{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.83398392652124,
  "eval_steps": 50,
  "global_step": 500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 1.112724234908819,
      "epoch": 0.018369690011481057,
      "grad_norm": 1.4453125,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.6026195526123047,
      "mean_token_accuracy": 0.8328000560402871,
      "num_tokens": 51776.0,
      "step": 5
    },
    {
      "entropy": 1.0621884115040303,
      "epoch": 0.03673938002296211,
      "grad_norm": 0.83984375,
      "learning_rate": 9e-06,
      "loss": 0.46164793968200685,
      "mean_token_accuracy": 0.8591593734920024,
      "num_tokens": 104874.0,
      "step": 10
    },
    {
      "entropy": 1.1294862307608127,
      "epoch": 0.05510907003444317,
      "grad_norm": 1.078125,
      "learning_rate": 1.4000000000000001e-05,
      "loss": 0.513739538192749,
      "mean_token_accuracy": 0.842011620849371,
      "num_tokens": 156280.0,
      "step": 15
    },
    {
      "entropy": 1.119742776453495,
      "epoch": 0.07347876004592423,
      "grad_norm": 0.48046875,
      "learning_rate": 1.9e-05,
      "loss": 0.3990218162536621,
      "mean_token_accuracy": 0.8733994543552399,
      "num_tokens": 207456.0,
      "step": 20
    },
    {
      "entropy": 1.1446169175207614,
      "epoch": 0.09184845005740529,
      "grad_norm": 0.6640625,
      "learning_rate": 2.4e-05,
      "loss": 0.36916563510894773,
      "mean_token_accuracy": 0.8824506208300591,
      "num_tokens": 255260.0,
      "step": 25
    },
    {
      "entropy": 1.1338598035275935,
      "epoch": 0.11021814006888633,
      "grad_norm": 0.640625,
      "learning_rate": 2.9e-05,
      "loss": 0.3296294927597046,
      "mean_token_accuracy": 0.8984391801059246,
      "num_tokens": 307920.0,
      "step": 30
    },
    {
      "entropy": 1.0991555109620095,
      "epoch": 0.1285878300803674,
      "grad_norm": 0.48828125,
      "learning_rate": 3.4000000000000007e-05,
      "loss": 0.3003449201583862,
      "mean_token_accuracy": 0.8993794552981853,
      "num_tokens": 360187.0,
      "step": 35
    },
    {
      "entropy": 1.0383110415190457,
      "epoch": 0.14695752009184845,
      "grad_norm": 0.38671875,
      "learning_rate": 3.9000000000000006e-05,
      "loss": 0.24448208808898925,
      "mean_token_accuracy": 0.9111124865710736,
      "num_tokens": 413522.0,
      "step": 40
    },
    {
      "entropy": 1.0760389782488347,
      "epoch": 0.1653272101033295,
      "grad_norm": 0.57421875,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.3537877321243286,
      "mean_token_accuracy": 0.8975837916135788,
      "num_tokens": 461168.0,
      "step": 45
    },
    {
      "entropy": 1.031023583188653,
      "epoch": 0.18369690011481057,
      "grad_norm": 0.5078125,
      "learning_rate": 4.9e-05,
      "loss": 0.251396632194519,
      "mean_token_accuracy": 0.9120446160435677,
      "num_tokens": 514162.0,
      "step": 50
    },
    {
      "epoch": 0.18369690011481057,
      "eval_entropy": 0.962059138391329,
      "eval_loss": 0.2778071165084839,
      "eval_mean_token_accuracy": 0.9210807476354682,
      "eval_num_tokens": 514162.0,
      "eval_runtime": 235.2955,
      "eval_samples_per_second": 0.977,
      "eval_steps_per_second": 0.977,
      "step": 50
    },
    {
      "entropy": 0.9742233645170927,
      "epoch": 0.2020665901262916,
      "grad_norm": 0.51171875,
      "learning_rate": 5.4000000000000005e-05,
      "loss": 0.23308825492858887,
      "mean_token_accuracy": 0.9161512136459351,
      "num_tokens": 567768.0,
      "step": 55
    },
    {
      "entropy": 1.006071785837412,
      "epoch": 0.22043628013777267,
      "grad_norm": 0.51171875,
      "learning_rate": 5.9e-05,
      "loss": 0.21275506019592286,
      "mean_token_accuracy": 0.9260130412876606,
      "num_tokens": 622058.0,
      "step": 60
    },
    {
      "entropy": 1.0143970489501952,
      "epoch": 0.23880597014925373,
      "grad_norm": 0.365234375,
      "learning_rate": 6.400000000000001e-05,
      "loss": 0.2733752965927124,
      "mean_token_accuracy": 0.917312602698803,
