{
  "best_global_step": 300,
  "best_metric": 0.1683121770620346,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 50,
  "global_step": 338,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 5.058100134134293,
      "epoch": 0.002962962962962963,
      "grad_norm": 22.178205490112305,
      "learning_rate": 0.0,
      "loss": 2.7352845668792725,
      "mean_token_accuracy": 0.6378202959895134,
      "num_tokens": 16384.0,
      "step": 1
    },
    {
      "entropy": 5.261495590209961,
      "epoch": 0.005925925925925926,
      "grad_norm": 22.13639259338379,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 2.801886558532715,
      "mean_token_accuracy": 0.6148578561842442,
      "num_tokens": 32768.0,
      "step": 2
    },
    {
      "entropy": 4.852016031742096,
      "epoch": 0.008888888888888889,
      "grad_norm": 24.148590087890625,
      "learning_rate": 4.000000000000001e-06,
      "loss": 2.492982864379883,
      "mean_token_accuracy": 0.607308741658926,
      "num_tokens": 49152.0,
      "step": 3
    },
    {
      "entropy": 4.487060964107513,
      "epoch": 0.011851851851851851,
      "grad_norm": 25.266876220703125,
      "learning_rate": 6e-06,
      "loss": 2.376826286315918,
      "mean_token_accuracy": 0.6445449814200401,
      "num_tokens": 65536.0,
      "step": 4
    },
    {
      "entropy": 4.639737457036972,
      "epoch": 0.014814814814814815,
      "grad_norm": 29.031770706176758,
      "learning_rate": 8.000000000000001e-06,
      "loss": 2.1553027629852295,
      "mean_token_accuracy": 0.6867448091506958,
      "num_tokens": 81920.0,
      "step": 5
    },
    {
      "entropy": 5.071612536907196,
      "epoch": 0.017777777777777778,
      "grad_norm": 28.10101890563965,
      "learning_rate": 1e-05,
      "loss": 2.3551838397979736,
      "mean_token_accuracy": 0.6524050980806351,
      "num_tokens": 98304.0,
      "step": 6
    },
    {
      "entropy": 4.82295498251915,
      "epoch": 0.02074074074074074,
      "grad_norm": 16.076705932617188,
      "learning_rate": 9.96996996996997e-06,
      "loss": 1.601116418838501,
      "mean_token_accuracy": 0.7177855372428894,
      "num_tokens": 114688.0,
      "step": 7
    },
    {
      "entropy": 4.47369259595871,
      "epoch": 0.023703703703703703,
      "grad_norm": 10.790403366088867,
      "learning_rate": 9.93993993993994e-06,
      "loss": 1.0933233499526978,
      "mean_token_accuracy": 0.8017240017652512,
      "num_tokens": 131072.0,
      "step": 8
    },
    {
      "entropy": 4.481221288442612,
      "epoch": 0.02666666666666667,
      "grad_norm": 7.170196533203125,
      "learning_rate": 9.90990990990991e-06,
      "loss": 1.0544596910476685,
      "mean_token_accuracy": 0.7790070474147797,
      "num_tokens": 147456.0,
      "step": 9
    },
    {
      "entropy": 4.763907015323639,
      "epoch": 0.02962962962962963,
      "grad_norm": 6.003873825073242,
      "learning_rate": 9.879879879879881e-06,
      "loss": 0.9390186071395874,
      "mean_token_accuracy": 0.8092629015445709,
      "num_tokens": 163840.0,
      "step": 10
    },
    {
      "entropy": 4.3083637952804565,
      "epoch": 0.03259259259259259,
      "grad_norm": 6.397047996520996,
      "learning_rate": 9.849849849849851e-06,
      "loss": 1.1286870241165161,
      "mean_token_accuracy": 0.8032371178269386,
      "num_tokens": 180224.0,
      "step": 11
    },
    {
      "entropy": 4.784777849912643,
      "epoch": 0.035555555555555556,
      "grad_norm": 4.655178070068359,
      "learning_rate": 9.81981981981982e-06,
      "loss": 1.00806725025177,
      "mean_token_accuracy": 0.7795897796750069,
      "num_tokens": 196608.0,
      "step": 12
    },
    {
      "entropy": 4.484914273023605,
      "epoch": 0.03851851851851852,
      "grad_norm": 2.666088819503784,
      "learning_rate": 9.78978978978979e-06,
      "loss": 0.48579731583595276,
      "mean_token_accuracy": 0.888950765132904,
      "num_tokens": 212992.0,
      "step": 13
    },
    {
      "entropy": 4.908732533454895,
      "epoch": 0.04148148148148148,
      "grad_norm": 3.5955660343170166,
      "learning_rate": 9.75975975975976e-06,
      "loss": 0.8036181926727295,
      "mean_token_accuracy": 0.8394324332475662,
      "num_tokens": 229376.0,
      "step": 14
    },
    {
      "entropy": 4.669357061386108,
      "epoch": 0.044444444444444446,
      "grad_norm": 3.167022943496704,
      "learning_rate": 9.729729729729732e-06,
      "loss": 0.5817256569862366,
      "mean_token_accuracy": 0.8610154986381531,
      "num_tokens": 245760.0,
      "step": 15
    },
    {
      "entropy": 4.967731356620789,
      "epoch": 0.047407407407407405,
      "grad_norm": 3.0831165313720703,
      "learning_rate": 9.699699699699701e-06,
      "loss": 0.6593126654624939,
      "mean_token_accuracy": 0.8428334817290306,
      "num_tokens": 262144.0,
      "step": 16
    },
    {
      "entropy": 4.689703971147537,
      "epoch": 0.05037037037037037,
      "grad_norm": 4.152491092681885,
      "learning_rate": 9.669669669669671e-06,
      "loss": 0.8403801918029785,
      "mean_token_accuracy": 0.8153181448578835,
      "num_tokens": 278528.0,
      "step": 17
    },
    {
      "entropy": 4.898822784423828,
      "epoch": 0.05333333333333334,
      "grad_norm": 2.1701176166534424,
      "learning_rate": 9.63963963963964e-06,
      "loss": 0.3927370309829712,
      "mean_token_accuracy": 0.8808837682008743,
      "num_tokens": 294912.0,
      "step": 18
    },
    {
      "entropy": 4.994785308837891,
      "epoch": 0.056296296296296296,
      "grad_norm": 2.7339329719543457,
      "learning_rate": 9.60960960960961e-06,
      "loss": 0.5464373230934143,
      "mean_token_accuracy": 0.8478362932801247,
      "num_tokens": 311296.0,
      "step": 19
    },
    {
      "entropy": 4.959144294261932,
      "epoch": 0.05925925925925926,
      "grad_norm": 2.9087443351745605,
      "learning_rate": 9.57957957957958e-06,
      "loss": 0.6418419480323792,
      "mean_token_accuracy": 0.8467107564210892,
      "num_tokens": 327680.0,
      "step": 20
    },
    {
      "entropy": 4.299957424402237,
      "epoch": 0.06222222222222222,
      "grad_norm": 1.9521111249923706,
      "learning_rate": 9.54954954954955e-06,
      "loss": 0.3496806025505066,
      "mean_token_accuracy": 0.9161608219146729,
      "num_tokens": 344064.0,
      "step": 21
    },
    {
      "entropy": 5.4237741231918335,
      "epoch": 0.06518518518518518,
      "grad_norm": 3.1053647994995117,
      "learning_rate": 9.51951951951952e-06,
      "loss": 0.6263900995254517,
      "mean_token_accuracy": 0.8462602198123932,
      "num_tokens": 360448.0,
      "step": 22
    },
    {
      "entropy": 5.040278404951096,
      "epoch": 0.06814814814814815,
      "grad_norm": 2.7072396278381348,
      "learning_rate": 9.489489489489491e-06,
      "loss": 0.6241140961647034,
      "mean_token_accuracy": 0.8455960974097252,
      "num_tokens": 376832.0,
      "step": 23
    },
    {
      "entropy": 4.836316138505936,
      "epoch": 0.07111111111111111,
      "grad_norm": 2.4444594383239746,
      "learning_rate": 9.45945945945946e-06,
      "loss": 0.5835846662521362,
      "mean_token_accuracy": 0.8459183797240257,
      "num_tokens": 393216.0,
      "step": 24
    },
    {
      "entropy": 4.762331336736679,
      "epoch": 0.07407407407407407,
      "grad_norm": 1.536238670349121,
      "learning_rate": 9.42942942942943e-06,
      "loss": 0.35127949714660645,
      "mean_token_accuracy": 0.9184967949986458,
      "num_tokens": 409600.0,
      "step": 25
    },
    {
      "entropy": 4.690429449081421,
      "epoch": 0.07703703703703704,
      "grad_norm": 2.179724931716919,
      "learning_rate": 9.3993993993994e-06,
      "loss": 0.4584190249443054,
      "mean_token_accuracy": 0.8972427546977997,
      "num_tokens": 425984.0,
      "step": 26
    },
    {
      "entropy": 5.287811815738678,
      "epoch": 0.08,
      "grad_norm": 2.173309087753296,
      "learning_rate": 9.36936936936937e-06,
      "loss": 0.5549678802490234,
      "mean_token_accuracy": 0.8874932527542114,
      "num_tokens": 442368.0,
      "step": 27
