, { "epoch": 3.0503529411764707, "grad_norm": 1.044011217453048, "learning_rate": 0.0001887028730676981, "loss": 0.2559, "step": 6482 }, { "epoch": 3.0508235294117645, "grad_norm": 0.5303339504545544, "learning_rate": 0.00018866051954066608, "loss": 0.1396, "step": 6483 }, { "epoch": 3.0512941176470587, "grad_norm": 0.39690493680898115, "learning_rate": 0.00018861816271183613, "loss": 0.1582, "step": 6484 }, { "epoch": 3.051764705882353, "grad_norm": 0.4378668121721406, "learning_rate": 0.00018857580258482557, "loss": 0.3652, "step": 6485 }, { "epoch": 3.052235294117647, "grad_norm": 0.5299466257147153, "learning_rate": 0.0001885334391632523, "loss": 0.377, "step": 6486 }, { "epoch": 3.0527058823529414, "grad_norm": 0.5646216313909733, "learning_rate": 0.00018849107245073432, "loss": 0.3379, "step": 6487 }, { "epoch": 3.053176470588235, "grad_norm": 0.9130884915044373, "learning_rate": 0.00018844870245088996, "loss": 0.4805, "step": 6488 }, { "epoch": 3.0536470588235294, "grad_norm": 0.6170628297532655, "learning_rate": 0.00018840632916733777, "loss": 0.3164, "step": 6489 }, { "epoch": 3.0541176470588236, "grad_norm": 0.24208205677890493, "learning_rate": 0.00018836395260369676, "loss": 0.0972, "step": 6490 }, { "epoch": 3.054588235294118, "grad_norm": 0.22070952917076309, "learning_rate": 0.000188321572763586, "loss": 0.0796, "step": 6491 }, { "epoch": 3.0550588235294116, "grad_norm": 0.2707800082979115, "learning_rate": 0.00018827918965062492, "loss": 0.0728, "step": 6492 }, { "epoch": 3.055529411764706, "grad_norm": 0.6073743023806126, "learning_rate": 0.00018823680326843334, "loss": 0.3262, "step": 6493 }, { "epoch": 3.056, "grad_norm": 0.3364596610013, "learning_rate": 0.00018819441362063122, "loss": 0.1719, "step": 6494 }, { "epoch": 3.0564705882352943, "grad_norm": 0.24543573956655704, "learning_rate": 0.00018815202071083878, "loss": 0.0859, "step": 6495 }, { "epoch": 3.056941176470588, "grad_norm": 0.7566139187609304, "learning_rate": 0.00018810962454267666, "loss": 0.2852, "step": 6496 }, { "epoch": 3.0574117647058823, "grad_norm": 0.3910026546114749, "learning_rate": 0.00018806722511976567, "loss": 0.209, "step": 6497 }, { "epoch": 3.0578823529411765, "grad_norm": 0.8424192637771343, "learning_rate": 0.00018802482244572693, "loss": 0.3906, "step": 6498 }, { "epoch": 3.0583529411764707, "grad_norm": 0.4192845806047651, "learning_rate": 0.00018798241652418182, "loss": 0.1226, "step": 6499 }, { "epoch": 3.0588235294117645, "grad_norm": 0.5249091959737395, "learning_rate": 0.00018794000735875208, "loss": 0.252, "step": 6500 }, { "epoch": 3.0592941176470587, "grad_norm": 0.6274259535235688, "learning_rate": 0.00018789759495305963, "loss": 0.4395, "step": 6501 }, { "epoch": 3.059764705882353, "grad_norm": 0.6052202023433294, "learning_rate": 0.00018785517931072668, "loss": 0.3223, "step": 6502 }, { "epoch": 3.060235294117647, "grad_norm": 0.44478323745610343, "learning_rate": 0.0001878127604353757, "loss": 0.2471, "step": 6503 }, { "epoch": 3.0607058823529414, "grad_norm": 0.630327081635497, "learning_rate": 0.00018777033833062954, "loss": 0.3281, "step": 6504 }, { "epoch": 3.061176470588235, "grad_norm": 0.43717338922379473, "learning_rate": 0.00018772791300011122, "loss": 0.2314, "step": 6505 }, { "epoch": 3.0616470588235294, "grad_norm": 0.46591374584472023, "learning_rate": 0.00018768548444744404, "loss": 0.1084, "step": 6506 }, { "epoch": 3.0621176470588236, "grad_norm": 0.7010435498046318, "learning_rate": 0.00018764305267625169, "loss": 0.3145, "step": 6507 }, { "epoch": 3.062588235294118, "grad_norm": 0.14874801648478012, "learning_rate": 0.000187600617690158, "loss": 0.0879, "step": 6508 }, { "epoch": 3.0630588235294116, "grad_norm": 0.4249815759335649, "learning_rate": 0.00018755817949278706, "loss": 0.1338, "step": 6509 }, { "epoch": 3.063529411764706, "grad_norm": 0.6250237289097489, "learning_rate": 0.00018751573808776346, "loss": 0.5312, "step": 6510 }, { "epoch": 3.064, "grad_norm": 0.620951429314564, "learning_rate": 0.00018747329347871178, "loss": 0.4355, "step": 6511 }, { "epoch": 3.0644705882352943, "grad_norm": 0.574986586679259, "learning_rate": 0.00018743084566925702, "loss": 0.2891, "step": 6512 }, { "epoch": 3.064941176470588, "grad_norm": 0.6774856798025118, "learning_rate": 0.00018738839466302443, "loss": 0.3809, "step": 6513 }, { "epoch": 3.0654117647058823, "grad_norm": 0.38834702718429814, "learning_rate": 0.00018734594046363956, "loss": 0.1523, "step": 6514 }, { "epoch": 3.0658823529411765, "grad_norm": 0.46402010837366275, "learning_rate": 0.00018730348307472824, "loss": 0.2949, "step": 6515 }, { "epoch": 3.0663529411764707, "grad_norm": 0.3423217997044851, "learning_rate": 0.0001872610224999164, "loss": 0.1758, "step": 6516 }, { "epoch": 3.0668235294117645, "grad_norm": 0.7324362948701127, "learning_rate": 0.00018721855874283058, "loss": 0.4297, "step": 6517 }, { "epoch": 3.0672941176470587, "grad_norm": 0.6292044620502845, "learning_rate": 0.00018717609180709721, "loss": 0.4004, "step": 6518 }, { "epoch": 3.067764705882353, "grad_norm": 0.5525658134298908, "learning_rate": 0.00018713362169634325, "loss": 0.3477, "step": 6519 }, { "epoch": 3.068235294117647, "grad_norm": 0.6452523353944897, "learning_rate": 0.00018709114841419588, "loss": 0.3711, "step": 6520 }, { "epoch": 3.0687058823529414, "grad_norm": 0.523344785235683, "learning_rate": 0.0001870486719642825, "loss": 0.248, "step": 6521 }, { "epoch": 3.069176470588235, "grad_norm": 0.6271030175649926, "learning_rate": 0.00018700619235023085, "loss": 0.3867, "step": 6522 }, { "epoch": 3.0696470588235294, "grad_norm": 0.6382255333165703, "learning_rate": 0.00018696370957566882, "loss": 0.3301, "step": 6523 }, { "epoch": 3.0701176470588236, "grad_norm": 0.27057635255670165, "learning_rate": 0.0001869212236442247, "loss": 0.1729, "step": 6524 }, { "epoch": 3.070588235294118, "grad_norm": 0.7859649503110561, "learning_rate": 0.00018687873455952697, "loss": 0.373, "step": 6525 }, { "epoch": 3.0710588235294116, "grad_norm": 0.29057465292926776, "learning_rate": 0.00018683624232520446, "loss": 0.1157, "step": 6526 }, { "epoch": 3.071529411764706, "grad_norm": 0.43586131520699967, "learning_rate": 0.00018679374694488608, "loss": 0.1699, "step": 6527 }, { "epoch": 3.072, "grad_norm": 0.5394876412306118, "learning_rate": 0.00018675124842220135, "loss": 0.3652, "step": 6528 }, { "epoch": 3.0724705882352943, "grad_norm": 0.40118103167797076, "learning_rate": 0.0001867087467607797, "loss": 0.1641, "step": 6529 }, { "epoch": 3.072941176470588, "grad_norm": 0.4840976366125743, "learning_rate": 0.00018666624196425098, "loss": 0.334, "step": 6530 }, { "epoch": 3.0734117647058823, "grad_norm": 0.6739131491611549, "learning_rate": 0.00018662373403624538, "loss": 0.3574, "step": 6531 }, { "epoch": 3.0738823529411765, "grad_norm": 0.5792986208958016, "learning_rate": 0.00018658122298039327, "loss": 0.4082, "step": 6532 }, { "epoch": 3.0743529411764707, "grad_norm": 0.279022779480288, "learning_rate": 0.00018653870880032527, "loss": 0.0723, "step": 6533 }, { "epoch": 3.0748235294117645, "grad_norm": 0.506254958678793, "learning_rate": 0.00018649619149967232, "loss": 0.3301, "step": 6534 }, { "epoch": 3.0752941176470587, "grad_norm": 0.6668054013414385, "learning_rate": 0.00018645367108206565, "loss": 0.3906, "step": 6535 }, { "epoch": 3.075764705882353, "grad_norm": 0.8384729395266781, "learning_rate": 0.00018641114755113658, "loss": 0.334, "step": 6536 }, { "epoch": 3.076235294117647, "grad_norm": 0.4499304194247741, "learning_rate": 0.00018636862091051697, "loss": 0.1338, "step": 6537 }, { "epoch": 3.0767058823529414, "grad_norm": 0.33331969167751296, "learning_rate": 0.00018632609116383874, "loss": 0.2285, "step": 6538 }, { "epoch": 3.077176470588235, "grad_norm": 0.31791743046233933, "learning_rate": 0.00018628355831473416, "loss": 0.0967, "step": 6539 }, { "epoch": 3.0776470588235294, "grad_norm": 0.6030167329079379, "learning_rate": 0.00018624102236683567, "loss": 0.3379, "step": 6540 }, { "epoch": 3.0781176470588236, "grad_norm": 0.8650117618385382, "learning_rate": 0.00018619848332377615, "loss": 0.5117, "step": 6541 }, { "epoch": 3.078588235294118, "grad_norm": 0.5375978330775988, "learning_rate": 0.00018615594118918862, "loss": 0.3613, "step": 6542 }, { "epoch": 3.0790588235294116, "grad_norm": 0.4745997086619456, "learning_rate": 0.00018611339596670632, "loss": 0.2988, "step": 6543 }, { "epoch": 3.079529411764706, "grad_norm": 0.409943914489747, "learning_rate": 0.00018607084765996285, "loss": 0.2812, "step": 6544 }, { "epoch": 3.08, "grad_norm": 0.5547034174551547, "learning_rate": 0.00018602829627259213, "loss": 0.3184, "step": 6545 }, { "epoch": 3.0804705882352943, "grad_norm": 0.42698842189805203, "learning_rate": 0.00018598574180822818, "loss": 0.2119, "step": 6546 }, { "epoch": 3.080941176470588, "grad_norm": 0.5244193451108793, "learning_rate": 0.00018594318427050537, "loss": 0.2969, "step": 6547 }, { "epoch": 3.0814117647058823, "grad_norm": 0.3402434017418359, "learning_rate": 0.00018590062366305828, "loss": 0.1621, "step": 6548 }, { "epoch": 3.0818823529411765, "grad_norm": 0.37907691248631775, "learning_rate": 0.0001858580599895219, "loss": 0.1885, "step": 6549 }, { "epoch": 3.0823529411764707, "grad_norm": 0.340408011704375, "learning_rate": 0.00018581549325353126, "loss": 0.1807, "step": 6550 }, { "epoch": 3.0828235294117645, "grad_norm": 0.35889342455536954, "learning_rate": 0.00018577292345872188, "loss": 0.126, "step": 6551 }, { "epoch": 3.0832941176470587, "grad_norm": 0.5553114329096284, "learning_rate": 0.00018573035060872933, "loss": 0.3242, "step": 6552 }, { "epoch": 3.083764705882353, "grad_norm": 0.472441317488486, "learning_rate": 0.0001856877747071896, "loss": 0.2295, "step": 6553 }, { "epoch": 3.084235294117647, "grad_norm": 0.7123317307491065, "learning_rate": 0.00018564519575773884, "loss": 0.3242, "step": 6554 }, { "epoch": 3.084705882352941, "grad_norm": 0.49920486053708835, "learning_rate": 0.00018560261376401356, "loss": 0.3203, "step": 6555 }, { "epoch": 3.085176470588235, "grad_norm": 0.7907396061325629, "learning_rate": 0.0001855600287296504, "loss": 0.5039, "step": 6556 }, { "epoch": 3.0856470588235294, "grad_norm": 0.641452204741701, "learning_rate": 0.00018551744065828634, "loss": 0.3809, "step": 6557 }, { "epoch": 3.0861176470588236, "grad_norm": 0.6954897528091821, "learning_rate": 0.0001854748495535587, "loss": 0.4375, "step": 6558 }, { "epoch": 3.086588235294118, "grad_norm": 0.3061487385834582, "learning_rate": 0.0001854322554191049, "loss": 0.1484, "step": 6559 }, { "epoch": 3.0870588235294116, "grad_norm": 0.7557157158588318, "learning_rate": 0.00018538965825856264, "loss": 0.5938, "step": 6560 }, { "epoch": 3.087529411764706, "grad_norm": 0.5599330857573245, "learning_rate": 0.00018534705807557003, "loss": 0.2617, "step": 6561 }, { "epoch": 3.088, "grad_norm": 0.4218518963150494, "learning_rate": 0.00018530445487376524, "loss": 0.2061, "step": 6562 }, { "epoch": 3.0884705882352943, "grad_norm": 0.8546371651566584, "learning_rate": 0.0001852618486567869, "loss": 0.582, "step": 6563 }, { "epoch": 3.088941176470588, "grad_norm": 0.4386743939998672, "learning_rate": 0.00018521923942827362, "loss": 0.208, "step": 6564 }, { "epoch": 3.0894117647058823, "grad_norm": 0.6174685311121344, "learning_rate": 0.00018517662719186464, "loss": 0.3984, "step": 6565 }, { "epoch": 3.0898823529411765, "grad_norm": 0.6943649995884215, "learning_rate": 0.0001851340119511991, "loss": 0.3574, "step": 6566 }, { "epoch": 3.0903529411764707, "grad_norm": 0.5924588819194189, "learning_rate": 0.00018509139370991666, "loss": 0.373, "step": 6567 }, { "epoch": 3.0908235294117645, "grad_norm": 0.3159610280651491, "learning_rate": 0.000185048772471657, "loss": 0.1426, "step": 6568 }, { "epoch": 3.0912941176470587, "grad_norm": 0.50956946637702, "learning_rate": 0.00018500614824006035, "loss": 0.2773, "step": 6569 }, { "epoch": 3.091764705882353, "grad_norm": 0.6762305697610481, "learning_rate": 0.0001849635210187669, "loss": 0.4727, "step": 6570 }, { "epoch": 3.092235294117647, "grad_norm": 0.4061476874271581, "learning_rate": 0.00018492089081141724, "loss": 0.1758, "step": 6571 }, { "epoch": 3.092705882352941, "grad_norm": 0.3203814386224028, "learning_rate": 0.00018487825762165223, "loss": 0.1299, "step": 6572 }, { "epoch": 3.093176470588235, "grad_norm": 0.8320434140230003, "learning_rate": 0.00018483562145311296, "loss": 0.5156, "step": 6573 }, { "epoch": 3.0936470588235294, "grad_norm": 0.5975537040403047, "learning_rate": 0.00018479298230944071, "loss": 0.1973, "step": 6574 }, { "epoch": 3.0941176470588236, "grad_norm": 0.5377307485751357, "learning_rate": 0.00018475034019427717, "loss": 0.1914, "step": 6575 }, { "epoch": 3.094588235294118, "grad_norm": 0.5842030217101107, "learning_rate": 0.00018470769511126409, "loss": 0.3477, "step": 6576 }, { "epoch": 3.0950588235294116, "grad_norm": 0.40124292791907507, "learning_rate": 0.00018466504706404362, "loss": 0.1689, "step": 6577 }, { "epoch": 3.095529411764706, "grad_norm": 0.5536287434713769, "learning_rate": 0.00018462239605625808, "loss": 0.2695, "step": 6578 }, { "epoch": 3.096, "grad_norm": 1.179465816298113, "learning_rate": 0.0001845797420915501, "loss": 0.6289, "step": 6579 }, { "epoch": 3.0964705882352943, "grad_norm": 0.7864134515287257, "learning_rate": 0.00018453708517356256, "loss": 0.8086, "step": 6580 }, { "epoch": 3.096941176470588, "grad_norm": 0.5374904740551315, "learning_rate": 0.0001844944253059385, "loss": 0.3906, "step": 6581 }, { "epoch": 3.0974117647058823, "grad_norm": 0.09036667872524483, "learning_rate": 0.00018445176249232134, "loss": 0.0144, "step": 6582 }, { "epoch": 3.0978823529411765, "grad_norm": 0.49519038212941563, "learning_rate": 0.0001844090967363547, "loss": 0.3086, "step": 6583 }, { "epoch": 3.0983529411764708, "grad_norm": 0.6470846468748344, "learning_rate": 0.0001843664280416824, "loss": 0.3105, "step": 6584 }, { "epoch": 3.0988235294117645, "grad_norm": 0.5006885230000726, "learning_rate": 0.0001843237564119486, "loss": 0.1875, "step": 6585 }, { "epoch": 3.0992941176470588, "grad_norm": 0.8787752715390501, "learning_rate": 0.00018428108185079763, "loss": 0.5703, "step": 6586 }, { "epoch": 3.099764705882353, "grad_norm": 0.44840841538828413, "learning_rate": 0.00018423840436187412, "loss": 0.1699, "step": 6587 }, { "epoch": 3.100235294117647, "grad_norm": 0.4665929048683248, "learning_rate": 0.0001841957239488229, "loss": 0.291, "step": 6588 }, { "epoch": 3.100705882352941, "grad_norm": 0.4520446838966815, "learning_rate": 0.00018415304061528918, "loss": 0.3125, "step": 6589 }, { "epoch": 3.101176470588235, "grad_norm": 0.4484106515565773, "learning_rate": 0.00018411035436491827, "loss": 0.2217, "step": 6590 }, { "epoch": 3.1016470588235294, "grad_norm": 0.5440152024024841, "learning_rate": 0.0001840676652013558, "loss": 0.3594, "step": 6591 }, { "epoch": 3.1021176470588236, "grad_norm": 0.3178179722896369, "learning_rate": 0.00018402497312824758, "loss": 0.207, "step": 6592 }, { "epoch": 3.1025882352941174, "grad_norm": 0.6052081386655035, "learning_rate": 0.0001839822781492398, "loss": 0.3828, "step": 6593 }, { "epoch": 3.1030588235294116, "grad_norm": 0.3090679675995084, "learning_rate": 0.00018393958026797878, "loss": 0.1318, "step": 6594 }, { "epoch": 3.103529411764706, "grad_norm": 0.361852808117419, "learning_rate": 0.0001838968794881111, "loss": 0.1807, "step": 6595 }, { "epoch": 3.104, "grad_norm": 1.0414123111418838, "learning_rate": 0.00018385417581328364, "loss": 0.4609, "step": 6596 }, { "epoch": 3.1044705882352943, "grad_norm": 0.4815385978782181, "learning_rate": 0.00018381146924714356, "loss": 0.1309, "step": 6597 }, { "epoch": 3.104941176470588, "grad_norm": 0.4474516870855463, "learning_rate": 0.00018376875979333812, "loss": 0.2812, "step": 6598 }, { "epoch": 3.1054117647058823, "grad_norm": 0.5501700740985166, "learning_rate": 0.000183726047455515, "loss": 0.1738, "step": 6599 }, { "epoch": 3.1058823529411765, "grad_norm": 1.0576637933739006, "learning_rate": 0.000183683332237322, "loss": 0.4668, "step": 6600 }, { "epoch": 3.1063529411764708, "grad_norm": 0.5040836639785794, "learning_rate": 0.00018364061414240717, "loss": 0.168, "step": 6601 }, { "epoch": 3.1068235294117645, "grad_norm": 0.38788094878877766, "learning_rate": 0.00018359789317441892, "loss": 0.1553, "step": 6602 }, { "epoch": 3.1072941176470588, "grad_norm": 0.6522463395975883, "learning_rate": 0.00018355516933700578, "loss": 0.3457, "step": 6603 }, { "epoch": 3.107764705882353, "grad_norm": 0.5488094814972616, "learning_rate": 0.0001835124426338166, "loss": 0.3203, "step": 6604 }, { "epoch": 3.108235294117647, "grad_norm": 0.5036272401712825, "learning_rate": 0.00018346971306850045, "loss": 0.2012, "step": 6605 }, { "epoch": 3.108705882352941, "grad_norm": 0.3013908076749158, "learning_rate": 0.00018342698064470662, "loss": 0.0806, "step": 6606 }, { "epoch": 3.109176470588235, "grad_norm": 0.6839552748820141, "learning_rate": 0.0001833842453660847, "loss": 0.2949, "step": 6607 }, { "epoch": 3.1096470588235294, "grad_norm": 0.29430078382249636, "learning_rate": 0.00018334150723628442, "loss": 0.0574, "step": 6608 }, { "epoch": 3.1101176470588237, "grad_norm": 0.19749762572296956, "learning_rate": 0.00018329876625895592, "loss": 0.0442, "step": 6609 }, { "epoch": 3.1105882352941174, "grad_norm": 0.5419031935248405, "learning_rate": 0.0001832560224377495, "loss": 0.2852, "step": 6610 }, { "epoch": 3.1110588235294117, "grad_norm": 0.7756397153971821, "learning_rate": 0.0001832132757763156, "loss": 0.4043, "step": 6611 }, { "epoch": 3.111529411764706, "grad_norm": 0.5170196999477605, "learning_rate": 0.000183170526278305, "loss": 0.2617, "step": 6612 }, { "epoch": 3.112, "grad_norm": 0.3632612821610256, "learning_rate": 0.0001831277739473688, "loss": 0.1602, "step": 6613 }, { "epoch": 3.1124705882352943, "grad_norm": 0.687108542777202, "learning_rate": 0.00018308501878715823, "loss": 0.4277, "step": 6614 }, { "epoch": 3.112941176470588, "grad_norm": 0.92038745919505, "learning_rate": 0.00018304226080132472, "loss": 0.6016, "step": 6615 }, { "epoch": 3.1134117647058823, "grad_norm": 0.6503758930183519, "learning_rate": 0.00018299949999352013, "loss": 0.3027, "step": 6616 }, { "epoch": 3.1138823529411765, "grad_norm": 0.6340260036245524, "learning_rate": 0.00018295673636739635, "loss": 0.2637, "step": 6617 }, { "epoch": 3.1143529411764708, "grad_norm": 0.5230832483993433, "learning_rate": 0.00018291396992660565, "loss": 0.3027, "step": 6618 }, { "epoch": 3.1148235294117645, "grad_norm": 0.6828325554936061, "learning_rate": 0.00018287120067480046, "loss": 0.3086, "step": 6619 }, { "epoch": 3.1152941176470588, "grad_norm": 0.7068521992936733, "learning_rate": 0.0001828284286156335, "loss": 0.4688, "step": 6620 }, { "epoch": 3.115764705882353, "grad_norm": 0.3864418183973687, "learning_rate": 0.0001827856537527577, "loss": 0.2021, "step": 6621 }, { "epoch": 3.116235294117647, "grad_norm": 0.7261467573138207, "learning_rate": 0.00018274287608982626, "loss": 0.4258, "step": 6622 }, { "epoch": 3.116705882352941, "grad_norm": 0.3887109589245598, "learning_rate": 0.00018270009563049264, "loss": 0.1338, "step": 6623 }, { "epoch": 3.117176470588235, "grad_norm": 0.4507552840697575, "learning_rate": 0.00018265731237841043, "loss": 0.1758, "step": 6624 }, { "epoch": 3.1176470588235294, "grad_norm": 0.5637967489747361, "learning_rate": 0.00018261452633723356, "loss": 0.2891, "step": 6625 }, { "epoch": 3.1181176470588237, "grad_norm": 0.5072626077184754, "learning_rate": 0.00018257173751061613, "loss": 0.2393, "step": 6626 }, { "epoch": 3.1185882352941174, "grad_norm": 0.8358664144844098, "learning_rate": 0.00018252894590221262, "loss": 0.3398, "step": 6627 }, { "epoch": 3.1190588235294117, "grad_norm": 0.5675430480864957, "learning_rate": 0.00018248615151567754, "loss": 0.3008, "step": 6628 }, { "epoch": 3.119529411764706, "grad_norm": 1.2690928950547733, "learning_rate": 0.00018244335435466577, "loss": 0.4648, "step": 6629 }, { "epoch": 3.12, "grad_norm": 0.5262931319655684, "learning_rate": 0.00018240055442283242, "loss": 0.3867, "step": 6630 }, { "epoch": 3.1204705882352943, "grad_norm": 0.6817474767439877, "learning_rate": 0.00018235775172383277, "loss": 0.3945, "step": 6631 }, { "epoch": 3.120941176470588, "grad_norm": 0.7451875035338655, "learning_rate": 0.00018231494626132243, "loss": 0.3828, "step": 6632 }, { "epoch": 3.1214117647058823, "grad_norm": 0.33587018737145974, "learning_rate": 0.00018227213803895717, "loss": 0.123, "step": 6633 }, { "epoch": 3.1218823529411766, "grad_norm": 0.3921729552231156, "learning_rate": 0.00018222932706039307, "loss": 0.2236, "step": 6634 }, { "epoch": 3.1223529411764708, "grad_norm": 0.8041168665627388, "learning_rate": 0.0001821865133292863, "loss": 0.4434, "step": 6635 }, { "epoch": 3.1228235294117646, "grad_norm": 0.5403388590451066, "learning_rate": 0.00018214369684929347, "loss": 0.3203, "step": 6636 }, { "epoch": 3.1232941176470588, "grad_norm": 0.5782098220952173, "learning_rate": 0.00018210087762407122, "loss": 0.4121, "step": 6637 }, { "epoch": 3.123764705882353, "grad_norm": 0.46534987209262246, "learning_rate": 0.00018205805565727658, "loss": 0.2217, "step": 6638 }, { "epoch": 3.124235294117647, "grad_norm": 0.6068388713565591, "learning_rate": 0.00018201523095256672, "loss": 0.4004, "step": 6639 }, { "epoch": 3.124705882352941, "grad_norm": 0.6074199192585185, "learning_rate": 0.00018197240351359913, "loss": 0.3789, "step": 6640 }, { "epoch": 3.1251764705882352, "grad_norm": 0.3140392674767089, "learning_rate": 0.0001819295733440315, "loss": 0.053, "step": 6641 }, { "epoch": 3.1256470588235294, "grad_norm": 0.6296386786921557, "learning_rate": 0.00018188674044752163, "loss": 0.3574, "step": 6642 }, { "epoch": 3.1261176470588237, "grad_norm": 0.367245918692605, "learning_rate": 0.00018184390482772774, "loss": 0.2559, "step": 6643 }, { "epoch": 3.1265882352941174, "grad_norm": 0.7353043548354321, "learning_rate": 0.0001818010664883082, "loss": 0.4395, "step": 6644 }, { "epoch": 3.1270588235294117, "grad_norm": 0.6840961916585133, "learning_rate": 0.0001817582254329216, "loss": 0.2695, "step": 6645 }, { "epoch": 3.127529411764706, "grad_norm": 0.28857488901926326, "learning_rate": 0.0001817153816652267, "loss": 0.085, "step": 6646 }, { "epoch": 3.128, "grad_norm": 0.4737635667861489, "learning_rate": 0.0001816725351888827, "loss": 0.1206, "step": 6647 }, { "epoch": 3.1284705882352943, "grad_norm": 0.49202499333429134, "learning_rate": 0.00018162968600754888, "loss": 0.3066, "step": 6648 }, { "epoch": 3.128941176470588, "grad_norm": 0.394641241103853, "learning_rate": 0.00018158683412488466, "loss": 0.1689, "step": 6649 }, { "epoch": 3.1294117647058823, "grad_norm": 0.952562204738647, "learning_rate": 0.00018154397954454993, "loss": 0.5234, "step": 6650 }, { "epoch": 3.1298823529411766, "grad_norm": 0.7539740846916329, "learning_rate": 0.0001815011222702046, "loss": 0.457, "step": 6651 }, { "epoch": 3.130352941176471, "grad_norm": 0.45654088239803914, "learning_rate": 0.00018145826230550892, "loss": 0.2178, "step": 6652 }, { "epoch": 3.1308235294117646, "grad_norm": 0.09594111435276581, "learning_rate": 0.00018141539965412334, "loss": 0.0194, "step": 6653 }, { "epoch": 3.131294117647059, "grad_norm": 0.8234995837584868, "learning_rate": 0.00018137253431970853, "loss": 0.4062, "step": 6654 }, { "epoch": 3.131764705882353, "grad_norm": 0.3743487466430164, "learning_rate": 0.0001813296663059254, "loss": 0.1992, "step": 6655 }, { "epoch": 3.1322352941176472, "grad_norm": 0.7421339979184501, "learning_rate": 0.00018128679561643508, "loss": 0.5195, "step": 6656 }, { "epoch": 3.132705882352941, "grad_norm": 0.6725673889072117, "learning_rate": 0.000181243922254899, "loss": 0.3086, "step": 6657 }, { "epoch": 3.1331764705882352, "grad_norm": 0.4515125154064136, "learning_rate": 0.0001812010462249787, "loss": 0.2852, "step": 6658 }, { "epoch": 3.1336470588235295, "grad_norm": 0.2743708903450769, "learning_rate": 0.00018115816753033601, "loss": 0.0903, "step": 6659 }, { "epoch": 3.1341176470588237, "grad_norm": 0.6380717917976693, "learning_rate": 0.00018111528617463296, "loss": 0.3926, "step": 6660 }, { "epoch": 3.1345882352941175, "grad_norm": 0.8013447560143737, "learning_rate": 0.00018107240216153184, "loss": 0.3633, "step": 6661 }, { "epoch": 3.1350588235294117, "grad_norm": 0.7920835938723196, "learning_rate": 0.00018102951549469522, "loss": 0.4707, "step": 6662 }, { "epoch": 3.135529411764706, "grad_norm": 0.4642635307306665, "learning_rate": 0.00018098662617778573, "loss": 0.3457, "step": 6663 }, { "epoch": 3.136, "grad_norm": 0.43934921221832973, "learning_rate": 0.0001809437342144664, "loss": 0.1689, "step": 6664 }, { "epoch": 3.1364705882352943, "grad_norm": 0.3379914920196547, "learning_rate": 0.00018090083960840042, "loss": 0.1235, "step": 6665 }, { "epoch": 3.136941176470588, "grad_norm": 0.6054350399817963, "learning_rate": 0.0001808579423632511, "loss": 0.3594, "step": 6666 }, { "epoch": 3.1374117647058823, "grad_norm": 0.6990018509791817, "learning_rate": 0.0001808150424826822, "loss": 0.3418, "step": 6667 }, { "epoch": 3.1378823529411766, "grad_norm": 0.4851711274188685, "learning_rate": 0.0001807721399703575, "loss": 0.2461, "step": 6668 }, { "epoch": 3.138352941176471, "grad_norm": 0.42026131736707595, "learning_rate": 0.0001807292348299411, "loss": 0.2227, "step": 6669 }, { "epoch": 3.1388235294117646, "grad_norm": 1.317016794273592, "learning_rate": 0.00018068632706509733, "loss": 0.4355, "step": 6670 }, { "epoch": 3.139294117647059, "grad_norm": 0.6123726977748295, "learning_rate": 0.00018064341667949074, "loss": 0.334, "step": 6671 }, { "epoch": 3.139764705882353, "grad_norm": 0.40190786462125017, "learning_rate": 0.00018060050367678607, "loss": 0.248, "step": 6672 }, { "epoch": 3.1402352941176472, "grad_norm": 0.2867163181316533, "learning_rate": 0.00018055758806064822, "loss": 0.103, "step": 6673 }, { "epoch": 3.140705882352941, "grad_norm": 0.25355159349807055, "learning_rate": 0.00018051466983474252, "loss": 0.0564, "step": 6674 }, { "epoch": 3.1411764705882352, "grad_norm": 0.3381316080838937, "learning_rate": 0.00018047174900273435, "loss": 0.103, "step": 6675 }, { "epoch": 3.1416470588235295, "grad_norm": 0.604602171561247, "learning_rate": 0.00018042882556828927, "loss": 0.4336, "step": 6676 }, { "epoch": 3.1421176470588237, "grad_norm": 0.4637464967247812, "learning_rate": 0.0001803858995350733, "loss": 0.2227, "step": 6677 }, { "epoch": 3.1425882352941175, "grad_norm": 0.5525084673461044, "learning_rate": 0.00018034297090675246, "loss": 0.3086, "step": 6678 }, { "epoch": 3.1430588235294117, "grad_norm": 0.5255010610239805, "learning_rate": 0.00018030003968699308, "loss": 0.2363, "step": 6679 }, { "epoch": 3.143529411764706, "grad_norm": 0.736002172011312, "learning_rate": 0.00018025710587946166, "loss": 0.2207, "step": 6680 }, { "epoch": 3.144, "grad_norm": 0.32499427345079984, "learning_rate": 0.00018021416948782503, "loss": 0.1807, "step": 6681 }, { "epoch": 3.1444705882352944, "grad_norm": 0.3158131037402795, "learning_rate": 0.00018017123051575012, "loss": 0.0972, "step": 6682 }, { "epoch": 3.144941176470588, "grad_norm": 0.6021985007846347, "learning_rate": 0.0001801282889669041, "loss": 0.334, "step": 6683 }, { "epoch": 3.1454117647058824, "grad_norm": 0.6537964757017659, "learning_rate": 0.00018008534484495445, "loss": 0.4785, "step": 6684 }, { "epoch": 3.1458823529411766, "grad_norm": 0.5861601783491763, "learning_rate": 0.0001800423981535688, "loss": 0.3066, "step": 6685 }, { "epoch": 3.1463529411764704, "grad_norm": 0.6969169865009395, "learning_rate": 0.000179999448896415, "loss": 0.3828, "step": 6686 }, { "epoch": 3.1468235294117646, "grad_norm": 0.6005443170891864, "learning_rate": 0.00017995649707716103, "loss": 0.3477, "step": 6687 }, { "epoch": 3.147294117647059, "grad_norm": 0.46863612409780137, "learning_rate": 0.00017991354269947533, "loss": 0.1914, "step": 6688 }, { "epoch": 3.147764705882353, "grad_norm": 0.3233873271660857, "learning_rate": 0.00017987058576702638, "loss": 0.1338, "step": 6689 }, { "epoch": 3.1482352941176472, "grad_norm": 0.6592973193503557, "learning_rate": 0.00017982762628348286, "loss": 0.3203, "step": 6690 }, { "epoch": 3.148705882352941, "grad_norm": 0.40598724630438565, "learning_rate": 0.00017978466425251375, "loss": 0.1367, "step": 6691 }, { "epoch": 3.1491764705882352, "grad_norm": 0.29741737900570253, "learning_rate": 0.00017974169967778825, "loss": 0.2354, "step": 6692 }, { "epoch": 3.1496470588235295, "grad_norm": 0.2934710377304927, "learning_rate": 0.00017969873256297572, "loss": 0.0972, "step": 6693 }, { "epoch": 3.1501176470588237, "grad_norm": 0.34571506705311783, "learning_rate": 0.0001796557629117457, "loss": 0.1533, "step": 6694 }, { "epoch": 3.1505882352941175, "grad_norm": 0.8281247408205242, "learning_rate": 0.00017961279072776812, "loss": 0.4473, "step": 6695 }, { "epoch": 3.1510588235294117, "grad_norm": 0.36881894505090457, "learning_rate": 0.00017956981601471294, "loss": 0.3262, "step": 6696 }, { "epoch": 3.151529411764706, "grad_norm": 0.6282637287330378, "learning_rate": 0.00017952683877625042, "loss": 0.2578, "step": 6697 }, { "epoch": 3.152, "grad_norm": 0.256854547432011, "learning_rate": 0.00017948385901605106, "loss": 0.0703, "step": 6698 }, { "epoch": 3.152470588235294, "grad_norm": 0.5365493489891253, "learning_rate": 0.0001794408767377855, "loss": 0.4023, "step": 6699 }, { "epoch": 3.152941176470588, "grad_norm": 0.7038293024357521, "learning_rate": 0.00017939789194512472, "loss": 0.3574, "step": 6700 }, { "epoch": 3.1534117647058824, "grad_norm": 0.4688010204686515, "learning_rate": 0.0001793549046417397, "loss": 0.3086, "step": 6701 }, { "epoch": 3.1538823529411766, "grad_norm": 0.2262749125765161, "learning_rate": 0.00017931191483130186, "loss": 0.0552, "step": 6702 }, { "epoch": 3.1543529411764704, "grad_norm": 0.5489204172496697, "learning_rate": 0.00017926892251748274, "loss": 0.2197, "step": 6703 }, { "epoch": 3.1548235294117646, "grad_norm": 0.3714203183958291, "learning_rate": 0.00017922592770395405, "loss": 0.2139, "step": 6704 }, { "epoch": 3.155294117647059, "grad_norm": 0.3680071444005107, "learning_rate": 0.00017918293039438781, "loss": 0.165, "step": 6705 }, { "epoch": 3.155764705882353, "grad_norm": 0.45846426352703357, "learning_rate": 0.00017913993059245616, "loss": 0.249, "step": 6706 }, { "epoch": 3.1562352941176472, "grad_norm": 0.49648876612975384, "learning_rate": 0.00017909692830183152, "loss": 0.1377, "step": 6707 }, { "epoch": 3.156705882352941, "grad_norm": 0.6807966243443124, "learning_rate": 0.00017905392352618647, "loss": 0.5391, "step": 6708 }, { "epoch": 3.1571764705882353, "grad_norm": 0.4781745465607009, "learning_rate": 0.00017901091626919383, "loss": 0.2793, "step": 6709 }, { "epoch": 3.1576470588235295, "grad_norm": 0.9094368108097561, "learning_rate": 0.00017896790653452668, "loss": 0.5938, "step": 6710 }, { "epoch": 3.1581176470588237, "grad_norm": 0.29960183203934704, "learning_rate": 0.00017892489432585817, "loss": 0.064, "step": 6711 }, { "epoch": 3.1585882352941175, "grad_norm": 0.27718565064019884, "learning_rate": 0.00017888187964686182, "loss": 0.0942, "step": 6712 }, { "epoch": 3.1590588235294117, "grad_norm": 0.37556743010045296, "learning_rate": 0.00017883886250121133, "loss": 0.1904, "step": 6713 }, { "epoch": 3.159529411764706, "grad_norm": 0.7443234846390908, "learning_rate": 0.00017879584289258048, "loss": 0.3008, "step": 6714 }, { "epoch": 3.16, "grad_norm": 0.4143152120198416, "learning_rate": 0.00017875282082464343, "loss": 0.2363, "step": 6715 }, { "epoch": 3.160470588235294, "grad_norm": 0.5314640105639981, "learning_rate": 0.00017870979630107446, "loss": 0.2451, "step": 6716 }, { "epoch": 3.160941176470588, "grad_norm": 0.37517879630637757, "learning_rate": 0.00017866676932554802, "loss": 0.165, "step": 6717 }, { "epoch": 3.1614117647058824, "grad_norm": 0.5357170717057718, "learning_rate": 0.0001786237399017389, "loss": 0.25, "step": 6718 }, { "epoch": 3.1618823529411766, "grad_norm": 0.37861194396459247, "learning_rate": 0.000178580708033322, "loss": 0.2139, "step": 6719 }, { "epoch": 3.1623529411764704, "grad_norm": 0.15529913889494326, "learning_rate": 0.00017853767372397243, "loss": 0.042, "step": 6720 }, { "epoch": 3.1628235294117646, "grad_norm": 0.4855494892841346, "learning_rate": 0.00017849463697736554, "loss": 0.2656, "step": 6721 }, { "epoch": 3.163294117647059, "grad_norm": 0.8046632347744088, "learning_rate": 0.0001784515977971769, "loss": 0.4023, "step": 6722 }, { "epoch": 3.163764705882353, "grad_norm": 0.17793562026534931, "learning_rate": 0.00017840855618708227, "loss": 0.0459, "step": 6723 }, { "epoch": 3.1642352941176473, "grad_norm": 0.6589789753845029, "learning_rate": 0.00017836551215075758, "loss": 0.3145, "step": 6724 }, { "epoch": 3.164705882352941, "grad_norm": 0.6149758910877448, "learning_rate": 0.000178322465691879, "loss": 0.3516, "step": 6725 }, { "epoch": 3.1651764705882353, "grad_norm": 0.5254785306052644, "learning_rate": 0.00017827941681412298, "loss": 0.2598, "step": 6726 }, { "epoch": 3.1656470588235295, "grad_norm": 0.4279424804229367, "learning_rate": 0.00017823636552116603, "loss": 0.1367, "step": 6727 }, { "epoch": 3.1661176470588237, "grad_norm": 0.6198846991749661, "learning_rate": 0.00017819331181668496, "loss": 0.2119, "step": 6728 }, { "epoch": 3.1665882352941175, "grad_norm": 0.5246618963693881, "learning_rate": 0.00017815025570435677, "loss": 0.2314, "step": 6729 }, { "epoch": 3.1670588235294117, "grad_norm": 0.6925292297909943, "learning_rate": 0.0001781071971878587, "loss": 0.4219, "step": 6730 }, { "epoch": 3.167529411764706, "grad_norm": 0.34972474682851223, "learning_rate": 0.00017806413627086807, "loss": 0.1484, "step": 6731 }, { "epoch": 3.168, "grad_norm": 0.3698151282284785, "learning_rate": 0.0001780210729570626, "loss": 0.1562, "step": 6732 }, { "epoch": 3.168470588235294, "grad_norm": 0.6681731733765094, "learning_rate": 0.00017797800725012002, "loss": 0.3984, "step": 6733 }, { "epoch": 3.168941176470588, "grad_norm": 0.7492928566537592, "learning_rate": 0.00017793493915371842, "loss": 0.3438, "step": 6734 }, { "epoch": 3.1694117647058824, "grad_norm": 0.5177459738606585, "learning_rate": 0.00017789186867153594, "loss": 0.2715, "step": 6735 }, { "epoch": 3.1698823529411766, "grad_norm": 0.615042943286852, "learning_rate": 0.00017784879580725113, "loss": 0.3105, "step": 6736 }, { "epoch": 3.1703529411764704, "grad_norm": 0.5503711180273864, "learning_rate": 0.00017780572056454256, "loss": 0.3633, "step": 6737 }, { "epoch": 3.1708235294117646, "grad_norm": 0.4766372001799862, "learning_rate": 0.00017776264294708905, "loss": 0.1963, "step": 6738 }, { "epoch": 3.171294117647059, "grad_norm": 0.5999746677833615, "learning_rate": 0.00017771956295856964, "loss": 0.291, "step": 6739 }, { "epoch": 3.171764705882353, "grad_norm": 0.5378481006058509, "learning_rate": 0.0001776764806026636, "loss": 0.3398, "step": 6740 }, { "epoch": 3.1722352941176473, "grad_norm": 0.08551066389849342, "learning_rate": 0.00017763339588305037, "loss": 0.051, "step": 6741 }, { "epoch": 3.172705882352941, "grad_norm": 0.5796369884944175, "learning_rate": 0.00017759030880340956, "loss": 0.3672, "step": 6742 }, { "epoch": 3.1731764705882353, "grad_norm": 0.07779329260070475, "learning_rate": 0.00017754721936742108, "loss": 0.01, "step": 6743 }, { "epoch": 3.1736470588235295, "grad_norm": 1.1433746461718646, "learning_rate": 0.00017750412757876492, "loss": 0.4609, "step": 6744 }, { "epoch": 3.1741176470588237, "grad_norm": 0.3873139755632322, "learning_rate": 0.00017746103344112136, "loss": 0.1157, "step": 6745 }, { "epoch": 3.1745882352941175, "grad_norm": 0.6045129507654463, "learning_rate": 0.00017741793695817093, "loss": 0.3301, "step": 6746 }, { "epoch": 3.1750588235294117, "grad_norm": 0.5016101669745572, "learning_rate": 0.00017737483813359413, "loss": 0.1689, "step": 6747 }, { "epoch": 3.175529411764706, "grad_norm": 0.40743750518957833, "learning_rate": 0.00017733173697107187, "loss": 0.1001, "step": 6748 }, { "epoch": 3.176, "grad_norm": 0.2992025937128683, "learning_rate": 0.00017728863347428524, "loss": 0.1494, "step": 6749 }, { "epoch": 3.176470588235294, "grad_norm": 0.1910096621707912, "learning_rate": 0.00017724552764691545, "loss": 0.0508, "step": 6750 }, { "epoch": 3.176941176470588, "grad_norm": 0.7429919622338792, "learning_rate": 0.000177202419492644, "loss": 0.2471, "step": 6751 }, { "epoch": 3.1774117647058824, "grad_norm": 0.8513679142178058, "learning_rate": 0.00017715930901515242, "loss": 0.2832, "step": 6752 }, { "epoch": 3.1778823529411766, "grad_norm": 0.9022658798629385, "learning_rate": 0.0001771161962181227, "loss": 0.4102, "step": 6753 }, { "epoch": 3.1783529411764704, "grad_norm": 0.6072575945728165, "learning_rate": 0.0001770730811052368, "loss": 0.3809, "step": 6754 }, { "epoch": 3.1788235294117646, "grad_norm": 0.42165605256762245, "learning_rate": 0.00017702996368017693, "loss": 0.2754, "step": 6755 }, { "epoch": 3.179294117647059, "grad_norm": 0.26259834637683765, "learning_rate": 0.00017698684394662563, "loss": 0.0747, "step": 6756 }, { "epoch": 3.179764705882353, "grad_norm": 1.0588857327421637, "learning_rate": 0.0001769437219082655, "loss": 0.5508, "step": 6757 }, { "epoch": 3.1802352941176473, "grad_norm": 0.5254365238030652, "learning_rate": 0.00017690059756877933, "loss": 0.3691, "step": 6758 }, { "epoch": 3.180705882352941, "grad_norm": 0.6222034101551099, "learning_rate": 0.00017685747093185017, "loss": 0.1885, "step": 6759 }, { "epoch": 3.1811764705882353, "grad_norm": 0.732536897699304, "learning_rate": 0.00017681434200116132, "loss": 0.3633, "step": 6760 }, { "epoch": 3.1816470588235295, "grad_norm": 0.5646689254772959, "learning_rate": 0.0001767712107803961, "loss": 0.2119, "step": 6761 }, { "epoch": 3.1821176470588237, "grad_norm": 0.41918515438012455, "learning_rate": 0.00017672807727323818, "loss": 0.2168, "step": 6762 }, { "epoch": 3.1825882352941175, "grad_norm": 0.047055931694626324, "learning_rate": 0.00017668494148337133, "loss": 0.0048, "step": 6763 }, { "epoch": 3.1830588235294117, "grad_norm": 0.5888167153050828, "learning_rate": 0.00017664180341447964, "loss": 0.4258, "step": 6764 }, { "epoch": 3.183529411764706, "grad_norm": 0.34893333215538475, "learning_rate": 0.0001765986630702472, "loss": 0.1787, "step": 6765 }, { "epoch": 3.184, "grad_norm": 0.45547172383486195, "learning_rate": 0.00017655552045435846, "loss": 0.1943, "step": 6766 }, { "epoch": 3.184470588235294, "grad_norm": 0.48705732088517584, "learning_rate": 0.00017651237557049808, "loss": 0.3066, "step": 6767 }, { "epoch": 3.184941176470588, "grad_norm": 0.8388934058691507, "learning_rate": 0.00017646922842235076, "loss": 0.5234, "step": 6768 }, { "epoch": 3.1854117647058824, "grad_norm": 0.401733439272016, "learning_rate": 0.00017642607901360147, "loss": 0.2236, "step": 6769 }, { "epoch": 3.1858823529411766, "grad_norm": 0.8182706275717412, "learning_rate": 0.00017638292734793547, "loss": 0.582, "step": 6770 }, { "epoch": 3.1863529411764704, "grad_norm": 0.812303281579024, "learning_rate": 0.00017633977342903807, "loss": 0.377, "step": 6771 }, { "epoch": 3.1868235294117646, "grad_norm": 0.6391668279267578, "learning_rate": 0.0001762966172605948, "loss": 0.3164, "step": 6772 }, { "epoch": 3.187294117647059, "grad_norm": 0.5309622562645622, "learning_rate": 0.00017625345884629141, "loss": 0.2734, "step": 6773 }, { "epoch": 3.187764705882353, "grad_norm": 0.5240981785117075, "learning_rate": 0.00017621029818981387, "loss": 0.3496, "step": 6774 }, { "epoch": 3.1882352941176473, "grad_norm": 0.482031101282759, "learning_rate": 0.00017616713529484835, "loss": 0.1123, "step": 6775 }, { "epoch": 3.188705882352941, "grad_norm": 0.38095689291577967, "learning_rate": 0.00017612397016508104, "loss": 0.1279, "step": 6776 }, { "epoch": 3.1891764705882353, "grad_norm": 1.2379595907463803, "learning_rate": 0.0001760808028041986, "loss": 0.8047, "step": 6777 }, { "epoch": 3.1896470588235295, "grad_norm": 0.5758630954272957, "learning_rate": 0.0001760376332158877, "loss": 0.2852, "step": 6778 }, { "epoch": 3.1901176470588237, "grad_norm": 0.32643339592586335, "learning_rate": 0.00017599446140383512, "loss": 0.1021, "step": 6779 }, { "epoch": 3.1905882352941175, "grad_norm": 0.7413574596697369, "learning_rate": 0.00017595128737172812, "loss": 0.3047, "step": 6780 }, { "epoch": 3.1910588235294117, "grad_norm": 0.3325851966528811, "learning_rate": 0.00017590811112325384, "loss": 0.1099, "step": 6781 }, { "epoch": 3.191529411764706, "grad_norm": 0.6224641921273695, "learning_rate": 0.00017586493266209978, "loss": 0.1572, "step": 6782 }, { "epoch": 3.192, "grad_norm": 0.7138238265363944, "learning_rate": 0.00017582175199195365, "loss": 0.2734, "step": 6783 }, { "epoch": 3.192470588235294, "grad_norm": 0.8483718376419619, "learning_rate": 0.00017577856911650324, "loss": 0.5078, "step": 6784 }, { "epoch": 3.192941176470588, "grad_norm": 0.4275837073238545, "learning_rate": 0.00017573538403943656, "loss": 0.1875, "step": 6785 }, { "epoch": 3.1934117647058824, "grad_norm": 0.49852977607280735, "learning_rate": 0.00017569219676444187, "loss": 0.1885, "step": 6786 }, { "epoch": 3.1938823529411766, "grad_norm": 0.5472234951141147, "learning_rate": 0.00017564900729520751, "loss": 0.4746, "step": 6787 }, { "epoch": 3.1943529411764704, "grad_norm": 0.7493452442170432, "learning_rate": 0.0001756058156354222, "loss": 0.459, "step": 6788 }, { "epoch": 3.1948235294117646, "grad_norm": 1.037472358023225, "learning_rate": 0.00017556262178877459, "loss": 0.7539, "step": 6789 }, { "epoch": 3.195294117647059, "grad_norm": 0.7007148251413611, "learning_rate": 0.00017551942575895368, "loss": 0.3984, "step": 6790 }, { "epoch": 3.195764705882353, "grad_norm": 0.42005508876224285, "learning_rate": 0.00017547622754964868, "loss": 0.1299, "step": 6791 }, { "epoch": 3.1962352941176473, "grad_norm": 0.3909734976581166, "learning_rate": 0.0001754330271645489, "loss": 0.0216, "step": 6792 }, { "epoch": 3.196705882352941, "grad_norm": 0.21058748051032933, "learning_rate": 0.00017538982460734375, "loss": 0.0732, "step": 6793 }, { "epoch": 3.1971764705882353, "grad_norm": 0.6749989481916979, "learning_rate": 0.00017534661988172315, "loss": 0.4609, "step": 6794 }, { "epoch": 3.1976470588235295, "grad_norm": 0.48569514540454106, "learning_rate": 0.00017530341299137685, "loss": 0.2461, "step": 6795 }, { "epoch": 3.1981176470588237, "grad_norm": 1.0510638655971132, "learning_rate": 0.00017526020393999496, "loss": 0.6484, "step": 6796 }, { "epoch": 3.1985882352941175, "grad_norm": 0.24222260628150025, "learning_rate": 0.00017521699273126778, "loss": 0.1201, "step": 6797 }, { "epoch": 3.1990588235294117, "grad_norm": 0.624227719723349, "learning_rate": 0.00017517377936888568, "loss": 0.4316, "step": 6798 }, { "epoch": 3.199529411764706, "grad_norm": 0.3438180409729147, "learning_rate": 0.0001751305638565394, "loss": 0.1582, "step": 6799 }, { "epoch": 3.2, "grad_norm": 0.6867275433438039, "learning_rate": 0.00017508734619791966, "loss": 0.5, "step": 6800 }, { "epoch": 3.200470588235294, "grad_norm": 0.18857064605897625, "learning_rate": 0.00017504412639671754, "loss": 0.062, "step": 6801 }, { "epoch": 3.200941176470588, "grad_norm": 0.0652479181639932, "learning_rate": 0.00017500090445662415, "loss": 0.0126, "step": 6802 }, { "epoch": 3.2014117647058824, "grad_norm": 0.6020036564945873, "learning_rate": 0.0001749576803813309, "loss": 0.2637, "step": 6803 }, { "epoch": 3.2018823529411766, "grad_norm": 0.6085893570969655, "learning_rate": 0.00017491445417452934, "loss": 0.3477, "step": 6804 }, { "epoch": 3.2023529411764704, "grad_norm": 0.6506028334787615, "learning_rate": 0.00017487122583991123, "loss": 0.4375, "step": 6805 }, { "epoch": 3.2028235294117646, "grad_norm": 0.3451786564106098, "learning_rate": 0.00017482799538116834, "loss": 0.1328, "step": 6806 }, { "epoch": 3.203294117647059, "grad_norm": 0.39930684780814457, "learning_rate": 0.00017478476280199294, "loss": 0.1934, "step": 6807 }, { "epoch": 3.203764705882353, "grad_norm": 0.8426730580307555, "learning_rate": 0.0001747415281060772, "loss": 0.5977, "step": 6808 }, { "epoch": 3.2042352941176473, "grad_norm": 1.0510849078240638, "learning_rate": 0.00017469829129711361, "loss": 0.6172, "step": 6809 }, { "epoch": 3.204705882352941, "grad_norm": 0.2084558684838963, "learning_rate": 0.00017465505237879476, "loss": 0.0603, "step": 6810 }, { "epoch": 3.2051764705882353, "grad_norm": 0.5655583164384346, "learning_rate": 0.00017461181135481355, "loss": 0.3242, "step": 6811 }, { "epoch": 3.2056470588235295, "grad_norm": 0.6848972259638202, "learning_rate": 0.00017456856822886297, "loss": 0.2832, "step": 6812 }, { "epoch": 3.2061176470588233, "grad_norm": 0.9693165335421285, "learning_rate": 0.00017452532300463609, "loss": 0.6484, "step": 6813 }, { "epoch": 3.2065882352941175, "grad_norm": 0.579385791345723, "learning_rate": 0.00017448207568582634, "loss": 0.3105, "step": 6814 }, { "epoch": 3.2070588235294117, "grad_norm": 0.5005573921825478, "learning_rate": 0.00017443882627612726, "loss": 0.2656, "step": 6815 }, { "epoch": 3.207529411764706, "grad_norm": 0.504362255625395, "learning_rate": 0.00017439557477923255, "loss": 0.167, "step": 6816 }, { "epoch": 3.208, "grad_norm": 0.6142570743246842, "learning_rate": 0.00017435232119883604, "loss": 0.2871, "step": 6817 }, { "epoch": 3.208470588235294, "grad_norm": 0.2620838501212683, "learning_rate": 0.0001743090655386319, "loss": 0.1035, "step": 6818 }, { "epoch": 3.208941176470588, "grad_norm": 1.036188022092125, "learning_rate": 0.00017426580780231435, "loss": 0.6055, "step": 6819 }, { "epoch": 3.2094117647058824, "grad_norm": 0.8079578550053337, "learning_rate": 0.0001742225479935777, "loss": 0.4629, "step": 6820 }, { "epoch": 3.2098823529411766, "grad_norm": 0.5730670948574654, "learning_rate": 0.00017417928611611675, "loss": 0.2793, "step": 6821 }, { "epoch": 3.2103529411764704, "grad_norm": 0.5404782276264514, "learning_rate": 0.0001741360221736261, "loss": 0.3086, "step": 6822 }, { "epoch": 3.2108235294117646, "grad_norm": 0.4287884056172611, "learning_rate": 0.0001740927561698008, "loss": 0.1416, "step": 6823 }, { "epoch": 3.211294117647059, "grad_norm": 0.5485221621524594, "learning_rate": 0.00017404948810833594, "loss": 0.4473, "step": 6824 }, { "epoch": 3.211764705882353, "grad_norm": 0.89809236050684, "learning_rate": 0.0001740062179929268, "loss": 0.5234, "step": 6825 }, { "epoch": 3.212235294117647, "grad_norm": 0.37634067780633895, "learning_rate": 0.00017396294582726896, "loss": 0.1348, "step": 6826 }, { "epoch": 3.212705882352941, "grad_norm": 0.49570877524695833, "learning_rate": 0.00017391967161505795, "loss": 0.2793, "step": 6827 }, { "epoch": 3.2131764705882353, "grad_norm": 0.299076247949998, "learning_rate": 0.00017387639535998967, "loss": 0.0771, "step": 6828 }, { "epoch": 3.2136470588235295, "grad_norm": 0.5705248935724611, "learning_rate": 0.00017383311706576016, "loss": 0.21, "step": 6829 }, { "epoch": 3.2141176470588233, "grad_norm": 0.8888541176436354, "learning_rate": 0.00017378983673606548, "loss": 0.3965, "step": 6830 }, { "epoch": 3.2145882352941175, "grad_norm": 0.8321340897787249, "learning_rate": 0.00017374655437460206, "loss": 0.332, "step": 6831 }, { "epoch": 3.2150588235294117, "grad_norm": 0.6941992951042073, "learning_rate": 0.00017370326998506646, "loss": 0.3828, "step": 6832 }, { "epoch": 3.215529411764706, "grad_norm": 0.6750860199618905, "learning_rate": 0.00017365998357115535, "loss": 0.1865, "step": 6833 }, { "epoch": 3.216, "grad_norm": 0.5229036914004273, "learning_rate": 0.00017361669513656558, "loss": 0.1943, "step": 6834 }, { "epoch": 3.216470588235294, "grad_norm": 0.7682084989782689, "learning_rate": 0.0001735734046849942, "loss": 0.3926, "step": 6835 }, { "epoch": 3.216941176470588, "grad_norm": 1.0094546694777944, "learning_rate": 0.0001735301122201384, "loss": 0.3359, "step": 6836 }, { "epoch": 3.2174117647058824, "grad_norm": 0.6332714587435833, "learning_rate": 0.00017348681774569563, "loss": 0.3223, "step": 6837 }, { "epoch": 3.2178823529411766, "grad_norm": 0.5219996844865402, "learning_rate": 0.00017344352126536343, "loss": 0.2334, "step": 6838 }, { "epoch": 3.2183529411764704, "grad_norm": 0.32121999366010023, "learning_rate": 0.00017340022278283953, "loss": 0.1108, "step": 6839 }, { "epoch": 3.2188235294117646, "grad_norm": 0.6673365585101071, "learning_rate": 0.00017335692230182183, "loss": 0.2051, "step": 6840 }, { "epoch": 3.219294117647059, "grad_norm": 0.6950816835893752, "learning_rate": 0.00017331361982600836, "loss": 0.3887, "step": 6841 }, { "epoch": 3.219764705882353, "grad_norm": 0.514598257574992, "learning_rate": 0.00017327031535909745, "loss": 0.2617, "step": 6842 }, { "epoch": 3.220235294117647, "grad_norm": 0.49278849719353, "learning_rate": 0.00017322700890478744, "loss": 0.2412, "step": 6843 }, { "epoch": 3.220705882352941, "grad_norm": 0.7946560432999632, "learning_rate": 0.00017318370046677693, "loss": 0.3984, "step": 6844 }, { "epoch": 3.2211764705882353, "grad_norm": 0.4533025040900146, "learning_rate": 0.00017314039004876473, "loss": 0.2266, "step": 6845 }, { "epoch": 3.2216470588235295, "grad_norm": 0.36276568481699106, "learning_rate": 0.0001730970776544497, "loss": 0.1855, "step": 6846 }, { "epoch": 3.2221176470588233, "grad_norm": 0.6112844415162689, "learning_rate": 0.00017305376328753092, "loss": 0.4355, "step": 6847 }, { "epoch": 3.2225882352941175, "grad_norm": 0.29818903096525906, "learning_rate": 0.00017301044695170769, "loss": 0.1221, "step": 6848 }, { "epoch": 3.2230588235294118, "grad_norm": 0.05181989021164563, "learning_rate": 0.00017296712865067944, "loss": 0.0156, "step": 6849 }, { "epoch": 3.223529411764706, "grad_norm": 0.5377164619134239, "learning_rate": 0.00017292380838814577, "loss": 0.2793, "step": 6850 }, { "epoch": 3.224, "grad_norm": 0.6219017849091516, "learning_rate": 0.00017288048616780636, "loss": 0.3359, "step": 6851 }, { "epoch": 3.224470588235294, "grad_norm": 0.8353770948901248, "learning_rate": 0.00017283716199336123, "loss": 0.332, "step": 6852 }, { "epoch": 3.224941176470588, "grad_norm": 0.3168315550715862, "learning_rate": 0.0001727938358685105, "loss": 0.1455, "step": 6853 }, { "epoch": 3.2254117647058824, "grad_norm": 0.8654082030035508, "learning_rate": 0.0001727505077969543, "loss": 0.4961, "step": 6854 }, { "epoch": 3.2258823529411766, "grad_norm": 0.58232265421113, "learning_rate": 0.00017270717778239316, "loss": 0.3457, "step": 6855 }, { "epoch": 3.2263529411764704, "grad_norm": 0.6055189927874267, "learning_rate": 0.0001726638458285277, "loss": 0.2734, "step": 6856 }, { "epoch": 3.2268235294117646, "grad_norm": 0.8382435629692799, "learning_rate": 0.00017262051193905865, "loss": 0.5664, "step": 6857 }, { "epoch": 3.227294117647059, "grad_norm": 0.3690793527739182, "learning_rate": 0.00017257717611768685, "loss": 0.166, "step": 6858 }, { "epoch": 3.227764705882353, "grad_norm": 0.29056173413023834, "learning_rate": 0.00017253383836811356, "loss": 0.1177, "step": 6859 }, { "epoch": 3.228235294117647, "grad_norm": 0.1845076205227799, "learning_rate": 0.0001724904986940399, "loss": 0.0415, "step": 6860 }, { "epoch": 3.228705882352941, "grad_norm": 0.2823070162686618, "learning_rate": 0.00017244715709916733, "loss": 0.1182, "step": 6861 }, { "epoch": 3.2291764705882353, "grad_norm": 0.6355941833549317, "learning_rate": 0.00017240381358719746, "loss": 0.377, "step": 6862 }, { "epoch": 3.2296470588235295, "grad_norm": 0.06629618380840277, "learning_rate": 0.000172360468161832, "loss": 0.0096, "step": 6863 }, { "epoch": 3.2301176470588233, "grad_norm": 0.3518273960081721, "learning_rate": 0.00017231712082677291, "loss": 0.1367, "step": 6864 }, { "epoch": 3.2305882352941175, "grad_norm": 0.5879772659000315, "learning_rate": 0.0001722737715857222, "loss": 0.3516, "step": 6865 }, { "epoch": 3.2310588235294118, "grad_norm": 0.32241787838789193, "learning_rate": 0.00017223042044238219, "loss": 0.0708, "step": 6866 }, { "epoch": 3.231529411764706, "grad_norm": 0.21254105789141725, "learning_rate": 0.0001721870674004552, "loss": 0.0579, "step": 6867 }, { "epoch": 3.232, "grad_norm": 1.188551063142079, "learning_rate": 0.00017214371246364384, "loss": 0.4141, "step": 6868 }, { "epoch": 3.232470588235294, "grad_norm": 0.431409389841431, "learning_rate": 0.00017210035563565082, "loss": 0.1709, "step": 6869 }, { "epoch": 3.232941176470588, "grad_norm": 0.48021061033938156, "learning_rate": 0.00017205699692017902, "loss": 0.125, "step": 6870 }, { "epoch": 3.2334117647058824, "grad_norm": 0.565779378123247, "learning_rate": 0.0001720136363209315, "loss": 0.3711, "step": 6871 }, { "epoch": 3.2338823529411767, "grad_norm": 0.5005906660819961, "learning_rate": 0.0001719702738416115, "loss": 0.1738, "step": 6872 }, { "epoch": 3.2343529411764704, "grad_norm": 0.6546779802828723, "learning_rate": 0.0001719269094859223, "loss": 0.5742, "step": 6873 }, { "epoch": 3.2348235294117647, "grad_norm": 0.3342925160154493, "learning_rate": 0.0001718835432575675, "loss": 0.2188, "step": 6874 }, { "epoch": 3.235294117647059, "grad_norm": 0.4273755417020273, "learning_rate": 0.00017184017516025076, "loss": 0.1562, "step": 6875 }, { "epoch": 3.235764705882353, "grad_norm": 0.4753322175826824, "learning_rate": 0.0001717968051976759, "loss": 0.207, "step": 6876 }, { "epoch": 3.236235294117647, "grad_norm": 0.526580324679593, "learning_rate": 0.00017175343337354698, "loss": 0.3301, "step": 6877 }, { "epoch": 3.236705882352941, "grad_norm": 0.39848458150166194, "learning_rate": 0.00017171005969156817, "loss": 0.106, "step": 6878 }, { "epoch": 3.2371764705882353, "grad_norm": 0.3365346201258115, "learning_rate": 0.00017166668415544376, "loss": 0.2598, "step": 6879 }, { "epoch": 3.2376470588235295, "grad_norm": 0.5882681807241691, "learning_rate": 0.0001716233067688783, "loss": 0.2656, "step": 6880 }, { "epoch": 3.2381176470588233, "grad_norm": 0.6798750556332334, "learning_rate": 0.00017157992753557633, "loss": 0.3301, "step": 6881 }, { "epoch": 3.2385882352941175, "grad_norm": 0.6756826481014455, "learning_rate": 0.0001715365464592427, "loss": 0.3711, "step": 6882 }, { "epoch": 3.2390588235294118, "grad_norm": 0.3589039898658324, "learning_rate": 0.00017149316354358237, "loss": 0.1963, "step": 6883 }, { "epoch": 3.239529411764706, "grad_norm": 0.4790617951126278, "learning_rate": 0.00017144977879230046, "loss": 0.165, "step": 6884 }, { "epoch": 3.24, "grad_norm": 0.30898589248423297, "learning_rate": 0.00017140639220910222, "loss": 0.1777, "step": 6885 }, { "epoch": 3.240470588235294, "grad_norm": 0.5395734194464411, "learning_rate": 0.00017136300379769312, "loss": 0.1934, "step": 6886 }, { "epoch": 3.240941176470588, "grad_norm": 0.5602152820248494, "learning_rate": 0.00017131961356177869, "loss": 0.2832, "step": 6887 }, { "epoch": 3.2414117647058824, "grad_norm": 0.4597195099753366, "learning_rate": 0.00017127622150506472, "loss": 0.249, "step": 6888 }, { "epoch": 3.2418823529411767, "grad_norm": 0.4310711649009983, "learning_rate": 0.000171232827631257, "loss": 0.252, "step": 6889 }, { "epoch": 3.2423529411764704, "grad_norm": 0.8244720835583491, "learning_rate": 0.00017118943194406169, "loss": 0.3867, "step": 6890 }, { "epoch": 3.2428235294117647, "grad_norm": 0.6094629506026558, "learning_rate": 0.00017114603444718496, "loss": 0.3438, "step": 6891 }, { "epoch": 3.243294117647059, "grad_norm": 0.6367068779225122, "learning_rate": 0.00017110263514433317, "loss": 0.3262, "step": 6892 }, { "epoch": 3.243764705882353, "grad_norm": 0.43205232970402196, "learning_rate": 0.00017105923403921283, "loss": 0.2236, "step": 6893 }, { "epoch": 3.244235294117647, "grad_norm": 0.9776792319734134, "learning_rate": 0.00017101583113553063, "loss": 0.377, "step": 6894 }, { "epoch": 3.244705882352941, "grad_norm": 0.5494457791832611, "learning_rate": 0.00017097242643699333, "loss": 0.3125, "step": 6895 }, { "epoch": 3.2451764705882353, "grad_norm": 0.7936269289816381, "learning_rate": 0.00017092901994730797, "loss": 0.3281, "step": 6896 }, { "epoch": 3.2456470588235296, "grad_norm": 0.47800729780981727, "learning_rate": 0.00017088561167018164, "loss": 0.1787, "step": 6897 }, { "epoch": 3.2461176470588233, "grad_norm": 0.45608432619223793, "learning_rate": 0.0001708422016093216, "loss": 0.1592, "step": 6898 }, { "epoch": 3.2465882352941176, "grad_norm": 0.7868409212036618, "learning_rate": 0.00017079878976843533, "loss": 0.4043, "step": 6899 }, { "epoch": 3.2470588235294118, "grad_norm": 0.7626070074399368, "learning_rate": 0.00017075537615123042, "loss": 0.375, "step": 6900 }, { "epoch": 3.247529411764706, "grad_norm": 0.28490357568698443, "learning_rate": 0.0001707119607614146, "loss": 0.0874, "step": 6901 }, { "epoch": 3.248, "grad_norm": 0.5910625051592984, "learning_rate": 0.0001706685436026957, "loss": 0.2432, "step": 6902 }, { "epoch": 3.248470588235294, "grad_norm": 0.323862429918688, "learning_rate": 0.0001706251246787818, "loss": 0.1484, "step": 6903 }, { "epoch": 3.248941176470588, "grad_norm": 0.6699404528016536, "learning_rate": 0.0001705817039933811, "loss": 0.3906, "step": 6904 }, { "epoch": 3.2494117647058824, "grad_norm": 0.7375782477950431, "learning_rate": 0.00017053828155020193, "loss": 0.5391, "step": 6905 }, { "epoch": 3.2498823529411767, "grad_norm": 0.49842662079864125, "learning_rate": 0.00017049485735295275, "loss": 0.2793, "step": 6906 }, { "epoch": 3.2503529411764704, "grad_norm": 0.28803217236531775, "learning_rate": 0.00017045143140534226, "loss": 0.1001, "step": 6907 }, { "epoch": 3.2508235294117647, "grad_norm": 0.6500071581548732, "learning_rate": 0.0001704080037110792, "loss": 0.2178, "step": 6908 }, { "epoch": 3.251294117647059, "grad_norm": 0.6766324739740346, "learning_rate": 0.00017036457427387248, "loss": 0.4355, "step": 6909 }, { "epoch": 3.251764705882353, "grad_norm": 0.4760894470835952, "learning_rate": 0.00017032114309743128, "loss": 0.3594, "step": 6910 }, { "epoch": 3.252235294117647, "grad_norm": 0.34458551135943677, "learning_rate": 0.0001702777101854648, "loss": 0.1943, "step": 6911 }, { "epoch": 3.252705882352941, "grad_norm": 0.12959016557650185, "learning_rate": 0.0001702342755416824, "loss": 0.0796, "step": 6912 }, { "epoch": 3.2531764705882353, "grad_norm": 0.5310409364293628, "learning_rate": 0.00017019083916979357, "loss": 0.1348, "step": 6913 }, { "epoch": 3.2536470588235296, "grad_norm": 0.5488209428204385, "learning_rate": 0.0001701474010735081, "loss": 0.2773, "step": 6914 }, { "epoch": 3.2541176470588233, "grad_norm": 0.7652332143760782, "learning_rate": 0.00017010396125653572, "loss": 0.5664, "step": 6915 }, { "epoch": 3.2545882352941176, "grad_norm": 0.6074598138546924, "learning_rate": 0.00017006051972258645, "loss": 0.3711, "step": 6916 }, { "epoch": 3.255058823529412, "grad_norm": 0.7067718454354676, "learning_rate": 0.0001700170764753704, "loss": 0.3379, "step": 6917 }, { "epoch": 3.255529411764706, "grad_norm": 1.1080382666254822, "learning_rate": 0.00016997363151859785, "loss": 0.7031, "step": 6918 }, { "epoch": 3.2560000000000002, "grad_norm": 0.4703501087142581, "learning_rate": 0.00016993018485597915, "loss": 0.2988, "step": 6919 }, { "epoch": 3.256470588235294, "grad_norm": 0.31894473939851314, "learning_rate": 0.00016988673649122492, "loss": 0.1514, "step": 6920 }, { "epoch": 3.2569411764705882, "grad_norm": 0.29734082230193637, "learning_rate": 0.00016984328642804587, "loss": 0.0559, "step": 6921 }, { "epoch": 3.2574117647058825, "grad_norm": 0.4741152138303745, "learning_rate": 0.00016979983467015284, "loss": 0.1758, "step": 6922 }, { "epoch": 3.2578823529411762, "grad_norm": 0.3847853889623287, "learning_rate": 0.00016975638122125678, "loss": 0.1426, "step": 6923 }, { "epoch": 3.2583529411764705, "grad_norm": 0.4004365063696311, "learning_rate": 0.00016971292608506888, "loss": 0.1465, "step": 6924 }, { "epoch": 3.2588235294117647, "grad_norm": 0.7240964939636311, "learning_rate": 0.0001696694692653004, "loss": 0.2598, "step": 6925 }, { "epoch": 3.259294117647059, "grad_norm": 0.4985336472504809, "learning_rate": 0.00016962601076566275, "loss": 0.1836, "step": 6926 }, { "epoch": 3.259764705882353, "grad_norm": 0.5746848225601944, "learning_rate": 0.00016958255058986753, "loss": 0.334, "step": 6927 }, { "epoch": 3.260235294117647, "grad_norm": 0.9793918018039877, "learning_rate": 0.00016953908874162645, "loss": 0.6172, "step": 6928 }, { "epoch": 3.260705882352941, "grad_norm": 1.0003141127831714, "learning_rate": 0.00016949562522465135, "loss": 0.4531, "step": 6929 }, { "epoch": 3.2611764705882353, "grad_norm": 0.726846375272455, "learning_rate": 0.00016945216004265418, "loss": 0.3672, "step": 6930 }, { "epoch": 3.2616470588235296, "grad_norm": 0.3519494706347084, "learning_rate": 0.0001694086931993472, "loss": 0.1562, "step": 6931 }, { "epoch": 3.2621176470588233, "grad_norm": 0.37659151580424405, "learning_rate": 0.0001693652246984426, "loss": 0.1895, "step": 6932 }, { "epoch": 3.2625882352941176, "grad_norm": 0.4593187958810782, "learning_rate": 0.0001693217545436528, "loss": 0.1475, "step": 6933 }, { "epoch": 3.263058823529412, "grad_norm": 0.5674308238226998, "learning_rate": 0.0001692782827386904, "loss": 0.2539, "step": 6934 }, { "epoch": 3.263529411764706, "grad_norm": 0.2799133959749482, "learning_rate": 0.00016923480928726813, "loss": 0.0781, "step": 6935 }, { "epoch": 3.2640000000000002, "grad_norm": 0.21865647547910572, "learning_rate": 0.00016919133419309875, "loss": 0.0552, "step": 6936 }, { "epoch": 3.264470588235294, "grad_norm": 0.651483972129275, "learning_rate": 0.00016914785745989532, "loss": 0.3926, "step": 6937 }, { "epoch": 3.2649411764705882, "grad_norm": 0.21234935913186864, "learning_rate": 0.00016910437909137098, "loss": 0.0767, "step": 6938 }, { "epoch": 3.2654117647058825, "grad_norm": 0.5509430920755304, "learning_rate": 0.00016906089909123892, "loss": 0.3574, "step": 6939 }, { "epoch": 3.2658823529411762, "grad_norm": 0.4152046908181496, "learning_rate": 0.00016901741746321253, "loss": 0.3418, "step": 6940 }, { "epoch": 3.2663529411764705, "grad_norm": 0.5898159292692694, "learning_rate": 0.0001689739342110055, "loss": 0.2949, "step": 6941 }, { "epoch": 3.2668235294117647, "grad_norm": 0.44417077490898915, "learning_rate": 0.00016893044933833135, "loss": 0.2656, "step": 6942 }, { "epoch": 3.267294117647059, "grad_norm": 0.5609629368760666, "learning_rate": 0.00016888696284890402, "loss": 0.2383, "step": 6943 }, { "epoch": 3.267764705882353, "grad_norm": 0.6558061233548746, "learning_rate": 0.0001688434747464374, "loss": 0.4082, "step": 6944 }, { "epoch": 3.268235294117647, "grad_norm": 0.6437474009542737, "learning_rate": 0.00016879998503464561, "loss": 0.3984, "step": 6945 }, { "epoch": 3.268705882352941, "grad_norm": 0.6860645582121062, "learning_rate": 0.00016875649371724292, "loss": 0.3574, "step": 6946 }, { "epoch": 3.2691764705882354, "grad_norm": 0.16138716288635932, "learning_rate": 0.00016871300079794363, "loss": 0.0269, "step": 6947 }, { "epoch": 3.2696470588235296, "grad_norm": 0.65953473983688, "learning_rate": 0.0001686695062804623, "loss": 0.3828, "step": 6948 }, { "epoch": 3.2701176470588234, "grad_norm": 0.44984605944311806, "learning_rate": 0.00016862601016851358, "loss": 0.2275, "step": 6949 }, { "epoch": 3.2705882352941176, "grad_norm": 0.4716222697809442, "learning_rate": 0.00016858251246581216, "loss": 0.2314, "step": 6950 }, { "epoch": 3.271058823529412, "grad_norm": 0.4624191988554584, "learning_rate": 0.0001685390131760731, "loss": 0.3184, "step": 6951 }, { "epoch": 3.271529411764706, "grad_norm": 0.5905987073017802, "learning_rate": 0.00016849551230301135, "loss": 0.3418, "step": 6952 }, { "epoch": 3.2720000000000002, "grad_norm": 0.7612945538094761, "learning_rate": 0.00016845200985034215, "loss": 0.373, "step": 6953 }, { "epoch": 3.272470588235294, "grad_norm": 0.8755372244303153, "learning_rate": 0.0001684085058217808, "loss": 0.5234, "step": 6954 }, { "epoch": 3.2729411764705882, "grad_norm": 0.8911979180059927, "learning_rate": 0.00016836500022104274, "loss": 0.4434, "step": 6955 }, { "epoch": 3.2734117647058825, "grad_norm": 0.16293411983127107, "learning_rate": 0.00016832149305184363, "loss": 0.0388, "step": 6956 }, { "epoch": 3.2738823529411762, "grad_norm": 0.3392912053246539, "learning_rate": 0.00016827798431789907, "loss": 0.2158, "step": 6957 }, { "epoch": 3.2743529411764705, "grad_norm": 0.8953470226796757, "learning_rate": 0.00016823447402292504, "loss": 0.2754, "step": 6958 }, { "epoch": 3.2748235294117647, "grad_norm": 1.8165477009791327, "learning_rate": 0.00016819096217063752, "loss": 0.2891, "step": 6959 }, { "epoch": 3.275294117647059, "grad_norm": 0.9857378726662468, "learning_rate": 0.00016814744876475252, "loss": 0.3066, "step": 6960 }, { "epoch": 3.275764705882353, "grad_norm": 0.43054394806477086, "learning_rate": 0.00016810393380898645, "loss": 0.293, "step": 6961 }, { "epoch": 3.276235294117647, "grad_norm": 0.4623679285561159, "learning_rate": 0.0001680604173070556, "loss": 0.2266, "step": 6962 }, { "epoch": 3.276705882352941, "grad_norm": 0.5222715595469296, "learning_rate": 0.00016801689926267653, "loss": 0.249, "step": 6963 }, { "epoch": 3.2771764705882354, "grad_norm": 0.9648238187113897, "learning_rate": 0.00016797337967956592, "loss": 0.6641, "step": 6964 }, { "epoch": 3.2776470588235296, "grad_norm": 0.4304226388315321, "learning_rate": 0.00016792985856144049, "loss": 0.2949, "step": 6965 }, { "epoch": 3.2781176470588234, "grad_norm": 0.6099260620015938, "learning_rate": 0.00016788633591201723, "loss": 0.2578, "step": 6966 }, { "epoch": 3.2785882352941176, "grad_norm": 0.4658452135573616, "learning_rate": 0.0001678428117350131, "loss": 0.165, "step": 6967 }, { "epoch": 3.279058823529412, "grad_norm": 0.4736142653092246, "learning_rate": 0.0001677992860341454, "loss": 0.1973, "step": 6968 }, { "epoch": 3.279529411764706, "grad_norm": 0.7592135513424719, "learning_rate": 0.00016775575881313132, "loss": 0.4531, "step": 6969 }, { "epoch": 3.2800000000000002, "grad_norm": 0.7356661808331539, "learning_rate": 0.00016771223007568836, "loss": 0.4082, "step": 6970 }, { "epoch": 3.280470588235294, "grad_norm": 0.19908551700952293, "learning_rate": 0.00016766869982553405, "loss": 0.0476, "step": 6971 }, { "epoch": 3.2809411764705882, "grad_norm": 0.39970312386068785, "learning_rate": 0.0001676251680663861, "loss": 0.1953, "step": 6972 }, { "epoch": 3.2814117647058825, "grad_norm": 0.3945212298980273, "learning_rate": 0.0001675816348019624, "loss": 0.1318, "step": 6973 }, { "epoch": 3.2818823529411763, "grad_norm": 0.4342003375267216, "learning_rate": 0.0001675381000359808, "loss": 0.2148, "step": 6974 }, { "epoch": 3.2823529411764705, "grad_norm": 0.4817098618756812, "learning_rate": 0.0001674945637721595, "loss": 0.2051, "step": 6975 }, { "epoch": 3.2828235294117647, "grad_norm": 0.24028118915139968, "learning_rate": 0.00016745102601421663, "loss": 0.1338, "step": 6976 }, { "epoch": 3.283294117647059, "grad_norm": 0.6193576755381418, "learning_rate": 0.00016740748676587045, "loss": 0.3281, "step": 6977 }, { "epoch": 3.283764705882353, "grad_norm": 0.6990319749133708, "learning_rate": 0.0001673639460308396, "loss": 0.6406, "step": 6978 }, { "epoch": 3.284235294117647, "grad_norm": 0.3861371501852891, "learning_rate": 0.00016732040381284258, "loss": 0.252, "step": 6979 }, { "epoch": 3.284705882352941, "grad_norm": 0.7899446224670873, "learning_rate": 0.00016727686011559813, "loss": 0.6875, "step": 6980 }, { "epoch": 3.2851764705882354, "grad_norm": 0.3022640780350071, "learning_rate": 0.00016723331494282502, "loss": 0.1465, "step": 6981 }, { "epoch": 3.2856470588235296, "grad_norm": 0.9937713263340326, "learning_rate": 0.00016718976829824234, "loss": 0.6719, "step": 6982 }, { "epoch": 3.2861176470588234, "grad_norm": 0.6967464490135789, "learning_rate": 0.00016714622018556915, "loss": 0.3301, "step": 6983 }, { "epoch": 3.2865882352941176, "grad_norm": 0.3773774704352534, "learning_rate": 0.0001671026706085246, "loss": 0.1953, "step": 6984 }, { "epoch": 3.287058823529412, "grad_norm": 0.2912296335403466, "learning_rate": 0.0001670591195708281, "loss": 0.1094, "step": 6985 }, { "epoch": 3.287529411764706, "grad_norm": 0.39563621934080284, "learning_rate": 0.00016701556707619914, "loss": 0.1895, "step": 6986 }, { "epoch": 3.288, "grad_norm": 0.3270996422476466, "learning_rate": 0.00016697201312835727, "loss": 0.1484, "step": 6987 }, { "epoch": 3.288470588235294, "grad_norm": 0.7789001450125531, "learning_rate": 0.00016692845773102222, "loss": 0.3672, "step": 6988 }, { "epoch": 3.2889411764705883, "grad_norm": 1.068851823931102, "learning_rate": 0.0001668849008879139, "loss": 0.6211, "step": 6989 }, { "epoch": 3.2894117647058825, "grad_norm": 0.2188075471139014, "learning_rate": 0.00016684134260275217, "loss": 0.0874, "step": 6990 }, { "epoch": 3.2898823529411763, "grad_norm": 0.7665150797578224, "learning_rate": 0.0001667977828792572, "loss": 0.3594, "step": 6991 }, { "epoch": 3.2903529411764705, "grad_norm": 0.9782220121168077, "learning_rate": 0.00016675422172114917, "loss": 0.7266, "step": 6992 }, { "epoch": 3.2908235294117647, "grad_norm": 0.5530960079114378, "learning_rate": 0.00016671065913214845, "loss": 0.1846, "step": 6993 }, { "epoch": 3.291294117647059, "grad_norm": 1.0321147962363653, "learning_rate": 0.00016666709511597546, "loss": 0.6797, "step": 6994 }, { "epoch": 3.291764705882353, "grad_norm": 0.42511721603184444, "learning_rate": 0.0001666235296763508, "loss": 0.1846, "step": 6995 }, { "epoch": 3.292235294117647, "grad_norm": 0.5266494562419805, "learning_rate": 0.00016657996281699514, "loss": 0.2852, "step": 6996 }, { "epoch": 3.292705882352941, "grad_norm": 0.38314310116060757, "learning_rate": 0.00016653639454162938, "loss": 0.1504, "step": 6997 }, { "epoch": 3.2931764705882354, "grad_norm": 0.6331914078161207, "learning_rate": 0.00016649282485397437, "loss": 0.3184, "step": 6998 }, { "epoch": 3.2936470588235296, "grad_norm": 0.4422960484471317, "learning_rate": 0.00016644925375775124, "loss": 0.2119, "step": 6999 }, { "epoch": 3.2941176470588234, "grad_norm": 0.6652545151540153, "learning_rate": 0.00016640568125668117, "loss": 0.2617, "step": 7000 }, { "epoch": 3.2945882352941176, "grad_norm": 0.42817793619757544, "learning_rate": 0.00016636210735448539, "loss": 0.127, "step": 7001 }, { "epoch": 3.295058823529412, "grad_norm": 0.4157423961599269, "learning_rate": 0.00016631853205488546, "loss": 0.0981, "step": 7002 }, { "epoch": 3.295529411764706, "grad_norm": 2.574699002767094, "learning_rate": 0.00016627495536160284, "loss": 0.127, "step": 7003 }, { "epoch": 3.296, "grad_norm": 0.41714552938493077, "learning_rate": 0.0001662313772783592, "loss": 0.1279, "step": 7004 }, { "epoch": 3.296470588235294, "grad_norm": 1.1196650110464867, "learning_rate": 0.00016618779780887627, "loss": 0.293, "step": 7005 }, { "epoch": 3.2969411764705883, "grad_norm": 0.408493756724447, "learning_rate": 0.00016614421695687606, "loss": 0.1309, "step": 7006 }, { "epoch": 3.2974117647058825, "grad_norm": 0.1208108905264505, "learning_rate": 0.00016610063472608053, "loss": 0.0225, "step": 7007 }, { "epoch": 3.2978823529411763, "grad_norm": 0.6511940091291445, "learning_rate": 0.0001660570511202118, "loss": 0.293, "step": 7008 }, { "epoch": 3.2983529411764705, "grad_norm": 0.8883927446660347, "learning_rate": 0.00016601346614299214, "loss": 0.4766, "step": 7009 }, { "epoch": 3.2988235294117647, "grad_norm": 0.32289483151123316, "learning_rate": 0.00016596987979814395, "loss": 0.082, "step": 7010 }, { "epoch": 3.299294117647059, "grad_norm": 0.12961226885138533, "learning_rate": 0.0001659262920893897, "loss": 0.0427, "step": 7011 }, { "epoch": 3.299764705882353, "grad_norm": 1.255512299199102, "learning_rate": 0.00016588270302045197, "loss": 0.8281, "step": 7012 }, { "epoch": 3.300235294117647, "grad_norm": 0.5016377570192497, "learning_rate": 0.00016583911259505352, "loss": 0.2871, "step": 7013 }, { "epoch": 3.300705882352941, "grad_norm": 0.2692493846140139, "learning_rate": 0.00016579552081691715, "loss": 0.0913, "step": 7014 }, { "epoch": 3.3011764705882354, "grad_norm": 0.8905146276648879, "learning_rate": 0.0001657519276897658, "loss": 0.377, "step": 7015 }, { "epoch": 3.3016470588235296, "grad_norm": 0.86458748006218, "learning_rate": 0.0001657083332173226, "loss": 0.4004, "step": 7016 }, { "epoch": 3.3021176470588234, "grad_norm": 0.49527634278267985, "learning_rate": 0.0001656647374033107, "loss": 0.2598, "step": 7017 }, { "epoch": 3.3025882352941176, "grad_norm": 0.2487661120734922, "learning_rate": 0.00016562114025145332, "loss": 0.0703, "step": 7018 }, { "epoch": 3.303058823529412, "grad_norm": 0.3961337510872942, "learning_rate": 0.00016557754176547404, "loss": 0.085, "step": 7019 }, { "epoch": 3.303529411764706, "grad_norm": 0.3263959679651903, "learning_rate": 0.0001655339419490962, "loss": 0.2119, "step": 7020 }, { "epoch": 3.304, "grad_norm": 0.6078836923531182, "learning_rate": 0.00016549034080604359, "loss": 0.3789, "step": 7021 }, { "epoch": 3.304470588235294, "grad_norm": 0.48761773375370854, "learning_rate": 0.00016544673834003985, "loss": 0.1357, "step": 7022 }, { "epoch": 3.3049411764705883, "grad_norm": 0.3570128627712842, "learning_rate": 0.00016540313455480895, "loss": 0.207, "step": 7023 }, { "epoch": 3.3054117647058825, "grad_norm": 0.41688091004112066, "learning_rate": 0.00016535952945407473, "loss": 0.1631, "step": 7024 }, { "epoch": 3.3058823529411763, "grad_norm": 0.5707074966788677, "learning_rate": 0.00016531592304156137, "loss": 0.2676, "step": 7025 }, { "epoch": 3.3063529411764705, "grad_norm": 0.2726739796524973, "learning_rate": 0.0001652723153209931, "loss": 0.0991, "step": 7026 }, { "epoch": 3.3068235294117647, "grad_norm": 0.5151405544242853, "learning_rate": 0.0001652287062960942, "loss": 0.209, "step": 7027 }, { "epoch": 3.307294117647059, "grad_norm": 0.8369315995030719, "learning_rate": 0.00016518509597058902, "loss": 0.5625, "step": 7028 }, { "epoch": 3.307764705882353, "grad_norm": 0.8411126905385553, "learning_rate": 0.0001651414843482022, "loss": 0.2773, "step": 7029 }, { "epoch": 3.308235294117647, "grad_norm": 0.5809754265323965, "learning_rate": 0.00016509787143265836, "loss": 0.2119, "step": 7030 }, { "epoch": 3.308705882352941, "grad_norm": 0.2624858128584193, "learning_rate": 0.0001650542572276822, "loss": 0.0747, "step": 7031 }, { "epoch": 3.3091764705882354, "grad_norm": 0.3881522758352401, "learning_rate": 0.00016501064173699866, "loss": 0.1582, "step": 7032 }, { "epoch": 3.309647058823529, "grad_norm": 0.7776871426149641, "learning_rate": 0.0001649670249643327, "loss": 0.4453, "step": 7033 }, { "epoch": 3.3101176470588234, "grad_norm": 0.3353862913676695, "learning_rate": 0.0001649234069134094, "loss": 0.1235, "step": 7034 }, { "epoch": 3.3105882352941176, "grad_norm": 0.39056473355072513, "learning_rate": 0.00016487978758795397, "loss": 0.1177, "step": 7035 }, { "epoch": 3.311058823529412, "grad_norm": 0.7786364563931695, "learning_rate": 0.00016483616699169163, "loss": 0.6211, "step": 7036 }, { "epoch": 3.311529411764706, "grad_norm": 0.4335072608604567, "learning_rate": 0.0001647925451283479, "loss": 0.2178, "step": 7037 }, { "epoch": 3.312, "grad_norm": 0.5488814080330358, "learning_rate": 0.00016474892200164825, "loss": 0.208, "step": 7038 }, { "epoch": 3.312470588235294, "grad_norm": 0.5346862742659451, "learning_rate": 0.0001647052976153183, "loss": 0.293, "step": 7039 }, { "epoch": 3.3129411764705883, "grad_norm": 0.4409288843200718, "learning_rate": 0.00016466167197308384, "loss": 0.2246, "step": 7040 }, { "epoch": 3.3134117647058825, "grad_norm": 0.43949399019750723, "learning_rate": 0.00016461804507867067, "loss": 0.25, "step": 7041 }, { "epoch": 3.3138823529411763, "grad_norm": 0.5928678756287546, "learning_rate": 0.00016457441693580472, "loss": 0.2832, "step": 7042 }, { "epoch": 3.3143529411764705, "grad_norm": 0.393906818440926, "learning_rate": 0.0001645307875482121, "loss": 0.1738, "step": 7043 }, { "epoch": 3.3148235294117647, "grad_norm": 0.5906995099811206, "learning_rate": 0.00016448715691961898, "loss": 0.2314, "step": 7044 }, { "epoch": 3.315294117647059, "grad_norm": 0.43003867439215937, "learning_rate": 0.00016444352505375155, "loss": 0.1758, "step": 7045 }, { "epoch": 3.315764705882353, "grad_norm": 0.3983772640513409, "learning_rate": 0.00016439989195433626, "loss": 0.106, "step": 7046 }, { "epoch": 3.316235294117647, "grad_norm": 0.7335682712955142, "learning_rate": 0.00016435625762509955, "loss": 0.4668, "step": 7047 }, { "epoch": 3.316705882352941, "grad_norm": 0.37012358832207737, "learning_rate": 0.00016431262206976805, "loss": 0.1963, "step": 7048 }, { "epoch": 3.3171764705882354, "grad_norm": 0.542449316997818, "learning_rate": 0.00016426898529206836, "loss": 0.3672, "step": 7049 }, { "epoch": 3.317647058823529, "grad_norm": 0.42200149911747864, "learning_rate": 0.00016422534729572738, "loss": 0.1582, "step": 7050 }, { "epoch": 3.3181176470588234, "grad_norm": 0.4380235361999844, "learning_rate": 0.00016418170808447198, "loss": 0.1885, "step": 7051 }, { "epoch": 3.3185882352941176, "grad_norm": 0.22315478473609343, "learning_rate": 0.00016413806766202912, "loss": 0.0593, "step": 7052 }, { "epoch": 3.319058823529412, "grad_norm": 0.7452276012842978, "learning_rate": 0.00016409442603212594, "loss": 0.3809, "step": 7053 }, { "epoch": 3.319529411764706, "grad_norm": 0.3502166720462001, "learning_rate": 0.00016405078319848965, "loss": 0.1196, "step": 7054 }, { "epoch": 3.32, "grad_norm": 0.29401520149278115, "learning_rate": 0.0001640071391648476, "loss": 0.1553, "step": 7055 }, { "epoch": 3.320470588235294, "grad_norm": 0.5823470996065537, "learning_rate": 0.0001639634939349271, "loss": 0.2969, "step": 7056 }, { "epoch": 3.3209411764705883, "grad_norm": 0.72446894808298, "learning_rate": 0.00016391984751245572, "loss": 0.3242, "step": 7057 }, { "epoch": 3.3214117647058825, "grad_norm": 0.3231629476704146, "learning_rate": 0.00016387619990116116, "loss": 0.0923, "step": 7058 }, { "epoch": 3.3218823529411763, "grad_norm": 0.7510385383058088, "learning_rate": 0.00016383255110477102, "loss": 0.377, "step": 7059 }, { "epoch": 3.3223529411764705, "grad_norm": 0.40569705746443685, "learning_rate": 0.00016378890112701312, "loss": 0.1338, "step": 7060 }, { "epoch": 3.3228235294117647, "grad_norm": 0.7848162891419003, "learning_rate": 0.0001637452499716155, "loss": 0.3242, "step": 7061 }, { "epoch": 3.323294117647059, "grad_norm": 0.21610065365995068, "learning_rate": 0.0001637015976423061, "loss": 0.0659, "step": 7062 }, { "epoch": 3.323764705882353, "grad_norm": 0.5569134919982648, "learning_rate": 0.00016365794414281303, "loss": 0.2656, "step": 7063 }, { "epoch": 3.324235294117647, "grad_norm": 6.757739593947518, "learning_rate": 0.00016361428947686457, "loss": 0.2471, "step": 7064 }, { "epoch": 3.324705882352941, "grad_norm": 0.29561458005889224, "learning_rate": 0.00016357063364818902, "loss": 0.0559, "step": 7065 }, { "epoch": 3.3251764705882354, "grad_norm": 0.30819121701214053, "learning_rate": 0.00016352697666051474, "loss": 0.1426, "step": 7066 }, { "epoch": 3.325647058823529, "grad_norm": 0.3640853622940434, "learning_rate": 0.0001634833185175703, "loss": 0.1445, "step": 7067 }, { "epoch": 3.3261176470588234, "grad_norm": 2.99118689455736, "learning_rate": 0.00016343965922308438, "loss": 0.4043, "step": 7068 }, { "epoch": 3.3265882352941176, "grad_norm": 0.47712863704389025, "learning_rate": 0.00016339599878078563, "loss": 0.1455, "step": 7069 }, { "epoch": 3.327058823529412, "grad_norm": 0.8391953846073088, "learning_rate": 0.00016335233719440282, "loss": 0.0674, "step": 7070 }, { "epoch": 3.327529411764706, "grad_norm": 1.3958310841375, "learning_rate": 0.000163308674467665, "loss": 0.3535, "step": 7071 }, { "epoch": 3.328, "grad_norm": 0.9925163134498823, "learning_rate": 0.00016326501060430105, "loss": 0.457, "step": 7072 }, { "epoch": 3.328470588235294, "grad_norm": 0.850673024140944, "learning_rate": 0.00016322134560804017, "loss": 0.5352, "step": 7073 }, { "epoch": 3.3289411764705883, "grad_norm": 0.6995384144069765, "learning_rate": 0.00016317767948261148, "loss": 0.2871, "step": 7074 }, { "epoch": 3.3294117647058825, "grad_norm": 0.9246104814047489, "learning_rate": 0.00016313401223174434, "loss": 0.3379, "step": 7075 }, { "epoch": 3.3298823529411763, "grad_norm": 0.9776834234690303, "learning_rate": 0.00016309034385916815, "loss": 0.6836, "step": 7076 }, { "epoch": 3.3303529411764705, "grad_norm": 0.16368345282855332, "learning_rate": 0.00016304667436861233, "loss": 0.0464, "step": 7077 }, { "epoch": 3.3308235294117647, "grad_norm": 0.49778915954929154, "learning_rate": 0.00016300300376380657, "loss": 0.3379, "step": 7078 }, { "epoch": 3.331294117647059, "grad_norm": 0.379424538221969, "learning_rate": 0.00016295933204848052, "loss": 0.1348, "step": 7079 }, { "epoch": 3.331764705882353, "grad_norm": 0.5627189878893899, "learning_rate": 0.00016291565922636392, "loss": 0.1836, "step": 7080 }, { "epoch": 3.332235294117647, "grad_norm": 0.771037672671278, "learning_rate": 0.0001628719853011867, "loss": 0.2715, "step": 7081 }, { "epoch": 3.332705882352941, "grad_norm": 0.3842929057742154, "learning_rate": 0.00016282831027667881, "loss": 0.1689, "step": 7082 }, { "epoch": 3.3331764705882354, "grad_norm": 0.5640369759680947, "learning_rate": 0.00016278463415657027, "loss": 0.1973, "step": 7083 }, { "epoch": 3.333647058823529, "grad_norm": 0.0007275904800067898, "learning_rate": 0.00016274095694459128, "loss": 0.0, "step": 7084 }, { "epoch": 3.3341176470588234, "grad_norm": 0.4237833439562285, "learning_rate": 0.00016269727864447206, "loss": 0.1865, "step": 7085 }, { "epoch": 3.3345882352941176, "grad_norm": 0.8032753282808437, "learning_rate": 0.00016265359925994298, "loss": 0.4219, "step": 7086 }, { "epoch": 3.335058823529412, "grad_norm": 0.484030673872847, "learning_rate": 0.00016260991879473441, "loss": 0.208, "step": 7087 }, { "epoch": 3.335529411764706, "grad_norm": 0.5061954141622121, "learning_rate": 0.000162566237252577, "loss": 0.2734, "step": 7088 }, { "epoch": 3.336, "grad_norm": 0.48758067601827565, "learning_rate": 0.00016252255463720125, "loss": 0.207, "step": 7089 }, { "epoch": 3.336470588235294, "grad_norm": 0.42693029629200585, "learning_rate": 0.0001624788709523379, "loss": 0.1436, "step": 7090 }, { "epoch": 3.3369411764705883, "grad_norm": 0.26530769710156876, "learning_rate": 0.00016243518620171778, "loss": 0.1191, "step": 7091 }, { "epoch": 3.3374117647058825, "grad_norm": 0.5447609306901599, "learning_rate": 0.00016239150038907177, "loss": 0.332, "step": 7092 }, { "epoch": 3.3378823529411763, "grad_norm": 0.46268325710641633, "learning_rate": 0.00016234781351813079, "loss": 0.3027, "step": 7093 }, { "epoch": 3.3383529411764705, "grad_norm": 0.4618674742449585, "learning_rate": 0.00016230412559262606, "loss": 0.377, "step": 7094 }, { "epoch": 3.3388235294117647, "grad_norm": 0.20923262533132936, "learning_rate": 0.00016226043661628862, "loss": 0.083, "step": 7095 }, { "epoch": 3.339294117647059, "grad_norm": 0.3699043384304116, "learning_rate": 0.00016221674659284976, "loss": 0.2148, "step": 7096 }, { "epoch": 3.339764705882353, "grad_norm": 0.4880888203035864, "learning_rate": 0.00016217305552604076, "loss": 0.4297, "step": 7097 }, { "epoch": 3.340235294117647, "grad_norm": 0.1737118059628495, "learning_rate": 0.00016212936341959312, "loss": 0.0513, "step": 7098 }, { "epoch": 3.340705882352941, "grad_norm": 0.5353304703642774, "learning_rate": 0.0001620856702772384, "loss": 0.2656, "step": 7099 }, { "epoch": 3.3411764705882354, "grad_norm": 0.5599505614023799, "learning_rate": 0.00016204197610270816, "loss": 0.3203, "step": 7100 }, { "epoch": 3.341647058823529, "grad_norm": 0.5911699753149892, "learning_rate": 0.00016199828089973406, "loss": 0.5547, "step": 7101 }, { "epoch": 3.3421176470588234, "grad_norm": 0.7650369055877317, "learning_rate": 0.00016195458467204793, "loss": 0.3652, "step": 7102 }, { "epoch": 3.3425882352941176, "grad_norm": 0.33367144282807565, "learning_rate": 0.00016191088742338166, "loss": 0.1011, "step": 7103 }, { "epoch": 3.343058823529412, "grad_norm": 0.8831585363242845, "learning_rate": 0.00016186718915746714, "loss": 0.4492, "step": 7104 }, { "epoch": 3.343529411764706, "grad_norm": 0.6292094707562076, "learning_rate": 0.00016182348987803648, "loss": 0.4902, "step": 7105 }, { "epoch": 3.344, "grad_norm": 0.6278557107435555, "learning_rate": 0.0001617797895888218, "loss": 0.373, "step": 7106 }, { "epoch": 3.344470588235294, "grad_norm": 0.6061685878236454, "learning_rate": 0.00016173608829355527, "loss": 0.3105, "step": 7107 }, { "epoch": 3.3449411764705883, "grad_norm": 0.5326480506156079, "learning_rate": 0.00016169238599596927, "loss": 0.4375, "step": 7108 }, { "epoch": 3.3454117647058825, "grad_norm": 0.600771373369712, "learning_rate": 0.00016164868269979619, "loss": 0.2969, "step": 7109 }, { "epoch": 3.3458823529411763, "grad_norm": 0.5245164638704798, "learning_rate": 0.0001616049784087684, "loss": 0.2334, "step": 7110 }, { "epoch": 3.3463529411764705, "grad_norm": 0.5340271893313457, "learning_rate": 0.00016156127312661855, "loss": 0.2637, "step": 7111 }, { "epoch": 3.3468235294117648, "grad_norm": 0.7200929897927695, "learning_rate": 0.0001615175668570793, "loss": 0.4551, "step": 7112 }, { "epoch": 3.347294117647059, "grad_norm": 0.5686586274633442, "learning_rate": 0.0001614738596038833, "loss": 0.2617, "step": 7113 }, { "epoch": 3.3477647058823528, "grad_norm": 0.4352858323073677, "learning_rate": 0.00016143015137076345, "loss": 0.1436, "step": 7114 }, { "epoch": 3.348235294117647, "grad_norm": 0.30582645766197936, "learning_rate": 0.0001613864421614526, "loss": 0.0986, "step": 7115 }, { "epoch": 3.348705882352941, "grad_norm": 0.5176403750621354, "learning_rate": 0.00016134273197968374, "loss": 0.2852, "step": 7116 }, { "epoch": 3.3491764705882354, "grad_norm": 0.2982569214862104, "learning_rate": 0.0001612990208291899, "loss": 0.124, "step": 7117 }, { "epoch": 3.349647058823529, "grad_norm": 0.32485264584236534, "learning_rate": 0.0001612553087137043, "loss": 0.1089, "step": 7118 }, { "epoch": 3.3501176470588234, "grad_norm": 0.7244807635776045, "learning_rate": 0.0001612115956369601, "loss": 0.3496, "step": 7119 }, { "epoch": 3.3505882352941176, "grad_norm": 0.5632043208288138, "learning_rate": 0.00016116788160269065, "loss": 0.4375, "step": 7120 }, { "epoch": 3.351058823529412, "grad_norm": 0.7524999121764152, "learning_rate": 0.0001611241666146293, "loss": 0.4141, "step": 7121 }, { "epoch": 3.351529411764706, "grad_norm": 0.48660249177396014, "learning_rate": 0.00016108045067650962, "loss": 0.207, "step": 7122 }, { "epoch": 3.352, "grad_norm": 0.4637923396164858, "learning_rate": 0.0001610367337920651, "loss": 0.2891, "step": 7123 }, { "epoch": 3.352470588235294, "grad_norm": 0.5789894839319367, "learning_rate": 0.00016099301596502936, "loss": 0.377, "step": 7124 }, { "epoch": 3.3529411764705883, "grad_norm": 0.45970872461824586, "learning_rate": 0.00016094929719913612, "loss": 0.2324, "step": 7125 }, { "epoch": 3.3534117647058825, "grad_norm": 0.5044138589838174, "learning_rate": 0.0001609055774981192, "loss": 0.3027, "step": 7126 }, { "epoch": 3.3538823529411763, "grad_norm": 0.40208859694137755, "learning_rate": 0.0001608618568657125, "loss": 0.1689, "step": 7127 }, { "epoch": 3.3543529411764705, "grad_norm": 0.8063901019062623, "learning_rate": 0.0001608181353056499, "loss": 0.5312, "step": 7128 }, { "epoch": 3.3548235294117648, "grad_norm": 0.9681719585211618, "learning_rate": 0.0001607744128216655, "loss": 0.6172, "step": 7129 }, { "epoch": 3.355294117647059, "grad_norm": 0.5134595157404362, "learning_rate": 0.00016073068941749343, "loss": 0.291, "step": 7130 }, { "epoch": 3.3557647058823528, "grad_norm": 0.3990870420909823, "learning_rate": 0.0001606869650968678, "loss": 0.1797, "step": 7131 }, { "epoch": 3.356235294117647, "grad_norm": 0.4171477967622275, "learning_rate": 0.00016064323986352295, "loss": 0.2002, "step": 7132 }, { "epoch": 3.356705882352941, "grad_norm": 0.14207355711454958, "learning_rate": 0.00016059951372119323, "loss": 0.0381, "step": 7133 }, { "epoch": 3.3571764705882354, "grad_norm": 0.23069573062685922, "learning_rate": 0.000160555786673613, "loss": 0.0474, "step": 7134 }, { "epoch": 3.357647058823529, "grad_norm": 0.1307529119703625, "learning_rate": 0.00016051205872451686, "loss": 0.0493, "step": 7135 }, { "epoch": 3.3581176470588234, "grad_norm": 0.5701394537573335, "learning_rate": 0.00016046832987763935, "loss": 0.2148, "step": 7136 }, { "epoch": 3.3585882352941177, "grad_norm": 0.40714539651244336, "learning_rate": 0.0001604246001367151, "loss": 0.1807, "step": 7137 }, { "epoch": 3.359058823529412, "grad_norm": 0.64539016577736, "learning_rate": 0.00016038086950547885, "loss": 0.4668, "step": 7138 }, { "epoch": 3.359529411764706, "grad_norm": 1.0441926846844125, "learning_rate": 0.00016033713798766546, "loss": 0.2812, "step": 7139 }, { "epoch": 3.36, "grad_norm": 0.3322055116450624, "learning_rate": 0.00016029340558700977, "loss": 0.1143, "step": 7140 }, { "epoch": 3.360470588235294, "grad_norm": 0.6895599407314299, "learning_rate": 0.00016024967230724672, "loss": 0.3789, "step": 7141 }, { "epoch": 3.3609411764705883, "grad_norm": 0.5868307895380315, "learning_rate": 0.00016020593815211138, "loss": 0.2061, "step": 7142 }, { "epoch": 3.3614117647058825, "grad_norm": 0.3540890815210815, "learning_rate": 0.00016016220312533888, "loss": 0.1367, "step": 7143 }, { "epoch": 3.3618823529411763, "grad_norm": 0.31811151701983703, "learning_rate": 0.00016011846723066445, "loss": 0.1226, "step": 7144 }, { "epoch": 3.3623529411764705, "grad_norm": 0.3558233502706568, "learning_rate": 0.0001600747304718232, "loss": 0.1768, "step": 7145 }, { "epoch": 3.3628235294117648, "grad_norm": 0.18978139663936625, "learning_rate": 0.00016003099285255058, "loss": 0.0557, "step": 7146 }, { "epoch": 3.363294117647059, "grad_norm": 0.7050622078719778, "learning_rate": 0.000159987254376582, "loss": 0.375, "step": 7147 }, { "epoch": 3.3637647058823528, "grad_norm": 0.2553350194449692, "learning_rate": 0.00015994351504765287, "loss": 0.1035, "step": 7148 }, { "epoch": 3.364235294117647, "grad_norm": 0.7955051092232615, "learning_rate": 0.00015989977486949883, "loss": 0.5469, "step": 7149 }, { "epoch": 3.364705882352941, "grad_norm": 0.5486903915975732, "learning_rate": 0.00015985603384585542, "loss": 0.3672, "step": 7150 }, { "epoch": 3.3651764705882354, "grad_norm": 0.24966599849103574, "learning_rate": 0.0001598122919804584, "loss": 0.0908, "step": 7151 }, { "epoch": 3.365647058823529, "grad_norm": 0.4680561448235897, "learning_rate": 0.0001597685492770435, "loss": 0.2012, "step": 7152 }, { "epoch": 3.3661176470588234, "grad_norm": 0.15052971294640474, "learning_rate": 0.00015972480573934656, "loss": 0.0242, "step": 7153 }, { "epoch": 3.3665882352941177, "grad_norm": 0.6019198366361755, "learning_rate": 0.00015968106137110352, "loss": 0.5195, "step": 7154 }, { "epoch": 3.367058823529412, "grad_norm": 0.7884020572370112, "learning_rate": 0.00015963731617605034, "loss": 0.3262, "step": 7155 }, { "epoch": 3.367529411764706, "grad_norm": 0.4243506870732535, "learning_rate": 0.00015959357015792313, "loss": 0.1895, "step": 7156 }, { "epoch": 3.368, "grad_norm": 0.7669092900771762, "learning_rate": 0.00015954982332045794, "loss": 0.3906, "step": 7157 }, { "epoch": 3.368470588235294, "grad_norm": 0.1906470157482051, "learning_rate": 0.000159506075667391, "loss": 0.0486, "step": 7158 }, { "epoch": 3.3689411764705883, "grad_norm": 0.266962799157908, "learning_rate": 0.0001594623272024586, "loss": 0.0869, "step": 7159 }, { "epoch": 3.369411764705882, "grad_norm": 0.4408123526378185, "learning_rate": 0.000159418577929397, "loss": 0.3184, "step": 7160 }, { "epoch": 3.3698823529411763, "grad_norm": 0.4382501359950681, "learning_rate": 0.0001593748278519427, "loss": 0.168, "step": 7161 }, { "epoch": 3.3703529411764706, "grad_norm": 0.2795526475485585, "learning_rate": 0.00015933107697383206, "loss": 0.0957, "step": 7162 }, { "epoch": 3.3708235294117648, "grad_norm": 0.3963748604145277, "learning_rate": 0.0001592873252988017, "loss": 0.1641, "step": 7163 }, { "epoch": 3.371294117647059, "grad_norm": 0.38852520097492277, "learning_rate": 0.0001592435728305882, "loss": 0.1299, "step": 7164 }, { "epoch": 3.3717647058823528, "grad_norm": 0.5552991697090979, "learning_rate": 0.00015919981957292825, "loss": 0.2695, "step": 7165 }, { "epoch": 3.372235294117647, "grad_norm": 0.5027911452553239, "learning_rate": 0.00015915606552955856, "loss": 0.3027, "step": 7166 }, { "epoch": 3.372705882352941, "grad_norm": 0.3003277452493772, "learning_rate": 0.000159112310704216, "loss": 0.1191, "step": 7167 }, { "epoch": 3.3731764705882354, "grad_norm": 0.24622052470270483, "learning_rate": 0.00015906855510063744, "loss": 0.1167, "step": 7168 }, { "epoch": 3.373647058823529, "grad_norm": 0.6823163990555848, "learning_rate": 0.0001590247987225597, "loss": 0.4336, "step": 7169 }, { "epoch": 3.3741176470588234, "grad_norm": 1.2964064571097995, "learning_rate": 0.00015898104157371993, "loss": 0.5664, "step": 7170 }, { "epoch": 3.3745882352941177, "grad_norm": 0.5971431082329515, "learning_rate": 0.00015893728365785518, "loss": 0.3711, "step": 7171 }, { "epoch": 3.375058823529412, "grad_norm": 0.27941818251900435, "learning_rate": 0.00015889352497870253, "loss": 0.0967, "step": 7172 }, { "epoch": 3.375529411764706, "grad_norm": 0.6107634603318597, "learning_rate": 0.00015884976553999925, "loss": 0.2129, "step": 7173 }, { "epoch": 3.376, "grad_norm": 0.31159346936036914, "learning_rate": 0.0001588060053454826, "loss": 0.0806, "step": 7174 }, { "epoch": 3.376470588235294, "grad_norm": 0.862288395500027, "learning_rate": 0.0001587622443988899, "loss": 0.7227, "step": 7175 }, { "epoch": 3.3769411764705883, "grad_norm": 0.5937801581915498, "learning_rate": 0.0001587184827039585, "loss": 0.3184, "step": 7176 }, { "epoch": 3.377411764705882, "grad_norm": 0.4151825849652547, "learning_rate": 0.00015867472026442596, "loss": 0.1377, "step": 7177 }, { "epoch": 3.3778823529411763, "grad_norm": 0.7704978098076591, "learning_rate": 0.00015863095708402972, "loss": 0.3711, "step": 7178 }, { "epoch": 3.3783529411764706, "grad_norm": 2.5491268811754884, "learning_rate": 0.00015858719316650744, "loss": 0.5664, "step": 7179 }, { "epoch": 3.378823529411765, "grad_norm": 0.5023714101346123, "learning_rate": 0.00015854342851559673, "loss": 0.3066, "step": 7180 }, { "epoch": 3.379294117647059, "grad_norm": 0.11224274497976336, "learning_rate": 0.00015849966313503537, "loss": 0.0315, "step": 7181 }, { "epoch": 3.379764705882353, "grad_norm": 0.414675247752209, "learning_rate": 0.00015845589702856102, "loss": 0.1099, "step": 7182 }, { "epoch": 3.380235294117647, "grad_norm": 2.3834708146965533, "learning_rate": 0.0001584121301999116, "loss": 0.3223, "step": 7183 }, { "epoch": 3.3807058823529412, "grad_norm": 0.9299961604473931, "learning_rate": 0.00015836836265282502, "loss": 0.3652, "step": 7184 }, { "epoch": 3.3811764705882354, "grad_norm": 0.8009118783736866, "learning_rate": 0.0001583245943910392, "loss": 0.3164, "step": 7185 }, { "epoch": 3.3816470588235292, "grad_norm": 1.120969461612084, "learning_rate": 0.0001582808254182922, "loss": 0.459, "step": 7186 }, { "epoch": 3.3821176470588235, "grad_norm": 0.44684706598085555, "learning_rate": 0.0001582370557383221, "loss": 0.1855, "step": 7187 }, { "epoch": 3.3825882352941177, "grad_norm": 0.9265835460004818, "learning_rate": 0.00015819328535486704, "loss": 0.5977, "step": 7188 }, { "epoch": 3.383058823529412, "grad_norm": 0.48371086962075044, "learning_rate": 0.0001581495142716652, "loss": 0.2402, "step": 7189 }, { "epoch": 3.383529411764706, "grad_norm": 0.6834359038724175, "learning_rate": 0.00015810574249245488, "loss": 0.5, "step": 7190 }, { "epoch": 3.384, "grad_norm": 0.749018681520497, "learning_rate": 0.00015806197002097438, "loss": 0.3027, "step": 7191 }, { "epoch": 3.384470588235294, "grad_norm": 0.40526785510195795, "learning_rate": 0.00015801819686096213, "loss": 0.1299, "step": 7192 }, { "epoch": 3.3849411764705883, "grad_norm": 0.5629088743190653, "learning_rate": 0.00015797442301615652, "loss": 0.3184, "step": 7193 }, { "epoch": 3.385411764705882, "grad_norm": 11.512038084183528, "learning_rate": 0.00015793064849029608, "loss": 0.3945, "step": 7194 }, { "epoch": 3.3858823529411763, "grad_norm": 0.5156040160622514, "learning_rate": 0.00015788687328711933, "loss": 0.3105, "step": 7195 }, { "epoch": 3.3863529411764706, "grad_norm": 0.3268426078190216, "learning_rate": 0.00015784309741036492, "loss": 0.1055, "step": 7196 }, { "epoch": 3.386823529411765, "grad_norm": 0.708462103308451, "learning_rate": 0.00015779932086377153, "loss": 0.3555, "step": 7197 }, { "epoch": 3.387294117647059, "grad_norm": 0.5068465319804019, "learning_rate": 0.00015775554365107791, "loss": 0.2031, "step": 7198 }, { "epoch": 3.387764705882353, "grad_norm": 0.6132147451575559, "learning_rate": 0.0001577117657760228, "loss": 0.4023, "step": 7199 }, { "epoch": 3.388235294117647, "grad_norm": 0.8211603826416939, "learning_rate": 0.00015766798724234506, "loss": 0.4453, "step": 7200 }, { "epoch": 3.3887058823529412, "grad_norm": 0.602743951493823, "learning_rate": 0.00015762420805378363, "loss": 0.3496, "step": 7201 }, { "epoch": 3.3891764705882355, "grad_norm": 0.3840692245131978, "learning_rate": 0.00015758042821407742, "loss": 0.1133, "step": 7202 }, { "epoch": 3.3896470588235292, "grad_norm": 0.9155183564179683, "learning_rate": 0.00015753664772696544, "loss": 0.4902, "step": 7203 }, { "epoch": 3.3901176470588235, "grad_norm": 0.5085699010131034, "learning_rate": 0.00015749286659618678, "loss": 0.291, "step": 7204 }, { "epoch": 3.3905882352941177, "grad_norm": 0.48811750224657996, "learning_rate": 0.00015744908482548059, "loss": 0.2539, "step": 7205 }, { "epoch": 3.391058823529412, "grad_norm": 0.42637126217229493, "learning_rate": 0.00015740530241858597, "loss": 0.1514, "step": 7206 }, { "epoch": 3.391529411764706, "grad_norm": 0.36069892022511457, "learning_rate": 0.00015736151937924224, "loss": 0.1543, "step": 7207 }, { "epoch": 3.392, "grad_norm": 0.7817160390219959, "learning_rate": 0.00015731773571118864, "loss": 0.375, "step": 7208 }, { "epoch": 3.392470588235294, "grad_norm": 0.4981848218760675, "learning_rate": 0.00015727395141816448, "loss": 0.3086, "step": 7209 }, { "epoch": 3.3929411764705883, "grad_norm": 0.8312691616416193, "learning_rate": 0.00015723016650390922, "loss": 0.3203, "step": 7210 }, { "epoch": 3.393411764705882, "grad_norm": 0.7990481198019503, "learning_rate": 0.0001571863809721623, "loss": 0.2578, "step": 7211 }, { "epoch": 3.3938823529411764, "grad_norm": 0.4738930455674323, "learning_rate": 0.00015714259482666317, "loss": 0.1523, "step": 7212 }, { "epoch": 3.3943529411764706, "grad_norm": 0.911697302386067, "learning_rate": 0.00015709880807115138, "loss": 0.582, "step": 7213 }, { "epoch": 3.394823529411765, "grad_norm": 0.3832445714918439, "learning_rate": 0.0001570550207093666, "loss": 0.1064, "step": 7214 }, { "epoch": 3.395294117647059, "grad_norm": 0.6689169812940111, "learning_rate": 0.00015701123274504845, "loss": 0.3516, "step": 7215 }, { "epoch": 3.395764705882353, "grad_norm": 0.28846226781850604, "learning_rate": 0.00015696744418193664, "loss": 0.0977, "step": 7216 }, { "epoch": 3.396235294117647, "grad_norm": 0.5339732942903332, "learning_rate": 0.00015692365502377088, "loss": 0.2871, "step": 7217 }, { "epoch": 3.3967058823529412, "grad_norm": 0.768296787882603, "learning_rate": 0.00015687986527429103, "loss": 0.3887, "step": 7218 }, { "epoch": 3.3971764705882355, "grad_norm": 0.6776257883495901, "learning_rate": 0.00015683607493723699, "loss": 0.5312, "step": 7219 }, { "epoch": 3.3976470588235292, "grad_norm": 0.5430409693662176, "learning_rate": 0.00015679228401634855, "loss": 0.3242, "step": 7220 }, { "epoch": 3.3981176470588235, "grad_norm": 0.8138991543700603, "learning_rate": 0.0001567484925153658, "loss": 0.5469, "step": 7221 }, { "epoch": 3.3985882352941177, "grad_norm": 0.5385005538189631, "learning_rate": 0.00015670470043802867, "loss": 0.1611, "step": 7222 }, { "epoch": 3.399058823529412, "grad_norm": 0.3542219066883203, "learning_rate": 0.0001566609077880772, "loss": 0.1187, "step": 7223 }, { "epoch": 3.3995294117647057, "grad_norm": 0.25602752486126373, "learning_rate": 0.00015661711456925158, "loss": 0.0845, "step": 7224 }, { "epoch": 3.4, "grad_norm": 0.5630270191461845, "learning_rate": 0.0001565733207852919, "loss": 0.2871, "step": 7225 }, { "epoch": 3.400470588235294, "grad_norm": 0.4556289827329815, "learning_rate": 0.00015652952643993841, "loss": 0.1553, "step": 7226 }, { "epoch": 3.4009411764705884, "grad_norm": 0.4114743060708863, "learning_rate": 0.00015648573153693128, "loss": 0.2617, "step": 7227 }, { "epoch": 3.401411764705882, "grad_norm": 0.3272724389938714, "learning_rate": 0.0001564419360800109, "loss": 0.1514, "step": 7228 }, { "epoch": 3.4018823529411764, "grad_norm": 0.4680102998653721, "learning_rate": 0.0001563981400729176, "loss": 0.2139, "step": 7229 }, { "epoch": 3.4023529411764706, "grad_norm": 0.7132662883028192, "learning_rate": 0.0001563543435193917, "loss": 0.4531, "step": 7230 }, { "epoch": 3.402823529411765, "grad_norm": 0.7247103541111147, "learning_rate": 0.0001563105464231737, "loss": 0.3652, "step": 7231 }, { "epoch": 3.403294117647059, "grad_norm": 0.7078525424340353, "learning_rate": 0.0001562667487880041, "loss": 0.3477, "step": 7232 }, { "epoch": 3.403764705882353, "grad_norm": 0.4489344736836266, "learning_rate": 0.00015622295061762346, "loss": 0.1562, "step": 7233 }, { "epoch": 3.404235294117647, "grad_norm": 0.4021622205561301, "learning_rate": 0.00015617915191577222, "loss": 0.2119, "step": 7234 }, { "epoch": 3.4047058823529412, "grad_norm": 0.18044448312436634, "learning_rate": 0.0001561353526861912, "loss": 0.0325, "step": 7235 }, { "epoch": 3.4051764705882355, "grad_norm": 0.37019976760064244, "learning_rate": 0.0001560915529326209, "loss": 0.1826, "step": 7236 }, { "epoch": 3.4056470588235292, "grad_norm": 0.58851016942093, "learning_rate": 0.00015604775265880212, "loss": 0.373, "step": 7237 }, { "epoch": 3.4061176470588235, "grad_norm": 0.270840519174454, "learning_rate": 0.0001560039518684756, "loss": 0.1162, "step": 7238 }, { "epoch": 3.4065882352941177, "grad_norm": 0.35026419617701915, "learning_rate": 0.00015596015056538215, "loss": 0.1504, "step": 7239 }, { "epoch": 3.407058823529412, "grad_norm": 1.080156826423091, "learning_rate": 0.0001559163487532626, "loss": 0.5938, "step": 7240 }, { "epoch": 3.4075294117647057, "grad_norm": 0.39533741835512504, "learning_rate": 0.00015587254643585784, "loss": 0.1748, "step": 7241 }, { "epoch": 3.408, "grad_norm": 0.33667988058998866, "learning_rate": 0.00015582874361690885, "loss": 0.104, "step": 7242 }, { "epoch": 3.408470588235294, "grad_norm": 0.6009524747448202, "learning_rate": 0.00015578494030015652, "loss": 0.5859, "step": 7243 }, { "epoch": 3.4089411764705884, "grad_norm": 0.5886733220892202, "learning_rate": 0.00015574113648934192, "loss": 0.3828, "step": 7244 }, { "epoch": 3.409411764705882, "grad_norm": 0.40315073478226, "learning_rate": 0.0001556973321882061, "loss": 0.2129, "step": 7245 }, { "epoch": 3.4098823529411764, "grad_norm": 0.39954901131589654, "learning_rate": 0.0001556535274004902, "loss": 0.1689, "step": 7246 }, { "epoch": 3.4103529411764706, "grad_norm": 0.33406531643977944, "learning_rate": 0.00015560972212993527, "loss": 0.1235, "step": 7247 }, { "epoch": 3.410823529411765, "grad_norm": 0.4660748334724586, "learning_rate": 0.0001555659163802826, "loss": 0.2412, "step": 7248 }, { "epoch": 3.411294117647059, "grad_norm": 0.7355371236986379, "learning_rate": 0.00015552211015527336, "loss": 0.4219, "step": 7249 }, { "epoch": 3.411764705882353, "grad_norm": 0.2647670978636765, "learning_rate": 0.00015547830345864885, "loss": 0.0527, "step": 7250 }, { "epoch": 3.412235294117647, "grad_norm": 0.23239585456668535, "learning_rate": 0.00015543449629415028, "loss": 0.0645, "step": 7251 }, { "epoch": 3.4127058823529413, "grad_norm": 0.5138382510007575, "learning_rate": 0.00015539068866551916, "loss": 0.2236, "step": 7252 }, { "epoch": 3.4131764705882355, "grad_norm": 0.2697557356629609, "learning_rate": 0.00015534688057649673, "loss": 0.0718, "step": 7253 }, { "epoch": 3.4136470588235293, "grad_norm": 0.6616479452036416, "learning_rate": 0.0001553030720308245, "loss": 0.4199, "step": 7254 }, { "epoch": 3.4141176470588235, "grad_norm": 0.352010629752778, "learning_rate": 0.00015525926303224387, "loss": 0.2061, "step": 7255 }, { "epoch": 3.4145882352941177, "grad_norm": 0.6196838314161225, "learning_rate": 0.0001552154535844964, "loss": 0.3711, "step": 7256 }, { "epoch": 3.415058823529412, "grad_norm": 0.501570102269485, "learning_rate": 0.00015517164369132364, "loss": 0.2188, "step": 7257 }, { "epoch": 3.4155294117647057, "grad_norm": 0.5615986967947508, "learning_rate": 0.0001551278333564671, "loss": 0.4102, "step": 7258 }, { "epoch": 3.416, "grad_norm": 0.32701781490167847, "learning_rate": 0.00015508402258366846, "loss": 0.2383, "step": 7259 }, { "epoch": 3.416470588235294, "grad_norm": 0.5175582643097345, "learning_rate": 0.00015504021137666938, "loss": 0.3516, "step": 7260 }, { "epoch": 3.4169411764705884, "grad_norm": 0.48588973768292565, "learning_rate": 0.00015499639973921148, "loss": 0.4141, "step": 7261 }, { "epoch": 3.417411764705882, "grad_norm": 0.13451673062736932, "learning_rate": 0.00015495258767503657, "loss": 0.0378, "step": 7262 }, { "epoch": 3.4178823529411764, "grad_norm": 0.3106261098645371, "learning_rate": 0.0001549087751878864, "loss": 0.1865, "step": 7263 }, { "epoch": 3.4183529411764706, "grad_norm": 0.3436609019248025, "learning_rate": 0.00015486496228150266, "loss": 0.103, "step": 7264 }, { "epoch": 3.418823529411765, "grad_norm": 0.5519861962174162, "learning_rate": 