{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 512,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01953125,
      "grad_norm": 1.3515021800994873,
      "learning_rate": 1.875e-06,
      "loss": 1.333,
      "step": 5
    },
    {
      "epoch": 0.0390625,
      "grad_norm": 0.9634742736816406,
      "learning_rate": 4.21875e-06,
      "loss": 1.3235,
      "step": 10
    },
    {
      "epoch": 0.05859375,
      "grad_norm": 0.7603744268417358,
      "learning_rate": 6.5625e-06,
      "loss": 1.2374,
      "step": 15
    },
    {
      "epoch": 0.078125,
      "grad_norm": 0.898777186870575,
      "learning_rate": 8.90625e-06,
      "loss": 1.2363,
      "step": 20
    },
    {
      "epoch": 0.09765625,
      "grad_norm": 0.5078862309455872,
      "learning_rate": 1.125e-05,
      "loss": 1.2713,
      "step": 25
    },
    {
      "epoch": 0.1171875,
      "grad_norm": 0.6618434190750122,
      "learning_rate": 1.359375e-05,
      "loss": 1.2592,
      "step": 30
    },
    {
      "epoch": 0.13671875,
      "grad_norm": 0.4789685904979706,
      "learning_rate": 1.59375e-05,
      "loss": 1.1962,
      "step": 35
    },
    {
      "epoch": 0.15625,
      "grad_norm": 0.4443208873271942,
      "learning_rate": 1.828125e-05,
      "loss": 1.2079,
      "step": 40
    },
    {
      "epoch": 0.17578125,
      "grad_norm": 0.42774105072021484,
      "learning_rate": 2.0625e-05,
      "loss": 1.1485,
      "step": 45
    },
    {
      "epoch": 0.1953125,
      "grad_norm": 0.4273070991039276,
      "learning_rate": 2.296875e-05,
      "loss": 1.1809,
      "step": 50
    },
    {
      "epoch": 0.21484375,
      "grad_norm": 0.4687409996986389,
      "learning_rate": 2.5312500000000002e-05,
      "loss": 1.1016,
      "step": 55
    },
    {
      "epoch": 0.234375,
      "grad_norm": 0.4467330873012543,
      "learning_rate": 2.765625e-05,
      "loss": 1.1695,
      "step": 60
    },
    {
      "epoch": 0.25390625,
      "grad_norm": 0.6110517978668213,
      "learning_rate": 3e-05,
      "loss": 1.1325,
      "step": 65
    },
    {
      "epoch": 0.2734375,
      "grad_norm": 0.75448077917099,
      "learning_rate": 2.9998748508718575e-05,
      "loss": 1.093,
      "step": 70
    },
    {
      "epoch": 0.29296875,
      "grad_norm": 0.6111339330673218,
      "learning_rate": 2.9994994243705013e-05,
      "loss": 1.1134,
      "step": 75
    },
    {
      "epoch": 0.3125,
      "grad_norm": 0.5295779705047607,
      "learning_rate": 2.9988737831416642e-05,
      "loss": 1.0628,
      "step": 80
    },
    {
      "epoch": 0.33203125,
      "grad_norm": 0.538719654083252,
      "learning_rate": 2.9979980315832853e-05,
      "loss": 1.0402,
      "step": 85
    },
    {
      "epoch": 0.3515625,
      "grad_norm": 0.554957389831543,
      "learning_rate": 2.9968723158280906e-05,
      "loss": 1.0623,
      "step": 90
    },
    {
      "epoch": 0.37109375,
      "grad_norm": 0.5269491076469421,
      "learning_rate": 2.995496823719206e-05,
      "loss": 1.0775,
      "step": 95
    },
    {
      "epoch": 0.390625,
      "grad_norm": 0.5827288031578064,
      "learning_rate": 2.9938717847788167e-05,
      "loss": 0.9854,
      "step": 100
    },
    {
      "epoch": 0.41015625,
      "grad_norm": 0.5621783137321472,
      "learning_rate": 2.9919974701698638e-05,
      "loss": 0.9554,
      "step": 105