      "num_tokens": 671671.0,
      "step": 65
    },
    {
      "entropy": 0.942653215304017,
      "epoch": 0.2571756601607348,
      "grad_norm": 0.490234375,
      "learning_rate": 6.9e-05,
      "loss": 0.1964206337928772,
      "mean_token_accuracy": 0.9264056660234928,
      "num_tokens": 720265.0,
      "step": 70
    },
    {
      "entropy": 0.9703110054135322,
      "epoch": 0.2755453501722158,
      "grad_norm": 0.447265625,
      "learning_rate": 7.4e-05,
      "loss": 0.22070631980895997,
      "mean_token_accuracy": 0.9224855802953243,
      "num_tokens": 774485.0,
      "step": 75
    },
    {
      "entropy": 1.068749053031206,
      "epoch": 0.2939150401836969,
      "grad_norm": 0.6015625,
      "learning_rate": 7.900000000000001e-05,
      "loss": 0.22988514900207518,
      "mean_token_accuracy": 0.9242509074509144,
      "num_tokens": 823901.0,
      "step": 80
    },
    {
      "entropy": 1.0351632162928581,
      "epoch": 0.31228473019517794,
      "grad_norm": 0.294921875,
      "learning_rate": 8.4e-05,
      "loss": 0.20756268501281738,
      "mean_token_accuracy": 0.9241402253508568,
      "num_tokens": 878046.0,
      "step": 85
    },
    {
      "entropy": 1.0976531580090523,
      "epoch": 0.330654420206659,
      "grad_norm": 0.326171875,
      "learning_rate": 8.900000000000001e-05,
      "loss": 0.2149263858795166,
      "mean_token_accuracy": 0.9239596940577031,
      "num_tokens": 929136.0,
      "step": 90
    },
    {
      "entropy": 0.9986054591834546,
      "epoch": 0.34902411021814006,
      "grad_norm": 0.5703125,
      "learning_rate": 9.4e-05,
      "loss": 0.20242960453033448,
      "mean_token_accuracy": 0.9288195371627808,
      "num_tokens": 981017.0,
      "step": 95
    },
    {
      "entropy": 1.048951267451048,
      "epoch": 0.36739380022962115,
      "grad_norm": 0.423828125,
      "learning_rate": 9.900000000000001e-05,
      "loss": 0.23832230567932128,
      "mean_token_accuracy": 0.918797142058611,
      "num_tokens": 1031240.0,
      "step": 100
    },
    {
      "epoch": 0.36739380022962115,
      "eval_entropy": 1.050680076168931,
      "eval_loss": 0.24488472938537598,
      "eval_mean_token_accuracy": 0.92657623031865,
      "eval_num_tokens": 1031240.0,
      "eval_runtime": 235.5698,
      "eval_samples_per_second": 0.976,
      "eval_steps_per_second": 0.976,
      "step": 100
    },
    {
      "entropy": 1.11647514924407,
      "epoch": 0.3857634902411022,
      "grad_norm": 0.326171875,
      "learning_rate": 9.999236356075722e-05,
      "loss": 0.2082956314086914,
      "mean_token_accuracy": 0.9249477311968803,
      "num_tokens": 1080925.0,
      "step": 105
    },
    {
      "entropy": 1.1468311317265034,
      "epoch": 0.4041331802525832,
      "grad_norm": 0.40625,
      "learning_rate": 9.996134452407077e-05,
      "loss": 0.18167414665222167,
      "mean_token_accuracy": 0.9367795646190643,
      "num_tokens": 1135955.0,
      "step": 110
    },
    {
      "entropy": 1.1528719030320644,
      "epoch": 0.4225028702640643,
      "grad_norm": 0.36328125,
      "learning_rate": 9.990648040557267e-05,
      "loss": 0.20406742095947267,
      "mean_token_accuracy": 0.9238176062703133,
      "num_tokens": 1189197.0,
      "step": 115
    },
    {
      "entropy": 1.11902012526989,
      "epoch": 0.44087256027554533,
      "grad_norm": 0.419921875,
      "learning_rate": 9.982779739029468e-05,
      "loss": 0.21374931335449218,
      "mean_token_accuracy": 0.9228075712919235,
      "num_tokens": 1237467.0,
      "step": 120
    },
    {
      "entropy": 1.0910739868879318,
      "epoch": 0.4592422502870264,
      "grad_norm": 0.33984375,
      "learning_rate": 9.972533303132822e-05,
      "loss": 0.20214686393737794,
      "mean_token_accuracy": 0.9237420402467251,
      "num_tokens": 1288789.0,
      "step": 125
    },
    {
      "entropy": 1.0787541776895524,
      "epoch": 0.47761194029850745,
      "grad_norm": 0.400390625,
      "learning_rate": 9.959913623190127e-05,
      "loss": 0.17503974437713624,
      "mean_token_accuracy": 0.9302031256258487,
      "num_tokens": 1340899.0,