0.00015482114895962735, "loss": 0.3145, "step": 7265 }, { "epoch": 3.419294117647059, "grad_norm": 0.9168013537795231, "learning_rate": 0.00015477733522600227, "loss": 0.6875, "step": 7266 }, { "epoch": 3.419764705882353, "grad_norm": 0.3356127540250867, "learning_rate": 0.00015473352108436934, "loss": 0.208, "step": 7267 }, { "epoch": 3.420235294117647, "grad_norm": 0.6999504896040765, "learning_rate": 0.00015468970653847043, "loss": 0.373, "step": 7268 }, { "epoch": 3.4207058823529413, "grad_norm": 0.6478566783722581, "learning_rate": 0.00015464589159204764, "loss": 0.3633, "step": 7269 }, { "epoch": 3.4211764705882355, "grad_norm": 0.34072931040815346, "learning_rate": 0.0001546020762488429, "loss": 0.1338, "step": 7270 }, { "epoch": 3.4216470588235293, "grad_norm": 0.4144350460738508, "learning_rate": 0.00015455826051259826, "loss": 0.1973, "step": 7271 }, { "epoch": 3.4221176470588235, "grad_norm": 0.4395655670045259, "learning_rate": 0.00015451444438705584, "loss": 0.1787, "step": 7272 }, { "epoch": 3.4225882352941177, "grad_norm": 0.29753168640725636, "learning_rate": 0.00015447062787595772, "loss": 0.1011, "step": 7273 }, { "epoch": 3.423058823529412, "grad_norm": 0.695394110715235, "learning_rate": 0.00015442681098304603, "loss": 0.3594, "step": 7274 }, { "epoch": 3.4235294117647057, "grad_norm": 0.7380401889527597, "learning_rate": 0.000154382993712063, "loss": 0.3242, "step": 7275 }, { "epoch": 3.424, "grad_norm": 0.5988596827063879, "learning_rate": 0.00015433917606675075, "loss": 0.3477, "step": 7276 }, { "epoch": 3.424470588235294, "grad_norm": 0.37773724302519057, "learning_rate": 0.00015429535805085162, "loss": 0.1582, "step": 7277 }, { "epoch": 3.4249411764705884, "grad_norm": 0.8728704231738188, "learning_rate": 0.00015425153966810783, "loss": 0.5352, "step": 7278 }, { "epoch": 3.425411764705882, "grad_norm": 0.37235740485217794, "learning_rate": 0.00015420772092226173, "loss": 0.1152, "step": 7279 }, { "epoch": 3.4258823529411764, "grad_norm": 0.5146564951362309, "learning_rate": 0.00015416390181705562, "loss": 0.3906, "step": 7280 }, { "epoch": 3.4263529411764706, "grad_norm": 0.5326660116706448, "learning_rate": 0.00015412008235623186, "loss": 0.2637, "step": 7281 }, { "epoch": 3.426823529411765, "grad_norm": 0.49736363887586743, "learning_rate": 0.00015407626254353282, "loss": 0.3535, "step": 7282 }, { "epoch": 3.427294117647059, "grad_norm": 0.20666475821387156, "learning_rate": 0.00015403244238270102, "loss": 0.082, "step": 7283 }, { "epoch": 3.427764705882353, "grad_norm": 0.5089687773996671, "learning_rate": 0.00015398862187747888, "loss": 0.334, "step": 7284 }, { "epoch": 3.428235294117647, "grad_norm": 0.5299782154502009, "learning_rate": 0.0001539448010316088, "loss": 0.2002, "step": 7285 }, { "epoch": 3.4287058823529413, "grad_norm": 0.5263814235813535, "learning_rate": 0.0001539009798488334, "loss": 0.3789, "step": 7286 }, { "epoch": 3.429176470588235, "grad_norm": 0.4922666977681504, "learning_rate": 0.0001538571583328952, "loss": 0.2852, "step": 7287 }, { "epoch": 3.4296470588235293, "grad_norm": 0.116204132342137, "learning_rate": 0.0001538133364875367, "loss": 0.0248, "step": 7288 }, { "epoch": 3.4301176470588235, "grad_norm": 0.7675913675391185, "learning_rate": 0.00015376951431650063, "loss": 0.5117, "step": 7289 }, { "epoch": 3.4305882352941177, "grad_norm": 0.7955047242674853, "learning_rate": 0.00015372569182352952, "loss": 0.6172, "step": 7290 }, { "epoch": 3.431058823529412, "grad_norm": 0.4204336348677775, "learning_rate": 0.00015368186901236605, "loss": 0.3281, "step": 7291 }, { "epoch": 3.4315294117647057, "grad_norm": 0.34742060013414183, "learning_rate": 0.00015363804588675292, "loss": 0.1533, "step": 7292 }, { "epoch": 3.432, "grad_norm": 0.07989926695113853, "learning_rate": 0.00015359422245043278, "loss": 0.0112, "step": 7293 }, { "epoch": 3.432470588235294, "grad_norm": 0.4644409618787535, "learning_rate": 0.00015355039870714848, "loss": 0.2832, "step": 7294 }, { "epoch": 3.4329411764705884, "grad_norm": 0.36622369338890487, "learning_rate": 0.00015350657466064264, "loss": 0.1592, "step": 7295 }, { "epoch": 3.433411764705882, "grad_norm": 0.3856869787073076, "learning_rate": 0.00015346275031465818, "loss": 0.1484, "step": 7296 }, { "epoch": 3.4338823529411764, "grad_norm": 0.29026131526942167, "learning_rate": 0.00015341892567293785, "loss": 0.1875, "step": 7297 }, { "epoch": 3.4343529411764706, "grad_norm": 0.5927337994824701, "learning_rate": 0.0001533751007392245, "loss": 0.334, "step": 7298 }, { "epoch": 3.434823529411765, "grad_norm": 0.62495204270941, "learning_rate": 0.00015333127551726096, "loss": 0.5039, "step": 7299 }, { "epoch": 3.435294117647059, "grad_norm": 0.9115680699469239, "learning_rate": 0.0001532874500107902, "loss": 0.6055, "step": 7300 }, { "epoch": 3.435764705882353, "grad_norm": 0.3065712660708337, "learning_rate": 0.00015324362422355507, "loss": 0.1494, "step": 7301 }, { "epoch": 3.436235294117647, "grad_norm": 0.3644928632856316, "learning_rate": 0.00015319979815929847, "loss": 0.166, "step": 7302 }, { "epoch": 3.4367058823529413, "grad_norm": 0.4289191957945955, "learning_rate": 0.00015315597182176348, "loss": 0.252, "step": 7303 }, { "epoch": 3.437176470588235, "grad_norm": 0.32936930266564884, "learning_rate": 0.000153112145214693, "loss": 0.2832, "step": 7304 }, { "epoch": 3.4376470588235293, "grad_norm": 0.9045929105309265, "learning_rate": 0.00015306831834183008, "loss": 0.625, "step": 7305 }, { "epoch": 3.4381176470588235, "grad_norm": 0.44457626211609896, "learning_rate": 0.00015302449120691765, "loss": 0.1836, "step": 7306 }, { "epoch": 3.4385882352941177, "grad_norm": 0.603969376751115, "learning_rate": 0.00015298066381369888, "loss": 0.3711, "step": 7307 }, { "epoch": 3.439058823529412, "grad_norm": 0.3206406704189857, "learning_rate": 0.0001529368361659168, "loss": 0.1235, "step": 7308 }, { "epoch": 3.4395294117647057, "grad_norm": 0.41228156709355607, "learning_rate": 0.0001528930082673145, "loss": 0.2617, "step": 7309 }, { "epoch": 3.44, "grad_norm": 0.6903406749835459, "learning_rate": 0.0001528491801216351, "loss": 0.3613, "step": 7310 }, { "epoch": 3.440470588235294, "grad_norm": 0.33374382786922124, "learning_rate": 0.00015280535173262176, "loss": 0.1865, "step": 7311 }, { "epoch": 3.4409411764705884, "grad_norm": 0.4327745202464663, "learning_rate": 0.0001527615231040176, "loss": 0.3184, "step": 7312 }, { "epoch": 3.441411764705882, "grad_norm": 0.5129932299539427, "learning_rate": 0.00015271769423956584, "loss": 0.3711, "step": 7313 }, { "epoch": 3.4418823529411764, "grad_norm": 0.5836407820105541, "learning_rate": 0.00015267386514300965, "loss": 0.4609, "step": 7314 }, { "epoch": 3.4423529411764706, "grad_norm": 0.9574790389707711, "learning_rate": 0.00015263003581809226, "loss": 0.5273, "step": 7315 }, { "epoch": 3.442823529411765, "grad_norm": 0.466952610041872, "learning_rate": 0.00015258620626855691, "loss": 0.1807, "step": 7316 }, { "epoch": 3.443294117647059, "grad_norm": 0.8306240670642702, "learning_rate": 0.00015254237649814687, "loss": 0.5234, "step": 7317 }, { "epoch": 3.443764705882353, "grad_norm": 0.4794598065675017, "learning_rate": 0.00015249854651060542, "loss": 0.3203, "step": 7318 }, { "epoch": 3.444235294117647, "grad_norm": 0.3427104787745371, "learning_rate": 0.00015245471630967583, "loss": 0.0781, "step": 7319 }, { "epoch": 3.4447058823529413, "grad_norm": 0.6872712309984439, "learning_rate": 0.00015241088589910138, "loss": 0.3848, "step": 7320 }, { "epoch": 3.445176470588235, "grad_norm": 0.3691352403681497, "learning_rate": 0.0001523670552826255, "loss": 0.1592, "step": 7321 }, { "epoch": 3.4456470588235293, "grad_norm": 0.036450223682229964, "learning_rate": 0.00015232322446399147, "loss": 0.0044, "step": 7322 }, { "epoch": 3.4461176470588235, "grad_norm": 0.7658134013263547, "learning_rate": 0.00015227939344694268, "loss": 0.5898, "step": 7323 }, { "epoch": 3.4465882352941177, "grad_norm": 0.9478578008723978, "learning_rate": 0.0001522355622352225, "loss": 0.5664, "step": 7324 }, { "epoch": 3.447058823529412, "grad_norm": 0.33142148412166805, "learning_rate": 0.00015219173083257437, "loss": 0.0767, "step": 7325 }, { "epoch": 3.4475294117647057, "grad_norm": 0.7637689137365786, "learning_rate": 0.00015214789924274168, "loss": 0.3516, "step": 7326 }, { "epoch": 3.448, "grad_norm": 1.5988703298426554, "learning_rate": 0.00015210406746946783, "loss": 1.4609, "step": 7327 }, { "epoch": 3.448470588235294, "grad_norm": 0.3005767763035715, "learning_rate": 0.0001520602355164963, "loss": 0.1138, "step": 7328 }, { "epoch": 3.4489411764705884, "grad_norm": 0.6276447588448494, "learning_rate": 0.00015201640338757054, "loss": 0.3379, "step": 7329 }, { "epoch": 3.449411764705882, "grad_norm": 0.7671464015420071, "learning_rate": 0.00015197257108643406, "loss": 0.4902, "step": 7330 }, { "epoch": 3.4498823529411764, "grad_norm": 0.24307038224781402, "learning_rate": 0.00015192873861683036, "loss": 0.0859, "step": 7331 }, { "epoch": 3.4503529411764706, "grad_norm": 0.4305110453449784, "learning_rate": 0.00015188490598250288, "loss": 0.2178, "step": 7332 }, { "epoch": 3.450823529411765, "grad_norm": 0.5933637134235178, "learning_rate": 0.0001518410731871952, "loss": 0.5469, "step": 7333 }, { "epoch": 3.451294117647059, "grad_norm": 0.37315426079475095, "learning_rate": 0.00015179724023465084, "loss": 0.1533, "step": 7334 }, { "epoch": 3.451764705882353, "grad_norm": 0.46079481968529457, "learning_rate": 0.00015175340712861337, "loss": 0.1846, "step": 7335 }, { "epoch": 3.452235294117647, "grad_norm": 0.6172385748902328, "learning_rate": 0.00015170957387282627, "loss": 0.3652, "step": 7336 }, { "epoch": 3.4527058823529413, "grad_norm": 0.6178378717203035, "learning_rate": 0.00015166574047103322, "loss": 0.2158, "step": 7337 }, { "epoch": 3.453176470588235, "grad_norm": 0.5163384770436421, "learning_rate": 0.00015162190692697778, "loss": 0.332, "step": 7338 }, { "epoch": 3.4536470588235293, "grad_norm": 0.7442031083201949, "learning_rate": 0.00015157807324440347, "loss": 0.4824, "step": 7339 }, { "epoch": 3.4541176470588235, "grad_norm": 0.20155199299835033, "learning_rate": 0.00015153423942705402, "loss": 0.0771, "step": 7340 }, { "epoch": 3.4545882352941177, "grad_norm": 0.5719635830200004, "learning_rate": 0.000151490405478673, "loss": 0.2197, "step": 7341 }, { "epoch": 3.455058823529412, "grad_norm": 0.880711736994779, "learning_rate": 0.000151446571403004, "loss": 0.5977, "step": 7342 }, { "epoch": 3.4555294117647057, "grad_norm": 0.43143006236110804, "learning_rate": 0.0001514027372037907, "loss": 0.2559, "step": 7343 }, { "epoch": 3.456, "grad_norm": 0.38612303633960265, "learning_rate": 0.0001513589028847768, "loss": 0.1777, "step": 7344 }, { "epoch": 3.456470588235294, "grad_norm": 0.8052424725150349, "learning_rate": 0.00015131506844970596, "loss": 0.5938, "step": 7345 }, { "epoch": 3.4569411764705884, "grad_norm": 0.22444135375045068, "learning_rate": 0.00015127123390232177, "loss": 0.0408, "step": 7346 }, { "epoch": 3.457411764705882, "grad_norm": 0.9378692715119614, "learning_rate": 0.00015122739924636796, "loss": 0.543, "step": 7347 }, { "epoch": 3.4578823529411764, "grad_norm": 0.5079111247069883, "learning_rate": 0.00015118356448558828, "loss": 0.2949, "step": 7348 }, { "epoch": 3.4583529411764706, "grad_norm": 0.36978200807198414, "learning_rate": 0.00015113972962372635, "loss": 0.1465, "step": 7349 }, { "epoch": 3.458823529411765, "grad_norm": 0.18222009809604722, "learning_rate": 0.00015109589466452594, "loss": 0.0354, "step": 7350 }, { "epoch": 3.4592941176470586, "grad_norm": 0.34078718077964304, "learning_rate": 0.00015105205961173074, "loss": 0.1475, "step": 7351 }, { "epoch": 3.459764705882353, "grad_norm": 0.24984433882396495, "learning_rate": 0.00015100822446908452, "loss": 0.0835, "step": 7352 }, { "epoch": 3.460235294117647, "grad_norm": 0.39697150950899124, "learning_rate": 0.00015096438924033095, "loss": 0.1064, "step": 7353 }, { "epoch": 3.4607058823529413, "grad_norm": 0.8923580865743443, "learning_rate": 0.00015092055392921383, "loss": 0.4824, "step": 7354 }, { "epoch": 3.461176470588235, "grad_norm": 0.5189835893381602, "learning_rate": 0.00015087671853947684, "loss": 0.1895, "step": 7355 }, { "epoch": 3.4616470588235293, "grad_norm": 0.7768577206362404, "learning_rate": 0.00015083288307486388, "loss": 0.4668, "step": 7356 }, { "epoch": 3.4621176470588235, "grad_norm": 0.35916857475268865, "learning_rate": 0.00015078904753911853, "loss": 0.1826, "step": 7357 }, { "epoch": 3.4625882352941177, "grad_norm": 0.6997848063629188, "learning_rate": 0.0001507452119359847, "loss": 0.2656, "step": 7358 }, { "epoch": 3.463058823529412, "grad_norm": 0.2939890897842972, "learning_rate": 0.0001507013762692061, "loss": 0.1055, "step": 7359 }, { "epoch": 3.4635294117647057, "grad_norm": 0.589090452217865, "learning_rate": 0.0001506575405425265, "loss": 0.2324, "step": 7360 }, { "epoch": 3.464, "grad_norm": 0.16406682942360376, "learning_rate": 0.00015061370475968975, "loss": 0.0334, "step": 7361 }, { "epoch": 3.464470588235294, "grad_norm": 0.800165389047552, "learning_rate": 0.0001505698689244396, "loss": 0.3086, "step": 7362 }, { "epoch": 3.4649411764705884, "grad_norm": 0.492720191073766, "learning_rate": 0.0001505260330405198, "loss": 0.2637, "step": 7363 }, { "epoch": 3.465411764705882, "grad_norm": 0.49690235453132703, "learning_rate": 0.00015048219711167422, "loss": 0.2305, "step": 7364 }, { "epoch": 3.4658823529411764, "grad_norm": 0.12842548387338196, "learning_rate": 0.00015043836114164667, "loss": 0.0295, "step": 7365 }, { "epoch": 3.4663529411764706, "grad_norm": 0.2632075268857782, "learning_rate": 0.00015039452513418094, "loss": 0.1226, "step": 7366 }, { "epoch": 3.466823529411765, "grad_norm": 0.742344654636554, "learning_rate": 0.00015035068909302079, "loss": 0.2988, "step": 7367 }, { "epoch": 3.4672941176470586, "grad_norm": 0.6545278174732996, "learning_rate": 0.00015030685302191011, "loss": 0.293, "step": 7368 }, { "epoch": 3.467764705882353, "grad_norm": 0.5453679670318203, "learning_rate": 0.00015026301692459265, "loss": 0.1895, "step": 7369 }, { "epoch": 3.468235294117647, "grad_norm": 0.4236671050336758, "learning_rate": 0.00015021918080481225, "loss": 0.1904, "step": 7370 }, { "epoch": 3.4687058823529413, "grad_norm": 0.4197102394784948, "learning_rate": 0.00015017534466631272, "loss": 0.1807, "step": 7371 }, { "epoch": 3.469176470588235, "grad_norm": 0.9283927194161959, "learning_rate": 0.00015013150851283795, "loss": 0.6289, "step": 7372 }, { "epoch": 3.4696470588235293, "grad_norm": 0.3298918046025511, "learning_rate": 0.0001500876723481317, "loss": 0.1514, "step": 7373 }, { "epoch": 3.4701176470588235, "grad_norm": 0.6162536631151355, "learning_rate": 0.00015004383617593772, "loss": 0.3809, "step": 7374 }, { "epoch": 3.4705882352941178, "grad_norm": 0.5123769608969128, "learning_rate": 0.00015, "loss": 0.2871, "step": 7375 }, { "epoch": 3.471058823529412, "grad_norm": 0.5441855547170245, "learning_rate": 0.00014995616382406225, "loss": 0.2559, "step": 7376 }, { "epoch": 3.4715294117647058, "grad_norm": 0.38926972526423315, "learning_rate": 0.00014991232765186832, "loss": 0.1494, "step": 7377 }, { "epoch": 3.472, "grad_norm": 0.8047805243082046, "learning_rate": 0.00014986849148716205, "loss": 0.3359, "step": 7378 }, { "epoch": 3.472470588235294, "grad_norm": 0.504668652862156, "learning_rate": 0.00014982465533368725, "loss": 0.2393, "step": 7379 }, { "epoch": 3.4729411764705884, "grad_norm": 0.7842562859011156, "learning_rate": 0.00014978081919518772, "loss": 0.5547, "step": 7380 }, { "epoch": 3.473411764705882, "grad_norm": 0.3889732966022988, "learning_rate": 0.00014973698307540735, "loss": 0.2266, "step": 7381 }, { "epoch": 3.4738823529411764, "grad_norm": 0.7332826687216667, "learning_rate": 0.00014969314697808991, "loss": 0.3145, "step": 7382 }, { "epoch": 3.4743529411764706, "grad_norm": 0.8434746094542703, "learning_rate": 0.00014964931090697922, "loss": 0.3945, "step": 7383 }, { "epoch": 3.474823529411765, "grad_norm": 0.4303999721328547, "learning_rate": 0.00014960547486581906, "loss": 0.1768, "step": 7384 }, { "epoch": 3.4752941176470586, "grad_norm": 0.4882860472334698, "learning_rate": 0.0001495616388583533, "loss": 0.3477, "step": 7385 }, { "epoch": 3.475764705882353, "grad_norm": 0.42282410672466925, "learning_rate": 0.00014951780288832575, "loss": 0.1943, "step": 7386 }, { "epoch": 3.476235294117647, "grad_norm": 0.37797257562612874, "learning_rate": 0.0001494739669594802, "loss": 0.168, "step": 7387 }, { "epoch": 3.4767058823529413, "grad_norm": 0.9626394486852451, "learning_rate": 0.00014943013107556044, "loss": 0.2949, "step": 7388 }, { "epoch": 3.477176470588235, "grad_norm": 0.42840306710058573, "learning_rate": 0.00014938629524031028, "loss": 0.1855, "step": 7389 }, { "epoch": 3.4776470588235293, "grad_norm": 0.34841627154503824, "learning_rate": 0.00014934245945747349, "loss": 0.1118, "step": 7390 }, { "epoch": 3.4781176470588235, "grad_norm": 0.7182041552524872, "learning_rate": 0.00014929862373079392, "loss": 0.2793, "step": 7391 }, { "epoch": 3.4785882352941178, "grad_norm": 0.4239256855800998, "learning_rate": 0.00014925478806401533, "loss": 0.1953, "step": 7392 }, { "epoch": 3.479058823529412, "grad_norm": 0.33861133523363296, "learning_rate": 0.0001492109524608815, "loss": 0.1357, "step": 7393 }, { "epoch": 3.4795294117647058, "grad_norm": 0.2518886669380219, "learning_rate": 0.00014916711692513618, "loss": 0.085, "step": 7394 }, { "epoch": 3.48, "grad_norm": 0.23542160447858781, "learning_rate": 0.00014912328146052316, "loss": 0.1118, "step": 7395 }, { "epoch": 3.480470588235294, "grad_norm": 0.3634074962587284, "learning_rate": 0.00014907944607078623, "loss": 0.1738, "step": 7396 }, { "epoch": 3.4809411764705884, "grad_norm": 0.4531529888195629, "learning_rate": 0.00014903561075966908, "loss": 0.4336, "step": 7397 }, { "epoch": 3.481411764705882, "grad_norm": 0.5974160573770988, "learning_rate": 0.00014899177553091543, "loss": 0.252, "step": 7398 }, { "epoch": 3.4818823529411764, "grad_norm": 0.5259756590913195, "learning_rate": 0.0001489479403882692, "loss": 0.2656, "step": 7399 }, { "epoch": 3.4823529411764707, "grad_norm": 0.4236345332947406, "learning_rate": 0.00014890410533547404, "loss": 0.1406, "step": 7400 }, { "epoch": 3.482823529411765, "grad_norm": 0.3274238227892056, "learning_rate": 0.00014886027037627362, "loss": 0.1143, "step": 7401 }, { "epoch": 3.4832941176470587, "grad_norm": 0.3537904756498093, "learning_rate": 0.0001488164355144117, "loss": 0.1201, "step": 7402 }, { "epoch": 3.483764705882353, "grad_norm": 0.695190044208733, "learning_rate": 0.00014877260075363199, "loss": 0.3906, "step": 7403 }, { "epoch": 3.484235294117647, "grad_norm": 0.23211142422823075, "learning_rate": 0.00014872876609767823, "loss": 0.1572, "step": 7404 }, { "epoch": 3.4847058823529413, "grad_norm": 0.6063717127373401, "learning_rate": 0.00014868493155029402, "loss": 0.2637, "step": 7405 }, { "epoch": 3.485176470588235, "grad_norm": 0.607842834973418, "learning_rate": 0.00014864109711522315, "loss": 0.3906, "step": 7406 }, { "epoch": 3.4856470588235293, "grad_norm": 0.4767502830465428, "learning_rate": 0.00014859726279620926, "loss": 0.2773, "step": 7407 }, { "epoch": 3.4861176470588235, "grad_norm": 0.6147425740590834, "learning_rate": 0.00014855342859699597, "loss": 0.4336, "step": 7408 }, { "epoch": 3.4865882352941178, "grad_norm": 0.608854803718036, "learning_rate": 0.000148509594521327, "loss": 0.3574, "step": 7409 }, { "epoch": 3.487058823529412, "grad_norm": 0.347249019129702, "learning_rate": 0.00014846576057294596, "loss": 0.1167, "step": 7410 }, { "epoch": 3.4875294117647058, "grad_norm": 0.43282541187106166, "learning_rate": 0.0001484219267555965, "loss": 0.1885, "step": 7411 }, { "epoch": 3.488, "grad_norm": 0.2109621754868006, "learning_rate": 0.0001483780930730222, "loss": 0.0525, "step": 7412 }, { "epoch": 3.488470588235294, "grad_norm": 0.20030953055080872, "learning_rate": 0.00014833425952896675, "loss": 0.0583, "step": 7413 }, { "epoch": 3.488941176470588, "grad_norm": 0.37137970005929494, "learning_rate": 0.00014829042612717373, "loss": 0.2051, "step": 7414 }, { "epoch": 3.489411764705882, "grad_norm": 0.6008410151454414, "learning_rate": 0.00014824659287138663, "loss": 0.543, "step": 7415 }, { "epoch": 3.4898823529411764, "grad_norm": 0.49379065043739606, "learning_rate": 0.00014820275976534916, "loss": 0.1885, "step": 7416 }, { "epoch": 3.4903529411764707, "grad_norm": 0.31779925850483154, "learning_rate": 0.0001481589268128048, "loss": 0.1016, "step": 7417 }, { "epoch": 3.490823529411765, "grad_norm": 0.25288375476234703, "learning_rate": 0.0001481150940174971, "loss": 0.1133, "step": 7418 }, { "epoch": 3.4912941176470587, "grad_norm": 0.7095369364668592, "learning_rate": 0.00014807126138316964, "loss": 0.4297, "step": 7419 }, { "epoch": 3.491764705882353, "grad_norm": 0.5610103249289692, "learning_rate": 0.00014802742891356592, "loss": 0.2266, "step": 7420 }, { "epoch": 3.492235294117647, "grad_norm": 0.7817753392726605, "learning_rate": 0.00014798359661242943, "loss": 0.373, "step": 7421 }, { "epoch": 3.4927058823529413, "grad_norm": 0.24164277743946116, "learning_rate": 0.00014793976448350366, "loss": 0.1191, "step": 7422 }, { "epoch": 3.493176470588235, "grad_norm": 0.8634384811587388, "learning_rate": 0.00014789593253053217, "loss": 0.4102, "step": 7423 }, { "epoch": 3.4936470588235293, "grad_norm": 0.14847382149537788, "learning_rate": 0.00014785210075725835, "loss": 0.0574, "step": 7424 }, { "epoch": 3.4941176470588236, "grad_norm": 0.6752347735804141, "learning_rate": 0.0001478082691674256, "loss": 0.334, "step": 7425 }, { "epoch": 3.4945882352941178, "grad_norm": 0.6041649741992117, "learning_rate": 0.00014776443776477746, "loss": 0.2891, "step": 7426 }, { "epoch": 3.495058823529412, "grad_norm": 0.5874171228797924, "learning_rate": 0.00014772060655305732, "loss": 0.3516, "step": 7427 }, { "epoch": 3.4955294117647058, "grad_norm": 0.6079287922754621, "learning_rate": 0.0001476767755360085, "loss": 0.2793, "step": 7428 }, { "epoch": 3.496, "grad_norm": 0.5672166064456815, "learning_rate": 0.00014763294471737448, "loss": 0.3047, "step": 7429 }, { "epoch": 3.496470588235294, "grad_norm": 0.6564117137739369, "learning_rate": 0.00014758911410089862, "loss": 0.4492, "step": 7430 }, { "epoch": 3.496941176470588, "grad_norm": 0.5981539134121222, "learning_rate": 0.0001475452836903242, "loss": 0.373, "step": 7431 }, { "epoch": 3.497411764705882, "grad_norm": 0.22630628311755227, "learning_rate": 0.0001475014534893946, "loss": 0.0645, "step": 7432 }, { "epoch": 3.4978823529411764, "grad_norm": 0.8244777994584972, "learning_rate": 0.00014745762350185316, "loss": 0.4922, "step": 7433 }, { "epoch": 3.4983529411764707, "grad_norm": 0.5671638627210518, "learning_rate": 0.0001474137937314431, "loss": 0.3145, "step": 7434 }, { "epoch": 3.498823529411765, "grad_norm": 0.5331190281926227, "learning_rate": 0.00014736996418190774, "loss": 0.2363, "step": 7435 }, { "epoch": 3.4992941176470587, "grad_norm": 0.6161258018200284, "learning_rate": 0.00014732613485699035, "loss": 0.4199, "step": 7436 }, { "epoch": 3.499764705882353, "grad_norm": 0.30491919717380106, "learning_rate": 0.0001472823057604342, "loss": 0.1562, "step": 7437 }, { "epoch": 3.500235294117647, "grad_norm": 0.3028361990159035, "learning_rate": 0.0001472384768959824, "loss": 0.2129, "step": 7438 }, { "epoch": 3.5007058823529413, "grad_norm": 0.2984896829036563, "learning_rate": 0.00014719464826737827, "loss": 0.1494, "step": 7439 }, { "epoch": 3.501176470588235, "grad_norm": 0.40207462210496403, "learning_rate": 0.0001471508198783649, "loss": 0.1436, "step": 7440 }, { "epoch": 3.5016470588235293, "grad_norm": 0.24198054910031144, "learning_rate": 0.0001471069917326855, "loss": 0.083, "step": 7441 }, { "epoch": 3.5021176470588236, "grad_norm": 0.6761736593688386, "learning_rate": 0.0001470631638340832, "loss": 0.3711, "step": 7442 }, { "epoch": 3.502588235294118, "grad_norm": 0.7286720205219767, "learning_rate": 0.00014701933618630107, "loss": 0.416, "step": 7443 }, { "epoch": 3.503058823529412, "grad_norm": 0.7288693767729851, "learning_rate": 0.00014697550879308233, "loss": 0.4648, "step": 7444 }, { "epoch": 3.503529411764706, "grad_norm": 0.6529724955740291, "learning_rate": 0.00014693168165816995, "loss": 0.3008, "step": 7445 }, { "epoch": 3.504, "grad_norm": 0.2224704258834395, "learning_rate": 0.00014688785478530697, "loss": 0.0479, "step": 7446 }, { "epoch": 3.5044705882352942, "grad_norm": 0.4699186234160071, "learning_rate": 0.00014684402817823647, "loss": 0.1709, "step": 7447 }, { "epoch": 3.504941176470588, "grad_norm": 0.3046683712585886, "learning_rate": 0.00014680020184070148, "loss": 0.0649, "step": 7448 }, { "epoch": 3.5054117647058822, "grad_norm": 0.24346601705482157, "learning_rate": 0.0001467563757764449, "loss": 0.1201, "step": 7449 }, { "epoch": 3.5058823529411764, "grad_norm": 0.3951533583006946, "learning_rate": 0.00014671254998920976, "loss": 0.1426, "step": 7450 }, { "epoch": 3.5063529411764707, "grad_norm": 0.5706030704837164, "learning_rate": 0.000146668724482739, "loss": 0.3418, "step": 7451 }, { "epoch": 3.506823529411765, "grad_norm": 1.0406881938539412, "learning_rate": 0.0001466248992607755, "loss": 0.4023, "step": 7452 }, { "epoch": 3.5072941176470587, "grad_norm": 0.2229011487040093, "learning_rate": 0.00014658107432706212, "loss": 0.0698, "step": 7453 }, { "epoch": 3.507764705882353, "grad_norm": 0.6313747324834659, "learning_rate": 0.0001465372496853418, "loss": 0.3691, "step": 7454 }, { "epoch": 3.508235294117647, "grad_norm": 0.17580959835946902, "learning_rate": 0.00014649342533935733, "loss": 0.0366, "step": 7455 }, { "epoch": 3.5087058823529413, "grad_norm": 0.6336700209688402, "learning_rate": 0.00014644960129285152, "loss": 0.2891, "step": 7456 }, { "epoch": 3.509176470588235, "grad_norm": 0.29670367457674474, "learning_rate": 0.0001464057775495672, "loss": 0.0913, "step": 7457 }, { "epoch": 3.5096470588235293, "grad_norm": 0.7836395485399673, "learning_rate": 0.0001463619541132471, "loss": 0.5234, "step": 7458 }, { "epoch": 3.5101176470588236, "grad_norm": 0.5285123858230746, "learning_rate": 0.00014631813098763392, "loss": 0.2773, "step": 7459 }, { "epoch": 3.510588235294118, "grad_norm": 1.0342071264259995, "learning_rate": 0.00014627430817647048, "loss": 0.4824, "step": 7460 }, { "epoch": 3.511058823529412, "grad_norm": 0.28837636453388354, "learning_rate": 0.00014623048568349937, "loss": 0.0598, "step": 7461 }, { "epoch": 3.511529411764706, "grad_norm": 0.11305162749110478, "learning_rate": 0.0001461866635124633, "loss": 0.027, "step": 7462 }, { "epoch": 3.512, "grad_norm": 0.6140809068057774, "learning_rate": 0.0001461428416671048, "loss": 0.252, "step": 7463 }, { "epoch": 3.5124705882352942, "grad_norm": 0.6063955543603217, "learning_rate": 0.0001460990201511666, "loss": 0.3281, "step": 7464 }, { "epoch": 3.512941176470588, "grad_norm": 0.21871720927225224, "learning_rate": 0.0001460551989683912, "loss": 0.0776, "step": 7465 }, { "epoch": 3.5134117647058822, "grad_norm": 0.5322676493943964, "learning_rate": 0.00014601137812252112, "loss": 0.2188, "step": 7466 }, { "epoch": 3.5138823529411765, "grad_norm": 0.6434364595105068, "learning_rate": 0.00014596755761729895, "loss": 0.3145, "step": 7467 }, { "epoch": 3.5143529411764707, "grad_norm": 0.9610064216591544, "learning_rate": 0.00014592373745646715, "loss": 0.6602, "step": 7468 }, { "epoch": 3.514823529411765, "grad_norm": 0.39743793327780996, "learning_rate": 0.00014587991764376812, "loss": 0.2578, "step": 7469 }, { "epoch": 3.5152941176470587, "grad_norm": 0.3662067158115943, "learning_rate": 0.00014583609818294438, "loss": 0.1787, "step": 7470 }, { "epoch": 3.515764705882353, "grad_norm": 0.5851240197457293, "learning_rate": 0.00014579227907773827, "loss": 0.2412, "step": 7471 }, { "epoch": 3.516235294117647, "grad_norm": 0.5617337254533348, "learning_rate": 0.00014574846033189217, "loss": 0.2871, "step": 7472 }, { "epoch": 3.516705882352941, "grad_norm": 0.5993820297964947, "learning_rate": 0.00014570464194914835, "loss": 0.3672, "step": 7473 }, { "epoch": 3.517176470588235, "grad_norm": 0.3538532859438315, "learning_rate": 0.00014566082393324925, "loss": 0.1641, "step": 7474 }, { "epoch": 3.5176470588235293, "grad_norm": 0.43634933172949175, "learning_rate": 0.00014561700628793706, "loss": 0.2793, "step": 7475 }, { "epoch": 3.5181176470588236, "grad_norm": 0.281552267657537, "learning_rate": 0.000145573189016954, "loss": 0.0776, "step": 7476 }, { "epoch": 3.518588235294118, "grad_norm": 0.6884128568427125, "learning_rate": 0.0001455293721240423, "loss": 0.5117, "step": 7477 }, { "epoch": 3.519058823529412, "grad_norm": 0.46186700869078096, "learning_rate": 0.0001454855556129442, "loss": 0.2891, "step": 7478 }, { "epoch": 3.519529411764706, "grad_norm": 0.5606854207499008, "learning_rate": 0.00014544173948740174, "loss": 0.334, "step": 7479 }, { "epoch": 3.52, "grad_norm": 0.3110890907368806, "learning_rate": 0.00014539792375115713, "loss": 0.1611, "step": 7480 }, { "epoch": 3.5204705882352942, "grad_norm": 0.4384480221298332, "learning_rate": 0.0001453541084079524, "loss": 0.2236, "step": 7481 }, { "epoch": 3.520941176470588, "grad_norm": 0.4561781407665828, "learning_rate": 0.0001453102934615296, "loss": 0.291, "step": 7482 }, { "epoch": 3.5214117647058822, "grad_norm": 0.38565551044594426, "learning_rate": 0.0001452664789156307, "loss": 0.1245, "step": 7483 }, { "epoch": 3.5218823529411765, "grad_norm": 0.4754959896417893, "learning_rate": 0.00014522266477399776, "loss": 0.373, "step": 7484 }, { "epoch": 3.5223529411764707, "grad_norm": 0.1911865672382976, "learning_rate": 0.00014517885104037268, "loss": 0.0801, "step": 7485 }, { "epoch": 3.522823529411765, "grad_norm": 0.4836618515735618, "learning_rate": 0.00014513503771849734, "loss": 0.2256, "step": 7486 }, { "epoch": 3.5232941176470587, "grad_norm": 0.6860011061649406, "learning_rate": 0.00014509122481211358, "loss": 0.1436, "step": 7487 }, { "epoch": 3.523764705882353, "grad_norm": 0.35673773819810856, "learning_rate": 0.0001450474123249634, "loss": 0.1318, "step": 7488 }, { "epoch": 3.524235294117647, "grad_norm": 0.3411326781760001, "learning_rate": 0.00014500360026078847, "loss": 0.1514, "step": 7489 }, { "epoch": 3.524705882352941, "grad_norm": 0.9016984584162527, "learning_rate": 0.00014495978862333063, "loss": 0.5625, "step": 7490 }, { "epoch": 3.525176470588235, "grad_norm": 0.3727787517529819, "learning_rate": 0.00014491597741633148, "loss": 0.1885, "step": 7491 }, { "epoch": 3.5256470588235294, "grad_norm": 0.6177476130465916, "learning_rate": 0.00014487216664353287, "loss": 0.3066, "step": 7492 }, { "epoch": 3.5261176470588236, "grad_norm": 0.4092193391000257, "learning_rate": 0.00014482835630867636, "loss": 0.1963, "step": 7493 }, { "epoch": 3.526588235294118, "grad_norm": 0.3489354210315971, "learning_rate": 0.00014478454641550354, "loss": 0.1699, "step": 7494 }, { "epoch": 3.527058823529412, "grad_norm": 0.26698319482028804, "learning_rate": 0.0001447407369677561, "loss": 0.1172, "step": 7495 }, { "epoch": 3.527529411764706, "grad_norm": 0.597195859885424, "learning_rate": 0.0001446969279691755, "loss": 0.1875, "step": 7496 }, { "epoch": 3.528, "grad_norm": 0.9480495729898599, "learning_rate": 0.00014465311942350324, "loss": 0.4238, "step": 7497 }, { "epoch": 3.5284705882352942, "grad_norm": 0.4980108514581985, "learning_rate": 0.00014460931133448084, "loss": 0.2539, "step": 7498 }, { "epoch": 3.528941176470588, "grad_norm": 0.973758967544018, "learning_rate": 0.0001445655037058497, "loss": 0.5898, "step": 7499 }, { "epoch": 3.5294117647058822, "grad_norm": 0.5235169487072548, "learning_rate": 0.00014452169654135115, "loss": 0.2305, "step": 7500 }, { "epoch": 3.5298823529411765, "grad_norm": 0.7433734732010981, "learning_rate": 0.0001444778898447266, "loss": 0.4629, "step": 7501 }, { "epoch": 3.5303529411764707, "grad_norm": 0.7918573831858077, "learning_rate": 0.00014443408361971739, "loss": 0.3535, "step": 7502 }, { "epoch": 3.530823529411765, "grad_norm": 0.4619229680032333, "learning_rate": 0.0001443902778700647, "loss": 0.2656, "step": 7503 }, { "epoch": 3.5312941176470587, "grad_norm": 0.577551205408211, "learning_rate": 0.0001443464725995098, "loss": 0.4082, "step": 7504 }, { "epoch": 3.531764705882353, "grad_norm": 0.4251606670321753, "learning_rate": 0.0001443026678117939, "loss": 0.2656, "step": 7505 }, { "epoch": 3.532235294117647, "grad_norm": 0.46903369968939407, "learning_rate": 0.00014425886351065808, "loss": 0.3105, "step": 7506 }, { "epoch": 3.532705882352941, "grad_norm": 0.3322068187547496, "learning_rate": 0.00014421505969984346, "loss": 0.1475, "step": 7507 }, { "epoch": 3.533176470588235, "grad_norm": 0.14465526461215814, "learning_rate": 0.00014417125638309115, "loss": 0.0347, "step": 7508 }, { "epoch": 3.5336470588235294, "grad_norm": 0.3762302285357426, "learning_rate": 0.00014412745356414213, "loss": 0.1348, "step": 7509 }, { "epoch": 3.5341176470588236, "grad_norm": 0.5804828203681338, "learning_rate": 0.00014408365124673737, "loss": 0.3438, "step": 7510 }, { "epoch": 3.534588235294118, "grad_norm": 0.45729966261914096, "learning_rate": 0.00014403984943461783, "loss": 0.2373, "step": 7511 }, { "epoch": 3.535058823529412, "grad_norm": 0.11592952810933034, "learning_rate": 0.0001439960481315244, "loss": 0.0292, "step": 7512 }, { "epoch": 3.535529411764706, "grad_norm": 0.602461834562906, "learning_rate": 0.00014395224734119788, "loss": 0.3359, "step": 7513 }, { "epoch": 3.536, "grad_norm": 0.2449101192211691, "learning_rate": 0.00014390844706737907, "loss": 0.0596, "step": 7514 }, { "epoch": 3.5364705882352943, "grad_norm": 0.7689760788471467, "learning_rate": 0.00014386464731380881, "loss": 0.4336, "step": 7515 }, { "epoch": 3.536941176470588, "grad_norm": 0.32221187564715675, "learning_rate": 0.00014382084808422775, "loss": 0.1689, "step": 7516 }, { "epoch": 3.5374117647058823, "grad_norm": 0.3801512092424246, "learning_rate": 0.00014377704938237654, "loss": 0.124, "step": 7517 }, { "epoch": 3.5378823529411765, "grad_norm": 0.48376929490860543, "learning_rate": 0.00014373325121199587, "loss": 0.2158, "step": 7518 }, { "epoch": 3.5383529411764707, "grad_norm": 0.665126126839154, "learning_rate": 0.00014368945357682628, "loss": 0.5, "step": 7519 }, { "epoch": 3.538823529411765, "grad_norm": 0.6167123995988958, "learning_rate": 0.0001436456564806083, "loss": 0.3555, "step": 7520 }, { "epoch": 3.5392941176470587, "grad_norm": 0.7369905329587739, "learning_rate": 0.00014360185992708243, "loss": 0.4785, "step": 7521 }, { "epoch": 3.539764705882353, "grad_norm": 0.45561243808085616, "learning_rate": 0.00014355806391998913, "loss": 0.2656, "step": 7522 }, { "epoch": 3.540235294117647, "grad_norm": 0.15549476412091426, "learning_rate": 0.00014351426846306875, "loss": 0.0442, "step": 7523 }, { "epoch": 3.540705882352941, "grad_norm": 0.506948559208105, "learning_rate": 0.00014347047356006161, "loss": 0.3105, "step": 7524 }, { "epoch": 3.541176470588235, "grad_norm": 0.7100725907212047, "learning_rate": 0.00014342667921470812, "loss": 0.7852, "step": 7525 }, { "epoch": 3.5416470588235294, "grad_norm": 0.5938392589420749, "learning_rate": 0.00014338288543074845, "loss": 0.4062, "step": 7526 }, { "epoch": 3.5421176470588236, "grad_norm": 0.5541794382554792, "learning_rate": 0.00014333909221192278, "loss": 0.1797, "step": 7527 }, { "epoch": 3.542588235294118, "grad_norm": 0.44664476290156363, "learning_rate": 0.00014329529956197135, "loss": 0.1621, "step": 7528 }, { "epoch": 3.543058823529412, "grad_norm": 0.8326107913012162, "learning_rate": 0.00014325150748463423, "loss": 0.332, "step": 7529 }, { "epoch": 3.543529411764706, "grad_norm": 0.4621466907214265, "learning_rate": 0.00014320771598365142, "loss": 0.2363, "step": 7530 }, { "epoch": 3.544, "grad_norm": 0.4715085691016033, "learning_rate": 0.00014316392506276302, "loss": 0.1982, "step": 7531 }, { "epoch": 3.5444705882352943, "grad_norm": 0.36386338186134476, "learning_rate": 0.0001431201347257089, "loss": 0.1943, "step": 7532 }, { "epoch": 3.544941176470588, "grad_norm": 0.37942211304981355, "learning_rate": 0.00014307634497622907, "loss": 0.1367, "step": 7533 }, { "epoch": 3.5454117647058823, "grad_norm": 0.43950880740109755, "learning_rate": 0.00014303255581806336, "loss": 0.3047, "step": 7534 }, { "epoch": 3.5458823529411765, "grad_norm": 0.41316686226094995, "learning_rate": 0.0001429887672549515, "loss": 0.2334, "step": 7535 }, { "epoch": 3.5463529411764707, "grad_norm": 0.6590032936351659, "learning_rate": 0.00014294497929063336, "loss": 0.2432, "step": 7536 }, { "epoch": 3.546823529411765, "grad_norm": 0.6091378288509837, "learning_rate": 0.0001429011919288486, "loss": 0.3066, "step": 7537 }, { "epoch": 3.5472941176470587, "grad_norm": 0.5159509412573887, "learning_rate": 0.0001428574051733368, "loss": 0.1562, "step": 7538 }, { "epoch": 3.547764705882353, "grad_norm": 0.3690697574339191, "learning_rate": 0.0001428136190278377, "loss": 0.1377, "step": 7539 }, { "epoch": 3.548235294117647, "grad_norm": 0.5714320895686184, "learning_rate": 0.00014276983349609075, "loss": 0.3301, "step": 7540 }, { "epoch": 3.548705882352941, "grad_norm": 0.6281803006309314, "learning_rate": 0.0001427260485818355, "loss": 0.3359, "step": 7541 }, { "epoch": 3.549176470588235, "grad_norm": 0.7439998656532882, "learning_rate": 0.00014268226428881136, "loss": 0.6211, "step": 7542 }, { "epoch": 3.5496470588235294, "grad_norm": 0.8133531370476403, "learning_rate": 0.00014263848062075776, "loss": 0.6016, "step": 7543 }, { "epoch": 3.5501176470588236, "grad_norm": 0.29224999732210805, "learning_rate": 0.00014259469758141403, "loss": 0.0933, "step": 7544 }, { "epoch": 3.550588235294118, "grad_norm": 0.6208931900698705, "learning_rate": 0.00014255091517451941, "loss": 0.3535, "step": 7545 }, { "epoch": 3.5510588235294116, "grad_norm": 0.46706393747338487, "learning_rate": 0.0001425071334038132, "loss": 0.2012, "step": 7546 }, { "epoch": 3.551529411764706, "grad_norm": 0.3632403584281262, "learning_rate": 0.00014246335227303456, "loss": 0.209, "step": 7547 }, { "epoch": 3.552, "grad_norm": 0.6084779385287228, "learning_rate": 0.00014241957178592258, "loss": 0.3379, "step": 7548 }, { "epoch": 3.5524705882352943, "grad_norm": 0.5199131676238836, "learning_rate": 0.00014237579194621637, "loss": 0.3418, "step": 7549 }, { "epoch": 3.552941176470588, "grad_norm": 0.4250129193913078, "learning_rate": 0.00014233201275765494, "loss": 0.125, "step": 7550 }, { "epoch": 3.5534117647058823, "grad_norm": 0.22649991315931417, "learning_rate": 0.0001422882342239772, "loss": 0.0811, "step": 7551 }, { "epoch": 3.5538823529411765, "grad_norm": 0.6464074922479648, "learning_rate": 0.00014224445634892206, "loss": 0.2578, "step": 7552 }, { "epoch": 3.5543529411764707, "grad_norm": 0.35690234799457615, "learning_rate": 0.00014220067913622844, "loss": 0.1973, "step": 7553 }, { "epoch": 3.554823529411765, "grad_norm": 0.4998744909232499, "learning_rate": 0.00014215690258963508, "loss": 0.2832, "step": 7554 }, { "epoch": 3.5552941176470587, "grad_norm": 0.2925366785889896, "learning_rate": 0.00014211312671288064, "loss": 0.0576, "step": 7555 }, { "epoch": 3.555764705882353, "grad_norm": 0.4222410903631462, "learning_rate": 0.00014206935150970392, "loss": 0.124, "step": 7556 }, { "epoch": 3.556235294117647, "grad_norm": 0.7788900113475093, "learning_rate": 0.00014202557698384348, "loss": 0.4844, "step": 7557 }, { "epoch": 3.556705882352941, "grad_norm": 0.7129401538260274, "learning_rate": 0.00014198180313903784, "loss": 0.3711, "step": 7558 }, { "epoch": 3.557176470588235, "grad_norm": 0.2875610716338223, "learning_rate": 0.0001419380299790256, "loss": 0.0791, "step": 7559 }, { "epoch": 3.5576470588235294, "grad_norm": 0.5189322265967374, "learning_rate": 0.00014189425750754512, "loss": 0.2676, "step": 7560 }, { "epoch": 3.5581176470588236, "grad_norm": 0.9469894031940349, "learning_rate": 0.00014185048572833478, "loss": 0.5391, "step": 7561 }, { "epoch": 3.558588235294118, "grad_norm": 0.34943193882071205, "learning_rate": 0.00014180671464513296, "loss": 0.1211, "step": 7562 }, { "epoch": 3.5590588235294116, "grad_norm": 0.6427841018858298, "learning_rate": 0.0001417629442616779, "loss": 0.2988, "step": 7563 }, { "epoch": 3.559529411764706, "grad_norm": 0.5725655616047397, "learning_rate": 0.00014171917458170782, "loss": 0.3555, "step": 7564 }, { "epoch": 3.56, "grad_norm": 0.5085744665085183, "learning_rate": 0.0001416754056089608, "loss": 0.2676, "step": 7565 }, { "epoch": 3.5604705882352943, "grad_norm": 0.22151209911002523, "learning_rate": 0.000141631637347175, "loss": 0.0505, "step": 7566 }, { "epoch": 3.560941176470588, "grad_norm": 0.7884941371421211, "learning_rate": 0.0001415878698000884, "loss": 0.2617, "step": 7567 }, { "epoch": 3.5614117647058823, "grad_norm": 0.6281212762182609, "learning_rate": 0.00014154410297143898, "loss": 0.2715, "step": 7568 }, { "epoch": 3.5618823529411765, "grad_norm": 0.28434552206516095, "learning_rate": 0.00014150033686496469, "loss": 0.1187, "step": 7569 }, { "epoch": 3.5623529411764707, "grad_norm": 0.2184597296132763, "learning_rate": 0.0001414565714844033, "loss": 0.064, "step": 7570 }, { "epoch": 3.562823529411765, "grad_norm": 0.384590285960324, "learning_rate": 0.00014141280683349253, "loss": 0.1787, "step": 7571 }, { "epoch": 3.5632941176470587, "grad_norm": 0.7241914522860058, "learning_rate": 0.00014136904291597028, "loss": 0.332, "step": 7572 }, { "epoch": 3.563764705882353, "grad_norm": 0.4430436525390199, "learning_rate": 0.00014132527973557407, "loss": 0.2305, "step": 7573 }, { "epoch": 3.564235294117647, "grad_norm": 0.5421813972728421, "learning_rate": 0.00014128151729604152, "loss": 0.21, "step": 7574 }, { "epoch": 3.564705882352941, "grad_norm": 0.5256613108939372, "learning_rate": 0.00014123775560111009, "loss": 0.125, "step": 7575 }, { "epoch": 3.565176470588235, "grad_norm": 0.34017850838173547, "learning_rate": 0.00014119399465451738, "loss": 0.126, "step": 7576 }, { "epoch": 3.5656470588235294, "grad_norm": 0.6124681819031266, "learning_rate": 0.0001411502344600007, "loss": 0.3555, "step": 7577 }, { "epoch": 3.5661176470588236, "grad_norm": 0.4545194776280319, "learning_rate": 0.00014110647502129742, "loss": 0.166, "step": 7578 }, { "epoch": 3.566588235294118, "grad_norm": 0.6592100893142065, "learning_rate": 0.00014106271634214477, "loss": 0.4141, "step": 7579 }, { "epoch": 3.5670588235294116, "grad_norm": 0.6577689684020016, "learning_rate": 0.00014101895842628001, "loss": 0.3711, "step": 7580 }, { "epoch": 3.567529411764706, "grad_norm": 0.6475024200840916, "learning_rate": 0.00014097520127744024, "loss": 0.3926, "step": 7581 }, { "epoch": 3.568, "grad_norm": 0.3569262621227825, "learning_rate": 0.00014093144489936259, "loss": 0.1191, "step": 7582 }, { "epoch": 3.5684705882352943, "grad_norm": 0.4341496489314862, "learning_rate": 0.00014088768929578397, "loss": 0.252, "step": 7583 }, { "epoch": 3.568941176470588, "grad_norm": 0.2611993944798286, "learning_rate": 0.00014084393447044138, "loss": 0.1113, "step": 7584 }, { "epoch": 3.5694117647058823, "grad_norm": 0.3439935480947578, "learning_rate": 0.00014080018042707173, "loss": 0.1543, "step": 7585 }, { "epoch": 3.5698823529411765, "grad_norm": 0.8965615725596398, "learning_rate": 0.00014075642716941176, "loss": 0.4961, "step": 7586 }, { "epoch": 3.5703529411764707, "grad_norm": 0.9396915750901729, "learning_rate": 0.0001407126747011983, "loss": 0.5039, "step": 7587 }, { "epoch": 3.570823529411765, "grad_norm": 0.8007890401643619, "learning_rate": 0.00014066892302616794, "loss": 0.6523, "step": 7588 }, { "epoch": 3.5712941176470587, "grad_norm": 0.31681381900012046, "learning_rate": 0.0001406251721480573, "loss": 0.1924, "step": 7589 }, { "epoch": 3.571764705882353, "grad_norm": 0.6645369312085051, "learning_rate": 0.000140581422070603, "loss": 0.3184, "step": 7590 }, { "epoch": 3.572235294117647, "grad_norm": 0.32705479440584995, "learning_rate": 0.0001405376727975414, "loss": 0.0869, "step": 7591 }, { "epoch": 3.572705882352941, "grad_norm": 0.6174031530920081, "learning_rate": 0.000140493924332609, "loss": 0.1572, "step": 7592 }, { "epoch": 3.573176470588235, "grad_norm": 0.6944314800059213, "learning_rate": 0.00014045017667954203, "loss": 0.3848, "step": 7593 }, { "epoch": 3.5736470588235294, "grad_norm": 0.4972742703166666, "learning_rate": 0.00014040642984207687, "loss": 0.207, "step": 7594 }, { "epoch": 3.5741176470588236, "grad_norm": 0.47270827059966475, "learning_rate": 0.00014036268382394964, "loss": 0.2275, "step": 7595 }, { "epoch": 3.574588235294118, "grad_norm": 0.519529884070352, "learning_rate": 0.00014031893862889646, "loss": 0.1709, "step": 7596 }, { "epoch": 3.5750588235294116, "grad_norm": 0.5196606533746296, "learning_rate": 0.00014027519426065344, "loss": 0.2852, "step": 7597 }, { "epoch": 3.575529411764706, "grad_norm": 0.2717172921875987, "learning_rate": 0.00014023145072295654, "loss": 0.0908, "step": 7598 }, { "epoch": 3.576, "grad_norm": 0.5581306907842905, "learning_rate": 0.0001401877080195416, "loss": 0.2871, "step": 7599 }, { "epoch": 3.576470588235294, "grad_norm": 0.855097177851949, "learning_rate": 0.00014014396615414458, "loss": 0.4062, "step": 7600 }, { "epoch": 3.576941176470588, "grad_norm": 0.37340406813898774, "learning_rate": 0.0001401002251305012, "loss": 0.126, "step": 7601 }, { "epoch": 3.5774117647058823, "grad_norm": 0.047533730744264, "learning_rate": 0.0001400564849523471, "loss": 0.0043, "step": 7602 }, { "epoch": 3.5778823529411765, "grad_norm": 0.06984995921055126, "learning_rate": 0.00014001274562341798, "loss": 0.0096, "step": 7603 }, { "epoch": 3.5783529411764707, "grad_norm": 0.5552005501331355, "learning_rate": 0.0001399690071474494, "loss": 0.2197, "step": 7604 }, { "epoch": 3.578823529411765, "grad_norm": 1.0607046673603995, "learning_rate": 0.0001399252695281768, "loss": 0.418, "step": 7605 }, { "epoch": 3.5792941176470587, "grad_norm": 0.7220590955882381, "learning_rate": 0.00013988153276933555, "loss": 0.3457, "step": 7606 }, { "epoch": 3.579764705882353, "grad_norm": 0.5910742351078138, "learning_rate": 0.0001398377968746611, "loss": 0.2617, "step": 7607 }, { "epoch": 3.580235294117647, "grad_norm": 0.543911735535666, "learning_rate": 0.0001397940618478886, "loss": 0.1885, "step": 7608 }, { "epoch": 3.580705882352941, "grad_norm": 0.41595135381300535, "learning_rate": 0.00013975032769275326, "loss": 0.1592, "step": 7609 }, { "epoch": 3.581176470588235, "grad_norm": 0.38331540520780827, "learning_rate": 0.00013970659441299026, "loss": 0.2178, "step": 7610 }, { "epoch": 3.5816470588235294, "grad_norm": 0.533782423496375, "learning_rate": 0.00013966286201233457, "loss": 0.2676, "step": 7611 }, { "epoch": 3.5821176470588236, "grad_norm": 0.40748523943966886, "learning_rate": 0.00013961913049452115, "loss": 0.1592, "step": 7612 }, { "epoch": 3.582588235294118, "grad_norm": 0.7646014250938068, "learning_rate": 0.00013957539986328492, "loss": 0.3574, "step": 7613 }, { "epoch": 3.5830588235294116, "grad_norm": 0.35214009445188643, "learning_rate": 0.00013953167012236068, "loss": 0.1196, "step": 7614 }, { "epoch": 3.583529411764706, "grad_norm": 1.0554678612056867, "learning_rate": 0.00013948794127548316, "loss": 0.4434, "step": 7615 }, { "epoch": 3.584, "grad_norm": 0.5954903838623075, "learning_rate": 0.00013944421332638698, "loss": 0.293, "step": 7616 }, { "epoch": 3.584470588235294, "grad_norm": 0.5449018040562187, "learning_rate": 0.0001394004862788068, "loss": 0.2754, "step": 7617 }, { "epoch": 3.584941176470588, "grad_norm": 0.7287350557634409, "learning_rate": 0.00013935676013647708, "loss": 0.3398, "step": 7618 }, { "epoch": 3.5854117647058823, "grad_norm": 0.4171545055960734, "learning_rate": 0.0001393130349031322, "loss": 0.1787, "step": 7619 }, { "epoch": 3.5858823529411765, "grad_norm": 0.6061477561351927, "learning_rate": 0.00013926931058250657, "loss": 0.2891, "step": 7620 }, { "epoch": 3.5863529411764707, "grad_norm": 0.5864185837623277, "learning_rate": 0.00013922558717833445, "loss": 0.293, "step": 7621 }, { "epoch": 3.586823529411765, "grad_norm": 0.7160503542828158, "learning_rate": 0.00013918186469435005, "loss": 0.3301, "step": 7622 }, { "epoch": 3.5872941176470587, "grad_norm": 0.5507392137245307, "learning_rate": 0.00013913814313428748, "loss": 0.3887, "step": 7623 }, { "epoch": 3.587764705882353, "grad_norm": 0.6249714431551722, "learning_rate": 0.00013909442250188074, "loss": 0.4609, "step": 7624 }, { "epoch": 3.588235294117647, "grad_norm": 0.3611836351685308, "learning_rate": 0.00013905070280086386, "loss": 0.1406, "step": 7625 }, { "epoch": 3.588705882352941, "grad_norm": 0.7727792864621266, "learning_rate": 0.00013900698403497062, "loss": 0.3438, "step": 7626 }, { "epoch": 3.589176470588235, "grad_norm": 0.3253915433053675, "learning_rate": 0.00013896326620793488, "loss": 0.105, "step": 7627 }, { "epoch": 3.5896470588235294, "grad_norm": 0.6868971162263935, "learning_rate": 0.00013891954932349033, "loss": 0.3672, "step": 7628 }, { "epoch": 3.5901176470588236, "grad_norm": 1.028079929609729, "learning_rate": 0.00013887583338537064, "loss": 0.707, "step": 7629 }, { "epoch": 3.590588235294118, "grad_norm": 0.4385914581592976, "learning_rate": 0.0001388321183973093, "loss": 0.2539, "step": 7630 }, { "epoch": 3.5910588235294116, "grad_norm": 0.3209226614123197, "learning_rate": 0.0001387884043630399, "loss": 0.1279, "step": 7631 }, { "epoch": 3.591529411764706, "grad_norm": 0.46214236393609204, "learning_rate": 0.0001387446912862957, "loss": 0.3008, "step": 7632 }, { "epoch": 3.592, "grad_norm": 0.5216112231187967, "learning_rate": 0.0001387009791708101, "loss": 0.3535, "step": 7633 }, { "epoch": 3.592470588235294, "grad_norm": 0.6128888817559102, "learning_rate": 0.00013865726802031627, "loss": 0.3086, "step": 7634 }, { "epoch": 3.592941176470588, "grad_norm": 0.7957285797896753, "learning_rate": 0.0001386135578385474, "loss": 0.6016, "step": 7635 }, { "epoch": 3.5934117647058823, "grad_norm": 0.3316544200104317, "learning_rate": 0.00013856984862923655, "loss": 0.166, "step": 7636 }, { "epoch": 3.5938823529411765, "grad_norm": 0.3830893467149226, "learning_rate": 0.00013852614039611667, "loss": 0.0947, "step": 7637 }, { "epoch": 3.5943529411764708, "grad_norm": 0.541658727518224, "learning_rate": 0.0001384824331429207, "loss": 0.3145, "step": 7638 }, { "epoch": 3.594823529411765, "grad_norm": 0.45237749156401896, "learning_rate": 0.00013843872687338142, "loss": 0.2129, "step": 7639 }, { "epoch": 3.5952941176470588, "grad_norm": 0.3040482118482517, "learning_rate": 0.00013839502159123158, "loss": 0.0894, "step": 7640 }, { "epoch": 3.595764705882353, "grad_norm": 0.4189527399282918, "learning_rate": 0.00013835131730020382, "loss": 0.1367, "step": 7641 }, { "epoch": 3.596235294117647, "grad_norm": 0.2898608817167641, "learning_rate": 0.0001383076140040307, "loss": 0.1035, "step": 7642 }, { "epoch": 3.596705882352941, "grad_norm": 0.6000709582714665, "learning_rate": 0.0001382639117064447, "loss": 0.3672, "step": 7643 }, { "epoch": 3.597176470588235, "grad_norm": 0.44727511230337025, "learning_rate": 0.0001382202104111782, "loss": 0.2344, "step": 7644 }, { "epoch": 3.5976470588235294, "grad_norm": 0.8605560602934567, "learning_rate": 0.0001381765101219635, "loss": 0.582, "step": 7645 }, { "epoch": 3.5981176470588236, "grad_norm": 0.2968583179671894, "learning_rate": 0.00013813281084253286, "loss": 0.1592, "step": 7646 }, { "epoch": 3.598588235294118, "grad_norm": 0.4147019140568095, "learning_rate": 0.00013808911257661834, "loss": 0.1982, "step": 7647 }, { "epoch": 3.5990588235294116, "grad_norm": 0.39249197170316663, "learning_rate": 0.00013804541532795204, "loss": 0.1748, "step": 7648 }, { "epoch": 3.599529411764706, "grad_norm": 0.456901899670845, "learning_rate": 0.00013800171910026594, "loss": 0.2715, "step": 7649 }, { "epoch": 3.6, "grad_norm": 0.5466510121659389, "learning_rate": 0.00013795802389729184, "loss": 0.2129, "step": 7650 }, { "epoch": 3.600470588235294, "grad_norm": 0.28704008118801533, "learning_rate": 0.0001379143297227616, "loss": 0.1221, "step": 7651 }, { "epoch": 3.600941176470588, "grad_norm": 0.2762216093830515, "learning_rate": 0.00013787063658040686, "loss": 0.1699, "step": 7652 }, { "epoch": 3.6014117647058823, "grad_norm": 0.35193914588677555, "learning_rate": 0.00013782694447395926, "loss": 0.1406, "step": 7653 }, { "epoch": 3.6018823529411765, "grad_norm": 0.5293672025657473, "learning_rate": 0.0001377832534071503, "loss": 0.3105, "step": 7654 }, { "epoch": 3.6023529411764708, "grad_norm": 0.3125321416989755, "learning_rate": 0.0001377395633837114, "loss": 0.0986, "step": 7655 }, { "epoch": 3.602823529411765, "grad_norm": 0.6885826816349323, "learning_rate": 0.000137695874407374, "loss": 0.5078, "step": 7656 }, { "epoch": 3.6032941176470588, "grad_norm": 0.5015317928989953, "learning_rate": 0.0001376521864818692, "loss": 0.5078, "step": 7657 }, { "epoch": 3.603764705882353, "grad_norm": 0.40149896259701834, "learning_rate": 0.00013760849961092826, "loss": 0.2754, "step": 7658 }, { "epoch": 3.604235294117647, "grad_norm": 0.914866563325158, "learning_rate": 0.00013756481379828225, "loss": 0.582, "step": 7659 }, { "epoch": 3.604705882352941, "grad_norm": 0.8074584508382338, "learning_rate": 0.00013752112904766208, "loss": 0.5391, "step": 7660 }, { "epoch": 3.605176470588235, "grad_norm": 0.6389329049028412, "learning_rate": 0.00013747744536279878, "loss": 0.334, "step": 7661 }, { "epoch": 3.6056470588235294, "grad_norm": 0.6767618536377464, "learning_rate": 0.00013743376274742304, "loss": 0.4648, "step": 7662 }, { "epoch": 3.6061176470588236, "grad_norm": 0.28454550888275987, "learning_rate": 0.00013739008120526559, "loss": 0.1387, "step": 7663 }, { "epoch": 3.606588235294118, "grad_norm": 0.5124706607554745, "learning_rate": 0.00013734640074005702, "loss": 0.3418, "step": 7664 }, { "epoch": 3.6070588235294117, "grad_norm": 0.5517434726468653, "learning_rate": 0.0001373027213555279, "loss": 0.2734, "step": 7665 }, { "epoch": 3.607529411764706, "grad_norm": 0.4008152830592808, "learning_rate": 0.0001372590430554087, "loss": 0.1245, "step": 7666 }, { "epoch": 3.608, "grad_norm": 0.7262420341124444, "learning_rate": 0.0001372153658434297, "loss": 0.418, "step": 7667 }, { "epoch": 3.608470588235294, "grad_norm": 0.42309465567616106, "learning_rate": 0.00013717168972332116, "loss": 0.2656, "step": 7668 }, { "epoch": 3.608941176470588, "grad_norm": 0.730858966016894, "learning_rate": 0.00013712801469881324, "loss": 0.5078, "step": 7669 }, { "epoch": 3.6094117647058823, "grad_norm": 0.4663031637035016, "learning_rate": 0.00013708434077363603, "loss": 0.25, "step": 7670 }, { "epoch": 3.6098823529411765, "grad_norm": 0.5055834448735307, "learning_rate": 0.00013704066795151946, "loss": 0.248, "step": 7671 }, { "epoch": 3.6103529411764708, "grad_norm": 0.568202584142969, "learning_rate": 0.00013699699623619338, "loss": 0.2715, "step": 7672 }, { "epoch": 3.6108235294117645, "grad_norm": 0.593360421174278, "learning_rate": 0.00013695332563138762, "loss": 0.2559, "step": 7673 }, { "epoch": 3.6112941176470588, "grad_norm": 0.5909383340502764, "learning_rate": 0.00013690965614083185, "loss": 0.4805, "step": 7674 }, { "epoch": 3.611764705882353, "grad_norm": 0.5194674260001708, "learning_rate": 0.00013686598776825563, "loss": 0.2949, "step": 7675 }, { "epoch": 3.612235294117647, "grad_norm": 0.378226976311988, "learning_rate": 0.00013682232051738852, "loss": 0.1816, "step": 7676 }, { "epoch": 3.612705882352941, "grad_norm": 0.3613272274562029, "learning_rate": 0.00013677865439195986, "loss": 0.1328, "step": 7677 }, { "epoch": 3.613176470588235, "grad_norm": 0.5075107001772082, "learning_rate": 0.00013673498939569892, "loss": 0.2812, "step": 7678 }, { "epoch": 3.6136470588235294, "grad_norm": 0.66324441614165, "learning_rate": 0.000136691325532335, "loss": 0.375, "step": 7679 }, { "epoch": 3.6141176470588237, "grad_norm": 0.24140978795818474, "learning_rate": 0.00013664766280559716, "loss": 0.0747, "step": 7680 }, { "epoch": 3.614588235294118, "grad_norm": 0.35447247360599365, "learning_rate": 0.00013660400121921434, "loss": 0.208, "step": 7681 }, { "epoch": 3.6150588235294117, "grad_norm": 0.2941245461941088, "learning_rate": 0.00013656034077691562, "loss": 0.0898, "step": 7682 }, { "epoch": 3.615529411764706, "grad_norm": 0.7613583392436081, "learning_rate": 0.00013651668148242967, "loss": 0.4473, "step": 7683 }, { "epoch": 3.616, "grad_norm": 0.6475988860025153, "learning_rate": 0.00013647302333948528, "loss": 0.3066, "step": 7684 }, { "epoch": 3.616470588235294, "grad_norm": 0.49579864112690075, "learning_rate": 0.000136429366351811, "loss": 0.3594, "step": 7685 }, { "epoch": 3.616941176470588, "grad_norm": 0.7786359662677816, "learning_rate": 0.00013638571052313543, "loss": 0.4551, "step": 7686 }, { "epoch": 3.6174117647058823, "grad_norm": 0.6585943367989882, "learning_rate": 0.00013634205585718698, "loss": 0.3184, "step": 7687 }, { "epoch": 3.6178823529411765, "grad_norm": 0.5368766557649367, "learning_rate": 0.00013629840235769388, "loss": 0.2451, "step": 7688 }, { "epoch": 3.6183529411764708, "grad_norm": 0.3872764613495153, "learning_rate": 0.0001362547500283845, "loss": 0.167, "step": 7689 }, { "epoch": 3.6188235294117646, "grad_norm": 0.63607305204666, "learning_rate": 0.00013621109887298685, "loss": 0.3008, "step": 7690 }, { "epoch": 3.6192941176470588, "grad_norm": 0.5468852336910317, "learning_rate": 0.00013616744889522898, "loss": 0.3945, "step": 7691 }, { "epoch": 3.619764705882353, "grad_norm": 0.8178550920263048, "learning_rate": 0.00013612380009883886, "loss": 0.4609, "step": 7692 }, { "epoch": 3.620235294117647, "grad_norm": 0.3368157140018008, "learning_rate": 0.00013608015248754425, "loss": 0.2393, "step": 7693 }, { "epoch": 3.620705882352941, "grad_norm": 0.35261262111095076, "learning_rate": 0.00013603650606507293, "loss": 0.1157, "step": 7694 }, { "epoch": 3.621176470588235, "grad_norm": 0.7069990777980274, "learning_rate": 0.0001359928608351524, "loss": 0.4004, "step": 7695 }, { "epoch": 3.6216470588235294, "grad_norm": 0.4133301597749221, "learning_rate": 0.00013594921680151032, "loss": 0.1621, "step": 7696 }, { "epoch": 3.6221176470588237, "grad_norm": 0.846479652477771, "learning_rate": 0.00013590557396787406, "loss": 0.4785, "step": 7697 }, { "epoch": 3.622588235294118, "grad_norm": 0.897195895894082, "learning_rate": 0.00013586193233797086, "loss": 0.5312, "step": 7698 }, { "epoch": 3.6230588235294117, "grad_norm": 0.47413763922022534, "learning_rate": 0.00013581829191552802, "loss": 0.1865, "step": 7699 }, { "epoch": 3.623529411764706, "grad_norm": 0.5440043528137697, "learning_rate": 0.00013577465270427262, "loss": 0.2695, "step": 7700 }, { "epoch": 3.624, "grad_norm": 0.3997771594451842, "learning_rate": 0.00013573101470793161, "loss": 0.1885, "step": 7701 }, { "epoch": 3.624470588235294, "grad_norm": 0.49026449433303876, "learning_rate": 0.00013568737793023198, "loss": 0.2246, "step": 7702 }, { "epoch": 3.624941176470588, "grad_norm": 0.3994360680158722, "learning_rate": 0.00013564374237490047, "loss": 0.1748, "step": 7703 }, { "epoch": 3.6254117647058823, "grad_norm": 0.5734466048174096, "learning_rate": 0.0001356001080456638, "loss": 0.249, "step": 7704 }, { "epoch": 3.6258823529411766, "grad_norm": 0.8474536772917695, "learning_rate": 0.00013555647494624846, "loss": 0.5469, "step": 7705 }, { "epoch": 3.626352941176471, "grad_norm": 0.5822041284421906, "learning_rate": 0.00013551284308038105, "loss": 0.4336, "step": 7706 }, { "epoch": 3.6268235294117646, "grad_norm": 0.5114849960365322, "learning_rate": 0.00013546921245178792, "loss": 0.2188, "step": 7707 }, { "epoch": 3.627294117647059, "grad_norm": 0.3544697958461108, "learning_rate": 0.00013542558306419526, "loss": 0.2285, "step": 7708 }, { "epoch": 3.627764705882353, "grad_norm": 1.0639017544203169, "learning_rate": 0.00013538195492132928, "loss": 0.4902, "step": 7709 }, { "epoch": 3.6282352941176472, "grad_norm": 0.4931918747931083, "learning_rate": 0.00013533832802691613, "loss": 0.3398, "step": 7710 }, { "epoch": 3.628705882352941, "grad_norm": 0.564823751396047, "learning_rate": 0.00013529470238468166, "loss": 0.3086, "step": 7711 }, { "epoch": 3.6291764705882352, "grad_norm": 0.9189971784375511, "learning_rate": 0.00013525107799835172, "loss": 0.5469, "step": 7712 }, { "epoch": 3.6296470588235294, "grad_norm": 0.5022754113259564, "learning_rate": 0.00013520745487165207, "loss": 0.2598, "step": 7713 }, { "epoch": 3.6301176470588237, "grad_norm": 0.5103993374292398, "learning_rate": 0.00013516383300830835, "loss": 0.21, "step": 7714 }, { "epoch": 3.630588235294118, "grad_norm": 0.9893120063631956, "learning_rate": 0.00013512021241204604, "loss": 0.4668, "step": 7715 }, { "epoch": 3.6310588235294117, "grad_norm": 0.658606502262212, "learning_rate": 0.00013507659308659053, "loss": 0.373, "step": 7716 }, { "epoch": 3.631529411764706, "grad_norm": 0.3756877439783238, "learning_rate": 0.00013503297503566727, "loss": 0.1377, "step": 7717 }, { "epoch": 3.632, "grad_norm": 0.7582271029950013, "learning_rate": 0.00013498935826300132, "loss": 0.543, "step": 7718 }, { "epoch": 3.632470588235294, "grad_norm": 0.33781808552628295, "learning_rate": 0.00013494574277231772, "loss": 0.0942, "step": 7719 }, { "epoch": 3.632941176470588, "grad_norm": 0.5571346606864215, "learning_rate": 0.00013490212856734162, "loss": 0.4199, "step": 7720 }, { "epoch": 3.6334117647058823, "grad_norm": 0.7539329306239632, "learning_rate": 0.00013485851565179778, "loss": 0.4258, "step": 7721 }, { "epoch": 3.6338823529411766, "grad_norm": 0.6316368214545411, "learning_rate": 0.00013481490402941098, "loss": 0.4102, "step": 7722 }, { "epoch": 3.634352941176471, "grad_norm": 0.49891038466703985, "learning_rate": 0.00013477129370390582, "loss": 0.3047, "step": 7723 }, { "epoch": 3.6348235294117646, "grad_norm": 0.6500088912664421, "learning_rate": 0.00013472768467900686, "loss": 0.4629, "step": 7724 }, { "epoch": 3.635294117647059, "grad_norm": 0.4691463231411614, "learning_rate": 0.0001346840769584386, "loss": 0.3496, "step": 7725 }, { "epoch": 3.635764705882353, "grad_norm": 0.5633666453491636, "learning_rate": 0.00013464047054592525, "loss": 0.2773, "step": 7726 }, { "epoch": 3.636235294117647, "grad_norm": 0.5109208754670382, "learning_rate": 0.00013459686544519108, "loss": 0.2344, "step": 7727 }, { "epoch": 3.636705882352941, "grad_norm": 0.17784354844205189, "learning_rate": 0.00013455326165996015, "loss": 0.033, "step": 7728 }, { "epoch": 3.6371764705882352, "grad_norm": 0.33267017219468703, "learning_rate": 0.0001345096591939564, "loss": 0.1787, "step": 7729 }, { "epoch": 3.6376470588235295, "grad_norm": 0.7167387134359502, "learning_rate": 0.00013446605805090376, "loss": 0.3398, "step": 7730 }, { "epoch": 3.6381176470588237, "grad_norm": 0.49073620087101083, "learning_rate": 0.000134422458234526, "loss": 0.291, "step": 7731 }, { "epoch": 3.638588235294118, "grad_norm": 0.5242772695009448, "learning_rate": 0.00013437885974854665, "loss": 0.2539, "step": 7732 }, { "epoch": 3.6390588235294117, "grad_norm": 0.49606484152721025, "learning_rate": 0.00013433526259668931, "loss": 0.208, "step": 7733 }, { "epoch": 3.639529411764706, "grad_norm": 0.47307434663674247, "learning_rate": 0.0001342916667826774, "loss": 0.1572, "step": 7734 }, { "epoch": 3.64, "grad_norm": 0.7150126075598064, "learning_rate": 0.0001342480723102342, "loss": 0.3477, "step": 7735 }, { "epoch": 3.640470588235294, "grad_norm": 0.4515805368242314, "learning_rate": 0.00013420447918308285, "loss": 0.1895, "step": 7736 }, { "epoch": 3.640941176470588, "grad_norm": 0.6662003604158684, "learning_rate": 0.00013416088740494649, "loss": 0.4434, "step": 7737 }, { "epoch": 3.6414117647058823, "grad_norm": 0.4742272466870986, "learning_rate": 0.00013411729697954803, "loss": 0.2197, "step": 7738 }, { "epoch": 3.6418823529411766, "grad_norm": 0.6730797170415344, "learning_rate": 0.00013407370791061027, "loss": 0.3535, "step": 7739 }, { "epoch": 3.642352941176471, "grad_norm": 0.3004405411456621, "learning_rate": 0.00013403012020185603, "loss": 0.1021, "step": 7740 }, { "epoch": 3.6428235294117646, "grad_norm": 0.5257753975063075, "learning_rate": 0.00013398653385700784, "loss": 0.1865, "step": 7741 }, { "epoch": 3.643294117647059, "grad_norm": 0.3929749247799959, "learning_rate": 0.00013394294887978818, "loss": 0.1602, "step": 7742 }, { "epoch": 3.643764705882353, "grad_norm": 0.3021274908864397, "learning_rate": 0.0001338993652739195, "loss": 0.1167, "step": 7743 }, { "epoch": 3.644235294117647, "grad_norm": 0.5243730904147774, "learning_rate": 0.00013385578304312397, "loss": 0.2539, "step": 7744 }, { "epoch": 3.644705882352941, "grad_norm": 0.6443563606148799, "learning_rate": 0.00013381220219112373, "loss": 0.2373, "step": 7745 }, { "epoch": 3.6451764705882352, "grad_norm": 0.6209850567902315, "learning_rate": 0.00013376862272164084, "loss": 0.3203, "step": 7746 }, { "epoch": 3.6456470588235295, "grad_norm": 0.6146797403963477, "learning_rate": 0.00013372504463839718, "loss": 0.2715, "step": 7747 }, { "epoch": 3.6461176470588237, "grad_norm": 0.9632028554020052, "learning_rate": 0.00013368146794511457, "loss": 0.6484, "step": 7748 }, { "epoch": 3.646588235294118, "grad_norm": 0.6171669495720925, "learning_rate": 0.0001336378926455146, "loss": 0.3496, "step": 7749 }, { "epoch": 3.6470588235294117, "grad_norm": 0.714818555024602, "learning_rate": 0.00013359431874331886, "loss": 0.3184, "step": 7750 }, { "epoch": 3.647529411764706, "grad_norm": 0.3625100755745974, "learning_rate": 0.0001335507462422488, "loss": 0.1328, "step": 7751 }, { "epoch": 3.648, "grad_norm": 0.49268024869631033, "learning_rate": 0.00013350717514602563, "loss": 0.207, "step": 7752 }, { "epoch": 3.648470588235294, "grad_norm": 0.24618600522351622, "learning_rate": 0.00013346360545837062, "loss": 0.0947, "step": 7753 }, { "epoch": 3.648941176470588, "grad_norm": 0.4064462272392357, "learning_rate": 0.0001334200371830048, "loss": 0.1826, "step": 7754 }, { "epoch": 3.6494117647058824, "grad_norm": 0.4195809157737068, "learning_rate": 0.00013337647032364919, "loss": 0.1865, "step": 7755 }, { "epoch": 3.6498823529411766, "grad_norm": 0.7459274978255007, "learning_rate": 0.0001333329048840245, "loss": 0.4746, "step": 7756 }, { "epoch": 3.650352941176471, "grad_norm": 0.40197122096847626, "learning_rate": 0.0001332893408678515, "loss": 0.2471, "step": 7757 }, { "epoch": 3.6508235294117646, "grad_norm": 0.5873161636448581, "learning_rate": 0.00013324577827885078, "loss": 0.1992, "step": 7758 }, { "epoch": 3.651294117647059, "grad_norm": 0.5500110539345513, "learning_rate": 0.00013320221712074276, "loss": 0.2314, "step": 7759 }, { "epoch": 3.651764705882353, "grad_norm": 0.7364018988617177, "learning_rate": 0.00013315865739724778, "loss": 0.6484, "step": 7760 }, { "epoch": 3.652235294117647, "grad_norm": 0.7666042802836591, "learning_rate": 0.00013311509911208608, "loss": 0.3633, "step": 7761 }, { "epoch": 3.652705882352941, "grad_norm": 0.5346654027398954, "learning_rate": 0.00013307154226897773, "loss": 0.209, "step": 7762 }, { "epoch": 3.6531764705882352, "grad_norm": 0.5792691664432107, "learning_rate": 0.0001330279868716427, "loss": 0.2734, "step": 7763 }, { "epoch": 3.6536470588235295, "grad_norm": 0.7977386971926678, "learning_rate": 0.00013298443292380084, "loss": 0.332, "step": 7764 }, { "epoch": 3.6541176470588237, "grad_norm": 0.4476412417678736, "learning_rate": 0.00013294088042917188, "loss": 0.1406, "step": 7765 }, { "epoch": 3.654588235294118, "grad_norm": 0.36168950967753893, "learning_rate": 0.0001328973293914754, "loss": 0.1582, "step": 7766 }, { "epoch": 3.6550588235294117, "grad_norm": 0.6680012326911308, "learning_rate": 0.00013285377981443085, "loss": 0.3789, "step": 7767 }, { "epoch": 3.655529411764706, "grad_norm": 0.9265184044550404, "learning_rate": 0.00013281023170175766, "loss": 0.5977, "step": 7768 }, { "epoch": 3.656, "grad_norm": 0.5909511189297129, "learning_rate": 0.00013276668505717498, "loss": 0.4297, "step": 7769 }, { "epoch": 3.656470588235294, "grad_norm": 0.7168290717840783, "learning_rate": 0.00013272313988440187, "loss": 0.3652, "step": 7770 }, { "epoch": 3.656941176470588, "grad_norm": 0.9021585692895188, "learning_rate": 0.0001326795961871574, "loss": 0.3418, "step": 7771 }, { "epoch": 3.6574117647058824, "grad_norm": 0.5612540782892792, "learning_rate": 0.00013263605396916038, "loss": 0.1523, "step": 7772 }, { "epoch": 3.6578823529411766, "grad_norm": 0.24868077101950806, "learning_rate": 0.00013259251323412952, "loss": 0.0786, "step": 7773 }, { "epoch": 3.658352941176471, "grad_norm": 0.2657213246965765, "learning_rate": 0.0001325489739857834, "loss": 0.1235, "step": 7774 }, { "epoch": 3.6588235294117646, "grad_norm": 0.6675136716712647, "learning_rate": 0.0001325054362278405, "loss": 0.3262, "step": 7775 }, { "epoch": 3.659294117647059, "grad_norm": 0.5506604071802167, "learning_rate": 0.00013246189996401916, "loss": 0.2119, "step": 7776 }, { "epoch": 3.659764705882353, "grad_norm": 0.7109031089693407, "learning_rate": 0.00013241836519803756, "loss": 0.4883, "step": 7777 }, { "epoch": 3.660235294117647, "grad_norm": 0.6920763431159362, "learning_rate": 0.00013237483193361387, "loss": 0.3418, "step": 7778 }, { "epoch": 3.660705882352941, "grad_norm": 0.4057410961218393, "learning_rate": 0.00013233130017446595, "loss": 0.2041, "step": 7779 }, { "epoch": 3.6611764705882353, "grad_norm": 0.7268565477582417, "learning_rate": 0.00013228776992431165, "loss": 0.5195, "step": 7780 }, { "epoch": 3.6616470588235295, "grad_norm": 0.6004508879605062, "learning_rate": 0.00013224424118686868, "loss": 0.3887, "step": 7781 }, { "epoch": 3.6621176470588237, "grad_norm": 0.44660022761202645, "learning_rate": 0.00013220071396585463, "loss": 0.2578, "step": 7782 }, { "epoch": 3.662588235294118, "grad_norm": 0.727616097732452, "learning_rate": 0.0001321571882649869, "loss": 0.2871, "step": 7783 }, { "epoch": 3.6630588235294117, "grad_norm": 0.5462880827108197, "learning_rate": 0.00013211366408798277, "loss": 0.2969, "step": 7784 }, { "epoch": 3.663529411764706, "grad_norm": 0.5648658303476901, "learning_rate": 0.00013207014143855951, "loss": 0.3105, "step": 7785 }, { "epoch": 3.664, "grad_norm": 0.5692615407520445, "learning_rate": 0.00013202662032043408, "loss": 0.2988, "step": 7786 }, { "epoch": 3.664470588235294, "grad_norm": 0.36930826911190023, "learning_rate": 0.00013198310073732344, "loss": 0.1318, "step": 7787 }, { "epoch": 3.664941176470588, "grad_norm": 0.448224447889691, "learning_rate": 0.0001319395826929444, "loss": 0.2451, "step": 7788 }, { "epoch": 3.6654117647058824, "grad_norm": 0.49323529784215997, "learning_rate": 0.00013189606619101358, "loss": 0.2451, "step": 7789 }, { "epoch": 3.6658823529411766, "grad_norm": 0.20838702619443253, "learning_rate": 0.00013185255123524745, "loss": 0.0574, "step": 7790 }, { "epoch": 3.666352941176471, "grad_norm": 0.7768695651175346, "learning_rate": 0.00013180903782936254, "loss": 0.5352, "step": 7791 }, { "epoch": 3.6668235294117646, "grad_norm": 0.7663987991302343, "learning_rate": 0.000131765525977075, "loss": 0.4785, "step": 7792 }, { "epoch": 3.667294117647059, "grad_norm": 0.313995868640876, "learning_rate": 0.0001317220156821009, "loss": 0.1641, "step": 7793 }, { "epoch": 3.667764705882353, "grad_norm": 0.5453567893666169, "learning_rate": 0.00013167850694815642, "loss": 0.2754, "step": 7794 }, { "epoch": 3.668235294117647, "grad_norm": 0.48127138742266456, "learning_rate": 0.0001316349997789573, "loss": 0.2559, "step": 7795 }, { "epoch": 3.668705882352941, "grad_norm": 0.3225586693175006, "learning_rate": 0.00013159149417821923, "loss": 0.166, "step": 7796 }, { "epoch": 3.6691764705882353, "grad_norm": 1.0274705510573563, "learning_rate": 0.00013154799014965785, "loss": 0.7969, "step": 7797 }, { "epoch": 3.6696470588235295, "grad_norm": 0.2387335415638724, "learning_rate": 0.00013150448769698857, "loss": 0.1025, "step": 7798 }, { "epoch": 3.6701176470588237, "grad_norm": 0.24910618406975205, "learning_rate": 0.00013146098682392685, "loss": 0.0767, "step": 7799 }, { "epoch": 3.6705882352941175, "grad_norm": 0.33245165095049867, "learning_rate": 0.0001314174875341878, "loss": 0.1138, "step": 7800 }, { "epoch": 3.6710588235294117, "grad_norm": 0.3294683355974251, "learning_rate": 0.00013137398983148642, "loss": 0.1289, "step": 7801 }, { "epoch": 3.671529411764706, "grad_norm": 0.8037393137182145, "learning_rate": 0.00013133049371953767, "loss": 0.4102, "step": 7802 }, { "epoch": 3.672, "grad_norm": 0.4398795508685683, "learning_rate": 0.00013128699920205634, "loss": 0.2354, "step": 7803 }, { "epoch": 3.672470588235294, "grad_norm": 0.6592399524917383, "learning_rate": 0.00013124350628275708, "loss": 0.4258, "step": 7804 }, { "epoch": 3.672941176470588, "grad_norm": 0.6567217818293443, "learning_rate": 0.00013120001496535433, "loss": 0.3203, "step": 7805 }, { "epoch": 3.6734117647058824, "grad_norm": 0.5518361206189047, "learning_rate": 0.00013115652525356257, "loss": 0.2949, "step": 7806 }, { "epoch": 3.6738823529411766, "grad_norm": 0.28404085190191897, "learning_rate": 0.00013111303715109598, "loss": 0.083, "step": 7807 }, { "epoch": 3.674352941176471, "grad_norm": 0.129340715305411, "learning_rate": 0.0001310695506616686, "loss": 0.0425, "step": 7808 }, { "epoch": 3.6748235294117646, "grad_norm": 0.45694840609337806, "learning_rate": 0.0001310260657889945, "loss": 0.0796, "step": 7809 }, { "epoch": 3.675294117647059, "grad_norm": 0.33177678352155593, "learning_rate": 0.00013098258253678744, "loss": 0.1318, "step": 7810 }, { "epoch": 3.675764705882353, "grad_norm": 0.30262446195494536, "learning_rate": 0.00013093910090876108, "loss": 0.0845, "step": 7811 }, { "epoch": 3.676235294117647, "grad_norm": 0.4159285768929776, "learning_rate": 0.00013089562090862905, "loss": 0.2793, "step": 7812 }, { "epoch": 3.676705882352941, "grad_norm": 0.4647846529377477, "learning_rate": 0.00013085214254010465, "loss": 0.2256, "step": 7813 }, { "epoch": 3.6771764705882353, "grad_norm": 0.7160581692219404, "learning_rate": 0.00013080866580690126, "loss": 0.3672, "step": 7814 }, { "epoch": 3.6776470588235295, "grad_norm": 0.626299368005403, "learning_rate": 0.00013076519071273187, "loss": 0.3789, "step": 7815 }, { "epoch": 3.6781176470588237, "grad_norm": 0.4503419899523573, "learning_rate": 0.00013072171726130957, "loss": 0.2275, "step": 7816 }, { "epoch": 3.6785882352941175, "grad_norm": 0.702044783546471, "learning_rate": 0.00013067824545634718, "loss": 0.4531, "step": 7817 }, { "epoch": 3.6790588235294117, "grad_norm": 0.25644323732648644, "learning_rate": 0.00013063477530155738, "loss": 0.0684, "step": 7818 }, { "epoch": 3.679529411764706, "grad_norm": 0.5567560316868203, "learning_rate": 0.00013059130680065278, "loss": 0.4082, "step": 7819 }, { "epoch": 3.68, "grad_norm": 0.5538033971652121, "learning_rate": 0.0001305478399573458, "loss": 0.3008, "step": 7820 }, { "epoch": 3.680470588235294, "grad_norm": 0.3068531170512361, "learning_rate": 0.00013050437477534862, "loss": 0.106, "step": 7821 }, { "epoch": 3.680941176470588, "grad_norm": 0.3622241860949758, "learning_rate": 0.00013046091125837352, "loss": 0.1406, "step": 7822 }, { "epoch": 3.6814117647058824, "grad_norm": 0.7174768700170155, "learning_rate": 0.00013041744941013245, "loss": 0.2236, "step": 7823 }, { "epoch": 3.6818823529411766, "grad_norm": 0.46621919548005497, "learning_rate": 0.00013037398923433725, "loss": 0.1562, "step": 7824 }, { "epoch": 3.682352941176471, "grad_norm": 0.1915621460481547, "learning_rate": 0.00013033053073469958, "loss": 0.0483, "step": 7825 }, { "epoch": 3.6828235294117646, "grad_norm": 0.7536685660900576, "learning_rate": 0.00013028707391493112, "loss": 0.3516, "step": 7826 }, { "epoch": 3.683294117647059, "grad_norm": 0.4097147567519564, "learning_rate": 0.00013024361877874324, "loss": 0.165, "step": 7827 }, { "epoch": 3.683764705882353, "grad_norm": 0.8253297113841743, "learning_rate": 0.00013020016532984714, "loss": 0.418, "step": 7828 }, { "epoch": 3.684235294117647, "grad_norm": 0.38515319757515765, "learning_rate": 0.00013015671357195413, "loss": 0.1357, "step": 7829 }, { "epoch": 3.684705882352941, "grad_norm": 0.8021838188611635, "learning_rate": 0.00013011326350877508, "loss": 0.5508, "step": 7830 }, { "epoch": 3.6851764705882353, "grad_norm": 0.6546562366249318, "learning_rate": 0.00013006981514402085, "loss": 0.3066, "step": 7831 }, { "epoch": 3.6856470588235295, "grad_norm": 0.7150705526030158, "learning_rate": 0.00013002636848140217, "loss": 0.375, "step": 7832 }, { "epoch": 3.6861176470588237, "grad_norm": 0.6695053230966981, "learning_rate": 0.00012998292352462963, "loss": 0.4004, "step": 7833 }, { "epoch": 3.6865882352941175, "grad_norm": 0.37865092340715006, "learning_rate": 0.00012993948027741357, "loss": 0.1494, "step": 7834 }, { "epoch": 3.6870588235294117, "grad_norm": 0.4495129532393885, "learning_rate": 0.0001298960387434643, "loss": 0.2715, "step": 7835 }, { "epoch": 3.687529411764706, "grad_norm": 0.2581837053669576, "learning_rate": 0.00012985259892649191, "loss": 0.0771, "step": 7836 }, { "epoch": 3.6879999999999997, "grad_norm": 0.2718791300922167, "learning_rate": 0.00012980916083020646, "loss": 0.0967, "step": 7837 }, { "epoch": 3.688470588235294, "grad_norm": 0.547296773993828, "learning_rate": 0.0001297657244583176, "loss": 0.2695, "step": 7838 }, { "epoch": 3.688941176470588, "grad_norm": 0.4742035002724371, "learning_rate": 0.00012972228981453522, "loss": 0.252, "step": 7839 }, { "epoch": 3.6894117647058824, "grad_norm": 0.49045908024349016, "learning_rate": 0.00012967885690256872, "loss": 0.1416, "step": 7840 }, { "epoch": 3.6898823529411766, "grad_norm": 0.919661500821608, "learning_rate": 0.0001296354257261275, "loss": 0.5, "step": 7841 }, { "epoch": 3.690352941176471, "grad_norm": 0.6580422420228057, "learning_rate": 0.0001295919962889208, "loss": 0.3555, "step": 7842 }, { "epoch": 3.6908235294117646, "grad_norm": 0.8016834094010293, "learning_rate": 0.0001295485685946577, "loss": 0.6328, "step": 7843 }, { "epoch": 3.691294117647059, "grad_norm": 0.6070142795411162, "learning_rate": 0.0001295051426470472, "loss": 0.3125, "step": 7844 }, { "epoch": 3.691764705882353, "grad_norm": 0.3988640897195185, "learning_rate": 0.00012946171844979804, "loss": 0.2139, "step": 7845 }, { "epoch": 3.692235294117647, "grad_norm": 0.13843304791999514, "learning_rate": 0.00012941829600661886, "loss": 0.0374, "step": 7846 }, { "epoch": 3.692705882352941, "grad_norm": 0.6351020204683144, "learning_rate": 0.00012937487532121818, "loss": 0.3398, "step": 7847 }, { "epoch": 3.6931764705882353, "grad_norm": 0.7886957025791144, "learning_rate": 0.00012933145639730428, "loss": 0.6016, "step": 7848 }, { "epoch": 3.6936470588235295, "grad_norm": 0.6821068795146604, "learning_rate": 0.00012928803923858536, "loss": 0.3359, "step": 7849 }, { "epoch": 3.6941176470588237, "grad_norm": 0.34861773688858894, "learning_rate": 0.00012924462384876953, "loss": 0.123, "step": 7850 }, { "epoch": 3.6945882352941175, "grad_norm": 0.5315778569607074, "learning_rate": 0.00012920121023156462, "loss": 0.2676, "step": 7851 }, { "epoch": 3.6950588235294117, "grad_norm": 0.6694640680570282, "learning_rate": 0.00012915779839067838, "loss": 0.3633, "step": 7852 }, { "epoch": 3.695529411764706, "grad_norm": 0.4917554929058661, "learning_rate": 0.00012911438832981836, "loss": 0.2021, "step": 7853 }, { "epoch": 3.6959999999999997, "grad_norm": 0.32508310392567, "learning_rate": 0.00012907098005269203, "loss": 0.127, "step": 7854 }, { "epoch": 3.696470588235294, "grad_norm": 0.5892899238687673, "learning_rate": 0.00012902757356300667, "loss": 0.3906, "step": 7855 }, { "epoch": 3.696941176470588, "grad_norm": 0.930270757733015, "learning_rate": 0.00012898416886446937, "loss": 0.5352, "step": 7856 }, { "epoch": 3.6974117647058824, "grad_norm": 0.7843302423397025, "learning_rate": 0.00012894076596078717, "loss": 0.3984, "step": 7857 }, { "epoch": 3.6978823529411766, "grad_norm": 0.6097571621023404, "learning_rate": 0.00012889736485566683, "loss": 0.3398, "step": 7858 }, { "epoch": 3.698352941176471, "grad_norm": 0.6080681465954495, "learning_rate": 0.000128853965552815, "loss": 0.3184, "step": 7859 }, { "epoch": 3.6988235294117646, "grad_norm": 0.6349305009196607, "learning_rate": 0.0001288105680559383, "loss": 0.3691, "step": 7860 }, { "epoch": 3.699294117647059, "grad_norm": 0.6231410358480964, "learning_rate": 0.000128767172368743, "loss": 0.4102, "step": 7861 }, { "epoch": 3.699764705882353, "grad_norm": 0.6137558767451354, "learning_rate": 0.0001287237784949353, "loss": 0.2988, "step": 7862 }, { "epoch": 3.700235294117647, "grad_norm": 0.3925611384885454, "learning_rate": 0.0001286803864382213, "loss": 0.1523, "step": 7863 }, { "epoch": 3.700705882352941, "grad_norm": 0.524347652254961, "learning_rate": 0.00012863699620230688, "loss": 0.2422, "step": 7864 }, { "epoch": 3.7011764705882353, "grad_norm": 0.4520660869652994, "learning_rate": 0.00012859360779089775, "loss": 0.2402, "step": 7865 }, { "epoch": 3.7016470588235295, "grad_norm": 0.1858528191522451, "learning_rate": 0.00012855022120769952, "loss": 0.05, "step": 7866 }, { "epoch": 3.7021176470588237, "grad_norm": 0.5109342092743941, "learning_rate": 0.0001285068364564176, "loss": 0.2207, "step": 7867 }, { "epoch": 3.7025882352941175, "grad_norm": 0.45847655374976554, "learning_rate": 0.0001284634535407573, "loss": 0.3555, "step": 7868 }, { "epoch": 3.7030588235294117, "grad_norm": 0.31862986084255235, "learning_rate": 0.00012842007246442367, "loss": 0.1006, "step": 7869 }, { "epoch": 3.703529411764706, "grad_norm": 0.4835809410003715, "learning_rate": 0.0001283766932311217, "loss": 0.2266, "step": 7870 }, { "epoch": 3.7039999999999997, "grad_norm": 0.3907481646027432, "learning_rate": 0.00012833331584455622, "loss": 0.1875, "step": 7871 }, { "epoch": 3.704470588235294, "grad_norm": 0.49004718743295367, "learning_rate": 0.0001282899403084318, "loss": 0.3262, "step": 7872 }, { "epoch": 3.704941176470588, "grad_norm": 0.3653924613013968, "learning_rate": 0.000128246566626453, "loss": 0.1348, "step": 7873 }, { "epoch": 3.7054117647058824, "grad_norm": 0.48869825256360494, "learning_rate": 0.0001282031948023241, "loss": 0.2949, "step": 7874 }, { "epoch": 3.7058823529411766, "grad_norm": 0.6262020889524692, "learning_rate": 0.0001281598248397493, "loss": 0.4863, "step": 7875 }, { "epoch": 3.706352941176471, "grad_norm": 0.15250237094908006, "learning_rate": 0.00012811645674243254, "loss": 0.0347, "step": 7876 }, { "epoch": 3.7068235294117646, "grad_norm": 0.6276358298990582, "learning_rate": 0.00012807309051407774, "loss": 0.252, "step": 7877 }, { "epoch": 3.707294117647059, "grad_norm": 0.9832512603997307, "learning_rate": 0.00012802972615838856, "loss": 0.9844, "step": 7878 }, { "epoch": 3.707764705882353, "grad_norm": 0.23267838058793944, "learning_rate": 0.00012798636367906849, "loss": 0.0452, "step": 7879 }, { "epoch": 3.708235294117647, "grad_norm": 0.5316600196632242, "learning_rate": 0.00012794300307982098, "loss": 0.3242, "step": 7880 }, { "epoch": 3.708705882352941, "grad_norm": 0.5318316411178056, "learning_rate": 0.0001278996443643492, "loss": 0.3145, "step": 7881 }, { "epoch": 3.7091764705882353, "grad_norm": 0.4069688291791237, "learning_rate": 0.00012785628753635617, "loss": 0.0918, "step": 7882 }, { "epoch": 3.7096470588235295, "grad_norm": 0.4733812727518818, "learning_rate": 0.0001278129325995448, "loss": 0.2432, "step": 7883 }, { "epoch": 3.7101176470588237, "grad_norm": 0.637028770323177, "learning_rate": 0.00012776957955761784, "loss": 0.3809, "step": 7884 }, { "epoch": 3.7105882352941175, "grad_norm": 0.5946644394446888, "learning_rate": 0.0001277262284142778, "loss": 0.2461, "step": 7885 }, { "epoch": 3.7110588235294117, "grad_norm": 0.5180489340572029, "learning_rate": 0.00012768287917322706, "loss": 0.2852, "step": 7886 }, { "epoch": 3.711529411764706, "grad_norm": 0.575975025128031, "learning_rate": 0.00012763953183816794, "loss": 0.2656, "step": 7887 }, { "epoch": 3.7119999999999997, "grad_norm": 0.6178949853772236, "learning_rate": 0.00012759618641280251, "loss": 0.2295, "step": 7888 }, { "epoch": 3.712470588235294, "grad_norm": 0.3198140803928315, "learning_rate": 0.00012755284290083264, "loss": 0.126, "step": 7889 }, { "epoch": 3.712941176470588, "grad_norm": 0.4303712559922566, "learning_rate": 0.00012750950130596006, "loss": 0.1387, "step": 7890 }, { "epoch": 3.7134117647058824, "grad_norm": 0.57635839271137, "learning_rate": 0.0001274661616318864, "loss": 0.5664, "step": 7891 }, { "epoch": 3.7138823529411766, "grad_norm": 0.5993709043060375, "learning_rate": 0.0001274228238823131, "loss": 0.2852, "step": 7892 }, { "epoch": 3.714352941176471, "grad_norm": 0.43869013053102046, "learning_rate": 0.00012737948806094135, "loss": 0.2559, "step": 7893 }, { "epoch": 3.7148235294117646, "grad_norm": 0.3212834199879089, "learning_rate": 0.00012733615417147228, "loss": 0.2109, "step": 7894 }, { "epoch": 3.715294117647059, "grad_norm": 0.8043877060718306, "learning_rate": 0.0001272928222176068, "loss": 0.6055, "step": 7895 }, { "epoch": 3.715764705882353, "grad_norm": 0.4223613912216541, "learning_rate": 0.0001272494922030457, "loss": 0.1787, "step": 7896 }, { "epoch": 3.716235294117647, "grad_norm": 0.49144665600776116, "learning_rate": 0.0001272061641314895, "loss": 0.2031, "step": 7897 }, { "epoch": 3.716705882352941, "grad_norm": 0.614424005035296, "learning_rate": 0.00012716283800663874, "loss": 0.3633, "step": 7898 }, { "epoch": 3.7171764705882353, "grad_norm": 0.3799012017907965, "learning_rate": 0.00012711951383219362, "loss": 0.2402, "step": 7899 }, { "epoch": 3.7176470588235295, "grad_norm": 0.4789590665873662, "learning_rate": 0.00012707619161185423, "loss": 0.3008, "step": 7900 }, { "epoch": 3.7181176470588237, "grad_norm": 0.3168900568697174, "learning_rate": 0.00012703287134932054, "loss": 0.1309, "step": 7901 }, { "epoch": 3.7185882352941175, "grad_norm": 0.3444695585799633, "learning_rate": 0.0001269895530482923, "loss": 0.1523, "step": 7902 }, { "epoch": 3.7190588235294118, "grad_norm": 0.5314863245539381, "learning_rate": 0.00012694623671246908, "loss": 0.3438, "step": 7903 }, { "epoch": 3.719529411764706, "grad_norm": 0.6887294527929412, "learning_rate": 0.0001269029223455503, "loss": 0.4512, "step": 7904 }, { "epoch": 3.7199999999999998, "grad_norm": 0.5744146021086939, "learning_rate": 0.00012685960995123527, "loss": 0.2471, "step": 7905 }, { "epoch": 3.720470588235294, "grad_norm": 0.6643493449887877, "learning_rate": 0.00012681629953322307, "loss": 0.3496, "step": 7906 }, { "epoch": 3.720941176470588, "grad_norm": 0.5513499754878531, "learning_rate": 0.00012677299109521256, "loss": 0.4121, "step": 7907 }, { "epoch": 3.7214117647058824, "grad_norm": 0.5972664905868311, "learning_rate": 0.00012672968464090255, "loss": 0.1865, "step": 7908 }, { "epoch": 3.7218823529411766, "grad_norm": 0.04460394663783399, "learning_rate": 0.00012668638017399164, "loss": 0.0157, "step": 7909 }, { "epoch": 3.722352941176471, "grad_norm": 0.9299176598029744, "learning_rate": 0.00012664307769817815, "loss": 0.5898, "step": 7910 }, { "epoch": 3.7228235294117646, "grad_norm": 1.7109396523958, "learning_rate": 0.00012659977721716045, "loss": 1.2734, "step": 7911 }, { "epoch": 3.723294117647059, "grad_norm": 0.7657265300353648, "learning_rate": 0.00012655647873463657, "loss": 0.3457, "step": 7912 }, { "epoch": 3.723764705882353, "grad_norm": 0.39454921754213884, "learning_rate": 0.00012651318225430437, "loss": 0.1865, "step": 7913 }, { "epoch": 3.724235294117647, "grad_norm": 0.873095330079431, "learning_rate": 0.00012646988777986157, "loss": 0.498, "step": 7914 }, { "epoch": 3.724705882352941, "grad_norm": 0.2675582781808734, "learning_rate": 0.00012642659531500583, "loss": 0.0679, "step": 7915 }, { "epoch": 3.7251764705882353, "grad_norm": 0.5313596554453528, "learning_rate": 0.00012638330486343445, "loss": 0.1787, "step": 7916 }, { "epoch": 3.7256470588235295, "grad_norm": 0.5537315761071597, "learning_rate": 0.00012634001642884465, "loss": 0.2852, "step": 7917 }, { "epoch": 3.7261176470588238, "grad_norm": 0.8293151689052235, "learning_rate": 0.0001262967300149335, "loss": 0.2598, "step": 7918 }, { "epoch": 3.7265882352941175, "grad_norm": 0.6051996542684296, "learning_rate": 0.0001262534456253979, "loss": 0.3496, "step": 7919 }, { "epoch": 3.7270588235294118, "grad_norm": 0.6180959896335708, "learning_rate": 0.0001262101632639345, "loss": 0.3945, "step": 7920 }, { "epoch": 3.727529411764706, "grad_norm": 0.3150820526559812, "learning_rate": 0.00012616688293423987, "loss": 0.1099, "step": 7921 }, { "epoch": 3.7279999999999998, "grad_norm": 0.3245634912237643, "learning_rate": 0.00012612360464001033, "loss": 0.1348, "step": 7922 }, { "epoch": 3.728470588235294, "grad_norm": 0.8958109625670703, "learning_rate": 0.00012608032838494205, "loss": 0.75, "step": 7923 }, { "epoch": 3.728941176470588, "grad_norm": 0.6774794405813038, "learning_rate": 0.00012603705417273107, "loss": 0.2598, "step": 7924 }, { "epoch": 3.7294117647058824, "grad_norm": 0.3805896079525426, "learning_rate": 0.0001259937820070732, "loss": 0.2168, "step": 7925 }, { "epoch": 3.7298823529411766, "grad_norm": 0.32400001853011806, "learning_rate": 0.0001259505118916641, "loss": 0.1562, "step": 7926 }, { "epoch": 3.7303529411764704, "grad_norm": 0.830225773989846, "learning_rate": 0.0001259072438301992, "loss": 0.6055, "step": 7927 }, { "epoch": 3.7308235294117646, "grad_norm": 0.5564564232437784, "learning_rate": 0.00012586397782637392, "loss": 0.3281, "step": 7928 }, { "epoch": 3.731294117647059, "grad_norm": 0.6772994783553093, "learning_rate": 0.0001258207138838833, "loss": 0.2715, "step": 7929 }, { "epoch": 3.731764705882353, "grad_norm": 0.5854898859042317, "learning_rate": 0.00012577745200642227, "loss": 0.3789, "step": 7930 }, { "epoch": 3.732235294117647, "grad_norm": 0.6918932888891782, "learning_rate": 0.00012573419219768562, "loss": 0.4609, "step": 7931 }, { "epoch": 3.732705882352941, "grad_norm": 0.8325379377536535, "learning_rate": 0.00012569093446136806, "loss": 0.3555, "step": 7932 }, { "epoch": 3.7331764705882353, "grad_norm": 0.30111270917195115, "learning_rate": 0.0001256476788011639, "loss": 0.1191, "step": 7933 }, { "epoch": 3.7336470588235295, "grad_norm": 0.375242119516279, "learning_rate": 0.00012560442522076745, "loss": 0.1357, "step": 7934 }, { "epoch": 3.7341176470588238, "grad_norm": 0.1960579786290169, "learning_rate": 0.00012556117372387272, "loss": 0.062, "step": 7935 }, { "epoch": 3.7345882352941175, "grad_norm": 0.4408986400732444, "learning_rate": 0.00012551792431417363, "loss": 0.2324, "step": 7936 }, { "epoch": 3.7350588235294118, "grad_norm": 0.4179373803880469, "learning_rate": 0.0001254746769953639, "loss": 0.1719, "step": 7937 }, { "epoch": 3.735529411764706, "grad_norm": 0.6570636303659178, "learning_rate": 0.000125431431771137, "loss": 0.1699, "step": 7938 }, { "epoch": 3.7359999999999998, "grad_norm": 0.9286082229421389, "learning_rate": 0.0001253881886451864, "loss": 0.3184, "step": 7939 }, { "epoch": 3.736470588235294, "grad_norm": 0.36127611353664735, "learning_rate": 0.0001253449476212052, "loss": 0.1982, "step": 7940 }, { "epoch": 3.736941176470588, "grad_norm": 0.40027134596422626, "learning_rate": 0.00012530170870288636, "loss": 0.2676, "step": 7941 }, { "epoch": 3.7374117647058824, "grad_norm": 0.5313122870548683, "learning_rate": 0.00012525847189392278, "loss": 0.332, "step": 7942 }, { "epoch": 3.7378823529411767, "grad_norm": 0.41362921153668725, "learning_rate": 0.00012521523719800706, "loss": 0.1592, "step": 7943 }, { "epoch": 3.7383529411764704, "grad_norm": 0.2943897106073625, "learning_rate": 0.00012517200461883163, "loss": 0.1245, "step": 7944 }, { "epoch": 3.7388235294117647, "grad_norm": 0.4961867470819411, "learning_rate": 0.00012512877416008878, "loss": 0.1299, "step": 7945 }, { "epoch": 3.739294117647059, "grad_norm": 0.8676046853135367, "learning_rate": 0.00012508554582547064, "loss": 0.4023, "step": 7946 }, { "epoch": 3.739764705882353, "grad_norm": 0.6309383015987924, "learning_rate": 0.00012504231961866908, "loss": 0.498, "step": 7947 }, { "epoch": 3.740235294117647, "grad_norm": 0.7618880443243309, "learning_rate": 0.00012499909554337582, "loss": 0.5273, "step": 7948 }, { "epoch": 3.740705882352941, "grad_norm": 0.44847500190357625, "learning_rate": 0.00012495587360328246, "loss": 0.2734, "step": 7949 }, { "epoch": 3.7411764705882353, "grad_norm": 0.22566831911005497, "learning_rate": 0.00012491265380208032, "loss": 0.0864, "step": 7950 }, { "epoch": 3.7416470588235295, "grad_norm": 0.3097401560258763, "learning_rate": 0.0001248694361434606, "loss": 0.1157, "step": 7951 }, { "epoch": 3.7421176470588238, "grad_norm": 0.94832284927541, "learning_rate": 0.0001248262206311143, "loss": 0.4473, "step": 7952 }, { "epoch": 3.7425882352941175, "grad_norm": 0.6315323783581529, "learning_rate": 0.00012478300726873225, "loss": 0.3047, "step": 7953 }, { "epoch": 3.7430588235294118, "grad_norm": 0.3648243537313486, "learning_rate": 0.00012473979606000504, "loss": 0.126, "step": 7954 }, { "epoch": 3.743529411764706, "grad_norm": 0.42415654929646973, "learning_rate": 0.00012469658700862315, "loss": 0.2295, "step": 7955 }, { "epoch": 3.7439999999999998, "grad_norm": 0.6454195024901261, "learning_rate": 0.00012465338011827685, "loss": 0.3691, "step": 7956 }, { "epoch": 3.744470588235294, "grad_norm": 0.20721408486563703, "learning_rate": 0.00012461017539265622, "loss": 0.0522, "step": 7957 }, { "epoch": 3.744941176470588, "grad_norm": 0.70249104332071, "learning_rate": 0.00012456697283545114, "loss": 0.5195, "step": 7958 }, { "epoch": 3.7454117647058824, "grad_norm": 0.531756336181589, "learning_rate": 0.00012452377245035132, "loss": 0.3809, "step": 7959 }, { "epoch": 3.7458823529411767, "grad_norm": 0.5076656709328901, "learning_rate": 0.00012448057424104632, "loss": 0.21, "step": 7960 }, { "epoch": 3.7463529411764704, "grad_norm": 0.5323391291612396, "learning_rate": 0.00012443737821122541, "loss": 0.3887, "step": 7961 }, { "epoch": 3.7468235294117647, "grad_norm": 0.2009490900878427, "learning_rate": 0.0001243941843645778, "loss": 0.105, "step": 7962 }, { "epoch": 3.747294117647059, "grad_norm": 0.6295420524171415, "learning_rate": 0.00012435099270479246, "loss": 0.2676, "step": 7963 }, { "epoch": 3.7477647058823527, "grad_norm": 0.5210273398713319, "learning_rate": 0.00012430780323555816, "loss": 0.3281, "step": 7964 }, { "epoch": 3.748235294117647, "grad_norm": 0.4829768386193262, "learning_rate": 0.00012426461596056344, "loss": 0.2539, "step": 7965 }, { "epoch": 3.748705882352941, "grad_norm": 0.5879782346382378, "learning_rate": 0.00012422143088349679, "loss": 0.2949, "step": 7966 }, { "epoch": 3.7491764705882353, "grad_norm": 0.7685326234860641, "learning_rate": 0.00012417824800804638, "loss": 0.4375, "step": 7967 }, { "epoch": 3.7496470588235296, "grad_norm": 0.4815926585865583, "learning_rate": 0.0001241350673379002, "loss": 0.2578, "step": 7968 }, { "epoch": 3.7501176470588238, "grad_norm": 1.4146176649888313, "learning_rate": 0.0001240918888767462, "loss": 0.1523, "step": 7969 }, { "epoch": 3.7505882352941176, "grad_norm": 0.6764836074842301, "learning_rate": 0.00012404871262827193, "loss": 0.3281, "step": 7970 }, { "epoch": 3.751058823529412, "grad_norm": 0.46421965364145745, "learning_rate": 0.00012400553859616488, "loss": 0.1992, "step": 7971 }, { "epoch": 3.751529411764706, "grad_norm": 0.5414319033728667, "learning_rate": 0.00012396236678411236, "loss": 0.3438, "step": 7972 }, { "epoch": 3.752, "grad_norm": 0.5351661363250958, "learning_rate": 0.00012391919719580144, "loss": 0.3105, "step": 7973 }, { "epoch": 3.752470588235294, "grad_norm": 0.24518000410571872, "learning_rate": 0.00012387602983491894, "loss": 0.0571, "step": 7974 }, { "epoch": 3.7529411764705882, "grad_norm": 1.274343667383266, "learning_rate": 0.00012383286470515166, "loss": 0.6602, "step": 7975 }, { "epoch": 3.7534117647058824, "grad_norm": 0.47870896877074975, "learning_rate": 0.0001237897018101861, "loss": 0.1152, "step": 7976 }, { "epoch": 3.7538823529411767, "grad_norm": 0.716587049679995, "learning_rate": 0.00012374654115370856, "loss": 0.3301, "step": 7977 }, { "epoch": 3.7543529411764704, "grad_norm": 0.40005363801275884, "learning_rate": 0.0001237033827394052, "loss": 0.2012, "step": 7978 }, { "epoch": 3.7548235294117647, "grad_norm": 0.8828156966818513, "learning_rate": 0.0001236602265709619, "loss": 0.543, "step": 7979 }, { "epoch": 3.755294117647059, "grad_norm": 0.5024160720550331, "learning_rate": 0.00012361707265206448, "loss": 0.2344, "step": 7980 }, { "epoch": 3.7557647058823527, "grad_norm": 0.46357322002054185, "learning_rate": 0.00012357392098639848, "loss": 0.2207, "step": 7981 }, { "epoch": 3.756235294117647, "grad_norm": 0.48625409013953796, "learning_rate": 0.00012353077157764922, "loss": 0.248, "step": 7982 }, { "epoch": 3.756705882352941, "grad_norm": 0.7652672373735675, "learning_rate": 0.00012348762442950187, "loss": 0.3867, "step": 7983 }, { "epoch": 3.7571764705882353, "grad_norm": 0.5205973129310106, "learning_rate": 0.00012344447954564148, "loss": 0.2598, "step": 7984 }, { "epoch": 3.7576470588235296, "grad_norm": 0.3038090294205329, "learning_rate": 0.0001234013369297528, "loss": 0.084, "step": 7985 }, { "epoch": 3.758117647058824, "grad_norm": 0.9049246536087997, "learning_rate": 0.00012335819658552036, "loss": 0.5078, "step": 7986 }, { "epoch": 3.7585882352941176, "grad_norm": 0.9971049813115321, "learning_rate": 0.00012331505851662864, "loss": 0.5547, "step": 7987 }, { "epoch": 3.759058823529412, "grad_norm": 0.5298785177808807, "learning_rate": 0.00012327192272676183, "loss": 0.3281, "step": 7988 }, { "epoch": 3.759529411764706, "grad_norm": 0.6992207252961857, "learning_rate": 0.0001232287892196039, "loss": 0.3203, "step": 7989 }, { "epoch": 3.76, "grad_norm": 0.30431920358754283, "learning_rate": 0.00012318565799883865, "loss": 0.1689, "step": 7990 }, { "epoch": 3.760470588235294, "grad_norm": 0.27111951418688474, "learning_rate": 0.0001231425290681498, "loss": 0.0894, "step": 7991 }, { "epoch": 3.7609411764705882, "grad_norm": 0.4314799838232285, "learning_rate": 0.00012309940243122064, "loss": 0.2119, "step": 7992 }, { "epoch": 3.7614117647058825, "grad_norm": 0.46472919709255506, "learning_rate": 0.00012305627809173448, "loss": 0.2578, "step": 7993 }, { "epoch": 3.7618823529411767, "grad_norm": 0.34611501413242146, "learning_rate": 0.00012301315605337435, "loss": 0.1553, "step": 7994 }, { "epoch": 3.7623529411764705, "grad_norm": 0.5091296683770244, "learning_rate": 0.00012297003631982304, "loss": 0.2393, "step": 7995 }, { "epoch": 3.7628235294117647, "grad_norm": 0.5258719361242611, "learning_rate": 0.0001229269188947632, "loss": 0.416, "step": 7996 }, { "epoch": 3.763294117647059, "grad_norm": 0.8212051177519523, "learning_rate": 0.0001228838037818773, "loss": 0.3008, "step": 7997 }, { "epoch": 3.7637647058823527, "grad_norm": 0.8182406097155099, "learning_rate": 0.00012284069098484758, "loss": 0.3906, "step": 7998 }, { "epoch": 3.764235294117647, "grad_norm": 0.55732821546789, "learning_rate": 0.000122797580507356, "loss": 0.4004, "step": 7999 }, { "epoch": 3.764705882352941, "grad_norm": 0.7255015937037783, "learning_rate": 0.00012275447235308453, "loss": 0.3359, "step": 8000 }, { "epoch": 3.7651764705882353, "grad_norm": 0.5497656107969205, "learning_rate": 0.00012271136652571474, "loss": 0.2832, "step": 8001 }, { "epoch": 3.7656470588235296, "grad_norm": 0.20253369452634531, "learning_rate": 0.0001226682630289281, "loss": 0.0459, "step": 8002 }, { "epoch": 3.766117647058824, "grad_norm": 0.5846368165186486, "learning_rate": 0.00012262516186640587, "loss": 0.3027, "step": 8003 }, { "epoch": 3.7665882352941176, "grad_norm": 0.5251856374374932, "learning_rate": 0.0001225820630418291, "loss": 0.2578, "step": 8004 }, { "epoch": 3.767058823529412, "grad_norm": 0.4308301243856378, "learning_rate": 0.00012253896655887861, "loss": 0.1357, "step": 8005 }, { "epoch": 3.767529411764706, "grad_norm": 0.6052472342113724, "learning_rate": 0.00012249587242123505, "loss": 0.2139, "step": 8006 }, { "epoch": 3.768, "grad_norm": 0.4771604242091295, "learning_rate": 0.00012245278063257892, "loss": 0.252, "step": 8007 }, { "epoch": 3.768470588235294, "grad_norm": 0.31750201336416933, "learning_rate": 0.00012240969119659044, "loss": 0.082, "step": 8008 }, { "epoch": 3.7689411764705882, "grad_norm": 0.6220344730896141, "learning_rate": 0.00012236660411694964, "loss": 0.3164, "step": 8009 }, { "epoch": 3.7694117647058825, "grad_norm": 0.3086770312961515, "learning_rate": 0.00012232351939733642, "loss": 0.1138, "step": 8010 }, { "epoch": 3.7698823529411767, "grad_norm": 0.7053818643787992, "learning_rate": 0.0001222804370414304, "loss": 0.3477, "step": 8011 }, { "epoch": 3.7703529411764705, "grad_norm": 0.687851956297001, "learning_rate": 0.00012223735705291098, "loss": 0.5195, "step": 8012 }, { "epoch": 3.7708235294117647, "grad_norm": 0.48062967591970013, "learning_rate": 0.00012219427943545744, "loss": 0.2119, "step": 8013 }, { "epoch": 3.771294117647059, "grad_norm": 0.6338233719765511, "learning_rate": 0.00012215120419274887, "loss": 0.2373, "step": 8014 }, { "epoch": 3.7717647058823527, "grad_norm": 0.6460408231506579, "learning_rate": 0.00012210813132846406, "loss": 0.2949, "step": 8015 }, { "epoch": 3.772235294117647, "grad_norm": 1.296679984403052, "learning_rate": 0.0001220650608462816, "loss": 0.6016, "step": 8016 }, { "epoch": 3.772705882352941, "grad_norm": 0.5585158086211667, "learning_rate": 0.00012202199274987999, "loss": 0.2656, "step": 8017 }, { "epoch": 3.7731764705882354, "grad_norm": 0.28546359342302796, "learning_rate": 0.00012197892704293743, "loss": 0.1006, "step": 8018 }, { "epoch": 3.7736470588235296, "grad_norm": 0.48305904387487164, "learning_rate": 0.00012193586372913192, "loss": 0.1875, "step": 8019 }, { "epoch": 3.774117647058824, "grad_norm": 0.4282321901320762, "learning_rate": 0.00012189280281214126, "loss": 0.1553, "step": 8020 }, { "epoch": 3.7745882352941176, "grad_norm": 0.6409825009203286, "learning_rate": 0.00012184974429564318, "loss": 0.2451, "step": 8021 }, { "epoch": 3.775058823529412, "grad_norm": 0.2271147841712004, "learning_rate": 0.000121806688183315, "loss": 0.0464, "step": 8022 }, { "epoch": 3.775529411764706, "grad_norm": 0.8679180396518213, "learning_rate": 0.00012176363447883397, "loss": 0.4941, "step": 8023 }, { "epoch": 3.776, "grad_norm": 0.6446820359832055, "learning_rate": 0.00012172058318587698, "loss": 0.3594, "step": 8024 }, { "epoch": 3.776470588235294, "grad_norm": 0.46221958635442806, "learning_rate": 0.00012167753430812093, "loss": 0.3379, "step": 8025 }, { "epoch": 3.7769411764705882, "grad_norm": 0.5282565654941134, "learning_rate": 0.0001216344878492424, "loss": 0.3105, "step": 8026 }, { "epoch": 3.7774117647058825, "grad_norm": 0.5984822780545116, "learning_rate": 0.0001215914438129177, "loss": 0.2695, "step": 8027 }, { "epoch": 3.7778823529411767, "grad_norm": 0.6876173966383534, "learning_rate": 0.00012154840220282306, "loss": 0.3164, "step": 8028 }, { "epoch": 3.7783529411764705, "grad_norm": 0.3481544669401202, "learning_rate": 0.00012150536302263444, "loss": 0.1553, "step": 8029 }, { "epoch": 3.7788235294117647, "grad_norm": 0.3832041703694936, "learning_rate": 0.00012146232627602753, "loss": 0.1113, "step": 8030 }, { "epoch": 3.779294117647059, "grad_norm": 0.48539262861108967, "learning_rate": 0.00012141929196667798, "loss": 0.3145, "step": 8031 }, { "epoch": 3.7797647058823527, "grad_norm": 0.43310753887791664, "learning_rate": 0.00012137626009826108, "loss": 0.3105, "step": 8032 }, { "epoch": 3.780235294117647, "grad_norm": 0.4714277002979071, "learning_rate": 0.00012133323067445198, "loss": 0.2812, "step": 8033 }, { "epoch": 3.780705882352941, "grad_norm": 0.7163596036109596, "learning_rate": 0.00012129020369892555, "loss": 0.4922, "step": 8034 }, { "epoch": 3.7811764705882354, "grad_norm": 0.526470932156885, "learning_rate": 0.00012124717917535656, "loss": 0.2676, "step": 8035 }, { "epoch": 3.7816470588235296, "grad_norm": 0.4950505198960151, "learning_rate": 0.00012120415710741951, "loss": 0.3555, "step": 8036 }, { "epoch": 3.782117647058824, "grad_norm": 0.8686144610930366, "learning_rate": 0.00012116113749878867, "loss": 0.5273, "step": 8037 }, { "epoch": 3.7825882352941176, "grad_norm": 0.747229643627429, "learning_rate": 0.00012111812035313814, "loss": 0.3574, "step": 8038 }, { "epoch": 3.783058823529412, "grad_norm": 0.48177169893129795, "learning_rate": 0.00012107510567414183, "loss": 0.1611, "step": 8039 }, { "epoch": 3.783529411764706, "grad_norm": 0.23752932812499505, "learning_rate": 0.00012103209346547332, "loss": 0.0542, "step": 8040 }, { "epoch": 3.784, "grad_norm": 0.5788396868446409, "learning_rate": 0.00012098908373080617, "loss": 0.3184, "step": 8041 }, { "epoch": 3.784470588235294, "grad_norm": 0.40832375854704483, "learning_rate": 0.00012094607647381355, "loss": 0.1553, "step": 8042 }, { "epoch": 3.7849411764705883, "grad_norm": 0.39720675209483863, "learning_rate": 0.00012090307169816847, "loss": 0.1021, "step": 8043 }, { "epoch": 3.7854117647058825, "grad_norm": 0.451752762988628, "learning_rate": 0.00012086006940754382, "loss": 0.2793, "step": 8044 }, { "epoch": 3.7858823529411767, "grad_norm": 0.3795091085545391, "learning_rate": 0.00012081706960561217, "loss": 0.1172, "step": 8045 }, { "epoch": 3.7863529411764705, "grad_norm": 0.6098461365653021, "learning_rate": 0.00012077407229604593, "loss": 0.2949, "step": 8046 }, { "epoch": 3.7868235294117647, "grad_norm": 0.41847056368193003, "learning_rate": 0.00012073107748251724, "loss": 0.1758, "step": 8047 }, { "epoch": 3.787294117647059, "grad_norm": 0.675355496267361, "learning_rate": 0.00012068808516869812, "loss": 0.3086, "step": 8048 }, { "epoch": 3.7877647058823527, "grad_norm": 