{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 468,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0106951871657754,
      "grad_norm": 5.869461536407471,
      "learning_rate": 9.30232558139535e-07,
      "loss": 0.3054,
      "step": 5
    },
    {
      "epoch": 0.0213903743315508,
      "grad_norm": 4.7690653800964355,
      "learning_rate": 2.0930232558139536e-06,
      "loss": 0.2661,
      "step": 10
    },
    {
      "epoch": 0.03208556149732621,
      "grad_norm": 4.248436450958252,
      "learning_rate": 3.2558139534883724e-06,
      "loss": 0.2409,
      "step": 15
    },
    {
      "epoch": 0.0427807486631016,
      "grad_norm": 4.08921480178833,
      "learning_rate": 4.418604651162791e-06,
      "loss": 0.2228,
      "step": 20
    },
    {
      "epoch": 0.053475935828877004,
      "grad_norm": 4.846882343292236,
      "learning_rate": 5.58139534883721e-06,
      "loss": 0.2258,
      "step": 25
    },
    {
      "epoch": 0.06417112299465241,
      "grad_norm": 4.230620384216309,
      "learning_rate": 6.744186046511628e-06,
      "loss": 0.2103,
      "step": 30
    },
    {
      "epoch": 0.0748663101604278,
      "grad_norm": 4.4064130783081055,
      "learning_rate": 7.906976744186048e-06,
      "loss": 0.2122,
      "step": 35
    },
    {
      "epoch": 0.0855614973262032,
      "grad_norm": 4.86531400680542,
      "learning_rate": 9.069767441860465e-06,
      "loss": 0.1996,
      "step": 40
    },
    {
      "epoch": 0.0962566844919786,
      "grad_norm": 5.53223991394043,
      "learning_rate": 9.999986679414613e-06,
      "loss": 0.2311,
      "step": 45
    },
    {
      "epoch": 0.10695187165775401,
      "grad_norm": 4.2916178703308105,
      "learning_rate": 9.999520466378376e-06,
      "loss": 0.2319,
      "step": 50
    },
    {
      "epoch": 0.11764705882352941,
      "grad_norm": 4.200897216796875,
      "learning_rate": 9.998388295046227e-06,
      "loss": 0.2368,
      "step": 55
    },
    {
      "epoch": 0.12834224598930483,
      "grad_norm": 4.808861255645752,
      "learning_rate": 9.996590316228402e-06,
      "loss": 0.2354,
      "step": 60
    },
    {
      "epoch": 0.13903743315508021,
      "grad_norm": 4.5134077072143555,
      "learning_rate": 9.994126769423656e-06,
      "loss": 0.234,
      "step": 65
    },
    {
      "epoch": 0.1497326203208556,
      "grad_norm": 4.348900318145752,
      "learning_rate": 9.990997982787348e-06,
      "loss": 0.2541,
      "step": 70
    },
    {
      "epoch": 0.16042780748663102,
      "grad_norm": 4.4039306640625,
      "learning_rate": 9.98720437308773e-06,
      "loss": 0.2496,
      "step": 75
    },
    {
      "epoch": 0.1711229946524064,
      "grad_norm": 4.726532936096191,
      "learning_rate": 9.982746445650437e-06,
      "loss": 0.2797,
      "step": 80
    },
    {
      "epoch": 0.18181818181818182,
      "grad_norm": 4.256592273712158,
      "learning_rate": 9.977624794291172e-06,
      "loss": 0.2518,
      "step": 85
    },
    {
      "epoch": 0.1925133689839572,
      "grad_norm": 4.330168724060059,
      "learning_rate": 9.97184010123661e-06,
      "loss": 0.261,
      "step": 90
    },
    {
      "epoch": 0.20320855614973263,
      "grad_norm": 4.281119346618652,
      "learning_rate": 9.965393137033512e-06,
      "loss": 0.2564,
      "step": 95
    },
    {
      "epoch": 0.21390374331550802,
      "grad_norm": 4.247673511505127,
      "learning_rate": 9.958284760446104e-06,
      "loss": 0.2501,
      "step": 100
    },
    {
      "epoch": 0.22459893048128343,
      "grad_norm": 3.930504322052002,
      "learning_rate": 9.950515918341666e-06,
      "loss": 0.2504,
      "step": 105
    },
    {
      "epoch": 0.23529411764705882,
      "grad_norm": 4.084280967712402,
      "learning_rate": 9.942087645564415e-06,
      "loss": 0.2706,
      "step": 110
    },
    {
      "epoch": 0.24598930481283424,
      "grad_norm": 3.869400978088379,
      "learning_rate": 9.93300106479766e-06,
      "loss": 0.2499,
      "step": 115
    },
    {
      "epoch": 0.25668449197860965,
