{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9894459102902374,
  "eval_steps": 500,
  "global_step": 1500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "learning_rate": 4.989006156552331e-05,
      "loss": 1.8206,
      "step": 10
    },
    {
      "epoch": 0.01,
      "learning_rate": 4.978012313104662e-05,
      "loss": 1.7673,
      "step": 20
    },
    {
      "epoch": 0.02,
      "learning_rate": 4.967018469656992e-05,
      "loss": 1.7575,
      "step": 30
    },
    {
      "epoch": 0.03,
      "learning_rate": 4.956024626209323e-05,
      "loss": 1.6743,
      "step": 40
    },
    {
      "epoch": 0.03,
      "learning_rate": 4.9450307827616535e-05,
      "loss": 1.5758,
      "step": 50
    },
    {
      "epoch": 0.04,
      "learning_rate": 4.934036939313985e-05,
      "loss": 1.615,
      "step": 60
    },
    {
      "epoch": 0.05,
      "learning_rate": 4.9230430958663146e-05,
      "loss": 1.5376,
      "step": 70
    },
    {
      "epoch": 0.05,
      "learning_rate": 4.912049252418646e-05,
      "loss": 1.6185,
      "step": 80
    },
    {
      "epoch": 0.06,
      "learning_rate": 4.901055408970976e-05,
      "loss": 1.5484,
      "step": 90
    },
    {
      "epoch": 0.07,
      "learning_rate": 4.8900615655233075e-05,
      "loss": 1.572,
      "step": 100
    },
    {
      "epoch": 0.07,
      "learning_rate": 4.879067722075638e-05,
      "loss": 1.5253,
      "step": 110
    },
    {
      "epoch": 0.08,
      "learning_rate": 4.8680738786279686e-05,
      "loss": 1.5206,
      "step": 120
    },
    {
      "epoch": 0.09,
      "learning_rate": 4.857080035180299e-05,
      "loss": 1.5172,
      "step": 130
    },
    {
      "epoch": 0.09,
      "learning_rate": 4.84608619173263e-05,
      "loss": 1.526,
      "step": 140
    },
    {
      "epoch": 0.1,
      "learning_rate": 4.835092348284961e-05,
      "loss": 1.5833,
      "step": 150
    },
    {
      "epoch": 0.11,
      "learning_rate": 4.824098504837291e-05,
      "loss": 1.5402,
      "step": 160
    },
    {
      "epoch": 0.11,
      "learning_rate": 4.813104661389622e-05,
      "loss": 1.5685,
      "step": 170
    },
    {
      "epoch": 0.12,
      "learning_rate": 4.802110817941953e-05,
      "loss": 1.5386,
      "step": 180
    },
    {
      "epoch": 0.13,
      "learning_rate": 4.7911169744942836e-05,
      "loss": 1.5592,
      "step": 190
    },
    {
      "epoch": 0.13,
      "learning_rate": 4.780123131046614e-05,
      "loss": 1.5198,
      "step": 200
    },
    {
      "epoch": 0.14,
      "learning_rate": 4.7691292875989446e-05,
      "loss": 1.6008,
      "step": 210
    },
    {
      "epoch": 0.15,
      "learning_rate": 4.758135444151275e-05,
      "loss": 1.5608,
      "step": 220
    },
    {
      "epoch": 0.15,
      "learning_rate": 4.7471416007036063e-05,
      "loss": 1.5535,
      "step": 230
    },
    {
      "epoch": 0.16,
      "learning_rate": 4.736147757255937e-05,
      "loss": 1.5633,
      "step": 240
    },
    {
      "epoch": 0.16,
      "learning_rate": 4.7251539138082674e-05,
      "loss": 1.56,
      "step": 250
    },
    {
      "epoch": 0.17,
      "learning_rate": 4.714160070360598e-05,
      "loss": 1.5407,
      "step": 260
    },
    {
      "epoch": 0.18,
      "learning_rate": 4.703166226912929e-05,
      "loss": 1.4851,
      "step": 270
    },
    {
      "epoch": 0.18,
      "learning_rate": 4.6921723834652596e-05,
      "loss": 1.5332,
      "step": 280
    },
    {
      "epoch": 0.19,
      "learning_rate": 4.681178540017591e-05,
      "loss": 1.5681,
      "step": 290
    },
    {
      "epoch": 0.2,
      "learning_rate": 4.670184696569921e-05,