    },
    {
      "epoch": 0.4296875,
      "grad_norm": 0.7156975269317627,
      "learning_rate": 2.989874192650801e-05,
      "loss": 0.9722,
      "step": 110
    },
    {
      "epoch": 0.44921875,
      "grad_norm": 0.5912815928459167,
      "learning_rate": 2.9875023065234003e-05,
      "loss": 0.9747,
      "step": 115
    },
    {
      "epoch": 0.46875,
      "grad_norm": 0.686764657497406,
      "learning_rate": 2.984882207573638e-05,
      "loss": 1.0,
      "step": 120
    },
    {
      "epoch": 0.48828125,
      "grad_norm": 0.7672634720802307,
      "learning_rate": 2.982014333005645e-05,
      "loss": 0.9985,
      "step": 125
    },
    {
      "epoch": 0.5078125,
      "grad_norm": 0.7488982677459717,
      "learning_rate": 2.9788991613687575e-05,
      "loss": 0.9438,
      "step": 130
    },
    {
      "epoch": 0.52734375,
      "grad_norm": 0.7190539240837097,
      "learning_rate": 2.9755372124776616e-05,
      "loss": 0.9301,
      "step": 135
    },
    {
      "epoch": 0.546875,
      "grad_norm": 0.6693652868270874,
      "learning_rate": 2.971929047325654e-05,
      "loss": 0.8901,
      "step": 140
    },
    {
      "epoch": 0.56640625,
      "grad_norm": 0.7850137948989868,
      "learning_rate": 2.968075267991032e-05,
      "loss": 0.8936,
      "step": 145
    },
    {
      "epoch": 0.5859375,
      "grad_norm": 0.9584252834320068,
      "learning_rate": 2.963976517536627e-05,
      "loss": 0.8915,
      "step": 150
    },
    {
      "epoch": 0.60546875,
      "grad_norm": 0.8579109907150269,
      "learning_rate": 2.9596334799025003e-05,
      "loss": 0.8889,
      "step": 155
    },
    {
      "epoch": 0.625,
      "grad_norm": 0.8218862414360046,
      "learning_rate": 2.9550468797918162e-05,
      "loss": 0.855,
      "step": 160
    },
    {
      "epoch": 0.64453125,
      "grad_norm": 0.7801545858383179,
      "learning_rate": 2.950217482549915e-05,
      "loss": 0.8265,
      "step": 165
    },
    {
      "epoch": 0.6640625,
      "grad_norm": 0.803713858127594,
      "learning_rate": 2.9451460940366025e-05,
      "loss": 0.8222,
      "step": 170
    },
    {
      "epoch": 0.68359375,
      "grad_norm": 0.8898375034332275,
      "learning_rate": 2.9398335604916797e-05,
      "loss": 0.7957,
      "step": 175
    },
    {
      "epoch": 0.703125,
      "grad_norm": 1.0227468013763428,
      "learning_rate": 2.9342807683937352e-05,
      "loss": 0.8001,
      "step": 180
    },
    {
      "epoch": 0.72265625,
      "grad_norm": 0.8922404050827026,
      "learning_rate": 2.928488644312222e-05,
      "loss": 0.7994,
      "step": 185
    },
    {
      "epoch": 0.7421875,
      "grad_norm": 0.9021320939064026,
      "learning_rate": 2.9224581547528453e-05,
      "loss": 0.7911,
      "step": 190
    },
    {
      "epoch": 0.76171875,
      "grad_norm": 0.8039846420288086,
      "learning_rate": 2.916190305996286e-05,
      "loss": 0.779,
      "step": 195
    },
    {
      "epoch": 0.78125,
      "grad_norm": 0.977204442024231,
      "learning_rate": 2.909686143930287e-05,
      "loss": 0.7827,
      "step": 200
    },
    {
      "epoch": 0.80078125,
      "grad_norm": 0.8399918675422668,
      "learning_rate": 2.902946753875131e-05,
      "loss": 0.7524,
      "step": 205
    },
    {
      "epoch": 0.8203125,
      "grad_norm": 0.8906209468841553,
      "learning_rate": 2.895973260402537e-05,
