{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9992659652556888,
  "eval_steps": 500,
  "global_step": 1021,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.009787129924149743,
      "grad_norm": 3.160645008087158,
      "learning_rate": 8.737864077669904e-07,
      "loss": 1.6202,
      "step": 10
    },
    {
      "epoch": 0.019574259848299486,
      "grad_norm": 1.8336282968521118,
      "learning_rate": 1.8446601941747574e-06,
      "loss": 1.5766,
      "step": 20
    },
    {
      "epoch": 0.029361389772449228,
      "grad_norm": 1.5718436241149902,
      "learning_rate": 2.8155339805825245e-06,
      "loss": 1.4962,
      "step": 30
    },
    {
      "epoch": 0.03914851969659897,
      "grad_norm": 1.2343661785125732,
      "learning_rate": 3.7864077669902915e-06,
      "loss": 1.3827,
      "step": 40
    },
    {
      "epoch": 0.048935649620748714,
      "grad_norm": 1.0965512990951538,
      "learning_rate": 4.7572815533980585e-06,
      "loss": 1.3328,
      "step": 50
    },
    {
      "epoch": 0.058722779544898455,
      "grad_norm": 1.0640696287155151,
      "learning_rate": 5.728155339805825e-06,
      "loss": 1.2826,
      "step": 60
    },
    {
      "epoch": 0.0685099094690482,
      "grad_norm": 1.076289415359497,
      "learning_rate": 6.6990291262135935e-06,
      "loss": 1.234,
      "step": 70
    },
    {
      "epoch": 0.07829703939319795,
      "grad_norm": 1.036766529083252,
      "learning_rate": 7.66990291262136e-06,
      "loss": 1.2398,
      "step": 80
    },
    {
      "epoch": 0.0880841693173477,
      "grad_norm": 1.0808024406433105,
      "learning_rate": 8.640776699029127e-06,
      "loss": 1.2031,
      "step": 90
    },
    {
      "epoch": 0.09787129924149743,
      "grad_norm": 1.1380928754806519,
      "learning_rate": 9.611650485436894e-06,
      "loss": 1.2039,
      "step": 100
    },
    {
      "epoch": 0.10765842916564718,
      "grad_norm": 1.0566169023513794,
      "learning_rate": 9.998945997517957e-06,
      "loss": 1.1918,
      "step": 110
    },
    {
      "epoch": 0.11744555908979691,
      "grad_norm": 1.080167293548584,
      "learning_rate": 9.992506480404137e-06,
      "loss": 1.1916,
      "step": 120
    },
    {
      "epoch": 0.12723268901394666,
      "grad_norm": 1.0032321214675903,
      "learning_rate": 9.980220534903889e-06,
      "loss": 1.1654,
      "step": 130
    },
    {
      "epoch": 0.1370198189380964,
      "grad_norm": 1.101294994354248,
      "learning_rate": 9.96210254835968e-06,
      "loss": 1.1627,
      "step": 140
    },
    {
      "epoch": 0.14680694886224616,
      "grad_norm": 1.1083616018295288,
      "learning_rate": 9.938173737671531e-06,
      "loss": 1.1622,
      "step": 150
    },
    {
      "epoch": 0.1565940787863959,
      "grad_norm": 0.9857867360115051,
      "learning_rate": 9.908462124451152e-06,
      "loss": 1.1588,
      "step": 160
    },
    {
      "epoch": 0.16638120871054562,
      "grad_norm": 1.0299389362335205,
      "learning_rate": 9.873002502207502e-06,
      "loss": 1.1774,
      "step": 170
    },
    {
      "epoch": 0.1761683386346954,
      "grad_norm": 1.0562578439712524,
      "learning_rate": 9.831836395602164e-06,
      "loss": 1.1632,
      "step": 180
    },
    {
      "epoch": 0.18595546855884512,
      "grad_norm": 1.0918824672698975,
      "learning_rate": 9.78501201182228e-06,
      "loss": 1.1575,
      "step": 190
    },
    {
      "epoch": 0.19574259848299486,
      "grad_norm": 0.9658213257789612,
      "learning_rate": 9.732584184127973e-06,
      "loss": 1.145,
      "step": 200
    },
    {
      "epoch": 0.20552972840714462,