      "grad_norm": 4.0887675285339355,
      "learning_rate": 9.923257386414253e-06,
      "loss": 0.2628,
      "step": 120
    },
    {
      "epoch": 0.26737967914438504,
      "grad_norm": 4.140570163726807,
      "learning_rate": 9.912857908315363e-06,
      "loss": 0.2642,
      "step": 125
    },
    {
      "epoch": 0.27807486631016043,
      "grad_norm": 4.731294631958008,
      "learning_rate": 9.901804015757588e-06,
      "loss": 0.2669,
      "step": 130
    },
    {
      "epoch": 0.2887700534759358,
      "grad_norm": 4.113650321960449,
      "learning_rate": 9.89009718116843e-06,
      "loss": 0.2809,
      "step": 135
    },
    {
      "epoch": 0.2994652406417112,
      "grad_norm": 3.8770499229431152,
      "learning_rate": 9.877738963950175e-06,
      "loss": 0.2705,
      "step": 140
    },
    {
      "epoch": 0.31016042780748665,
      "grad_norm": 4.077112197875977,
      "learning_rate": 9.864731010272152e-06,
      "loss": 0.2548,
      "step": 145
    },
    {
      "epoch": 0.32085561497326204,
      "grad_norm": 4.771427631378174,
      "learning_rate": 9.851075052851476e-06,
      "loss": 0.2534,
      "step": 150
    },
    {
      "epoch": 0.3315508021390374,
      "grad_norm": 3.84387469291687,
      "learning_rate": 9.83677291072223e-06,
      "loss": 0.2438,
      "step": 155
    },
    {
      "epoch": 0.3422459893048128,
      "grad_norm": 4.427935600280762,
      "learning_rate": 9.821826488993168e-06,
      "loss": 0.2791,
      "step": 160
    },
    {
      "epoch": 0.35294117647058826,
      "grad_norm": 4.1868767738342285,
      "learning_rate": 9.806237778593941e-06,
      "loss": 0.2734,
      "step": 165
    },
    {
      "epoch": 0.36363636363636365,
      "grad_norm": 3.8055779933929443,
      "learning_rate": 9.790008856009902e-06,
      "loss": 0.2463,
      "step": 170
    },
    {
      "epoch": 0.37433155080213903,
      "grad_norm": 4.041962623596191,
      "learning_rate": 9.773141883005507e-06,
      "loss": 0.2813,
      "step": 175
    },
    {
      "epoch": 0.3850267379679144,
      "grad_norm": 3.6363329887390137,
      "learning_rate": 9.755639106336347e-06,
      "loss": 0.2724,
      "step": 180
    },
    {
      "epoch": 0.39572192513368987,
      "grad_norm": 3.50486421585083,
      "learning_rate": 9.737502857449894e-06,
      "loss": 0.271,
      "step": 185
    },
    {
      "epoch": 0.40641711229946526,
      "grad_norm": 3.4967920780181885,
      "learning_rate": 9.718735552174923e-06,
      "loss": 0.2467,
      "step": 190
    },
    {
      "epoch": 0.41711229946524064,
      "grad_norm": 3.3797831535339355,
      "learning_rate": 9.699339690399717e-06,
      "loss": 0.2621,
      "step": 195
    },
    {
      "epoch": 0.42780748663101603,
      "grad_norm": 15.109771728515625,
      "learning_rate": 9.679317855739073e-06,
      "loss": 0.2791,
      "step": 200
    },
    {
      "epoch": 0.4385026737967914,
      "grad_norm": 3.828474760055542,
      "learning_rate": 9.658672715190151e-06,
      "loss": 0.2829,
      "step": 205
    },
    {
      "epoch": 0.44919786096256686,
      "grad_norm": 3.2750980854034424,
      "learning_rate": 9.637407018777224e-06,
      "loss": 0.2866,
      "step": 210
    },
    {
      "epoch": 0.45989304812834225,
      "grad_norm": 3.615290880203247,
      "learning_rate": 9.615523599185353e-06,
      "loss": 0.2422,
      "step": 215
    },
    {
      "epoch": 0.47058823529411764,
      "grad_norm": 3.5919339656829834,
      "learning_rate": 9.593025371383064e-06,
      "loss": 0.269,
      "step": 220
    },
    {
      "epoch": 0.48128342245989303,
      "grad_norm": 4.708985805511475,
      "learning_rate": 9.569915332234068e-06,
      "loss": 0.2686,
      "step": 225
    },
    {
      "epoch": 0.4919786096256685,
      "grad_norm": 3.5988545417785645,
      "learning_rate": 9.546196560098062e-06,
      "loss": 0.273,
      "step": 230
    },
    {
      "epoch": 0.5026737967914439,
      "grad_norm": 3.4772067070007324,
      "learning_rate": 9.521872214420668e-06,
