{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 100000.0,
  "global_step": 96,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.03125,
      "grad_norm": 189.0,
      "learning_rate": 0.0,
      "loss": 5.3761,
      "mean_token_accuracy": 0.4222588539123535,
      "step": 1
    },
    {
      "epoch": 0.0625,
      "grad_norm": 187.0,
      "learning_rate": 9.999999999999999e-06,
      "loss": 5.4398,
      "mean_token_accuracy": 0.42141932249069214,
      "step": 2
    },
    {
      "epoch": 0.09375,
      "grad_norm": 29.375,
      "learning_rate": 1.9999999999999998e-05,
      "loss": 3.0716,
      "mean_token_accuracy": 0.5135114192962646,
      "step": 3
    },
    {
      "epoch": 0.125,
      "grad_norm": 17.625,
      "learning_rate": 3e-05,
      "loss": 2.1269,
      "mean_token_accuracy": 0.646616518497467,
      "step": 4
    },
    {
      "epoch": 0.15625,
      "grad_norm": 40.0,
      "learning_rate": 2.967741935483871e-05,
      "loss": 1.5739,
      "mean_token_accuracy": 0.7027881145477295,
      "step": 5
    },
    {
      "epoch": 0.1875,
      "grad_norm": 7.5625,
      "learning_rate": 2.935483870967742e-05,
      "loss": 1.2517,
      "mean_token_accuracy": 0.7468280792236328,
      "step": 6
    },
    {
      "epoch": 0.21875,
      "grad_norm": 8.125,
      "learning_rate": 2.903225806451613e-05,
      "loss": 1.1719,
      "mean_token_accuracy": 0.7558277249336243,
      "step": 7
    },
    {
      "epoch": 0.25,
      "grad_norm": 6.84375,
      "learning_rate": 2.870967741935484e-05,
      "loss": 1.0179,
      "mean_token_accuracy": 0.7789369821548462,
      "step": 8
    },
    {
      "epoch": 0.28125,
      "grad_norm": 4.5625,
      "learning_rate": 2.8387096774193552e-05,
      "loss": 0.9696,
      "mean_token_accuracy": 0.7895718812942505,
      "step": 9
    },
    {
      "epoch": 0.3125,
      "grad_norm": 3.71875,
      "learning_rate": 2.806451612903226e-05,
      "loss": 0.9511,
      "mean_token_accuracy": 0.7805737853050232,
      "step": 10
    },
    {
      "epoch": 0.34375,
      "grad_norm": 4.40625,
      "learning_rate": 2.7741935483870968e-05,
      "loss": 0.9864,
      "mean_token_accuracy": 0.7709536552429199,
      "step": 11
    },
    {
      "epoch": 0.375,
      "grad_norm": 3.75,
      "learning_rate": 2.7419354838709678e-05,
      "loss": 0.8482,
      "mean_token_accuracy": 0.7989001274108887,
      "step": 12
    },
    {
      "epoch": 0.40625,
      "grad_norm": 10.1875,
      "learning_rate": 2.7096774193548387e-05,
      "loss": 0.8758,
      "mean_token_accuracy": 0.7906646728515625,
      "step": 13
    },
    {
      "epoch": 0.4375,
      "grad_norm": 4.375,
      "learning_rate": 2.6774193548387097e-05,
      "loss": 0.8406,
      "mean_token_accuracy": 0.7955039739608765,
      "step": 14
    },
    {
      "epoch": 0.46875,
      "grad_norm": 3.34375,
      "learning_rate": 2.6451612903225806e-05,
      "loss": 0.8676,
      "mean_token_accuracy": 0.7854324579238892,
      "step": 15
    },
    {
      "epoch": 0.5,
      "grad_norm": 2.9375,
      "learning_rate": 2.6129032258064516e-05,
      "loss": 0.8414,
      "mean_token_accuracy": 0.7912613749504089,
      "step": 16