      "loss": 0.7638,
      "step": 210
    },
    {
      "epoch": 0.83984375,
      "grad_norm": 1.2285107374191284,
      "learning_rate": 2.8887668271480098e-05,
      "loss": 0.7679,
      "step": 215
    },
    {
      "epoch": 0.859375,
      "grad_norm": 0.9521386027336121,
      "learning_rate": 2.8813286566166674e-05,
      "loss": 0.7347,
      "step": 220
    },
    {
      "epoch": 0.87890625,
      "grad_norm": 1.1255961656570435,
      "learning_rate": 2.873659989982586e-05,
      "loss": 0.7486,
      "step": 225
    },
    {
      "epoch": 0.8984375,
      "grad_norm": 0.9320322871208191,
      "learning_rate": 2.8657621068816903e-05,
      "loss": 0.7073,
      "step": 230
    },
    {
      "epoch": 0.91796875,
      "grad_norm": 1.0787020921707153,
      "learning_rate": 2.857636325198225e-05,
      "loss": 0.6956,
      "step": 235
    },
    {
      "epoch": 0.9375,
      "grad_norm": 0.9191898703575134,
      "learning_rate": 2.849284000844849e-05,
      "loss": 0.6815,
      "step": 240
    },
    {
      "epoch": 0.95703125,
      "grad_norm": 0.947468638420105,
      "learning_rate": 2.8407065275363756e-05,
      "loss": 0.6692,
      "step": 245
    },
    {
      "epoch": 0.9765625,
      "grad_norm": 0.8919875621795654,
      "learning_rate": 2.8319053365572135e-05,
      "loss": 0.688,
      "step": 250
    },
    {
      "epoch": 0.99609375,
      "grad_norm": 0.9745859503746033,
      "learning_rate": 2.8228818965225325e-05,
      "loss": 0.6103,
      "step": 255
    },
    {
      "epoch": 1.015625,
      "grad_norm": 1.2867947816848755,
      "learning_rate": 2.8136377131332043e-05,
      "loss": 0.6321,
      "step": 260
    },
    {
      "epoch": 1.03515625,
      "grad_norm": 1.055025577545166,
      "learning_rate": 2.8041743289245503e-05,
      "loss": 0.5343,
      "step": 265
    },
    {
      "epoch": 1.0546875,
      "grad_norm": 1.0192275047302246,
      "learning_rate": 2.7944933230089484e-05,
      "loss": 0.5867,
      "step": 270
    },
    {
      "epoch": 1.07421875,
      "grad_norm": 1.1829568147659302,
      "learning_rate": 2.784596310812331e-05,
      "loss": 0.5331,
      "step": 275
    },
    {
      "epoch": 1.09375,
      "grad_norm": 1.0421009063720703,
      "learning_rate": 2.774484943804629e-05,
      "loss": 0.5378,
      "step": 280
    },
    {
      "epoch": 1.11328125,
      "grad_norm": 1.039535641670227,
      "learning_rate": 2.764160909224196e-05,
      "loss": 0.5446,
      "step": 285
    },
    {
      "epoch": 1.1328125,
      "grad_norm": 1.2662440538406372,
      "learning_rate": 2.7536259297962674e-05,
      "loss": 0.4932,
      "step": 290
    },
    {
      "epoch": 1.15234375,
      "grad_norm": 1.0458778142929077,
      "learning_rate": 2.7428817634454973e-05,
      "loss": 0.5315,
      "step": 295
    },
    {
      "epoch": 1.171875,
      "grad_norm": 1.0862466096878052,
      "learning_rate": 2.7319302030026207e-05,
      "loss": 0.5626,
      "step": 300
    },
    {
      "epoch": 1.19140625,
      "grad_norm": 1.191480278968811,
      "learning_rate": 2.7207730759052925e-05,
      "loss": 0.6235,
      "step": 305
    },
    {
      "epoch": 1.2109375,
      "grad_norm": 1.3713099956512451,
      "learning_rate": 2.7094122438931513e-05,
      "loss": 0.4837,
      "step": 310
    },
    {
      "epoch": 1.23046875,