      "step": 130
    },
    {
      "entropy": 1.1493681058287621,
      "epoch": 0.49598163030998854,
      "grad_norm": 0.4765625,
      "learning_rate": 9.944926722203834e-05,
      "loss": 0.19960200786590576,
      "mean_token_accuracy": 0.93163458481431,
      "num_tokens": 1391976.0,
      "step": 135
    },
    {
      "entropy": 1.1304576590657234,
      "epoch": 0.5143513203214696,
      "grad_norm": 0.357421875,
      "learning_rate": 9.927579752981453e-05,
      "loss": 0.18575313091278076,
      "mean_token_accuracy": 0.9333173990249634,
      "num_tokens": 1445748.0,
      "step": 140
    },
    {
      "entropy": 1.2128599300980567,
      "epoch": 0.5327210103329506,
      "grad_norm": 0.33984375,
      "learning_rate": 9.907880994721714e-05,
      "loss": 0.23733038902282716,
      "mean_token_accuracy": 0.9256897248327732,
      "num_tokens": 1494377.0,
      "step": 145
    },
    {
      "entropy": 1.271282359212637,
      "epoch": 0.5510907003444316,
      "grad_norm": 0.35546875,
      "learning_rate": 9.885839849063163e-05,
      "loss": 0.21090853214263916,
      "mean_token_accuracy": 0.9234609276056289,
      "num_tokens": 1547736.0,
      "step": 150
    },
    {
      "epoch": 0.5510907003444316,
      "eval_entropy": 1.2982040446737537,
      "eval_loss": 0.23037733137607574,
      "eval_mean_token_accuracy": 0.9290674313254979,
      "eval_num_tokens": 1547736.0,
      "eval_runtime": 235.0469,
      "eval_samples_per_second": 0.979,
      "eval_steps_per_second": 0.979,
      "step": 150
    },
    {
      "entropy": 1.3101727493107318,
      "epoch": 0.5694603903559128,
      "grad_norm": 0.306640625,
      "learning_rate": 9.861466835597022e-05,
      "loss": 0.20799176692962645,
      "mean_token_accuracy": 0.9304812930524349,
      "num_tokens": 1598649.0,
      "step": 155
    },
    {
      "entropy": 1.247225718945265,
      "epoch": 0.5878300803673938,
      "grad_norm": 0.41796875,
      "learning_rate": 9.834773586846508e-05,
      "loss": 0.17320654392242432,
      "mean_token_accuracy": 0.9360279515385628,
      "num_tokens": 1648996.0,
      "step": 160
    },
    {
      "entropy": 1.1912323724478484,
      "epoch": 0.6061997703788748,
      "grad_norm": 0.296875,
      "learning_rate": 9.805772842714968e-05,
      "loss": 0.17429428100585936,
      "mean_token_accuracy": 0.9320318982005119,
      "num_tokens": 1699074.0,
      "step": 165
    },
    {
      "entropy": 1.1658439684659243,
      "epoch": 0.6245694603903559,
      "grad_norm": 0.416015625,
      "learning_rate": 9.774478444405495e-05,
      "loss": 0.19361705780029298,
      "mean_token_accuracy": 0.9293734088540078,
      "num_tokens": 1750705.0,
      "step": 170
    },
    {
      "entropy": 1.2298888713121414,
      "epoch": 0.642939150401837,
      "grad_norm": 0.34375,
      "learning_rate": 9.740905327814931e-05,
      "loss": 0.20352272987365722,
      "mean_token_accuracy": 0.9281675852835178,
      "num_tokens": 1801765.0,
      "step": 175
    },
    {
      "entropy": 1.2987987767904996,
      "epoch": 0.661308840413318,
      "grad_norm": 0.361328125,
      "learning_rate": 9.705069516405405e-05,
      "loss": 0.21955931186676025,
      "mean_token_accuracy": 0.9234688125550747,
      "num_tokens": 1852641.0,
      "step": 180
    },
    {
      "entropy": 1.2799828708171845,
      "epoch": 0.6796785304247991,
      "grad_norm": 0.302734375,
      "learning_rate": 9.666988113556798e-05,
      "loss": 0.19475481510162354,
      "mean_token_accuracy": 0.9299531005322933,
      "num_tokens": 1904888.0,
      "step": 185
    },
    {
      "entropy": 1.2942580103874206,
      "epoch": 0.6980482204362801,
      "grad_norm": 0.29296875,
      "learning_rate": 9.626679294403808e-05,
      "loss": 0.2028306245803833,
      "mean_token_accuracy": 0.9338345572352409,
      "num_tokens": 1955884.0,
      "step": 190
    },
    {
      "entropy": 1.2923168614506721,
      "epoch": 0.7164179104477612,
      "grad_norm": 0.30859375,
      "learning_rate": 9.584162297161495e-05,
      "loss": 0.2044360876083374,
