{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.9587628865979383,
  "eval_steps": 5,
  "global_step": 240,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.016494845360824743,
      "grad_norm": 5.5,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 0.569,
      "step": 1
    },
    {
      "epoch": 0.032989690721649485,
      "grad_norm": 3.21875,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 0.6647,
      "step": 2
    },
    {
      "epoch": 0.049484536082474224,
      "grad_norm": 2.46875,
      "learning_rate": 3e-06,
      "loss": 0.8035,
      "step": 3
    },
    {
      "epoch": 0.06597938144329897,
      "grad_norm": 5.0625,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.7259,
      "step": 4
    },
    {
      "epoch": 0.08247422680412371,
      "grad_norm": 3.984375,
      "learning_rate": 5e-06,
      "loss": 1.3804,
      "step": 5
    },
    {
      "epoch": 0.08247422680412371,
      "eval_loss": 0.5400243401527405,
      "eval_runtime": 1.246,
      "eval_samples_per_second": 20.866,
      "eval_steps_per_second": 20.866,
      "step": 5
    },
    {
      "epoch": 0.09896907216494845,
      "grad_norm": 5.71875,
      "learning_rate": 6e-06,
      "loss": 0.7447,
      "step": 6
    },
    {
      "epoch": 0.1154639175257732,
      "grad_norm": 4.65625,
      "learning_rate": 7e-06,
      "loss": 0.5788,
      "step": 7
    },
    {
      "epoch": 0.13195876288659794,
      "grad_norm": 4.3125,
      "learning_rate": 8.000000000000001e-06,
      "loss": 0.394,
      "step": 8
    },
    {
      "epoch": 0.14845360824742268,
      "grad_norm": 5.0,
      "learning_rate": 9e-06,
      "loss": 0.5634,
      "step": 9
    },
    {
      "epoch": 0.16494845360824742,
      "grad_norm": 6.21875,
      "learning_rate": 1e-05,
      "loss": 0.3369,
      "step": 10
    },
    {
      "epoch": 0.16494845360824742,
      "eval_loss": 0.5371499061584473,
      "eval_runtime": 1.2372,
      "eval_samples_per_second": 21.016,
      "eval_steps_per_second": 21.016,
      "step": 10
    },
    {
      "epoch": 0.18144329896907216,
      "grad_norm": 4.0,
      "learning_rate": 9.956521739130436e-06,
      "loss": 0.5154,
      "step": 11
    },
    {
      "epoch": 0.1979381443298969,
      "grad_norm": 6.21875,
      "learning_rate": 9.913043478260871e-06,
      "loss": 0.5452,
      "step": 12
    },
    {
      "epoch": 0.21443298969072164,
      "grad_norm": 7.09375,
      "learning_rate": 9.869565217391304e-06,
      "loss": 0.8781,
      "step": 13
    },
    {
      "epoch": 0.2309278350515464,
      "grad_norm": 4.5,
      "learning_rate": 9.82608695652174e-06,
      "loss": 0.8791,
      "step": 14
    },
    {
      "epoch": 0.24742268041237114,
      "grad_norm": 9.6875,
      "learning_rate": 9.782608695652175e-06,
      "loss": 0.7528,
      "step": 15
    },
    {
      "epoch": 0.24742268041237114,
      "eval_loss": 0.5349909663200378,
      "eval_runtime": 1.2584,
      "eval_samples_per_second": 20.66,
      "eval_steps_per_second": 20.66,
      "step": 15
    },
    {
      "epoch": 0.2639175257731959,
      "grad_norm": 3.03125,
      "learning_rate": 9.73913043478261e-06,
      "loss": 0.3554,
      "step": 16
    },
    {
      "epoch": 0.2804123711340206,
      "grad_norm": 3.609375,
      "learning_rate": 9.695652173913043e-06,
      "loss": 0.4604,
      "step": 17
    },
    {
      "epoch": 0.29690721649484536,
      "grad_norm": 4.40625,
      "learning_rate": 9.652173913043478e-06,
      "loss": 0.6079,
      "step": 18
    },
    {
      "epoch": 0.3134020618556701,
      "grad_norm": 4.96875,
      "learning_rate": 9.608695652173914e-06,
      "loss": 0.5466,
      "step": 19
    },
    {
      "epoch": 0.32989690721649484,
      "grad_norm": 6.5,
      "learning_rate": 9.565217391304349e-06,
      "loss": 0.6961,
      "step": 20
    },
    {
      "epoch": 0.32989690721649484,
      "eval_loss": 0.5351170897483826,
      "eval_runtime": 1.2599,
      "eval_samples_per_second": 20.636,
      "eval_steps_per_second": 20.636,
      "step": 20
    },
    {
      "epoch": 0.3463917525773196,
      "grad_norm": 6.0,
      "learning_rate": 9.521739130434784e-06,
      "loss": 0.9412,
      "step": 21
    },
    {
      "epoch": 0.3628865979381443,
      "grad_norm": 5.6875,
      "learning_rate": 9.478260869565217e-06,
      "loss": 0.5747,
      "step": 22
    },
    {
      "epoch": 0.37938144329896906,
      "grad_norm": 8.625,
      "learning_rate": 9.434782608695652e-06,
      "loss": 0.7224,
      "step": 23
    },
    {
      "epoch": 0.3958762886597938,
      "grad_norm": 6.84375,
      "learning_rate": 9.391304347826087e-06,
      "loss": 0.588,
      "step": 24
    },
    {
      "epoch": 0.41237113402061853,
      "grad_norm": 4.71875,
      "learning_rate": 9.347826086956523e-06,
      "loss": 0.4478,
      "step": 25
    },
    {
      "epoch": 0.41237113402061853,
      "eval_loss": 0.5326387882232666,
      "eval_runtime": 1.2718,
      "eval_samples_per_second": 20.444,
      "eval_steps_per_second": 20.444,
      "step": 25
    },
    {
      "epoch": 0.4288659793814433,
      "grad_norm": 8.125,
      "learning_rate": 9.304347826086956e-06,
      "loss": 0.5952,
      "step": 26
    },
    {
      "epoch": 0.44536082474226807,
      "grad_norm": 3.90625,
      "learning_rate": 9.260869565217391e-06,
      "loss": 0.5545,
      "step": 27
    },
    {
      "epoch": 0.4618556701030928,
      "grad_norm": 6.125,
      "learning_rate": 9.217391304347826e-06,
      "loss": 0.6788,
      "step": 28
    },
    {
      "epoch": 0.47835051546391755,
      "grad_norm": 7.5,
      "learning_rate": 9.173913043478261e-06,
      "loss": 0.5714,
      "step": 29
    },
    {
      "epoch": 0.4948453608247423,
      "grad_norm": 4.75,
      "learning_rate": 9.130434782608697e-06,
      "loss": 0.5826,
      "step": 30
    },
    {
      "epoch": 0.4948453608247423,
      "eval_loss": 0.53000408411026,
      "eval_runtime": 1.2766,
      "eval_samples_per_second": 20.366,
      "eval_steps_per_second": 20.366,
      "step": 30
    },
    {
      "epoch": 0.511340206185567,