    },
    {
      "entropy": 4.725773096084595,
      "epoch": 0.08296296296296296,
      "grad_norm": 1.9286603927612305,
      "learning_rate": 9.339339339339341e-06,
      "loss": 0.5437217950820923,
      "mean_token_accuracy": 0.8805192112922668,
      "num_tokens": 458752.0,
      "step": 28
    },
    {
      "entropy": 4.341046392917633,
      "epoch": 0.08592592592592592,
      "grad_norm": 1.6609255075454712,
      "learning_rate": 9.30930930930931e-06,
      "loss": 0.3974815607070923,
      "mean_token_accuracy": 0.9058798402547836,
      "num_tokens": 475136.0,
      "step": 29
    },
    {
      "entropy": 4.40281879901886,
      "epoch": 0.08888888888888889,
      "grad_norm": 1.395637035369873,
      "learning_rate": 9.27927927927928e-06,
      "loss": 0.30734431743621826,
      "mean_token_accuracy": 0.9118703082203865,
      "num_tokens": 491520.0,
      "step": 30
    },
    {
      "entropy": 4.363550394773483,
      "epoch": 0.09185185185185185,
      "grad_norm": 1.3851248025894165,
      "learning_rate": 9.24924924924925e-06,
      "loss": 0.2637123763561249,
      "mean_token_accuracy": 0.9236166179180145,
      "num_tokens": 507904.0,
      "step": 31
    },
    {
      "entropy": 3.8239974826574326,
      "epoch": 0.09481481481481481,
      "grad_norm": 1.16873300075531,
      "learning_rate": 9.21921921921922e-06,
      "loss": 0.19853506982326508,
      "mean_token_accuracy": 0.9410291537642479,
      "num_tokens": 524288.0,
      "step": 32
    },
    {
      "entropy": 4.0603585839271545,
      "epoch": 0.09777777777777778,
      "grad_norm": 1.1862430572509766,
      "learning_rate": 9.189189189189191e-06,
      "loss": 0.23707839846611023,
      "mean_token_accuracy": 0.9392589628696442,
      "num_tokens": 540672.0,
      "step": 33
    },
    {
      "entropy": 4.260178178548813,
      "epoch": 0.10074074074074074,
      "grad_norm": 1.308255672454834,
      "learning_rate": 9.15915915915916e-06,
      "loss": 0.269561767578125,
      "mean_token_accuracy": 0.9147831574082375,
      "num_tokens": 557056.0,
      "step": 34
    },
    {
      "entropy": 4.4169876873493195,
      "epoch": 0.1037037037037037,
      "grad_norm": 1.417231559753418,
      "learning_rate": 9.129129129129129e-06,
      "loss": 0.383775919675827,
      "mean_token_accuracy": 0.9040728136897087,
      "num_tokens": 573440.0,
      "step": 35
    },
    {
      "entropy": 4.547160506248474,
      "epoch": 0.10666666666666667,
      "grad_norm": 1.5031898021697998,
      "learning_rate": 9.0990990990991e-06,
      "loss": 0.28304705023765564,
      "mean_token_accuracy": 0.9317299649119377,
      "num_tokens": 589824.0,
      "step": 36
    },
    {
      "entropy": 4.741255044937134,
      "epoch": 0.10962962962962963,
      "grad_norm": 1.4280507564544678,
      "learning_rate": 9.06906906906907e-06,
      "loss": 0.363383948802948,
      "mean_token_accuracy": 0.9032057896256447,
      "num_tokens": 606208.0,
      "step": 37
    },
    {
      "entropy": 4.389887273311615,
      "epoch": 0.11259259259259259,
      "grad_norm": 1.3031010627746582,
      "learning_rate": 9.03903903903904e-06,
      "loss": 0.33205077052116394,
      "mean_token_accuracy": 0.9147423282265663,
      "num_tokens": 622592.0,
      "step": 38
    },
    {
      "entropy": 4.335622668266296,
      "epoch": 0.11555555555555555,
      "grad_norm": 1.1646226644515991,
      "learning_rate": 9.00900900900901e-06,
      "loss": 0.22627940773963928,
      "mean_token_accuracy": 0.9445070028305054,
      "num_tokens": 638976.0,
      "step": 39
    },
    {
      "entropy": 4.044681996107101,
      "epoch": 0.11851851851851852,
      "grad_norm": 1.2125204801559448,
      "learning_rate": 8.97897897897898e-06,
      "loss": 0.2382013499736786,
      "mean_token_accuracy": 0.9404822066426277,
      "num_tokens": 655360.0,
      "step": 40
    },
    {
      "entropy": 4.5097949504852295,
      "epoch": 0.12148148148148148,
      "grad_norm": 1.4368265867233276,
      "learning_rate": 8.94894894894895e-06,
      "loss": 0.34105437994003296,
      "mean_token_accuracy": 0.9247819036245346,
      "num_tokens": 671744.0,
      "step": 41
    },
    {
      "entropy": 4.808058381080627,
      "epoch": 0.12444444444444444,
      "grad_norm": 1.8107435703277588,
      "learning_rate": 8.91891891891892e-06,
      "loss": 0.4083002805709839,
      "mean_token_accuracy": 0.9025916680693626,
      "num_tokens": 688128.0,
      "step": 42
    },
    {
      "entropy": 4.513297438621521,
      "epoch": 0.1274074074074074,
      "grad_norm": 1.0778708457946777,
      "learning_rate": 8.888888888888888e-06,
      "loss": 0.20781882107257843,
      "mean_token_accuracy": 0.9481773301959038,
      "num_tokens": 704512.0,
      "step": 43
    },
    {
      "entropy": 4.035643100738525,
      "epoch": 0.13037037037037036,
      "grad_norm": 1.1798043251037598,
      "learning_rate": 8.85885885885886e-06,
      "loss": 0.21809439361095428,
      "mean_token_accuracy": 0.9438548609614372,
      "num_tokens": 720896.0,
      "step": 44
    },
    {
      "entropy": 4.418309658765793,
      "epoch": 0.13333333333333333,
      "grad_norm": 1.6771466732025146,
      "learning_rate": 8.82882882882883e-06,
      "loss": 0.34134653210639954,
      "mean_token_accuracy": 0.9118344038724899,
      "num_tokens": 737280.0,
      "step": 45
    },
    {
      "entropy": 4.977082014083862,
      "epoch": 0.1362962962962963,
      "grad_norm": 1.55361807346344,
      "learning_rate": 8.798798798798799e-06,
      "loss": 0.3535059988498688,
      "mean_token_accuracy": 0.9039807245135307,
      "num_tokens": 753664.0,
      "step": 46
    },
    {
      "entropy": 5.195175528526306,
      "epoch": 0.13925925925925925,
      "grad_norm": 1.7871145009994507,
      "learning_rate": 8.768768768768769e-06,
      "loss": 0.4435764253139496,
      "mean_token_accuracy": 0.8920075297355652,
      "num_tokens": 770048.0,
      "step": 47
    },
    {
      "entropy": 4.436403423547745,
      "epoch": 0.14222222222222222,
      "grad_norm": 1.1365272998809814,
      "learning_rate": 8.738738738738739e-06,
      "loss": 0.1749107986688614,
      "mean_token_accuracy": 0.953681156039238,
      "num_tokens": 786432.0,
      "step": 48
    },
    {
      "entropy": 4.25896155834198,
      "epoch": 0.1451851851851852,
      "grad_norm": 1.2131942510604858,
      "learning_rate": 8.70870870870871e-06,
      "loss": 0.2522778809070587,
      "mean_token_accuracy": 0.9327027127146721,
      "num_tokens": 802816.0,
      "step": 49
    },
    {
      "entropy": 3.960051655769348,
      "epoch": 0.14814814814814814,
      "grad_norm": 0.9915124177932739,
      "learning_rate": 8.67867867867868e-06,
      "loss": 0.21660354733467102,
      "mean_token_accuracy": 0.9381572529673576,
      "num_tokens": 819200.0,
      "step": 50
    },
    {
      "epoch": 0.14814814814814814,
      "eval_entropy": 4.4217639191945395,
      "eval_loss": 0.2387709617614746,
      "eval_mean_token_accuracy": 0.9391023286183675,
      "eval_num_tokens": 819200.0,
      "eval_runtime": 41.7137,
      "eval_samples_per_second": 14.384,
      "eval_steps_per_second": 1.798,
      "step": 50
    },
    {
      "entropy": 4.788861632347107,
      "epoch": 0.1511111111111111,
      "grad_norm": 1.2990758419036865,
      "learning_rate": 8.64864864864865e-06,
      "loss": 0.3283311724662781,
      "mean_token_accuracy": 0.9205455407500267,
      "num_tokens": 835584.0,
      "step": 51
    },
    {
      "entropy": 4.751463770866394,
      "epoch": 0.15407407407407409,
      "grad_norm": 1.3297243118286133,
      "learning_rate": 8.618618618618619e-06,
      "loss": 0.27133890986442566,
      "mean_token_accuracy": 0.9276576563715935,
      "num_tokens": 851968.0,
      "step": 52
    },
    {
      "entropy": 4.121861755847931,
      "epoch": 0.15703703703703703,
      "grad_norm": 1.4608443975448608,
      "learning_rate": 8.588588588588589e-06,
      "loss": 0.3538311719894409,
      "mean_token_accuracy": 0.9215891882777214,
      "num_tokens": 868352.0,
      "step": 53
    },
    {
      "entropy": 4.774166256189346,