    },
    {
      "epoch": 0.53125,
      "grad_norm": 2.671875,
      "learning_rate": 2.5806451612903226e-05,
      "loss": 0.8772,
      "mean_token_accuracy": 0.7842865586280823,
      "step": 17
    },
    {
      "epoch": 0.5625,
      "grad_norm": 3.078125,
      "learning_rate": 2.548387096774194e-05,
      "loss": 0.8225,
      "mean_token_accuracy": 0.7936803102493286,
      "step": 18
    },
    {
      "epoch": 0.59375,
      "grad_norm": 2.9375,
      "learning_rate": 2.5161290322580648e-05,
      "loss": 0.7779,
      "mean_token_accuracy": 0.8040180802345276,
      "step": 19
    },
    {
      "epoch": 0.625,
      "grad_norm": 2.6875,
      "learning_rate": 2.4838709677419358e-05,
      "loss": 0.7973,
      "mean_token_accuracy": 0.7983251810073853,
      "step": 20
    },
    {
      "epoch": 0.65625,
      "grad_norm": 2.734375,
      "learning_rate": 2.4516129032258067e-05,
      "loss": 0.7968,
      "mean_token_accuracy": 0.7994269132614136,
      "step": 21
    },
    {
      "epoch": 0.6875,
      "grad_norm": 2.703125,
      "learning_rate": 2.4193548387096773e-05,
      "loss": 0.7902,
      "mean_token_accuracy": 0.799203634262085,
      "step": 22
    },
    {
      "epoch": 0.71875,
      "grad_norm": 2.796875,
      "learning_rate": 2.3870967741935483e-05,
      "loss": 0.7776,
      "mean_token_accuracy": 0.8002063035964966,
      "step": 23
    },
    {
      "epoch": 0.75,
      "grad_norm": 2.578125,
      "learning_rate": 2.3548387096774193e-05,
      "loss": 0.7243,
      "mean_token_accuracy": 0.8160389065742493,
      "step": 24
    },
    {
      "epoch": 0.78125,
      "grad_norm": 2.453125,
      "learning_rate": 2.3225806451612902e-05,
      "loss": 0.7039,
      "mean_token_accuracy": 0.817916750907898,
      "step": 25
    },
    {
      "epoch": 0.8125,
      "grad_norm": 2.4375,
      "learning_rate": 2.2903225806451612e-05,
      "loss": 0.7417,
      "mean_token_accuracy": 0.8118934035301208,
      "step": 26
    },
    {
      "epoch": 0.84375,
      "grad_norm": 2.65625,
      "learning_rate": 2.258064516129032e-05,
      "loss": 0.7577,
      "mean_token_accuracy": 0.8054953217506409,
      "step": 27
    },
    {
      "epoch": 0.875,
      "grad_norm": 2.546875,
      "learning_rate": 2.2258064516129034e-05,
      "loss": 0.7387,
      "mean_token_accuracy": 0.8119359612464905,
      "step": 28
    },
    {
      "epoch": 0.90625,
      "grad_norm": 2.421875,
      "learning_rate": 2.1935483870967744e-05,
      "loss": 0.7541,
      "mean_token_accuracy": 0.8092722296714783,
      "step": 29
    },
    {
      "epoch": 0.9375,
      "grad_norm": 2.4375,
      "learning_rate": 2.1612903225806454e-05,
      "loss": 0.768,
      "mean_token_accuracy": 0.8078325390815735,
      "step": 30
    },
    {
      "epoch": 0.96875,
      "grad_norm": 2.3125,
      "learning_rate": 2.1290322580645163e-05,
      "loss": 0.6699,
      "mean_token_accuracy": 0.8237451910972595,
      "step": 31
    },
    {
      "epoch": 1.0,
      "grad_norm": 2.5,
      "learning_rate": 2.0967741935483873e-05,
      "loss": 0.721,
      "mean_token_accuracy": 0.8169918656349182,
      "step": 32
    },
    {
      "epoch": 1.03125,