      "grad_norm": 6.875,
      "learning_rate": 9.086956521739132e-06,
      "loss": 0.5211,
      "step": 31
    },
    {
      "epoch": 0.5278350515463918,
      "grad_norm": 3.734375,
      "learning_rate": 9.043478260869565e-06,
      "loss": 0.4774,
      "step": 32
    },
    {
      "epoch": 0.5443298969072164,
      "grad_norm": 6.4375,
      "learning_rate": 9e-06,
      "loss": 0.7318,
      "step": 33
    },
    {
      "epoch": 0.5608247422680412,
      "grad_norm": 5.75,
      "learning_rate": 8.956521739130435e-06,
      "loss": 0.5202,
      "step": 34
    },
    {
      "epoch": 0.5773195876288659,
      "grad_norm": 4.625,
      "learning_rate": 8.91304347826087e-06,
      "loss": 0.6316,
      "step": 35
    },
    {
      "epoch": 0.5773195876288659,
      "eval_loss": 0.5289856195449829,
      "eval_runtime": 1.2887,
      "eval_samples_per_second": 20.175,
      "eval_steps_per_second": 20.175,
      "step": 35
    },
    {
      "epoch": 0.5938144329896907,
      "grad_norm": 5.0,
      "learning_rate": 8.869565217391306e-06,
      "loss": 0.396,
      "step": 36
    },
    {
      "epoch": 0.6103092783505155,
      "grad_norm": 3.65625,
      "learning_rate": 8.82608695652174e-06,
      "loss": 0.2928,
      "step": 37
    },
    {
      "epoch": 0.6268041237113402,
      "grad_norm": 7.28125,
      "learning_rate": 8.782608695652174e-06,
      "loss": 0.6256,
      "step": 38
    },
    {
      "epoch": 0.643298969072165,
      "grad_norm": 4.8125,
      "learning_rate": 8.73913043478261e-06,
      "loss": 0.5751,
      "step": 39
    },
    {
      "epoch": 0.6597938144329897,
      "grad_norm": 3.015625,
      "learning_rate": 8.695652173913044e-06,
      "loss": 1.1084,
      "step": 40
    },
    {
      "epoch": 0.6597938144329897,
      "eval_loss": 0.5289974808692932,
      "eval_runtime": 1.2832,
      "eval_samples_per_second": 20.261,
      "eval_steps_per_second": 20.261,
      "step": 40
    },
    {
      "epoch": 0.6762886597938145,
      "grad_norm": 5.625,
      "learning_rate": 8.65217391304348e-06,
      "loss": 0.5896,
      "step": 41
    },
    {
      "epoch": 0.6927835051546392,
      "grad_norm": 5.46875,
      "learning_rate": 8.608695652173915e-06,
      "loss": 0.5278,
      "step": 42
    },
    {
      "epoch": 0.709278350515464,
      "grad_norm": 4.90625,
      "learning_rate": 8.56521739130435e-06,
      "loss": 0.4396,
      "step": 43
    },
    {
      "epoch": 0.7257731958762886,
      "grad_norm": 3.546875,
      "learning_rate": 8.521739130434783e-06,
      "loss": 0.6561,
      "step": 44
    },
    {
      "epoch": 0.7422680412371134,
      "grad_norm": 4.8125,
      "learning_rate": 8.478260869565218e-06,
      "loss": 0.3966,
      "step": 45
    },
    {
      "epoch": 0.7422680412371134,
      "eval_loss": 0.5284722447395325,
      "eval_runtime": 1.2831,
      "eval_samples_per_second": 20.264,
      "eval_steps_per_second": 20.264,
      "step": 45
    },
    {
      "epoch": 0.7587628865979381,
      "grad_norm": 4.375,
      "learning_rate": 8.434782608695653e-06,
      "loss": 0.2738,
      "step": 46
    },
    {
      "epoch": 0.7752577319587629,
      "grad_norm": 5.78125,
      "learning_rate": 8.391304347826089e-06,
      "loss": 0.7747,
      "step": 47
    },
    {
      "epoch": 0.7917525773195876,
      "grad_norm": 4.3125,
      "learning_rate": 8.347826086956522e-06,
      "loss": 1.089,
      "step": 48
    },
    {
      "epoch": 0.8082474226804124,
      "grad_norm": 3.578125,
      "learning_rate": 8.304347826086957e-06,
      "loss": 0.301,
      "step": 49
    },
    {
      "epoch": 0.8247422680412371,
      "grad_norm": 5.40625,
      "learning_rate": 8.260869565217392e-06,
      "loss": 0.6264,
      "step": 50
    },
    {
      "epoch": 0.8247422680412371,
      "eval_loss": 0.5277838110923767,
      "eval_runtime": 1.2877,
      "eval_samples_per_second": 20.192,
      "eval_steps_per_second": 20.192,
      "step": 50
    },
    {
      "epoch": 0.8412371134020619,
      "grad_norm": 4.71875,
      "learning_rate": 8.217391304347827e-06,
      "loss": 0.6022,
      "step": 51
    },
    {
      "epoch": 0.8577319587628865,
      "grad_norm": 4.96875,
      "learning_rate": 8.173913043478263e-06,
      "loss": 0.2717,
      "step": 52
    },
    {
      "epoch": 0.8742268041237113,
      "grad_norm": 4.1875,
      "learning_rate": 8.130434782608696e-06,
      "loss": 0.4995,
      "step": 53
    },
    {
      "epoch": 0.8907216494845361,
      "grad_norm": 4.90625,
      "learning_rate": 8.086956521739131e-06,
      "loss": 0.5654,
      "step": 54
    },
    {
      "epoch": 0.9072164948453608,
      "grad_norm": 6.78125,
      "learning_rate": 8.043478260869566e-06,
      "loss": 0.5308,
      "step": 55
    },
    {
      "epoch": 0.9072164948453608,
      "eval_loss": 0.5283468961715698,
      "eval_runtime": 1.2896,
      "eval_samples_per_second": 20.161,
      "eval_steps_per_second": 20.161,
      "step": 55
    },
    {
      "epoch": 0.9237113402061856,
      "grad_norm": 3.515625,
      "learning_rate": 8.000000000000001e-06,
      "loss": 0.6069,
      "step": 56
    },
    {
      "epoch": 0.9402061855670103,
      "grad_norm": 4.03125,
      "learning_rate": 7.956521739130435e-06,
      "loss": 0.3835,
      "step": 57
    },
    {
      "epoch": 0.9567010309278351,
      "grad_norm": 4.15625,
      "learning_rate": 7.91304347826087e-06,
      "loss": 0.5169,
      "step": 58
    },
    {
      "epoch": 0.9731958762886598,
      "grad_norm": 2.796875,
      "learning_rate": 7.869565217391305e-06,
      "loss": 0.4643,
      "step": 59
    },
    {
      "epoch": 0.9896907216494846,
      "grad_norm": 5.46875,
      "learning_rate": 7.82608695652174e-06,
      "loss": 0.6768,
      "step": 60
    },
    {
      "epoch": 0.9896907216494846,
      "eval_loss": 0.5277718305587769,
      "eval_runtime": 1.2735,
      "eval_samples_per_second": 20.416,
      "eval_steps_per_second": 20.416,
      "step": 60
    },
    {
      "epoch": 1.0061855670103093,
      "grad_norm": 8.625,
      "learning_rate": 7.782608695652174e-06,