      "epoch": 0.16,
      "grad_norm": 1.390952467918396,
      "learning_rate": 8.55855855855856e-06,
      "loss": 0.2998553216457367,
      "mean_token_accuracy": 0.9229848235845566,
      "num_tokens": 884736.0,
      "step": 54
    },
    {
      "entropy": 3.8496759235858917,
      "epoch": 0.16296296296296298,
      "grad_norm": 0.8479620218276978,
      "learning_rate": 8.52852852852853e-06,
      "loss": 0.15498237311840057,
      "mean_token_accuracy": 0.9652328491210938,
      "num_tokens": 901120.0,
      "step": 55
    },
    {
      "entropy": 4.928109407424927,
      "epoch": 0.16592592592592592,
      "grad_norm": 1.4525344371795654,
      "learning_rate": 8.4984984984985e-06,
      "loss": 0.2703397274017334,
      "mean_token_accuracy": 0.9255108162760735,
      "num_tokens": 917504.0,
      "step": 56
    },
    {
      "entropy": 4.435289919376373,
      "epoch": 0.1688888888888889,
      "grad_norm": 1.13084876537323,
      "learning_rate": 8.46846846846847e-06,
      "loss": 0.22037354111671448,
      "mean_token_accuracy": 0.9428698942065239,
      "num_tokens": 933888.0,
      "step": 57
    },
    {
      "entropy": 4.2903048396110535,
      "epoch": 0.17185185185185184,
      "grad_norm": 1.0052417516708374,
      "learning_rate": 8.438438438438439e-06,
      "loss": 0.18087778985500336,
      "mean_token_accuracy": 0.9510272592306137,
      "num_tokens": 950272.0,
      "step": 58
    },
    {
      "entropy": 4.594562739133835,
      "epoch": 0.1748148148148148,
      "grad_norm": 1.4043190479278564,
      "learning_rate": 8.408408408408409e-06,
      "loss": 0.31700533628463745,
      "mean_token_accuracy": 0.9114612638950348,
      "num_tokens": 966656.0,
      "step": 59
    },
    {
      "entropy": 4.491921782493591,
      "epoch": 0.17777777777777778,
      "grad_norm": 1.2546675205230713,
      "learning_rate": 8.378378378378378e-06,
      "loss": 0.2658831775188446,
      "mean_token_accuracy": 0.9242468029260635,
      "num_tokens": 983040.0,
      "step": 60
    },
    {
      "entropy": 4.467200607061386,
      "epoch": 0.18074074074074073,
      "grad_norm": 1.0783617496490479,
      "learning_rate": 8.348348348348348e-06,
      "loss": 0.22270521521568298,
      "mean_token_accuracy": 0.9390252754092216,
      "num_tokens": 999424.0,
      "step": 61
    },
    {
      "entropy": 4.204399555921555,
      "epoch": 0.1837037037037037,
      "grad_norm": 1.0276278257369995,
      "learning_rate": 8.31831831831832e-06,
      "loss": 0.19260820746421814,
      "mean_token_accuracy": 0.9505967274308205,
      "num_tokens": 1015808.0,
      "step": 62
    },
    {
      "entropy": 4.730292469263077,
      "epoch": 0.18666666666666668,
      "grad_norm": 1.3187460899353027,
      "learning_rate": 8.288288288288289e-06,
      "loss": 0.3181997239589691,
      "mean_token_accuracy": 0.9214338138699532,
      "num_tokens": 1032192.0,
      "step": 63
    },
    {
      "entropy": 4.260522335767746,
      "epoch": 0.18962962962962962,
      "grad_norm": 1.072191834449768,
      "learning_rate": 8.258258258258259e-06,
      "loss": 0.22660037875175476,
      "mean_token_accuracy": 0.9363192096352577,
      "num_tokens": 1048576.0,
      "step": 64
    },
    {
      "entropy": 4.483565002679825,
      "epoch": 0.1925925925925926,
      "grad_norm": 1.2490708827972412,
      "learning_rate": 8.228228228228229e-06,
      "loss": 0.27671581506729126,
      "mean_token_accuracy": 0.925926610827446,
      "num_tokens": 1064960.0,
      "step": 65
    },
    {
      "entropy": 4.48708564043045,
      "epoch": 0.19555555555555557,
      "grad_norm": 1.2734612226486206,
      "learning_rate": 8.198198198198198e-06,
      "loss": 0.23059041798114777,
      "mean_token_accuracy": 0.9401981234550476,
      "num_tokens": 1081344.0,
      "step": 66
    },
    {
      "entropy": 4.593912869691849,
      "epoch": 0.1985185185185185,
      "grad_norm": 1.340935230255127,
      "learning_rate": 8.16816816816817e-06,
      "loss": 0.30792826414108276,
      "mean_token_accuracy": 0.9217148572206497,
      "num_tokens": 1097728.0,
      "step": 67
    },
    {
      "entropy": 4.979928016662598,
      "epoch": 0.20148148148148148,
      "grad_norm": 1.8352488279342651,
      "learning_rate": 8.13813813813814e-06,
      "loss": 0.384459912776947,
      "mean_token_accuracy": 0.9019611701369286,
      "num_tokens": 1114112.0,
      "step": 68
    },
    {
      "entropy": 4.8273065984249115,
      "epoch": 0.20444444444444446,
      "grad_norm": 1.4885120391845703,
      "learning_rate": 8.108108108108109e-06,
      "loss": 0.279024213552475,
      "mean_token_accuracy": 0.9357851520180702,
      "num_tokens": 1130496.0,
      "step": 69
    },
    {
      "entropy": 3.957000643014908,
      "epoch": 0.2074074074074074,
      "grad_norm": 0.9765417575836182,
      "learning_rate": 8.078078078078079e-06,
      "loss": 0.18877030909061432,
      "mean_token_accuracy": 0.9573989510536194,
      "num_tokens": 1146880.0,
      "step": 70
    },
    {
      "entropy": 5.035370349884033,
      "epoch": 0.21037037037037037,
      "grad_norm": 1.3658390045166016,
      "learning_rate": 8.048048048048048e-06,
      "loss": 0.2599449157714844,
      "mean_token_accuracy": 0.9303188249468803,
      "num_tokens": 1163264.0,
      "step": 71
    },
    {
      "entropy": 4.666994631290436,
      "epoch": 0.21333333333333335,
      "grad_norm": 1.433559536933899,
      "learning_rate": 8.018018018018018e-06,
      "loss": 0.2785227298736572,
      "mean_token_accuracy": 0.9423833563923836,
      "num_tokens": 1179648.0,
      "step": 72
    },
    {
      "entropy": 4.430081635713577,
      "epoch": 0.2162962962962963,
      "grad_norm": 0.9990199208259583,
      "learning_rate": 7.987987987987988e-06,
      "loss": 0.17144463956356049,
      "mean_token_accuracy": 0.9521136060357094,
      "num_tokens": 1196032.0,
      "step": 73
    },
    {
      "entropy": 4.351767390966415,
      "epoch": 0.21925925925925926,
      "grad_norm": 1.0576039552688599,
      "learning_rate": 7.95795795795796e-06,
      "loss": 0.1420559138059616,
      "mean_token_accuracy": 0.9511761665344238,
      "num_tokens": 1212416.0,
      "step": 74
    },
    {
      "entropy": 4.669542849063873,
      "epoch": 0.2222222222222222,
      "grad_norm": 1.2082196474075317,
      "learning_rate": 7.927927927927929e-06,
      "loss": 0.19001160562038422,
      "mean_token_accuracy": 0.9446141794323921,
      "num_tokens": 1228800.0,
      "step": 75
    },
    {
      "entropy": 4.577369809150696,
      "epoch": 0.22518518518518518,
      "grad_norm": 1.236482858657837,
      "learning_rate": 7.897897897897899e-06,
      "loss": 0.21950268745422363,
      "mean_token_accuracy": 0.9380715191364288,
      "num_tokens": 1245184.0,
      "step": 76
    },
    {
      "entropy": 4.402379363775253,
      "epoch": 0.22814814814814816,
      "grad_norm": 1.4677314758300781,
      "learning_rate": 7.867867867867868e-06,
      "loss": 0.31606757640838623,
      "mean_token_accuracy": 0.9232503846287727,
      "num_tokens": 1261568.0,
      "step": 77
    },
    {
      "entropy": 4.475267738103867,
      "epoch": 0.2311111111111111,
      "grad_norm": 1.2310786247253418,
      "learning_rate": 7.837837837837838e-06,
      "loss": 0.23852695524692535,
      "mean_token_accuracy": 0.9350754916667938,
      "num_tokens": 1277952.0,
      "step": 78
    },
    {
      "entropy": 4.346428841352463,
      "epoch": 0.23407407407407407,
      "grad_norm": 0.9605655670166016,
      "learning_rate": 7.807807807807808e-06,
      "loss": 0.1685204952955246,
      "mean_token_accuracy": 0.9506700709462166,
      "num_tokens": 1294336.0,
      "step": 79
    },
    {
      "entropy": 4.07252049446106,
      "epoch": 0.23703703703703705,
      "grad_norm": 1.1857283115386963,
      "learning_rate": 7.77777777777778e-06,
      "loss": 0.22819997370243073,
      "mean_token_accuracy": 0.9420149400830269,
      "num_tokens": 1310720.0,
      "step": 80
    },
    {
      "entropy": 4.396164029836655,
      "epoch": 0.24,
      "grad_norm": 1.1349984407424927,
