{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.2,
  "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": 0.12638650834560394,
      "learning_rate": 3.6e-06,
      "loss": 1.1641,
      "step": 10
    },
    {
      "grad_norm": 0.1471242904663086,
      "learning_rate": 7.6e-06,
      "loss": 1.1535,
      "step": 20
    },
    {
      "grad_norm": 0.18666379153728485,
      "learning_rate": 1.16e-05,
      "loss": 1.1457,
      "step": 30
    },
    {
      "grad_norm": 0.14849230647087097,
      "learning_rate": 1.56e-05,
      "loss": 1.1223,
      "step": 40
    },
    {
      "grad_norm": 0.27684518694877625,
      "learning_rate": 1.9600000000000002e-05,
      "loss": 1.0957,
      "step": 50
    },
    {
      "grad_norm": 0.2600708305835724,
      "learning_rate": 2.36e-05,
      "loss": 1.0719,
      "step": 60
    },
    {
      "grad_norm": 0.5172604322433472,
      "learning_rate": 2.7600000000000003e-05,
      "loss": 1.0266,
      "step": 70
    },
    {
      "grad_norm": 0.44987690448760986,
      "learning_rate": 3.16e-05,
      "loss": 0.9682,
      "step": 80
    },
    {
      "grad_norm": 0.7133139967918396,
      "learning_rate": 3.56e-05,
      "loss": 0.9043,
      "step": 90
    },
    {
      "grad_norm": 0.652862548828125,
      "learning_rate": 3.960000000000001e-05,
      "loss": 0.8439,
      "step": 100
    },
    {
      "grad_norm": 0.6717215180397034,
      "learning_rate": 4.36e-05,
      "loss": 0.7908,
      "step": 110
    },
    {
      "grad_norm": 0.8351187109947205,
      "learning_rate": 4.76e-05,
      "loss": 0.7508,
      "step": 120
    },
    {
      "grad_norm": 0.951558530330658,
      "learning_rate": 5.16e-05,
      "loss": 0.7117,
      "step": 130
    },
    {
      "grad_norm": 0.7069472074508667,
      "learning_rate": 5.560000000000001e-05,
      "loss": 0.6824,
      "step": 140
    },
    {
      "grad_norm": 0.8377196788787842,
      "learning_rate": 5.96e-05,
      "loss": 0.6523,
      "step": 150
    },
    {
      "grad_norm": 0.7963162064552307,
      "learning_rate": 6.36e-05,
      "loss": 0.6193,
      "step": 160
    },
    {
      "grad_norm": 1.0922210216522217,
      "learning_rate": 6.76e-05,
      "loss": 0.5893,
      "step": 170
    },
    {
      "grad_norm": 1.2284756898880005,
      "learning_rate": 7.16e-05,
      "loss": 0.5648,
      "step": 180
    },
    {
      "grad_norm": 0.8342697024345398,
      "learning_rate": 7.560000000000001e-05,
      "loss": 0.5453,
      "step": 190
    },
    {
      "grad_norm": 1.121726155281067,
      "learning_rate": 7.960000000000001e-05,
      "loss": 0.5277,
      "step": 200
    },
    {
      "grad_norm": 0.8587872982025146,
      "learning_rate": 8.36e-05,
      "loss": 0.5082,
      "step": 210
    },
    {
      "grad_norm": 1.2177629470825195,
      "learning_rate": 8.76e-05,
      "loss": 0.4881,
      "step": 220
    },
    {
      "grad_norm": 1.0331931114196777,
      "learning_rate": 9.16e-05,
      "loss": 0.467,
      "step": 230
    },
    {
      "grad_norm": 1.178455114364624,
      "learning_rate": 9.56e-05,
      "loss": 0.4453,
      "step": 240
    },
    {
      "grad_norm": 1.0000393390655518,
      "learning_rate": 9.960000000000001e-05,
      "loss": 0.4235,
      "step": 250
    },
    {
      "grad_norm": 1.2453455924987793,
      "learning_rate": 9.999911419878559e-05,
      "loss": 0.4034,
      "step": 260
    },
    {