      "loss": 0.7185,
      "step": 61
    },
    {
      "epoch": 1.022680412371134,
      "grad_norm": 3.78125,
      "learning_rate": 7.739130434782609e-06,
      "loss": 0.5707,
      "step": 62
    },
    {
      "epoch": 1.0391752577319588,
      "grad_norm": 3.46875,
      "learning_rate": 7.695652173913044e-06,
      "loss": 0.4542,
      "step": 63
    },
    {
      "epoch": 1.0556701030927835,
      "grad_norm": 3.015625,
      "learning_rate": 7.652173913043479e-06,
      "loss": 0.3663,
      "step": 64
    },
    {
      "epoch": 1.0721649484536082,
      "grad_norm": 2.046875,
      "learning_rate": 7.608695652173914e-06,
      "loss": 0.4641,
      "step": 65
    },
    {
      "epoch": 1.0721649484536082,
      "eval_loss": 0.5268765091896057,
      "eval_runtime": 1.2807,
      "eval_samples_per_second": 20.302,
      "eval_steps_per_second": 20.302,
      "step": 65
    },
    {
      "epoch": 1.088659793814433,
      "grad_norm": 2.640625,
      "learning_rate": 7.565217391304348e-06,
      "loss": 0.6291,
      "step": 66
    },
    {
      "epoch": 1.1051546391752578,
      "grad_norm": 2.53125,
      "learning_rate": 7.5217391304347835e-06,
      "loss": 0.2755,
      "step": 67
    },
    {
      "epoch": 1.1216494845360825,
      "grad_norm": 2.734375,
      "learning_rate": 7.478260869565218e-06,
      "loss": 0.4472,
      "step": 68
    },
    {
      "epoch": 1.1381443298969072,
      "grad_norm": 2.96875,
      "learning_rate": 7.434782608695653e-06,
      "loss": 0.4451,
      "step": 69
    },
    {
      "epoch": 1.1546391752577319,
      "grad_norm": 5.34375,
      "learning_rate": 7.391304347826087e-06,
      "loss": 0.9232,
      "step": 70
    },
    {
      "epoch": 1.1546391752577319,
      "eval_loss": 0.5259441137313843,
      "eval_runtime": 1.2903,
      "eval_samples_per_second": 20.151,
      "eval_steps_per_second": 20.151,
      "step": 70
    },
    {
      "epoch": 1.1711340206185568,
      "grad_norm": 4.53125,
      "learning_rate": 7.347826086956522e-06,
      "loss": 0.3537,
      "step": 71
    },
    {
      "epoch": 1.1876288659793814,
      "grad_norm": 3.84375,
      "learning_rate": 7.304347826086957e-06,
      "loss": 0.3703,
      "step": 72
    },
    {
      "epoch": 1.2041237113402061,
      "grad_norm": 4.84375,
      "learning_rate": 7.2608695652173925e-06,
      "loss": 1.0391,
      "step": 73
    },
    {
      "epoch": 1.220618556701031,
      "grad_norm": 3.71875,
      "learning_rate": 7.217391304347827e-06,
      "loss": 0.5239,
      "step": 74
    },
    {
      "epoch": 1.2371134020618557,
      "grad_norm": 2.96875,
      "learning_rate": 7.173913043478261e-06,
      "loss": 0.2765,
      "step": 75
    },
    {
      "epoch": 1.2371134020618557,
      "eval_loss": 0.5257338285446167,
      "eval_runtime": 1.2846,
      "eval_samples_per_second": 20.24,
      "eval_steps_per_second": 20.24,
      "step": 75
    },
    {
      "epoch": 1.2536082474226804,
      "grad_norm": 2.96875,
      "learning_rate": 7.130434782608696e-06,
      "loss": 1.3695,
      "step": 76
    },
    {
      "epoch": 1.270103092783505,
      "grad_norm": 5.0,
      "learning_rate": 7.086956521739131e-06,
      "loss": 0.7998,
      "step": 77
    },
    {
      "epoch": 1.2865979381443298,
      "grad_norm": 2.546875,
      "learning_rate": 7.0434782608695665e-06,
      "loss": 0.3198,
      "step": 78
    },
    {
      "epoch": 1.3030927835051547,
      "grad_norm": 2.875,
      "learning_rate": 7e-06,
      "loss": 0.848,
      "step": 79
    },
    {
      "epoch": 1.3195876288659794,
      "grad_norm": 1.7578125,
      "learning_rate": 6.956521739130435e-06,
      "loss": 0.2294,
      "step": 80
    },
    {
      "epoch": 1.3195876288659794,
      "eval_loss": 0.5260586142539978,
      "eval_runtime": 1.2823,
      "eval_samples_per_second": 20.277,
      "eval_steps_per_second": 20.277,
      "step": 80
    },
    {
      "epoch": 1.3360824742268043,
      "grad_norm": 3.546875,
      "learning_rate": 6.91304347826087e-06,
      "loss": 1.289,
      "step": 81
    },
    {
      "epoch": 1.352577319587629,
      "grad_norm": 2.34375,
      "learning_rate": 6.869565217391305e-06,
      "loss": 0.4023,
      "step": 82
    },
    {
      "epoch": 1.3690721649484536,
      "grad_norm": 2.5625,
      "learning_rate": 6.8260869565217395e-06,
      "loss": 0.3378,
      "step": 83
    },
    {
      "epoch": 1.3855670103092783,
      "grad_norm": 6.03125,
      "learning_rate": 6.782608695652174e-06,
      "loss": 0.7623,
      "step": 84
    },
    {
      "epoch": 1.402061855670103,
      "grad_norm": 4.21875,
      "learning_rate": 6.739130434782609e-06,
      "loss": 0.7804,
      "step": 85
    },
    {
      "epoch": 1.402061855670103,
      "eval_loss": 0.5257543325424194,
      "eval_runtime": 1.2932,
      "eval_samples_per_second": 20.106,
      "eval_steps_per_second": 20.106,
      "step": 85
    },
    {
      "epoch": 1.418556701030928,
      "grad_norm": 3.46875,
      "learning_rate": 6.695652173913044e-06,
      "loss": 0.5337,
      "step": 86
    },
    {
      "epoch": 1.4350515463917526,
      "grad_norm": 2.71875,
      "learning_rate": 6.652173913043479e-06,
      "loss": 0.4078,
      "step": 87
    },
    {
      "epoch": 1.4515463917525773,
      "grad_norm": 2.3125,
      "learning_rate": 6.6086956521739135e-06,
      "loss": 0.3687,
      "step": 88
    },
    {
      "epoch": 1.4680412371134022,
      "grad_norm": 1.59375,
      "learning_rate": 6.565217391304349e-06,
      "loss": 0.1253,
      "step": 89
    },
    {
      "epoch": 1.4845360824742269,
      "grad_norm": 3.9375,
      "learning_rate": 6.521739130434783e-06,
      "loss": 0.8425,
      "step": 90
    },
    {
      "epoch": 1.4845360824742269,
      "eval_loss": 0.5253785848617554,
      "eval_runtime": 1.2875,
      "eval_samples_per_second": 20.194,
      "eval_steps_per_second": 20.194,
      "step": 90
    },
    {
      "epoch": 1.5010309278350515,
      "grad_norm": 3.53125,
      "learning_rate": 6.478260869565218e-06,
      "loss": 0.4739,