      "mean_token_accuracy": 0.9323657967150212,
      "num_tokens": 2006652.0,
      "step": 195
    },
    {
      "entropy": 1.2760241843760014,
      "epoch": 0.7347876004592423,
      "grad_norm": 0.2490234375,
      "learning_rate": 9.539457413943431e-05,
      "loss": 0.17776079177856446,
      "mean_token_accuracy": 0.9367670975625515,
      "num_tokens": 2058448.0,
      "step": 200
    },
    {
      "epoch": 0.7347876004592423,
      "eval_entropy": 1.256573446030202,
      "eval_loss": 0.22272691130638123,
      "eval_mean_token_accuracy": 0.9324424870636152,
      "eval_num_tokens": 2058448.0,
      "eval_runtime": 235.7133,
      "eval_samples_per_second": 0.976,
      "eval_steps_per_second": 0.976,
      "step": 200
    },
    {
      "entropy": 1.275668701529503,
      "epoch": 0.7531572904707233,
      "grad_norm": 0.279296875,
      "learning_rate": 9.492585981076892e-05,
      "loss": 0.19257676601409912,
      "mean_token_accuracy": 0.9327104166150093,
      "num_tokens": 2110182.0,
      "step": 205
    },
    {
      "entropy": 1.3027702417224645,
      "epoch": 0.7715269804822044,
      "grad_norm": 0.294921875,
      "learning_rate": 9.44357036891964e-05,
      "loss": 0.16913862228393556,
      "mean_token_accuracy": 0.9372122779488563,
      "num_tokens": 2162768.0,
      "step": 210
    },
    {
      "entropy": 1.282259364426136,
      "epoch": 0.7898966704936854,
      "grad_norm": 0.345703125,
      "learning_rate": 9.392433971183215e-05,
      "loss": 0.1728440999984741,
      "mean_token_accuracy": 0.9397348918020725,
      "num_tokens": 2212251.0,
      "step": 215
    },
    {
      "entropy": 1.2394653480499982,
      "epoch": 0.8082663605051664,
      "grad_norm": 0.19921875,
      "learning_rate": 9.339201193767804e-05,
      "loss": 0.17258710861206056,
      "mean_token_accuracy": 0.9361227460205555,
      "num_tokens": 2264915.0,
      "step": 220
    },
    {
      "entropy": 1.2013982728123664,
      "epoch": 0.8266360505166476,
      "grad_norm": 0.2734375,
      "learning_rate": 9.283897443114016e-05,
      "loss": 0.16289238929748534,
      "mean_token_accuracy": 0.9394523158669472,
      "num_tokens": 2319219.0,
      "step": 225
    },
    {
      "entropy": 1.2622935853898525,
      "epoch": 0.8450057405281286,
      "grad_norm": 0.365234375,
      "learning_rate": 9.226549114077143e-05,
      "loss": 0.21228764057159424,
      "mean_token_accuracy": 0.9301649205386638,
      "num_tokens": 2370888.0,
      "step": 230
    },
    {
      "entropy": 1.2306364908814431,
      "epoch": 0.8633754305396096,
      "grad_norm": 0.2578125,
      "learning_rate": 9.167183577329666e-05,
      "loss": 0.16442283391952514,
      "mean_token_accuracy": 0.936205918341875,
      "num_tokens": 2425361.0,
      "step": 235
    },
    {
      "entropy": 1.2999191500246525,
      "epoch": 0.8817451205510907,
      "grad_norm": 0.2578125,
      "learning_rate": 9.105829166298037e-05,
      "loss": 0.18101825714111328,
      "mean_token_accuracy": 0.9311671845614911,
      "num_tokens": 2476449.0,
      "step": 240
    },
    {
      "entropy": 1.299282594397664,
      "epoch": 0.9001148105625718,
      "grad_norm": 0.203125,
      "learning_rate": 9.042515163639966e-05,
      "loss": 0.16907334327697754,
      "mean_token_accuracy": 0.9381237253546715,
      "num_tokens": 2533097.0,
      "step": 245
    },
    {
      "entropy": 1.2447652645409106,
      "epoch": 0.9184845005740528,
      "grad_norm": 0.2333984375,
      "learning_rate": 8.977271787268677e-05,
      "loss": 0.1682221531867981,
      "mean_token_accuracy": 0.939639289677143,
      "num_tokens": 2588343.0,
      "step": 250
    },
    {
      "epoch": 0.9184845005740528,
      "eval_entropy": 1.267150087201077,
      "eval_loss": 0.21378357708454132,
      "eval_mean_token_accuracy": 0.9356894262458967,
      "eval_num_tokens": 2588343.0,
      "eval_runtime": 235.4683,
      "eval_samples_per_second": 0.977,
      "eval_steps_per_second": 0.977,
      "step": 250
    },
    {
      "entropy": 1.2884869419038296,
      "epoch": 0.9368541905855339,
      "grad_norm": 0.337890625,