      "loss": 1.5634,
      "step": 300
    },
    {
      "epoch": 0.2,
      "learning_rate": 4.659190853122252e-05,
      "loss": 1.5432,
      "step": 310
    },
    {
      "epoch": 0.21,
      "learning_rate": 4.6481970096745824e-05,
      "loss": 1.5003,
      "step": 320
    },
    {
      "epoch": 0.22,
      "learning_rate": 4.6372031662269136e-05,
      "loss": 1.5412,
      "step": 330
    },
    {
      "epoch": 0.22,
      "learning_rate": 4.6262093227792435e-05,
      "loss": 1.5273,
      "step": 340
    },
    {
      "epoch": 0.23,
      "learning_rate": 4.615215479331575e-05,
      "loss": 1.5083,
      "step": 350
    },
    {
      "epoch": 0.24,
      "learning_rate": 4.604221635883905e-05,
      "loss": 1.5037,
      "step": 360
    },
    {
      "epoch": 0.24,
      "learning_rate": 4.593227792436236e-05,
      "loss": 1.5416,
      "step": 370
    },
    {
      "epoch": 0.25,
      "learning_rate": 4.582233948988567e-05,
      "loss": 1.5093,
      "step": 380
    },
    {
      "epoch": 0.26,
      "learning_rate": 4.5712401055408974e-05,
      "loss": 1.5609,
      "step": 390
    },
    {
      "epoch": 0.26,
      "learning_rate": 4.560246262093228e-05,
      "loss": 1.5628,
      "step": 400
    },
    {
      "epoch": 0.27,
      "learning_rate": 4.5492524186455585e-05,
      "loss": 1.5398,
      "step": 410
    },
    {
      "epoch": 0.28,
      "learning_rate": 4.53825857519789e-05,
      "loss": 1.4759,
      "step": 420
    },
    {
      "epoch": 0.28,
      "learning_rate": 4.52726473175022e-05,
      "loss": 1.5367,
      "step": 430
    },
    {
      "epoch": 0.29,
      "learning_rate": 4.516270888302551e-05,
      "loss": 1.5037,
      "step": 440
    },
    {
      "epoch": 0.3,
      "learning_rate": 4.505277044854881e-05,
      "loss": 1.534,
      "step": 450
    },
    {
      "epoch": 0.3,
      "learning_rate": 4.4942832014072125e-05,
      "loss": 1.507,
      "step": 460
    },
    {
      "epoch": 0.31,
      "learning_rate": 4.483289357959543e-05,
      "loss": 1.491,
      "step": 470
    },
    {
      "epoch": 0.32,
      "learning_rate": 4.4722955145118735e-05,
      "loss": 1.4894,
      "step": 480
    },
    {
      "epoch": 0.32,
      "learning_rate": 4.461301671064204e-05,
      "loss": 1.5337,
      "step": 490
    },
    {
      "epoch": 0.33,
      "learning_rate": 4.450307827616535e-05,
      "loss": 1.5258,
      "step": 500
    },
    {
      "epoch": 0.34,
      "learning_rate": 4.439313984168866e-05,
      "loss": 1.537,
      "step": 510
    },
    {
      "epoch": 0.34,
      "learning_rate": 4.428320140721196e-05,
      "loss": 1.4971,
      "step": 520
    },
    {
      "epoch": 0.35,
      "learning_rate": 4.417326297273527e-05,
      "loss": 1.4865,
      "step": 530
    },
    {
      "epoch": 0.36,
      "learning_rate": 4.406332453825858e-05,
      "loss": 1.5169,
      "step": 540
    },
    {
      "epoch": 0.36,
      "learning_rate": 4.3953386103781885e-05,
      "loss": 1.5243,
      "step": 550
    },
    {
      "epoch": 0.37,
      "learning_rate": 4.384344766930519e-05,
      "loss": 1.5308,
      "step": 560
    },
    {
      "epoch": 0.38,
      "learning_rate": 4.3733509234828496e-05,
      "loss": 1.5089,
      "step": 570
    },
    {
      "epoch": 0.38,
      "learning_rate": 4.36235708003518e-05,
      "loss": 1.5039,
      "step": 580
    },
    {
      "epoch": 0.39,
      "learning_rate": 4.351363236587511e-05,
      "loss": 1.4936,
      "step": 590
    },
    {
      "epoch": 0.4,
      "learning_rate": 4.340369393139842e-05,
      "loss": 1.5111,
      "step": 600
    },
    {
      "epoch": 0.4,