      "learning_rate": 7.747747747747749e-06,
      "loss": 0.23652417957782745,
      "mean_token_accuracy": 0.9372891262173653,
      "num_tokens": 1327104.0,
      "step": 81
    },
    {
      "entropy": 4.5756478905677795,
      "epoch": 0.24296296296296296,
      "grad_norm": 1.2655242681503296,
      "learning_rate": 7.717717717717719e-06,
      "loss": 0.19343595206737518,
      "mean_token_accuracy": 0.9472563192248344,
      "num_tokens": 1343488.0,
      "step": 82
    },
    {
      "entropy": 4.306285113096237,
      "epoch": 0.24592592592592594,
      "grad_norm": 0.9519059658050537,
      "learning_rate": 7.687687687687688e-06,
      "loss": 0.16915273666381836,
      "mean_token_accuracy": 0.949285976588726,
      "num_tokens": 1359872.0,
      "step": 83
    },
    {
      "entropy": 4.455766886472702,
      "epoch": 0.24888888888888888,
      "grad_norm": 1.2601597309112549,
      "learning_rate": 7.657657657657658e-06,
      "loss": 0.23391816020011902,
      "mean_token_accuracy": 0.929872952401638,
      "num_tokens": 1376256.0,
      "step": 84
    },
    {
      "entropy": 4.698046267032623,
      "epoch": 0.2518518518518518,
      "grad_norm": 1.3096479177474976,
      "learning_rate": 7.6276276276276285e-06,
      "loss": 0.24509109556674957,
      "mean_token_accuracy": 0.9399605691432953,
      "num_tokens": 1392640.0,
      "step": 85
    },
    {
      "entropy": 4.206448048353195,
      "epoch": 0.2548148148148148,
      "grad_norm": 1.031250238418579,
      "learning_rate": 7.597597597597598e-06,
      "loss": 0.21270793676376343,
      "mean_token_accuracy": 0.9467888027429581,
      "num_tokens": 1409024.0,
      "step": 86
    },
    {
      "entropy": 4.716405630111694,
      "epoch": 0.2577777777777778,
      "grad_norm": 1.4683616161346436,
      "learning_rate": 7.567567567567569e-06,
      "loss": 0.3145734369754791,
      "mean_token_accuracy": 0.9134083241224289,
      "num_tokens": 1425408.0,
      "step": 87
    },
    {
      "entropy": 4.557429909706116,
      "epoch": 0.2607407407407407,
      "grad_norm": 1.3729525804519653,
      "learning_rate": 7.5375375375375385e-06,
      "loss": 0.2858428359031677,
      "mean_token_accuracy": 0.9254892840981483,
      "num_tokens": 1441792.0,
      "step": 88
    },
    {
      "entropy": 4.628496021032333,
      "epoch": 0.2637037037037037,
      "grad_norm": 1.0310251712799072,
      "learning_rate": 7.507507507507507e-06,
      "loss": 0.17122310400009155,
      "mean_token_accuracy": 0.9573537185788155,
      "num_tokens": 1458176.0,
      "step": 89
    },
    {
      "entropy": 4.517728209495544,
      "epoch": 0.26666666666666666,
      "grad_norm": 1.1787363290786743,
      "learning_rate": 7.477477477477479e-06,
      "loss": 0.23055516183376312,
      "mean_token_accuracy": 0.9414607807993889,
      "num_tokens": 1474560.0,
      "step": 90
    },
    {
      "entropy": 4.360023856163025,
      "epoch": 0.2696296296296296,
      "grad_norm": 1.1146780252456665,
      "learning_rate": 7.447447447447448e-06,
      "loss": 0.2493232786655426,
      "mean_token_accuracy": 0.9312975108623505,
      "num_tokens": 1490944.0,
      "step": 91
    },
    {
      "entropy": 4.418433308601379,
      "epoch": 0.2725925925925926,
      "grad_norm": 1.009095311164856,
      "learning_rate": 7.417417417417418e-06,
      "loss": 0.18512818217277527,
      "mean_token_accuracy": 0.9603502079844475,
      "num_tokens": 1507328.0,
      "step": 92
    },
    {
      "entropy": 4.048649460077286,
      "epoch": 0.27555555555555555,
      "grad_norm": 0.8700444102287292,
      "learning_rate": 7.387387387387388e-06,
      "loss": 0.14213146269321442,
      "mean_token_accuracy": 0.9653777033090591,
      "num_tokens": 1523712.0,
      "step": 93
    },
    {
      "entropy": 4.686308145523071,
      "epoch": 0.2785185185185185,
      "grad_norm": 1.4470664262771606,
      "learning_rate": 7.3573573573573575e-06,
      "loss": 0.3455600440502167,
      "mean_token_accuracy": 0.9082425460219383,
      "num_tokens": 1540096.0,
      "step": 94
    },
    {
      "entropy": 4.605021774768829,
      "epoch": 0.2814814814814815,
      "grad_norm": 1.116581916809082,
      "learning_rate": 7.327327327327328e-06,
      "loss": 0.2298290878534317,
      "mean_token_accuracy": 0.9413929283618927,
      "num_tokens": 1556480.0,
      "step": 95
    },
    {
      "entropy": 4.413563251495361,
      "epoch": 0.28444444444444444,
      "grad_norm": 1.2914295196533203,
      "learning_rate": 7.297297297297298e-06,
      "loss": 0.26630380749702454,
      "mean_token_accuracy": 0.9396533817052841,
      "num_tokens": 1572864.0,
      "step": 96
    },
    {
      "entropy": 4.79040065407753,
      "epoch": 0.2874074074074074,
      "grad_norm": 1.2909210920333862,
      "learning_rate": 7.267267267267268e-06,
      "loss": 0.2731528878211975,
      "mean_token_accuracy": 0.9223111271858215,
      "num_tokens": 1589248.0,
      "step": 97
    },
    {
      "entropy": 4.4953738152980804,
      "epoch": 0.2903703703703704,
      "grad_norm": 1.1834053993225098,
      "learning_rate": 7.237237237237238e-06,
      "loss": 0.22691267728805542,
      "mean_token_accuracy": 0.9453237801790237,
      "num_tokens": 1605632.0,
      "step": 98
    },
    {
      "entropy": 4.639296114444733,
      "epoch": 0.29333333333333333,
      "grad_norm": 1.3944916725158691,
      "learning_rate": 7.207207207207208e-06,
      "loss": 0.3488730788230896,
      "mean_token_accuracy": 0.9018916934728622,
      "num_tokens": 1622016.0,
      "step": 99
    },
    {
      "entropy": 4.078823536634445,
      "epoch": 0.2962962962962963,
      "grad_norm": 1.0764650106430054,
      "learning_rate": 7.177177177177178e-06,
      "loss": 0.22934724390506744,
      "mean_token_accuracy": 0.9346700236201286,
      "num_tokens": 1638400.0,
      "step": 100
    },
    {
      "epoch": 0.2962962962962963,
      "eval_entropy": 4.461662375132243,
      "eval_loss": 0.19545181095600128,
      "eval_mean_token_accuracy": 0.9480786522229513,
      "eval_num_tokens": 1638400.0,
      "eval_runtime": 41.708,
      "eval_samples_per_second": 14.386,
      "eval_steps_per_second": 1.798,
      "step": 100
    },
    {
      "entropy": 4.814678251743317,
      "epoch": 0.2992592592592593,
      "grad_norm": 1.0189930200576782,
      "learning_rate": 7.147147147147148e-06,
      "loss": 0.14823393523693085,
      "mean_token_accuracy": 0.9615615606307983,
      "num_tokens": 1654784.0,
      "step": 101
    },
    {
      "entropy": 4.98028501868248,
      "epoch": 0.3022222222222222,
      "grad_norm": 1.2111811637878418,
      "learning_rate": 7.117117117117117e-06,
      "loss": 0.2203434854745865,
      "mean_token_accuracy": 0.9475290551781654,
      "num_tokens": 1671168.0,
      "step": 102
    },
    {
      "entropy": 4.726410657167435,
      "epoch": 0.30518518518518517,
      "grad_norm": 1.4401518106460571,
      "learning_rate": 7.087087087087087e-06,
      "loss": 0.33535507321357727,
      "mean_token_accuracy": 0.9216739237308502,
      "num_tokens": 1687552.0,
      "step": 103
    },
    {
      "entropy": 4.530529856681824,
      "epoch": 0.30814814814814817,
      "grad_norm": 1.0770063400268555,
      "learning_rate": 7.057057057057057e-06,
      "loss": 0.1991444081068039,
      "mean_token_accuracy": 0.9454346299171448,
      "num_tokens": 1703936.0,
      "step": 104
    },
    {
      "entropy": 4.670209527015686,
      "epoch": 0.3111111111111111,
      "grad_norm": 1.15728759765625,
      "learning_rate": 7.027027027027028e-06,
      "loss": 0.2929231822490692,
      "mean_token_accuracy": 0.9210405126214027,
      "num_tokens": 1720320.0,
      "step": 105
    },
    {
      "entropy": 4.4161136746406555,
      "epoch": 0.31407407407407406,
      "grad_norm": 0.8282631039619446,
      "learning_rate": 6.996996996996997e-06,
      "loss": 0.1592385619878769,
      "mean_token_accuracy": 0.9598511978983879,
      "num_tokens": 1736704.0,
      "step": 106
    },
    {
      "entropy": 4.86144083738327,
      "epoch": 0.31703703703703706,
      "grad_norm": 1.263832926750183,