      "grad_norm": 0.9882791042327881,
      "learning_rate": 9.674614307640368e-06,
      "loss": 1.1595,
      "step": 210
    },
    {
      "epoch": 0.21531685833129435,
      "grad_norm": 1.0786079168319702,
      "learning_rate": 9.611170267445401e-06,
      "loss": 1.1383,
      "step": 220
    },
    {
      "epoch": 0.2251039882554441,
      "grad_norm": 1.0685032606124878,
      "learning_rate": 9.542326359097619e-06,
      "loss": 1.1329,
      "step": 230
    },
    {
      "epoch": 0.23489111817959382,
      "grad_norm": 1.0679073333740234,
      "learning_rate": 9.468163201617063e-06,
      "loss": 1.1323,
      "step": 240
    },
    {
      "epoch": 0.24467824810374358,
      "grad_norm": 0.9797915816307068,
      "learning_rate": 9.388767643081109e-06,
      "loss": 1.1332,
      "step": 250
    },
    {
      "epoch": 0.2544653780278933,
      "grad_norm": 1.055120587348938,
      "learning_rate": 9.30423265892184e-06,
      "loss": 1.1301,
      "step": 260
    },
    {
      "epoch": 0.2642525079520431,
      "grad_norm": 1.034970998764038,
      "learning_rate": 9.214657243048021e-06,
      "loss": 1.1483,
      "step": 270
    },
    {
      "epoch": 0.2740396378761928,
      "grad_norm": 1.105406641960144,
      "learning_rate": 9.120146291919206e-06,
      "loss": 1.1249,
      "step": 280
    },
    {
      "epoch": 0.28382676780034255,
      "grad_norm": 0.9189131259918213,
      "learning_rate": 9.020810481707709e-06,
      "loss": 1.1385,
      "step": 290
    },
    {
      "epoch": 0.2936138977244923,
      "grad_norm": 1.0855603218078613,
      "learning_rate": 8.916766138692303e-06,
      "loss": 1.1326,
      "step": 300
    },
    {
      "epoch": 0.303401027648642,
      "grad_norm": 0.9882815480232239,
      "learning_rate": 8.808135103035407e-06,
      "loss": 1.1223,
      "step": 310
    },
    {
      "epoch": 0.3131881575727918,
      "grad_norm": 0.9927927255630493,
      "learning_rate": 8.695044586103297e-06,
      "loss": 1.1314,
      "step": 320
    },
    {
      "epoch": 0.32297528749694154,
      "grad_norm": 1.0153377056121826,
      "learning_rate": 8.577627021496413e-06,
      "loss": 1.1254,
      "step": 330
    },
    {
      "epoch": 0.33276241742109125,
      "grad_norm": 0.9959344267845154,
      "learning_rate": 8.456019909964224e-06,
      "loss": 1.1115,
      "step": 340
    },
    {
      "epoch": 0.342549547345241,
      "grad_norm": 1.044042706489563,
      "learning_rate": 8.330365658386252e-06,
      "loss": 1.142,
      "step": 350
    },
    {
      "epoch": 0.3523366772693908,
      "grad_norm": 0.9094734787940979,
      "learning_rate": 8.200811413007808e-06,
      "loss": 1.1254,
      "step": 360
    },
    {
      "epoch": 0.3621238071935405,
      "grad_norm": 1.0200663805007935,
      "learning_rate": 8.06750888712576e-06,
      "loss": 1.127,
      "step": 370
    },
    {
      "epoch": 0.37191093711769024,
      "grad_norm": 1.0179405212402344,
      "learning_rate": 7.930614183426074e-06,
      "loss": 1.0985,
      "step": 380
    },
    {
      "epoch": 0.38169806704184,
      "grad_norm": 0.9346506595611572,
      "learning_rate": 7.790287611181217e-06,
      "loss": 1.1061,
      "step": 390
    },
    {
      "epoch": 0.3914851969659897,
      "grad_norm": 0.9131274223327637,
      "learning_rate": 7.646693498521472e-06,
      "loss": 1.1178,
      "step": 400
    },
    {
      "epoch": 0.4012723268901395,
      "grad_norm": 1.1208817958831787,
      "learning_rate": 7.500000000000001e-06,
      "loss": 1.1076,
      "step": 410
    },
    {
      "epoch": 0.41105945681428924,