0.6253394634205108, "learning_rate": 0.00012064509535826029, "loss": 0.418, "step": 8049 }, { "epoch": 3.788235294117647, "grad_norm": 0.39707328523386454, "learning_rate": 0.00012060210805487529, "loss": 0.1226, "step": 8050 }, { "epoch": 3.788705882352941, "grad_norm": 0.6519837045123174, "learning_rate": 0.00012055912326221448, "loss": 0.4043, "step": 8051 }, { "epoch": 3.7891764705882354, "grad_norm": 0.42462689221709365, "learning_rate": 0.00012051614098394894, "loss": 0.2324, "step": 8052 }, { "epoch": 3.7896470588235296, "grad_norm": 0.5770110765903075, "learning_rate": 0.00012047316122374957, "loss": 0.3633, "step": 8053 }, { "epoch": 3.7901176470588234, "grad_norm": 0.5894300483024437, "learning_rate": 0.00012043018398528707, "loss": 0.3672, "step": 8054 }, { "epoch": 3.7905882352941176, "grad_norm": 0.3079354910568531, "learning_rate": 0.0001203872092722319, "loss": 0.1299, "step": 8055 }, { "epoch": 3.791058823529412, "grad_norm": 0.6683428315470229, "learning_rate": 0.00012034423708825431, "loss": 0.5938, "step": 8056 }, { "epoch": 3.791529411764706, "grad_norm": 0.36684566824136217, "learning_rate": 0.00012030126743702431, "loss": 0.1455, "step": 8057 }, { "epoch": 3.792, "grad_norm": 0.49674385695943907, "learning_rate": 0.00012025830032221177, "loss": 0.1475, "step": 8058 }, { "epoch": 3.792470588235294, "grad_norm": 0.7369006611612015, "learning_rate": 0.00012021533574748625, "loss": 0.4746, "step": 8059 }, { "epoch": 3.7929411764705883, "grad_norm": 0.4824419866720755, "learning_rate": 0.00012017237371651714, "loss": 0.2812, "step": 8060 }, { "epoch": 3.7934117647058825, "grad_norm": 0.6224078110028998, "learning_rate": 0.00012012941423297363, "loss": 0.4062, "step": 8061 }, { "epoch": 3.7938823529411767, "grad_norm": 0.5576754199649893, "learning_rate": 0.00012008645730052468, "loss": 0.3125, "step": 8062 }, { "epoch": 3.7943529411764705, "grad_norm": 0.13045578233721622, "learning_rate": 0.00012004350292283896, "loss": 0.0334, "step": 8063 }, { "epoch": 3.7948235294117647, "grad_norm": 0.5067030899400393, "learning_rate": 0.00012000055110358502, "loss": 0.3105, "step": 8064 }, { "epoch": 3.795294117647059, "grad_norm": 0.7656175311997397, "learning_rate": 0.00011995760184643118, "loss": 0.4453, "step": 8065 }, { "epoch": 3.7957647058823527, "grad_norm": 0.45510432196643635, "learning_rate": 0.00011991465515504551, "loss": 0.291, "step": 8066 }, { "epoch": 3.796235294117647, "grad_norm": 0.39327041841757465, "learning_rate": 0.00011987171103309586, "loss": 0.1172, "step": 8067 }, { "epoch": 3.796705882352941, "grad_norm": 0.4574949927413775, "learning_rate": 0.00011982876948424986, "loss": 0.2793, "step": 8068 }, { "epoch": 3.7971764705882354, "grad_norm": 0.8179859413686936, "learning_rate": 0.00011978583051217493, "loss": 0.4082, "step": 8069 }, { "epoch": 3.7976470588235296, "grad_norm": 0.536108763156429, "learning_rate": 0.0001197428941205383, "loss": 0.2871, "step": 8070 }, { "epoch": 3.7981176470588234, "grad_norm": 0.544131799402651, "learning_rate": 0.00011969996031300687, "loss": 0.3457, "step": 8071 }, { "epoch": 3.7985882352941176, "grad_norm": 0.8151602797196057, "learning_rate": 0.00011965702909324751, "loss": 0.4453, "step": 8072 }, { "epoch": 3.799058823529412, "grad_norm": 0.21338989203852488, "learning_rate": 0.00011961410046492667, "loss": 0.0981, "step": 8073 }, { "epoch": 3.799529411764706, "grad_norm": 0.29413002113441783, "learning_rate": 0.0001195711744317107, "loss": 0.1309, "step": 8074 }, { "epoch": 3.8, "grad_norm": 0.14371762862880685, "learning_rate": 0.00011952825099726567, "loss": 0.033, "step": 8075 }, { "epoch": 3.800470588235294, "grad_norm": 0.27241987378256893, "learning_rate": 0.00011948533016525749, "loss": 0.1011, "step": 8076 }, { "epoch": 3.8009411764705883, "grad_norm": 0.632970738204834, "learning_rate": 0.00011944241193935178, "loss": 0.2793, "step": 8077 }, { "epoch": 3.8014117647058825, "grad_norm": 0.560237715499901, "learning_rate": 0.00011939949632321395, "loss": 0.2305, "step": 8078 }, { "epoch": 3.8018823529411767, "grad_norm": 0.28638225088250896, "learning_rate": 0.00011935658332050925, "loss": 0.0796, "step": 8079 }, { "epoch": 3.8023529411764705, "grad_norm": 0.13060305306425876, "learning_rate": 0.00011931367293490265, "loss": 0.0457, "step": 8080 }, { "epoch": 3.8028235294117647, "grad_norm": 0.2503764989293188, "learning_rate": 0.00011927076517005885, "loss": 0.0579, "step": 8081 }, { "epoch": 3.803294117647059, "grad_norm": 0.3878592287102578, "learning_rate": 0.0001192278600296425, "loss": 0.1846, "step": 8082 }, { "epoch": 3.8037647058823527, "grad_norm": 0.10069481282762588, "learning_rate": 0.00011918495751731781, "loss": 0.0168, "step": 8083 }, { "epoch": 3.804235294117647, "grad_norm": 0.5930976253745077, "learning_rate": 0.0001191420576367489, "loss": 0.293, "step": 8084 }, { "epoch": 3.804705882352941, "grad_norm": 0.42422549461537684, "learning_rate": 0.00011909916039159958, "loss": 0.2891, "step": 8085 }, { "epoch": 3.8051764705882354, "grad_norm": 0.6280681190905203, "learning_rate": 0.00011905626578553359, "loss": 0.3301, "step": 8086 }, { "epoch": 3.8056470588235296, "grad_norm": 0.7143412973381924, "learning_rate": 0.00011901337382221425, "loss": 0.2715, "step": 8087 }, { "epoch": 3.8061176470588234, "grad_norm": 0.8485571239215992, "learning_rate": 0.00011897048450530477, "loss": 0.4863, "step": 8088 }, { "epoch": 3.8065882352941176, "grad_norm": 0.25368529767880493, "learning_rate": 0.00011892759783846812, "loss": 0.0649, "step": 8089 }, { "epoch": 3.807058823529412, "grad_norm": 0.7800417245420461, "learning_rate": 0.00011888471382536704, "loss": 0.3262, "step": 8090 }, { "epoch": 3.8075294117647056, "grad_norm": 0.4606034175021041, "learning_rate": 0.00011884183246966399, "loss": 0.1709, "step": 8091 }, { "epoch": 3.808, "grad_norm": 0.7821928427806668, "learning_rate": 0.00011879895377502129, "loss": 0.4688, "step": 8092 }, { "epoch": 3.808470588235294, "grad_norm": 0.27318366038186587, "learning_rate": 0.00011875607774510099, "loss": 0.0752, "step": 8093 }, { "epoch": 3.8089411764705883, "grad_norm": 0.34728236780338184, "learning_rate": 0.00011871320438356491, "loss": 0.0942, "step": 8094 }, { "epoch": 3.8094117647058825, "grad_norm": 0.603093194149576, "learning_rate": 0.0001186703336940746, "loss": 0.2871, "step": 8095 }, { "epoch": 3.8098823529411767, "grad_norm": 0.3373383040065065, "learning_rate": 0.00011862746568029149, "loss": 0.1602, "step": 8096 }, { "epoch": 3.8103529411764705, "grad_norm": 0.7790973014523618, "learning_rate": 0.00011858460034587669, "loss": 0.3633, "step": 8097 }, { "epoch": 3.8108235294117647, "grad_norm": 0.3184637741908391, "learning_rate": 0.0001185417376944911, "loss": 0.0962, "step": 8098 }, { "epoch": 3.811294117647059, "grad_norm": 0.8088468367311415, "learning_rate": 0.00011849887772979541, "loss": 0.3633, "step": 8099 }, { "epoch": 3.8117647058823527, "grad_norm": 0.30774576270249104, "learning_rate": 0.00011845602045545008, "loss": 0.1133, "step": 8100 }, { "epoch": 3.812235294117647, "grad_norm": 0.3198422455185708, "learning_rate": 0.00011841316587511531, "loss": 0.0981, "step": 8101 }, { "epoch": 3.812705882352941, "grad_norm": 0.520890087913708, "learning_rate": 0.00011837031399245115, "loss": 0.2119, "step": 8102 }, { "epoch": 3.8131764705882354, "grad_norm": 0.2982363973823625, "learning_rate": 0.00011832746481111728, "loss": 0.1079, "step": 8103 }, { "epoch": 3.8136470588235296, "grad_norm": 0.49148832132072207, "learning_rate": 0.00011828461833477326, "loss": 0.1377, "step": 8104 }, { "epoch": 3.8141176470588234, "grad_norm": 0.22413004173604475, "learning_rate": 0.00011824177456707844, "loss": 0.0515, "step": 8105 }, { "epoch": 3.8145882352941176, "grad_norm": 0.7630973628434095, "learning_rate": 0.00011819893351169183, "loss": 0.2988, "step": 8106 }, { "epoch": 3.815058823529412, "grad_norm": 0.8584883976476764, "learning_rate": 0.00011815609517227228, "loss": 0.3574, "step": 8107 }, { "epoch": 3.8155294117647056, "grad_norm": 0.4674152728871924, "learning_rate": 0.00011811325955247833, "loss": 0.3398, "step": 8108 }, { "epoch": 3.816, "grad_norm": 0.43745896853673094, "learning_rate": 0.00011807042665596846, "loss": 0.1582, "step": 8109 }, { "epoch": 3.816470588235294, "grad_norm": 0.2846760477125476, "learning_rate": 0.0001180275964864008, "loss": 0.1709, "step": 8110 }, { "epoch": 3.8169411764705883, "grad_norm": 0.46141052736949384, "learning_rate": 0.00011798476904743321, "loss": 0.2969, "step": 8111 }, { "epoch": 3.8174117647058825, "grad_norm": 0.3541496575325523, "learning_rate": 0.00011794194434272336, "loss": 0.127, "step": 8112 }, { "epoch": 3.8178823529411767, "grad_norm": 0.5241547595756715, "learning_rate": 0.00011789912237592874, "loss": 0.1963, "step": 8113 }, { "epoch": 3.8183529411764705, "grad_norm": 0.5443379873773921, "learning_rate": 0.00011785630315070651, "loss": 0.3145, "step": 8114 }, { "epoch": 3.8188235294117647, "grad_norm": 0.25312141637082824, "learning_rate": 0.00011781348667071365, "loss": 0.1494, "step": 8115 }, { "epoch": 3.819294117647059, "grad_norm": 0.5872254397135007, "learning_rate": 0.0001177706729396069, "loss": 0.2695, "step": 8116 }, { "epoch": 3.8197647058823527, "grad_norm": 0.8956669804252549, "learning_rate": 0.00011772786196104276, "loss": 0.4551, "step": 8117 }, { "epoch": 3.820235294117647, "grad_norm": 0.6759622714483663, "learning_rate": 0.00011768505373867753, "loss": 0.3672, "step": 8118 }, { "epoch": 3.820705882352941, "grad_norm": 0.6838584042376861, "learning_rate": 0.00011764224827616717, "loss": 0.377, "step": 8119 }, { "epoch": 3.8211764705882354, "grad_norm": 0.4403761129374907, "learning_rate": 0.00011759944557716755, "loss": 0.2969, "step": 8120 }, { "epoch": 3.8216470588235296, "grad_norm": 0.5573930790941269, "learning_rate": 0.00011755664564533422, "loss": 0.3516, "step": 8121 }, { "epoch": 3.8221176470588234, "grad_norm": 0.8344378042881629, "learning_rate": 0.00011751384848432245, "loss": 0.498, "step": 8122 }, { "epoch": 3.8225882352941176, "grad_norm": 0.36975935157125356, "learning_rate": 0.00011747105409778738, "loss": 0.1543, "step": 8123 }, { "epoch": 3.823058823529412, "grad_norm": 0.6736800143853626, "learning_rate": 0.00011742826248938384, "loss": 0.3555, "step": 8124 }, { "epoch": 3.8235294117647056, "grad_norm": 0.49575028228321844, "learning_rate": 0.00011738547366276645, "loss": 0.1943, "step": 8125 }, { "epoch": 3.824, "grad_norm": 0.5286419597750693, "learning_rate": 0.00011734268762158957, "loss": 0.2949, "step": 8126 }, { "epoch": 3.824470588235294, "grad_norm": 0.5765329359137907, "learning_rate": 0.00011729990436950737, "loss": 0.3105, "step": 8127 }, { "epoch": 3.8249411764705883, "grad_norm": 0.5443559209471002, "learning_rate": 0.00011725712391017374, "loss": 0.3652, "step": 8128 }, { "epoch": 3.8254117647058825, "grad_norm": 0.6117537084089566, "learning_rate": 0.00011721434624724227, "loss": 0.3516, "step": 8129 }, { "epoch": 3.8258823529411767, "grad_norm": 0.2632130181423749, "learning_rate": 0.0001171715713843665, "loss": 0.1084, "step": 8130 }, { "epoch": 3.8263529411764705, "grad_norm": 0.40086283418223123, "learning_rate": 0.00011712879932519954, "loss": 0.2451, "step": 8131 }, { "epoch": 3.8268235294117647, "grad_norm": 0.3167650488241545, "learning_rate": 0.00011708603007339433, "loss": 0.1074, "step": 8132 }, { "epoch": 3.827294117647059, "grad_norm": 0.4813536093689656, "learning_rate": 0.00011704326363260365, "loss": 0.2617, "step": 8133 }, { "epoch": 3.8277647058823527, "grad_norm": 0.45303367463382305, "learning_rate": 0.00011700050000647986, "loss": 0.3438, "step": 8134 }, { "epoch": 3.828235294117647, "grad_norm": 0.7094993100000999, "learning_rate": 0.00011695773919867525, "loss": 0.6289, "step": 8135 }, { "epoch": 3.828705882352941, "grad_norm": 0.3714047740278238, "learning_rate": 0.00011691498121284176, "loss": 0.1211, "step": 8136 }, { "epoch": 3.8291764705882354, "grad_norm": 0.3712102417420187, "learning_rate": 0.00011687222605263118, "loss": 0.1377, "step": 8137 }, { "epoch": 3.8296470588235296, "grad_norm": 0.39472928525924816, "learning_rate": 0.000116829473721695, "loss": 0.2334, "step": 8138 }, { "epoch": 3.8301176470588234, "grad_norm": 0.6193715096949941, "learning_rate": 0.00011678672422368442, "loss": 0.3379, "step": 8139 }, { "epoch": 3.8305882352941176, "grad_norm": 0.5804828603132073, "learning_rate": 0.00011674397756225053, "loss": 0.416, "step": 8140 }, { "epoch": 3.831058823529412, "grad_norm": 0.7844733445935319, "learning_rate": 0.00011670123374104409, "loss": 0.4434, "step": 8141 }, { "epoch": 3.8315294117647056, "grad_norm": 0.4227224083601953, "learning_rate": 0.00011665849276371557, "loss": 0.1953, "step": 8142 }, { "epoch": 3.832, "grad_norm": 0.616431360144137, "learning_rate": 0.00011661575463391534, "loss": 0.3809, "step": 8143 }, { "epoch": 3.832470588235294, "grad_norm": 0.22724699190212008, "learning_rate": 0.0001165730193552934, "loss": 0.052, "step": 8144 }, { "epoch": 3.8329411764705883, "grad_norm": 0.899536060235808, "learning_rate": 0.00011653028693149961, "loss": 0.5547, "step": 8145 }, { "epoch": 3.8334117647058825, "grad_norm": 0.8590543865612988, "learning_rate": 0.00011648755736618342, "loss": 0.5234, "step": 8146 }, { "epoch": 3.8338823529411767, "grad_norm": 0.6229522570645304, "learning_rate": 0.00011644483066299424, "loss": 0.3965, "step": 8147 }, { "epoch": 3.8343529411764705, "grad_norm": 0.49555964970356686, "learning_rate": 0.00011640210682558112, "loss": 0.1631, "step": 8148 }, { "epoch": 3.8348235294117647, "grad_norm": 0.7522373002319354, "learning_rate": 0.00011635938585759284, "loss": 0.4512, "step": 8149 }, { "epoch": 3.835294117647059, "grad_norm": 0.6863502198844872, "learning_rate": 0.00011631666776267803, "loss": 0.3633, "step": 8150 }, { "epoch": 3.8357647058823527, "grad_norm": 0.531415585101756, "learning_rate": 0.00011627395254448504, "loss": 0.2715, "step": 8151 }, { "epoch": 3.836235294117647, "grad_norm": 0.5935674026317946, "learning_rate": 0.00011623124020666186, "loss": 0.2969, "step": 8152 }, { "epoch": 3.836705882352941, "grad_norm": 0.6616257356982554, "learning_rate": 0.00011618853075285644, "loss": 0.3438, "step": 8153 }, { "epoch": 3.8371764705882354, "grad_norm": 0.7045583570832634, "learning_rate": 0.00011614582418671629, "loss": 0.5391, "step": 8154 }, { "epoch": 3.8376470588235296, "grad_norm": 0.4507284048870639, "learning_rate": 0.00011610312051188888, "loss": 0.2354, "step": 8155 }, { "epoch": 3.8381176470588234, "grad_norm": 0.5835856186092931, "learning_rate": 0.00011606041973202123, "loss": 0.3184, "step": 8156 }, { "epoch": 3.8385882352941176, "grad_norm": 0.49727616098626104, "learning_rate": 0.00011601772185076018, "loss": 0.2773, "step": 8157 }, { "epoch": 3.839058823529412, "grad_norm": 0.7892142755437274, "learning_rate": 0.00011597502687175238, "loss": 0.4805, "step": 8158 }, { "epoch": 3.8395294117647056, "grad_norm": 0.3568037712486429, "learning_rate": 0.0001159323347986442, "loss": 0.1621, "step": 8159 }, { "epoch": 3.84, "grad_norm": 0.35849505981977087, "learning_rate": 0.00011588964563508167, "loss": 0.0742, "step": 8160 }, { "epoch": 3.840470588235294, "grad_norm": 0.3964017719869632, "learning_rate": 0.00011584695938471077, "loss": 0.1436, "step": 8161 }, { "epoch": 3.8409411764705883, "grad_norm": 0.32331761842362605, "learning_rate": 0.00011580427605117705, "loss": 0.1279, "step": 8162 }, { "epoch": 3.8414117647058825, "grad_norm": 0.5585584637646803, "learning_rate": 0.00011576159563812589, "loss": 0.1709, "step": 8163 }, { "epoch": 3.8418823529411767, "grad_norm": 0.3718015503799811, "learning_rate": 0.00011571891814920236, "loss": 0.1494, "step": 8164 }, { "epoch": 3.8423529411764705, "grad_norm": 0.6161594273751849, "learning_rate": 0.00011567624358805139, "loss": 0.3984, "step": 8165 }, { "epoch": 3.8428235294117647, "grad_norm": 0.7582939097269636, "learning_rate": 0.0001156335719583176, "loss": 0.416, "step": 8166 }, { "epoch": 3.843294117647059, "grad_norm": 0.49139865321548515, "learning_rate": 0.00011559090326364527, "loss": 0.1523, "step": 8167 }, { "epoch": 3.8437647058823528, "grad_norm": 0.5657433515118512, "learning_rate": 0.00011554823750767863, "loss": 0.3242, "step": 8168 }, { "epoch": 3.844235294117647, "grad_norm": 0.5912643640362155, "learning_rate": 0.0001155055746940615, "loss": 0.4688, "step": 8169 }, { "epoch": 3.844705882352941, "grad_norm": 0.4874268157683087, "learning_rate": 0.00011546291482643744, "loss": 0.25, "step": 8170 }, { "epoch": 3.8451764705882354, "grad_norm": 0.38860188465944406, "learning_rate": 0.00011542025790844988, "loss": 0.2227, "step": 8171 }, { "epoch": 3.8456470588235296, "grad_norm": 0.6309024833712826, "learning_rate": 0.00011537760394374192, "loss": 0.4414, "step": 8172 }, { "epoch": 3.8461176470588234, "grad_norm": 0.5836883542482753, "learning_rate": 0.00011533495293595637, "loss": 0.2949, "step": 8173 }, { "epoch": 3.8465882352941176, "grad_norm": 0.41474721046606505, "learning_rate": 0.0001152923048887359, "loss": 0.2197, "step": 8174 }, { "epoch": 3.847058823529412, "grad_norm": 0.3744824341025988, "learning_rate": 0.00011524965980572284, "loss": 0.2695, "step": 8175 }, { "epoch": 3.8475294117647056, "grad_norm": 0.6374864117426406, "learning_rate": 0.00011520701769055927, "loss": 0.332, "step": 8176 }, { "epoch": 3.848, "grad_norm": 0.30375995002863043, "learning_rate": 0.00011516437854688703, "loss": 0.1001, "step": 8177 }, { "epoch": 3.848470588235294, "grad_norm": 0.6318534852531288, "learning_rate": 0.00011512174237834776, "loss": 0.3574, "step": 8178 }, { "epoch": 3.8489411764705883, "grad_norm": 0.37553887403073466, "learning_rate": 0.00011507910918858276, "loss": 0.1973, "step": 8179 }, { "epoch": 3.8494117647058825, "grad_norm": 0.4327299020706276, "learning_rate": 0.0001150364789812331, "loss": 0.2148, "step": 8180 }, { "epoch": 3.8498823529411763, "grad_norm": 0.30402245324300725, "learning_rate": 0.00011499385175993965, "loss": 0.1011, "step": 8181 }, { "epoch": 3.8503529411764705, "grad_norm": 0.6819488888497427, "learning_rate": 0.00011495122752834296, "loss": 0.4414, "step": 8182 }, { "epoch": 3.8508235294117648, "grad_norm": 0.14450328776270532, "learning_rate": 0.00011490860629008333, "loss": 0.0273, "step": 8183 }, { "epoch": 3.851294117647059, "grad_norm": 0.5052561719705487, "learning_rate": 0.00011486598804880088, "loss": 0.3047, "step": 8184 }, { "epoch": 3.8517647058823528, "grad_norm": 0.41767723839062904, "learning_rate": 0.00011482337280813536, "loss": 0.2734, "step": 8185 }, { "epoch": 3.852235294117647, "grad_norm": 0.4547196731720571, "learning_rate": 0.00011478076057172638, "loss": 0.1973, "step": 8186 }, { "epoch": 3.852705882352941, "grad_norm": 0.34604345292784255, "learning_rate": 0.00011473815134321315, "loss": 0.1494, "step": 8187 }, { "epoch": 3.8531764705882354, "grad_norm": 1.0811991214111416, "learning_rate": 0.00011469554512623477, "loss": 0.5547, "step": 8188 }, { "epoch": 3.8536470588235296, "grad_norm": 0.650154551773605, "learning_rate": 0.00011465294192443001, "loss": 0.4141, "step": 8189 }, { "epoch": 3.8541176470588234, "grad_norm": 0.7466299353531201, "learning_rate": 0.00011461034174143736, "loss": 0.4492, "step": 8190 }, { "epoch": 3.8545882352941176, "grad_norm": 0.3056089293852044, "learning_rate": 0.00011456774458089515, "loss": 0.1001, "step": 8191 }, { "epoch": 3.855058823529412, "grad_norm": 0.5071031546771015, "learning_rate": 0.00011452515044644132, "loss": 0.2754, "step": 8192 }, { "epoch": 3.8555294117647056, "grad_norm": 0.6303149255225938, "learning_rate": 0.00011448255934171363, "loss": 0.4512, "step": 8193 }, { "epoch": 3.856, "grad_norm": 0.1681567199886074, "learning_rate": 0.00011443997127034962, "loss": 0.0374, "step": 8194 }, { "epoch": 3.856470588235294, "grad_norm": 0.6563726013426331, "learning_rate": 0.0001143973862359865, "loss": 0.4473, "step": 8195 }, { "epoch": 3.8569411764705883, "grad_norm": 0.6029349205374509, "learning_rate": 0.0001143548042422612, "loss": 0.3613, "step": 8196 }, { "epoch": 3.8574117647058825, "grad_norm": 0.47482070738365884, "learning_rate": 0.0001143122252928104, "loss": 0.3145, "step": 8197 }, { "epoch": 3.8578823529411763, "grad_norm": 0.5361677011296007, "learning_rate": 0.00011426964939127063, "loss": 0.3066, "step": 8198 }, { "epoch": 3.8583529411764705, "grad_norm": 0.6359726901767094, "learning_rate": 0.00011422707654127809, "loss": 0.4531, "step": 8199 }, { "epoch": 3.8588235294117648, "grad_norm": 0.4494388144739643, "learning_rate": 0.00011418450674646868, "loss": 0.127, "step": 8200 }, { "epoch": 3.859294117647059, "grad_norm": 0.5247850288302389, "learning_rate": 0.00011414194001047806, "loss": 0.2578, "step": 8201 }, { "epoch": 3.8597647058823528, "grad_norm": 0.47422070403825445, "learning_rate": 0.00011409937633694165, "loss": 0.2246, "step": 8202 }, { "epoch": 3.860235294117647, "grad_norm": 0.5808555426791175, "learning_rate": 0.0001140568157294946, "loss": 0.2676, "step": 8203 }, { "epoch": 3.860705882352941, "grad_norm": 0.725039540467553, "learning_rate": 0.00011401425819177179, "loss": 0.6016, "step": 8204 }, { "epoch": 3.8611764705882354, "grad_norm": 0.5583190628002582, "learning_rate": 0.00011397170372740782, "loss": 0.2617, "step": 8205 }, { "epoch": 3.8616470588235297, "grad_norm": 0.5244625786931265, "learning_rate": 0.0001139291523400371, "loss": 0.334, "step": 8206 }, { "epoch": 3.8621176470588234, "grad_norm": 0.40034161580210215, "learning_rate": 0.00011388660403329368, "loss": 0.1816, "step": 8207 }, { "epoch": 3.8625882352941177, "grad_norm": 0.7536348584677585, "learning_rate": 0.0001138440588108114, "loss": 0.4824, "step": 8208 }, { "epoch": 3.863058823529412, "grad_norm": 0.5577925494676232, "learning_rate": 0.00011380151667622384, "loss": 0.3672, "step": 8209 }, { "epoch": 3.8635294117647057, "grad_norm": 0.21147536053364768, "learning_rate": 0.00011375897763316433, "loss": 0.0776, "step": 8210 }, { "epoch": 3.864, "grad_norm": 0.5560943523717938, "learning_rate": 0.00011371644168526586, "loss": 0.3691, "step": 8211 }, { "epoch": 3.864470588235294, "grad_norm": 0.3836107126938781, "learning_rate": 0.00011367390883616126, "loss": 0.1777, "step": 8212 }, { "epoch": 3.8649411764705883, "grad_norm": 0.762134507437322, "learning_rate": 0.00011363137908948303, "loss": 0.3125, "step": 8213 }, { "epoch": 3.8654117647058825, "grad_norm": 0.39455157319118916, "learning_rate": 0.00011358885244886342, "loss": 0.167, "step": 8214 }, { "epoch": 3.8658823529411763, "grad_norm": 0.7074041471451468, "learning_rate": 0.00011354632891793437, "loss": 0.2988, "step": 8215 }, { "epoch": 3.8663529411764705, "grad_norm": 0.35690499715101537, "learning_rate": 0.00011350380850032766, "loss": 0.1338, "step": 8216 }, { "epoch": 3.8668235294117648, "grad_norm": 0.501434943232515, "learning_rate": 0.00011346129119967473, "loss": 0.2949, "step": 8217 }, { "epoch": 3.8672941176470585, "grad_norm": 0.2704956875348438, "learning_rate": 0.0001134187770196067, "loss": 0.0986, "step": 8218 }, { "epoch": 3.8677647058823528, "grad_norm": 0.5147913497870181, "learning_rate": 0.00011337626596375458, "loss": 0.1426, "step": 8219 }, { "epoch": 3.868235294117647, "grad_norm": 0.30721259123419636, "learning_rate": 0.000113333758035749, "loss": 0.1074, "step": 8220 }, { "epoch": 3.868705882352941, "grad_norm": 0.5746792761281503, "learning_rate": 0.00011329125323922031, "loss": 0.2891, "step": 8221 }, { "epoch": 3.8691764705882354, "grad_norm": 0.20773927212016471, "learning_rate": 0.00011324875157779865, "loss": 0.0825, "step": 8222 }, { "epoch": 3.8696470588235297, "grad_norm": 0.5198551733233665, "learning_rate": 0.00011320625305511388, "loss": 0.1885, "step": 8223 }, { "epoch": 3.8701176470588234, "grad_norm": 0.5125778408715137, "learning_rate": 0.00011316375767479554, "loss": 0.4062, "step": 8224 }, { "epoch": 3.8705882352941177, "grad_norm": 0.41702162997145586, "learning_rate": 0.00011312126544047302, "loss": 0.1045, "step": 8225 }, { "epoch": 3.871058823529412, "grad_norm": 0.47319486148602624, "learning_rate": 0.0001130787763557753, "loss": 0.2295, "step": 8226 }, { "epoch": 3.8715294117647057, "grad_norm": 0.31124932374304104, "learning_rate": 0.00011303629042433118, "loss": 0.209, "step": 8227 }, { "epoch": 3.872, "grad_norm": 0.3100198521444955, "learning_rate": 0.00011299380764976915, "loss": 0.1582, "step": 8228 }, { "epoch": 3.872470588235294, "grad_norm": 0.5737130127710405, "learning_rate": 0.00011295132803571747, "loss": 0.3809, "step": 8229 }, { "epoch": 3.8729411764705883, "grad_norm": 0.5589242164817606, "learning_rate": 0.00011290885158580411, "loss": 0.2539, "step": 8230 }, { "epoch": 3.8734117647058826, "grad_norm": 0.15238981405436844, "learning_rate": 0.00011286637830365672, "loss": 0.0383, "step": 8231 }, { "epoch": 3.8738823529411763, "grad_norm": 0.2633355676627995, "learning_rate": 0.0001128239081929028, "loss": 0.0928, "step": 8232 }, { "epoch": 3.8743529411764706, "grad_norm": 0.7406571175561624, "learning_rate": 0.00011278144125716946, "loss": 0.3711, "step": 8233 }, { "epoch": 3.8748235294117648, "grad_norm": 0.2158942913902589, "learning_rate": 0.00011273897750008358, "loss": 0.0559, "step": 8234 }, { "epoch": 3.8752941176470586, "grad_norm": 0.5100254618158804, "learning_rate": 0.0001126965169252718, "loss": 0.2412, "step": 8235 }, { "epoch": 3.875764705882353, "grad_norm": 0.6803194029803824, "learning_rate": 0.00011265405953636044, "loss": 0.3555, "step": 8236 }, { "epoch": 3.876235294117647, "grad_norm": 0.5330824663284414, "learning_rate": 0.0001126116053369756, "loss": 0.1865, "step": 8237 }, { "epoch": 3.8767058823529412, "grad_norm": 0.531382929139516, "learning_rate": 0.000112569154330743, "loss": 0.334, "step": 8238 }, { "epoch": 3.8771764705882354, "grad_norm": 0.8780775919269359, "learning_rate": 0.00011252670652128825, "loss": 0.6445, "step": 8239 }, { "epoch": 3.8776470588235297, "grad_norm": 0.39221186430743793, "learning_rate": 0.00011248426191223658, "loss": 0.2578, "step": 8240 }, { "epoch": 3.8781176470588234, "grad_norm": 0.6820740018529841, "learning_rate": 0.00011244182050721293, "loss": 0.3496, "step": 8241 }, { "epoch": 3.8785882352941177, "grad_norm": 0.5059581874391176, "learning_rate": 0.00011239938230984198, "loss": 0.2412, "step": 8242 }, { "epoch": 3.879058823529412, "grad_norm": 1.035178112973526, "learning_rate": 0.00011235694732374829, "loss": 0.4414, "step": 8243 }, { "epoch": 3.8795294117647057, "grad_norm": 0.6218629852114688, "learning_rate": 0.00011231451555255592, "loss": 0.2354, "step": 8244 }, { "epoch": 3.88, "grad_norm": 0.5502531828254241, "learning_rate": 0.00011227208699988879, "loss": 0.2832, "step": 8245 }, { "epoch": 3.880470588235294, "grad_norm": 0.4144586528228098, "learning_rate": 0.00011222966166937044, "loss": 0.3027, "step": 8246 }, { "epoch": 3.8809411764705883, "grad_norm": 0.303906590523202, "learning_rate": 0.00011218723956462427, "loss": 0.1377, "step": 8247 }, { "epoch": 3.8814117647058826, "grad_norm": 0.6119116270288429, "learning_rate": 0.00011214482068927332, "loss": 0.293, "step": 8248 }, { "epoch": 3.8818823529411763, "grad_norm": 0.39406781404912017, "learning_rate": 0.00011210240504694033, "loss": 0.2832, "step": 8249 }, { "epoch": 3.8823529411764706, "grad_norm": 0.7799351730330356, "learning_rate": 0.00011205999264124786, "loss": 0.4102, "step": 8250 }, { "epoch": 3.882823529411765, "grad_norm": 0.274652405218952, "learning_rate": 0.00011201758347581814, "loss": 0.0791, "step": 8251 }, { "epoch": 3.8832941176470586, "grad_norm": 0.7239743158406101, "learning_rate": 0.00011197517755427303, "loss": 0.4023, "step": 8252 }, { "epoch": 3.883764705882353, "grad_norm": 0.516308961560043, "learning_rate": 0.00011193277488023432, "loss": 0.3535, "step": 8253 }, { "epoch": 3.884235294117647, "grad_norm": 0.711642073123429, "learning_rate": 0.00011189037545732335, "loss": 0.2012, "step": 8254 }, { "epoch": 3.8847058823529412, "grad_norm": 0.7104165585581727, "learning_rate": 0.00011184797928916124, "loss": 0.5547, "step": 8255 }, { "epoch": 3.8851764705882355, "grad_norm": 0.7318918768972249, "learning_rate": 0.0001118055863793688, "loss": 0.2852, "step": 8256 }, { "epoch": 3.8856470588235297, "grad_norm": 0.48608711146533984, "learning_rate": 0.00011176319673156666, "loss": 0.2754, "step": 8257 }, { "epoch": 3.8861176470588235, "grad_norm": 0.313999334569218, "learning_rate": 0.00011172081034937507, "loss": 0.1396, "step": 8258 }, { "epoch": 3.8865882352941177, "grad_norm": 0.4889056623359122, "learning_rate": 0.00011167842723641398, "loss": 0.2715, "step": 8259 }, { "epoch": 3.887058823529412, "grad_norm": 0.2651085100820762, "learning_rate": 0.00011163604739630324, "loss": 0.0874, "step": 8260 }, { "epoch": 3.8875294117647057, "grad_norm": 0.7286992223850572, "learning_rate": 0.00011159367083266219, "loss": 0.3711, "step": 8261 }, { "epoch": 3.888, "grad_norm": 0.29362202848256275, "learning_rate": 0.00011155129754911002, "loss": 0.1416, "step": 8262 }, { "epoch": 3.888470588235294, "grad_norm": 0.40351656173495304, "learning_rate": 0.00011150892754926568, "loss": 0.2285, "step": 8263 }, { "epoch": 3.8889411764705883, "grad_norm": 0.5872554838442207, "learning_rate": 0.00011146656083674768, "loss": 0.3086, "step": 8264 }, { "epoch": 3.8894117647058826, "grad_norm": 0.2367159475253318, "learning_rate": 0.00011142419741517443, "loss": 0.0879, "step": 8265 }, { "epoch": 3.8898823529411763, "grad_norm": 0.8176801510655156, "learning_rate": 0.00011138183728816389, "loss": 0.5938, "step": 8266 }, { "epoch": 3.8903529411764706, "grad_norm": 0.4258677181839045, "learning_rate": 0.0001113394804593339, "loss": 0.2373, "step": 8267 }, { "epoch": 3.890823529411765, "grad_norm": 0.6515008260632195, "learning_rate": 0.00011129712693230191, "loss": 0.3008, "step": 8268 }, { "epoch": 3.8912941176470586, "grad_norm": 0.32820590389240073, "learning_rate": 0.00011125477671068506, "loss": 0.1172, "step": 8269 }, { "epoch": 3.891764705882353, "grad_norm": 0.37955455800579224, "learning_rate": 0.0001112124297981004, "loss": 0.1611, "step": 8270 }, { "epoch": 3.892235294117647, "grad_norm": 0.29586318662827515, "learning_rate": 0.00011117008619816445, "loss": 0.0708, "step": 8271 }, { "epoch": 3.8927058823529412, "grad_norm": 0.5693751705367079, "learning_rate": 0.00011112774591449356, "loss": 0.3457, "step": 8272 }, { "epoch": 3.8931764705882355, "grad_norm": 0.23334249623299325, "learning_rate": 0.00011108540895070388, "loss": 0.0898, "step": 8273 }, { "epoch": 3.8936470588235297, "grad_norm": 0.25185969654502044, "learning_rate": 0.00011104307531041117, "loss": 0.0786, "step": 8274 }, { "epoch": 3.8941176470588235, "grad_norm": 0.3962382569765075, "learning_rate": 0.0001110007449972309, "loss": 0.1533, "step": 8275 }, { "epoch": 3.8945882352941177, "grad_norm": 0.7138529626102591, "learning_rate": 0.00011095841801477826, "loss": 0.3867, "step": 8276 }, { "epoch": 3.895058823529412, "grad_norm": 0.3827423708484413, "learning_rate": 0.00011091609436666825, "loss": 0.1738, "step": 8277 }, { "epoch": 3.8955294117647057, "grad_norm": 0.4136801721525831, "learning_rate": 0.00011087377405651549, "loss": 0.1562, "step": 8278 }, { "epoch": 3.896, "grad_norm": 0.759695242059863, "learning_rate": 0.00011083145708793431, "loss": 0.4629, "step": 8279 }, { "epoch": 3.896470588235294, "grad_norm": 0.7388747171330817, "learning_rate": 0.00011078914346453883, "loss": 0.3164, "step": 8280 }, { "epoch": 3.8969411764705884, "grad_norm": 0.13811404874980218, "learning_rate": 0.00011074683318994286, "loss": 0.0231, "step": 8281 }, { "epoch": 3.8974117647058826, "grad_norm": 0.7759397223526745, "learning_rate": 0.00011070452626775981, "loss": 0.5664, "step": 8282 }, { "epoch": 3.8978823529411764, "grad_norm": 1.0385281043160715, "learning_rate": 0.000110662222701603, "loss": 0.3789, "step": 8283 }, { "epoch": 3.8983529411764706, "grad_norm": 0.5363885514626823, "learning_rate": 0.00011061992249508533, "loss": 0.1963, "step": 8284 }, { "epoch": 3.898823529411765, "grad_norm": 0.5163416452390589, "learning_rate": 0.0001105776256518194, "loss": 0.21, "step": 8285 }, { "epoch": 3.8992941176470586, "grad_norm": 0.37359476767193944, "learning_rate": 0.0001105353321754176, "loss": 0.1895, "step": 8286 }, { "epoch": 3.899764705882353, "grad_norm": 0.7623042864325409, "learning_rate": 0.000110493042069492, "loss": 0.4121, "step": 8287 }, { "epoch": 3.900235294117647, "grad_norm": 0.4538768227913089, "learning_rate": 0.00011045075533765444, "loss": 0.165, "step": 8288 }, { "epoch": 3.9007058823529412, "grad_norm": 0.6073277147830114, "learning_rate": 0.00011040847198351635, "loss": 0.2598, "step": 8289 }, { "epoch": 3.9011764705882355, "grad_norm": 0.9356726890032675, "learning_rate": 0.00011036619201068892, "loss": 0.4805, "step": 8290 }, { "epoch": 3.9016470588235292, "grad_norm": 0.9858961504522076, "learning_rate": 0.00011032391542278313, "loss": 0.3477, "step": 8291 }, { "epoch": 3.9021176470588235, "grad_norm": 0.7820188929063516, "learning_rate": 0.00011028164222340957, "loss": 0.6133, "step": 8292 }, { "epoch": 3.9025882352941177, "grad_norm": 0.3001276714460363, "learning_rate": 0.00011023937241617853, "loss": 0.1377, "step": 8293 }, { "epoch": 3.903058823529412, "grad_norm": 0.49674271713884033, "learning_rate": 0.00011019710600470016, "loss": 0.2695, "step": 8294 }, { "epoch": 3.9035294117647057, "grad_norm": 0.5628390802766098, "learning_rate": 0.00011015484299258417, "loss": 0.3145, "step": 8295 }, { "epoch": 3.904, "grad_norm": 0.4054610935125804, "learning_rate": 0.00011011258338344002, "loss": 0.123, "step": 8296 }, { "epoch": 3.904470588235294, "grad_norm": 0.7021509142221426, "learning_rate": 0.00011007032718087688, "loss": 0.3027, "step": 8297 }, { "epoch": 3.9049411764705884, "grad_norm": 0.34475728275211937, "learning_rate": 0.00011002807438850367, "loss": 0.0967, "step": 8298 }, { "epoch": 3.9054117647058826, "grad_norm": 0.2573717217051084, "learning_rate": 0.00010998582500992897, "loss": 0.0806, "step": 8299 }, { "epoch": 3.9058823529411764, "grad_norm": 0.6854121234466919, "learning_rate": 0.00010994357904876106, "loss": 0.291, "step": 8300 }, { "epoch": 3.9063529411764706, "grad_norm": 0.3294848276080086, "learning_rate": 0.000109901336508608, "loss": 0.1445, "step": 8301 }, { "epoch": 3.906823529411765, "grad_norm": 0.3643826669349351, "learning_rate": 0.00010985909739307748, "loss": 0.1406, "step": 8302 }, { "epoch": 3.9072941176470586, "grad_norm": 0.377883808497881, "learning_rate": 0.0001098168617057769, "loss": 0.1387, "step": 8303 }, { "epoch": 3.907764705882353, "grad_norm": 0.8606138446759711, "learning_rate": 0.00010977462945031347, "loss": 0.3496, "step": 8304 }, { "epoch": 3.908235294117647, "grad_norm": 0.6339529576252128, "learning_rate": 0.00010973240063029398, "loss": 0.3594, "step": 8305 }, { "epoch": 3.9087058823529413, "grad_norm": 0.4139533165263195, "learning_rate": 0.000109690175249325, "loss": 0.1523, "step": 8306 }, { "epoch": 3.9091764705882355, "grad_norm": 0.8258576327880368, "learning_rate": 0.00010964795331101273, "loss": 0.1982, "step": 8307 }, { "epoch": 3.9096470588235293, "grad_norm": 0.3942003071908298, "learning_rate": 0.00010960573481896322, "loss": 0.1533, "step": 8308 }, { "epoch": 3.9101176470588235, "grad_norm": 0.28576862889284516, "learning_rate": 0.00010956351977678206, "loss": 0.0688, "step": 8309 }, { "epoch": 3.9105882352941177, "grad_norm": 0.5915106755859845, "learning_rate": 0.00010952130818807465, "loss": 0.3125, "step": 8310 }, { "epoch": 3.911058823529412, "grad_norm": 0.5399866535318628, "learning_rate": 0.00010947910005644608, "loss": 0.1992, "step": 8311 }, { "epoch": 3.9115294117647057, "grad_norm": 0.48103379046446554, "learning_rate": 0.00010943689538550111, "loss": 0.2119, "step": 8312 }, { "epoch": 3.912, "grad_norm": 0.3883421888518301, "learning_rate": 0.00010939469417884424, "loss": 0.0864, "step": 8313 }, { "epoch": 3.912470588235294, "grad_norm": 0.7691921807444908, "learning_rate": 0.00010935249644007966, "loss": 0.3477, "step": 8314 }, { "epoch": 3.9129411764705884, "grad_norm": 0.8266974502531548, "learning_rate": 0.00010931030217281125, "loss": 0.2852, "step": 8315 }, { "epoch": 3.9134117647058826, "grad_norm": 0.24038669045006544, "learning_rate": 0.00010926811138064264, "loss": 0.0459, "step": 8316 }, { "epoch": 3.9138823529411764, "grad_norm": 0.4705631390437569, "learning_rate": 0.00010922592406717706, "loss": 0.1787, "step": 8317 }, { "epoch": 3.9143529411764706, "grad_norm": 0.4390641923066702, "learning_rate": 0.0001091837402360176, "loss": 0.2793, "step": 8318 }, { "epoch": 3.914823529411765, "grad_norm": 0.5764435136598154, "learning_rate": 0.00010914155989076692, "loss": 0.3555, "step": 8319 }, { "epoch": 3.9152941176470586, "grad_norm": 0.23926500546974638, "learning_rate": 0.00010909938303502741, "loss": 0.0801, "step": 8320 }, { "epoch": 3.915764705882353, "grad_norm": 0.2858590315549084, "learning_rate": 0.00010905720967240122, "loss": 0.0752, "step": 8321 }, { "epoch": 3.916235294117647, "grad_norm": 0.7130864830657414, "learning_rate": 0.00010901503980649017, "loss": 0.3633, "step": 8322 }, { "epoch": 3.9167058823529413, "grad_norm": 0.3993986373254025, "learning_rate": 0.0001089728734408957, "loss": 0.2539, "step": 8323 }, { "epoch": 3.9171764705882355, "grad_norm": 0.6165926620210285, "learning_rate": 0.00010893071057921911, "loss": 0.2637, "step": 8324 }, { "epoch": 3.9176470588235293, "grad_norm": 0.3673485672059885, "learning_rate": 0.0001088885512250613, "loss": 0.167, "step": 8325 }, { "epoch": 3.9181176470588235, "grad_norm": 0.8420339629422084, "learning_rate": 0.00010884639538202285, "loss": 0.3965, "step": 8326 }, { "epoch": 3.9185882352941177, "grad_norm": 0.7163209721406, "learning_rate": 0.00010880424305370406, "loss": 0.4629, "step": 8327 }, { "epoch": 3.919058823529412, "grad_norm": 0.19106656275362588, "learning_rate": 0.00010876209424370502, "loss": 0.0476, "step": 8328 }, { "epoch": 3.9195294117647057, "grad_norm": 0.2733726515677985, "learning_rate": 0.00010871994895562538, "loss": 0.127, "step": 8329 }, { "epoch": 3.92, "grad_norm": 0.8025783852184303, "learning_rate": 0.00010867780719306458, "loss": 0.3281, "step": 8330 }, { "epoch": 3.920470588235294, "grad_norm": 0.6932266574920652, "learning_rate": 0.00010863566895962166, "loss": 0.4414, "step": 8331 }, { "epoch": 3.9209411764705884, "grad_norm": 0.221259091655241, "learning_rate": 0.0001085935342588956, "loss": 0.0688, "step": 8332 }, { "epoch": 3.9214117647058826, "grad_norm": 0.37368455775180065, "learning_rate": 0.00010855140309448479, "loss": 0.1338, "step": 8333 }, { "epoch": 3.9218823529411764, "grad_norm": 0.753213342393004, "learning_rate": 0.00010850927546998744, "loss": 0.4023, "step": 8334 }, { "epoch": 3.9223529411764706, "grad_norm": 0.5712713444902614, "learning_rate": 0.00010846715138900148, "loss": 0.3438, "step": 8335 }, { "epoch": 3.922823529411765, "grad_norm": 0.36892720605321305, "learning_rate": 0.00010842503085512453, "loss": 0.1484, "step": 8336 }, { "epoch": 3.9232941176470586, "grad_norm": 0.5733143169637178, "learning_rate": 0.00010838291387195387, "loss": 0.2773, "step": 8337 }, { "epoch": 3.923764705882353, "grad_norm": 0.5171826560379047, "learning_rate": 0.00010834080044308646, "loss": 0.2031, "step": 8338 }, { "epoch": 3.924235294117647, "grad_norm": 0.4634601943350437, "learning_rate": 0.00010829869057211906, "loss": 0.2363, "step": 8339 }, { "epoch": 3.9247058823529413, "grad_norm": 0.5643383288814515, "learning_rate": 0.00010825658426264803, "loss": 0.3047, "step": 8340 }, { "epoch": 3.9251764705882355, "grad_norm": 0.7057337867660191, "learning_rate": 0.00010821448151826941, "loss": 0.2578, "step": 8341 }, { "epoch": 3.9256470588235293, "grad_norm": 0.5575026602171396, "learning_rate": 0.00010817238234257907, "loss": 0.377, "step": 8342 }, { "epoch": 3.9261176470588235, "grad_norm": 0.5032767030826881, "learning_rate": 0.00010813028673917245, "loss": 0.2598, "step": 8343 }, { "epoch": 3.9265882352941177, "grad_norm": 0.781254494904686, "learning_rate": 0.0001080881947116447, "loss": 0.5508, "step": 8344 }, { "epoch": 3.9270588235294115, "grad_norm": 0.5594856658651971, "learning_rate": 0.00010804610626359068, "loss": 0.3281, "step": 8345 }, { "epoch": 3.9275294117647057, "grad_norm": 0.6394534957051368, "learning_rate": 0.000108004021398605, "loss": 0.2832, "step": 8346 }, { "epoch": 3.928, "grad_norm": 0.46821094795650703, "learning_rate": 0.00010796194012028185, "loss": 0.2773, "step": 8347 }, { "epoch": 3.928470588235294, "grad_norm": 0.5869027481086905, "learning_rate": 0.00010791986243221521, "loss": 0.1338, "step": 8348 }, { "epoch": 3.9289411764705884, "grad_norm": 0.5680596578292728, "learning_rate": 0.00010787778833799875, "loss": 0.2461, "step": 8349 }, { "epoch": 3.9294117647058826, "grad_norm": 0.7058067458041911, "learning_rate": 0.00010783571784122577, "loss": 0.3438, "step": 8350 }, { "epoch": 3.9298823529411764, "grad_norm": 0.5889304522164049, "learning_rate": 0.00010779365094548926, "loss": 0.2969, "step": 8351 }, { "epoch": 3.9303529411764706, "grad_norm": 0.3881597800625646, "learning_rate": 0.00010775158765438202, "loss": 0.1475, "step": 8352 }, { "epoch": 3.930823529411765, "grad_norm": 0.035640376901042374, "learning_rate": 0.0001077095279714964, "loss": 0.0084, "step": 8353 }, { "epoch": 3.9312941176470586, "grad_norm": 0.34478043311556594, "learning_rate": 0.00010766747190042452, "loss": 0.1338, "step": 8354 }, { "epoch": 3.931764705882353, "grad_norm": 0.3089402490092659, "learning_rate": 0.00010762541944475821, "loss": 0.1318, "step": 8355 }, { "epoch": 3.932235294117647, "grad_norm": 0.9572785131526458, "learning_rate": 0.0001075833706080889, "loss": 0.4395, "step": 8356 }, { "epoch": 3.9327058823529413, "grad_norm": 0.638141302834633, "learning_rate": 0.00010754132539400782, "loss": 0.3438, "step": 8357 }, { "epoch": 3.9331764705882355, "grad_norm": 0.616850102296335, "learning_rate": 0.00010749928380610577, "loss": 0.4648, "step": 8358 }, { "epoch": 3.9336470588235293, "grad_norm": 0.697909442346403, "learning_rate": 0.00010745724584797338, "loss": 0.4375, "step": 8359 }, { "epoch": 3.9341176470588235, "grad_norm": 0.5808282906211811, "learning_rate": 0.00010741521152320088, "loss": 0.2002, "step": 8360 }, { "epoch": 3.9345882352941177, "grad_norm": 0.3684230717245512, "learning_rate": 0.00010737318083537813, "loss": 0.1079, "step": 8361 }, { "epoch": 3.9350588235294115, "grad_norm": 0.3305344581201009, "learning_rate": 0.00010733115378809485, "loss": 0.1777, "step": 8362 }, { "epoch": 3.9355294117647057, "grad_norm": 0.3293010406943207, "learning_rate": 0.00010728913038494038, "loss": 0.1074, "step": 8363 }, { "epoch": 3.936, "grad_norm": 0.4802390328201492, "learning_rate": 0.00010724711062950358, "loss": 0.1387, "step": 8364 }, { "epoch": 3.936470588235294, "grad_norm": 0.7228997052260958, "learning_rate": 0.00010720509452537329, "loss": 0.3418, "step": 8365 }, { "epoch": 3.9369411764705884, "grad_norm": 0.540766100611858, "learning_rate": 0.00010716308207613784, "loss": 0.2461, "step": 8366 }, { "epoch": 3.9374117647058826, "grad_norm": 1.039984886277148, "learning_rate": 0.00010712107328538528, "loss": 0.4863, "step": 8367 }, { "epoch": 3.9378823529411764, "grad_norm": 0.5986876743324554, "learning_rate": 0.00010707906815670337, "loss": 0.4492, "step": 8368 }, { "epoch": 3.9383529411764706, "grad_norm": 0.4840868380822453, "learning_rate": 0.00010703706669367956, "loss": 0.1465, "step": 8369 }, { "epoch": 3.938823529411765, "grad_norm": 0.5535745321674901, "learning_rate": 0.00010699506889990102, "loss": 0.3066, "step": 8370 }, { "epoch": 3.9392941176470586, "grad_norm": 0.4770394715242795, "learning_rate": 0.00010695307477895449, "loss": 0.332, "step": 8371 }, { "epoch": 3.939764705882353, "grad_norm": 0.8476272451639213, "learning_rate": 0.00010691108433442656, "loss": 0.4629, "step": 8372 }, { "epoch": 3.940235294117647, "grad_norm": 0.5270275595459654, "learning_rate": 0.00010686909756990337, "loss": 0.3145, "step": 8373 }, { "epoch": 3.9407058823529413, "grad_norm": 1.1328470188592736, "learning_rate": 0.00010682711448897077, "loss": 0.5312, "step": 8374 }, { "epoch": 3.9411764705882355, "grad_norm": 0.833438143788593, "learning_rate": 0.00010678513509521435, "loss": 0.5898, "step": 8375 }, { "epoch": 3.9416470588235293, "grad_norm": 0.5162980548251415, "learning_rate": 0.00010674315939221937, "loss": 0.2256, "step": 8376 }, { "epoch": 3.9421176470588235, "grad_norm": 0.37468124744382114, "learning_rate": 0.00010670118738357078, "loss": 0.0845, "step": 8377 }, { "epoch": 3.9425882352941177, "grad_norm": 0.17275296898001988, "learning_rate": 0.00010665921907285317, "loss": 0.0598, "step": 8378 }, { "epoch": 3.9430588235294115, "grad_norm": 0.763886674650563, "learning_rate": 0.0001066172544636508, "loss": 0.4551, "step": 8379 }, { "epoch": 3.9435294117647057, "grad_norm": 0.7893384241960392, "learning_rate": 0.00010657529355954771, "loss": 0.5039, "step": 8380 }, { "epoch": 3.944, "grad_norm": 0.8702723523384299, "learning_rate": 0.00010653333636412758, "loss": 0.4746, "step": 8381 }, { "epoch": 3.944470588235294, "grad_norm": 0.5025895856693678, "learning_rate": 0.00010649138288097366, "loss": 0.1572, "step": 8382 }, { "epoch": 3.9449411764705884, "grad_norm": 0.26424304931234416, "learning_rate": 0.00010644943311366911, "loss": 0.0977, "step": 8383 }, { "epoch": 3.9454117647058826, "grad_norm": 0.46945968642817754, "learning_rate": 0.00010640748706579658, "loss": 0.1865, "step": 8384 }, { "epoch": 3.9458823529411764, "grad_norm": 0.5045888392144594, "learning_rate": 0.00010636554474093846, "loss": 0.2383, "step": 8385 }, { "epoch": 3.9463529411764706, "grad_norm": 0.8905289291056361, "learning_rate": 0.00010632360614267683, "loss": 0.4473, "step": 8386 }, { "epoch": 3.946823529411765, "grad_norm": 0.4347290211427915, "learning_rate": 0.0001062816712745935, "loss": 0.1934, "step": 8387 }, { "epoch": 3.9472941176470586, "grad_norm": 0.31010686562522766, "learning_rate": 0.00010623974014026986, "loss": 0.0903, "step": 8388 }, { "epoch": 3.947764705882353, "grad_norm": 0.4387041214571648, "learning_rate": 0.00010619781274328703, "loss": 0.1689, "step": 8389 }, { "epoch": 3.948235294117647, "grad_norm": 0.4251514681944473, "learning_rate": 0.00010615588908722588, "loss": 0.1387, "step": 8390 }, { "epoch": 3.9487058823529413, "grad_norm": 0.398221530175294, "learning_rate": 0.00010611396917566685, "loss": 0.1885, "step": 8391 }, { "epoch": 3.9491764705882355, "grad_norm": 0.31087396320545424, "learning_rate": 0.00010607205301219008, "loss": 0.0869, "step": 8392 }, { "epoch": 3.9496470588235293, "grad_norm": 0.9061597157716541, "learning_rate": 0.00010603014060037549, "loss": 0.582, "step": 8393 }, { "epoch": 3.9501176470588235, "grad_norm": 0.6843251297012448, "learning_rate": 0.00010598823194380256, "loss": 0.3281, "step": 8394 }, { "epoch": 3.9505882352941177, "grad_norm": 0.8137692942377579, "learning_rate": 0.0001059463270460505, "loss": 0.6094, "step": 8395 }, { "epoch": 3.9510588235294115, "grad_norm": 0.4620003225322111, "learning_rate": 0.00010590442591069816, "loss": 0.3203, "step": 8396 }, { "epoch": 3.9515294117647057, "grad_norm": 0.2373067938036684, "learning_rate": 0.00010586252854132415, "loss": 0.0767, "step": 8397 }, { "epoch": 3.952, "grad_norm": 0.6624027002826995, "learning_rate": 0.0001058206349415067, "loss": 0.293, "step": 8398 }, { "epoch": 3.952470588235294, "grad_norm": 0.38814978745500056, "learning_rate": 0.00010577874511482371, "loss": 0.1953, "step": 8399 }, { "epoch": 3.9529411764705884, "grad_norm": 0.4097730647672444, "learning_rate": 0.00010573685906485282, "loss": 0.2285, "step": 8400 }, { "epoch": 3.9534117647058826, "grad_norm": 0.4916297014355408, "learning_rate": 0.00010569497679517126, "loss": 0.3027, "step": 8401 }, { "epoch": 3.9538823529411764, "grad_norm": 0.6054777031082232, "learning_rate": 0.00010565309830935599, "loss": 0.252, "step": 8402 }, { "epoch": 3.9543529411764706, "grad_norm": 0.551340806579296, "learning_rate": 0.00010561122361098365, "loss": 0.2676, "step": 8403 }, { "epoch": 3.954823529411765, "grad_norm": 0.5009291911761905, "learning_rate": 0.00010556935270363055, "loss": 0.1309, "step": 8404 }, { "epoch": 3.9552941176470586, "grad_norm": 0.4476499445764742, "learning_rate": 0.00010552748559087261, "loss": 0.1406, "step": 8405 }, { "epoch": 3.955764705882353, "grad_norm": 0.7009477565737562, "learning_rate": 0.00010548562227628559, "loss": 0.3281, "step": 8406 }, { "epoch": 3.956235294117647, "grad_norm": 0.38805653113429794, "learning_rate": 0.00010544376276344477, "loss": 0.2812, "step": 8407 }, { "epoch": 3.9567058823529413, "grad_norm": 0.5429289507164932, "learning_rate": 0.00010540190705592516, "loss": 0.127, "step": 8408 }, { "epoch": 3.9571764705882355, "grad_norm": 0.6371193515413084, "learning_rate": 0.00010536005515730141, "loss": 0.3828, "step": 8409 }, { "epoch": 3.9576470588235293, "grad_norm": 0.37588058147037495, "learning_rate": 0.00010531820707114793, "loss": 0.1133, "step": 8410 }, { "epoch": 3.9581176470588235, "grad_norm": 0.26631044340979537, "learning_rate": 0.00010527636280103872, "loss": 0.1045, "step": 8411 }, { "epoch": 3.9585882352941177, "grad_norm": 0.696378138735532, "learning_rate": 0.00010523452235054746, "loss": 0.3906, "step": 8412 }, { "epoch": 3.9590588235294115, "grad_norm": 0.22516707077701448, "learning_rate": 0.00010519268572324763, "loss": 0.0549, "step": 8413 }, { "epoch": 3.9595294117647057, "grad_norm": 0.7607738680591948, "learning_rate": 0.00010515085292271219, "loss": 0.3672, "step": 8414 }, { "epoch": 3.96, "grad_norm": 0.6091939641881985, "learning_rate": 0.00010510902395251387, "loss": 0.2852, "step": 8415 }, { "epoch": 3.960470588235294, "grad_norm": 0.923603751711177, "learning_rate": 0.00010506719881622514, "loss": 0.6289, "step": 8416 }, { "epoch": 3.9609411764705884, "grad_norm": 0.7937972242435195, "learning_rate": 0.00010502537751741801, "loss": 0.2559, "step": 8417 }, { "epoch": 3.961411764705882, "grad_norm": 0.4013029077788718, "learning_rate": 0.00010498356005966422, "loss": 0.1689, "step": 8418 }, { "epoch": 3.9618823529411764, "grad_norm": 0.558128116847131, "learning_rate": 0.00010494174644653518, "loss": 0.2578, "step": 8419 }, { "epoch": 3.9623529411764706, "grad_norm": 0.7545760008363006, "learning_rate": 0.000104899936681602, "loss": 0.3242, "step": 8420 }, { "epoch": 3.962823529411765, "grad_norm": 0.7303401202404963, "learning_rate": 0.0001048581307684355, "loss": 0.3828, "step": 8421 }, { "epoch": 3.9632941176470586, "grad_norm": 0.4114731217133911, "learning_rate": 0.00010481632871060605, "loss": 0.1367, "step": 8422 }, { "epoch": 3.963764705882353, "grad_norm": 0.3988856751763344, "learning_rate": 0.0001047745305116837, "loss": 0.1523, "step": 8423 }, { "epoch": 3.964235294117647, "grad_norm": 0.5524992978353944, "learning_rate": 0.00010473273617523831, "loss": 0.2695, "step": 8424 }, { "epoch": 3.9647058823529413, "grad_norm": 0.29055444454864765, "learning_rate": 0.00010469094570483928, "loss": 0.064, "step": 8425 }, { "epoch": 3.9651764705882355, "grad_norm": 0.30047429200107906, "learning_rate": 0.00010464915910405576, "loss": 0.0981, "step": 8426 }, { "epoch": 3.9656470588235293, "grad_norm": 0.6439424317055209, "learning_rate": 0.00010460737637645645, "loss": 0.4668, "step": 8427 }, { "epoch": 3.9661176470588235, "grad_norm": 0.41551991358119816, "learning_rate": 0.00010456559752560987, "loss": 0.1387, "step": 8428 }, { "epoch": 3.9665882352941177, "grad_norm": 0.455358205108404, "learning_rate": 0.00010452382255508412, "loss": 0.2617, "step": 8429 }, { "epoch": 3.9670588235294115, "grad_norm": 0.2829265084806012, "learning_rate": 0.00010448205146844696, "loss": 0.0854, "step": 8430 }, { "epoch": 3.9675294117647057, "grad_norm": 0.6184667752429527, "learning_rate": 0.00010444028426926588, "loss": 0.334, "step": 8431 }, { "epoch": 3.968, "grad_norm": 0.40324878089979654, "learning_rate": 0.000104398520961108, "loss": 0.0762, "step": 8432 }, { "epoch": 3.968470588235294, "grad_norm": 1.005880081273249, "learning_rate": 0.00010435676154754009, "loss": 0.2334, "step": 8433 }, { "epoch": 3.9689411764705884, "grad_norm": 0.40538083602983116, "learning_rate": 0.00010431500603212863, "loss": 0.1367, "step": 8434 }, { "epoch": 3.969411764705882, "grad_norm": 0.4369040350620584, "learning_rate": 0.00010427325441843976, "loss": 0.1689, "step": 8435 }, { "epoch": 3.9698823529411764, "grad_norm": 0.5552598101871299, "learning_rate": 0.00010423150671003924, "loss": 0.373, "step": 8436 }, { "epoch": 3.9703529411764706, "grad_norm": 0.5623667782228322, "learning_rate": 0.00010418976291049251, "loss": 0.3086, "step": 8437 }, { "epoch": 3.970823529411765, "grad_norm": 0.41316828869643757, "learning_rate": 0.00010414802302336474, "loss": 0.1748, "step": 8438 }, { "epoch": 3.9712941176470586, "grad_norm": 0.539298951635976, "learning_rate": 0.00010410628705222073, "loss": 0.3066, "step": 8439 }, { "epoch": 3.971764705882353, "grad_norm": 0.381660598984594, "learning_rate": 0.00010406455500062487, "loss": 0.1064, "step": 8440 }, { "epoch": 3.972235294117647, "grad_norm": 0.7001156851560476, "learning_rate": 0.00010402282687214137, "loss": 0.2061, "step": 8441 }, { "epoch": 3.9727058823529413, "grad_norm": 0.47252037383082895, "learning_rate": 0.00010398110267033396, "loss": 0.2256, "step": 8442 }, { "epoch": 3.9731764705882355, "grad_norm": 0.3846139721201833, "learning_rate": 0.00010393938239876608, "loss": 0.1709, "step": 8443 }, { "epoch": 3.9736470588235293, "grad_norm": 0.5241649259805432, "learning_rate": 0.0001038976660610009, "loss": 0.2812, "step": 8444 }, { "epoch": 3.9741176470588235, "grad_norm": 0.6664149521874447, "learning_rate": 0.00010385595366060116, "loss": 0.3906, "step": 8445 }, { "epoch": 3.9745882352941178, "grad_norm": 0.6448957456706181, "learning_rate": 0.0001038142452011293, "loss": 0.3828, "step": 8446 }, { "epoch": 3.9750588235294115, "grad_norm": 0.6039077927205125, "learning_rate": 0.00010377254068614743, "loss": 0.4395, "step": 8447 }, { "epoch": 3.9755294117647058, "grad_norm": 0.823239871792968, "learning_rate": 0.00010373084011921736, "loss": 0.5664, "step": 8448 }, { "epoch": 3.976, "grad_norm": 0.41466933612160123, "learning_rate": 0.00010368914350390048, "loss": 0.1084, "step": 8449 }, { "epoch": 3.976470588235294, "grad_norm": 0.47170554716860225, "learning_rate": 0.0001036474508437579, "loss": 0.2451, "step": 8450 }, { "epoch": 3.9769411764705884, "grad_norm": 0.5905256650275845, "learning_rate": 0.00010360576214235036, "loss": 0.2363, "step": 8451 }, { "epoch": 3.977411764705882, "grad_norm": 0.4815937773986685, "learning_rate": 0.00010356407740323832, "loss": 0.1836, "step": 8452 }, { "epoch": 3.9778823529411764, "grad_norm": 0.9985437805363883, "learning_rate": 0.00010352239662998182, "loss": 0.5586, "step": 8453 }, { "epoch": 3.9783529411764706, "grad_norm": 0.5339917127090094, "learning_rate": 0.00010348071982614063, "loss": 0.2832, "step": 8454 }, { "epoch": 3.978823529411765, "grad_norm": 0.45029303039841406, "learning_rate": 0.00010343904699527415, "loss": 0.1357, "step": 8455 }, { "epoch": 3.9792941176470586, "grad_norm": 0.08818556508665679, "learning_rate": 0.00010339737814094144, "loss": 0.0111, "step": 8456 }, { "epoch": 3.979764705882353, "grad_norm": 0.8483325269191758, "learning_rate": 0.00010335571326670117, "loss": 0.5625, "step": 8457 }, { "epoch": 3.980235294117647, "grad_norm": 0.6494056753162353, "learning_rate": 0.00010331405237611182, "loss": 0.2441, "step": 8458 }, { "epoch": 3.9807058823529413, "grad_norm": 0.49112709472230476, "learning_rate": 0.00010327239547273138, "loss": 0.1348, "step": 8459 }, { "epoch": 3.9811764705882355, "grad_norm": 0.5949068051890449, "learning_rate": 0.00010323074256011756, "loss": 0.3281, "step": 8460 }, { "epoch": 3.9816470588235293, "grad_norm": 0.38390905697416633, "learning_rate": 0.00010318909364182774, "loss": 0.2266, "step": 8461 }, { "epoch": 3.9821176470588235, "grad_norm": 0.1506739183954814, "learning_rate": 0.00010314744872141893, "loss": 0.0432, "step": 8462 }, { "epoch": 3.9825882352941178, "grad_norm": 0.6217263166140197, "learning_rate": 0.00010310580780244779, "loss": 0.2539, "step": 8463 }, { "epoch": 3.9830588235294115, "grad_norm": 0.36369281623199534, "learning_rate": 0.00010306417088847067, "loss": 0.0791, "step": 8464 }, { "epoch": 3.9835294117647058, "grad_norm": 0.49206318364436163, "learning_rate": 0.00010302253798304356, "loss": 0.3203, "step": 8465 }, { "epoch": 3.984, "grad_norm": 0.49879396974755014, "learning_rate": 0.00010298090908972217, "loss": 0.2422, "step": 8466 }, { "epoch": 3.984470588235294, "grad_norm": 0.3762279129459424, "learning_rate": 0.00010293928421206179, "loss": 0.2021, "step": 8467 }, { "epoch": 3.9849411764705884, "grad_norm": 0.7421831201838742, "learning_rate": 0.00010289766335361733, "loss": 0.4531, "step": 8468 }, { "epoch": 3.985411764705882, "grad_norm": 0.43758173502251346, "learning_rate": 0.00010285604651794347, "loss": 0.1816, "step": 8469 }, { "epoch": 3.9858823529411764, "grad_norm": 0.372558976003591, "learning_rate": 0.00010281443370859449, "loss": 0.207, "step": 8470 }, { "epoch": 3.9863529411764707, "grad_norm": 0.4102329902928578, "learning_rate": 0.00010277282492912429, "loss": 0.126, "step": 8471 }, { "epoch": 3.9868235294117644, "grad_norm": 0.7266556747241938, "learning_rate": 0.0001027312201830865, "loss": 0.4199, "step": 8472 }, { "epoch": 3.9872941176470587, "grad_norm": 0.41814685563163084, "learning_rate": 0.00010268961947403437, "loss": 0.2051, "step": 8473 }, { "epoch": 3.987764705882353, "grad_norm": 0.5056753513163323, "learning_rate": 0.00010264802280552077, "loss": 0.2832, "step": 8474 }, { "epoch": 3.988235294117647, "grad_norm": 0.27429557416742, "learning_rate": 0.0001026064301810983, "loss": 0.0938, "step": 8475 }, { "epoch": 3.9887058823529413, "grad_norm": 0.19232032509820293, "learning_rate": 0.00010256484160431917, "loss": 0.0591, "step": 8476 }, { "epoch": 3.9891764705882355, "grad_norm": 0.6122275841976514, "learning_rate": 0.00010252325707873523, "loss": 0.3105, "step": 8477 }, { "epoch": 3.9896470588235293, "grad_norm": 0.30149330922166667, "learning_rate": 0.00010248167660789797, "loss": 0.1167, "step": 8478 }, { "epoch": 3.9901176470588235, "grad_norm": 0.8276212522509188, "learning_rate": 0.00010244010019535864, "loss": 0.293, "step": 8479 }, { "epoch": 3.9905882352941178, "grad_norm": 0.5269375788427315, "learning_rate": 0.00010239852784466803, "loss": 0.2363, "step": 8480 }, { "epoch": 3.9910588235294115, "grad_norm": 0.3393364191565182, "learning_rate": 0.00010235695955937661, "loss": 0.0884, "step": 8481 }, { "epoch": 3.9915294117647058, "grad_norm": 0.38937826890934574, "learning_rate": 0.00010231539534303454, "loss": 0.1455, "step": 8482 }, { "epoch": 3.992, "grad_norm": 0.343197390110164, "learning_rate": 0.00010227383519919161, "loss": 0.1865, "step": 8483 }, { "epoch": 3.992470588235294, "grad_norm": 0.4540212517110544, "learning_rate": 0.00010223227913139722, "loss": 0.1426, "step": 8484 }, { "epoch": 3.9929411764705884, "grad_norm": 0.7621253044598979, "learning_rate": 0.0001021907271432005, "loss": 0.3145, "step": 8485 }, { "epoch": 3.993411764705882, "grad_norm": 0.8496417933275886, "learning_rate": 0.00010214917923815022, "loss": 0.2471, "step": 8486 }, { "epoch": 3.9938823529411764, "grad_norm": 0.5322845758046781, "learning_rate": 0.00010210763541979473, "loss": 0.1572, "step": 8487 }, { "epoch": 3.9943529411764707, "grad_norm": 0.49056942403304227, "learning_rate": 0.00010206609569168204, "loss": 0.1885, "step": 8488 }, { "epoch": 3.9948235294117644, "grad_norm": 0.3135612842950513, "learning_rate": 0.00010202456005735993, "loss": 0.1465, "step": 8489 }, { "epoch": 3.9952941176470587, "grad_norm": 0.31511281296105703, "learning_rate": 0.00010198302852037573, "loss": 0.1021, "step": 8490 }, { "epoch": 3.995764705882353, "grad_norm": 1.0270961809996495, "learning_rate": 0.00010194150108427637, "loss": 0.6211, "step": 8491 }, { "epoch": 3.996235294117647, "grad_norm": 0.5828325261414204, "learning_rate": 0.00010189997775260857, "loss": 0.332, "step": 8492 }, { "epoch": 3.9967058823529413, "grad_norm": 0.4806985899974469, "learning_rate": 0.00010185845852891858, "loss": 0.2178, "step": 8493 }, { "epoch": 3.9971764705882356, "grad_norm": 0.208193081731885, "learning_rate": 0.00010181694341675235, "loss": 0.0806, "step": 8494 }, { "epoch": 3.9976470588235293, "grad_norm": 0.10981169089862512, "learning_rate": 0.00010177543241965555, "loss": 0.0452, "step": 8495 }, { "epoch": 3.9981176470588236, "grad_norm": 0.619702418812001, "learning_rate": 0.00010173392554117333, "loss": 0.3887, "step": 8496 }, { "epoch": 3.9985882352941178, "grad_norm": 0.49983466365169615, "learning_rate": 0.00010169242278485063, "loss": 0.3398, "step": 8497 }, { "epoch": 3.9990588235294116, "grad_norm": 0.6688268766428671, "learning_rate": 0.00010165092415423193, "loss": 0.3457, "step": 8498 }, { "epoch": 3.9995294117647058, "grad_norm": 0.4752646936410088, "learning_rate": 0.00010160942965286148, "loss": 0.2139, "step": 8499 }, { "epoch": 4.0, "grad_norm": 0.3472190282646978, "learning_rate": 0.0001015679392842831, "loss": 0.1396, "step": 8500 }, { "epoch": 4.0, "eval_loss": 0.2586979269981384, "eval_runtime": 62.3417, "eval_samples_per_second": 12.03, "eval_steps_per_second": 6.015, "step": 8500 }, { "epoch": 4.000470588235294, "grad_norm": 0.4812893118530729, "learning_rate": 0.00010152645305204022, "loss": 0.3398, "step": 8501 }, { "epoch": 4.0009411764705884, "grad_norm": 1.0574858932950624, "learning_rate": 0.00010148497095967606, "loss": 0.543, "step": 8502 }, { "epoch": 4.001411764705883, "grad_norm": 1.2028217427370345, "learning_rate": 0.00010144349301073334, "loss": 1.3281, "step": 8503 }, { "epoch": 4.001882352941177, "grad_norm": 0.45899484978015764, "learning_rate": 0.00010140201920875445, "loss": 0.3535, "step": 8504 }, { "epoch": 4.00235294117647, "grad_norm": 0.3570849182883533, "learning_rate": 0.00010136054955728154, "loss": 0.1406, "step": 8505 }, { "epoch": 4.0028235294117644, "grad_norm": 0.8829921467813989, "learning_rate": 0.00010131908405985625, "loss": 0.4336, "step": 8506 }, { "epoch": 4.003294117647059, "grad_norm": 0.4645765265122216, "learning_rate": 0.00010127762272001996, "loss": 0.2344, "step": 8507 }, { "epoch": 4.003764705882353, "grad_norm": 0.06392960517235567, "learning_rate": 0.00010123616554131363, "loss": 0.0093, "step": 8508 }, { "epoch": 4.004235294117647, "grad_norm": 0.6957731331083626, "learning_rate": 0.00010119471252727796, "loss": 0.3809, "step": 8509 }, { "epoch": 4.004705882352941, "grad_norm": 0.4664501029164791, "learning_rate": 0.00010115326368145327, "loss": 0.1523, "step": 8510 }, { "epoch": 4.005176470588236, "grad_norm": 0.3055781322809599, "learning_rate": 0.00010111181900737945, "loss": 0.0767, "step": 8511 }, { "epoch": 4.00564705882353, "grad_norm": 0.6632544677917634, "learning_rate": 0.00010107037850859605, "loss": 0.2656, "step": 8512 }, { "epoch": 4.006117647058823, "grad_norm": 0.587103584422072, "learning_rate": 0.00010102894218864235, "loss": 0.2988, "step": 8513 }, { "epoch": 4.006588235294117, "grad_norm": 0.6242125971817478, "learning_rate": 0.00010098751005105719, "loss": 0.3125, "step": 8514 }, { "epoch": 4.007058823529412, "grad_norm": 0.38000301956351645, "learning_rate": 0.0001009460820993791, "loss": 0.1953, "step": 8515 }, { "epoch": 4.007529411764706, "grad_norm": 0.245464035942448, "learning_rate": 0.00010090465833714616, "loss": 0.0811, "step": 8516 }, { "epoch": 4.008, "grad_norm": 0.31414897706777295, "learning_rate": 0.00010086323876789621, "loss": 0.1475, "step": 8517 }, { "epoch": 4.008470588235294, "grad_norm": 0.3607713554164851, "learning_rate": 0.00010082182339516674, "loss": 0.2109, "step": 8518 }, { "epoch": 4.0089411764705885, "grad_norm": 0.5932740265937112, "learning_rate": 0.00010078041222249469, "loss": 0.3555, "step": 8519 }, { "epoch": 4.009411764705883, "grad_norm": 0.33633203758032715, "learning_rate": 0.00010073900525341692, "loss": 0.0781, "step": 8520 }, { "epoch": 4.009882352941177, "grad_norm": 0.23516633191749323, "learning_rate": 0.00010069760249146972, "loss": 0.0762, "step": 8521 }, { "epoch": 4.01035294117647, "grad_norm": 1.0434790753824472, "learning_rate": 0.00010065620394018904, "loss": 0.3516, "step": 8522 }, { "epoch": 4.0108235294117645, "grad_norm": 0.45150409826804894, "learning_rate": 0.00010061480960311061, "loss": 0.2988, "step": 8523 }, { "epoch": 4.011294117647059, "grad_norm": 0.6825216561574903, "learning_rate": 0.0001005734194837697, "loss": 0.3184, "step": 8524 }, { "epoch": 4.011764705882353, "grad_norm": 0.3716420309136899, "learning_rate": 0.00010053203358570115, "loss": 0.1279, "step": 8525 }, { "epoch": 4.012235294117647, "grad_norm": 0.6584068572981635, "learning_rate": 0.00010049065191243961, "loss": 0.3359, "step": 8526 }, { "epoch": 4.012705882352941, "grad_norm": 0.6385589385322971, "learning_rate": 0.00010044927446751923, "loss": 0.332, "step": 8527 }, { "epoch": 4.013176470588236, "grad_norm": 0.8224317002634555, "learning_rate": 0.00010040790125447386, "loss": 0.5703, "step": 8528 }, { "epoch": 4.01364705882353, "grad_norm": 0.5731534259487691, "learning_rate": 0.00010036653227683694, "loss": 0.2559, "step": 8529 }, { "epoch": 4.014117647058823, "grad_norm": 0.3135169043497091, "learning_rate": 0.00010032516753814166, "loss": 0.0981, "step": 8530 }, { "epoch": 4.014588235294117, "grad_norm": 0.7707659704611609, "learning_rate": 0.0001002838070419207, "loss": 0.2393, "step": 8531 }, { "epoch": 4.015058823529412, "grad_norm": 0.6878757637483525, "learning_rate": 0.00010024245079170645, "loss": 0.3965, "step": 8532 }, { "epoch": 4.015529411764706, "grad_norm": 0.32890795011353213, "learning_rate": 0.000100201098791031, "loss": 0.0889, "step": 8533 }, { "epoch": 4.016, "grad_norm": 0.3409653296683476, "learning_rate": 0.00010015975104342598, "loss": 0.1904, "step": 8534 }, { "epoch": 4.016470588235294, "grad_norm": 0.8472242450993195, "learning_rate": 0.00010011840755242266, "loss": 0.4414, "step": 8535 }, { "epoch": 4.0169411764705885, "grad_norm": 0.819286300469196, "learning_rate": 0.00010007706832155201, "loss": 0.332, "step": 8536 }, { "epoch": 4.017411764705883, "grad_norm": 0.795474001058558, "learning_rate": 0.00010003573335434462, "loss": 0.4902, "step": 8537 }, { "epoch": 4.017882352941177, "grad_norm": 0.4832604620314537, "learning_rate": 9.999440265433065e-05, "loss": 0.2598, "step": 8538 }, { "epoch": 4.01835294117647, "grad_norm": 0.5183111555492099, "learning_rate": 9.995307622503993e-05, "loss": 0.3242, "step": 8539 }, { "epoch": 4.0188235294117645, "grad_norm": 0.4826378050627949, "learning_rate": 9.991175407000206e-05, "loss": 0.2344, "step": 8540 }, { "epoch": 4.019294117647059, "grad_norm": 0.8529139877637107, "learning_rate": 9.987043619274604e-05, "loss": 0.4043, "step": 8541 }, { "epoch": 4.019764705882353, "grad_norm": 0.4094956766834612, "learning_rate": 9.982912259680063e-05, "loss": 0.1992, "step": 8542 }, { "epoch": 4.020235294117647, "grad_norm": 0.5251207709221359, "learning_rate": 9.978781328569427e-05, "loss": 0.2578, "step": 8543 }, { "epoch": 4.020705882352941, "grad_norm": 0.8307057918848515, "learning_rate": 9.974650826295493e-05, "loss": 0.6758, "step": 8544 }, { "epoch": 4.021176470588236, "grad_norm": 0.628775831876864, "learning_rate": 9.970520753211027e-05, "loss": 0.418, "step": 8545 }, { "epoch": 4.02164705882353, "grad_norm": 0.4824240176280166, "learning_rate": 9.966391109668758e-05, "loss": 0.2422, "step": 8546 }, { "epoch": 4.022117647058823, "grad_norm": 0.5480273362104535, "learning_rate": 9.96226189602138e-05, "loss": 0.4355, "step": 8547 }, { "epoch": 4.022588235294117, "grad_norm": 0.16965989151147015, "learning_rate": 9.958133112621547e-05, "loss": 0.025, "step": 8548 }, { "epoch": 4.023058823529412, "grad_norm": 0.2789984435289752, "learning_rate": 9.954004759821871e-05, "loss": 0.1094, "step": 8549 }, { "epoch": 4.023529411764706, "grad_norm": 0.6272360988576221, "learning_rate": 9.949876837974944e-05, "loss": 0.3945, "step": 8550 }, { "epoch": 4.024, "grad_norm": 0.8099401073990854, "learning_rate": 9.945749347433304e-05, "loss": 0.3496, "step": 8551 }, { "epoch": 4.024470588235294, "grad_norm": 0.40130302616597513, "learning_rate": 9.941622288549458e-05, "loss": 0.2266, "step": 8552 }, { "epoch": 4.0249411764705885, "grad_norm": 0.786460865688374, "learning_rate": 9.937495661675878e-05, "loss": 0.3438, "step": 8553 }, { "epoch": 4.025411764705883, "grad_norm": 0.32720315017999496, "learning_rate": 9.933369467165004e-05, "loss": 0.2002, "step": 8554 }, { "epoch": 4.025882352941177, "grad_norm": 0.7277096530212983, "learning_rate": 9.929243705369232e-05, "loss": 0.4258, "step": 8555 }, { "epoch": 4.02635294117647, "grad_norm": 0.7717459692259961, "learning_rate": 9.925118376640918e-05, "loss": 0.2715, "step": 8556 }, { "epoch": 4.0268235294117645, "grad_norm": 0.7210246583183529, "learning_rate": 9.920993481332383e-05, "loss": 0.3184, "step": 8557 }, { "epoch": 4.027294117647059, "grad_norm": 0.4072540172476609, "learning_rate": 9.916869019795922e-05, "loss": 0.1533, "step": 8558 }, { "epoch": 4.027764705882353, "grad_norm": 0.48332117134948843, "learning_rate": 9.912744992383777e-05, "loss": 0.2256, "step": 8559 }, { "epoch": 4.028235294117647, "grad_norm": 0.5125578058824768, "learning_rate": 9.90862139944816e-05, "loss": 0.2637, "step": 8560 }, { "epoch": 4.028705882352941, "grad_norm": 0.5865018194083402, "learning_rate": 9.904498241341252e-05, "loss": 0.3984, "step": 8561 }, { "epoch": 4.029176470588236, "grad_norm": 0.3874812329946635, "learning_rate": 9.900375518415188e-05, "loss": 0.1846, "step": 8562 }, { "epoch": 4.02964705882353, "grad_norm": 0.1871154734206204, "learning_rate": 9.896253231022066e-05, "loss": 0.0476, "step": 8563 }, { "epoch": 4.030117647058823, "grad_norm": 0.6144723144157699, "learning_rate": 9.892131379513953e-05, "loss": 0.3984, "step": 8564 }, { "epoch": 4.030588235294117, "grad_norm": 0.6230558738943991, "learning_rate": 9.888009964242876e-05, "loss": 0.3613, "step": 8565 }, { "epoch": 4.031058823529412, "grad_norm": 0.4656163947434519, "learning_rate": 9.883888985560822e-05, "loss": 0.1768, "step": 8566 }, { "epoch": 4.031529411764706, "grad_norm": 0.3019010994742417, "learning_rate": 9.879768443819737e-05, "loss": 0.0796, "step": 8567 }, { "epoch": 4.032, "grad_norm": 0.5742691032595975, "learning_rate": 9.875648339371546e-05, "loss": 0.2969, "step": 8568 }, { "epoch": 4.032470588235294, "grad_norm": 0.5580649328937305, "learning_rate": 9.871528672568124e-05, "loss": 0.3145, "step": 8569 }, { "epoch": 4.0329411764705885, "grad_norm": 0.5643993330511131, "learning_rate": 9.867409443761304e-05, "loss": 0.4102, "step": 8570 }, { "epoch": 4.033411764705883, "grad_norm": 0.6303812934228443, "learning_rate": 9.863290653302894e-05, "loss": 0.4141, "step": 8571 }, { "epoch": 4.033882352941176, "grad_norm": 0.14089723035240245, "learning_rate": 9.859172301544659e-05, "loss": 0.0337, "step": 8572 }, { "epoch": 4.03435294117647, "grad_norm": 0.5206799569457073, "learning_rate": 9.855054388838322e-05, "loss": 0.3184, "step": 8573 }, { "epoch": 4.0348235294117645, "grad_norm": 0.3231293130645712, "learning_rate": 9.85093691553558e-05, "loss": 0.1206, "step": 8574 }, { "epoch": 4.035294117647059, "grad_norm": 0.4694544779881911, "learning_rate": 9.84681988198808e-05, "loss": 0.21, "step": 8575 }, { "epoch": 4.035764705882353, "grad_norm": 0.4187514261101952, "learning_rate": 9.842703288547438e-05, "loss": 0.2119, "step": 8576 }, { "epoch": 4.036235294117647, "grad_norm": 0.38409611633320173, "learning_rate": 9.838587135565229e-05, "loss": 0.1367, "step": 8577 }, { "epoch": 4.036705882352941, "grad_norm": 0.7029241193389008, "learning_rate": 9.834471423393e-05, "loss": 0.4219, "step": 8578 }, { "epoch": 4.037176470588236, "grad_norm": 0.5382244385968717, "learning_rate": 9.830356152382245e-05, "loss": 0.2451, "step": 8579 }, { "epoch": 4.03764705882353, "grad_norm": 0.3944098153928631, "learning_rate": 9.826241322884432e-05, "loss": 0.1338, "step": 8580 }, { "epoch": 4.038117647058823, "grad_norm": 0.4009100192099433, "learning_rate": 9.822126935250987e-05, "loss": 0.1807, "step": 8581 }, { "epoch": 4.038588235294117, "grad_norm": 0.4711757282518455, "learning_rate": 9.818012989833301e-05, "loss": 0.208, "step": 8582 }, { "epoch": 4.039058823529412, "grad_norm": 0.6484830638773675, "learning_rate": 9.81389948698272e-05, "loss": 0.3105, "step": 8583 }, { "epoch": 4.039529411764706, "grad_norm": 1.0599901083361845, "learning_rate": 9.809786427050562e-05, "loss": 0.5898, "step": 8584 }, { "epoch": 4.04, "grad_norm": 0.5155680743851501, "learning_rate": 9.805673810388103e-05, "loss": 0.252, "step": 8585 }, { "epoch": 4.040470588235294, "grad_norm": 1.3615743524691355, "learning_rate": 9.801561637346575e-05, "loss": 0.5078, "step": 8586 }, { "epoch": 4.0409411764705885, "grad_norm": 0.3405620121550869, "learning_rate": 9.797449908277184e-05, "loss": 0.1875, "step": 8587 }, { "epoch": 4.041411764705883, "grad_norm": 0.7846664158990424, "learning_rate": 9.793338623531089e-05, "loss": 0.3496, "step": 8588 }, { "epoch": 4.041882352941176, "grad_norm": 0.8852235396548211, "learning_rate": 9.789227783459415e-05, "loss": 0.4082, "step": 8589 }, { "epoch": 4.04235294117647, "grad_norm": 0.38452242621598287, "learning_rate": 9.785117388413244e-05, "loss": 0.1113, "step": 8590 }, { "epoch": 4.0428235294117645, "grad_norm": 0.8478514616634786, "learning_rate": 9.781007438743629e-05, "loss": 0.5078, "step": 8591 }, { "epoch": 4.043294117647059, "grad_norm": 0.47651415513840717, "learning_rate": 9.776897934801577e-05, "loss": 0.2158, "step": 8592 }, { "epoch": 4.043764705882353, "grad_norm": 0.38936711963511145, "learning_rate": 9.77278887693806e-05, "loss": 0.1465, "step": 8593 }, { "epoch": 4.044235294117647, "grad_norm": 0.49075462192669467, "learning_rate": 9.768680265504013e-05, "loss": 0.209, "step": 8594 }, { "epoch": 4.044705882352941, "grad_norm": 0.4762658292793485, "learning_rate": 9.764572100850331e-05, "loss": 0.1279, "step": 8595 }, { "epoch": 4.045176470588236, "grad_norm": 0.608322348906237, "learning_rate": 9.760464383327868e-05, "loss": 0.2451, "step": 8596 }, { "epoch": 4.04564705882353, "grad_norm": 0.41025877318653586, "learning_rate": 9.756357113287449e-05, "loss": 0.1064, "step": 8597 }, { "epoch": 4.046117647058823, "grad_norm": 0.329132508851677, "learning_rate": 9.75225029107985e-05, "loss": 0.1328, "step": 8598 }, { "epoch": 4.046588235294117, "grad_norm": 0.5638951937452956, "learning_rate": 9.74814391705582e-05, "loss": 0.2812, "step": 8599 }, { "epoch": 4.047058823529412, "grad_norm": 0.38370765940954465, "learning_rate": 9.744037991566058e-05, "loss": 0.1689, "step": 8600 }, { "epoch": 4.047529411764706, "grad_norm": 0.9082002129929322, "learning_rate": 9.739932514961232e-05, "loss": 0.6289, "step": 8601 }, { "epoch": 4.048, "grad_norm": 0.5039621779805982, "learning_rate": 9.73582748759197e-05, "loss": 0.3105, "step": 8602 }, { "epoch": 4.048470588235294, "grad_norm": 0.27633202653524014, "learning_rate": 9.731722909808863e-05, "loss": 0.0864, "step": 8603 }, { "epoch": 4.0489411764705885, "grad_norm": 0.5249872381496962, "learning_rate": 9.727618781962457e-05, "loss": 0.2969, "step": 8604 }, { "epoch": 4.049411764705883, "grad_norm": 0.6703649333066787, "learning_rate": 9.723515104403268e-05, "loss": 0.3438, "step": 8605 }, { "epoch": 4.049882352941176, "grad_norm": 0.33975982499911067, "learning_rate": 9.719411877481776e-05, "loss": 0.1436, "step": 8606 }, { "epoch": 4.05035294117647, "grad_norm": 0.6169534657434976, "learning_rate": 9.715309101548406e-05, "loss": 0.3086, "step": 8607 }, { "epoch": 4.0508235294117645, "grad_norm": 0.352371786395396, "learning_rate": 9.711206776953559e-05, "loss": 0.1118, "step": 8608 }, { "epoch": 4.051294117647059, "grad_norm": 0.5490473651599408, "learning_rate": 9.707104904047596e-05, "loss": 0.3145, "step": 8609 }, { "epoch": 4.051764705882353, "grad_norm": 0.53422572915355, "learning_rate": 9.703003483180836e-05, "loss": 0.3242, "step": 8610 }, { "epoch": 4.052235294117647, "grad_norm": 0.49229843463474365, "learning_rate": 9.698902514703558e-05, "loss": 0.2715, "step": 8611 }, { "epoch": 4.052705882352941, "grad_norm": 0.29323641760977365, "learning_rate": 9.694801998966011e-05, "loss": 0.1074, "step": 8612 }, { "epoch": 4.053176470588236, "grad_norm": 0.529363978851336, "learning_rate": 9.690701936318395e-05, "loss": 0.252, "step": 8613 }, { "epoch": 4.05364705882353, "grad_norm": 0.3280231797197967, "learning_rate": 9.686602327110869e-05, "loss": 0.1279, "step": 8614 }, { "epoch": 4.054117647058823, "grad_norm": 0.45144584444418107, "learning_rate": 9.682503171693573e-05, "loss": 0.2383, "step": 8615 }, { "epoch": 4.054588235294117, "grad_norm": 0.8540616442844844, "learning_rate": 9.678404470416587e-05, "loss": 0.3555, "step": 8616 }, { "epoch": 4.055058823529412, "grad_norm": 0.5340639066082741, "learning_rate": 9.674306223629963e-05, "loss": 0.375, "step": 8617 }, { "epoch": 4.055529411764706, "grad_norm": 0.24092297603349477, "learning_rate": 9.670208431683704e-05, "loss": 0.084, "step": 8618 }, { "epoch": 4.056, "grad_norm": 0.18002892456055536, "learning_rate": 9.666111094927792e-05, "loss": 0.0608, "step": 8619 }, { "epoch": 4.056470588235294, "grad_norm": 0.4853843874827034, "learning_rate": 9.662014213712155e-05, "loss": 0.1709, "step": 8620 }, { "epoch": 4.0569411764705885, "grad_norm": 0.5750666648609658, "learning_rate": 9.657917788386684e-05, "loss": 0.3164, "step": 8621 }, { "epoch": 4.057411764705883, "grad_norm": 0.5985639403963983, "learning_rate": 9.65382181930124e-05, "loss": 0.3281, "step": 8622 }, { "epoch": 4.057882352941176, "grad_norm": 0.8174806304009572, "learning_rate": 9.649726306805635e-05, "loss": 0.4961, "step": 8623 }, { "epoch": 4.05835294117647, "grad_norm": 0.47482373869096056, "learning_rate": 9.645631251249643e-05, "loss": 0.2041, "step": 8624 }, { "epoch": 4.0588235294117645, "grad_norm": 0.5284884972302327, "learning_rate": 9.641536652983008e-05, "loss": 0.3008, "step": 8625 }, { "epoch": 4.059294117647059, "grad_norm": 0.6872501213078327, "learning_rate": 9.637442512355428e-05, "loss": 0.291, "step": 8626 }, { "epoch": 4.059764705882353, "grad_norm": 0.38522482832976596, "learning_rate": 9.63334882971656e-05, "loss": 0.2344, "step": 8627 }, { "epoch": 4.060235294117647, "grad_norm": 0.600015105529857, "learning_rate": 9.629255605416021e-05, "loss": 0.2715, "step": 8628 }, { "epoch": 4.060705882352941, "grad_norm": 0.6285752833120836, "learning_rate": 9.6251628398034e-05, "loss": 0.2656, "step": 8629 }, { "epoch": 4.061176470588236, "grad_norm": 0.4166893452701863, "learning_rate": 9.621070533228236e-05, "loss": 0.1621, "step": 8630 }, { "epoch": 4.06164705882353, "grad_norm": 0.29771462454062936, "learning_rate": 9.616978686040028e-05, "loss": 0.1099, "step": 8631 }, { "epoch": 4.062117647058823, "grad_norm": 0.6594431375863915, "learning_rate": 9.612887298588246e-05, "loss": 0.4199, "step": 8632 }, { "epoch": 4.062588235294117, "grad_norm": 0.5163515004894258, "learning_rate": 9.608796371222316e-05, "loss": 0.3438, "step": 8633 }, { "epoch": 4.063058823529412, "grad_norm": 0.580667805803247, "learning_rate": 9.604705904291614e-05, "loss": 0.3398, "step": 8634 }, { "epoch": 4.063529411764706, "grad_norm": 0.5077674003632355, "learning_rate": 9.600615898145496e-05, "loss": 0.2656, "step": 8635 }, { "epoch": 4.064, "grad_norm": 0.45612706577908824, "learning_rate": 9.596526353133261e-05, "loss": 0.1436, "step": 8636 }, { "epoch": 4.064470588235294, "grad_norm": 0.8652336044355797, "learning_rate": 9.59243726960418e-05, "loss": 0.5078, "step": 8637 }, { "epoch": 4.0649411764705885, "grad_norm": 0.682478084898737, "learning_rate": 9.588348647907481e-05, "loss": 0.4609, "step": 8638 }, { "epoch": 4.065411764705883, "grad_norm": 0.39322282339886394, "learning_rate": 9.584260488392353e-05, "loss": 0.1709, "step": 8639 }, { "epoch": 4.065882352941176, "grad_norm": 0.34740565312122035, "learning_rate": 9.580172791407942e-05, "loss": 0.1963, "step": 8640 }, { "epoch": 4.06635294117647, "grad_norm": 0.5641094984609225, "learning_rate": 9.576085557303355e-05, "loss": 0.2168, "step": 8641 }, { "epoch": 4.0668235294117645, "grad_norm": 0.32491057191748235, "learning_rate": 9.571998786427665e-05, "loss": 0.1289, "step": 8642 }, { "epoch": 4.067294117647059, "grad_norm": 0.7018491041405412, "learning_rate": 9.567912479129907e-05, "loss": 0.2969, "step": 8643 }, { "epoch": 4.067764705882353, "grad_norm": 0.43287781220689775, "learning_rate": 9.563826635759068e-05, "loss": 0.1699, "step": 8644 }, { "epoch": 4.068235294117647, "grad_norm": 0.3771455943838169, "learning_rate": 9.559741256664098e-05, "loss": 0.0957, "step": 8645 }, { "epoch": 4.068705882352941, "grad_norm": 0.8477376879855594, "learning_rate": 9.555656342193907e-05, "loss": 0.5547, "step": 8646 }, { "epoch": 4.069176470588236, "grad_norm": 0.34490727491102907, "learning_rate": 9.551571892697372e-05, "loss": 0.0977, "step": 8647 }, { "epoch": 4.06964705882353, "grad_norm": 0.4257701400202624, "learning_rate": 9.547487908523322e-05, "loss": 0.2441, "step": 8648 }, { "epoch": 4.070117647058823, "grad_norm": 0.4404894716622709, "learning_rate": 9.543404390020543e-05, "loss": 0.1904, "step": 8649 }, { "epoch": 4.070588235294117, "grad_norm": 0.24245512275467773, "learning_rate": 9.5393213375378e-05, "loss": 0.0869, "step": 8650 }, { "epoch": 4.071058823529412, "grad_norm": 0.4561982267856668, "learning_rate": 9.535238751423799e-05, "loss": 0.2295, "step": 8651 }, { "epoch": 4.071529411764706, "grad_norm": 0.35524257900103695, "learning_rate": 9.531156632027209e-05, "loss": 0.0869, "step": 8652 }, { "epoch": 4.072, "grad_norm": 0.48204483872671355, "learning_rate": 9.52707497969667e-05, "loss": 0.1602, "step": 8653 }, { "epoch": 4.072470588235294, "grad_norm": 0.35385195285072757, "learning_rate": 9.522993794780773e-05, "loss": 0.1191, "step": 8654 }, { "epoch": 4.0729411764705885, "grad_norm": 0.5509496959780539, "learning_rate": 9.518913077628066e-05, "loss": 0.291, "step": 8655 }, { "epoch": 4.073411764705883, "grad_norm": 0.4757026543054119, "learning_rate": 9.514832828587071e-05, "loss": 0.3418, "step": 8656 }, { "epoch": 4.073882352941176, "grad_norm": 0.46637814700484825, "learning_rate": 9.510753048006257e-05, "loss": 0.209, "step": 8657 }, { "epoch": 4.07435294117647, "grad_norm": 0.4609995488437375, "learning_rate": 9.506673736234057e-05, "loss": 0.1484, "step": 8658 }, { "epoch": 4.0748235294117645, "grad_norm": 0.5230355061907782, "learning_rate": 9.502594893618861e-05, "loss": 0.2559, "step": 8659 }, { "epoch": 4.075294117647059, "grad_norm": 0.43605719072940835, "learning_rate": 9.49851652050903e-05, "loss": 0.209, "step": 8660 }, { "epoch": 4.075764705882353, "grad_norm": 0.25924209812429927, "learning_rate": 9.494438617252875e-05, "loss": 0.1167, "step": 8661 }, { "epoch": 4.076235294117647, "grad_norm": 0.6914792089987897, "learning_rate": 9.490361184198659e-05, "loss": 0.3438, "step": 8662 }, { "epoch": 4.076705882352941, "grad_norm": 0.4575416550611837, "learning_rate": 9.486284221694626e-05, "loss": 0.2559, "step": 8663 }, { "epoch": 4.077176470588236, "grad_norm": 0.5391527777057924, "learning_rate": 9.482207730088968e-05, "loss": 0.2832, "step": 8664 }, { "epoch": 4.07764705882353, "grad_norm": 0.286427288446895, "learning_rate": 9.47813170972983e-05, "loss": 0.0981, "step": 8665 }, { "epoch": 4.078117647058823, "grad_norm": 0.5453563328352901, "learning_rate": 9.474056160965332e-05, "loss": 0.3926, "step": 8666 }, { "epoch": 4.078588235294117, "grad_norm": 0.698119602645962, "learning_rate": 9.469981084143541e-05, "loss": 0.4258, "step": 8667 }, { "epoch": 4.079058823529412, "grad_norm": 0.05000262738384607, "learning_rate": 9.465906479612492e-05, "loss": 0.0083, "step": 8668 }, { "epoch": 4.079529411764706, "grad_norm": 0.9485014551615156, "learning_rate": 9.461832347720168e-05, "loss": 0.3613, "step": 8669 }, { "epoch": 4.08, "grad_norm": 0.36056519105790685, "learning_rate": 9.457758688814533e-05, "loss": 0.1396, "step": 8670 }, { "epoch": 4.080470588235294, "grad_norm": 0.3505140736165327, "learning_rate": 9.453685503243489e-05, "loss": 0.1338, "step": 8671 }, { "epoch": 4.0809411764705885, "grad_norm": 0.7822098906741403, "learning_rate": 9.449612791354904e-05, "loss": 0.4043, "step": 8672 }, { "epoch": 4.081411764705883, "grad_norm": 0.2699364436065174, "learning_rate": 9.445540553496614e-05, "loss": 0.0645, "step": 8673 }, { "epoch": 4.081882352941176, "grad_norm": 0.2893553270132482, "learning_rate": 9.441468790016406e-05, "loss": 0.1299, "step": 8674 }, { "epoch": 4.08235294117647, "grad_norm": 0.28917771838676687, "learning_rate": 9.437397501262026e-05, "loss": 0.1602, "step": 8675 }, { "epoch": 4.0828235294117645, "grad_norm": 0.3208568205557452, "learning_rate": 9.433326687581183e-05, "loss": 0.1245, "step": 8676 }, { "epoch": 4.083294117647059, "grad_norm": 0.6317222701467895, "learning_rate": 9.429256349321548e-05, "loss": 0.2402, "step": 8677 }, { "epoch": 4.083764705882353, "grad_norm": 0.9864207108984383, "learning_rate": 9.425186486830745e-05, "loss": 0.4961, "step": 8678 }, { "epoch": 4.084235294117647, "grad_norm": 0.7021042889795853, "learning_rate": 9.421117100456356e-05, "loss": 0.2197, "step": 8679 }, { "epoch": 4.084705882352941, "grad_norm": 0.35421281319883813, "learning_rate": 9.417048190545935e-05, "loss": 0.1309, "step": 8680 }, { "epoch": 4.085176470588236, "grad_norm": 0.40556356923647296, "learning_rate": 9.412979757446983e-05, "loss": 0.166, "step": 8681 }, { "epoch": 4.085647058823529, "grad_norm": 0.34827597567313295, "learning_rate": 9.40891180150696e-05, "loss": 0.1543, "step": 8682 }, { "epoch": 4.086117647058823, "grad_norm": 0.23557341858124128, "learning_rate": 9.404844323073296e-05, "loss": 0.054, "step": 8683 }, { "epoch": 4.086588235294117, "grad_norm": 0.7370951131595292, "learning_rate": 9.400777322493373e-05, "loss": 0.4316, "step": 8684 }, { "epoch": 4.087058823529412, "grad_norm": 2.066339231810894, "learning_rate": 9.396710800114526e-05, "loss": 1.5469, "step": 8685 }, { "epoch": 4.087529411764706, "grad_norm": 0.5564550325794825, "learning_rate": 9.39264475628406e-05, "loss": 0.334, "step": 8686 }, { "epoch": 4.088, "grad_norm": 0.8188696770690596, "learning_rate": 9.388579191349237e-05, "loss": 0.3809, "step": 8687 }, { "epoch": 4.088470588235294, "grad_norm": 0.41748203436391773, "learning_rate": 9.384514105657277e-05, "loss": 0.1289, "step": 8688 }, { "epoch": 4.0889411764705885, "grad_norm": 0.7037010106642568, "learning_rate": 9.380449499555358e-05, "loss": 0.3301, "step": 8689 }, { "epoch": 4.089411764705883, "grad_norm": 0.39112939503406297, "learning_rate": 9.376385373390611e-05, "loss": 0.2041, "step": 8690 }, { "epoch": 4.089882352941176, "grad_norm": 0.8437877354484107, "learning_rate": 9.372321727510139e-05, "loss": 0.4102, "step": 8691 }, { "epoch": 4.09035294117647, "grad_norm": 0.28904641862564834, "learning_rate": 9.368258562260999e-05, "loss": 0.1152, "step": 8692 }, { "epoch": 4.0908235294117645, "grad_norm": 0.6285595017294633, "learning_rate": 9.364195877990195e-05, "loss": 0.4102, "step": 8693 }, { "epoch": 4.091294117647059, "grad_norm": 0.5318168282869571, "learning_rate": 9.360133675044708e-05, "loss": 0.2832, "step": 8694 }, { "epoch": 4.091764705882353, "grad_norm": 0.6512345892829144, "learning_rate": 9.356071953771472e-05, "loss": 0.2871, "step": 8695 }, { "epoch": 4.092235294117647, "grad_norm": 0.6874494691845394, "learning_rate": 9.352010714517371e-05, "loss": 0.2178, "step": 8696 }, { "epoch": 4.092705882352941, "grad_norm": 0.5033955457360769, "learning_rate": 9.347949957629258e-05, "loss": 0.3145, "step": 8697 }, { "epoch": 4.093176470588236, "grad_norm": 0.29078710854660317, "learning_rate": 9.343889683453945e-05, "loss": 0.0767, "step": 8698 }, { "epoch": 4.09364705882353, "grad_norm": 0.3574575092104462, "learning_rate": 9.339829892338193e-05, "loss": 0.1553, "step": 8699 }, { "epoch": 4.094117647058823, "grad_norm": 0.7078547633532865, "learning_rate": 9.33577058462873e-05, "loss": 0.4082, "step": 8700 }, { "epoch": 4.094588235294117, "grad_norm": 0.5396840828979016, "learning_rate": 9.331711760672246e-05, "loss": 0.334, "step": 8701 }, { "epoch": 4.095058823529412, "grad_norm": 0.8256498252592283, "learning_rate": 9.32765342081538e-05, "loss": 0.4395, "step": 8702 }, { "epoch": 4.095529411764706, "grad_norm": 0.17421518194368582, "learning_rate": 9.323595565404731e-05, "loss": 0.0518, "step": 8703 }, { "epoch": 4.096, "grad_norm": 0.3175327856444526, "learning_rate": 9.319538194786867e-05, "loss": 0.1143, "step": 8704 }, { "epoch": 4.096470588235294, "grad_norm": 0.1311292962806196, "learning_rate": 9.315481309308303e-05, "loss": 0.0288, "step": 8705 }, { "epoch": 4.0969411764705885, "grad_norm": 0.5495830145501832, "learning_rate": 9.311424909315514e-05, "loss": 0.3828, "step": 8706 }, { "epoch": 4.097411764705883, "grad_norm": 0.5049380036511806, "learning_rate": 9.307368995154945e-05, "loss": 0.3027, "step": 8707 }, { "epoch": 4.097882352941176, "grad_norm": 0.5074165980869134, "learning_rate": 9.303313567172985e-05, "loss": 0.3477, "step": 8708 }, { "epoch": 4.09835294117647, "grad_norm": 0.48309746900669664, "learning_rate": 9.299258625715988e-05, "loss": 0.2227, "step": 8709 }, { "epoch": 4.0988235294117645, "grad_norm": 0.5367113757544654, "learning_rate": 9.295204171130264e-05, "loss": 0.2656, "step": 8710 }, { "epoch": 4.099294117647059, "grad_norm": 0.5468495656361938, "learning_rate": 9.291150203762087e-05, "loss": 0.2871, "step": 8711 }, { "epoch": 4.099764705882353, "grad_norm": 0.5955578587364286, "learning_rate": 9.287096723957685e-05, "loss": 0.3262, "step": 8712 }, { "epoch": 4.100235294117647, "grad_norm": 0.16569004813498528, "learning_rate": 9.283043732063243e-05, "loss": 0.028, "step": 8713 }, { "epoch": 4.100705882352941, "grad_norm": 0.588689418417459, "learning_rate": 9.27899122842491e-05, "loss": 0.2559, "step": 8714 }, { "epoch": 4.101176470588236, "grad_norm": 0.26160015232393047, "learning_rate": 9.274939213388785e-05, "loss": 0.1748, "step": 8715 }, { "epoch": 4.101647058823529, "grad_norm": 0.6701029678230161, "learning_rate": 9.270887687300932e-05, "loss": 0.5, "step": 8716 }, { "epoch": 4.102117647058823, "grad_norm": 0.5487918441446569, "learning_rate": 9.266836650507375e-05, "loss": 0.3066, "step": 8717 }, { "epoch": 4.102588235294117, "grad_norm": 0.42349352276744384, "learning_rate": 9.26278610335409e-05, "loss": 0.166, "step": 8718 }, { "epoch": 4.103058823529412, "grad_norm": 0.6449431279185173, "learning_rate": 9.25873604618701e-05, "loss": 0.5, "step": 8719 }, { "epoch": 4.103529411764706, "grad_norm": 0.5271809891904579, "learning_rate": 9.254686479352029e-05, "loss": 0.2031, "step": 8720 }, { "epoch": 4.104, "grad_norm": 0.5213666408435035, "learning_rate": 9.250637403195008e-05, "loss": 0.2559, "step": 8721 }, { "epoch": 4.104470588235294, "grad_norm": 0.35160338930514273, "learning_rate": 9.246588818061753e-05, "loss": 0.1514, "step": 8722 }, { "epoch": 4.1049411764705885, "grad_norm": 0.5087252813444222, "learning_rate": 9.242540724298032e-05, "loss": 0.3242, "step": 8723 }, { "epoch": 4.105411764705883, "grad_norm": 0.6194980830596191, "learning_rate": 9.238493122249574e-05, "loss": 0.3262, "step": 8724 }, { "epoch": 4.105882352941176, "grad_norm": 0.45539153319853587, "learning_rate": 9.234446012262065e-05, "loss": 0.2256, "step": 8725 }, { "epoch": 4.10635294117647, "grad_norm": 0.5289490792772912, "learning_rate": 9.230399394681141e-05, "loss": 0.293, "step": 8726 }, { "epoch": 4.1068235294117645, "grad_norm": 0.38491786915898624, "learning_rate": 9.226353269852413e-05, "loss": 0.1177, "step": 8727 }, { "epoch": 4.107294117647059, "grad_norm": 0.31921584702087036, "learning_rate": 9.222307638121436e-05, "loss": 0.0781, "step": 8728 }, { "epoch": 4.107764705882353, "grad_norm": 0.28256810388012027, "learning_rate": 9.218262499833723e-05, "loss": 0.0977, "step": 8729 }, { "epoch": 4.108235294117647, "grad_norm": 0.7613775393716492, "learning_rate": 9.21421785533475e-05, "loss": 0.4355, "step": 8730 }, { "epoch": 4.108705882352941, "grad_norm": 0.299736090771427, "learning_rate": 9.210173704969952e-05, "loss": 0.1177, "step": 8731 }, { "epoch": 4.109176470588236, "grad_norm": 0.2735160134383197, "learning_rate": 9.206130049084721e-05, "loss": 0.1162, "step": 8732 }, { "epoch": 4.109647058823529, "grad_norm": 0.7758132754510996, "learning_rate": 9.202086888024404e-05, "loss": 0.4395, "step": 8733 }, { "epoch": 4.110117647058823, "grad_norm": 0.25288427468043373, "learning_rate": 9.1980442221343e-05, "loss": 0.0854, "step": 8734 }, { "epoch": 4.110588235294117, "grad_norm": 0.22523932689954218, "learning_rate": 9.194002051759685e-05, "loss": 0.0464, "step": 8735 }, { "epoch": 4.111058823529412, "grad_norm": 0.7111056948090776, "learning_rate": 9.189960377245768e-05, "loss": 0.5078, "step": 8736 }, { "epoch": 4.111529411764706, "grad_norm": 0.3663718094273956, "learning_rate": 9.185919198937736e-05, "loss": 0.1602, "step": 8737 }, { "epoch": 4.112, "grad_norm": 0.32658922663882195, "learning_rate": 9.181878517180719e-05, "loss": 0.0815, "step": 8738 }, { "epoch": 4.112470588235294, "grad_norm": 0.37239178846971366, "learning_rate": 9.177838332319818e-05, "loss": 0.1709, "step": 8739 }, { "epoch": 4.1129411764705885, "grad_norm": 0.3128695363866202, "learning_rate": 9.173798644700079e-05, "loss": 0.125, "step": 8740 }, { "epoch": 4.113411764705883, "grad_norm": 0.7211102932532828, "learning_rate": 9.169759454666511e-05, "loss": 0.3965, "step": 8741 }, { "epoch": 4.113882352941176, "grad_norm": 0.31079549928626243, "learning_rate": 9.165720762564086e-05, "loss": 0.2031, "step": 8742 }, { "epoch": 4.11435294117647, "grad_norm": 0.4392057098916457, "learning_rate": 9.161682568737726e-05, "loss": 0.2402, "step": 8743 }, { "epoch": 4.1148235294117645, "grad_norm": 0.2503508023665373, "learning_rate": 9.157644873532308e-05, "loss": 0.063, "step": 8744 }, { "epoch": 4.115294117647059, "grad_norm": 0.5214413216169468, "learning_rate": 9.153607677292676e-05, "loss": 0.3574, "step": 8745 }, { "epoch": 4.115764705882353, "grad_norm": 0.3042333433250561, "learning_rate": 9.149570980363624e-05, "loss": 0.1133, "step": 8746 }, { "epoch": 4.116235294117647, "grad_norm": 0.565619611127401, "learning_rate": 9.145534783089909e-05, "loss": 0.2891, "step": 8747 }, { "epoch": 4.116705882352941, "grad_norm": 0.4779564134220846, "learning_rate": 9.141499085816237e-05, "loss": 0.375, "step": 8748 }, { "epoch": 4.117176470588236, "grad_norm": 0.6153118054391666, "learning_rate": 9.137463888887279e-05, "loss": 0.248, "step": 8749 }, { "epoch": 4.117647058823529, "grad_norm": 0.5413771374245528, "learning_rate": 9.133429192647661e-05, "loss": 0.2178, "step": 8750 }, { "epoch": 4.118117647058823, "grad_norm": 0.7376286368067889, "learning_rate": 9.129394997441963e-05, "loss": 0.5234, "step": 8751 }, { "epoch": 4.118588235294117, "grad_norm": 0.6871168451715305, "learning_rate": 9.12536130361473e-05, "loss": 0.3867, "step": 8752 }, { "epoch": 4.119058823529412, "grad_norm": 0.3397007895711048, "learning_rate": 9.121328111510457e-05, "loss": 0.1396, "step": 8753 }, { "epoch": 4.119529411764706, "grad_norm": 0.7132457014954116, "learning_rate": 9.117295421473595e-05, "loss": 0.2295, "step": 8754 }, { "epoch": 4.12, "grad_norm": 0.6483800941492214, "learning_rate": 9.113263233848562e-05, "loss": 0.5156, "step": 8755 }, { "epoch": 4.120470588235294, "grad_norm": 0.27562944785379173, "learning_rate": 9.109231548979722e-05, "loss": 0.106, "step": 8756 }, { "epoch": 4.1209411764705886, "grad_norm": 0.25550206212436577, "learning_rate": 9.105200367211401e-05, "loss": 0.0864, "step": 8757 }, { "epoch": 4.121411764705883, "grad_norm": 0.3571249781646845, "learning_rate": 9.101169688887882e-05, "loss": 0.1455, "step": 8758 }, { "epoch": 4.121882352941176, "grad_norm": 0.7238968851319808, "learning_rate": 9.097139514353407e-05, "loss": 0.3203, "step": 8759 }, { "epoch": 4.12235294117647, "grad_norm": 1.1153407612268962, "learning_rate": 9.093109843952171e-05, "loss": 0.4473, "step": 8760 }, { "epoch": 4.1228235294117646, "grad_norm": 0.5049091414109813, "learning_rate": 9.089080678028325e-05, "loss": 0.2656, "step": 8761 }, { "epoch": 4.123294117647059, "grad_norm": 0.7050602559032766, "learning_rate": 9.085052016925984e-05, "loss": 0.3359, "step": 8762 }, { "epoch": 4.123764705882353, "grad_norm": 0.2557205356094414, "learning_rate": 9.081023860989213e-05, "loss": 0.0552, "step": 8763 }, { "epoch": 4.124235294117647, "grad_norm": 0.5354236936003256, "learning_rate": 9.076996210562035e-05, "loss": 0.3203, "step": 8764 }, { "epoch": 4.124705882352941, "grad_norm": 0.289004067700469, "learning_rate": 9.072969065988435e-05, "loss": 0.1436, "step": 8765 }, { "epoch": 4.125176470588236, "grad_norm": 0.7018480639814924, "learning_rate": 9.06894242761235e-05, "loss": 0.2422, "step": 8766 }, { "epoch": 4.125647058823529, "grad_norm": 0.6107778407549501, "learning_rate": 9.064916295777669e-05, "loss": 0.1787, "step": 8767 }, { "epoch": 4.126117647058823, "grad_norm": 0.45504127984886916, "learning_rate": 9.06089067082825e-05, "loss": 0.1768, "step": 8768 }, { "epoch": 4.1265882352941174, "grad_norm": 0.6707193613174751, "learning_rate": 9.0568655531079e-05, "loss": 0.3633, "step": 8769 }, { "epoch": 4.127058823529412, "grad_norm": 0.2733097587717274, "learning_rate": 9.052840942960383e-05, "loss": 0.0649, "step": 8770 }, { "epoch": 4.127529411764706, "grad_norm": 0.47399674646685375, "learning_rate": 9.048816840729415e-05, "loss": 0.2002, "step": 8771 }, { "epoch": 4.128, "grad_norm": 0.10750170673255337, "learning_rate": 9.044793246758685e-05, "loss": 0.0214, "step": 8772 }, { "epoch": 4.128470588235294, "grad_norm": 0.6363216004566439, "learning_rate": 9.04077016139182e-05, "loss": 0.4688, "step": 8773 }, { "epoch": 4.128941176470589, "grad_norm": 0.47310007395522274, "learning_rate": 9.03674758497241e-05, "loss": 0.3516, "step": 8774 }, { "epoch": 4.129411764705883, "grad_norm": 0.5284189438988571, "learning_rate": 9.032725517844004e-05, "loss": 0.3203, "step": 8775 }, { "epoch": 4.129882352941176, "grad_norm": 0.40830363279426823, "learning_rate": 9.028703960350113e-05, "loss": 0.1836, "step": 8776 }, { "epoch": 4.13035294117647, "grad_norm": 0.3959328740072792, "learning_rate": 9.024682912834196e-05, "loss": 0.1504, "step": 8777 }, { "epoch": 4.130823529411765, "grad_norm": 0.48958672195606684, "learning_rate": 9.020662375639663e-05, "loss": 0.293, "step": 8778 }, { "epoch": 4.131294117647059, "grad_norm": 0.5980066402639566, "learning_rate": 9.016642349109892e-05, "loss": 0.2793, "step": 8779 }, { "epoch": 4.131764705882353, "grad_norm": 0.4593179537378178, "learning_rate": 9.012622833588213e-05, "loss": 0.2061, "step": 8780 }, { "epoch": 4.132235294117647, "grad_norm": 0.39404692252807655, "learning_rate": 9.008603829417913e-05, "loss": 0.1592, "step": 8781 }, { "epoch": 4.1327058823529415, "grad_norm": 0.7471255144675482, "learning_rate": 9.00458533694223e-05, "loss": 0.3789, "step": 8782 }, { "epoch": 4.133176470588236, "grad_norm": 0.607406386908269, "learning_rate": 9.000567356504368e-05, "loss": 0.3867, "step": 8783 }, { "epoch": 4.133647058823529, "grad_norm": 0.4112594515663783, "learning_rate": 8.996549888447484e-05, "loss": 0.127, "step": 8784 }, { "epoch": 4.134117647058823, "grad_norm": 0.5162302149157441, "learning_rate": 8.99253293311468e-05, "loss": 0.3594, "step": 8785 }, { "epoch": 4.1345882352941175, "grad_norm": 0.5952553920848569, "learning_rate": 8.988516490849037e-05, "loss": 0.3145, "step": 8786 }, { "epoch": 4.135058823529412, "grad_norm": 0.5046976856400791, "learning_rate": 8.984500561993569e-05, "loss": 0.1562, "step": 8787 }, { "epoch": 4.135529411764706, "grad_norm": 0.5452764326745891, "learning_rate": 8.98048514689126e-05, "loss": 0.3164, "step": 8788 }, { "epoch": 4.136, "grad_norm": 0.33540951115170026, "learning_rate": 8.976470245885041e-05, "loss": 0.1416, "step": 8789 }, { "epoch": 4.136470588235294, "grad_norm": 0.5047133352989629, "learning_rate": 8.972455859317811e-05, "loss": 0.2949, "step": 8790 }, { "epoch": 4.136941176470589, "grad_norm": 0.5327760236642969, "learning_rate": 8.968441987532417e-05, "loss": 0.2637, "step": 8791 }, { "epoch": 4.137411764705883, "grad_norm": 0.5723163929417547, "learning_rate": 8.964428630871657e-05, "loss": 0.1289, "step": 8792 }, { "epoch": 4.137882352941176, "grad_norm": 0.6610292312141413, "learning_rate": 8.960415789678303e-05, "loss": 0.3457, "step": 8793 }, { "epoch": 4.13835294117647, "grad_norm": 0.6297122348047918, "learning_rate": 8.95640346429506e-05, "loss": 0.3438, "step": 8794 }, { "epoch": 4.138823529411765, "grad_norm": 0.418472995945334, "learning_rate": 8.952391655064605e-05, "loss": 0.1934, "step": 8795 }, { "epoch": 4.139294117647059, "grad_norm": 0.7936836735670405, "learning_rate": 8.948380362329567e-05, "loss": 0.6211, "step": 8796 }, { "epoch": 4.139764705882353, "grad_norm": 0.6028003923690434, "learning_rate": 8.944369586432531e-05, "loss": 0.2285, "step": 8797 }, { "epoch": 4.140235294117647, "grad_norm": 0.23161902848313676, "learning_rate": 8.940359327716032e-05, "loss": 0.0923, "step": 8798 }, { "epoch": 4.1407058823529415, "grad_norm": 0.13557422761683519, "learning_rate": 8.936349586522569e-05, "loss": 0.0265, "step": 8799 }, { "epoch": 4.141176470588236, "grad_norm": 0.38436385790848987, "learning_rate": 8.932340363194595e-05, "loss": 0.1328, "step": 8800 }, { "epoch": 4.141647058823529, "grad_norm": 0.25852244790983403, "learning_rate": 8.928331658074516e-05, "loss": 0.082, "step": 8801 }, { "epoch": 4.142117647058823, "grad_norm": 0.6448706729081398, "learning_rate": 8.92432347150469e-05, "loss": 0.3086, "step": 8802 }, { "epoch": 4.1425882352941175, "grad_norm": 0.41367878427180266, "learning_rate": 8.920315803827446e-05, "loss": 0.1582, "step": 8803 }, { "epoch": 4.143058823529412, "grad_norm": 0.4955340706040792, "learning_rate": 8.916308655385054e-05, "loss": 0.3027, "step": 8804 }, { "epoch": 4.143529411764706, "grad_norm": 0.5077536879396031, "learning_rate": 8.912302026519738e-05, "loss": 0.2119, "step": 8805 }, { "epoch": 4.144, "grad_norm": 0.6413407895909892, "learning_rate": 8.908295917573692e-05, "loss": 0.4355, "step": 8806 }, { "epoch": 4.144470588235294, "grad_norm": 0.6795373619488783, "learning_rate": 8.904290328889055e-05, "loss": 0.4551, "step": 8807 }, { "epoch": 4.144941176470589, "grad_norm": 0.6856346615111181, "learning_rate": 8.900285260807922e-05, "loss": 0.3125, "step": 8808 }, { "epoch": 4.145411764705882, "grad_norm": 0.2755945934992064, "learning_rate": 8.896280713672344e-05, "loss": 0.084, "step": 8809 }, { "epoch": 4.145882352941176, "grad_norm": 0.5679733408483056, "learning_rate": 8.892276687824335e-05, "loss": 0.4473, "step": 8810 }, { "epoch": 4.14635294117647, "grad_norm": 0.49295026575837747, "learning_rate": 8.888273183605856e-05, "loss": 0.3086, "step": 8811 }, { "epoch": 4.146823529411765, "grad_norm": 0.2940912088773981, "learning_rate": 8.884270201358821e-05, "loss": 0.0781, "step": 8812 }, { "epoch": 4.147294117647059, "grad_norm": 0.3989576673750348, "learning_rate": 8.880267741425111e-05, "loss": 0.1387, "step": 8813 }, { "epoch": 4.147764705882353, "grad_norm": 0.5388001187766404, "learning_rate": 8.876265804146554e-05, "loss": 0.3164, "step": 8814 }, { "epoch": 4.148235294117647, "grad_norm": 0.4793921403514286, "learning_rate": 8.872264389864929e-05, "loss": 0.2988, "step": 8815 }, { "epoch": 4.1487058823529415, "grad_norm": 0.04059876555800666, "learning_rate": 8.868263498921987e-05, "loss": 0.0068, "step": 8816 }, { "epoch": 4.149176470588236, "grad_norm": 0.8892810708036357, "learning_rate": 8.864263131659417e-05, "loss": 0.6562, "step": 8817 }, { "epoch": 4.149647058823529, "grad_norm": 0.7516711491344363, "learning_rate": 8.860263288418866e-05, "loss": 0.4688, "step": 8818 }, { "epoch": 4.150117647058823, "grad_norm": 0.41635129744202864, "learning_rate": 8.856263969541948e-05, "loss": 0.1157, "step": 8819 }, { "epoch": 4.1505882352941175, "grad_norm": 0.4353169355674776, "learning_rate": 8.852265175370223e-05, "loss": 0.2471, "step": 8820 }, { "epoch": 4.151058823529412, "grad_norm": 0.23760060400020805, "learning_rate": 8.848266906245207e-05, "loss": 0.1226, "step": 8821 }, { "epoch": 4.151529411764706, "grad_norm": 0.6838463120670538, "learning_rate": 8.844269162508373e-05, "loss": 0.416, "step": 8822 }, { "epoch": 4.152, "grad_norm": 0.6028925653617399, "learning_rate": 8.840271944501143e-05, "loss": 0.4238, "step": 8823 }, { "epoch": 4.152470588235294, "grad_norm": 0.2831476288087767, "learning_rate": 8.836275252564903e-05, "loss": 0.1113, "step": 8824 }, { "epoch": 4.152941176470589, "grad_norm": 0.5569027146840367, "learning_rate": 8.832279087040992e-05, "loss": 0.3379, "step": 8825 }, { "epoch": 4.153411764705883, "grad_norm": 0.5389234117012168, "learning_rate": 8.828283448270698e-05, "loss": 0.291, "step": 8826 }, { "epoch": 4.153882352941176, "grad_norm": 0.4840143717258405, "learning_rate": 8.824288336595267e-05, "loss": 0.2949, "step": 8827 }, { "epoch": 4.15435294117647, "grad_norm": 0.3343665773994474, "learning_rate": 8.820293752355906e-05, "loss": 0.1152, "step": 8828 }, { "epoch": 4.154823529411765, "grad_norm": 0.5137649617736334, "learning_rate": 8.816299695893773e-05, "loss": 0.2812, "step": 8829 }, { "epoch": 4.155294117647059, "grad_norm": 0.738723905644015, "learning_rate": 8.81230616754997e-05, "loss": 0.4707, "step": 8830 }, { "epoch": 4.155764705882353, "grad_norm": 0.6690411379444868, "learning_rate": 8.808313167665575e-05, "loss": 0.2969, "step": 8831 }, { "epoch": 4.156235294117647, "grad_norm": 0.4472457217581903, "learning_rate": 8.804320696581606e-05, "loss": 0.2266, "step": 8832 }, { "epoch": 4.1567058823529415, "grad_norm": 0.4492298227393551, "learning_rate": 8.800328754639038e-05, "loss": 