{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 10.0,
  "eval_steps": 500,
  "global_step": 510,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0196078431372549,
      "grad_norm": 494.0,
      "learning_rate": 3.92156862745098e-06,
      "loss": 48.0011,
      "step": 1
    },
    {
      "epoch": 0.09803921568627451,
      "grad_norm": 394.0,
      "learning_rate": 1.9607843137254903e-05,
      "loss": 46.237,
      "step": 5
    },
    {
      "epoch": 0.19607843137254902,
      "grad_norm": 131.0,
      "learning_rate": 3.9215686274509805e-05,
      "loss": 36.1065,
      "step": 10
    },
    {
      "epoch": 0.29411764705882354,
      "grad_norm": 36.25,
      "learning_rate": 5.882352941176471e-05,
      "loss": 23.9218,
      "step": 15
    },
    {
      "epoch": 0.39215686274509803,
      "grad_norm": 11.6875,
      "learning_rate": 7.843137254901961e-05,
      "loss": 20.3348,
      "step": 20
    },
    {
      "epoch": 0.49019607843137253,
      "grad_norm": 8.125,
      "learning_rate": 9.80392156862745e-05,
      "loss": 19.5012,
      "step": 25
    },
    {
      "epoch": 0.5882352941176471,
      "grad_norm": 20.375,
      "learning_rate": 0.00011764705882352942,
      "loss": 17.6271,
      "step": 30
    },
    {
      "epoch": 0.6862745098039216,
      "grad_norm": 60.0,
      "learning_rate": 0.0001372549019607843,
      "loss": 13.6754,
      "step": 35
    },
    {
      "epoch": 0.7843137254901961,
      "grad_norm": 11.0625,
      "learning_rate": 0.00015686274509803922,
      "loss": 4.1094,
      "step": 40
    },
    {
      "epoch": 0.8823529411764706,
      "grad_norm": 5.0,
      "learning_rate": 0.00017647058823529413,
      "loss": 2.136,
      "step": 45
    },
    {
      "epoch": 0.9803921568627451,
      "grad_norm": 3.375,
      "learning_rate": 0.000196078431372549,
      "loss": 1.7584,
      "step": 50
    },
    {
      "epoch": 1.0,
      "eval_loss": 2.7735681533813477,
      "eval_runtime": 2.0507,
      "eval_samples_per_second": 4.876,
      "eval_steps_per_second": 0.975,
      "step": 51
    },
    {
      "epoch": 1.0784313725490196,
      "grad_norm": 2.6875,
      "learning_rate": 0.00019996252538028507,
      "loss": 1.5368,
      "step": 55
    },
    {
      "epoch": 1.1764705882352942,
      "grad_norm": 23.125,
      "learning_rate": 0.00019981033287370443,
      "loss": 1.4659,
      "step": 60
    },
    {
      "epoch": 1.2745098039215685,
      "grad_norm": 3.640625,
      "learning_rate": 0.00019954125840299165,
      "loss": 1.3602,
      "step": 65
    },
    {
      "epoch": 1.3725490196078431,
      "grad_norm": 6.40625,
      "learning_rate": 0.00019915561706530883,
      "loss": 1.3083,
      "step": 70
    },
    {
      "epoch": 1.4705882352941178,
      "grad_norm": 2.53125,
      "learning_rate": 0.00019865386046236596,
      "loss": 1.2742,
      "step": 75
    },
    {
      "epoch": 1.5686274509803921,
      "grad_norm": 3.21875,
      "learning_rate": 0.0001980365761715769,
      "loss": 1.2384,
      "step": 80
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 3.8125,
      "learning_rate": 0.00019730448705798239,
      "loss": 1.1832,
      "step": 85
    },
    {
      "epoch": 1.7647058823529411,
      "grad_norm": 2.953125,
      "learning_rate": 0.00019645845042774553,
      "loss": 1.1402,
      "step": 90
    },
    {
      "epoch": 1.8627450980392157,
      "grad_norm": 4.6875,
      "learning_rate": 0.00019549945702421144,
      "loss": 1.0986,
      "step": 95
    },
    {
      "epoch": 1.9607843137254903,
      "grad_norm": 2.015625,
      "learning_rate": 0.00019442862986770646,
      "loss": 1.0801,
      "step": 100
    },
    {
      "epoch": 2.0,
      "eval_loss": 2.539125442504883,
      "eval_runtime": 2.0419,
      "eval_samples_per_second": 4.897,