      "step": 91
    },
    {
      "epoch": 1.5175257731958762,
      "grad_norm": 3.3125,
      "learning_rate": 6.434782608695652e-06,
      "loss": 0.9112,
      "step": 92
    },
    {
      "epoch": 1.534020618556701,
      "grad_norm": 3.984375,
      "learning_rate": 6.391304347826087e-06,
      "loss": 0.8557,
      "step": 93
    },
    {
      "epoch": 1.5505154639175258,
      "grad_norm": 3.0,
      "learning_rate": 6.3478260869565225e-06,
      "loss": 0.401,
      "step": 94
    },
    {
      "epoch": 1.5670103092783505,
      "grad_norm": 2.859375,
      "learning_rate": 6.304347826086958e-06,
      "loss": 0.3689,
      "step": 95
    },
    {
      "epoch": 1.5670103092783505,
      "eval_loss": 0.5263558030128479,
      "eval_runtime": 1.2892,
      "eval_samples_per_second": 20.168,
      "eval_steps_per_second": 20.168,
      "step": 95
    },
    {
      "epoch": 1.5835051546391754,
      "grad_norm": 4.28125,
      "learning_rate": 6.260869565217392e-06,
      "loss": 0.7863,
      "step": 96
    },
    {
      "epoch": 1.6,
      "grad_norm": 2.578125,
      "learning_rate": 6.217391304347826e-06,
      "loss": 0.3845,
      "step": 97
    },
    {
      "epoch": 1.6164948453608248,
      "grad_norm": 2.265625,
      "learning_rate": 6.173913043478261e-06,
      "loss": 0.8762,
      "step": 98
    },
    {
      "epoch": 1.6329896907216495,
      "grad_norm": 3.1875,
      "learning_rate": 6.1304347826086965e-06,
      "loss": 0.6063,
      "step": 99
    },
    {
      "epoch": 1.6494845360824741,
      "grad_norm": 3.5625,
      "learning_rate": 6.086956521739132e-06,
      "loss": 0.6211,
      "step": 100
    },
    {
      "epoch": 1.6494845360824741,
      "eval_loss": 0.5265222787857056,
      "eval_runtime": 1.2886,
      "eval_samples_per_second": 20.176,
      "eval_steps_per_second": 20.176,
      "step": 100
    },
    {
      "epoch": 1.6659793814432988,
      "grad_norm": 4.25,
      "learning_rate": 6.043478260869565e-06,
      "loss": 0.6451,
      "step": 101
    },
    {
      "epoch": 1.6824742268041237,
      "grad_norm": 3.1875,
      "learning_rate": 6e-06,
      "loss": 0.3286,
      "step": 102
    },
    {
      "epoch": 1.6989690721649484,
      "grad_norm": 3.4375,
      "learning_rate": 5.956521739130435e-06,
      "loss": 0.9278,
      "step": 103
    },
    {
      "epoch": 1.7154639175257733,
      "grad_norm": 3.734375,
      "learning_rate": 5.91304347826087e-06,
      "loss": 0.4669,
      "step": 104
    },
    {
      "epoch": 1.731958762886598,
      "grad_norm": 3.421875,
      "learning_rate": 5.8695652173913055e-06,
      "loss": 0.8641,
      "step": 105
    },
    {
      "epoch": 1.731958762886598,
      "eval_loss": 0.52596116065979,
      "eval_runtime": 1.2826,
      "eval_samples_per_second": 20.271,
      "eval_steps_per_second": 20.271,
      "step": 105
    },
    {
      "epoch": 1.7484536082474227,
      "grad_norm": 4.625,
      "learning_rate": 5.826086956521739e-06,
      "loss": 0.4438,
      "step": 106
    },
    {
      "epoch": 1.7649484536082474,
      "grad_norm": 3.3125,
      "learning_rate": 5.782608695652174e-06,
      "loss": 0.2845,
      "step": 107
    },
    {
      "epoch": 1.781443298969072,
      "grad_norm": 3.015625,
      "learning_rate": 5.739130434782609e-06,
      "loss": 0.2432,
      "step": 108
    },
    {
      "epoch": 1.797938144329897,
      "grad_norm": 4.3125,
      "learning_rate": 5.695652173913044e-06,
      "loss": 1.0055,
      "step": 109
    },
    {
      "epoch": 1.8144329896907216,
      "grad_norm": 4.15625,
      "learning_rate": 5.652173913043479e-06,
      "loss": 0.5496,
      "step": 110
    },
    {
      "epoch": 1.8144329896907216,
      "eval_loss": 0.5252729058265686,
      "eval_runtime": 1.2934,
      "eval_samples_per_second": 20.103,
      "eval_steps_per_second": 20.103,
      "step": 110
    },
    {
      "epoch": 1.8309278350515465,
      "grad_norm": 2.84375,
      "learning_rate": 5.608695652173914e-06,
      "loss": 0.2763,
      "step": 111
    },
    {
      "epoch": 1.8474226804123712,
      "grad_norm": 2.8125,
      "learning_rate": 5.565217391304348e-06,
      "loss": 0.6699,
      "step": 112
    },
    {
      "epoch": 1.863917525773196,
      "grad_norm": 3.515625,
      "learning_rate": 5.521739130434783e-06,
      "loss": 0.3332,
      "step": 113
    },
    {
      "epoch": 1.8804123711340206,
      "grad_norm": 3.375,
      "learning_rate": 5.478260869565217e-06,
      "loss": 0.5998,
      "step": 114
    },
    {
      "epoch": 1.8969072164948453,
      "grad_norm": 3.078125,
      "learning_rate": 5.4347826086956525e-06,
      "loss": 0.2939,
      "step": 115
    },
    {
      "epoch": 1.8969072164948453,
      "eval_loss": 0.5248008370399475,
      "eval_runtime": 1.2932,
      "eval_samples_per_second": 20.105,
      "eval_steps_per_second": 20.105,
      "step": 115
    },
    {
      "epoch": 1.91340206185567,
      "grad_norm": 4.03125,
      "learning_rate": 5.391304347826088e-06,
      "loss": 0.4714,
      "step": 116
    },
    {
      "epoch": 1.9298969072164949,
      "grad_norm": 3.609375,
      "learning_rate": 5.347826086956523e-06,
      "loss": 0.5091,
      "step": 117
    },
    {
      "epoch": 1.9463917525773196,
      "grad_norm": 4.3125,
      "learning_rate": 5.304347826086957e-06,
      "loss": 0.3588,
      "step": 118
    },
    {
      "epoch": 1.9628865979381445,
      "grad_norm": 4.40625,
      "learning_rate": 5.260869565217391e-06,
      "loss": 0.5583,
      "step": 119
    },
    {
      "epoch": 1.9793814432989691,
      "grad_norm": 3.09375,
      "learning_rate": 5.2173913043478265e-06,
      "loss": 0.3538,
      "step": 120
    },
    {
      "epoch": 1.9793814432989691,
      "eval_loss": 0.5252537131309509,
      "eval_runtime": 1.2955,
      "eval_samples_per_second": 20.07,
      "eval_steps_per_second": 20.07,
      "step": 120
    },
    {
      "epoch": 1.9958762886597938,
      "grad_norm": 3.140625,
      "learning_rate": 5.173913043478262e-06,
      "loss": 0.4421,
      "step": 121
    },
    {