0.1875, "step": 8833 }, { "epoch": 4.157176470588236, "grad_norm": 0.4188862800957806, "learning_rate": 8.796337342178804e-05, "loss": 0.1992, "step": 8834 }, { "epoch": 4.157647058823529, "grad_norm": 0.2651452485045242, "learning_rate": 8.792346459541789e-05, "loss": 0.1367, "step": 8835 }, { "epoch": 4.158117647058823, "grad_norm": 0.9106199101388979, "learning_rate": 8.788356107068833e-05, "loss": 0.5508, "step": 8836 }, { "epoch": 4.1585882352941175, "grad_norm": 0.5171525589225486, "learning_rate": 8.784366285100738e-05, "loss": 0.3184, "step": 8837 }, { "epoch": 4.159058823529412, "grad_norm": 0.4120849231373852, "learning_rate": 8.780376993978246e-05, "loss": 0.1357, "step": 8838 }, { "epoch": 4.159529411764706, "grad_norm": 0.47787276035374776, "learning_rate": 8.77638823404207e-05, "loss": 0.3262, "step": 8839 }, { "epoch": 4.16, "grad_norm": 0.6772623439121006, "learning_rate": 8.772400005632859e-05, "loss": 0.5664, "step": 8840 }, { "epoch": 4.160470588235294, "grad_norm": 0.4341307583824158, "learning_rate": 8.768412309091238e-05, "loss": 0.1177, "step": 8841 }, { "epoch": 4.160941176470589, "grad_norm": 0.17849709194355773, "learning_rate": 8.764425144757771e-05, "loss": 0.0491, "step": 8842 }, { "epoch": 4.161411764705882, "grad_norm": 0.5012563662459683, "learning_rate": 8.760438512972982e-05, "loss": 0.2969, "step": 8843 }, { "epoch": 4.161882352941176, "grad_norm": 0.45721790177709626, "learning_rate": 8.756452414077351e-05, "loss": 0.2109, "step": 8844 }, { "epoch": 4.16235294117647, "grad_norm": 0.51993230057698, "learning_rate": 8.752466848411305e-05, "loss": 0.2852, "step": 8845 }, { "epoch": 4.162823529411765, "grad_norm": 0.739701536138338, "learning_rate": 8.748481816315233e-05, "loss": 0.3398, "step": 8846 }, { "epoch": 4.163294117647059, "grad_norm": 0.29201142166317445, "learning_rate": 8.744497318129483e-05, "loss": 0.1289, "step": 8847 }, { "epoch": 4.163764705882353, "grad_norm": 0.490057495887631, "learning_rate": 8.74051335419434e-05, "loss": 0.293, "step": 8848 }, { "epoch": 4.164235294117647, "grad_norm": 0.5976653441754425, "learning_rate": 8.736529924850057e-05, "loss": 0.2432, "step": 8849 }, { "epoch": 4.1647058823529415, "grad_norm": 0.5244484701203532, "learning_rate": 8.73254703043685e-05, "loss": 0.3105, "step": 8850 }, { "epoch": 4.165176470588236, "grad_norm": 0.5045163052029106, "learning_rate": 8.72856467129486e-05, "loss": 0.2178, "step": 8851 }, { "epoch": 4.165647058823529, "grad_norm": 0.7116076435854531, "learning_rate": 8.72458284776421e-05, "loss": 0.3066, "step": 8852 }, { "epoch": 4.166117647058823, "grad_norm": 0.7513340650568893, "learning_rate": 8.72060156018497e-05, "loss": 0.2598, "step": 8853 }, { "epoch": 4.1665882352941175, "grad_norm": 0.5849287949647083, "learning_rate": 8.716620808897151e-05, "loss": 0.4355, "step": 8854 }, { "epoch": 4.167058823529412, "grad_norm": 0.6388092978304332, "learning_rate": 8.712640594240735e-05, "loss": 0.5469, "step": 8855 }, { "epoch": 4.167529411764706, "grad_norm": 0.5259153444492077, "learning_rate": 8.708660916555657e-05, "loss": 0.2217, "step": 8856 }, { "epoch": 4.168, "grad_norm": 0.9410763696889012, "learning_rate": 8.70468177618179e-05, "loss": 0.4707, "step": 8857 }, { "epoch": 4.168470588235294, "grad_norm": 0.3529008033025359, "learning_rate": 8.700703173458978e-05, "loss": 0.1533, "step": 8858 }, { "epoch": 4.168941176470589, "grad_norm": 0.32469737232793416, "learning_rate": 8.69672510872702e-05, "loss": 0.1084, "step": 8859 }, { "epoch": 4.169411764705882, "grad_norm": 0.6423598250581708, "learning_rate": 8.692747582325646e-05, "loss": 0.3945, "step": 8860 }, { "epoch": 4.169882352941176, "grad_norm": 0.3646310865703618, "learning_rate": 8.68877059459457e-05, "loss": 0.1768, "step": 8861 }, { "epoch": 4.17035294117647, "grad_norm": 0.468839777243969, "learning_rate": 8.684794145873443e-05, "loss": 0.3027, "step": 8862 }, { "epoch": 4.170823529411765, "grad_norm": 0.6062231455883131, "learning_rate": 8.680818236501867e-05, "loss": 0.3145, "step": 8863 }, { "epoch": 4.171294117647059, "grad_norm": 0.5138329937305907, "learning_rate": 8.67684286681941e-05, "loss": 0.2002, "step": 8864 }, { "epoch": 4.171764705882353, "grad_norm": 0.7354805498993252, "learning_rate": 8.672868037165587e-05, "loss": 0.5586, "step": 8865 }, { "epoch": 4.172235294117647, "grad_norm": 0.5278246197410879, "learning_rate": 8.668893747879869e-05, "loss": 0.2871, "step": 8866 }, { "epoch": 4.1727058823529415, "grad_norm": 0.5709690332055088, "learning_rate": 8.664919999301683e-05, "loss": 0.4434, "step": 8867 }, { "epoch": 4.173176470588236, "grad_norm": 0.628556371359245, "learning_rate": 8.660946791770397e-05, "loss": 0.498, "step": 8868 }, { "epoch": 4.173647058823529, "grad_norm": 0.5869494874382652, "learning_rate": 8.656974125625349e-05, "loss": 0.3438, "step": 8869 }, { "epoch": 4.174117647058823, "grad_norm": 0.5478527929950414, "learning_rate": 8.653002001205827e-05, "loss": 0.2812, "step": 8870 }, { "epoch": 4.1745882352941175, "grad_norm": 0.7114124812194677, "learning_rate": 8.649030418851061e-05, "loss": 0.5078, "step": 8871 }, { "epoch": 4.175058823529412, "grad_norm": 0.6629104159738645, "learning_rate": 8.645059378900255e-05, "loss": 0.3535, "step": 8872 }, { "epoch": 4.175529411764706, "grad_norm": 0.25356042358245245, "learning_rate": 8.641088881692542e-05, "loss": 0.1318, "step": 8873 }, { "epoch": 4.176, "grad_norm": 0.5046952489639024, "learning_rate": 8.637118927567028e-05, "loss": 0.2773, "step": 8874 }, { "epoch": 4.176470588235294, "grad_norm": 0.43381367222402906, "learning_rate": 8.633149516862775e-05, "loss": 0.2559, "step": 8875 }, { "epoch": 4.176941176470589, "grad_norm": 0.5773686414035243, "learning_rate": 8.629180649918775e-05, "loss": 0.3809, "step": 8876 }, { "epoch": 4.177411764705882, "grad_norm": 0.2958322305631198, "learning_rate": 8.625212327073998e-05, "loss": 0.1035, "step": 8877 }, { "epoch": 4.177882352941176, "grad_norm": 0.680968103694262, "learning_rate": 8.621244548667363e-05, "loss": 0.4609, "step": 8878 }, { "epoch": 4.17835294117647, "grad_norm": 0.5530558497196395, "learning_rate": 8.617277315037723e-05, "loss": 0.2402, "step": 8879 }, { "epoch": 4.178823529411765, "grad_norm": 0.9610772468872874, "learning_rate": 8.613310626523909e-05, "loss": 0.6875, "step": 8880 }, { "epoch": 4.179294117647059, "grad_norm": 0.8640537675111888, "learning_rate": 8.609344483464697e-05, "loss": 0.5352, "step": 8881 }, { "epoch": 4.179764705882353, "grad_norm": 0.7403549195893315, "learning_rate": 8.60537888619881e-05, "loss": 0.6094, "step": 8882 }, { "epoch": 4.180235294117647, "grad_norm": 0.8064697216952263, "learning_rate": 8.601413835064929e-05, "loss": 0.6719, "step": 8883 }, { "epoch": 4.1807058823529415, "grad_norm": 0.7634072136201935, "learning_rate": 8.597449330401698e-05, "loss": 0.4492, "step": 8884 }, { "epoch": 4.181176470588236, "grad_norm": 0.37272168018762336, "learning_rate": 8.593485372547693e-05, "loss": 0.1064, "step": 8885 }, { "epoch": 4.181647058823529, "grad_norm": 0.33174236842720817, "learning_rate": 8.589521961841462e-05, "loss": 0.124, "step": 8886 }, { "epoch": 4.182117647058823, "grad_norm": 0.8095292180563753, "learning_rate": 8.585559098621503e-05, "loss": 0.3242, "step": 8887 }, { "epoch": 4.1825882352941175, "grad_norm": 0.47957715498361464, "learning_rate": 8.581596783226254e-05, "loss": 0.2734, "step": 8888 }, { "epoch": 4.183058823529412, "grad_norm": 0.9719166531209852, "learning_rate": 8.577635015994122e-05, "loss": 0.6758, "step": 8889 }, { "epoch": 4.183529411764706, "grad_norm": 0.5424966323643307, "learning_rate": 8.573673797263466e-05, "loss": 0.3398, "step": 8890 }, { "epoch": 4.184, "grad_norm": 0.512219552053069, "learning_rate": 8.569713127372583e-05, "loss": 0.3301, "step": 8891 }, { "epoch": 4.184470588235294, "grad_norm": 0.38393305024118557, "learning_rate": 8.565753006659741e-05, "loss": 0.1484, "step": 8892 }, { "epoch": 4.184941176470589, "grad_norm": 0.5912149073616703, "learning_rate": 8.561793435463157e-05, "loss": 0.3691, "step": 8893 }, { "epoch": 4.185411764705882, "grad_norm": 0.7279656849479906, "learning_rate": 8.557834414120986e-05, "loss": 0.5, "step": 8894 }, { "epoch": 4.185882352941176, "grad_norm": 0.4080449957144196, "learning_rate": 8.553875942971362e-05, "loss": 0.167, "step": 8895 }, { "epoch": 4.18635294117647, "grad_norm": 0.5800267532591179, "learning_rate": 8.549918022352343e-05, "loss": 0.2949, "step": 8896 }, { "epoch": 4.186823529411765, "grad_norm": 0.44290932177025, "learning_rate": 8.545960652601964e-05, "loss": 0.1211, "step": 8897 }, { "epoch": 4.187294117647059, "grad_norm": 0.5100781117541195, "learning_rate": 8.542003834058207e-05, "loss": 0.3262, "step": 8898 }, { "epoch": 4.187764705882353, "grad_norm": 0.30956652002298385, "learning_rate": 8.538047567058994e-05, "loss": 0.1309, "step": 8899 }, { "epoch": 4.188235294117647, "grad_norm": 0.7291815318927923, "learning_rate": 8.534091851942214e-05, "loss": 0.5156, "step": 8900 }, { "epoch": 4.1887058823529415, "grad_norm": 0.22166892011864572, "learning_rate": 8.530136689045712e-05, "loss": 0.0552, "step": 8901 }, { "epoch": 4.189176470588236, "grad_norm": 0.7553243089646062, "learning_rate": 8.526182078707261e-05, "loss": 0.2891, "step": 8902 }, { "epoch": 4.189647058823529, "grad_norm": 0.8057444107049023, "learning_rate": 8.522228021264618e-05, "loss": 0.582, "step": 8903 }, { "epoch": 4.190117647058823, "grad_norm": 0.5586148658185494, "learning_rate": 8.518274517055482e-05, "loss": 0.1826, "step": 8904 }, { "epoch": 4.1905882352941175, "grad_norm": 0.5000084029149053, "learning_rate": 8.514321566417488e-05, "loss": 0.2676, "step": 8905 }, { "epoch": 4.191058823529412, "grad_norm": 0.5104388551811826, "learning_rate": 8.510369169688243e-05, "loss": 0.3301, "step": 8906 }, { "epoch": 4.191529411764706, "grad_norm": 0.4093586910461303, "learning_rate": 8.506417327205309e-05, "loss": 0.1572, "step": 8907 }, { "epoch": 4.192, "grad_norm": 0.35269651614510106, "learning_rate": 8.502466039306175e-05, "loss": 0.1162, "step": 8908 }, { "epoch": 4.192470588235294, "grad_norm": 0.4002030438579036, "learning_rate": 8.498515306328323e-05, "loss": 0.1338, "step": 8909 }, { "epoch": 4.192941176470589, "grad_norm": 0.75810512997597, "learning_rate": 8.494565128609146e-05, "loss": 0.291, "step": 8910 }, { "epoch": 4.193411764705882, "grad_norm": 0.5173728513834789, "learning_rate": 8.490615506486018e-05, "loss": 0.2773, "step": 8911 }, { "epoch": 4.193882352941176, "grad_norm": 0.8536921146007485, "learning_rate": 8.486666440296259e-05, "loss": 0.4785, "step": 8912 }, { "epoch": 4.19435294117647, "grad_norm": 0.8579698888049775, "learning_rate": 8.482717930377129e-05, "loss": 0.4297, "step": 8913 }, { "epoch": 4.194823529411765, "grad_norm": 0.13473471648616384, "learning_rate": 8.478769977065853e-05, "loss": 0.0386, "step": 8914 }, { "epoch": 4.195294117647059, "grad_norm": 0.8309031781032543, "learning_rate": 8.474822580699614e-05, "loss": 0.582, "step": 8915 }, { "epoch": 4.195764705882353, "grad_norm": 0.48109063857888107, "learning_rate": 8.470875741615527e-05, "loss": 0.3105, "step": 8916 }, { "epoch": 4.196235294117647, "grad_norm": 0.3826504165270635, "learning_rate": 8.466929460150678e-05, "loss": 0.0923, "step": 8917 }, { "epoch": 4.1967058823529415, "grad_norm": 0.29478736238409314, "learning_rate": 8.462983736642101e-05, "loss": 0.1875, "step": 8918 }, { "epoch": 4.197176470588236, "grad_norm": 0.6129594865481792, "learning_rate": 8.459038571426772e-05, "loss": 0.3379, "step": 8919 }, { "epoch": 4.197647058823529, "grad_norm": 0.20370284202477235, "learning_rate": 8.455093964841632e-05, "loss": 0.051, "step": 8920 }, { "epoch": 4.198117647058823, "grad_norm": 0.1449062556951238, "learning_rate": 8.451149917223576e-05, "loss": 0.0381, "step": 8921 }, { "epoch": 4.1985882352941175, "grad_norm": 0.5330673719655077, "learning_rate": 8.447206428909434e-05, "loss": 0.2949, "step": 8922 }, { "epoch": 4.199058823529412, "grad_norm": 0.4219988115043982, "learning_rate": 8.44326350023601e-05, "loss": 0.2393, "step": 8923 }, { "epoch": 4.199529411764706, "grad_norm": 0.6001576805109173, "learning_rate": 8.439321131540036e-05, "loss": 0.2334, "step": 8924 }, { "epoch": 4.2, "grad_norm": 0.3877015760479619, "learning_rate": 8.435379323158218e-05, "loss": 0.1553, "step": 8925 }, { "epoch": 4.200470588235294, "grad_norm": 0.5641850367841387, "learning_rate": 8.431438075427213e-05, "loss": 0.3613, "step": 8926 }, { "epoch": 4.200941176470589, "grad_norm": 0.5466679766796115, "learning_rate": 8.427497388683607e-05, "loss": 0.291, "step": 8927 }, { "epoch": 4.201411764705882, "grad_norm": 0.3606261195554683, "learning_rate": 8.423557263263961e-05, "loss": 0.1006, "step": 8928 }, { "epoch": 4.201882352941176, "grad_norm": 0.5447030364833653, "learning_rate": 8.419617699504791e-05, "loss": 0.375, "step": 8929 }, { "epoch": 4.20235294117647, "grad_norm": 0.5264944856711149, "learning_rate": 8.415678697742538e-05, "loss": 0.3516, "step": 8930 }, { "epoch": 4.202823529411765, "grad_norm": 0.635613155299587, "learning_rate": 8.41174025831362e-05, "loss": 0.3281, "step": 8931 }, { "epoch": 4.203294117647059, "grad_norm": 0.5958493314014799, "learning_rate": 8.407802381554406e-05, "loss": 0.3086, "step": 8932 }, { "epoch": 4.203764705882353, "grad_norm": 0.6726125962806235, "learning_rate": 8.403865067801197e-05, "loss": 0.6016, "step": 8933 }, { "epoch": 4.204235294117647, "grad_norm": 0.8196887716535964, "learning_rate": 8.399928317390265e-05, "loss": 0.582, "step": 8934 }, { "epoch": 4.2047058823529415, "grad_norm": 0.580295487295583, "learning_rate": 8.395992130657834e-05, "loss": 0.2373, "step": 8935 }, { "epoch": 4.205176470588235, "grad_norm": 0.6644298744103158, "learning_rate": 8.392056507940061e-05, "loss": 0.3457, "step": 8936 }, { "epoch": 4.205647058823529, "grad_norm": 0.45705714910583634, "learning_rate": 8.388121449573075e-05, "loss": 0.2217, "step": 8937 }, { "epoch": 4.206117647058823, "grad_norm": 0.7560274111063155, "learning_rate": 8.384186955892954e-05, "loss": 0.4258, "step": 8938 }, { "epoch": 4.2065882352941175, "grad_norm": 0.48523383483653254, "learning_rate": 8.380253027235709e-05, "loss": 0.2695, "step": 8939 }, { "epoch": 4.207058823529412, "grad_norm": 0.5352892075189731, "learning_rate": 8.37631966393733e-05, "loss": 0.2754, "step": 8940 }, { "epoch": 4.207529411764706, "grad_norm": 0.4945004040752701, "learning_rate": 8.372386866333743e-05, "loss": 0.416, "step": 8941 }, { "epoch": 4.208, "grad_norm": 0.3278240443311379, "learning_rate": 8.36845463476082e-05, "loss": 0.0874, "step": 8942 }, { "epoch": 4.208470588235294, "grad_norm": 0.22142788264360969, "learning_rate": 8.364522969554399e-05, "loss": 0.0518, "step": 8943 }, { "epoch": 4.208941176470589, "grad_norm": 0.36794213612833176, "learning_rate": 8.36059187105027e-05, "loss": 0.1592, "step": 8944 }, { "epoch": 4.209411764705882, "grad_norm": 0.7631181309083734, "learning_rate": 8.356661339584156e-05, "loss": 0.4648, "step": 8945 }, { "epoch": 4.209882352941176, "grad_norm": 0.3927138833042343, "learning_rate": 8.352731375491754e-05, "loss": 0.1934, "step": 8946 }, { "epoch": 4.21035294117647, "grad_norm": 0.34819486876130035, "learning_rate": 8.348801979108691e-05, "loss": 0.1235, "step": 8947 }, { "epoch": 4.210823529411765, "grad_norm": 0.7407345357161118, "learning_rate": 8.344873150770562e-05, "loss": 0.4668, "step": 8948 }, { "epoch": 4.211294117647059, "grad_norm": 0.5140667398248766, "learning_rate": 8.340944890812916e-05, "loss": 0.2715, "step": 8949 }, { "epoch": 4.211764705882353, "grad_norm": 0.4534252640675833, "learning_rate": 8.337017199571235e-05, "loss": 0.1533, "step": 8950 }, { "epoch": 4.212235294117647, "grad_norm": 0.10229004793135801, "learning_rate": 8.333090077380964e-05, "loss": 0.041, "step": 8951 }, { "epoch": 4.2127058823529415, "grad_norm": 0.8026443983498351, "learning_rate": 8.329163524577504e-05, "loss": 0.3301, "step": 8952 }, { "epoch": 4.213176470588236, "grad_norm": 0.7071558245992146, "learning_rate": 8.325237541496197e-05, "loss": 0.4102, "step": 8953 }, { "epoch": 4.213647058823529, "grad_norm": 0.5214798364266601, "learning_rate": 8.32131212847235e-05, "loss": 0.21, "step": 8954 }, { "epoch": 4.214117647058823, "grad_norm": 0.6106735966156417, "learning_rate": 8.3173872858412e-05, "loss": 0.2178, "step": 8955 }, { "epoch": 4.2145882352941175, "grad_norm": 0.3492168050485693, "learning_rate": 8.313463013937953e-05, "loss": 0.1396, "step": 8956 }, { "epoch": 4.215058823529412, "grad_norm": 1.097981015455599, "learning_rate": 8.309539313097769e-05, "loss": 0.2148, "step": 8957 }, { "epoch": 4.215529411764706, "grad_norm": 0.394048538407313, "learning_rate": 8.305616183655738e-05, "loss": 0.2002, "step": 8958 }, { "epoch": 4.216, "grad_norm": 0.6144917199187043, "learning_rate": 8.30169362594692e-05, "loss": 0.332, "step": 8959 }, { "epoch": 4.216470588235294, "grad_norm": 0.6080919458398982, "learning_rate": 8.29777164030633e-05, "loss": 0.4102, "step": 8960 }, { "epoch": 4.216941176470589, "grad_norm": 0.4837824405252441, "learning_rate": 8.293850227068908e-05, "loss": 0.2422, "step": 8961 }, { "epoch": 4.217411764705882, "grad_norm": 0.5478755349655584, "learning_rate": 8.289929386569569e-05, "loss": 0.248, "step": 8962 }, { "epoch": 4.217882352941176, "grad_norm": 0.25458773196547174, "learning_rate": 8.28600911914318e-05, "loss": 0.1055, "step": 8963 }, { "epoch": 4.21835294117647, "grad_norm": 0.22653663118172734, "learning_rate": 8.282089425124538e-05, "loss": 0.0544, "step": 8964 }, { "epoch": 4.218823529411765, "grad_norm": 0.47391370656980425, "learning_rate": 8.278170304848413e-05, "loss": 0.3086, "step": 8965 }, { "epoch": 4.219294117647059, "grad_norm": 0.5368209766732664, "learning_rate": 8.274251758649518e-05, "loss": 0.3633, "step": 8966 }, { "epoch": 4.219764705882353, "grad_norm": 1.0590224628876108, "learning_rate": 8.270333786862506e-05, "loss": 0.498, "step": 8967 }, { "epoch": 4.220235294117647, "grad_norm": 0.8440942032048213, "learning_rate": 8.266416389821997e-05, "loss": 0.5273, "step": 8968 }, { "epoch": 4.2207058823529415, "grad_norm": 0.2474509875713161, "learning_rate": 8.262499567862565e-05, "loss": 0.0771, "step": 8969 }, { "epoch": 4.221176470588235, "grad_norm": 0.7084195257549929, "learning_rate": 8.258583321318708e-05, "loss": 0.4707, "step": 8970 }, { "epoch": 4.221647058823529, "grad_norm": 0.5710885155453812, "learning_rate": 8.254667650524904e-05, "loss": 0.4121, "step": 8971 }, { "epoch": 4.222117647058823, "grad_norm": 0.44038577159803605, "learning_rate": 8.250752555815574e-05, "loss": 0.2734, "step": 8972 }, { "epoch": 4.2225882352941175, "grad_norm": 0.3629812931297279, "learning_rate": 8.246838037525073e-05, "loss": 0.2676, "step": 8973 }, { "epoch": 4.223058823529412, "grad_norm": 0.2868925666405666, "learning_rate": 8.242924095987736e-05, "loss": 0.0835, "step": 8974 }, { "epoch": 4.223529411764706, "grad_norm": 0.6930768461906897, "learning_rate": 8.239010731537816e-05, "loss": 0.4297, "step": 8975 }, { "epoch": 4.224, "grad_norm": 0.6471356097493632, "learning_rate": 8.235097944509545e-05, "loss": 0.3359, "step": 8976 }, { "epoch": 4.224470588235294, "grad_norm": 0.28397041958195574, "learning_rate": 8.231185735237096e-05, "loss": 0.0977, "step": 8977 }, { "epoch": 4.224941176470589, "grad_norm": 0.9745725213896242, "learning_rate": 8.22727410405458e-05, "loss": 0.3926, "step": 8978 }, { "epoch": 4.225411764705882, "grad_norm": 0.7019133042147438, "learning_rate": 8.223363051296074e-05, "loss": 0.3809, "step": 8979 }, { "epoch": 4.225882352941176, "grad_norm": 0.5592355550544025, "learning_rate": 8.21945257729561e-05, "loss": 0.3105, "step": 8980 }, { "epoch": 4.22635294117647, "grad_norm": 0.36731653475803877, "learning_rate": 8.215542682387148e-05, "loss": 0.1328, "step": 8981 }, { "epoch": 4.226823529411765, "grad_norm": 0.606360465624307, "learning_rate": 8.211633366904618e-05, "loss": 0.252, "step": 8982 }, { "epoch": 4.227294117647059, "grad_norm": 0.43264887886577225, "learning_rate": 8.207724631181904e-05, "loss": 0.1748, "step": 8983 }, { "epoch": 4.227764705882353, "grad_norm": 0.490343799860872, "learning_rate": 8.203816475552813e-05, "loss": 0.2793, "step": 8984 }, { "epoch": 4.228235294117647, "grad_norm": 0.3007512738630056, "learning_rate": 8.19990890035113e-05, "loss": 0.0825, "step": 8985 }, { "epoch": 4.2287058823529415, "grad_norm": 0.8447473061517531, "learning_rate": 8.196001905910589e-05, "loss": 0.5625, "step": 8986 }, { "epoch": 4.229176470588235, "grad_norm": 0.7371497266968109, "learning_rate": 8.192095492564851e-05, "loss": 0.3691, "step": 8987 }, { "epoch": 4.229647058823529, "grad_norm": 0.3496435677681237, "learning_rate": 8.188189660647554e-05, "loss": 0.1348, "step": 8988 }, { "epoch": 4.230117647058823, "grad_norm": 0.7471791001572633, "learning_rate": 8.184284410492274e-05, "loss": 0.3926, "step": 8989 }, { "epoch": 4.2305882352941175, "grad_norm": 0.6554164627223468, "learning_rate": 8.180379742432533e-05, "loss": 0.4121, "step": 8990 }, { "epoch": 4.231058823529412, "grad_norm": 0.8391590679571854, "learning_rate": 8.176475656801811e-05, "loss": 0.2197, "step": 8991 }, { "epoch": 4.231529411764706, "grad_norm": 0.36989259393450763, "learning_rate": 8.172572153933541e-05, "loss": 0.0669, "step": 8992 }, { "epoch": 4.232, "grad_norm": 0.44164211019338184, "learning_rate": 8.168669234161094e-05, "loss": 0.1689, "step": 8993 }, { "epoch": 4.232470588235294, "grad_norm": 0.6115445510547258, "learning_rate": 8.164766897817806e-05, "loss": 0.3281, "step": 8994 }, { "epoch": 4.232941176470589, "grad_norm": 0.49841291416199857, "learning_rate": 8.160865145236946e-05, "loss": 0.1777, "step": 8995 }, { "epoch": 4.233411764705882, "grad_norm": 0.5409785715351876, "learning_rate": 8.15696397675175e-05, "loss": 0.3008, "step": 8996 }, { "epoch": 4.233882352941176, "grad_norm": 0.5481524422977139, "learning_rate": 8.153063392695394e-05, "loss": 0.1787, "step": 8997 }, { "epoch": 4.23435294117647, "grad_norm": 0.6133647051041076, "learning_rate": 8.14916339340101e-05, "loss": 0.3184, "step": 8998 }, { "epoch": 4.234823529411765, "grad_norm": 0.9403254977436378, "learning_rate": 8.145263979201672e-05, "loss": 0.5859, "step": 8999 }, { "epoch": 4.235294117647059, "grad_norm": 0.6675261867409642, "learning_rate": 8.141365150430421e-05, "loss": 0.5234, "step": 9000 }, { "epoch": 4.235764705882353, "grad_norm": 0.34079393814436054, "learning_rate": 8.137466907420222e-05, "loss": 0.0645, "step": 9001 }, { "epoch": 4.236235294117647, "grad_norm": 0.3266357519398967, "learning_rate": 8.133569250504013e-05, "loss": 0.0894, "step": 9002 }, { "epoch": 4.2367058823529415, "grad_norm": 0.037075352657382266, "learning_rate": 8.129672180014666e-05, "loss": 0.0073, "step": 9003 }, { "epoch": 4.237176470588235, "grad_norm": 0.7168849947886334, "learning_rate": 8.125775696285013e-05, "loss": 0.3066, "step": 9004 }, { "epoch": 4.237647058823529, "grad_norm": 0.4455518749822619, "learning_rate": 8.121879799647839e-05, "loss": 0.2266, "step": 9005 }, { "epoch": 4.238117647058823, "grad_norm": 0.6600890037083439, "learning_rate": 8.117984490435861e-05, "loss": 0.2891, "step": 9006 }, { "epoch": 4.2385882352941175, "grad_norm": 0.5401131585901111, "learning_rate": 8.114089768981765e-05, "loss": 0.2734, "step": 9007 }, { "epoch": 4.239058823529412, "grad_norm": 0.36312895307109316, "learning_rate": 8.110195635618184e-05, "loss": 0.1396, "step": 9008 }, { "epoch": 4.239529411764706, "grad_norm": 0.46943047009794336, "learning_rate": 8.106302090677682e-05, "loss": 0.1953, "step": 9009 }, { "epoch": 4.24, "grad_norm": 0.5058851134069624, "learning_rate": 8.102409134492795e-05, "loss": 0.1934, "step": 9010 }, { "epoch": 4.240470588235294, "grad_norm": 0.6246860037073128, "learning_rate": 8.098516767396008e-05, "loss": 0.2207, "step": 9011 }, { "epoch": 4.240941176470589, "grad_norm": 0.43269516593121027, "learning_rate": 8.094624989719734e-05, "loss": 0.127, "step": 9012 }, { "epoch": 4.241411764705882, "grad_norm": 0.35186162103551266, "learning_rate": 8.090733801796356e-05, "loss": 0.1279, "step": 9013 }, { "epoch": 4.241882352941176, "grad_norm": 0.12828349831839944, "learning_rate": 8.086843203958207e-05, "loss": 0.0219, "step": 9014 }, { "epoch": 4.24235294117647, "grad_norm": 1.0686993372991478, "learning_rate": 8.082953196537553e-05, "loss": 0.4766, "step": 9015 }, { "epoch": 4.242823529411765, "grad_norm": 0.5121633583317508, "learning_rate": 8.079063779866622e-05, "loss": 0.2402, "step": 9016 }, { "epoch": 4.243294117647059, "grad_norm": 0.5519320636517432, "learning_rate": 8.075174954277598e-05, "loss": 0.2969, "step": 9017 }, { "epoch": 4.243764705882353, "grad_norm": 0.5174936112169368, "learning_rate": 8.071286720102594e-05, "loss": 0.1328, "step": 9018 }, { "epoch": 4.244235294117647, "grad_norm": 0.7934754774042961, "learning_rate": 8.067399077673688e-05, "loss": 0.4492, "step": 9019 }, { "epoch": 4.2447058823529416, "grad_norm": 0.5488600367775971, "learning_rate": 8.06351202732291e-05, "loss": 0.2363, "step": 9020 }, { "epoch": 4.245176470588235, "grad_norm": 0.5848504316720111, "learning_rate": 8.059625569382223e-05, "loss": 0.2119, "step": 9021 }, { "epoch": 4.245647058823529, "grad_norm": 0.34549542432289443, "learning_rate": 8.055739704183553e-05, "loss": 0.1128, "step": 9022 }, { "epoch": 4.246117647058823, "grad_norm": 0.5363566565211062, "learning_rate": 8.051854432058785e-05, "loss": 0.2539, "step": 9023 }, { "epoch": 4.2465882352941176, "grad_norm": 0.7535082314818339, "learning_rate": 8.047969753339717e-05, "loss": 0.2852, "step": 9024 }, { "epoch": 4.247058823529412, "grad_norm": 0.5603506640173364, "learning_rate": 8.04408566835814e-05, "loss": 0.3613, "step": 9025 }, { "epoch": 4.247529411764706, "grad_norm": 0.5629673720176507, "learning_rate": 8.040202177445761e-05, "loss": 0.3066, "step": 9026 }, { "epoch": 4.248, "grad_norm": 0.6599686636101043, "learning_rate": 8.036319280934251e-05, "loss": 0.249, "step": 9027 }, { "epoch": 4.248470588235294, "grad_norm": 0.48022314905955943, "learning_rate": 8.032436979155237e-05, "loss": 0.1885, "step": 9028 }, { "epoch": 4.248941176470589, "grad_norm": 0.5703027222402363, "learning_rate": 8.02855527244028e-05, "loss": 0.3066, "step": 9029 }, { "epoch": 4.249411764705882, "grad_norm": 0.4342831520921471, "learning_rate": 8.024674161120892e-05, "loss": 0.1523, "step": 9030 }, { "epoch": 4.249882352941176, "grad_norm": 0.40683137681729986, "learning_rate": 8.020793645528552e-05, "loss": 0.1934, "step": 9031 }, { "epoch": 4.2503529411764704, "grad_norm": 0.4918142480276696, "learning_rate": 8.016913725994661e-05, "loss": 0.1523, "step": 9032 }, { "epoch": 4.250823529411765, "grad_norm": 0.2614629473798107, "learning_rate": 8.01303440285059e-05, "loss": 0.085, "step": 9033 }, { "epoch": 4.251294117647059, "grad_norm": 0.6077686620099377, "learning_rate": 8.009155676427659e-05, "loss": 0.2373, "step": 9034 }, { "epoch": 4.251764705882353, "grad_norm": 0.373054628252334, "learning_rate": 8.005277547057114e-05, "loss": 0.1846, "step": 9035 }, { "epoch": 4.252235294117647, "grad_norm": 0.4625193520765679, "learning_rate": 8.001400015070178e-05, "loss": 0.334, "step": 9036 }, { "epoch": 4.252705882352942, "grad_norm": 0.6567374653294368, "learning_rate": 7.997523080798014e-05, "loss": 0.4746, "step": 9037 }, { "epoch": 4.253176470588235, "grad_norm": 0.37920518154958216, "learning_rate": 7.993646744571719e-05, "loss": 0.1689, "step": 9038 }, { "epoch": 4.253647058823529, "grad_norm": 0.2889821091701213, "learning_rate": 7.98977100672236e-05, "loss": 0.0752, "step": 9039 }, { "epoch": 4.254117647058823, "grad_norm": 0.8200244596159348, "learning_rate": 7.985895867580948e-05, "loss": 0.3691, "step": 9040 }, { "epoch": 4.254588235294118, "grad_norm": 1.052845534255674, "learning_rate": 7.982021327478419e-05, "loss": 0.543, "step": 9041 }, { "epoch": 4.255058823529412, "grad_norm": 0.39727381128861894, "learning_rate": 7.978147386745705e-05, "loss": 0.1035, "step": 9042 }, { "epoch": 4.255529411764706, "grad_norm": 0.6361802259352816, "learning_rate": 7.974274045713638e-05, "loss": 0.332, "step": 9043 }, { "epoch": 4.256, "grad_norm": 0.5036477934582727, "learning_rate": 7.97040130471303e-05, "loss": 0.1924, "step": 9044 }, { "epoch": 4.2564705882352944, "grad_norm": 0.8165968772075634, "learning_rate": 7.966529164074636e-05, "loss": 0.3105, "step": 9045 }, { "epoch": 4.256941176470589, "grad_norm": 0.5961169809808896, "learning_rate": 7.962657624129147e-05, "loss": 0.3691, "step": 9046 }, { "epoch": 4.257411764705882, "grad_norm": 0.6808450417087344, "learning_rate": 7.958786685207212e-05, "loss": 0.3848, "step": 9047 }, { "epoch": 4.257882352941176, "grad_norm": 0.6849292618978224, "learning_rate": 7.954916347639438e-05, "loss": 0.3848, "step": 9048 }, { "epoch": 4.2583529411764705, "grad_norm": 0.21577192418403346, "learning_rate": 7.951046611756358e-05, "loss": 0.0488, "step": 9049 }, { "epoch": 4.258823529411765, "grad_norm": 0.28601249888937125, "learning_rate": 7.947177477888472e-05, "loss": 0.0684, "step": 9050 }, { "epoch": 4.259294117647059, "grad_norm": 0.21576807136804735, "learning_rate": 7.94330894636623e-05, "loss": 0.1152, "step": 9051 }, { "epoch": 4.259764705882353, "grad_norm": 0.3013628434180105, "learning_rate": 7.939441017520011e-05, "loss": 0.1611, "step": 9052 }, { "epoch": 4.260235294117647, "grad_norm": 0.20771390688324065, "learning_rate": 7.935573691680166e-05, "loss": 0.0535, "step": 9053 }, { "epoch": 4.260705882352942, "grad_norm": 0.6572501627977126, "learning_rate": 7.931706969176972e-05, "loss": 0.3672, "step": 9054 }, { "epoch": 4.261176470588235, "grad_norm": 0.4460739184282708, "learning_rate": 7.927840850340673e-05, "loss": 0.166, "step": 9055 }, { "epoch": 4.261647058823529, "grad_norm": 0.37191153332616605, "learning_rate": 7.92397533550146e-05, "loss": 0.2021, "step": 9056 }, { "epoch": 4.262117647058823, "grad_norm": 0.7302041550642252, "learning_rate": 7.920110424989455e-05, "loss": 0.4492, "step": 9057 }, { "epoch": 4.262588235294118, "grad_norm": 0.35813953561399775, "learning_rate": 7.916246119134748e-05, "loss": 0.1147, "step": 9058 }, { "epoch": 4.263058823529412, "grad_norm": 0.2834824668556548, "learning_rate": 7.912382418267374e-05, "loss": 0.1631, "step": 9059 }, { "epoch": 4.263529411764706, "grad_norm": 0.5477450463735268, "learning_rate": 7.908519322717297e-05, "loss": 0.2988, "step": 9060 }, { "epoch": 4.264, "grad_norm": 0.5396420691559408, "learning_rate": 7.904656832814457e-05, "loss": 0.2295, "step": 9061 }, { "epoch": 4.2644705882352945, "grad_norm": 0.6607665762173224, "learning_rate": 7.900794948888728e-05, "loss": 0.3867, "step": 9062 }, { "epoch": 4.264941176470588, "grad_norm": 0.5343354020369014, "learning_rate": 7.896933671269929e-05, "loss": 0.2969, "step": 9063 }, { "epoch": 4.265411764705882, "grad_norm": 0.26252731299404064, "learning_rate": 7.893073000287834e-05, "loss": 0.0903, "step": 9064 }, { "epoch": 4.265882352941176, "grad_norm": 0.44169704925355135, "learning_rate": 7.88921293627217e-05, "loss": 0.2637, "step": 9065 }, { "epoch": 4.2663529411764705, "grad_norm": 0.5376145949029002, "learning_rate": 7.885353479552592e-05, "loss": 0.2793, "step": 9066 }, { "epoch": 4.266823529411765, "grad_norm": 0.4287930872328107, "learning_rate": 7.881494630458726e-05, "loss": 0.1748, "step": 9067 }, { "epoch": 4.267294117647059, "grad_norm": 0.5609646652443442, "learning_rate": 7.87763638932014e-05, "loss": 0.2617, "step": 9068 }, { "epoch": 4.267764705882353, "grad_norm": 0.5941609853377348, "learning_rate": 7.873778756466335e-05, "loss": 0.2578, "step": 9069 }, { "epoch": 4.268235294117647, "grad_norm": 0.6388599128818762, "learning_rate": 7.869921732226779e-05, "loss": 0.332, "step": 9070 }, { "epoch": 4.268705882352942, "grad_norm": 0.652848581274838, "learning_rate": 7.866065316930886e-05, "loss": 0.3496, "step": 9071 }, { "epoch": 4.269176470588235, "grad_norm": 0.49393172807274394, "learning_rate": 7.862209510907999e-05, "loss": 0.2676, "step": 9072 }, { "epoch": 4.269647058823529, "grad_norm": 0.5491869766444687, "learning_rate": 7.858354314487432e-05, "loss": 0.2256, "step": 9073 }, { "epoch": 4.270117647058823, "grad_norm": 0.2945035684955607, "learning_rate": 7.854499727998442e-05, "loss": 0.0889, "step": 9074 }, { "epoch": 4.270588235294118, "grad_norm": 0.7074885116168674, "learning_rate": 7.850645751770217e-05, "loss": 0.3203, "step": 9075 }, { "epoch": 4.271058823529412, "grad_norm": 0.8682921527887698, "learning_rate": 7.846792386131917e-05, "loss": 0.5117, "step": 9076 }, { "epoch": 4.271529411764706, "grad_norm": 0.8546461721546477, "learning_rate": 7.84293963141263e-05, "loss": 0.4551, "step": 9077 }, { "epoch": 4.272, "grad_norm": 0.34699633698998855, "learning_rate": 7.839087487941403e-05, "loss": 0.1426, "step": 9078 }, { "epoch": 4.2724705882352945, "grad_norm": 0.7085114340452039, "learning_rate": 7.835235956047234e-05, "loss": 0.3594, "step": 9079 }, { "epoch": 4.272941176470589, "grad_norm": 0.5865964602289889, "learning_rate": 7.83138503605905e-05, "loss": 0.2314, "step": 9080 }, { "epoch": 4.273411764705882, "grad_norm": 0.5042055704399568, "learning_rate": 7.827534728305747e-05, "loss": 0.3535, "step": 9081 }, { "epoch": 4.273882352941176, "grad_norm": 0.5747188926971422, "learning_rate": 7.823685033116166e-05, "loss": 0.334, "step": 9082 }, { "epoch": 4.2743529411764705, "grad_norm": 0.5453357005135818, "learning_rate": 7.819835950819078e-05, "loss": 0.248, "step": 9083 }, { "epoch": 4.274823529411765, "grad_norm": 0.45794245587372456, "learning_rate": 7.815987481743215e-05, "loss": 0.21, "step": 9084 }, { "epoch": 4.275294117647059, "grad_norm": 0.8884429339591345, "learning_rate": 7.812139626217266e-05, "loss": 0.4355, "step": 9085 }, { "epoch": 4.275764705882353, "grad_norm": 0.5731535312413188, "learning_rate": 7.808292384569838e-05, "loss": 0.2812, "step": 9086 }, { "epoch": 4.276235294117647, "grad_norm": 0.2703251916672987, "learning_rate": 7.804445757129525e-05, "loss": 0.0991, "step": 9087 }, { "epoch": 4.276705882352942, "grad_norm": 0.43939901757666167, "learning_rate": 7.800599744224834e-05, "loss": 0.2051, "step": 9088 }, { "epoch": 4.277176470588235, "grad_norm": 1.02997634468222, "learning_rate": 7.796754346184238e-05, "loss": 0.7539, "step": 9089 }, { "epoch": 4.277647058823529, "grad_norm": 0.7815665048269633, "learning_rate": 7.792909563336157e-05, "loss": 0.4609, "step": 9090 }, { "epoch": 4.278117647058823, "grad_norm": 0.5581990642202262, "learning_rate": 7.789065396008945e-05, "loss": 0.3184, "step": 9091 }, { "epoch": 4.278588235294118, "grad_norm": 0.7222585042802484, "learning_rate": 7.785221844530918e-05, "loss": 0.291, "step": 9092 }, { "epoch": 4.279058823529412, "grad_norm": 0.49631284967704214, "learning_rate": 7.78137890923034e-05, "loss": 0.1826, "step": 9093 }, { "epoch": 4.279529411764706, "grad_norm": 0.20022270792227215, "learning_rate": 7.777536590435404e-05, "loss": 0.0522, "step": 9094 }, { "epoch": 4.28, "grad_norm": 0.5839237811130534, "learning_rate": 7.773694888474267e-05, "loss": 0.418, "step": 9095 }, { "epoch": 4.2804705882352945, "grad_norm": 0.4556218503794098, "learning_rate": 7.769853803675041e-05, "loss": 0.1289, "step": 9096 }, { "epoch": 4.280941176470588, "grad_norm": 0.5634962937964654, "learning_rate": 7.766013336365755e-05, "loss": 0.208, "step": 9097 }, { "epoch": 4.281411764705882, "grad_norm": 0.6585725865870455, "learning_rate": 7.762173486874414e-05, "loss": 0.3242, "step": 9098 }, { "epoch": 4.281882352941176, "grad_norm": 0.3440710854725616, "learning_rate": 7.758334255528963e-05, "loss": 0.1553, "step": 9099 }, { "epoch": 4.2823529411764705, "grad_norm": 0.4516620629124544, "learning_rate": 7.754495642657282e-05, "loss": 0.2119, "step": 9100 }, { "epoch": 4.282823529411765, "grad_norm": 0.7364384636495852, "learning_rate": 7.750657648587213e-05, "loss": 0.3613, "step": 9101 }, { "epoch": 4.283294117647059, "grad_norm": 0.31760726800231925, "learning_rate": 7.746820273646542e-05, "loss": 0.124, "step": 9102 }, { "epoch": 4.283764705882353, "grad_norm": 0.6304222195374659, "learning_rate": 7.742983518162993e-05, "loss": 0.291, "step": 9103 }, { "epoch": 4.284235294117647, "grad_norm": 0.42888929560759, "learning_rate": 7.73914738246425e-05, "loss": 0.249, "step": 9104 }, { "epoch": 4.284705882352942, "grad_norm": 0.3362300824370131, "learning_rate": 7.735311866877933e-05, "loss": 0.0723, "step": 9105 }, { "epoch": 4.285176470588235, "grad_norm": 0.4812533294825419, "learning_rate": 7.731476971731612e-05, "loss": 0.2715, "step": 9106 }, { "epoch": 4.285647058823529, "grad_norm": 0.3504106286083848, "learning_rate": 7.727642697352817e-05, "loss": 0.0718, "step": 9107 }, { "epoch": 4.286117647058823, "grad_norm": 0.28723590949376726, "learning_rate": 7.723809044069e-05, "loss": 0.127, "step": 9108 }, { "epoch": 4.286588235294118, "grad_norm": 0.5010491966907097, "learning_rate": 7.719976012207579e-05, "loss": 0.2637, "step": 9109 }, { "epoch": 4.287058823529412, "grad_norm": 0.23174062819239638, "learning_rate": 7.716143602095924e-05, "loss": 0.0806, "step": 9110 }, { "epoch": 4.287529411764706, "grad_norm": 0.5281899904059499, "learning_rate": 7.712311814061323e-05, "loss": 0.1377, "step": 9111 }, { "epoch": 4.288, "grad_norm": 0.41184331525389084, "learning_rate": 7.708480648431043e-05, "loss": 0.1504, "step": 9112 }, { "epoch": 4.2884705882352945, "grad_norm": 1.086877379338915, "learning_rate": 7.704650105532284e-05, "loss": 0.5352, "step": 9113 }, { "epoch": 4.288941176470589, "grad_norm": 0.6200465328169995, "learning_rate": 7.700820185692184e-05, "loss": 0.2734, "step": 9114 }, { "epoch": 4.289411764705882, "grad_norm": 0.4143293272066671, "learning_rate": 7.69699088923784e-05, "loss": 0.2227, "step": 9115 }, { "epoch": 4.289882352941176, "grad_norm": 0.25193624759594974, "learning_rate": 7.693162216496303e-05, "loss": 0.0728, "step": 9116 }, { "epoch": 4.2903529411764705, "grad_norm": 0.35746606679938764, "learning_rate": 7.689334167794547e-05, "loss": 0.2305, "step": 9117 }, { "epoch": 4.290823529411765, "grad_norm": 0.5859387077002653, "learning_rate": 7.685506743459511e-05, "loss": 0.3359, "step": 9118 }, { "epoch": 4.291294117647059, "grad_norm": 0.3897112788915612, "learning_rate": 7.681679943818082e-05, "loss": 0.1035, "step": 9119 }, { "epoch": 4.291764705882353, "grad_norm": 0.21699685448863507, "learning_rate": 7.677853769197078e-05, "loss": 0.0588, "step": 9120 }, { "epoch": 4.292235294117647, "grad_norm": 0.532583091454966, "learning_rate": 7.674028219923275e-05, "loss": 0.2354, "step": 9121 }, { "epoch": 4.292705882352941, "grad_norm": 0.845315065068755, "learning_rate": 7.670203296323403e-05, "loss": 0.543, "step": 9122 }, { "epoch": 4.293176470588235, "grad_norm": 0.7793451616171837, "learning_rate": 7.666378998724116e-05, "loss": 0.3945, "step": 9123 }, { "epoch": 4.293647058823529, "grad_norm": 0.555868634483214, "learning_rate": 7.662555327452031e-05, "loss": 0.3496, "step": 9124 }, { "epoch": 4.294117647058823, "grad_norm": 0.4393476520119044, "learning_rate": 7.65873228283372e-05, "loss": 0.1846, "step": 9125 }, { "epoch": 4.294588235294118, "grad_norm": 0.4595812677041494, "learning_rate": 7.654909865195675e-05, "loss": 0.2539, "step": 9126 }, { "epoch": 4.295058823529412, "grad_norm": 0.10772378034879007, "learning_rate": 7.65108807486436e-05, "loss": 0.0184, "step": 9127 }, { "epoch": 4.295529411764706, "grad_norm": 0.38406817815648714, "learning_rate": 7.647266912166162e-05, "loss": 0.1035, "step": 9128 }, { "epoch": 4.296, "grad_norm": 0.47059058349262994, "learning_rate": 7.643446377427438e-05, "loss": 0.1523, "step": 9129 }, { "epoch": 4.2964705882352945, "grad_norm": 0.7624756530225332, "learning_rate": 7.639626470974474e-05, "loss": 0.1865, "step": 9130 }, { "epoch": 4.296941176470588, "grad_norm": 0.41279337754024653, "learning_rate": 7.635807193133512e-05, "loss": 0.1426, "step": 9131 }, { "epoch": 4.297411764705882, "grad_norm": 0.3332307159284424, "learning_rate": 7.631988544230745e-05, "loss": 0.0957, "step": 9132 }, { "epoch": 4.297882352941176, "grad_norm": 0.6797656573167932, "learning_rate": 7.628170524592288e-05, "loss": 0.2969, "step": 9133 }, { "epoch": 4.2983529411764705, "grad_norm": 0.5747217933006371, "learning_rate": 7.624353134544229e-05, "loss": 0.2695, "step": 9134 }, { "epoch": 4.298823529411765, "grad_norm": 0.33615216098781425, "learning_rate": 7.620536374412594e-05, "loss": 0.125, "step": 9135 }, { "epoch": 4.299294117647059, "grad_norm": 0.3333316409801652, "learning_rate": 7.616720244523342e-05, "loss": 0.1191, "step": 9136 }, { "epoch": 4.299764705882353, "grad_norm": 0.6710613510047397, "learning_rate": 7.612904745202399e-05, "loss": 0.3457, "step": 9137 }, { "epoch": 4.300235294117647, "grad_norm": 0.4318910719672696, "learning_rate": 7.609089876775628e-05, "loss": 0.1953, "step": 9138 }, { "epoch": 4.300705882352942, "grad_norm": 0.3478333416222562, "learning_rate": 7.605275639568828e-05, "loss": 0.1279, "step": 9139 }, { "epoch": 4.301176470588235, "grad_norm": 0.6296741635205466, "learning_rate": 7.60146203390776e-05, "loss": 0.3379, "step": 9140 }, { "epoch": 4.301647058823529, "grad_norm": 0.32203566229883873, "learning_rate": 7.59764906011813e-05, "loss": 0.0742, "step": 9141 }, { "epoch": 4.302117647058823, "grad_norm": 1.007442490665481, "learning_rate": 7.593836718525572e-05, "loss": 0.5664, "step": 9142 }, { "epoch": 4.302588235294118, "grad_norm": 0.3911469644363012, "learning_rate": 7.590025009455684e-05, "loss": 0.1963, "step": 9143 }, { "epoch": 4.303058823529412, "grad_norm": 0.7277962422614543, "learning_rate": 7.586213933234013e-05, "loss": 0.4238, "step": 9144 }, { "epoch": 4.303529411764706, "grad_norm": 0.4015014614363717, "learning_rate": 7.58240349018603e-05, "loss": 0.1738, "step": 9145 }, { "epoch": 4.304, "grad_norm": 0.7264022622621354, "learning_rate": 7.578593680637173e-05, "loss": 0.373, "step": 9146 }, { "epoch": 4.3044705882352945, "grad_norm": 0.4224704252740412, "learning_rate": 7.574784504912823e-05, "loss": 0.1768, "step": 9147 }, { "epoch": 4.304941176470588, "grad_norm": 0.2549742347305858, "learning_rate": 7.57097596333829e-05, "loss": 0.085, "step": 9148 }, { "epoch": 4.305411764705882, "grad_norm": 0.12122742184371835, "learning_rate": 7.567168056238851e-05, "loss": 0.0261, "step": 9149 }, { "epoch": 4.305882352941176, "grad_norm": 0.45680169468704557, "learning_rate": 7.563360783939722e-05, "loss": 0.2773, "step": 9150 }, { "epoch": 4.3063529411764705, "grad_norm": 0.5207296411273168, "learning_rate": 7.559554146766054e-05, "loss": 0.2969, "step": 9151 }, { "epoch": 4.306823529411765, "grad_norm": 0.5487678346345966, "learning_rate": 7.555748145042957e-05, "loss": 0.377, "step": 9152 }, { "epoch": 4.307294117647059, "grad_norm": 0.5291631804175146, "learning_rate": 7.551942779095486e-05, "loss": 0.2637, "step": 9153 }, { "epoch": 4.307764705882353, "grad_norm": 0.5471249173431293, "learning_rate": 7.54813804924863e-05, "loss": 0.3594, "step": 9154 }, { "epoch": 4.308235294117647, "grad_norm": 0.43399926425329105, "learning_rate": 7.544333955827342e-05, "loss": 0.1953, "step": 9155 }, { "epoch": 4.308705882352941, "grad_norm": 0.7649526454254741, "learning_rate": 7.540530499156496e-05, "loss": 0.249, "step": 9156 }, { "epoch": 4.309176470588235, "grad_norm": 0.42574306543120144, "learning_rate": 7.536727679560937e-05, "loss": 0.1045, "step": 9157 }, { "epoch": 4.309647058823529, "grad_norm": 0.700511542851854, "learning_rate": 7.532925497365444e-05, "loss": 0.4062, "step": 9158 }, { "epoch": 4.310117647058823, "grad_norm": 0.48722439300728393, "learning_rate": 7.529123952894734e-05, "loss": 0.2402, "step": 9159 }, { "epoch": 4.310588235294118, "grad_norm": 0.5978958995505912, "learning_rate": 7.525323046473485e-05, "loss": 0.2773, "step": 9160 }, { "epoch": 4.311058823529412, "grad_norm": 0.5726351734573978, "learning_rate": 7.521522778426315e-05, "loss": 0.1943, "step": 9161 }, { "epoch": 4.311529411764706, "grad_norm": 0.38147157028908746, "learning_rate": 7.517723149077776e-05, "loss": 0.1206, "step": 9162 }, { "epoch": 4.312, "grad_norm": 0.5930187369946983, "learning_rate": 7.513924158752383e-05, "loss": 0.2812, "step": 9163 }, { "epoch": 4.3124705882352945, "grad_norm": 0.7013751542292203, "learning_rate": 7.510125807774589e-05, "loss": 0.4199, "step": 9164 }, { "epoch": 4.312941176470588, "grad_norm": 0.6191299317685424, "learning_rate": 7.506328096468782e-05, "loss": 0.2119, "step": 9165 }, { "epoch": 4.313411764705882, "grad_norm": 0.7104906329805049, "learning_rate": 7.502531025159313e-05, "loss": 0.332, "step": 9166 }, { "epoch": 4.313882352941176, "grad_norm": 0.22203392784460033, "learning_rate": 7.498734594170474e-05, "loss": 0.0732, "step": 9167 }, { "epoch": 4.3143529411764705, "grad_norm": 0.3677057795446683, "learning_rate": 7.49493880382649e-05, "loss": 0.1406, "step": 9168 }, { "epoch": 4.314823529411765, "grad_norm": 0.6632883342439878, "learning_rate": 7.491143654451544e-05, "loss": 0.3496, "step": 9169 }, { "epoch": 4.315294117647059, "grad_norm": 0.4450799252331369, "learning_rate": 7.487349146369766e-05, "loss": 0.2441, "step": 9170 }, { "epoch": 4.315764705882353, "grad_norm": 0.37294204633819633, "learning_rate": 7.483555279905215e-05, "loss": 0.1758, "step": 9171 }, { "epoch": 4.316235294117647, "grad_norm": 0.3726371298727479, "learning_rate": 7.479762055381909e-05, "loss": 0.1328, "step": 9172 }, { "epoch": 4.316705882352942, "grad_norm": 0.5418584969544099, "learning_rate": 7.475969473123819e-05, "loss": 0.293, "step": 9173 }, { "epoch": 4.317176470588235, "grad_norm": 1.0362723144179327e-05, "learning_rate": 7.472177533454833e-05, "loss": 0.0, "step": 9174 }, { "epoch": 4.317647058823529, "grad_norm": 0.5091011538525302, "learning_rate": 7.46838623669881e-05, "loss": 0.3027, "step": 9175 }, { "epoch": 4.318117647058823, "grad_norm": 0.8145714976887418, "learning_rate": 7.464595583179544e-05, "loss": 0.5781, "step": 9176 }, { "epoch": 4.318588235294118, "grad_norm": 0.5876053680424222, "learning_rate": 7.460805573220779e-05, "loss": 0.291, "step": 9177 }, { "epoch": 4.319058823529412, "grad_norm": 0.7230163729399554, "learning_rate": 7.4570162071462e-05, "loss": 0.4141, "step": 9178 }, { "epoch": 4.319529411764706, "grad_norm": 0.4917411029748024, "learning_rate": 7.453227485279432e-05, "loss": 0.168, "step": 9179 }, { "epoch": 4.32, "grad_norm": 0.5716023019821004, "learning_rate": 7.449439407944052e-05, "loss": 0.2451, "step": 9180 }, { "epoch": 4.3204705882352945, "grad_norm": 0.7917200096776561, "learning_rate": 7.445651975463588e-05, "loss": 0.457, "step": 9181 }, { "epoch": 4.320941176470588, "grad_norm": 0.34761756940135125, "learning_rate": 7.441865188161494e-05, "loss": 0.1396, "step": 9182 }, { "epoch": 4.321411764705882, "grad_norm": 0.3608096575192048, "learning_rate": 7.43807904636119e-05, "loss": 0.0889, "step": 9183 }, { "epoch": 4.321882352941176, "grad_norm": 0.4836471562668657, "learning_rate": 7.434293550386021e-05, "loss": 0.3965, "step": 9184 }, { "epoch": 4.3223529411764705, "grad_norm": 0.09611869650685585, "learning_rate": 7.430508700559292e-05, "loss": 0.0481, "step": 9185 }, { "epoch": 4.322823529411765, "grad_norm": 0.6701424496149512, "learning_rate": 7.426724497204254e-05, "loss": 0.3301, "step": 9186 }, { "epoch": 4.323294117647059, "grad_norm": 0.46311915114907237, "learning_rate": 7.422940940644085e-05, "loss": 0.2314, "step": 9187 }, { "epoch": 4.323764705882353, "grad_norm": 0.26144532893304645, "learning_rate": 7.419158031201924e-05, "loss": 0.0898, "step": 9188 }, { "epoch": 4.324235294117647, "grad_norm": 0.2912072575145247, "learning_rate": 7.415375769200859e-05, "loss": 0.0825, "step": 9189 }, { "epoch": 4.324705882352941, "grad_norm": 0.5914883455321039, "learning_rate": 7.411594154963897e-05, "loss": 0.373, "step": 9190 }, { "epoch": 4.325176470588235, "grad_norm": 0.7008193112050818, "learning_rate": 7.407813188814017e-05, "loss": 0.4609, "step": 9191 }, { "epoch": 4.325647058823529, "grad_norm": 0.48344303242784264, "learning_rate": 7.404032871074135e-05, "loss": 0.1826, "step": 9192 }, { "epoch": 4.326117647058823, "grad_norm": 0.5465532106862547, "learning_rate": 7.400253202067101e-05, "loss": 0.3418, "step": 9193 }, { "epoch": 4.326588235294118, "grad_norm": 0.5986482743598233, "learning_rate": 7.396474182115717e-05, "loss": 0.2871, "step": 9194 }, { "epoch": 4.327058823529412, "grad_norm": 0.3081564717689121, "learning_rate": 7.392695811542739e-05, "loss": 0.0977, "step": 9195 }, { "epoch": 4.327529411764706, "grad_norm": 0.536073323493063, "learning_rate": 7.388918090670847e-05, "loss": 0.2168, "step": 9196 }, { "epoch": 4.328, "grad_norm": 0.3597146007854429, "learning_rate": 7.385141019822683e-05, "loss": 0.252, "step": 9197 }, { "epoch": 4.3284705882352945, "grad_norm": 0.2670593162115762, "learning_rate": 7.381364599320831e-05, "loss": 0.1089, "step": 9198 }, { "epoch": 4.328941176470588, "grad_norm": 0.6102911998742807, "learning_rate": 7.377588829487806e-05, "loss": 0.377, "step": 9199 }, { "epoch": 4.329411764705882, "grad_norm": 0.35975186773905, "learning_rate": 7.373813710646083e-05, "loss": 0.1543, "step": 9200 }, { "epoch": 4.329882352941176, "grad_norm": 0.33066998631964717, "learning_rate": 7.370039243118082e-05, "loss": 0.1055, "step": 9201 }, { "epoch": 4.3303529411764705, "grad_norm": 0.6336488738400042, "learning_rate": 7.366265427226148e-05, "loss": 0.2734, "step": 9202 }, { "epoch": 4.330823529411765, "grad_norm": 0.6407320319086267, "learning_rate": 7.36249226329259e-05, "loss": 0.4609, "step": 9203 }, { "epoch": 4.331294117647059, "grad_norm": 0.506341152818082, "learning_rate": 7.358719751639659e-05, "loss": 0.2969, "step": 9204 }, { "epoch": 4.331764705882353, "grad_norm": 0.389245221069039, "learning_rate": 7.354947892589536e-05, "loss": 0.1875, "step": 9205 }, { "epoch": 4.332235294117647, "grad_norm": 0.42103733665562904, "learning_rate": 7.351176686464368e-05, "loss": 0.2422, "step": 9206 }, { "epoch": 4.332705882352942, "grad_norm": 0.9561782001541168, "learning_rate": 7.347406133586223e-05, "loss": 0.3945, "step": 9207 }, { "epoch": 4.333176470588235, "grad_norm": 0.5813506377300343, "learning_rate": 7.34363623427713e-05, "loss": 0.3438, "step": 9208 }, { "epoch": 4.333647058823529, "grad_norm": 0.519243477740694, "learning_rate": 7.339866988859061e-05, "loss": 0.3164, "step": 9209 }, { "epoch": 4.334117647058823, "grad_norm": 0.4906428245774491, "learning_rate": 7.33609839765392e-05, "loss": 0.252, "step": 9210 }, { "epoch": 4.334588235294118, "grad_norm": 0.2836235346260309, "learning_rate": 7.332330460983565e-05, "loss": 0.0615, "step": 9211 }, { "epoch": 4.335058823529412, "grad_norm": 0.16065402419052, "learning_rate": 7.328563179169806e-05, "loss": 0.0747, "step": 9212 }, { "epoch": 4.335529411764706, "grad_norm": 0.6056715592127644, "learning_rate": 7.324796552534373e-05, "loss": 0.3984, "step": 9213 }, { "epoch": 4.336, "grad_norm": 0.3558003743595292, "learning_rate": 7.321030581398961e-05, "loss": 0.127, "step": 9214 }, { "epoch": 4.3364705882352945, "grad_norm": 0.5064990724728697, "learning_rate": 7.317265266085209e-05, "loss": 0.2949, "step": 9215 }, { "epoch": 4.336941176470588, "grad_norm": 0.35777526483232613, "learning_rate": 7.31350060691468e-05, "loss": 0.1963, "step": 9216 }, { "epoch": 4.337411764705882, "grad_norm": 0.5222023283902254, "learning_rate": 7.309736604208902e-05, "loss": 0.4766, "step": 9217 }, { "epoch": 4.337882352941176, "grad_norm": 0.22484043073587268, "learning_rate": 7.305973258289343e-05, "loss": 0.0479, "step": 9218 }, { "epoch": 4.3383529411764705, "grad_norm": 0.44614642387629266, "learning_rate": 7.302210569477397e-05, "loss": 0.1602, "step": 9219 }, { "epoch": 4.338823529411765, "grad_norm": 0.9383343170666801, "learning_rate": 7.298448538094436e-05, "loss": 0.6445, "step": 9220 }, { "epoch": 4.339294117647059, "grad_norm": 0.47279886009232974, "learning_rate": 7.294687164461741e-05, "loss": 0.3027, "step": 9221 }, { "epoch": 4.339764705882353, "grad_norm": 0.7668695491810285, "learning_rate": 7.290926448900557e-05, "loss": 0.543, "step": 9222 }, { "epoch": 4.340235294117647, "grad_norm": 0.6256059180232798, "learning_rate": 7.287166391732071e-05, "loss": 0.3555, "step": 9223 }, { "epoch": 4.340705882352941, "grad_norm": 0.5897571287403109, "learning_rate": 7.283406993277401e-05, "loss": 0.3887, "step": 9224 }, { "epoch": 4.341176470588235, "grad_norm": 0.5535672750271472, "learning_rate": 7.279648253857626e-05, "loss": 0.3125, "step": 9225 }, { "epoch": 4.341647058823529, "grad_norm": 0.47150691840682474, "learning_rate": 7.275890173793761e-05, "loss": 0.2305, "step": 9226 }, { "epoch": 4.342117647058823, "grad_norm": 0.9203271933291488, "learning_rate": 7.272132753406757e-05, "loss": 0.5469, "step": 9227 }, { "epoch": 4.342588235294118, "grad_norm": 0.37115539680767745, "learning_rate": 7.26837599301752e-05, "loss": 0.1514, "step": 9228 }, { "epoch": 4.343058823529412, "grad_norm": 0.22370637592713874, "learning_rate": 7.264619892946904e-05, "loss": 0.0874, "step": 9229 }, { "epoch": 4.343529411764706, "grad_norm": 0.7648666154578615, "learning_rate": 7.260864453515683e-05, "loss": 0.4023, "step": 9230 }, { "epoch": 4.344, "grad_norm": 0.40058614696266814, "learning_rate": 7.2571096750446e-05, "loss": 0.1973, "step": 9231 }, { "epoch": 4.3444705882352945, "grad_norm": 0.3936102087148313, "learning_rate": 7.253355557854334e-05, "loss": 0.1982, "step": 9232 }, { "epoch": 4.344941176470588, "grad_norm": 0.249566293511172, "learning_rate": 7.249602102265496e-05, "loss": 0.0757, "step": 9233 }, { "epoch": 4.345411764705882, "grad_norm": 0.5667501801486573, "learning_rate": 7.245849308598658e-05, "loss": 0.2637, "step": 9234 }, { "epoch": 4.345882352941176, "grad_norm": 0.6803731778892615, "learning_rate": 7.242097177174318e-05, "loss": 0.3047, "step": 9235 }, { "epoch": 4.3463529411764705, "grad_norm": 0.1988899358396329, "learning_rate": 7.238345708312931e-05, "loss": 0.0505, "step": 9236 }, { "epoch": 4.346823529411765, "grad_norm": 0.4848192166016626, "learning_rate": 7.234594902334897e-05, "loss": 0.2578, "step": 9237 }, { "epoch": 4.347294117647059, "grad_norm": 0.3620498614216146, "learning_rate": 7.230844759560544e-05, "loss": 0.1914, "step": 9238 }, { "epoch": 4.347764705882353, "grad_norm": 0.2873305770405433, "learning_rate": 7.227095280310152e-05, "loss": 0.1201, "step": 9239 }, { "epoch": 4.348235294117647, "grad_norm": 0.7543847770753987, "learning_rate": 7.223346464903958e-05, "loss": 0.4434, "step": 9240 }, { "epoch": 4.348705882352942, "grad_norm": 0.8310719323168684, "learning_rate": 7.219598313662113e-05, "loss": 0.6211, "step": 9241 }, { "epoch": 4.349176470588235, "grad_norm": 0.3356912617255582, "learning_rate": 7.215850826904734e-05, "loss": 0.0796, "step": 9242 }, { "epoch": 4.349647058823529, "grad_norm": 0.29662434199294135, "learning_rate": 7.212104004951883e-05, "loss": 0.0903, "step": 9243 }, { "epoch": 4.350117647058823, "grad_norm": 0.46061312787949066, "learning_rate": 7.208357848123543e-05, "loss": 0.2422, "step": 9244 }, { "epoch": 4.350588235294118, "grad_norm": 0.6025491811476753, "learning_rate": 7.204612356739658e-05, "loss": 0.2773, "step": 9245 }, { "epoch": 4.351058823529412, "grad_norm": 0.5046830346236382, "learning_rate": 7.200867531120124e-05, "loss": 0.3984, "step": 9246 }, { "epoch": 4.351529411764706, "grad_norm": 0.2100503160584034, "learning_rate": 7.197123371584749e-05, "loss": 0.0708, "step": 9247 }, { "epoch": 4.352, "grad_norm": 0.13674858658196973, "learning_rate": 7.193379878453312e-05, "loss": 0.0625, "step": 9248 }, { "epoch": 4.352470588235294, "grad_norm": 0.8712326816241946, "learning_rate": 7.189637052045529e-05, "loss": 0.4746, "step": 9249 }, { "epoch": 4.352941176470588, "grad_norm": 0.05732185430512384, "learning_rate": 7.185894892681048e-05, "loss": 0.0127, "step": 9250 }, { "epoch": 4.353411764705882, "grad_norm": 0.641142190814798, "learning_rate": 7.182153400679468e-05, "loss": 0.3105, "step": 9251 }, { "epoch": 4.353882352941176, "grad_norm": 0.13173951180991453, "learning_rate": 7.17841257636034e-05, "loss": 0.0208, "step": 9252 }, { "epoch": 4.3543529411764705, "grad_norm": 0.49520647488304426, "learning_rate": 7.17467242004314e-05, "loss": 0.2236, "step": 9253 }, { "epoch": 4.354823529411765, "grad_norm": 0.48110235254890804, "learning_rate": 7.170932932047296e-05, "loss": 0.2314, "step": 9254 }, { "epoch": 4.355294117647059, "grad_norm": 0.7068979398590688, "learning_rate": 7.167194112692186e-05, "loss": 0.5391, "step": 9255 }, { "epoch": 4.355764705882353, "grad_norm": 0.8681377795516265, "learning_rate": 7.163455962297116e-05, "loss": 0.5703, "step": 9256 }, { "epoch": 4.356235294117647, "grad_norm": 0.681372993488395, "learning_rate": 7.159718481181346e-05, "loss": 0.3281, "step": 9257 }, { "epoch": 4.356705882352941, "grad_norm": 0.20989707207159777, "learning_rate": 7.155981669664072e-05, "loss": 0.1143, "step": 9258 }, { "epoch": 4.357176470588235, "grad_norm": 0.45546230288508194, "learning_rate": 7.152245528064437e-05, "loss": 0.2197, "step": 9259 }, { "epoch": 4.357647058823529, "grad_norm": 0.78978482821722, "learning_rate": 7.148510056701532e-05, "loss": 0.4961, "step": 9260 }, { "epoch": 4.358117647058823, "grad_norm": 0.41552137932889616, "learning_rate": 7.144775255894375e-05, "loss": 0.2119, "step": 9261 }, { "epoch": 4.358588235294118, "grad_norm": 0.28754575319954606, "learning_rate": 7.14104112596194e-05, "loss": 0.0786, "step": 9262 }, { "epoch": 4.359058823529412, "grad_norm": 0.6944138152461625, "learning_rate": 7.137307667223141e-05, "loss": 0.2773, "step": 9263 }, { "epoch": 4.359529411764706, "grad_norm": 0.5734882109546493, "learning_rate": 7.133574879996833e-05, "loss": 0.1855, "step": 9264 }, { "epoch": 4.36, "grad_norm": 0.3933726757711709, "learning_rate": 7.129842764601821e-05, "loss": 0.2012, "step": 9265 }, { "epoch": 4.3604705882352945, "grad_norm": 0.7399345095130783, "learning_rate": 7.126111321356834e-05, "loss": 0.3613, "step": 9266 }, { "epoch": 4.360941176470588, "grad_norm": 0.4069674615293755, "learning_rate": 7.122380550580562e-05, "loss": 0.2314, "step": 9267 }, { "epoch": 4.361411764705882, "grad_norm": 0.7716745391567951, "learning_rate": 7.118650452591634e-05, "loss": 0.4746, "step": 9268 }, { "epoch": 4.361882352941176, "grad_norm": 0.5388650158791841, "learning_rate": 7.114921027708613e-05, "loss": 0.2617, "step": 9269 }, { "epoch": 4.3623529411764705, "grad_norm": 0.5984316510547804, "learning_rate": 7.111192276250011e-05, "loss": 0.4707, "step": 9270 }, { "epoch": 4.362823529411765, "grad_norm": 0.44973687440492843, "learning_rate": 7.107464198534287e-05, "loss": 0.332, "step": 9271 }, { "epoch": 4.363294117647059, "grad_norm": 0.4403017123809024, "learning_rate": 7.103736794879829e-05, "loss": 0.2236, "step": 9272 }, { "epoch": 4.363764705882353, "grad_norm": 0.4030810704509298, "learning_rate": 7.100010065604978e-05, "loss": 0.1553, "step": 9273 }, { "epoch": 4.364235294117647, "grad_norm": 0.4162149829717885, "learning_rate": 7.096284011028024e-05, "loss": 0.1494, "step": 9274 }, { "epoch": 4.364705882352941, "grad_norm": 0.6968974021726674, "learning_rate": 7.092558631467177e-05, "loss": 0.3711, "step": 9275 }, { "epoch": 4.365176470588235, "grad_norm": 0.4944826700806735, "learning_rate": 7.088833927240608e-05, "loss": 0.2285, "step": 9276 }, { "epoch": 4.365647058823529, "grad_norm": 0.9144316564897826, "learning_rate": 7.085109898666433e-05, "loss": 0.5117, "step": 9277 }, { "epoch": 4.366117647058823, "grad_norm": 0.7961866122664318, "learning_rate": 7.081386546062687e-05, "loss": 0.3691, "step": 9278 }, { "epoch": 4.366588235294118, "grad_norm": 0.42421922028420267, "learning_rate": 7.077663869747372e-05, "loss": 0.1855, "step": 9279 }, { "epoch": 4.367058823529412, "grad_norm": 0.6395518803396927, "learning_rate": 7.073941870038427e-05, "loss": 0.3555, "step": 9280 }, { "epoch": 4.367529411764706, "grad_norm": 0.5429391708085433, "learning_rate": 7.070220547253717e-05, "loss": 0.3105, "step": 9281 }, { "epoch": 4.368, "grad_norm": 0.5009628392570605, "learning_rate": 7.066499901711068e-05, "loss": 0.25, "step": 9282 }, { "epoch": 4.368470588235294, "grad_norm": 0.7688747812094653, "learning_rate": 7.062779933728247e-05, "loss": 0.6562, "step": 9283 }, { "epoch": 4.368941176470588, "grad_norm": 0.6088496917702062, "learning_rate": 7.059060643622944e-05, "loss": 0.3867, "step": 9284 }, { "epoch": 4.369411764705882, "grad_norm": 0.8943698008243984, "learning_rate": 7.05534203171282e-05, "loss": 0.5469, "step": 9285 }, { "epoch": 4.369882352941176, "grad_norm": 0.3342375473895122, "learning_rate": 7.051624098315449e-05, "loss": 0.1025, "step": 9286 }, { "epoch": 4.3703529411764706, "grad_norm": 0.46518029530996846, "learning_rate": 7.047906843748365e-05, "loss": 0.1953, "step": 9287 }, { "epoch": 4.370823529411765, "grad_norm": 0.39408137606176125, "learning_rate": 7.044190268329049e-05, "loss": 0.2773, "step": 9288 }, { "epoch": 4.371294117647059, "grad_norm": 0.20518146080220234, "learning_rate": 7.0404743723749e-05, "loss": 0.0635, "step": 9289 }, { "epoch": 4.371764705882353, "grad_norm": 0.47469066398617693, "learning_rate": 7.036759156203281e-05, "loss": 0.2354, "step": 9290 }, { "epoch": 4.372235294117647, "grad_norm": 0.43764428885018386, "learning_rate": 7.033044620131495e-05, "loss": 0.1338, "step": 9291 }, { "epoch": 4.372705882352941, "grad_norm": 0.39966065599138617, "learning_rate": 7.02933076447677e-05, "loss": 0.1592, "step": 9292 }, { "epoch": 4.373176470588235, "grad_norm": 0.39935728646108987, "learning_rate": 7.025617589556297e-05, "loss": 0.1152, "step": 9293 }, { "epoch": 4.373647058823529, "grad_norm": 0.4905468990249713, "learning_rate": 7.021905095687198e-05, "loss": 0.2598, "step": 9294 }, { "epoch": 4.374117647058823, "grad_norm": 0.8144505848684112, "learning_rate": 7.018193283186531e-05, "loss": 0.5508, "step": 9295 }, { "epoch": 4.374588235294118, "grad_norm": 0.47985147851713816, "learning_rate": 7.01448215237131e-05, "loss": 0.2109, "step": 9296 }, { "epoch": 4.375058823529412, "grad_norm": 0.2799592708115643, "learning_rate": 7.010771703558488e-05, "loss": 0.1064, "step": 9297 }, { "epoch": 4.375529411764706, "grad_norm": 0.18457419339087083, "learning_rate": 7.007061937064941e-05, "loss": 0.0444, "step": 9298 }, { "epoch": 4.376, "grad_norm": 0.5388639874198842, "learning_rate": 7.00335285320751e-05, "loss": 0.3145, "step": 9299 }, { "epoch": 4.376470588235295, "grad_norm": 0.3388047654377608, "learning_rate": 6.999644452302975e-05, "loss": 0.1289, "step": 9300 }, { "epoch": 4.376941176470588, "grad_norm": 0.7045281033660539, "learning_rate": 6.995936734668042e-05, "loss": 0.4102, "step": 9301 }, { "epoch": 4.377411764705882, "grad_norm": 0.39585616323365386, "learning_rate": 6.992229700619369e-05, "loss": 0.1777, "step": 9302 }, { "epoch": 4.377882352941176, "grad_norm": 0.49212011721682747, "learning_rate": 6.988523350473563e-05, "loss": 0.1245, "step": 9303 }, { "epoch": 4.378352941176471, "grad_norm": 0.44149845724814557, "learning_rate": 6.984817684547155e-05, "loss": 0.1992, "step": 9304 }, { "epoch": 4.378823529411765, "grad_norm": 0.47478595432903803, "learning_rate": 6.981112703156628e-05, "loss": 0.1963, "step": 9305 }, { "epoch": 4.379294117647059, "grad_norm": 0.7953936943742043, "learning_rate": 6.977408406618418e-05, "loss": 0.3398, "step": 9306 }, { "epoch": 4.379764705882353, "grad_norm": 0.5005964730970724, "learning_rate": 6.973704795248871e-05, "loss": 0.2812, "step": 9307 }, { "epoch": 4.3802352941176474, "grad_norm": 0.18417533712731532, "learning_rate": 6.970001869364306e-05, "loss": 0.0583, "step": 9308 }, { "epoch": 4.380705882352941, "grad_norm": 0.2249463350334831, "learning_rate": 6.966299629280968e-05, "loss": 0.0581, "step": 9309 }, { "epoch": 4.381176470588235, "grad_norm": 0.6474055952283455, "learning_rate": 6.962598075315046e-05, "loss": 0.4883, "step": 9310 }, { "epoch": 4.381647058823529, "grad_norm": 0.530885941385125, "learning_rate": 6.958897207782677e-05, "loss": 0.3086, "step": 9311 }, { "epoch": 4.3821176470588235, "grad_norm": 0.4782221075115947, "learning_rate": 6.955197026999925e-05, "loss": 0.2168, "step": 9312 }, { "epoch": 4.382588235294118, "grad_norm": 0.4263345542612314, "learning_rate": 6.951497533282809e-05, "loss": 0.1455, "step": 9313 }, { "epoch": 4.383058823529412, "grad_norm": 0.5809744422649642, "learning_rate": 6.947798726947277e-05, "loss": 0.4141, "step": 9314 }, { "epoch": 4.383529411764706, "grad_norm": 0.8306526890645644, "learning_rate": 6.94410060830923e-05, "loss": 0.3418, "step": 9315 }, { "epoch": 4.384, "grad_norm": 0.4291496858674137, "learning_rate": 6.940403177684509e-05, "loss": 0.2061, "step": 9316 }, { "epoch": 4.384470588235294, "grad_norm": 0.7346311778042436, "learning_rate": 6.936706435388882e-05, "loss": 0.3301, "step": 9317 }, { "epoch": 4.384941176470588, "grad_norm": 0.413337824100398, "learning_rate": 6.93301038173808e-05, "loss": 0.1836, "step": 9318 }, { "epoch": 4.385411764705882, "grad_norm": 0.3463167459054849, "learning_rate": 6.929315017047762e-05, "loss": 0.1299, "step": 9319 }, { "epoch": 4.385882352941176, "grad_norm": 0.10931355316423208, "learning_rate": 6.925620341633522e-05, "loss": 0.0233, "step": 9320 }, { "epoch": 4.386352941176471, "grad_norm": 0.5223463824543905, "learning_rate": 6.921926355810911e-05, "loss": 0.2119, "step": 9321 }, { "epoch": 4.386823529411765, "grad_norm": 0.2541122565088808, "learning_rate": 6.918233059895418e-05, "loss": 0.1079, "step": 9322 }, { "epoch": 4.387294117647059, "grad_norm": 0.2714209172007904, "learning_rate": 6.914540454202457e-05, "loss": 0.0781, "step": 9323 }, { "epoch": 4.387764705882353, "grad_norm": 0.6826247387232764, "learning_rate": 6.910848539047399e-05, "loss": 0.3164, "step": 9324 }, { "epoch": 4.3882352941176475, "grad_norm": 0.05101169003204101, "learning_rate": 6.907157314745558e-05, "loss": 0.0085, "step": 9325 }, { "epoch": 4.388705882352941, "grad_norm": 0.8502315633806811, "learning_rate": 6.903466781612173e-05, "loss": 0.3105, "step": 9326 }, { "epoch": 4.389176470588235, "grad_norm": 0.34687530153936663, "learning_rate": 6.899776939962438e-05, "loss": 0.0659, "step": 9327 }, { "epoch": 4.389647058823529, "grad_norm": 0.45902264181997887, "learning_rate": 6.89608779011149e-05, "loss": 0.1826, "step": 9328 }, { "epoch": 4.3901176470588235, "grad_norm": 0.6464488475547244, "learning_rate": 6.892399332374387e-05, "loss": 0.2871, "step": 9329 }, { "epoch": 4.390588235294118, "grad_norm": 0.4355587708951472, "learning_rate": 6.888711567066149e-05, "loss": 0.1582, "step": 9330 }, { "epoch": 4.391058823529412, "grad_norm": 0.5331602593275359, "learning_rate": 6.885024494501735e-05, "loss": 0.2871, "step": 9331 }, { "epoch": 4.391529411764706, "grad_norm": 0.46136487136929316, "learning_rate": 6.881338114996027e-05, "loss": 0.1953, "step": 9332 }, { "epoch": 4.392, "grad_norm": 0.4420748102805973, "learning_rate": 6.877652428863866e-05, "loss": 0.105, "step": 9333 }, { "epoch": 4.392470588235295, "grad_norm": 0.5921947646618461, "learning_rate": 6.873967436420033e-05, "loss": 0.2471, "step": 9334 }, { "epoch": 4.392941176470588, "grad_norm": 0.6035892477829727, "learning_rate": 6.870283137979233e-05, "loss": 0.2695, "step": 9335 }, { "epoch": 4.393411764705882, "grad_norm": 0.2529281340295786, "learning_rate": 6.866599533856135e-05, "loss": 0.0815, "step": 9336 }, { "epoch": 4.393882352941176, "grad_norm": 0.6346918829719042, "learning_rate": 6.862916624365325e-05, "loss": 0.3008, "step": 9337 }, { "epoch": 4.394352941176471, "grad_norm": 0.1553062216011237, "learning_rate": 6.859234409821347e-05, "loss": 0.0391, "step": 9338 }, { "epoch": 4.394823529411765, "grad_norm": 0.15224033072147938, "learning_rate": 6.855552890538688e-05, "loss": 0.0457, "step": 9339 }, { "epoch": 4.395294117647059, "grad_norm": 0.6420654410236134, "learning_rate": 6.851872066831751e-05, "loss": 0.2734, "step": 9340 }, { "epoch": 4.395764705882353, "grad_norm": 0.5883290870001415, "learning_rate": 6.848191939014909e-05, "loss": 0.3594, "step": 9341 }, { "epoch": 4.3962352941176475, "grad_norm": 0.3900624756303962, "learning_rate": 6.844512507402465e-05, "loss": 0.127, "step": 9342 }, { "epoch": 4.396705882352941, "grad_norm": 0.22032295394515625, "learning_rate": 6.84083377230865e-05, "loss": 0.0302, "step": 9343 }, { "epoch": 4.397176470588235, "grad_norm": 0.3198790758731355, "learning_rate": 6.837155734047652e-05, "loss": 0.1064, "step": 9344 }, { "epoch": 4.397647058823529, "grad_norm": 0.3901902385825157, "learning_rate": 6.833478392933597e-05, "loss": 0.1523, "step": 9345 }, { "epoch": 4.3981176470588235, "grad_norm": 0.2670490810522658, "learning_rate": 6.829801749280537e-05, "loss": 0.1079, "step": 9346 }, { "epoch": 4.398588235294118, "grad_norm": 0.5773058618011315, "learning_rate": 6.826125803402484e-05, "loss": 0.3418, "step": 9347 }, { "epoch": 4.399058823529412, "grad_norm": 0.21469248697693855, "learning_rate": 6.822450555613385e-05, "loss": 0.0381, "step": 9348 }, { "epoch": 4.399529411764706, "grad_norm": 0.5535946723911145, "learning_rate": 6.818776006227114e-05, "loss": 0.2676, "step": 9349 }, { "epoch": 4.4, "grad_norm": 0.607535041121294, "learning_rate": 6.815102155557501e-05, "loss": 0.293, "step": 9350 }, { "epoch": 4.400470588235294, "grad_norm": 0.62085566134953, "learning_rate": 6.811429003918317e-05, "loss": 0.3535, "step": 9351 }, { "epoch": 4.400941176470588, "grad_norm": 0.3222767631297417, "learning_rate": 6.807756551623248e-05, "loss": 0.1245, "step": 9352 }, { "epoch": 4.401411764705882, "grad_norm": 0.6159429610534873, "learning_rate": 6.804084798985964e-05, "loss": 0.373, "step": 9353 }, { "epoch": 4.401882352941176, "grad_norm": 0.19303049998683663, "learning_rate": 6.800413746320034e-05, "loss": 0.0442, "step": 9354 }, { "epoch": 4.402352941176471, "grad_norm": 0.4796715786022741, "learning_rate": 6.796743393938987e-05, "loss": 0.2363, "step": 9355 }, { "epoch": 4.402823529411765, "grad_norm": 0.5109320660028108, "learning_rate": 6.793073742156298e-05, "loss": 0.2871, "step": 9356 }, { "epoch": 4.403294117647059, "grad_norm": 0.44798938205661765, "learning_rate": 6.78940479128536e-05, "loss": 0.3281, "step": 9357 }, { "epoch": 4.403764705882353, "grad_norm": 0.3568457948960714, "learning_rate": 6.785736541639526e-05, "loss": 0.1206, "step": 9358 }, { "epoch": 4.4042352941176475, "grad_norm": 0.22176011792819103, "learning_rate": 6.782068993532085e-05, "loss": 0.0918, "step": 9359 }, { "epoch": 4.404705882352941, "grad_norm": 0.2863782598913893, "learning_rate": 6.778402147276259e-05, "loss": 0.1021, "step": 9360 }, { "epoch": 4.405176470588235, "grad_norm": 0.6616988624173928, "learning_rate": 6.774736003185215e-05, "loss": 0.3242, "step": 9361 }, { "epoch": 4.405647058823529, "grad_norm": 0.8221671612007638, "learning_rate": 6.771070561572067e-05, "loss": 0.4707, "step": 9362 }, { "epoch": 4.4061176470588235, "grad_norm": 0.308185783913737, "learning_rate": 6.767405822749848e-05, "loss": 0.1641, "step": 9363 }, { "epoch": 4.406588235294118, "grad_norm": 0.4836888756177243, "learning_rate": 6.76374178703156e-05, "loss": 0.2734, "step": 9364 }, { "epoch": 4.407058823529412, "grad_norm": 0.26592206216795344, "learning_rate": 6.760078454730115e-05, "loss": 0.0635, "step": 9365 }, { "epoch": 4.407529411764706, "grad_norm": 0.8720292280092303, "learning_rate": 6.756415826158388e-05, "loss": 0.6211, "step": 9366 }, { "epoch": 4.408, "grad_norm": 0.42998002159550813, "learning_rate": 6.752753901629189e-05, "loss": 0.2715, "step": 9367 }, { "epoch": 4.408470588235295, "grad_norm": 0.404836465662432, "learning_rate": 6.749092681455255e-05, "loss": 0.1216, "step": 9368 }, { "epoch": 4.408941176470588, "grad_norm": 0.6104215024565696, "learning_rate": 6.745432165949274e-05, "loss": 0.3613, "step": 9369 }, { "epoch": 4.409411764705882, "grad_norm": 0.6052260278647597, "learning_rate": 6.741772355423882e-05, "loss": 0.3418, "step": 9370 }, { "epoch": 4.409882352941176, "grad_norm": 0.4009939094019624, "learning_rate": 6.73811325019163e-05, "loss": 0.1865, "step": 9371 }, { "epoch": 4.410352941176471, "grad_norm": 0.16042200698270753, "learning_rate": 6.734454850565031e-05, "loss": 0.0649, "step": 9372 }, { "epoch": 4.410823529411765, "grad_norm": 0.5590709093522105, "learning_rate": 6.730797156856537e-05, "loss": 0.2676, "step": 9373 }, { "epoch": 4.411294117647059, "grad_norm": 0.5639837646268923, "learning_rate": 6.727140169378521e-05, "loss": 0.3516, "step": 9374 }, { "epoch": 4.411764705882353, "grad_norm": 0.44116590653359405, "learning_rate": 6.72348388844331e-05, "loss": 0.208, "step": 9375 }, { "epoch": 4.412235294117647, "grad_norm": 0.6357890859850766, "learning_rate": 6.719828314363178e-05, "loss": 0.373, "step": 9376 }, { "epoch": 4.412705882352941, "grad_norm": 0.5318072979107712, "learning_rate": 6.716173447450317e-05, "loss": 0.2812, "step": 9377 }, { "epoch": 4.413176470588235, "grad_norm": 0.5192131874945483, "learning_rate": 6.712519288016875e-05, "loss": 0.3652, "step": 9378 }, { "epoch": 4.413647058823529, "grad_norm": 0.27351478195043416, "learning_rate": 6.70886583637494e-05, "loss": 0.0962, "step": 9379 }, { "epoch": 4.4141176470588235, "grad_norm": 0.3695901473522507, "learning_rate": 6.705213092836527e-05, "loss": 0.1553, "step": 9380 }, { "epoch": 4.414588235294118, "grad_norm": 0.25012880859399866, "learning_rate": 6.701561057713599e-05, "loss": 0.0767, "step": 9381 }, { "epoch": 4.415058823529412, "grad_norm": 0.690775184408281, "learning_rate": 6.697909731318069e-05, "loss": 0.2285, "step": 9382 }, { "epoch": 4.415529411764706, "grad_norm": 0.4844536026093978, "learning_rate": 6.694259113961761e-05, "loss": 0.2559, "step": 9383 }, { "epoch": 4.416, "grad_norm": 0.495313901637238, "learning_rate": 6.690609205956464e-05, "loss": 0.1924, "step": 9384 }, { "epoch": 4.416470588235294, "grad_norm": 0.6280752353383184, "learning_rate": 6.686960007613906e-05, "loss": 0.2949, "step": 9385 }, { "epoch": 4.416941176470588, "grad_norm": 0.633322896658181, "learning_rate": 6.683311519245733e-05, "loss": 0.4414, "step": 9386 }, { "epoch": 4.417411764705882, "grad_norm": 0.5882989930086744, "learning_rate": 6.679663741163553e-05, "loss": 0.4531, "step": 9387 }, { "epoch": 4.417882352941176, "grad_norm": 0.9249664245039179, "learning_rate": 6.676016673678896e-05, "loss": 0.5547, "step": 9388 }, { "epoch": 4.418352941176471, "grad_norm": 0.4690244341521223, "learning_rate": 6.672370317103245e-05, "loss": 0.1543, "step": 9389 }, { "epoch": 4.418823529411765, "grad_norm": 0.30251470186597124, "learning_rate": 6.668724671748018e-05, "loss": 0.1592, "step": 9390 }, { "epoch": 4.419294117647059, "grad_norm": 0.8579066916704083, "learning_rate": 6.665079737924566e-05, "loss": 0.7617, "step": 9391 }, { "epoch": 4.419764705882353, "grad_norm": 0.4290680252776103, "learning_rate": 6.661435515944186e-05, "loss": 0.3457, "step": 9392 }, { "epoch": 4.4202352941176475, "grad_norm": 0.81596211210525, "learning_rate": 6.65779200611812e-05, "loss": 0.4277, "step": 9393 }, { "epoch": 4.420705882352941, "grad_norm": 0.5110491918777127, "learning_rate": 6.654149208757529e-05, "loss": 0.2002, "step": 9394 }, { "epoch": 4.421176470588235, "grad_norm": 0.4741573464770203, "learning_rate": 6.65050712417353e-05, "loss": 0.2324, "step": 9395 }, { "epoch": 4.421647058823529, "grad_norm": 0.3343046833105304, "learning_rate": 6.646865752677186e-05, "loss": 0.124, "step": 9396 }, { "epoch": 4.4221176470588235, "grad_norm": 0.42030893770626626, "learning_rate": 6.643225094579462e-05, "loss": 0.1592, "step": 9397 }, { "epoch": 4.422588235294118, "grad_norm": 0.5581559145181765, "learning_rate": 6.63958515019132e-05, "loss": 0.2715, "step": 9398 }, { "epoch": 4.423058823529412, "grad_norm": 0.6968127020519523, "learning_rate": 6.635945919823607e-05, "loss": 0.4375, "step": 9399 }, { "epoch": 4.423529411764706, "grad_norm": 0.23800722308282607, "learning_rate": 6.632307403787138e-05, "loss": 0.0513, "step": 9400 }, { "epoch": 4.424, "grad_norm": 0.34233642077605714, "learning_rate": 6.628669602392665e-05, "loss": 0.1221, "step": 9401 }, { "epoch": 4.424470588235294, "grad_norm": 0.6316210273927121, "learning_rate": 6.625032515950865e-05, "loss": 0.3047, "step": 9402 }, { "epoch": 4.424941176470588, "grad_norm": 0.2731806458844101, "learning_rate": 6.621396144772365e-05, "loss": 0.123, "step": 9403 }, { "epoch": 4.425411764705882, "grad_norm": 0.5121376113235985, "learning_rate": 6.617760489167738e-05, "loss": 0.2598, "step": 9404 }, { "epoch": 4.425882352941176, "grad_norm": 0.42572753598398355, "learning_rate": 6.614125549447475e-05, "loss": 0.291, "step": 9405 }, { "epoch": 4.426352941176471, "grad_norm": 0.4339501843028379, "learning_rate": 6.610491325922022e-05, "loss": 0.2129, "step": 9406 }, { "epoch": 4.426823529411765, "grad_norm": 0.32614093213086315, "learning_rate": 6.606857818901765e-05, "loss": 0.1406, "step": 9407 }, { "epoch": 4.427294117647059, "grad_norm": 0.9393967595011568, "learning_rate": 6.603225028697013e-05, "loss": 0.7891, "step": 9408 }, { "epoch": 4.427764705882353, "grad_norm": 0.2687979682491395, "learning_rate": 6.59959295561803e-05, "loss": 0.0723, "step": 9409 }, { "epoch": 4.428235294117647, "grad_norm": 0.4587240443019533, "learning_rate": 6.595961599975016e-05, "loss": 0.2969, "step": 9410 }, { "epoch": 4.428705882352941, "grad_norm": 0.46843459289328854, "learning_rate": 6.5923309620781e-05, "loss": 0.1943, "step": 9411 }, { "epoch": 4.429176470588235, "grad_norm": 0.523665974521767, "learning_rate": 6.588701042237358e-05, "loss": 0.2559, "step": 9412 }, { "epoch": 4.429647058823529, "grad_norm": 0.6427215294931894, "learning_rate": 6.585071840762809e-05, "loss": 0.2432, "step": 9413 }, { "epoch": 4.4301176470588235, "grad_norm": 0.4530684414601278, "learning_rate": 6.581443357964395e-05, "loss": 0.1934, "step": 9414 }, { "epoch": 4.430588235294118, "grad_norm": 0.39816178274985237, "learning_rate": 6.577815594152017e-05, "loss": 0.1289, "step": 9415 }, { "epoch": 4.431058823529412, "grad_norm": 0.4114559769473215, "learning_rate": 6.574188549635493e-05, "loss": 0.2061, "step": 9416 }, { "epoch": 4.431529411764706, "grad_norm": 0.6105286874874106, "learning_rate": 6.570562224724593e-05, "loss": 0.2754, "step": 9417 }, { "epoch": 4.432, "grad_norm": 0.39497467685716664, "learning_rate": 6.566936619729033e-05, "loss": 0.291, "step": 9418 }, { "epoch": 4.432470588235294, "grad_norm": 0.6642261385261338, "learning_rate": 6.563311734958444e-05, "loss": 0.4473, "step": 9419 }, { "epoch": 4.432941176470588, "grad_norm": 0.27792207279201564, "learning_rate": 6.559687570722413e-05, "loss": 0.0806, "step": 9420 }, { "epoch": 4.433411764705882, "grad_norm": 0.3210644307649734, "learning_rate": 6.556064127330467e-05, "loss": 0.1387, "step": 9421 }, { "epoch": 4.433882352941176, "grad_norm": 0.391080302794867, "learning_rate": 6.552441405092058e-05, "loss": 0.1279, "step": 9422 }, { "epoch": 4.434352941176471, "grad_norm": 0.45750149901302223, "learning_rate": 6.54881940431659e-05, "loss": 0.1592, "step": 9423 }, { "epoch": 4.434823529411765, "grad_norm": 0.4255945711612097, "learning_rate": 6.5451981253134e-05, "loss": 0.165, "step": 9424 }, { "epoch": 4.435294117647059, "grad_norm": 0.37424807074536476, "learning_rate": 6.541577568391758e-05, "loss": 0.1079, "step": 9425 }, { "epoch": 4.435764705882353, "grad_norm": 0.35490241523363414, "learning_rate": 6.537957733860879e-05, "loss": 0.1348, "step": 9426 }, { "epoch": 4.4362352941176475, "grad_norm": 0.7565625739841163, "learning_rate": 6.53433862202992e-05, "loss": 0.373, "step": 9427 }, { "epoch": 4.436705882352941, "grad_norm": 0.259562443795505, "learning_rate": 6.530720233207962e-05, "loss": 0.0767, "step": 9428 }, { "epoch": 4.437176470588235, "grad_norm": 0.49562598560756665, "learning_rate": 6.527102567704036e-05, "loss": 0.165, "step": 9429 }, { "epoch": 4.437647058823529, "grad_norm": 0.1048571674890776, "learning_rate": 6.523485625827119e-05, "loss": 0.0238, "step": 9430 }, { "epoch": 4.4381176470588235, "grad_norm": 0.3750900902186328, "learning_rate": 6.519869407886098e-05, "loss": 0.2031, "step": 9431 }, { "epoch": 4.438588235294118, "grad_norm": 0.5545907994409138, "learning_rate": 6.516253914189825e-05, "loss": 0.377, "step": 9432 }, { "epoch": 4.439058823529412, "grad_norm": 0.5784494495236259, "learning_rate": 6.512639145047086e-05, "loss": 0.3867, "step": 9433 }, { "epoch": 4.439529411764706, "grad_norm": 0.6296233963562151, "learning_rate": 6.509025100766591e-05, "loss": 0.5195, "step": 9434 }, { "epoch": 4.44, "grad_norm": 0.6417542606343897, "learning_rate": 6.505411781656997e-05, "loss": 0.4766, "step": 9435 }, { "epoch": 4.440470588235294, "grad_norm": 0.32031848738210245, "learning_rate": 6.50179918802691e-05, "loss": 0.1396, "step": 9436 }, { "epoch": 4.440941176470588, "grad_norm": 0.617263954006819, "learning_rate": 6.49818732018485e-05, "loss": 0.3242, "step": 9437 }, { "epoch": 4.441411764705882, "grad_norm": 0.2273064805810352, "learning_rate": 6.494576178439298e-05, "loss": 0.0835, "step": 9438 }, { "epoch": 4.441882352941176, "grad_norm": 0.5257498217978026, "learning_rate": 6.490965763098654e-05, "loss": 0.3262, "step": 9439 }, { "epoch": 4.442352941176471, "grad_norm": 0.18057790512170066, "learning_rate": 6.487356074471272e-05, "loss": 0.063, "step": 9440 }, { "epoch": 4.442823529411765, "grad_norm": 0.5485214969658813, "learning_rate": 6.483747112865436e-05, "loss": 0.2471, "step": 9441 }, { "epoch": 4.443294117647059, "grad_norm": 0.24016612477216448, "learning_rate": 6.480138878589368e-05, "loss": 0.0659, "step": 9442 }, { "epoch": 4.443764705882353, "grad_norm": 0.6500412629691988, "learning_rate": 6.476531371951229e-05, "loss": 0.3594, "step": 9443 }, { "epoch": 4.444235294117647, "grad_norm": 0.3487012161819711, "learning_rate": 6.472924593259122e-05, "loss": 0.1318, "step": 9444 }, { "epoch": 4.444705882352941, "grad_norm": 0.679095544193463, "learning_rate": 6.469318542821078e-05, "loss": 0.4121, "step": 9445 }, { "epoch": 4.445176470588235, "grad_norm": 0.5532057372011356, "learning_rate": 6.465713220945075e-05, "loss": 0.2949, "step": 9446 }, { "epoch": 4.445647058823529, "grad_norm": 0.5993448107866345, "learning_rate": 6.46210862793902e-05, "loss": 0.2695, "step": 9447 }, { "epoch": 4.4461176470588235, "grad_norm": 0.5941109894708212, "learning_rate": 6.458504764110764e-05, "loss": 0.3848, "step": 9448 }, { "epoch": 4.446588235294118, "grad_norm": 0.36476144684349604, "learning_rate": 6.454901629768105e-05, "loss": 0.1328, "step": 9449 }, { "epoch": 4.447058823529412, "grad_norm": 0.7906652027319938, "learning_rate": 6.451299225218754e-05, "loss": 0.4609, "step": 9450 }, { "epoch": 4.447529411764706, "grad_norm": 0.5440940402671827, "learning_rate": 6.447697550770381e-05, "loss": 0.2676, "step": 9451 }, { "epoch": 4.448, "grad_norm": 0.586544968816369, "learning_rate": 6.44409660673059e-05, "loss": 0.2354, "step": 9452 }, { "epoch": 4.448470588235294, "grad_norm": 0.24627391711941504, "learning_rate": 6.440496393406912e-05, "loss": 0.0466, "step": 9453 }, { "epoch": 4.448941176470588, "grad_norm": 0.5328248798700002, "learning_rate": 6.436896911106824e-05, "loss": 0.2559, "step": 9454 }, { "epoch": 4.449411764705882, "grad_norm": 0.7228094209478513, "learning_rate": 6.433298160137749e-05, "loss": 0.5117, "step": 9455 }, { "epoch": 4.449882352941176, "grad_norm": 0.3664785013916711, "learning_rate": 6.429700140807023e-05, "loss": 0.2139, "step": 9456 }, { "epoch": 4.450352941176471, "grad_norm": 0.816128036464765, "learning_rate": 6.426102853421942e-05, "loss": 0.5938, "step": 9457 }, { "epoch": 4.450823529411765, "grad_norm": 0.47224030008792844, "learning_rate": 6.42250629828974e-05, "loss": 0.2734, "step": 9458 }, { "epoch": 4.451294117647059, "grad_norm": 0.46076133327615554, "learning_rate": 6.418910475717565e-05, "loss": 0.3125, "step": 9459 }, { "epoch": 4.451764705882353, "grad_norm": 0.6775062795728537, "learning_rate": 6.415315386012525e-05, "loss": 0.4082, "step": 9460 }, { "epoch": 4.4522352941176475, "grad_norm": 0.40271448334311505, "learning_rate": 6.411721029481665e-05, "loss": 0.1729, "step": 9461 }, { "epoch": 4.452705882352941, "grad_norm": 0.23202334826894852, "learning_rate": 6.40812740643195e-05, "loss": 0.0757, "step": 9462 }, { "epoch": 4.453176470588235, "grad_norm": 0.6860677232280071, "learning_rate": 6.404534517170296e-05, "loss": 0.3398, "step": 9463 }, { "epoch": 4.453647058823529, "grad_norm": 0.6587944040330166, "learning_rate": 6.40094236200356e-05, "loss": 0.2637, "step": 9464 }, { "epoch": 4.4541176470588235, "grad_norm": 0.6170643205974364, "learning_rate": 6.397350941238519e-05, "loss": 0.3711, "step": 9465 }, { "epoch": 4.454588235294118, "grad_norm": 0.4405716145395754, "learning_rate": 6.39376025518191e-05, "loss": 0.2832, "step": 9466 }, { "epoch": 4.455058823529412, "grad_norm": 0.5786237593223229, "learning_rate": 6.390170304140382e-05, "loss": 0.2236, "step": 9467 }, { "epoch": 4.455529411764706, "grad_norm": 0.2550199618731656, "learning_rate": 6.386581088420542e-05, "loss": 0.064, "step": 9468 }, { "epoch": 4.456, "grad_norm": 0.3539803317809865, "learning_rate": 6.382992608328929e-05, "loss": 0.0879, "step": 9469 }, { "epoch": 4.456470588235294, "grad_norm": 0.7705082194531818, "learning_rate": 6.37940486417201e-05, "loss": 0.5625, "step": 9470 }, { "epoch": 4.456941176470588, "grad_norm": 0.8957710677137316, "learning_rate": 6.375817856256199e-05, "loss": 0.4766, "step": 9471 }, { "epoch": 4.457411764705882, "grad_norm": 0.5965822018950325, "learning_rate": 6.37223158488785e-05, "loss": 0.6836, "step": 9472 }, { "epoch": 4.457882352941176, "grad_norm": 0.3746011100706906, "learning_rate": 6.368646050373236e-05, "loss": 0.1582, "step": 9473 }, { "epoch": 4.458352941176471, "grad_norm": 0.9383110464556323, "learning_rate": 6.365061253018586e-05, "loss": 0.918, "step": 9474 }, { "epoch": 4.458823529411765, "grad_norm": 0.3612485263008603, "learning_rate": 6.361477193130065e-05, "loss": 0.21, "step": 9475 }, { "epoch": 4.459294117647059, "grad_norm": 0.33099074450264165, "learning_rate": 6.35789387101376e-05, "loss": 0.0674, "step": 9476 }, { "epoch": 4.459764705882353, "grad_norm": 0.3468628886440834, "learning_rate": 6.354311286975705e-05, "loss": 0.1079, "step": 9477 }, { "epoch": 4.460235294117647, "grad_norm": 0.5501577789689402, "learning_rate": 6.35072944132188e-05, "loss": 0.2578, "step": 9478 }, { "epoch": 4.460705882352941, "grad_norm": 0.3256686879449916, "learning_rate": 6.34714833435818e-05, "loss": 0.0757, "step": 9479 }, { "epoch": 4.461176470588235, "grad_norm": 0.5707136309106693, "learning_rate": 6.343567966390452e-05, "loss": 0.3262, "step": 9480 }, { "epoch": 4.461647058823529, "grad_norm": 0.6728680640476162, "learning_rate": 6.339988337724487e-05, "loss": 0.3418, "step": 9481 }, { "epoch": 4.4621176470588235, "grad_norm": 0.8175218994389303, "learning_rate": 6.336409448665988e-05, "loss": 0.2256, "step": 9482 }, { "epoch": 4.462588235294118, "grad_norm": 0.5487465764977887, "learning_rate": 6.332831299520617e-05, "loss": 0.2891, "step": 9483 }, { "epoch": 4.463058823529412, "grad_norm": 0.35022323841719216, "learning_rate": 6.32925389059397e-05, "loss": 0.1719, "step": 9484 }, { "epoch": 4.463529411764706, "grad_norm": 0.40651858301696175, "learning_rate": 6.325677222191568e-05, "loss": 0.2373, "step": 9485 }, { "epoch": 4.464, "grad_norm": 0.7426302593189008, "learning_rate": 6.322101294618875e-05, "loss": 0.3828, "step": 9486 }, { "epoch": 4.464470588235294, "grad_norm": 0.7105046478664997, "learning_rate": 6.318526108181299e-05, "loss": 0.4766, "step": 9487 }, { "epoch": 4.464941176470588, "grad_norm": 0.5358364829488416, "learning_rate": 6.314951663184175e-05, "loss": 0.332, "step": 9488 }, { "epoch": 4.465411764705882, "grad_norm": 0.622386457896282, "learning_rate": 6.311377959932783e-05, "loss": 0.3594, "step": 9489 }, { "epoch": 4.465882352941176, "grad_norm": 0.4264135407036143, "learning_rate": 6.307804998732328e-05, "loss": 0.127, "step": 9490 }, { "epoch": 4.466352941176471, "grad_norm": 0.46752338011583205, "learning_rate": 6.304232779887958e-05, "loss": 0.1562, "step": 9491 }, { "epoch": 4.466823529411765, "grad_norm": 0.43153094693698524, "learning_rate": 6.300661303704768e-05, "loss": 0.1299, "step": 9492 }, { "epoch": 4.467294117647059, "grad_norm": 0.24693114209704506, "learning_rate": 6.297090570487766e-05, "loss": 0.0981, "step": 9493 }, { "epoch": 4.467764705882353, "grad_norm": 0.7167042317892033, "learning_rate": 6.293520580541925e-05, "loss": 0.5039, "step": 9494 }, { "epoch": 4.4682352941176475, "grad_norm": 0.4223022513185077, "learning_rate": 6.289951334172123e-05, "loss": 0.2715, "step": 9495 }, { "epoch": 4.468705882352941, "grad_norm": 0.6878453794494347, "learning_rate": 6.2863828316832e-05, "loss": 0.4785, "step": 9496 }, { "epoch": 4.469176470588235, "grad_norm": 0.3089160636435543, "learning_rate": 6.282815073379932e-05, "loss": 0.1221, "step": 9497 }, { "epoch": 4.469647058823529, "grad_norm": 0.7490626102226761, "learning_rate": 6.279248059567006e-05, "loss": 0.373, "step": 9498 }, { "epoch": 4.4701176470588235, "grad_norm": 0.5995547147504471, "learning_rate": 6.275681790549072e-05, "loss": 0.2988, "step": 9499 }, { "epoch": 4.470588235294118, "grad_norm": 0.07803426199813865, "learning_rate": 6.27211626663071e-05, "loss": 0.0223, "step": 9500 }, { "epoch": 4.471058823529412, "grad_norm": 0.5003635751686201, "learning_rate": 6.268551488116423e-05, "loss": 0.3398, "step": 9501 }, { "epoch": 4.471529411764706, "grad_norm": 0.38204368024529345, "learning_rate": 6.264987455310667e-05, "loss": 0.105, "step": 9502 }, { "epoch": 4.4719999999999995, "grad_norm": 0.9263995202589016, "learning_rate": 6.261424168517832e-05, "loss": 0.5391, "step": 9503 }, { "epoch": 4.472470588235294, "grad_norm": 0.5218029797190058, "learning_rate": 6.257861628042233e-05, "loss": 0.1523, "step": 9504 }, { "epoch": 4.472941176470588, "grad_norm": 0.7162648760054842, "learning_rate": 6.254299834188128e-05, "loss": 0.3477, "step": 9505 }, { "epoch": 4.473411764705882, "grad_norm": 0.31856146519707096, "learning_rate": 6.25073878725972e-05, "loss": 0.1436, "step": 9506 }, { "epoch": 4.473882352941176, "grad_norm": 0.759592814323525, "learning_rate": 6.24717848756113e-05, "loss": 0.4648, "step": 9507 }, { "epoch": 4.474352941176471, "grad_norm": 0.6111971376741397, "learning_rate": 6.24361893539643e-05, "loss": 0.4199, "step": 9508 }, { "epoch": 4.474823529411765, "grad_norm": 0.7324755769762313, "learning_rate": 6.240060131069628e-05, "loss": 0.4766, "step": 9509 }, { "epoch": 4.475294117647059, "grad_norm": 0.844613179117576, "learning_rate": 6.23650207488465e-05, "loss": 0.4121, "step": 9510 }, { "epoch": 4.475764705882353, "grad_norm": 0.48846206925479685, "learning_rate": 6.232944767145384e-05, "loss": 0.3359, "step": 9511 }, { "epoch": 4.476235294117647, "grad_norm": 0.14281119523322203, "learning_rate": 6.229388208155639e-05, "loss": 0.0339, "step": 9512 }, { "epoch": 4.476705882352941, "grad_norm": 0.8028321338768528, "learning_rate": 6.225832398219158e-05, "loss": 0.5703, "step": 9513 }, { "epoch": 4.477176470588235, "grad_norm": 0.3071884780422912, "learning_rate": 6.222277337639624e-05, "loss": 0.0815, "step": 9514 }, { "epoch": 4.477647058823529, "grad_norm": 0.37725592301485517, "learning_rate": 6.218723026720668e-05, "loss": 0.1436, "step": 9515 }, { "epoch": 4.4781176470588235, "grad_norm": 0.6365589863099157, "learning_rate": 6.21516946576583e-05, "loss": 0.3281, "step": 9516 }, { "epoch": 4.478588235294118, "grad_norm": 0.6901176029024038, "learning_rate": 6.211616655078616e-05, "loss": 0.3477, "step": 9517 }, { "epoch": 4.479058823529412, "grad_norm": 0.28909178049098594, "learning_rate": 6.208064594962441e-05, "loss": 0.0947, "step": 9518 }, { "epoch": 4.479529411764706, "grad_norm": 0.5186410454751322, "learning_rate": 6.204513285720673e-05, "loss": 0.2773, "step": 9519 }, { "epoch": 4.48, "grad_norm": 0.6621134166657348, "learning_rate": 6.200962727656616e-05, "loss": 0.4355, "step": 9520 }, { "epoch": 4.480470588235294, "grad_norm": 0.5569036794223191, "learning_rate": 6.197412921073498e-05, "loss": 0.167, "step": 9521 }, { "epoch": 4.480941176470588, "grad_norm": 0.5380104432511702, "learning_rate": 6.193863866274492e-05, "loss": 0.3145, "step": 9522 }, { "epoch": 4.481411764705882, "grad_norm": 0.4170371137968993, "learning_rate": 6.190315563562711e-05, "loss": 0.1758, "step": 9523 }, { "epoch": 4.481882352941176, "grad_norm": 0.6624832610714912, "learning_rate": 6.186768013241188e-05, "loss": 0.3867, "step": 9524 }, { "epoch": 4.482352941176471, "grad_norm": 0.7118587519872324, "learning_rate": 6.183221215612904e-05, "loss": 0.3418, "step": 9525 }, { "epoch": 4.482823529411765, "grad_norm": 0.252797874760666, "learning_rate": 6.179675170980779e-05, "loss": 0.083, "step": 9526 }, { "epoch": 4.483294117647059, "grad_norm": 0.40687278900304785, "learning_rate": 6.176129879647654e-05, "loss": 0.2119, "step": 9527 }, { "epoch": 4.483764705882353, "grad_norm": 0.3773718595949566, "learning_rate": 6.172585341916317e-05, "loss": 0.1426, "step": 9528 }, { "epoch": 4.484235294117647, "grad_norm": 0.21654040763340138, "learning_rate": 6.169041558089496e-05, "loss": 0.0601, "step": 9529 }, { "epoch": 4.484705882352941, "grad_norm": 0.7070386145702742, "learning_rate": 6.16549852846983e-05, "loss": 0.5234, "step": 9530 }, { "epoch": 4.485176470588235, "grad_norm": 0.5149689020132039, "learning_rate": 6.161956253359935e-05, "loss": 0.3066, "step": 9531 }, { "epoch": 4.485647058823529, "grad_norm": 0.383478342910587, "learning_rate": 6.15841473306232e-05, "loss": 0.2051, "step": 9532 }, { "epoch": 4.4861176470588235, "grad_norm": 0.5671630044504119, "learning_rate": 6.154873967879455e-05, "loss": 0.2969, "step": 9533 }, { "epoch": 4.486588235294118, "grad_norm": 0.2138202823243058, "learning_rate": 6.151333958113743e-05, "loss": 0.0615, "step": 9534 }, { "epoch": 4.487058823529412, "grad_norm": 0.6316521972434687, "learning_rate": 6.147794704067508e-05, "loss": 0.3477, "step": 9535 }, { "epoch": 4.487529411764706, "grad_norm": 0.4803515681839956, "learning_rate": 6.144256206043025e-05, "loss": 0.3301, "step": 9536 }, { "epoch": 4.4879999999999995, "grad_norm": 0.22985732279088547, "learning_rate": 6.140718464342505e-05, "loss": 0.0693, "step": 9537 }, { "epoch": 4.488470588235294, "grad_norm": 0.3255148436497507, "learning_rate": 6.137181479268077e-05, "loss": 0.125, "step": 9538 }, { "epoch": 4.488941176470588, "grad_norm": 0.39068895397103776, "learning_rate": 6.133645251121823e-05, "loss": 0.1387, "step": 9539 }, { "epoch": 4.489411764705882, "grad_norm": 0.3839375448798466, "learning_rate": 6.130109780205759e-05, "loss": 0.1465, "step": 9540 }, { "epoch": 4.489882352941176, "grad_norm": 0.6537299233167277, "learning_rate": 6.126575066821821e-05, "loss": 0.4023, "step": 9541 }, { "epoch": 4.490352941176471, "grad_norm": 0.39485740460395496, "learning_rate": 6.123041111271896e-05, "loss": 0.1201, "step": 9542 }, { "epoch": 4.490823529411765, "grad_norm": 0.6169170795967465, "learning_rate": 6.119507913857807e-05, "loss": 0.3672, "step": 9543 }, { "epoch": 4.491294117647059, "grad_norm": 0.36832338509685025, "learning_rate": 6.115975474881296e-05, "loss": 0.1865, "step": 9544 }, { "epoch": 4.491764705882353, "grad_norm": 0.2985744015176895, "learning_rate": 6.11244379464406e-05, "loss": 0.1108, "step": 9545 }, { "epoch": 4.492235294117647, "grad_norm": 0.4274985721237089, "learning_rate": 6.108912873447712e-05, "loss": 0.1069, "step": 9546 }, { "epoch": 4.492705882352941, "grad_norm": 0.651251612945121, "learning_rate": 6.105382711593814e-05, "loss": 0.4023, "step": 9547 }, { "epoch": 4.493176470588235, "grad_norm": 0.5279053949339005, "learning_rate": 6.101853309383866e-05, "loss": 0.3418, "step": 9548 }, { "epoch": 4.493647058823529, "grad_norm": 0.683610479692629, "learning_rate": 6.0983246671192866e-05, "loss": 0.3945, "step": 9549 }, { "epoch": 4.4941176470588236, "grad_norm": 0.5124603870175883, "learning_rate": 6.0947967851014405e-05, "loss": 0.3535, "step": 9550 }, { "epoch": 4.494588235294118, "grad_norm": 0.44139123091021454, "learning_rate": 6.091269663631636e-05, "loss": 0.2109, "step": 9551 }, { "epoch": 4.495058823529412, "grad_norm": 0.6311392787023903, "learning_rate": 6.0877433030110935e-05, "loss": 0.3555, "step": 9552 }, { "epoch": 4.495529411764706, "grad_norm": 0.3191421998692287, "learning_rate": 6.084217703540988e-05, "loss": 0.1289, "step": 9553 }, { "epoch": 4.496, "grad_norm": 0.5543250721162771, "learning_rate": 6.0806928655224266e-05, "loss": 0.2539, "step": 9554 }, { "epoch": 4.496470588235294, "grad_norm": 0.5748561934500858, "learning_rate": 6.0771687892564394e-05, "loss": 0.2422, "step": 9555 }, { "epoch": 4.496941176470588, "grad_norm": 0.3225389864960418, "learning_rate": 6.0736454750440025e-05, "loss": 0.1226, "step": 9556 }, { "epoch": 4.497411764705882, "grad_norm": 0.3239313519357596, "learning_rate": 6.070122923186032e-05, "loss": 0.1069, "step": 9557 }, { "epoch": 4.497882352941176, "grad_norm": 0.4510530681942572, "learning_rate": 6.0666011339833584e-05, "loss": 0.1562, "step": 9558 }, { "epoch": 4.498352941176471, "grad_norm": 0.5787049071718988, "learning_rate": 6.063080107736767e-05, "loss": 0.3125, "step": 9559 }, { "epoch": 4.498823529411765, "grad_norm": 0.07101064713324731, "learning_rate": 6.0595598447469726e-05, "loss": 0.0106, "step": 9560 }, { "epoch": 4.499294117647059, "grad_norm": 0.7854180440344583, "learning_rate": 6.056040345314618e-05, "loss": 0.4141, "step": 9561 }, { "epoch": 4.499764705882353, "grad_norm": 0.5169276714574025, "learning_rate": 6.0525216097402865e-05, "loss": 0.2617, "step": 9562 }, { "epoch": 4.5002352941176476, "grad_norm": 0.5156331586118386, "learning_rate": 6.049003638324501e-05, "loss": 0.3184, "step": 9563 }, { "epoch": 4.500705882352941, "grad_norm": 0.575904848763304, "learning_rate": 6.0454864313677054e-05, "loss": 0.3359, "step": 9564 }, { "epoch": 4.501176470588235, "grad_norm": 0.5018467887564894, "learning_rate": 6.041969989170289e-05, "loss": 0.1895, "step": 9565 }, { "epoch": 4.501647058823529, "grad_norm": 0.6840745944857116, "learning_rate": 6.0384543120325796e-05, "loss": 0.3574, "step": 9566 }, { "epoch": 4.502117647058824, "grad_norm": 0.590757610122789, "learning_rate": 6.034939400254824e-05, "loss": 0.167, "step": 9567 }, { "epoch": 4.502588235294118, "grad_norm": 0.6585285153181272, "learning_rate": 6.031425254137222e-05, "loss": 0.3164, "step": 9568 }, { "epoch": 4.503058823529412, "grad_norm": 0.16714777058966315, "learning_rate": 6.0279118739798874e-05, "loss": 0.0417, "step": 9569 }, { "epoch": 4.503529411764706, "grad_norm": 0.2268893453743229, "learning_rate": 6.0243992600828875e-05, "loss": 0.0957, "step": 9570 }, { "epoch": 4.504, "grad_norm": 0.4037479678821848, "learning_rate": 6.020887412746221e-05, "loss": 0.1309, "step": 9571 }, { "epoch": 4.504470588235294, "grad_norm": 0.10336086900007442, "learning_rate": 6.017376332269807e-05, "loss": 0.024, "step": 9572 }, { "epoch": 4.504941176470588, "grad_norm": 0.3840780976088968, "learning_rate": 6.0138660189535126e-05, "loss": 0.166, "step": 9573 }, { "epoch": 4.505411764705882, "grad_norm": 0.46123219090686235, "learning_rate": 6.0103564730971375e-05, "loss": 0.1895, "step": 9574 }, { "epoch": 4.5058823529411764, "grad_norm": 0.6757687797564835, "learning_rate": 6.006847695000415e-05, "loss": 0.3535, "step": 9575 }, { "epoch": 4.506352941176471, "grad_norm": 0.3954012356091031, "learning_rate": 6.0033396849630136e-05, "loss": 0.1602, "step": 9576 }, { "epoch": 4.506823529411765, "grad_norm": 0.6450334484374596, "learning_rate": 5.999832443284527e-05, "loss": 0.3809, "step": 9577 }, { "epoch": 4.507294117647059, "grad_norm": 0.46443069811017584, "learning_rate": 5.9963259702644963e-05, "loss": 0.2031, "step": 9578 }, { "epoch": 4.507764705882353, "grad_norm": 0.4706671060710227, "learning_rate": 5.992820266202394e-05, "loss": 0.2578, "step": 9579 }, { "epoch": 4.508235294117647, "grad_norm": 0.43273499317558495, "learning_rate": 5.989315331397616e-05, "loss": 0.1973, "step": 9580 }, { "epoch": 4.508705882352941, "grad_norm": 0.3964283888859936, "learning_rate": 5.985811166149507e-05, "loss": 0.1543, "step": 9581 }, { "epoch": 4.509176470588235, "grad_norm": 0.24388167671915312, "learning_rate": 5.982307770757343e-05, "loss": 0.0537, "step": 9582 }, { "epoch": 4.509647058823529, "grad_norm": 0.6542869439382238, "learning_rate": 5.978805145520322e-05, "loss": 0.4746, "step": 9583 }, { "epoch": 4.510117647058824, "grad_norm": 0.5997566639741498, "learning_rate": 5.97530329073759e-05, "loss": 0.3535, "step": 9584 }, { "epoch": 4.510588235294118, "grad_norm": 0.5207116438041153, "learning_rate": 5.971802206708229e-05, "loss": 0.2275, "step": 9585 }, { "epoch": 4.511058823529412, "grad_norm": 0.9923710114849971, "learning_rate": 5.968301893731237e-05, "loss": 0.7578, "step": 9586 }, { "epoch": 4.511529411764706, "grad_norm": 0.56051397038145, "learning_rate": 5.964802352105564e-05, "loss": 0.3223, "step": 9587 }, { "epoch": 4.5120000000000005, "grad_norm": 0.4663176029262292, "learning_rate": 5.961303582130092e-05, "loss": 0.1924, "step": 9588 }, { "epoch": 4.512470588235294, "grad_norm": 0.5879499688939314, "learning_rate": 5.957805584103623e-05, "loss": 0.3164, "step": 9589 }, { "epoch": 4.512941176470588, "grad_norm": 0.8724331404464265, "learning_rate": 5.954308358324909e-05, "loss": 0.3496, "step": 9590 }, { "epoch": 4.513411764705882, "grad_norm": 0.3505330328312402, "learning_rate": 5.950811905092636e-05, "loss": 0.2168, "step": 9591 }, { "epoch": 4.5138823529411765, "grad_norm": 0.5590041985059039, "learning_rate": 5.947316224705407e-05, "loss": 0.2432, "step": 9592 }, { "epoch": 4.514352941176471, "grad_norm": 0.31698872703808295, "learning_rate": 5.943821317461775e-05, "loss": 0.1406, "step": 9593 }, { "epoch": 4.514823529411765, "grad_norm": 0.5077153811510186, "learning_rate": 5.940327183660228e-05, "loss": 0.3242, "step": 9594 }, { "epoch": 4.515294117647059, "grad_norm": 0.4526278941310566, "learning_rate": 5.936833823599172e-05, "loss": 0.2441, "step": 9595 }, { "epoch": 4.5157647058823525, "grad_norm": 0.5022154678447349, "learning_rate": 5.933341237576968e-05, "loss": 0.2354, "step": 9596 }, { "epoch": 4.516235294117647, "grad_norm": 0.25344819199733903, "learning_rate": 5.9298494258918875e-05, "loss": 0.0854, "step": 9597 }, { "epoch": 4.516705882352941, "grad_norm": 0.6607222573224222, "learning_rate": 5.926358388842156e-05, "loss": 0.2812, "step": 9598 }, { "epoch": 4.517176470588235, "grad_norm": 0.5669591201872408, "learning_rate": 5.922868126725929e-05, "loss": 0.3066, "step": 9599 }, { "epoch": 4.517647058823529, "grad_norm": 0.4151014239090608, "learning_rate": 5.919378639841281e-05, "loss": 0.2119, "step": 9600 }, { "epoch": 4.518117647058824, "grad_norm": 0.33585005745490143, "learning_rate": 5.915889928486239e-05, "loss": 0.1484, "step": 9601 }, { "epoch": 4.518588235294118, "grad_norm": 0.57328856735966, "learning_rate": 5.912401992958757e-05, "loss": 0.3184, "step": 9602 }, { "epoch": 4.519058823529412, "grad_norm": 0.451187788361841, "learning_rate": 5.908914833556715e-05, "loss": 0.2119, "step": 9603 }, { "epoch": 4.519529411764706, "grad_norm": 0.6478554709140939, "learning_rate": 5.9054284505779396e-05, "loss": 0.4668, "step": 9604 }, { "epoch": 4.52, "grad_norm": 0.3174869332743717, "learning_rate": 5.901942844320187e-05, "loss": 0.105, "step": 9605 }, { "epoch": 4.520470588235294, "grad_norm": 0.4336849692598017, "learning_rate": 5.898458015081136e-05, "loss": 0.2891, "step": 9606 }, { "epoch": 4.520941176470588, "grad_norm": 0.35802679233855295, "learning_rate": 5.8949739631584136e-05, "loss": 0.2334, "step": 9607 }, { "epoch": 4.521411764705882, "grad_norm": 0.8287811258905299, "learning_rate": 5.8914906888495797e-05, "loss": 0.5312, "step": 9608 }, { "epoch": 4.5218823529411765, "grad_norm": 0.45108769627855105, "learning_rate": 5.888008192452112e-05, "loss": 0.2188, "step": 9609 }, { "epoch": 4.522352941176471, "grad_norm": 0.8206557744061994, "learning_rate": 5.8845264742634394e-05, "loss": 0.4805, "step": 9610 }, { "epoch": 4.522823529411765, "grad_norm": 0.5695528574848079, "learning_rate": 5.881045534580922e-05, "loss": 0.1943, "step": 9611 }, { "epoch": 4.523294117647059, "grad_norm": 0.6170247703284018, "learning_rate": 5.87756537370184e-05, "loss": 0.3438, "step": 9612 }, { "epoch": 4.523764705882353, "grad_norm": 0.5874356927486006, "learning_rate": 5.874085991923419e-05, "loss": 0.2949, "step": 9613 }, { "epoch": 4.524235294117647, "grad_norm": 0.37820437348063785, "learning_rate": 5.8706073895428225e-05, "loss": 0.168, "step": 9614 }, { "epoch": 4.524705882352941, "grad_norm": 0.6136699106952527, "learning_rate": 5.867129566857129e-05, "loss": 0.127, "step": 9615 }, { "epoch": 4.525176470588235, "grad_norm": 0.1861776102111986, "learning_rate": 5.8636525241633673e-05, "loss": 0.0427, "step": 9616 }, { "epoch": 4.525647058823529, "grad_norm": 0.22690761325189093, "learning_rate": 5.860176261758499e-05, "loss": 0.0767, "step": 9617 }, { "epoch": 4.526117647058824, "grad_norm": 0.6804997637902043, "learning_rate": 5.8567007799394015e-05, "loss": 0.3203, "step": 9618 }, { "epoch": 4.526588235294118, "grad_norm": 0.6684519254863635, "learning_rate": 5.853226079002906e-05, "loss": 0.4004, "step": 9619 }, { "epoch": 4.527058823529412, "grad_norm": 0.46816194257580174, "learning_rate": 5.849752159245769e-05, "loss": 0.1514, "step": 9620 }, { "epoch": 4.527529411764706, "grad_norm": 0.5204592493666331, "learning_rate": 5.84627902096468e-05, "loss": 0.1699, "step": 9621 }, { "epoch": 4.5280000000000005, "grad_norm": 0.6855394363102942, "learning_rate": 5.842806664456264e-05, "loss": 0.3828, "step": 9622 }, { "epoch": 4.528470588235294, "grad_norm": 0.7200608292940687, "learning_rate": 5.839335090017072e-05, "loss": 0.5078, "step": 9623 }, { "epoch": 4.528941176470588, "grad_norm": 0.3072913481744889, "learning_rate": 