      "grad_norm": 1.474942684173584,
      "learning_rate": 9.999605221019081e-05,
      "loss": 0.3801,
      "step": 270
    },
    {
      "grad_norm": 0.9545984864234924,
      "learning_rate": 9.999080323230761e-05,
      "loss": 0.3551,
      "step": 280
    },
    {
      "grad_norm": 1.0123811960220337,
      "learning_rate": 9.998336749474329e-05,
      "loss": 0.329,
      "step": 290
    },
    {
      "grad_norm": 1.214898705482483,
      "learning_rate": 9.997374532276107e-05,
      "loss": 0.3039,
      "step": 300
    },
    {
      "grad_norm": 1.2700718641281128,
      "learning_rate": 9.996193713726596e-05,
      "loss": 0.2804,
      "step": 310
    },
    {
      "grad_norm": 1.0745455026626587,
      "learning_rate": 9.994794345478624e-05,
      "loss": 0.2604,
      "step": 320
    },
    {
      "grad_norm": 1.0220052003860474,
      "learning_rate": 9.99317648874509e-05,
      "loss": 0.2303,
      "step": 330
    },
    {
      "grad_norm": 0.9410139918327332,
      "learning_rate": 9.991340214296292e-05,
      "loss": 0.2094,
      "step": 340
    },
    {
      "grad_norm": 1.046053171157837,
      "learning_rate": 9.989285602456819e-05,
      "loss": 0.1971,
      "step": 350
    },
    {
      "grad_norm": 1.174741268157959,
      "learning_rate": 9.98701274310205e-05,
      "loss": 0.1773,
      "step": 360
    },
    {
      "grad_norm": 1.0372276306152344,
      "learning_rate": 9.984521735654218e-05,
      "loss": 0.1615,
      "step": 370
    },
    {
      "grad_norm": 1.0307488441467285,
      "learning_rate": 9.981812689078057e-05,
      "loss": 0.1429,
      "step": 380
    },
    {
      "grad_norm": 2.1818134784698486,
      "learning_rate": 9.978885721876041e-05,
      "loss": 0.1284,
      "step": 390
    },
    {
      "grad_norm": 0.9964090585708618,
      "learning_rate": 9.975740962083198e-05,
      "loss": 0.1258,
      "step": 400
    },
    {
      "grad_norm": 0.9651419520378113,
      "learning_rate": 9.972378547261504e-05,
      "loss": 0.1093,
      "step": 410
    },
    {
      "grad_norm": 1.0263780355453491,
      "learning_rate": 9.968798624493885e-05,
      "loss": 0.1062,
      "step": 420
    },
    {
      "grad_norm": 0.9092751741409302,
      "learning_rate": 9.965001350377753e-05,
      "loss": 0.1009,
      "step": 430
    },
    {
      "grad_norm": 1.1471643447875977,
      "learning_rate": 9.960986891018183e-05,
      "loss": 0.0967,
      "step": 440
    },
    {
      "grad_norm": 0.95041823387146,
      "learning_rate": 9.95675542202063e-05,
      "loss": 0.0941,
      "step": 450
    },
    {
      "grad_norm": 0.9787719249725342,
      "learning_rate": 9.952307128483256e-05,
      "loss": 0.0942,
      "step": 460
    },
    {
      "grad_norm": 0.9564359188079834,
      "learning_rate": 9.947642204988835e-05,
      "loss": 0.096,
      "step": 470
    },
    {
      "grad_norm": 0.9926328659057617,
      "learning_rate": 9.942760855596226e-05,
      "loss": 0.0931,
      "step": 480
    },
    {
      "grad_norm": 0.8731317520141602,
      "learning_rate": 9.937663293831471e-05,
      "loss": 0.0901,
      "step": 490
    },
    {
      "grad_norm": 1.1485458612442017,
      "learning_rate": 9.932349742678433e-05,
      "loss": 0.0875,
      "step": 500
    },
    {
      "grad_norm": 0.9461956024169922,
      "learning_rate": 9.926820434569051e-05,
      "loss": 0.0892,
      "step": 510
    },
    {
