{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.0,
  "eval_steps": 500,
  "global_step": 1660,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.060350030175015085,
      "grad_norm": 0.17640693485736847,
      "learning_rate": 9.599999999999999e-05,
      "loss": 1.7813,
      "mean_token_accuracy": 0.6300852990150452,
      "num_tokens": 156656.0,
      "step": 25
    },
    {
      "epoch": 0.12070006035003017,
      "grad_norm": 0.24193964898586273,
      "learning_rate": 0.00019599999999999997,
      "loss": 0.9197,
      "mean_token_accuracy": 0.768382026553154,
      "num_tokens": 282982.0,
      "step": 50
    },
    {
      "epoch": 0.18105009052504525,
      "grad_norm": 0.16499435901641846,
      "learning_rate": 0.000296,
      "loss": 0.5906,
      "mean_token_accuracy": 0.8341364151239395,
      "num_tokens": 441181.0,
      "step": 75
    },
    {
      "epoch": 0.24140012070006034,
      "grad_norm": 0.23563049733638763,
      "learning_rate": 0.0002999269005776963,
      "loss": 0.4832,
      "mean_token_accuracy": 0.8592967188358307,
      "num_tokens": 567644.0,
      "step": 100
    },
    {
      "epoch": 0.30175015087507545,
      "grad_norm": 0.22556591033935547,
      "learning_rate": 0.0002996953705789175,
      "loss": 0.3612,
      "mean_token_accuracy": 0.8925437909364701,
      "num_tokens": 725987.0,
      "step": 125
    },
    {
      "epoch": 0.3621001810500905,
      "grad_norm": 0.33429527282714844,
      "learning_rate": 0.00029930552794275785,
      "loss": 0.3126,
      "mean_token_accuracy": 0.9086851555109025,
      "num_tokens": 853185.0,
      "step": 150
    },
    {
      "epoch": 0.4224502112251056,
      "grad_norm": 0.27340370416641235,
      "learning_rate": 0.0002987577849532824,
      "loss": 0.2343,
      "mean_token_accuracy": 0.9301495373249054,
      "num_tokens": 1011232.0,
      "step": 175
    },
    {
      "epoch": 0.4828002414001207,
      "grad_norm": 0.2711191475391388,
      "learning_rate": 0.00029805272088449905,
      "loss": 0.2021,
      "mean_token_accuracy": 0.9406860828399658,
      "num_tokens": 1138074.0,
      "step": 200
    },
    {
      "epoch": 0.5431502715751357,
      "grad_norm": 0.19240038096904755,
      "learning_rate": 0.00029719108138773827,
      "loss": 0.1508,
      "mean_token_accuracy": 0.9550948125123978,
      "num_tokens": 1293601.0,
      "step": 225
    },
    {
      "epoch": 0.6035003017501509,
      "grad_norm": 0.27221739292144775,
      "learning_rate": 0.00029617377770307837,
      "loss": 0.1563,
      "mean_token_accuracy": 0.9542003554105759,
      "num_tokens": 1418074.0,
      "step": 250
    },
    {
      "epoch": 0.663850331925166,
      "grad_norm": 0.25977134704589844,
      "learning_rate": 0.0002950018856956494,
      "loss": 0.1228,
      "mean_token_accuracy": 0.9640595990419388,
      "num_tokens": 1577856.0,
      "step": 275
    },
    {
      "epoch": 0.724200362100181,
      "grad_norm": 0.2311161458492279,
      "learning_rate": 0.0002936766447178356,
      "loss": 0.1229,
      "mean_token_accuracy": 0.9646531140804291,
      "num_tokens": 1704393.0,
      "step": 300
    },
    {
      "epoch": 0.7845503922751962,
      "grad_norm": 0.1375264674425125,
      "learning_rate": 0.0002921994562985788,
      "loss": 0.0972,
      "mean_token_accuracy": 0.9722524845600128,
      "num_tokens": 1860935.0,
      "step": 325
    },
    {
      "epoch": 0.8449004224502112,
      "grad_norm": 0.2860631048679352,
      "learning_rate": 0.0002905718826611708,
      "loss": 0.0853,
      "mean_token_accuracy": 0.9756521546840667,
      "num_tokens": 1988266.0,
      "step": 350