      "learning_rate": 6.966966966966967e-06,
      "loss": 0.19089697301387787,
      "mean_token_accuracy": 0.9483430981636047,
      "num_tokens": 1753088.0,
      "step": 107
    },
    {
      "entropy": 4.153488516807556,
      "epoch": 0.32,
      "grad_norm": 1.0922951698303223,
      "learning_rate": 6.936936936936938e-06,
      "loss": 0.2128307819366455,
      "mean_token_accuracy": 0.9392627403140068,
      "num_tokens": 1769472.0,
      "step": 108
    },
    {
      "entropy": 4.55321004986763,
      "epoch": 0.32296296296296295,
      "grad_norm": 1.2877445220947266,
      "learning_rate": 6.906906906906907e-06,
      "loss": 0.25118085741996765,
      "mean_token_accuracy": 0.9339887574315071,
      "num_tokens": 1785856.0,
      "step": 109
    },
    {
      "entropy": 4.488452315330505,
      "epoch": 0.32592592592592595,
      "grad_norm": 1.0957287549972534,
      "learning_rate": 6.876876876876878e-06,
      "loss": 0.22158771753311157,
      "mean_token_accuracy": 0.934796467423439,
      "num_tokens": 1802240.0,
      "step": 110
    },
    {
      "entropy": 4.805920362472534,
      "epoch": 0.3288888888888889,
      "grad_norm": 1.364698886871338,
      "learning_rate": 6.846846846846848e-06,
      "loss": 0.29956161975860596,
      "mean_token_accuracy": 0.9315352365374565,
      "num_tokens": 1818624.0,
      "step": 111
    },
    {
      "entropy": 4.508695632219315,
      "epoch": 0.33185185185185184,
      "grad_norm": 1.1932005882263184,
      "learning_rate": 6.816816816816817e-06,
      "loss": 0.2309732735157013,
      "mean_token_accuracy": 0.9365206658840179,
      "num_tokens": 1835008.0,
      "step": 112
    },
    {
      "entropy": 4.294718265533447,
      "epoch": 0.3348148148148148,
      "grad_norm": 1.1498323678970337,
      "learning_rate": 6.786786786786788e-06,
      "loss": 0.19367162883281708,
      "mean_token_accuracy": 0.9481090977787971,
      "num_tokens": 1851392.0,
      "step": 113
    },
    {
      "entropy": 4.655509650707245,
      "epoch": 0.3377777777777778,
      "grad_norm": 1.2988401651382446,
      "learning_rate": 6.7567567567567575e-06,
      "loss": 0.27142661809921265,
      "mean_token_accuracy": 0.9328557252883911,
      "num_tokens": 1867776.0,
      "step": 114
    },
    {
      "entropy": 4.659626722335815,
      "epoch": 0.34074074074074073,
      "grad_norm": 1.1721936464309692,
      "learning_rate": 6.726726726726728e-06,
      "loss": 0.20252208411693573,
      "mean_token_accuracy": 0.9363461658358574,
      "num_tokens": 1884160.0,
      "step": 115
    },
    {
      "entropy": 5.089058935642242,
      "epoch": 0.3437037037037037,
      "grad_norm": 1.9151562452316284,
      "learning_rate": 6.696696696696697e-06,
      "loss": 0.3615419268608093,
      "mean_token_accuracy": 0.9023440852761269,
      "num_tokens": 1900544.0,
      "step": 116
    },
    {
      "entropy": 4.381918042898178,
      "epoch": 0.3466666666666667,
      "grad_norm": 1.1667546033859253,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.19516335427761078,
      "mean_token_accuracy": 0.9462241530418396,
      "num_tokens": 1916928.0,
      "step": 117
    },
    {
      "entropy": 4.843257635831833,
      "epoch": 0.3496296296296296,
      "grad_norm": 1.1661940813064575,
      "learning_rate": 6.636636636636637e-06,
      "loss": 0.283145934343338,
      "mean_token_accuracy": 0.9199336618185043,
      "num_tokens": 1933312.0,
      "step": 118
    },
    {
      "entropy": 4.623775035142899,
      "epoch": 0.35259259259259257,
      "grad_norm": 1.254604697227478,
      "learning_rate": 6.606606606606607e-06,
      "loss": 0.2810809910297394,
      "mean_token_accuracy": 0.925496369600296,
      "num_tokens": 1949696.0,
      "step": 119
    },
    {
      "entropy": 4.221776694059372,
      "epoch": 0.35555555555555557,
      "grad_norm": 1.0797185897827148,
      "learning_rate": 6.5765765765765775e-06,
      "loss": 0.2070639282464981,
      "mean_token_accuracy": 0.9441968277096748,
      "num_tokens": 1966080.0,
      "step": 120
    },
    {
      "entropy": 4.430790603160858,
      "epoch": 0.3585185185185185,
      "grad_norm": 1.0618715286254883,
      "learning_rate": 6.546546546546547e-06,
      "loss": 0.22707627713680267,
      "mean_token_accuracy": 0.9264308586716652,
      "num_tokens": 1982464.0,
      "step": 121
    },
    {
      "entropy": 4.429312914609909,
      "epoch": 0.36148148148148146,
      "grad_norm": 1.106501579284668,
      "learning_rate": 6.516516516516517e-06,
      "loss": 0.20258170366287231,
      "mean_token_accuracy": 0.9558736011385918,
      "num_tokens": 1998848.0,
      "step": 122
    },
    {
      "entropy": 4.722253382205963,
      "epoch": 0.36444444444444446,
      "grad_norm": 1.2203396558761597,
      "learning_rate": 6.486486486486487e-06,
      "loss": 0.2102155089378357,
      "mean_token_accuracy": 0.946428619325161,
      "num_tokens": 2015232.0,
      "step": 123
    },
    {
      "entropy": 5.0044931173324585,
      "epoch": 0.3674074074074074,
      "grad_norm": 1.4379228353500366,
      "learning_rate": 6.456456456456457e-06,
      "loss": 0.3243516683578491,
      "mean_token_accuracy": 0.9254569783806801,
      "num_tokens": 2031616.0,
      "step": 124
    },
    {
      "entropy": 4.44594606757164,
      "epoch": 0.37037037037037035,
      "grad_norm": 1.2764923572540283,
      "learning_rate": 6.426426426426427e-06,
      "loss": 0.22515040636062622,
      "mean_token_accuracy": 0.9395722597837448,
      "num_tokens": 2048000.0,
      "step": 125
    },
    {
      "entropy": 4.9530656933784485,
      "epoch": 0.37333333333333335,
      "grad_norm": 1.3866773843765259,
      "learning_rate": 6.396396396396397e-06,
      "loss": 0.2734587490558624,
      "mean_token_accuracy": 0.9305318221449852,
      "num_tokens": 2064384.0,
      "step": 126
    },
    {
      "entropy": 4.404450595378876,
      "epoch": 0.3762962962962963,
      "grad_norm": 1.0615060329437256,
      "learning_rate": 6.366366366366366e-06,
      "loss": 0.18650338053703308,
      "mean_token_accuracy": 0.9417764097452164,
      "num_tokens": 2080768.0,
      "step": 127
    },
    {
      "entropy": 4.358800292015076,
      "epoch": 0.37925925925925924,
      "grad_norm": 0.9904522895812988,
      "learning_rate": 6.336336336336338e-06,
      "loss": 0.1557006686925888,
      "mean_token_accuracy": 0.9499502927064896,
      "num_tokens": 2097152.0,
      "step": 128
    },
    {
      "entropy": 4.347220629453659,
      "epoch": 0.38222222222222224,
      "grad_norm": 1.1018481254577637,
      "learning_rate": 6.3063063063063065e-06,
      "loss": 0.17383213341236115,
      "mean_token_accuracy": 0.9520699456334114,
      "num_tokens": 2113536.0,
      "step": 129
    },
    {
      "entropy": 4.678435325622559,
      "epoch": 0.3851851851851852,
      "grad_norm": 1.077415943145752,
      "learning_rate": 6.276276276276276e-06,
      "loss": 0.1535685956478119,
      "mean_token_accuracy": 0.9551054239273071,
      "num_tokens": 2129920.0,
      "step": 130
    },
    {
      "entropy": 4.648881524801254,
      "epoch": 0.38814814814814813,
      "grad_norm": 1.2710124254226685,
      "learning_rate": 6.246246246246247e-06,
      "loss": 0.2373301386833191,
      "mean_token_accuracy": 0.9301018118858337,
      "num_tokens": 2146304.0,
      "step": 131
    },
    {
      "entropy": 4.318471074104309,
      "epoch": 0.39111111111111113,
      "grad_norm": 1.0050504207611084,
      "learning_rate": 6.2162162162162164e-06,
      "loss": 0.18840913474559784,
      "mean_token_accuracy": 0.9467770904302597,
      "num_tokens": 2162688.0,
      "step": 132
    },
    {
      "entropy": 4.847321391105652,
      "epoch": 0.3940740740740741,
      "grad_norm": 1.339849829673767,
      "learning_rate": 6.186186186186187e-06,
      "loss": 0.27330562472343445,
      "mean_token_accuracy": 0.9201410859823227,
      "num_tokens": 2179072.0,
      "step": 133
    },
    {
      "entropy": 4.795216828584671,
      "epoch": 0.397037037037037,
      "grad_norm": 1.403182864189148,