      "grad_norm": 0.9498858451843262,
      "learning_rate": 7.35037889967702e-06,
      "loss": 1.123,
      "step": 420
    },
    {
      "epoch": 0.42084658673843894,
      "grad_norm": 0.9704530239105225,
      "learning_rate": 7.19800540995367e-06,
      "loss": 1.1142,
      "step": 430
    },
    {
      "epoch": 0.4306337166625887,
      "grad_norm": 1.083723545074463,
      "learning_rate": 7.043057966391158e-06,
      "loss": 1.1068,
      "step": 440
    },
    {
      "epoch": 0.4404208465867384,
      "grad_norm": 0.959192156791687,
      "learning_rate": 6.885718018755448e-06,
      "loss": 1.1024,
      "step": 450
    },
    {
      "epoch": 0.4502079765108882,
      "grad_norm": 0.9829457402229309,
      "learning_rate": 6.7261698185322e-06,
      "loss": 1.0918,
      "step": 460
    },
    {
      "epoch": 0.45999510643503794,
      "grad_norm": 0.9769061207771301,
      "learning_rate": 6.5646002031607726e-06,
      "loss": 1.1092,
      "step": 470
    },
    {
      "epoch": 0.46978223635918764,
      "grad_norm": 0.9248567223548889,
      "learning_rate": 6.401198377239979e-06,
      "loss": 1.1096,
      "step": 480
    },
    {
      "epoch": 0.4795693662833374,
      "grad_norm": 1.0321615934371948,
      "learning_rate": 6.236155690961795e-06,
      "loss": 1.1118,
      "step": 490
    },
    {
      "epoch": 0.48935649620748717,
      "grad_norm": 0.9511229395866394,
      "learning_rate": 6.0696654160324875e-06,
      "loss": 1.0969,
      "step": 500
    },
    {
      "epoch": 0.4991436261316369,
      "grad_norm": 0.9163072109222412,
      "learning_rate": 5.901922519343586e-06,
      "loss": 1.107,
      "step": 510
    },
    {
      "epoch": 0.5089307560557866,
      "grad_norm": 0.9588350653648376,
      "learning_rate": 5.733123434657704e-06,
      "loss": 1.1008,
      "step": 520
    },
    {
      "epoch": 0.5187178859799364,
      "grad_norm": 0.9725273251533508,
      "learning_rate": 5.5634658325766066e-06,
      "loss": 1.1122,
      "step": 530
    },
    {
      "epoch": 0.5285050159040862,
      "grad_norm": 0.954433798789978,
      "learning_rate": 5.393148389060893e-06,
      "loss": 1.1052,
      "step": 540
    },
    {
      "epoch": 0.5382921458282359,
      "grad_norm": 1.1059950590133667,
      "learning_rate": 5.222370552772353e-06,
      "loss": 1.1014,
      "step": 550
    },
    {
      "epoch": 0.5480792757523856,
      "grad_norm": 0.9268523454666138,
      "learning_rate": 5.05133231151145e-06,
      "loss": 1.0981,
      "step": 560
    },
    {
      "epoch": 0.5578664056765353,
      "grad_norm": 0.9672321677207947,
      "learning_rate": 4.880233958023486e-06,
      "loss": 1.1086,
      "step": 570
    },
    {
      "epoch": 0.5676535356006851,
      "grad_norm": 1.0383234024047852,
      "learning_rate": 4.7092758554476215e-06,
      "loss": 1.0938,
      "step": 580
    },
    {
      "epoch": 0.5774406655248349,
      "grad_norm": 0.9663616418838501,
      "learning_rate": 4.53865820268349e-06,
      "loss": 1.1038,
      "step": 590
    },
    {
      "epoch": 0.5872277954489846,
      "grad_norm": 0.9746555685997009,
      "learning_rate": 4.368580799950133e-06,
      "loss": 1.0784,
      "step": 600
    },
    {
      "epoch": 0.5970149253731343,
      "grad_norm": 0.8926970362663269,
      "learning_rate": 4.199242814811807e-06,
      "loss": 1.1028,
      "step": 610
    },
    {
      "epoch": 0.606802055297284,
      "grad_norm": 0.9124487638473511,
      "learning_rate": 4.03084254894465e-06,
      "loss": 1.0938,
      "step": 620
    },
    {
      "epoch": 0.6165891852214338,
      "grad_norm": 0.9407256841659546,