      "epoch": 2.0123711340206185,
      "grad_norm": 3.875,
      "learning_rate": 5.130434782608697e-06,
      "loss": 0.7688,
      "step": 122
    },
    {
      "epoch": 2.028865979381443,
      "grad_norm": 2.125,
      "learning_rate": 5.08695652173913e-06,
      "loss": 0.2562,
      "step": 123
    },
    {
      "epoch": 2.045360824742268,
      "grad_norm": 2.078125,
      "learning_rate": 5.043478260869565e-06,
      "loss": 0.3267,
      "step": 124
    },
    {
      "epoch": 2.0618556701030926,
      "grad_norm": 2.171875,
      "learning_rate": 5e-06,
      "loss": 0.3511,
      "step": 125
    },
    {
      "epoch": 2.0618556701030926,
      "eval_loss": 0.5247462391853333,
      "eval_runtime": 1.2904,
      "eval_samples_per_second": 20.149,
      "eval_steps_per_second": 20.149,
      "step": 125
    },
    {
      "epoch": 2.0783505154639177,
      "grad_norm": 2.0625,
      "learning_rate": 4.9565217391304355e-06,
      "loss": 0.3862,
      "step": 126
    },
    {
      "epoch": 2.0948453608247424,
      "grad_norm": 2.78125,
      "learning_rate": 4.91304347826087e-06,
      "loss": 0.5712,
      "step": 127
    },
    {
      "epoch": 2.111340206185567,
      "grad_norm": 3.203125,
      "learning_rate": 4.869565217391305e-06,
      "loss": 0.9281,
      "step": 128
    },
    {
      "epoch": 2.1278350515463917,
      "grad_norm": 1.703125,
      "learning_rate": 4.826086956521739e-06,
      "loss": 0.3723,
      "step": 129
    },
    {
      "epoch": 2.1443298969072164,
      "grad_norm": 2.21875,
      "learning_rate": 4.782608695652174e-06,
      "loss": 0.544,
      "step": 130
    },
    {
      "epoch": 2.1443298969072164,
      "eval_loss": 0.525088906288147,
      "eval_runtime": 1.2916,
      "eval_samples_per_second": 20.131,
      "eval_steps_per_second": 20.131,
      "step": 130
    },
    {
      "epoch": 2.160824742268041,
      "grad_norm": 2.703125,
      "learning_rate": 4.739130434782609e-06,
      "loss": 0.4491,
      "step": 131
    },
    {
      "epoch": 2.177319587628866,
      "grad_norm": 3.875,
      "learning_rate": 4.695652173913044e-06,
      "loss": 0.5799,
      "step": 132
    },
    {
      "epoch": 2.193814432989691,
      "grad_norm": 3.171875,
      "learning_rate": 4.652173913043478e-06,
      "loss": 0.6968,
      "step": 133
    },
    {
      "epoch": 2.2103092783505156,
      "grad_norm": 2.1875,
      "learning_rate": 4.608695652173913e-06,
      "loss": 0.5588,
      "step": 134
    },
    {
      "epoch": 2.2268041237113403,
      "grad_norm": 2.03125,
      "learning_rate": 4.565217391304348e-06,
      "loss": 0.2793,
      "step": 135
    },
    {
      "epoch": 2.2268041237113403,
      "eval_loss": 0.5259795784950256,
      "eval_runtime": 1.2926,
      "eval_samples_per_second": 20.115,
      "eval_steps_per_second": 20.115,
      "step": 135
    },
    {
      "epoch": 2.243298969072165,
      "grad_norm": 2.109375,
      "learning_rate": 4.5217391304347826e-06,
      "loss": 1.2616,
      "step": 136
    },
    {
      "epoch": 2.2597938144329897,
      "grad_norm": 2.015625,
      "learning_rate": 4.478260869565218e-06,
      "loss": 0.2996,
      "step": 137
    },
    {
      "epoch": 2.2762886597938143,
      "grad_norm": 2.640625,
      "learning_rate": 4.434782608695653e-06,
      "loss": 0.4981,
      "step": 138
    },
    {
      "epoch": 2.292783505154639,
      "grad_norm": 4.25,
      "learning_rate": 4.391304347826087e-06,
      "loss": 0.5537,
      "step": 139
    },
    {
      "epoch": 2.3092783505154637,
      "grad_norm": 2.90625,
      "learning_rate": 4.347826086956522e-06,
      "loss": 1.1338,
      "step": 140
    },
    {
      "epoch": 2.3092783505154637,
      "eval_loss": 0.5254911184310913,
      "eval_runtime": 1.2896,
      "eval_samples_per_second": 20.162,
      "eval_steps_per_second": 20.162,
      "step": 140
    },
    {
      "epoch": 2.325773195876289,
      "grad_norm": 2.046875,
      "learning_rate": 4.304347826086957e-06,
      "loss": 0.4052,
      "step": 141
    },
    {
      "epoch": 2.3422680412371135,
      "grad_norm": 2.046875,
      "learning_rate": 4.260869565217392e-06,
      "loss": 0.3079,
      "step": 142
    },
    {
      "epoch": 2.358762886597938,
      "grad_norm": 2.421875,
      "learning_rate": 4.217391304347827e-06,
      "loss": 0.3265,
      "step": 143
    },
    {
      "epoch": 2.375257731958763,
      "grad_norm": 2.65625,
      "learning_rate": 4.173913043478261e-06,
      "loss": 0.4978,
      "step": 144
    },
    {
      "epoch": 2.3917525773195876,
      "grad_norm": 2.890625,
      "learning_rate": 4.130434782608696e-06,
      "loss": 0.4199,
      "step": 145
    },
    {
      "epoch": 2.3917525773195876,
      "eval_loss": 0.5256755352020264,
      "eval_runtime": 1.2989,
      "eval_samples_per_second": 20.017,
      "eval_steps_per_second": 20.017,
      "step": 145
    },
    {
      "epoch": 2.4082474226804123,
      "grad_norm": 2.453125,
      "learning_rate": 4.086956521739131e-06,
      "loss": 0.3124,
      "step": 146
    },
    {
      "epoch": 2.424742268041237,
      "grad_norm": 2.828125,
      "learning_rate": 4.0434782608695655e-06,
      "loss": 0.5702,
      "step": 147
    },
    {
      "epoch": 2.441237113402062,
      "grad_norm": 1.953125,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.4731,
      "step": 148
    },
    {
      "epoch": 2.4577319587628867,
      "grad_norm": 2.921875,
      "learning_rate": 3.956521739130435e-06,
      "loss": 0.4377,
      "step": 149
    },
    {
      "epoch": 2.4742268041237114,
      "grad_norm": 2.421875,
      "learning_rate": 3.91304347826087e-06,
      "loss": 0.2762,
      "step": 150
    },
    {
      "epoch": 2.4742268041237114,
      "eval_loss": 0.525678813457489,
      "eval_runtime": 1.2937,
      "eval_samples_per_second": 20.097,
      "eval_steps_per_second": 20.097,
      "step": 150
    },
    {
      "epoch": 2.490721649484536,
      "grad_norm": 2.046875,
      "learning_rate": 3.869565217391304e-06,
      "loss": 0.4265,
      "step": 151
    },
    {
      "epoch": 2.507216494845361,