    },
    {
      "epoch": 0.9052504526252263,
      "grad_norm": 0.11123040318489075,
      "learning_rate": 0.00028879564507109946,
      "loss": 0.0814,
      "mean_token_accuracy": 0.9769123244285584,
      "num_tokens": 2146122.0,
      "step": 375
    },
    {
      "epoch": 0.9656004828002414,
      "grad_norm": 0.239139586687088,
      "learning_rate": 0.0002868726220156981,
      "loss": 0.0696,
      "mean_token_accuracy": 0.9802538651227951,
      "num_tokens": 2273996.0,
      "step": 400
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.07211296260356903,
      "eval_mean_token_accuracy": 0.979879263285044,
      "eval_num_tokens": 2354180.0,
      "eval_runtime": 29.4819,
      "eval_samples_per_second": 12.516,
      "eval_steps_per_second": 6.275,
      "step": 415
    },
    {
      "epoch": 1.024140012070006,
      "grad_norm": 0.13852086663246155,
      "learning_rate": 0.0002848048472175225,
      "loss": 0.0764,
      "mean_token_accuracy": 0.9782544571099822,
      "num_tokens": 2422077.0,
      "step": 425
    },
    {
      "epoch": 1.0844900422450212,
      "grad_norm": 0.17753440141677856,
      "learning_rate": 0.00028259450748355637,
      "loss": 0.0527,
      "mean_token_accuracy": 0.9847306323051452,
      "num_tokens": 2564180.0,
      "step": 450
    },
    {
      "epoch": 1.1448400724200363,
      "grad_norm": 0.13473442196846008,
      "learning_rate": 0.00028024394039252005,
      "loss": 0.0697,
      "mean_token_accuracy": 0.9803690612316132,
      "num_tokens": 2705737.0,
      "step": 475
    },
    {
      "epoch": 1.2051901025950513,
      "grad_norm": 0.04998031258583069,
      "learning_rate": 0.0002777556318227281,
      "loss": 0.0452,
      "mean_token_accuracy": 0.987179564833641,
      "num_tokens": 2848782.0,
      "step": 500
    },
    {
      "epoch": 1.2655401327700664,
      "grad_norm": 0.09412259608507156,
      "learning_rate": 0.00027513221332311073,
      "loss": 0.0615,
      "mean_token_accuracy": 0.9825534737110138,
      "num_tokens": 2991024.0,
      "step": 525
    },
    {
      "epoch": 1.3258901629450814,
      "grad_norm": 0.12740211188793182,
      "learning_rate": 0.0002723764593301788,
      "loss": 0.0452,
      "mean_token_accuracy": 0.9870287185907364,
      "num_tokens": 3132622.0,
      "step": 550
    },
    {
      "epoch": 1.3862401931200965,
      "grad_norm": 0.11777028441429138,
      "learning_rate": 0.0002694912842338756,
      "loss": 0.0568,
      "mean_token_accuracy": 0.9838440799713135,
      "num_tokens": 3273143.0,
      "step": 575
    },
    {
      "epoch": 1.4465902232951118,
      "grad_norm": 0.09758122265338898,
      "learning_rate": 0.0002664797392954194,
      "loss": 0.0444,
      "mean_token_accuracy": 0.986908946633339,
      "num_tokens": 3416381.0,
      "step": 600
    },
    {
      "epoch": 1.5069402534701268,
      "grad_norm": 0.0658789575099945,
      "learning_rate": 0.0002633450094203953,
      "loss": 0.0535,
      "mean_token_accuracy": 0.9848115313053131,
      "num_tokens": 3558249.0,
      "step": 625
    },
    {
      "epoch": 1.567290283645142,
      "grad_norm": 0.07210762798786163,
      "learning_rate": 0.000260090409790509,
      "loss": 0.0409,
      "mean_token_accuracy": 0.9878959685564042,
      "num_tokens": 3700350.0,
      "step": 650
    },
    {
      "epoch": 1.627640313820157,
      "grad_norm": 0.09006072580814362,
      "learning_rate": 0.000256719382357566,
      "loss": 0.052,
      "mean_token_accuracy": 0.9852559435367584,
      "num_tokens": 3842983.0,
      "step": 675
    },
    {
      "epoch": 1.687990343995172,
      "grad_norm": 0.07121206820011139,