      "grad_norm": 2.140625,
      "learning_rate": 2.0645161290322582e-05,
      "loss": 0.564,
      "mean_token_accuracy": 0.848415732383728,
      "step": 33
    },
    {
      "epoch": 1.0625,
      "grad_norm": 2.171875,
      "learning_rate": 2.032258064516129e-05,
      "loss": 0.5677,
      "mean_token_accuracy": 0.8471966981887817,
      "step": 34
    },
    {
      "epoch": 1.09375,
      "grad_norm": 2.09375,
      "learning_rate": 1.9999999999999998e-05,
      "loss": 0.5395,
      "mean_token_accuracy": 0.8564569354057312,
      "step": 35
    },
    {
      "epoch": 1.125,
      "grad_norm": 2.125,
      "learning_rate": 1.9677419354838708e-05,
      "loss": 0.5333,
      "mean_token_accuracy": 0.8557505011558533,
      "step": 36
    },
    {
      "epoch": 1.15625,
      "grad_norm": 2.1875,
      "learning_rate": 1.935483870967742e-05,
      "loss": 0.5176,
      "mean_token_accuracy": 0.859412670135498,
      "step": 37
    },
    {
      "epoch": 1.1875,
      "grad_norm": 2.4375,
      "learning_rate": 1.903225806451613e-05,
      "loss": 0.5255,
      "mean_token_accuracy": 0.8575270771980286,
      "step": 38
    },
    {
      "epoch": 1.21875,
      "grad_norm": 2.5,
      "learning_rate": 1.870967741935484e-05,
      "loss": 0.533,
      "mean_token_accuracy": 0.8552377820014954,
      "step": 39
    },
    {
      "epoch": 1.25,
      "grad_norm": 2.6875,
      "learning_rate": 1.838709677419355e-05,
      "loss": 0.5193,
      "mean_token_accuracy": 0.8584305047988892,
      "step": 40
    },
    {
      "epoch": 1.28125,
      "grad_norm": 2.46875,
      "learning_rate": 1.806451612903226e-05,
      "loss": 0.484,
      "mean_token_accuracy": 0.8667942881584167,
      "step": 41
    },
    {
      "epoch": 1.3125,
      "grad_norm": 2.671875,
      "learning_rate": 1.774193548387097e-05,
      "loss": 0.536,
      "mean_token_accuracy": 0.8559280633926392,
      "step": 42
    },
    {
      "epoch": 1.34375,
      "grad_norm": 2.765625,
      "learning_rate": 1.741935483870968e-05,
      "loss": 0.5303,
      "mean_token_accuracy": 0.8593308329582214,
      "step": 43
    },
    {
      "epoch": 1.375,
      "grad_norm": 2.53125,
      "learning_rate": 1.7096774193548388e-05,
      "loss": 0.506,
      "mean_token_accuracy": 0.8613560199737549,
      "step": 44
    },
    {
      "epoch": 1.40625,
      "grad_norm": 2.609375,
      "learning_rate": 1.6774193548387094e-05,
      "loss": 0.51,
      "mean_token_accuracy": 0.8609563708305359,
      "step": 45
    },
    {
      "epoch": 1.4375,
      "grad_norm": 2.5625,
      "learning_rate": 1.6451612903225807e-05,
      "loss": 0.5152,
      "mean_token_accuracy": 0.8623529672622681,
      "step": 46
    },
    {
      "epoch": 1.46875,
      "grad_norm": 2.375,
      "learning_rate": 1.6129032258064517e-05,
      "loss": 0.4736,
      "mean_token_accuracy": 0.8748390078544617,
      "step": 47
    },
    {
      "epoch": 1.5,
      "grad_norm": 2.484375,
      "learning_rate": 1.5806451612903226e-05,
      "loss": 0.525,
      "mean_token_accuracy": 0.8599056005477905,
      "step": 48
    },
    {
      "epoch": 1.53125,
      "grad_norm": 2.546875,
      "learning_rate": 1.5483870967741936e-05,