      "learning_rate": 4.329375549692173e-05,
      "loss": 1.4898,
      "step": 610
    },
    {
      "epoch": 0.41,
      "learning_rate": 4.318381706244503e-05,
      "loss": 1.5017,
      "step": 620
    },
    {
      "epoch": 0.42,
      "learning_rate": 4.307387862796834e-05,
      "loss": 1.5017,
      "step": 630
    },
    {
      "epoch": 0.42,
      "learning_rate": 4.2963940193491646e-05,
      "loss": 1.5331,
      "step": 640
    },
    {
      "epoch": 0.43,
      "learning_rate": 4.285400175901496e-05,
      "loss": 1.5442,
      "step": 650
    },
    {
      "epoch": 0.44,
      "learning_rate": 4.2744063324538256e-05,
      "loss": 1.5269,
      "step": 660
    },
    {
      "epoch": 0.44,
      "learning_rate": 4.263412489006157e-05,
      "loss": 1.5329,
      "step": 670
    },
    {
      "epoch": 0.45,
      "learning_rate": 4.2524186455584874e-05,
      "loss": 1.5572,
      "step": 680
    },
    {
      "epoch": 0.46,
      "learning_rate": 4.2414248021108186e-05,
      "loss": 1.5299,
      "step": 690
    },
    {
      "epoch": 0.46,
      "learning_rate": 4.230430958663149e-05,
      "loss": 1.4894,
      "step": 700
    },
    {
      "epoch": 0.47,
      "learning_rate": 4.2194371152154796e-05,
      "loss": 1.5005,
      "step": 710
    },
    {
      "epoch": 0.47,
      "learning_rate": 4.20844327176781e-05,
      "loss": 1.5133,
      "step": 720
    },
    {
      "epoch": 0.48,
      "learning_rate": 4.1974494283201407e-05,
      "loss": 1.5079,
      "step": 730
    },
    {
      "epoch": 0.49,
      "learning_rate": 4.186455584872472e-05,
      "loss": 1.5001,
      "step": 740
    },
    {
      "epoch": 0.49,
      "learning_rate": 4.1754617414248024e-05,
      "loss": 1.5151,
      "step": 750
    },
    {
      "epoch": 0.5,
      "learning_rate": 4.164467897977133e-05,
      "loss": 1.4863,
      "step": 760
    },
    {
      "epoch": 0.51,
      "learning_rate": 4.1534740545294634e-05,
      "loss": 1.5193,
      "step": 770
    },
    {
      "epoch": 0.51,
      "learning_rate": 4.1424802110817946e-05,
      "loss": 1.4906,
      "step": 780
    },
    {
      "epoch": 0.52,
      "learning_rate": 4.131486367634125e-05,
      "loss": 1.544,
      "step": 790
    },
    {
      "epoch": 0.53,
      "learning_rate": 4.120492524186456e-05,
      "loss": 1.5458,
      "step": 800
    },
    {
      "epoch": 0.53,
      "learning_rate": 4.109498680738786e-05,
      "loss": 1.5614,
      "step": 810
    },
    {
      "epoch": 0.54,
      "learning_rate": 4.0985048372911174e-05,
      "loss": 1.5536,
      "step": 820
    },
    {
      "epoch": 0.55,
      "learning_rate": 4.087510993843448e-05,
      "loss": 1.5204,
      "step": 830
    },
    {
      "epoch": 0.55,
      "learning_rate": 4.0765171503957784e-05,
      "loss": 1.4918,
      "step": 840
    },
    {
      "epoch": 0.56,
      "learning_rate": 4.065523306948109e-05,
      "loss": 1.4721,
      "step": 850
    },
    {
      "epoch": 0.57,
      "learning_rate": 4.05452946350044e-05,
      "loss": 1.5353,
      "step": 860
    },
    {
      "epoch": 0.57,
      "learning_rate": 4.043535620052771e-05,
      "loss": 1.5162,
      "step": 870
    },
    {
      "epoch": 0.58,
      "learning_rate": 4.032541776605102e-05,
      "loss": 1.5408,
      "step": 880
    },
    {
      "epoch": 0.59,
      "learning_rate": 4.021547933157432e-05,
      "loss": 1.5148,
      "step": 890
    },
    {
      "epoch": 0.59,
      "learning_rate": 4.010554089709763e-05,
      "loss": 1.5481,
      "step": 900
    },
    {
      "epoch": 0.6,
      "learning_rate": 3.9995602462620935e-05,
      "loss": 1.4869,