5.835864297943594e-05, "loss": 0.1289, "step": 9624 }, { "epoch": 4.529411764705882, "grad_norm": 0.7597924360148114, "learning_rate": 5.8323942885322605e-05, "loss": 0.5156, "step": 9625 }, { "epoch": 4.5298823529411765, "grad_norm": 0.5538578199077148, "learning_rate": 5.828925062079416e-05, "loss": 0.4863, "step": 9626 }, { "epoch": 4.530352941176471, "grad_norm": 0.2497288211315633, "learning_rate": 5.8254566188813615e-05, "loss": 0.1221, "step": 9627 }, { "epoch": 4.530823529411765, "grad_norm": 0.4072740893795518, "learning_rate": 5.821988959234308e-05, "loss": 0.1611, "step": 9628 }, { "epoch": 4.531294117647059, "grad_norm": 0.4582077793175891, "learning_rate": 5.8185220834344124e-05, "loss": 0.2354, "step": 9629 }, { "epoch": 4.5317647058823525, "grad_norm": 0.7090384250125369, "learning_rate": 5.815055991777773e-05, "loss": 0.3496, "step": 9630 }, { "epoch": 4.532235294117647, "grad_norm": 0.3369099914903562, "learning_rate": 5.811590684560398e-05, "loss": 0.1406, "step": 9631 }, { "epoch": 4.532705882352941, "grad_norm": 0.6881675222245927, "learning_rate": 5.808126162078249e-05, "loss": 0.4062, "step": 9632 }, { "epoch": 4.533176470588235, "grad_norm": 0.3574535938002339, "learning_rate": 5.8046624246272146e-05, "loss": 0.1777, "step": 9633 }, { "epoch": 4.533647058823529, "grad_norm": 0.15631643944774964, "learning_rate": 5.801199472503109e-05, "loss": 0.0378, "step": 9634 }, { "epoch": 4.534117647058824, "grad_norm": 0.6052032337759977, "learning_rate": 5.797737306001686e-05, "loss": 0.4492, "step": 9635 }, { "epoch": 4.534588235294118, "grad_norm": 0.9512276767231513, "learning_rate": 5.794275925418641e-05, "loss": 0.3613, "step": 9636 }, { "epoch": 4.535058823529412, "grad_norm": 0.6354071952842348, "learning_rate": 5.79081533104958e-05, "loss": 0.4863, "step": 9637 }, { "epoch": 4.535529411764706, "grad_norm": 0.19347508550973513, "learning_rate": 5.787355523190059e-05, "loss": 0.0825, "step": 9638 }, { "epoch": 4.536, "grad_norm": 0.7723690571054359, "learning_rate": 5.783896502135572e-05, "loss": 0.4824, "step": 9639 }, { "epoch": 4.536470588235294, "grad_norm": 0.5331940081959435, "learning_rate": 5.780438268181522e-05, "loss": 0.2412, "step": 9640 }, { "epoch": 4.536941176470588, "grad_norm": 0.6532044962439537, "learning_rate": 5.776980821623264e-05, "loss": 0.2715, "step": 9641 }, { "epoch": 4.537411764705882, "grad_norm": 0.34712349077118254, "learning_rate": 5.7735241627560905e-05, "loss": 0.1226, "step": 9642 }, { "epoch": 4.5378823529411765, "grad_norm": 0.4820746963956806, "learning_rate": 5.770068291875202e-05, "loss": 0.2393, "step": 9643 }, { "epoch": 4.538352941176471, "grad_norm": 0.6438858362794911, "learning_rate": 5.766613209275754e-05, "loss": 0.2949, "step": 9644 }, { "epoch": 4.538823529411765, "grad_norm": 0.25811741696614876, "learning_rate": 5.763158915252833e-05, "loss": 0.0449, "step": 9645 }, { "epoch": 4.539294117647059, "grad_norm": 0.6064202590860788, "learning_rate": 5.759705410101443e-05, "loss": 0.248, "step": 9646 }, { "epoch": 4.539764705882353, "grad_norm": 0.533331745110573, "learning_rate": 5.756252694116538e-05, "loss": 0.3281, "step": 9647 }, { "epoch": 4.540235294117647, "grad_norm": 0.5901605801218222, "learning_rate": 5.752800767592988e-05, "loss": 0.3867, "step": 9648 }, { "epoch": 4.540705882352941, "grad_norm": 0.3841271172451237, "learning_rate": 5.74934963082561e-05, "loss": 0.1138, "step": 9649 }, { "epoch": 4.541176470588235, "grad_norm": 0.4517875514300741, "learning_rate": 5.745899284109154e-05, "loss": 0.2432, "step": 9650 }, { "epoch": 4.541647058823529, "grad_norm": 0.5798481519370328, "learning_rate": 5.742449727738285e-05, "loss": 0.2363, "step": 9651 }, { "epoch": 4.542117647058824, "grad_norm": 0.5478487027817153, "learning_rate": 5.739000962007616e-05, "loss": 0.2871, "step": 9652 }, { "epoch": 4.542588235294118, "grad_norm": 0.5966911573797413, "learning_rate": 5.7355529872116957e-05, "loss": 0.291, "step": 9653 }, { "epoch": 4.543058823529412, "grad_norm": 0.42276401439447936, "learning_rate": 5.732105803644986e-05, "loss": 0.1943, "step": 9654 }, { "epoch": 4.543529411764706, "grad_norm": 0.40368191790791524, "learning_rate": 5.7286594116019e-05, "loss": 0.1602, "step": 9655 }, { "epoch": 4.5440000000000005, "grad_norm": 0.3041190508218837, "learning_rate": 5.725213811376782e-05, "loss": 0.1875, "step": 9656 }, { "epoch": 4.544470588235294, "grad_norm": 0.3739645015017847, "learning_rate": 5.7217690032638916e-05, "loss": 0.1992, "step": 9657 }, { "epoch": 4.544941176470588, "grad_norm": 0.4804244612423824, "learning_rate": 5.718324987557438e-05, "loss": 0.2754, "step": 9658 }, { "epoch": 4.545411764705882, "grad_norm": 0.7450556465401457, "learning_rate": 5.714881764551561e-05, "loss": 0.1992, "step": 9659 }, { "epoch": 4.5458823529411765, "grad_norm": 0.3893030850210197, "learning_rate": 5.7114393345403194e-05, "loss": 0.1758, "step": 9660 }, { "epoch": 4.546352941176471, "grad_norm": 0.4264638405961057, "learning_rate": 5.707997697817719e-05, "loss": 0.1128, "step": 9661 }, { "epoch": 4.546823529411765, "grad_norm": 0.5629064530014289, "learning_rate": 5.704556854677698e-05, "loss": 0.1611, "step": 9662 }, { "epoch": 4.547294117647059, "grad_norm": 0.746739342493187, "learning_rate": 5.701116805414104e-05, "loss": 0.3418, "step": 9663 }, { "epoch": 4.5477647058823525, "grad_norm": 0.9225462975868485, "learning_rate": 5.6976775503207574e-05, "loss": 0.5742, "step": 9664 }, { "epoch": 4.548235294117647, "grad_norm": 0.33165983128143633, "learning_rate": 5.6942390896913686e-05, "loss": 0.0559, "step": 9665 }, { "epoch": 4.548705882352941, "grad_norm": 0.6205273170264931, "learning_rate": 5.690801423819608e-05, "loss": 0.332, "step": 9666 }, { "epoch": 4.549176470588235, "grad_norm": 0.636942490839442, "learning_rate": 5.687364552999071e-05, "loss": 0.332, "step": 9667 }, { "epoch": 4.549647058823529, "grad_norm": 0.8005063461222142, "learning_rate": 5.683928477523275e-05, "loss": 0.4492, "step": 9668 }, { "epoch": 4.550117647058824, "grad_norm": 0.24503140516906274, "learning_rate": 5.680493197685681e-05, "loss": 0.1045, "step": 9669 }, { "epoch": 4.550588235294118, "grad_norm": 0.34213192642481677, "learning_rate": 5.677058713779687e-05, "loss": 0.0942, "step": 9670 }, { "epoch": 4.551058823529412, "grad_norm": 0.7039742372114693, "learning_rate": 5.6736250260986e-05, "loss": 0.4062, "step": 9671 }, { "epoch": 4.551529411764706, "grad_norm": 0.5117338036507387, "learning_rate": 5.6701921349356826e-05, "loss": 0.1777, "step": 9672 }, { "epoch": 4.552, "grad_norm": 0.440128370358165, "learning_rate": 5.666760040584126e-05, "loss": 0.207, "step": 9673 }, { "epoch": 4.552470588235294, "grad_norm": 0.2701899720419863, "learning_rate": 5.663328743337034e-05, "loss": 0.0603, "step": 9674 }, { "epoch": 4.552941176470588, "grad_norm": 0.4744839390157916, "learning_rate": 5.659898243487463e-05, "loss": 0.1953, "step": 9675 }, { "epoch": 4.553411764705882, "grad_norm": 0.6152292992761501, "learning_rate": 5.6564685413284014e-05, "loss": 0.2139, "step": 9676 }, { "epoch": 4.5538823529411765, "grad_norm": 0.2576690972525783, "learning_rate": 5.6530396371527515e-05, "loss": 0.0535, "step": 9677 }, { "epoch": 4.554352941176471, "grad_norm": 0.5812826682834689, "learning_rate": 5.649611531253366e-05, "loss": 0.2656, "step": 9678 }, { "epoch": 4.554823529411765, "grad_norm": 0.5075838725409808, "learning_rate": 5.646184223923017e-05, "loss": 0.2891, "step": 9679 }, { "epoch": 4.555294117647059, "grad_norm": 0.6367165321983362, "learning_rate": 5.642757715454413e-05, "loss": 0.3262, "step": 9680 }, { "epoch": 4.555764705882353, "grad_norm": 0.3986494576275112, "learning_rate": 5.639332006140203e-05, "loss": 0.1963, "step": 9681 }, { "epoch": 4.556235294117647, "grad_norm": 0.5392148483084694, "learning_rate": 5.635907096272949e-05, "loss": 0.3672, "step": 9682 }, { "epoch": 4.556705882352941, "grad_norm": 0.4549706968750075, "learning_rate": 5.632482986145159e-05, "loss": 0.2041, "step": 9683 }, { "epoch": 4.557176470588235, "grad_norm": 0.5135098136484051, "learning_rate": 5.6290596760492745e-05, "loss": 0.3086, "step": 9684 }, { "epoch": 4.557647058823529, "grad_norm": 0.6369742711630735, "learning_rate": 5.625637166277654e-05, "loss": 0.4102, "step": 9685 }, { "epoch": 4.558117647058824, "grad_norm": 0.427030754061467, "learning_rate": 5.622215457122598e-05, "loss": 0.166, "step": 9686 }, { "epoch": 4.558588235294118, "grad_norm": 0.5494019780652277, "learning_rate": 5.618794548876346e-05, "loss": 0.3848, "step": 9687 }, { "epoch": 4.559058823529412, "grad_norm": 0.15637501948798121, "learning_rate": 5.6153744418310505e-05, "loss": 0.0403, "step": 9688 }, { "epoch": 4.559529411764705, "grad_norm": 0.4597848877564854, "learning_rate": 5.611955136278808e-05, "loss": 0.2402, "step": 9689 }, { "epoch": 4.5600000000000005, "grad_norm": 0.4937632299277097, "learning_rate": 5.608536632511649e-05, "loss": 0.2734, "step": 9690 }, { "epoch": 4.560470588235294, "grad_norm": 0.29575869063857185, "learning_rate": 5.6051189308215225e-05, "loss": 0.1182, "step": 9691 }, { "epoch": 4.560941176470588, "grad_norm": 0.5838818088317431, "learning_rate": 5.6017020315003195e-05, "loss": 0.2773, "step": 9692 }, { "epoch": 4.561411764705882, "grad_norm": 0.6123500932668566, "learning_rate": 5.598285934839868e-05, "loss": 0.3711, "step": 9693 }, { "epoch": 4.5618823529411765, "grad_norm": 0.5197151934495552, "learning_rate": 5.594870641131907e-05, "loss": 0.2031, "step": 9694 }, { "epoch": 4.562352941176471, "grad_norm": 0.6192567074988755, "learning_rate": 5.591456150668126e-05, "loss": 0.4336, "step": 9695 }, { "epoch": 4.562823529411765, "grad_norm": 0.6610120527423362, "learning_rate": 5.588042463740143e-05, "loss": 0.2051, "step": 9696 }, { "epoch": 4.563294117647059, "grad_norm": 0.594797893397094, "learning_rate": 5.584629580639494e-05, "loss": 0.4121, "step": 9697 }, { "epoch": 4.5637647058823525, "grad_norm": 0.5495703231465039, "learning_rate": 5.581217501657666e-05, "loss": 0.2158, "step": 9698 }, { "epoch": 4.564235294117647, "grad_norm": 0.4976436973772436, "learning_rate": 5.577806227086057e-05, "loss": 0.168, "step": 9699 }, { "epoch": 4.564705882352941, "grad_norm": 0.4506193312462409, "learning_rate": 5.57439575721601e-05, "loss": 0.2461, "step": 9700 }, { "epoch": 4.565176470588235, "grad_norm": 0.45297644402378223, "learning_rate": 5.570986092338806e-05, "loss": 0.1914, "step": 9701 }, { "epoch": 4.565647058823529, "grad_norm": 0.48740056212217203, "learning_rate": 5.5675772327456313e-05, "loss": 0.2178, "step": 9702 }, { "epoch": 4.566117647058824, "grad_norm": 0.9793359943914861, "learning_rate": 5.5641691787276286e-05, "loss": 0.5156, "step": 9703 }, { "epoch": 4.566588235294118, "grad_norm": 0.1662549527712098, "learning_rate": 5.560761930575864e-05, "loss": 0.0396, "step": 9704 }, { "epoch": 4.567058823529412, "grad_norm": 0.5477788242219244, "learning_rate": 5.557355488581325e-05, "loss": 0.375, "step": 9705 }, { "epoch": 4.567529411764706, "grad_norm": 0.46792429135539804, "learning_rate": 5.553949853034943e-05, "loss": 0.2344, "step": 9706 }, { "epoch": 4.568, "grad_norm": 0.26988829026481165, "learning_rate": 5.550545024227582e-05, "loss": 0.1113, "step": 9707 }, { "epoch": 4.568470588235294, "grad_norm": 0.1835022975226479, "learning_rate": 5.5471410024500156e-05, "loss": 0.0459, "step": 9708 }, { "epoch": 4.568941176470588, "grad_norm": 0.27892721185791647, "learning_rate": 5.5437377879929826e-05, "loss": 0.0654, "step": 9709 }, { "epoch": 4.569411764705882, "grad_norm": 1.1100564546260516, "learning_rate": 5.5403353811471194e-05, "loss": 0.582, "step": 9710 }, { "epoch": 4.5698823529411765, "grad_norm": 0.7480177288847585, "learning_rate": 5.5369337822030154e-05, "loss": 0.4238, "step": 9711 }, { "epoch": 4.570352941176471, "grad_norm": 0.3133686318766129, "learning_rate": 5.533532991451187e-05, "loss": 0.1226, "step": 9712 }, { "epoch": 4.570823529411765, "grad_norm": 0.4738407058114026, "learning_rate": 5.53013300918207e-05, "loss": 0.209, "step": 9713 }, { "epoch": 4.571294117647059, "grad_norm": 0.28691817144177656, "learning_rate": 5.526733835686041e-05, "loss": 0.0869, "step": 9714 }, { "epoch": 4.571764705882353, "grad_norm": 0.5194720501095881, "learning_rate": 5.523335471253416e-05, "loss": 0.2061, "step": 9715 }, { "epoch": 4.572235294117647, "grad_norm": 0.5854465568024281, "learning_rate": 5.519937916174419e-05, "loss": 0.1562, "step": 9716 }, { "epoch": 4.572705882352941, "grad_norm": 0.41211881878413326, "learning_rate": 5.516541170739223e-05, "loss": 0.1533, "step": 9717 }, { "epoch": 4.573176470588235, "grad_norm": 0.5467895391106551, "learning_rate": 5.5131452352379314e-05, "loss": 0.3516, "step": 9718 }, { "epoch": 4.573647058823529, "grad_norm": 0.308646456895736, "learning_rate": 5.509750109960566e-05, "loss": 0.125, "step": 9719 }, { "epoch": 4.574117647058824, "grad_norm": 0.4436859372569293, "learning_rate": 5.5063557951970885e-05, "loss": 0.2539, "step": 9720 }, { "epoch": 4.574588235294118, "grad_norm": 0.4842284070771475, "learning_rate": 5.502962291237399e-05, "loss": 0.2598, "step": 9721 }, { "epoch": 4.575058823529412, "grad_norm": 0.7464162878483871, "learning_rate": 5.4995695983713076e-05, "loss": 0.2832, "step": 9722 }, { "epoch": 4.575529411764705, "grad_norm": 0.5301878840784605, "learning_rate": 5.496177716888571e-05, "loss": 0.3047, "step": 9723 }, { "epoch": 4.576, "grad_norm": 0.49180677791899785, "learning_rate": 5.4927866470788786e-05, "loss": 0.2676, "step": 9724 }, { "epoch": 4.576470588235294, "grad_norm": 0.5129201090662877, "learning_rate": 5.489396389231834e-05, "loss": 0.2266, "step": 9725 }, { "epoch": 4.576941176470588, "grad_norm": 0.390835486431908, "learning_rate": 5.4860069436369916e-05, "loss": 0.2051, "step": 9726 }, { "epoch": 4.577411764705882, "grad_norm": 0.31428483644307015, "learning_rate": 5.4826183105838184e-05, "loss": 0.103, "step": 9727 }, { "epoch": 4.5778823529411765, "grad_norm": 0.4900398622883494, "learning_rate": 5.4792304903617226e-05, "loss": 0.2148, "step": 9728 }, { "epoch": 4.578352941176471, "grad_norm": 0.4143997091581354, "learning_rate": 5.4758434832600475e-05, "loss": 0.1494, "step": 9729 }, { "epoch": 4.578823529411765, "grad_norm": 0.5437070650833222, "learning_rate": 5.47245728956805e-05, "loss": 0.2793, "step": 9730 }, { "epoch": 4.579294117647059, "grad_norm": 0.5129718094202216, "learning_rate": 5.469071909574933e-05, "loss": 0.1885, "step": 9731 }, { "epoch": 4.5797647058823525, "grad_norm": 0.610487642886021, "learning_rate": 5.465687343569828e-05, "loss": 0.4805, "step": 9732 }, { "epoch": 4.580235294117647, "grad_norm": 0.47342159098311476, "learning_rate": 5.462303591841783e-05, "loss": 0.2305, "step": 9733 }, { "epoch": 4.580705882352941, "grad_norm": 0.5588725221161335, "learning_rate": 5.458920654679795e-05, "loss": 0.3789, "step": 9734 }, { "epoch": 4.581176470588235, "grad_norm": 0.6489096054912867, "learning_rate": 5.4555385323727876e-05, "loss": 0.3047, "step": 9735 }, { "epoch": 4.581647058823529, "grad_norm": 0.3658418492050967, "learning_rate": 5.4521572252095985e-05, "loss": 0.1553, "step": 9736 }, { "epoch": 4.582117647058824, "grad_norm": 0.5039169677972687, "learning_rate": 5.4487767334790145e-05, "loss": 0.2832, "step": 9737 }, { "epoch": 4.582588235294118, "grad_norm": 0.6795628435275382, "learning_rate": 5.445397057469751e-05, "loss": 0.457, "step": 9738 }, { "epoch": 4.583058823529412, "grad_norm": 0.5797090989298285, "learning_rate": 5.4420181974704386e-05, "loss": 0.373, "step": 9739 }, { "epoch": 4.583529411764706, "grad_norm": 0.46882164137899723, "learning_rate": 5.4386401537696536e-05, "loss": 0.2598, "step": 9740 }, { "epoch": 4.584, "grad_norm": 0.43382218682015194, "learning_rate": 5.435262926655903e-05, "loss": 0.1504, "step": 9741 }, { "epoch": 4.584470588235294, "grad_norm": 0.46073749531769514, "learning_rate": 5.43188651641761e-05, "loss": 0.2275, "step": 9742 }, { "epoch": 4.584941176470588, "grad_norm": 0.6554465474850021, "learning_rate": 5.428510923343136e-05, "loss": 0.3672, "step": 9743 }, { "epoch": 4.585411764705882, "grad_norm": 0.5568230371343111, "learning_rate": 5.425136147720785e-05, "loss": 0.3496, "step": 9744 }, { "epoch": 4.5858823529411765, "grad_norm": 0.26332769525583244, "learning_rate": 5.421762189838765e-05, "loss": 0.1011, "step": 9745 }, { "epoch": 4.586352941176471, "grad_norm": 0.7880815680668616, "learning_rate": 5.4183890499852364e-05, "loss": 0.4512, "step": 9746 }, { "epoch": 4.586823529411765, "grad_norm": 0.493540171212808, "learning_rate": 5.4150167284482844e-05, "loss": 0.2852, "step": 9747 }, { "epoch": 4.587294117647059, "grad_norm": 0.6146466592782235, "learning_rate": 5.4116452255159126e-05, "loss": 0.291, "step": 9748 }, { "epoch": 4.587764705882353, "grad_norm": 0.6312374846551604, "learning_rate": 5.408274541476075e-05, "loss": 0.2969, "step": 9749 }, { "epoch": 4.588235294117647, "grad_norm": 0.052363344833212075, "learning_rate": 5.4049046766166335e-05, "loss": 0.0079, "step": 9750 }, { "epoch": 4.588705882352941, "grad_norm": 0.6062571489474424, "learning_rate": 5.401535631225396e-05, "loss": 0.4023, "step": 9751 }, { "epoch": 4.589176470588235, "grad_norm": 0.38632969644317067, "learning_rate": 5.398167405590096e-05, "loss": 0.2148, "step": 9752 }, { "epoch": 4.589647058823529, "grad_norm": 0.22895668269628883, "learning_rate": 5.3947999999983967e-05, "loss": 0.0598, "step": 9753 }, { "epoch": 4.590117647058824, "grad_norm": 0.5975448301481556, "learning_rate": 5.3914334147378894e-05, "loss": 0.4512, "step": 9754 }, { "epoch": 4.590588235294118, "grad_norm": 0.4600376759978994, "learning_rate": 5.388067650096106e-05, "loss": 0.2852, "step": 9755 }, { "epoch": 4.591058823529412, "grad_norm": 0.3438147306343464, "learning_rate": 5.384702706360483e-05, "loss": 0.0874, "step": 9756 }, { "epoch": 4.591529411764705, "grad_norm": 0.8963597761006419, "learning_rate": 5.3813385838184185e-05, "loss": 0.1855, "step": 9757 }, { "epoch": 4.592, "grad_norm": 0.7026144608351209, "learning_rate": 5.3779752827572126e-05, "loss": 0.3223, "step": 9758 }, { "epoch": 4.592470588235294, "grad_norm": 0.30125134026942174, "learning_rate": 5.374612803464115e-05, "loss": 0.124, "step": 9759 }, { "epoch": 4.592941176470588, "grad_norm": 0.5675989541325146, "learning_rate": 5.3712511462262994e-05, "loss": 0.2461, "step": 9760 }, { "epoch": 4.593411764705882, "grad_norm": 0.44991430104327795, "learning_rate": 5.36789031133086e-05, "loss": 0.166, "step": 9761 }, { "epoch": 4.5938823529411765, "grad_norm": 0.6804397692653882, "learning_rate": 5.364530299064834e-05, "loss": 0.3457, "step": 9762 }, { "epoch": 4.594352941176471, "grad_norm": 0.46661122601284344, "learning_rate": 5.361171109715186e-05, "loss": 0.2139, "step": 9763 }, { "epoch": 4.594823529411765, "grad_norm": 0.552284119685769, "learning_rate": 5.357812743568799e-05, "loss": 0.3105, "step": 9764 }, { "epoch": 4.595294117647059, "grad_norm": 0.4551601979913073, "learning_rate": 5.354455200912498e-05, "loss": 0.2715, "step": 9765 }, { "epoch": 4.5957647058823525, "grad_norm": 0.4677862630938948, "learning_rate": 5.35109848203304e-05, "loss": 0.209, "step": 9766 }, { "epoch": 4.596235294117647, "grad_norm": 0.698222842351985, "learning_rate": 5.3477425872170926e-05, "loss": 0.5469, "step": 9767 }, { "epoch": 4.596705882352941, "grad_norm": 0.5185199945230173, "learning_rate": 5.3443875167512736e-05, "loss": 0.1797, "step": 9768 }, { "epoch": 4.597176470588235, "grad_norm": 0.42654378608491444, "learning_rate": 5.341033270922126e-05, "loss": 0.21, "step": 9769 }, { "epoch": 4.597647058823529, "grad_norm": 0.232222085185807, "learning_rate": 5.337679850016109e-05, "loss": 0.0928, "step": 9770 }, { "epoch": 4.598117647058824, "grad_norm": 0.6467795868471523, "learning_rate": 5.3343272543196264e-05, "loss": 0.2793, "step": 9771 }, { "epoch": 4.598588235294118, "grad_norm": 0.13185742053745297, "learning_rate": 5.33097548411901e-05, "loss": 0.031, "step": 9772 }, { "epoch": 4.599058823529412, "grad_norm": 0.6593041580435013, "learning_rate": 5.3276245397005094e-05, "loss": 0.3574, "step": 9773 }, { "epoch": 4.599529411764706, "grad_norm": 0.590355489592481, "learning_rate": 5.324274421350316e-05, "loss": 0.4004, "step": 9774 }, { "epoch": 4.6, "grad_norm": 0.03294095888985242, "learning_rate": 5.3209251293545514e-05, "loss": 0.0056, "step": 9775 }, { "epoch": 4.600470588235294, "grad_norm": 0.5505189572923286, "learning_rate": 5.317576663999251e-05, "loss": 0.3535, "step": 9776 }, { "epoch": 4.600941176470588, "grad_norm": 0.33079564489140345, "learning_rate": 5.314229025570401e-05, "loss": 0.0566, "step": 9777 }, { "epoch": 4.601411764705882, "grad_norm": 0.2556071153269508, "learning_rate": 5.3108822143538944e-05, "loss": 0.0811, "step": 9778 }, { "epoch": 4.6018823529411765, "grad_norm": 0.5629350263628895, "learning_rate": 5.307536230635571e-05, "loss": 0.2402, "step": 9779 }, { "epoch": 4.602352941176471, "grad_norm": 0.5471689358983514, "learning_rate": 5.304191074701201e-05, "loss": 0.1523, "step": 9780 }, { "epoch": 4.602823529411765, "grad_norm": 0.305817373469018, "learning_rate": 5.3008467468364675e-05, "loss": 0.0825, "step": 9781 }, { "epoch": 4.603294117647059, "grad_norm": 0.7444847428430765, "learning_rate": 5.297503247326993e-05, "loss": 0.4551, "step": 9782 }, { "epoch": 4.603764705882353, "grad_norm": 0.26417997826524375, "learning_rate": 5.294160576458339e-05, "loss": 0.1338, "step": 9783 }, { "epoch": 4.604235294117647, "grad_norm": 0.47015178965368654, "learning_rate": 5.290818734515972e-05, "loss": 0.25, "step": 9784 }, { "epoch": 4.604705882352941, "grad_norm": 0.17101819964433077, "learning_rate": 5.287477721785309e-05, "loss": 0.0457, "step": 9785 }, { "epoch": 4.605176470588235, "grad_norm": 0.4881350337884972, "learning_rate": 5.284137538551693e-05, "loss": 0.207, "step": 9786 }, { "epoch": 4.605647058823529, "grad_norm": 0.3942238102510486, "learning_rate": 5.2807981851003825e-05, "loss": 0.2207, "step": 9787 }, { "epoch": 4.606117647058824, "grad_norm": 0.5978266592421604, "learning_rate": 5.2774596617165793e-05, "loss": 0.3691, "step": 9788 }, { "epoch": 4.606588235294118, "grad_norm": 0.48047894920781553, "learning_rate": 5.2741219686854146e-05, "loss": 0.2148, "step": 9789 }, { "epoch": 4.607058823529412, "grad_norm": 0.6169640180777018, "learning_rate": 5.270785106291934e-05, "loss": 0.2832, "step": 9790 }, { "epoch": 4.607529411764705, "grad_norm": 0.2985142108785133, "learning_rate": 5.2674490748211265e-05, "loss": 0.0952, "step": 9791 }, { "epoch": 4.608, "grad_norm": 0.5872541978574496, "learning_rate": 5.2641138745579126e-05, "loss": 0.3887, "step": 9792 }, { "epoch": 4.608470588235294, "grad_norm": 0.571628460435306, "learning_rate": 5.2607795057871216e-05, "loss": 0.373, "step": 9793 }, { "epoch": 4.608941176470588, "grad_norm": 0.9274664296978434, "learning_rate": 5.257445968793531e-05, "loss": 0.498, "step": 9794 }, { "epoch": 4.609411764705882, "grad_norm": 0.17466134326425267, "learning_rate": 5.254113263861848e-05, "loss": 0.0366, "step": 9795 }, { "epoch": 4.6098823529411765, "grad_norm": 0.6515337231807107, "learning_rate": 5.250781391276684e-05, "loss": 0.3125, "step": 9796 }, { "epoch": 4.610352941176471, "grad_norm": 0.30865761639428585, "learning_rate": 5.24745035132262e-05, "loss": 0.0913, "step": 9797 }, { "epoch": 4.610823529411765, "grad_norm": 0.3911961259683366, "learning_rate": 5.2441201442841264e-05, "loss": 0.1738, "step": 9798 }, { "epoch": 4.611294117647059, "grad_norm": 0.20937322901276326, "learning_rate": 5.240790770445625e-05, "loss": 0.0898, "step": 9799 }, { "epoch": 4.6117647058823525, "grad_norm": 0.6374767697570433, "learning_rate": 5.237462230091467e-05, "loss": 0.3867, "step": 9800 }, { "epoch": 4.612235294117647, "grad_norm": 0.4028323854419321, "learning_rate": 5.234134523505912e-05, "loss": 0.2969, "step": 9801 }, { "epoch": 4.612705882352941, "grad_norm": 0.7428424260833827, "learning_rate": 5.230807650973172e-05, "loss": 0.4668, "step": 9802 }, { "epoch": 4.613176470588235, "grad_norm": 0.5808071555395136, "learning_rate": 5.227481612777379e-05, "loss": 0.2324, "step": 9803 }, { "epoch": 4.613647058823529, "grad_norm": 0.7039039271493562, "learning_rate": 5.224156409202587e-05, "loss": 0.4531, "step": 9804 }, { "epoch": 4.614117647058824, "grad_norm": 0.3694719608021469, "learning_rate": 5.2208320405327865e-05, "loss": 0.1221, "step": 9805 }, { "epoch": 4.614588235294118, "grad_norm": 0.5497467546786171, "learning_rate": 5.2175085070519024e-05, "loss": 0.1797, "step": 9806 }, { "epoch": 4.615058823529412, "grad_norm": 0.5333748152141147, "learning_rate": 5.2141858090437696e-05, "loss": 0.2754, "step": 9807 }, { "epoch": 4.615529411764706, "grad_norm": 0.41288013473025437, "learning_rate": 5.210863946792172e-05, "loss": 0.1523, "step": 9808 }, { "epoch": 4.616, "grad_norm": 0.36992623939875763, "learning_rate": 5.207542920580805e-05, "loss": 0.106, "step": 9809 }, { "epoch": 4.616470588235294, "grad_norm": 0.5218419810668814, "learning_rate": 5.204222730693303e-05, "loss": 0.2324, "step": 9810 }, { "epoch": 4.616941176470588, "grad_norm": 0.27760896139647023, "learning_rate": 5.2009033774132345e-05, "loss": 0.1118, "step": 9811 }, { "epoch": 4.617411764705882, "grad_norm": 0.5021973511578862, "learning_rate": 5.197584861024076e-05, "loss": 0.3066, "step": 9812 }, { "epoch": 4.6178823529411765, "grad_norm": 0.6446685588891663, "learning_rate": 5.194267181809251e-05, "loss": 0.4512, "step": 9813 }, { "epoch": 4.618352941176471, "grad_norm": 0.40620905913482086, "learning_rate": 5.1909503400521104e-05, "loss": 0.1875, "step": 9814 }, { "epoch": 4.618823529411765, "grad_norm": 0.9084918306305719, "learning_rate": 5.187634336035918e-05, "loss": 0.2539, "step": 9815 }, { "epoch": 4.619294117647058, "grad_norm": 0.28794180847108214, "learning_rate": 5.184319170043884e-05, "loss": 0.0933, "step": 9816 }, { "epoch": 4.619764705882353, "grad_norm": 0.6783870921663961, "learning_rate": 5.1810048423591424e-05, "loss": 0.3027, "step": 9817 }, { "epoch": 4.620235294117647, "grad_norm": 0.16450271462025598, "learning_rate": 5.177691353264746e-05, "loss": 0.0413, "step": 9818 }, { "epoch": 4.620705882352941, "grad_norm": 0.3019411808398021, "learning_rate": 5.174378703043684e-05, "loss": 0.0679, "step": 9819 }, { "epoch": 4.621176470588235, "grad_norm": 0.19043714119120836, "learning_rate": 5.171066891978881e-05, "loss": 0.0352, "step": 9820 }, { "epoch": 4.621647058823529, "grad_norm": 0.7426878561580242, "learning_rate": 5.167755920353171e-05, "loss": 0.2754, "step": 9821 }, { "epoch": 4.622117647058824, "grad_norm": 0.6426997167666519, "learning_rate": 5.164445788449334e-05, "loss": 0.2656, "step": 9822 }, { "epoch": 4.622588235294118, "grad_norm": 0.7248842734821453, "learning_rate": 5.161136496550072e-05, "loss": 0.334, "step": 9823 }, { "epoch": 4.623058823529412, "grad_norm": 0.3984861309661198, "learning_rate": 5.15782804493801e-05, "loss": 0.1357, "step": 9824 }, { "epoch": 4.623529411764705, "grad_norm": 0.43898845759647526, "learning_rate": 5.1545204338957066e-05, "loss": 0.1641, "step": 9825 }, { "epoch": 4.624, "grad_norm": 0.3905138206340677, "learning_rate": 5.151213663705655e-05, "loss": 0.1396, "step": 9826 }, { "epoch": 4.624470588235294, "grad_norm": 0.6953886229171506, "learning_rate": 5.14790773465026e-05, "loss": 0.459, "step": 9827 }, { "epoch": 4.624941176470588, "grad_norm": 0.5902157443108789, "learning_rate": 5.144602647011874e-05, "loss": 0.418, "step": 9828 }, { "epoch": 4.625411764705882, "grad_norm": 0.7213650300839368, "learning_rate": 5.141298401072758e-05, "loss": 0.4004, "step": 9829 }, { "epoch": 4.625882352941177, "grad_norm": 0.5216641480959171, "learning_rate": 5.137994997115115e-05, "loss": 0.3027, "step": 9830 }, { "epoch": 4.626352941176471, "grad_norm": 0.5501283156739551, "learning_rate": 5.134692435421077e-05, "loss": 0.3066, "step": 9831 }, { "epoch": 4.626823529411765, "grad_norm": 0.5932981935533008, "learning_rate": 5.1313907162726883e-05, "loss": 0.2656, "step": 9832 }, { "epoch": 4.627294117647059, "grad_norm": 0.4198834926165777, "learning_rate": 5.128089839951937e-05, "loss": 0.1436, "step": 9833 }, { "epoch": 4.627764705882353, "grad_norm": 0.8693925638547365, "learning_rate": 5.124789806740742e-05, "loss": 0.5742, "step": 9834 }, { "epoch": 4.628235294117647, "grad_norm": 0.8659972148888191, "learning_rate": 5.12149061692093e-05, "loss": 0.0417, "step": 9835 }, { "epoch": 4.628705882352941, "grad_norm": 0.3226735275703307, "learning_rate": 5.118192270774271e-05, "loss": 0.1084, "step": 9836 }, { "epoch": 4.629176470588235, "grad_norm": 0.5574498726071192, "learning_rate": 5.114894768582468e-05, "loss": 0.1826, "step": 9837 }, { "epoch": 4.6296470588235294, "grad_norm": 1.2927956696468494, "learning_rate": 5.111598110627133e-05, "loss": 0.6523, "step": 9838 }, { "epoch": 4.630117647058824, "grad_norm": 0.5792734872056028, "learning_rate": 5.108302297189822e-05, "loss": 0.2139, "step": 9839 }, { "epoch": 4.630588235294118, "grad_norm": 0.1252944030163092, "learning_rate": 5.105007328552019e-05, "loss": 0.0253, "step": 9840 }, { "epoch": 4.631058823529412, "grad_norm": 0.6386354927876211, "learning_rate": 5.101713204995114e-05, "loss": 0.3379, "step": 9841 }, { "epoch": 4.631529411764706, "grad_norm": 0.4012009012733273, "learning_rate": 5.0984199268004615e-05, "loss": 0.1128, "step": 9842 }, { "epoch": 4.632, "grad_norm": 0.6591565857146313, "learning_rate": 5.095127494249309e-05, "loss": 0.2305, "step": 9843 }, { "epoch": 4.632470588235294, "grad_norm": 0.039888596240443164, "learning_rate": 5.091835907622854e-05, "loss": 0.0034, "step": 9844 }, { "epoch": 4.632941176470588, "grad_norm": 0.6403875634475753, "learning_rate": 5.0885451672022156e-05, "loss": 0.3027, "step": 9845 }, { "epoch": 4.633411764705882, "grad_norm": 0.43970292774637093, "learning_rate": 5.0852552732684296e-05, "loss": 0.3418, "step": 9846 }, { "epoch": 4.633882352941177, "grad_norm": 0.528155709895793, "learning_rate": 5.081966226102474e-05, "loss": 0.2871, "step": 9847 }, { "epoch": 4.634352941176471, "grad_norm": 0.3756175391226481, "learning_rate": 5.078678025985257e-05, "loss": 0.1748, "step": 9848 }, { "epoch": 4.634823529411765, "grad_norm": 0.6692714123561194, "learning_rate": 5.0753906731975954e-05, "loss": 0.3027, "step": 9849 }, { "epoch": 4.635294117647058, "grad_norm": 0.43366516816844236, "learning_rate": 5.07210416802025e-05, "loss": 0.1973, "step": 9850 }, { "epoch": 4.635764705882353, "grad_norm": 0.2046408925672608, "learning_rate": 5.0688185107339084e-05, "loss": 0.0381, "step": 9851 }, { "epoch": 4.636235294117647, "grad_norm": 0.33103540520342245, "learning_rate": 5.065533701619175e-05, "loss": 0.1504, "step": 9852 }, { "epoch": 4.636705882352941, "grad_norm": 0.3255459865259578, "learning_rate": 5.06224974095659e-05, "loss": 0.1162, "step": 9853 }, { "epoch": 4.637176470588235, "grad_norm": 0.8906070689684338, "learning_rate": 5.058966629026629e-05, "loss": 0.3984, "step": 9854 }, { "epoch": 4.6376470588235295, "grad_norm": 0.4721587511408613, "learning_rate": 5.0556843661096714e-05, "loss": 0.25, "step": 9855 }, { "epoch": 4.638117647058824, "grad_norm": 0.3422125016499709, "learning_rate": 5.052402952486046e-05, "loss": 0.1348, "step": 9856 }, { "epoch": 4.638588235294118, "grad_norm": 0.756979659289943, "learning_rate": 5.0491223884360054e-05, "loss": 0.4629, "step": 9857 }, { "epoch": 4.639058823529412, "grad_norm": 0.4228503254601729, "learning_rate": 5.0458426742397174e-05, "loss": 0.2314, "step": 9858 }, { "epoch": 4.6395294117647055, "grad_norm": 0.5305786844689457, "learning_rate": 5.042563810177296e-05, "loss": 0.418, "step": 9859 }, { "epoch": 4.64, "grad_norm": 0.45104914665382156, "learning_rate": 5.039285796528759e-05, "loss": 0.252, "step": 9860 }, { "epoch": 4.640470588235294, "grad_norm": 0.34468025740171737, "learning_rate": 5.0360086335740734e-05, "loss": 0.1299, "step": 9861 }, { "epoch": 4.640941176470588, "grad_norm": 0.43333282275164087, "learning_rate": 5.032732321593128e-05, "loss": 0.0952, "step": 9862 }, { "epoch": 4.641411764705882, "grad_norm": 0.6441366454488036, "learning_rate": 5.029456860865727e-05, "loss": 0.2168, "step": 9863 }, { "epoch": 4.641882352941177, "grad_norm": 0.2802087415842361, "learning_rate": 5.0261822516716146e-05, "loss": 0.0908, "step": 9864 }, { "epoch": 4.642352941176471, "grad_norm": 0.2321922052733604, "learning_rate": 5.022908494290465e-05, "loss": 0.0732, "step": 9865 }, { "epoch": 4.642823529411765, "grad_norm": 0.3189728621191964, "learning_rate": 5.0196355890018605e-05, "loss": 0.1406, "step": 9866 }, { "epoch": 4.643294117647059, "grad_norm": 0.3259528746267433, "learning_rate": 5.016363536085333e-05, "loss": 0.1865, "step": 9867 }, { "epoch": 4.643764705882353, "grad_norm": 0.6595494956694231, "learning_rate": 5.0130923358203316e-05, "loss": 0.4102, "step": 9868 }, { "epoch": 4.644235294117647, "grad_norm": 0.33792425065155246, "learning_rate": 5.0098219884862265e-05, "loss": 0.1328, "step": 9869 }, { "epoch": 4.644705882352941, "grad_norm": 0.0873952955207008, "learning_rate": 5.006552494362325e-05, "loss": 0.0098, "step": 9870 }, { "epoch": 4.645176470588235, "grad_norm": 0.42849013517101897, "learning_rate": 5.0032838537278627e-05, "loss": 0.1494, "step": 9871 }, { "epoch": 4.6456470588235295, "grad_norm": 0.4407499227684533, "learning_rate": 5.000016066861988e-05, "loss": 0.2422, "step": 9872 }, { "epoch": 4.646117647058824, "grad_norm": 0.6173097422663223, "learning_rate": 4.9967491340437904e-05, "loss": 0.3457, "step": 9873 }, { "epoch": 4.646588235294118, "grad_norm": 0.5738974615918134, "learning_rate": 4.993483055552289e-05, "loss": 0.3027, "step": 9874 }, { "epoch": 4.647058823529412, "grad_norm": 0.26985524578317716, "learning_rate": 4.990217831666409e-05, "loss": 0.0991, "step": 9875 }, { "epoch": 4.647529411764706, "grad_norm": 0.5766791183940998, "learning_rate": 4.986953462665026e-05, "loss": 0.2793, "step": 9876 }, { "epoch": 4.648, "grad_norm": 0.28673952612426434, "learning_rate": 4.9836899488269345e-05, "loss": 0.1426, "step": 9877 }, { "epoch": 4.648470588235294, "grad_norm": 0.3099367362882904, "learning_rate": 4.9804272904308465e-05, "loss": 0.1279, "step": 9878 }, { "epoch": 4.648941176470588, "grad_norm": 0.660454397337816, "learning_rate": 4.97716548775542e-05, "loss": 0.2109, "step": 9879 }, { "epoch": 4.649411764705882, "grad_norm": 0.7241862119378376, "learning_rate": 4.9739045410792164e-05, "loss": 0.4863, "step": 9880 }, { "epoch": 4.649882352941177, "grad_norm": 0.40480564253690654, "learning_rate": 4.970644450680742e-05, "loss": 0.0898, "step": 9881 }, { "epoch": 4.650352941176471, "grad_norm": 0.5059502615723481, "learning_rate": 4.9673852168384316e-05, "loss": 0.2676, "step": 9882 }, { "epoch": 4.650823529411765, "grad_norm": 0.5486740142480805, "learning_rate": 4.964126839830628e-05, "loss": 0.2852, "step": 9883 }, { "epoch": 4.651294117647058, "grad_norm": 0.6199204925674983, "learning_rate": 4.960869319935616e-05, "loss": 0.3535, "step": 9884 }, { "epoch": 4.651764705882353, "grad_norm": 0.48667518187931186, "learning_rate": 4.9576126574316086e-05, "loss": 0.2656, "step": 9885 }, { "epoch": 4.652235294117647, "grad_norm": 0.4788636240309436, "learning_rate": 4.9543568525967376e-05, "loss": 0.2695, "step": 9886 }, { "epoch": 4.652705882352941, "grad_norm": 0.5325461644867372, "learning_rate": 4.9511019057090684e-05, "loss": 0.2178, "step": 9887 }, { "epoch": 4.653176470588235, "grad_norm": 0.45871478624906414, "learning_rate": 4.947847817046582e-05, "loss": 0.1357, "step": 9888 }, { "epoch": 4.6536470588235295, "grad_norm": 0.4634415032148188, "learning_rate": 4.9445945868871974e-05, "loss": 0.2461, "step": 9889 }, { "epoch": 4.654117647058824, "grad_norm": 0.6931702455351229, "learning_rate": 4.941342215508761e-05, "loss": 0.3105, "step": 9890 }, { "epoch": 4.654588235294118, "grad_norm": 0.30738266793659597, "learning_rate": 4.9380907031890315e-05, "loss": 0.1436, "step": 9891 }, { "epoch": 4.655058823529412, "grad_norm": 0.4853493473654799, "learning_rate": 4.934840050205711e-05, "loss": 0.2324, "step": 9892 }, { "epoch": 4.6555294117647055, "grad_norm": 0.8227227555162354, "learning_rate": 4.931590256836424e-05, "loss": 0.4844, "step": 9893 }, { "epoch": 4.656, "grad_norm": 0.595717467616225, "learning_rate": 4.928341323358708e-05, "loss": 0.2832, "step": 9894 }, { "epoch": 4.656470588235294, "grad_norm": 0.37539005035793077, "learning_rate": 4.925093250050045e-05, "loss": 0.207, "step": 9895 }, { "epoch": 4.656941176470588, "grad_norm": 0.7152500649933797, "learning_rate": 4.921846037187841e-05, "loss": 0.4766, "step": 9896 }, { "epoch": 4.657411764705882, "grad_norm": 0.2766500710765095, "learning_rate": 4.918599685049414e-05, "loss": 0.1104, "step": 9897 }, { "epoch": 4.657882352941177, "grad_norm": 0.652988758631053, "learning_rate": 4.915354193912021e-05, "loss": 0.4023, "step": 9898 }, { "epoch": 4.658352941176471, "grad_norm": 0.7167214889332147, "learning_rate": 4.9121095640528516e-05, "loss": 0.4395, "step": 9899 }, { "epoch": 4.658823529411765, "grad_norm": 0.6549142425557687, "learning_rate": 4.908865795748999e-05, "loss": 0.3926, "step": 9900 }, { "epoch": 4.659294117647059, "grad_norm": 0.645598676209748, "learning_rate": 4.905622889277506e-05, "loss": 0.25, "step": 9901 }, { "epoch": 4.659764705882353, "grad_norm": 0.4002658516393328, "learning_rate": 4.902380844915337e-05, "loss": 0.1553, "step": 9902 }, { "epoch": 4.660235294117647, "grad_norm": 0.6364154320417302, "learning_rate": 4.899139662939366e-05, "loss": 0.5117, "step": 9903 }, { "epoch": 4.660705882352941, "grad_norm": 0.23008399525261286, "learning_rate": 4.895899343626412e-05, "loss": 0.0747, "step": 9904 }, { "epoch": 4.661176470588235, "grad_norm": 0.5705140479022571, "learning_rate": 4.892659887253219e-05, "loss": 0.2324, "step": 9905 }, { "epoch": 4.6616470588235295, "grad_norm": 0.522632501063971, "learning_rate": 4.8894212940964445e-05, "loss": 0.2754, "step": 9906 }, { "epoch": 4.662117647058824, "grad_norm": 0.9644937019638196, "learning_rate": 4.8861835644326866e-05, "loss": 0.6016, "step": 9907 }, { "epoch": 4.662588235294118, "grad_norm": 0.4971062323926041, "learning_rate": 4.882946698538457e-05, "loss": 0.2598, "step": 9908 }, { "epoch": 4.663058823529412, "grad_norm": 0.6490364401109449, "learning_rate": 4.879710696690202e-05, "loss": 0.3887, "step": 9909 }, { "epoch": 4.663529411764706, "grad_norm": 0.3070172543306949, "learning_rate": 4.876475559164299e-05, "loss": 0.1162, "step": 9910 }, { "epoch": 4.664, "grad_norm": 0.5609553656945747, "learning_rate": 4.873241286237033e-05, "loss": 0.2139, "step": 9911 }, { "epoch": 4.664470588235294, "grad_norm": 0.3505571620093968, "learning_rate": 4.8700078781846326e-05, "loss": 0.1494, "step": 9912 }, { "epoch": 4.664941176470588, "grad_norm": 0.46555404743115836, "learning_rate": 4.866775335283253e-05, "loss": 0.2168, "step": 9913 }, { "epoch": 4.665411764705882, "grad_norm": 0.47737137954965103, "learning_rate": 4.863543657808956e-05, "loss": 0.166, "step": 9914 }, { "epoch": 4.665882352941177, "grad_norm": 0.9068223440182758, "learning_rate": 4.860312846037753e-05, "loss": 0.4727, "step": 9915 }, { "epoch": 4.666352941176471, "grad_norm": 0.4900495763190218, "learning_rate": 4.857082900245569e-05, "loss": 0.2676, "step": 9916 }, { "epoch": 4.666823529411765, "grad_norm": 0.3519592872928928, "learning_rate": 4.8538538207082534e-05, "loss": 0.1309, "step": 9917 }, { "epoch": 4.667294117647058, "grad_norm": 0.38589001053801736, "learning_rate": 4.850625607701588e-05, "loss": 0.1387, "step": 9918 }, { "epoch": 4.667764705882353, "grad_norm": 0.36520108016762126, "learning_rate": 4.8473982615012816e-05, "loss": 0.1416, "step": 9919 }, { "epoch": 4.668235294117647, "grad_norm": 0.44413927614316623, "learning_rate": 4.844171782382959e-05, "loss": 0.1348, "step": 9920 }, { "epoch": 4.668705882352941, "grad_norm": 2.4843259612809666e-05, "learning_rate": 4.840946170622179e-05, "loss": 0.0, "step": 9921 }, { "epoch": 4.669176470588235, "grad_norm": 0.5574737821797798, "learning_rate": 4.8377214264944296e-05, "loss": 0.2988, "step": 9922 }, { "epoch": 4.6696470588235295, "grad_norm": 0.5034037459498772, "learning_rate": 4.834497550275114e-05, "loss": 0.2715, "step": 9923 }, { "epoch": 4.670117647058824, "grad_norm": 0.43903826097601056, "learning_rate": 4.831274542239566e-05, "loss": 0.1367, "step": 9924 }, { "epoch": 4.670588235294118, "grad_norm": 0.25748784411402936, "learning_rate": 4.8280524026630565e-05, "loss": 0.0718, "step": 9925 }, { "epoch": 4.671058823529412, "grad_norm": 0.6129687258002099, "learning_rate": 4.824831131820759e-05, "loss": 0.3633, "step": 9926 }, { "epoch": 4.6715294117647055, "grad_norm": 0.48718298305373275, "learning_rate": 4.821610729987792e-05, "loss": 0.3633, "step": 9927 }, { "epoch": 4.672, "grad_norm": 0.37443784473175795, "learning_rate": 4.818391197439199e-05, "loss": 0.2461, "step": 9928 }, { "epoch": 4.672470588235294, "grad_norm": 0.3370978979657199, "learning_rate": 4.815172534449933e-05, "loss": 0.1245, "step": 9929 }, { "epoch": 4.672941176470588, "grad_norm": 0.6419901359638963, "learning_rate": 4.811954741294889e-05, "loss": 0.3555, "step": 9930 }, { "epoch": 4.673411764705882, "grad_norm": 0.37262419373421873, "learning_rate": 4.808737818248882e-05, "loss": 0.0928, "step": 9931 }, { "epoch": 4.673882352941177, "grad_norm": 0.5358544860697086, "learning_rate": 4.805521765586653e-05, "loss": 0.3105, "step": 9932 }, { "epoch": 4.674352941176471, "grad_norm": 0.43253577511268393, "learning_rate": 4.802306583582873e-05, "loss": 0.1455, "step": 9933 }, { "epoch": 4.674823529411765, "grad_norm": 0.7609271045739304, "learning_rate": 4.7990922725121266e-05, "loss": 0.2949, "step": 9934 }, { "epoch": 4.675294117647059, "grad_norm": 0.5211022958950342, "learning_rate": 4.795878832648935e-05, "loss": 0.291, "step": 9935 }, { "epoch": 4.675764705882353, "grad_norm": 0.7073367396996822, "learning_rate": 4.7926662642677466e-05, "loss": 0.5469, "step": 9936 }, { "epoch": 4.676235294117647, "grad_norm": 0.18152072802157923, "learning_rate": 4.7894545676429206e-05, "loss": 0.0347, "step": 9937 }, { "epoch": 4.676705882352941, "grad_norm": 0.20422555401693748, "learning_rate": 4.7862437430487635e-05, "loss": 0.0488, "step": 9938 }, { "epoch": 4.677176470588235, "grad_norm": 0.04363463910023587, "learning_rate": 4.783033790759481e-05, "loss": 0.0034, "step": 9939 }, { "epoch": 4.6776470588235295, "grad_norm": 0.4912908541222362, "learning_rate": 4.77982471104923e-05, "loss": 0.1914, "step": 9940 }, { "epoch": 4.678117647058824, "grad_norm": 0.5100109370982328, "learning_rate": 4.776616504192082e-05, "loss": 0.2676, "step": 9941 }, { "epoch": 4.678588235294118, "grad_norm": 0.4199720574040564, "learning_rate": 4.7734091704620244e-05, "loss": 0.1748, "step": 9942 }, { "epoch": 4.679058823529411, "grad_norm": 0.5098675262979715, "learning_rate": 4.7702027101329846e-05, "loss": 0.2539, "step": 9943 }, { "epoch": 4.679529411764706, "grad_norm": 0.5269451494926594, "learning_rate": 4.7669971234788165e-05, "loss": 0.208, "step": 9944 }, { "epoch": 4.68, "grad_norm": 0.35474171396127474, "learning_rate": 4.763792410773281e-05, "loss": 0.1729, "step": 9945 }, { "epoch": 4.680470588235294, "grad_norm": 0.5188392311871013, "learning_rate": 4.760588572290082e-05, "loss": 0.3164, "step": 9946 }, { "epoch": 4.680941176470588, "grad_norm": 0.2619578648096342, "learning_rate": 4.7573856083028475e-05, "loss": 0.0625, "step": 9947 }, { "epoch": 4.681411764705882, "grad_norm": 0.5648005398768315, "learning_rate": 4.7541835190851175e-05, "loss": 0.3223, "step": 9948 }, { "epoch": 4.681882352941177, "grad_norm": 0.7243172143906298, "learning_rate": 4.7509823049103706e-05, "loss": 0.3848, "step": 9949 }, { "epoch": 4.682352941176471, "grad_norm": 0.5430226950062359, "learning_rate": 4.74778196605201e-05, "loss": 0.2734, "step": 9950 }, { "epoch": 4.682823529411765, "grad_norm": 0.6661268979673647, "learning_rate": 4.744582502783352e-05, "loss": 0.4238, "step": 9951 }, { "epoch": 4.683294117647058, "grad_norm": 0.7642772336411726, "learning_rate": 4.7413839153776504e-05, "loss": 0.7969, "step": 9952 }, { "epoch": 4.683764705882353, "grad_norm": 0.18147315316062368, "learning_rate": 4.7381862041080865e-05, "loss": 0.0242, "step": 9953 }, { "epoch": 4.684235294117647, "grad_norm": 0.821041083761168, "learning_rate": 4.7349893692477495e-05, "loss": 0.5234, "step": 9954 }, { "epoch": 4.684705882352941, "grad_norm": 0.7625105055154435, "learning_rate": 4.7317934110696685e-05, "loss": 0.498, "step": 9955 }, { "epoch": 4.685176470588235, "grad_norm": 0.557476237289603, "learning_rate": 4.728598329846802e-05, "loss": 0.207, "step": 9956 }, { "epoch": 4.6856470588235295, "grad_norm": 0.3197590408026067, "learning_rate": 4.725404125852013e-05, "loss": 0.1582, "step": 9957 }, { "epoch": 4.686117647058824, "grad_norm": 0.7210802339596967, "learning_rate": 4.7222107993581135e-05, "loss": 0.3535, "step": 9958 }, { "epoch": 4.686588235294118, "grad_norm": 0.6357052425005657, "learning_rate": 4.7190183506378205e-05, "loss": 0.4863, "step": 9959 }, { "epoch": 4.687058823529412, "grad_norm": 0.6420016770591491, "learning_rate": 4.7158267799637864e-05, "loss": 0.4004, "step": 9960 }, { "epoch": 4.6875294117647055, "grad_norm": 0.4325553126936778, "learning_rate": 4.712636087608594e-05, "loss": 0.3125, "step": 9961 }, { "epoch": 4.688, "grad_norm": 0.27485468068827995, "learning_rate": 4.709446273844736e-05, "loss": 0.1108, "step": 9962 }, { "epoch": 4.688470588235294, "grad_norm": 0.4667736798081324, "learning_rate": 4.7062573389446395e-05, "loss": 0.2109, "step": 9963 }, { "epoch": 4.688941176470588, "grad_norm": 0.41288943663575445, "learning_rate": 4.70306928318066e-05, "loss": 0.1318, "step": 9964 }, { "epoch": 4.689411764705882, "grad_norm": 0.5513512065138421, "learning_rate": 4.6998821068250675e-05, "loss": 0.2471, "step": 9965 }, { "epoch": 4.689882352941177, "grad_norm": 0.650348474010942, "learning_rate": 4.696695810150065e-05, "loss": 0.3809, "step": 9966 }, { "epoch": 4.690352941176471, "grad_norm": 0.4337323918572262, "learning_rate": 4.693510393427781e-05, "loss": 0.249, "step": 9967 }, { "epoch": 4.690823529411765, "grad_norm": 0.3193789925610585, "learning_rate": 4.690325856930258e-05, "loss": 0.1074, "step": 9968 }, { "epoch": 4.691294117647059, "grad_norm": 0.39641164265921325, "learning_rate": 4.6871422009294764e-05, "loss": 0.1709, "step": 9969 }, { "epoch": 4.691764705882353, "grad_norm": 0.1693930293358786, "learning_rate": 4.6839594256973406e-05, "loss": 0.0466, "step": 9970 }, { "epoch": 4.692235294117647, "grad_norm": 1.1530803972362997, "learning_rate": 4.680777531505662e-05, "loss": 0.5352, "step": 9971 }, { "epoch": 4.692705882352941, "grad_norm": 0.6418512005892503, "learning_rate": 4.6775965186262e-05, "loss": 0.3105, "step": 9972 }, { "epoch": 4.693176470588235, "grad_norm": 0.9741750783994502, "learning_rate": 4.6744163873306314e-05, "loss": 0.5312, "step": 9973 }, { "epoch": 4.6936470588235295, "grad_norm": 0.2553458824188249, "learning_rate": 4.671237137890539e-05, "loss": 0.0454, "step": 9974 }, { "epoch": 4.694117647058824, "grad_norm": 0.7515165426973781, "learning_rate": 4.668058770577468e-05, "loss": 0.4062, "step": 9975 }, { "epoch": 4.694588235294118, "grad_norm": 0.5607273157440661, "learning_rate": 4.6648812856628515e-05, "loss": 0.2471, "step": 9976 }, { "epoch": 4.695058823529411, "grad_norm": 0.6995765948752912, "learning_rate": 4.6617046834180656e-05, "loss": 0.416, "step": 9977 }, { "epoch": 4.6955294117647055, "grad_norm": 0.5055467549387277, "learning_rate": 4.658528964114413e-05, "loss": 0.2246, "step": 9978 }, { "epoch": 4.696, "grad_norm": 0.6855304796552134, "learning_rate": 4.655354128023107e-05, "loss": 0.3457, "step": 9979 }, { "epoch": 4.696470588235294, "grad_norm": 0.4085356762857177, "learning_rate": 4.6521801754152995e-05, "loss": 0.1553, "step": 9980 }, { "epoch": 4.696941176470588, "grad_norm": 0.46282565350540034, "learning_rate": 4.649007106562066e-05, "loss": 0.2432, "step": 9981 }, { "epoch": 4.697411764705882, "grad_norm": 0.5267189281985502, "learning_rate": 4.645834921734392e-05, "loss": 0.3438, "step": 9982 }, { "epoch": 4.697882352941177, "grad_norm": 0.7162854052782943, "learning_rate": 4.642663621203201e-05, "loss": 0.2422, "step": 9983 }, { "epoch": 4.698352941176471, "grad_norm": 0.7246166770330074, "learning_rate": 4.639493205239346e-05, "loss": 0.2422, "step": 9984 }, { "epoch": 4.698823529411765, "grad_norm": 0.2558233072942622, "learning_rate": 4.636323674113583e-05, "loss": 0.0815, "step": 9985 }, { "epoch": 4.699294117647058, "grad_norm": 0.4228412338498343, "learning_rate": 4.6331550280966126e-05, "loss": 0.1475, "step": 9986 }, { "epoch": 4.699764705882353, "grad_norm": 0.5136777197192838, "learning_rate": 4.629987267459056e-05, "loss": 0.291, "step": 9987 }, { "epoch": 4.700235294117647, "grad_norm": 0.19315312254960806, "learning_rate": 4.626820392471446e-05, "loss": 0.0623, "step": 9988 }, { "epoch": 4.700705882352941, "grad_norm": 0.3227459746696226, "learning_rate": 4.6236544034042604e-05, "loss": 0.1001, "step": 9989 }, { "epoch": 4.701176470588235, "grad_norm": 0.6898729367395183, "learning_rate": 4.62048930052788e-05, "loss": 0.3379, "step": 9990 }, { "epoch": 4.7016470588235295, "grad_norm": 0.7496806396479947, "learning_rate": 4.6173250841126245e-05, "loss": 0.3691, "step": 9991 }, { "epoch": 4.702117647058824, "grad_norm": 0.5235911108726297, "learning_rate": 4.614161754428737e-05, "loss": 0.2637, "step": 9992 }, { "epoch": 4.702588235294118, "grad_norm": 0.49207822781419824, "learning_rate": 4.6109993117463756e-05, "loss": 0.2656, "step": 9993 }, { "epoch": 4.703058823529412, "grad_norm": 0.35842456233173003, "learning_rate": 4.607837756335629e-05, "loss": 0.1387, "step": 9994 }, { "epoch": 4.7035294117647055, "grad_norm": 0.5343507179632322, "learning_rate": 4.604677088466516e-05, "loss": 0.2539, "step": 9995 }, { "epoch": 4.704, "grad_norm": 0.6278319690311693, "learning_rate": 4.6015173084089635e-05, "loss": 0.5938, "step": 9996 }, { "epoch": 4.704470588235294, "grad_norm": 0.26592474666785404, "learning_rate": 4.598358416432838e-05, "loss": 0.1206, "step": 9997 }, { "epoch": 4.704941176470588, "grad_norm": 0.48373902912616207, "learning_rate": 4.595200412807927e-05, "loss": 0.2539, "step": 9998 }, { "epoch": 4.705411764705882, "grad_norm": 0.5736373865566319, "learning_rate": 4.592043297803932e-05, "loss": 0.3555, "step": 9999 }, { "epoch": 4.705882352941177, "grad_norm": 0.5868128246352966, "learning_rate": 4.58888707169049e-05, "loss": 0.4668, "step": 10000 }, { "epoch": 4.706352941176471, "grad_norm": 0.9088623927816765, "learning_rate": 4.585731734737164e-05, "loss": 0.6133, "step": 10001 }, { "epoch": 4.706823529411765, "grad_norm": 0.3622057583125326, "learning_rate": 4.5825772872134234e-05, "loss": 0.1533, "step": 10002 }, { "epoch": 4.707294117647059, "grad_norm": 0.5998844822234628, "learning_rate": 4.579423729388681e-05, "loss": 0.3145, "step": 10003 }, { "epoch": 4.707764705882353, "grad_norm": 0.6164584985154299, "learning_rate": 4.576271061532268e-05, "loss": 0.3262, "step": 10004 }, { "epoch": 4.708235294117647, "grad_norm": 0.3404056815272993, "learning_rate": 4.573119283913431e-05, "loss": 0.1157, "step": 10005 }, { "epoch": 4.708705882352941, "grad_norm": 0.6387996075775673, "learning_rate": 4.569968396801351e-05, "loss": 0.3125, "step": 10006 }, { "epoch": 4.709176470588235, "grad_norm": 0.7076185548007098, "learning_rate": 4.566818400465133e-05, "loss": 0.543, "step": 10007 }, { "epoch": 4.7096470588235295, "grad_norm": 0.621648328568631, "learning_rate": 4.563669295173793e-05, "loss": 0.4941, "step": 10008 }, { "epoch": 4.710117647058824, "grad_norm": 0.29306373329499846, "learning_rate": 4.560521081196289e-05, "loss": 0.0645, "step": 10009 }, { "epoch": 4.710588235294118, "grad_norm": 0.5930662682937736, "learning_rate": 4.5573737588014875e-05, "loss": 0.3594, "step": 10010 }, { "epoch": 4.711058823529411, "grad_norm": 0.7012877910383705, "learning_rate": 4.554227328258185e-05, "loss": 0.3984, "step": 10011 }, { "epoch": 4.7115294117647055, "grad_norm": 0.31776714618470425, "learning_rate": 4.551081789835113e-05, "loss": 0.0918, "step": 10012 }, { "epoch": 4.712, "grad_norm": 0.7158851657500291, "learning_rate": 4.547937143800901e-05, "loss": 0.3633, "step": 10013 }, { "epoch": 4.712470588235294, "grad_norm": 0.5624748426685129, "learning_rate": 4.544793390424124e-05, "loss": 0.2246, "step": 10014 }, { "epoch": 4.712941176470588, "grad_norm": 0.18842431142401805, "learning_rate": 4.541650529973279e-05, "loss": 0.0518, "step": 10015 }, { "epoch": 4.713411764705882, "grad_norm": 0.4426713627231844, "learning_rate": 4.5385085627167717e-05, "loss": 0.1758, "step": 10016 }, { "epoch": 4.713882352941177, "grad_norm": 0.4953048941338949, "learning_rate": 4.535367488922947e-05, "loss": 0.2383, "step": 10017 }, { "epoch": 4.714352941176471, "grad_norm": 0.6706129679961649, "learning_rate": 4.5322273088600714e-05, "loss": 0.4961, "step": 10018 }, { "epoch": 4.714823529411765, "grad_norm": 0.29703070635240136, "learning_rate": 4.5290880227963186e-05, "loss": 0.0874, "step": 10019 }, { "epoch": 4.715294117647058, "grad_norm": 0.5703069263200918, "learning_rate": 4.525949630999818e-05, "loss": 0.248, "step": 10020 }, { "epoch": 4.715764705882353, "grad_norm": 0.2710259675570953, "learning_rate": 4.522812133738589e-05, "loss": 0.1143, "step": 10021 }, { "epoch": 4.716235294117647, "grad_norm": 0.6013177669108024, "learning_rate": 4.519675531280594e-05, "loss": 0.2168, "step": 10022 }, { "epoch": 4.716705882352941, "grad_norm": 0.5190189468788499, "learning_rate": 4.5165398238937194e-05, "loss": 0.249, "step": 10023 }, { "epoch": 4.717176470588235, "grad_norm": 0.7570291735177, "learning_rate": 4.51340501184576e-05, "loss": 0.4609, "step": 10024 }, { "epoch": 4.7176470588235295, "grad_norm": 0.3943002715328735, "learning_rate": 4.510271095404449e-05, "loss": 0.1582, "step": 10025 }, { "epoch": 4.718117647058824, "grad_norm": 0.5441839938545512, "learning_rate": 4.507138074837443e-05, "loss": 0.2363, "step": 10026 }, { "epoch": 4.718588235294118, "grad_norm": 0.5851459150776168, "learning_rate": 4.504005950412308e-05, "loss": 0.3203, "step": 10027 }, { "epoch": 4.719058823529412, "grad_norm": 0.4728706989256064, "learning_rate": 4.5008747223965456e-05, "loss": 0.2441, "step": 10028 }, { "epoch": 4.7195294117647055, "grad_norm": 0.2976532716335119, "learning_rate": 4.4977443910575866e-05, "loss": 0.106, "step": 10029 }, { "epoch": 4.72, "grad_norm": 0.5437283165743555, "learning_rate": 4.494614956662765e-05, "loss": 0.1875, "step": 10030 }, { "epoch": 4.720470588235294, "grad_norm": 0.6739177905088829, "learning_rate": 4.491486419479355e-05, "loss": 0.3789, "step": 10031 }, { "epoch": 4.720941176470588, "grad_norm": 0.7158473818161912, "learning_rate": 4.488358779774551e-05, "loss": 0.4922, "step": 10032 }, { "epoch": 4.721411764705882, "grad_norm": 0.38601138616747, "learning_rate": 4.485232037815465e-05, "loss": 0.1465, "step": 10033 }, { "epoch": 4.721882352941177, "grad_norm": 0.22741372954627426, "learning_rate": 4.4821061938691355e-05, "loss": 0.0464, "step": 10034 }, { "epoch": 4.722352941176471, "grad_norm": 0.42275217040120405, "learning_rate": 4.478981248202533e-05, "loss": 0.2373, "step": 10035 }, { "epoch": 4.722823529411765, "grad_norm": 0.5835514177966419, "learning_rate": 4.4758572010825324e-05, "loss": 0.2969, "step": 10036 }, { "epoch": 4.723294117647059, "grad_norm": 0.3972034501331763, "learning_rate": 4.472734052775947e-05, "loss": 0.1309, "step": 10037 }, { "epoch": 4.723764705882353, "grad_norm": 0.23614076515062302, "learning_rate": 4.4696118035495136e-05, "loss": 0.0574, "step": 10038 }, { "epoch": 4.724235294117647, "grad_norm": 0.617513090716443, "learning_rate": 4.4664904536698796e-05, "loss": 0.3594, "step": 10039 }, { "epoch": 4.724705882352941, "grad_norm": 0.5189818018509402, "learning_rate": 4.463370003403632e-05, "loss": 0.252, "step": 10040 }, { "epoch": 4.725176470588235, "grad_norm": 0.5187574763086539, "learning_rate": 4.4602504530172636e-05, "loss": 0.3262, "step": 10041 }, { "epoch": 4.7256470588235295, "grad_norm": 0.3294475179203682, "learning_rate": 4.457131802777202e-05, "loss": 0.1729, "step": 10042 }, { "epoch": 4.726117647058824, "grad_norm": 0.502368547755052, "learning_rate": 4.454014052949804e-05, "loss": 0.3945, "step": 10043 }, { "epoch": 4.726588235294118, "grad_norm": 0.43529077042347464, "learning_rate": 4.4508972038013274e-05, "loss": 0.1934, "step": 10044 }, { "epoch": 4.727058823529411, "grad_norm": 0.7282834473683907, "learning_rate": 4.4477812555979724e-05, "loss": 0.252, "step": 10045 }, { "epoch": 4.7275294117647055, "grad_norm": 0.13597966776111314, "learning_rate": 4.4446662086058634e-05, "loss": 0.0311, "step": 10046 }, { "epoch": 4.728, "grad_norm": 0.4876989321109881, "learning_rate": 4.4415520630910265e-05, "loss": 0.3555, "step": 10047 }, { "epoch": 4.728470588235294, "grad_norm": 0.5932282238912567, "learning_rate": 4.438438819319434e-05, "loss": 0.2715, "step": 10048 }, { "epoch": 4.728941176470588, "grad_norm": 0.31584299155039675, "learning_rate": 4.435326477556975e-05, "loss": 0.0708, "step": 10049 }, { "epoch": 4.729411764705882, "grad_norm": 0.49008928410364005, "learning_rate": 4.432215038069449e-05, "loss": 0.332, "step": 10050 }, { "epoch": 4.729882352941177, "grad_norm": 0.6691518851095524, "learning_rate": 4.4291045011225934e-05, "loss": 0.332, "step": 10051 }, { "epoch": 4.730352941176471, "grad_norm": 0.7193784545533057, "learning_rate": 4.4259948669820676e-05, "loss": 0.5039, "step": 10052 }, { "epoch": 4.730823529411765, "grad_norm": 0.38760634834190005, "learning_rate": 4.422886135913441e-05, "loss": 0.1465, "step": 10053 }, { "epoch": 4.731294117647058, "grad_norm": 0.4779198064505155, "learning_rate": 4.419778308182219e-05, "loss": 0.3652, "step": 10054 }, { "epoch": 4.731764705882353, "grad_norm": 0.7373812716612148, "learning_rate": 4.41667138405383e-05, "loss": 0.5938, "step": 10055 }, { "epoch": 4.732235294117647, "grad_norm": 0.7647151232743926, "learning_rate": 4.413565363793611e-05, "loss": 0.291, "step": 10056 }, { "epoch": 4.732705882352941, "grad_norm": 0.7291513555339256, "learning_rate": 4.4104602476668364e-05, "loss": 0.4824, "step": 10057 }, { "epoch": 4.733176470588235, "grad_norm": 0.6085548679351886, "learning_rate": 4.4073560359387035e-05, "loss": 0.252, "step": 10058 }, { "epoch": 4.7336470588235295, "grad_norm": 0.42486569782349193, "learning_rate": 4.4042527288743155e-05, "loss": 0.1543, "step": 10059 }, { "epoch": 4.734117647058824, "grad_norm": 0.18890180651159705, "learning_rate": 4.401150326738722e-05, "loss": 0.0417, "step": 10060 }, { "epoch": 4.734588235294118, "grad_norm": 0.47549445332384066, "learning_rate": 4.3980488297968726e-05, "loss": 0.3281, "step": 10061 }, { "epoch": 4.735058823529412, "grad_norm": 0.601883891455633, "learning_rate": 4.394948238313655e-05, "loss": 0.3301, "step": 10062 }, { "epoch": 4.7355294117647055, "grad_norm": 0.1676069709487698, "learning_rate": 4.391848552553875e-05, "loss": 0.0442, "step": 10063 }, { "epoch": 4.736, "grad_norm": 0.2249702258463268, "learning_rate": 4.3887497727822616e-05, "loss": 0.0674, "step": 10064 }, { "epoch": 4.736470588235294, "grad_norm": 0.4833170539787526, "learning_rate": 4.385651899263467e-05, "loss": 0.2539, "step": 10065 }, { "epoch": 4.736941176470588, "grad_norm": 0.5392803261858156, "learning_rate": 4.3825549322620657e-05, "loss": 0.2441, "step": 10066 }, { "epoch": 4.737411764705882, "grad_norm": 1.0673137304748521, "learning_rate": 4.3794588720425486e-05, "loss": 0.5195, "step": 10067 }, { "epoch": 4.737882352941177, "grad_norm": 0.7180520370596964, "learning_rate": 4.3763637188693404e-05, "loss": 0.3398, "step": 10068 }, { "epoch": 4.738352941176471, "grad_norm": 0.5073131547773291, "learning_rate": 4.3732694730067766e-05, "loss": 0.1426, "step": 10069 }, { "epoch": 4.738823529411764, "grad_norm": 0.6927587899427693, "learning_rate": 4.370176134719121e-05, "loss": 0.3945, "step": 10070 }, { "epoch": 4.739294117647058, "grad_norm": 0.348154325802613, "learning_rate": 4.36708370427057e-05, "loss": 0.1895, "step": 10071 }, { "epoch": 4.739764705882353, "grad_norm": 0.2814806815961958, "learning_rate": 4.363992181925218e-05, "loss": 0.0854, "step": 10072 }, { "epoch": 4.740235294117647, "grad_norm": 0.3942915745844693, "learning_rate": 4.3609015679471036e-05, "loss": 0.082, "step": 10073 }, { "epoch": 4.740705882352941, "grad_norm": 0.5280222919825972, "learning_rate": 4.357811862600185e-05, "loss": 0.2715, "step": 10074 }, { "epoch": 4.741176470588235, "grad_norm": 0.6111696993912317, "learning_rate": 4.3547230661483276e-05, "loss": 0.4648, "step": 10075 }, { "epoch": 4.7416470588235295, "grad_norm": 0.6086016602185081, "learning_rate": 4.351635178855336e-05, "loss": 0.3574, "step": 10076 }, { "epoch": 4.742117647058824, "grad_norm": 0.3167000314619259, "learning_rate": 4.348548200984934e-05, "loss": 0.1162, "step": 10077 }, { "epoch": 4.742588235294118, "grad_norm": 0.45688543896065376, "learning_rate": 4.345462132800756e-05, "loss": 0.165, "step": 10078 }, { "epoch": 4.743058823529411, "grad_norm": 0.12618008317091903, "learning_rate": 4.342376974566372e-05, "loss": 0.0554, "step": 10079 }, { "epoch": 4.7435294117647056, "grad_norm": 0.5441913737372541, "learning_rate": 4.339292726545275e-05, "loss": 0.2949, "step": 10080 }, { "epoch": 4.744, "grad_norm": 0.5072484989050091, "learning_rate": 4.3362093890008645e-05, "loss": 0.3027, "step": 10081 }, { "epoch": 4.744470588235294, "grad_norm": 0.35101846295281985, "learning_rate": 4.333126962196478e-05, "loss": 0.1152, "step": 10082 }, { "epoch": 4.744941176470588, "grad_norm": 0.45453372276795934, "learning_rate": 4.330045446395374e-05, "loss": 0.21, "step": 10083 }, { "epoch": 4.745411764705882, "grad_norm": 0.45846323975398773, "learning_rate": 4.3269648418607194e-05, "loss": 0.2402, "step": 10084 }, { "epoch": 4.745882352941177, "grad_norm": 0.5425484912260214, "learning_rate": 4.32388514885562e-05, "loss": 0.2891, "step": 10085 }, { "epoch": 4.746352941176471, "grad_norm": 0.5944413891936877, "learning_rate": 4.320806367643098e-05, "loss": 0.3027, "step": 10086 }, { "epoch": 4.746823529411765, "grad_norm": 0.38159668605037705, "learning_rate": 4.3177284984860895e-05, "loss": 0.125, "step": 10087 }, { "epoch": 4.747294117647058, "grad_norm": 0.2688805856189537, "learning_rate": 4.314651541647468e-05, "loss": 0.1211, "step": 10088 }, { "epoch": 4.747764705882353, "grad_norm": 0.4700097234286701, "learning_rate": 4.3115754973900104e-05, "loss": 0.2012, "step": 10089 }, { "epoch": 4.748235294117647, "grad_norm": 0.7921811757439158, "learning_rate": 4.308500365976434e-05, "loss": 0.377, "step": 10090 }, { "epoch": 4.748705882352941, "grad_norm": 0.5468526011196323, "learning_rate": 4.305426147669371e-05, "loss": 0.2793, "step": 10091 }, { "epoch": 4.749176470588235, "grad_norm": 0.7364832475057179, "learning_rate": 4.302352842731368e-05, "loss": 0.4883, "step": 10092 }, { "epoch": 4.7496470588235296, "grad_norm": 0.4099645169685785, "learning_rate": 4.299280451424903e-05, "loss": 0.1572, "step": 10093 }, { "epoch": 4.750117647058824, "grad_norm": 0.5471101334163602, "learning_rate": 4.296208974012379e-05, "loss": 0.1758, "step": 10094 }, { "epoch": 4.750588235294118, "grad_norm": 0.5459655168696138, "learning_rate": 4.2931384107561065e-05, "loss": 0.2852, "step": 10095 }, { "epoch": 4.751058823529412, "grad_norm": 0.8161008985373212, "learning_rate": 4.290068761918329e-05, "loss": 0.5352, "step": 10096 }, { "epoch": 4.751529411764706, "grad_norm": 0.35710150014266345, "learning_rate": 4.287000027761217e-05, "loss": 0.1768, "step": 10097 }, { "epoch": 4.752, "grad_norm": 0.6601849796004498, "learning_rate": 4.2839322085468445e-05, "loss": 0.4785, "step": 10098 }, { "epoch": 4.752470588235294, "grad_norm": 0.48534354628485576, "learning_rate": 4.280865304537223e-05, "loss": 0.3496, "step": 10099 }, { "epoch": 4.752941176470588, "grad_norm": 0.5889152038046396, "learning_rate": 4.277799315994286e-05, "loss": 0.2949, "step": 10100 }, { "epoch": 4.7534117647058824, "grad_norm": 0.8142962281176324, "learning_rate": 4.2747342431798767e-05, "loss": 0.4297, "step": 10101 }, { "epoch": 4.753882352941177, "grad_norm": 0.38375510465486745, "learning_rate": 4.271670086355768e-05, "loss": 0.1338, "step": 10102 }, { "epoch": 4.754352941176471, "grad_norm": 0.3321435165957533, "learning_rate": 4.268606845783663e-05, "loss": 0.1514, "step": 10103 }, { "epoch": 4.754823529411764, "grad_norm": 0.5813449773133608, "learning_rate": 4.265544521725165e-05, "loss": 0.3887, "step": 10104 }, { "epoch": 4.7552941176470584, "grad_norm": 0.5713686607491364, "learning_rate": 4.2624831144418166e-05, "loss": 0.3184, "step": 10105 }, { "epoch": 4.755764705882353, "grad_norm": 0.2747556295175158, "learning_rate": 4.259422624195082e-05, "loss": 0.0378, "step": 10106 }, { "epoch": 4.756235294117647, "grad_norm": 0.5068359478453813, "learning_rate": 4.2563630512463284e-05, "loss": 0.2217, "step": 10107 }, { "epoch": 4.756705882352941, "grad_norm": 0.37646655621249564, "learning_rate": 4.253304395856876e-05, "loss": 0.1465, "step": 10108 }, { "epoch": 4.757176470588235, "grad_norm": 0.5108102358881313, "learning_rate": 4.2502466582879375e-05, "loss": 0.3066, "step": 10109 }, { "epoch": 4.75764705882353, "grad_norm": 0.6372869174819221, "learning_rate": 4.2471898388006606e-05, "loss": 0.3066, "step": 10110 }, { "epoch": 4.758117647058824, "grad_norm": 0.44185520353726443, "learning_rate": 4.244133937656118e-05, "loss": 0.21, "step": 10111 }, { "epoch": 4.758588235294118, "grad_norm": 0.4363427492490562, "learning_rate": 4.2410789551152894e-05, "loss": 0.3145, "step": 10112 }, { "epoch": 4.759058823529411, "grad_norm": 0.6593250254183808, "learning_rate": 4.238024891439092e-05, "loss": 0.3008, "step": 10113 }, { "epoch": 4.759529411764706, "grad_norm": 0.6994048786918771, "learning_rate": 4.234971746888359e-05, "loss": 0.4824, "step": 10114 }, { "epoch": 4.76, "grad_norm": 0.49091930061716793, "learning_rate": 4.231919521723836e-05, "loss": 0.2451, "step": 10115 }, { "epoch": 4.760470588235294, "grad_norm": 0.3798422518453997, "learning_rate": 4.228868216206206e-05, "loss": 0.1885, "step": 10116 }, { "epoch": 4.760941176470588, "grad_norm": 0.3991943598626123, "learning_rate": 4.225817830596064e-05, "loss": 0.1553, "step": 10117 }, { "epoch": 4.7614117647058825, "grad_norm": 0.624468052486757, "learning_rate": 4.2227683651539236e-05, "loss": 0.3164, "step": 10118 }, { "epoch": 4.761882352941177, "grad_norm": 0.3720046541722495, "learning_rate": 4.21971982014023e-05, "loss": 0.126, "step": 10119 }, { "epoch": 4.762352941176471, "grad_norm": 0.5552143782233091, "learning_rate": 4.2166721958153365e-05, "loss": 0.252, "step": 10120 }, { "epoch": 4.762823529411765, "grad_norm": 0.5693206162357249, "learning_rate": 4.2136254924395305e-05, "loss": 0.4219, "step": 10121 }, { "epoch": 4.7632941176470585, "grad_norm": 0.37357468629702945, "learning_rate": 4.210579710273018e-05, "loss": 0.207, "step": 10122 }, { "epoch": 4.763764705882353, "grad_norm": 0.49646652066951047, "learning_rate": 4.207534849575916e-05, "loss": 0.3125, "step": 10123 }, { "epoch": 4.764235294117647, "grad_norm": 0.4381961705703733, "learning_rate": 4.204490910608274e-05, "loss": 0.2031, "step": 10124 }, { "epoch": 4.764705882352941, "grad_norm": 0.7802282129300172, "learning_rate": 4.201447893630065e-05, "loss": 0.4863, "step": 10125 }, { "epoch": 4.765176470588235, "grad_norm": 0.5033523362755743, "learning_rate": 4.1984057989011674e-05, "loss": 0.2441, "step": 10126 }, { "epoch": 4.76564705882353, "grad_norm": 0.35698235719397087, "learning_rate": 4.195364626681396e-05, "loss": 0.166, "step": 10127 }, { "epoch": 4.766117647058824, "grad_norm": 0.1994451492018584, "learning_rate": 4.192324377230487e-05, "loss": 0.0391, "step": 10128 }, { "epoch": 4.766588235294118, "grad_norm": 0.5659260364258039, "learning_rate": 4.1892850508080824e-05, "loss": 0.4102, "step": 10129 }, { "epoch": 4.767058823529412, "grad_norm": 0.3314844616241438, "learning_rate": 4.18624664767376e-05, "loss": 0.1426, "step": 10130 }, { "epoch": 4.767529411764706, "grad_norm": 0.14411620150451718, "learning_rate": 4.1832091680870214e-05, "loss": 0.0334, "step": 10131 }, { "epoch": 4.768, "grad_norm": 0.8014630288581122, "learning_rate": 4.1801726123072696e-05, "loss": 0.4004, "step": 10132 }, { "epoch": 4.768470588235294, "grad_norm": 0.4868591987901477, "learning_rate": 4.1771369805938486e-05, "loss": 0.3164, "step": 10133 }, { "epoch": 4.768941176470588, "grad_norm": 0.45793083171398546, "learning_rate": 4.1741022732060204e-05, "loss": 0.1455, "step": 10134 }, { "epoch": 4.7694117647058825, "grad_norm": 0.4127277797930258, "learning_rate": 4.171068490402954e-05, "loss": 0.125, "step": 10135 }, { "epoch": 4.769882352941177, "grad_norm": 0.42848856060894647, "learning_rate": 4.168035632443754e-05, "loss": 0.1572, "step": 10136 }, { "epoch": 4.770352941176471, "grad_norm": 0.660544455796012, "learning_rate": 4.165003699587445e-05, "loss": 0.2598, "step": 10137 }, { "epoch": 4.770823529411764, "grad_norm": 0.6855585900798078, "learning_rate": 4.161972692092962e-05, "loss": 0.2793, "step": 10138 }, { "epoch": 4.7712941176470585, "grad_norm": 0.6157921419772259, "learning_rate": 4.158942610219175e-05, "loss": 0.3066, "step": 10139 }, { "epoch": 4.771764705882353, "grad_norm": 0.2959171096106696, "learning_rate": 4.155913454224862e-05, "loss": 0.0952, "step": 10140 }, { "epoch": 4.772235294117647, "grad_norm": 0.36426027956682133, "learning_rate": 4.1528852243687276e-05, "loss": 0.1885, "step": 10141 }, { "epoch": 4.772705882352941, "grad_norm": 0.18810142031278698, "learning_rate": 4.1498579209094046e-05, "loss": 0.04, "step": 10142 }, { "epoch": 4.773176470588235, "grad_norm": 0.4862064240388657, "learning_rate": 4.146831544105431e-05, "loss": 0.2598, "step": 10143 }, { "epoch": 4.77364705882353, "grad_norm": 0.5100971259965547, "learning_rate": 4.143806094215277e-05, "loss": 0.3223, "step": 10144 }, { "epoch": 4.774117647058824, "grad_norm": 0.643082190393194, "learning_rate": 4.140781571497336e-05, "loss": 0.1914, "step": 10145 }, { "epoch": 4.774588235294118, "grad_norm": 0.35813501160142, "learning_rate": 4.137757976209911e-05, "loss": 0.1328, "step": 10146 }, { "epoch": 4.775058823529411, "grad_norm": 0.48168435009321026, "learning_rate": 4.134735308611232e-05, "loss": 0.2676, "step": 10147 }, { "epoch": 4.775529411764706, "grad_norm": 0.5949960811041644, "learning_rate": 4.131713568959457e-05, "loss": 0.1953, "step": 10148 }, { "epoch": 4.776, "grad_norm": 0.4557143333570478, "learning_rate": 4.128692757512646e-05, "loss": 0.1797, "step": 10149 }, { "epoch": 4.776470588235294, "grad_norm": 0.2512201973206024, "learning_rate": 4.125672874528797e-05, "loss": 0.1079, "step": 10150 }, { "epoch": 4.776941176470588, "grad_norm": 0.3081912316348812, "learning_rate": 4.122653920265828e-05, "loss": 0.1328, "step": 10151 }, { "epoch": 4.7774117647058825, "grad_norm": 0.4284765931239098, "learning_rate": 4.119635894981558e-05, "loss": 0.208, "step": 10152 }, { "epoch": 4.777882352941177, "grad_norm": 0.38545224574107045, "learning_rate": 4.1166187989337596e-05, "loss": 0.249, "step": 10153 }, { "epoch": 4.778352941176471, "grad_norm": 0.3011589712023289, "learning_rate": 4.113602632380094e-05, "loss": 0.0549, "step": 10154 }, { "epoch": 4.778823529411765, "grad_norm": 0.6760337164346164, "learning_rate": 4.110587395578161e-05, "loss": 0.3516, "step": 10155 }, { "epoch": 4.7792941176470585, "grad_norm": 0.379351732877705, "learning_rate": 4.1075730887854805e-05, "loss": 0.1152, "step": 10156 }, { "epoch": 4.779764705882353, "grad_norm": 0.3666188901241578, "learning_rate": 4.104559712259482e-05, "loss": 0.1895, "step": 10157 }, { "epoch": 4.780235294117647, "grad_norm": 0.5294094021581963, "learning_rate": 4.101547266257526e-05, "loss": 0.2637, "step": 10158 }, { "epoch": 4.780705882352941, "grad_norm": 0.5000443572517385, "learning_rate": 4.098535751036894e-05, "loss": 0.2949, "step": 10159 }, { "epoch": 4.781176470588235, "grad_norm": 0.3666681882513906, "learning_rate": 4.0955251668547754e-05, "loss": 0.0674, "step": 10160 }, { "epoch": 4.78164705882353, "grad_norm": 0.6740472370182558, "learning_rate": 4.092515513968294e-05, "loss": 0.2969, "step": 10161 }, { "epoch": 4.782117647058824, "grad_norm": 1.055419106917281, "learning_rate": 4.0895067926344935e-05, "loss": 0.6328, "step": 10162 }, { "epoch": 4.782588235294118, "grad_norm": 0.3363989061654442, "learning_rate": 4.0864990031103235e-05, "loss": 0.1064, "step": 10163 }, { "epoch": 4.783058823529412, "grad_norm": 0.24756713013087916, "learning_rate": 4.0834921456526684e-05, "loss": 0.1182, "step": 10164 }, { "epoch": 4.783529411764706, "grad_norm": 0.686191687918933, "learning_rate": 4.080486220518336e-05, "loss": 0.4062, "step": 10165 }, { "epoch": 4.784, "grad_norm": 0.22874566333866825, "learning_rate": 4.0774812279640325e-05, "loss": 0.0879, "step": 10166 }, { "epoch": 4.784470588235294, "grad_norm": 0.5659773254761393, "learning_rate": 4.074477168246409e-05, "loss": 0.457, "step": 10167 }, { "epoch": 4.784941176470588, "grad_norm": 0.7441930097925997, "learning_rate": 4.071474041622029e-05, "loss": 0.5352, "step": 10168 }, { "epoch": 4.7854117647058825, "grad_norm": 0.28712930281542187, "learning_rate": 4.0684718483473656e-05, "loss": 0.0732, "step": 10169 }, { "epoch": 4.785882352941177, "grad_norm": 0.554821392174569, "learning_rate": 4.06547058867883e-05, "loss": 0.2969, "step": 10170 }, { "epoch": 4.786352941176471, "grad_norm": 0.5714540601514089, "learning_rate": 4.0624702628727345e-05, "loss": 0.3281, "step": 10171 }, { "epoch": 4.786823529411764, "grad_norm": 0.6578152779001696, "learning_rate": 4.0594708711853267e-05, "loss": 0.5312, "step": 10172 }, { "epoch": 4.7872941176470585, "grad_norm": 0.6529324315722582, "learning_rate": 4.0564724138727734e-05, "loss": 0.249, "step": 10173 }, { "epoch": 4.787764705882353, "grad_norm": 0.7350983670866336, "learning_rate": 4.0534748911911505e-05, "loss": 0.5508, "step": 10174 }, { "epoch": 4.788235294117647, "grad_norm": 0.3440396429106516, "learning_rate": 4.0504783033964645e-05, "loss": 0.1553, "step": 10175 }, { "epoch": 4.788705882352941, "grad_norm": 0.11547845980601082, "learning_rate": 4.047482650744641e-05, "loss": 0.0303, "step": 10176 }, { "epoch": 4.789176470588235, "grad_norm": 0.5398826376054684, "learning_rate": 4.044487933491518e-05, "loss": 0.3105, "step": 10177 }, { "epoch": 4.78964705882353, "grad_norm": 0.3771299428145957, "learning_rate": 4.041494151892862e-05, "loss": 0.1089, "step": 10178 }, { "epoch": 4.790117647058824, "grad_norm": 0.4013222315405901, "learning_rate": 4.038501306204361e-05, "loss": 0.1602, "step": 10179 }, { "epoch": 4.790588235294118, "grad_norm": 0.592586801926179, "learning_rate": 4.0355093966816095e-05, "loss": 0.3262, "step": 10180 }, { "epoch": 4.791058823529411, "grad_norm": 0.5105585242223802, "learning_rate": 4.032518423580135e-05, "loss": 0.2988, "step": 10181 }, { "epoch": 4.791529411764706, "grad_norm": 0.6423644052633076, "learning_rate": 4.029528387155389e-05, "loss": 0.3496, "step": 10182 }, { "epoch": 4.792, "grad_norm": 0.7184943194234852, "learning_rate": 4.026539287662722e-05, "loss": 0.5039, "step": 10183 }, { "epoch": 4.792470588235294, "grad_norm": 0.8436891161332909, "learning_rate": 4.0235511253574256e-05, "loss": 0.167, "step": 10184 }, { "epoch": 4.792941176470588, "grad_norm": 0.5765509173945238, "learning_rate": 4.020563900494707e-05, "loss": 0.3457, "step": 10185 }, { "epoch": 4.7934117647058825, "grad_norm": 0.7837698656206555, "learning_rate": 4.01757761332968e-05, "loss": 0.2812, "step": 10186 }, { "epoch": 4.793882352941177, "grad_norm": 0.40293376538697545, "learning_rate": 4.0145922641173934e-05, "loss": 0.1455, "step": 10187 }, { "epoch": 4.794352941176471, "grad_norm": 0.5473353884717198, "learning_rate": 4.011607853112815e-05, "loss": 0.2314, "step": 10188 }, { "epoch": 4.794823529411765, "grad_norm": 0.5149786234434788, "learning_rate": 4.008624380570819e-05, "loss": 0.2285, "step": 10189 }, { "epoch": 4.7952941176470585, "grad_norm": 0.4069873406016419, "learning_rate": 4.0056418467462184e-05, "loss": 0.1484, "step": 10190 }, { "epoch": 4.795764705882353, "grad_norm": 0.5217325281068971, "learning_rate": 4.002660251893727e-05, "loss": 0.2656, "step": 10191 }, { "epoch": 4.796235294117647, "grad_norm": 0.6167239305769846, "learning_rate": 3.9996795962679896e-05, "loss": 0.293, "step": 10192 }, { "epoch": 4.796705882352941, "grad_norm": 0.31109802910434536, "learning_rate": 3.9966998801235766e-05, "loss": 0.1167, "step": 10193 }, { "epoch": 4.797176470588235, "grad_norm": 0.5321865596496287, "learning_rate": 3.9937211037149616e-05, "loss": 0.3203, "step": 10194 }, { "epoch": 4.79764705882353, "grad_norm": 0.5518847789343289, "learning_rate": 3.9907432672965485e-05, "loss": 0.2695, "step": 10195 }, { "epoch": 4.798117647058824, "grad_norm": 0.5695243916383673, "learning_rate": 3.987766371122661e-05, "loss": 0.3047, "step": 10196 }, { "epoch": 4.798588235294117, "grad_norm": 0.506250650772478, "learning_rate": 3.984790415447539e-05, "loss": 0.207, "step": 10197 }, { "epoch": 4.799058823529411, "grad_norm": 0.18740159210149782, "learning_rate": 3.9818154005253514e-05, "loss": 0.1699, "step": 10198 }, { "epoch": 4.799529411764706, "grad_norm": 0.3464427539041273, "learning_rate": 3.978841326610166e-05, "loss": 0.167, "step": 10199 }, { "epoch": 4.8, "grad_norm": 0.6915887739850037, "learning_rate": 3.97586819395599e-05, "loss": 0.5859, "step": 10200 }, { "epoch": 4.800470588235294, "grad_norm": 0.3105029417875797, "learning_rate": 3.972896002816749e-05, "loss": 0.1318, "step": 10201 }, { "epoch": 4.800941176470588, "grad_norm": 0.26298515313289006, "learning_rate": 3.969924753446271e-05, "loss": 0.0981, "step": 10202 }, { "epoch": 4.8014117647058825, "grad_norm": 0.6023489631789526, "learning_rate": 3.966954446098322e-05, "loss": 0.3223, "step": 10203 }, { "epoch": 4.801882352941177, "grad_norm": 0.5916878051100731, "learning_rate": 3.9639850810265845e-05, "loss": 0.3164, "step": 10204 }, { "epoch": 4.802352941176471, "grad_norm": 0.47874842103814425, "learning_rate": 3.961016658484648e-05, "loss": 0.2734, "step": 10205 }, { "epoch": 4.802823529411764, "grad_norm": 0.46713320997500113, "learning_rate": 3.9580491787260346e-05, "loss": 0.1484, "step": 10206 }, { "epoch": 4.8032941176470585, "grad_norm": 0.49151365955201787, "learning_rate": 3.955082642004187e-05, "loss": 0.209, "step": 10207 }, { "epoch": 4.803764705882353, "grad_norm": 0.1684077427560434, "learning_rate": 3.952117048572452e-05, "loss": 0.0297, "step": 10208 }, { "epoch": 4.804235294117647, "grad_norm": 0.5606764216333724, "learning_rate": 3.94915239868411e-05, "loss": 0.3301, "step": 10209 }, { "epoch": 4.804705882352941, "grad_norm": 0.549324333149064, "learning_rate": 3.946188692592362e-05, "loss": 0.291, "step": 10210 }, { "epoch": 4.805176470588235, "grad_norm": 0.6688449229707567, "learning_rate": 3.943225930550315e-05, "loss": 0.4062, "step": 10211 }, { "epoch": 4.80564705882353, "grad_norm": 0.44207466865693495, "learning_rate": 3.9402641128110054e-05, "loss": 0.2773, "step": 10212 }, { "epoch": 4.806117647058824, "grad_norm": 0.43307135490535204, "learning_rate": 3.937303239627392e-05, "loss": 0.2109, "step": 10213 }, { "epoch": 4.806588235294118, "grad_norm": 0.5036195945354657, "learning_rate": 3.934343311252341e-05, "loss": 0.2832, "step": 10214 }, { "epoch": 4.807058823529411, "grad_norm": 0.5739187108483116, "learning_rate": 3.9313843279386486e-05, "loss": 0.3262, "step": 10215 }, { "epoch": 4.807529411764706, "grad_norm": 0.8881561235091734, "learning_rate": 3.9284262899390285e-05, "loss": 0.5156, "step": 10216 }, { "epoch": 4.808, "grad_norm": 0.3123064766544124, "learning_rate": 3.925469197506107e-05, "loss": 0.1196, "step": 10217 }, { "epoch": 4.808470588235294, "grad_norm": 0.6965768455086383, "learning_rate": 3.922513050892434e-05, "loss": 0.4336, "step": 10218 }, { "epoch": 4.808941176470588, "grad_norm": 0.5322728154835961, "learning_rate": 3.919557850350486e-05, "loss": 0.1973, "step": 10219 }, { "epoch": 4.8094117647058825, "grad_norm": 0.32275534358789454, "learning_rate": 3.916603596132642e-05, "loss": 0.1182, "step": 10220 }, { "epoch": 4.809882352941177, "grad_norm": 0.314692011406404, "learning_rate": 3.91365028849122e-05, "loss": 0.1895, "step": 10221 }, { "epoch": 4.810352941176471, "grad_norm": 0.5304723292293098, "learning_rate": 3.910697927678437e-05, "loss": 0.2871, "step": 10222 }, { "epoch": 4.810823529411765, "grad_norm": 0.4617980156609613, "learning_rate": 3.9077465139464446e-05, "loss": 0.25, "step": 10223 }, { "epoch": 4.8112941176470585, "grad_norm": 0.3081631131692529, "learning_rate": 3.9047960475473095e-05, "loss": 0.1177, "step": 10224 }, { "epoch": 4.811764705882353, "grad_norm": 0.2926959372698411, "learning_rate": 3.901846528733011e-05, "loss": 0.0869, "step": 10225 }, { "epoch": 4.812235294117647, "grad_norm": 0.5960864366136411, "learning_rate": 3.898897957755454e-05, "loss": 0.4043, "step": 10226 }, { "epoch": 4.812705882352941, "grad_norm": 0.5530292031684554, "learning_rate": 3.895950334866468e-05, "loss": 0.2832, "step": 10227 }, { "epoch": 4.813176470588235, "grad_norm": 0.4674935850932742, "learning_rate": 3.893003660317783e-05, "loss": 0.166, "step": 10228 }, { "epoch": 4.81364705882353, "grad_norm": 0.49702445555018576, "learning_rate": 3.8900579343610646e-05, "loss": 0.2021, "step": 10229 }, { "epoch": 4.814117647058824, "grad_norm": 0.30990528727675787, "learning_rate": 3.887113157247897e-05, "loss": 0.1118, "step": 10230 }, { "epoch": 4.814588235294117, "grad_norm": 0.2271577540881775, "learning_rate": 3.884169329229769e-05, "loss": 0.1235, "step": 10231 }, { "epoch": 4.815058823529411, "grad_norm": 0.5852124248573084, "learning_rate": 3.881226450558104e-05, "loss": 0.2695, "step": 10232 }, { "epoch": 4.815529411764706, "grad_norm": 0.12765783278718645, "learning_rate": 3.878284521484242e-05, "loss": 0.0251, "step": 10233 }, { "epoch": 4.816, "grad_norm": 0.2448685497637796, "learning_rate": 3.8753435422594295e-05, "loss": 0.0613, "step": 10234 }, { "epoch": 4.816470588235294, "grad_norm": 0.31691610367834355, "learning_rate": 3.8724035131348437e-05, "loss": 0.0923, "step": 10235 }, { "epoch": 4.816941176470588, "grad_norm": 0.525738257214034, "learning_rate": 3.8694644343615836e-05, "loss": 0.1553, "step": 10236 }, { "epoch": 4.8174117647058825, "grad_norm": 0.5471520778463935, "learning_rate": 3.866526306190651e-05, "loss": 0.1611, "step": 10237 }, { "epoch": 4.817882352941177, "grad_norm": 0.4190170359385134, "learning_rate": 3.863589128872981e-05, "loss": 0.2275, "step": 10238 }, { "epoch": 4.818352941176471, "grad_norm": 0.5905033028413987, "learning_rate": 3.8606529026594276e-05, "loss": 0.3945, "step": 10239 }, { "epoch": 4.818823529411764, "grad_norm": 0.5905852948094934, "learning_rate": 3.85771762780075e-05, "loss": 0.3301, "step": 10240 }, { "epoch": 4.8192941176470585, "grad_norm": 0.5645101435980092, "learning_rate": 3.854783304547639e-05, "loss": 0.2676, "step": 10241 }, { "epoch": 4.819764705882353, "grad_norm": 0.15366288676063886, "learning_rate": 3.851849933150703e-05, "loss": 0.0223, "step": 10242 }, { "epoch": 4.820235294117647, "grad_norm": 0.39619455119449287, "learning_rate": 3.848917513860463e-05, "loss": 0.1748, "step": 10243 }, { "epoch": 4.820705882352941, "grad_norm": 0.41312320710683725, "learning_rate": 3.8459860469273666e-05, "loss": 0.1523, "step": 10244 }, { "epoch": 4.821176470588235, "grad_norm": 0.7111769231058923, "learning_rate": 3.8430555326017684e-05, "loss": 0.4219, "step": 10245 }, { "epoch": 4.82164705882353, "grad_norm": 0.39191945377494686, "learning_rate": 3.840125971133952e-05, "loss": 0.0972, "step": 10246 }, { "epoch": 4.822117647058824, "grad_norm": 0.4924627456756279, "learning_rate": 3.8371973627741204e-05, "loss": 0.2295, "step": 10247 }, { "epoch": 4.822588235294118, "grad_norm": 0.4724488076310831, "learning_rate": 3.8342697077723845e-05, "loss": 0.1299, "step": 10248 }, { "epoch": 4.823058823529411, "grad_norm": 0.21400751186612785, "learning_rate": 3.8313430063787856e-05, "loss": 0.0708, "step": 10249 }, { "epoch": 4.823529411764706, "grad_norm": 0.4975493006100508, "learning_rate": 3.8284172588432716e-05, "loss": 0.2432, "step": 10250 }, { "epoch": 4.824, "grad_norm": 0.7396028232031235, "learning_rate": 3.8254924654157214e-05, "loss": 0.3164, "step": 10251 }, { "epoch": 4.824470588235294, "grad_norm": 0.8999027899719916, "learning_rate": 3.8225686263459296e-05, "loss": 0.5312, "step": 10252 }, { "epoch": 4.824941176470588, "grad_norm": 0.506813655200325, "learning_rate": 3.819645741883597e-05, "loss": 0.2188, "step": 10253 }, { "epoch": 4.8254117647058825, "grad_norm": 0.5340430815243484, "learning_rate": 3.816723812278358e-05, "loss": 0.2227, "step": 10254 }, { "epoch": 4.825882352941177, "grad_norm": 0.5005092139495448, "learning_rate": 3.813802837779762e-05, "loss": 0.126, "step": 10255 }, { "epoch": 4.826352941176471, "grad_norm": 0.5070271205812387, "learning_rate": 3.810882818637268e-05, "loss": 0.2969, "step": 10256 }, { "epoch": 4.826823529411765, "grad_norm": 0.7616470863885524, "learning_rate": 3.807963755100263e-05, "loss": 0.5195, "step": 10257 }, { "epoch": 4.8272941176470585, "grad_norm": 0.5919851077480636, "learning_rate": 3.8050456474180554e-05, "loss": 0.375, "step": 10258 }, { "epoch": 4.827764705882353, "grad_norm": 0.3845389131330123, "learning_rate": 3.802128495839856e-05, "loss": 0.1611, "step": 10259 }, { "epoch": 4.828235294117647, "grad_norm": 0.8776461226175615, "learning_rate": 3.799212300614809e-05, "loss": 0.5938, "step": 10260 }, { "epoch": 4.828705882352941, "grad_norm": 0.49798517719798985, "learning_rate": 3.7962970619919745e-05, "loss": 0.2363, "step": 10261 }, { "epoch": 4.829176470588235, "grad_norm": 0.5522509023510924, "learning_rate": 3.7933827802203207e-05, "loss": 0.4668, "step": 10262 }, { "epoch": 4.82964705882353, "grad_norm": 0.46317132855782744, "learning_rate": 3.790469455548745e-05, "loss": 0.1973, "step": 10263 }, { "epoch": 4.830117647058824, "grad_norm": 0.634091003894086, "learning_rate": 3.787557088226065e-05, "loss": 0.4258, "step": 10264 }, { "epoch": 4.830588235294117, "grad_norm": 0.7448873451001501, "learning_rate": 3.7846456785010014e-05, "loss": 0.4434, "step": 10265 }, { "epoch": 4.831058823529411, "grad_norm": 0.6448387766853627, "learning_rate": 3.7817352266222086e-05, "loss": 0.3008, "step": 10266 }, { "epoch": 4.831529411764706, "grad_norm": 0.72253611345198, "learning_rate": 3.778825732838257e-05, "loss": 0.3516, "step": 10267 }, { "epoch": 4.832, "grad_norm": 0.301687049141082, "learning_rate": 3.775917197397622e-05, "loss": 0.0874, "step": 10268 }, { "epoch": 4.832470588235294, "grad_norm": 0.43804250443037024, "learning_rate": 3.773009620548718e-05, "loss": 0.1221, "step": 10269 }, { "epoch": 4.832941176470588, "grad_norm": 0.6681916469075466, "learning_rate": 3.770103002539857e-05, "loss": 0.4648, "step": 10270 }, { "epoch": 4.8334117647058825, "grad_norm": 0.5805469607269577, "learning_rate": 3.767197343619279e-05, "loss": 0.3496, "step": 10271 }, { "epoch": 4.833882352941177, "grad_norm": 0.5030768313827545, "learning_rate": 3.7642926440351514e-05, "loss": 0.2188, "step": 10272 }, { "epoch": 4.834352941176471, "grad_norm": 0.29250173729464446, "learning_rate": 3.761388904035537e-05, "loss": 0.0747, "step": 10273 }, { "epoch": 4.834823529411764, "grad_norm": 0.7985424733656722, "learning_rate": 3.758486123868435e-05, "loss": 0.2637, "step": 10274 }, { "epoch": 4.8352941176470585, "grad_norm": 0.3606596240686745, "learning_rate": 3.755584303781763e-05, "loss": 0.0938, "step": 10275 }, { "epoch": 4.835764705882353, "grad_norm": 0.6091041776508532, "learning_rate": 3.7526834440233395e-05, "loss": 0.3242, "step": 10276 }, { "epoch": 4.836235294117647, "grad_norm": 0.4467436315873043, "learning_rate": 3.749783544840918e-05, "loss": 0.1289, "step": 10277 }, { "epoch": 4.836705882352941, "grad_norm": 0.6233803705913299, "learning_rate": 3.746884606482169e-05, "loss": 0.4004, "step": 10278 }, { "epoch": 4.837176470588235, "grad_norm": 0.04516783487088627, "learning_rate": 3.743986629194666e-05, "loss": 0.0056, "step": 10279 }, { "epoch": 4.83764705882353, "grad_norm": 0.7247995150371889, "learning_rate": 3.741089613225914e-05, "loss": 0.2734, "step": 10280 }, { "epoch": 4.838117647058824, "grad_norm": 0.49998713740938616, "learning_rate": 3.7381935588233385e-05, "loss": 0.2354, "step": 10281 }, { "epoch": 4.838588235294118, "grad_norm": 0.47811590566029805, "learning_rate": 3.735298466234269e-05, "loss": 0.2334, "step": 10282 }, { "epoch": 4.839058823529411, "grad_norm": 0.359537873648783, "learning_rate": 3.732404335705962e-05, "loss": 0.1895, "step": 10283 }, { "epoch": 4.839529411764706, "grad_norm": 0.5036520878247916, "learning_rate": 3.729511167485597e-05, "loss": 0.2266, "step": 10284 }, { "epoch": 4.84, "grad_norm": 0.49428192563563733, "learning_rate": 3.72661896182025e-05, "loss": 0.2734, "step": 10285 }, { "epoch": 4.840470588235294, "grad_norm": 0.7286550686833142, "learning_rate": 3.7237277189569505e-05, "loss": 0.4668, "step": 10286 }, { "epoch": 4.840941176470588, "grad_norm": 0.623753256869191, "learning_rate": 3.720837439142608e-05, "loss": 0.3242, "step": 10287 }, { "epoch": 4.8414117647058825, "grad_norm": 0.6524870870706805, "learning_rate": 3.717948122624072e-05, "loss": 0.373, "step": 10288 }, { "epoch": 4.841882352941177, "grad_norm": 0.5957053286845694, "learning_rate": 3.7150597696481076e-05, "loss": 0.3945, "step": 10289 }, { "epoch": 4.842352941176471, "grad_norm": 0.7181971355554717, "learning_rate": 3.712172380461387e-05, "loss": 0.3965, "step": 10290 }, { "epoch": 4.842823529411765, "grad_norm": 0.529703139616152, "learning_rate": 3.709285955310513e-05, "loss": 0.2852, "step": 10291 }, { "epoch": 4.8432941176470585, "grad_norm": 0.5007844081067768, "learning_rate": 3.7064004944420026e-05, "loss": 0.3027, "step": 10292 }, { "epoch": 4.843764705882353, "grad_norm": 0.12409587943433369, "learning_rate": 3.703515998102281e-05, "loss": 0.0315, "step": 10293 }, { "epoch": 4.844235294117647, "grad_norm": 0.4507270808076696, "learning_rate": 3.7006324665377025e-05, "loss": 0.2246, "step": 10294 }, { "epoch": 4.844705882352941, "grad_norm": 0.4176810940508414, "learning_rate": 3.6977498999945384e-05, "loss": 0.252, "step": 10295 }, { "epoch": 4.845176470588235, "grad_norm": 0.5646295678922222, "learning_rate": 3.694868298718965e-05, "loss": 0.3398, "step": 10296 }, { "epoch": 4.84564705882353, "grad_norm": 0.43332314458818566, "learning_rate": 3.6919876629570905e-05, "loss": 0.2676, "step": 10297 }, { "epoch": 4.846117647058824, "grad_norm": 0.21292026557157123, "learning_rate": 3.68910799295494e-05, "loss": 0.0552, "step": 10298 }, { "epoch": 4.846588235294117, "grad_norm": 0.49558627045401044, "learning_rate": 3.686229288958442e-05, "loss": 0.4082, "step": 10299 }, { "epoch": 4.847058823529411, "grad_norm": 0.20628495094836238, "learning_rate": 3.6833515512134606e-05, "loss": 0.0613, "step": 10300 }, { "epoch": 4.847529411764706, "grad_norm": 0.7112224336398583, "learning_rate": 3.6804747799657594e-05, "loss": 0.3418, "step": 10301 }, { "epoch": 4.848, "grad_norm": 0.2861087329275562, "learning_rate": 3.677598975461036e-05, "loss": 0.0947, "step": 10302 }, { "epoch": 4.848470588235294, "grad_norm": 0.6306905373299666, "learning_rate": 3.6747241379448986e-05, "loss": 0.2812, "step": 10303 }, { "epoch": 4.848941176470588, "grad_norm": 0.5448826830891103, "learning_rate": 3.671850267662866e-05, "loss": 0.3203, "step": 10304 }, { "epoch": 4.8494117647058825, "grad_norm": 0.3255282811080747, "learning_rate": 3.668977364860386e-05, "loss": 0.1045, "step": 10305 }, { "epoch": 4.849882352941177, "grad_norm": 0.43300458031592043, "learning_rate": 3.666105429782821e-05, "loss": 0.1836, "step": 10306 }, { "epoch": 4.850352941176471, "grad_norm": 0.7578388145782942, "learning_rate": 3.6632344626754405e-05, "loss": 0.0786, "step": 10307 }, { "epoch": 4.850823529411764, "grad_norm": 0.45143124091839126, "learning_rate": 3.660364463783443e-05, "loss": 0.1157, "step": 10308 }, { "epoch": 4.8512941176470585, "grad_norm": 0.505209590591841, "learning_rate": 3.657495433351945e-05, "loss": 0.3594, "step": 10309 }, { "epoch": 4.851764705882353, "grad_norm": 0.4474107642284152, "learning_rate": 3.6546273716259685e-05, "loss": 0.2412, "step": 10310 }, { "epoch": 4.852235294117647, "grad_norm": 1.0164484722182023, "learning_rate": 3.6517602788504613e-05, "loss": 0.4805, "step": 10311 }, { "epoch": 4.852705882352941, "grad_norm": 0.3451688800257848, "learning_rate": 3.648894155270295e-05, "loss": 0.1133, "step": 10312 }, { "epoch": 4.853176470588235, "grad_norm": 1.1114338411401894, "learning_rate": 3.64602900113024e-05, "loss": 0.1543, "step": 10313 }, { "epoch": 4.85364705882353, "grad_norm": 0.4343287388233276, "learning_rate": 3.643164816674999e-05, "loss": 0.2559, "step": 10314 }, { "epoch": 4.854117647058824, "grad_norm": 0.8737141431681436, "learning_rate": 3.6403016021491924e-05, "loss": 0.6484, "step": 10315 }, { "epoch": 4.854588235294118, "grad_norm": 0.5021926674667965, "learning_rate": 3.637439357797343e-05, "loss": 0.3203, "step": 10316 }, { "epoch": 4.855058823529411, "grad_norm": 0.43936551020785475, "learning_rate": 3.634578083863905e-05, "loss": 0.2041, "step": 10317 }, { "epoch": 4.855529411764706, "grad_norm": 0.7795483216455502, "learning_rate": 3.631717780593252e-05, "loss": 0.4531, "step": 10318 }, { "epoch": 4.856, "grad_norm": 0.6142184740247002, "learning_rate": 3.628858448229655e-05, "loss": 0.2305, "step": 10319 }, { "epoch": 4.856470588235294, "grad_norm": 0.5298938058699083, "learning_rate": 3.6260000870173266e-05, "loss": 0.3027, "step": 10320 }, { "epoch": 4.856941176470588, "grad_norm": 0.5262251535084154, "learning_rate": 3.6231426972003766e-05, "loss": 0.2275, "step": 10321 }, { "epoch": 4.8574117647058825, "grad_norm": 0.2434261081653557, "learning_rate": 3.620286279022843e-05, "loss": 0.0664, "step": 10322 }, { "epoch": 4.857882352941177, "grad_norm": 0.3299800290502977, "learning_rate": 3.617430832728682e-05, "loss": 0.1465, "step": 10323 }, { "epoch": 4.85835294117647, "grad_norm": 0.8308133707976572, "learning_rate": 3.614576358561757e-05, "loss": 0.3867, "step": 10324 }, { "epoch": 4.858823529411764, "grad_norm": 0.147893502602297, "learning_rate": 3.611722856765855e-05, "loss": 0.0356, "step": 10325 }, { "epoch": 4.8592941176470585, "grad_norm": 0.6302460008944444, "learning_rate": 3.608870327584683e-05, "loss": 0.4375, "step": 10326 }, { "epoch": 4.859764705882353, "grad_norm": 0.6398310032097232, "learning_rate": 3.606018771261856e-05, "loss": 0.4375, "step": 10327 }, { "epoch": 4.860235294117647, "grad_norm": 0.5630160611262078, "learning_rate": 3.6031681880409106e-05, "loss": 0.2656, "step": 10328 }, { "epoch": 4.860705882352941, "grad_norm": 0.7542320552095712, "learning_rate": 3.6003185781653097e-05, "loss": 0.4277, "step": 10329 }, { "epoch": 4.861176470588235, "grad_norm": 0.21413566256042008, "learning_rate": 3.5974699418784084e-05, "loss": 0.0703, "step": 10330 }, { "epoch": 4.86164705882353, "grad_norm": 0.49172210084851603, "learning_rate": 3.5946222794235105e-05, "loss": 0.2148, "step": 10331 }, { "epoch": 4.862117647058824, "grad_norm": 0.5804450579281912, "learning_rate": 3.5917755910438094e-05, "loss": 0.3574, "step": 10332 }, { "epoch": 4.862588235294117, "grad_norm": 0.7390776336027618, "learning_rate": 3.58892987698243e-05, "loss": 0.4316, "step": 10333 }, { "epoch": 4.863058823529411, "grad_norm": 0.3423544731518597, "learning_rate": 3.5860851374824134e-05, "loss": 0.0767, "step": 10334 }, { "epoch": 4.863529411764706, "grad_norm": 0.5406391964189533, "learning_rate": 3.5832413727867075e-05, "loss": 0.332, "step": 10335 }, { "epoch": 4.864, "grad_norm": 0.3939911044877195, "learning_rate": 3.580398583138188e-05, "loss": 0.1289, "step": 10336 }, { "epoch": 4.864470588235294, "grad_norm": 0.6301061709915876, "learning_rate": 3.577556768779646e-05, "loss": 0.3145, "step": 10337 }, { "epoch": 4.864941176470588, "grad_norm": 0.364510656109253, "learning_rate": 3.574715929953778e-05, "loss": 0.1387, "step": 10338 }, { "epoch": 4.8654117647058825, "grad_norm": 0.5414590868951956, "learning_rate": 3.571876066903211e-05, "loss": 0.3496, "step": 10339 }, { "epoch": 4.865882352941177, "grad_norm": 0.39459544095900995, "learning_rate": 3.569037179870487e-05, "loss": 0.1416, "step": 10340 }, { "epoch": 4.866352941176471, "grad_norm": 0.49361932329971525, "learning_rate": 3.5661992690980506e-05, "loss": 0.168, "step": 10341 }, { "epoch": 4.866823529411764, "grad_norm": 0.6534559233375916, "learning_rate": 3.56336233482828e-05, "loss": 0.3848, "step": 10342 }, { "epoch": 4.8672941176470585, "grad_norm": 0.2846218522667248, "learning_rate": 3.560526377303467e-05, "loss": 0.0522, "step": 10343 }, { "epoch": 4.867764705882353, "grad_norm": 0.48777342196631834, "learning_rate": 3.557691396765807e-05, "loss": 0.2168, "step": 10344 }, { "epoch": 4.868235294117647, "grad_norm": 0.11950153169150697, "learning_rate": 3.554857393457426e-05, "loss": 0.0273, "step": 10345 }, { "epoch": 4.868705882352941, "grad_norm": 0.5215313432330921, "learning_rate": 3.5520243676203664e-05, "loss": 0.1963, "step": 10346 }, { "epoch": 4.869176470588235, "grad_norm": 0.7398703791404384, "learning_rate": 3.5491923194965735e-05, "loss": 0.2559, "step": 10347 }, { "epoch": 4.86964705882353, "grad_norm": 0.3852432157562847, "learning_rate": 3.546361249327922e-05, "loss": 0.2188, "step": 10348 }, { "epoch": 4.870117647058824, "grad_norm": 0.5823562862742546, "learning_rate": 3.5435311573562024e-05, "loss": 0.2324, "step": 10349 }, { "epoch": 4.870588235294118, "grad_norm": 0.4122771512879172, "learning_rate": 3.540702043823113e-05, "loss": 0.1602, "step": 10350 }, { "epoch": 4.871058823529411, "grad_norm": 0.480834536536057, "learning_rate": 3.537873908970282e-05, "loss": 0.4062, "step": 10351 }, { "epoch": 4.871529411764706, "grad_norm": 0.6142405044428948, "learning_rate": 3.535046753039235e-05, "loss": 0.2598, "step": 10352 }, { "epoch": 4.872, "grad_norm": 0.5983231871787963, "learning_rate": 3.53222057627143e-05, "loss": 0.3262, "step": 10353 }, { "epoch": 4.872470588235294, "grad_norm": 0.5853112737815577, "learning_rate": 3.5293953789082425e-05, "loss": 0.2578, "step": 10354 }, { "epoch": 4.872941176470588, "grad_norm": 0.7055084433104747, "learning_rate": 3.5265711611909476e-05, "loss": 0.3633, "step": 10355 }, { "epoch": 4.8734117647058826, "grad_norm": 0.5784185566322234, "learning_rate": 3.5237479233607525e-05, "loss": 0.1592, "step": 10356 }, { "epoch": 4.873882352941177, "grad_norm": 0.5525174556180127, "learning_rate": 3.520925665658779e-05, "loss": 0.3125, "step": 10357 }, { "epoch": 4.87435294117647, "grad_norm": 0.4884565273268703, "learning_rate": 3.518104388326055e-05, "loss": 0.2559, "step": 10358 }, { "epoch": 4.874823529411764, "grad_norm": 0.5844247550182199, "learning_rate": 3.5152840916035344e-05, "loss": 0.1943, "step": 10359 }, { "epoch": 4.875294117647059, "grad_norm": 0.7808477043993722, "learning_rate": 3.5124647757320883e-05, "loss": 0.3223, "step": 10360 }, { "epoch": 4.875764705882353, "grad_norm": 0.6163578096253097, "learning_rate": 3.509646440952492e-05, "loss": 0.3789, "step": 10361 }, { "epoch": 4.876235294117647, "grad_norm": 0.8236807016056199, "learning_rate": 3.506829087505451e-05, "loss": 0.5, "step": 10362 }, { "epoch": 4.876705882352941, "grad_norm": 0.39505278491543366, "learning_rate": 3.504012715631583e-05, "loss": 0.2178, "step": 10363 }, { "epoch": 4.8771764705882354, "grad_norm": 0.8917076222839474, "learning_rate": 3.501197325571411e-05, "loss": 0.2754, "step": 10364 }, { "epoch": 4.87764705882353, "grad_norm": 0.279340559071795, "learning_rate": 3.49838291756539e-05, "loss": 0.0679, "step": 10365 }, { "epoch": 4.878117647058824, "grad_norm": 0.5665490333037911, "learning_rate": 3.495569491853887e-05, "loss": 0.2695, "step": 10366 }, { "epoch": 4.878588235294117, "grad_norm": 0.23314571880732635, "learning_rate": 3.492757048677175e-05, "loss": 0.0688, "step": 10367 }, { "epoch": 4.8790588235294114, "grad_norm": 0.5775202472496388, "learning_rate": 3.4899455882754526e-05, "loss": 0.3418, "step": 10368 }, { "epoch": 4.879529411764706, "grad_norm": 0.8191883979681748, "learning_rate": 3.4871351108888375e-05, "loss": 0.3398, "step": 10369 }, { "epoch": 4.88, "grad_norm": 0.3498715164584374, "learning_rate": 3.4843256167573514e-05, "loss": 0.1035, "step": 10370 }, { "epoch": 4.880470588235294, "grad_norm": 0.4453404744571417, "learning_rate": 3.481517106120944e-05, "loss": 0.1855, "step": 10371 }, { "epoch": 4.880941176470588, "grad_norm": 0.43632270392271105, "learning_rate": 3.478709579219472e-05, "loss": 0.1475, "step": 10372 }, { "epoch": 4.881411764705883, "grad_norm": 0.46004762625175427, "learning_rate": 3.4759030362927115e-05, "loss": 0.1748, "step": 10373 }, { "epoch": 4.881882352941177, "grad_norm": 0.5120108254776587, "learning_rate": 3.4730974775803574e-05, "loss": 0.293, "step": 10374 }, { "epoch": 4.882352941176471, "grad_norm": 0.10613598058381095, "learning_rate": 3.4702929033220174e-05, "loss": 0.0291, "step": 10375 }, { "epoch": 4.882823529411764, "grad_norm": 0.2583472311260073, "learning_rate": 3.467489313757217e-05, "loss": 0.0977, "step": 10376 }, { "epoch": 4.883294117647059, "grad_norm": 0.5401665563454469, "learning_rate": 3.4646867091254e-05, "loss": 0.3008, "step": 10377 }, { "epoch": 4.883764705882353, "grad_norm": 0.6061660719756358, "learning_rate": 3.4618850896659135e-05, "loss": 0.3066, "step": 10378 }, { "epoch": 4.884235294117647, "grad_norm": 0.5746927708918427, "learning_rate": 3.4590844556180395e-05, "loss": 0.3262, "step": 10379 }, { "epoch": 4.884705882352941, "grad_norm": 0.24359299552546565, "learning_rate": 3.456284807220958e-05, "loss": 0.1118, "step": 10380 }, { "epoch": 4.8851764705882355, "grad_norm": 0.38293654127571825, "learning_rate": 3.453486144713772e-05, "loss": 0.1475, "step": 10381 }, { "epoch": 4.88564705882353, "grad_norm": 0.3301678056244885, "learning_rate": 3.450688468335512e-05, "loss": 0.1348, "step": 10382 }, { "epoch": 4.886117647058824, "grad_norm": 0.546969869821031, "learning_rate": 3.4478917783251e-05, "loss": 0.3027, "step": 10383 }, { "epoch": 4.886588235294118, "grad_norm": 0.44485211758627086, "learning_rate": 3.445096074921393e-05, "loss": 0.166, "step": 10384 }, { "epoch": 4.8870588235294115, "grad_norm": 0.5963438830264981, "learning_rate": 3.442301358363163e-05, "loss": 0.3555, "step": 10385 }, { "epoch": 4.887529411764706, "grad_norm": 0.6268051756354784, "learning_rate": 3.4395076288890825e-05, "loss": 0.2354, "step": 10386 }, { "epoch": 4.888, "grad_norm": 0.6549198775710231, "learning_rate": 3.436714886737753e-05, "loss": 0.3633, "step": 10387 }, { "epoch": 4.888470588235294, "grad_norm": 0.5018692571266001, "learning_rate": 3.4339231321476965e-05, "loss": 0.2246, "step": 10388 }, { "epoch": 4.888941176470588, "grad_norm": 0.3993847240691699, "learning_rate": 3.43113236535733e-05, "loss": 0.168, "step": 10389 }, { "epoch": 4.889411764705883, "grad_norm": 0.30122727821395207, "learning_rate": 3.428342586605005e-05, "loss": 0.0781, "step": 10390 }, { "epoch": 4.889882352941177, "grad_norm": 0.6355653359334381, "learning_rate": 3.425553796128985e-05, "loss": 0.3496, "step": 10391 }, { "epoch": 4.89035294117647, "grad_norm": 0.4614835979208299, "learning_rate": 3.42276599416744e-05, "loss": 0.1729, "step": 10392 }, { "epoch": 4.890823529411764, "grad_norm": 0.64868150105223, "learning_rate": 3.419979180958464e-05, "loss": 0.3164, "step": 10393 }, { "epoch": 4.891294117647059, "grad_norm": 0.4067550416413396, "learning_rate": 3.417193356740071e-05, "loss": 0.2305, "step": 10394 }, { "epoch": 4.891764705882353, "grad_norm": 0.3424449145170838, "learning_rate": 3.414408521750173e-05, "loss": 0.085, "step": 10395 }, { "epoch": 4.892235294117647, "grad_norm": 0.26883168767658533, "learning_rate": 3.411624676226616e-05, "loss": 0.1084, "step": 10396 }, { "epoch": 4.892705882352941, "grad_norm": 0.2157350436947271, "learning_rate": 3.408841820407157e-05, "loss": 0.0623, "step": 10397 }, { "epoch": 4.8931764705882355, "grad_norm": 0.28609201186005323, "learning_rate": 3.406059954529454e-05, "loss": 0.1157, "step": 10398 }, { "epoch": 4.89364705882353, "grad_norm": 1.1962711083407882, "learning_rate": 3.403279078831101e-05, "loss": 0.166, "step": 10399 }, { "epoch": 4.894117647058824, "grad_norm": 0.920474017002178, "learning_rate": 3.4004991935496004e-05, "loss": 0.5117, "step": 10400 }, { "epoch": 4.894588235294117, "grad_norm": 0.35888847908346433, "learning_rate": 3.397720298922358e-05, "loss": 0.1729, "step": 10401 }, { "epoch": 4.8950588235294115, "grad_norm": 1.1837030989837645, "learning_rate": 3.394942395186716e-05, "loss": 0.7031, "step": 10402 }, { "epoch": 4.895529411764706, "grad_norm": 0.35389444625563854, "learning_rate": 3.392165482579912e-05, "loss": 0.1436, "step": 10403 }, { "epoch": 4.896, "grad_norm": 0.5573109786804448, "learning_rate": 3.389389561339112e-05, "loss": 0.1748, "step": 10404 }, { "epoch": 4.896470588235294, "grad_norm": 0.5667285260326019, "learning_rate": 3.3866146317013976e-05, "loss": 0.249, "step": 10405 }, { "epoch": 4.896941176470588, "grad_norm": 0.6235565796018003, "learning_rate": 3.3838406939037525e-05, "loss": 0.3438, "step": 10406 }, { "epoch": 4.897411764705883, "grad_norm": 0.48928115606666134, "learning_rate": 3.381067748183087e-05, "loss": 0.1953, "step": 10407 }, { "epoch": 4.897882352941177, "grad_norm": 0.6198249790886098, "learning_rate": 3.3782957947762336e-05, "loss": 0.3926, "step": 10408 }, { "epoch": 4.898352941176471, "grad_norm": 0.4906025615774955, "learning_rate": 3.375524833919918e-05, "loss": 0.2393, "step": 10409 }, { "epoch": 