      "learning_rate": 6.156156156156157e-06,
      "loss": 0.27707362174987793,
      "mean_token_accuracy": 0.9330373629927635,
      "num_tokens": 2195456.0,
      "step": 134
    },
    {
      "entropy": 4.491364538669586,
      "epoch": 0.4,
      "grad_norm": 1.0648137331008911,
      "learning_rate": 6.126126126126126e-06,
      "loss": 0.19971750676631927,
      "mean_token_accuracy": 0.9492718502879143,
      "num_tokens": 2211840.0,
      "step": 135
    },
    {
      "entropy": 4.5878481566905975,
      "epoch": 0.40296296296296297,
      "grad_norm": 1.0499775409698486,
      "learning_rate": 6.096096096096097e-06,
      "loss": 0.16306909918785095,
      "mean_token_accuracy": 0.9506752043962479,
      "num_tokens": 2228224.0,
      "step": 136
    },
    {
      "entropy": 4.106632500886917,
      "epoch": 0.4059259259259259,
      "grad_norm": 0.9085140228271484,
      "learning_rate": 6.066066066066067e-06,
      "loss": 0.11960416287183762,
      "mean_token_accuracy": 0.965574748814106,
      "num_tokens": 2244608.0,
      "step": 137
    },
    {
      "entropy": 4.460581988096237,
      "epoch": 0.4088888888888889,
      "grad_norm": 1.2903225421905518,
      "learning_rate": 6.036036036036037e-06,
      "loss": 0.22525092959403992,
      "mean_token_accuracy": 0.9429330825805664,
      "num_tokens": 2260992.0,
      "step": 138
    },
    {
      "entropy": 4.3643467128276825,
      "epoch": 0.41185185185185186,
      "grad_norm": 1.2166695594787598,
      "learning_rate": 6.006006006006007e-06,
      "loss": 0.2448989450931549,
      "mean_token_accuracy": 0.9346435517072678,
      "num_tokens": 2277376.0,
      "step": 139
    },
    {
      "entropy": 4.949205368757248,
      "epoch": 0.4148148148148148,
      "grad_norm": 1.4846951961517334,
      "learning_rate": 5.975975975975976e-06,
      "loss": 0.3188098073005676,
      "mean_token_accuracy": 0.9188738688826561,
      "num_tokens": 2293760.0,
      "step": 140
    },
    {
      "entropy": 4.4434521198272705,
      "epoch": 0.4177777777777778,
      "grad_norm": 1.1468238830566406,
      "learning_rate": 5.945945945945947e-06,
      "loss": 0.25719913840293884,
      "mean_token_accuracy": 0.9296797513961792,
      "num_tokens": 2310144.0,
      "step": 141
    },
    {
      "entropy": 4.3049706518650055,
      "epoch": 0.42074074074074075,
      "grad_norm": 1.0389187335968018,
      "learning_rate": 5.915915915915916e-06,
      "loss": 0.16033510863780975,
      "mean_token_accuracy": 0.954895943403244,
      "num_tokens": 2326528.0,
      "step": 142
    },
    {
      "entropy": 4.707719027996063,
      "epoch": 0.4237037037037037,
      "grad_norm": 1.4716471433639526,
      "learning_rate": 5.885885885885886e-06,
      "loss": 0.3065488338470459,
      "mean_token_accuracy": 0.9253052920103073,
      "num_tokens": 2342912.0,
      "step": 143
    },
    {
      "entropy": 4.36312934756279,
      "epoch": 0.4266666666666667,
      "grad_norm": 1.1793159246444702,
      "learning_rate": 5.855855855855856e-06,
      "loss": 0.22046490013599396,
      "mean_token_accuracy": 0.9440506026148796,
      "num_tokens": 2359296.0,
      "step": 144
    },
    {
      "entropy": 4.659156382083893,
      "epoch": 0.42962962962962964,
      "grad_norm": 0.9858176112174988,
      "learning_rate": 5.825825825825826e-06,
      "loss": 0.15407533943653107,
      "mean_token_accuracy": 0.9536504298448563,
      "num_tokens": 2375680.0,
      "step": 145
    },
    {
      "entropy": 3.868691772222519,
      "epoch": 0.4325925925925926,
      "grad_norm": 0.8139008283615112,
      "learning_rate": 5.7957957957957965e-06,
      "loss": 0.11436738818883896,
      "mean_token_accuracy": 0.9670137017965317,
      "num_tokens": 2392064.0,
      "step": 146
    },
    {
      "entropy": 4.071219593286514,
      "epoch": 0.43555555555555553,
      "grad_norm": 1.0085606575012207,
      "learning_rate": 5.765765765765766e-06,
      "loss": 0.163216695189476,
      "mean_token_accuracy": 0.962848387658596,
      "num_tokens": 2408448.0,
      "step": 147
    },
    {
      "entropy": 4.502195358276367,
      "epoch": 0.43851851851851853,
      "grad_norm": 1.225859522819519,
      "learning_rate": 5.735735735735736e-06,
      "loss": 0.20846611261367798,
      "mean_token_accuracy": 0.9376749396324158,
      "num_tokens": 2424832.0,
      "step": 148
    },
    {
      "entropy": 4.737431079149246,
      "epoch": 0.4414814814814815,
      "grad_norm": 1.1979032754898071,
      "learning_rate": 5.7057057057057065e-06,
      "loss": 0.23248976469039917,
      "mean_token_accuracy": 0.9266073629260063,
      "num_tokens": 2441216.0,
      "step": 149
    },
    {
      "entropy": 4.504071295261383,
      "epoch": 0.4444444444444444,
      "grad_norm": 1.3558123111724854,
      "learning_rate": 5.675675675675676e-06,
      "loss": 0.27378231287002563,
      "mean_token_accuracy": 0.9267841130495071,
      "num_tokens": 2457600.0,
      "step": 150
    },
    {
      "epoch": 0.4444444444444444,
      "eval_entropy": 4.427574087778727,
      "eval_loss": 0.1820228397846222,
      "eval_mean_token_accuracy": 0.9505827856063843,
      "eval_num_tokens": 2457600.0,
      "eval_runtime": 41.7241,
      "eval_samples_per_second": 14.38,
      "eval_steps_per_second": 1.798,
      "step": 150
    },
    {
      "entropy": 4.998361647129059,
      "epoch": 0.4474074074074074,
      "grad_norm": 1.4872044324874878,
      "learning_rate": 5.645645645645647e-06,
      "loss": 0.2821654677391052,
      "mean_token_accuracy": 0.928920067846775,
      "num_tokens": 2473984.0,
      "step": 151
    },
    {
      "entropy": 4.445061385631561,
      "epoch": 0.45037037037037037,
      "grad_norm": 1.0061935186386108,
      "learning_rate": 5.615615615615616e-06,
      "loss": 0.16303198039531708,
      "mean_token_accuracy": 0.9549186900258064,
      "num_tokens": 2490368.0,
      "step": 152
    },
    {
      "entropy": 4.636731952428818,
      "epoch": 0.4533333333333333,
      "grad_norm": 1.3043243885040283,
      "learning_rate": 5.585585585585585e-06,
      "loss": 0.21270065009593964,
      "mean_token_accuracy": 0.9298705905675888,
      "num_tokens": 2506752.0,
      "step": 153
    },
    {
      "entropy": 4.824311137199402,
      "epoch": 0.4562962962962963,
      "grad_norm": 1.2768608331680298,
      "learning_rate": 5.555555555555557e-06,
      "loss": 0.20686385035514832,
      "mean_token_accuracy": 0.9410912245512009,
      "num_tokens": 2523136.0,
      "step": 154
    },
    {
      "entropy": 4.417184978723526,
      "epoch": 0.45925925925925926,
      "grad_norm": 0.9344239234924316,
      "learning_rate": 5.5255255255255255e-06,
      "loss": 0.12936189770698547,
      "mean_token_accuracy": 0.9643698781728745,
      "num_tokens": 2539520.0,
      "step": 155
    },
    {
      "entropy": 3.8669098168611526,
      "epoch": 0.4622222222222222,
      "grad_norm": 0.7095533013343811,
      "learning_rate": 5.495495495495496e-06,
      "loss": 0.06662072241306305,
      "mean_token_accuracy": 0.9827630147337914,
      "num_tokens": 2555904.0,
      "step": 156
    },
    {
      "entropy": 4.7329277992248535,
      "epoch": 0.4651851851851852,
      "grad_norm": 1.485074758529663,
      "learning_rate": 5.465465465465466e-06,
      "loss": 0.2506575286388397,
      "mean_token_accuracy": 0.9287298172712326,
      "num_tokens": 2572288.0,
      "step": 157
    },
    {
      "entropy": 4.442395329475403,
      "epoch": 0.46814814814814815,
      "grad_norm": 0.875119686126709,
      "learning_rate": 5.4354354354354355e-06,
      "loss": 0.15853841602802277,
      "mean_token_accuracy": 0.9585354551672935,
      "num_tokens": 2588672.0,
      "step": 158
    },
    {
      "entropy": 4.53130978345871,
      "epoch": 0.4711111111111111,
      "grad_norm": 1.1801632642745972,
      "learning_rate": 5.405405405405406e-06,
      "loss": 0.2645067870616913,
      "mean_token_accuracy": 0.9257083982229233,
      "num_tokens": 2605056.0,
      "step": 159
    },
    {
      "entropy": 4.549875617027283,
      "epoch": 0.4740740740740741,
      "grad_norm": 1.2281057834625244,