      "grad_norm": 1.104356288909912,
      "learning_rate": 2.697849602697159e-05,
      "loss": 0.5028,
      "step": 315
    },
    {
      "epoch": 1.25,
      "grad_norm": 1.2873727083206177,
      "learning_rate": 2.6860870817232682e-05,
      "loss": 0.5134,
      "step": 320
    },
    {
      "epoch": 1.26953125,
      "grad_norm": 1.0831677913665771,
      "learning_rate": 2.6741266437304718e-05,
      "loss": 0.475,
      "step": 325
    },
    {
      "epoch": 1.2890625,
      "grad_norm": 1.1841942071914673,
      "learning_rate": 2.661970284503286e-05,
      "loss": 0.5094,
      "step": 330
    },
    {
      "epoch": 1.30859375,
      "grad_norm": 1.1621878147125244,
      "learning_rate": 2.6496200325187223e-05,
      "loss": 0.5235,
      "step": 335
    },
    {
      "epoch": 1.328125,
      "grad_norm": 1.1662148237228394,
      "learning_rate": 2.6370779486078047e-05,
      "loss": 0.5086,
      "step": 340
    },
    {
      "epoch": 1.34765625,
      "grad_norm": 1.1327015161514282,
      "learning_rate": 2.6243461256116892e-05,
      "loss": 0.5031,
      "step": 345
    },
    {
      "epoch": 1.3671875,
      "grad_norm": 1.211604118347168,
      "learning_rate": 2.611426688032439e-05,
      "loss": 0.4841,
      "step": 350
    },
    {
      "epoch": 1.38671875,
      "grad_norm": 1.0225025415420532,
      "learning_rate": 2.598321791678519e-05,
      "loss": 0.4889,
      "step": 355
    },
    {
      "epoch": 1.40625,
      "grad_norm": 1.1432801485061646,
      "learning_rate": 2.5850336233050677e-05,
      "loss": 0.4618,
      "step": 360
    },
    {
      "epoch": 1.42578125,
      "grad_norm": 1.3656413555145264,
      "learning_rate": 2.5715644002489998e-05,
      "loss": 0.4594,
      "step": 365
    },
    {
      "epoch": 1.4453125,
      "grad_norm": 1.383625864982605,
      "learning_rate": 2.557916370059012e-05,
      "loss": 0.5009,
      "step": 370
    },
    {
      "epoch": 1.46484375,
      "grad_norm": 1.2634763717651367,
      "learning_rate": 2.544091810120543e-05,
      "loss": 0.5046,
      "step": 375
    },
    {
      "epoch": 1.484375,
      "grad_norm": 1.156620979309082,
      "learning_rate": 2.530093027275757e-05,
      "loss": 0.4219,
      "step": 380
    },
    {
      "epoch": 1.50390625,
      "grad_norm": 1.0806770324707031,
      "learning_rate": 2.5159223574386117e-05,
      "loss": 0.4786,
      "step": 385
    },
    {
      "epoch": 1.5234375,
      "grad_norm": 1.1003456115722656,
      "learning_rate": 2.501582165205074e-05,
      "loss": 0.4232,
      "step": 390
    },
    {
      "epoch": 1.54296875,
      "grad_norm": 1.1384750604629517,
      "learning_rate": 2.4870748434585514e-05,
      "loss": 0.499,
      "step": 395
    },
    {
      "epoch": 1.5625,
      "grad_norm": 1.0580660104751587,
      "learning_rate": 2.4724028129706012e-05,
      "loss": 0.4066,
      "step": 400
    },
    {
      "epoch": 1.58203125,
      "grad_norm": 1.2567460536956787,
      "learning_rate": 2.4575685219969884e-05,
      "loss": 0.4367,
      "step": 405
    },
    {
      "epoch": 1.6015625,
      "grad_norm": 1.1812260150909424,
      "learning_rate": 2.442574445869156e-05,
      "loss": 0.4303,
      "step": 410
    },
    {
      "epoch": 1.62109375,
      "grad_norm": 1.2039210796356201,
      "learning_rate": 2.4274230865811763e-05,
      "loss": 0.4385,