4.898823529411764, "grad_norm": 0.3862609660814856, "learning_rate": 3.372754865850799e-05, "loss": 0.1504, "step": 10410 }, { "epoch": 4.899294117647059, "grad_norm": 0.3193039180147177, "learning_rate": 3.3699858908054506e-05, "loss": 0.1377, "step": 10411 }, { "epoch": 4.899764705882353, "grad_norm": 0.6403836468260378, "learning_rate": 3.367217909020348e-05, "loss": 0.3652, "step": 10412 }, { "epoch": 4.900235294117647, "grad_norm": 0.36846238018928673, "learning_rate": 3.364450920731894e-05, "loss": 0.2256, "step": 10413 }, { "epoch": 4.900705882352941, "grad_norm": 0.5986217821444241, "learning_rate": 3.361684926176408e-05, "loss": 0.3848, "step": 10414 }, { "epoch": 4.9011764705882355, "grad_norm": 0.5822084276645224, "learning_rate": 3.35891992559011e-05, "loss": 0.3789, "step": 10415 }, { "epoch": 4.90164705882353, "grad_norm": 0.0780092661426812, "learning_rate": 3.35615591920915e-05, "loss": 0.0155, "step": 10416 }, { "epoch": 4.902117647058824, "grad_norm": 0.3352518062797979, "learning_rate": 3.353392907269591e-05, "loss": 0.0977, "step": 10417 }, { "epoch": 4.902588235294118, "grad_norm": 0.5053239094452343, "learning_rate": 3.3506308900073964e-05, "loss": 0.2715, "step": 10418 }, { "epoch": 4.9030588235294115, "grad_norm": 0.5224194645247954, "learning_rate": 3.3478698676584714e-05, "loss": 0.2363, "step": 10419 }, { "epoch": 4.903529411764706, "grad_norm": 0.42997706427963384, "learning_rate": 3.34510984045861e-05, "loss": 0.2285, "step": 10420 }, { "epoch": 4.904, "grad_norm": 0.5301921173262583, "learning_rate": 3.3423508086435326e-05, "loss": 0.373, "step": 10421 }, { "epoch": 4.904470588235294, "grad_norm": 0.3534398446854311, "learning_rate": 3.339592772448883e-05, "loss": 0.0752, "step": 10422 }, { "epoch": 4.904941176470588, "grad_norm": 0.2455558924057039, "learning_rate": 3.3368357321102e-05, "loss": 0.0742, "step": 10423 }, { "epoch": 4.905411764705883, "grad_norm": 0.10258136106799688, "learning_rate": 3.334079687862952e-05, "loss": 0.0123, "step": 10424 }, { "epoch": 4.905882352941177, "grad_norm": 0.317549793101039, "learning_rate": 3.331324639942526e-05, "loss": 0.104, "step": 10425 }, { "epoch": 4.90635294117647, "grad_norm": 0.4347603474438576, "learning_rate": 3.328570588584204e-05, "loss": 0.1641, "step": 10426 }, { "epoch": 4.906823529411764, "grad_norm": 0.3096462584635498, "learning_rate": 3.325817534023202e-05, "loss": 0.1221, "step": 10427 }, { "epoch": 4.907294117647059, "grad_norm": 0.6970564076218065, "learning_rate": 3.32306547649465e-05, "loss": 0.3574, "step": 10428 }, { "epoch": 4.907764705882353, "grad_norm": 0.4559442239519607, "learning_rate": 3.320314416233574e-05, "loss": 0.1562, "step": 10429 }, { "epoch": 4.908235294117647, "grad_norm": 0.48167589538289274, "learning_rate": 3.317564353474943e-05, "loss": 0.1631, "step": 10430 }, { "epoch": 4.908705882352941, "grad_norm": 0.40560004896805935, "learning_rate": 3.314815288453613e-05, "loss": 0.2676, "step": 10431 }, { "epoch": 4.9091764705882355, "grad_norm": 0.777817896942339, "learning_rate": 3.312067221404371e-05, "loss": 0.4727, "step": 10432 }, { "epoch": 4.90964705882353, "grad_norm": 0.5186099128378637, "learning_rate": 3.309320152561923e-05, "loss": 0.2812, "step": 10433 }, { "epoch": 4.910117647058824, "grad_norm": 0.5658805726040343, "learning_rate": 3.3065740821608725e-05, "loss": 0.2852, "step": 10434 }, { "epoch": 4.910588235294117, "grad_norm": 0.6702932151769889, "learning_rate": 3.303829010435751e-05, "loss": 0.3418, "step": 10435 }, { "epoch": 4.9110588235294115, "grad_norm": 0.47118491949005453, "learning_rate": 3.3010849376210056e-05, "loss": 0.3066, "step": 10436 }, { "epoch": 4.911529411764706, "grad_norm": 0.5154836375033973, "learning_rate": 3.298341863950986e-05, "loss": 0.1982, "step": 10437 }, { "epoch": 4.912, "grad_norm": 0.6098982465982997, "learning_rate": 3.295599789659967e-05, "loss": 0.3496, "step": 10438 }, { "epoch": 4.912470588235294, "grad_norm": 0.8631451357664714, "learning_rate": 3.2928587149821385e-05, "loss": 0.4199, "step": 10439 }, { "epoch": 4.912941176470588, "grad_norm": 0.4815619111037709, "learning_rate": 3.290118640151597e-05, "loss": 0.2773, "step": 10440 }, { "epoch": 4.913411764705883, "grad_norm": 0.3530477748675569, "learning_rate": 3.287379565402358e-05, "loss": 0.0996, "step": 10441 }, { "epoch": 4.913882352941177, "grad_norm": 0.28089247753106994, "learning_rate": 3.284641490968359e-05, "loss": 0.0547, "step": 10442 }, { "epoch": 4.914352941176471, "grad_norm": 0.4685218533494734, "learning_rate": 3.281904417083436e-05, "loss": 0.2031, "step": 10443 }, { "epoch": 4.914823529411764, "grad_norm": 0.572528361009662, "learning_rate": 3.279168343981355e-05, "loss": 0.3594, "step": 10444 }, { "epoch": 4.915294117647059, "grad_norm": 0.36276154684641565, "learning_rate": 3.2764332718957894e-05, "loss": 0.2158, "step": 10445 }, { "epoch": 4.915764705882353, "grad_norm": 0.29101808619696123, "learning_rate": 3.2736992010603226e-05, "loss": 0.1836, "step": 10446 }, { "epoch": 4.916235294117647, "grad_norm": 0.704470399285778, "learning_rate": 3.270966131708461e-05, "loss": 0.4023, "step": 10447 }, { "epoch": 4.916705882352941, "grad_norm": 0.5528045375161575, "learning_rate": 3.268234064073628e-05, "loss": 0.3457, "step": 10448 }, { "epoch": 4.9171764705882355, "grad_norm": 0.2245375615232661, "learning_rate": 3.265502998389143e-05, "loss": 0.1104, "step": 10449 }, { "epoch": 4.91764705882353, "grad_norm": 0.9307786616898913, "learning_rate": 3.262772934888265e-05, "loss": 0.5703, "step": 10450 }, { "epoch": 4.918117647058823, "grad_norm": 0.3384154612268414, "learning_rate": 3.260043873804147e-05, "loss": 0.1211, "step": 10451 }, { "epoch": 4.918588235294117, "grad_norm": 0.7669549508256914, "learning_rate": 3.257315815369863e-05, "loss": 0.3867, "step": 10452 }, { "epoch": 4.9190588235294115, "grad_norm": 0.28691062641977066, "learning_rate": 3.2545887598184114e-05, "loss": 0.1074, "step": 10453 }, { "epoch": 4.919529411764706, "grad_norm": 0.5935630397003021, "learning_rate": 3.251862707382686e-05, "loss": 0.252, "step": 10454 }, { "epoch": 4.92, "grad_norm": 0.7133985873032482, "learning_rate": 3.249137658295511e-05, "loss": 0.3848, "step": 10455 }, { "epoch": 4.920470588235294, "grad_norm": 0.36920202961607035, "learning_rate": 3.246413612789622e-05, "loss": 0.1533, "step": 10456 }, { "epoch": 4.920941176470588, "grad_norm": 0.4621267205884877, "learning_rate": 3.243690571097656e-05, "loss": 0.1621, "step": 10457 }, { "epoch": 4.921411764705883, "grad_norm": 0.6892817946712709, "learning_rate": 3.24096853345218e-05, "loss": 0.2539, "step": 10458 }, { "epoch": 4.921882352941177, "grad_norm": 0.3446992869377807, "learning_rate": 3.2382475000856725e-05, "loss": 0.1338, "step": 10459 }, { "epoch": 4.92235294117647, "grad_norm": 0.5457004269512635, "learning_rate": 3.2355274712305154e-05, "loss": 0.291, "step": 10460 }, { "epoch": 4.922823529411764, "grad_norm": 0.6032904479759975, "learning_rate": 3.232808447119018e-05, "loss": 0.3398, "step": 10461 }, { "epoch": 4.923294117647059, "grad_norm": 0.5198852862982741, "learning_rate": 3.2300904279834e-05, "loss": 0.2832, "step": 10462 }, { "epoch": 4.923764705882353, "grad_norm": 0.2701231787974937, "learning_rate": 3.227373414055782e-05, "loss": 0.1064, "step": 10463 }, { "epoch": 4.924235294117647, "grad_norm": 0.20494150188277255, "learning_rate": 3.224657405568229e-05, "loss": 0.0603, "step": 10464 }, { "epoch": 4.924705882352941, "grad_norm": 0.35341871731784646, "learning_rate": 3.221942402752685e-05, "loss": 0.0908, "step": 10465 }, { "epoch": 4.9251764705882355, "grad_norm": 0.5760193958711288, "learning_rate": 3.2192284058410324e-05, "loss": 0.3359, "step": 10466 }, { "epoch": 4.92564705882353, "grad_norm": 0.4417582116690977, "learning_rate": 3.216515415065061e-05, "loss": 0.1592, "step": 10467 }, { "epoch": 4.926117647058824, "grad_norm": 1.0072436019760063, "learning_rate": 3.213803430656468e-05, "loss": 0.7305, "step": 10468 }, { "epoch": 4.926588235294117, "grad_norm": 0.4152930152477202, "learning_rate": 3.2110924528468725e-05, "loss": 0.123, "step": 10469 }, { "epoch": 4.9270588235294115, "grad_norm": 0.38551551543090284, "learning_rate": 3.208382481867809e-05, "loss": 0.1357, "step": 10470 }, { "epoch": 4.927529411764706, "grad_norm": 0.29368771225594054, "learning_rate": 3.205673517950716e-05, "loss": 0.0444, "step": 10471 }, { "epoch": 4.928, "grad_norm": 0.29388057881588225, "learning_rate": 3.202965561326955e-05, "loss": 0.1758, "step": 10472 }, { "epoch": 4.928470588235294, "grad_norm": 0.49292025046038646, "learning_rate": 3.200258612227803e-05, "loss": 0.2246, "step": 10473 }, { "epoch": 4.928941176470588, "grad_norm": 0.7319291182467809, "learning_rate": 3.19755267088444e-05, "loss": 0.3027, "step": 10474 }, { "epoch": 4.929411764705883, "grad_norm": 0.6243779677938596, "learning_rate": 3.194847737527969e-05, "loss": 0.3867, "step": 10475 }, { "epoch": 4.929882352941177, "grad_norm": 0.5424305358759359, "learning_rate": 3.1921438123894086e-05, "loss": 0.3047, "step": 10476 }, { "epoch": 4.930352941176471, "grad_norm": 0.8449466082205752, "learning_rate": 3.189440895699681e-05, "loss": 0.4258, "step": 10477 }, { "epoch": 4.930823529411764, "grad_norm": 0.4360582944972737, "learning_rate": 3.1867389876896296e-05, "loss": 0.2041, "step": 10478 }, { "epoch": 4.931294117647059, "grad_norm": 0.3247496437626847, "learning_rate": 3.184038088590019e-05, "loss": 0.1089, "step": 10479 }, { "epoch": 4.931764705882353, "grad_norm": 0.8660849519605673, "learning_rate": 3.181338198631506e-05, "loss": 0.6484, "step": 10480 }, { "epoch": 4.932235294117647, "grad_norm": 0.36029354439639955, "learning_rate": 3.178639318044687e-05, "loss": 0.1201, "step": 10481 }, { "epoch": 4.932705882352941, "grad_norm": 0.30390015119148217, "learning_rate": 3.175941447060048e-05, "loss": 0.1416, "step": 10482 }, { "epoch": 4.9331764705882355, "grad_norm": 0.5208589970334849, "learning_rate": 3.173244585908007e-05, "loss": 0.4199, "step": 10483 }, { "epoch": 4.93364705882353, "grad_norm": 0.7360900760688079, "learning_rate": 3.170548734818893e-05, "loss": 0.4902, "step": 10484 }, { "epoch": 4.934117647058823, "grad_norm": 0.47379001684236843, "learning_rate": 3.1678538940229356e-05, "loss": 0.2617, "step": 10485 }, { "epoch": 4.934588235294117, "grad_norm": 0.4611666721679544, "learning_rate": 3.16516006375029e-05, "loss": 0.2451, "step": 10486 }, { "epoch": 4.9350588235294115, "grad_norm": 0.5785781215067348, "learning_rate": 3.1624672442310304e-05, "loss": 0.2295, "step": 10487 }, { "epoch": 4.935529411764706, "grad_norm": 0.42045706954533124, "learning_rate": 3.159775435695126e-05, "loss": 0.2773, "step": 10488 }, { "epoch": 4.936, "grad_norm": 0.4925393671474178, "learning_rate": 3.1570846383724775e-05, "loss": 0.2715, "step": 10489 }, { "epoch": 4.936470588235294, "grad_norm": 0.5539581356346261, "learning_rate": 3.154394852492892e-05, "loss": 0.3574, "step": 10490 }, { "epoch": 4.936941176470588, "grad_norm": 0.8906089937987032, "learning_rate": 3.1517060782860865e-05, "loss": 0.5078, "step": 10491 }, { "epoch": 4.937411764705883, "grad_norm": 0.3465393135250126, "learning_rate": 3.1490183159816956e-05, "loss": 0.105, "step": 10492 }, { "epoch": 4.937882352941177, "grad_norm": 0.329320641902476, "learning_rate": 3.146331565809273e-05, "loss": 0.1318, "step": 10493 }, { "epoch": 4.93835294117647, "grad_norm": 0.536686159504106, "learning_rate": 3.143645827998274e-05, "loss": 0.3457, "step": 10494 }, { "epoch": 4.938823529411764, "grad_norm": 0.6607485332161341, "learning_rate": 3.140961102778075e-05, "loss": 0.4883, "step": 10495 }, { "epoch": 4.939294117647059, "grad_norm": 0.5499692382931858, "learning_rate": 3.138277390377971e-05, "loss": 0.3516, "step": 10496 }, { "epoch": 4.939764705882353, "grad_norm": 0.7471011567184491, "learning_rate": 3.135594691027155e-05, "loss": 0.3887, "step": 10497 }, { "epoch": 4.940235294117647, "grad_norm": 0.49886402945213865, "learning_rate": 3.132913004954747e-05, "loss": 0.1533, "step": 10498 }, { "epoch": 4.940705882352941, "grad_norm": 0.7816305890045904, "learning_rate": 3.130232332389781e-05, "loss": 0.5664, "step": 10499 }, { "epoch": 4.9411764705882355, "grad_norm": 0.4990179022870621, "learning_rate": 3.1275526735611896e-05, "loss": 0.2227, "step": 10500 }, { "epoch": 4.94164705882353, "grad_norm": 0.7940179281411508, "learning_rate": 3.1248740286978375e-05, "loss": 0.5312, "step": 10501 }, { "epoch": 4.942117647058824, "grad_norm": 0.34402990780549253, "learning_rate": 3.1221963980284876e-05, "loss": 0.1523, "step": 10502 }, { "epoch": 4.942588235294117, "grad_norm": 0.4882911286591624, "learning_rate": 3.119519781781825e-05, "loss": 0.2256, "step": 10503 }, { "epoch": 4.9430588235294115, "grad_norm": 0.7750342110028301, "learning_rate": 3.1168441801864505e-05, "loss": 0.4316, "step": 10504 }, { "epoch": 4.943529411764706, "grad_norm": 0.3819329197767815, "learning_rate": 3.1141695934708664e-05, "loss": 0.1465, "step": 10505 }, { "epoch": 4.944, "grad_norm": 0.4268147486326239, "learning_rate": 3.1114960218634986e-05, "loss": 0.1953, "step": 10506 }, { "epoch": 4.944470588235294, "grad_norm": 0.2234477116402669, "learning_rate": 3.1088234655926815e-05, "loss": 0.0432, "step": 10507 }, { "epoch": 4.944941176470588, "grad_norm": 0.6965265791952805, "learning_rate": 3.106151924886668e-05, "loss": 0.5391, "step": 10508 }, { "epoch": 4.945411764705883, "grad_norm": 0.40092257489619987, "learning_rate": 3.103481399973622e-05, "loss": 0.0977, "step": 10509 }, { "epoch": 4.945882352941177, "grad_norm": 0.470079493833352, "learning_rate": 3.100811891081614e-05, "loss": 0.2061, "step": 10510 }, { "epoch": 4.946352941176471, "grad_norm": 0.44329325253752266, "learning_rate": 3.098143398438635e-05, "loss": 0.2734, "step": 10511 }, { "epoch": 4.946823529411764, "grad_norm": 0.9264107881329653, "learning_rate": 3.0954759222725916e-05, "loss": 0.4844, "step": 10512 }, { "epoch": 4.947294117647059, "grad_norm": 0.3875277262591783, "learning_rate": 3.092809462811292e-05, "loss": 0.1855, "step": 10513 }, { "epoch": 4.947764705882353, "grad_norm": 0.6303152091173443, "learning_rate": 3.090144020282469e-05, "loss": 0.3281, "step": 10514 }, { "epoch": 4.948235294117647, "grad_norm": 0.5123996746534462, "learning_rate": 3.087479594913767e-05, "loss": 0.2715, "step": 10515 }, { "epoch": 4.948705882352941, "grad_norm": 0.4744296580971649, "learning_rate": 3.0848161869327356e-05, "loss": 0.3516, "step": 10516 }, { "epoch": 4.9491764705882355, "grad_norm": 0.36364320327619376, "learning_rate": 3.0821537965668454e-05, "loss": 0.126, "step": 10517 }, { "epoch": 4.94964705882353, "grad_norm": 0.5948772087212181, "learning_rate": 3.079492424043481e-05, "loss": 0.3809, "step": 10518 }, { "epoch": 4.950117647058823, "grad_norm": 0.759350974340654, "learning_rate": 3.0768320695899294e-05, "loss": 0.5391, "step": 10519 }, { "epoch": 4.950588235294117, "grad_norm": 0.3501147770783712, "learning_rate": 3.0741727334334026e-05, "loss": 0.0967, "step": 10520 }, { "epoch": 4.9510588235294115, "grad_norm": 0.5636111727152456, "learning_rate": 3.071514415801024e-05, "loss": 0.2852, "step": 10521 }, { "epoch": 4.951529411764706, "grad_norm": 0.617527203174801, "learning_rate": 3.068857116919819e-05, "loss": 0.2314, "step": 10522 }, { "epoch": 4.952, "grad_norm": 0.33548327730242705, "learning_rate": 3.066200837016737e-05, "loss": 0.1289, "step": 10523 }, { "epoch": 4.952470588235294, "grad_norm": 0.3366826846850967, "learning_rate": 3.063545576318644e-05, "loss": 0.1289, "step": 10524 }, { "epoch": 4.952941176470588, "grad_norm": 0.5590948513265426, "learning_rate": 3.060891335052302e-05, "loss": 0.2773, "step": 10525 }, { "epoch": 4.953411764705883, "grad_norm": 0.6648846816958447, "learning_rate": 3.058238113444401e-05, "loss": 0.3711, "step": 10526 }, { "epoch": 4.953882352941177, "grad_norm": 0.3310587732635644, "learning_rate": 3.0555859117215446e-05, "loss": 0.0791, "step": 10527 }, { "epoch": 4.95435294117647, "grad_norm": 0.5916996001288984, "learning_rate": 3.052934730110232e-05, "loss": 0.1855, "step": 10528 }, { "epoch": 4.954823529411764, "grad_norm": 0.3459005818597526, "learning_rate": 3.050284568836895e-05, "loss": 0.1572, "step": 10529 }, { "epoch": 4.955294117647059, "grad_norm": 0.5291823418766812, "learning_rate": 3.0476354281278726e-05, "loss": 0.2002, "step": 10530 }, { "epoch": 4.955764705882353, "grad_norm": 0.7657459178190806, "learning_rate": 3.044987308209407e-05, "loss": 0.4551, "step": 10531 }, { "epoch": 4.956235294117647, "grad_norm": 0.4455373478385642, "learning_rate": 3.0423402093076683e-05, "loss": 0.2451, "step": 10532 }, { "epoch": 4.956705882352941, "grad_norm": 0.513802539425996, "learning_rate": 3.039694131648723e-05, "loss": 0.1396, "step": 10533 }, { "epoch": 4.9571764705882355, "grad_norm": 0.3190107510304483, "learning_rate": 3.0370490754585636e-05, "loss": 0.123, "step": 10534 }, { "epoch": 4.95764705882353, "grad_norm": 0.6310694292519641, "learning_rate": 3.034405040963095e-05, "loss": 0.1562, "step": 10535 }, { "epoch": 4.958117647058824, "grad_norm": 0.47595792568167505, "learning_rate": 3.0317620283881243e-05, "loss": 0.2168, "step": 10536 }, { "epoch": 4.958588235294117, "grad_norm": 0.5144739784283323, "learning_rate": 3.0291200379593782e-05, "loss": 0.2227, "step": 10537 }, { "epoch": 4.9590588235294115, "grad_norm": 0.770399811824734, "learning_rate": 3.0264790699025027e-05, "loss": 0.5977, "step": 10538 }, { "epoch": 4.959529411764706, "grad_norm": 0.42163458473250365, "learning_rate": 3.0238391244430405e-05, "loss": 0.1465, "step": 10539 }, { "epoch": 4.96, "grad_norm": 0.5822758432271864, "learning_rate": 3.0212002018064614e-05, "loss": 0.3086, "step": 10540 }, { "epoch": 4.960470588235294, "grad_norm": 0.8001437641622026, "learning_rate": 3.0185623022181415e-05, "loss": 0.5547, "step": 10541 }, { "epoch": 4.960941176470588, "grad_norm": 0.3555714924279799, "learning_rate": 3.015925425903368e-05, "loss": 0.2158, "step": 10542 }, { "epoch": 4.961411764705883, "grad_norm": 0.7363299871758113, "learning_rate": 3.0132895730873448e-05, "loss": 0.3535, "step": 10543 }, { "epoch": 4.961882352941177, "grad_norm": 0.5427579495234546, "learning_rate": 3.010654743995189e-05, "loss": 0.2949, "step": 10544 }, { "epoch": 4.962352941176471, "grad_norm": 0.8899302465065783, "learning_rate": 3.0080209388519222e-05, "loss": 0.5234, "step": 10545 }, { "epoch": 4.962823529411764, "grad_norm": 0.46521673362086685, "learning_rate": 3.005388157882488e-05, "loss": 0.2256, "step": 10546 }, { "epoch": 4.963294117647059, "grad_norm": 0.7585592909444199, "learning_rate": 3.002756401311741e-05, "loss": 0.2969, "step": 10547 }, { "epoch": 4.963764705882353, "grad_norm": 0.6411901780532786, "learning_rate": 3.0001256693644405e-05, "loss": 0.3027, "step": 10548 }, { "epoch": 4.964235294117647, "grad_norm": 0.5887828651091169, "learning_rate": 2.997495962265266e-05, "loss": 0.4023, "step": 10549 }, { "epoch": 4.964705882352941, "grad_norm": 0.36006391449621816, "learning_rate": 2.9948672802388135e-05, "loss": 0.1177, "step": 10550 }, { "epoch": 4.9651764705882355, "grad_norm": 0.5494479906056148, "learning_rate": 2.9922396235095762e-05, "loss": 0.3965, "step": 10551 }, { "epoch": 4.96564705882353, "grad_norm": 0.2568211392712389, "learning_rate": 2.9896129923019717e-05, "loss": 0.1045, "step": 10552 }, { "epoch": 4.966117647058823, "grad_norm": 0.5200213604115137, "learning_rate": 2.9869873868403287e-05, "loss": 0.1904, "step": 10553 }, { "epoch": 4.966588235294117, "grad_norm": 0.45975276320897723, "learning_rate": 2.9843628073488863e-05, "loss": 0.2314, "step": 10554 }, { "epoch": 4.9670588235294115, "grad_norm": 0.2736586909249571, "learning_rate": 2.9817392540518014e-05, "loss": 0.0903, "step": 10555 }, { "epoch": 4.967529411764706, "grad_norm": 0.5699424845141797, "learning_rate": 2.9791167271731287e-05, "loss": 0.3281, "step": 10556 }, { "epoch": 4.968, "grad_norm": 0.22730717664743305, "learning_rate": 2.9764952269368487e-05, "loss": 0.04, "step": 10557 }, { "epoch": 4.968470588235294, "grad_norm": 0.9041072576033621, "learning_rate": 2.973874753566857e-05, "loss": 0.4609, "step": 10558 }, { "epoch": 4.968941176470588, "grad_norm": 0.2672242804284862, "learning_rate": 2.9712553072869437e-05, "loss": 0.1328, "step": 10559 }, { "epoch": 4.969411764705883, "grad_norm": 0.4681528493163673, "learning_rate": 2.968636888320833e-05, "loss": 0.3652, "step": 10560 }, { "epoch": 4.969882352941177, "grad_norm": 0.09462624456148061, "learning_rate": 2.9660194968921403e-05, "loss": 0.0187, "step": 10561 }, { "epoch": 4.97035294117647, "grad_norm": 0.49481653259708275, "learning_rate": 2.963403133224409e-05, "loss": 0.2754, "step": 10562 }, { "epoch": 4.970823529411764, "grad_norm": 0.5229990527639509, "learning_rate": 2.9607877975410936e-05, "loss": 0.2871, "step": 10563 }, { "epoch": 4.971294117647059, "grad_norm": 0.2340620271572714, "learning_rate": 2.958173490065549e-05, "loss": 0.0403, "step": 10564 }, { "epoch": 4.971764705882353, "grad_norm": 0.4539308173199538, "learning_rate": 2.955560211021053e-05, "loss": 0.0972, "step": 10565 }, { "epoch": 4.972235294117647, "grad_norm": 0.5532601157250279, "learning_rate": 2.952947960630797e-05, "loss": 0.2109, "step": 10566 }, { "epoch": 4.972705882352941, "grad_norm": 0.45573398295013473, "learning_rate": 2.9503367391178713e-05, "loss": 0.2969, "step": 10567 }, { "epoch": 4.9731764705882355, "grad_norm": 0.3553525240182448, "learning_rate": 2.94772654670529e-05, "loss": 0.1367, "step": 10568 }, { "epoch": 4.97364705882353, "grad_norm": 0.5838976650863659, "learning_rate": 2.9451173836159826e-05, "loss": 0.3359, "step": 10569 }, { "epoch": 4.974117647058824, "grad_norm": 0.6693003387997065, "learning_rate": 2.9425092500727764e-05, "loss": 0.3203, "step": 10570 }, { "epoch": 4.974588235294117, "grad_norm": 0.3949883556481234, "learning_rate": 2.9399021462984206e-05, "loss": 0.2383, "step": 10571 }, { "epoch": 4.9750588235294115, "grad_norm": 0.7435411204236606, "learning_rate": 2.9372960725155804e-05, "loss": 0.2363, "step": 10572 }, { "epoch": 4.975529411764706, "grad_norm": 0.2854300134971011, "learning_rate": 2.934691028946819e-05, "loss": 0.1338, "step": 10573 }, { "epoch": 4.976, "grad_norm": 0.7682881017521832, "learning_rate": 2.9320870158146235e-05, "loss": 0.3633, "step": 10574 }, { "epoch": 4.976470588235294, "grad_norm": 0.2783014108337884, "learning_rate": 2.9294840333413932e-05, "loss": 0.1328, "step": 10575 }, { "epoch": 4.976941176470588, "grad_norm": 0.40235197781360466, "learning_rate": 2.926882081749428e-05, "loss": 0.1895, "step": 10576 }, { "epoch": 4.977411764705883, "grad_norm": 0.4587128278320114, "learning_rate": 2.9242811612609522e-05, "loss": 0.2266, "step": 10577 }, { "epoch": 4.977882352941176, "grad_norm": 0.45598468577198714, "learning_rate": 2.9216812720980998e-05, "loss": 0.2246, "step": 10578 }, { "epoch": 4.97835294117647, "grad_norm": 0.2934928484254834, "learning_rate": 2.919082414482908e-05, "loss": 0.0723, "step": 10579 }, { "epoch": 4.978823529411764, "grad_norm": 0.30701518471451106, "learning_rate": 2.9164845886373323e-05, "loss": 0.0693, "step": 10580 }, { "epoch": 4.979294117647059, "grad_norm": 0.3492239977147479, "learning_rate": 2.9138877947832474e-05, "loss": 0.1787, "step": 10581 }, { "epoch": 4.979764705882353, "grad_norm": 0.5336703743141239, "learning_rate": 2.911292033142422e-05, "loss": 0.2617, "step": 10582 }, { "epoch": 4.980235294117647, "grad_norm": 0.6397425316456876, "learning_rate": 2.9086973039365562e-05, "loss": 0.3301, "step": 10583 }, { "epoch": 4.980705882352941, "grad_norm": 0.3426957197421009, "learning_rate": 2.906103607387244e-05, "loss": 0.1221, "step": 10584 }, { "epoch": 4.9811764705882355, "grad_norm": 0.26478745037366397, "learning_rate": 2.9035109437160047e-05, "loss": 0.0781, "step": 10585 }, { "epoch": 4.98164705882353, "grad_norm": 0.3520608719287261, "learning_rate": 2.9009193131442665e-05, "loss": 0.0596, "step": 10586 }, { "epoch": 4.982117647058823, "grad_norm": 0.5819622024794582, "learning_rate": 2.8983287158933622e-05, "loss": 0.1338, "step": 10587 }, { "epoch": 4.982588235294117, "grad_norm": 0.3145683733091583, "learning_rate": 2.895739152184543e-05, "loss": 0.0825, "step": 10588 }, { "epoch": 4.9830588235294115, "grad_norm": 0.5181275092793949, "learning_rate": 2.8931506222389756e-05, "loss": 0.2871, "step": 10589 }, { "epoch": 4.983529411764706, "grad_norm": 0.1454377269161476, "learning_rate": 2.8905631262777245e-05, "loss": 0.0349, "step": 10590 }, { "epoch": 4.984, "grad_norm": 0.34969167938352963, "learning_rate": 2.8879766645217784e-05, "loss": 0.1177, "step": 10591 }, { "epoch": 4.984470588235294, "grad_norm": 0.42016677751821513, "learning_rate": 2.8853912371920384e-05, "loss": 0.1689, "step": 10592 }, { "epoch": 4.984941176470588, "grad_norm": 0.6116495690351279, "learning_rate": 2.882806844509305e-05, "loss": 0.3262, "step": 10593 }, { "epoch": 4.985411764705883, "grad_norm": 0.6139831343152409, "learning_rate": 2.8802234866943012e-05, "loss": 0.3203, "step": 10594 }, { "epoch": 4.985882352941177, "grad_norm": 0.6951900415282138, "learning_rate": 2.8776411639676644e-05, "loss": 0.3105, "step": 10595 }, { "epoch": 4.98635294117647, "grad_norm": 0.5023923523644043, "learning_rate": 2.8750598765499226e-05, "loss": 0.1357, "step": 10596 }, { "epoch": 4.986823529411764, "grad_norm": 0.5170184783460053, "learning_rate": 2.8724796246615485e-05, "loss": 0.1348, "step": 10597 }, { "epoch": 4.987294117647059, "grad_norm": 0.35007587325067924, "learning_rate": 2.8699004085228948e-05, "loss": 0.1367, "step": 10598 }, { "epoch": 4.987764705882353, "grad_norm": 0.3200712125372799, "learning_rate": 2.867322228354243e-05, "loss": 0.0762, "step": 10599 }, { "epoch": 4.988235294117647, "grad_norm": 0.7150488040194355, "learning_rate": 2.8647450843757897e-05, "loss": 0.418, "step": 10600 }, { "epoch": 4.988705882352941, "grad_norm": 0.7608822442829785, "learning_rate": 2.862168976807624e-05, "loss": 0.5273, "step": 10601 }, { "epoch": 4.9891764705882355, "grad_norm": 0.3405052910020563, "learning_rate": 2.859593905869763e-05, "loss": 0.1338, "step": 10602 }, { "epoch": 4.98964705882353, "grad_norm": 0.5855531866647394, "learning_rate": 2.8570198717821362e-05, "loss": 0.2773, "step": 10603 }, { "epoch": 4.990117647058824, "grad_norm": 0.47171629194721815, "learning_rate": 2.8544468747645676e-05, "loss": 0.2012, "step": 10604 }, { "epoch": 4.990588235294117, "grad_norm": 0.3787269016987795, "learning_rate": 2.8518749150368114e-05, "loss": 0.1631, "step": 10605 }, { "epoch": 4.9910588235294115, "grad_norm": 0.23757559552788418, "learning_rate": 2.849303992818527e-05, "loss": 0.0515, "step": 10606 }, { "epoch": 4.991529411764706, "grad_norm": 0.4108976170098304, "learning_rate": 2.846734108329275e-05, "loss": 0.1729, "step": 10607 }, { "epoch": 4.992, "grad_norm": 0.49255479305168526, "learning_rate": 2.8441652617885437e-05, "loss": 0.126, "step": 10608 }, { "epoch": 4.992470588235294, "grad_norm": 0.425727125938719, "learning_rate": 2.8415974534157255e-05, "loss": 0.2158, "step": 10609 }, { "epoch": 4.992941176470588, "grad_norm": 0.6877343020442881, "learning_rate": 2.8390306834301186e-05, "loss": 0.668, "step": 10610 }, { "epoch": 4.993411764705883, "grad_norm": 0.49598284837064965, "learning_rate": 2.836464952050944e-05, "loss": 0.3145, "step": 10611 }, { "epoch": 4.993882352941176, "grad_norm": 0.37978242475607493, "learning_rate": 2.8339002594973204e-05, "loss": 0.1094, "step": 10612 }, { "epoch": 4.99435294117647, "grad_norm": 0.3718241633956861, "learning_rate": 2.8313366059882887e-05, "loss": 0.1133, "step": 10613 }, { "epoch": 4.994823529411764, "grad_norm": 0.48422373059124124, "learning_rate": 2.828773991742802e-05, "loss": 0.1934, "step": 10614 }, { "epoch": 4.995294117647059, "grad_norm": 0.9134837565848188, "learning_rate": 2.8262124169797106e-05, "loss": 0.4258, "step": 10615 }, { "epoch": 4.995764705882353, "grad_norm": 0.4518093155019249, "learning_rate": 2.823651881917793e-05, "loss": 0.2129, "step": 10616 }, { "epoch": 4.996235294117647, "grad_norm": 0.19069575218060505, "learning_rate": 2.8210923867757317e-05, "loss": 0.0491, "step": 10617 }, { "epoch": 4.996705882352941, "grad_norm": 0.33686547085930735, "learning_rate": 2.8185339317721146e-05, "loss": 0.0986, "step": 10618 }, { "epoch": 4.9971764705882356, "grad_norm": 0.5491034809461997, "learning_rate": 2.8159765171254495e-05, "loss": 0.3281, "step": 10619 }, { "epoch": 4.99764705882353, "grad_norm": 0.4184551702807492, "learning_rate": 2.813420143054157e-05, "loss": 0.1748, "step": 10620 }, { "epoch": 4.998117647058823, "grad_norm": 0.4575386758768401, "learning_rate": 2.810864809776554e-05, "loss": 0.1943, "step": 10621 }, { "epoch": 4.998588235294117, "grad_norm": 0.512469320322719, "learning_rate": 2.8083105175108845e-05, "loss": 0.2754, "step": 10622 }, { "epoch": 4.9990588235294116, "grad_norm": 0.4233619891317078, "learning_rate": 2.8057572664752986e-05, "loss": 0.1572, "step": 10623 }, { "epoch": 4.999529411764706, "grad_norm": 0.5940945820892406, "learning_rate": 2.8032050568878524e-05, "loss": 0.2305, "step": 10624 }, { "epoch": 5.0, "grad_norm": 0.8594818428193332, "learning_rate": 2.800653888966519e-05, "loss": 0.291, "step": 10625 }, { "epoch": 5.0, "eval_loss": 0.24538134038448334, "eval_runtime": 62.2005, "eval_samples_per_second": 12.058, "eval_steps_per_second": 6.029, "step": 10625 }, { "epoch": 5.000470588235294, "grad_norm": 0.4960210853377702, "learning_rate": 2.7981037629291852e-05, "loss": 0.2578, "step": 10626 }, { "epoch": 5.0009411764705884, "grad_norm": 0.2841019598557179, "learning_rate": 2.7955546789936363e-05, "loss": 0.1387, "step": 10627 }, { "epoch": 5.001411764705883, "grad_norm": 0.6705276920130286, "learning_rate": 2.7930066373775794e-05, "loss": 0.5117, "step": 10628 }, { "epoch": 5.001882352941177, "grad_norm": 0.32767589118630647, "learning_rate": 2.7904596382986343e-05, "loss": 0.1348, "step": 10629 }, { "epoch": 5.00235294117647, "grad_norm": 0.3911608176497156, "learning_rate": 2.7879136819743203e-05, "loss": 0.1084, "step": 10630 }, { "epoch": 5.0028235294117644, "grad_norm": 0.4851984374950218, "learning_rate": 2.7853687686220766e-05, "loss": 0.2793, "step": 10631 }, { "epoch": 5.003294117647059, "grad_norm": 0.4843092654553376, "learning_rate": 2.782824898459258e-05, "loss": 0.3594, "step": 10632 }, { "epoch": 5.003764705882353, "grad_norm": 0.28398610210244773, "learning_rate": 2.7802820717031132e-05, "loss": 0.0942, "step": 10633 }, { "epoch": 5.004235294117647, "grad_norm": 0.7030681875738259, "learning_rate": 2.7777402885708212e-05, "loss": 0.3164, "step": 10634 }, { "epoch": 5.004705882352941, "grad_norm": 0.5247583347955698, "learning_rate": 2.7751995492794544e-05, "loss": 0.3848, "step": 10635 }, { "epoch": 5.005176470588236, "grad_norm": 0.3777142962936626, "learning_rate": 2.7726598540460076e-05, "loss": 0.1572, "step": 10636 }, { "epoch": 5.00564705882353, "grad_norm": 0.47240792872904075, "learning_rate": 2.770121203087389e-05, "loss": 0.2793, "step": 10637 }, { "epoch": 5.006117647058823, "grad_norm": 0.31438961426424555, "learning_rate": 2.7675835966204024e-05, "loss": 0.1025, "step": 10638 }, { "epoch": 5.006588235294117, "grad_norm": 0.1516928266140992, "learning_rate": 2.7650470348617772e-05, "loss": 0.0356, "step": 10639 }, { "epoch": 5.007058823529412, "grad_norm": 0.6185123784530746, "learning_rate": 2.7625115180281464e-05, "loss": 0.5039, "step": 10640 }, { "epoch": 5.007529411764706, "grad_norm": 0.529262180509733, "learning_rate": 2.7599770463360576e-05, "loss": 0.3262, "step": 10641 }, { "epoch": 5.008, "grad_norm": 0.71133574145346, "learning_rate": 2.7574436200019688e-05, "loss": 0.4199, "step": 10642 }, { "epoch": 5.008470588235294, "grad_norm": 0.24467614147440953, "learning_rate": 2.7549112392422405e-05, "loss": 0.0598, "step": 10643 }, { "epoch": 5.0089411764705885, "grad_norm": 0.49687849055972905, "learning_rate": 2.7523799042731548e-05, "loss": 0.2812, "step": 10644 }, { "epoch": 5.009411764705883, "grad_norm": 0.6208412704735974, "learning_rate": 2.749849615310902e-05, "loss": 0.2871, "step": 10645 }, { "epoch": 5.009882352941177, "grad_norm": 0.5237782306116197, "learning_rate": 2.7473203725715754e-05, "loss": 0.2656, "step": 10646 }, { "epoch": 5.01035294117647, "grad_norm": 0.5231922459759388, "learning_rate": 2.7447921762711894e-05, "loss": 0.2178, "step": 10647 }, { "epoch": 5.0108235294117645, "grad_norm": 0.5921461079671496, "learning_rate": 2.7422650266256663e-05, "loss": 0.1895, "step": 10648 }, { "epoch": 5.011294117647059, "grad_norm": 0.3899831585191797, "learning_rate": 2.7397389238508282e-05, "loss": 0.1729, "step": 10649 }, { "epoch": 5.011764705882353, "grad_norm": 0.1485837588895252, "learning_rate": 2.7372138681624244e-05, "loss": 0.0327, "step": 10650 }, { "epoch": 5.012235294117647, "grad_norm": 0.4666628935506141, "learning_rate": 2.7346898597761073e-05, "loss": 0.2324, "step": 10651 }, { "epoch": 5.012705882352941, "grad_norm": 0.5679893774299052, "learning_rate": 2.7321668989074347e-05, "loss": 0.3906, "step": 10652 }, { "epoch": 5.013176470588236, "grad_norm": 0.6204695456610658, "learning_rate": 2.7296449857718806e-05, "loss": 0.5117, "step": 10653 }, { "epoch": 5.01364705882353, "grad_norm": 0.5694612967028262, "learning_rate": 2.727124120584836e-05, "loss": 0.3242, "step": 10654 }, { "epoch": 5.014117647058823, "grad_norm": 0.39091367315104514, "learning_rate": 2.724604303561584e-05, "loss": 0.1211, "step": 10655 }, { "epoch": 5.014588235294117, "grad_norm": 0.5411453227142492, "learning_rate": 2.7220855349173353e-05, "loss": 0.2812, "step": 10656 }, { "epoch": 5.015058823529412, "grad_norm": 0.5952381527686657, "learning_rate": 2.71956781486721e-05, "loss": 0.3184, "step": 10657 }, { "epoch": 5.015529411764706, "grad_norm": 0.5566068373113183, "learning_rate": 2.7170511436262243e-05, "loss": 0.3945, "step": 10658 }, { "epoch": 5.016, "grad_norm": 0.202355842303565, "learning_rate": 2.7145355214093178e-05, "loss": 0.052, "step": 10659 }, { "epoch": 5.016470588235294, "grad_norm": 0.4139467972044241, "learning_rate": 2.7120209484313422e-05, "loss": 0.1455, "step": 10660 }, { "epoch": 5.0169411764705885, "grad_norm": 0.512304570656353, "learning_rate": 2.7095074249070455e-05, "loss": 0.2559, "step": 10661 }, { "epoch": 5.017411764705883, "grad_norm": 0.19511821818880667, "learning_rate": 2.7069949510511023e-05, "loss": 0.0361, "step": 10662 }, { "epoch": 5.017882352941177, "grad_norm": 0.4110895461420505, "learning_rate": 2.7044835270780836e-05, "loss": 0.2217, "step": 10663 }, { "epoch": 5.01835294117647, "grad_norm": 0.47094930851463856, "learning_rate": 2.701973153202482e-05, "loss": 0.2227, "step": 10664 }, { "epoch": 5.0188235294117645, "grad_norm": 0.5933128804137208, "learning_rate": 2.6994638296386977e-05, "loss": 0.4121, "step": 10665 }, { "epoch": 5.019294117647059, "grad_norm": 0.6780391864681581, "learning_rate": 2.6969555566010316e-05, "loss": 0.3887, "step": 10666 }, { "epoch": 5.019764705882353, "grad_norm": 0.6647644825827476, "learning_rate": 2.694448334303708e-05, "loss": 0.248, "step": 10667 }, { "epoch": 5.020235294117647, "grad_norm": 0.666603653895031, "learning_rate": 2.6919421629608584e-05, "loss": 0.3828, "step": 10668 }, { "epoch": 5.020705882352941, "grad_norm": 0.5530263119562012, "learning_rate": 2.6894370427865157e-05, "loss": 0.3672, "step": 10669 }, { "epoch": 5.021176470588236, "grad_norm": 0.5215421472495341, "learning_rate": 2.6869329739946312e-05, "loss": 0.3398, "step": 10670 }, { "epoch": 5.02164705882353, "grad_norm": 0.41099957696097644, "learning_rate": 2.6844299567990713e-05, "loss": 0.1934, "step": 10671 }, { "epoch": 5.022117647058823, "grad_norm": 0.39857719299899813, "learning_rate": 2.6819279914135956e-05, "loss": 0.2207, "step": 10672 }, { "epoch": 5.022588235294117, "grad_norm": 0.5476042237548767, "learning_rate": 2.6794270780518895e-05, "loss": 0.3242, "step": 10673 }, { "epoch": 5.023058823529412, "grad_norm": 0.09447474423730384, "learning_rate": 2.6769272169275474e-05, "loss": 0.0117, "step": 10674 }, { "epoch": 5.023529411764706, "grad_norm": 0.6766271256345552, "learning_rate": 2.67442840825406e-05, "loss": 0.3379, "step": 10675 }, { "epoch": 5.024, "grad_norm": 0.3913983047323971, "learning_rate": 2.671930652244845e-05, "loss": 0.2539, "step": 10676 }, { "epoch": 5.024470588235294, "grad_norm": 0.2081947024002447, "learning_rate": 2.6694339491132237e-05, "loss": 0.0684, "step": 10677 }, { "epoch": 5.0249411764705885, "grad_norm": 0.5235589159809965, "learning_rate": 2.666938299072422e-05, "loss": 0.3613, "step": 10678 }, { "epoch": 5.025411764705883, "grad_norm": 0.669469340999744, "learning_rate": 2.664443702335581e-05, "loss": 0.4727, "step": 10679 }, { "epoch": 5.025882352941177, "grad_norm": 0.5706153095444599, "learning_rate": 2.6619501591157583e-05, "loss": 0.3301, "step": 10680 }, { "epoch": 5.02635294117647, "grad_norm": 0.7615300150567444, "learning_rate": 2.6594576696259055e-05, "loss": 0.3867, "step": 10681 }, { "epoch": 5.0268235294117645, "grad_norm": 0.4081562601647798, "learning_rate": 2.6569662340789012e-05, "loss": 0.1729, "step": 10682 }, { "epoch": 5.027294117647059, "grad_norm": 0.4591417830135297, "learning_rate": 2.6544758526875193e-05, "loss": 0.2422, "step": 10683 }, { "epoch": 5.027764705882353, "grad_norm": 0.32365929706177665, "learning_rate": 2.651986525664455e-05, "loss": 0.1279, "step": 10684 }, { "epoch": 5.028235294117647, "grad_norm": 0.24982152179415854, "learning_rate": 2.649498253222307e-05, "loss": 0.0513, "step": 10685 }, { "epoch": 5.028705882352941, "grad_norm": 0.6629761756673664, "learning_rate": 2.647011035573588e-05, "loss": 0.3555, "step": 10686 }, { "epoch": 5.029176470588236, "grad_norm": 0.29985999742589287, "learning_rate": 2.6445248729307165e-05, "loss": 0.1104, "step": 10687 }, { "epoch": 5.02964705882353, "grad_norm": 0.30276909529810464, "learning_rate": 2.6420397655060282e-05, "loss": 0.0806, "step": 10688 }, { "epoch": 5.030117647058823, "grad_norm": 0.5217616053158215, "learning_rate": 2.639555713511756e-05, "loss": 0.2969, "step": 10689 }, { "epoch": 5.030588235294117, "grad_norm": 0.49114722377304687, "learning_rate": 2.637072717160056e-05, "loss": 0.2275, "step": 10690 }, { "epoch": 5.031058823529412, "grad_norm": 0.6398222908818298, "learning_rate": 2.6345907766629843e-05, "loss": 0.249, "step": 10691 }, { "epoch": 5.031529411764706, "grad_norm": 0.5187789113571807, "learning_rate": 2.6321098922325102e-05, "loss": 0.2363, "step": 10692 }, { "epoch": 5.032, "grad_norm": 0.5179675560105442, "learning_rate": 2.6296300640805206e-05, "loss": 0.2373, "step": 10693 }, { "epoch": 5.032470588235294, "grad_norm": 0.7696340453860929, "learning_rate": 2.627151292418795e-05, "loss": 0.3281, "step": 10694 }, { "epoch": 5.0329411764705885, "grad_norm": 0.42180195649773955, "learning_rate": 2.6246735774590383e-05, "loss": 0.1133, "step": 10695 }, { "epoch": 5.033411764705883, "grad_norm": 0.45551128320944817, "learning_rate": 2.6221969194128617e-05, "loss": 0.2148, "step": 10696 }, { "epoch": 5.033882352941176, "grad_norm": 0.4076264618903258, "learning_rate": 2.6197213184917775e-05, "loss": 0.1396, "step": 10697 }, { "epoch": 5.03435294117647, "grad_norm": 0.36865632555549405, "learning_rate": 2.6172467749072185e-05, "loss": 0.1357, "step": 10698 }, { "epoch": 5.0348235294117645, "grad_norm": 0.1903524561015312, "learning_rate": 2.6147732888705235e-05, "loss": 0.0461, "step": 10699 }, { "epoch": 5.035294117647059, "grad_norm": 1.0419554358454277, "learning_rate": 2.6123008605929375e-05, "loss": 0.9453, "step": 10700 }, { "epoch": 5.035764705882353, "grad_norm": 0.7192218745633434, "learning_rate": 2.6098294902856177e-05, "loss": 0.4297, "step": 10701 }, { "epoch": 5.036235294117647, "grad_norm": 0.3406440684229602, "learning_rate": 2.607359178159637e-05, "loss": 0.0894, "step": 10702 }, { "epoch": 5.036705882352941, "grad_norm": 0.43053652170834544, "learning_rate": 2.604889924425965e-05, "loss": 0.1914, "step": 10703 }, { "epoch": 5.037176470588236, "grad_norm": 0.2849656918963705, "learning_rate": 2.6024217292954917e-05, "loss": 0.0581, "step": 10704 }, { "epoch": 5.03764705882353, "grad_norm": 0.6877187941529221, "learning_rate": 2.599954592979015e-05, "loss": 0.3887, "step": 10705 }, { "epoch": 5.038117647058823, "grad_norm": 0.31437957326177224, "learning_rate": 2.5974885156872345e-05, "loss": 0.1338, "step": 10706 }, { "epoch": 5.038588235294117, "grad_norm": 0.25353980615394567, "learning_rate": 2.5950234976307682e-05, "loss": 0.1001, "step": 10707 }, { "epoch": 5.039058823529412, "grad_norm": 0.5563588012120548, "learning_rate": 2.5925595390201466e-05, "loss": 0.2051, "step": 10708 }, { "epoch": 5.039529411764706, "grad_norm": 0.22431891321259562, "learning_rate": 2.5900966400657923e-05, "loss": 0.0461, "step": 10709 }, { "epoch": 5.04, "grad_norm": 0.9411278272701188, "learning_rate": 2.5876348009780564e-05, "loss": 0.5625, "step": 10710 }, { "epoch": 5.040470588235294, "grad_norm": 0.551283082329778, "learning_rate": 2.5851740219671945e-05, "loss": 0.293, "step": 10711 }, { "epoch": 5.0409411764705885, "grad_norm": 0.5682559907160784, "learning_rate": 2.5827143032433627e-05, "loss": 0.2197, "step": 10712 }, { "epoch": 5.041411764705883, "grad_norm": 0.6200537315298968, "learning_rate": 2.5802556450166396e-05, "loss": 0.3711, "step": 10713 }, { "epoch": 5.041882352941176, "grad_norm": 0.4475628291866397, "learning_rate": 2.5777980474969983e-05, "loss": 0.2031, "step": 10714 }, { "epoch": 5.04235294117647, "grad_norm": 0.7223736430366438, "learning_rate": 2.5753415108943332e-05, "loss": 0.4648, "step": 10715 }, { "epoch": 5.0428235294117645, "grad_norm": 0.4565046674203293, "learning_rate": 2.572886035418451e-05, "loss": 0.1504, "step": 10716 }, { "epoch": 5.043294117647059, "grad_norm": 0.5096318845716294, "learning_rate": 2.5704316212790527e-05, "loss": 0.2178, "step": 10717 }, { "epoch": 5.043764705882353, "grad_norm": 0.14048019338297538, "learning_rate": 2.5679782686857586e-05, "loss": 0.0287, "step": 10718 }, { "epoch": 5.044235294117647, "grad_norm": 0.5487854293019129, "learning_rate": 2.5655259778481045e-05, "loss": 0.3281, "step": 10719 }, { "epoch": 5.044705882352941, "grad_norm": 0.6001098245553977, "learning_rate": 2.5630747489755182e-05, "loss": 0.3008, "step": 10720 }, { "epoch": 5.045176470588236, "grad_norm": 0.8307617993988313, "learning_rate": 2.560624582277352e-05, "loss": 0.4336, "step": 10721 }, { "epoch": 5.04564705882353, "grad_norm": 0.26965907919604953, "learning_rate": 2.558175477962865e-05, "loss": 0.0586, "step": 10722 }, { "epoch": 5.046117647058823, "grad_norm": 0.5846547142910504, "learning_rate": 2.5557274362412132e-05, "loss": 0.3418, "step": 10723 }, { "epoch": 5.046588235294117, "grad_norm": 0.40230794885339816, "learning_rate": 2.55328045732148e-05, "loss": 0.1553, "step": 10724 }, { "epoch": 5.047058823529412, "grad_norm": 0.4653278819866272, "learning_rate": 2.5508345414126474e-05, "loss": 0.2051, "step": 10725 }, { "epoch": 5.047529411764706, "grad_norm": 0.39626740935521587, "learning_rate": 2.5483896887236056e-05, "loss": 0.1533, "step": 10726 }, { "epoch": 5.048, "grad_norm": 0.7911426669787612, "learning_rate": 2.5459458994631604e-05, "loss": 0.6836, "step": 10727 }, { "epoch": 5.048470588235294, "grad_norm": 0.15641572307057833, "learning_rate": 2.5435031738400234e-05, "loss": 0.0454, "step": 10728 }, { "epoch": 5.0489411764705885, "grad_norm": 0.6098898265581282, "learning_rate": 2.541061512062808e-05, "loss": 0.3301, "step": 10729 }, { "epoch": 5.049411764705883, "grad_norm": 0.20140169548692038, "learning_rate": 2.538620914340057e-05, "loss": 0.0557, "step": 10730 }, { "epoch": 5.049882352941176, "grad_norm": 0.7682854643335831, "learning_rate": 2.5361813808802007e-05, "loss": 0.4453, "step": 10731 }, { "epoch": 5.05035294117647, "grad_norm": 0.2088161170090097, "learning_rate": 2.533742911891587e-05, "loss": 0.0493, "step": 10732 }, { "epoch": 5.0508235294117645, "grad_norm": 0.5029701833356176, "learning_rate": 2.531305507582481e-05, "loss": 0.1396, "step": 10733 }, { "epoch": 5.051294117647059, "grad_norm": 0.6373724210714676, "learning_rate": 2.5288691681610383e-05, "loss": 0.2754, "step": 10734 }, { "epoch": 5.051764705882353, "grad_norm": 0.23677607415270552, "learning_rate": 2.526433893835339e-05, "loss": 0.0359, "step": 10735 }, { "epoch": 5.052235294117647, "grad_norm": 0.37424016015828754, "learning_rate": 2.5239996848133715e-05, "loss": 0.1182, "step": 10736 }, { "epoch": 5.052705882352941, "grad_norm": 0.4935402039951176, "learning_rate": 2.5215665413030213e-05, "loss": 0.1406, "step": 10737 }, { "epoch": 5.053176470588236, "grad_norm": 0.39073323060823767, "learning_rate": 2.519134463512095e-05, "loss": 0.1377, "step": 10738 }, { "epoch": 5.05364705882353, "grad_norm": 0.5632036613661531, "learning_rate": 2.516703451648309e-05, "loss": 0.3008, "step": 10739 }, { "epoch": 5.054117647058823, "grad_norm": 0.8840065999095205, "learning_rate": 2.5142735059192737e-05, "loss": 0.5547, "step": 10740 }, { "epoch": 5.054588235294117, "grad_norm": 0.4251921987812962, "learning_rate": 2.511844626532526e-05, "loss": 0.2559, "step": 10741 }, { "epoch": 5.055058823529412, "grad_norm": 0.5770532454605758, "learning_rate": 2.5094168136954985e-05, "loss": 0.2031, "step": 10742 }, { "epoch": 5.055529411764706, "grad_norm": 0.5269028603886347, "learning_rate": 2.506990067615542e-05, "loss": 0.208, "step": 10743 }, { "epoch": 5.056, "grad_norm": 0.625076801623206, "learning_rate": 2.504564388499914e-05, "loss": 0.3262, "step": 10744 }, { "epoch": 5.056470588235294, "grad_norm": 0.42148976755767825, "learning_rate": 2.502139776555772e-05, "loss": 0.125, "step": 10745 }, { "epoch": 5.0569411764705885, "grad_norm": 0.5878355473201107, "learning_rate": 2.4997162319901977e-05, "loss": 0.3398, "step": 10746 }, { "epoch": 5.057411764705883, "grad_norm": 0.4521722014138965, "learning_rate": 2.497293755010173e-05, "loss": 0.2354, "step": 10747 }, { "epoch": 5.057882352941176, "grad_norm": 0.2873489680039184, "learning_rate": 2.4948723458225827e-05, "loss": 0.0645, "step": 10748 }, { "epoch": 5.05835294117647, "grad_norm": 0.3453947302672894, "learning_rate": 2.4924520046342317e-05, "loss": 0.0747, "step": 10749 }, { "epoch": 5.0588235294117645, "grad_norm": 0.4524208691468014, "learning_rate": 2.4900327316518326e-05, "loss": 0.1914, "step": 10750 }, { "epoch": 5.059294117647059, "grad_norm": 0.32252429808331967, "learning_rate": 2.4876145270819954e-05, "loss": 0.0928, "step": 10751 }, { "epoch": 5.059764705882353, "grad_norm": 0.31362702961205685, "learning_rate": 2.485197391131251e-05, "loss": 0.0938, "step": 10752 }, { "epoch": 5.060235294117647, "grad_norm": 0.37883148100737885, "learning_rate": 2.4827813240060367e-05, "loss": 0.0889, "step": 10753 }, { "epoch": 5.060705882352941, "grad_norm": 0.5954372551998012, "learning_rate": 2.480366325912692e-05, "loss": 0.2695, "step": 10754 }, { "epoch": 5.061176470588236, "grad_norm": 0.4511136116074445, "learning_rate": 2.4779523970574703e-05, "loss": 0.2441, "step": 10755 }, { "epoch": 5.06164705882353, "grad_norm": 0.5359003537913898, "learning_rate": 2.4755395376465386e-05, "loss": 0.2461, "step": 10756 }, { "epoch": 5.062117647058823, "grad_norm": 0.4824243839716635, "learning_rate": 2.4731277478859586e-05, "loss": 0.2949, "step": 10757 }, { "epoch": 5.062588235294117, "grad_norm": 0.4000594116282829, "learning_rate": 2.4707170279817118e-05, "loss": 0.1924, "step": 10758 }, { "epoch": 5.063058823529412, "grad_norm": 0.5140718556071855, "learning_rate": 2.4683073781396924e-05, "loss": 0.2236, "step": 10759 }, { "epoch": 5.063529411764706, "grad_norm": 0.3868360244083183, "learning_rate": 2.465898798565685e-05, "loss": 0.1797, "step": 10760 }, { "epoch": 5.064, "grad_norm": 0.4569977872167045, "learning_rate": 2.4634912894653996e-05, "loss": 0.208, "step": 10761 }, { "epoch": 5.064470588235294, "grad_norm": 0.1849370038023739, "learning_rate": 2.461084851044454e-05, "loss": 0.0581, "step": 10762 }, { "epoch": 5.0649411764705885, "grad_norm": 0.5721717614371572, "learning_rate": 2.4586794835083613e-05, "loss": 0.3477, "step": 10763 }, { "epoch": 5.065411764705883, "grad_norm": 0.26535991940240905, "learning_rate": 2.4562751870625597e-05, "loss": 0.0615, "step": 10764 }, { "epoch": 5.065882352941176, "grad_norm": 0.4622562996494797, "learning_rate": 2.4538719619123786e-05, "loss": 0.2754, "step": 10765 }, { "epoch": 5.06635294117647, "grad_norm": 0.9382509544899814, "learning_rate": 2.4514698082630706e-05, "loss": 0.5312, "step": 10766 }, { "epoch": 5.0668235294117645, "grad_norm": 0.3553507731913693, "learning_rate": 2.449068726319795e-05, "loss": 0.124, "step": 10767 }, { "epoch": 5.067294117647059, "grad_norm": 0.5019780352414317, "learning_rate": 2.4466687162876093e-05, "loss": 0.249, "step": 10768 }, { "epoch": 5.067764705882353, "grad_norm": 0.3940456439633505, "learning_rate": 2.4442697783714894e-05, "loss": 0.1748, "step": 10769 }, { "epoch": 5.068235294117647, "grad_norm": 0.5796722994269728, "learning_rate": 2.4418719127763183e-05, "loss": 0.1211, "step": 10770 }, { "epoch": 5.068705882352941, "grad_norm": 0.4083091052347279, "learning_rate": 2.43947511970688e-05, "loss": 0.1816, "step": 10771 }, { "epoch": 5.069176470588236, "grad_norm": 0.24188143744889198, "learning_rate": 2.437079399367875e-05, "loss": 0.1089, "step": 10772 }, { "epoch": 5.06964705882353, "grad_norm": 0.5776136216475338, "learning_rate": 2.434684751963914e-05, "loss": 0.332, "step": 10773 }, { "epoch": 5.070117647058823, "grad_norm": 0.7826054507300292, "learning_rate": 2.4322911776995002e-05, "loss": 0.4297, "step": 10774 }, { "epoch": 5.070588235294117, "grad_norm": 0.40385385977648974, "learning_rate": 2.429898676779072e-05, "loss": 0.0938, "step": 10775 }, { "epoch": 5.071058823529412, "grad_norm": 0.4409317152658309, "learning_rate": 2.4275072494069487e-05, "loss": 0.1904, "step": 10776 }, { "epoch": 5.071529411764706, "grad_norm": 0.7577547033664289, "learning_rate": 2.425116895787374e-05, "loss": 0.5234, "step": 10777 }, { "epoch": 5.072, "grad_norm": 0.4078138804215132, "learning_rate": 2.4227276161244995e-05, "loss": 0.2754, "step": 10778 }, { "epoch": 5.072470588235294, "grad_norm": 0.5551669993762264, "learning_rate": 2.4203394106223755e-05, "loss": 0.1973, "step": 10779 }, { "epoch": 5.0729411764705885, "grad_norm": 0.5326397987119587, "learning_rate": 2.417952279484967e-05, "loss": 0.2871, "step": 10780 }, { "epoch": 5.073411764705883, "grad_norm": 0.5110967549962859, "learning_rate": 2.4155662229161527e-05, "loss": 0.1953, "step": 10781 }, { "epoch": 5.073882352941176, "grad_norm": 0.5404353379854597, "learning_rate": 2.4131812411197045e-05, "loss": 0.3457, "step": 10782 }, { "epoch": 5.07435294117647, "grad_norm": 0.5724384202575291, "learning_rate": 2.4107973342993174e-05, "loss": 0.3633, "step": 10783 }, { "epoch": 5.0748235294117645, "grad_norm": 0.52975397468527, "learning_rate": 2.408414502658591e-05, "loss": 0.2676, "step": 10784 }, { "epoch": 5.075294117647059, "grad_norm": 0.7437851784334532, "learning_rate": 2.4060327464010254e-05, "loss": 0.3516, "step": 10785 }, { "epoch": 5.075764705882353, "grad_norm": 0.22367352217190548, "learning_rate": 2.4036520657300346e-05, "loss": 0.0747, "step": 10786 }, { "epoch": 5.076235294117647, "grad_norm": 0.20433228428792974, "learning_rate": 2.401272460848946e-05, "loss": 0.063, "step": 10787 }, { "epoch": 5.076705882352941, "grad_norm": 0.33912585611405904, "learning_rate": 2.3988939319609817e-05, "loss": 0.0674, "step": 10788 }, { "epoch": 5.077176470588236, "grad_norm": 0.5693085613449756, "learning_rate": 2.3965164792692833e-05, "loss": 0.3281, "step": 10789 }, { "epoch": 5.07764705882353, "grad_norm": 0.27402705738073346, "learning_rate": 2.3941401029769024e-05, "loss": 0.0625, "step": 10790 }, { "epoch": 5.078117647058823, "grad_norm": 0.5934844612971127, "learning_rate": 2.3917648032867844e-05, "loss": 0.2793, "step": 10791 }, { "epoch": 5.078588235294117, "grad_norm": 0.2041144566376979, "learning_rate": 2.3893905804017977e-05, "loss": 0.0623, "step": 10792 }, { "epoch": 5.079058823529412, "grad_norm": 0.5521883760554688, "learning_rate": 2.387017434524707e-05, "loss": 0.2012, "step": 10793 }, { "epoch": 5.079529411764706, "grad_norm": 0.3707565828948299, "learning_rate": 2.3846453658581927e-05, "loss": 0.125, "step": 10794 }, { "epoch": 5.08, "grad_norm": 0.3466436058121791, "learning_rate": 2.382274374604847e-05, "loss": 0.1963, "step": 10795 }, { "epoch": 5.080470588235294, "grad_norm": 0.5960171664931627, "learning_rate": 2.3799044609671558e-05, "loss": 0.3242, "step": 10796 }, { "epoch": 5.0809411764705885, "grad_norm": 0.29360434690312043, "learning_rate": 2.3775356251475248e-05, "loss": 0.084, "step": 10797 }, { "epoch": 5.081411764705883, "grad_norm": 0.34803020894686376, "learning_rate": 2.375167867348268e-05, "loss": 0.1494, "step": 10798 }, { "epoch": 5.081882352941176, "grad_norm": 0.5479210128453773, "learning_rate": 2.3728011877715947e-05, "loss": 0.3633, "step": 10799 }, { "epoch": 5.08235294117647, "grad_norm": 0.6387633099198431, "learning_rate": 2.3704355866196373e-05, "loss": 0.4941, "step": 10800 }, { "epoch": 5.0828235294117645, "grad_norm": 0.4317325227267386, "learning_rate": 2.3680710640944328e-05, "loss": 0.1572, "step": 10801 }, { "epoch": 5.083294117647059, "grad_norm": 0.5985880309679, "learning_rate": 2.3657076203979132e-05, "loss": 0.3848, "step": 10802 }, { "epoch": 5.083764705882353, "grad_norm": 0.5328862757757794, "learning_rate": 2.363345255731935e-05, "loss": 0.3672, "step": 10803 }, { "epoch": 5.084235294117647, "grad_norm": 0.43026401663885017, "learning_rate": 2.360983970298257e-05, "loss": 0.208, "step": 10804 }, { "epoch": 5.084705882352941, "grad_norm": 0.4258842603821898, "learning_rate": 2.3586237642985396e-05, "loss": 0.1465, "step": 10805 }, { "epoch": 5.085176470588236, "grad_norm": 0.48048151052522586, "learning_rate": 2.356264637934358e-05, "loss": 0.2363, "step": 10806 }, { "epoch": 5.085647058823529, "grad_norm": 0.4035639541759444, "learning_rate": 2.3539065914071963e-05, "loss": 0.1982, "step": 10807 }, { "epoch": 5.086117647058823, "grad_norm": 0.6616241202847312, "learning_rate": 2.351549624918437e-05, "loss": 0.5664, "step": 10808 }, { "epoch": 5.086588235294117, "grad_norm": 0.4074780824029582, "learning_rate": 2.34919373866938e-05, "loss": 0.1885, "step": 10809 }, { "epoch": 5.087058823529412, "grad_norm": 0.7334158520372096, "learning_rate": 2.3468389328612348e-05, "loss": 0.3438, "step": 10810 }, { "epoch": 5.087529411764706, "grad_norm": 0.4714569015900343, "learning_rate": 2.344485207695102e-05, "loss": 0.2695, "step": 10811 }, { "epoch": 5.088, "grad_norm": 0.858221303302265, "learning_rate": 2.3421325633720094e-05, "loss": 0.4551, "step": 10812 }, { "epoch": 5.088470588235294, "grad_norm": 0.33773215915876514, "learning_rate": 2.3397810000928845e-05, "loss": 0.1455, "step": 10813 }, { "epoch": 5.0889411764705885, "grad_norm": 0.5152462492258187, "learning_rate": 2.337430518058557e-05, "loss": 0.3105, "step": 10814 }, { "epoch": 5.089411764705883, "grad_norm": 0.1877933415410546, "learning_rate": 2.335081117469777e-05, "loss": 0.0815, "step": 10815 }, { "epoch": 5.089882352941176, "grad_norm": 0.7694493280181822, "learning_rate": 2.3327327985271865e-05, "loss": 0.3867, "step": 10816 }, { "epoch": 5.09035294117647, "grad_norm": 0.5154779065266094, "learning_rate": 2.3303855614313483e-05, "loss": 0.3398, "step": 10817 }, { "epoch": 5.0908235294117645, "grad_norm": 0.3960935413307188, "learning_rate": 2.328039406382727e-05, "loss": 0.1455, "step": 10818 }, { "epoch": 5.091294117647059, "grad_norm": 0.48385251994655104, "learning_rate": 2.325694333581695e-05, "loss": 0.1719, "step": 10819 }, { "epoch": 5.091764705882353, "grad_norm": 0.39884763849616556, "learning_rate": 2.32335034322854e-05, "loss": 0.1875, "step": 10820 }, { "epoch": 5.092235294117647, "grad_norm": 0.716469108747327, "learning_rate": 2.3210074355234392e-05, "loss": 0.4355, "step": 10821 }, { "epoch": 5.092705882352941, "grad_norm": 0.3282544612096081, "learning_rate": 2.318665610666493e-05, "loss": 0.0762, "step": 10822 }, { "epoch": 5.093176470588236, "grad_norm": 0.5394296902389943, "learning_rate": 2.3163248688577098e-05, "loss": 0.3086, "step": 10823 }, { "epoch": 5.09364705882353, "grad_norm": 0.5144351492177718, "learning_rate": 2.3139852102969906e-05, "loss": 0.2754, "step": 10824 }, { "epoch": 5.094117647058823, "grad_norm": 0.33068477673312774, "learning_rate": 2.31164663518416e-05, "loss": 0.1504, "step": 10825 }, { "epoch": 5.094588235294117, "grad_norm": 0.5306836330663792, "learning_rate": 2.3093091437189476e-05, "loss": 0.3242, "step": 10826 }, { "epoch": 5.095058823529412, "grad_norm": 0.6015286692401843, "learning_rate": 2.306972736100977e-05, "loss": 0.4023, "step": 10827 }, { "epoch": 5.095529411764706, "grad_norm": 0.4257575379748502, "learning_rate": 2.304637412529794e-05, "loss": 0.2061, "step": 10828 }, { "epoch": 5.096, "grad_norm": 0.4825313304033824, "learning_rate": 2.3023031732048504e-05, "loss": 0.1846, "step": 10829 }, { "epoch": 5.096470588235294, "grad_norm": 0.4471291931081528, "learning_rate": 2.2999700183254934e-05, "loss": 0.1797, "step": 10830 }, { "epoch": 5.0969411764705885, "grad_norm": 0.5798886250897671, "learning_rate": 2.29763794809099e-05, "loss": 0.2598, "step": 10831 }, { "epoch": 5.097411764705883, "grad_norm": 0.5346870978243927, "learning_rate": 2.2953069627005132e-05, "loss": 0.3184, "step": 10832 }, { "epoch": 5.097882352941176, "grad_norm": 0.657700290343881, "learning_rate": 2.2929770623531368e-05, "loss": 0.4688, "step": 10833 }, { "epoch": 5.09835294117647, "grad_norm": 0.34398429149740756, "learning_rate": 2.290648247247845e-05, "loss": 0.1631, "step": 10834 }, { "epoch": 5.0988235294117645, "grad_norm": 0.6908474258517915, "learning_rate": 2.2883205175835368e-05, "loss": 0.4258, "step": 10835 }, { "epoch": 5.099294117647059, "grad_norm": 0.5699881503348629, "learning_rate": 2.2859938735590017e-05, "loss": 0.3047, "step": 10836 }, { "epoch": 5.099764705882353, "grad_norm": 0.4800233476231812, "learning_rate": 2.2836683153729522e-05, "loss": 0.2812, "step": 10837 }, { "epoch": 5.100235294117647, "grad_norm": 0.9147946146684863, "learning_rate": 2.2813438432240067e-05, "loss": 0.418, "step": 10838 }, { "epoch": 5.100705882352941, "grad_norm": 0.8337405990300918, "learning_rate": 2.2790204573106756e-05, "loss": 0.5, "step": 10839 }, { "epoch": 5.101176470588236, "grad_norm": 0.37777608418471, "learning_rate": 2.276698157831396e-05, "loss": 0.1621, "step": 10840 }, { "epoch": 5.101647058823529, "grad_norm": 0.18856190490606675, "learning_rate": 2.2743769449845035e-05, "loss": 0.0432, "step": 10841 }, { "epoch": 5.102117647058823, "grad_norm": 0.6752724922047423, "learning_rate": 2.272056818968235e-05, "loss": 0.3594, "step": 10842 }, { "epoch": 5.102588235294117, "grad_norm": 0.16064504239188482, "learning_rate": 2.269737779980748e-05, "loss": 0.0403, "step": 10843 }, { "epoch": 5.103058823529412, "grad_norm": 0.5661698306503583, "learning_rate": 2.2674198282200934e-05, "loss": 0.3145, "step": 10844 }, { "epoch": 5.103529411764706, "grad_norm": 0.5313007106414877, "learning_rate": 2.2651029638842372e-05, "loss": 0.2891, "step": 10845 }, { "epoch": 5.104, "grad_norm": 0.4124037525845359, "learning_rate": 2.262787187171058e-05, "loss": 0.1494, "step": 10846 }, { "epoch": 5.104470588235294, "grad_norm": 0.0001203967057777799, "learning_rate": 2.2604724982783224e-05, "loss": 0.0, "step": 10847 }, { "epoch": 5.1049411764705885, "grad_norm": 0.5491532965193301, "learning_rate": 2.258158897403724e-05, "loss": 0.1611, "step": 10848 }, { "epoch": 5.105411764705883, "grad_norm": 0.4790494234009002, "learning_rate": 2.2558463847448566e-05, "loss": 0.2637, "step": 10849 }, { "epoch": 5.105882352941176, "grad_norm": 0.7170142390251054, "learning_rate": 2.2535349604992153e-05, "loss": 0.3281, "step": 10850 }, { "epoch": 5.10635294117647, "grad_norm": 0.7580885181381597, "learning_rate": 2.251224624864209e-05, "loss": 0.5312, "step": 10851 }, { "epoch": 5.1068235294117645, "grad_norm": 0.5175872592914895, "learning_rate": 2.2489153780371566e-05, "loss": 0.2637, "step": 10852 }, { "epoch": 5.107294117647059, "grad_norm": 0.549060119308075, "learning_rate": 2.2466072202152702e-05, "loss": 0.4531, "step": 10853 }, { "epoch": 5.107764705882353, "grad_norm": 0.4095563613263302, "learning_rate": 2.2443001515956823e-05, "loss": 0.1318, "step": 10854 }, { "epoch": 5.108235294117647, "grad_norm": 0.19198065998055303, "learning_rate": 2.24199417237543e-05, "loss": 0.0388, "step": 10855 }, { "epoch": 5.108705882352941, "grad_norm": 0.5202777611601215, "learning_rate": 2.2396892827514507e-05, "loss": 0.3594, "step": 10856 }, { "epoch": 5.109176470588236, "grad_norm": 0.35402213891308404, "learning_rate": 2.2373854829205957e-05, "loss": 0.1133, "step": 10857 }, { "epoch": 5.109647058823529, "grad_norm": 0.48008549879429624, "learning_rate": 2.235082773079624e-05, "loss": 0.2275, "step": 10858 }, { "epoch": 5.110117647058823, "grad_norm": 0.45979289733595574, "learning_rate": 2.2327811534251933e-05, "loss": 0.1953, "step": 10859 }, { "epoch": 5.110588235294117, "grad_norm": 0.542129401284842, "learning_rate": 2.230480624153873e-05, "loss": 0.3125, "step": 10860 }, { "epoch": 5.111058823529412, "grad_norm": 0.6367878215607884, "learning_rate": 2.2281811854621445e-05, "loss": 0.2734, "step": 10861 }, { "epoch": 5.111529411764706, "grad_norm": 0.7022261240914044, "learning_rate": 2.2258828375463855e-05, "loss": 0.4043, "step": 10862 }, { "epoch": 5.112, "grad_norm": 0.5771029637934317, "learning_rate": 2.223585580602889e-05, "loss": 0.3652, "step": 10863 }, { "epoch": 5.112470588235294, "grad_norm": 0.3991761756536847, "learning_rate": 2.221289414827853e-05, "loss": 0.1064, "step": 10864 }, { "epoch": 5.1129411764705885, "grad_norm": 0.49469240767944883, "learning_rate": 2.218994340417378e-05, "loss": 0.373, "step": 10865 }, { "epoch": 5.113411764705883, "grad_norm": 0.5538214399504163, "learning_rate": 2.2167003575674824e-05, "loss": 0.2598, "step": 10866 }, { "epoch": 5.113882352941176, "grad_norm": 0.42911710454433993, "learning_rate": 2.2144074664740723e-05, "loss": 0.2412, "step": 10867 }, { "epoch": 5.11435294117647, "grad_norm": 0.49585266645873616, "learning_rate": 2.2121156673329794e-05, "loss": 0.1309, "step": 10868 }, { "epoch": 5.1148235294117645, "grad_norm": 0.6295952393067141, "learning_rate": 2.2098249603399348e-05, "loss": 0.3242, "step": 10869 }, { "epoch": 5.115294117647059, "grad_norm": 0.3819975634743551, "learning_rate": 2.2075353456905727e-05, "loss": 0.2002, "step": 10870 }, { "epoch": 5.115764705882353, "grad_norm": 0.32647680903001197, "learning_rate": 2.2052468235804393e-05, "loss": 0.1807, "step": 10871 }, { "epoch": 5.116235294117647, "grad_norm": 0.430776034349812, "learning_rate": 2.2029593942049838e-05, "loss": 0.1963, "step": 10872 }, { "epoch": 5.116705882352941, "grad_norm": 0.8399022144826024, "learning_rate": 2.200673057759565e-05, "loss": 0.625, "step": 10873 }, { "epoch": 5.117176470588236, "grad_norm": 0.7662423014478015, "learning_rate": 2.1983878144394497e-05, "loss": 0.4395, "step": 10874 }, { "epoch": 5.117647058823529, "grad_norm": 0.26573682527883447, "learning_rate": 2.1961036644398035e-05, "loss": 0.127, "step": 10875 }, { "epoch": 5.118117647058823, "grad_norm": 0.31727267044741114, "learning_rate": 2.193820607955707e-05, "loss": 0.0737, "step": 10876 }, { "epoch": 5.118588235294117, "grad_norm": 0.3235452369564877, "learning_rate": 2.1915386451821494e-05, "loss": 0.0972, "step": 10877 }, { "epoch": 5.119058823529412, "grad_norm": 0.6479938861784551, "learning_rate": 2.1892577763140112e-05, "loss": 0.3711, "step": 10878 }, { "epoch": 5.119529411764706, "grad_norm": 0.3279825877876936, "learning_rate": 2.186978001546097e-05, "loss": 0.1494, "step": 10879 }, { "epoch": 5.12, "grad_norm": 0.7494661821040925, "learning_rate": 2.1846993210731108e-05, "loss": 0.4746, "step": 10880 }, { "epoch": 5.120470588235294, "grad_norm": 0.548572699560124, "learning_rate": 2.1824217350896585e-05, "loss": 0.2461, "step": 10881 }, { "epoch": 5.1209411764705886, "grad_norm": 0.6582325702567311, "learning_rate": 2.180145243790261e-05, "loss": 0.3594, "step": 10882 }, { "epoch": 5.121411764705883, "grad_norm": 0.6915701315463517, "learning_rate": 2.1778698473693435e-05, "loss": 0.3281, "step": 10883 }, { "epoch": 5.121882352941176, "grad_norm": 0.42320979517177376, "learning_rate": 2.1755955460212282e-05, "loss": 0.1904, "step": 10884 }, { "epoch": 5.12235294117647, "grad_norm": 0.6661056563375338, "learning_rate": 2.1733223399401585e-05, "loss": 0.4336, "step": 10885 }, { "epoch": 5.1228235294117646, "grad_norm": 0.5906199114699362, "learning_rate": 2.1710502293202786e-05, "loss": 0.1289, "step": 10886 }, { "epoch": 5.123294117647059, "grad_norm": 0.33175770956319806, "learning_rate": 2.1687792143556303e-05, "loss": 0.1021, "step": 10887 }, { "epoch": 5.123764705882353, "grad_norm": 0.5420169791230801, "learning_rate": 2.1665092952401754e-05, "loss": 0.3359, "step": 10888 }, { "epoch": 5.124235294117647, "grad_norm": 0.3717197705321309, "learning_rate": 2.1642404721677765e-05, "loss": 0.1177, "step": 10889 }, { "epoch": 5.124705882352941, "grad_norm": 0.6325251257276567, "learning_rate": 2.1619727453321974e-05, "loss": 0.4277, "step": 10890 }, { "epoch": 5.125176470588236, "grad_norm": 0.5127337232191082, "learning_rate": 2.159706114927116e-05, "loss": 0.3984, "step": 10891 }, { "epoch": 5.125647058823529, "grad_norm": 0.3557441420245907, "learning_rate": 2.157440581146118e-05, "loss": 0.1592, "step": 10892 }, { "epoch": 5.126117647058823, "grad_norm": 0.3842426532742999, "learning_rate": 2.155176144182682e-05, "loss": 0.1621, "step": 10893 }, { "epoch": 5.1265882352941174, "grad_norm": 0.5056838894819328, "learning_rate": 2.1529128042302113e-05, "loss": 0.2637, "step": 10894 }, { "epoch": 5.127058823529412, "grad_norm": 0.5120761249815794, "learning_rate": 2.150650561481998e-05, "loss": 0.2949, "step": 10895 }, { "epoch": 5.127529411764706, "grad_norm": 0.6505976323927586, "learning_rate": 2.1483894161312525e-05, "loss": 0.3164, "step": 10896 }, { "epoch": 5.128, "grad_norm": 0.5131017210723205, "learning_rate": 2.1461293683710905e-05, "loss": 0.3574, "step": 10897 }, { "epoch": 5.128470588235294, "grad_norm": 0.6240206927958083, "learning_rate": 2.143870418394526e-05, "loss": 0.1416, "step": 10898 }, { "epoch": 5.128941176470589, "grad_norm": 0.7073038051054391, "learning_rate": 2.141612566394486e-05, "loss": 0.498, "step": 10899 }, { "epoch": 5.129411764705883, "grad_norm": 0.4059671982120817, "learning_rate": 2.1393558125638066e-05, "loss": 0.2451, "step": 10900 }, { "epoch": 5.129882352941176, "grad_norm": 0.7854435191744628, "learning_rate": 2.1371001570952186e-05, "loss": 0.3984, "step": 10901 }, { "epoch": 5.13035294117647, "grad_norm": 0.208011715239912, "learning_rate": 2.1348456001813674e-05, "loss": 0.0486, "step": 10902 }, { "epoch": 5.130823529411765, "grad_norm": 0.8211228566881239, "learning_rate": 2.132592142014811e-05, "loss": 0.2949, "step": 10903 }, { "epoch": 5.131294117647059, "grad_norm": 0.7492813362006886, "learning_rate": 2.130339782787996e-05, "loss": 0.5586, "step": 10904 }, { "epoch": 5.131764705882353, "grad_norm": 0.535644258458962, "learning_rate": 2.1280885226932877e-05, "loss": 0.207, "step": 10905 }, { "epoch": 5.132235294117647, "grad_norm": 0.36470793226854303, "learning_rate": 2.125838361922961e-05, "loss": 0.1113, "step": 10906 }, { "epoch": 5.1327058823529415, "grad_norm": 0.43881266632451343, "learning_rate": 2.1235893006691786e-05, "loss": 0.166, "step": 10907 }, { "epoch": 5.133176470588236, "grad_norm": 0.7542787900232684, "learning_rate": 2.1213413391240335e-05, "loss": 0.5547, "step": 10908 }, { "epoch": 5.133647058823529, "grad_norm": 0.5631476559082268, "learning_rate": 2.119094477479506e-05, "loss": 0.3613, "step": 10909 }, { "epoch": 5.134117647058823, "grad_norm": 0.47882432345809106, "learning_rate": 2.116848715927489e-05, "loss": 0.2285, "step": 10910 }, { "epoch": 5.1345882352941175, "grad_norm": 0.4215765534263241, "learning_rate": 2.114604054659787e-05, "loss": 0.1367, "step": 10911 }, { "epoch": 5.135058823529412, "grad_norm": 0.8269767338050953, "learning_rate": 2.1123604938680987e-05, "loss": 0.5469, "step": 10912 }, { "epoch": 5.135529411764706, "grad_norm": 0.7627086287990757, "learning_rate": 2.1101180337440378e-05, "loss": 0.4668, "step": 10913 }, { "epoch": 5.136, "grad_norm": 0.7346948248265396, "learning_rate": 2.1078766744791238e-05, "loss": 0.2871, "step": 10914 }, { "epoch": 5.136470588235294, "grad_norm": 0.25367111074139853, "learning_rate": 2.1056364162647736e-05, "loss": 0.0522, "step": 10915 }, { "epoch": 5.136941176470589, "grad_norm": 0.28373381531851827, "learning_rate": 2.1033972592923214e-05, "loss": 0.1426, "step": 10916 }, { "epoch": 5.137411764705883, "grad_norm": 0.48982245084262444, "learning_rate": 2.1011592037530023e-05, "loss": 0.2871, "step": 10917 }, { "epoch": 5.137882352941176, "grad_norm": 0.3138670659914013, "learning_rate": 2.098922249837953e-05, "loss": 0.1191, "step": 10918 }, { "epoch": 5.13835294117647, "grad_norm": 0.5597516153644221, "learning_rate": 2.096686397738223e-05, "loss": 0.3027, "step": 10919 }, { "epoch": 5.138823529411765, "grad_norm": 0.6913254215691722, "learning_rate": 2.094451647644769e-05, "loss": 0.4961, "step": 10920 }, { "epoch": 5.139294117647059, "grad_norm": 0.6035116259898997, "learning_rate": 2.0922179997484424e-05, "loss": 0.4043, "step": 10921 }, { "epoch": 5.139764705882353, "grad_norm": 0.5901208401642476, "learning_rate": 2.0899854542400147e-05, "loss": 0.1758, "step": 10922 }, { "epoch": 5.140235294117647, "grad_norm": 0.3407354323693526, "learning_rate": 2.087754011310147e-05, "loss": 0.1611, "step": 10923 }, { "epoch": 5.1407058823529415, "grad_norm": 0.7063328195691393, "learning_rate": 2.0855236711494223e-05, "loss": 0.5234, "step": 10924 }, { "epoch": 5.141176470588236, "grad_norm": 0.3050568822455041, "learning_rate": 2.083294433948324e-05, "loss": 0.1387, "step": 10925 }, { "epoch": 5.141647058823529, "grad_norm": 0.6986538643359422, "learning_rate": 2.0810662998972345e-05, "loss": 0.3008, "step": 10926 }, { "epoch": 5.142117647058823, "grad_norm": 0.3575729528429337, "learning_rate": 2.0788392691864504e-05, "loss": 0.1318, "step": 10927 }, { "epoch": 5.1425882352941175, "grad_norm": 0.37290220527158574, "learning_rate": 2.0766133420061732e-05, "loss": 0.1406, "step": 10928 }, { "epoch": 5.143058823529412, "grad_norm": 0.26306950475952245, "learning_rate": 2.0743885185465025e-05, "loss": 0.1128, "step": 10929 }, { "epoch": 5.143529411764706, "grad_norm": 0.42854929035285005, "learning_rate": 2.0721647989974514e-05, "loss": 0.1953, "step": 10930 }, { "epoch": 5.144, "grad_norm": 0.48138457642905175, "learning_rate": 2.0699421835489417e-05, "loss": 0.3105, "step": 10931 }, { "epoch": 5.144470588235294, "grad_norm": 0.3676996902646503, "learning_rate": 2.0677206723907857e-05, "loss": 0.125, "step": 10932 }, { "epoch": 5.144941176470589, "grad_norm": 0.502348955152868, "learning_rate": 2.0655002657127188e-05, "loss": 0.252, "step": 10933 }, { "epoch": 5.145411764705882, "grad_norm": 0.3548436927053814, "learning_rate": 2.0632809637043747e-05, "loss": 0.1943, "step": 10934 }, { "epoch": 5.145882352941176, "grad_norm": 0.5799840965289939, "learning_rate": 2.0610627665552875e-05, "loss": 0.3379, "step": 10935 }, { "epoch": 5.14635294117647, "grad_norm": 0.6411439301849725, "learning_rate": 2.058845674454905e-05, "loss": 0.4492, "step": 10936 }, { "epoch": 5.146823529411765, "grad_norm": 0.9729692282084957, "learning_rate": 2.0566296875925813e-05, "loss": 0.167, "step": 10937 }, { "epoch": 5.147294117647059, "grad_norm": 0.48826847008683383, "learning_rate": 2.054414806157566e-05, "loss": 0.2168, "step": 10938 }, { "epoch": 5.147764705882353, "grad_norm": 0.5406341565301532, "learning_rate": 2.052201030339024e-05, "loss": 0.209, "step": 10939 }, { "epoch": 5.148235294117647, "grad_norm": 0.6905064674179232, "learning_rate": 2.0499883603260274e-05, "loss": 0.4199, "step": 10940 }, { "epoch": 5.1487058823529415, "grad_norm": 0.1598457722793033, "learning_rate": 2.0477767963075386e-05, "loss": 0.0281, "step": 10941 }, { "epoch": 5.149176470588236, "grad_norm": 0.6980962799972082, "learning_rate": 2.045566338472443e-05, "loss": 0.5938, "step": 10942 }, { "epoch": 5.149647058823529, "grad_norm": 0.5287852152335807, "learning_rate": 2.043356987009526e-05, "loss": 0.3438, "step": 10943 }, { "epoch": 5.150117647058823, "grad_norm": 0.5737793905125697, "learning_rate": 2.041148742107471e-05, "loss": 0.2598, "step": 10944 }, { "epoch": 5.1505882352941175, "grad_norm": 0.4019593803771482, "learning_rate": 2.038941603954879e-05, "loss": 0.125, "step": 10945 }, { "epoch": 5.151058823529412, "grad_norm": 0.41100183820971015, "learning_rate": 2.036735572740245e-05, "loss": 0.2227, "step": 10946 }, { "epoch": 5.151529411764706, "grad_norm": 0.7794124433190628, "learning_rate": 2.0345306486519787e-05, "loss": 0.4941, "step": 10947 }, { "epoch": 5.152, "grad_norm": 0.41935466122214277, "learning_rate": 2.032326831878392e-05, "loss": 0.1318, "step": 10948 }, { "epoch": 5.152470588235294, "grad_norm": 0.6015181379978991, "learning_rate": 2.0301241226076963e-05, "loss": 0.373, "step": 10949 }, { "epoch": 5.152941176470589, "grad_norm": 0.3111209469700211, "learning_rate": 2.027922521028018e-05, "loss": 0.1328, "step": 10950 }, { "epoch": 5.153411764705883, "grad_norm": 0.4840168320236139, "learning_rate": 2.025722027327385e-05, "loss": 0.2559, "step": 10951 }, { "epoch": 5.153882352941176, "grad_norm": 0.5050581277195414, "learning_rate": 2.0235226416937284e-05, "loss": 0.2559, "step": 10952 }, { "epoch": 5.15435294117647, "grad_norm": 0.7057026741412697, "learning_rate": 2.0213243643148914e-05, "loss": 0.4727, "step": 10953 }, { "epoch": 5.154823529411765, "grad_norm": 0.7217598101408372, "learning_rate": 2.0191271953786104e-05, "loss": 0.4258, "step": 10954 }, { "epoch": 5.155294117647059, "grad_norm": 0.4471503812156745, "learning_rate": 2.016931135072537e-05, "loss": 0.208, "step": 10955 }, { "epoch": 5.155764705882353, "grad_norm": 0.4047101388274418, "learning_rate": 2.0147361835842296e-05, "loss": 0.1455, "step": 10956 }, { "epoch": 5.156235294117647, "grad_norm": 0.6764790646287634, "learning_rate": 2.0125423411011416e-05, "loss": 0.4004, "step": 10957 }, { "epoch": 5.1567058823529415, "grad_norm": 0.2964054591081315, "learning_rate": 2.0103496078106396e-05, "loss": 0.0972, "step": 10958 }, { "epoch": 5.157176470588236, "grad_norm": 0.7581434433085361, "learning_rate": 2.0081579839000005e-05, "loss": 0.3711, "step": 10959 }, { "epoch": 5.157647058823529, "grad_norm": 0.33350526266629743, "learning_rate": 2.00596746955639e-05, "loss": 0.1021, "step": 10960 }, { "epoch": 5.158117647058823, "grad_norm": 0.21294366189717356, "learning_rate": 2.0037780649668912e-05, "loss": 0.0544, "step": 10961 }, { "epoch": 5.1585882352941175, "grad_norm": 0.43766626086217175, "learning_rate": 2.001589770318498e-05, "loss": 0.1514, "step": 10962 }, { "epoch": 5.159058823529412, "grad_norm": 0.4085998316730532, "learning_rate": 1.9994025857980896e-05, "loss": 0.1836, "step": 10963 }, { "epoch": 5.159529411764706, "grad_norm": 0.6381985125855419, "learning_rate": 1.9972165115924695e-05, "loss": 0.332, "step": 10964 }, { "epoch": 5.16, "grad_norm": 0.6829534819325667, "learning_rate": 1.99503154788834e-05, "loss": 0.3789, "step": 10965 }, { "epoch": 5.160470588235294, "grad_norm": 0.5909935214489015, "learning_rate": 1.9928476948723022e-05, "loss": 0.2891, "step": 10966 }, { "epoch": 5.160941176470589, "grad_norm": 0.5671638320570583, "learning_rate": 1.99066495273087e-05, "loss": 0.332, "step": 10967 }, { "epoch": 5.161411764705882, "grad_norm": 0.44359483615897816, "learning_rate": 1.9884833216504657e-05, "loss": 0.2734, "step": 10968 }, { "epoch": 5.161882352941176, "grad_norm": 0.43105354433519416, "learning_rate": 1.9863028018174043e-05, "loss": 0.2344, "step": 10969 }, { "epoch": 5.16235294117647, "grad_norm": 0.6509822164291087, "learning_rate": 1.9841233934179145e-05, "loss": 0.4141, "step": 10970 }, { "epoch": 5.162823529411765, "grad_norm": 0.6693317394104717, "learning_rate": 1.9819450966381328e-05, "loss": 0.3633, "step": 10971 }, { "epoch": 5.163294117647059, "grad_norm": 0.5001097973741658, "learning_rate": 1.9797679116640903e-05, "loss": 0.2207, "step": 10972 }, { "epoch": 5.163764705882353, "grad_norm": 0.7691573899996698, "learning_rate": 1.9775918386817346e-05, "loss": 0.6328, "step": 10973 }, { "epoch": 5.164235294117647, "grad_norm": 0.5631328105440062, "learning_rate": 1.9754168778769058e-05, "loss": 0.2734, "step": 10974 }, { "epoch": 5.1647058823529415, "grad_norm": 0.2950566771931655, "learning_rate": 1.9732430294353636e-05, "loss": 0.0991, "step": 10975 }, { "epoch": 5.165176470588236, "grad_norm": 0.3837874369714999, "learning_rate": 1.9710702935427642e-05, "loss": 0.1191, "step": 10976 }, { "epoch": 5.165647058823529, "grad_norm": 0.5187250826880748, "learning_rate": 1.9688986703846658e-05, "loss": 0.2891, "step": 10977 }, { "epoch": 5.166117647058823, "grad_norm": 0.6111766306252314, "learning_rate": 1.9667281601465372e-05, "loss": 0.3164, "step": 10978 }, { "epoch": 5.1665882352941175, "grad_norm": 0.48236442279357167, "learning_rate": 1.964558763013755e-05, "loss": 0.1758, "step": 10979 }, { "epoch": 5.167058823529412, "grad_norm": 0.7318676271915407, "learning_rate": 1.9623904791715907e-05, "loss": 0.4941, "step": 10980 }, { "epoch": 5.167529411764706, "grad_norm": 0.6254504084676075, "learning_rate": 1.9602233088052277e-05, "loss": 0.2598, "step": 10981 }, { "epoch": 5.168, "grad_norm": 0.5635282621681862, "learning_rate": 1.9580572520997566e-05, "loss": 0.1572, "step": 10982 }, { "epoch": 5.168470588235294, "grad_norm": 0.6319184382693207, "learning_rate": 1.9558923092401645e-05, "loss": 0.3281, "step": 10983 }, { "epoch": 5.168941176470589, "grad_norm": 0.430150141847672, "learning_rate": 1.95372848041135e-05, "loss": 0.1138, "step": 10984 }, { "epoch": 5.169411764705882, "grad_norm": 0.6541346068231193, "learning_rate": 1.951565765798117e-05, "loss": 0.3145, "step": 10985 }, { "epoch": 5.169882352941176, "grad_norm": 0.6170746440110085, "learning_rate": 1.9494041655851677e-05, "loss": 0.3828, "step": 10986 }, { "epoch": 5.17035294117647, "grad_norm": 0.4812967965644192, "learning_rate": 1.9472436799571145e-05, "loss": 0.2852, "step": 10987 }, { "epoch": 5.170823529411765, "grad_norm": 0.5181891979617955, "learning_rate": 1.9450843090984797e-05, "loss": 0.2676, "step": 10988 }, { "epoch": 5.171294117647059, "grad_norm": 0.26511639662285225, "learning_rate": 1.9429260531936743e-05, "loss": 0.0486, "step": 10989 }, { "epoch": 5.171764705882353, "grad_norm": 0.40473902312489546, "learning_rate": 1.940768912427027e-05, "loss": 0.2061, "step": 10990 }, { "epoch": 5.172235294117647, "grad_norm": 0.4391106385371334, "learning_rate": 1.9386128869827743e-05, "loss": 0.1245, "step": 10991 }, { "epoch": 5.1727058823529415, "grad_norm": 0.4987153407244807, "learning_rate": 1.936457977045044e-05, "loss": 0.2852, "step": 10992 }, { "epoch": 5.173176470588236, "grad_norm": 0.36050471553133934, "learning_rate": 1.9343041827978785e-05, "loss": 0.1279, "step": 10993 }, { "epoch": 5.173647058823529, "grad_norm": 0.5551522321568246, "learning_rate": 1.9321515044252273e-05, "loss": 0.2852, "step": 10994 }, { "epoch": 5.174117647058823, "grad_norm": 0.3502305262768381, "learning_rate": 1.9299999421109303e-05, "loss": 0.1562, "step": 10995 }, { "epoch": 5.1745882352941175, "grad_norm": 0.5777030032705299, "learning_rate": 1.9278494960387452e-05, "loss": 0.3457, "step": 10996 }, { "epoch": 5.175058823529412, "grad_norm": 0.34061866549052927, "learning_rate": 1.9257001663923338e-05, "loss": 0.1196, "step": 10997 }, { "epoch": 5.175529411764706, "grad_norm": 0.2827796631712738, "learning_rate": 1.9235519533552556e-05, "loss": 0.106, "step": 10998 }, { "epoch": 5.176, "grad_norm": 0.4191020254410094, "learning_rate": 1.921404857110984e-05, "loss": 0.2539, "step": 10999 }, { "epoch": 5.176470588235294, "grad_norm": 0.5384304230313963, "learning_rate": 1.9192588778428842e-05, "loss": 0.3164, "step": 11000 }, { "epoch": 5.176941176470589, "grad_norm": 0.37851733922234465, "learning_rate": 1.9171140157342407e-05, "loss": 0.1426, "step": 11001 }, { "epoch": 5.177411764705882, "grad_norm": 0.6030976898506717, "learning_rate": 1.9149702709682292e-05, "loss": 0.3906, "step": 11002 }, { "epoch": 5.177882352941176, "grad_norm": 0.499956192628437, "learning_rate": 1.9128276437279362e-05, "loss": 0.2637, "step": 11003 }, { "epoch": 5.17835294117647, "grad_norm": 0.5549682732399184, "learning_rate": 1.9106861341963587e-05, "loss": 0.3359, "step": 11004 }, { "epoch": 5.178823529411765, "grad_norm": 0.3563184974020185, "learning_rate": 1.9085457425563855e-05, "loss": 0.1104, "step": 11005 }, { "epoch": 5.179294117647059, "grad_norm": 0.2788437203711393, "learning_rate": 1.9064064689908187e-05, "loss": 0.0908, "step": 11006 }, { "epoch": 5.179764705882353, "grad_norm": 0.24054462914096275, "learning_rate": 1.9042683136823654e-05, "loss": 0.0383, "step": 11007 }, { "epoch": 5.180235294117647, "grad_norm": 0.15527356600860323, "learning_rate": 1.9021312768136314e-05, "loss": 0.0312, "step": 11008 }, { "epoch": 5.1807058823529415, "grad_norm": 0.44302967299702006, "learning_rate": 1.899995358567129e-05, "loss": 0.2617, "step": 11009 }, { "epoch": 5.181176470588236, "grad_norm": 0.6175390211572375, "learning_rate": 1.897860559125282e-05, "loss": 0.3867, "step": 11010 }, { "epoch": 5.181647058823529, "grad_norm": 0.6165413652077674, "learning_rate": 1.895726878670405e-05, "loss": 0.3633, "step": 11011 }, { "epoch": 5.182117647058823, "grad_norm": 0.4358945206189115, "learning_rate": 1.8935943173847302e-05, "loss": 0.1377, "step": 11012 }, { "epoch": 5.1825882352941175, "grad_norm": 0.5086277404816077, "learning_rate": 1.8914628754503885e-05, "loss": 0.2158, "step": 11013 }, { "epoch": 5.183058823529412, "grad_norm": 0.5308781834178485, "learning_rate": 1.889332553049413e-05, "loss": 0.2637, "step": 11014 }, { "epoch": 5.183529411764706, "grad_norm": 0.6473615628260719, "learning_rate": 1.887203350363743e-05, "loss": 0.4824, "step": 11015 }, { "epoch": 5.184, "grad_norm": 0.41187469595051274, "learning_rate": 1.8850752675752295e-05, "loss": 0.1465, "step": 11016 }, { "epoch": 5.184470588235294, "grad_norm": 0.35062365683446284, "learning_rate": 1.8829483048656124e-05, "loss": 0.0928, "step": 11017 }, { "epoch": 5.184941176470589, "grad_norm": 0.37810053076411626, "learning_rate": 1.8808224624165497e-05, "loss": 0.0918, "step": 11018 }, { "epoch": 5.185411764705882, "grad_norm": 0.46858694860117955, "learning_rate": 1.878697740409601e-05, "loss": 0.1602, "step": 11019 }, { "epoch": 5.185882352941176, "grad_norm": 0.8392844708068052, "learning_rate": 1.8765741390262212e-05, "loss": 0.1177, "step": 11020 }, { "epoch": 5.18635294117647, "grad_norm": 0.4838246698091082, "learning_rate": 1.8744516584477803e-05, "loss": 0.2051, "step": 11021 }, { "epoch": 5.186823529411765, "grad_norm": 0.2298394511181167, "learning_rate": 1.87233029885555e-05, "loss": 0.0479, "step": 11022 }, { "epoch": 5.187294117647059, "grad_norm": 0.378931017266467, "learning_rate": 1.8702100604307018e-05, "loss": 0.1729, "step": 11023 }, { "epoch": 5.187764705882353, "grad_norm": 0.5221223906170993, "learning_rate": 1.868090943354318e-05, "loss": 0.3125, "step": 11024 }, { "epoch": 5.188235294117647, "grad_norm": 0.6856062261530209, "learning_rate": 1.8659729478073764e-05, "loss": 0.3359, "step": 11025 }, { "epoch": 5.1887058823529415, "grad_norm": 0.5414769588308866, "learning_rate": 1.8638560739707664e-05, "loss": 0.2949, "step": 11026 }, { "epoch": 5.189176470588236, "grad_norm": 0.33910468024812157, "learning_rate": 1.8617403220252846e-05, "loss": 0.1455, "step": 11027 }, { "epoch": 5.189647058823529, "grad_norm": 0.58237241704804, "learning_rate": 1.8596256921516166e-05, "loss": 0.3418, "step": 11028 }, { "epoch": 5.190117647058823, "grad_norm": 0.4729180663658501, "learning_rate": 1.8575121845303698e-05, "loss": 0.1807, "step": 11029 }, { "epoch": 5.1905882352941175, "grad_norm": 0.3860655246950132, "learning_rate": 1.8553997993420495e-05, "loss": 0.1992, "step": 11030 }, { "epoch": 5.191058823529412, "grad_norm": 0.41683232856854846, "learning_rate": 1.8532885367670547e-05, "loss": 0.1533, "step": 11031 }, { "epoch": 5.191529411764706, "grad_norm": 0.6061741688920248, "learning_rate": 1.8511783969857046e-05, "loss": 0.3184, "step": 11032 }, { "epoch": 5.192, "grad_norm": 0.4075525237890654, "learning_rate": 1.849069380178217e-05, "loss": 0.1895, "step": 11033 }, { "epoch": 5.192470588235294, "grad_norm": 0.6761945203571816, "learning_rate": 1.8469614865247075e-05, "loss": 0.3262, "step": 11034 }, { "epoch": 5.192941176470589, "grad_norm": 0.6001242212478131, "learning_rate": 1.8448547162052002e-05, "loss": 0.375, "step": 11035 }, { "epoch": 5.193411764705882, "grad_norm": 0.30297960414079494, "learning_rate": 1.8427490693996298e-05, "loss": 0.1367, "step": 11036 }, { "epoch": 5.193882352941176, "grad_norm": 0.526776086363489, "learning_rate": 1.840644546287821e-05, "loss": 0.3906, "step": 11037 }, { "epoch": 5.19435294117647, "grad_norm": 0.30153592299864035, "learning_rate": 1.838541147049515e-05, "loss": 0.0845, "step": 11038 }, { "epoch": 5.194823529411765, "grad_norm": 0.6875446703989818, "learning_rate": 1.8364388718643542e-05, "loss": 0.2637, "step": 11039 }, { "epoch": 5.195294117647059, "grad_norm": 0.6086774594417909, "learning_rate": 1.8343377209118755e-05, "loss": 0.2441, "step": 11040 }, { "epoch": 5.195764705882353, "grad_norm": 0.529872564879949, "learning_rate": 1.8322376943715382e-05, "loss": 0.3926, "step": 11041 }, { "epoch": 5.196235294117647, "grad_norm": 0.31031277484822817, "learning_rate": 1.8301387924226863e-05, "loss": 0.1289, "step": 11042 }, { "epoch": 5.1967058823529415, "grad_norm": 0.5515921385377917, "learning_rate": 1.828041015244579e-05, "loss": 0.2773, "step": 11043 }, { "epoch": 5.197176470588236, "grad_norm": 0.40921084813661895, "learning_rate": 1.82594436301638e-05, "loss": 0.1865, "step": 11044 }, { "epoch": 5.197647058823529, "grad_norm": 0.664980176110861, "learning_rate": 1.8238488359171476e-05, "loss": 0.3652, "step": 11045 }, { "epoch": 5.198117647058823, "grad_norm": 0.30265348834688294, "learning_rate": 1.821754434125854e-05, "loss": 0.1387, "step": 11046 }, { "epoch": 5.1985882352941175, "grad_norm": 0.2553479569582728, "learning_rate": 1.8196611578213738e-05, "loss": 0.0752, "step": 11047 }, { "epoch": 5.199058823529412, "grad_norm": 0.4065876628622822, "learning_rate": 1.817569007182476e-05, "loss": 0.2256, "step": 11048 }, { "epoch": 5.199529411764706, "grad_norm": 0.486489831470078, "learning_rate": 1.815477982387846e-05, "loss": 0.125, "step": 11049 }, { "epoch": 5.2, "grad_norm": 0.7595066687667247, "learning_rate": 1.813388083616068e-05, "loss": 0.4121, "step": 11050 }, { "epoch": 5.200470588235294, "grad_norm": 0.6326799396929674, "learning_rate": 1.8112993110456235e-05, "loss": 0.252, "step": 11051 }, { "epoch": 5.200941176470589, "grad_norm": 0.497302021951895, "learning_rate": 1.8092116648549136e-05, "loss": 0.2314, "step": 11052 }, { "epoch": 5.201411764705882, "grad_norm": 0.4136301032218314, "learning_rate": 1.8071251452222224e-05, "loss": 0.207, "step": 11053 }, { "epoch": 5.201882352941176, "grad_norm": 0.9838377940450506, "learning_rate": 1.8050397523257555e-05, "loss": 0.5039, "step": 11054 }, { "epoch": 5.20235294117647, "grad_norm": 0.4591138184560018, "learning_rate": 1.8029554863436186e-05, "loss": 0.2461, "step": 11055 }, { "epoch": 5.202823529411765, "grad_norm": 0.4724503384631203, "learning_rate": 1.8008723474538095e-05, "loss": 0.2334, "step": 11056 }, { "epoch": 5.203294117647059, "grad_norm": 0.4263871930195246, "learning_rate": 1.798790335834244e-05, "loss": 0.1611, "step": 11057 }, { "epoch": 5.203764705882353, "grad_norm": 0.5621246921578794, "learning_rate": 1.796709451662739e-05, "loss": 0.2637, "step": 11058 }, { "epoch": 5.204235294117647, "grad_norm": 0.16418872719996971, "learning_rate": 1.7946296951170065e-05, "loss": 0.0287, "step": 11059 }, { "epoch": 5.2047058823529415, "grad_norm": 0.5592850824629169, "learning_rate": 1.792551066374668e-05, "loss": 0.3008, "step": 11060 }, { "epoch": 5.205176470588235, "grad_norm": 0.6324166453414495, "learning_rate": 1.7904735656132545e-05, "loss": 0.4043, "step": 11061 }, { "epoch": 5.205647058823529, "grad_norm": 0.588360205378604, "learning_rate": 1.7883971930101883e-05, "loss": 0.3945, "step": 11062 }, { "epoch": 5.206117647058823, "grad_norm": 0.6283472522528736, "learning_rate": 1.786321948742803e-05, "loss": 0.4629, "step": 11063 }, { "epoch": 5.2065882352941175, "grad_norm": 0.27706459529441896, "learning_rate": 1.78424783298834e-05, "loss": 0.0713, "step": 11064 }, { "epoch": 5.207058823529412, "grad_norm": 0.28325386745072184, "learning_rate": 1.782174845923931e-05, "loss": 0.1118, "step": 11065 }, { "epoch": 5.207529411764706, "grad_norm": 0.6916969183243291, "learning_rate": 1.7801029877266237e-05, "loss": 0.373, "step": 11066 }, { "epoch": 5.208, "grad_norm": 0.9051839888656615, "learning_rate": 1.778032258573368e-05, "loss": 0.5312, "step": 11067 }, { "epoch": 5.208470588235294, "grad_norm": 0.2783716326851588, "learning_rate": 1.7759626586410074e-05, "loss": 0.1167, "step": 11068 }, { "epoch": 5.208941176470589, "grad_norm": 0.5543552303887129, "learning_rate": 1.7738941881063007e-05, "loss": 0.2676, "step": 11069 }, { "epoch": 5.209411764705882, "grad_norm": 1.3268125869942125, "learning_rate": 1.7718268471459045e-05, "loss": 0.5625, "step": 11070 }, { "epoch": 5.209882352941176, "grad_norm": 0.30745829848291273, "learning_rate": 1.769760635936378e-05, "loss": 0.0894, "step": 11071 }, { "epoch": 5.21035294117647, "grad_norm": 0.5717543307449346, "learning_rate": 1.7676955546541867e-05, "loss": 0.3457, "step": 11072 }, { "epoch": 5.210823529411765, "grad_norm": 0.6254307763705277, "learning_rate": 1.7656316034757024e-05, "loss": 0.3613, "step": 11073 }, { "epoch": 5.211294117647059, "grad_norm": 0.5053980220892674, "learning_rate": 1.7635687825771894e-05, "loss": 0.2676, "step": 11074 }, { "epoch": 5.211764705882353, "grad_norm": 0.5214499180969752, "learning_rate": 1.7615070921348316e-05, "loss": 0.293, "step": 11075 }, { "epoch": 5.212235294117647, "grad_norm": 0.33890984971091503, "learning_rate": 1.7594465323246987e-05, "loss": 0.2471, "step": 11076 }, { "epoch": 5.2127058823529415, "grad_norm": 0.49488232148235467, "learning_rate": 1.757387103322776e-05, "loss": 0.2412, "step": 11077 }, { "epoch": 5.213176470588236, "grad_norm": 0.5818143051992359, "learning_rate": 1.7553288053049515e-05, "loss": 0.4453, "step": 11078 }, { "epoch": 5.213647058823529, "grad_norm": 0.47979611925818455, "learning_rate": 1.7532716384470108e-05, "loss": 0.291, "step": 11079 }, { "epoch": 5.214117647058823, "grad_norm": 0.4265577259138019, "learning_rate": 1.7512156029246453e-05, "loss": 0.25, "step": 11080 }, { "epoch": 5.2145882352941175, "grad_norm": 0.5810588991437382, "learning_rate": 1.7491606989134566e-05, "loss": 0.2832, "step": 11081 }, { "epoch": 5.215058823529412, "grad_norm": 0.35703356954221693, "learning_rate": 1.7471069265889353e-05, "loss": 0.0864, "step": 11082 }, { "epoch": 5.215529411764706, "grad_norm": 0.5041528524685925, "learning_rate": 1.7450542861264883e-05, "loss": 0.165, "step": 11083 }, { "epoch": 5.216, "grad_norm": 0.812775334039842, "learning_rate": 1.7430027777014217e-05, "loss": 0.373, "step": 11084 }, { "epoch": 5.216470588235294, "grad_norm": 0.5426719736425157, "learning_rate": 1.7409524014889388e-05, "loss": 0.208, "step": 11085 }, { "epoch": 5.216941176470589, "grad_norm": 0.6384450299425035, "learning_rate": 1.738903157664161e-05, "loss": 0.3594, "step": 11086 }, { "epoch": 5.217411764705882, "grad_norm": 0.6177168358885222, "learning_rate": 1.736855046402097e-05, "loss": 0.4844, "step": 11087 }, { "epoch": 5.217882352941176, "grad_norm": 0.3838849749575654, "learning_rate": 1.7348080678776665e-05, "loss": 0.0684, "step": 11088 }, { "epoch": 5.21835294117647, "grad_norm": 0.6980135601878689, "learning_rate": 1.732762222265695e-05, "loss": 0.3477, "step": 11089 }, { "epoch": 5.218823529411765, "grad_norm": 0.38701471528724923, "learning_rate": 1.7307175097409044e-05, "loss": 0.1338, "step": 11090 }, { "epoch": 5.219294117647059, "grad_norm": 0.3045977318063695, "learning_rate": 1.7286739304779214e-05, "loss": 0.1289, "step": 11091 }, { "epoch": 5.219764705882353, "grad_norm": 0.35797255401848166, "learning_rate": 1.7266314846512848e-05, "loss": 0.1914, "step": 11092 }, { "epoch": 5.220235294117647, "grad_norm": 0.6494071050856394, "learning_rate": 1.7245901724354218e-05, "loss": 0.3027, "step": 11093 }, { "epoch": 5.2207058823529415, "grad_norm": 0.5281612117783803, "learning_rate": 1.7225499940046726e-05, "loss": 0.3281, "step": 11094 }, { "epoch": 5.221176470588235, "grad_norm": 0.3225906854532867, "learning_rate": 1.720510949533282e-05, "loss": 0.0781, "step": 11095 }, { "epoch": 5.221647058823529, "grad_norm": 0.574518891367864, "learning_rate": 1.71847303919539e-05, "loss": 0.2695, "step": 11096 }, { "epoch": 5.222117647058823, "grad_norm": 0.6221043235319252, "learning_rate": 1.7164362631650457e-05, "loss": 0.4629, "step": 11097 }, { "epoch": 5.2225882352941175, "grad_norm": 0.30765491562133235, "learning_rate": 1.7144006216162036e-05, "loss": 0.0913, "step": 11098 }, { "epoch": 5.223058823529412, "grad_norm": 0.7990750763069119, "learning_rate": 1.7123661147227096e-05, "loss": 0.3594, "step": 11099 }, { "epoch": 5.223529411764706, "grad_norm": 0.36559883132410925, "learning_rate": 1.7103327426583265e-05, "loss": 0.1455, "step": 11100 }, { "epoch": 5.224, "grad_norm": 0.43706737700431175, "learning_rate": 1.7083005055967146e-05, "loss": 0.1416, "step": 11101 }, { "epoch": 5.224470588235294, "grad_norm": 0.5982057158121205, "learning_rate": 1.7062694037114326e-05, "loss": 0.418, "step": 11102 }, { "epoch": 5.224941176470589, "grad_norm": 0.47203573093657447, "learning_rate": 1.7042394371759516e-05, "loss": 0.2178, "step": 11103 }, { "epoch": 5.225411764705882, "grad_norm": 0.7265401566137805, "learning_rate": 1.7022106061636342e-05, "loss": 0.3477, "step": 11104 }, { "epoch": 5.225882352941176, "grad_norm": 0.44610785186114216, "learning_rate": 1.700182910847757e-05, "loss": 0.1011, "step": 11105 }, { "epoch": 5.22635294117647, "grad_norm": 0.7700097673771189, "learning_rate": 1.6981563514014967e-05, "loss": 0.3262, "step": 11106 }, { "epoch": 5.226823529411765, "grad_norm": 1.027958771955523, "learning_rate": 1.6961309279979278e-05, "loss": 0.3965, "step": 11107 }, { "epoch": 5.227294117647059, "grad_norm": 0.7137409259406567, "learning_rate": 1.6941066408100308e-05, "loss": 0.4434, "step": 11108 }, { "epoch": 5.227764705882353, "grad_norm": 0.6270290464061551, "learning_rate": 1.6920834900106967e-05, "loss": 0.2695, "step": 11109 }, { "epoch": 5.228235294117647, "grad_norm": 0.5728371367229876, "learning_rate": 1.690061475772703e-05, "loss": 0.3496, "step": 11110 }, { "epoch": 5.2287058823529415, "grad_norm": 0.3918985185422062, "learning_rate": 1.6880405982687455e-05, "loss": 0.1406, "step": 11111 }, { "epoch": 5.229176470588235, "grad_norm": 0.6455447473699856, "learning_rate": 1.6860208576714173e-05, "loss": 0.2578, "step": 11112 }, { "epoch": 5.229647058823529, "grad_norm": 0.7421980432664831, "learning_rate": 1.68400225415321e-05, "loss": 0.3691, "step": 11113 }, { "epoch": 5.230117647058823, "grad_norm": 0.7059900399172873, "learning_rate": 1.6819847878865255e-05, "loss": 0.5, "step": 11114 }, { "epoch": 5.2305882352941175, "grad_norm": 0.719571647293878, "learning_rate": 1.6799684590436663e-05, "loss": 0.2129, "step": 11115 }, { "epoch": 5.231058823529412, "grad_norm": 0.4843301016717478, "learning_rate": 1.6779532677968327e-05, "loss": 0.207, "step": 11116 }, { "epoch": 5.231529411764706, "grad_norm": 0.7748613293909498, "learning_rate": 1.6759392143181354e-05, "loss": 0.5859, "step": 11117 }, { "epoch": 5.232, "grad_norm": 0.3965166012367888, "learning_rate": 1.6739262987795854e-05, "loss": 0.0854, "step": 11118 }, { "epoch": 5.232470588235294, "grad_norm": 0.35620438168023694, "learning_rate": 1.6719145213530898e-05, "loss": 0.1436, "step": 11119 }, { "epoch": 5.232941176470589, "grad_norm": 0.5422682923973979, "learning_rate": 1.669903882210468e-05, "loss": 0.3008, "step": 11120 }, { "epoch": 5.233411764705882, "grad_norm": 0.4493248430953017, "learning_rate": 1.667894381523443e-05, "loss": 0.1611, "step": 11121 }, { "epoch": 5.233882352941176, "grad_norm": 0.3894582412651863, "learning_rate": 1.665886019463627e-05, "loss": 0.1855, "step": 11122 }, { "epoch": 5.23435294117647, "grad_norm": 0.4824454571335452, "learning_rate": 1.6638787962025495e-05, "loss": 0.3105, "step": 11123 }, { "epoch": 5.234823529411765, "grad_norm": 0.7331578387768722, "learning_rate": 1.6618727119116388e-05, "loss": 0.3594, "step": 11124 }, { "epoch": 5.235294117647059, "grad_norm": 0.54843189514206, "learning_rate": 1.6598677667622175e-05, "loss": 0.2812, "step": 11125 }, { "epoch": 5.235764705882353, "grad_norm": 0.5002946001958744, "learning_rate": 1.657863960925525e-05, "loss": 0.2617, "step": 11126 }, { "epoch": 5.236235294117647, "grad_norm": 0.5327667212258809, "learning_rate": 1.655861294572692e-05, "loss": 0.3223, "step": 11127 }, { "epoch": 5.2367058823529415, "grad_norm": 0.24681669408206294, "learning_rate": 1.6538597678747555e-05, "loss": 0.0486, "step": 11128 }, { "epoch": 5.237176470588235, "grad_norm": 0.708182708553355, "learning_rate": 1.651859381002658e-05, "loss": 0.3828, "step": 11129 }, { "epoch": 5.237647058823529, "grad_norm": 0.38678347801186513, "learning_rate": 1.649860134127241e-05, "loss": 0.0815, "step": 11130 }, { "epoch": 5.238117647058823, "grad_norm": 0.8382742257229862, "learning_rate": 1.647862027419251e-05, "loss": 0.5312, "step": 11131 }, { "epoch": 5.2385882352941175, "grad_norm": 0.40376788506371153, "learning_rate": 1.6458650610493385e-05, "loss": 0.0923, "step": 11132 }, { "epoch": 5.239058823529412, "grad_norm": 0.46486767944742513, "learning_rate": 1.6438692351880482e-05, "loss": 0.3145, "step": 11133 }, { "epoch": 5.239529411764706, "grad_norm": 0.13141163843887627, "learning_rate": 1.6418745500058395e-05, "loss": 0.0233, "step": 11134 }, { "epoch": 5.24, "grad_norm": 0.8542085225432041, "learning_rate": 1.639881005673064e-05, "loss": 0.4668, "step": 11135 }, { "epoch": 5.240470588235294, "grad_norm": 0.5514382873481231, "learning_rate": 1.6378886023599803e-05, "loss": 0.167, "step": 11136 }, { "epoch": 5.240941176470589, "grad_norm": 0.8700622107978425, "learning_rate": 1.6358973402367524e-05, "loss": 0.3594, "step": 11137 }, { "epoch": 5.241411764705882, "grad_norm": 0.3864180771729962, "learning_rate": 1.6339072194734414e-05, "loss": 0.2734, "step": 11138 }, { "epoch": 5.241882352941176, "grad_norm": 0.11086151530246754, "learning_rate": 1.6319182402400144e-05, "loss": 0.0283, "step": 11139 }, { "epoch": 5.24235294117647, "grad_norm": 0.6281131549612438, "learning_rate": 1.629930402706342e-05, "loss": 0.3848, "step": 11140 }, { "epoch": 5.242823529411765, "grad_norm": 0.4624119406202885, "learning_rate": 1.6279437070421885e-05, "loss": 0.2734, "step": 11141 }, { "epoch": 5.243294117647059, "grad_norm": 0.4221166821232096, "learning_rate": 1.625958153417235e-05, "loss": 0.2002, "step": 11142 }, { "epoch": 5.243764705882353, "grad_norm": 0.4005139297499571, "learning_rate": 1.6239737420010558e-05, "loss": 0.1094, "step": 11143 }, { "epoch": 5.244235294117647, "grad_norm": 0.32502059637763153, "learning_rate": 1.621990472963125e-05, "loss": 0.0596, "step": 11144 }, { "epoch": 5.2447058823529416, "grad_norm": 0.5570980803356085, "learning_rate": 1.620008346472826e-05, "loss": 0.2695, "step": 11145 }, { "epoch": 5.245176470588235, "grad_norm": 0.816908881470293, "learning_rate": 1.618027362699446e-05, "loss": 0.4453, "step": 11146 }, { "epoch": 5.245647058823529, "grad_norm": 0.3774399819979487, "learning_rate": 1.6160475218121654e-05, "loss": 0.0874, "step": 11147 }, { "epoch": 5.246117647058823, "grad_norm": 0.9264679131989849, "learning_rate": 1.6140688239800736e-05, "loss": 0.5312, "step": 11148 }, { "epoch": 5.2465882352941176, "grad_norm": 1.053198202728969, "learning_rate": 1.6120912693721654e-05, "loss": 0.8945, "step": 11149 }, { "epoch": 5.247058823529412, "grad_norm": 0.5632409826506611, "learning_rate": 1.6101148581573274e-05, "loss": 0.3203, "step": 11150 }, { "epoch": 5.247529411764706, "grad_norm": 0.5679977415367025, "learning_rate": 1.6081395905043565e-05, "loss": 0.1328, "step": 11151 }, { "epoch": 5.248, "grad_norm": 0.5573096246689637, "learning_rate": 1.606165466581954e-05, "loss": 0.1846, "step": 11152 }, { "epoch": 5.248470588235294, "grad_norm": 0.3974406557889323, "learning_rate": 1.6041924865587152e-05, "loss": 0.1216, "step": 11153 }, { "epoch": 5.248941176470589, "grad_norm": 0.6175804724508328, "learning_rate": 1.602220650603147e-05, "loss": 0.375, "step": 11154 }, { "epoch": 5.249411764705882, "grad_norm": 0.30128608495025744, "learning_rate": 1.600249958883648e-05, "loss": 0.1523, "step": 11155 }, { "epoch": 5.249882352941176, "grad_norm": 0.6884448337138384, "learning_rate": 1.5982804115685273e-05, "loss": 0.4805, "step": 11156 }, { "epoch": 5.2503529411764704, "grad_norm": 0.5448769789688502, "learning_rate": 1.5963120088259977e-05, "loss": 0.168, "step": 11157 }, { "epoch": 5.250823529411765, "grad_norm": 0.8084554036943129, "learning_rate": 1.594344750824164e-05, "loss": 0.4902, "step": 11158 }, { "epoch": 5.251294117647059, "grad_norm": 0.7004824325183553, "learning_rate": 1.5923786377310433e-05, "loss": 0.3457, "step": 11159 }, { "epoch": 5.251764705882353, "grad_norm": 0.5514506148898018, "learning_rate": 1.5904136697145542e-05, "loss": 0.3086, "step": 11160 }, { "epoch": 5.252235294117647, "grad_norm": 0.4781961626169468, "learning_rate": 1.5884498469425078e-05, "loss": 0.126, "step": 11161 }, { "epoch": 5.252705882352942, "grad_norm": 0.320629335899679, "learning_rate": 1.5864871695826282e-05, "loss": 0.0977, "step": 11162 }, { "epoch": 5.253176470588235, "grad_norm": 0.6443603855694662, "learning_rate": 1.5845256378025395e-05, "loss": 0.3711, "step": 11163 }, { "epoch": 5.253647058823529, "grad_norm": 0.4812006817630589, "learning_rate": 1.582565251769761e-05, "loss": 0.127, "step": 11164 }, { "epoch": 5.254117647058823, "grad_norm": 0.6033579064888271, "learning_rate": 1.5806060116517227e-05, "loss": 0.2559, "step": 11165 }, { "epoch": 5.254588235294118, "grad_norm": 0.5202048361445872, "learning_rate": 1.578647917615757e-05, "loss": 0.2188, "step": 11166 }, { "epoch": 5.255058823529412, "grad_norm": 0.23128709852347426, "learning_rate": 1.5766909698290863e-05, "loss": 0.0688, "step": 11167 }, { "epoch": 5.255529411764706, "grad_norm": 0.3852332015858808, "learning_rate": 1.574735168458846e-05, "loss": 0.2051, "step": 11168 }, { "epoch": 5.256, "grad_norm": 0.4254646829875708, "learning_rate": 1.572780513672079e-05, "loss": 0.1299, "step": 11169 }, { "epoch": 5.2564705882352944, "grad_norm": 0.12468015149956528, "learning_rate": 1.5708270056357115e-05, "loss": 0.0747, "step": 11170 }, { "epoch": 5.256941176470589, "grad_norm": 0.6049236627747978, "learning_rate": 1.5688746445165867e-05, "loss": 0.3047, "step": 11171 }, { "epoch": 5.257411764705882, "grad_norm": 0.6105310797119738, "learning_rate": 1.5669234304814515e-05, "loss": 0.3184, "step": 11172 }, { "epoch": 5.257882352941176, "grad_norm": 0.4825367985281931, "learning_rate": 1.564973363696939e-05, "loss": 0.3457, "step": 11173 }, { "epoch": 5.2583529411764705, "grad_norm": 0.44086339295458776, "learning_rate": 1.5630244443296013e-05, "loss": 0.3086, "step": 11174 }, { "epoch": 5.258823529411765, "grad_norm": 0.5325138180232273, "learning_rate": 1.5610766725458834e-05, "loss": 0.3125, "step": 11175 }, { "epoch": 5.259294117647059, "grad_norm": 0.28469820567295473, "learning_rate": 1.5591300485121354e-05, "loss": 0.0835, "step": 11176 }, { "epoch": 5.259764705882353, "grad_norm": 0.5569388344814742, "learning_rate": 1.5571845723946113e-05, "loss": 0.2578, "step": 11177 }, { "epoch": 5.260235294117647, "grad_norm": 0.611388230959829, "learning_rate": 1.555240244359458e-05, "loss": 0.373, "step": 11178 }, { "epoch": 5.260705882352942, "grad_norm": 0.7285370152045846, "learning_rate": 1.5532970645727344e-05, "loss": 0.6328, "step": 11179 }, { "epoch": 5.261176470588235, "grad_norm": 0.40065927407220625, "learning_rate": 1.551355033200401e-05, "loss": 0.2422, "step": 11180 }, { "epoch": 5.261647058823529, "grad_norm": 0.4200788632404193, "learning_rate": 1.5494141504083087e-05, "loss": 0.2021, "step": 11181 }, { "epoch": 5.262117647058823, "grad_norm": 0.6791486972879347, "learning_rate": 1.547474416362226e-05, "loss": 0.2246, "step": 11182 }, { "epoch": 5.262588235294118, "grad_norm": 0.1548749443677098, "learning_rate": 1.5455358312278093e-05, "loss": 0.0403, "step": 11183 }, { "epoch": 5.263058823529412, "grad_norm": 0.35011515190500114, "learning_rate": 1.5435983951706276e-05, "loss": 0.1504, "step": 11184 }, { "epoch": 5.263529411764706, "grad_norm": 0.6035856923979652, "learning_rate": 1.5416621083561486e-05, "loss": 0.3027, "step": 11185 }, { "epoch": 5.264, "grad_norm": 0.45059579243316494, "learning_rate": 1.5397269709497352e-05, "loss": 0.2637, "step": 11186 }, { "epoch": 5.2644705882352945, "grad_norm": 0.12274828267006992, "learning_rate": 1.5377929831166597e-05, "loss": 0.0197, "step": 11187 }, { "epoch": 5.264941176470588, "grad_norm": 0.7961281363177565, "learning_rate": 1.5358601450220988e-05, "loss": 0.4746, "step": 11188 }, { "epoch": 5.265411764705882, "grad_norm": 0.3116452179286585, "learning_rate": 1.53392845683112e-05, "loss": 0.1582, "step": 11189 }, { "epoch": 5.265882352941176, "grad_norm": 