      "learning_rate": 0.0002532354922033823,
      "loss": 0.0401,
      "mean_token_accuracy": 0.9882262688875199,
      "num_tokens": 3985789.0,
      "step": 700
    },
    {
      "epoch": 1.748340374170187,
      "grad_norm": 0.049150578677654266,
      "learning_rate": 0.00024964242376947747,
      "loss": 0.0514,
      "mean_token_accuracy": 0.9852595126628876,
      "num_tokens": 4128158.0,
      "step": 725
    },
    {
      "epoch": 1.8086904043452021,
      "grad_norm": 0.08217272907495499,
      "learning_rate": 0.000245943976960537,
      "loss": 0.0373,
      "mean_token_accuracy": 0.9887230151891708,
      "num_tokens": 4270705.0,
      "step": 750
    },
    {
      "epoch": 1.8690404345202172,
      "grad_norm": 0.060381677001714706,
      "learning_rate": 0.00024214406312576472,
      "loss": 0.051,
      "mean_token_accuracy": 0.9850554609298706,
      "num_tokens": 4412064.0,
      "step": 775
    },
    {
      "epoch": 1.9293904646952322,
      "grad_norm": 0.07599000632762909,
      "learning_rate": 0.00023824670092237557,
      "loss": 0.0385,
      "mean_token_accuracy": 0.9883499753475189,
      "num_tokens": 4554646.0,
      "step": 800
    },
    {
      "epoch": 1.9897404948702473,
      "grad_norm": 0.06878010928630829,
      "learning_rate": 0.00023425601206560257,
      "loss": 0.0432,
      "mean_token_accuracy": 0.9873087042570114,
      "num_tokens": 4688134.0,
      "step": 825
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.04542902857065201,
      "eval_mean_token_accuracy": 0.9870176924241556,
      "eval_num_tokens": 4708360.0,
      "eval_runtime": 29.4629,
      "eval_samples_per_second": 12.524,
      "eval_steps_per_second": 6.279,
      "step": 830
    },
    {
      "epoch": 2.048280024140012,
      "grad_norm": 0.07973352819681168,
      "learning_rate": 0.00023017621696971407,
      "loss": 0.0424,
      "mean_token_accuracy": 0.9869058310371084,
      "num_tokens": 4837256.0,
      "step": 850
    },
    {
      "epoch": 2.1086300543150274,
      "grad_norm": 0.3017246723175049,
      "learning_rate": 0.0002260116302846495,
      "loss": 0.0294,
      "mean_token_accuracy": 0.9908489334583283,
      "num_tokens": 4969729.0,
      "step": 875
    },
    {
      "epoch": 2.1689800844900424,
      "grad_norm": 0.05929604917764664,
      "learning_rate": 0.0002217666563329952,
      "loss": 0.0407,
      "mean_token_accuracy": 0.9875844532251358,
      "num_tokens": 5120477.0,
      "step": 900
    },
    {
      "epoch": 2.2293301146650575,
      "grad_norm": 0.10643558949232101,
      "learning_rate": 0.00021744578445212544,
      "loss": 0.03,
      "mean_token_accuracy": 0.9906252521276474,
      "num_tokens": 5253578.0,
      "step": 925
    },
    {
      "epoch": 2.2896801448400725,
      "grad_norm": 0.07998061180114746,
      "learning_rate": 0.0002130535842464348,
      "loss": 0.0405,
      "mean_token_accuracy": 0.9873760217428207,
      "num_tokens": 5406645.0,
      "step": 950
    },
    {
      "epoch": 2.3500301750150876,
      "grad_norm": 0.03785248100757599,
      "learning_rate": 0.0002085947007546829,
      "loss": 0.0286,
      "mean_token_accuracy": 0.9912718170881272,
      "num_tokens": 5540521.0,
      "step": 975
    },
    {
      "epoch": 2.4103802051901027,
      "grad_norm": 0.038521163165569305,
      "learning_rate": 0.00020407384953756216,
      "loss": 0.0402,
      "mean_token_accuracy": 0.9876785135269165,
      "num_tokens": 5691357.0,
      "step": 1000
    },
    {
      "epoch": 2.4707302353651177,
      "grad_norm": 0.08188804239034653,
      "learning_rate": 0.00019949581169068456,