      "grad_norm": 6.75,
      "learning_rate": 3.8260869565217395e-06,
      "loss": 0.4723,
      "step": 152
    },
    {
      "epoch": 2.5237113402061855,
      "grad_norm": 2.53125,
      "learning_rate": 3.782608695652174e-06,
      "loss": 0.2729,
      "step": 153
    },
    {
      "epoch": 2.54020618556701,
      "grad_norm": 2.296875,
      "learning_rate": 3.739130434782609e-06,
      "loss": 0.3146,
      "step": 154
    },
    {
      "epoch": 2.556701030927835,
      "grad_norm": 2.90625,
      "learning_rate": 3.6956521739130436e-06,
      "loss": 0.4397,
      "step": 155
    },
    {
      "epoch": 2.556701030927835,
      "eval_loss": 0.5252581834793091,
      "eval_runtime": 1.2942,
      "eval_samples_per_second": 20.09,
      "eval_steps_per_second": 20.09,
      "step": 155
    },
    {
      "epoch": 2.5731958762886595,
      "grad_norm": 2.890625,
      "learning_rate": 3.6521739130434787e-06,
      "loss": 0.3224,
      "step": 156
    },
    {
      "epoch": 2.5896907216494847,
      "grad_norm": 2.5625,
      "learning_rate": 3.6086956521739134e-06,
      "loss": 0.7719,
      "step": 157
    },
    {
      "epoch": 2.6061855670103093,
      "grad_norm": 2.59375,
      "learning_rate": 3.565217391304348e-06,
      "loss": 0.4424,
      "step": 158
    },
    {
      "epoch": 2.622680412371134,
      "grad_norm": 2.421875,
      "learning_rate": 3.5217391304347832e-06,
      "loss": 0.3581,
      "step": 159
    },
    {
      "epoch": 2.6391752577319587,
      "grad_norm": 3.671875,
      "learning_rate": 3.4782608695652175e-06,
      "loss": 0.5688,
      "step": 160
    },
    {
      "epoch": 2.6391752577319587,
      "eval_loss": 0.5254755616188049,
      "eval_runtime": 1.2943,
      "eval_samples_per_second": 20.088,
      "eval_steps_per_second": 20.088,
      "step": 160
    },
    {
      "epoch": 2.6556701030927834,
      "grad_norm": 3.203125,
      "learning_rate": 3.4347826086956526e-06,
      "loss": 0.6329,
      "step": 161
    },
    {
      "epoch": 2.6721649484536085,
      "grad_norm": 2.84375,
      "learning_rate": 3.391304347826087e-06,
      "loss": 0.5551,
      "step": 162
    },
    {
      "epoch": 2.688659793814433,
      "grad_norm": 2.4375,
      "learning_rate": 3.347826086956522e-06,
      "loss": 0.2958,
      "step": 163
    },
    {
      "epoch": 2.705154639175258,
      "grad_norm": 4.4375,
      "learning_rate": 3.3043478260869567e-06,
      "loss": 1.1335,
      "step": 164
    },
    {
      "epoch": 2.7216494845360826,
      "grad_norm": 2.921875,
      "learning_rate": 3.2608695652173914e-06,
      "loss": 0.6526,
      "step": 165
    },
    {
      "epoch": 2.7216494845360826,
      "eval_loss": 0.526247501373291,
      "eval_runtime": 1.2946,
      "eval_samples_per_second": 20.083,
      "eval_steps_per_second": 20.083,
      "step": 165
    },
    {
      "epoch": 2.7381443298969073,
      "grad_norm": 2.484375,
      "learning_rate": 3.217391304347826e-06,
      "loss": 0.2683,
      "step": 166
    },
    {
      "epoch": 2.754639175257732,
      "grad_norm": 2.875,
      "learning_rate": 3.1739130434782613e-06,
      "loss": 0.427,
      "step": 167
    },
    {
      "epoch": 2.7711340206185566,
      "grad_norm": 2.046875,
      "learning_rate": 3.130434782608696e-06,
      "loss": 0.3583,
      "step": 168
    },
    {
      "epoch": 2.7876288659793813,
      "grad_norm": 3.078125,
      "learning_rate": 3.0869565217391307e-06,
      "loss": 1.0284,
      "step": 169
    },
    {
      "epoch": 2.804123711340206,
      "grad_norm": 1.71875,
      "learning_rate": 3.043478260869566e-06,
      "loss": 0.3519,
      "step": 170
    },
    {
      "epoch": 2.804123711340206,
      "eval_loss": 0.5257400870323181,
      "eval_runtime": 1.2971,
      "eval_samples_per_second": 20.044,
      "eval_steps_per_second": 20.044,
      "step": 170
    },
    {
      "epoch": 2.8206185567010307,
      "grad_norm": 3.09375,
      "learning_rate": 3e-06,
      "loss": 0.4442,
      "step": 171
    },
    {
      "epoch": 2.837113402061856,
      "grad_norm": 2.296875,
      "learning_rate": 2.956521739130435e-06,
      "loss": 0.3421,
      "step": 172
    },
    {
      "epoch": 2.8536082474226805,
      "grad_norm": 3.1875,
      "learning_rate": 2.9130434782608695e-06,
      "loss": 0.5583,
      "step": 173
    },
    {
      "epoch": 2.870103092783505,
      "grad_norm": 3.109375,
      "learning_rate": 2.8695652173913046e-06,
      "loss": 0.44,
      "step": 174
    },
    {
      "epoch": 2.88659793814433,
      "grad_norm": 2.328125,
      "learning_rate": 2.8260869565217393e-06,
      "loss": 0.3957,
      "step": 175
    },
    {
      "epoch": 2.88659793814433,
      "eval_loss": 0.525533139705658,
      "eval_runtime": 1.2952,
      "eval_samples_per_second": 20.075,
      "eval_steps_per_second": 20.075,
      "step": 175
    },
    {
      "epoch": 2.9030927835051545,
      "grad_norm": 1.8359375,
      "learning_rate": 2.782608695652174e-06,
      "loss": 0.3214,
      "step": 176
    },
    {
      "epoch": 2.9195876288659792,
      "grad_norm": 2.234375,
      "learning_rate": 2.7391304347826087e-06,
      "loss": 0.681,
      "step": 177
    },
    {
      "epoch": 2.9360824742268044,
      "grad_norm": 1.8125,
      "learning_rate": 2.695652173913044e-06,
      "loss": 0.3143,
      "step": 178
    },
    {
      "epoch": 2.952577319587629,
      "grad_norm": 2.921875,
      "learning_rate": 2.6521739130434785e-06,
      "loss": 0.4294,
      "step": 179
    },
    {
      "epoch": 2.9690721649484537,
      "grad_norm": 3.25,
      "learning_rate": 2.6086956521739132e-06,
      "loss": 0.9113,
      "step": 180
    },
    {
      "epoch": 2.9690721649484537,
      "eval_loss": 0.5252847075462341,
      "eval_runtime": 1.2957,
      "eval_samples_per_second": 20.066,
      "eval_steps_per_second": 20.066,
      "step": 180
    },
    {
      "epoch": 2.9855670103092784,
      "grad_norm": 3.171875,
      "learning_rate": 2.5652173913043484e-06,
      "loss": 0.6406,
      "step": 181
    },
    {
      "epoch": 3.002061855670103,