      "loss": 0.2395,
      "step": 235
    },
    {
      "epoch": 0.5133689839572193,
      "grad_norm": 3.7974278926849365,
      "learning_rate": 9.496945535312597e-06,
      "loss": 0.2682,
      "step": 240
    },
    {
      "epoch": 0.5240641711229946,
      "grad_norm": 3.69970440864563,
      "learning_rate": 9.471419843118036e-06,
      "loss": 0.2557,
      "step": 245
    },
    {
      "epoch": 0.5347593582887701,
      "grad_norm": 3.470737934112549,
      "learning_rate": 9.44529853797238e-06,
      "loss": 0.266,
      "step": 250
    },
    {
      "epoch": 0.5454545454545454,
      "grad_norm": 3.733236312866211,
      "learning_rate": 9.418585099349306e-06,
      "loss": 0.2779,
      "step": 255
    },
    {
      "epoch": 0.5561497326203209,
      "grad_norm": 3.434349298477173,
      "learning_rate": 9.391283085597299e-06,
      "loss": 0.2508,
      "step": 260
    },
    {
      "epoch": 0.5668449197860963,
      "grad_norm": 3.5163183212280273,
      "learning_rate": 9.36339613346565e-06,
      "loss": 0.2727,
      "step": 265
    },
    {
      "epoch": 0.5775401069518716,
      "grad_norm": 4.07936429977417,
      "learning_rate": 9.33492795762005e-06,
      "loss": 0.2714,
      "step": 270
    },
    {
      "epoch": 0.5882352941176471,
      "grad_norm": 3.171149492263794,
      "learning_rate": 9.305882350147763e-06,
      "loss": 0.2585,
      "step": 275
    },
    {
      "epoch": 0.5989304812834224,
      "grad_norm": 3.424302101135254,
      "learning_rate": 9.276263180052498e-06,
      "loss": 0.2584,
      "step": 280
    },
    {
      "epoch": 0.6096256684491979,
      "grad_norm": 3.800421714782715,
      "learning_rate": 9.246074392739057e-06,
      "loss": 0.2771,
      "step": 285
    },
    {
      "epoch": 0.6203208556149733,
      "grad_norm": 3.4711737632751465,
      "learning_rate": 9.21532000948778e-06,
      "loss": 0.265,
      "step": 290
    },
    {
      "epoch": 0.6310160427807486,
      "grad_norm": 3.5598926544189453,
      "learning_rate": 9.184004126918891e-06,
      "loss": 0.2653,
      "step": 295
    },
    {
      "epoch": 0.6417112299465241,
      "grad_norm": 3.320875644683838,
      "learning_rate": 9.152130916446817e-06,
      "loss": 0.2373,
      "step": 300
    },
    {
      "epoch": 0.6524064171122995,
      "grad_norm": 3.606067419052124,
      "learning_rate": 9.119704623724528e-06,
      "loss": 0.2779,
      "step": 305
    },
    {
      "epoch": 0.6631016042780749,
      "grad_norm": 3.189316511154175,
      "learning_rate": 9.086729568078006e-06,
      "loss": 0.2813,
      "step": 310
    },
    {
      "epoch": 0.6737967914438503,
      "grad_norm": 3.725550651550293,
      "learning_rate": 9.05321014193089e-06,
      "loss": 0.284,
      "step": 315
    },
    {
      "epoch": 0.6844919786096256,
      "grad_norm": 3.6915953159332275,
      "learning_rate": 9.019150810219376e-06,
      "loss": 0.2646,
      "step": 320
    },
    {
      "epoch": 0.6951871657754011,
      "grad_norm": 3.499095916748047,
      "learning_rate": 8.984556109797484e-06,
      "loss": 0.2693,
      "step": 325
    },
    {
      "epoch": 0.7058823529411765,
      "grad_norm": 3.671689510345459,
      "learning_rate": 8.949430648832716e-06,
      "loss": 0.2662,
      "step": 330
    },
    {
      "epoch": 0.7165775401069518,
      "grad_norm": 3.549403429031372,
      "learning_rate": 8.91377910619223e-06,
      "loss": 0.2745,
      "step": 335
    },
    {
      "epoch": 0.7272727272727273,
      "grad_norm": 3.2107975482940674,
      "learning_rate": 8.8776062308196e-06,
      "loss": 0.2462,
      "step": 340
    },
    {
      "epoch": 0.7379679144385026,
      "grad_norm": 3.040982246398926,
      "learning_rate": 8.84091684110223e-06,
      "loss": 0.2663,
      "step": 345
    },
    {
      "epoch": 0.7486631016042781,
      "grad_norm": 3.460141658782959,
      "learning_rate": 8.803715824229525e-06,
      "loss": 0.2595,
      "step": 350
    },
    {
      "epoch": 0.7593582887700535,
      "grad_norm": 3.5115394592285156,