      "loss": 0.0286,
      "mean_token_accuracy": 0.991096887588501,
      "num_tokens": 5824399.0,
      "step": 1025
    },
    {
      "epoch": 2.5310802655401328,
      "grad_norm": 0.07109837234020233,
      "learning_rate": 0.0001948654287882601,
      "loss": 0.0388,
      "mean_token_accuracy": 0.9885295808315278,
      "num_tokens": 5975841.0,
      "step": 1050
    },
    {
      "epoch": 2.591430295715148,
      "grad_norm": 0.06275477260351181,
      "learning_rate": 0.00019018759776281605,
      "loss": 0.0261,
      "mean_token_accuracy": 0.9916540479660034,
      "num_tokens": 6108567.0,
      "step": 1075
    },
    {
      "epoch": 2.651780325890163,
      "grad_norm": 0.04038365185260773,
      "learning_rate": 0.00018546726572637065,
      "loss": 0.0352,
      "mean_token_accuracy": 0.9892910522222519,
      "num_tokens": 6259991.0,
      "step": 1100
    },
    {
      "epoch": 2.712130356065178,
      "grad_norm": 0.09150233864784241,
      "learning_rate": 0.00018070942473853873,
      "loss": 0.0255,
      "mean_token_accuracy": 0.9921514791250229,
      "num_tokens": 6393402.0,
      "step": 1125
    },
    {
      "epoch": 2.772480386240193,
      "grad_norm": 0.04401474818587303,
      "learning_rate": 0.00017591910652710262,
      "loss": 0.0355,
      "mean_token_accuracy": 0.9891881144046784,
      "num_tokens": 6544386.0,
      "step": 1150
    },
    {
      "epoch": 2.832830416415208,
      "grad_norm": 0.16860009729862213,
      "learning_rate": 0.00017110137716663107,
      "loss": 0.026,
      "mean_token_accuracy": 0.9918778198957443,
      "num_tokens": 6674740.0,
      "step": 1175
    },
    {
      "epoch": 2.8931804465902236,
      "grad_norm": 0.05165860429406166,
      "learning_rate": 0.0001662613317207742,
      "loss": 0.0403,
      "mean_token_accuracy": 0.9879327750205994,
      "num_tokens": 6826652.0,
      "step": 1200
    },
    {
      "epoch": 2.9535304767652386,
      "grad_norm": 0.06370134651660919,
      "learning_rate": 0.00016140408885390107,
      "loss": 0.0273,
      "mean_token_accuracy": 0.9915495270490646,
      "num_tokens": 6960481.0,
      "step": 1225
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.038739174604415894,
      "eval_mean_token_accuracy": 0.9888446437345969,
      "eval_num_tokens": 7062540.0,
      "eval_runtime": 29.4618,
      "eval_samples_per_second": 12.525,
      "eval_steps_per_second": 6.279,
      "step": 1245
    },
    {
      "epoch": 3.012070006035003,
      "grad_norm": 0.05011030286550522,
      "learning_rate": 0.0001565347854177771,
      "loss": 0.0303,
      "mean_token_accuracy": 0.9901853019429236,
      "num_tokens": 7098331.0,
      "step": 1250
    },
    {
      "epoch": 3.0724200362100182,
      "grad_norm": 0.05125705525279045,
      "learning_rate": 0.00015165857101900816,
      "loss": 0.0233,
      "mean_token_accuracy": 0.9924298238754272,
      "num_tokens": 7246298.0,
      "step": 1275
    },
    {
      "epoch": 3.1327700663850333,
      "grad_norm": 0.059397757053375244,
      "learning_rate": 0.00014678060257299454,
      "loss": 0.027,
      "mean_token_accuracy": 0.991271983385086,
      "num_tokens": 7381191.0,
      "step": 1300
    },
    {
      "epoch": 3.1931200965600484,
      "grad_norm": 0.031168634071946144,
      "learning_rate": 0.00014190603885015624,
      "loss": 0.025,
      "mean_token_accuracy": 0.9918223685026168,
      "num_tokens": 7529646.0,
      "step": 1325
    },
    {
      "epoch": 3.2534701267350634,
      "grad_norm": 0.06622699648141861,
      "learning_rate": 0.00013704003502019595,
      "loss": 0.0274,
      "mean_token_accuracy": 0.9909968906641007,