      "learning_rate": 5.375375375375376e-06,
      "loss": 0.16900277137756348,
      "mean_token_accuracy": 0.959097295999527,
      "num_tokens": 2621440.0,
      "step": 160
    },
    {
      "entropy": 4.628339350223541,
      "epoch": 0.47703703703703704,
      "grad_norm": 1.3054790496826172,
      "learning_rate": 5.345345345345346e-06,
      "loss": 0.2433934360742569,
      "mean_token_accuracy": 0.9287422746419907,
      "num_tokens": 2637824.0,
      "step": 161
    },
    {
      "entropy": 4.346887558698654,
      "epoch": 0.48,
      "grad_norm": 0.9419753551483154,
      "learning_rate": 5.315315315315316e-06,
      "loss": 0.13523709774017334,
      "mean_token_accuracy": 0.9573786482214928,
      "num_tokens": 2654208.0,
      "step": 162
    },
    {
      "entropy": 4.233735829591751,
      "epoch": 0.482962962962963,
      "grad_norm": 1.1294695138931274,
      "learning_rate": 5.285285285285286e-06,
      "loss": 0.2086874544620514,
      "mean_token_accuracy": 0.9446921944618225,
      "num_tokens": 2670592.0,
      "step": 163
    },
    {
      "entropy": 4.735133767127991,
      "epoch": 0.48592592592592593,
      "grad_norm": 1.1127125024795532,
      "learning_rate": 5.255255255255256e-06,
      "loss": 0.13167983293533325,
      "mean_token_accuracy": 0.9564699530601501,
      "num_tokens": 2686976.0,
      "step": 164
    },
    {
      "entropy": 4.497480601072311,
      "epoch": 0.4888888888888889,
      "grad_norm": 1.060238242149353,
      "learning_rate": 5.225225225225226e-06,
      "loss": 0.2078431397676468,
      "mean_token_accuracy": 0.9467749446630478,
      "num_tokens": 2703360.0,
      "step": 165
    },
    {
      "entropy": 4.46263587474823,
      "epoch": 0.4918518518518519,
      "grad_norm": 0.9650236368179321,
      "learning_rate": 5.195195195195195e-06,
      "loss": 0.15210066735744476,
      "mean_token_accuracy": 0.9565599709749222,
      "num_tokens": 2719744.0,
      "step": 166
    },
    {
      "entropy": 4.654350996017456,
      "epoch": 0.4948148148148148,
      "grad_norm": 1.076202154159546,
      "learning_rate": 5.165165165165165e-06,
      "loss": 0.174457848072052,
      "mean_token_accuracy": 0.9490634500980377,
      "num_tokens": 2736128.0,
      "step": 167
    },
    {
      "entropy": 5.074929893016815,
      "epoch": 0.49777777777777776,
      "grad_norm": 1.5211771726608276,
      "learning_rate": 5.135135135135135e-06,
      "loss": 0.34576427936553955,
      "mean_token_accuracy": 0.9180677607655525,
      "num_tokens": 2752512.0,
      "step": 168
    },
    {
      "entropy": 4.191278040409088,
      "epoch": 0.5007407407407407,
      "grad_norm": 1.2762339115142822,
      "learning_rate": 5.105105105105106e-06,
      "loss": 0.2191346287727356,
      "mean_token_accuracy": 0.9453108981251717,
      "num_tokens": 2768896.0,
      "step": 169
    },
    {
      "entropy": 4.65677535533905,
      "epoch": 0.5037037037037037,
      "grad_norm": 1.3585889339447021,
      "learning_rate": 5.075075075075075e-06,
      "loss": 0.25393176078796387,
      "mean_token_accuracy": 0.9185363203287125,
      "num_tokens": 2785280.0,
      "step": 170
    },
    {
      "entropy": 4.767172336578369,
      "epoch": 0.5066666666666667,
      "grad_norm": 1.282403588294983,
      "learning_rate": 5.045045045045045e-06,
      "loss": 0.23821371793746948,
      "mean_token_accuracy": 0.9363728985190392,
      "num_tokens": 2801664.0,
      "step": 171
    },
    {
      "entropy": 4.913634359836578,
      "epoch": 0.5096296296296297,
      "grad_norm": 1.370779037475586,
      "learning_rate": 5.0150150150150156e-06,
      "loss": 0.26560261845588684,
      "mean_token_accuracy": 0.9290853068232536,
      "num_tokens": 2818048.0,
      "step": 172
    },
    {
      "entropy": 4.962270259857178,
      "epoch": 0.5125925925925926,
      "grad_norm": 1.2277225255966187,
      "learning_rate": 4.984984984984985e-06,
      "loss": 0.2335178256034851,
      "mean_token_accuracy": 0.9330927804112434,
      "num_tokens": 2834432.0,
      "step": 173
    },
    {
      "entropy": 4.958651542663574,
      "epoch": 0.5155555555555555,
      "grad_norm": 1.1668320894241333,
      "learning_rate": 4.954954954954955e-06,
      "loss": 0.22480204701423645,
      "mean_token_accuracy": 0.9291108474135399,
      "num_tokens": 2850816.0,
      "step": 174
    },
    {
      "entropy": 4.368990540504456,
      "epoch": 0.5185185185185185,
      "grad_norm": 1.2106925249099731,
      "learning_rate": 4.9249249249249255e-06,
      "loss": 0.20308136940002441,
      "mean_token_accuracy": 0.9393906965851784,
      "num_tokens": 2867200.0,
      "step": 175
    },
    {
      "entropy": 4.659591734409332,
      "epoch": 0.5214814814814814,
      "grad_norm": 1.3593008518218994,
      "learning_rate": 4.894894894894895e-06,
      "loss": 0.24384461343288422,
      "mean_token_accuracy": 0.9311445951461792,
      "num_tokens": 2883584.0,
      "step": 176
    },
    {
      "entropy": 4.498990893363953,
      "epoch": 0.5244444444444445,
      "grad_norm": 0.914821207523346,
      "learning_rate": 4.864864864864866e-06,
      "loss": 0.13378006219863892,
      "mean_token_accuracy": 0.9584524184465408,
      "num_tokens": 2899968.0,
      "step": 177
    },
    {
      "entropy": 4.774127125740051,
      "epoch": 0.5274074074074074,
      "grad_norm": 1.1247864961624146,
      "learning_rate": 4.8348348348348355e-06,
      "loss": 0.20536412298679352,
      "mean_token_accuracy": 0.933859184384346,
      "num_tokens": 2916352.0,
      "step": 178
    },
    {
      "entropy": 4.800918698310852,
      "epoch": 0.5303703703703704,
      "grad_norm": 1.2831913232803345,
      "learning_rate": 4.804804804804805e-06,
      "loss": 0.2352910041809082,
      "mean_token_accuracy": 0.933259554207325,
      "num_tokens": 2932736.0,
      "step": 179
    },
    {
      "entropy": 4.0972049832344055,
      "epoch": 0.5333333333333333,
      "grad_norm": 1.1364777088165283,
      "learning_rate": 4.774774774774775e-06,
      "loss": 0.18518929183483124,
      "mean_token_accuracy": 0.9431184753775597,
      "num_tokens": 2949120.0,
      "step": 180
    },
    {
      "entropy": 3.7471932768821716,
      "epoch": 0.5362962962962963,
      "grad_norm": 0.9751579761505127,
      "learning_rate": 4.7447447447447454e-06,
      "loss": 0.18314646184444427,
      "mean_token_accuracy": 0.947534404695034,
      "num_tokens": 2965504.0,
      "step": 181
    },
    {
      "entropy": 4.793573826551437,
      "epoch": 0.5392592592592592,
      "grad_norm": 1.8164482116699219,
      "learning_rate": 4.714714714714715e-06,
      "loss": 0.20572032034397125,
      "mean_token_accuracy": 0.9426435232162476,
      "num_tokens": 2981888.0,
      "step": 182
    },
    {
      "entropy": 4.3548590540885925,
      "epoch": 0.5422222222222223,
      "grad_norm": 1.1246410608291626,
      "learning_rate": 4.684684684684685e-06,
      "loss": 0.15288400650024414,
      "mean_token_accuracy": 0.9575978219509125,
      "num_tokens": 2998272.0,
      "step": 183
    },
    {
      "entropy": 4.571796417236328,
      "epoch": 0.5451851851851852,
      "grad_norm": 1.290729284286499,
      "learning_rate": 4.654654654654655e-06,
      "loss": 0.2657758593559265,
      "mean_token_accuracy": 0.9306245073676109,
      "num_tokens": 3014656.0,
      "step": 184
    },
    {
      "entropy": 4.777157008647919,
      "epoch": 0.5481481481481482,
      "grad_norm": 0.9829837083816528,
      "learning_rate": 4.624624624624625e-06,
      "loss": 0.14354024827480316,
      "mean_token_accuracy": 0.9566553011536598,
      "num_tokens": 3031040.0,
      "step": 185
    },
    {
      "entropy": 4.889033555984497,
      "epoch": 0.5511111111111111,
      "grad_norm": 1.4265416860580444,
      "learning_rate": 4.594594594594596e-06,
      "loss": 0.24842345714569092,
      "mean_token_accuracy": 0.9338478893041611,
      "num_tokens": 3047424.0,
      "step": 186
    },
    {
      "entropy": 4.8326007425785065,
      "epoch": 0.554074074074074,
      "grad_norm": 1.6100798845291138,
      "learning_rate": 4.5645645645645645e-06,