      "learning_rate": 8.910130175930773e-05,
      "loss": 0.17831521034240722,
      "mean_token_accuracy": 0.935041431337595,
      "num_tokens": 2639320.0,
      "step": 255
    },
    {
      "entropy": 1.2763290710747242,
      "epoch": 0.9552238805970149,
      "grad_norm": 0.306640625,
      "learning_rate": 8.841122374344641e-05,
      "loss": 0.1962747097015381,
      "mean_token_accuracy": 0.932129617780447,
      "num_tokens": 2690965.0,
      "step": 260
    },
    {
      "entropy": 1.2064955301582814,
      "epoch": 0.9735935706084959,
      "grad_norm": 0.232421875,
      "learning_rate": 8.770281317906435e-05,
      "loss": 0.17078809738159179,
      "mean_token_accuracy": 0.9380322515964508,
      "num_tokens": 2746178.0,
      "step": 265
    },
    {
      "entropy": 1.26394734531641,
      "epoch": 0.9919632606199771,
      "grad_norm": 0.25,
      "learning_rate": 8.697640816970992e-05,
      "loss": 0.18415504693984985,
      "mean_token_accuracy": 0.9337026454508305,
      "num_tokens": 2794347.0,
      "step": 270
    },
    {
      "entropy": 1.2361332547308794,
      "epoch": 1.0073478760045924,
      "grad_norm": 0.330078125,
      "learning_rate": 8.623235540715135e-05,
      "loss": 0.1533940076828003,
      "mean_token_accuracy": 0.94200008129006,
      "num_tokens": 2839475.0,
      "step": 275
    },
    {
      "entropy": 1.2564052909612655,
      "epoch": 1.0257175660160736,
      "grad_norm": 0.28515625,
      "learning_rate": 8.547101000591096e-05,
      "loss": 0.1843060851097107,
      "mean_token_accuracy": 0.9352358430624008,
      "num_tokens": 2890291.0,
      "step": 280
    },
    {
      "entropy": 1.14183191023767,
      "epoch": 1.0440872560275545,
      "grad_norm": 0.2392578125,
      "learning_rate": 8.469273533377944e-05,
      "loss": 0.14344700574874877,
      "mean_token_accuracy": 0.9390346795320511,
      "num_tokens": 2945309.0,
      "step": 285
    },
    {
      "entropy": 1.229520983248949,
      "epoch": 1.0624569460390356,
      "grad_norm": 0.294921875,
      "learning_rate": 8.38979028383911e-05,
      "loss": 0.1814602017402649,
      "mean_token_accuracy": 0.9284747093915939,
      "num_tokens": 2996937.0,
      "step": 290
    },
    {
      "entropy": 1.1049077600240707,
      "epoch": 1.0808266360505168,
      "grad_norm": 0.1962890625,
      "learning_rate": 8.308689186994272e-05,
      "loss": 0.1314596176147461,
      "mean_token_accuracy": 0.9438463270664215,
      "num_tokens": 3052523.0,
      "step": 295
    },
    {
      "entropy": 1.1849625088274478,
      "epoch": 1.0991963260619977,
      "grad_norm": 0.25,
      "learning_rate": 8.2260089500141e-05,
      "loss": 0.1606202483177185,
      "mean_token_accuracy": 0.9425138957798481,
      "num_tokens": 3107108.0,
      "step": 300
    },
    {
      "epoch": 1.0991963260619977,
      "eval_entropy": 1.2064458042383195,
      "eval_loss": 0.21299803256988525,
      "eval_mean_token_accuracy": 0.9329434130502784,
      "eval_num_tokens": 3107108.0,
      "eval_runtime": 235.2994,
      "eval_samples_per_second": 0.977,
      "eval_steps_per_second": 0.977,
      "step": 300
    },
    {
      "entropy": 1.226095075905323,
      "epoch": 1.1175660160734788,
      "grad_norm": 0.2109375,
      "learning_rate": 8.14178903374645e-05,
      "loss": 0.1701970100402832,
      "mean_token_accuracy": 0.9386482208967208,
      "num_tokens": 3162372.0,
      "step": 305
    },
    {
      "entropy": 1.2645030312240124,
      "epoch": 1.1359357060849598,
      "grad_norm": 0.283203125,
      "learning_rate": 8.056069633882868e-05,
      "loss": 0.18268531560897827,
      "mean_token_accuracy": 0.9372545197606087,
      "num_tokens": 3211574.0,
      "step": 310
    },
    {
      "entropy": 1.2390486162155867,
      "epoch": 1.154305396096441,
      "grad_norm": 0.326171875,
      "learning_rate": 7.968891661774357e-05,
      "loss": 0.15707379579544067,
      "mean_token_accuracy": 0.9365985207259655,
      "num_tokens": 3264778.0,
      "step": 315
    },
    {
      "entropy": 1.2300870068371297,
      "epoch": 1.172675086107922,