      "loss": 0.5388,
      "mean_token_accuracy": 0.8566693067550659,
      "step": 49
    },
    {
      "epoch": 1.5625,
      "grad_norm": 2.53125,
      "learning_rate": 1.5161290322580646e-05,
      "loss": 0.5067,
      "mean_token_accuracy": 0.8631064295768738,
      "step": 50
    },
    {
      "epoch": 1.59375,
      "grad_norm": 2.453125,
      "learning_rate": 1.4838709677419355e-05,
      "loss": 0.5125,
      "mean_token_accuracy": 0.8618575930595398,
      "step": 51
    },
    {
      "epoch": 1.625,
      "grad_norm": 2.46875,
      "learning_rate": 1.4516129032258065e-05,
      "loss": 0.4827,
      "mean_token_accuracy": 0.8688983917236328,
      "step": 52
    },
    {
      "epoch": 1.65625,
      "grad_norm": 2.375,
      "learning_rate": 1.4193548387096776e-05,
      "loss": 0.4968,
      "mean_token_accuracy": 0.8661786317825317,
      "step": 53
    },
    {
      "epoch": 1.6875,
      "grad_norm": 2.390625,
      "learning_rate": 1.3870967741935484e-05,
      "loss": 0.5088,
      "mean_token_accuracy": 0.8644407391548157,
      "step": 54
    },
    {
      "epoch": 1.71875,
      "grad_norm": 2.3125,
      "learning_rate": 1.3548387096774194e-05,
      "loss": 0.4772,
      "mean_token_accuracy": 0.8696879148483276,
      "step": 55
    },
    {
      "epoch": 1.75,
      "grad_norm": 2.4375,
      "learning_rate": 1.3225806451612903e-05,
      "loss": 0.5017,
      "mean_token_accuracy": 0.8670803308486938,
      "step": 56
    },
    {
      "epoch": 1.78125,
      "grad_norm": 2.328125,
      "learning_rate": 1.2903225806451613e-05,
      "loss": 0.5076,
      "mean_token_accuracy": 0.8628109693527222,
      "step": 57
    },
    {
      "epoch": 1.8125,
      "grad_norm": 2.28125,
      "learning_rate": 1.2580645161290324e-05,
      "loss": 0.482,
      "mean_token_accuracy": 0.8687414526939392,
      "step": 58
    },
    {
      "epoch": 1.84375,
      "grad_norm": 2.3125,
      "learning_rate": 1.2258064516129034e-05,
      "loss": 0.4701,
      "mean_token_accuracy": 0.8710319399833679,
      "step": 59
    },
    {
      "epoch": 1.875,
      "grad_norm": 2.28125,
      "learning_rate": 1.1935483870967742e-05,
      "loss": 0.4713,
      "mean_token_accuracy": 0.8721932768821716,
      "step": 60
    },
    {
      "epoch": 1.90625,
      "grad_norm": 2.25,
      "learning_rate": 1.1612903225806451e-05,
      "loss": 0.4893,
      "mean_token_accuracy": 0.8665592074394226,
      "step": 61
    },
    {
      "epoch": 1.9375,
      "grad_norm": 2.34375,
      "learning_rate": 1.129032258064516e-05,
      "loss": 0.4741,
      "mean_token_accuracy": 0.8702882528305054,
      "step": 62
    },
    {
      "epoch": 1.96875,
      "grad_norm": 2.234375,
      "learning_rate": 1.0967741935483872e-05,
      "loss": 0.4385,
      "mean_token_accuracy": 0.8789963722229004,
      "step": 63
    },
    {
      "epoch": 2.0,
      "grad_norm": 2.109375,
      "learning_rate": 1.0645161290322582e-05,
      "loss": 0.4023,
      "mean_token_accuracy": 0.8905370235443115,
      "step": 64
    },
    {
      "epoch": 2.03125,
      "grad_norm": 2.125,
      "learning_rate": 1.0322580645161291e-05,
      "loss": 0.3732,