      "eval_steps_per_second": 0.979,
      "step": 102
    },
    {
      "epoch": 2.0588235294117645,
      "grad_norm": 42.0,
      "learning_rate": 0.00019324722294043558,
      "loss": 1.0191,
      "step": 105
    },
    {
      "epoch": 2.156862745098039,
      "grad_norm": 3.546875,
      "learning_rate": 0.00019195661971801827,
      "loss": 0.9673,
      "step": 110
    },
    {
      "epoch": 2.2549019607843137,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00019055833154938207,
      "loss": 0.9712,
      "step": 115
    },
    {
      "epoch": 2.3529411764705883,
      "grad_norm": 1.671875,
      "learning_rate": 0.00018905399588691163,
      "loss": 0.9414,
      "step": 120
    },
    {
      "epoch": 2.450980392156863,
      "grad_norm": 1.46875,
      "learning_rate": 0.00018744537436892516,
      "loss": 0.9309,
      "step": 125
    },
    {
      "epoch": 2.549019607843137,
      "grad_norm": 2.359375,
      "learning_rate": 0.00018573435075672424,
      "loss": 0.9278,
      "step": 130
    },
    {
      "epoch": 2.6470588235294117,
      "grad_norm": 1.5625,
      "learning_rate": 0.00018392292872863267,
      "loss": 0.928,
      "step": 135
    },
    {
      "epoch": 2.7450980392156863,
      "grad_norm": 4.6875,
      "learning_rate": 0.00018201322953360758,
      "loss": 0.9431,
      "step": 140
    },
    {
      "epoch": 2.843137254901961,
      "grad_norm": 1.7578125,
      "learning_rate": 0.00018000748950717038,
      "loss": 0.934,
      "step": 145
    },
    {
      "epoch": 2.9411764705882355,
      "grad_norm": 4.8125,
      "learning_rate": 0.00017790805745256704,
      "loss": 0.9341,
      "step": 150
    },
    {
      "epoch": 3.0,
      "eval_loss": 2.587712049484253,
      "eval_runtime": 2.0445,
      "eval_samples_per_second": 4.891,
      "eval_steps_per_second": 0.978,
      "step": 153
    },
    {
      "epoch": 3.0392156862745097,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00017571739189022365,
      "loss": 0.8748,
      "step": 155
    },
    {
      "epoch": 3.1372549019607843,
      "grad_norm": 1.765625,
      "learning_rate": 0.00017343805817871886,
      "loss": 0.805,
      "step": 160
    },
    {
      "epoch": 3.235294117647059,
      "grad_norm": 1.828125,
      "learning_rate": 0.00017107272551064473,
      "loss": 0.7843,
      "step": 165
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 5.40625,
      "learning_rate": 0.0001686241637868734,
      "loss": 0.7917,
      "step": 170
    },
    {
      "epoch": 3.431372549019608,
      "grad_norm": 1.21875,
      "learning_rate": 0.00016609524037289019,
      "loss": 0.8003,
      "step": 175
    },
    {
      "epoch": 3.5294117647058822,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0001634889167409923,
      "loss": 0.7639,
      "step": 180
    },
    {
      "epoch": 3.627450980392157,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00016080824500228367,
      "loss": 0.7796,
      "step": 185
    },
    {
      "epoch": 3.7254901960784315,
      "grad_norm": 1.8515625,
      "learning_rate": 0.00015805636433252891,
      "loss": 0.7783,
      "step": 190
    },
    {
      "epoch": 3.8235294117647056,
      "grad_norm": 1.265625,
      "learning_rate": 0.0001552364972960506,
      "loss": 0.7854,
      "step": 195
    },
    {
      "epoch": 3.9215686274509802,
      "grad_norm": 2.9375,
      "learning_rate": 0.00015235194607197508,
      "loss": 0.7796,
      "step": 200
    },
    {
      "epoch": 4.0,
      "eval_loss": 2.6408610343933105,
      "eval_runtime": 2.0419,
      "eval_samples_per_second": 4.897,
      "eval_steps_per_second": 0.979,
      "step": 204
    },
    {
      "epoch": 4.019607843137255,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0001494060885872464,
