{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.1,
  "eval_steps": 500,
  "global_step": 1000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "grad_norm": 1.3495701551437378,
      "learning_rate": 1.8e-06,
      "loss": 1.1757,
      "step": 10
    },
    {
      "grad_norm": 0.7409749031066895,
      "learning_rate": 3.8e-06,
      "loss": 1.1588,
      "step": 20
    },
    {
      "grad_norm": 0.5010238885879517,
      "learning_rate": 5.8e-06,
      "loss": 1.118,
      "step": 30
    },
    {
      "grad_norm": 0.38745883107185364,
      "learning_rate": 7.8e-06,
      "loss": 1.071,
      "step": 40
    },
    {
      "grad_norm": 0.2737957239151001,
      "learning_rate": 9.800000000000001e-06,
      "loss": 1.0505,
      "step": 50
    },
    {
      "grad_norm": 0.17909298837184906,
      "learning_rate": 1.18e-05,
      "loss": 1.0511,
      "step": 60
    },
    {
      "grad_norm": 0.15505068004131317,
      "learning_rate": 1.3800000000000002e-05,
      "loss": 1.0386,
      "step": 70
    },
    {
      "grad_norm": 0.19511660933494568,
      "learning_rate": 1.58e-05,
      "loss": 1.0291,
      "step": 80
    },
    {
      "grad_norm": 0.17100006341934204,
      "learning_rate": 1.78e-05,
      "loss": 1.0237,
      "step": 90
    },
    {
      "grad_norm": 0.1620642989873886,
      "learning_rate": 1.9800000000000004e-05,
      "loss": 1.0225,
      "step": 100
    },
    {
      "grad_norm": 0.1624162495136261,
      "learning_rate": 2.18e-05,
      "loss": 1.0158,
      "step": 110
    },
    {
      "grad_norm": 0.6077130436897278,
      "learning_rate": 2.38e-05,
      "loss": 0.9931,
      "step": 120
    },
    {
      "grad_norm": 0.43021947145462036,
      "learning_rate": 2.58e-05,
      "loss": 0.9595,
      "step": 130
    },
    {
      "grad_norm": 0.5613481402397156,
      "learning_rate": 2.7800000000000005e-05,
      "loss": 0.9041,
      "step": 140
    },
    {
      "grad_norm": 0.6526139378547668,
      "learning_rate": 2.98e-05,
      "loss": 0.8396,
      "step": 150
    },
    {
      "grad_norm": 0.7447936534881592,
      "learning_rate": 3.18e-05,
      "loss": 0.7885,
      "step": 160
    },
    {
      "grad_norm": 0.7026267051696777,
      "learning_rate": 3.38e-05,
      "loss": 0.7439,
      "step": 170
    },
    {
      "grad_norm": 0.7374712824821472,
      "learning_rate": 3.58e-05,
      "loss": 0.6911,
      "step": 180
    },
    {
      "grad_norm": 0.7992081642150879,
      "learning_rate": 3.7800000000000004e-05,
      "loss": 0.6503,
      "step": 190
    },
    {
      "grad_norm": 1.095856785774231,
      "learning_rate": 3.9800000000000005e-05,
      "loss": 0.6026,
      "step": 200
    },
    {
      "grad_norm": 1.0275039672851562,
      "learning_rate": 4.18e-05,
      "loss": 0.5638,
      "step": 210
    },
    {
      "grad_norm": 0.8001906871795654,
      "learning_rate": 4.38e-05,
      "loss": 0.5221,
      "step": 220
    },
    {
      "grad_norm": 1.0533819198608398,
      "learning_rate": 4.58e-05,
      "loss": 0.4761,
      "step": 230
    },
    {
      "grad_norm": 0.9747286438941956,
      "learning_rate": 4.78e-05,
      "loss": 0.4439,
      "step": 240
    },
    {
      "grad_norm": 0.8441274762153625,
      "learning_rate": 4.9800000000000004e-05,
      "loss": 0.412,
      "step": 250
    },
    {
      "grad_norm": 1.0516191720962524,