      "num_tokens": 7665384.0,
      "step": 1350
    },
    {
      "epoch": 3.3138201569100785,
      "grad_norm": 0.040858324617147446,
      "learning_rate": 0.0001321877372001702,
      "loss": 0.0234,
      "mean_token_accuracy": 0.992581251859665,
      "num_tokens": 7814867.0,
      "step": 1375
    },
    {
      "epoch": 3.3741701870850935,
      "grad_norm": 0.06217949092388153,
      "learning_rate": 0.00012735427701213444,
      "loss": 0.0264,
      "mean_token_accuracy": 0.9915199714899063,
      "num_tokens": 7951694.0,
      "step": 1400
    },
    {
      "epoch": 3.4345202172601086,
      "grad_norm": 0.044205911457538605,
      "learning_rate": 0.00012254476615611694,
      "loss": 0.0229,
      "mean_token_accuracy": 0.9923870205879212,
      "num_tokens": 8099509.0,
      "step": 1425
    },
    {
      "epoch": 3.4948702474351236,
      "grad_norm": 0.042222440242767334,
      "learning_rate": 0.00011776429100416252,
      "loss": 0.0283,
      "mean_token_accuracy": 0.9907158309221268,
      "num_tokens": 8235602.0,
      "step": 1450
    },
    {
      "epoch": 3.5552202776101387,
      "grad_norm": 0.036372531205415726,
      "learning_rate": 0.00011301790722116113,
      "loss": 0.0227,
      "mean_token_accuracy": 0.9924897265434265,
      "num_tokens": 8384445.0,
      "step": 1475
    },
    {
      "epoch": 3.6155703077851538,
      "grad_norm": 0.06498222798109055,
      "learning_rate": 0.00010831063441815225,
      "loss": 0.0254,
      "mean_token_accuracy": 0.9921332412958145,
      "num_tokens": 8520372.0,
      "step": 1500
    },
    {
      "epoch": 3.675920337960169,
      "grad_norm": 0.03929189220070839,
      "learning_rate": 0.0001036474508437579,
      "loss": 0.0221,
      "mean_token_accuracy": 0.9927822852134705,
      "num_tokens": 8668318.0,
      "step": 1525
    },
    {
      "epoch": 3.736270368135184,
      "grad_norm": 0.06582839041948318,
      "learning_rate": 9.903328811935959e-05,
      "loss": 0.0245,
      "mean_token_accuracy": 0.9920618611574173,
      "num_tokens": 8805749.0,
      "step": 1550
    },
    {
      "epoch": 3.796620398310199,
      "grad_norm": 0.04050470143556595,
      "learning_rate": 9.447302602358619e-05,
      "loss": 0.0219,
      "mean_token_accuracy": 0.9927525413036347,
      "num_tokens": 8953240.0,
      "step": 1575
    },
    {
      "epoch": 3.856970428485214,
      "grad_norm": 0.06876744329929352,
      "learning_rate": 8.997148733162942e-05,
      "loss": 0.0263,
      "mean_token_accuracy": 0.991570799946785,
      "num_tokens": 9088072.0,
      "step": 1600
    },
    {
      "epoch": 3.9173204586602295,
      "grad_norm": 0.026705719530582428,
      "learning_rate": 8.553343271484368e-05,
      "loss": 0.0226,
      "mean_token_accuracy": 0.9928545600175858,
      "num_tokens": 9236720.0,
      "step": 1625
    },
    {
      "epoch": 3.9776704888352445,
      "grad_norm": 0.03888670355081558,
      "learning_rate": 8.116355570602482e-05,
      "loss": 0.0228,
      "mean_token_accuracy": 0.9928395706415176,
      "num_tokens": 9369354.0,
      "step": 1650
    },
    {
      "epoch": 4.0,
      "eval_loss": 0.03669499605894089,
      "eval_mean_token_accuracy": 0.9898246610486829,
      "eval_num_tokens": 9416720.0,
      "eval_runtime": 29.4478,
      "eval_samples_per_second": 12.531,
      "eval_steps_per_second": 6.282,
      "step": 1660
    }
  ],
  "logging_steps": 25,
  "max_steps": 2490,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 6,
  "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": 7.992703903390188e+17,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}