      "grad_norm": 3.546875,
      "learning_rate": 2.5217391304347826e-06,
      "loss": 0.8178,
      "step": 182
    },
    {
      "epoch": 3.0185567010309278,
      "grad_norm": 2.21875,
      "learning_rate": 2.4782608695652178e-06,
      "loss": 0.325,
      "step": 183
    },
    {
      "epoch": 3.0350515463917525,
      "grad_norm": 2.796875,
      "learning_rate": 2.4347826086956525e-06,
      "loss": 0.5324,
      "step": 184
    },
    {
      "epoch": 3.051546391752577,
      "grad_norm": 2.515625,
      "learning_rate": 2.391304347826087e-06,
      "loss": 0.442,
      "step": 185
    },
    {
      "epoch": 3.051546391752577,
      "eval_loss": 0.5247930288314819,
      "eval_runtime": 1.2956,
      "eval_samples_per_second": 20.069,
      "eval_steps_per_second": 20.069,
      "step": 185
    },
    {
      "epoch": 3.0680412371134023,
      "grad_norm": 2.03125,
      "learning_rate": 2.347826086956522e-06,
      "loss": 0.5274,
      "step": 186
    },
    {
      "epoch": 3.084536082474227,
      "grad_norm": 2.4375,
      "learning_rate": 2.3043478260869566e-06,
      "loss": 1.2468,
      "step": 187
    },
    {
      "epoch": 3.1010309278350516,
      "grad_norm": 2.484375,
      "learning_rate": 2.2608695652173913e-06,
      "loss": 0.4707,
      "step": 188
    },
    {
      "epoch": 3.1175257731958763,
      "grad_norm": 3.265625,
      "learning_rate": 2.2173913043478264e-06,
      "loss": 1.0451,
      "step": 189
    },
    {
      "epoch": 3.134020618556701,
      "grad_norm": 2.96875,
      "learning_rate": 2.173913043478261e-06,
      "loss": 0.8531,
      "step": 190
    },
    {
      "epoch": 3.134020618556701,
      "eval_loss": 0.5254663825035095,
      "eval_runtime": 1.2902,
      "eval_samples_per_second": 20.152,
      "eval_steps_per_second": 20.152,
      "step": 190
    },
    {
      "epoch": 3.1505154639175257,
      "grad_norm": 1.7421875,
      "learning_rate": 2.130434782608696e-06,
      "loss": 0.3128,
      "step": 191
    },
    {
      "epoch": 3.1670103092783504,
      "grad_norm": 1.953125,
      "learning_rate": 2.0869565217391305e-06,
      "loss": 0.5937,
      "step": 192
    },
    {
      "epoch": 3.183505154639175,
      "grad_norm": 2.4375,
      "learning_rate": 2.0434782608695656e-06,
      "loss": 0.5541,
      "step": 193
    },
    {
      "epoch": 3.2,
      "grad_norm": 2.421875,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 0.6821,
      "step": 194
    },
    {
      "epoch": 3.216494845360825,
      "grad_norm": 2.125,
      "learning_rate": 1.956521739130435e-06,
      "loss": 0.6514,
      "step": 195
    },
    {
      "epoch": 3.216494845360825,
      "eval_loss": 0.5250334739685059,
      "eval_runtime": 1.2956,
      "eval_samples_per_second": 20.068,
      "eval_steps_per_second": 20.068,
      "step": 195
    },
    {
      "epoch": 3.2329896907216495,
      "grad_norm": 1.4140625,
      "learning_rate": 1.9130434782608697e-06,
      "loss": 0.2944,
      "step": 196
    },
    {
      "epoch": 3.2494845360824742,
      "grad_norm": 2.65625,
      "learning_rate": 1.8695652173913044e-06,
      "loss": 0.4221,
      "step": 197
    },
    {
      "epoch": 3.265979381443299,
      "grad_norm": 2.75,
      "learning_rate": 1.8260869565217394e-06,
      "loss": 0.4516,
      "step": 198
    },
    {
      "epoch": 3.2824742268041236,
      "grad_norm": 2.15625,
      "learning_rate": 1.782608695652174e-06,
      "loss": 0.2816,
      "step": 199
    },
    {
      "epoch": 3.2989690721649483,
      "grad_norm": 1.765625,
      "learning_rate": 1.7391304347826088e-06,
      "loss": 0.2407,
      "step": 200
    },
    {
      "epoch": 3.2989690721649483,
      "eval_loss": 0.5253295302391052,
      "eval_runtime": 1.2931,
      "eval_samples_per_second": 20.107,
      "eval_steps_per_second": 20.107,
      "step": 200
    },
    {
      "epoch": 3.315463917525773,
      "grad_norm": 1.859375,
      "learning_rate": 1.6956521739130435e-06,
      "loss": 0.3107,
      "step": 201
    },
    {
      "epoch": 3.331958762886598,
      "grad_norm": 3.625,
      "learning_rate": 1.6521739130434784e-06,
      "loss": 1.2809,
      "step": 202
    },
    {
      "epoch": 3.3484536082474228,
      "grad_norm": 2.328125,
      "learning_rate": 1.608695652173913e-06,
      "loss": 0.3474,
      "step": 203
    },
    {
      "epoch": 3.3649484536082475,
      "grad_norm": 1.6015625,
      "learning_rate": 1.565217391304348e-06,
      "loss": 0.217,
      "step": 204
    },
    {
      "epoch": 3.381443298969072,
      "grad_norm": 2.46875,
      "learning_rate": 1.521739130434783e-06,
      "loss": 0.5715,
      "step": 205
    },
    {
      "epoch": 3.381443298969072,
      "eval_loss": 0.5251511931419373,
      "eval_runtime": 1.2966,
      "eval_samples_per_second": 20.053,
      "eval_steps_per_second": 20.053,
      "step": 205
    },
    {
      "epoch": 3.397938144329897,
      "grad_norm": 3.09375,
      "learning_rate": 1.4782608695652176e-06,
      "loss": 0.3722,
      "step": 206
    },
    {
      "epoch": 3.4144329896907215,
      "grad_norm": 1.78125,
      "learning_rate": 1.4347826086956523e-06,
      "loss": 0.3074,
      "step": 207
    },
    {
      "epoch": 3.4309278350515466,
      "grad_norm": 2.125,
      "learning_rate": 1.391304347826087e-06,
      "loss": 0.3416,
      "step": 208
    },
    {
      "epoch": 3.4474226804123713,
      "grad_norm": 3.0625,
      "learning_rate": 1.347826086956522e-06,
      "loss": 0.7473,
      "step": 209
    },
    {
      "epoch": 3.463917525773196,
      "grad_norm": 2.4375,
      "learning_rate": 1.3043478260869566e-06,
      "loss": 0.3372,
      "step": 210
    },
    {
      "epoch": 3.463917525773196,
      "eval_loss": 0.5255177617073059,
      "eval_runtime": 1.2948,
      "eval_samples_per_second": 20.08,
      "eval_steps_per_second": 20.08,
      "step": 210
    },
    {
      "epoch": 3.4804123711340207,
      "grad_norm": 1.6171875,
      "learning_rate": 1.2608695652173913e-06,
      "loss": 0.2201,
      "step": 211
    },
    {
      "epoch": 3.4969072164948454,