      "loss": 0.7823,
      "step": 205
    },
    {
      "epoch": 4.117647058823529,
      "grad_norm": 16.875,
      "learning_rate": 0.00014640237456093634,
      "loss": 0.6465,
      "step": 210
    },
    {
      "epoch": 4.215686274509804,
      "grad_norm": 1.0625,
      "learning_rate": 0.00014334432146448272,
      "loss": 0.621,
      "step": 215
    },
    {
      "epoch": 4.313725490196078,
      "grad_norm": 1.5625,
      "learning_rate": 0.00014023551040258725,
      "loss": 0.612,
      "step": 220
    },
    {
      "epoch": 4.411764705882353,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00013707958191959608,
      "loss": 0.6456,
      "step": 225
    },
    {
      "epoch": 4.509803921568627,
      "grad_norm": 1.375,
      "learning_rate": 0.00013388023173627414,
      "loss": 0.6361,
      "step": 230
    },
    {
      "epoch": 4.607843137254902,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00013064120642196548,
      "loss": 0.6487,
      "step": 235
    },
    {
      "epoch": 4.705882352941177,
      "grad_norm": 0.921875,
      "learning_rate": 0.0001273662990072083,
      "loss": 0.6424,
      "step": 240
    },
    {
      "epoch": 4.803921568627451,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00012405934454194146,
      "loss": 0.653,
      "step": 245
    },
    {
      "epoch": 4.901960784313726,
      "grad_norm": 0.953125,
      "learning_rate": 0.00012072421560450497,
      "loss": 0.6493,
      "step": 250
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00011736481776669306,
      "loss": 0.6583,
      "step": 255
    },
    {
      "epoch": 5.0,
      "eval_loss": 2.8397467136383057,
      "eval_runtime": 2.0446,
      "eval_samples_per_second": 4.891,
      "eval_steps_per_second": 0.978,
      "step": 255
    },
    {
      "epoch": 5.098039215686274,
      "grad_norm": 1.9609375,
      "learning_rate": 0.00011398508502017047,
      "loss": 0.5107,
      "step": 260
    },
    {
      "epoch": 5.196078431372549,
      "grad_norm": 1.09375,
      "learning_rate": 0.00011058897516960816,
      "loss": 0.4939,
      "step": 265
    },
    {
      "epoch": 5.294117647058823,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00010718046519793276,
      "loss": 0.4931,
      "step": 270
    },
    {
      "epoch": 5.392156862745098,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00010376354660911771,
      "loss": 0.4983,
      "step": 275
    },
    {
      "epoch": 5.490196078431373,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00010034222075396953,
      "loss": 0.5112,
      "step": 280
    },
    {
      "epoch": 5.588235294117647,
      "grad_norm": 1.0546875,
      "learning_rate": 9.692049414438299e-05,
      "loss": 0.4841,
      "step": 285
    },
    {
      "epoch": 5.686274509803922,
      "grad_norm": 1.0078125,
      "learning_rate": 9.350237376155269e-05,
      "loss": 0.4976,
      "step": 290
    },
    {
      "epoch": 5.784313725490196,
      "grad_norm": 1.0390625,
      "learning_rate": 9.00918623636349e-05,
      "loss": 0.4954,
      "step": 295
    },
    {
      "epoch": 5.882352941176471,
      "grad_norm": 1.125,
      "learning_rate": 8.669295379835467e-05,
      "loss": 0.4989,
      "step": 300
    },
    {
      "epoch": 5.980392156862745,
      "grad_norm": 0.96484375,
      "learning_rate": 8.330962832604747e-05,
      "loss": 0.4945,
      "step": 305
    },
    {
      "epoch": 6.0,
      "eval_loss": 3.189112424850464,
      "eval_runtime": 2.0474,
      "eval_samples_per_second": 4.884,
      "eval_steps_per_second": 0.977,
      "step": 306
    },
    {
      "epoch": 6.078431372549019,
      "grad_norm": 1.03125,
      "learning_rate": 7.994584795861249e-05,
      "loss": 0.4204,
      "step": 310
    },
    {
      "epoch": 6.176470588235294,