      "learning_rate": 5.1800000000000005e-05,
      "loss": 0.3905,
      "step": 260
    },
    {
      "grad_norm": 1.1816160678863525,
      "learning_rate": 5.380000000000001e-05,
      "loss": 0.3687,
      "step": 270
    },
    {
      "grad_norm": 1.1618976593017578,
      "learning_rate": 5.580000000000001e-05,
      "loss": 0.3411,
      "step": 280
    },
    {
      "grad_norm": 1.151973843574524,
      "learning_rate": 5.7799999999999995e-05,
      "loss": 0.3124,
      "step": 290
    },
    {
      "grad_norm": 1.1495264768600464,
      "learning_rate": 5.9800000000000003e-05,
      "loss": 0.2993,
      "step": 300
    },
    {
      "grad_norm": 1.0407155752182007,
      "learning_rate": 6.18e-05,
      "loss": 0.2866,
      "step": 310
    },
    {
      "grad_norm": 0.9749088883399963,
      "learning_rate": 6.38e-05,
      "loss": 0.2567,
      "step": 320
    },
    {
      "grad_norm": 1.688358187675476,
      "learning_rate": 6.58e-05,
      "loss": 0.2514,
      "step": 330
    },
    {
      "grad_norm": 1.1689329147338867,
      "learning_rate": 6.780000000000001e-05,
      "loss": 0.2367,
      "step": 340
    },
    {
      "grad_norm": 1.1307363510131836,
      "learning_rate": 6.98e-05,
      "loss": 0.2104,
      "step": 350
    },
    {
      "grad_norm": 1.1584901809692383,
      "learning_rate": 7.18e-05,
      "loss": 0.2008,
      "step": 360
    },
    {
      "grad_norm": 1.0391298532485962,
      "learning_rate": 7.38e-05,
      "loss": 0.1792,
      "step": 370
    },
    {
      "grad_norm": 0.9876890182495117,
      "learning_rate": 7.58e-05,
      "loss": 0.1607,
      "step": 380
    },
    {
      "grad_norm": 0.9350799322128296,
      "learning_rate": 7.780000000000001e-05,
      "loss": 0.1535,
      "step": 390
    },
    {
      "grad_norm": 1.1527447700500488,
      "learning_rate": 7.98e-05,
      "loss": 0.1479,
      "step": 400
    },
    {
      "grad_norm": 1.156333327293396,
      "learning_rate": 8.18e-05,
      "loss": 0.1359,
      "step": 410
    },
    {
      "grad_norm": 0.9062078595161438,
      "learning_rate": 8.38e-05,
      "loss": 0.1159,
      "step": 420
    },
    {
      "grad_norm": 1.0753194093704224,
      "learning_rate": 8.58e-05,
      "loss": 0.1107,
      "step": 430
    },
    {
      "grad_norm": 1.0954129695892334,
      "learning_rate": 8.78e-05,
      "loss": 0.0974,
      "step": 440
    },
    {
      "grad_norm": 0.8616589903831482,
      "learning_rate": 8.98e-05,
      "loss": 0.0986,
      "step": 450
    },
    {
      "grad_norm": 1.0874602794647217,
      "learning_rate": 9.180000000000001e-05,
      "loss": 0.0956,
      "step": 460
    },
    {
      "grad_norm": 0.8882458209991455,
      "learning_rate": 9.38e-05,
      "loss": 0.0896,
      "step": 470
    },
    {
      "grad_norm": 1.0427426099777222,
      "learning_rate": 9.58e-05,
      "loss": 0.087,
      "step": 480
    },
    {
      "grad_norm": 1.2230194807052612,
      "learning_rate": 9.78e-05,
      "loss": 0.0768,
      "step": 490
    },
    {
      "grad_norm": 0.847732663154602,
      "learning_rate": 9.98e-05,
      "loss": 0.0704,
      "step": 500
    },
    {
      "grad_norm": 0.9105225801467896,
      "learning_rate": 9.9999778549206e-05,
      "loss": 0.0626,
      "step": 510
    },
    {
      "grad_norm": 1.1461974382400513,
      "learning_rate": 9.999901304280685e-05,