      "grad_norm": 0.310546875,
      "learning_rate": 7.880296724905605e-05,
      "loss": 0.14626569747924806,
      "mean_token_accuracy": 0.9456780076026916,
      "num_tokens": 3315172.0,
      "step": 320
    },
    {
      "entropy": 1.2197400435805321,
      "epoch": 1.191044776119403,
      "grad_norm": 0.25390625,
      "learning_rate": 7.790327107036943e-05,
      "loss": 0.16742393970489503,
      "mean_token_accuracy": 0.9381367027759552,
      "num_tokens": 3365850.0,
      "step": 325
    },
    {
      "entropy": 1.2276966467499733,
      "epoch": 1.209414466130884,
      "grad_norm": 0.283203125,
      "learning_rate": 7.699025748023553e-05,
      "loss": 0.1622040867805481,
      "mean_token_accuracy": 0.9411144189536571,
      "num_tokens": 3417982.0,
      "step": 330
    },
    {
      "entropy": 1.2253488399088384,
      "epoch": 1.227784156142365,
      "grad_norm": 0.302734375,
      "learning_rate": 7.60643622332154e-05,
      "loss": 0.16812814474105836,
      "mean_token_accuracy": 0.9322121366858482,
      "num_tokens": 3468705.0,
      "step": 335
    },
    {
      "entropy": 1.2625637173652648,
      "epoch": 1.2461538461538462,
      "grad_norm": 0.287109375,
      "learning_rate": 7.512602723190643e-05,
      "loss": 0.17029796838760375,
      "mean_token_accuracy": 0.9385799586772918,
      "num_tokens": 3520293.0,
      "step": 340
    },
    {
      "entropy": 1.2795540489256383,
      "epoch": 1.2645235361653273,
      "grad_norm": 0.173828125,
      "learning_rate": 7.417570031603525e-05,
      "loss": 0.1681541085243225,
      "mean_token_accuracy": 0.9350247129797935,
      "num_tokens": 3571467.0,
      "step": 345
    },
    {
      "entropy": 1.3206652700901031,
      "epoch": 1.2828932261768082,
      "grad_norm": 0.32421875,
      "learning_rate": 7.321383504871692e-05,
      "loss": 0.15741283893585206,
      "mean_token_accuracy": 0.9466758988797664,
      "num_tokens": 3621229.0,
      "step": 350
    },
    {
      "epoch": 1.2828932261768082,
      "eval_entropy": 1.2608893347823102,
      "eval_loss": 0.21446794271469116,
      "eval_mean_token_accuracy": 0.9339708304923514,
      "eval_num_tokens": 3621229.0,
      "eval_runtime": 235.2074,
      "eval_samples_per_second": 0.978,
      "eval_steps_per_second": 0.978,
      "step": 350
    },
    {
      "entropy": 1.3131505466997624,
      "epoch": 1.3012629161882894,
      "grad_norm": 0.34765625,
      "learning_rate": 7.22408904999827e-05,
      "loss": 0.17548441886901855,
      "mean_token_accuracy": 0.9359305046498776,
      "num_tokens": 3668811.0,
      "step": 355
    },
    {
      "entropy": 1.2512770250439644,
      "epoch": 1.3196326061997703,
      "grad_norm": 0.423828125,
      "learning_rate": 7.125733102767929e-05,
      "loss": 0.18522970676422118,
      "mean_token_accuracy": 0.9294097900390625,
      "num_tokens": 3719762.0,
      "step": 360
    },
    {
      "entropy": 1.224590115621686,
      "epoch": 1.3380022962112514,
      "grad_norm": 0.2734375,
      "learning_rate": 7.02636260558446e-05,
      "loss": 0.14064464569091797,
      "mean_token_accuracy": 0.9407533340156078,
      "num_tokens": 3773677.0,
      "step": 365
    },
    {
      "entropy": 1.2963713470846414,
      "epoch": 1.3563719862227326,
      "grad_norm": 0.2578125,
      "learning_rate": 6.926024985066529e-05,
      "loss": 0.173943030834198,
      "mean_token_accuracy": 0.9347950771450997,
      "num_tokens": 3823285.0,
      "step": 370
    },
    {
      "entropy": 1.2688597746193409,
      "epoch": 1.3747416762342135,
      "grad_norm": 0.216796875,
      "learning_rate": 6.824768129412361e-05,
      "loss": 0.1274775266647339,
      "mean_token_accuracy": 0.9493763312697411,
      "num_tokens": 3876999.0,
      "step": 375
    },
    {
      "entropy": 1.3514295771718026,
      "epoch": 1.3931113662456946,
      "grad_norm": 0.251953125,
      "learning_rate": 6.722640365544099e-05,
      "loss": 0.16237456798553468,
      "mean_token_accuracy": 0.9394098974764347,
      "num_tokens": 3927167.0,
      "step": 380
    },
    {