0.29078020031511137, "learning_rate": 1.5319979187087018e-05, "loss": 0.0962, "step": 11190 }, { "epoch": 5.2663529411764705, "grad_norm": 0.299929197111846, "learning_rate": 1.5300685308197254e-05, "loss": 0.1328, "step": 11191 }, { "epoch": 5.266823529411765, "grad_norm": 0.573486275907177, "learning_rate": 1.528140293328962e-05, "loss": 0.2578, "step": 11192 }, { "epoch": 5.267294117647059, "grad_norm": 0.28061775613173856, "learning_rate": 1.5262132064010956e-05, "loss": 0.1367, "step": 11193 }, { "epoch": 5.267764705882353, "grad_norm": 0.6451851003645951, "learning_rate": 1.5242872702007142e-05, "loss": 0.3594, "step": 11194 }, { "epoch": 5.268235294117647, "grad_norm": 0.7660404102828571, "learning_rate": 1.522362484892296e-05, "loss": 0.2891, "step": 11195 }, { "epoch": 5.268705882352942, "grad_norm": 0.5203257276609585, "learning_rate": 1.5204388506402281e-05, "loss": 0.3223, "step": 11196 }, { "epoch": 5.269176470588235, "grad_norm": 0.18247902137503222, "learning_rate": 1.5185163676088007e-05, "loss": 0.0452, "step": 11197 }, { "epoch": 5.269647058823529, "grad_norm": 0.7464085012554965, "learning_rate": 1.5165950359622009e-05, "loss": 0.5039, "step": 11198 }, { "epoch": 5.270117647058823, "grad_norm": 0.658853735847459, "learning_rate": 1.5146748558645205e-05, "loss": 0.3633, "step": 11199 }, { "epoch": 5.270588235294118, "grad_norm": 0.4059729789437556, "learning_rate": 1.5127558274797535e-05, "loss": 0.1729, "step": 11200 }, { "epoch": 5.271058823529412, "grad_norm": 0.5265188444549131, "learning_rate": 1.5108379509717922e-05, "loss": 0.3613, "step": 11201 }, { "epoch": 5.271529411764706, "grad_norm": 0.3681818357695783, "learning_rate": 1.5089212265044336e-05, "loss": 0.1064, "step": 11202 }, { "epoch": 5.272, "grad_norm": 0.5199724539543976, "learning_rate": 1.5070056542413788e-05, "loss": 0.2695, "step": 11203 }, { "epoch": 5.2724705882352945, "grad_norm": 0.33931652623457076, "learning_rate": 1.5050912343462202e-05, "loss": 0.0898, "step": 11204 }, { "epoch": 5.272941176470589, "grad_norm": 0.7478836807898719, "learning_rate": 1.5031779669824668e-05, "loss": 0.4824, "step": 11205 }, { "epoch": 5.273411764705882, "grad_norm": 0.6028056099807863, "learning_rate": 1.5012658523135113e-05, "loss": 0.2734, "step": 11206 }, { "epoch": 5.273882352941176, "grad_norm": 0.564666753919917, "learning_rate": 1.4993548905026648e-05, "loss": 0.25, "step": 11207 }, { "epoch": 5.2743529411764705, "grad_norm": 0.26910780157063546, "learning_rate": 1.4974450817131334e-05, "loss": 0.1289, "step": 11208 }, { "epoch": 5.274823529411765, "grad_norm": 0.559721139567493, "learning_rate": 1.4955364261080183e-05, "loss": 0.2051, "step": 11209 }, { "epoch": 5.275294117647059, "grad_norm": 0.38720641954600643, "learning_rate": 1.4936289238503324e-05, "loss": 0.1289, "step": 11210 }, { "epoch": 5.275764705882353, "grad_norm": 0.42216784524935697, "learning_rate": 1.491722575102987e-05, "loss": 0.1699, "step": 11211 }, { "epoch": 5.276235294117647, "grad_norm": 0.4358579800114484, "learning_rate": 1.4898173800287888e-05, "loss": 0.1523, "step": 11212 }, { "epoch": 5.276705882352942, "grad_norm": 0.36201199338834994, "learning_rate": 1.4879133387904523e-05, "loss": 0.106, "step": 11213 }, { "epoch": 5.277176470588235, "grad_norm": 0.2763550809352479, "learning_rate": 1.4860104515505977e-05, "loss": 0.0806, "step": 11214 }, { "epoch": 5.277647058823529, "grad_norm": 0.4555513037170299, "learning_rate": 1.4841087184717315e-05, "loss": 0.1631, "step": 11215 }, { "epoch": 5.278117647058823, "grad_norm": 0.4356841286446586, "learning_rate": 1.4822081397162771e-05, "loss": 0.2314, "step": 11216 }, { "epoch": 5.278588235294118, "grad_norm": 0.6531491874010267, "learning_rate": 1.4803087154465548e-05, "loss": 0.418, "step": 11217 }, { "epoch": 5.279058823529412, "grad_norm": 0.35755373050302297, "learning_rate": 1.4784104458247764e-05, "loss": 0.1523, "step": 11218 }, { "epoch": 5.279529411764706, "grad_norm": 0.73603790667164, "learning_rate": 1.476513331013074e-05, "loss": 0.3867, "step": 11219 }, { "epoch": 5.28, "grad_norm": 0.5959747493527459, "learning_rate": 1.4746173711734627e-05, "loss": 0.3418, "step": 11220 }, { "epoch": 5.2804705882352945, "grad_norm": 0.37563577658471586, "learning_rate": 1.4727225664678699e-05, "loss": 0.1748, "step": 11221 }, { "epoch": 5.280941176470588, "grad_norm": 0.3568154316348021, "learning_rate": 1.4708289170581227e-05, "loss": 0.167, "step": 11222 }, { "epoch": 5.281411764705882, "grad_norm": 0.6766274034939146, "learning_rate": 1.4689364231059436e-05, "loss": 0.4766, "step": 11223 }, { "epoch": 5.281882352941176, "grad_norm": 0.29516584521517764, "learning_rate": 1.4670450847729648e-05, "loss": 0.1001, "step": 11224 }, { "epoch": 5.2823529411764705, "grad_norm": 0.4127297353895936, "learning_rate": 1.465154902220717e-05, "loss": 0.2266, "step": 11225 }, { "epoch": 5.282823529411765, "grad_norm": 0.7945788422986689, "learning_rate": 1.4632658756106247e-05, "loss": 0.3887, "step": 11226 }, { "epoch": 5.283294117647059, "grad_norm": 0.5692384468654703, "learning_rate": 1.4613780051040236e-05, "loss": 0.2539, "step": 11227 }, { "epoch": 5.283764705882353, "grad_norm": 0.5287746222249952, "learning_rate": 1.4594912908621515e-05, "loss": 0.3164, "step": 11228 }, { "epoch": 5.284235294117647, "grad_norm": 0.6355317085317308, "learning_rate": 1.4576057330461361e-05, "loss": 0.3477, "step": 11229 }, { "epoch": 5.284705882352942, "grad_norm": 0.4590665475258074, "learning_rate": 1.4557213318170152e-05, "loss": 0.1885, "step": 11230 }, { "epoch": 5.285176470588235, "grad_norm": 0.5014314321752873, "learning_rate": 1.453838087335732e-05, "loss": 0.2471, "step": 11231 }, { "epoch": 5.285647058823529, "grad_norm": 0.499467998453463, "learning_rate": 1.4519559997631142e-05, "loss": 0.3164, "step": 11232 }, { "epoch": 5.286117647058823, "grad_norm": 0.5235439069351012, "learning_rate": 1.45007506925991e-05, "loss": 0.3652, "step": 11233 }, { "epoch": 5.286588235294118, "grad_norm": 0.09000343013804744, "learning_rate": 1.448195295986756e-05, "loss": 0.0183, "step": 11234 }, { "epoch": 5.287058823529412, "grad_norm": 0.555001531563147, "learning_rate": 1.4463166801041936e-05, "loss": 0.2949, "step": 11235 }, { "epoch": 5.287529411764706, "grad_norm": 0.3981426773359987, "learning_rate": 1.4444392217726697e-05, "loss": 0.2344, "step": 11236 }, { "epoch": 5.288, "grad_norm": 0.4299718144565046, "learning_rate": 1.442562921152524e-05, "loss": 0.209, "step": 11237 }, { "epoch": 5.2884705882352945, "grad_norm": 0.3250144943634691, "learning_rate": 1.4406877784040021e-05, "loss": 0.126, "step": 11238 }, { "epoch": 5.288941176470589, "grad_norm": 0.40957436166615546, "learning_rate": 1.4388137936872556e-05, "loss": 0.1904, "step": 11239 }, { "epoch": 5.289411764705882, "grad_norm": 0.48062069313162115, "learning_rate": 1.4369409671623267e-05, "loss": 0.2451, "step": 11240 }, { "epoch": 5.289882352941176, "grad_norm": 0.5120447589585977, "learning_rate": 1.4350692989891638e-05, "loss": 0.2598, "step": 11241 }, { "epoch": 5.2903529411764705, "grad_norm": 0.39854412592745575, "learning_rate": 1.4331987893276225e-05, "loss": 0.2002, "step": 11242 }, { "epoch": 5.290823529411765, "grad_norm": 0.5288923919350682, "learning_rate": 1.4313294383374468e-05, "loss": 0.2617, "step": 11243 }, { "epoch": 5.291294117647059, "grad_norm": 0.4744578140706647, "learning_rate": 1.429461246178289e-05, "loss": 0.2715, "step": 11244 }, { "epoch": 5.291764705882353, "grad_norm": 0.34564286262886357, "learning_rate": 1.4275942130097096e-05, "loss": 0.126, "step": 11245 }, { "epoch": 5.292235294117647, "grad_norm": 0.2559799410546592, "learning_rate": 1.4257283389911512e-05, "loss": 0.0569, "step": 11246 }, { "epoch": 5.292705882352941, "grad_norm": 0.49604531869562596, "learning_rate": 1.4238636242819761e-05, "loss": 0.2852, "step": 11247 }, { "epoch": 5.293176470588235, "grad_norm": 0.6603642554675923, "learning_rate": 1.422000069041439e-05, "loss": 0.3848, "step": 11248 }, { "epoch": 5.293647058823529, "grad_norm": 0.3731531934144553, "learning_rate": 1.420137673428694e-05, "loss": 0.167, "step": 11249 }, { "epoch": 5.294117647058823, "grad_norm": 0.7327927080716935, "learning_rate": 1.4182764376028006e-05, "loss": 0.3477, "step": 11250 }, { "epoch": 5.294588235294118, "grad_norm": 0.5679410929366653, "learning_rate": 1.4164163617227197e-05, "loss": 0.2305, "step": 11251 }, { "epoch": 5.295058823529412, "grad_norm": 0.3980147708671807, "learning_rate": 1.4145574459473064e-05, "loss": 0.1592, "step": 11252 }, { "epoch": 5.295529411764706, "grad_norm": 0.2913056323236744, "learning_rate": 1.4126996904353233e-05, "loss": 0.0879, "step": 11253 }, { "epoch": 5.296, "grad_norm": 0.4361558610012703, "learning_rate": 1.4108430953454352e-05, "loss": 0.2256, "step": 11254 }, { "epoch": 5.2964705882352945, "grad_norm": 0.762852563774783, "learning_rate": 1.4089876608361971e-05, "loss": 0.5234, "step": 11255 }, { "epoch": 5.296941176470588, "grad_norm": 0.3613950152714777, "learning_rate": 1.4071333870660807e-05, "loss": 0.0918, "step": 11256 }, { "epoch": 5.297411764705882, "grad_norm": 0.2875694171970186, "learning_rate": 1.4052802741934404e-05, "loss": 0.1074, "step": 11257 }, { "epoch": 5.297882352941176, "grad_norm": 0.8603029528582248, "learning_rate": 1.4034283223765485e-05, "loss": 0.6133, "step": 11258 }, { "epoch": 5.2983529411764705, "grad_norm": 0.23285722480901988, "learning_rate": 1.4015775317735717e-05, "loss": 0.0488, "step": 11259 }, { "epoch": 5.298823529411765, "grad_norm": 0.5935698548687357, "learning_rate": 1.3997279025425701e-05, "loss": 0.3066, "step": 11260 }, { "epoch": 5.299294117647059, "grad_norm": 0.5483744723519631, "learning_rate": 1.397879434841514e-05, "loss": 0.3906, "step": 11261 }, { "epoch": 5.299764705882353, "grad_norm": 0.479659873912032, "learning_rate": 1.3960321288282722e-05, "loss": 0.3027, "step": 11262 }, { "epoch": 5.300235294117647, "grad_norm": 0.480000011095285, "learning_rate": 1.3941859846606152e-05, "loss": 0.2344, "step": 11263 }, { "epoch": 5.300705882352942, "grad_norm": 0.44773192705199394, "learning_rate": 1.3923410024962117e-05, "loss": 0.1934, "step": 11264 }, { "epoch": 5.301176470588235, "grad_norm": 0.3973279928212218, "learning_rate": 1.390497182492631e-05, "loss": 0.1963, "step": 11265 }, { "epoch": 5.301647058823529, "grad_norm": 0.7561199479656756, "learning_rate": 1.388654524807345e-05, "loss": 0.4043, "step": 11266 }, { "epoch": 5.302117647058823, "grad_norm": 0.4242725283185892, "learning_rate": 1.3868130295977282e-05, "loss": 0.1719, "step": 11267 }, { "epoch": 5.302588235294118, "grad_norm": 0.6508975761352322, "learning_rate": 1.3849726970210479e-05, "loss": 0.252, "step": 11268 }, { "epoch": 5.303058823529412, "grad_norm": 0.5635890844682643, "learning_rate": 1.3831335272344801e-05, "loss": 0.3086, "step": 11269 }, { "epoch": 5.303529411764706, "grad_norm": 0.467509942415674, "learning_rate": 1.3812955203951026e-05, "loss": 0.2695, "step": 11270 }, { "epoch": 5.304, "grad_norm": 0.30584725861206946, "learning_rate": 1.3794586766598848e-05, "loss": 0.0618, "step": 11271 }, { "epoch": 5.3044705882352945, "grad_norm": 0.36282142530417894, "learning_rate": 1.3776229961857043e-05, "loss": 0.1484, "step": 11272 }, { "epoch": 5.304941176470588, "grad_norm": 0.26459751298896933, "learning_rate": 1.3757884791293394e-05, "loss": 0.0408, "step": 11273 }, { "epoch": 5.305411764705882, "grad_norm": 0.514103609512788, "learning_rate": 1.373955125647463e-05, "loss": 0.3379, "step": 11274 }, { "epoch": 5.305882352941176, "grad_norm": 0.502653594845265, "learning_rate": 1.3721229358966545e-05, "loss": 0.2559, "step": 11275 }, { "epoch": 5.3063529411764705, "grad_norm": 0.34749867222673636, "learning_rate": 1.3702919100333925e-05, "loss": 0.1855, "step": 11276 }, { "epoch": 5.306823529411765, "grad_norm": 0.4906385029641534, "learning_rate": 1.3684620482140534e-05, "loss": 0.2852, "step": 11277 }, { "epoch": 5.307294117647059, "grad_norm": 0.40134546373210017, "learning_rate": 1.3666333505949173e-05, "loss": 0.2275, "step": 11278 }, { "epoch": 5.307764705882353, "grad_norm": 0.5421614598051815, "learning_rate": 1.3648058173321675e-05, "loss": 0.2334, "step": 11279 }, { "epoch": 5.308235294117647, "grad_norm": 0.6192460002738382, "learning_rate": 1.3629794485818791e-05, "loss": 0.3691, "step": 11280 }, { "epoch": 5.308705882352941, "grad_norm": 0.6308771198580538, "learning_rate": 1.361154244500034e-05, "loss": 0.3574, "step": 11281 }, { "epoch": 5.309176470588235, "grad_norm": 0.40837873276337927, "learning_rate": 1.3593302052425159e-05, "loss": 0.1475, "step": 11282 }, { "epoch": 5.309647058823529, "grad_norm": 1.119579324831274, "learning_rate": 1.3575073309651052e-05, "loss": 0.625, "step": 11283 }, { "epoch": 5.310117647058823, "grad_norm": 0.5283596846023982, "learning_rate": 1.3556856218234858e-05, "loss": 0.2598, "step": 11284 }, { "epoch": 5.310588235294118, "grad_norm": 0.8460327511177088, "learning_rate": 1.3538650779732362e-05, "loss": 0.2354, "step": 11285 }, { "epoch": 5.311058823529412, "grad_norm": 0.25857353798500293, "learning_rate": 1.3520456995698443e-05, "loss": 0.0879, "step": 11286 }, { "epoch": 5.311529411764706, "grad_norm": 0.8719700566702707, "learning_rate": 1.3502274867686936e-05, "loss": 0.5156, "step": 11287 }, { "epoch": 5.312, "grad_norm": 0.6213620190212411, "learning_rate": 1.348410439725065e-05, "loss": 0.3691, "step": 11288 }, { "epoch": 5.3124705882352945, "grad_norm": 0.49745840296280425, "learning_rate": 1.3465945585941462e-05, "loss": 0.2305, "step": 11289 }, { "epoch": 5.312941176470588, "grad_norm": 0.35008805869176707, "learning_rate": 1.3447798435310247e-05, "loss": 0.0942, "step": 11290 }, { "epoch": 5.313411764705882, "grad_norm": 0.41337079841857216, "learning_rate": 1.3429662946906799e-05, "loss": 0.1475, "step": 11291 }, { "epoch": 5.313882352941176, "grad_norm": 0.5945434832381588, "learning_rate": 1.3411539122279996e-05, "loss": 0.3594, "step": 11292 }, { "epoch": 5.3143529411764705, "grad_norm": 0.2314884497621144, "learning_rate": 1.3393426962977765e-05, "loss": 0.0664, "step": 11293 }, { "epoch": 5.314823529411765, "grad_norm": 0.4860536069283776, "learning_rate": 1.3375326470546872e-05, "loss": 0.291, "step": 11294 }, { "epoch": 5.315294117647059, "grad_norm": 0.3138015424907396, "learning_rate": 1.3357237646533258e-05, "loss": 0.0874, "step": 11295 }, { "epoch": 5.315764705882353, "grad_norm": 0.648444030388054, "learning_rate": 1.3339160492481793e-05, "loss": 0.3496, "step": 11296 }, { "epoch": 5.316235294117647, "grad_norm": 0.5691006038319308, "learning_rate": 1.3321095009936321e-05, "loss": 0.3555, "step": 11297 }, { "epoch": 5.316705882352942, "grad_norm": 0.5125944971076678, "learning_rate": 1.3303041200439729e-05, "loss": 0.1572, "step": 11298 }, { "epoch": 5.317176470588235, "grad_norm": 0.49131354154365736, "learning_rate": 1.3284999065533946e-05, "loss": 0.1875, "step": 11299 }, { "epoch": 5.317647058823529, "grad_norm": 0.470962177464538, "learning_rate": 1.326696860675981e-05, "loss": 0.1328, "step": 11300 }, { "epoch": 5.318117647058823, "grad_norm": 0.41873897609619165, "learning_rate": 1.3248949825657217e-05, "loss": 0.2168, "step": 11301 }, { "epoch": 5.318588235294118, "grad_norm": 0.8261463386162377, "learning_rate": 1.3230942723765092e-05, "loss": 0.4785, "step": 11302 }, { "epoch": 5.319058823529412, "grad_norm": 0.6110290960197432, "learning_rate": 1.3212947302621284e-05, "loss": 0.3184, "step": 11303 }, { "epoch": 5.319529411764706, "grad_norm": 0.25451705967302024, "learning_rate": 1.3194963563762718e-05, "loss": 0.0559, "step": 11304 }, { "epoch": 5.32, "grad_norm": 0.4890349799534655, "learning_rate": 1.317699150872531e-05, "loss": 0.1826, "step": 11305 }, { "epoch": 5.3204705882352945, "grad_norm": 0.47933135428921486, "learning_rate": 1.3159031139043918e-05, "loss": 0.2373, "step": 11306 }, { "epoch": 5.320941176470588, "grad_norm": 0.3562085971584924, "learning_rate": 1.3141082456252467e-05, "loss": 0.1396, "step": 11307 }, { "epoch": 5.321411764705882, "grad_norm": 0.7479969023503864, "learning_rate": 1.312314546188386e-05, "loss": 0.2471, "step": 11308 }, { "epoch": 5.321882352941176, "grad_norm": 0.465104327319147, "learning_rate": 1.3105220157470009e-05, "loss": 0.2871, "step": 11309 }, { "epoch": 5.3223529411764705, "grad_norm": 0.8632743063049645, "learning_rate": 1.3087306544541836e-05, "loss": 0.5234, "step": 11310 }, { "epoch": 5.322823529411765, "grad_norm": 0.396014043864393, "learning_rate": 1.306940462462922e-05, "loss": 0.208, "step": 11311 }, { "epoch": 5.323294117647059, "grad_norm": 0.4983909549174012, "learning_rate": 1.3051514399261104e-05, "loss": 0.209, "step": 11312 }, { "epoch": 5.323764705882353, "grad_norm": 0.40249285252125516, "learning_rate": 1.3033635869965398e-05, "loss": 0.1328, "step": 11313 }, { "epoch": 5.324235294117647, "grad_norm": 0.567247891049821, "learning_rate": 1.3015769038268981e-05, "loss": 0.3086, "step": 11314 }, { "epoch": 5.324705882352941, "grad_norm": 0.5508285352956063, "learning_rate": 1.2997913905697816e-05, "loss": 0.3613, "step": 11315 }, { "epoch": 5.325176470588235, "grad_norm": 0.38172809118367795, "learning_rate": 1.2980070473776782e-05, "loss": 0.1157, "step": 11316 }, { "epoch": 5.325647058823529, "grad_norm": 0.6758389692472963, "learning_rate": 1.2962238744029795e-05, "loss": 0.3672, "step": 11317 }, { "epoch": 5.326117647058823, "grad_norm": 0.35419454871233846, "learning_rate": 1.2944418717979804e-05, "loss": 0.1099, "step": 11318 }, { "epoch": 5.326588235294118, "grad_norm": 0.5246726551918972, "learning_rate": 1.2926610397148707e-05, "loss": 0.249, "step": 11319 }, { "epoch": 5.327058823529412, "grad_norm": 0.6976374002473406, "learning_rate": 1.2908813783057404e-05, "loss": 0.5664, "step": 11320 }, { "epoch": 5.327529411764706, "grad_norm": 0.4796438403517786, "learning_rate": 1.2891028877225863e-05, "loss": 0.126, "step": 11321 }, { "epoch": 5.328, "grad_norm": 0.15592094165626286, "learning_rate": 1.287325568117295e-05, "loss": 0.0283, "step": 11322 }, { "epoch": 5.3284705882352945, "grad_norm": 0.6930649989144901, "learning_rate": 1.2855494196416604e-05, "loss": 0.4219, "step": 11323 }, { "epoch": 5.328941176470588, "grad_norm": 0.5421532925219494, "learning_rate": 1.2837744424473761e-05, "loss": 0.3477, "step": 11324 }, { "epoch": 5.329411764705882, "grad_norm": 0.42252377062610225, "learning_rate": 1.2820006366860292e-05, "loss": 0.1631, "step": 11325 }, { "epoch": 5.329882352941176, "grad_norm": 0.19450879731309115, "learning_rate": 1.2802280025091165e-05, "loss": 0.0444, "step": 11326 }, { "epoch": 5.3303529411764705, "grad_norm": 0.5039109846036159, "learning_rate": 1.2784565400680291e-05, "loss": 0.2236, "step": 11327 }, { "epoch": 5.330823529411765, "grad_norm": 0.7556114565582419, "learning_rate": 1.276686249514054e-05, "loss": 0.4004, "step": 11328 }, { "epoch": 5.331294117647059, "grad_norm": 0.5203870315559556, "learning_rate": 1.2749171309983852e-05, "loss": 0.2246, "step": 11329 }, { "epoch": 5.331764705882353, "grad_norm": 0.5414840326805576, "learning_rate": 1.273149184672117e-05, "loss": 0.2656, "step": 11330 }, { "epoch": 5.332235294117647, "grad_norm": 0.37822871089163224, "learning_rate": 1.2713824106862369e-05, "loss": 0.1328, "step": 11331 }, { "epoch": 5.332705882352942, "grad_norm": 0.6191339603352127, "learning_rate": 1.2696168091916376e-05, "loss": 0.2275, "step": 11332 }, { "epoch": 5.333176470588235, "grad_norm": 0.5051396754773269, "learning_rate": 1.2678523803391116e-05, "loss": 0.2285, "step": 11333 }, { "epoch": 5.333647058823529, "grad_norm": 0.7456832256773718, "learning_rate": 1.2660891242793453e-05, "loss": 0.457, "step": 11334 }, { "epoch": 5.334117647058823, "grad_norm": 0.3521563076059768, "learning_rate": 1.2643270411629346e-05, "loss": 0.1104, "step": 11335 }, { "epoch": 5.334588235294118, "grad_norm": 0.88228717618106, "learning_rate": 1.2625661311403662e-05, "loss": 0.3809, "step": 11336 }, { "epoch": 5.335058823529412, "grad_norm": 0.432755893743632, "learning_rate": 1.2608063943620312e-05, "loss": 0.21, "step": 11337 }, { "epoch": 5.335529411764706, "grad_norm": 0.5889447040219732, "learning_rate": 1.2590478309782243e-05, "loss": 0.3594, "step": 11338 }, { "epoch": 5.336, "grad_norm": 0.7237524024681289, "learning_rate": 1.2572904411391272e-05, "loss": 0.3652, "step": 11339 }, { "epoch": 5.3364705882352945, "grad_norm": 0.8403301049800656, "learning_rate": 1.2555342249948347e-05, "loss": 0.4414, "step": 11340 }, { "epoch": 5.336941176470588, "grad_norm": 0.27969311038727357, "learning_rate": 1.2537791826953387e-05, "loss": 0.0996, "step": 11341 }, { "epoch": 5.337411764705882, "grad_norm": 0.2346451802975148, "learning_rate": 1.2520253143905223e-05, "loss": 0.0505, "step": 11342 }, { "epoch": 5.337882352941176, "grad_norm": 0.0943039317818322, "learning_rate": 1.2502726202301777e-05, "loss": 0.0161, "step": 11343 }, { "epoch": 5.3383529411764705, "grad_norm": 0.47319638890352644, "learning_rate": 1.2485211003639967e-05, "loss": 0.168, "step": 11344 }, { "epoch": 5.338823529411765, "grad_norm": 0.6074016012328389, "learning_rate": 1.2467707549415612e-05, "loss": 0.3535, "step": 11345 }, { "epoch": 5.339294117647059, "grad_norm": 0.4202729036999127, "learning_rate": 1.2450215841123618e-05, "loss": 0.2334, "step": 11346 }, { "epoch": 5.339764705882353, "grad_norm": 0.6249484408645658, "learning_rate": 1.2432735880257904e-05, "loss": 0.2754, "step": 11347 }, { "epoch": 5.340235294117647, "grad_norm": 0.42190603303816193, "learning_rate": 1.2415267668311263e-05, "loss": 0.1426, "step": 11348 }, { "epoch": 5.340705882352941, "grad_norm": 0.49376953415789615, "learning_rate": 1.2397811206775632e-05, "loss": 0.1777, "step": 11349 }, { "epoch": 5.341176470588235, "grad_norm": 0.3455046595401191, "learning_rate": 1.2380366497141886e-05, "loss": 0.1387, "step": 11350 }, { "epoch": 5.341647058823529, "grad_norm": 0.38088328521019904, "learning_rate": 1.23629335408998e-05, "loss": 0.2402, "step": 11351 }, { "epoch": 5.342117647058823, "grad_norm": 0.9091513943525433, "learning_rate": 1.234551233953835e-05, "loss": 0.4922, "step": 11352 }, { "epoch": 5.342588235294118, "grad_norm": 0.5145803225127398, "learning_rate": 1.232810289454531e-05, "loss": 0.2695, "step": 11353 }, { "epoch": 5.343058823529412, "grad_norm": 0.41583954071162377, "learning_rate": 1.2310705207407561e-05, "loss": 0.2109, "step": 11354 }, { "epoch": 5.343529411764706, "grad_norm": 0.2524256503552606, "learning_rate": 1.229331927961098e-05, "loss": 0.0496, "step": 11355 }, { "epoch": 5.344, "grad_norm": 0.5540594740167696, "learning_rate": 1.2275945112640363e-05, "loss": 0.2695, "step": 11356 }, { "epoch": 5.3444705882352945, "grad_norm": 0.46250511776907005, "learning_rate": 1.2258582707979554e-05, "loss": 0.3008, "step": 11357 }, { "epoch": 5.344941176470588, "grad_norm": 0.4443407500256309, "learning_rate": 1.2241232067111418e-05, "loss": 0.1836, "step": 11358 }, { "epoch": 5.345411764705882, "grad_norm": 0.49001388217396263, "learning_rate": 1.2223893191517753e-05, "loss": 0.2539, "step": 11359 }, { "epoch": 5.345882352941176, "grad_norm": 0.3208027900092499, "learning_rate": 1.2206566082679392e-05, "loss": 0.0903, "step": 11360 }, { "epoch": 5.3463529411764705, "grad_norm": 0.578371292031877, "learning_rate": 1.2189250742076184e-05, "loss": 0.2832, "step": 11361 }, { "epoch": 5.346823529411765, "grad_norm": 0.5033698348448362, "learning_rate": 1.2171947171186897e-05, "loss": 0.2168, "step": 11362 }, { "epoch": 5.347294117647059, "grad_norm": 0.3058835968641122, "learning_rate": 1.2154655371489397e-05, "loss": 0.0603, "step": 11363 }, { "epoch": 5.347764705882353, "grad_norm": 0.46165146225233156, "learning_rate": 1.213737534446042e-05, "loss": 0.1758, "step": 11364 }, { "epoch": 5.348235294117647, "grad_norm": 0.7835673195752949, "learning_rate": 1.2120107091575803e-05, "loss": 0.1426, "step": 11365 }, { "epoch": 5.348705882352942, "grad_norm": 0.3061323154374222, "learning_rate": 1.2102850614310366e-05, "loss": 0.1299, "step": 11366 }, { "epoch": 5.349176470588235, "grad_norm": 0.4028469384845583, "learning_rate": 1.2085605914137847e-05, "loss": 0.2393, "step": 11367 }, { "epoch": 5.349647058823529, "grad_norm": 0.12331685524761424, "learning_rate": 1.206837299253105e-05, "loss": 0.0415, "step": 11368 }, { "epoch": 5.350117647058823, "grad_norm": 0.41392452867324964, "learning_rate": 1.2051151850961766e-05, "loss": 0.2637, "step": 11369 }, { "epoch": 5.350588235294118, "grad_norm": 0.572886420069135, "learning_rate": 1.2033942490900716e-05, "loss": 0.3027, "step": 11370 }, { "epoch": 5.351058823529412, "grad_norm": 0.3234743412163235, "learning_rate": 1.201674491381771e-05, "loss": 0.1348, "step": 11371 }, { "epoch": 5.351529411764706, "grad_norm": 0.22999667348309075, "learning_rate": 1.1999559121181523e-05, "loss": 0.0806, "step": 11372 }, { "epoch": 5.352, "grad_norm": 0.6543817665010515, "learning_rate": 1.1982385114459846e-05, "loss": 0.5586, "step": 11373 }, { "epoch": 5.352470588235294, "grad_norm": 0.5130780919398938, "learning_rate": 1.1965222895119442e-05, "loss": 0.4219, "step": 11374 }, { "epoch": 5.352941176470588, "grad_norm": 0.5782890439431997, "learning_rate": 1.1948072464626101e-05, "loss": 0.2432, "step": 11375 }, { "epoch": 5.353411764705882, "grad_norm": 0.6842211468775649, "learning_rate": 1.1930933824444472e-05, "loss": 0.3438, "step": 11376 }, { "epoch": 5.353882352941176, "grad_norm": 0.6879997369146246, "learning_rate": 1.1913806976038331e-05, "loss": 0.3613, "step": 11377 }, { "epoch": 5.3543529411764705, "grad_norm": 0.6396172474404727, "learning_rate": 1.1896691920870394e-05, "loss": 0.3047, "step": 11378 }, { "epoch": 5.354823529411765, "grad_norm": 0.5964091582055079, "learning_rate": 1.1879588660402356e-05, "loss": 0.2734, "step": 11379 }, { "epoch": 5.355294117647059, "grad_norm": 0.485299471244616, "learning_rate": 1.18624971960949e-05, "loss": 0.3105, "step": 11380 }, { "epoch": 5.355764705882353, "grad_norm": 0.3198049983871363, "learning_rate": 1.1845417529407791e-05, "loss": 0.0569, "step": 11381 }, { "epoch": 5.356235294117647, "grad_norm": 0.744317144155818, "learning_rate": 1.1828349661799629e-05, "loss": 0.4531, "step": 11382 }, { "epoch": 5.356705882352941, "grad_norm": 0.591261428772992, "learning_rate": 1.1811293594728133e-05, "loss": 0.2949, "step": 11383 }, { "epoch": 5.357176470588235, "grad_norm": 0.45879593948369635, "learning_rate": 1.179424932965002e-05, "loss": 0.2207, "step": 11384 }, { "epoch": 5.357647058823529, "grad_norm": 0.36032766908157426, "learning_rate": 1.1777216868020872e-05, "loss": 0.1118, "step": 11385 }, { "epoch": 5.358117647058823, "grad_norm": 0.6789150685865688, "learning_rate": 1.1760196211295398e-05, "loss": 0.4492, "step": 11386 }, { "epoch": 5.358588235294118, "grad_norm": 0.4303365077550088, "learning_rate": 1.1743187360927214e-05, "loss": 0.1611, "step": 11387 }, { "epoch": 5.359058823529412, "grad_norm": 0.20797350065731807, "learning_rate": 1.1726190318368977e-05, "loss": 0.0811, "step": 11388 }, { "epoch": 5.359529411764706, "grad_norm": 0.43628475249037, "learning_rate": 1.1709205085072344e-05, "loss": 0.1455, "step": 11389 }, { "epoch": 5.36, "grad_norm": 0.09081669230722852, "learning_rate": 1.1692231662487866e-05, "loss": 0.015, "step": 11390 }, { "epoch": 5.3604705882352945, "grad_norm": 0.4733574168781901, "learning_rate": 1.1675270052065222e-05, "loss": 0.2119, "step": 11391 }, { "epoch": 5.360941176470588, "grad_norm": 0.5289431516086815, "learning_rate": 1.1658320255253e-05, "loss": 0.2832, "step": 11392 }, { "epoch": 5.361411764705882, "grad_norm": 0.6509847495407557, "learning_rate": 1.1641382273498777e-05, "loss": 0.3457, "step": 11393 }, { "epoch": 5.361882352941176, "grad_norm": 0.4628760973304793, "learning_rate": 1.1624456108249164e-05, "loss": 0.2637, "step": 11394 }, { "epoch": 5.3623529411764705, "grad_norm": 0.43916939301781804, "learning_rate": 1.1607541760949735e-05, "loss": 0.125, "step": 11395 }, { "epoch": 5.362823529411765, "grad_norm": 0.4693641121753045, "learning_rate": 1.1590639233045002e-05, "loss": 0.1699, "step": 11396 }, { "epoch": 5.363294117647059, "grad_norm": 0.6870537091794436, "learning_rate": 1.1573748525978626e-05, "loss": 0.4297, "step": 11397 }, { "epoch": 5.363764705882353, "grad_norm": 0.441569092000621, "learning_rate": 1.155686964119309e-05, "loss": 0.252, "step": 11398 }, { "epoch": 5.364235294117647, "grad_norm": 0.43222171400935055, "learning_rate": 1.154000258012992e-05, "loss": 0.1904, "step": 11399 }, { "epoch": 5.364705882352941, "grad_norm": 0.2053028167743946, "learning_rate": 1.1523147344229716e-05, "loss": 0.064, "step": 11400 }, { "epoch": 5.365176470588235, "grad_norm": 0.32143731382290996, "learning_rate": 1.1506303934931926e-05, "loss": 0.127, "step": 11401 }, { "epoch": 5.365647058823529, "grad_norm": 0.5478042180843401, "learning_rate": 1.1489472353675083e-05, "loss": 0.2754, "step": 11402 }, { "epoch": 5.366117647058823, "grad_norm": 0.7103460004176684, "learning_rate": 1.147265260189672e-05, "loss": 0.4453, "step": 11403 }, { "epoch": 5.366588235294118, "grad_norm": 0.3154819168239938, "learning_rate": 1.1455844681033272e-05, "loss": 0.0781, "step": 11404 }, { "epoch": 5.367058823529412, "grad_norm": 0.24668649294847603, "learning_rate": 1.1439048592520256e-05, "loss": 0.0884, "step": 11405 }, { "epoch": 5.367529411764706, "grad_norm": 0.445177808860924, "learning_rate": 1.1422264337792125e-05, "loss": 0.25, "step": 11406 }, { "epoch": 5.368, "grad_norm": 0.6920904154843194, "learning_rate": 1.1405491918282333e-05, "loss": 0.3711, "step": 11407 }, { "epoch": 5.368470588235294, "grad_norm": 0.3024613283069956, "learning_rate": 1.1388731335423334e-05, "loss": 0.104, "step": 11408 }, { "epoch": 5.368941176470588, "grad_norm": 0.42623258907146516, "learning_rate": 1.1371982590646584e-05, "loss": 0.2852, "step": 11409 }, { "epoch": 5.369411764705882, "grad_norm": 0.3776773011598995, "learning_rate": 1.135524568538247e-05, "loss": 0.1748, "step": 11410 }, { "epoch": 5.369882352941176, "grad_norm": 0.46356040083271854, "learning_rate": 1.1338520621060415e-05, "loss": 0.2246, "step": 11411 }, { "epoch": 5.3703529411764706, "grad_norm": 0.49923373981893965, "learning_rate": 1.1321807399108845e-05, "loss": 0.1982, "step": 11412 }, { "epoch": 5.370823529411765, "grad_norm": 0.6515399940347311, "learning_rate": 1.1305106020955135e-05, "loss": 0.6562, "step": 11413 }, { "epoch": 5.371294117647059, "grad_norm": 0.10121239997938003, "learning_rate": 1.1288416488025675e-05, "loss": 0.0139, "step": 11414 }, { "epoch": 5.371764705882353, "grad_norm": 0.22380871556514656, "learning_rate": 1.1271738801745795e-05, "loss": 0.0728, "step": 11415 }, { "epoch": 5.372235294117647, "grad_norm": 0.25773685268637186, "learning_rate": 1.1255072963539885e-05, "loss": 0.0684, "step": 11416 }, { "epoch": 5.372705882352941, "grad_norm": 0.5498769686969299, "learning_rate": 1.123841897483131e-05, "loss": 0.2988, "step": 11417 }, { "epoch": 5.373176470588235, "grad_norm": 0.5101023710289482, "learning_rate": 1.1221776837042345e-05, "loss": 0.2363, "step": 11418 }, { "epoch": 5.373647058823529, "grad_norm": 0.362195745039149, "learning_rate": 1.1205146551594357e-05, "loss": 0.1245, "step": 11419 }, { "epoch": 5.374117647058823, "grad_norm": 0.3242442016534588, "learning_rate": 1.1188528119907642e-05, "loss": 0.1279, "step": 11420 }, { "epoch": 5.374588235294118, "grad_norm": 0.5371043239726324, "learning_rate": 1.117192154340148e-05, "loss": 0.2852, "step": 11421 }, { "epoch": 5.375058823529412, "grad_norm": 0.20717941422589706, "learning_rate": 1.115532682349417e-05, "loss": 0.0581, "step": 11422 }, { "epoch": 5.375529411764706, "grad_norm": 0.41163127825407336, "learning_rate": 1.1138743961602998e-05, "loss": 0.1914, "step": 11423 }, { "epoch": 5.376, "grad_norm": 0.6663163182044933, "learning_rate": 1.1122172959144177e-05, "loss": 0.332, "step": 11424 }, { "epoch": 5.376470588235295, "grad_norm": 0.4547215961222794, "learning_rate": 1.1105613817532976e-05, "loss": 0.2852, "step": 11425 }, { "epoch": 5.376941176470588, "grad_norm": 0.6255595967695149, "learning_rate": 1.108906653818365e-05, "loss": 0.4883, "step": 11426 }, { "epoch": 5.377411764705882, "grad_norm": 0.6563853162915735, "learning_rate": 1.1072531122509381e-05, "loss": 0.4004, "step": 11427 }, { "epoch": 5.377882352941176, "grad_norm": 0.5275096355805623, "learning_rate": 1.1056007571922377e-05, "loss": 0.249, "step": 11428 }, { "epoch": 5.378352941176471, "grad_norm": 0.5357121410567568, "learning_rate": 1.1039495887833872e-05, "loss": 0.3906, "step": 11429 }, { "epoch": 5.378823529411765, "grad_norm": 0.44068164870975585, "learning_rate": 1.1022996071653988e-05, "loss": 0.1992, "step": 11430 }, { "epoch": 5.379294117647059, "grad_norm": 0.33046818828247665, "learning_rate": 1.1006508124791935e-05, "loss": 0.1108, "step": 11431 }, { "epoch": 5.379764705882353, "grad_norm": 0.3631793209993492, "learning_rate": 1.099003204865585e-05, "loss": 0.1455, "step": 11432 }, { "epoch": 5.3802352941176474, "grad_norm": 0.5272136609093274, "learning_rate": 1.0973567844652858e-05, "loss": 0.209, "step": 11433 }, { "epoch": 5.380705882352941, "grad_norm": 0.5241561865730422, "learning_rate": 1.0957115514189086e-05, "loss": 0.2832, "step": 11434 }, { "epoch": 5.381176470588235, "grad_norm": 0.6496356420350528, "learning_rate": 1.0940675058669673e-05, "loss": 0.3652, "step": 11435 }, { "epoch": 5.381647058823529, "grad_norm": 0.6417680191912346, "learning_rate": 1.0924246479498682e-05, "loss": 0.2559, "step": 11436 }, { "epoch": 5.3821176470588235, "grad_norm": 0.742523976701559, "learning_rate": 1.0907829778079207e-05, "loss": 0.5234, "step": 11437 }, { "epoch": 5.382588235294118, "grad_norm": 0.492327594747315, "learning_rate": 1.089142495581331e-05, "loss": 0.2793, "step": 11438 }, { "epoch": 5.383058823529412, "grad_norm": 0.39161564729513004, "learning_rate": 1.0875032014102036e-05, "loss": 0.1387, "step": 11439 }, { "epoch": 5.383529411764706, "grad_norm": 0.6115723567984207, "learning_rate": 1.0858650954345433e-05, "loss": 0.2812, "step": 11440 }, { "epoch": 5.384, "grad_norm": 0.4538569322353663, "learning_rate": 1.0842281777942514e-05, "loss": 0.2178, "step": 11441 }, { "epoch": 5.384470588235294, "grad_norm": 0.7978681559762395, "learning_rate": 1.082592448629131e-05, "loss": 0.3789, "step": 11442 }, { "epoch": 5.384941176470588, "grad_norm": 0.23711086947736, "learning_rate": 1.0809579080788805e-05, "loss": 0.0762, "step": 11443 }, { "epoch": 5.385411764705882, "grad_norm": 0.3401789184326865, "learning_rate": 1.0793245562830966e-05, "loss": 0.1021, "step": 11444 }, { "epoch": 5.385882352941176, "grad_norm": 0.4630773925571603, "learning_rate": 1.0776923933812775e-05, "loss": 0.1904, "step": 11445 }, { "epoch": 5.386352941176471, "grad_norm": 0.4442290684519348, "learning_rate": 1.0760614195128153e-05, "loss": 0.2676, "step": 11446 }, { "epoch": 5.386823529411765, "grad_norm": 0.34307710460297, "learning_rate": 1.0744316348170035e-05, "loss": 0.1187, "step": 11447 }, { "epoch": 5.387294117647059, "grad_norm": 0.46117559135472175, "learning_rate": 1.0728030394330373e-05, "loss": 0.1738, "step": 11448 }, { "epoch": 5.387764705882353, "grad_norm": 0.5925993571913817, "learning_rate": 1.0711756335000005e-05, "loss": 0.3984, "step": 11449 }, { "epoch": 5.3882352941176475, "grad_norm": 0.5436702628260486, "learning_rate": 1.069549417156887e-05, "loss": 0.3457, "step": 11450 }, { "epoch": 5.388705882352941, "grad_norm": 0.6475638098304243, "learning_rate": 1.0679243905425822e-05, "loss": 0.3086, "step": 11451 }, { "epoch": 5.389176470588235, "grad_norm": 0.3012269073441476, "learning_rate": 1.0663005537958702e-05, "loss": 0.1787, "step": 11452 }, { "epoch": 5.389647058823529, "grad_norm": 0.9971654792180324, "learning_rate": 1.0646779070554335e-05, "loss": 0.7539, "step": 11453 }, { "epoch": 5.3901176470588235, "grad_norm": 0.2216039204975713, "learning_rate": 1.0630564504598593e-05, "loss": 0.0258, "step": 11454 }, { "epoch": 5.390588235294118, "grad_norm": 0.3824239496731766, "learning_rate": 1.0614361841476238e-05, "loss": 0.1309, "step": 11455 }, { "epoch": 5.391058823529412, "grad_norm": 0.7170336413637938, "learning_rate": 1.0598171082571044e-05, "loss": 0.4668, "step": 11456 }, { "epoch": 5.391529411764706, "grad_norm": 0.5770476144933397, "learning_rate": 1.0581992229265823e-05, "loss": 0.2275, "step": 11457 }, { "epoch": 5.392, "grad_norm": 0.5206366882864071, "learning_rate": 1.0565825282942303e-05, "loss": 0.2314, "step": 11458 }, { "epoch": 5.392470588235295, "grad_norm": 0.6356875300248523, "learning_rate": 1.0549670244981211e-05, "loss": 0.3613, "step": 11459 }, { "epoch": 5.392941176470588, "grad_norm": 0.5794341534934904, "learning_rate": 1.0533527116762296e-05, "loss": 0.2793, "step": 11460 }, { "epoch": 5.393411764705882, "grad_norm": 0.8206715517153694, "learning_rate": 1.0517395899664239e-05, "loss": 0.5273, "step": 11461 }, { "epoch": 5.393882352941176, "grad_norm": 0.43927756782566246, "learning_rate": 1.0501276595064716e-05, "loss": 0.1484, "step": 11462 }, { "epoch": 5.394352941176471, "grad_norm": 0.7679467176729623, "learning_rate": 1.0485169204340432e-05, "loss": 0.5039, "step": 11463 }, { "epoch": 5.394823529411765, "grad_norm": 0.7225399029889366, "learning_rate": 1.0469073728867e-05, "loss": 0.3496, "step": 11464 }, { "epoch": 5.395294117647059, "grad_norm": 0.25804958315180926, "learning_rate": 1.0452990170019088e-05, "loss": 0.0437, "step": 11465 }, { "epoch": 5.395764705882353, "grad_norm": 0.5194409387179355, "learning_rate": 1.0436918529170264e-05, "loss": 0.2969, "step": 11466 }, { "epoch": 5.3962352941176475, "grad_norm": 0.4616302415397864, "learning_rate": 1.0420858807693162e-05, "loss": 0.252, "step": 11467 }, { "epoch": 5.396705882352941, "grad_norm": 0.23861249801366605, "learning_rate": 1.0404811006959352e-05, "loss": 0.0576, "step": 11468 }, { "epoch": 5.397176470588235, "grad_norm": 0.49439512913283507, "learning_rate": 1.0388775128339389e-05, "loss": 0.2539, "step": 11469 }, { "epoch": 5.397647058823529, "grad_norm": 0.39409254687843925, "learning_rate": 1.037275117320281e-05, "loss": 0.2451, "step": 11470 }, { "epoch": 5.3981176470588235, "grad_norm": 0.4246210901778168, "learning_rate": 1.0356739142918169e-05, "loss": 0.2451, "step": 11471 }, { "epoch": 5.398588235294118, "grad_norm": 0.589943346532801, "learning_rate": 1.0340739038852924e-05, "loss": 0.2656, "step": 11472 }, { "epoch": 5.399058823529412, "grad_norm": 1.06420133746066, "learning_rate": 1.0324750862373598e-05, "loss": 0.6016, "step": 11473 }, { "epoch": 5.399529411764706, "grad_norm": 0.4309519776612356, "learning_rate": 1.0308774614845683e-05, "loss": 0.126, "step": 11474 }, { "epoch": 5.4, "grad_norm": 0.5491954139723978, "learning_rate": 1.0292810297633569e-05, "loss": 0.291, "step": 11475 }, { "epoch": 5.400470588235294, "grad_norm": 0.2961011704058787, "learning_rate": 1.02768579121007e-05, "loss": 0.0747, "step": 11476 }, { "epoch": 5.400941176470588, "grad_norm": 0.4936315677769389, "learning_rate": 1.0260917459609535e-05, "loss": 0.2285, "step": 11477 }, { "epoch": 5.401411764705882, "grad_norm": 0.63711592183697, "learning_rate": 1.024498894152142e-05, "loss": 0.291, "step": 11478 }, { "epoch": 5.401882352941176, "grad_norm": 0.8160002549480206, "learning_rate": 1.0229072359196749e-05, "loss": 0.4863, "step": 11479 }, { "epoch": 5.402352941176471, "grad_norm": 0.6200960113773526, "learning_rate": 1.0213167713994885e-05, "loss": 0.2617, "step": 11480 }, { "epoch": 5.402823529411765, "grad_norm": 0.918511683250976, "learning_rate": 1.0197275007274126e-05, "loss": 0.5781, "step": 11481 }, { "epoch": 5.403294117647059, "grad_norm": 0.364363318635954, "learning_rate": 1.0181394240391798e-05, "loss": 0.1445, "step": 11482 }, { "epoch": 5.403764705882353, "grad_norm": 0.4754283956441361, "learning_rate": 1.016552541470424e-05, "loss": 0.1914, "step": 11483 }, { "epoch": 5.4042352941176475, "grad_norm": 0.25689477821618684, "learning_rate": 1.0149668531566662e-05, "loss": 0.0381, "step": 11484 }, { "epoch": 5.404705882352941, "grad_norm": 0.9069896892202963, "learning_rate": 1.0133823592333385e-05, "loss": 0.6562, "step": 11485 }, { "epoch": 5.405176470588235, "grad_norm": 0.7458619785679951, "learning_rate": 1.0117990598357589e-05, "loss": 0.3281, "step": 11486 }, { "epoch": 5.405647058823529, "grad_norm": 0.4945521358576834, "learning_rate": 1.0102169550991529e-05, "loss": 0.2578, "step": 11487 }, { "epoch": 5.4061176470588235, "grad_norm": 0.6296152365202747, "learning_rate": 1.008636045158639e-05, "loss": 0.2207, "step": 11488 }, { "epoch": 5.406588235294118, "grad_norm": 0.5202276348557903, "learning_rate": 1.0070563301492323e-05, "loss": 0.0898, "step": 11489 }, { "epoch": 5.407058823529412, "grad_norm": 0.587624405850376, "learning_rate": 1.0054778102058502e-05, "loss": 0.3086, "step": 11490 }, { "epoch": 5.407529411764706, "grad_norm": 0.5587803233418269, "learning_rate": 1.0039004854633082e-05, "loss": 0.2617, "step": 11491 }, { "epoch": 5.408, "grad_norm": 0.308916680640731, "learning_rate": 1.0023243560563132e-05, "loss": 0.0635, "step": 11492 }, { "epoch": 5.408470588235295, "grad_norm": 0.47740038339653434, "learning_rate": 1.0007494221194762e-05, "loss": 0.2246, "step": 11493 }, { "epoch": 5.408941176470588, "grad_norm": 0.7119456234301378, "learning_rate": 9.991756837873061e-06, "loss": 0.4688, "step": 11494 }, { "epoch": 5.409411764705882, "grad_norm": 0.4423457909667018, "learning_rate": 9.976031411942054e-06, "loss": 0.1953, "step": 11495 }, { "epoch": 5.409882352941176, "grad_norm": 0.4736756919182226, "learning_rate": 9.960317944744783e-06, "loss": 0.2305, "step": 11496 }, { "epoch": 5.410352941176471, "grad_norm": 0.7646105288091932, "learning_rate": 9.944616437623243e-06, "loss": 0.4746, "step": 11497 }, { "epoch": 5.410823529411765, "grad_norm": 0.37970822165306467, "learning_rate": 9.928926891918426e-06, "loss": 0.1396, "step": 11498 }, { "epoch": 5.411294117647059, "grad_norm": 0.45039934014411004, "learning_rate": 9.913249308970312e-06, "loss": 0.2412, "step": 11499 }, { "epoch": 5.411764705882353, "grad_norm": 0.40418798678373025, "learning_rate": 9.89758369011781e-06, "loss": 0.2041, "step": 11500 }, { "epoch": 5.412235294117647, "grad_norm": 0.5391641912675228, "learning_rate": 9.881930036698854e-06, "loss": 0.3789, "step": 11501 }, { "epoch": 5.412705882352941, "grad_norm": 0.5301692921649331, "learning_rate": 9.866288350050372e-06, "loss": 0.3242, "step": 11502 }, { "epoch": 5.413176470588235, "grad_norm": 0.5561113472449595, "learning_rate": 9.850658631508196e-06, "loss": 0.3418, "step": 11503 }, { "epoch": 5.413647058823529, "grad_norm": 0.25836704292127194, "learning_rate": 9.835040882407207e-06, "loss": 0.0942, "step": 11504 }, { "epoch": 5.4141176470588235, "grad_norm": 0.4021993811391637, "learning_rate": 9.819435104081242e-06, "loss": 0.2178, "step": 11505 }, { "epoch": 5.414588235294118, "grad_norm": 0.5737912048981996, "learning_rate": 9.803841297863096e-06, "loss": 0.3145, "step": 11506 }, { "epoch": 5.415058823529412, "grad_norm": 1.3158710343813735, "learning_rate": 9.788259465084541e-06, "loss": 0.6172, "step": 11507 }, { "epoch": 5.415529411764706, "grad_norm": 0.314114352916037, "learning_rate": 9.772689607076394e-06, "loss": 0.0923, "step": 11508 }, { "epoch": 5.416, "grad_norm": 1.067384872543277, "learning_rate": 9.757131725168343e-06, "loss": 0.6758, "step": 11509 }, { "epoch": 5.416470588235294, "grad_norm": 0.3484392126829448, "learning_rate": 9.74158582068914e-06, "loss": 0.0674, "step": 11510 }, { "epoch": 5.416941176470588, "grad_norm": 0.49268364564761546, "learning_rate": 9.726051894966492e-06, "loss": 0.1631, "step": 11511 }, { "epoch": 5.417411764705882, "grad_norm": 0.4960532376536087, "learning_rate": 9.710529949327018e-06, "loss": 0.3262, "step": 11512 }, { "epoch": 5.417882352941176, "grad_norm": 0.6773303956872855, "learning_rate": 9.695019985096409e-06, "loss": 0.3301, "step": 11513 }, { "epoch": 5.418352941176471, "grad_norm": 0.16409595281869563, "learning_rate": 9.679522003599305e-06, "loss": 0.0349, "step": 11514 }, { "epoch": 5.418823529411765, "grad_norm": 0.5167594737788125, "learning_rate": 9.664036006159266e-06, "loss": 0.2061, "step": 11515 }, { "epoch": 5.419294117647059, "grad_norm": 0.3138185859254565, "learning_rate": 9.648561994098913e-06, "loss": 0.0767, "step": 11516 }, { "epoch": 5.419764705882353, "grad_norm": 0.41163414431502104, "learning_rate": 9.633099968739777e-06, "loss": 0.1162, "step": 11517 }, { "epoch": 5.4202352941176475, "grad_norm": 0.39688703917036317, "learning_rate": 9.617649931402399e-06, "loss": 0.1797, "step": 11518 }, { "epoch": 5.420705882352941, "grad_norm": 0.6738432521250094, "learning_rate": 9.602211883406308e-06, "loss": 0.3008, "step": 11519 }, { "epoch": 5.421176470588235, "grad_norm": 0.14712245668117752, "learning_rate": 9.586785826069932e-06, "loss": 0.0206, "step": 11520 }, { "epoch": 5.421647058823529, "grad_norm": 0.4745768191311179, "learning_rate": 9.571371760710783e-06, "loss": 0.1895, "step": 11521 }, { "epoch": 5.4221176470588235, "grad_norm": 0.20392643251957027, "learning_rate": 9.55596968864531e-06, "loss": 0.0603, "step": 11522 }, { "epoch": 5.422588235294118, "grad_norm": 0.5170702469307303, "learning_rate": 9.540579611188875e-06, "loss": 0.3223, "step": 11523 }, { "epoch": 5.423058823529412, "grad_norm": 0.44960266453360503, "learning_rate": 9.52520152965588e-06, "loss": 0.207, "step": 11524 }, { "epoch": 5.423529411764706, "grad_norm": 0.3640356661915313, "learning_rate": 9.509835445359736e-06, "loss": 0.1235, "step": 11525 }, { "epoch": 5.424, "grad_norm": 0.4278553186753964, "learning_rate": 9.494481359612715e-06, "loss": 0.1738, "step": 11526 }, { "epoch": 5.424470588235294, "grad_norm": 0.565578787421228, "learning_rate": 9.479139273726166e-06, "loss": 0.2812, "step": 11527 }, { "epoch": 5.424941176470588, "grad_norm": 0.9264023364857568, "learning_rate": 9.46380918901039e-06, "loss": 0.6562, "step": 11528 }, { "epoch": 5.425411764705882, "grad_norm": 0.7689428232861926, "learning_rate": 9.448491106774591e-06, "loss": 0.2852, "step": 11529 }, { "epoch": 5.425882352941176, "grad_norm": 0.539232476502466, "learning_rate": 9.433185028327106e-06, "loss": 0.3262, "step": 11530 }, { "epoch": 5.426352941176471, "grad_norm": 0.3834291161970205, "learning_rate": 9.417890954975071e-06, "loss": 0.1738, "step": 11531 }, { "epoch": 5.426823529411765, "grad_norm": 0.5372795894486077, "learning_rate": 9.402608888024692e-06, "loss": 0.2422, "step": 11532 }, { "epoch": 5.427294117647059, "grad_norm": 0.3921770457503792, "learning_rate": 9.38733882878116e-06, "loss": 0.1777, "step": 11533 }, { "epoch": 5.427764705882353, "grad_norm": 0.38931567268274025, "learning_rate": 9.37208077854858e-06, "loss": 0.1572, "step": 11534 }, { "epoch": 5.428235294117647, "grad_norm": 0.534324046645192, "learning_rate": 9.356834738630075e-06, "loss": 0.3203, "step": 11535 }, { "epoch": 5.428705882352941, "grad_norm": 0.5411061521534016, "learning_rate": 9.341600710327773e-06, "loss": 0.2344, "step": 11536 }, { "epoch": 5.429176470588235, "grad_norm": 0.31354361524927693, "learning_rate": 9.32637869494265e-06, "loss": 0.1221, "step": 11537 }, { "epoch": 5.429647058823529, "grad_norm": 0.4872726990651753, "learning_rate": 9.311168693774796e-06, "loss": 0.2363, "step": 11538 }, { "epoch": 5.4301176470588235, "grad_norm": 0.5522770974230821, "learning_rate": 9.295970708123223e-06, "loss": 0.1826, "step": 11539 }, { "epoch": 5.430588235294118, "grad_norm": 0.57489361228107, "learning_rate": 9.280784739285896e-06, "loss": 0.2227, "step": 11540 }, { "epoch": 5.431058823529412, "grad_norm": 0.3528671863283307, "learning_rate": 9.265610788559758e-06, "loss": 0.0942, "step": 11541 }, { "epoch": 5.431529411764706, "grad_norm": 0.45409353350239184, "learning_rate": 9.250448857240789e-06, "loss": 0.209, "step": 11542 }, { "epoch": 5.432, "grad_norm": 0.452693830642993, "learning_rate": 9.235298946623838e-06, "loss": 0.1572, "step": 11543 }, { "epoch": 5.432470588235294, "grad_norm": 0.533410657491791, "learning_rate": 9.220161058002834e-06, "loss": 0.4238, "step": 11544 }, { "epoch": 5.432941176470588, "grad_norm": 0.7026549541009474, "learning_rate": 9.205035192670563e-06, "loss": 0.5859, "step": 11545 }, { "epoch": 5.433411764705882, "grad_norm": 0.5987036184215382, "learning_rate": 9.189921351918888e-06, "loss": 0.2637, "step": 11546 }, { "epoch": 5.433882352941176, "grad_norm": 0.8676093644619806, "learning_rate": 9.174819537038614e-06, "loss": 0.5742, "step": 11547 }, { "epoch": 5.434352941176471, "grad_norm": 0.7875119443543641, "learning_rate": 9.159729749319473e-06, "loss": 0.3945, "step": 11548 }, { "epoch": 5.434823529411765, "grad_norm": 0.8739324094577109, "learning_rate": 9.144651990050234e-06, "loss": 0.7148, "step": 11549 }, { "epoch": 5.435294117647059, "grad_norm": 0.518711717800814, "learning_rate": 9.129586260518634e-06, "loss": 0.4102, "step": 11550 }, { "epoch": 5.435764705882353, "grad_norm": 0.41987927302316996, "learning_rate": 9.114532562011312e-06, "loss": 0.1914, "step": 11551 }, { "epoch": 5.4362352941176475, "grad_norm": 0.5803247777373889, "learning_rate": 9.099490895813938e-06, "loss": 0.3086, "step": 11552 }, { "epoch": 5.436705882352941, "grad_norm": 0.1382077813773649, "learning_rate": 9.084461263211168e-06, "loss": 0.0408, "step": 11553 }, { "epoch": 5.437176470588235, "grad_norm": 0.35791262655410444, "learning_rate": 9.069443665486593e-06, "loss": 0.2559, "step": 11554 }, { "epoch": 5.437647058823529, "grad_norm": 0.29340304352573104, "learning_rate": 9.054438103922767e-06, "loss": 0.0859, "step": 11555 }, { "epoch": 5.4381176470588235, "grad_norm": 0.1890586289361848, "learning_rate": 9.039444579801302e-06, "loss": 0.0515, "step": 11556 }, { "epoch": 5.438588235294118, "grad_norm": 0.4460942044868828, "learning_rate": 9.024463094402656e-06, "loss": 0.1709, "step": 11557 }, { "epoch": 5.439058823529412, "grad_norm": 0.36813823129676176, "learning_rate": 9.009493649006322e-06, "loss": 0.1357, "step": 11558 }, { "epoch": 5.439529411764706, "grad_norm": 0.3626262622515887, "learning_rate": 8.994536244890827e-06, "loss": 0.0889, "step": 11559 }, { "epoch": 5.44, "grad_norm": 0.3613017520118771, "learning_rate": 8.979590883333532e-06, "loss": 0.2168, "step": 11560 }, { "epoch": 5.440470588235294, "grad_norm": 0.487638395928107, "learning_rate": 8.964657565610884e-06, "loss": 0.2051, "step": 11561 }, { "epoch": 5.440941176470588, "grad_norm": 0.4201212774258218, "learning_rate": 8.949736292998277e-06, "loss": 0.0645, "step": 11562 }, { "epoch": 5.441411764705882, "grad_norm": 0.3748022758055811, "learning_rate": 8.934827066770012e-06, "loss": 0.1816, "step": 11563 }, { "epoch": 5.441882352941176, "grad_norm": 0.3633190090863944, "learning_rate": 8.919929888199435e-06, "loss": 0.2012, "step": 11564 }, { "epoch": 5.442352941176471, "grad_norm": 0.3704417610446727, "learning_rate": 8.90504475855886e-06, "loss": 0.0801, "step": 11565 }, { "epoch": 5.442823529411765, "grad_norm": 0.5564075041336924, "learning_rate": 8.890171679119506e-06, "loss": 0.291, "step": 11566 }, { "epoch": 5.443294117647059, "grad_norm": 0.6037818697787803, "learning_rate": 8.875310651151657e-06, "loss": 0.4375, "step": 11567 }, { "epoch": 5.443764705882353, "grad_norm": 0.608730703605801, "learning_rate": 8.86046167592448e-06, "loss": 0.293, "step": 11568 }, { "epoch": 5.444235294117647, "grad_norm": 0.5601851601559868, "learning_rate": 8.845624754706144e-06, "loss": 0.4648, "step": 11569 }, { "epoch": 5.444705882352941, "grad_norm": 0.4322093612536987, "learning_rate": 8.830799888763834e-06, "loss": 0.2256, "step": 11570 }, { "epoch": 5.445176470588235, "grad_norm": 0.5780883158224855, "learning_rate": 8.81598707936364e-06, "loss": 0.252, "step": 11571 }, { "epoch": 5.445647058823529, "grad_norm": 0.4581333320191277, "learning_rate": 8.801186327770632e-06, "loss": 0.2256, "step": 11572 }, { "epoch": 5.4461176470588235, "grad_norm": 0.5336188342407736, "learning_rate": 8.7863976352489e-06, "loss": 0.2617, "step": 11573 }, { "epoch": 5.446588235294118, "grad_norm": 0.30894580130760435, "learning_rate": 8.771621003061463e-06, "loss": 0.1504, "step": 11574 }, { "epoch": 5.447058823529412, "grad_norm": 0.6870263417278493, "learning_rate": 8.756856432470315e-06, "loss": 0.3008, "step": 11575 }, { "epoch": 5.447529411764706, "grad_norm": 0.5531454412138876, "learning_rate": 8.742103924736416e-06, "loss": 0.3105, "step": 11576 }, { "epoch": 5.448, "grad_norm": 0.6647389562076739, "learning_rate": 8.727363481119704e-06, "loss": 0.3203, "step": 11577 }, { "epoch": 5.448470588235294, "grad_norm": 0.5617485603036696, "learning_rate": 8.712635102879124e-06, "loss": 0.3105, "step": 11578 }, { "epoch": 5.448941176470588, "grad_norm": 0.4572206991878405, "learning_rate": 8.697918791272485e-06, "loss": 0.2656, "step": 11579 }, { "epoch": 5.449411764705882, "grad_norm": 0.5180653504860703, "learning_rate": 8.68321454755665e-06, "loss": 0.2373, "step": 11580 }, { "epoch": 5.449882352941176, "grad_norm": 0.367833447795061, "learning_rate": 8.668522372987479e-06, "loss": 0.1758, "step": 11581 }, { "epoch": 5.450352941176471, "grad_norm": 0.3577127953121611, "learning_rate": 8.653842268819723e-06, "loss": 0.0835, "step": 11582 }, { "epoch": 5.450823529411765, "grad_norm": 0.3360195925528823, "learning_rate": 8.639174236307106e-06, "loss": 0.1123, "step": 11583 }, { "epoch": 5.451294117647059, "grad_norm": 0.5163254216709986, "learning_rate": 8.624518276702414e-06, "loss": 0.2598, "step": 11584 }, { "epoch": 5.451764705882353, "grad_norm": 0.45338516635221665, "learning_rate": 8.609874391257293e-06, "loss": 0.2793, "step": 11585 }, { "epoch": 5.4522352941176475, "grad_norm": 0.4699239918112135, "learning_rate": 8.595242581222394e-06, "loss": 0.1758, "step": 11586 }, { "epoch": 5.452705882352941, "grad_norm": 0.47459085804066176, "learning_rate": 8.580622847847402e-06, "loss": 0.1484, "step": 11587 }, { "epoch": 5.453176470588235, "grad_norm": 0.6364271709420132, "learning_rate": 8.566015192380849e-06, "loss": 0.1855, "step": 11588 }, { "epoch": 5.453647058823529, "grad_norm": 0.6222051824130752, "learning_rate": 8.551419616070322e-06, "loss": 0.2578, "step": 11589 }, { "epoch": 5.4541176470588235, "grad_norm": 0.8322855106719159, "learning_rate": 8.536836120162389e-06, "loss": 0.5273, "step": 11590 }, { "epoch": 5.454588235294118, "grad_norm": 0.6522280586543422, "learning_rate": 8.522264705902504e-06, "loss": 0.3789, "step": 11591 }, { "epoch": 5.455058823529412, "grad_norm": 0.4779462539586162, "learning_rate": 8.507705374535157e-06, "loss": 0.2129, "step": 11592 }, { "epoch": 5.455529411764706, "grad_norm": 0.4162711035848519, "learning_rate": 8.4931581273038e-06, "loss": 0.2002, "step": 11593 }, { "epoch": 5.456, "grad_norm": 0.5624754601145121, "learning_rate": 8.47862296545081e-06, "loss": 0.3203, "step": 11594 }, { "epoch": 5.456470588235294, "grad_norm": 0.47104465034043125, "learning_rate": 8.464099890217607e-06, "loss": 0.291, "step": 11595 }, { "epoch": 5.456941176470588, "grad_norm": 0.5969661671053922, "learning_rate": 8.449588902844484e-06, "loss": 0.332, "step": 11596 }, { "epoch": 5.457411764705882, "grad_norm": 0.434416483737698, "learning_rate": 8.43509000457075e-06, "loss": 0.1592, "step": 11597 }, { "epoch": 5.457882352941176, "grad_norm": 0.41955458200294354, "learning_rate": 8.420603196634745e-06, "loss": 0.123, "step": 11598 }, { "epoch": 5.458352941176471, "grad_norm": 0.39865029919689304, "learning_rate": 8.406128480273633e-06, "loss": 0.1641, "step": 11599 }, { "epoch": 5.458823529411765, "grad_norm": 0.5630639759853231, "learning_rate": 8.391665856723655e-06, "loss": 0.2637, "step": 11600 }, { "epoch": 5.459294117647059, "grad_norm": 0.5427916036575283, "learning_rate": 8.377215327220043e-06, "loss": 0.3633, "step": 11601 }, { "epoch": 5.459764705882353, "grad_norm": 0.5624771655487618, "learning_rate": 8.362776892996858e-06, "loss": 0.4902, "step": 11602 }, { "epoch": 5.460235294117647, "grad_norm": 0.6927266509947422, "learning_rate": 8.348350555287264e-06, "loss": 0.3984, "step": 11603 }, { "epoch": 5.460705882352941, "grad_norm": 0.37964148159585176, "learning_rate": 8.333936315323343e-06, "loss": 0.1436, "step": 11604 }, { "epoch": 5.461176470588235, "grad_norm": 0.25553372473923885, "learning_rate": 8.319534174336106e-06, "loss": 0.1016, "step": 11605 }, { "epoch": 5.461647058823529, "grad_norm": 0.4193149093170249, "learning_rate": 8.305144133555608e-06, "loss": 0.2578, "step": 11606 }, { "epoch": 5.4621176470588235, "grad_norm": 0.6010441374086128, "learning_rate": 8.290766194210813e-06, "loss": 0.2871, "step": 11607 }, { "epoch": 5.462588235294118, "grad_norm": 0.6611978198473878, "learning_rate": 8.276400357529655e-06, "loss": 0.4102, "step": 11608 }, { "epoch": 5.463058823529412, "grad_norm": 0.4580814288756175, "learning_rate": 8.262046624739038e-06, "loss": 0.2637, "step": 11609 }, { "epoch": 5.463529411764706, "grad_norm": 0.4564290233393328, "learning_rate": 8.247704997064896e-06, "loss": 0.1914, "step": 11610 }, { "epoch": 5.464, "grad_norm": 0.2742941245392114, "learning_rate": 8.233375475732018e-06, "loss": 0.0889, "step": 11611 }, { "epoch": 5.464470588235294, "grad_norm": 0.686799391319325, "learning_rate": 8.21905806196424e-06, "loss": 0.4414, "step": 11612 }, { "epoch": 5.464941176470588, "grad_norm": 0.08354394816552073, "learning_rate": 8.204752756984351e-06, "loss": 0.0162, "step": 11613 }, { "epoch": 5.465411764705882, "grad_norm": 0.5151601056940567, "learning_rate": 8.19045956201404e-06, "loss": 0.2334, "step": 11614 }, { "epoch": 5.465882352941176, "grad_norm": 0.44778862512841233, "learning_rate": 8.176178478274082e-06, "loss": 0.2305, "step": 11615 }, { "epoch": 5.466352941176471, "grad_norm": 0.5365792151082702, "learning_rate": 8.161909506984132e-06, "loss": 0.21, "step": 11616 }, { "epoch": 5.466823529411765, "grad_norm": 0.5010376385901474, "learning_rate": 8.147652649362802e-06, "loss": 0.1816, "step": 11617 }, { "epoch": 5.467294117647059, "grad_norm": 0.4392320583900988, "learning_rate": 8.133407906627715e-06, "loss": 0.2324, "step": 11618 }, { "epoch": 5.467764705882353, "grad_norm": 0.4327532705538617, "learning_rate": 8.119175279995448e-06, "loss": 0.1484, "step": 11619 }, { "epoch": 5.4682352941176475, "grad_norm": 0.13600985317480177, "learning_rate": 8.104954770681532e-06, "loss": 0.0403, "step": 11620 }, { "epoch": 5.468705882352941, "grad_norm": 0.5103795459614114, "learning_rate": 8.090746379900475e-06, "loss": 0.252, "step": 11621 }, { "epoch": 5.469176470588235, "grad_norm": 0.2470225828555253, "learning_rate": 8.076550108865743e-06, "loss": 0.0688, "step": 11622 }, { "epoch": 5.469647058823529, "grad_norm": 0.6195650886811032, "learning_rate": 8.062365958789746e-06, "loss": 0.2559, "step": 11623 }, { "epoch": 5.4701176470588235, "grad_norm": 0.5173758011743418, "learning_rate": 8.048193930883933e-06, "loss": 0.1934, "step": 11624 }, { "epoch": 5.470588235294118, "grad_norm": 0.6282217936680373, "learning_rate": 8.034034026358587e-06, "loss": 0.3223, "step": 11625 }, { "epoch": 5.471058823529412, "grad_norm": 0.6400643816900389, "learning_rate": 8.019886246423107e-06, "loss": 0.3477, "step": 11626 }, { "epoch": 5.471529411764706, "grad_norm": 0.6972296544361677, "learning_rate": 8.005750592285725e-06, "loss": 0.3906, "step": 11627 }, { "epoch": 5.4719999999999995, "grad_norm": 0.5150013218556013, "learning_rate": 7.991627065153727e-06, "loss": 0.3438, "step": 11628 }, { "epoch": 5.472470588235294, "grad_norm": 0.4069732020020468, "learning_rate": 7.977515666233347e-06, "loss": 0.1128, "step": 11629 }, { "epoch": 5.472941176470588, "grad_norm": 0.6801744090314111, "learning_rate": 7.963416396729722e-06, "loss": 0.3613, "step": 11630 }, { "epoch": 5.473411764705882, "grad_norm": 0.5890390264242185, "learning_rate": 7.94932925784702e-06, "loss": 0.2363, "step": 11631 }, { "epoch": 5.473882352941176, "grad_norm": 0.23941166435657607, "learning_rate": 7.935254250788364e-06, "loss": 0.0615, "step": 11632 }, { "epoch": 5.474352941176471, "grad_norm": 0.501684971153529, "learning_rate": 7.921191376755808e-06, "loss": 0.4043, "step": 11633 }, { "epoch": 5.474823529411765, "grad_norm": 0.34040160798905944, "learning_rate": 7.90714063695041e-06, "loss": 0.1182, "step": 11634 }, { "epoch": 5.475294117647059, "grad_norm": 0.35316656417789977, "learning_rate": 7.893102032572174e-06, "loss": 0.1396, "step": 11635 }, { "epoch": 5.475764705882353, "grad_norm": 0.5501203994855831, "learning_rate": 7.879075564820025e-06, "loss": 0.2617, "step": 11636 }, { "epoch": 5.476235294117647, "grad_norm": 0.15895306390661468, "learning_rate": 7.86506123489194e-06, "loss": 0.0383, "step": 11637 }, { "epoch": 5.476705882352941, "grad_norm": 0.48609665233001303, "learning_rate": 7.85105904398481e-06, "loss": 0.2578, "step": 11638 }, { "epoch": 5.477176470588235, "grad_norm": 0.5798194298342632, "learning_rate": 7.837068993294443e-06, "loss": 0.1895, "step": 11639 }, { "epoch": 5.477647058823529, "grad_norm": 0.44434256731732996, "learning_rate": 7.823091084015687e-06, "loss": 0.1562, "step": 11640 }, { "epoch": 5.4781176470588235, "grad_norm": 0.5627201009925423, "learning_rate": 7.809125317342369e-06, "loss": 0.3281, "step": 11641 }, { "epoch": 5.478588235294118, "grad_norm": 0.5349100444504419, "learning_rate": 7.79517169446715e-06, "loss": 0.3516, "step": 11642 }, { "epoch": 5.479058823529412, "grad_norm": 0.3198803049394244, "learning_rate": 7.781230216581779e-06, "loss": 0.083, "step": 11643 }, { "epoch": 5.479529411764706, "grad_norm": 0.28058048159821997, "learning_rate": 7.767300884876971e-06, "loss": 0.0669, "step": 11644 }, { "epoch": 5.48, "grad_norm": 0.25049000905497404, "learning_rate": 7.753383700542271e-06, "loss": 0.0405, "step": 11645 }, { "epoch": 5.480470588235294, "grad_norm": 0.3016466206466344, "learning_rate": 7.739478664766347e-06, "loss": 0.0879, "step": 11646 }, { "epoch": 5.480941176470588, "grad_norm": 0.31396422833067317, "learning_rate": 7.725585778736698e-06, "loss": 0.0786, "step": 11647 }, { "epoch": 5.481411764705882, "grad_norm": 0.477371163493958, "learning_rate": 7.711705043639893e-06, "loss": 0.3477, "step": 11648 }, { "epoch": 5.481882352941176, "grad_norm": 0.6203541032947327, "learning_rate": 7.697836460661399e-06, "loss": 0.3672, "step": 11649 }, { "epoch": 5.482352941176471, "grad_norm": 0.33392418716177685, "learning_rate": 7.683980030985654e-06, "loss": 0.1777, "step": 11650 }, { "epoch": 5.482823529411765, "grad_norm": 0.31566491616728254, "learning_rate": 7.670135755796074e-06, "loss": 0.127, "step": 11651 }, { "epoch": 5.483294117647059, "grad_norm": 0.2877185627572803, "learning_rate": 7.656303636275036e-06, "loss": 0.0894, "step": 11652 }, { "epoch": 5.483764705882353, "grad_norm": 0.22883008125780377, "learning_rate": 7.64248367360384e-06, "loss": 0.0588, "step": 11653 }, { "epoch": 5.484235294117647, "grad_norm": 0.604135875367978, "learning_rate": 7.6286758689627915e-06, "loss": 0.3984, "step": 11654 }, { "epoch": 5.484705882352941, "grad_norm": 0.2826897328363916, "learning_rate": 7.614880223531183e-06, "loss": 0.0996, "step": 11655 }, { "epoch": 5.485176470588235, "grad_norm": 0.6272782002308428, "learning_rate": 7.6010967384871725e-06, "loss": 0.373, "step": 11656 }, { "epoch": 5.485647058823529, "grad_norm": 0.31367817374076784, "learning_rate": 7.587325415007967e-06, "loss": 0.1357, "step": 11657 }, { "epoch": 5.4861176470588235, "grad_norm": 0.44179428869857035, "learning_rate": 7.573566254269708e-06, "loss": 0.2314, "step": 11658 }, { "epoch": 5.486588235294118, "grad_norm": 0.5914406435005941, "learning_rate": 7.559819257447492e-06, "loss": 0.373, "step": 11659 }, { "epoch": 5.487058823529412, "grad_norm": 0.5604360948065611, "learning_rate": 7.546084425715359e-06, "loss": 0.2773, "step": 11660 }, { "epoch": 5.487529411764706, "grad_norm": 0.37519462260267405, "learning_rate": 7.532361760246375e-06, "loss": 0.1348, "step": 11661 }, { "epoch": 5.4879999999999995, "grad_norm": 0.3101455500517905, "learning_rate": 7.5186512622124495e-06, "loss": 0.0732, "step": 11662 }, { "epoch": 5.488470588235294, "grad_norm": 0.18810555445351315, "learning_rate": 7.5049529327846135e-06, "loss": 0.0471, "step": 11663 }, { "epoch": 5.488941176470588, "grad_norm": 0.6117546393650511, "learning_rate": 7.491266773132715e-06, "loss": 0.2734, "step": 11664 }, { "epoch": 5.489411764705882, "grad_norm": 0.4769659963932166, "learning_rate": 7.477592784425634e-06, "loss": 0.1924, "step": 11665 }, { "epoch": 5.489882352941176, "grad_norm": 0.38767242608988356, "learning_rate": 7.463930967831222e-06, "loss": 0.1221, "step": 11666 }, { "epoch": 5.490352941176471, "grad_norm": 0.2423294802949501, "learning_rate": 7.450281324516228e-06, "loss": 0.0986, "step": 11667 }, { "epoch": 5.490823529411765, "grad_norm": 0.3738980037756671, "learning_rate": 7.436643855646402e-06, "loss": 0.1553, "step": 11668 }, { "epoch": 5.491294117647059, "grad_norm": 0.40087186269971187, "learning_rate": 7.423018562386479e-06, "loss": 0.1108, "step": 11669 }, { "epoch": 5.491764705882353, "grad_norm": 0.47174750048009, "learning_rate": 7.40940544590008e-06, "loss": 0.1875, "step": 11670 }, { "epoch": 5.492235294117647, "grad_norm": 0.4518432766325655, "learning_rate": 7.395804507349873e-06, "loss": 0.2178, "step": 11671 }, { "epoch": 5.492705882352941, "grad_norm": 0.30577170511101676, "learning_rate": 7.3822157478974465e-06, "loss": 0.1118, "step": 11672 }, { "epoch": 5.493176470588235, "grad_norm": 0.5955333842591112, "learning_rate": 7.368639168703306e-06, "loss": 0.2324, "step": 11673 }, { "epoch": 5.493647058823529, "grad_norm": 0.516060471605374, "learning_rate": 7.3550747709269885e-06, "loss": 0.1826, "step": 11674 }, { "epoch": 5.4941176470588236, "grad_norm": 0.6237450929431569, "learning_rate": 7.34152255572697e-06, "loss": 0.3613, "step": 11675 }, { "epoch": 5.494588235294118, "grad_norm": 0.5157048697808934, "learning_rate": 7.327982524260639e-06, "loss": 0.2217, "step": 11676 }, { "epoch": 5.495058823529412, "grad_norm": 0.27306405352594293, "learning_rate": 7.314454677684439e-06, "loss": 0.0679, "step": 11677 }, { "epoch": 5.495529411764706, "grad_norm": 0.7336319639070132, "learning_rate": 7.300939017153645e-06, "loss": 0.3711, "step": 11678 }, { "epoch": 5.496, "grad_norm": 0.554005658490469, "learning_rate": 7.287435543822584e-06, "loss": 0.2891, "step": 11679 }, { "epoch": 5.496470588235294, "grad_norm": 0.5287613398947846, "learning_rate": 7.2739442588445666e-06, "loss": 0.3535, "step": 11680 }, { "epoch": 5.496941176470588, "grad_norm": 0.5723523155773507, "learning_rate": 7.260465163371754e-06, "loss": 0.1299, "step": 11681 }, { "epoch": 5.497411764705882, "grad_norm": 0.7417945850278711, "learning_rate": 7.246998258555342e-06, "loss": 0.5547, "step": 11682 }, { "epoch": 5.497882352941176, "grad_norm": 0.6347494074122606, "learning_rate": 7.233543545545511e-06, "loss": 0.4141, "step": 11683 }, { "epoch": 5.498352941176471, "grad_norm": 0.5778481767574314, "learning_rate": 7.220101025491293e-06, "loss": 0.2246, "step": 11684 }, { "epoch": 5.498823529411765, "grad_norm": 0.7787702014667334, "learning_rate": 7.2066706995407845e-06, "loss": 0.4629, "step": 11685 }, { "epoch": 5.499294117647059, "grad_norm": 0.24779736411417874, "learning_rate": 7.193252568841018e-06, "loss": 0.0811, "step": 11686 }, { "epoch": 5.499764705882353, "grad_norm": 0.4512794977113227, "learning_rate": 7.1798466345379115e-06, "loss": 0.2256, "step": 11687 }, { "epoch": 5.5002352941176476, "grad_norm": 0.5782315002035183, "learning_rate": 7.166452897776448e-06, "loss": 0.373, "step": 11688 }, { "epoch": 5.500705882352941, "grad_norm": 0.7053805636477395, "learning_rate": 7.153071359700513e-06, "loss": 0.3438, "step": 11689 }, { "epoch": 5.501176470588235, "grad_norm": 0.5603717528937857, "learning_rate": 7.1397020214529255e-06, "loss": 0.2715, "step": 11690 }, { "epoch": 5.501647058823529, "grad_norm": 0.5883031253866566, "learning_rate": 7.126344884175522e-06, "loss": 0.2051, "step": 11691 }, { "epoch": 5.502117647058824, "grad_norm": 0.4518962011610364, "learning_rate": 7.1129999490090735e-06, "loss": 0.1572, "step": 11692 }, { "epoch": 5.502588235294118, "grad_norm": 0.5990832371005348, "learning_rate": 7.099667217093252e-06, "loss": 0.332, "step": 11693 }, { "epoch": 5.503058823529412, "grad_norm": 0.5737461105432177, "learning_rate": 7.08634668956678e-06, "loss": 0.2314, "step": 11694 }, { "epoch": 5.503529411764706, "grad_norm": 0.6315155969840999, "learning_rate": 7.073038367567313e-06, "loss": 0.2598, "step": 11695 }, { "epoch": 5.504, "grad_norm": 0.7704983473759693, "learning_rate": 7.059742252231393e-06, "loss": 0.5078, "step": 11696 }, { "epoch": 5.504470588235294, "grad_norm": 0.7105226615392333, "learning_rate": 7.046458344694628e-06, "loss": 0.2305, "step": 11697 }, { "epoch": 5.504941176470588, "grad_norm": 0.6537480752261892, "learning_rate": 7.033186646091477e-06, "loss": 0.3574, "step": 11698 }, { "epoch": 5.505411764705882, "grad_norm": 0.4772572415415801, "learning_rate": 7.019927157555449e-06, "loss": 0.252, "step": 11699 }, { "epoch": 5.5058823529411764, "grad_norm": 0.6227309032733477, "learning_rate": 7.006679880218974e-06, "loss": 0.5039, "step": 11700 }, { "epoch": 5.506352941176471, "grad_norm": 0.3829216657662003, "learning_rate": 6.993444815213378e-06, "loss": 0.1069, "step": 11701 }, { "epoch": 5.506823529411765, "grad_norm": 0.555881335233902, "learning_rate": 6.9802219636690585e-06, "loss": 0.3164, "step": 11702 }, { "epoch": 5.507294117647059, "grad_norm": 0.4593910843859698, "learning_rate": 6.967011326715294e-06, "loss": 0.1738, "step": 11703 }, { "epoch": 5.507764705882353, "grad_norm": 0.37454023986008794, "learning_rate": 6.953812905480316e-06, "loss": 0.1826, "step": 11704 }, { "epoch": 5.508235294117647, "grad_norm": 0.2856941304699046, "learning_rate": 6.940626701091356e-06, "loss": 0.0767, "step": 11705 }, { "epoch": 5.508705882352941, "grad_norm": 0.6201436617707808, "learning_rate": 6.927452714674614e-06, "loss": 0.3496, "step": 11706 }, { "epoch": 5.509176470588235, "grad_norm": 0.3997399135970611, "learning_rate": 6.914290947355122e-06, "loss": 0.1572, "step": 11707 }, { "epoch": 5.509647058823529, "grad_norm": 0.4555260400562399, "learning_rate": 6.901141400257048e-06, "loss": 0.2988, "step": 11708 }, { "epoch": 5.510117647058824, "grad_norm": 0.4787438206450843, "learning_rate": 6.888004074503395e-06, "loss": 0.2949, "step": 11709 }, { "epoch": 5.510588235294118, "grad_norm": 0.3466740956317288, "learning_rate": 6.87487897121613e-06, "loss": 0.1523, "step": 11710 }, { "epoch": 5.511058823529412, "grad_norm": 0.32456666005470275, "learning_rate": 6.861766091516274e-06, "loss": 0.1104, "step": 11711 }, { "epoch": 5.511529411764706, "grad_norm": 0.2888280320182928, "learning_rate": 6.848665436523631e-06, "loss": 0.0757, "step": 11712 }, { "epoch": 5.5120000000000005, "grad_norm": 0.4481741426098084, "learning_rate": 6.83557700735714e-06, "loss": 0.1689, "step": 11713 }, { "epoch": 5.512470588235294, "grad_norm": 0.8270305047780658, "learning_rate": 6.82250080513459e-06, "loss": 0.375, "step": 11714 }, { "epoch": 5.512941176470588, "grad_norm": 0.5687958555021, "learning_rate": 6.809436830972753e-06, "loss": 0.1963, "step": 11715 }, { "epoch": 5.513411764705882, "grad_norm": 0.2626540225769925, "learning_rate": 6.796385085987338e-06, "loss": 0.0776, "step": 11716 }, { "epoch": 5.5138823529411765, "grad_norm": 0.6268468945230462, "learning_rate": 6.783345571293069e-06, "loss": 0.2617, "step": 11717 }, { "epoch": 5.514352941176471, "grad_norm": 0.5911678273180159, "learning_rate": 6.770318288003556e-06, "loss": 0.416, "step": 11718 }, { "epoch": 5.514823529411765, "grad_norm": 0.7367181418080171, "learning_rate": 6.757303237231393e-06, "loss": 0.252, "step": 11719 }, { "epoch": 5.515294117647059, "grad_norm": 0.7929453356211601, "learning_rate": 6.744300420088156e-06, "loss": 0.5273, "step": 11720 }, { "epoch": 5.5157647058823525, "grad_norm": 0.6268255201809528, "learning_rate": 6.73130983768429e-06, "loss": 0.291, "step": 11721 }, { "epoch": 5.516235294117647, "grad_norm": 0.2574838537682978, "learning_rate": 6.718331491129309e-06, "loss": 0.0693, "step": 11722 }, { "epoch": 5.516705882352941, "grad_norm": 0.4077205699368032, "learning_rate": 6.705365381531641e-06, "loss": 0.1592, "step": 11723 }, { "epoch": 5.517176470588235, "grad_norm": 0.5111739667346258, "learning_rate": 6.6924115099985856e-06, "loss": 0.1748, "step": 11724 }, { "epoch": 5.517647058823529, "grad_norm": 0.3280116180148843, "learning_rate": 6.679469877636523e-06, "loss": 0.082, "step": 11725 }, { "epoch": 5.518117647058824, "grad_norm": 0.290358479609755, "learning_rate": 6.6665404855507195e-06, "loss": 0.0947, "step": 11726 }, { "epoch": 5.518588235294118, "grad_norm": 0.41368706393599075, "learning_rate": 6.6536233348453925e-06, "loss": 0.291, "step": 11727 }, { "epoch": 5.519058823529412, "grad_norm": 0.37946039406739135, "learning_rate": 6.640718426623759e-06, "loss": 0.1196, "step": 11728 }, { "epoch": 5.519529411764706, "grad_norm": 0.608913136155475, "learning_rate": 6.627825761987937e-06, "loss": 0.3633, "step": 11729 }, { "epoch": 5.52, "grad_norm": 0.37994185484978626, "learning_rate": 6.61494534203903e-06, "loss": 0.249, "step": 11730 }, { "epoch": 5.520470588235294, "grad_norm": 0.44536064323916325, "learning_rate": 6.6020771678771066e-06, "loss": 0.1523, "step": 11731 }, { "epoch": 5.520941176470588, "grad_norm": 0.37801637848843184, "learning_rate": 6.589221240601123e-06, "loss": 0.168, "step": 11732 }, { "epoch": 5.521411764705882, "grad_norm": 0.4886950572209865, "learning_rate": 6.576377561309082e-06, "loss": 0.2988, "step": 11733 }, { "epoch": 5.5218823529411765, "grad_norm": 0.15043612018342287, "learning_rate": 6.563546131097907e-06, "loss": 0.0266, "step": 11734 }, { "epoch": 5.522352941176471, "grad_norm": 0.6490949902968872, "learning_rate": 6.550726951063423e-06, "loss": 0.3711, "step": 11735 }, { "epoch": 5.522823529411765, "grad_norm": 0.42091135218539916, "learning_rate": 6.5379200223004855e-06, "loss": 0.2266, "step": 11736 }, { "epoch": 5.523294117647059, "grad_norm": 0.4472372939892886, "learning_rate": 6.5251253459028545e-06, "loss": 0.1621, "step": 11737 }, { "epoch": 5.523764705882353, "grad_norm": 0.521651101884515, "learning_rate": 6.512342922963254e-06, "loss": 0.2217, "step": 11738 }, { "epoch": 5.524235294117647, "grad_norm": 0.5539383913395105, "learning_rate": 6.499572754573362e-06, "loss": 0.2734, "step": 11739 }, { "epoch": 5.524705882352941, "grad_norm": 0.30908884921859786, "learning_rate": 6.48681484182384e-06, "loss": 0.1543, "step": 11740 }, { "epoch": 5.525176470588235, "grad_norm": 0.6602678713763693, "learning_rate": 6.474069185804248e-06, "loss": 0.3203, "step": 11741 }, { "epoch": 5.525647058823529, "grad_norm": 0.36442728493832766, "learning_rate": 6.461335787603134e-06, "loss": 0.1348, "step": 11742 }, { "epoch": 5.526117647058824, "grad_norm": 0.6236246824510218, "learning_rate": 6.448614648308009e-06, "loss": 0.3359, "step": 11743 }, { "epoch": 5.526588235294118, "grad_norm": 0.8352328484059313, "learning_rate": 6.435905769005306e-06, "loss": 0.5312, "step": 11744 }, { "epoch": 5.527058823529412, "grad_norm": 0.4570987155419161, "learning_rate": 6.423209150780423e-06, "loss": 0.2119, "step": 11745 }, { "epoch": 5.527529411764706, "grad_norm": 0.31063690302025443, "learning_rate": 6.4105247947177256e-06, "loss": 0.0952, "step": 11746 }, { "epoch": 5.5280000000000005, "grad_norm": 0.31135585403202565, "learning_rate": 6.3978527019004964e-06, "loss": 0.0859, "step": 11747 }, { "epoch": 5.528470588235294, "grad_norm": 0.29753861448289953, "learning_rate": 6.385192873411038e-06, "loss": 0.0669, "step": 11748 }, { "epoch": 5.528941176470588, "grad_norm": 0.24785021458581857, "learning_rate": 6.372545310330518e-06, "loss": 0.0542, "step": 11749 }, { "epoch": 5.529411764705882, "grad_norm": 0.32996117897128413, "learning_rate": 6.359910013739122e-06, "loss": 0.1689, "step": 11750 }, { "epoch": 5.5298823529411765, "grad_norm": 0.39910055502032266, "learning_rate": 6.347286984715938e-06, "loss": 0.1377, "step": 11751 }, { "epoch": 5.530352941176471, "grad_norm": 0.4606097727912005, "learning_rate": 6.334676224339086e-06, "loss": 0.2578, "step": 11752 }, { "epoch": 5.530823529411765, "grad_norm": 0.45006614141747925, "learning_rate": 6.322077733685538e-06, "loss": 0.1543, "step": 11753 }, { "epoch": 5.531294117647059, "grad_norm": 0.7079247054028751, "learning_rate": 6.309491513831317e-06, "loss": 0.457, "step": 11754 }, { "epoch": 5.5317647058823525, "grad_norm": 0.7295885144079636, "learning_rate": 6.296917565851295e-06, "loss": 0.4668, "step": 11755 }, { "epoch": 5.532235294117647, "grad_norm": 0.2390185188122923, "learning_rate": 6.284355890819398e-06, "loss": 0.064, "step": 11756 }, { "epoch": 5.532705882352941, "grad_norm": 0.615188570162541, "learning_rate": 6.271806489808401e-06, "loss": 0.4102, "step": 11757 }, { "epoch": 5.533176470588235, "grad_norm": 0.5666442433608002, "learning_rate": 6.2592693638901125e-06, "loss": 0.3555, "step": 11758 }, { "epoch": 5.533647058823529, "grad_norm": 0.24439951157058637, "learning_rate": 6.246744514135293e-06, "loss": 0.0557, "step": 11759 }, { "epoch": 5.534117647058824, "grad_norm": 0.32161663455965656, "learning_rate": 6.234231941613571e-06, "loss": 0.0889, "step": 11760 }, { "epoch": 5.534588235294118, "grad_norm": 0.5731216886551795, "learning_rate": 6.221731647393607e-06, "loss": 0.3242, "step": 11761 }, { "epoch": 5.535058823529412, "grad_norm": 0.1262783073532806, "learning_rate": 6.209243632543015e-06, "loss": 0.016, "step": 11762 }, { "epoch": 5.535529411764706, "grad_norm": 1.085876448464295, "learning_rate": 6.196767898128291e-06, "loss": 0.4824, "step": 11763 }, { "epoch": 5.536, "grad_norm": 0.6689888224067309, "learning_rate": 6.184304445214932e-06, "loss": 0.2793, "step": 11764 }, { "epoch": 5.536470588235294, "grad_norm": 0.362660103123064, "learning_rate": 6.171853274867406e-06, "loss": 0.0977, "step": 11765 }, { "epoch": 5.536941176470588, "grad_norm": 0.3282332293809932, "learning_rate": 6.159414388149059e-06, "loss": 0.0737, "step": 11766 }, { "epoch": 5.537411764705882, "grad_norm": 0.49158875195150126, "learning_rate": 6.146987786122259e-06, "loss": 0.168, "step": 11767 }, { "epoch": 5.5378823529411765, "grad_norm": 0.31792193173400196, "learning_rate": 6.134573469848325e-06, "loss": 0.085, "step": 11768 }, { "epoch": 5.538352941176471, "grad_norm": 0.4486192819083438, "learning_rate": 6.1221714403874415e-06, "loss": 0.3281, "step": 11769 }, { "epoch": 5.538823529411765, "grad_norm": 0.2663353462439939, "learning_rate": 6.109781698798843e-06, "loss": 0.0938, "step": 11770 }, { "epoch": 5.539294117647059, "grad_norm": 0.502039743316479, "learning_rate": 6.097404246140669e-06, "loss": 0.2656, "step": 11771 }, { "epoch": 5.539764705882353, "grad_norm": 0.5028232348048273, "learning_rate": 6.085039083470006e-06, "loss": 0.293, "step": 11772 }, { "epoch": 5.540235294117647, "grad_norm": 0.32013462720881797, "learning_rate": 6.072686211842909e-06, "loss": 0.1143, "step": 11773 }, { "epoch": 5.540705882352941, "grad_norm": 0.49785201567948445, "learning_rate": 6.060345632314384e-06, "loss": 0.166, "step": 11774 }, { "epoch": 5.541176470588235, "grad_norm": 0.42244795862858003, "learning_rate": 6.048017345938339e-06, "loss": 0.1206, "step": 11775 }, { "epoch": 5.541647058823529, "grad_norm": 0.36383280690241954, "learning_rate": 6.035701353767713e-06, "loss": 0.0894, "step": 11776 }, { "epoch": 5.542117647058824, "grad_norm": 0.4266392163542559, "learning_rate": 6.023397656854301e-06, "loss": 0.0889, "step": 11777 }, { "epoch": 5.542588235294118, "grad_norm": 0.39040193165886045, "learning_rate": 6.011106256248943e-06, "loss": 0.3008, "step": 11778 }, { "epoch": 5.543058823529412, "grad_norm": 0.5027387215291864, "learning_rate": 5.998827153001401e-06, "loss": 0.1924, "step": 11779 }, { "epoch": 5.543529411764706, "grad_norm": 0.516184078184786, "learning_rate": 5.986560348160302e-06, "loss": 0.3574, "step": 11780 }, { "epoch": 5.5440000000000005, "grad_norm": 0.53996424136403, "learning_rate": 5.9743058427733414e-06, "loss": 0.1523, "step": 11781 }, { "epoch": 5.544470588235294, "grad_norm": 0.3977039424603941, "learning_rate": 5.9620636378871325e-06, "loss": 0.1309, "step": 11782 }, { "epoch": 5.544941176470588, "grad_norm": 0.72389698590378, "learning_rate": 5.9498337345471544e-06, "loss": 0.3379, "step": 11783 }, { "epoch": 5.545411764705882, "grad_norm": 0.6018172781749764, "learning_rate": 5.93761613379794e-06, "loss": 0.3691, "step": 11784 }, { "epoch": 5.5458823529411765, "grad_norm": 0.15355264278674105, "learning_rate": 5.925410836682953e-06, "loss": 0.0272, "step": 11785 }, { "epoch": 5.546352941176471, "grad_norm": 0.34193152447490893, "learning_rate": 5.913217844244545e-06, "loss": 0.1289, "step": 11786 }, { "epoch": 5.546823529411765, "grad_norm": 0.4461813021358526, "learning_rate": 5.90103715752408e-06, "loss": 0.1279, "step": 11787 }, { "epoch": 5.547294117647059, "grad_norm": 0.3240780374446045, "learning_rate": 5.888868777561862e-06, "loss": 0.0874, "step": 11788 }, { "epoch": 5.5477647058823525, "grad_norm": 0.7111503867676489, "learning_rate": 5.8767127053970745e-06, "loss": 0.4316, "step": 11789 }, { "epoch": 5.548235294117647, "grad_norm": 0.6087240545756595, "learning_rate": 5.864568942067954e-06, "loss": 0.2012, "step": 11790 }, { "epoch": 5.548705882352941, "grad_norm": 0.49531072570205664, "learning_rate": 5.852437488611656e-06, "loss": 0.1953, "step": 11791 }, { "epoch": 5.549176470588235, "grad_norm": 0.645334598030671, "learning_rate": 5.840318346064199e-06, "loss": 0.4102, "step": 11792 }, { "epoch": 5.549647058823529, "grad_norm": 0.5917511224877761, "learning_rate": 5.828211515460673e-06, "loss": 0.3242, "step": 11793 }, { "epoch": 5.550117647058824, "grad_norm": 0.41445475799438714, "learning_rate": 5.816116997835052e-06, "loss": 0.1816, "step": 11794 }, { "epoch": 5.550588235294118, "grad_norm": 0.4816804981156664, "learning_rate": 5.804034794220208e-06, "loss": 0.21, "step": 11795 }, { "epoch": 5.551058823529412, "grad_norm": 0.27415769509570326, "learning_rate": 5.791964905648116e-06, "loss": 0.1328, "step": 11796 }, { "epoch": 5.551529411764706, "grad_norm": 0.5246234998063952, "learning_rate": 5.779907333149553e-06, "loss": 0.3184, "step": 11797 }, { "epoch": 5.552, "grad_norm": 0.30901019743892055, "learning_rate": 5.767862077754276e-06, "loss": 0.0869, "step": 11798 }, { "epoch": 5.552470588235294, "grad_norm": 0.37048280691127156, "learning_rate": 5.755829140491064e-06, "loss": 0.0586, "step": 11799 }, { "epoch": 5.552941176470588, "grad_norm": 0.3470516369979744, "learning_rate": 5.743808522387544e-06, "loss": 0.126, "step": 11800 }, { "epoch": 5.553411764705882, "grad_norm": 0.35906110037168854, "learning_rate": 5.7318002244703285e-06, "loss": 0.0913, "step": 11801 }, { "epoch": 5.5538823529411765, "grad_norm": 0.7030207552253284, "learning_rate": 5.7198042477650475e-06, "loss": 0.5469, "step": 11802 }, { "epoch": 5.554352941176471, "grad_norm": 0.6724943066126331, "learning_rate": 5.7078205932961476e-06, "loss": 0.3164, "step": 11803 }, { "epoch": 5.554823529411765, "grad_norm": 0.5238961062735781, "learning_rate": 5.69584926208711e-06, "loss": 0.3516, "step": 11804 }, { "epoch": 5.555294117647059, "grad_norm": 0.48737599933961595, "learning_rate": 5.683890255160367e-06, "loss": 0.2109, "step": 11805 }, { "epoch": 5.555764705882353, "grad_norm": 0.4416394676722258, "learning_rate": 5.671943573537269e-06, "loss": 0.3379, "step": 11806 }, { "epoch": 5.556235294117647, "grad_norm": 0.20214538498269047, "learning_rate": 5.6600092182381156e-06, "loss": 0.0388, "step": 11807 }, { "epoch": 5.556705882352941, "grad_norm": 0.06603732480786843, "learning_rate": 5.648087190282141e-06, "loss": 0.0211, "step": 11808 }, { "epoch": 5.557176470588235, "grad_norm": 0.4836746537595512, "learning_rate": 5.636177490687565e-06, "loss": 0.2217, "step": 11809 }, { "epoch": 5.557647058823529, "grad_norm": 0.2672134235819322, "learning_rate": 5.624280120471542e-06, "loss": 0.0757, "step": 11810 }, { "epoch": 5.558117647058824, "grad_norm": 0.7398281949694607, "learning_rate": 5.612395080650156e-06, "loss": 0.4219, "step": 11811 }, { "epoch": 5.558588235294118, "grad_norm": 0.3903921908422866, "learning_rate": 5.600522372238431e-06, "loss": 0.1934, "step": 11812 }, { "epoch": 5.559058823529412, "grad_norm": 0.28706043246242546, "learning_rate": 5.588661996250387e-06, "loss": 0.0898, "step": 11813 }, { "epoch": 5.559529411764705, "grad_norm": 0.4205685110253374, "learning_rate": 5.576813953698933e-06, "loss": 0.2354, "step": 11814 }, { "epoch": 5.5600000000000005, "grad_norm": 0.644048841194377, "learning_rate": 5.564978245595958e-06, "loss": 0.3223, "step": 11815 }, { "epoch": 5.560470588235294, "grad_norm": 0.3060684409230354, "learning_rate": 5.553154872952303e-06, "loss": 0.0928, "step": 11816 }, { "epoch": 5.560941176470588, "grad_norm": 0.5659402153706793, "learning_rate": 5.541343836777729e-06, "loss": 0.2314, "step": 11817 }, { "epoch": 5.561411764705882, "grad_norm": 0.6439445193506642, "learning_rate": 5.529545138080943e-06, "loss": 0.3984, "step": 11818 }, { "epoch": 5.5618823529411765, "grad_norm": 0.05796031878169002, "learning_rate": 5.517758777869641e-06, "loss": 0.0072, "step": 11819 }, { "epoch": 5.562352941176471, "grad_norm": 0.5381806521478818, "learning_rate": 5.5059847571504e-06, "loss": 0.1904, "step": 11820 }, { "epoch": 5.562823529411765, "grad_norm": 0.365458260585991, "learning_rate": 5.494223076928817e-06, "loss": 0.125, "step": 11821 }, { "epoch": 5.563294117647059, "grad_norm": 0.36647185343308386, "learning_rate": 5.482473738209386e-06, "loss": 0.1309, "step": 11822 }, { "epoch": 5.5637647058823525, "grad_norm": 0.5674942847776965, "learning_rate": 5.470736741995541e-06, "loss": 0.2676, "step": 11823 }, { "epoch": 5.564235294117647, "grad_norm": 0.4475607177117173, "learning_rate": 5.4590120892896775e-06, "loss": 0.1406, "step": 11824 }, { "epoch": 5.564705882352941, "grad_norm": 0.36831163356383484, "learning_rate": 5.447299781093178e-06, "loss": 0.1104, "step": 11825 }, { "epoch": 5.565176470588235, "grad_norm": 0.33851355132268784, "learning_rate": 5.435599818406278e-06, "loss": 0.0972, "step": 11826 }, { "epoch": 5.565647058823529, "grad_norm": 0.31173262308620625, "learning_rate": 5.423912202228275e-06, "loss": 0.1196, "step": 11827 }, { "epoch": 5.566117647058824, "grad_norm": 0.6754531399722657, "learning_rate": 5.412236933557274e-06, "loss": 0.2852, "step": 11828 }, { "epoch": 5.566588235294118, "grad_norm": 0.6801261970411111, "learning_rate": 5.400574013390441e-06, "loss": 0.3711, "step": 11829 }, { "epoch": 5.567058823529412, "grad_norm": 0.7351179633639242, "learning_rate": 5.388923442723864e-06, "loss": 0.4297, "step": 11830 }, { "epoch": 5.567529411764706, "grad_norm": 0.6362094062785009, "learning_rate": 5.37728522255253e-06, "loss": 0.457, "step": 11831 }, { "epoch": 5.568, "grad_norm": 0.3531442650026017, "learning_rate": 5.365659353870394e-06, "loss": 0.1592, "step": 11832 }, { "epoch": 5.568470588235294, "grad_norm": 0.4628882835170806, "learning_rate": 5.354045837670412e-06, "loss": 0.1953, "step": 11833 }, { "epoch": 5.568941176470588, "grad_norm": 0.4397571223260448, "learning_rate": 5.342444674944374e-06, "loss": 0.2295, "step": 11834 }, { "epoch": 5.569411764705882, "grad_norm": 0.3256380618954241, "learning_rate": 5.330855866683087e-06, "loss": 0.1187, "step": 11835 }, { "epoch": 5.5698823529411765, "grad_norm": 0.5258966009575197, "learning_rate": 5.319279413876343e-06, "loss": 0.3359, "step": 11836 }, { "epoch": 5.570352941176471, "grad_norm": 0.5409329712662094, "learning_rate": 5.307715317512767e-06, "loss": 0.2324, "step": 11837 }, { "epoch": 5.570823529411765, "grad_norm": 0.5559358516034967, "learning_rate": 5.29616357858002e-06, "loss": 0.2832, "step": 11838 }, { "epoch": 5.571294117647059, "grad_norm": 0.580019782372817, "learning_rate": 5.2846241980646795e-06, "loss": 0.2461, "step": 11839 }, { "epoch": 5.571764705882353, "grad_norm": 0.5780609826382404, "learning_rate": 5.273097176952224e-06, "loss": 0.2754, "step": 11840 }, { "epoch": 5.572235294117647, "grad_norm": 0.500653166622737, "learning_rate": 5.261582516227197e-06, "loss": 0.2871, "step": 11841 }, { "epoch": 5.572705882352941, "grad_norm": 0.5625282052309221, "learning_rate": 5.250080216872932e-06, "loss": 0.1943, "step": 11842 }, { "epoch": 5.573176470588235, "grad_norm": 0.43974836782045057, "learning_rate": 5.238590279871824e-06, "loss": 0.2793, "step": 11843 }, { "epoch": 5.573647058823529, "grad_norm": 0.5243620496547904, "learning_rate": 5.227112706205155e-06, "loss": 0.4238, "step": 11844 }, { "epoch": 5.574117647058824, "grad_norm": 0.6345307163795992, "learning_rate": 5.2156474968531595e-06, "loss": 0.3086, "step": 11845 }, { "epoch": 5.574588235294118, "grad_norm": 0.3591252850689306, "learning_rate": 5.204194652795035e-06, "loss": 0.1289, "step": 11846 }, { "epoch": 5.575058823529412, "grad_norm": 0.5017589410391408, "learning_rate": 5.1927541750089175e-06, "loss": 0.2695, "step": 11847 }, { "epoch": 5.575529411764705, "grad_norm": 0.48235715214826724, "learning_rate": 5.181326064471858e-06, "loss": 0.3047, "step": 11848 }, { "epoch": 5.576, "grad_norm": 0.639750142577468, "learning_rate": 5.169910322159876e-06, "loss": 0.3945, "step": 11849 }, { "epoch": 5.576470588235294, "grad_norm": 0.6235688579737477, "learning_rate": 5.158506949047975e-06, "loss": 0.2637, "step": 11850 }, { "epoch": 5.576941176470588, "grad_norm": 0.4411362374265664, "learning_rate": 5.147115946109992e-06, "loss": 0.3145, "step": 11851 }, { "epoch": 5.577411764705882, "grad_norm": 0.4957856303614027, "learning_rate": 5.1357373143188e-06, "loss": 0.2617, "step": 11852 }, { "epoch": 5.5778823529411765, "grad_norm": 0.4985034754939516, "learning_rate": 5.124371054646237e-06, "loss": 0.2207, "step": 11853 }, { "epoch": 5.578352941176471, "grad_norm": 0.5601507328018587, "learning_rate": 5.11301716806296e-06, "loss": 0.3242, "step": 11854 }, { "epoch": 5.578823529411765, "grad_norm": 0.2948268354644197, "learning_rate": 5.101675655538695e-06, "loss": 0.0898, "step": 11855 }, { "epoch": 5.579294117647059, "grad_norm": 0.3311610689701111, "learning_rate": 5.090346518042049e-06, "loss": 0.0752, "step": 11856 }, { "epoch": 5.5797647058823525, "grad_norm": 0.38215199569594965, "learning_rate": 5.079029756540598e-06, "loss": 0.1553, "step": 11857 }, { "epoch": 5.580235294117647, "grad_norm": 0.49518008175602685, "learning_rate": 5.067725372000836e-06, "loss": 0.2832, "step": 11858 }, { "epoch": 5.580705882352941, "grad_norm": 0.7287709609466846, "learning_rate": 5.056433365388207e-06, "loss": 0.4668, "step": 11859 }, { "epoch": 5.581176470588235, "grad_norm": 0.36061673563458274, "learning_rate": 5.045153737667107e-06, "loss": 0.1816, "step": 11860 }, { "epoch": 5.581647058823529, "grad_norm": 0.36069015114303854, "learning_rate": 5.033886489800898e-06, "loss": 0.1436, "step": 11861 }, { "epoch": 5.582117647058824, "grad_norm": 0.4728976073001548, "learning_rate": 5.022631622751827e-06, "loss": 0.1611, "step": 11862 }, { "epoch": 5.582588235294118, "grad_norm": 0.7632926879502748, "learning_rate": 5.011389137481109e-06, "loss": 0.4805, "step": 11863 }, { "epoch": 5.583058823529412, "grad_norm": 0.6640721321375628, "learning_rate": 5.000159034948958e-06, "loss": 0.4668, "step": 11864 }, { "epoch": 5.583529411764706, "grad_norm": 0.18552153121940068, "learning_rate": 4.988941316114425e-06, "loss": 0.0194, "step": 11865 }, { "epoch": 5.584, "grad_norm": 0.5153378623395892, "learning_rate": 4.9777359819355765e-06, "loss": 0.3027, "step": 11866 }, { "epoch": 5.584470588235294, "grad_norm": 0.4844629648330756, "learning_rate": 4.9665430333694135e-06, "loss": 0.291, "step": 11867 }, { "epoch": 5.584941176470588, "grad_norm": 0.5018183201322708, "learning_rate": 4.955362471371854e-06, "loss": 0.2832, "step": 11868 }, { "epoch": 5.585411764705882, "grad_norm": 0.3534174757200154, "learning_rate": 4.944194296897785e-06, "loss": 0.168, "step": 11869 }, { "epoch": 5.5858823529411765, "grad_norm": 0.3834386139941228, "learning_rate": 4.933038510901027e-06, "loss": 0.1592, "step": 11870 }, { "epoch": 5.586352941176471, "grad_norm": 0.41941548775159376, "learning_rate": 4.921895114334301e-06, "loss": 0.1836, "step": 11871 }, { "epoch": 5.586823529411765, "grad_norm": 0.5893459233067653, "learning_rate": 4.910764108149346e-06, "loss": 0.2119, "step": 11872 }, { "epoch": 5.587294117647059, "grad_norm": 0.9048463044484576, "learning_rate": 4.899645493296817e-06, "loss": 0.6289, "step": 11873 }, { "epoch": 5.587764705882353, "grad_norm": 0.4695596325661457, "learning_rate": 4.888539270726255e-06, "loss": 0.2217, "step": 11874 }, { "epoch": 5.588235294117647, "grad_norm": 0.4277829424825327, "learning_rate": 4.877445441386218e-06, "loss": 0.1289, "step": 11875 }, { "epoch": 5.588705882352941, "grad_norm": 0.5712960780657128, "learning_rate": 4.866364006224166e-06, "loss": 0.3711, "step": 11876 }, { "epoch": 5.589176470588235, "grad_norm": 0.5186254097631752, "learning_rate": 4.85529496618649e-06, "loss": 0.2559, "step": 11877 }, { "epoch": 5.589647058823529, "grad_norm": 0.5700150504206171, "learning_rate": 4.844238322218568e-06, "loss": 0.4043, "step": 11878 }, { "epoch": 5.590117647058824, "grad_norm": 0.6539721611713467, "learning_rate": 4.833194075264663e-06, "loss": 0.1084, "step": 11879 }, { "epoch": 5.590588235294118, "grad_norm": 0.4093573215291365, "learning_rate": 4.8221622262680336e-06, "loss": 0.1377, "step": 11880 }, { "epoch": 5.591058823529412, "grad_norm": 0.9581345284293353, "learning_rate": 4.811142776170845e-06, "loss": 0.5859, "step": 11881 }, { "epoch": 5.591529411764705, "grad_norm": 0.4227576731608875, "learning_rate": 4.800135725914195e-06, "loss": 0.1118, "step": 11882 }, { "epoch": 5.592, "grad_norm": 0.48062329082951705, "learning_rate": 4.789141076438163e-06, "loss": 0.249, "step": 11883 }, { "epoch": 5.592470588235294, "grad_norm": 0.47927002441581124, "learning_rate": 4.778158828681733e-06, "loss": 0.2812, "step": 11884 }, { "epoch": 5.592941176470588, "grad_norm": 0.3625782124545113, "learning_rate": 4.767188983582854e-06, "loss": 0.0742, "step": 11885 }, { "epoch": 5.593411764705882, "grad_norm": 0.4290048110944784, "learning_rate": 4.75623154207841e-06, "loss": 0.2188, "step": 11886 }, { "epoch": 5.5938823529411765, "grad_norm": 0.10924746287040336, "learning_rate": 4.745286505104201e-06, "loss": 0.022, "step": 11887 }, { "epoch": 5.594352941176471, "grad_norm": 0.7249087603150053, "learning_rate": 4.734353873594998e-06, "loss": 0.1475, "step": 11888 }, { "epoch": 5.594823529411765, "grad_norm": 0.3536954045751934, "learning_rate": 4.723433648484503e-06, "loss": 0.0815, "step": 11889 }, { "epoch": 5.595294117647059, "grad_norm": 0.5554050648434158, "learning_rate": 4.712525830705338e-06, "loss": 0.2236, "step": 11890 }, { "epoch": 5.5957647058823525, "grad_norm": 0.5046311347105098, "learning_rate": 4.70163042118909e-06, "loss": 0.2969, "step": 11891 }, { "epoch": 5.596235294117647, "grad_norm": 0.4730325099934611, "learning_rate": 4.690747420866314e-06, "loss": 0.1592, "step": 11892 }, { "epoch": 5.596705882352941, "grad_norm": 0.7873834581185841, "learning_rate": 4.679876830666434e-06, "loss": 0.4531, "step": 11893 }, { "epoch": 5.597176470588235, "grad_norm": 0.4681116546068257, "learning_rate": 4.669018651517859e-06, "loss": 0.3516, "step": 11894 }, { "epoch": 5.597647058823529, "grad_norm": 0.42639986977918715, "learning_rate": 4.658172884347943e-06, "loss": 0.1611, "step": 11895 }, { "epoch": 5.598117647058824, "grad_norm": 0.47538631897405015, "learning_rate": 4.6473395300829495e-06, "loss": 0.1904, "step": 11896 }, { "epoch": 5.598588235294118, "grad_norm": 0.7142871337318758, "learning_rate": 4.636518589648119e-06, "loss": 0.5273, "step": 11897 }, { "epoch": 5.599058823529412, "grad_norm": 0.42925164810526684, "learning_rate": 4.625710063967613e-06, "loss": 0.2539, "step": 11898 }, { "epoch": 5.599529411764706, "grad_norm": 0.5134732363907614, "learning_rate": 4.614913953964511e-06, "loss": 0.2793, "step": 11899 }, { "epoch": 5.6, "grad_norm": 0.4632497317745016, "learning_rate": 4.604130260560873e-06, "loss": 0.2344, "step": 11900 }, { "epoch": 5.600470588235294, "grad_norm": 0.34921381429361736, "learning_rate": 4.593358984677681e-06, "loss": 0.1592, "step": 11901 }, { "epoch": 5.600941176470588, "grad_norm": 0.6233839985116717, "learning_rate": 4.5826001272348315e-06, "loss": 0.3086, "step": 11902 }, { "epoch": 5.601411764705882, "grad_norm": 0.47430256254216585, "learning_rate": 4.571853689151206e-06, "loss": 0.1885, "step": 11903 }, { "epoch": 5.6018823529411765, "grad_norm": 0.48606831926876887, "learning_rate": 4.5611196713446045e-06, "loss": 0.2256, "step": 11904 }, { "epoch": 5.602352941176471, "grad_norm": 0.4358509327133377, "learning_rate": 4.5503980747317425e-06, "loss": 0.1875, "step": 11905 }, { "epoch": 5.602823529411765, "grad_norm": 0.4912146592838434, "learning_rate": 4.539688900228322e-06, "loss": 0.2676, "step": 11906 }, { "epoch": 5.603294117647059, "grad_norm": 0.49390782690076285, "learning_rate": 4.528992148748944e-06, "loss": 0.2354, "step": 11907 }, { "epoch": 5.603764705882353, "grad_norm": 0.2833255956172555, "learning_rate": 4.518307821207162e-06, "loss": 0.1128, "step": 11908 }, { "epoch": 5.604235294117647, "grad_norm": 0.5327374556475447, "learning_rate": 4.507635918515496e-06, "loss": 0.2988, "step": 11909 }, { "epoch": 5.604705882352941, "grad_norm": 0.5514250808602886, "learning_rate": 4.496976441585348e-06, "loss": 0.2402, "step": 11910 }, { "epoch": 5.605176470588235, "grad_norm": 0.4755794163873126, "learning_rate": 4.486329391327093e-06, "loss": 0.3008, "step": 11911 }, { "epoch": 5.605647058823529, "grad_norm": 4.207881957249293, "learning_rate": 4.475694768650051e-06, "loss": 0.0791, "step": 11912 }, { "epoch": 5.606117647058824, "grad_norm": 0.6288081511557871, "learning_rate": 4.4650725744624635e-06, "loss": 0.291, "step": 11913 }, { "epoch": 5.606588235294118, "grad_norm": 0.5280244451290491, "learning_rate": 4.454462809671538e-06, "loss": 0.2949, "step": 11914 }, { "epoch": 5.607058823529412, "grad_norm": 0.4126507434483193, "learning_rate": 4.443865475183384e-06, "loss": 0.1455, "step": 11915 }, { "epoch": 5.607529411764705, "grad_norm": 0.45845411250860185, "learning_rate": 4.4332805719030586e-06, "loss": 0.2217, "step": 11916 }, { "epoch": 5.608, "grad_norm": 0.5246307013394707, "learning_rate": 4.4227081007345565e-06, "loss": 0.2656, "step": 11917 }, { "epoch": 5.608470588235294, "grad_norm": 0.47105147171627343, "learning_rate": 4.412148062580873e-06, "loss": 0.2002, "step": 11918 }, { "epoch": 5.608941176470588, "grad_norm": 0.36517135529568717, "learning_rate": 4.401600458343818e-06, "loss": 0.1367, "step": 11919 }, { "epoch": 5.609411764705882, "grad_norm": 0.3937866098296848, "learning_rate": 4.39106528892424e-06, "loss": 0.2139, "step": 11920 }, { "epoch": 5.6098823529411765, "grad_norm": 0.5306990095673977, "learning_rate": 4.380542555221917e-06, "loss": 0.2891, "step": 11921 }, { "epoch": 5.610352941176471, "grad_norm": 0.5665863611114734, "learning_rate": 4.370032258135514e-06, "loss": 0.3223, "step": 11922 }, { "epoch": 5.610823529411765, "grad_norm": 0.40555418176728736, "learning_rate": 4.3595343985626476e-06, "loss": 0.105, "step": 11923 }, { "epoch": 5.611294117647059, "grad_norm": 0.16270193526620597, "learning_rate": 4.349048977399933e-06, "loss": 0.0723, "step": 11924 }, { "epoch": 5.6117647058823525, "grad_norm": 0.6295279805429919, "learning_rate": 4.33857599554282e-06, "loss": 0.3633, "step": 11925 }, { "epoch": 5.612235294117647, "grad_norm": 0.44016411075476153, "learning_rate": 4.328115453885794e-06, "loss": 0.1738, "step": 11926 }, { "epoch": 5.612705882352941, "grad_norm": 0.5509670805876695, "learning_rate": 4.317667353322241e-06, "loss": 0.2539, "step": 11927 }, { "epoch": 5.613176470588235, "grad_norm": 0.3519678275280698, "learning_rate": 4.307231694744445e-06, "loss": 0.1318, "step": 11928 }, { "epoch": 5.613647058823529, "grad_norm": 0.635252790592284, "learning_rate": 4.296808479043678e-06, "loss": 0.2637, "step": 11929 }, { "epoch": 5.614117647058824, "grad_norm": 0.31298547724415954, "learning_rate": 4.286397707110145e-06, "loss": 0.1035, "step": 11930 }, { "epoch": 5.614588235294118, "grad_norm": 0.5067626418686096, "learning_rate": 4.275999379832967e-06, "loss": 0.2695, "step": 11931 }, { "epoch": 5.615058823529412, "grad_norm": 0.1955236790754691, "learning_rate": 4.265613498100234e-06, "loss": 0.0276, "step": 11932 }, { "epoch": 5.615529411764706, "grad_norm": 0.6452690902148941, "learning_rate": 4.255240062798904e-06, "loss": 0.4023, "step": 11933 }, { "epoch": 5.616, "grad_norm": 0.7155093470751943, "learning_rate": 4.24487907481495e-06, "loss": 0.293, "step": 11934 }, { "epoch": 5.616470588235294, "grad_norm": 0.1718244160496397, "learning_rate": 4.234530535033265e-06, "loss": 0.026, "step": 11935 }, { "epoch": 5.616941176470588, "grad_norm": 0.5095982882649258, "learning_rate": 4.224194444337642e-06, "loss": 0.3047, "step": 11936 }, { "epoch": 5.617411764705882, "grad_norm": 0.3234015230605961, "learning_rate": 4.213870803610842e-06, "loss": 0.1641, "step": 11937 }, { "epoch": 5.6178823529411765, "grad_norm": 0.3982360904503772, "learning_rate": 4.203559613734542e-06, "loss": 0.1143, "step": 11938 }, { "epoch": 5.618352941176471, "grad_norm": 0.48871332461701417, "learning_rate": 4.193260875589371e-06, "loss": 0.249, "step": 11939 }, { "epoch": 5.618823529411765, "grad_norm": 0.5095553400424746, "learning_rate": 4.1829745900549275e-06, "loss": 0.1855, "step": 11940 }, { "epoch": 5.619294117647058, "grad_norm": 0.4503307305865121, "learning_rate": 4.1727007580096565e-06, "loss": 0.1982, "step": 11941 }, { "epoch": 5.619764705882353, "grad_norm": 0.21483210338550746, "learning_rate": 4.162439380331023e-06, "loss": 0.1104, "step": 11942 }, { "epoch": 5.620235294117647, "grad_norm": 0.7756441660547612, "learning_rate": 4.15219045789541e-06, "loss": 0.5664, "step": 11943 }, { "epoch": 5.620705882352941, "grad_norm": 0.5586850409611118, "learning_rate": 4.141953991578101e-06, "loss": 0.3301, "step": 11944 }, { "epoch": 5.621176470588235, "grad_norm": 0.46476663555027287, "learning_rate": 4.131729982253345e-06, "loss": 0.2617, "step": 11945 }, { "epoch": 5.621647058823529, "grad_norm": 0.31730969946047993, "learning_rate": 4.121518430794346e-06, "loss": 0.125, "step": 11946 }, { "epoch": 5.622117647058824, "grad_norm": 0.482433371348432, "learning_rate": 4.111319338073188e-06, "loss": 0.291, "step": 11947 }, { "epoch": 5.622588235294118, "grad_norm": 0.6046103639458513, "learning_rate": 4.101132704960941e-06, "loss": 0.3672, "step": 11948 }, { "epoch": 5.623058823529412, "grad_norm": 0.5602137122204798, "learning_rate": 4.09095853232761e-06, "loss": 0.332, "step": 11949 }, { "epoch": 5.623529411764705, "grad_norm": 0.3857483775662322, "learning_rate": 4.080796821042082e-06, "loss": 0.1719, "step": 11950 }, { "epoch": 5.624, "grad_norm": 0.7012405683490552, "learning_rate": 4.07064757197223e-06, "loss": 0.3906, "step": 11951 }, { "epoch": 5.624470588235294, "grad_norm": 0.3662215509749287, "learning_rate": 4.0605107859848604e-06, "loss": 0.1406, "step": 11952 }, { "epoch": 5.624941176470588, "grad_norm": 0.8137243765029856, "learning_rate": 4.050386463945699e-06, "loss": 0.5742, "step": 11953 }, { "epoch": 5.625411764705882, "grad_norm": 0.35500252009002603, "learning_rate": 4.040274606719418e-06, "loss": 0.1021, "step": 11954 }, { "epoch": 5.625882352941177, "grad_norm": 0.4781193631719423, "learning_rate": 4.030175215169612e-06, "loss": 0.1885, "step": 11955 }, { "epoch": 5.626352941176471, "grad_norm": 0.6094290989953212, "learning_rate": 4.020088290158807e-06, "loss": 0.3809, "step": 11956 }, { "epoch": 5.626823529411765, "grad_norm": 0.3831109006532591, "learning_rate": 4.010013832548514e-06, "loss": 0.1807, "step": 11957 }, { "epoch": 5.627294117647059, "grad_norm": 0.5135717940366191, "learning_rate": 3.999951843199095e-06, "loss": 0.2812, "step": 11958 }, { "epoch": 5.627764705882353, "grad_norm": 1.410109386256376, "learning_rate": 3.989902322969929e-06, "loss": 0.1108, "step": 11959 }, { "epoch": 5.628235294117647, "grad_norm": 0.4972228092344468, "learning_rate": 3.979865272719279e-06, "loss": 0.2207, "step": 11960 }, { "epoch": 5.628705882352941, "grad_norm": 0.23614751359377484, "learning_rate": 3.969840693304344e-06, "loss": 0.0603, "step": 11961 }, { "epoch": 5.629176470588235, "grad_norm": 0.7160613486843599, "learning_rate": 3.959828585581287e-06, "loss": 0.5117, "step": 11962 }, { "epoch": 5.6296470588235294, "grad_norm": 0.5425741080278834, "learning_rate": 3.949828950405226e-06, "loss": 0.1465, "step": 11963 }, { "epoch": 5.630117647058824, "grad_norm": 0.2472645691071011, "learning_rate": 3.939841788630111e-06, "loss": 0.1016, "step": 11964 }, { "epoch": 5.630588235294118, "grad_norm": 0.513726937163324, "learning_rate": 3.9298671011089245e-06, "loss": 0.2773, "step": 11965 }, { "epoch": 5.631058823529412, "grad_norm": 0.46386771106840685, "learning_rate": 3.9199048886935865e-06, "loss": 0.2197, "step": 11966 }, { "epoch": 5.631529411764706, "grad_norm": 0.4599667177583749, "learning_rate": 3.909955152234867e-06, "loss": 0.1338, "step": 11967 }, { "epoch": 5.632, "grad_norm": 0.40633232587778456, "learning_rate": 3.900017892582552e-06, "loss": 0.1689, "step": 11968 }, { "epoch": 5.632470588235294, "grad_norm": 0.35724664569673537, "learning_rate": 3.8900931105853294e-06, "loss": 0.1196, "step": 11969 }, { "epoch": 5.632941176470588, "grad_norm": 0.5233861767476283, "learning_rate": 3.8801808070908215e-06, "loss": 0.1523, "step": 11970 }, { "epoch": 5.633411764705882, "grad_norm": 0.5765886007377837, "learning_rate": 3.870280982945567e-06, "loss": 0.4023, "step": 11971 }, { "epoch": 5.633882352941177, "grad_norm": 