      "learning_rate": 8.766008135541896e-06,
      "loss": 0.2476,
      "step": 355
    },
    {
      "epoch": 0.7700534759358288,
      "grad_norm": 3.2345032691955566,
      "learning_rate": 8.727798797870688e-06,
      "loss": 0.2399,
      "step": 360
    },
    {
      "epoch": 0.7807486631016043,
      "grad_norm": 3.489104986190796,
      "learning_rate": 8.689092900869112e-06,
      "loss": 0.2568,
      "step": 365
    },
    {
      "epoch": 0.7914438502673797,
      "grad_norm": 3.4822070598602295,
      "learning_rate": 8.649895600334284e-06,
      "loss": 0.2779,
      "step": 370
    },
    {
      "epoch": 0.8021390374331551,
      "grad_norm": 3.6014904975891113,
      "learning_rate": 8.610212117520453e-06,
      "loss": 0.2738,
      "step": 375
    },
    {
      "epoch": 0.8128342245989305,
      "grad_norm": 3.3353493213653564,
      "learning_rate": 8.570047738443502e-06,
      "loss": 0.2564,
      "step": 380
    },
    {
      "epoch": 0.8235294117647058,
      "grad_norm": 3.1294198036193848,
      "learning_rate": 8.52940781317683e-06,
      "loss": 0.2459,
      "step": 385
    },
    {
      "epoch": 0.8342245989304813,
      "grad_norm": 3.2122206687927246,
      "learning_rate": 8.48829775513869e-06,
      "loss": 0.2639,
      "step": 390
    },
    {
      "epoch": 0.8449197860962567,
      "grad_norm": 3.2475640773773193,
      "learning_rate": 8.446723040371114e-06,
      "loss": 0.2545,
      "step": 395
    },
    {
      "epoch": 0.8556149732620321,
      "grad_norm": 3.113609552383423,
      "learning_rate": 8.40468920681047e-06,
      "loss": 0.2712,
      "step": 400
    },
    {
      "epoch": 0.8663101604278075,
      "grad_norm": 3.231935501098633,
      "learning_rate": 8.362201853549777e-06,
      "loss": 0.27,
      "step": 405
    },
    {
      "epoch": 0.8770053475935828,
      "grad_norm": 3.3826744556427,
      "learning_rate": 8.319266640092899e-06,
      "loss": 0.2823,
      "step": 410
    },
    {
      "epoch": 0.8877005347593583,
      "grad_norm": 3.3691983222961426,
      "learning_rate": 8.275889285600656e-06,
      "loss": 0.2555,
      "step": 415
    },
    {
      "epoch": 0.8983957219251337,
      "grad_norm": 3.4460556507110596,
      "learning_rate": 8.23207556812902e-06,
      "loss": 0.2615,
      "step": 420
    },
    {
      "epoch": 0.9090909090909091,
      "grad_norm": 3.0713820457458496,
      "learning_rate": 8.187831323859445e-06,
      "loss": 0.2512,
      "step": 425
    },
    {
      "epoch": 0.9197860962566845,
      "grad_norm": 2.781736373901367,
      "learning_rate": 8.143162446321465e-06,
      "loss": 0.2568,
      "step": 430
    },
    {
      "epoch": 0.93048128342246,
      "grad_norm": 3.0802292823791504,
      "learning_rate": 8.098074885607646e-06,
      "loss": 0.2506,
      "step": 435
    },
    {
      "epoch": 0.9411764705882353,
      "grad_norm": 3.0498411655426025,
      "learning_rate": 8.052574647581009e-06,
      "loss": 0.2586,
      "step": 440
    },
    {
      "epoch": 0.9518716577540107,
      "grad_norm": 3.0598437786102295,
      "learning_rate": 8.006667793075026e-06,
      "loss": 0.2576,
      "step": 445
    },
    {
      "epoch": 0.9625668449197861,
      "grad_norm": 2.935920238494873,
      "learning_rate": 7.960360437086287e-06,
      "loss": 0.2363,
      "step": 450
    },
    {
      "epoch": 0.9732620320855615,
      "grad_norm": 3.2348110675811768,
      "learning_rate": 7.91365874795995e-06,
      "loss": 0.2677,
      "step": 455
    },
    {
      "epoch": 0.983957219251337,
      "grad_norm": 3.4138996601104736,
      "learning_rate": 7.866568946568107e-06,
      "loss": 0.2702,
      "step": 460
    },
    {
      "epoch": 0.9946524064171123,
      "grad_norm": 2.949201822280884,
      "learning_rate": 7.819097305481112e-06,
      "loss": 0.2487,
      "step": 465
    }
  ],
  "logging_steps": 5,
  "max_steps": 1404,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 3.619093694675681e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}