      "entropy": 1.301701857894659,
      "epoch": 1.4114810562571756,
      "grad_norm": 0.322265625,
      "learning_rate": 6.619690436042792e-05,
      "loss": 0.1615502953529358,
      "mean_token_accuracy": 0.9405923426151276,
      "num_tokens": 3979023.0,
      "step": 385
    },
    {
      "entropy": 1.252206278592348,
      "epoch": 1.4298507462686567,
      "grad_norm": 0.421875,
      "learning_rate": 6.515967475884987e-05,
      "loss": 0.18379870653152466,
      "mean_token_accuracy": 0.9334723077714443,
      "num_tokens": 4025708.0,
      "step": 390
    },
    {
      "entropy": 1.2386855103075505,
      "epoch": 1.4482204362801379,
      "grad_norm": 0.43359375,
      "learning_rate": 6.41152098899205e-05,
      "loss": 0.17962586879730225,
      "mean_token_accuracy": 0.9347062438726426,
      "num_tokens": 4076707.0,
      "step": 395
    },
    {
      "entropy": 1.1456173967570067,
      "epoch": 1.4665901262916188,
      "grad_norm": 0.267578125,
      "learning_rate": 6.3064008246034e-05,
      "loss": 0.13056639432907105,
      "mean_token_accuracy": 0.9435105569660663,
      "num_tokens": 4133447.0,
      "step": 400
    },
    {
      "epoch": 1.4665901262916188,
      "eval_entropy": 1.1722814697286357,
      "eval_loss": 0.2111370861530304,
      "eval_mean_token_accuracy": 0.9337613523006439,
      "eval_num_tokens": 4133447.0,
      "eval_runtime": 235.3086,
      "eval_samples_per_second": 0.977,
      "eval_steps_per_second": 0.977,
      "step": 400
    },
    {
      "entropy": 1.1656159494072198,
      "epoch": 1.4849598163031,
      "grad_norm": 0.2099609375,
      "learning_rate": 6.200657153484934e-05,
      "loss": 0.11656948328018188,
      "mean_token_accuracy": 0.9497443072497844,
      "num_tokens": 4187492.0,
      "step": 405
    },
    {
      "entropy": 1.206179552152753,
      "epoch": 1.5033295063145808,
      "grad_norm": 0.2421875,
      "learning_rate": 6.094340443983988e-05,
      "loss": 0.15037946701049804,
      "mean_token_accuracy": 0.9415765330195427,
      "num_tokens": 4240075.0,
      "step": 410
    },
    {
      "entropy": 1.222862060368061,
      "epoch": 1.521699196326062,
      "grad_norm": 0.275390625,
      "learning_rate": 5.987501437942282e-05,
      "loss": 0.15529404878616332,
      "mean_token_accuracy": 0.9389708809554577,
      "num_tokens": 4292551.0,
      "step": 415
    },
    {
      "entropy": 1.2309008352458477,
      "epoch": 1.5400688863375431,
      "grad_norm": 0.287109375,
      "learning_rate": 5.880191126478331e-05,
      "loss": 0.1716233730316162,
      "mean_token_accuracy": 0.9391154661774636,
      "num_tokens": 4344511.0,
      "step": 420
    },
    {
      "entropy": 1.2401153899729251,
      "epoch": 1.5584385763490243,
      "grad_norm": 0.2216796875,
      "learning_rate": 5.772460725650879e-05,
      "loss": 0.15457810163497926,
      "mean_token_accuracy": 0.9414997152984143,
      "num_tokens": 4394757.0,
      "step": 425
    },
    {
      "entropy": 1.2463274285197259,
      "epoch": 1.5768082663605052,
      "grad_norm": 0.318359375,
      "learning_rate": 5.664361652014981e-05,
      "loss": 0.15999503135681153,
      "mean_token_accuracy": 0.9394483931362629,
      "num_tokens": 4446148.0,
      "step": 430
    },
    {
      "entropy": 1.21002069786191,
      "epoch": 1.5951779563719861,
      "grad_norm": 0.37109375,
      "learning_rate": 5.555945498082391e-05,
      "loss": 0.1874553918838501,
      "mean_token_accuracy": 0.9351649850606918,
      "num_tokens": 4498541.0,
      "step": 435
    },
    {
      "entropy": 1.2327654764056206,
      "epoch": 1.6135476463834673,
      "grad_norm": 0.34375,
      "learning_rate": 5.447264007697973e-05,
      "loss": 0.16420402526855468,
      "mean_token_accuracy": 0.939295843988657,
      "num_tokens": 4548920.0,
      "step": 440
    },
    {
      "entropy": 1.2176817886531353,
      "epoch": 1.6319173363949484,
      "grad_norm": 0.263671875,
      "learning_rate": 5.3383690513438834e-05,
      "loss": 0.14642031192779542,
      "mean_token_accuracy": 0.9439929507672786,
      "num_tokens": 4603416.0,