      "grad_norm": 1.109375,
      "learning_rate": 7.660555181983518e-05,
      "loss": 0.3844,
      "step": 315
    },
    {
      "epoch": 6.2745098039215685,
      "grad_norm": 1.21875,
      "learning_rate": 7.329265153251285e-05,
      "loss": 0.3839,
      "step": 320
    },
    {
      "epoch": 6.372549019607844,
      "grad_norm": 0.9453125,
      "learning_rate": 7.001102663778533e-05,
      "loss": 0.3772,
      "step": 325
    },
    {
      "epoch": 6.470588235294118,
      "grad_norm": 0.99609375,
      "learning_rate": 6.676452005203406e-05,
      "loss": 0.3877,
      "step": 330
    },
    {
      "epoch": 6.568627450980392,
      "grad_norm": 1.0078125,
      "learning_rate": 6.355693356667065e-05,
      "loss": 0.388,
      "step": 335
    },
    {
      "epoch": 6.666666666666667,
      "grad_norm": 0.98046875,
      "learning_rate": 6.039202339608432e-05,
      "loss": 0.3832,
      "step": 340
    },
    {
      "epoch": 6.764705882352941,
      "grad_norm": 1.0859375,
      "learning_rate": 5.727349577896194e-05,
      "loss": 0.3926,
      "step": 345
    },
    {
      "epoch": 6.862745098039216,
      "grad_norm": 1.046875,
      "learning_rate": 5.4205002638131406e-05,
      "loss": 0.387,
      "step": 350
    },
    {
      "epoch": 6.96078431372549,
      "grad_norm": 0.98046875,
      "learning_rate": 5.119013730401152e-05,
      "loss": 0.3891,
      "step": 355
    },
    {
      "epoch": 7.0,
      "eval_loss": 3.7245430946350098,
      "eval_runtime": 2.0488,
      "eval_samples_per_second": 4.881,
      "eval_steps_per_second": 0.976,
      "step": 357
    },
    {
      "epoch": 7.0588235294117645,
      "grad_norm": 0.81640625,
      "learning_rate": 4.823243030667576e-05,
      "loss": 0.3491,
      "step": 360
    },
    {
      "epoch": 7.1568627450980395,
      "grad_norm": 1.0546875,
      "learning_rate": 4.533534524145756e-05,
      "loss": 0.3038,
      "step": 365
    },
    {
      "epoch": 7.254901960784314,
      "grad_norm": 0.91015625,
      "learning_rate": 4.250227471293935e-05,
      "loss": 0.3093,
      "step": 370
    },
    {
      "epoch": 7.352941176470588,
      "grad_norm": 0.99609375,
      "learning_rate": 3.973653636207437e-05,
      "loss": 0.3096,
      "step": 375
    },
    {
      "epoch": 7.450980392156863,
      "grad_norm": 0.94921875,
      "learning_rate": 3.7041368981094024e-05,
      "loss": 0.3137,
      "step": 380
    },
    {
      "epoch": 7.549019607843137,
      "grad_norm": 0.984375,
      "learning_rate": 3.441992872075027e-05,
      "loss": 0.3073,
      "step": 385
    },
    {
      "epoch": 7.647058823529412,
      "grad_norm": 0.9375,
      "learning_rate": 3.187528539433458e-05,
      "loss": 0.2992,
      "step": 390
    },
    {
      "epoch": 7.745098039215686,
      "grad_norm": 0.93359375,
      "learning_rate": 2.9410418882801683e-05,
      "loss": 0.3063,
      "step": 395
    },
    {
      "epoch": 7.8431372549019605,
      "grad_norm": 0.94140625,
      "learning_rate": 2.702821564520732e-05,
      "loss": 0.308,
      "step": 400
    },
    {
      "epoch": 7.9411764705882355,
      "grad_norm": 0.8984375,
      "learning_rate": 2.4731465338547556e-05,
      "loss": 0.3029,
      "step": 405
    },
    {
      "epoch": 8.0,
      "eval_loss": 4.195121765136719,
      "eval_runtime": 2.0472,
      "eval_samples_per_second": 4.885,
      "eval_steps_per_second": 0.977,
      "step": 408
    },
    {
      "epoch": 8.03921568627451,
      "grad_norm": 0.796875,
      "learning_rate": 2.252285755095652e-05,
      "loss": 0.2897,
      "step": 410
    },
    {
      "epoch": 8.137254901960784,
      "grad_norm": 0.80078125,
      "learning_rate": 2.0404978652089323e-05,
      "loss": 0.2649,
      "step": 415
    },
    {
      "epoch": 8.235294117647058,