      "grad_norm": 0.8998910784721375,
      "learning_rate": 9.921075611373179e-05,
      "loss": 0.0853,
      "step": 520
    },
    {
      "grad_norm": 1.0842655897140503,
      "learning_rate": 9.915115524387988e-05,
      "loss": 0.0841,
      "step": 530
    },
    {
      "grad_norm": 1.0050004720687866,
      "learning_rate": 9.908940434326997e-05,
      "loss": 0.0864,
      "step": 540
    },
    {
      "grad_norm": 1.1228138208389282,
      "learning_rate": 9.902550611308645e-05,
      "loss": 0.0839,
      "step": 550
    },
    {
      "grad_norm": 0.8371630907058716,
      "learning_rate": 9.895946334844494e-05,
      "loss": 0.087,
      "step": 560
    },
    {
      "grad_norm": 0.9920855760574341,
      "learning_rate": 9.889127893826989e-05,
      "loss": 0.0833,
      "step": 570
    },
    {
      "grad_norm": 0.8540536165237427,
      "learning_rate": 9.882095586516831e-05,
      "loss": 0.0808,
      "step": 580
    },
    {
      "grad_norm": 1.0284672975540161,
      "learning_rate": 9.874849720529921e-05,
      "loss": 0.0778,
      "step": 590
    },
    {
      "grad_norm": 0.7409518361091614,
      "learning_rate": 9.867390612823914e-05,
      "loss": 0.0776,
      "step": 600
    },
    {
      "grad_norm": 0.7265949249267578,
      "learning_rate": 9.859718589684344e-05,
      "loss": 0.0775,
      "step": 610
    },
    {
      "grad_norm": 0.8303801417350769,
      "learning_rate": 9.851833986710353e-05,
      "loss": 0.0766,
      "step": 620
    },
    {
      "grad_norm": 0.846651017665863,
      "learning_rate": 9.843737148800023e-05,
      "loss": 0.0737,
      "step": 630
    },
    {
      "grad_norm": 0.8616825342178345,
      "learning_rate": 9.835428430135271e-05,
      "loss": 0.0761,
      "step": 640
    },
    {
      "grad_norm": 0.6212893724441528,
      "learning_rate": 9.82690819416637e-05,
      "loss": 0.0734,
      "step": 650
    },
    {
      "grad_norm": 0.950407862663269,
      "learning_rate": 9.818176813596041e-05,
      "loss": 0.0733,
      "step": 660
    },
    {
      "grad_norm": 0.8584088683128357,
      "learning_rate": 9.809234670363159e-05,
      "loss": 0.0729,
      "step": 670
    },
    {
      "grad_norm": 0.8692848682403564,
      "learning_rate": 9.800082155626034e-05,
      "loss": 0.0715,
      "step": 680
    },
    {
      "grad_norm": 0.813220202922821,
      "learning_rate": 9.790719669745312e-05,
      "loss": 0.0705,
      "step": 690
    },
    {
      "grad_norm": 0.9423041939735413,
      "learning_rate": 9.781147622266455e-05,
      "loss": 0.0696,
      "step": 700
    },
    {
      "grad_norm": 0.8096789121627808,
      "learning_rate": 9.771366431901831e-05,
      "loss": 0.0729,
      "step": 710
    },
    {
      "grad_norm": 0.949595034122467,
      "learning_rate": 9.761376526512394e-05,
      "loss": 0.0704,
      "step": 720
    },
    {
      "grad_norm": 0.8399985432624817,
      "learning_rate": 9.751178343088963e-05,
      "loss": 0.07,
      "step": 730
    },
    {
      "grad_norm": 0.708450198173523,
      "learning_rate": 9.740772327733123e-05,
      "loss": 0.0701,
      "step": 740
    },
    {
      "grad_norm": 0.8553755879402161,
      "learning_rate": 9.730158935637697e-05,
      "loss": 0.0696,
      "step": 750
    },
    {
      "grad_norm": 0.7043846845626831,
      "learning_rate": 9.719338631066834e-05,
      "loss": 0.0675,
      "step": 760
    },
    {