      "step": 445
    },
    {
      "entropy": 1.2352366156876087,
      "epoch": 1.6502870264064295,
      "grad_norm": 0.31640625,
      "learning_rate": 5.2293126013833105e-05,
      "loss": 0.14479190111160278,
      "mean_token_accuracy": 0.943158283084631,
      "num_tokens": 4654316.0,
      "step": 450
    },
    {
      "epoch": 1.6502870264064295,
      "eval_entropy": 1.2289481560821118,
      "eval_loss": 0.20621241629123688,
      "eval_mean_token_accuracy": 0.9355152529218923,
      "eval_num_tokens": 4654316.0,
      "eval_runtime": 235.1441,
      "eval_samples_per_second": 0.978,
      "eval_steps_per_second": 0.978,
      "step": 450
    },
    {
      "entropy": 1.2598364517092704,
      "epoch": 1.6686567164179105,
      "grad_norm": 0.205078125,
      "learning_rate": 5.120146707255603e-05,
      "loss": 0.1650448441505432,
      "mean_token_accuracy": 0.9434025213122368,
      "num_tokens": 4706512.0,
      "step": 455
    },
    {
      "entropy": 1.260502176731825,
      "epoch": 1.6870264064293914,
      "grad_norm": 0.25390625,
      "learning_rate": 5.0109234706345895e-05,
      "loss": 0.15825694799423218,
      "mean_token_accuracy": 0.9416167691349984,
      "num_tokens": 4757604.0,
      "step": 460
    },
    {
      "entropy": 1.259766686707735,
      "epoch": 1.7053960964408725,
      "grad_norm": 0.1904296875,
      "learning_rate": 4.9016950205619896e-05,
      "loss": 0.13869506120681763,
      "mean_token_accuracy": 0.9492659881711006,
      "num_tokens": 4808517.0,
      "step": 465
    },
    {
      "entropy": 1.234454058855772,
      "epoch": 1.7237657864523537,
      "grad_norm": 0.31640625,
      "learning_rate": 4.7925134885677436e-05,
      "loss": 0.15484601259231567,
      "mean_token_accuracy": 0.9401533454656601,
      "num_tokens": 4862803.0,
      "step": 470
    },
    {
      "entropy": 1.2563340097665787,
      "epoch": 1.7421354764638348,
      "grad_norm": 0.37109375,
      "learning_rate": 4.683430983789182e-05,
      "loss": 0.18182371854782103,
      "mean_token_accuracy": 0.9343613557517528,
      "num_tokens": 4912418.0,
      "step": 475
    },
    {
      "entropy": 1.285528728365898,
      "epoch": 1.7605051664753157,
      "grad_norm": 0.3828125,
      "learning_rate": 4.574499568100843e-05,
      "loss": 0.19415501356124878,
      "mean_token_accuracy": 0.9363782249391079,
      "num_tokens": 4960195.0,
      "step": 480
    },
    {
      "entropy": 1.2490490935742855,
      "epoch": 1.7788748564867967,
      "grad_norm": 0.23828125,
      "learning_rate": 4.4657712312668915e-05,
      "loss": 0.14608356952667237,
      "mean_token_accuracy": 0.9376209229230881,
      "num_tokens": 5008503.0,
      "step": 485
    },
    {
      "entropy": 1.1854878295212985,
      "epoch": 1.7972445464982778,
      "grad_norm": 0.2412109375,
      "learning_rate": 4.357297866127917e-05,
      "loss": 0.1179052472114563,
      "mean_token_accuracy": 0.9505260616540909,
      "num_tokens": 5065315.0,
      "step": 490
    },
    {
      "entropy": 1.2593422457575798,
      "epoch": 1.815614236509759,
      "grad_norm": 0.357421875,
      "learning_rate": 4.2491312438340076e-05,
      "loss": 0.16701943874359132,
      "mean_token_accuracy": 0.9410144165158272,
      "num_tokens": 5113414.0,
      "step": 495
    },
    {
      "entropy": 1.19709093272686,
      "epoch": 1.83398392652124,
      "grad_norm": 0.1796875,
      "learning_rate": 4.1413229891359114e-05,
      "loss": 0.1293657422065735,
      "mean_token_accuracy": 0.9477272056043148,
      "num_tokens": 5165889.0,
      "step": 500
    },
    {
      "epoch": 1.83398392652124,
      "eval_entropy": 1.1967732807864313,
      "eval_loss": 0.20430247485637665,
      "eval_mean_token_accuracy": 0.9354635331941688,
      "eval_num_tokens": 5165889.0,
      "eval_runtime": 234.8347,
      "eval_samples_per_second": 0.979,
      "eval_steps_per_second": 0.979,
      "step": 500
    }
  ],
  "logging_steps": 5,
  "max_steps": 819,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 4.17590881015081e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}