      "learning_rate": 3.863577205917356e-06,
      "loss": 1.0917,
      "step": 630
    },
    {
      "epoch": 0.6263763151455836,
      "grad_norm": 0.9204623103141785,
      "learning_rate": 3.6976426602577565e-06,
      "loss": 1.0837,
      "step": 640
    },
    {
      "epoch": 0.6361634450697333,
      "grad_norm": 0.9785788059234619,
      "learning_rate": 3.5332332280757706e-06,
      "loss": 1.0964,
      "step": 650
    },
    {
      "epoch": 0.6459505749938831,
      "grad_norm": 0.9679555892944336,
      "learning_rate": 3.3705414395113354e-06,
      "loss": 1.0897,
      "step": 660
    },
    {
      "epoch": 0.6557377049180327,
      "grad_norm": 0.8785240054130554,
      "learning_rate": 3.2097578132737716e-06,
      "loss": 1.0986,
      "step": 670
    },
    {
      "epoch": 0.6655248348421825,
      "grad_norm": 0.9295753240585327,
      "learning_rate": 3.0510706335366034e-06,
      "loss": 1.0877,
      "step": 680
    },
    {
      "epoch": 0.6753119647663323,
      "grad_norm": 0.8944317102432251,
      "learning_rate": 2.8946657294491452e-06,
      "loss": 1.0941,
      "step": 690
    },
    {
      "epoch": 0.685099094690482,
      "grad_norm": 0.9147707223892212,
      "learning_rate": 2.740726257522987e-06,
      "loss": 1.0857,
      "step": 700
    },
    {
      "epoch": 0.6948862246146318,
      "grad_norm": 0.889441967010498,
      "learning_rate": 2.5894324871482557e-06,
      "loss": 1.0781,
      "step": 710
    },
    {
      "epoch": 0.7046733545387815,
      "grad_norm": 0.8948488235473633,
      "learning_rate": 2.4409615894908407e-06,
      "loss": 1.0713,
      "step": 720
    },
    {
      "epoch": 0.7144604844629312,
      "grad_norm": 0.881840169429779,
      "learning_rate": 2.2954874300177197e-06,
      "loss": 1.0765,
      "step": 730
    },
    {
      "epoch": 0.724247614387081,
      "grad_norm": 0.9054113030433655,
      "learning_rate": 2.1531803648934333e-06,
      "loss": 1.0835,
      "step": 740
    },
    {
      "epoch": 0.7340347443112307,
      "grad_norm": 0.8451863527297974,
      "learning_rate": 2.0142070414860704e-06,
      "loss": 1.0836,
      "step": 750
    },
    {
      "epoch": 0.7438218742353805,
      "grad_norm": 0.8607450127601624,
      "learning_rate": 1.8787302032164168e-06,
      "loss": 1.0816,
      "step": 760
    },
    {
      "epoch": 0.7536090041595302,
      "grad_norm": 0.8773781657218933,
      "learning_rate": 1.746908498978791e-06,
      "loss": 1.0807,
      "step": 770
    },
    {
      "epoch": 0.76339613408368,
      "grad_norm": 0.9455496668815613,
      "learning_rate": 1.6188962973567068e-06,
      "loss": 1.0828,
      "step": 780
    },
    {
      "epoch": 0.7731832640078297,
      "grad_norm": 0.9180036783218384,
      "learning_rate": 1.4948435058510036e-06,
      "loss": 1.1013,
      "step": 790
    },
    {
      "epoch": 0.7829703939319794,
      "grad_norm": 0.900719165802002,
      "learning_rate": 1.374895395332037e-06,
      "loss": 1.07,
      "step": 800
    },
    {
      "epoch": 0.7927575238561292,
      "grad_norm": 0.8657832145690918,
      "learning_rate": 1.259192429921584e-06,
      "loss": 1.0796,
      "step": 810
    },
    {
      "epoch": 0.802544653780279,
      "grad_norm": 1.0004676580429077,
      "learning_rate": 1.1478701025036359e-06,
      "loss": 1.1017,
      "step": 820
    },
    {
      "epoch": 0.8123317837044287,
      "grad_norm": 0.9102425575256348,
      "learning_rate": 1.0410587760567104e-06,
      "loss": 1.0805,
      "step": 830
    },
    {
      "epoch": 0.8221189136285785,
      "grad_norm": 0.8790053129196167,