      "loss": 0.0607,
      "step": 520
    },
    {
      "grad_norm": 0.9256789684295654,
      "learning_rate": 9.999770075521164e-05,
      "loss": 0.0596,
      "step": 530
    },
    {
      "grad_norm": 0.929280161857605,
      "learning_rate": 9.99958417007713e-05,
      "loss": 0.0615,
      "step": 540
    },
    {
      "grad_norm": 0.8615525364875793,
      "learning_rate": 9.999343589981615e-05,
      "loss": 0.0506,
      "step": 550
    },
    {
      "grad_norm": 0.6592100262641907,
      "learning_rate": 9.999048337865568e-05,
      "loss": 0.057,
      "step": 560
    },
    {
      "grad_norm": 1.0078480243682861,
      "learning_rate": 9.998698416957815e-05,
      "loss": 0.0622,
      "step": 570
    },
    {
      "grad_norm": 0.9626633524894714,
      "learning_rate": 9.998293831085037e-05,
      "loss": 0.0583,
      "step": 580
    },
    {
      "grad_norm": 0.8204993009567261,
      "learning_rate": 9.997834584671719e-05,
      "loss": 0.0549,
      "step": 590
    },
    {
      "grad_norm": 1.0736879110336304,
      "learning_rate": 9.997320682740107e-05,
      "loss": 0.0596,
      "step": 600
    },
    {
      "grad_norm": 0.5788105726242065,
      "learning_rate": 9.996752130910149e-05,
      "loss": 0.0537,
      "step": 610
    },
    {
      "grad_norm": 0.9862315654754639,
      "learning_rate": 9.99612893539944e-05,
      "loss": 0.0523,
      "step": 620
    },
    {
      "grad_norm": 0.8640528917312622,
      "learning_rate": 9.995451103023144e-05,
      "loss": 0.052,
      "step": 630
    },
    {
      "grad_norm": 0.7985650897026062,
      "learning_rate": 9.994718641193928e-05,
      "loss": 0.0519,
      "step": 640
    },
    {
      "grad_norm": 0.6485480070114136,
      "learning_rate": 9.993931557921874e-05,
      "loss": 0.053,
      "step": 650
    },
    {
      "grad_norm": 0.944926381111145,
      "learning_rate": 9.993089861814402e-05,
      "loss": 0.0512,
      "step": 660
    },
    {
      "grad_norm": 0.9629592299461365,
      "learning_rate": 9.992193562076166e-05,
      "loss": 0.0499,
      "step": 670
    },
    {
      "grad_norm": 0.8933172821998596,
      "learning_rate": 9.991242668508954e-05,
      "loss": 0.0454,
      "step": 680
    },
    {
      "grad_norm": 0.8176608681678772,
      "learning_rate": 9.990237191511587e-05,
      "loss": 0.042,
      "step": 690
    },
    {
      "grad_norm": 0.8295159935951233,
      "learning_rate": 9.989177142079802e-05,
      "loss": 0.044,
      "step": 700
    },
    {
      "grad_norm": 1.0152912139892578,
      "learning_rate": 9.988062531806126e-05,
      "loss": 0.0453,
      "step": 710
    },
    {
      "grad_norm": 0.9435839056968689,
      "learning_rate": 9.986893372879762e-05,
      "loss": 0.0473,
      "step": 720
    },
    {
      "grad_norm": 0.846592903137207,
      "learning_rate": 9.985669678086443e-05,
      "loss": 0.0475,
      "step": 730
    },
    {
      "grad_norm": 0.7264214754104614,
      "learning_rate": 9.984391460808298e-05,
      "loss": 0.0498,
      "step": 740
    },
    {
      "grad_norm": 0.7152071595191956,
      "learning_rate": 9.983058735023709e-05,
      "loss": 0.0493,
      "step": 750
    },
    {
      "grad_norm": 0.7937959432601929,
      "learning_rate": 9.98167151530715e-05,
      "loss": 0.0445,
      "step": 760
    },
    {
      "grad_norm": 0.7876198887825012,
      "learning_rate": 9.980229816829034e-05,