      "grad_norm": 0.87890625,
      "learning_rate": 1.8380308764377842e-05,
      "loss": 0.2647,
      "step": 420
    },
    {
      "epoch": 8.333333333333334,
      "grad_norm": 0.828125,
      "learning_rate": 1.6451218858706374e-05,
      "loss": 0.2644,
      "step": 425
    },
    {
      "epoch": 8.431372549019608,
      "grad_norm": 0.77734375,
      "learning_rate": 1.4619967977908156e-05,
      "loss": 0.2709,
      "step": 430
    },
    {
      "epoch": 8.529411764705882,
      "grad_norm": 0.79296875,
      "learning_rate": 1.2888700591334223e-05,
      "loss": 0.2621,
      "step": 435
    },
    {
      "epoch": 8.627450980392156,
      "grad_norm": 0.8125,
      "learning_rate": 1.1259444083592585e-05,
      "loss": 0.2628,
      "step": 440
    },
    {
      "epoch": 8.72549019607843,
      "grad_norm": 0.81640625,
      "learning_rate": 9.734106380398023e-06,
      "loss": 0.2556,
      "step": 445
    },
    {
      "epoch": 8.823529411764707,
      "grad_norm": 0.765625,
      "learning_rate": 8.314473714313719e-06,
      "loss": 0.2611,
      "step": 450
    },
    {
      "epoch": 8.92156862745098,
      "grad_norm": 0.80859375,
      "learning_rate": 7.002208532999932e-06,
      "loss": 0.259,
      "step": 455
    },
    {
      "epoch": 9.0,
      "eval_loss": 4.499588966369629,
      "eval_runtime": 2.0494,
      "eval_samples_per_second": 4.88,
      "eval_steps_per_second": 0.976,
      "step": 459
    },
    {
      "epoch": 9.019607843137255,
      "grad_norm": 0.765625,
      "learning_rate": 5.798847552420183e-06,
      "loss": 0.2564,
      "step": 460
    },
    {
      "epoch": 9.117647058823529,
      "grad_norm": 0.74609375,
      "learning_rate": 4.705799957284351e-06,
      "loss": 0.2573,
      "step": 465
    },
    {
      "epoch": 9.215686274509803,
      "grad_norm": 0.75,
      "learning_rate": 3.724345750835878e-06,
      "loss": 0.2564,
      "step": 470
    },
    {
      "epoch": 9.313725490196079,
      "grad_norm": 0.78515625,
      "learning_rate": 2.855634255915951e-06,
      "loss": 0.2487,
      "step": 475
    },
    {
      "epoch": 9.411764705882353,
      "grad_norm": 0.7421875,
      "learning_rate": 2.100682769059548e-06,
      "loss": 0.2552,
      "step": 480
    },
    {
      "epoch": 9.509803921568627,
      "grad_norm": 0.765625,
      "learning_rate": 1.4603753691998735e-06,
      "loss": 0.253,
      "step": 485
    },
    {
      "epoch": 9.607843137254902,
      "grad_norm": 0.7578125,
      "learning_rate": 9.354618823758655e-07,
      "loss": 0.251,
      "step": 490
    },
    {
      "epoch": 9.705882352941176,
      "grad_norm": 0.734375,
      "learning_rate": 5.265570036553813e-07,
      "loss": 0.2528,
      "step": 495
    },
    {
      "epoch": 9.803921568627452,
      "grad_norm": 0.76171875,
      "learning_rate": 2.3413957730226144e-07,
      "loss": 0.2577,
      "step": 500
    },
    {
      "epoch": 9.901960784313726,
      "grad_norm": 0.7890625,
      "learning_rate": 5.8552036030179445e-08,
      "loss": 0.2508,
      "step": 505
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0,
      "loss": 0.2532,
      "step": 510
    },
    {
      "epoch": 10.0,
      "eval_loss": 4.513559818267822,
      "eval_runtime": 2.0473,
      "eval_samples_per_second": 4.884,
      "eval_steps_per_second": 0.977,
      "step": 510
    },
    {
      "epoch": 10.0,
      "step": 510,
      "total_flos": 3.8986265891805594e+17,
      "train_loss": 2.3579357918571024,
      "train_runtime": 3338.7641,
      "train_samples_per_second": 2.441,
      "train_steps_per_second": 0.153
    }
  ],
  "logging_steps": 5,
  "max_steps": 510,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 100,
  "total_flos": 3.8986265891805594e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}