      "step": 910
    },
    {
      "epoch": 0.61,
      "learning_rate": 3.988566402814424e-05,
      "loss": 1.556,
      "step": 920
    },
    {
      "epoch": 0.61,
      "learning_rate": 3.9775725593667545e-05,
      "loss": 1.5087,
      "step": 930
    },
    {
      "epoch": 0.62,
      "learning_rate": 3.966578715919085e-05,
      "loss": 1.5089,
      "step": 940
    },
    {
      "epoch": 0.63,
      "learning_rate": 3.955584872471416e-05,
      "loss": 1.512,
      "step": 950
    },
    {
      "epoch": 0.63,
      "learning_rate": 3.944591029023747e-05,
      "loss": 1.4822,
      "step": 960
    },
    {
      "epoch": 0.64,
      "learning_rate": 3.933597185576078e-05,
      "loss": 1.4931,
      "step": 970
    },
    {
      "epoch": 0.65,
      "learning_rate": 3.922603342128408e-05,
      "loss": 1.5033,
      "step": 980
    },
    {
      "epoch": 0.65,
      "learning_rate": 3.911609498680739e-05,
      "loss": 1.545,
      "step": 990
    },
    {
      "epoch": 0.66,
      "learning_rate": 3.9006156552330695e-05,
      "loss": 1.4808,
      "step": 1000
    },
    {
      "epoch": 0.67,
      "learning_rate": 3.889621811785401e-05,
      "loss": 1.5541,
      "step": 1010
    },
    {
      "epoch": 0.67,
      "learning_rate": 3.8786279683377306e-05,
      "loss": 1.5157,
      "step": 1020
    },
    {
      "epoch": 0.68,
      "learning_rate": 3.867634124890062e-05,
      "loss": 1.502,
      "step": 1030
    },
    {
      "epoch": 0.69,
      "learning_rate": 3.856640281442392e-05,
      "loss": 1.531,
      "step": 1040
    },
    {
      "epoch": 0.69,
      "learning_rate": 3.8456464379947235e-05,
      "loss": 1.5537,
      "step": 1050
    },
    {
      "epoch": 0.7,
      "learning_rate": 3.834652594547054e-05,
      "loss": 1.5171,
      "step": 1060
    },
    {
      "epoch": 0.71,
      "learning_rate": 3.8236587510993846e-05,
      "loss": 1.4988,
      "step": 1070
    },
    {
      "epoch": 0.71,
      "learning_rate": 3.812664907651715e-05,
      "loss": 1.5139,
      "step": 1080
    },
    {
      "epoch": 0.72,
      "learning_rate": 3.801671064204046e-05,
      "loss": 1.5116,
      "step": 1090
    },
    {
      "epoch": 0.73,
      "learning_rate": 3.790677220756377e-05,
      "loss": 1.5541,
      "step": 1100
    },
    {
      "epoch": 0.73,
      "learning_rate": 3.779683377308707e-05,
      "loss": 1.5128,
      "step": 1110
    },
    {
      "epoch": 0.74,
      "learning_rate": 3.768689533861038e-05,
      "loss": 1.5127,
      "step": 1120
    },
    {
      "epoch": 0.75,
      "learning_rate": 3.7576956904133684e-05,
      "loss": 1.5146,
      "step": 1130
    },
    {
      "epoch": 0.75,
      "learning_rate": 3.7467018469656996e-05,
      "loss": 1.5715,
      "step": 1140
    },
    {
      "epoch": 0.76,
      "learning_rate": 3.73570800351803e-05,
      "loss": 1.5222,
      "step": 1150
    },
    {
      "epoch": 0.77,
      "learning_rate": 3.7247141600703606e-05,
      "loss": 1.4698,
      "step": 1160
    },
    {
      "epoch": 0.77,
      "learning_rate": 3.713720316622691e-05,
      "loss": 1.5553,
      "step": 1170
    },
    {
      "epoch": 0.78,
      "learning_rate": 3.7027264731750223e-05,
      "loss": 1.5197,
      "step": 1180
    },
    {
      "epoch": 0.78,
      "learning_rate": 3.691732629727353e-05,
      "loss": 1.5245,
      "step": 1190
    },
    {
      "epoch": 0.79,
      "learning_rate": 3.680738786279684e-05,
      "loss": 1.526,
      "step": 1200
    },
    {
      "epoch": 0.8,
      "learning_rate": 3.669744942832014e-05,
      "loss": 1.5052,
      "step": 1210
    },
    {
      "epoch": 0.8,