      "loss": 0.0486,
      "step": 770
    },
    {
      "grad_norm": 0.8420520424842834,
      "learning_rate": 9.978733655355544e-05,
      "loss": 0.0465,
      "step": 780
    },
    {
      "grad_norm": 0.7807585597038269,
      "learning_rate": 9.977183047248464e-05,
      "loss": 0.0415,
      "step": 790
    },
    {
      "grad_norm": 0.8610442280769348,
      "learning_rate": 9.975578009464992e-05,
      "loss": 0.0437,
      "step": 800
    },
    {
      "grad_norm": 0.844108521938324,
      "learning_rate": 9.97391855955757e-05,
      "loss": 0.0424,
      "step": 810
    },
    {
      "grad_norm": 0.6607022881507874,
      "learning_rate": 9.972204715673669e-05,
      "loss": 0.0404,
      "step": 820
    },
    {
      "grad_norm": 0.7759642004966736,
      "learning_rate": 9.970436496555617e-05,
      "loss": 0.042,
      "step": 830
    },
    {
      "grad_norm": 0.7415374517440796,
      "learning_rate": 9.968613921540373e-05,
      "loss": 0.0452,
      "step": 840
    },
    {
      "grad_norm": 0.7532877922058105,
      "learning_rate": 9.966737010559326e-05,
      "loss": 0.0438,
      "step": 850
    },
    {
      "grad_norm": 0.5179683566093445,
      "learning_rate": 9.964805784138072e-05,
      "loss": 0.0432,
      "step": 860
    },
    {
      "grad_norm": 0.7171632051467896,
      "learning_rate": 9.962820263396195e-05,
      "loss": 0.0387,
      "step": 870
    },
    {
      "grad_norm": 0.7029691338539124,
      "learning_rate": 9.960780470047033e-05,
      "loss": 0.0414,
      "step": 880
    },
    {
      "grad_norm": 0.7791739702224731,
      "learning_rate": 9.958686426397437e-05,
      "loss": 0.0384,
      "step": 890
    },
    {
      "grad_norm": 0.8373865485191345,
      "learning_rate": 9.956538155347534e-05,
      "loss": 0.0433,
      "step": 900
    },
    {
      "grad_norm": 0.5835863351821899,
      "learning_rate": 9.95433568039047e-05,
      "loss": 0.0382,
      "step": 910
    },
    {
      "grad_norm": 0.5688396692276001,
      "learning_rate": 9.952079025612162e-05,
      "loss": 0.0367,
      "step": 920
    },
    {
      "grad_norm": 0.6643704175949097,
      "learning_rate": 9.949768215691022e-05,
      "loss": 0.0383,
      "step": 930
    },
    {
      "grad_norm": 0.6345815062522888,
      "learning_rate": 9.9474032758977e-05,
      "loss": 0.0387,
      "step": 940
    },
    {
      "grad_norm": 0.6888556480407715,
      "learning_rate": 9.944984232094794e-05,
      "loss": 0.0404,
      "step": 950
    },
    {
      "grad_norm": 0.6753697395324707,
      "learning_rate": 9.942511110736584e-05,
      "loss": 0.0405,
      "step": 960
    },
    {
      "grad_norm": 0.7859911322593689,
      "learning_rate": 9.939983938868726e-05,
      "loss": 0.0406,
      "step": 970
    },
    {
      "grad_norm": 0.6559446454048157,
      "learning_rate": 9.93740274412797e-05,
      "loss": 0.0376,
      "step": 980
    },
    {
      "grad_norm": 0.7269642353057861,
      "learning_rate": 9.934767554741846e-05,
      "loss": 0.0459,
      "step": 990
    },
    {
      "grad_norm": 0.632786214351654,
      "learning_rate": 9.932078399528361e-05,
      "loss": 0.0361,
      "step": 1000
    }
  ],
  "logging_steps": 10,
  "max_steps": 10000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9223372036854775807,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}