      "step": 415
    },
    {
      "epoch": 1.640625,
      "grad_norm": 1.2550849914550781,
      "learning_rate": 2.4121169723722566e-05,
      "loss": 0.3703,
      "step": 420
    },
    {
      "epoch": 1.66015625,
      "grad_norm": 1.1235885620117188,
      "learning_rate": 2.396658657304861e-05,
      "loss": 0.4242,
      "step": 425
    },
    {
      "epoch": 1.6796875,
      "grad_norm": 1.2263010740280151,
      "learning_rate": 2.381050720838528e-05,
      "loss": 0.3919,
      "step": 430
    },
    {
      "epoch": 1.69921875,
      "grad_norm": 1.1357097625732422,
      "learning_rate": 2.3652957673994448e-05,
      "loss": 0.4487,
      "step": 435
    },
    {
      "epoch": 1.71875,
      "grad_norm": 1.271737813949585,
      "learning_rate": 2.3493964259458603e-05,
      "loss": 0.3659,
      "step": 440
    },
    {
      "epoch": 1.73828125,
      "grad_norm": 1.257322072982788,
      "learning_rate": 2.3333553495294033e-05,
      "loss": 0.3847,
      "step": 445
    },
    {
      "epoch": 1.7578125,
      "grad_norm": 1.1676826477050781,
      "learning_rate": 2.317175214852377e-05,
      "loss": 0.3531,
      "step": 450
    },
    {
      "epoch": 1.77734375,
      "grad_norm": 1.032481074333191,
      "learning_rate": 2.3008587218211127e-05,
      "loss": 0.3849,
      "step": 455
    },
    {
      "epoch": 1.796875,
      "grad_norm": 1.195465087890625,
      "learning_rate": 2.284408593095446e-05,
      "loss": 0.4086,
      "step": 460
    },
    {
      "epoch": 1.81640625,
      "grad_norm": 1.3418141603469849,
      "learning_rate": 2.2678275736344014e-05,
      "loss": 0.3851,
      "step": 465
    },
    {
      "epoch": 1.8359375,
      "grad_norm": 1.1920610666275024,
      "learning_rate": 2.251118430238151e-05,
      "loss": 0.3678,
      "step": 470
    },
    {
      "epoch": 1.85546875,
      "grad_norm": 1.2250503301620483,
      "learning_rate": 2.2342839510863323e-05,
      "loss": 0.367,
      "step": 475
    },
    {
      "epoch": 1.875,
      "grad_norm": 1.1807481050491333,
      "learning_rate": 2.2173269452727965e-05,
      "loss": 0.3344,
      "step": 480
    },
    {
      "epoch": 1.89453125,
      "grad_norm": 1.2924655675888062,
      "learning_rate": 2.200250242336868e-05,
      "loss": 0.338,
      "step": 485
    },
    {
      "epoch": 1.9140625,
      "grad_norm": 1.1538184881210327,
      "learning_rate": 2.183056691791193e-05,
      "loss": 0.3683,
      "step": 490
    },
    {
      "epoch": 1.93359375,
      "grad_norm": 1.335293173789978,
      "learning_rate": 2.1657491626462514e-05,
      "loss": 0.3779,
      "step": 495
    },
    {
      "epoch": 1.953125,
      "grad_norm": 1.2173998355865479,
      "learning_rate": 2.1483305429316208e-05,
      "loss": 0.3728,
      "step": 500
    },
    {
      "epoch": 1.97265625,
      "grad_norm": 1.1745744943618774,
      "learning_rate": 2.1308037392140613e-05,
      "loss": 0.3268,
      "step": 505
    },
    {
      "epoch": 1.9921875,
      "grad_norm": 1.1856087446212769,
      "learning_rate": 2.113171676112513e-05,
      "loss": 0.3501,
      "step": 510
    }
  ],
  "logging_steps": 5,
  "max_steps": 1280,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 2000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 7.521082542632468e+17,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}