      "learning_rate": 3.658751099384345e-05,
      "loss": 1.5379,
      "step": 1220
    },
    {
      "epoch": 0.81,
      "learning_rate": 3.6477572559366756e-05,
      "loss": 1.5566,
      "step": 1230
    },
    {
      "epoch": 0.82,
      "learning_rate": 3.636763412489007e-05,
      "loss": 1.5293,
      "step": 1240
    },
    {
      "epoch": 0.82,
      "learning_rate": 3.625769569041337e-05,
      "loss": 1.5149,
      "step": 1250
    },
    {
      "epoch": 0.83,
      "learning_rate": 3.614775725593668e-05,
      "loss": 1.5626,
      "step": 1260
    },
    {
      "epoch": 0.84,
      "learning_rate": 3.6037818821459984e-05,
      "loss": 1.5603,
      "step": 1270
    },
    {
      "epoch": 0.84,
      "learning_rate": 3.592788038698329e-05,
      "loss": 1.5378,
      "step": 1280
    },
    {
      "epoch": 0.85,
      "learning_rate": 3.58179419525066e-05,
      "loss": 1.4977,
      "step": 1290
    },
    {
      "epoch": 0.86,
      "learning_rate": 3.5708003518029907e-05,
      "loss": 1.476,
      "step": 1300
    },
    {
      "epoch": 0.86,
      "learning_rate": 3.559806508355321e-05,
      "loss": 1.4655,
      "step": 1310
    },
    {
      "epoch": 0.87,
      "learning_rate": 3.548812664907652e-05,
      "loss": 1.5385,
      "step": 1320
    },
    {
      "epoch": 0.88,
      "learning_rate": 3.537818821459983e-05,
      "loss": 1.4823,
      "step": 1330
    },
    {
      "epoch": 0.88,
      "learning_rate": 3.526824978012313e-05,
      "loss": 1.5179,
      "step": 1340
    },
    {
      "epoch": 0.89,
      "learning_rate": 3.515831134564644e-05,
      "loss": 1.5145,
      "step": 1350
    },
    {
      "epoch": 0.9,
      "learning_rate": 3.5048372911169745e-05,
      "loss": 1.5374,
      "step": 1360
    },
    {
      "epoch": 0.9,
      "learning_rate": 3.493843447669306e-05,
      "loss": 1.5339,
      "step": 1370
    },
    {
      "epoch": 0.91,
      "learning_rate": 3.482849604221636e-05,
      "loss": 1.5322,
      "step": 1380
    },
    {
      "epoch": 0.92,
      "learning_rate": 3.471855760773967e-05,
      "loss": 1.5059,
      "step": 1390
    },
    {
      "epoch": 0.92,
      "learning_rate": 3.460861917326297e-05,
      "loss": 1.5242,
      "step": 1400
    },
    {
      "epoch": 0.93,
      "learning_rate": 3.4498680738786285e-05,
      "loss": 1.513,
      "step": 1410
    },
    {
      "epoch": 0.94,
      "learning_rate": 3.438874230430959e-05,
      "loss": 1.5219,
      "step": 1420
    },
    {
      "epoch": 0.94,
      "learning_rate": 3.4278803869832895e-05,
      "loss": 1.5234,
      "step": 1430
    },
    {
      "epoch": 0.95,
      "learning_rate": 3.41688654353562e-05,
      "loss": 1.4968,
      "step": 1440
    },
    {
      "epoch": 0.96,
      "learning_rate": 3.405892700087951e-05,
      "loss": 1.5396,
      "step": 1450
    },
    {
      "epoch": 0.96,
      "learning_rate": 3.394898856640282e-05,
      "loss": 1.5557,
      "step": 1460
    },
    {
      "epoch": 0.97,
      "learning_rate": 3.383905013192612e-05,
      "loss": 1.4818,
      "step": 1470
    },
    {
      "epoch": 0.98,
      "learning_rate": 3.372911169744943e-05,
      "loss": 1.4661,
      "step": 1480
    },
    {
      "epoch": 0.98,
      "learning_rate": 3.361917326297273e-05,
      "loss": 1.5072,
      "step": 1490
    },
    {
      "epoch": 0.99,
      "learning_rate": 3.3509234828496045e-05,
      "loss": 1.5356,
      "step": 1500
    }
  ],
  "logging_steps": 10,
  "max_steps": 4548,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "total_flos": 9.4803723288576e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}