      "learning_rate": 9.388835309934985e-07,
      "loss": 1.068,
      "step": 840
    },
    {
      "epoch": 0.8319060435527281,
      "grad_norm": 0.8961747288703918,
      "learning_rate": 8.414640186866063e-07,
      "loss": 1.0832,
      "step": 850
    },
    {
      "epoch": 0.8416931734768779,
      "grad_norm": 0.874332845211029,
      "learning_rate": 7.489143213519301e-07,
      "loss": 1.0858,
      "step": 860
    },
    {
      "epoch": 0.8514803034010276,
      "grad_norm": 0.9413198828697205,
      "learning_rate": 6.613428184537235e-07,
      "loss": 1.0938,
      "step": 870
    },
    {
      "epoch": 0.8612674333251774,
      "grad_norm": 0.8453653454780579,
      "learning_rate": 5.788520597878477e-07,
      "loss": 1.0722,
      "step": 880
    },
    {
      "epoch": 0.8710545632493272,
      "grad_norm": 0.880970299243927,
      "learning_rate": 5.015386453917742e-07,
      "loss": 1.0721,
      "step": 890
    },
    {
      "epoch": 0.8808416931734768,
      "grad_norm": 0.8894344568252563,
      "learning_rate": 4.29493112422007e-07,
      "loss": 1.0691,
      "step": 900
    },
    {
      "epoch": 0.8906288230976266,
      "grad_norm": 0.8694501519203186,
      "learning_rate": 3.627998291313939e-07,
      "loss": 1.086,
      "step": 910
    },
    {
      "epoch": 0.9004159530217763,
      "grad_norm": 0.8791433572769165,
      "learning_rate": 3.015368960704584e-07,
      "loss": 1.0739,
      "step": 920
    },
    {
      "epoch": 0.9102030829459261,
      "grad_norm": 0.9409059882164001,
      "learning_rate": 2.4577605462847764e-07,
      "loss": 1.0684,
      "step": 930
    },
    {
      "epoch": 0.9199902128700759,
      "grad_norm": 0.9185978770256042,
      "learning_rate": 1.9558260302139642e-07,
      "loss": 1.0857,
      "step": 940
    },
    {
      "epoch": 0.9297773427942256,
      "grad_norm": 0.8623930811882019,
      "learning_rate": 1.510153198249531e-07,
      "loss": 1.0742,
      "step": 950
    },
    {
      "epoch": 0.9395644727183753,
      "grad_norm": 0.9032447338104248,
      "learning_rate": 1.1212639514257829e-07,
      "loss": 1.0909,
      "step": 960
    },
    {
      "epoch": 0.949351602642525,
      "grad_norm": 0.9127726554870605,
      "learning_rate": 7.896136948865429e-08,
      "loss": 1.0675,
      "step": 970
    },
    {
      "epoch": 0.9591387325666748,
      "grad_norm": 0.8940075635910034,
      "learning_rate": 5.155908045872349e-08,
      "loss": 1.0865,
      "step": 980
    },
    {
      "epoch": 0.9689258624908246,
      "grad_norm": 0.8495497703552246,
      "learning_rate": 2.995161724907658e-08,
      "loss": 1.0787,
      "step": 990
    },
    {
      "epoch": 0.9787129924149743,
      "grad_norm": 0.8723552227020264,
      "learning_rate": 1.4164283079001196e-08,
      "loss": 1.0988,
      "step": 1000
    },
    {
      "epoch": 0.9885001223391241,
      "grad_norm": 0.8567104935646057,
      "learning_rate": 4.2155655596809455e-09,
      "loss": 1.0825,
      "step": 1010
    },
    {
      "epoch": 0.9982872522632737,
      "grad_norm": 0.8691814541816711,
      "learning_rate": 1.1711504444733567e-10,
      "loss": 1.0754,
      "step": 1020
    },
    {
      "epoch": 0.9992659652556888,
      "step": 1021,
      "total_flos": 2.8896472925732864e+18,
      "train_loss": 1.1314670071662583,
      "train_runtime": 7504.4933,
      "train_samples_per_second": 17.424,
      "train_steps_per_second": 0.136
    }
  ],
  "logging_steps": 10,
  "max_steps": 1021,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.8896472925732864e+18,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}