0.8175318552174063, "learning_rate": 3.860393638995107e-06, "loss": 0.5195, "step": 11972 }, { "epoch": 5.634352941176471, "grad_norm": 0.40974831161522807, "learning_rate": 3.850518776083817e-06, "loss": 0.1719, "step": 11973 }, { "epoch": 5.634823529411765, "grad_norm": 0.34289607522141635, "learning_rate": 3.840656395055103e-06, "loss": 0.1104, "step": 11974 }, { "epoch": 5.635294117647058, "grad_norm": 0.5245586093355272, "learning_rate": 3.8308064967512095e-06, "loss": 0.2256, "step": 11975 }, { "epoch": 5.635764705882353, "grad_norm": 0.34976551414374574, "learning_rate": 3.820969082013414e-06, "loss": 0.1094, "step": 11976 }, { "epoch": 5.636235294117647, "grad_norm": 0.29566893262080035, "learning_rate": 3.8111441516818776e-06, "loss": 0.1055, "step": 11977 }, { "epoch": 5.636705882352941, "grad_norm": 0.4725221452827667, "learning_rate": 3.8013317065956784e-06, "loss": 0.1963, "step": 11978 }, { "epoch": 5.637176470588235, "grad_norm": 0.4779881314215456, "learning_rate": 3.7915317475928296e-06, "loss": 0.2754, "step": 11979 }, { "epoch": 5.6376470588235295, "grad_norm": 0.6259851948986198, "learning_rate": 3.7817442755103457e-06, "loss": 0.2422, "step": 11980 }, { "epoch": 5.638117647058824, "grad_norm": 0.6927631065969871, "learning_rate": 3.7719692911840905e-06, "loss": 0.5859, "step": 11981 }, { "epoch": 5.638588235294118, "grad_norm": 0.3949640754166719, "learning_rate": 3.76220679544888e-06, "loss": 0.2021, "step": 11982 }, { "epoch": 5.639058823529412, "grad_norm": 0.33724402653877633, "learning_rate": 3.7524567891385305e-06, "loss": 0.1279, "step": 11983 }, { "epoch": 5.6395294117647055, "grad_norm": 0.477678085661578, "learning_rate": 3.7427192730856927e-06, "loss": 0.2119, "step": 11984 }, { "epoch": 5.64, "grad_norm": 0.635980705457827, "learning_rate": 3.7329942481220176e-06, "loss": 0.2432, "step": 11985 }, { "epoch": 5.640470588235294, "grad_norm": 0.46191629118628375, "learning_rate": 3.723281715078091e-06, "loss": 0.127, "step": 11986 }, { "epoch": 5.640941176470588, "grad_norm": 0.6044853807064937, "learning_rate": 3.7135816747833667e-06, "loss": 0.2715, "step": 11987 }, { "epoch": 5.641411764705882, "grad_norm": 0.5001476532574517, "learning_rate": 3.7038941280663147e-06, "loss": 0.2363, "step": 11988 }, { "epoch": 5.641882352941177, "grad_norm": 0.40955729380482786, "learning_rate": 3.6942190757542735e-06, "loss": 0.1709, "step": 11989 }, { "epoch": 5.642352941176471, "grad_norm": 0.4729724718456568, "learning_rate": 3.684556518673548e-06, "loss": 0.2451, "step": 11990 }, { "epoch": 5.642823529411765, "grad_norm": 0.5607342283008184, "learning_rate": 3.6749064576493793e-06, "loss": 0.3691, "step": 11991 }, { "epoch": 5.643294117647059, "grad_norm": 0.7047858287949033, "learning_rate": 3.6652688935058904e-06, "loss": 0.4551, "step": 11992 }, { "epoch": 5.643764705882353, "grad_norm": 0.8653538834818639, "learning_rate": 3.6556438270662236e-06, "loss": 0.6367, "step": 11993 }, { "epoch": 5.644235294117647, "grad_norm": 0.5571241301198953, "learning_rate": 3.6460312591524044e-06, "loss": 0.25, "step": 11994 }, { "epoch": 5.644705882352941, "grad_norm": 0.500963866944255, "learning_rate": 3.6364311905853594e-06, "loss": 0.252, "step": 11995 }, { "epoch": 5.645176470588235, "grad_norm": 0.4155707101135161, "learning_rate": 3.6268436221849994e-06, "loss": 0.1748, "step": 11996 }, { "epoch": 5.6456470588235295, "grad_norm": 0.5297725539191543, "learning_rate": 3.6172685547701695e-06, "loss": 0.2969, "step": 11997 }, { "epoch": 5.646117647058824, "grad_norm": 0.81735018127893, "learning_rate": 3.6077059891585826e-06, "loss": 0.5078, "step": 11998 }, { "epoch": 5.646588235294118, "grad_norm": 0.6349241051997072, "learning_rate": 3.598155926166968e-06, "loss": 0.4902, "step": 11999 }, { "epoch": 5.647058823529412, "grad_norm": 0.4396637376140548, "learning_rate": 3.5886183666109405e-06, "loss": 0.1348, "step": 12000 }, { "epoch": 5.647529411764706, "grad_norm": 0.4832368890353388, "learning_rate": 3.579093311305031e-06, "loss": 0.1758, "step": 12001 }, { "epoch": 5.648, "grad_norm": 0.5228566639552101, "learning_rate": 3.5695807610627402e-06, "loss": 0.1855, "step": 12002 }, { "epoch": 5.648470588235294, "grad_norm": 0.3960247027504582, "learning_rate": 3.5600807166965173e-06, "loss": 0.1533, "step": 12003 }, { "epoch": 5.648941176470588, "grad_norm": 0.4644065963817827, "learning_rate": 3.5505931790176635e-06, "loss": 0.1895, "step": 12004 }, { "epoch": 5.649411764705882, "grad_norm": 0.498731975192612, "learning_rate": 3.5411181488364802e-06, "loss": 0.2207, "step": 12005 }, { "epoch": 5.649882352941177, "grad_norm": 0.3980895004563765, "learning_rate": 3.531655626962204e-06, "loss": 0.25, "step": 12006 }, { "epoch": 5.650352941176471, "grad_norm": 0.4564647284718263, "learning_rate": 3.5222056142029387e-06, "loss": 0.2617, "step": 12007 }, { "epoch": 5.650823529411765, "grad_norm": 0.3565254672567311, "learning_rate": 3.5127681113658044e-06, "loss": 0.1826, "step": 12008 }, { "epoch": 5.651294117647058, "grad_norm": 0.6298248714632023, "learning_rate": 3.503343119256774e-06, "loss": 0.4785, "step": 12009 }, { "epoch": 5.651764705882353, "grad_norm": 0.5703304782914945, "learning_rate": 3.493930638680803e-06, "loss": 0.2197, "step": 12010 }, { "epoch": 5.652235294117647, "grad_norm": 0.40961587932954086, "learning_rate": 3.484530670441782e-06, "loss": 0.1367, "step": 12011 }, { "epoch": 5.652705882352941, "grad_norm": 0.22885709240181723, "learning_rate": 3.4751432153424687e-06, "loss": 0.0625, "step": 12012 }, { "epoch": 5.653176470588235, "grad_norm": 0.6325168579043832, "learning_rate": 3.465768274184638e-06, "loss": 0.2891, "step": 12013 }, { "epoch": 5.6536470588235295, "grad_norm": 0.5039589108359657, "learning_rate": 3.4564058477689503e-06, "loss": 0.1221, "step": 12014 }, { "epoch": 5.654117647058824, "grad_norm": 0.5801765003584817, "learning_rate": 3.4470559368949825e-06, "loss": 0.375, "step": 12015 }, { "epoch": 5.654588235294118, "grad_norm": 0.48620568742193887, "learning_rate": 3.4377185423612785e-06, "loss": 0.2461, "step": 12016 }, { "epoch": 5.655058823529412, "grad_norm": 0.45833398676603476, "learning_rate": 3.428393664965301e-06, "loss": 0.2393, "step": 12017 }, { "epoch": 5.6555294117647055, "grad_norm": 0.3774507791409837, "learning_rate": 3.4190813055034127e-06, "loss": 0.1309, "step": 12018 }, { "epoch": 5.656, "grad_norm": 0.33770644622514473, "learning_rate": 3.4097814647709775e-06, "loss": 0.126, "step": 12019 }, { "epoch": 5.656470588235294, "grad_norm": 0.5461424043417854, "learning_rate": 3.400494143562227e-06, "loss": 0.3027, "step": 12020 }, { "epoch": 5.656941176470588, "grad_norm": 0.3789632842941396, "learning_rate": 3.391219342670326e-06, "loss": 0.2119, "step": 12021 }, { "epoch": 5.657411764705882, "grad_norm": 0.4261590567369713, "learning_rate": 3.3819570628874414e-06, "loss": 0.1167, "step": 12022 }, { "epoch": 5.657882352941177, "grad_norm": 0.5331898713271266, "learning_rate": 3.3727073050045406e-06, "loss": 0.3613, "step": 12023 }, { "epoch": 5.658352941176471, "grad_norm": 0.46086162799669905, "learning_rate": 3.3634700698116577e-06, "loss": 0.1719, "step": 12024 }, { "epoch": 5.658823529411765, "grad_norm": 0.4999577774711279, "learning_rate": 3.354245358097679e-06, "loss": 0.1748, "step": 12025 }, { "epoch": 5.659294117647059, "grad_norm": 0.6204596154283015, "learning_rate": 3.345033170650441e-06, "loss": 0.3184, "step": 12026 }, { "epoch": 5.659764705882353, "grad_norm": 0.26968443083233107, "learning_rate": 3.3358335082566966e-06, "loss": 0.1543, "step": 12027 }, { "epoch": 5.660235294117647, "grad_norm": 0.47293853097122707, "learning_rate": 3.326646371702185e-06, "loss": 0.2539, "step": 12028 }, { "epoch": 5.660705882352941, "grad_norm": 0.5204053074237791, "learning_rate": 3.317471761771462e-06, "loss": 0.207, "step": 12029 }, { "epoch": 5.661176470588235, "grad_norm": 0.3313031685582311, "learning_rate": 3.30830967924815e-06, "loss": 0.0723, "step": 12030 }, { "epoch": 5.6616470588235295, "grad_norm": 0.43575834371521877, "learning_rate": 3.299160124914707e-06, "loss": 0.2422, "step": 12031 }, { "epoch": 5.662117647058824, "grad_norm": 0.9586901970487285, "learning_rate": 3.290023099552541e-06, "loss": 0.6172, "step": 12032 }, { "epoch": 5.662588235294118, "grad_norm": 0.5568542572510847, "learning_rate": 3.2808986039420115e-06, "loss": 0.249, "step": 12033 }, { "epoch": 5.663058823529412, "grad_norm": 0.44597698228813176, "learning_rate": 3.2717866388624114e-06, "loss": 0.2598, "step": 12034 }, { "epoch": 5.663529411764706, "grad_norm": 0.5047894223328173, "learning_rate": 3.2626872050919185e-06, "loss": 0.3555, "step": 12035 }, { "epoch": 5.664, "grad_norm": 0.3490068911295876, "learning_rate": 3.2536003034076774e-06, "loss": 0.085, "step": 12036 }, { "epoch": 5.664470588235294, "grad_norm": 0.497521085840229, "learning_rate": 3.2445259345857676e-06, "loss": 0.2695, "step": 12037 }, { "epoch": 5.664941176470588, "grad_norm": 0.6640350007550233, "learning_rate": 3.2354640994011693e-06, "loss": 0.1816, "step": 12038 }, { "epoch": 5.665411764705882, "grad_norm": 0.7115662412411268, "learning_rate": 3.22641479862783e-06, "loss": 0.5117, "step": 12039 }, { "epoch": 5.665882352941177, "grad_norm": 0.6841474904629444, "learning_rate": 3.2173780330385647e-06, "loss": 0.4082, "step": 12040 }, { "epoch": 5.666352941176471, "grad_norm": 0.45029651504187773, "learning_rate": 3.2083538034051726e-06, "loss": 0.2295, "step": 12041 }, { "epoch": 5.666823529411765, "grad_norm": 0.3744911645482491, "learning_rate": 3.199342110498404e-06, "loss": 0.1494, "step": 12042 }, { "epoch": 5.667294117647058, "grad_norm": 0.3283369671963902, "learning_rate": 3.1903429550878435e-06, "loss": 0.103, "step": 12043 }, { "epoch": 5.667764705882353, "grad_norm": 0.507131900894679, "learning_rate": 3.181356337942109e-06, "loss": 0.1875, "step": 12044 }, { "epoch": 5.668235294117647, "grad_norm": 0.5151086059557081, "learning_rate": 3.1723822598286875e-06, "loss": 0.1523, "step": 12045 }, { "epoch": 5.668705882352941, "grad_norm": 0.4033259054797192, "learning_rate": 3.163420721513982e-06, "loss": 0.1299, "step": 12046 }, { "epoch": 5.669176470588235, "grad_norm": 0.5174839367785453, "learning_rate": 3.1544717237633965e-06, "loss": 0.249, "step": 12047 }, { "epoch": 5.6696470588235295, "grad_norm": 0.3245579901723683, "learning_rate": 3.145535267341204e-06, "loss": 0.0903, "step": 12048 }, { "epoch": 5.670117647058824, "grad_norm": 0.7112848911392132, "learning_rate": 3.1366113530106097e-06, "loss": 0.332, "step": 12049 }, { "epoch": 5.670588235294118, "grad_norm": 0.6476774273661293, "learning_rate": 3.1276999815337544e-06, "loss": 0.3242, "step": 12050 }, { "epoch": 5.671058823529412, "grad_norm": 0.6899695444819193, "learning_rate": 3.118801153671746e-06, "loss": 0.5742, "step": 12051 }, { "epoch": 5.6715294117647055, "grad_norm": 0.5251030852583568, "learning_rate": 3.1099148701845435e-06, "loss": 0.3457, "step": 12052 }, { "epoch": 5.672, "grad_norm": 0.3473720444404334, "learning_rate": 3.1010411318311226e-06, "loss": 0.1206, "step": 12053 }, { "epoch": 5.672470588235294, "grad_norm": 0.35915798153776096, "learning_rate": 3.092179939369327e-06, "loss": 0.0767, "step": 12054 }, { "epoch": 5.672941176470588, "grad_norm": 0.5608086995261089, "learning_rate": 3.083331293555935e-06, "loss": 0.2412, "step": 12055 }, { "epoch": 5.673411764705882, "grad_norm": 0.5220484202890671, "learning_rate": 3.074495195146659e-06, "loss": 0.2969, "step": 12056 }, { "epoch": 5.673882352941177, "grad_norm": 0.5826354947020783, "learning_rate": 3.0656716448961784e-06, "loss": 0.1196, "step": 12057 }, { "epoch": 5.674352941176471, "grad_norm": 0.5126175380222956, "learning_rate": 3.05686064355804e-06, "loss": 0.1328, "step": 12058 }, { "epoch": 5.674823529411765, "grad_norm": 0.4969434858928136, "learning_rate": 3.048062191884759e-06, "loss": 0.2969, "step": 12059 }, { "epoch": 5.675294117647059, "grad_norm": 0.47277018268537413, "learning_rate": 3.039276290627768e-06, "loss": 0.1279, "step": 12060 }, { "epoch": 5.675764705882353, "grad_norm": 0.42738987941738266, "learning_rate": 3.0305029405373993e-06, "loss": 0.1777, "step": 12061 }, { "epoch": 5.676235294117647, "grad_norm": 0.4524331982770989, "learning_rate": 3.0217421423629706e-06, "loss": 0.3594, "step": 12062 }, { "epoch": 5.676705882352941, "grad_norm": 0.40933875318360613, "learning_rate": 3.0129938968526836e-06, "loss": 0.2656, "step": 12063 }, { "epoch": 5.677176470588235, "grad_norm": 0.44637236883649695, "learning_rate": 3.00425820475369e-06, "loss": 0.2051, "step": 12064 }, { "epoch": 5.6776470588235295, "grad_norm": 0.49375722009796563, "learning_rate": 2.9955350668120605e-06, "loss": 0.1553, "step": 12065 }, { "epoch": 5.678117647058824, "grad_norm": 0.5097468885611969, "learning_rate": 2.9868244837727815e-06, "loss": 0.3516, "step": 12066 }, { "epoch": 5.678588235294118, "grad_norm": 0.34058240198736195, "learning_rate": 2.978126456379809e-06, "loss": 0.0947, "step": 12067 }, { "epoch": 5.679058823529411, "grad_norm": 0.13956681426440057, "learning_rate": 2.969440985375948e-06, "loss": 0.0306, "step": 12068 }, { "epoch": 5.679529411764706, "grad_norm": 0.5979235157710421, "learning_rate": 2.960768071503006e-06, "loss": 0.5898, "step": 12069 }, { "epoch": 5.68, "grad_norm": 0.7350398377617864, "learning_rate": 2.9521077155017224e-06, "loss": 0.4219, "step": 12070 }, { "epoch": 5.680470588235294, "grad_norm": 0.48131342277594524, "learning_rate": 2.9434599181116736e-06, "loss": 0.2002, "step": 12071 }, { "epoch": 5.680941176470588, "grad_norm": 0.8113244595630071, "learning_rate": 2.934824680071468e-06, "loss": 0.6406, "step": 12072 }, { "epoch": 5.681411764705882, "grad_norm": 0.6844909334730904, "learning_rate": 2.9262020021185993e-06, "loss": 0.5586, "step": 12073 }, { "epoch": 5.681882352941177, "grad_norm": 0.24209282479538705, "learning_rate": 2.917591884989462e-06, "loss": 0.0613, "step": 12074 }, { "epoch": 5.682352941176471, "grad_norm": 0.7399393080918008, "learning_rate": 2.908994329419401e-06, "loss": 0.3008, "step": 12075 }, { "epoch": 5.682823529411765, "grad_norm": 0.5575794582505937, "learning_rate": 2.9004093361427282e-06, "loss": 0.2256, "step": 12076 }, { "epoch": 5.683294117647058, "grad_norm": 0.25712859864501486, "learning_rate": 2.891836905892592e-06, "loss": 0.0698, "step": 12077 }, { "epoch": 5.683764705882353, "grad_norm": 0.7103069067871496, "learning_rate": 2.883277039401155e-06, "loss": 0.375, "step": 12078 }, { "epoch": 5.684235294117647, "grad_norm": 0.5000488037998555, "learning_rate": 2.874729737399467e-06, "loss": 0.3359, "step": 12079 }, { "epoch": 5.684705882352941, "grad_norm": 0.35855518585090723, "learning_rate": 2.866195000617494e-06, "loss": 0.1211, "step": 12080 }, { "epoch": 5.685176470588235, "grad_norm": 0.6108628240255384, "learning_rate": 2.857672829784169e-06, "loss": 0.3398, "step": 12081 }, { "epoch": 5.6856470588235295, "grad_norm": 0.4469657967824938, "learning_rate": 2.84916322562731e-06, "loss": 0.2676, "step": 12082 }, { "epoch": 5.686117647058824, "grad_norm": 0.6575351338439097, "learning_rate": 2.840666188873669e-06, "loss": 0.4219, "step": 12083 }, { "epoch": 5.686588235294118, "grad_norm": 0.26647952972429295, "learning_rate": 2.8321817202489493e-06, "loss": 0.05, "step": 12084 }, { "epoch": 5.687058823529412, "grad_norm": 0.5770155379224885, "learning_rate": 2.8237098204777874e-06, "loss": 0.2949, "step": 12085 }, { "epoch": 5.6875294117647055, "grad_norm": 0.3551813511996362, "learning_rate": 2.815250490283688e-06, "loss": 0.1484, "step": 12086 }, { "epoch": 5.688, "grad_norm": 0.5250741827245735, "learning_rate": 2.8068037303891233e-06, "loss": 0.3184, "step": 12087 }, { "epoch": 5.688470588235294, "grad_norm": 0.5941053759749808, "learning_rate": 2.7983695415155162e-06, "loss": 0.1396, "step": 12088 }, { "epoch": 5.688941176470588, "grad_norm": 0.2802524537155189, "learning_rate": 2.789947924383157e-06, "loss": 0.1147, "step": 12089 }, { "epoch": 5.689411764705882, "grad_norm": 0.5384898174898404, "learning_rate": 2.78153887971132e-06, "loss": 0.2578, "step": 12090 }, { "epoch": 5.689882352941177, "grad_norm": 0.43251574724184044, "learning_rate": 2.773142408218165e-06, "loss": 0.1816, "step": 12091 }, { "epoch": 5.690352941176471, "grad_norm": 0.5728740399359794, "learning_rate": 2.7647585106207682e-06, "loss": 0.293, "step": 12092 }, { "epoch": 5.690823529411765, "grad_norm": 0.6073149882614269, "learning_rate": 2.7563871876352063e-06, "loss": 0.3379, "step": 12093 }, { "epoch": 5.691294117647059, "grad_norm": 0.5461617454699248, "learning_rate": 2.7480284399763907e-06, "loss": 0.2676, "step": 12094 }, { "epoch": 5.691764705882353, "grad_norm": 0.46604349679922924, "learning_rate": 2.7396822683582177e-06, "loss": 0.2852, "step": 12095 }, { "epoch": 5.692235294117647, "grad_norm": 0.4530844302784051, "learning_rate": 2.731348673493483e-06, "loss": 0.1895, "step": 12096 }, { "epoch": 5.692705882352941, "grad_norm": 0.28574419216535385, "learning_rate": 2.723027656093918e-06, "loss": 0.0796, "step": 12097 }, { "epoch": 5.693176470588235, "grad_norm": 0.8061415909181343, "learning_rate": 2.714719216870187e-06, "loss": 0.3438, "step": 12098 }, { "epoch": 5.6936470588235295, "grad_norm": 0.4943771645667207, "learning_rate": 2.706423356531873e-06, "loss": 0.2461, "step": 12099 }, { "epoch": 5.694117647058824, "grad_norm": 0.21984690198858198, "learning_rate": 2.6981400757874584e-06, "loss": 0.0815, "step": 12100 }, { "epoch": 5.694588235294118, "grad_norm": 0.65317167375083, "learning_rate": 2.6898693753443945e-06, "loss": 0.3223, "step": 12101 }, { "epoch": 5.695058823529411, "grad_norm": 0.6866031271435872, "learning_rate": 2.68161125590905e-06, "loss": 0.3184, "step": 12102 }, { "epoch": 5.6955294117647055, "grad_norm": 0.11733247779144308, "learning_rate": 2.673365718186693e-06, "loss": 0.0243, "step": 12103 }, { "epoch": 5.696, "grad_norm": 0.29942170751773306, "learning_rate": 2.6651327628815277e-06, "loss": 0.0586, "step": 12104 }, { "epoch": 5.696470588235294, "grad_norm": 0.6303434058092715, "learning_rate": 2.656912390696708e-06, "loss": 0.2441, "step": 12105 }, { "epoch": 5.696941176470588, "grad_norm": 0.3646874097742741, "learning_rate": 2.6487046023342563e-06, "loss": 0.0996, "step": 12106 }, { "epoch": 5.697411764705882, "grad_norm": 0.24697093205776474, "learning_rate": 2.6405093984952275e-06, "loss": 0.0427, "step": 12107 }, { "epoch": 5.697882352941177, "grad_norm": 0.5022137239008211, "learning_rate": 2.6323267798794625e-06, "loss": 0.2275, "step": 12108 }, { "epoch": 5.698352941176471, "grad_norm": 0.7523895939902674, "learning_rate": 2.624156747185835e-06, "loss": 0.4922, "step": 12109 }, { "epoch": 5.698823529411765, "grad_norm": 0.3239804582667825, "learning_rate": 2.615999301112104e-06, "loss": 0.0942, "step": 12110 }, { "epoch": 5.699294117647058, "grad_norm": 0.37476778593665966, "learning_rate": 2.6078544423549286e-06, "loss": 0.1992, "step": 12111 }, { "epoch": 5.699764705882353, "grad_norm": 0.562216957019733, "learning_rate": 2.599722171609936e-06, "loss": 0.3242, "step": 12112 }, { "epoch": 5.700235294117647, "grad_norm": 0.5481714086855868, "learning_rate": 2.591602489571687e-06, "loss": 0.3125, "step": 12113 }, { "epoch": 5.700705882352941, "grad_norm": 0.29186057818694316, "learning_rate": 2.5834953969336105e-06, "loss": 0.0449, "step": 12114 }, { "epoch": 5.701176470588235, "grad_norm": 0.7809346718612641, "learning_rate": 2.5754008943880855e-06, "loss": 0.457, "step": 12115 }, { "epoch": 5.7016470588235295, "grad_norm": 0.5379887543773632, "learning_rate": 2.5673189826264596e-06, "loss": 0.2305, "step": 12116 }, { "epoch": 5.702117647058824, "grad_norm": 0.7797124503499385, "learning_rate": 2.559249662338947e-06, "loss": 0.2344, "step": 12117 }, { "epoch": 5.702588235294118, "grad_norm": 0.410093746773438, "learning_rate": 2.5511929342146964e-06, "loss": 0.1074, "step": 12118 }, { "epoch": 5.703058823529412, "grad_norm": 0.16151903785552704, "learning_rate": 2.543148798941791e-06, "loss": 0.0669, "step": 12119 }, { "epoch": 5.7035294117647055, "grad_norm": 0.5781681793728657, "learning_rate": 2.535117257207264e-06, "loss": 0.3125, "step": 12120 }, { "epoch": 5.704, "grad_norm": 0.7246777956751975, "learning_rate": 2.5270983096970344e-06, "loss": 0.3262, "step": 12121 }, { "epoch": 5.704470588235294, "grad_norm": 0.7753739733342225, "learning_rate": 2.5190919570959533e-06, "loss": 0.4512, "step": 12122 }, { "epoch": 5.704941176470588, "grad_norm": 0.6468680234081458, "learning_rate": 2.511098200087808e-06, "loss": 0.4727, "step": 12123 }, { "epoch": 5.705411764705882, "grad_norm": 0.3552072051756507, "learning_rate": 2.503117039355318e-06, "loss": 0.1475, "step": 12124 }, { "epoch": 5.705882352941177, "grad_norm": 0.7609821889367934, "learning_rate": 2.4951484755800886e-06, "loss": 0.4609, "step": 12125 }, { "epoch": 5.706352941176471, "grad_norm": 0.322024128829307, "learning_rate": 2.487192509442676e-06, "loss": 0.1377, "step": 12126 }, { "epoch": 5.706823529411765, "grad_norm": 0.30444562371773926, "learning_rate": 2.4792491416226024e-06, "loss": 0.1172, "step": 12127 }, { "epoch": 5.707294117647059, "grad_norm": 0.32847763143357867, "learning_rate": 2.471318372798209e-06, "loss": 0.0732, "step": 12128 }, { "epoch": 5.707764705882353, "grad_norm": 0.32958061154823587, "learning_rate": 2.4634002036468536e-06, "loss": 0.1543, "step": 12129 }, { "epoch": 5.708235294117647, "grad_norm": 0.6378842906499143, "learning_rate": 2.4554946348447958e-06, "loss": 0.1729, "step": 12130 }, { "epoch": 5.708705882352941, "grad_norm": 0.29706520226565425, "learning_rate": 2.447601667067195e-06, "loss": 0.0972, "step": 12131 }, { "epoch": 5.709176470588235, "grad_norm": 0.3508010401657439, "learning_rate": 2.4397213009881456e-06, "loss": 0.1006, "step": 12132 }, { "epoch": 5.7096470588235295, "grad_norm": 0.6088054209394933, "learning_rate": 2.4318535372806757e-06, "loss": 0.332, "step": 12133 }, { "epoch": 5.710117647058824, "grad_norm": 0.5292000965123478, "learning_rate": 2.4239983766167315e-06, "loss": 0.1777, "step": 12134 }, { "epoch": 5.710588235294118, "grad_norm": 0.4018747398165247, "learning_rate": 2.416155819667193e-06, "loss": 0.1562, "step": 12135 }, { "epoch": 5.711058823529411, "grad_norm": 0.8352781482740472, "learning_rate": 2.4083258671018403e-06, "loss": 0.5977, "step": 12136 }, { "epoch": 5.7115294117647055, "grad_norm": 0.3388025069322309, "learning_rate": 2.4005085195893725e-06, "loss": 0.0898, "step": 12137 }, { "epoch": 5.712, "grad_norm": 0.39697493074634804, "learning_rate": 2.3927037777974722e-06, "loss": 0.0859, "step": 12138 }, { "epoch": 5.712470588235294, "grad_norm": 0.45630520642085, "learning_rate": 2.384911642392673e-06, "loss": 0.2373, "step": 12139 }, { "epoch": 5.712941176470588, "grad_norm": 0.519053885852047, "learning_rate": 2.377132114040459e-06, "loss": 0.1826, "step": 12140 }, { "epoch": 5.713411764705882, "grad_norm": 0.5210005661005087, "learning_rate": 2.369365193405248e-06, "loss": 0.3633, "step": 12141 }, { "epoch": 5.713882352941177, "grad_norm": 0.5221783429852981, "learning_rate": 2.3616108811503775e-06, "loss": 0.2305, "step": 12142 }, { "epoch": 5.714352941176471, "grad_norm": 0.5110271065649945, "learning_rate": 2.3538691779381003e-06, "loss": 0.2852, "step": 12143 }, { "epoch": 5.714823529411765, "grad_norm": 0.33722124934496756, "learning_rate": 2.346140084429604e-06, "loss": 0.1523, "step": 12144 }, { "epoch": 5.715294117647058, "grad_norm": 0.27107632710298835, "learning_rate": 2.3384236012849778e-06, "loss": 0.0635, "step": 12145 }, { "epoch": 5.715764705882353, "grad_norm": 0.2613963404493113, "learning_rate": 2.330719729163244e-06, "loss": 0.0771, "step": 12146 }, { "epoch": 5.716235294117647, "grad_norm": 0.45943566212921844, "learning_rate": 2.32302846872236e-06, "loss": 0.167, "step": 12147 }, { "epoch": 5.716705882352941, "grad_norm": 0.6635706261787108, "learning_rate": 2.3153498206192e-06, "loss": 0.5039, "step": 12148 }, { "epoch": 5.717176470588235, "grad_norm": 0.20921119481689726, "learning_rate": 2.3076837855095563e-06, "loss": 0.0527, "step": 12149 }, { "epoch": 5.7176470588235295, "grad_norm": 0.7359950976557599, "learning_rate": 2.3000303640481386e-06, "loss": 0.582, "step": 12150 }, { "epoch": 5.718117647058824, "grad_norm": 0.33628535446463625, "learning_rate": 2.292389556888591e-06, "loss": 0.0898, "step": 12151 }, { "epoch": 5.718588235294118, "grad_norm": 0.7443342613973044, "learning_rate": 2.2847613646834905e-06, "loss": 0.4453, "step": 12152 }, { "epoch": 5.719058823529412, "grad_norm": 0.43931448869420736, "learning_rate": 2.2771457880842834e-06, "loss": 0.1904, "step": 12153 }, { "epoch": 5.7195294117647055, "grad_norm": 0.32357027039921454, "learning_rate": 2.2695428277414165e-06, "loss": 0.1377, "step": 12154 }, { "epoch": 5.72, "grad_norm": 0.6622277377573879, "learning_rate": 2.261952484304219e-06, "loss": 0.3555, "step": 12155 }, { "epoch": 5.720470588235294, "grad_norm": 0.2710427323437237, "learning_rate": 2.254374758420924e-06, "loss": 0.1279, "step": 12156 }, { "epoch": 5.720941176470588, "grad_norm": 0.37547754073168593, "learning_rate": 2.2468096507386956e-06, "loss": 0.1396, "step": 12157 }, { "epoch": 5.721411764705882, "grad_norm": 0.4647327981940749, "learning_rate": 2.239257161903668e-06, "loss": 0.1855, "step": 12158 }, { "epoch": 5.721882352941177, "grad_norm": 0.5115990514247123, "learning_rate": 2.231717292560825e-06, "loss": 0.2031, "step": 12159 }, { "epoch": 5.722352941176471, "grad_norm": 0.5444664453639295, "learning_rate": 2.224190043354135e-06, "loss": 0.2832, "step": 12160 }, { "epoch": 5.722823529411765, "grad_norm": 0.03522604095842789, "learning_rate": 2.2166754149264667e-06, "loss": 0.0027, "step": 12161 }, { "epoch": 5.723294117647059, "grad_norm": 0.7551433930330766, "learning_rate": 2.2091734079195565e-06, "loss": 0.5117, "step": 12162 }, { "epoch": 5.723764705882353, "grad_norm": 0.13016570211544318, "learning_rate": 2.2016840229741593e-06, "loss": 0.0332, "step": 12163 }, { "epoch": 5.724235294117647, "grad_norm": 0.29574720178387975, "learning_rate": 2.1942072607299135e-06, "loss": 0.0859, "step": 12164 }, { "epoch": 5.724705882352941, "grad_norm": 0.45320392467371506, "learning_rate": 2.1867431218253406e-06, "loss": 0.2637, "step": 12165 }, { "epoch": 5.725176470588235, "grad_norm": 0.4718538406439423, "learning_rate": 2.179291606897915e-06, "loss": 0.3242, "step": 12166 }, { "epoch": 5.7256470588235295, "grad_norm": 0.483417656645847, "learning_rate": 2.171852716584044e-06, "loss": 0.2314, "step": 12167 }, { "epoch": 5.726117647058824, "grad_norm": 0.745773368467488, "learning_rate": 2.1644264515190524e-06, "loss": 0.2734, "step": 12168 }, { "epoch": 5.726588235294118, "grad_norm": 0.4771025498572797, "learning_rate": 2.1570128123371667e-06, "loss": 0.2412, "step": 12169 }, { "epoch": 5.727058823529411, "grad_norm": 0.5223363604702245, "learning_rate": 2.1496117996715636e-06, "loss": 0.3184, "step": 12170 }, { "epoch": 5.7275294117647055, "grad_norm": 0.4847401383721601, "learning_rate": 2.142223414154287e-06, "loss": 0.2559, "step": 12171 }, { "epoch": 5.728, "grad_norm": 0.5648549265622917, "learning_rate": 2.1348476564163997e-06, "loss": 0.3398, "step": 12172 }, { "epoch": 5.728470588235294, "grad_norm": 0.5072341316818811, "learning_rate": 2.1274845270877804e-06, "loss": 0.2812, "step": 12173 }, { "epoch": 5.728941176470588, "grad_norm": 0.5640977063497379, "learning_rate": 2.1201340267972933e-06, "loss": 0.2871, "step": 12174 }, { "epoch": 5.729411764705882, "grad_norm": 0.38041716382247204, "learning_rate": 2.1127961561727193e-06, "loss": 0.1211, "step": 12175 }, { "epoch": 5.729882352941177, "grad_norm": 0.26500916784767076, "learning_rate": 2.1054709158407235e-06, "loss": 0.0845, "step": 12176 }, { "epoch": 5.730352941176471, "grad_norm": 0.7424068268133911, "learning_rate": 2.0981583064269226e-06, "loss": 0.4609, "step": 12177 }, { "epoch": 5.730823529411765, "grad_norm": 0.3920752974949613, "learning_rate": 2.0908583285558834e-06, "loss": 0.1934, "step": 12178 }, { "epoch": 5.731294117647058, "grad_norm": 0.5893258451164466, "learning_rate": 2.083570982851007e-06, "loss": 0.4316, "step": 12179 }, { "epoch": 5.731764705882353, "grad_norm": 0.4691805797327912, "learning_rate": 2.0762962699346965e-06, "loss": 0.3203, "step": 12180 }, { "epoch": 5.732235294117647, "grad_norm": 0.5623807655134564, "learning_rate": 2.06903419042827e-06, "loss": 0.2695, "step": 12181 }, { "epoch": 5.732705882352941, "grad_norm": 0.23733882101549544, "learning_rate": 2.0617847449519e-06, "loss": 0.0659, "step": 12182 }, { "epoch": 5.733176470588235, "grad_norm": 0.4126952444511235, "learning_rate": 2.054547934124756e-06, "loss": 0.1147, "step": 12183 }, { "epoch": 5.7336470588235295, "grad_norm": 0.3766685520168285, "learning_rate": 2.047323758564878e-06, "loss": 0.1602, "step": 12184 }, { "epoch": 5.734117647058824, "grad_norm": 0.5583455522338165, "learning_rate": 2.040112218889256e-06, "loss": 0.2832, "step": 12185 }, { "epoch": 5.734588235294118, "grad_norm": 0.4541276560048125, "learning_rate": 2.0329133157137965e-06, "loss": 0.1787, "step": 12186 }, { "epoch": 5.735058823529412, "grad_norm": 0.5256644901530253, "learning_rate": 2.0257270496533086e-06, "loss": 0.1514, "step": 12187 }, { "epoch": 5.7355294117647055, "grad_norm": 0.4986742277720515, "learning_rate": 2.0185534213215506e-06, "loss": 0.2715, "step": 12188 }, { "epoch": 5.736, "grad_norm": 0.3034618749817238, "learning_rate": 2.011392431331166e-06, "loss": 0.1758, "step": 12189 }, { "epoch": 5.736470588235294, "grad_norm": 0.11141349999359632, "learning_rate": 2.0042440802937485e-06, "loss": 0.0225, "step": 12190 }, { "epoch": 5.736941176470588, "grad_norm": 0.2813496422771534, "learning_rate": 1.9971083688197943e-06, "loss": 0.0991, "step": 12191 }, { "epoch": 5.737411764705882, "grad_norm": 0.14204868083062136, "learning_rate": 1.989985297518748e-06, "loss": 0.0515, "step": 12192 }, { "epoch": 5.737882352941177, "grad_norm": 0.4478902815698624, "learning_rate": 1.9828748669989405e-06, "loss": 0.1562, "step": 12193 }, { "epoch": 5.738352941176471, "grad_norm": 0.5039963967393996, "learning_rate": 1.9757770778676183e-06, "loss": 0.1348, "step": 12194 }, { "epoch": 5.738823529411764, "grad_norm": 0.1675478250989818, "learning_rate": 1.9686919307310145e-06, "loss": 0.0474, "step": 12195 }, { "epoch": 5.739294117647058, "grad_norm": 0.7159731606963625, "learning_rate": 1.961619426194194e-06, "loss": 0.3984, "step": 12196 }, { "epoch": 5.739764705882353, "grad_norm": 0.4449707249163169, "learning_rate": 1.9545595648612078e-06, "loss": 0.1396, "step": 12197 }, { "epoch": 5.740235294117647, "grad_norm": 0.2918061400347441, "learning_rate": 1.9475123473349897e-06, "loss": 0.1377, "step": 12198 }, { "epoch": 5.740705882352941, "grad_norm": 0.6977217151188925, "learning_rate": 1.9404777742174083e-06, "loss": 0.4336, "step": 12199 }, { "epoch": 5.741176470588235, "grad_norm": 0.46585686963662787, "learning_rate": 1.9334558461092663e-06, "loss": 0.2129, "step": 12200 }, { "epoch": 5.7416470588235295, "grad_norm": 0.6234445120519253, "learning_rate": 1.9264465636102342e-06, "loss": 0.3184, "step": 12201 }, { "epoch": 5.742117647058824, "grad_norm": 0.6014594761725082, "learning_rate": 1.9194499273189834e-06, "loss": 0.2432, "step": 12202 }, { "epoch": 5.742588235294118, "grad_norm": 0.19004406966578155, "learning_rate": 1.912465937833035e-06, "loss": 0.0309, "step": 12203 }, { "epoch": 5.743058823529411, "grad_norm": 0.5880890690760342, "learning_rate": 1.9054945957488466e-06, "loss": 0.3184, "step": 12204 }, { "epoch": 5.7435294117647056, "grad_norm": 0.6369521811291807, "learning_rate": 1.898535901661824e-06, "loss": 0.2324, "step": 12205 }, { "epoch": 5.744, "grad_norm": 0.15462785939622553, "learning_rate": 1.8915898561662923e-06, "loss": 0.0322, "step": 12206 }, { "epoch": 5.744470588235294, "grad_norm": 0.3729162100418688, "learning_rate": 1.884656459855427e-06, "loss": 0.1221, "step": 12207 }, { "epoch": 5.744941176470588, "grad_norm": 0.6685456244167198, "learning_rate": 1.8777357133214044e-06, "loss": 0.3848, "step": 12208 }, { "epoch": 5.745411764705882, "grad_norm": 0.7737473067911042, "learning_rate": 1.8708276171553017e-06, "loss": 0.3262, "step": 12209 }, { "epoch": 5.745882352941177, "grad_norm": 0.6143145038349767, "learning_rate": 1.86393217194708e-06, "loss": 0.1621, "step": 12210 }, { "epoch": 5.746352941176471, "grad_norm": 0.4245535563446793, "learning_rate": 1.857049378285652e-06, "loss": 0.2461, "step": 12211 }, { "epoch": 5.746823529411765, "grad_norm": 0.3429117567492918, "learning_rate": 1.8501792367588642e-06, "loss": 0.1147, "step": 12212 }, { "epoch": 5.747294117647058, "grad_norm": 0.7705674512696286, "learning_rate": 1.84332174795343e-06, "loss": 0.457, "step": 12213 }, { "epoch": 5.747764705882353, "grad_norm": 0.6287704953058229, "learning_rate": 1.8364769124550149e-06, "loss": 0.4219, "step": 12214 }, { "epoch": 5.748235294117647, "grad_norm": 0.5944119031425492, "learning_rate": 1.8296447308482176e-06, "loss": 0.2002, "step": 12215 }, { "epoch": 5.748705882352941, "grad_norm": 0.8384113585925869, "learning_rate": 1.8228252037165215e-06, "loss": 0.3359, "step": 12216 }, { "epoch": 5.749176470588235, "grad_norm": 0.5041982327860667, "learning_rate": 1.8160183316423605e-06, "loss": 0.2119, "step": 12217 }, { "epoch": 5.7496470588235296, "grad_norm": 0.7438117870881507, "learning_rate": 1.8092241152070696e-06, "loss": 0.5586, "step": 12218 }, { "epoch": 5.750117647058824, "grad_norm": 0.7457573776538415, "learning_rate": 1.8024425549909183e-06, "loss": 0.3789, "step": 12219 }, { "epoch": 5.750588235294118, "grad_norm": 0.45708263839906493, "learning_rate": 1.795673651573093e-06, "loss": 0.2021, "step": 12220 }, { "epoch": 5.751058823529412, "grad_norm": 0.5447227122786078, "learning_rate": 1.788917405531648e-06, "loss": 0.3145, "step": 12221 }, { "epoch": 5.751529411764706, "grad_norm": 0.6536702907075699, "learning_rate": 1.7821738174436384e-06, "loss": 0.4551, "step": 12222 }, { "epoch": 5.752, "grad_norm": 0.31020081651991277, "learning_rate": 1.7754428878849868e-06, "loss": 0.1045, "step": 12223 }, { "epoch": 5.752470588235294, "grad_norm": 0.44921901989390556, "learning_rate": 1.7687246174305502e-06, "loss": 0.1973, "step": 12224 }, { "epoch": 5.752941176470588, "grad_norm": 0.4728506689203958, "learning_rate": 1.762019006654103e-06, "loss": 0.1455, "step": 12225 }, { "epoch": 5.7534117647058824, "grad_norm": 0.48622422602581494, "learning_rate": 1.7553260561283533e-06, "loss": 0.2256, "step": 12226 }, { "epoch": 5.753882352941177, "grad_norm": 0.23684325472463363, "learning_rate": 1.748645766424861e-06, "loss": 0.0486, "step": 12227 }, { "epoch": 5.754352941176471, "grad_norm": 0.4672625894198133, "learning_rate": 1.7419781381142028e-06, "loss": 0.2969, "step": 12228 }, { "epoch": 5.754823529411764, "grad_norm": 0.4842341073729376, "learning_rate": 1.7353231717658233e-06, "loss": 0.2793, "step": 12229 }, { "epoch": 5.7552941176470584, "grad_norm": 0.3901441218373375, "learning_rate": 1.728680867948068e-06, "loss": 0.1855, "step": 12230 }, { "epoch": 5.755764705882353, "grad_norm": 0.4074650239295036, "learning_rate": 1.7220512272282328e-06, "loss": 0.1699, "step": 12231 }, { "epoch": 5.756235294117647, "grad_norm": 0.5150630750519997, "learning_rate": 1.7154342501725316e-06, "loss": 0.3711, "step": 12232 }, { "epoch": 5.756705882352941, "grad_norm": 0.33534879698899384, "learning_rate": 1.708829937346079e-06, "loss": 0.1133, "step": 12233 }, { "epoch": 5.757176470588235, "grad_norm": 0.7976485207925256, "learning_rate": 1.7022382893129072e-06, "loss": 0.5742, "step": 12234 }, { "epoch": 5.75764705882353, "grad_norm": 0.47585639974484273, "learning_rate": 1.6956593066359825e-06, "loss": 0.3262, "step": 12235 }, { "epoch": 5.758117647058824, "grad_norm": 0.43518081648335166, "learning_rate": 1.6890929898771722e-06, "loss": 0.293, "step": 12236 }, { "epoch": 5.758588235294118, "grad_norm": 0.557704660070932, "learning_rate": 1.6825393395972942e-06, "loss": 0.2676, "step": 12237 }, { "epoch": 5.759058823529411, "grad_norm": 0.31752294935082265, "learning_rate": 1.6759983563560508e-06, "loss": 0.1182, "step": 12238 }, { "epoch": 5.759529411764706, "grad_norm": 0.48509471206083543, "learning_rate": 1.669470040712062e-06, "loss": 0.2363, "step": 12239 }, { "epoch": 5.76, "grad_norm": 0.4677025449428635, "learning_rate": 1.6629543932228817e-06, "loss": 0.1289, "step": 12240 }, { "epoch": 5.760470588235294, "grad_norm": 0.5634749991828569, "learning_rate": 1.6564514144449982e-06, "loss": 0.1943, "step": 12241 }, { "epoch": 5.760941176470588, "grad_norm": 0.5462680440984228, "learning_rate": 1.6499611049337835e-06, "loss": 0.3203, "step": 12242 }, { "epoch": 5.7614117647058825, "grad_norm": 0.6266636970205784, "learning_rate": 1.6434834652435447e-06, "loss": 0.3184, "step": 12243 }, { "epoch": 5.761882352941177, "grad_norm": 0.4309019837957245, "learning_rate": 1.6370184959275058e-06, "loss": 0.2217, "step": 12244 }, { "epoch": 5.762352941176471, "grad_norm": 0.13169019969533222, "learning_rate": 1.6305661975377916e-06, "loss": 0.0228, "step": 12245 }, { "epoch": 5.762823529411765, "grad_norm": 0.4515051724777316, "learning_rate": 1.6241265706254947e-06, "loss": 0.1729, "step": 12246 }, { "epoch": 5.7632941176470585, "grad_norm": 0.29917164833507953, "learning_rate": 1.617699615740542e-06, "loss": 0.0693, "step": 12247 }, { "epoch": 5.763764705882353, "grad_norm": 0.32857847025021103, "learning_rate": 1.6112853334318775e-06, "loss": 0.167, "step": 12248 }, { "epoch": 5.764235294117647, "grad_norm": 0.1465583171977418, "learning_rate": 1.6048837242472634e-06, "loss": 0.0294, "step": 12249 }, { "epoch": 5.764705882352941, "grad_norm": 0.4438370696090344, "learning_rate": 1.598494788733462e-06, "loss": 0.1758, "step": 12250 }, { "epoch": 5.765176470588235, "grad_norm": 0.3349671979980706, "learning_rate": 1.5921185274361203e-06, "loss": 0.0747, "step": 12251 }, { "epoch": 5.76564705882353, "grad_norm": 0.6404510572605647, "learning_rate": 1.585754940899786e-06, "loss": 0.2637, "step": 12252 }, { "epoch": 5.766117647058824, "grad_norm": 0.2589914944854131, "learning_rate": 1.5794040296679411e-06, "loss": 0.0889, "step": 12253 }, { "epoch": 5.766588235294118, "grad_norm": 0.6252455215191695, "learning_rate": 1.5730657942830183e-06, "loss": 0.4609, "step": 12254 }, { "epoch": 5.767058823529412, "grad_norm": 0.4160446722684372, "learning_rate": 1.5667402352862846e-06, "loss": 0.1182, "step": 12255 }, { "epoch": 5.767529411764706, "grad_norm": 0.6752942823463716, "learning_rate": 1.5604273532179913e-06, "loss": 0.4688, "step": 12256 }, { "epoch": 5.768, "grad_norm": 0.7377441149457123, "learning_rate": 1.554127148617307e-06, "loss": 0.5273, "step": 12257 }, { "epoch": 5.768470588235294, "grad_norm": 0.4734693786555287, "learning_rate": 1.5478396220222678e-06, "loss": 0.2812, "step": 12258 }, { "epoch": 5.768941176470588, "grad_norm": 0.47622426654800903, "learning_rate": 1.5415647739698944e-06, "loss": 0.2617, "step": 12259 }, { "epoch": 5.7694117647058825, "grad_norm": 0.7375281417466965, "learning_rate": 1.5353026049960748e-06, "loss": 0.3789, "step": 12260 }, { "epoch": 5.769882352941177, "grad_norm": 0.5835535223038892, "learning_rate": 1.5290531156356146e-06, "loss": 0.418, "step": 12261 }, { "epoch": 5.770352941176471, "grad_norm": 0.46196712000703305, "learning_rate": 1.5228163064222698e-06, "loss": 0.1924, "step": 12262 }, { "epoch": 5.770823529411764, "grad_norm": 0.654344038069269, "learning_rate": 1.516592177888698e-06, "loss": 0.4922, "step": 12263 }, { "epoch": 5.7712941176470585, "grad_norm": 0.5377026821351462, "learning_rate": 1.5103807305664573e-06, "loss": 0.3672, "step": 12264 }, { "epoch": 5.771764705882353, "grad_norm": 0.5205469660209572, "learning_rate": 1.5041819649860399e-06, "loss": 0.2598, "step": 12265 }, { "epoch": 5.772235294117647, "grad_norm": 0.4205451163334439, "learning_rate": 1.4979958816768556e-06, "loss": 0.1738, "step": 12266 }, { "epoch": 5.772705882352941, "grad_norm": 0.36070385166324376, "learning_rate": 1.4918224811672153e-06, "loss": 0.2217, "step": 12267 }, { "epoch": 5.773176470588235, "grad_norm": 0.622044825597766, "learning_rate": 1.485661763984347e-06, "loss": 0.3535, "step": 12268 }, { "epoch": 5.77364705882353, "grad_norm": 0.4644966781753564, "learning_rate": 1.479513730654447e-06, "loss": 0.1016, "step": 12269 }, { "epoch": 5.774117647058824, "grad_norm": 0.413708547478666, "learning_rate": 1.4733783817025613e-06, "loss": 0.1504, "step": 12270 }, { "epoch": 5.774588235294118, "grad_norm": 0.5589088331525032, "learning_rate": 1.4672557176526712e-06, "loss": 0.2988, "step": 12271 }, { "epoch": 5.775058823529411, "grad_norm": 0.6455315115046895, "learning_rate": 1.4611457390276915e-06, "loss": 0.3203, "step": 12272 }, { "epoch": 5.775529411764706, "grad_norm": 0.4695099281499, "learning_rate": 1.4550484463494384e-06, "loss": 0.2246, "step": 12273 }, { "epoch": 5.776, "grad_norm": 0.298627827883168, "learning_rate": 1.4489638401386783e-06, "loss": 0.0801, "step": 12274 }, { "epoch": 5.776470588235294, "grad_norm": 0.5639782450763676, "learning_rate": 1.442891920915029e-06, "loss": 0.3359, "step": 12275 }, { "epoch": 5.776941176470588, "grad_norm": 0.5010623966373654, "learning_rate": 1.436832689197076e-06, "loss": 0.3438, "step": 12276 }, { "epoch": 5.7774117647058825, "grad_norm": 0.392129256019912, "learning_rate": 1.4307861455023218e-06, "loss": 0.1191, "step": 12277 }, { "epoch": 5.777882352941177, "grad_norm": 0.20286754578159508, "learning_rate": 1.4247522903471531e-06, "loss": 0.1514, "step": 12278 }, { "epoch": 5.778352941176471, "grad_norm": 0.6758256329139608, "learning_rate": 1.4187311242468914e-06, "loss": 0.5, "step": 12279 }, { "epoch": 5.778823529411765, "grad_norm": 1.059648039492914, "learning_rate": 1.4127226477157915e-06, "loss": 0.4297, "step": 12280 }, { "epoch": 5.7792941176470585, "grad_norm": 0.7652947195995252, "learning_rate": 1.4067268612669769e-06, "loss": 0.3887, "step": 12281 }, { "epoch": 5.779764705882353, "grad_norm": 0.4502536457931738, "learning_rate": 1.4007437654125375e-06, "loss": 0.1963, "step": 12282 }, { "epoch": 5.780235294117647, "grad_norm": 0.6747923871898855, "learning_rate": 1.3947733606634647e-06, "loss": 0.3984, "step": 12283 }, { "epoch": 5.780705882352941, "grad_norm": 0.7156926870965356, "learning_rate": 1.388815647529634e-06, "loss": 0.2734, "step": 12284 }, { "epoch": 5.781176470588235, "grad_norm": 0.3363944906152679, "learning_rate": 1.3828706265199051e-06, "loss": 0.1182, "step": 12285 }, { "epoch": 5.78164705882353, "grad_norm": 0.7265657947482526, "learning_rate": 1.376938298141972e-06, "loss": 0.2324, "step": 12286 }, { "epoch": 5.782117647058824, "grad_norm": 0.2542165424143516, "learning_rate": 1.3710186629025121e-06, "loss": 0.0684, "step": 12287 }, { "epoch": 5.782588235294118, "grad_norm": 0.6079685939555073, "learning_rate": 1.3651117213070716e-06, "loss": 0.3359, "step": 12288 }, { "epoch": 5.783058823529412, "grad_norm": 0.20827148326692124, "learning_rate": 1.3592174738601302e-06, "loss": 0.0732, "step": 12289 }, { "epoch": 5.783529411764706, "grad_norm": 0.2960601214749938, "learning_rate": 1.3533359210651184e-06, "loss": 0.0972, "step": 12290 }, { "epoch": 5.784, "grad_norm": 0.5378781805320401, "learning_rate": 1.3474670634243178e-06, "loss": 0.1982, "step": 12291 }, { "epoch": 5.784470588235294, "grad_norm": 0.649660980447561, "learning_rate": 1.341610901438961e-06, "loss": 0.3672, "step": 12292 }, { "epoch": 5.784941176470588, "grad_norm": 0.5742060154749482, "learning_rate": 1.3357674356091974e-06, "loss": 0.3008, "step": 12293 }, { "epoch": 5.7854117647058825, "grad_norm": 0.231153425631554, "learning_rate": 1.3299366664340948e-06, "loss": 0.106, "step": 12294 }, { "epoch": 5.785882352941177, "grad_norm": 0.5324378215673486, "learning_rate": 1.3241185944116217e-06, "loss": 0.2871, "step": 12295 }, { "epoch": 5.786352941176471, "grad_norm": 0.32047887909506806, "learning_rate": 1.3183132200386637e-06, "loss": 0.1406, "step": 12296 }, { "epoch": 5.786823529411764, "grad_norm": 0.5602922791243781, "learning_rate": 1.3125205438110408e-06, "loss": 0.3145, "step": 12297 }, { "epoch": 5.7872941176470585, "grad_norm": 0.2761446403404217, "learning_rate": 1.3067405662234743e-06, "loss": 0.0977, "step": 12298 }, { "epoch": 5.787764705882353, "grad_norm": 0.7169216367717013, "learning_rate": 1.3009732877696022e-06, "loss": 0.4785, "step": 12299 }, { "epoch": 5.788235294117647, "grad_norm": 0.5428074718448047, "learning_rate": 1.2952187089419642e-06, "loss": 0.4102, "step": 12300 }, { "epoch": 5.788705882352941, "grad_norm": 0.4494246923726104, "learning_rate": 1.28947683023205e-06, "loss": 0.1846, "step": 12301 }, { "epoch": 5.789176470588235, "grad_norm": 0.38869034940542097, "learning_rate": 1.2837476521302514e-06, "loss": 0.2295, "step": 12302 }, { "epoch": 5.78964705882353, "grad_norm": 0.597121416693656, "learning_rate": 1.2780311751258265e-06, "loss": 0.2656, "step": 12303 }, { "epoch": 5.790117647058824, "grad_norm": 0.451128665479003, "learning_rate": 1.2723273997070182e-06, "loss": 0.2471, "step": 12304 }, { "epoch": 5.790588235294118, "grad_norm": 0.4601263417405088, "learning_rate": 1.2666363263609702e-06, "loss": 0.3867, "step": 12305 }, { "epoch": 5.791058823529411, "grad_norm": 0.5205974681843512, "learning_rate": 1.2609579555737104e-06, "loss": 0.252, "step": 12306 }, { "epoch": 5.791529411764706, "grad_norm": 0.6562883767558788, "learning_rate": 1.2552922878301841e-06, "loss": 0.3652, "step": 12307 }, { "epoch": 5.792, "grad_norm": 0.336093320835396, "learning_rate": 1.2496393236143044e-06, "loss": 0.1123, "step": 12308 }, { "epoch": 5.792470588235294, "grad_norm": 0.488079873542176, "learning_rate": 1.2439990634088348e-06, "loss": 0.2363, "step": 12309 }, { "epoch": 5.792941176470588, "grad_norm": 0.24776030057531703, "learning_rate": 1.2383715076954902e-06, "loss": 0.105, "step": 12310 }, { "epoch": 5.7934117647058825, "grad_norm": 0.5975644847941222, "learning_rate": 1.232756656954903e-06, "loss": 0.3047, "step": 12311 }, { "epoch": 5.793882352941177, "grad_norm": 0.1619429828430695, "learning_rate": 1.2271545116665727e-06, "loss": 0.027, "step": 12312 }, { "epoch": 5.794352941176471, "grad_norm": 0.7382311227538874, "learning_rate": 1.2215650723089832e-06, "loss": 0.168, "step": 12313 }, { "epoch": 5.794823529411765, "grad_norm": 0.5953328464814851, "learning_rate": 1.2159883393595026e-06, "loss": 0.3184, "step": 12314 }, { "epoch": 5.7952941176470585, "grad_norm": 0.6282191790133389, "learning_rate": 1.2104243132943836e-06, "loss": 0.2021, "step": 12315 }, { "epoch": 5.795764705882353, "grad_norm": 0.5322692656702028, "learning_rate": 1.2048729945888458e-06, "loss": 0.2109, "step": 12316 }, { "epoch": 5.796235294117647, "grad_norm": 0.5938889639518297, "learning_rate": 1.1993343837169933e-06, "loss": 0.3184, "step": 12317 }, { "epoch": 5.796705882352941, "grad_norm": 0.5479897197427024, "learning_rate": 1.1938084811518311e-06, "loss": 0.2402, "step": 12318 }, { "epoch": 5.797176470588235, "grad_norm": 0.48856111034867433, "learning_rate": 1.1882952873653317e-06, "loss": 0.2314, "step": 12319 }, { "epoch": 5.79764705882353, "grad_norm": 0.4800269271833241, "learning_rate": 1.1827948028283352e-06, "loss": 0.2734, "step": 12320 }, { "epoch": 5.798117647058824, "grad_norm": 0.5459015856210387, "learning_rate": 1.1773070280105822e-06, "loss": 0.3496, "step": 12321 }, { "epoch": 5.798588235294117, "grad_norm": 0.5143383380976172, "learning_rate": 1.171831963380815e-06, "loss": 0.1523, "step": 12322 }, { "epoch": 5.799058823529411, "grad_norm": 0.5251163003483073, "learning_rate": 1.1663696094065755e-06, "loss": 0.293, "step": 12323 }, { "epoch": 5.799529411764706, "grad_norm": 0.5115473280709102, "learning_rate": 1.160919966554391e-06, "loss": 0.1846, "step": 12324 }, { "epoch": 5.8, "grad_norm": 0.5146727245657363, "learning_rate": 1.155483035289706e-06, "loss": 0.248, "step": 12325 }, { "epoch": 5.800470588235294, "grad_norm": 0.43029358349976937, "learning_rate": 1.150058816076832e-06, "loss": 0.1826, "step": 12326 }, { "epoch": 5.800941176470588, "grad_norm": 0.630501669265558, "learning_rate": 1.1446473093790487e-06, "loss": 0.3809, "step": 12327 }, { "epoch": 5.8014117647058825, "grad_norm": 0.5324598222345284, "learning_rate": 1.1392485156585196e-06, "loss": 0.5117, "step": 12328 }, { "epoch": 5.801882352941177, "grad_norm": 0.3188064154819375, "learning_rate": 1.1338624353763093e-06, "loss": 0.0767, "step": 12329 }, { "epoch": 5.802352941176471, "grad_norm": 0.5690421364447107, "learning_rate": 1.1284890689924497e-06, "loss": 0.2676, "step": 12330 }, { "epoch": 5.802823529411764, "grad_norm": 0.4415725710396058, "learning_rate": 1.1231284169658239e-06, "loss": 0.2109, "step": 12331 }, { "epoch": 5.8032941176470585, "grad_norm": 0.7915605658724715, "learning_rate": 1.1177804797542821e-06, "loss": 0.1582, "step": 12332 }, { "epoch": 5.803764705882353, "grad_norm": 0.5687359053521336, "learning_rate": 1.1124452578145425e-06, "loss": 0.2617, "step": 12333 }, { "epoch": 5.804235294117647, "grad_norm": 0.6514213863373576, "learning_rate": 1.1071227516022574e-06, "loss": 0.3242, "step": 12334 }, { "epoch": 5.804705882352941, "grad_norm": 0.10777597878119669, "learning_rate": 1.1018129615720129e-06, "loss": 0.0121, "step": 12335 }, { "epoch": 5.805176470588235, "grad_norm": 0.26160784148838456, "learning_rate": 1.0965158881772961e-06, "loss": 0.0986, "step": 12336 }, { "epoch": 5.80564705882353, "grad_norm": 0.5623293934370752, "learning_rate": 1.091231531870479e-06, "loss": 0.2217, "step": 12337 }, { "epoch": 5.806117647058824, "grad_norm": 0.2508836919600701, "learning_rate": 1.0859598931028668e-06, "loss": 0.0737, "step": 12338 }, { "epoch": 5.806588235294118, "grad_norm": 0.516770024038876, "learning_rate": 1.0807009723247328e-06, "loss": 0.2295, "step": 12339 }, { "epoch": 5.807058823529411, "grad_norm": 0.4626731440691716, "learning_rate": 1.0754547699851513e-06, "loss": 0.127, "step": 12340 }, { "epoch": 5.807529411764706, "grad_norm": 0.525983703048826, "learning_rate": 1.0702212865322135e-06, "loss": 0.3535, "step": 12341 }, { "epoch": 5.808, "grad_norm": 0.5414022529579552, "learning_rate": 1.0650005224128788e-06, "loss": 0.3906, "step": 12342 }, { "epoch": 5.808470588235294, "grad_norm": 0.4402990074895921, "learning_rate": 1.059792478073007e-06, "loss": 0.252, "step": 12343 }, { "epoch": 5.808941176470588, "grad_norm": 0.34409254567348013, "learning_rate": 1.0545971539574094e-06, "loss": 0.1582, "step": 12344 }, { "epoch": 5.8094117647058825, "grad_norm": 0.5302314467763498, "learning_rate": 1.0494145505097973e-06, "loss": 0.3574, "step": 12345 }, { "epoch": 5.809882352941177, "grad_norm": 0.46243901768378515, "learning_rate": 1.0442446681727834e-06, "loss": 0.2168, "step": 12346 }, { "epoch": 5.810352941176471, "grad_norm": 0.16506981342494806, "learning_rate": 1.0390875073878812e-06, "loss": 0.0447, "step": 12347 }, { "epoch": 5.810823529411765, "grad_norm": 0.22169949356464858, "learning_rate": 1.0339430685955719e-06, "loss": 0.0312, "step": 12348 }, { "epoch": 5.8112941176470585, "grad_norm": 0.3810935847281926, "learning_rate": 1.0288113522352038e-06, "loss": 0.1572, "step": 12349 }, { "epoch": 5.811764705882353, "grad_norm": 0.06505284603134712, "learning_rate": 1.0236923587450263e-06, "loss": 0.0095, "step": 12350 }, { "epoch": 5.812235294117647, "grad_norm": 0.38864369180130964, "learning_rate": 1.0185860885622565e-06, "loss": 0.1484, "step": 12351 }, { "epoch": 5.812705882352941, "grad_norm": 0.4908381618517256, "learning_rate": 1.013492542122979e-06, "loss": 0.2969, "step": 12352 }, { "epoch": 5.813176470588235, "grad_norm": 0.4978344993473245, "learning_rate": 1.0084117198622287e-06, "loss": 0.2051, "step": 12353 }, { "epoch": 5.81364705882353, "grad_norm": 0.6181442782413784, "learning_rate": 1.0033436222138924e-06, "loss": 0.3926, "step": 12354 }, { "epoch": 5.814117647058824, "grad_norm": 0.4050301190281266, "learning_rate": 9.98288249610857e-07, "loss": 0.1318, "step": 12355 }, { "epoch": 5.814588235294117, "grad_norm": 0.6877141122407759, "learning_rate": 9.932456024848435e-07, "loss": 0.2812, "step": 12356 }, { "epoch": 5.815058823529411, "grad_norm": 0.37751685991300593, "learning_rate": 9.882156812665244e-07, "loss": 0.1572, "step": 12357 }, { "epoch": 5.815529411764706, "grad_norm": 0.7547133027216671, "learning_rate": 9.831984863854891e-07, "loss": 0.4785, "step": 12358 }, { "epoch": 5.816, "grad_norm": 0.7462737654467669, "learning_rate": 9.781940182702451e-07, "loss": 0.2412, "step": 12359 }, { "epoch": 5.816470588235294, "grad_norm": 0.301820696748373, "learning_rate": 9.732022773481506e-07, "loss": 0.0923, "step": 12360 }, { "epoch": 5.816941176470588, "grad_norm": 0.46275017171142474, "learning_rate": 9.682232640455644e-07, "loss": 0.2285, "step": 12361 }, { "epoch": 5.8174117647058825, "grad_norm": 0.6166136650901919, "learning_rate": 9.632569787877298e-07, "loss": 0.3203, "step": 12362 }, { "epoch": 5.817882352941177, "grad_norm": 0.4301453373340576, "learning_rate": 9.583034219987406e-07, "loss": 0.1699, "step": 12363 }, { "epoch": 5.818352941176471, "grad_norm": 0.4446911580725922, "learning_rate": 9.53362594101692e-07, "loss": 0.2812, "step": 12364 }, { "epoch": 5.818823529411764, "grad_norm": 0.4809482882552322, "learning_rate": 9.48434495518563e-07, "loss": 0.293, "step": 12365 }, { "epoch": 5.8192941176470585, "grad_norm": 0.2054507526271747, "learning_rate": 9.435191266702003e-07, "loss": 0.0403, "step": 12366 }, { "epoch": 5.819764705882353, "grad_norm": 0.30421572232960464, "learning_rate": 9.386164879764346e-07, "loss": 0.0908, "step": 12367 }, { "epoch": 5.820235294117647, "grad_norm": 0.5779048837155588, "learning_rate": 9.337265798559645e-07, "loss": 0.2637, "step": 12368 }, { "epoch": 5.820705882352941, "grad_norm": 0.5057274202020086, "learning_rate": 9.28849402726406e-07, "loss": 0.2734, "step": 12369 }, { "epoch": 5.821176470588235, "grad_norm": 0.5380667748731102, "learning_rate": 9.239849570042924e-07, "loss": 0.3906, "step": 12370 }, { "epoch": 5.82164705882353, "grad_norm": 0.605797733408016, "learning_rate": 9.191332431050747e-07, "loss": 0.3535, "step": 12371 }, { "epoch": 5.822117647058824, "grad_norm": 0.46563884676464146, "learning_rate": 9.142942614431215e-07, "loss": 0.167, "step": 12372 }, { "epoch": 5.822588235294118, "grad_norm": 0.6202408825541781, "learning_rate": 9.094680124317022e-07, "loss": 0.4102, "step": 12373 }, { "epoch": 5.823058823529411, "grad_norm": 0.4527309215066827, "learning_rate": 9.046544964830038e-07, "loss": 0.25, "step": 12374 }, { "epoch": 5.823529411764706, "grad_norm": 0.516369190349363, "learning_rate": 8.998537140081141e-07, "loss": 0.2188, "step": 12375 }, { "epoch": 5.824, "grad_norm": 0.48017675296336265, "learning_rate": 8.950656654170552e-07, "loss": 0.2969, "step": 12376 }, { "epoch": 5.824470588235294, "grad_norm": 0.5627222530895307, "learning_rate": 8.902903511187332e-07, "loss": 0.293, "step": 12377 }, { "epoch": 5.824941176470588, "grad_norm": 0.5764533796804174, "learning_rate": 8.855277715210052e-07, "loss": 0.3125, "step": 12378 }, { "epoch": 5.8254117647058825, "grad_norm": 0.043921037987403294, "learning_rate": 8.807779270305959e-07, "loss": 0.0037, "step": 12379 }, { "epoch": 5.825882352941177, "grad_norm": 0.774453840419, "learning_rate": 8.760408180531809e-07, "loss": 0.4141, "step": 12380 }, { "epoch": 5.826352941176471, "grad_norm": 0.5819927731320632, "learning_rate": 8.713164449933364e-07, "loss": 0.3926, "step": 12381 }, { "epoch": 5.826823529411765, "grad_norm": 0.3295541075989036, "learning_rate": 8.666048082545396e-07, "loss": 0.1465, "step": 12382 }, { "epoch": 5.8272941176470585, "grad_norm": 0.5076870064515834, "learning_rate": 8.619059082391689e-07, "loss": 0.2471, "step": 12383 }, { "epoch": 5.827764705882353, "grad_norm": 0.609605298559079, "learning_rate": 8.572197453485696e-07, "loss": 0.3535, "step": 12384 }, { "epoch": 5.828235294117647, "grad_norm": 0.8901673381209988, "learning_rate": 8.525463199829385e-07, "loss": 0.4453, "step": 12385 }, { "epoch": 5.828705882352941, "grad_norm": 0.42595173527707875, "learning_rate": 8.47885632541423e-07, "loss": 0.1475, "step": 12386 }, { "epoch": 5.829176470588235, "grad_norm": 0.4566678848798674, "learning_rate": 8.432376834220544e-07, "loss": 0.2656, "step": 12387 }, { "epoch": 5.82964705882353, "grad_norm": 0.36625978957232574, "learning_rate": 8.386024730217988e-07, "loss": 0.0605, "step": 12388 }, { "epoch": 5.830117647058824, "grad_norm": 0.36617780048058574, "learning_rate": 8.339800017365228e-07, "loss": 0.1455, "step": 12389 }, { "epoch": 5.830588235294117, "grad_norm": 0.6520173481101861, "learning_rate": 8.293702699610272e-07, "loss": 0.4219, "step": 12390 }, { "epoch": 5.831058823529411, "grad_norm": 0.9674544182280278, "learning_rate": 8.247732780889637e-07, "loss": 0.5391, "step": 12391 }, { "epoch": 5.831529411764706, "grad_norm": 0.5691122924615348, "learning_rate": 8.201890265129684e-07, "loss": 0.3555, "step": 12392 }, { "epoch": 5.832, "grad_norm": 0.4179770148548359, "learning_rate": 8.156175156245614e-07, "loss": 0.207, "step": 12393 }, { "epoch": 5.832470588235294, "grad_norm": 0.47139460075780215, "learning_rate": 8.110587458141637e-07, "loss": 0.2051, "step": 12394 }, { "epoch": 5.832941176470588, "grad_norm": 0.5798967526619433, "learning_rate": 8.065127174711139e-07, "loss": 0.3027, "step": 12395 }, { "epoch": 5.8334117647058825, "grad_norm": 0.6154512406696658, "learning_rate": 8.01979430983668e-07, "loss": 0.1973, "step": 12396 }, { "epoch": 5.833882352941177, "grad_norm": 0.30426091548875794, "learning_rate": 7.97458886738983e-07, "loss": 0.0884, "step": 12397 }, { "epoch": 5.834352941176471, "grad_norm": 0.4863403211325887, "learning_rate": 7.929510851231502e-07, "loss": 0.1904, "step": 12398 }, { "epoch": 5.834823529411764, "grad_norm": 0.6124354398916286, "learning_rate": 7.884560265211615e-07, "loss": 0.3555, "step": 12399 }, { "epoch": 5.8352941176470585, "grad_norm": 0.4118893521960676, "learning_rate": 7.839737113168931e-07, "loss": 0.082, "step": 12400 }, { "epoch": 5.835764705882353, "grad_norm": 0.24947683099827186, "learning_rate": 7.795041398931723e-07, "loss": 0.0708, "step": 12401 }, { "epoch": 5.836235294117647, "grad_norm": 0.3446114979107569, "learning_rate": 7.750473126317269e-07, "loss": 0.1128, "step": 12402 }, { "epoch": 5.836705882352941, "grad_norm": 0.5879700866672943, "learning_rate": 7.706032299131692e-07, "loss": 0.3477, "step": 12403 }, { "epoch": 5.837176470588235, "grad_norm": 0.41541052032907594, "learning_rate": 7.66171892117079e-07, "loss": 0.1367, "step": 12404 }, { "epoch": 5.83764705882353, "grad_norm": 0.36964349789572054, "learning_rate": 7.617532996218867e-07, "loss": 0.1523, "step": 12405 }, { "epoch": 5.838117647058824, "grad_norm": 0.39198937971350306, "learning_rate": 7.573474528049739e-07, "loss": 0.2227, "step": 12406 }, { "epoch": 5.838588235294118, "grad_norm": 0.4624731711174657, "learning_rate": 7.529543520426228e-07, "loss": 0.2637, "step": 12407 }, { "epoch": 5.839058823529411, "grad_norm": 0.7475771679651849, "learning_rate": 7.485739977100336e-07, "loss": 0.375, "step": 12408 }, { "epoch": 5.839529411764706, "grad_norm": 0.40335285833433787, "learning_rate": 7.442063901812734e-07, "loss": 0.1602, "step": 12409 }, { "epoch": 5.84, "grad_norm": 0.5375502291310037, "learning_rate": 7.398515298294106e-07, "loss": 0.1719, "step": 12410 }, { "epoch": 5.840470588235294, "grad_norm": 0.15003254216277626, "learning_rate": 7.355094170263143e-07, "loss": 0.1235, "step": 12411 }, { "epoch": 5.840941176470588, "grad_norm": 0.23545097331015544, "learning_rate": 7.311800521428712e-07, "loss": 0.0605, "step": 12412 }, { "epoch": 5.8414117647058825, "grad_norm": 0.5402531722341465, "learning_rate": 7.26863435548819e-07, "loss": 0.3398, "step": 12413 }, { "epoch": 5.841882352941177, "grad_norm": 0.7772364222345806, "learning_rate": 7.22559567612796e-07, "loss": 0.4238, "step": 12414 }, { "epoch": 5.842352941176471, "grad_norm": 0.503019731604588, "learning_rate": 7.182684487023748e-07, "loss": 0.2188, "step": 12415 }, { "epoch": 5.842823529411765, "grad_norm": 0.6067437294810548, "learning_rate": 7.139900791840792e-07, "loss": 0.2334, "step": 12416 }, { "epoch": 5.8432941176470585, "grad_norm": 0.19749433158643645, "learning_rate": 7.097244594232498e-07, "loss": 0.0559, "step": 12417 }, { "epoch": 5.843764705882353, "grad_norm": 0.3747167043629897, "learning_rate": 7.054715897842289e-07, "loss": 0.1416, "step": 12418 }, { "epoch": 5.844235294117647, "grad_norm": 0.3639765226753595, "learning_rate": 7.012314706302257e-07, "loss": 0.1465, "step": 12419 }, { "epoch": 5.844705882352941, "grad_norm": 0.5865279983084968, "learning_rate": 6.970041023233508e-07, "loss": 0.3379, "step": 12420 }, { "epoch": 5.845176470588235, "grad_norm": 0.5618484900126043, "learning_rate": 6.927894852246485e-07, "loss": 0.3398, "step": 12421 }, { "epoch": 5.84564705882353, "grad_norm": 0.5621551974391987, "learning_rate": 6.88587619694081e-07, "loss": 0.2891, "step": 12422 }, { "epoch": 5.846117647058824, "grad_norm": 0.6653733927285488, "learning_rate": 6.843985060904944e-07, "loss": 0.3438, "step": 12423 }, { "epoch": 5.846588235294117, "grad_norm": 0.5584518723622238, "learning_rate": 6.802221447716694e-07, "loss": 0.2949, "step": 12424 }, { "epoch": 5.847058823529411, "grad_norm": 0.3310945303204313, "learning_rate": 6.760585360942872e-07, "loss": 0.1011, "step": 12425 }, { "epoch": 5.847529411764706, "grad_norm": 0.3527460470554208, "learning_rate": 6.719076804139134e-07, "loss": 0.2246, "step": 12426 }, { "epoch": 5.848, "grad_norm": 0.28146674392689386, "learning_rate": 6.677695780850978e-07, "loss": 0.1357, "step": 12427 }, { "epoch": 5.848470588235294, "grad_norm": 0.4870941898415211, "learning_rate": 6.636442294612243e-07, "loss": 0.2578, "step": 12428 }, { "epoch": 5.848941176470588, "grad_norm": 0.32924726515900055, "learning_rate": 6.595316348946278e-07, "loss": 0.1514, "step": 12429 }, { "epoch": 5.8494117647058825, "grad_norm": 0.48151307783511654, "learning_rate": 6.554317947365273e-07, "loss": 0.2559, "step": 12430 }, { "epoch": 5.849882352941177, "grad_norm": 0.7092221633434048, "learning_rate": 6.513447093370927e-07, "loss": 0.6758, "step": 12431 }, { "epoch": 5.850352941176471, "grad_norm": 0.3806765479341151, "learning_rate": 6.472703790453781e-07, "loss": 0.1309, "step": 12432 }, { "epoch": 5.850823529411764, "grad_norm": 0.4087250131536979, "learning_rate": 6.43208804209322e-07, "loss": 0.1875, "step": 12433 }, { "epoch": 5.8512941176470585, "grad_norm": 0.5334258040655194, "learning_rate": 6.391599851758466e-07, "loss": 0.3027, "step": 12434 }, { "epoch": 5.851764705882353, "grad_norm": 0.6365563222191586, "learning_rate": 6.351239222907089e-07, "loss": 0.3008, "step": 12435 }, { "epoch": 5.852235294117647, "grad_norm": 0.6896971678297432, "learning_rate": 6.311006158986165e-07, "loss": 0.4746, "step": 12436 }, { "epoch": 5.852705882352941, "grad_norm": 0.44330159887017667, "learning_rate": 6.270900663431777e-07, "loss": 0.2617, "step": 12437 }, { "epoch": 5.853176470588235, "grad_norm": 0.45398204206002063, "learning_rate": 6.230922739669353e-07, "loss": 0.21, "step": 12438 }, { "epoch": 5.85364705882353, "grad_norm": 0.614527894448414, "learning_rate": 6.191072391112828e-07, "loss": 0.4805, "step": 12439 }, { "epoch": 5.854117647058824, "grad_norm": 0.7259976976434664, "learning_rate": 6.151349621165979e-07, "loss": 0.377, "step": 12440 }, { "epoch": 5.854588235294118, "grad_norm": 0.5165227844751609, "learning_rate": 6.111754433221094e-07, "loss": 0.2891, "step": 12441 }, { "epoch": 5.855058823529411, "grad_norm": 0.43111320427660016, "learning_rate": 6.0722868306598e-07, "loss": 0.1865, "step": 12442 }, { "epoch": 5.855529411764706, "grad_norm": 0.5605395179322366, "learning_rate": 6.032946816852902e-07, "loss": 0.2871, "step": 12443 }, { "epoch": 5.856, "grad_norm": 0.732176751532028, "learning_rate": 5.99373439516021e-07, "loss": 0.6367, "step": 12444 }, { "epoch": 5.856470588235294, "grad_norm": 0.3906542113543985, "learning_rate": 5.954649568930714e-07, "loss": 0.1953, "step": 12445 }, { "epoch": 5.856941176470588, "grad_norm": 0.7470877247032026, "learning_rate": 5.91569234150241e-07, "loss": 0.4375, "step": 12446 }, { "epoch": 5.8574117647058825, "grad_norm": 0.43443579309600305, "learning_rate": 5.876862716202302e-07, "loss": 0.1455, "step": 12447 }, { "epoch": 5.857882352941177, "grad_norm": 0.6110517077937555, "learning_rate": 5.838160696346906e-07, "loss": 0.3711, "step": 12448 }, { "epoch": 5.85835294117647, "grad_norm": 0.5101462591814108, "learning_rate": 5.799586285241242e-07, "loss": 0.4141, "step": 12449 }, { "epoch": 5.858823529411764, "grad_norm": 0.6450072711275068, "learning_rate": 5.761139486180178e-07, "loss": 0.4219, "step": 12450 }, { "epoch": 5.8592941176470585, "grad_norm": 0.5766066657837436, "learning_rate": 5.722820302446751e-07, "loss": 0.2949, "step": 12451 }, { "epoch": 5.859764705882353, "grad_norm": 0.8378165434541831, "learning_rate": 5.684628737314012e-07, "loss": 0.3516, "step": 12452 }, { "epoch": 5.860235294117647, "grad_norm": 0.42854696283858434, "learning_rate": 5.646564794043518e-07, "loss": 0.2061, "step": 12453 }, { "epoch": 5.860705882352941, "grad_norm": 0.5035791984342334, "learning_rate": 5.608628475886001e-07, "loss": 0.2139, "step": 12454 }, { "epoch": 5.861176470588235, "grad_norm": 0.5713572095525562, "learning_rate": 5.570819786081704e-07, "loss": 0.3164, "step": 12455 }, { "epoch": 5.86164705882353, "grad_norm": 0.4579330600102511, "learning_rate": 5.533138727859543e-07, "loss": 0.1445, "step": 12456 }, { "epoch": 5.862117647058824, "grad_norm": 0.49382342575893523, "learning_rate": 5.495585304437611e-07, "loss": 0.3027, "step": 12457 }, { "epoch": 5.862588235294117, "grad_norm": 0.5640491980100869, "learning_rate": 5.458159519023175e-07, "loss": 0.3242, "step": 12458 }, { "epoch": 5.863058823529411, "grad_norm": 0.6113276886797141, "learning_rate": 5.42086137481268e-07, "loss": 0.249, "step": 12459 }, { "epoch": 5.863529411764706, "grad_norm": 0.42115092573787405, "learning_rate": 5.383690874991408e-07, "loss": 0.1787, "step": 12460 }, { "epoch": 5.864, "grad_norm": 0.6540600885950056, "learning_rate": 5.34664802273399e-07, "loss": 0.2695, "step": 12461 }, { "epoch": 5.864470588235294, "grad_norm": 0.42368764595183467, "learning_rate": 5.30973282120406e-07, "loss": 0.293, "step": 12462 }, { "epoch": 5.864941176470588, "grad_norm": 0.19450538143953405, "learning_rate": 5.272945273554429e-07, "loss": 0.0483, "step": 12463 }, { "epoch": 5.8654117647058825, "grad_norm": 0.3580768994590732, "learning_rate": 5.236285382926752e-07, "loss": 0.085, "step": 12464 }, { "epoch": 5.865882352941177, "grad_norm": 0.45281820497530906, "learning_rate": 5.19975315245219e-07, "loss": 0.1504, "step": 12465 }, { "epoch": 5.866352941176471, "grad_norm": 0.4518453861370524, "learning_rate": 5.163348585250748e-07, "loss": 0.1465, "step": 12466 }, { "epoch": 5.866823529411764, "grad_norm": 0.6943004934207432, "learning_rate": 5.127071684431272e-07, "loss": 0.2793, "step": 12467 }, { "epoch": 5.8672941176470585, "grad_norm": 0.6817798382529837, "learning_rate": 5.090922453092283e-07, "loss": 0.4238, "step": 12468 }, { "epoch": 5.867764705882353, "grad_norm": 0.677798383746864, "learning_rate": 5.054900894320979e-07, "loss": 0.4121, "step": 12469 }, { "epoch": 5.868235294117647, "grad_norm": 0.11448370695497863, "learning_rate": 5.019007011193732e-07, "loss": 0.0186, "step": 12470 }, { "epoch": 5.868705882352941, "grad_norm": 0.65086150523622, "learning_rate": 4.983240806776257e-07, "loss": 0.1108, "step": 12471 }, { "epoch": 5.869176470588235, "grad_norm": 0.27855981189789997, "learning_rate": 4.947602284122942e-07, "loss": 0.0718, "step": 12472 }, { "epoch": 5.86964705882353, "grad_norm": 0.5350775192363266, "learning_rate": 4.912091446277688e-07, "loss": 0.3262, "step": 12473 }, { "epoch": 5.870117647058824, "grad_norm": 0.44998173504030015, "learning_rate": 4.876708296273068e-07, "loss": 0.25, "step": 12474 }, { "epoch": 5.870588235294118, "grad_norm": 0.5941275432757592, "learning_rate": 4.841452837131166e-07, "loss": 0.2461, "step": 12475 }, { "epoch": 5.871058823529411, "grad_norm": 0.2620300400788011, "learning_rate": 4.806325071862905e-07, "loss": 0.0571, "step": 12476 }, { "epoch": 5.871529411764706, "grad_norm": 0.4510477708297639, "learning_rate": 4.771325003468385e-07, "loss": 0.1836, "step": 12477 }, { "epoch": 5.872, "grad_norm": 0.6247751944251303, "learning_rate": 4.736452634936716e-07, "loss": 0.3809, "step": 12478 }, { "epoch": 5.872470588235294, "grad_norm": 0.45325999210105694, "learning_rate": 4.701707969246182e-07, "loss": 0.3047, "step": 12479 }, { "epoch": 5.872941176470588, "grad_norm": 0.4696141049305684, "learning_rate": 4.6670910093642435e-07, "loss": 0.21, "step": 12480 }, { "epoch": 5.8734117647058826, "grad_norm": 0.6455202341733783, "learning_rate": 4.6326017582473674e-07, "loss": 0.3633, "step": 12481 }, { "epoch": 5.873882352941177, "grad_norm": 0.5620670011673954, "learning_rate": 4.5982402188408653e-07, "loss": 0.2373, "step": 12482 }, { "epoch": 5.87435294117647, "grad_norm": 0.3071782223177839, "learning_rate": 4.5640063940797223e-07, "loss": 0.1533, "step": 12483 }, { "epoch": 5.874823529411764, "grad_norm": 0.24894317551474493, "learning_rate": 4.529900286887433e-07, "loss": 0.0396, "step": 12484 }, { "epoch": 5.875294117647059, "grad_norm": 0.3296954786853032, "learning_rate": 4.4959219001768355e-07, "loss": 0.0942, "step": 12485 }, { "epoch": 5.875764705882353, "grad_norm": 0.5475345057652402, "learning_rate": 4.4620712368501065e-07, "loss": 0.2275, "step": 12486 }, { "epoch": 5.876235294117647, "grad_norm": 0.38451676924136097, "learning_rate": 4.428348299797768e-07, "loss": 0.1069, "step": 12487 }, { "epoch": 5.876705882352941, "grad_norm": 0.6448769014601785, "learning_rate": 4.394753091900516e-07, "loss": 0.5312, "step": 12488 }, { "epoch": 5.8771764705882354, "grad_norm": 0.5212976824041955, "learning_rate": 4.361285616027055e-07, "loss": 0.3281, "step": 12489 }, { "epoch": 5.87764705882353, "grad_norm": 0.6987912106017322, "learning_rate": 4.327945875035932e-07, "loss": 0.2695, "step": 12490 }, { "epoch": 5.878117647058824, "grad_norm": 0.421461329283368, "learning_rate": 4.2947338717743694e-07, "loss": 0.1162, "step": 12491 }, { "epoch": 5.878588235294117, "grad_norm": 0.515501226016352, "learning_rate": 4.2616496090790983e-07, "loss": 0.2734, "step": 12492 }, { "epoch": 5.8790588235294114, "grad_norm": 0.2744939705661699, "learning_rate": 4.228693089775526e-07, "loss": 0.0437, "step": 12493 }, { "epoch": 5.879529411764706, "grad_norm": 0.6282879197761351, "learning_rate": 4.195864316678066e-07, "loss": 0.334, "step": 12494 }, { "epoch": 5.88, "grad_norm": 0.5348941980850046, "learning_rate": 4.16316329259081e-07, "loss": 0.3223, "step": 12495 }, { "epoch": 5.880470588235294, "grad_norm": 0.6221305498413094, "learning_rate": 4.130590020306523e-07, "loss": 0.1611, "step": 12496 }, { "epoch": 5.880941176470588, "grad_norm": 0.6747430860868469, "learning_rate": 4.0981445026068147e-07, "loss": 0.5039, "step": 12497 }, { "epoch": 5.881411764705883, "grad_norm": 0.6099598880939514, "learning_rate": 4.0658267422631343e-07, "loss": 0.332, "step": 12498 }, { "epoch": 5.881882352941177, "grad_norm": 0.27796277720611334, "learning_rate": 4.033636742035273e-07, "loss": 0.0703, "step": 12499 }, { "epoch": 5.882352941176471, "grad_norm": 0.5066506003859514, "learning_rate": 4.0015745046725336e-07, "loss": 0.2412, "step": 12500 }, { "epoch": 5.882823529411764, "grad_norm": 0.42497396039158125, "learning_rate": 3.9696400329132236e-07, "loss": 0.2031, "step": 12501 }, { "epoch": 5.883294117647059, "grad_norm": 0.47881599712559797, "learning_rate": 3.9378333294844967e-07, "loss": 0.1875, "step": 12502 }, { "epoch": 5.883764705882353, "grad_norm": 0.3989682480714736, "learning_rate": 3.906154397103012e-07, "loss": 0.1631, "step": 12503 }, { "epoch": 5.884235294117647, "grad_norm": 0.5005974181684166, "learning_rate": 3.874603238474272e-07, "loss": 0.3555, "step": 12504 }, { "epoch": 5.884705882352941, "grad_norm": 0.6957871193352756, "learning_rate": 3.8431798562927887e-07, "loss": 0.4492, "step": 12505 }, { "epoch": 5.8851764705882355, "grad_norm": 0.32303931174518913, "learning_rate": 3.8118842532424145e-07, "loss": 0.1758, "step": 12506 }, { "epoch": 5.88564705882353, "grad_norm": 0.46302788623764646, "learning_rate": 3.780716431995845e-07, "loss": 0.1299, "step": 12507 }, { "epoch": 5.886117647058824, "grad_norm": 0.18373056891923628, "learning_rate": 3.7496763952151177e-07, "loss": 0.0391, "step": 12508 }, { "epoch": 5.886588235294118, "grad_norm": 0.6192315899043009, "learning_rate": 3.7187641455509455e-07, "loss": 0.2344, "step": 12509 }, { "epoch": 5.8870588235294115, "grad_norm": 0.0001910393892288199, "learning_rate": 3.687979685643549e-07, "loss": 0.0, "step": 12510 }, { "epoch": 5.887529411764706, "grad_norm": 0.2777420558742751, "learning_rate": 3.6573230181221604e-07, "loss": 0.1025, "step": 12511 }, { "epoch": 5.888, "grad_norm": 0.552025296590338, "learning_rate": 3.62679414560485e-07, "loss": 0.334, "step": 12512 }, { "epoch": 5.888470588235294, "grad_norm": 0.6051508036478922, "learning_rate": 3.596393070699033e-07, "loss": 0.2559, "step": 12513 }, { "epoch": 5.888941176470588, "grad_norm": 0.3774043077581923, "learning_rate": 3.566119796000966e-07, "loss": 0.1187, "step": 12514 }, { "epoch": 5.889411764705883, "grad_norm": 0.46698914627272725, "learning_rate": 3.535974324096413e-07, "loss": 0.2598, "step": 12515 }, { "epoch": 5.889882352941177, "grad_norm": 0.45108132460497496, "learning_rate": 3.5059566575596497e-07, "loss": 0.1982, "step": 12516 }, { "epoch": 5.89035294117647, "grad_norm": 0.46707161329250757, "learning_rate": 3.4760667989544575e-07, "loss": 0.1738, "step": 12517 }, { "epoch": 5.890823529411764, "grad_norm": 0.505158837689941, "learning_rate": 3.446304750833462e-07, "loss": 0.3105, "step": 12518 }, { "epoch": 5.891294117647059, "grad_norm": 0.5229364830659168, "learning_rate": 3.416670515738629e-07, "loss": 0.377, "step": 12519 }, { "epoch": 5.891764705882353, "grad_norm": 0.6343101066824747, "learning_rate": 3.387164096200934e-07, "loss": 0.5078, "step": 12520 }, { "epoch": 5.892235294117647, "grad_norm": 0.5036732236660639, "learning_rate": 3.3577854947403615e-07, "loss": 0.3008, "step": 12521 }, { "epoch": 5.892705882352941, "grad_norm": 0.5661905749151351, "learning_rate": 3.3285347138657383e-07, "loss": 0.2949, "step": 12522 }, { "epoch": 5.8931764705882355, "grad_norm": 0.47714734746712134, "learning_rate": 3.2994117560753986e-07, "loss": 0.25, "step": 12523 }, { "epoch": 5.89364705882353, "grad_norm": 0.5571744439653344, "learning_rate": 3.2704166238565197e-07, "loss": 0.2773, "step": 12524 }, { "epoch": 5.894117647058824, "grad_norm": 0.5813521314254713, "learning_rate": 3.241549319685621e-07, "loss": 0.332, "step": 12525 }, { "epoch": 5.894588235294117, "grad_norm": 0.5522103704143189, "learning_rate": 3.212809846027897e-07, "loss": 0.1377, "step": 12526 }, { "epoch": 5.8950588235294115, "grad_norm": 0.452401173344547, "learning_rate": 3.1841982053378846e-07, "loss": 0.2012, "step": 12527 }, { "epoch": 5.895529411764706, "grad_norm": 0.4235224434750772, "learning_rate": 3.155714400059128e-07, "loss": 0.1533, "step": 12528 }, { "epoch": 5.896, "grad_norm": 0.3429048460664385, "learning_rate": 3.127358432624516e-07, "loss": 0.105, "step": 12529 }, { "epoch": 5.896470588235294, "grad_norm": 0.45062686100365346, "learning_rate": 3.099130305455444e-07, "loss": 0.25, "step": 12530 }, { "epoch": 5.896941176470588, "grad_norm": 0.32039989419699993, "learning_rate": 3.0710300209629836e-07, "loss": 0.1318, "step": 12531 }, { "epoch": 5.897411764705883, "grad_norm": 0.639877186643441, "learning_rate": 3.043057581547048e-07, "loss": 0.2793, "step": 12532 }, { "epoch": 5.897882352941177, "grad_norm": 0.5369552084707697, "learning_rate": 3.0152129895963937e-07, "loss": 0.1729, "step": 12533 }, { "epoch": 5.898352941176471, "grad_norm": 0.8150489989885833, "learning_rate": 2.987496247489285e-07, "loss": 0.375, "step": 12534 }, { "epoch": 5.898823529411764, "grad_norm": 0.8264548511212948, "learning_rate": 2.959907357592661e-07, "loss": 0.5742, "step": 12535 }, { "epoch": 5.899294117647059, "grad_norm": 0.3630666526123028, "learning_rate": 2.9324463222629715e-07, "loss": 0.165, "step": 12536 }, { "epoch": 5.899764705882353, "grad_norm": 0.19558688996305482, "learning_rate": 2.9051131438455055e-07, "loss": 0.0564, "step": 12537 }, { "epoch": 5.900235294117647, "grad_norm": 0.7721101191628867, "learning_rate": 2.8779078246745635e-07, "loss": 0.5156, "step": 12538 }, { "epoch": 5.900705882352941, "grad_norm": 0.543434718231728, "learning_rate": 2.85083036707362e-07, "loss": 0.3926, "step": 12539 }, { "epoch": 5.9011764705882355, "grad_norm": 0.6412544782780774, "learning_rate": 2.8238807733551583e-07, "loss": 0.4121, "step": 12540 }, { "epoch": 5.90164705882353, "grad_norm": 0.5750826482909283, "learning_rate": 2.797059045821004e-07, "loss": 0.2256, "step": 12541 }, { "epoch": 5.902117647058824, "grad_norm": 0.4242922371540763, "learning_rate": 2.7703651867616583e-07, "loss": 0.1895, "step": 12542 }, { "epoch": 5.902588235294118, "grad_norm": 0.5918746043336207, "learning_rate": 2.743799198457131e-07, "loss": 0.1201, "step": 12543 }, { "epoch": 5.9030588235294115, "grad_norm": 0.36602764572728047, "learning_rate": 2.71736108317594e-07, "loss": 0.1484, "step": 12544 }, { "epoch": 5.903529411764706, "grad_norm": 0.5427919199435401, "learning_rate": 2.6910508431764457e-07, "loss": 0.2773, "step": 12545 }, { "epoch": 5.904, "grad_norm": 1.6244465422972427, "learning_rate": 2.6648684807053513e-07, "loss": 1.3594, "step": 12546 }, { "epoch": 5.904470588235294, "grad_norm": 0.6180228487280379, "learning_rate": 2.6388139979988677e-07, "loss": 0.332, "step": 12547 }, { "epoch": 5.904941176470588, "grad_norm": 0.4126430834175369, "learning_rate": 2.612887397282215e-07, "loss": 0.2344, "step": 12548 }, { "epoch": 5.905411764705883, "grad_norm": 0.4001269951484147, "learning_rate": 2.5870886807696225e-07, "loss": 0.1201, "step": 12549 }, { "epoch": 5.905882352941177, "grad_norm": 0.6691403880682392, "learning_rate": 2.5614178506644934e-07, "loss": 0.2344, "step": 12550 }, { "epoch": 5.90635294117647, "grad_norm": 0.7315603702986664, "learning_rate": 2.5358749091590746e-07, "loss": 0.5586, "step": 12551 }, { "epoch": 5.906823529411764, "grad_norm": 0.4720543484532558, "learning_rate": 2.5104598584349543e-07, "loss": 0.2617, "step": 12552 }, { "epoch": 5.907294117647059, "grad_norm": 0.4962407419322232, "learning_rate": 2.4851727006627297e-07, "loss": 0.1973, "step": 12553 }, { "epoch": 5.907764705882353, "grad_norm": 0.5776113899188174, "learning_rate": 2.460013438002173e-07, "loss": 0.2656, "step": 12554 }, { "epoch": 5.908235294117647, "grad_norm": 0.32136059344083023, "learning_rate": 2.434982072601732e-07, "loss": 0.0908, "step": 12555 }, { "epoch": 5.908705882352941, "grad_norm": 0.5443491647659363, "learning_rate": 2.4100786065991975e-07, "loss": 0.3086, "step": 12556 }, { "epoch": 5.9091764705882355, "grad_norm": 0.6578333340806463, "learning_rate": 2.3853030421218667e-07, "loss": 0.3848, "step": 12557 }, { "epoch": 5.90964705882353, "grad_norm": 0.3647005041301363, "learning_rate": 2.3606553812852146e-07, "loss": 0.1631, "step": 12558 }, { "epoch": 5.910117647058824, "grad_norm": 0.6693441811217876, "learning_rate": 2.3361356261945575e-07, "loss": 0.3398, "step": 12559 }, { "epoch": 5.910588235294117, "grad_norm": 0.4203691599600658, "learning_rate": 2.3117437789438865e-07, "loss": 0.1328, "step": 12560 }, { "epoch": 5.9110588235294115, "grad_norm": 0.5900371487799744, "learning_rate": 2.2874798416163686e-07, "loss": 0.2676, "step": 12561 }, { "epoch": 5.911529411764706, "grad_norm": 0.3648327116817486, "learning_rate": 2.2633438162843465e-07, "loss": 0.085, "step": 12562 }, { "epoch": 5.912, "grad_norm": 0.7435373682144825, "learning_rate": 2.239335705009171e-07, "loss": 0.3594, "step": 12563 }, { "epoch": 5.912470588235294, "grad_norm": 0.596784663465486, "learning_rate": 2.215455509841202e-07, "loss": 0.457, "step": 12564 }, { "epoch": 5.912941176470588, "grad_norm": 0.7394522410858827, "learning_rate": 2.1917032328199747e-07, "loss": 0.4316, "step": 12565 }, { "epoch": 5.913411764705883, "grad_norm": 0.3319521816012554, "learning_rate": 2.1680788759738664e-07, "loss": 0.0845, "step": 12566 }, { "epoch": 5.913882352941177, "grad_norm": 1.27206445288379, "learning_rate": 2.144582441320597e-07, "loss": 1.2031, "step": 12567 }, { "epoch": 5.914352941176471, "grad_norm": 0.4349352847971498, "learning_rate": 2.1212139308670607e-07, "loss": 0.2402, "step": 12568 }, { "epoch": 5.914823529411764, "grad_norm": 0.5798270591820089, "learning_rate": 2.0979733466088277e-07, "loss": 0.4062, "step": 12569 }, { "epoch": 5.915294117647059, "grad_norm": 0.3344140010594876, "learning_rate": 2.0748606905309774e-07, "loss": 0.0947, "step": 12570 }, { "epoch": 5.915764705882353, "grad_norm": 0.24873420733352022, "learning_rate": 2.051875964607097e-07, "loss": 0.0718, "step": 12571 }, { "epoch": 5.916235294117647, "grad_norm": 0.26933407118167696, "learning_rate": 2.0290191708006165e-07, "loss": 0.0544, "step": 12572 }, { "epoch": 5.916705882352941, "grad_norm": 0.4997225963584571, "learning_rate": 2.0062903110633077e-07, "loss": 0.2061, "step": 12573 }, { "epoch": 5.9171764705882355, "grad_norm": 0.5382763383884814, "learning_rate": 1.9836893873364513e-07, "loss": 0.2695, "step": 12574 }, { "epoch": 5.91764705882353, "grad_norm": 0.5219134177157799, "learning_rate": 1.9612164015501696e-07, "loss": 0.2598, "step": 12575 }, { "epoch": 5.918117647058823, "grad_norm": 0.35879432668494493, "learning_rate": 1.9388713556239277e-07, "loss": 0.1006, "step": 12576 }, { "epoch": 5.918588235294117, "grad_norm": 0.41969111242941626, "learning_rate": 1.9166542514658656e-07, "loss": 0.1338, "step": 12577 }, { "epoch": 5.9190588235294115, "grad_norm": 0.6521483977682097, "learning_rate": 1.8945650909737985e-07, "loss": 0.2656, "step": 12578 }, { "epoch": 5.919529411764706, "grad_norm": 0.39669067759119947, "learning_rate": 1.8726038760338847e-07, "loss": 0.207, "step": 12579 }, { "epoch": 5.92, "grad_norm": 0.4577758402582968, "learning_rate": 1.8507706085219566e-07, "loss": 0.1699, "step": 12580 }, { "epoch": 5.920470588235294, "grad_norm": 0.592631972792001, "learning_rate": 1.8290652903025228e-07, "loss": 0.249, "step": 12581 }, { "epoch": 5.920941176470588, "grad_norm": 0.7426709889288475, "learning_rate": 1.8074879232296002e-07, "loss": 0.3145, "step": 12582 }, { "epoch": 5.921411764705883, "grad_norm": 0.6012527792097627, "learning_rate": 1.7860385091455488e-07, "loss": 0.2598, "step": 12583 }, { "epoch": 5.921882352941177, "grad_norm": 0.41971246056213735, "learning_rate": 1.764717049882569e-07, "loss": 0.2344, "step": 12584 }, { "epoch": 5.92235294117647, "grad_norm": 0.67043413676988, "learning_rate": 1.743523547261705e-07, "loss": 0.3281, "step": 12585 }, { "epoch": 5.922823529411764, "grad_norm": 0.49117439783698036, "learning_rate": 1.722458003092675e-07, "loss": 0.2988, "step": 12586 }, { "epoch": 5.923294117647059, "grad_norm": 0.49043941786457496, "learning_rate": 1.701520419174873e-07, "loss": 0.1475, "step": 12587 }, { "epoch": 5.923764705882353, "grad_norm": 0.33423340471284524, "learning_rate": 1.680710797296203e-07, "loss": 0.1318, "step": 12588 }, { "epoch": 5.924235294117647, "grad_norm": 0.4240938844789537, "learning_rate": 1.660029139234076e-07, "loss": 0.2139, "step": 12589 }, { "epoch": 5.924705882352941, "grad_norm": 0.29763178051432787, "learning_rate": 1.639475446754912e-07, "loss": 0.0718, "step": 12590 }, { "epoch": 5.9251764705882355, "grad_norm": 0.6455110369080879, "learning_rate": 1.6190497216138076e-07, "loss": 0.2656, "step": 12591 }, { "epoch": 5.92564705882353, "grad_norm": 0.828882577023011, "learning_rate": 1.598751965555367e-07, "loss": 0.4102, "step": 12592 }, { "epoch": 5.926117647058824, "grad_norm": 0.4729666830081644, "learning_rate": 1.5785821803132037e-07, "loss": 0.2871, "step": 12593 }, { "epoch": 5.926588235294117, "grad_norm": 0.25258709494552234, "learning_rate": 1.558540367609773e-07, "loss": 0.0654, "step": 12594 }, { "epoch": 5.9270588235294115, "grad_norm": 0.37148785831658887, "learning_rate": 1.5386265291568723e-07, "loss": 0.0947, "step": 12595 }, { "epoch": 5.927529411764706, "grad_norm": 0.3330924408479114, "learning_rate": 1.5188406666551412e-07, "loss": 0.1377, "step": 12596 }, { "epoch": 5.928, "grad_norm": 0.4911077060966843, "learning_rate": 1.4991827817943947e-07, "loss": 0.3359, "step": 12597 }, { "epoch": 5.928470588235294, "grad_norm": 0.265095632173157, "learning_rate": 1.479652876253623e-07, "loss": 0.0552, "step": 12598 }, { "epoch": 5.928941176470588, "grad_norm": 0.6259373702472495, "learning_rate": 1.4602509517006588e-07, "loss": 0.25, "step": 12599 }, { "epoch": 5.929411764705883, "grad_norm": 0.33087888627291107, "learning_rate": 1.4409770097926765e-07, "loss": 0.1245, "step": 12600 }, { "epoch": 5.929882352941177, "grad_norm": 0.44378860617478216, "learning_rate": 1.421831052175526e-07, "loss": 0.2773, "step": 12601 }, { "epoch": 5.930352941176471, "grad_norm": 0.6651885132373561, "learning_rate": 1.4028130804843995e-07, "loss": 0.2988, "step": 12602 }, { "epoch": 5.930823529411764, "grad_norm": 0.6346059022101846, "learning_rate": 1.3839230963436642e-07, "loss": 0.2715, "step": 12603 }, { "epoch": 5.931294117647059, "grad_norm": 0.7211090282397091, "learning_rate": 1.3651611013666964e-07, "loss": 0.3574, "step": 12604 }, { "epoch": 5.931764705882353, "grad_norm": 0.30668218117516294, "learning_rate": 1.346527097155714e-07, "loss": 0.0796, "step": 12605 }, { "epoch": 5.932235294117647, "grad_norm": 0.27035324499121527, "learning_rate": 1.3280210853019447e-07, "loss": 0.061, "step": 12606 }, { "epoch": 5.932705882352941, "grad_norm": 0.3700729016958049, "learning_rate": 1.3096430673862901e-07, "loss": 0.1157, "step": 12607 }, { "epoch": 5.9331764705882355, "grad_norm": 0.5024502512190288, "learning_rate": 1.2913930449781619e-07, "loss": 0.1973, "step": 12608 }, { "epoch": 5.93364705882353, "grad_norm": 0.08266149212724117, "learning_rate": 1.2732710196361463e-07, "loss": 0.0121, "step": 12609 }, { "epoch": 5.934117647058823, "grad_norm": 0.47588652812885707, "learning_rate": 1.2552769929081717e-07, "loss": 0.2373, "step": 12610 }, { "epoch": 5.934588235294117, "grad_norm": 0.8854643486033688, "learning_rate": 1.2374109663306763e-07, "loss": 0.3672, "step": 12611 }, { "epoch": 5.9350588235294115, "grad_norm": 0.45031785162800847, "learning_rate": 1.2196729414296058e-07, "loss": 0.1934, "step": 12612 }, { "epoch": 5.935529411764706, "grad_norm": 0.796345991794611, "learning_rate": 1.2020629197200814e-07, "loss": 0.4922, "step": 12613 }, { "epoch": 5.936, "grad_norm": 0.5863723735526535, "learning_rate": 1.1845809027059006e-07, "loss": 0.3496, "step": 12614 }, { "epoch": 5.936470588235294, "grad_norm": 0.7377409875398377, "learning_rate": 1.1672268918800353e-07, "loss": 0.4141, "step": 12615 }, { "epoch": 5.936941176470588, "grad_norm": 0.8630947436558543, "learning_rate": 1.1500008887249668e-07, "loss": 0.5312, "step": 12616 }, { "epoch": 5.937411764705883, "grad_norm": 0.2823626060745228, "learning_rate": 1.1329028947115181e-07, "loss": 0.083, "step": 12617 }, { "epoch": 5.937882352941177, "grad_norm": 0.6936198992699287, "learning_rate": 1.1159329113000215e-07, "loss": 0.3262, "step": 12618 }, { "epoch": 5.93835294117647, "grad_norm": 0.2072520933908833, "learning_rate": 1.0990909399398173e-07, "loss": 0.0737, "step": 12619 }, { "epoch": 5.938823529411764, "grad_norm": 0.24403532478620168, "learning_rate": 1.0823769820694218e-07, "loss": 0.0569, "step": 12620 }, { "epoch": 5.939294117647059, "grad_norm": 0.5675530876004454, "learning_rate": 1.0657910391161928e-07, "loss": 0.3965, "step": 12621 }, { "epoch": 5.939764705882353, "grad_norm": 0.589149770777491, "learning_rate": 1.0493331124964976e-07, "loss": 0.418, "step": 12622 }, { "epoch": 5.940235294117647, "grad_norm": 0.5004970467175277, "learning_rate": 1.0330032036162117e-07, "loss": 0.2637, "step": 12623 }, { "epoch": 5.940705882352941, "grad_norm": 0.1876275727936758, "learning_rate": 1.0168013138697196e-07, "loss": 0.0315, "step": 12624 }, { "epoch": 5.9411764705882355, "grad_norm": 0.4560351438188064, "learning_rate": 1.0007274446409141e-07, "loss": 0.1924, "step": 12625 }, { "epoch": 5.94164705882353, "grad_norm": 0.48687132001769096, "learning_rate": 9.847815973025308e-08, "loss": 0.1699, "step": 12626 }, { "epoch": 5.942117647058824, "grad_norm": 0.45150032567891274, "learning_rate": 9.689637732164802e-08, "loss": 0.166, "step": 12627 }, { "epoch": 5.942588235294117, "grad_norm": 0.4935291281756161, "learning_rate": 9.532739737336814e-08, "loss": 0.2793, "step": 12628 }, { "epoch": 5.9430588235294115, "grad_norm": 0.4225205922714356, "learning_rate": 9.377122001938963e-08, "loss": 0.3184, "step": 12629 }, { "epoch": 5.943529411764706, "grad_norm": 0.3977836460126386, "learning_rate": 9.222784539263949e-08, "loss": 0.1699, "step": 12630 }, { "epoch": 5.944, "grad_norm": 0.45292770247985015, "learning_rate": 9.069727362492896e-08, "loss": 0.2119, "step": 12631 }, { "epoch": 5.944470588235294, "grad_norm": 0.3667294544201088, "learning_rate": 8.917950484697012e-08, "loss": 0.1787, "step": 12632 }, { "epoch": 5.944941176470588, "grad_norm": 0.401453336771987, "learning_rate": 8.767453918839263e-08, "loss": 0.1187, "step": 12633 }, { "epoch": 5.945411764705883, "grad_norm": 0.3383439987256194, "learning_rate": 8.618237677772699e-08, "loss": 0.0952, "step": 12634 }, { "epoch": 5.945882352941177, "grad_norm": 0.6038058407062272, "learning_rate": 8.470301774240463e-08, "loss": 0.4531, "step": 12635 }, { "epoch": 5.946352941176471, "grad_norm": 0.36535055560129026, "learning_rate": 8.323646220877444e-08, "loss": 0.1113, "step": 12636 }, { "epoch": 5.946823529411764, "grad_norm": 0.5292176100207335, "learning_rate": 8.178271030210292e-08, "loss": 0.3457, "step": 12637 }, { "epoch": 5.947294117647059, "grad_norm": 0.35086533731231073, "learning_rate": 8.034176214650745e-08, "loss": 0.0913, "step": 12638 }, { "epoch": 5.947764705882353, "grad_norm": 0.36176572762406867, "learning_rate": 7.891361786508953e-08, "loss": 0.0762, "step": 12639 }, { "epoch": 5.948235294117647, "grad_norm": 0.6533934681166462, "learning_rate": 7.749827757981831e-08, "loss": 0.3906, "step": 12640 }, { "epoch": 5.948705882352941, "grad_norm": 0.40687015133234367, "learning_rate": 7.609574141154706e-08, "loss": 0.1465, "step": 12641 }, { "epoch": 5.9491764705882355, "grad_norm": 0.4030703647424811, "learning_rate": 7.470600948006333e-08, "loss": 0.1846, "step": 12642 }, { "epoch": 5.94964705882353, "grad_norm": 0.5058823980851609, "learning_rate": 7.332908190408882e-08, "loss": 0.293, "step": 12643 }, { "epoch": 5.950117647058823, "grad_norm": 0.44978276786633375, "learning_rate": 7.196495880117947e-08, "loss": 0.2021, "step": 12644 }, { "epoch": 5.950588235294117, "grad_norm": 0.5177088855847849, "learning_rate": 7.061364028785876e-08, "loss": 0.2773, "step": 12645 }, { "epoch": 5.9510588235294115, "grad_norm": 0.7896895299027985, "learning_rate": 6.927512647953437e-08, "loss": 0.373, "step": 12646 }, { "epoch": 5.951529411764706, "grad_norm": 0.6561476994200932, "learning_rate": 6.794941749053151e-08, "loss": 0.4277, "step": 12647 }, { "epoch": 5.952, "grad_norm": 0.552297450968013, "learning_rate": 6.663651343404297e-08, "loss": 0.3672, "step": 12648 }, { "epoch": 5.952470588235294, "grad_norm": 0.23731689555784982, "learning_rate": 6.533641442222904e-08, "loss": 0.0718, "step": 12649 }, { "epoch": 5.952941176470588, "grad_norm": 0.5980629363980458, "learning_rate": 6.40491205661009e-08, "loss": 0.2412, "step": 12650 }, { "epoch": 5.953411764705883, "grad_norm": 0.5628321899031378, "learning_rate": 6.277463197560395e-08, "loss": 0.3984, "step": 12651 }, { "epoch": 5.953882352941177, "grad_norm": 0.4741311958615409, "learning_rate": 6.151294875960111e-08, "loss": 0.2871, "step": 12652 }, { "epoch": 5.95435294117647, "grad_norm": 0.2544468812572502, "learning_rate": 6.026407102583952e-08, "loss": 0.0559, "step": 12653 }, { "epoch": 5.954823529411764, "grad_norm": 0.531664348967858, "learning_rate": 5.902799888096721e-08, "loss": 0.2148, "step": 12654 }, { "epoch": 5.955294117647059, "grad_norm": 0.6377564640788302, "learning_rate": 5.780473243056638e-08, "loss": 0.2305, "step": 12655 }, { "epoch": 5.955764705882353, "grad_norm": 0.48145568460479937, "learning_rate": 5.659427177910347e-08, "loss": 0.3691, "step": 12656 }, { "epoch": 5.956235294117647, "grad_norm": 0.3644994766806513, "learning_rate": 5.539661702996245e-08, "loss": 0.1338, "step": 12657 }, { "epoch": 5.956705882352941, "grad_norm": 0.21621629095110206, "learning_rate": 5.421176828541152e-08, "loss": 0.0569, "step": 12658 }, { "epoch": 5.9571764705882355, "grad_norm": 0.3416598843908404, "learning_rate": 5.30397256466697e-08, "loss": 0.0737, "step": 12659 }, { "epoch": 5.95764705882353, "grad_norm": 0.5017944404853989, "learning_rate": 5.1880489213806944e-08, "loss": 0.3633, "step": 12660 }, { "epoch": 5.958117647058824, "grad_norm": 0.7732903299813751, "learning_rate": 5.0734059085860704e-08, "loss": 0.1797, "step": 12661 }, { "epoch": 5.958588235294117, "grad_norm": 0.6892904210497292, "learning_rate": 4.9600435360702685e-08, "loss": 0.4023, "step": 12662 }, { "epoch": 5.9590588235294115, "grad_norm": 0.6303644044065808, "learning_rate": 4.847961813517209e-08, "loss": 0.3086, "step": 12663 }, { "epoch": 5.959529411764706, "grad_norm": 0.6082955054347328, "learning_rate": 4.737160750500901e-08, "loss": 0.2178, "step": 12664 }, { "epoch": 5.96, "grad_norm": 0.35382416793199817, "learning_rate": 4.6276403564804444e-08, "loss": 0.1235, "step": 12665 }, { "epoch": 5.960470588235294, "grad_norm": 0.48660935111410203, "learning_rate": 4.519400640811688e-08, "loss": 0.2412, "step": 12666 }, { "epoch": 5.960941176470588, "grad_norm": 0.36288879492418097, "learning_rate": 4.412441612738904e-08, "loss": 0.2246, "step": 12667 }, { "epoch": 5.961411764705883, "grad_norm": 0.4728999567146886, "learning_rate": 4.3067632813964524e-08, "loss": 0.1553, "step": 12668 }, { "epoch": 5.961882352941177, "grad_norm": 0.5570834183922817, "learning_rate": 4.2023656558104466e-08, "loss": 0.3125, "step": 12669 }, { "epoch": 5.962352941176471, "grad_norm": 0.5613041859828329, "learning_rate": 4.099248744897088e-08, "loss": 0.2285, "step": 12670 }, { "epoch": 5.962823529411764, "grad_norm": 0.5535888253124523, "learning_rate": 3.9974125574609994e-08, "loss": 0.2734, "step": 12671 }, { "epoch": 5.963294117647059, "grad_norm": 0.6271929136019732, "learning_rate": 3.896857102200224e-08, "loss": 0.3672, "step": 12672 }, { "epoch": 5.963764705882353, "grad_norm": 0.48849498715998957, "learning_rate": 3.797582387704556e-08, "loss": 0.2695, "step": 12673 }, { "epoch": 5.964235294117647, "grad_norm": 0.5989846062900204, "learning_rate": 3.69958842245055e-08, "loss": 0.3184, "step": 12674 }, { "epoch": 5.964705882352941, "grad_norm": 0.4832300739466724, "learning_rate": 3.6028752148081766e-08, "loss": 0.2373, "step": 12675 }, { "epoch": 5.9651764705882355, "grad_norm": 0.3200119617840182, "learning_rate": 3.5074427730374945e-08, "loss": 0.103, "step": 12676 }, { "epoch": 5.96564705882353, "grad_norm": 0.2406198319269537, "learning_rate": 3.4132911052886514e-08, "loss": 0.1045, "step": 12677 }, { "epoch": 5.966117647058823, "grad_norm": 0.5807383972198895, "learning_rate": 3.320420219603548e-08, "loss": 0.4102, "step": 12678 }, { "epoch": 5.966588235294117, "grad_norm": 0.36892530335979246, "learning_rate": 3.228830123910842e-08, "loss": 0.1157, "step": 12679 }, { "epoch": 5.9670588235294115, "grad_norm": 0.4289198228900091, "learning_rate": 3.13852082603594e-08, "loss": 0.1357, "step": 12680 }, { "epoch": 5.967529411764706, "grad_norm": 0.7148804718159829, "learning_rate": 3.049492333691006e-08, "loss": 0.5039, "step": 12681 }, { "epoch": 5.968, "grad_norm": 0.5004998360194735, "learning_rate": 2.961744654478293e-08, "loss": 0.3086, "step": 12682 }, { "epoch": 5.968470588235294, "grad_norm": 0.5372551777440544, "learning_rate": 2.875277795891806e-08, "loss": 0.2988, "step": 12683 }, { "epoch": 5.968941176470588, "grad_norm": 0.36194496652165065, "learning_rate": 2.7900917653189692e-08, "loss": 0.168, "step": 12684 }, { "epoch": 5.969411764705883, "grad_norm": 0.2536065341950233, "learning_rate": 2.706186570032298e-08, "loss": 0.0693, "step": 12685 }, { "epoch": 5.969882352941177, "grad_norm": 0.4779485657841658, "learning_rate": 2.6235622171977278e-08, "loss": 0.1982, "step": 12686 }, { "epoch": 5.97035294117647, "grad_norm": 0.44688910434050644, "learning_rate": 2.542218713874611e-08, "loss": 0.2285, "step": 12687 }, { "epoch": 5.970823529411764, "grad_norm": 0.7804115983157166, "learning_rate": 2.462156067005727e-08, "loss": 0.5352, "step": 12688 }, { "epoch": 5.971294117647059, "grad_norm": 0.559678832711546, "learning_rate": 2.3833742834339365e-08, "loss": 0.3047, "step": 12689 }, { "epoch": 5.971764705882353, "grad_norm": 0.5243798010369524, "learning_rate": 2.305873369882194e-08, "loss": 0.2754, "step": 12690 }, { "epoch": 5.972235294117647, "grad_norm": 0.46597374623091886, "learning_rate": 2.2296533329735356e-08, "loss": 0.25, "step": 12691 }, { "epoch": 5.972705882352941, "grad_norm": 0.5060963781677861, "learning_rate": 2.1547141792160882e-08, "loss": 0.1157, "step": 12692 }, { "epoch": 5.9731764705882355, "grad_norm": 0.9791176582238444, "learning_rate": 2.0810559150097327e-08, "loss": 0.6094, "step": 12693 }, { "epoch": 5.97364705882353, "grad_norm": 0.49050991312986375, "learning_rate": 2.008678546646103e-08, "loss": 0.293, "step": 12694 }, { "epoch": 5.974117647058824, "grad_norm": 0.47917283754871776, "learning_rate": 1.9375820803052555e-08, "loss": 0.2051, "step": 12695 }, { "epoch": 5.974588235294117, "grad_norm": 0.7461191378155299, "learning_rate": 1.8677665220606653e-08, "loss": 0.2852, "step": 12696 }, { "epoch": 5.9750588235294115, "grad_norm": 0.778936868360311, "learning_rate": 1.7992318778742298e-08, "loss": 0.3398, "step": 12697 }, { "epoch": 5.975529411764706, "grad_norm": 0.305804503857753, "learning_rate": 1.7319781535996002e-08, "loss": 0.0991, "step": 12698 }, { "epoch": 5.976, "grad_norm": 0.6069051845972413, "learning_rate": 1.6660053549788498e-08, "loss": 0.3496, "step": 12699 }, { "epoch": 5.976470588235294, "grad_norm": 0.36769621269697583, "learning_rate": 1.6013134876491362e-08, "loss": 0.165, "step": 12700 }, { "epoch": 5.976941176470588, "grad_norm": 0.6217934733393986, "learning_rate": 1.5379025571327088e-08, "loss": 0.2305, "step": 12701 }, { "epoch": 5.977411764705883, "grad_norm": 0.42117335065198536, "learning_rate": 1.4757725688469002e-08, "loss": 0.2471, "step": 12702 }, { "epoch": 5.977882352941176, "grad_norm": 0.5624617564836059, "learning_rate": 1.414923528097467e-08, "loss": 0.2754, "step": 12703 }, { "epoch": 5.97835294117647, "grad_norm": 0.37201356627808996, "learning_rate": 1.3553554400802524e-08, "loss": 0.1562, "step": 12704 }, { "epoch": 5.978823529411764, "grad_norm": 0.37764221881488386, "learning_rate": 1.2970683098845192e-08, "loss": 0.2002, "step": 12705 }, { "epoch": 5.979294117647059, "grad_norm": 0.30748459966370195, "learning_rate": 1.240062142487952e-08, "loss": 0.103, "step": 12706 }, { "epoch": 5.979764705882353, "grad_norm": 0.28517457716128036, "learning_rate": 1.1843369427583238e-08, "loss": 0.0845, "step": 12707 }, { "epoch": 5.980235294117647, "grad_norm": 0.5644324357842699, "learning_rate": 1.1298927154551608e-08, "loss": 0.416, "step": 12708 }, { "epoch": 5.980705882352941, "grad_norm": 0.05114337213272146, "learning_rate": 1.0767294652280767e-08, "loss": 0.0087, "step": 12709 }, { "epoch": 5.9811764705882355, "grad_norm": 0.13171881028627344, "learning_rate": 1.0248471966167737e-08, "loss": 0.0125, "step": 12710 }, { "epoch": 5.98164705882353, "grad_norm": 0.4503238784295067, "learning_rate": 9.74245914052707e-09, "loss": 0.2217, "step": 12711 }, { "epoch": 5.982117647058823, "grad_norm": 0.664110226363492, "learning_rate": 9.24925621859085e-09, "loss": 0.3633, "step": 12712 }, { "epoch": 5.982588235294117, "grad_norm": 0.5792498082813332, "learning_rate": 8.768863242475389e-09, "loss": 0.1973, "step": 12713 }, { "epoch": 5.9830588235294115, "grad_norm": 0.336626405428433, "learning_rate": 8.30128025318122e-09, "loss": 0.1367, "step": 12714 }, { "epoch": 5.983529411764706, "grad_norm": 0.6958613971507637, "learning_rate": 7.846507290676373e-09, "loss": 0.5547, "step": 12715 }, { "epoch": 5.984, "grad_norm": 0.3330949225569356, "learning_rate": 7.404544393779799e-09, "loss": 0.1797, "step": 12716 }, { "epoch": 5.984470588235294, "grad_norm": 0.7265141204966064, "learning_rate": 6.975391600261282e-09, "loss": 0.2266, "step": 12717 }, { "epoch": 5.984941176470588, "grad_norm": 0.34458136306448517, "learning_rate": 6.55904894674153e-09, "loss": 0.1245, "step": 12718 }, { "epoch": 5.985411764705883, "grad_norm": 0.2576953216931293, "learning_rate": 6.155516468792088e-09, "loss": 0.0815, "step": 12719 }, { "epoch": 5.985882352941177, "grad_norm": 0.2714658761781946, "learning_rate": 5.764794200885381e-09, "loss": 0.064, "step": 12720 }, { "epoch": 5.98635294117647, "grad_norm": 0.36749800444782976, "learning_rate": 5.38688217637806e-09, "loss": 0.103, "step": 12721 }, { "epoch": 5.986823529411764, "grad_norm": 0.4111196071344931, "learning_rate": 5.0217804275443084e-09, "loss": 0.2598, "step": 12722 }, { "epoch": 5.987294117647059, "grad_norm": 0.43931468294110837, "learning_rate": 4.669488985575842e-09, "loss": 0.3027, "step": 12723 }, { "epoch": 5.987764705882353, "grad_norm": 0.2774467494620989, "learning_rate": 4.3300078805652564e-09, "loss": 0.0659, "step": 12724 }, { "epoch": 5.988235294117647, "grad_norm": 0.286444076882457, "learning_rate": 4.003337141489371e-09, "loss": 0.0923, "step": 12725 }, { "epoch": 5.988705882352941, "grad_norm": 0.38403321281348957, "learning_rate": 3.689476796242541e-09, "loss": 0.1602, "step": 12726 }, { "epoch": 5.9891764705882355, "grad_norm": 0.43484552269393806, "learning_rate": 3.388426871653305e-09, "loss": 0.1982, "step": 12727 }, { "epoch": 5.98964705882353, "grad_norm": 0.6316532288062314, "learning_rate": 3.100187393434428e-09, "loss": 0.3887, "step": 12728 }, { "epoch": 5.990117647058824, "grad_norm": 0.4226299625892364, "learning_rate": 2.8247583861662484e-09, "loss": 0.1855, "step": 12729 }, { "epoch": 5.990588235294117, "grad_norm": 0.4266961228673246, "learning_rate": 2.562139873413249e-09, "loss": 0.1182, "step": 12730 }, { "epoch": 5.9910588235294115, "grad_norm": 0.28626912086598544, "learning_rate": 2.3123318775741805e-09, "loss": 0.084, "step": 12731 }, { "epoch": 5.991529411764706, "grad_norm": 0.5252047551699398, "learning_rate": 2.075334419998631e-09, "loss": 0.2334, "step": 12732 }, { "epoch": 5.992, "grad_norm": 0.5984837988388658, "learning_rate": 1.8511475209370685e-09, "loss": 0.2754, "step": 12733 }, { "epoch": 5.992470588235294, "grad_norm": 0.6033204223196117, "learning_rate": 1.6397711995241869e-09, "loss": 0.4727, "step": 12734 }, { "epoch": 5.992941176470588, "grad_norm": 0.17007345055710543, "learning_rate": 1.441205473795559e-09, "loss": 0.0383, "step": 12735 }, { "epoch": 5.993411764705883, "grad_norm": 0.48271825701712606, "learning_rate": 1.2554503607375976e-09, "loss": 0.2793, "step": 12736 }, { "epoch": 5.993882352941176, "grad_norm": 0.3697127500575653, "learning_rate": 1.082505876204287e-09, "loss": 0.1216, "step": 12737 }, { "epoch": 5.99435294117647, "grad_norm": 0.3391161869643749, "learning_rate": 9.223720349671449e-10, "loss": 0.1328, "step": 12738 }, { "epoch": 5.994823529411764, "grad_norm": 0.5375892479264835, "learning_rate": 7.750488506985674e-10, "loss": 0.2246, "step": 12739 }, { "epoch": 5.995294117647059, "grad_norm": 0.9289108992725962, "learning_rate": 6.405363359718308e-10, "loss": 0.6836, "step": 12740 }, { "epoch": 5.995764705882353, "grad_norm": 0.4601347488726678, "learning_rate": 5.188345022943963e-10, "loss": 0.2754, "step": 12741 }, { "epoch": 5.996235294117647, "grad_norm": 0.18083754284202183, "learning_rate": 4.099433600579516e-10, "loss": 0.0378, "step": 12742 }, { "epoch": 5.996705882352941, "grad_norm": 0.20523896538156558, "learning_rate": 3.138629185550634e-10, "loss": 0.0571, "step": 12743 }, { "epoch": 5.9971764705882356, "grad_norm": 0.4170472848623673, "learning_rate": 2.3059318597917763e-10, "loss": 0.1309, "step": 12744 }, { "epoch": 5.99764705882353, "grad_norm": 0.46713271657304684, "learning_rate": 1.6013416947457946e-10, "loss": 0.1641, "step": 12745 }, { "epoch": 5.998117647058823, "grad_norm": 0.6446400367807935, "learning_rate": 1.0248587501981986e-10, "loss": 0.3867, "step": 12746 }, { "epoch": 5.998588235294117, "grad_norm": 0.2524479514965446, "learning_rate": 5.7648307577595756e-11, "loss": 0.0908, "step": 12747 }, { "epoch": 5.9990588235294116, "grad_norm": 0.3799780149256822, "learning_rate": 2.5621470944869883e-11, "loss": 0.1191, "step": 12748 }, { "epoch": 5.999529411764706, "grad_norm": 0.45216947453654455, "learning_rate": 6.4053678694442335e-12, "loss": 0.3574, "step": 12749 }, { "epoch": 6.0, "grad_norm": 0.3187440048189806, "learning_rate": 0.0, "loss": 0.0952, "step": 12750 }, { "epoch": 6.0, "eval_loss": 0.2419528067111969, "eval_runtime": 62.5339, "eval_samples_per_second": 11.993, "eval_steps_per_second": 5.997, "step": 12750 }, { "epoch": 6.0, "step": 12750, "total_flos": 35696972513280.0, "train_loss": 0.3865196811496189, "train_runtime": 430060.8609, "train_samples_per_second": 0.059, "train_steps_per_second": 0.03 } ], "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": 35696972513280.0, "train_batch_size": 2, "trial_name": null, "trial_params": null }