{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 200,
  "global_step": 891,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.03368421052631579,
      "grad_norm": 0.39124977588653564,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.6411091327667237,
      "step": 10
    },
    {
      "epoch": 0.06736842105263158,
      "grad_norm": 0.292263388633728,
      "learning_rate": 8.444444444444446e-06,
      "loss": 0.5484030246734619,
      "step": 20
    },
    {
      "epoch": 0.10105263157894737,
      "grad_norm": 0.25442785024642944,
      "learning_rate": 1.288888888888889e-05,
      "loss": 0.5427792072296143,
      "step": 30
    },
    {
      "epoch": 0.13473684210526315,
      "grad_norm": 0.2947239875793457,
      "learning_rate": 1.7333333333333336e-05,
      "loss": 0.5606600761413574,
      "step": 40
    },
    {
      "epoch": 0.16842105263157894,
      "grad_norm": 0.2775092124938965,
      "learning_rate": 1.9998896833611603e-05,
      "loss": 0.5333517074584961,
      "step": 50
    },
    {
      "epoch": 0.20210526315789473,
      "grad_norm": 0.29785287380218506,
      "learning_rate": 1.9986489006814454e-05,
      "loss": 0.5225426197052002,
      "step": 60
    },
    {
      "epoch": 0.23578947368421052,
      "grad_norm": 0.2655927836894989,
      "learning_rate": 1.9960311560501457e-05,
      "loss": 0.5233192920684815,
      "step": 70
    },
    {
      "epoch": 0.2694736842105263,
      "grad_norm": 0.26086923480033875,
      "learning_rate": 1.99204005887869e-05,
      "loss": 0.5474831581115722,
      "step": 80
    },
    {
      "epoch": 0.3031578947368421,
      "grad_norm": 0.28363120555877686,
      "learning_rate": 1.986681112191161e-05,
      "loss": 0.5189789772033692,
      "step": 90
    },
    {
      "epoch": 0.3368421052631579,
      "grad_norm": 0.27852076292037964,
      "learning_rate": 1.979961705036587e-05,
      "loss": 0.5073096752166748,
      "step": 100
    },
    {
      "epoch": 0.3705263157894737,
      "grad_norm": 0.2481539398431778,
      "learning_rate": 1.9718911023007382e-05,
      "loss": 0.5242129325866699,
      "step": 110
    },
    {
      "epoch": 0.40421052631578946,
      "grad_norm": 0.29133835434913635,
      "learning_rate": 1.9624804319314704e-05,
      "loss": 0.5255519866943359,
      "step": 120
    },
    {
      "epoch": 0.4378947368421053,
      "grad_norm": 0.24475380778312683,
      "learning_rate": 1.9517426695952358e-05,
      "loss": 0.5238783836364747,
      "step": 130
    },
    {
      "epoch": 0.47157894736842104,
      "grad_norm": 0.20989426970481873,
      "learning_rate": 1.9396926207859085e-05,
      "loss": 0.5111160278320312,
      "step": 140
    },
    {
      "epoch": 0.5052631578947369,
      "grad_norm": 0.26935264468193054,
      "learning_rate": 1.926346900410604e-05,
      "loss": 0.5124940872192383,
      "step": 150
    },
    {
      "epoch": 0.5389473684210526,
      "grad_norm": 0.2176002562046051,
      "learning_rate": 1.9117239098806296e-05,
      "loss": 0.5246289253234864,
      "step": 160
    },
    {
      "epoch": 0.5726315789473684,
      "grad_norm": 0.24210523068904877,
      "learning_rate": 1.895843811739162e-05,
      "loss": 0.5537409305572509,
      "step": 170
    },
    {
      "epoch": 0.6063157894736843,
      "grad_norm": 0.257548451423645,
      "learning_rate": 1.87872850186063e-05,
      "loss": 0.513533639907837,
      "step": 180
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.23951275646686554,
      "learning_rate": 1.8604015792601395e-05,
      "loss": 0.48828821182250975,
      "step": 190
    },
    {
      "epoch": 0.6736842105263158,
      "grad_norm": 0.2258734107017517,
      "learning_rate": 1.8408883135545634e-05,
      "loss": 0.5039851188659668,
      "step": 200
    },
    {
      "epoch": 0.7073684210526315,
      "grad_norm": 0.19968342781066895,
      "learning_rate": 1.8202156101201646e-05,
      "loss": 0.521153450012207,
      "step": 210
    },
    {
      "epoch": 0.7410526315789474,
      "grad_norm": 0.2292439192533493,
      "learning_rate": 1.7984119729947944e-05,
      "loss": 0.5093928337097168,
      "step": 220
    },
    {
      "epoch": 0.7747368421052632,