      "grad_norm": 2.0625,
      "learning_rate": 1.2173913043478262e-06,
      "loss": 0.4559,
      "step": 212
    },
    {
      "epoch": 3.51340206185567,
      "grad_norm": 1.625,
      "learning_rate": 1.173913043478261e-06,
      "loss": 0.278,
      "step": 213
    },
    {
      "epoch": 3.5298969072164947,
      "grad_norm": 2.484375,
      "learning_rate": 1.1304347826086956e-06,
      "loss": 0.4145,
      "step": 214
    },
    {
      "epoch": 3.5463917525773194,
      "grad_norm": 1.9140625,
      "learning_rate": 1.0869565217391306e-06,
      "loss": 0.3477,
      "step": 215
    },
    {
      "epoch": 3.5463917525773194,
      "eval_loss": 0.5260653495788574,
      "eval_runtime": 1.297,
      "eval_samples_per_second": 20.046,
      "eval_steps_per_second": 20.046,
      "step": 215
    },
    {
      "epoch": 3.562886597938144,
      "grad_norm": 2.0625,
      "learning_rate": 1.0434782608695653e-06,
      "loss": 0.2958,
      "step": 216
    },
    {
      "epoch": 3.5793814432989692,
      "grad_norm": 1.8828125,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 0.2438,
      "step": 217
    },
    {
      "epoch": 3.595876288659794,
      "grad_norm": 2.859375,
      "learning_rate": 9.565217391304349e-07,
      "loss": 0.526,
      "step": 218
    },
    {
      "epoch": 3.6123711340206186,
      "grad_norm": 2.171875,
      "learning_rate": 9.130434782608697e-07,
      "loss": 0.345,
      "step": 219
    },
    {
      "epoch": 3.6288659793814433,
      "grad_norm": 3.03125,
      "learning_rate": 8.695652173913044e-07,
      "loss": 0.5634,
      "step": 220
    },
    {
      "epoch": 3.6288659793814433,
      "eval_loss": 0.5257369875907898,
      "eval_runtime": 1.2957,
      "eval_samples_per_second": 20.066,
      "eval_steps_per_second": 20.066,
      "step": 220
    },
    {
      "epoch": 3.645360824742268,
      "grad_norm": 3.296875,
      "learning_rate": 8.260869565217392e-07,
      "loss": 0.598,
      "step": 221
    },
    {
      "epoch": 3.6618556701030927,
      "grad_norm": 2.15625,
      "learning_rate": 7.82608695652174e-07,
      "loss": 0.3963,
      "step": 222
    },
    {
      "epoch": 3.678350515463918,
      "grad_norm": 2.453125,
      "learning_rate": 7.391304347826088e-07,
      "loss": 0.4503,
      "step": 223
    },
    {
      "epoch": 3.6948453608247425,
      "grad_norm": 2.4375,
      "learning_rate": 6.956521739130435e-07,
      "loss": 0.3006,
      "step": 224
    },
    {
      "epoch": 3.711340206185567,
      "grad_norm": 2.546875,
      "learning_rate": 6.521739130434783e-07,
      "loss": 0.7533,
      "step": 225
    },
    {
      "epoch": 3.711340206185567,
      "eval_loss": 0.5261620879173279,
      "eval_runtime": 1.2946,
      "eval_samples_per_second": 20.084,
      "eval_steps_per_second": 20.084,
      "step": 225
    },
    {
      "epoch": 3.727835051546392,
      "grad_norm": 2.640625,
      "learning_rate": 6.086956521739131e-07,
      "loss": 0.4221,
      "step": 226
    },
    {
      "epoch": 3.7443298969072165,
      "grad_norm": 3.15625,
      "learning_rate": 5.652173913043478e-07,
      "loss": 0.7248,
      "step": 227
    },
    {
      "epoch": 3.760824742268041,
      "grad_norm": 2.65625,
      "learning_rate": 5.217391304347826e-07,
      "loss": 0.4017,
      "step": 228
    },
    {
      "epoch": 3.777319587628866,
      "grad_norm": 2.71875,
      "learning_rate": 4.782608695652174e-07,
      "loss": 0.366,
      "step": 229
    },
    {
      "epoch": 3.7938144329896906,
      "grad_norm": 2.171875,
      "learning_rate": 4.347826086956522e-07,
      "loss": 0.3177,
      "step": 230
    },
    {
      "epoch": 3.7938144329896906,
      "eval_loss": 0.5259535908699036,
      "eval_runtime": 1.3026,
      "eval_samples_per_second": 19.96,
      "eval_steps_per_second": 19.96,
      "step": 230
    },
    {
      "epoch": 3.8103092783505152,
      "grad_norm": 2.546875,
      "learning_rate": 3.91304347826087e-07,
      "loss": 0.5557,
      "step": 231
    },
    {
      "epoch": 3.82680412371134,
      "grad_norm": 1.6875,
      "learning_rate": 3.4782608695652175e-07,
      "loss": 0.8832,
      "step": 232
    },
    {
      "epoch": 3.843298969072165,
      "grad_norm": 1.7734375,
      "learning_rate": 3.0434782608695656e-07,
      "loss": 0.4498,
      "step": 233
    },
    {
      "epoch": 3.8597938144329897,
      "grad_norm": 2.875,
      "learning_rate": 2.608695652173913e-07,
      "loss": 0.348,
      "step": 234
    },
    {
      "epoch": 3.8762886597938144,
      "grad_norm": 2.5625,
      "learning_rate": 2.173913043478261e-07,
      "loss": 0.5632,
      "step": 235
    },
    {
      "epoch": 3.8762886597938144,
      "eval_loss": 0.5255343914031982,
      "eval_runtime": 1.2946,
      "eval_samples_per_second": 20.084,
      "eval_steps_per_second": 20.084,
      "step": 235
    },
    {
      "epoch": 3.892783505154639,
      "grad_norm": 2.28125,
      "learning_rate": 1.7391304347826088e-07,
      "loss": 0.8766,
      "step": 236
    },
    {
      "epoch": 3.909278350515464,
      "grad_norm": 2.5,
      "learning_rate": 1.3043478260869566e-07,
      "loss": 0.4997,
      "step": 237
    },
    {
      "epoch": 3.925773195876289,
      "grad_norm": 2.390625,
      "learning_rate": 8.695652173913044e-08,
      "loss": 0.7791,
      "step": 238
    },
    {
      "epoch": 3.9422680412371136,
      "grad_norm": 2.890625,
      "learning_rate": 4.347826086956522e-08,
      "loss": 0.7023,
      "step": 239
    },
    {
      "epoch": 3.9587628865979383,
      "grad_norm": 3.0625,
      "learning_rate": 0.0,
      "loss": 0.6532,
      "step": 240
    },
    {
      "epoch": 3.9587628865979383,
      "eval_loss": 0.5261483192443848,
      "eval_runtime": 1.2982,
      "eval_samples_per_second": 20.028,
      "eval_steps_per_second": 20.028,
      "step": 240
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 240,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "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": 2.175841849337856e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}