      "mean_token_accuracy": 0.9010424613952637,
      "step": 65
    },
    {
      "epoch": 2.0625,
      "grad_norm": 2.015625,
      "learning_rate": 9.999999999999999e-06,
      "loss": 0.3639,
      "mean_token_accuracy": 0.903084397315979,
      "step": 66
    },
    {
      "epoch": 2.09375,
      "grad_norm": 2.15625,
      "learning_rate": 9.67741935483871e-06,
      "loss": 0.338,
      "mean_token_accuracy": 0.9060129523277283,
      "step": 67
    },
    {
      "epoch": 2.125,
      "grad_norm": 2.09375,
      "learning_rate": 9.35483870967742e-06,
      "loss": 0.368,
      "mean_token_accuracy": 0.902858555316925,
      "step": 68
    },
    {
      "epoch": 2.15625,
      "grad_norm": 2.171875,
      "learning_rate": 9.03225806451613e-06,
      "loss": 0.3557,
      "mean_token_accuracy": 0.9019851088523865,
      "step": 69
    },
    {
      "epoch": 2.1875,
      "grad_norm": 2.1875,
      "learning_rate": 8.70967741935484e-06,
      "loss": 0.3477,
      "mean_token_accuracy": 0.9044142961502075,
      "step": 70
    },
    {
      "epoch": 2.21875,
      "grad_norm": 2.34375,
      "learning_rate": 8.387096774193547e-06,
      "loss": 0.3691,
      "mean_token_accuracy": 0.8983471989631653,
      "step": 71
    },
    {
      "epoch": 2.25,
      "grad_norm": 2.28125,
      "learning_rate": 8.064516129032258e-06,
      "loss": 0.3374,
      "mean_token_accuracy": 0.9074879884719849,
      "step": 72
    },
    {
      "epoch": 2.28125,
      "grad_norm": 2.34375,
      "learning_rate": 7.741935483870968e-06,
      "loss": 0.3373,
      "mean_token_accuracy": 0.9076383709907532,
      "step": 73
    },
    {
      "epoch": 2.3125,
      "grad_norm": 2.4375,
      "learning_rate": 7.419354838709678e-06,
      "loss": 0.3518,
      "mean_token_accuracy": 0.9008965492248535,
      "step": 74
    },
    {
      "epoch": 2.34375,
      "grad_norm": 2.3125,
      "learning_rate": 7.096774193548388e-06,
      "loss": 0.311,
      "mean_token_accuracy": 0.9127882719039917,
      "step": 75
    },
    {
      "epoch": 2.375,
      "grad_norm": 2.34375,
      "learning_rate": 6.774193548387097e-06,
      "loss": 0.3072,
      "mean_token_accuracy": 0.9146369099617004,
      "step": 76
    },
    {
      "epoch": 2.40625,
      "grad_norm": 2.578125,
      "learning_rate": 6.451612903225806e-06,
      "loss": 0.3627,
      "mean_token_accuracy": 0.9014822840690613,
      "step": 77
    },
    {
      "epoch": 2.4375,
      "grad_norm": 2.484375,
      "learning_rate": 6.129032258064517e-06,
      "loss": 0.3219,
      "mean_token_accuracy": 0.90999835729599,
      "step": 78
    },
    {
      "epoch": 2.46875,
      "grad_norm": 2.390625,
      "learning_rate": 5.8064516129032256e-06,
      "loss": 0.3146,
      "mean_token_accuracy": 0.9131765365600586,
      "step": 79
    },
    {
      "epoch": 2.5,
      "grad_norm": 2.625,
      "learning_rate": 5.483870967741936e-06,
      "loss": 0.3407,
      "mean_token_accuracy": 0.906125545501709,
      "step": 80
    },
    {
      "epoch": 2.53125,
      "grad_norm": 2.625,
      "learning_rate": 5.161290322580646e-06,
      "loss": 0.3365,
      "mean_token_accuracy": 0.9048438668251038,
      "step": 81
    },