      "grad_norm": 0.8093163371086121,
      "learning_rate": 9.708311887335713e-05,
      "loss": 0.0656,
      "step": 770
    },
    {
      "grad_norm": 0.7058002352714539,
      "learning_rate": 9.697079186789823e-05,
      "loss": 0.0665,
      "step": 780
    },
    {
      "grad_norm": 0.6831558346748352,
      "learning_rate": 9.685641020783876e-05,
      "loss": 0.0641,
      "step": 790
    },
    {
      "grad_norm": 0.799716055393219,
      "learning_rate": 9.67399788966031e-05,
      "loss": 0.0627,
      "step": 800
    },
    {
      "grad_norm": 0.982138454914093,
      "learning_rate": 9.662150302727395e-05,
      "loss": 0.0637,
      "step": 810
    },
    {
      "grad_norm": 0.761420726776123,
      "learning_rate": 9.650098778236968e-05,
      "loss": 0.0653,
      "step": 820
    },
    {
      "grad_norm": 0.7208728790283203,
      "learning_rate": 9.637843843361749e-05,
      "loss": 0.0636,
      "step": 830
    },
    {
      "grad_norm": 0.8718505501747131,
      "learning_rate": 9.62538603417229e-05,
      "loss": 0.0619,
      "step": 840
    },
    {
      "grad_norm": 0.969348132610321,
      "learning_rate": 9.612725895613526e-05,
      "loss": 0.0615,
      "step": 850
    },
    {
      "grad_norm": 0.8718401789665222,
      "learning_rate": 9.599863981480926e-05,
      "loss": 0.0623,
      "step": 860
    },
    {
      "grad_norm": 0.6654447317123413,
      "learning_rate": 9.586800854396283e-05,
      "loss": 0.0608,
      "step": 870
    },
    {
      "grad_norm": 0.7654517889022827,
      "learning_rate": 9.573537085783095e-05,
      "loss": 0.0602,
      "step": 880
    },
    {
      "grad_norm": 0.8219175934791565,
      "learning_rate": 9.560073255841571e-05,
      "loss": 0.0594,
      "step": 890
    },
    {
      "grad_norm": 0.6851817965507507,
      "learning_rate": 9.546409953523247e-05,
      "loss": 0.0591,
      "step": 900
    },
    {
      "grad_norm": 0.7167078256607056,
      "learning_rate": 9.532547776505229e-05,
      "loss": 0.0574,
      "step": 910
    },
    {
      "grad_norm": 0.8798813819885254,
      "learning_rate": 9.518487331164048e-05,
      "loss": 0.0591,
      "step": 920
    },
    {
      "grad_norm": 0.5818827748298645,
      "learning_rate": 9.504229232549134e-05,
      "loss": 0.0574,
      "step": 930
    },
    {
      "grad_norm": 0.749271810054779,
      "learning_rate": 9.489774104355909e-05,
      "loss": 0.0574,
      "step": 940
    },
    {
      "grad_norm": 0.6833415031433105,
      "learning_rate": 9.475122578898507e-05,
      "loss": 0.0572,
      "step": 950
    },
    {
      "grad_norm": 0.6324511766433716,
      "learning_rate": 9.460275297082119e-05,
      "loss": 0.0554,
      "step": 960
    },
    {
      "grad_norm": 0.7919363975524902,
      "learning_rate": 9.445232908374948e-05,
      "loss": 0.0586,
      "step": 970
    },
    {
      "grad_norm": 0.6597112417221069,
      "learning_rate": 9.429996070779808e-05,
      "loss": 0.0561,
      "step": 980
    },
    {
      "grad_norm": 0.6228717565536499,
      "learning_rate": 9.414565450805333e-05,
      "loss": 0.0558,
      "step": 990
    },
    {
      "grad_norm": 0.7159019708633423,
      "learning_rate": 9.398941723436831e-05,
      "loss": 0.0575,
      "step": 1000
    }
  ],
  "logging_steps": 10,
  "max_steps": 5000,
  "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": 768,
  "trial_name": null,
  "trial_params": null
}