      "grad_norm": 0.2542964518070221,
      "learning_rate": 1.7755074655758174e-05,
      "loss": 0.5263922691345215,
      "step": 230
    },
    {
      "epoch": 0.8084210526315789,
      "grad_norm": 0.2533174455165863,
      "learning_rate": 1.7515336691679478e-05,
      "loss": 0.49692983627319337,
      "step": 240
    },
    {
      "epoch": 0.8421052631578947,
      "grad_norm": 0.2821718156337738,
      "learning_rate": 1.7265236394381634e-05,
      "loss": 0.518037223815918,
      "step": 250
    },
    {
      "epoch": 0.8757894736842106,
      "grad_norm": 0.2320486307144165,
      "learning_rate": 1.7005118608377288e-05,
      "loss": 0.49869470596313475,
      "step": 260
    },
    {
      "epoch": 0.9094736842105263,
      "grad_norm": 0.23216022551059723,
      "learning_rate": 1.6735341990541766e-05,
      "loss": 0.496633243560791,
      "step": 270
    },
    {
      "epoch": 0.9431578947368421,
      "grad_norm": 0.23232285678386688,
      "learning_rate": 1.6456278515588023e-05,
      "loss": 0.4894012451171875,
      "step": 280
    },
    {
      "epoch": 0.9768421052631578,
      "grad_norm": 0.23754945397377014,
      "learning_rate": 1.61683129631787e-05,
      "loss": 0.4862018585205078,
      "step": 290
    },
    {
      "epoch": 1.0101052631578948,
      "grad_norm": 0.23209305107593536,
      "learning_rate": 1.5871842387382307e-05,
      "loss": 0.4496741771697998,
      "step": 300
    },
    {
      "epoch": 1.0437894736842106,
      "grad_norm": 0.230687215924263,
      "learning_rate": 1.5567275569205216e-05,
      "loss": 0.3736781120300293,
      "step": 310
    },
    {
      "epoch": 1.0774736842105264,
      "grad_norm": 0.26011544466018677,
      "learning_rate": 1.5255032452954246e-05,
      "loss": 0.3811728239059448,
      "step": 320
    },
    {
      "epoch": 1.1111578947368421,
      "grad_norm": 0.2397359162569046,
      "learning_rate": 1.4935543567206984e-05,
      "loss": 0.3847402572631836,
      "step": 330
    },
    {
      "epoch": 1.1448421052631579,
      "grad_norm": 0.24537348747253418,
      "learning_rate": 1.460924943118828e-05,
      "loss": 0.4140181541442871,
      "step": 340
    },
    {
      "epoch": 1.1785263157894736,
      "grad_norm": 0.29034218192100525,
      "learning_rate": 1.4276599947371388e-05,
      "loss": 0.40039353370666503,
      "step": 350
    },
    {
      "epoch": 1.2122105263157894,
      "grad_norm": 0.23461507260799408,
      "learning_rate": 1.3938053781141224e-05,
      "loss": 0.38444128036499026,
      "step": 360
    },
    {
      "epoch": 1.2458947368421052,
      "grad_norm": 0.278699666261673,
      "learning_rate": 1.3594077728375129e-05,
      "loss": 0.4034078598022461,
      "step": 370
    },
    {
      "epoch": 1.279578947368421,
      "grad_norm": 0.2430262416601181,
      "learning_rate": 1.3245146071813114e-05,
      "loss": 0.3933774471282959,
      "step": 380
    },
    {
      "epoch": 1.313263157894737,
      "grad_norm": 0.24822789430618286,
      "learning_rate": 1.2891739927104992e-05,
      "loss": 0.37237229347229006,
      "step": 390
    },
    {
      "epoch": 1.3469473684210527,
      "grad_norm": 0.23200945556163788,
      "learning_rate": 1.2534346579436158e-05,
      "loss": 0.3913832187652588,
      "step": 400
    },
    {
      "epoch": 1.3806315789473684,
      "grad_norm": 0.22769705951213837,
      "learning_rate": 1.217345881164667e-05,
      "loss": 0.3784568071365356,
      "step": 410
    },
    {
      "epoch": 1.4143157894736842,
      "grad_norm": 0.24226728081703186,
      "learning_rate": 1.1809574224769983e-05,
      "loss": 0.39284965991973875,
      "step": 420
    },
    {
      "epoch": 1.448,
      "grad_norm": 0.25648000836372375,
      "learning_rate": 1.1443194551928267e-05,
      "loss": 0.38346447944641116,
      "step": 430
    },
    {
      "epoch": 1.4816842105263157,
      "grad_norm": 0.26635488867759705,
      "learning_rate": 1.1074824966530312e-05,
      "loss": 0.4148388385772705,
      "step": 440
    },
    {
      "epoch": 1.5153684210526315,
      "grad_norm": 0.25696197152137756,
      "learning_rate": 1.0704973385725853e-05,
      "loss": 0.37754507064819337,