    {
      "epoch": 2.5625,
      "grad_norm": 2.5625,
      "learning_rate": 4.838709677419355e-06,
      "loss": 0.336,
      "mean_token_accuracy": 0.9078318476676941,
      "step": 82
    },
    {
      "epoch": 2.59375,
      "grad_norm": 2.59375,
      "learning_rate": 4.516129032258065e-06,
      "loss": 0.3293,
      "mean_token_accuracy": 0.9083393812179565,
      "step": 83
    },
    {
      "epoch": 2.625,
      "grad_norm": 2.71875,
      "learning_rate": 4.1935483870967736e-06,
      "loss": 0.3619,
      "mean_token_accuracy": 0.9003615379333496,
      "step": 84
    },
    {
      "epoch": 2.65625,
      "grad_norm": 2.515625,
      "learning_rate": 3.870967741935484e-06,
      "loss": 0.348,
      "mean_token_accuracy": 0.9053718447685242,
      "step": 85
    },
    {
      "epoch": 2.6875,
      "grad_norm": 2.578125,
      "learning_rate": 3.548387096774194e-06,
      "loss": 0.3351,
      "mean_token_accuracy": 0.9073967933654785,
      "step": 86
    },
    {
      "epoch": 2.71875,
      "grad_norm": 2.578125,
      "learning_rate": 3.225806451612903e-06,
      "loss": 0.3425,
      "mean_token_accuracy": 0.9027067422866821,
      "step": 87
    },
    {
      "epoch": 2.75,
      "grad_norm": 2.484375,
      "learning_rate": 2.9032258064516128e-06,
      "loss": 0.3282,
      "mean_token_accuracy": 0.9083059430122375,
      "step": 88
    },
    {
      "epoch": 2.78125,
      "grad_norm": 2.53125,
      "learning_rate": 2.580645161290323e-06,
      "loss": 0.3159,
      "mean_token_accuracy": 0.9134910106658936,
      "step": 89
    },
    {
      "epoch": 2.8125,
      "grad_norm": 2.515625,
      "learning_rate": 2.2580645161290324e-06,
      "loss": 0.3426,
      "mean_token_accuracy": 0.9068294167518616,
      "step": 90
    },
    {
      "epoch": 2.84375,
      "grad_norm": 2.609375,
      "learning_rate": 1.935483870967742e-06,
      "loss": 0.3427,
      "mean_token_accuracy": 0.9076163172721863,
      "step": 91
    },
    {
      "epoch": 2.875,
      "grad_norm": 2.609375,
      "learning_rate": 1.6129032258064516e-06,
      "loss": 0.3513,
      "mean_token_accuracy": 0.9038658142089844,
      "step": 92
    },
    {
      "epoch": 2.90625,
      "grad_norm": 2.640625,
      "learning_rate": 1.2903225806451614e-06,
      "loss": 0.3378,
      "mean_token_accuracy": 0.9042639136314392,
      "step": 93
    },
    {
      "epoch": 2.9375,
      "grad_norm": 2.515625,
      "learning_rate": 9.67741935483871e-07,
      "loss": 0.3308,
      "mean_token_accuracy": 0.9092385172843933,
      "step": 94
    },
    {
      "epoch": 2.96875,
      "grad_norm": 2.59375,
      "learning_rate": 6.451612903225807e-07,
      "loss": 0.3369,
      "mean_token_accuracy": 0.9074664115905762,
      "step": 95
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.453125,
      "learning_rate": 3.2258064516129035e-07,
      "loss": 0.331,
      "mean_token_accuracy": 0.9075567722320557,
      "step": 96
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 96,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 100000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 3.63102607549399e+16,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}