      "step": 450
    },
    {
      "epoch": 1.5490526315789475,
      "grad_norm": 0.2661338746547699,
      "learning_rate": 1.0334149770076747e-05,
      "loss": 0.3888374090194702,
      "step": 460
    },
    {
      "epoch": 1.582736842105263,
      "grad_norm": 0.26346883177757263,
      "learning_rate": 9.962865420410702e-06,
      "loss": 0.3881821632385254,
      "step": 470
    },
    {
      "epoch": 1.616421052631579,
      "grad_norm": 0.23709534108638763,
      "learning_rate": 9.591632272826935e-06,
      "loss": 0.3894140958786011,
      "step": 480
    },
    {
      "epoch": 1.6501052631578947,
      "grad_norm": 0.2380121797323227,
      "learning_rate": 9.22096219282597e-06,
      "loss": 0.38091790676116943,
      "step": 490
    },
    {
      "epoch": 1.6837894736842105,
      "grad_norm": 0.2552591562271118,
      "learning_rate": 8.85136626953673e-06,
      "loss": 0.3866596698760986,
      "step": 500
    },
    {
      "epoch": 1.7174736842105263,
      "grad_norm": 0.2706044018268585,
      "learning_rate": 8.483354111014142e-06,
      "loss": 0.34658288955688477,
      "step": 510
    },
    {
      "epoch": 1.751157894736842,
      "grad_norm": 0.238291934132576,
      "learning_rate": 8.117433141578865e-06,
      "loss": 0.3751774549484253,
      "step": 520
    },
    {
      "epoch": 1.784842105263158,
      "grad_norm": 0.2516458332538605,
      "learning_rate": 7.75410790216802e-06,
      "loss": 0.3806126356124878,
      "step": 530
    },
    {
      "epoch": 1.8185263157894735,
      "grad_norm": 0.2464667409658432,
      "learning_rate": 7.3938793546615775e-06,
      "loss": 0.3747535705566406,
      "step": 540
    },
    {
      "epoch": 1.8522105263157895,
      "grad_norm": 0.23317350447177887,
      "learning_rate": 7.037244191143662e-06,
      "loss": 0.39618306159973143,
      "step": 550
    },
    {
      "epoch": 1.8858947368421053,
      "grad_norm": 0.23871706426143646,
      "learning_rate": 6.684694149051156e-06,
      "loss": 0.39216933250427244,
      "step": 560
    },
    {
      "epoch": 1.919578947368421,
      "grad_norm": 0.25860607624053955,
      "learning_rate": 6.336715333153869e-06,
      "loss": 0.3773647308349609,
      "step": 570
    },
    {
      "epoch": 1.9532631578947368,
      "grad_norm": 0.24600288271903992,
      "learning_rate": 5.993787545301204e-06,
      "loss": 0.3795567274093628,
      "step": 580
    },
    {
      "epoch": 1.9869473684210526,
      "grad_norm": 0.263887882232666,
      "learning_rate": 5.656383622859418e-06,
      "loss": 0.3699490070343018,
      "step": 590
    },
    {
      "epoch": 2.0202105263157897,
      "grad_norm": 0.24102307856082916,
      "learning_rate": 5.3249687867517095e-06,
      "loss": 0.33141674995422366,
      "step": 600
    },
    {
      "epoch": 2.053894736842105,
      "grad_norm": 0.2210257202386856,
      "learning_rate": 5.000000000000003e-06,
      "loss": 0.2707043647766113,
      "step": 610
    },
    {
      "epoch": 2.087578947368421,
      "grad_norm": 0.25143396854400635,
      "learning_rate": 4.681925337653006e-06,
      "loss": 0.27166755199432374,
      "step": 620
    },
    {
      "epoch": 2.1212631578947367,
      "grad_norm": 0.23633138835430145,
      "learning_rate": 4.371183368969165e-06,
      "loss": 0.28992409706115724,
      "step": 630
    },
    {
      "epoch": 2.1549473684210527,
      "grad_norm": 0.248697429895401,
      "learning_rate": 4.0682025527064486e-06,
      "loss": 0.28457343578338623,
      "step": 640
    },
    {
      "epoch": 2.1886315789473683,
      "grad_norm": 0.26528167724609375,
      "learning_rate": 3.7734006463527695e-06,
      "loss": 0.2810338497161865,
      "step": 650
    },
    {
      "epoch": 2.2223157894736842,
      "grad_norm": 0.28675395250320435,
      "learning_rate": 3.4871841301115615e-06,
      "loss": 0.28777623176574707,
      "step": 660
    },
    {
      "epoch": 2.2560000000000002,
      "grad_norm": 0.24856451153755188,
      "learning_rate": 3.209947646436752e-06,
      "loss": 0.2675544500350952,
      "step": 670
    },
    {
      "epoch": 2.2896842105263158,
      "grad_norm": 0.27366021275520325,
      "learning_rate": 2.9420734558899323e-06,
      "loss": 0.27961390018463134,
      "step": 680
    },
    {
      "epoch": 2.3233684210526317,
      "grad_norm": 0.26319196820259094,
      "learning_rate": 2.6839309100699975e-06,
      "loss": 0.2918535232543945,
      "step": 690
    },
    {
      "epoch": 2.3570526315789473,
      "grad_norm": 0.3220144808292389,
      "learning_rate": 2.4358759423419476e-06,
      "loss": 0.2924020767211914,
      "step": 700
    },
    {
      "epoch": 2.3907368421052633,
      "grad_norm": 0.2410717010498047,
      "learning_rate": 2.1982505770671303e-06,
      "loss": 0.2743253707885742,
      "step": 710
    },
    {
      "epoch": 2.424421052631579,
      "grad_norm": 0.30828380584716797,
      "learning_rate": 1.9713824580115336e-06,
      "loss": 0.27377402782440186,
      "step": 720
    },
    {
      "epoch": 2.458105263157895,
      "grad_norm": 0.20561999082565308,
      "learning_rate": 1.7555843965823992e-06,
      "loss": 0.27103970050811765,
      "step": 730
    },
    {
      "epoch": 2.4917894736842103,
      "grad_norm": 0.26168638467788696,
      "learning_rate": 1.5511539405160824e-06,
      "loss": 0.2865826845169067,
      "step": 740
    },
    {
      "epoch": 2.5254736842105263,
      "grad_norm": 0.27809929847717285,
      "learning_rate": 1.3583729636118359e-06,
      "loss": 0.2839942455291748,
      "step": 750
    },
    {
      "epoch": 2.559157894736842,
      "grad_norm": 0.27653998136520386,
      "learning_rate": 1.1775072770771833e-06,
      "loss": 0.28238811492919924,
      "step": 760
    },
    {
      "epoch": 2.592842105263158,
      "grad_norm": 0.23926998674869537,
      "learning_rate": 1.0088062630208272e-06,
      "loss": 0.2802205324172974,
      "step": 770
    },
    {
      "epoch": 2.626526315789474,
      "grad_norm": 0.23364897072315216,
      "learning_rate": 8.525025305983936e-07,
      "loss": 0.2747425317764282,
      "step": 780
    },
    {
      "epoch": 2.6602105263157894,
      "grad_norm": 0.26213040947914124,
      "learning_rate": 7.088115952851238e-07,
      "loss": 0.264876651763916,
      "step": 790
    },
    {
      "epoch": 2.6938947368421053,
      "grad_norm": 0.23854820430278778,
      "learning_rate": 5.779315817178e-07,
      "loss": 0.27306721210479734,
      "step": 800
    },
    {
      "epoch": 2.7275789473684213,
      "grad_norm": 0.26059138774871826,
      "learning_rate": 4.6004295051554236e-07,
      "loss": 0.27785990238189695,
      "step": 810
    },
    {
      "epoch": 2.761263157894737,
      "grad_norm": 0.22004592418670654,
      "learning_rate": 3.553082494562354e-07,
      "loss": 0.28950111865997313,
      "step": 820
    },
    {
      "epoch": 2.7949473684210524,
      "grad_norm": 0.24607206881046295,
      "learning_rate": 2.638718893515946e-07,
      "loss": 0.28249540328979494,
      "step": 830
    },
    {
      "epoch": 2.8286315789473684,
      "grad_norm": 0.2633419930934906,
      "learning_rate": 1.8585994492995917e-07,
      "loss": 0.2748891353607178,
      "step": 840
    },
    {
      "epoch": 2.8623157894736844,
      "grad_norm": 0.2897525429725647,
      "learning_rate": 1.21379981001305e-07,
      "loss": 0.27536725997924805,
      "step": 850
    },
    {
      "epoch": 2.896,
      "grad_norm": 0.25982382893562317,
      "learning_rate": 7.052090414422119e-08,
      "loss": 0.2835198163986206,
      "step": 860
    },
    {
      "epoch": 2.929684210526316,
      "grad_norm": 0.2790670096874237,
      "learning_rate": 3.335284011929951e-08,
      "loss": 0.2774952411651611,
      "step": 870
    },
    {
      "epoch": 2.9633684210526314,
      "grad_norm": 0.24153731763362885,
      "learning_rate": 9.927037177993593e-09,
      "loss": 0.25984973907470704,
      "step": 880
    },
    {
      "epoch": 2.9970526315789474,
      "grad_norm": 0.215993270277977,
      "learning_rate": 2.75795400255241e-10,
      "loss": 0.28056533336639405,
      "step": 890
    }
  ],
  "logging_steps": 10,
  "max_steps": 891,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 200,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 4.040324108884378e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}