      "loss": 0.30461499094963074,
      "mean_token_accuracy": 0.9184162020683289,
      "num_tokens": 3063808.0,
      "step": 187
    },
    {
      "entropy": 3.95138019323349,
      "epoch": 0.557037037037037,
      "grad_norm": 1.0334235429763794,
      "learning_rate": 4.534534534534535e-06,
      "loss": 0.19512997567653656,
      "mean_token_accuracy": 0.9504513815045357,
      "num_tokens": 3080192.0,
      "step": 188
    },
    {
      "entropy": 4.767197489738464,
      "epoch": 0.56,
      "grad_norm": 1.4528827667236328,
      "learning_rate": 4.504504504504505e-06,
      "loss": 0.24135306477546692,
      "mean_token_accuracy": 0.9408271014690399,
      "num_tokens": 3096576.0,
      "step": 189
    },
    {
      "entropy": 4.60133022069931,
      "epoch": 0.562962962962963,
      "grad_norm": 1.3094948530197144,
      "learning_rate": 4.474474474474475e-06,
      "loss": 0.2210572212934494,
      "mean_token_accuracy": 0.9462493360042572,
      "num_tokens": 3112960.0,
      "step": 190
    },
    {
      "entropy": 4.119889825582504,
      "epoch": 0.5659259259259259,
      "grad_norm": 1.1088842153549194,
      "learning_rate": 4.444444444444444e-06,
      "loss": 0.22152382135391235,
      "mean_token_accuracy": 0.9464557394385338,
      "num_tokens": 3129344.0,
      "step": 191
    },
    {
      "entropy": 4.496746778488159,
      "epoch": 0.5688888888888889,
      "grad_norm": 0.8695931434631348,
      "learning_rate": 4.414414414414415e-06,
      "loss": 0.13890038430690765,
      "mean_token_accuracy": 0.9619273692369461,
      "num_tokens": 3145728.0,
      "step": 192
    },
    {
      "entropy": 4.323371887207031,
      "epoch": 0.5718518518518518,
      "grad_norm": 1.016728401184082,
      "learning_rate": 4.384384384384384e-06,
      "loss": 0.24737483263015747,
      "mean_token_accuracy": 0.9274903908371925,
      "num_tokens": 3162112.0,
      "step": 193
    },
    {
      "entropy": 4.250603348016739,
      "epoch": 0.5748148148148148,
      "grad_norm": 1.087903618812561,
      "learning_rate": 4.354354354354355e-06,
      "loss": 0.24808856844902039,
      "mean_token_accuracy": 0.9316486865282059,
      "num_tokens": 3178496.0,
      "step": 194
    },
    {
      "entropy": 4.634637236595154,
      "epoch": 0.5777777777777777,
      "grad_norm": 0.9575849175453186,
      "learning_rate": 4.324324324324325e-06,
      "loss": 0.15280787646770477,
      "mean_token_accuracy": 0.9524153172969818,
      "num_tokens": 3194880.0,
      "step": 195
    },
    {
      "entropy": 4.202795147895813,
      "epoch": 0.5807407407407408,
      "grad_norm": 1.0395658016204834,
      "learning_rate": 4.294294294294294e-06,
      "loss": 0.18110454082489014,
      "mean_token_accuracy": 0.9497261047363281,
      "num_tokens": 3211264.0,
      "step": 196
    },
    {
      "entropy": 4.339707612991333,
      "epoch": 0.5837037037037037,
      "grad_norm": 1.0732710361480713,
      "learning_rate": 4.264264264264265e-06,
      "loss": 0.224087655544281,
      "mean_token_accuracy": 0.9398151263594627,
      "num_tokens": 3227648.0,
      "step": 197
    },
    {
      "entropy": 4.744602203369141,
      "epoch": 0.5866666666666667,
      "grad_norm": 1.0211141109466553,
      "learning_rate": 4.234234234234235e-06,
      "loss": 0.20520702004432678,
      "mean_token_accuracy": 0.9335606023669243,
      "num_tokens": 3244032.0,
      "step": 198
    },
    {
      "entropy": 4.193909078836441,
      "epoch": 0.5896296296296296,
      "grad_norm": 1.1632949113845825,
      "learning_rate": 4.204204204204204e-06,
      "loss": 0.1872227042913437,
      "mean_token_accuracy": 0.9562376439571381,
      "num_tokens": 3260416.0,
      "step": 199
    },
    {
      "entropy": 4.291110724210739,
      "epoch": 0.5925925925925926,
      "grad_norm": 0.9554310441017151,
      "learning_rate": 4.174174174174174e-06,
      "loss": 0.15306523442268372,
      "mean_token_accuracy": 0.9542046561837196,
      "num_tokens": 3276800.0,
      "step": 200
    },
    {
      "epoch": 0.5925925925925926,
      "eval_entropy": 4.396466445922852,
      "eval_loss": 0.1749386340379715,
      "eval_mean_token_accuracy": 0.9518819769223531,
      "eval_num_tokens": 3276800.0,
      "eval_runtime": 41.7244,
      "eval_samples_per_second": 14.38,
      "eval_steps_per_second": 1.798,
      "step": 200
    },
    {
      "entropy": 4.5472937524318695,
      "epoch": 0.5955555555555555,
      "grad_norm": 1.2405363321304321,
      "learning_rate": 4.1441441441441446e-06,
      "loss": 0.27398329973220825,
      "mean_token_accuracy": 0.9217006787657738,
      "num_tokens": 3293184.0,
      "step": 201
    },
    {
      "entropy": 4.4060549437999725,
      "epoch": 0.5985185185185186,
      "grad_norm": 1.0400793552398682,
      "learning_rate": 4.114114114114114e-06,
      "loss": 0.20156747102737427,
      "mean_token_accuracy": 0.9446389377117157,
      "num_tokens": 3309568.0,
      "step": 202
    },
    {
      "entropy": 4.185441434383392,
      "epoch": 0.6014814814814815,
      "grad_norm": 0.8990085124969482,
      "learning_rate": 4.084084084084085e-06,
      "loss": 0.1516772210597992,
      "mean_token_accuracy": 0.9569917395710945,
      "num_tokens": 3325952.0,
      "step": 203
    },
    {
      "entropy": 4.80301433801651,
      "epoch": 0.6044444444444445,
      "grad_norm": 1.384344458580017,
      "learning_rate": 4.0540540540540545e-06,
      "loss": 0.283430278301239,
      "mean_token_accuracy": 0.9350188449025154,
      "num_tokens": 3342336.0,
      "step": 204
    },
    {
      "entropy": 4.536960780620575,
      "epoch": 0.6074074074074074,
      "grad_norm": 1.2103245258331299,
      "learning_rate": 4.024024024024024e-06,
      "loss": 0.235305055975914,
      "mean_token_accuracy": 0.932244710624218,
      "num_tokens": 3358720.0,
      "step": 205
    },
    {
      "entropy": 4.044310688972473,
      "epoch": 0.6103703703703703,
      "grad_norm": 0.8922559022903442,
      "learning_rate": 3.993993993993994e-06,
      "loss": 0.1615198254585266,
      "mean_token_accuracy": 0.9466530829668045,
      "num_tokens": 3375104.0,
      "step": 206
    },
    {
      "entropy": 4.553136348724365,
      "epoch": 0.6133333333333333,
      "grad_norm": 0.9324111342430115,
      "learning_rate": 3.9639639639639645e-06,
      "loss": 0.1158020943403244,
      "mean_token_accuracy": 0.9715171456336975,
      "num_tokens": 3391488.0,
      "step": 207
    },
    {
      "entropy": 3.9612647891044617,
      "epoch": 0.6162962962962963,
      "grad_norm": 1.1383168697357178,
      "learning_rate": 3.933933933933934e-06,
      "loss": 0.18634311854839325,
      "mean_token_accuracy": 0.9486203193664551,
      "num_tokens": 3407872.0,
      "step": 208
    },
    {
      "entropy": 4.339573085308075,
      "epoch": 0.6192592592592593,
      "grad_norm": 1.042952537536621,
      "learning_rate": 3.903903903903904e-06,
      "loss": 0.22029240429401398,
      "mean_token_accuracy": 0.9371641352772713,
      "num_tokens": 3424256.0,
      "step": 209
    },
    {
      "entropy": 4.470686435699463,
      "epoch": 0.6222222222222222,
      "grad_norm": 1.1252963542938232,
      "learning_rate": 3.8738738738738744e-06,
      "loss": 0.2142847180366516,
      "mean_token_accuracy": 0.937769889831543,
      "num_tokens": 3440640.0,
      "step": 210
    },
    {
      "entropy": 4.301492214202881,
      "epoch": 0.6251851851851852,
      "grad_norm": 1.1598789691925049,
      "learning_rate": 3.843843843843844e-06,
      "loss": 0.21656247973442078,
      "mean_token_accuracy": 0.9341708347201347,
      "num_tokens": 3457024.0,
      "step": 211
    },
    {
      "entropy": 4.2081300020217896,
      "epoch": 0.6281481481481481,
      "grad_norm": 0.9859828948974609,
      "learning_rate": 3.8138138138138143e-06,
      "loss": 0.1877346932888031,
      "mean_token_accuracy": 0.9473938569426537,
      "num_tokens": 3473408.0,
      "step": 212
    },
    {
      "entropy": 4.661305487155914,
      "epoch": 0.6311111111111111,
      "grad_norm": 1.2205841541290283,
      "learning_rate": 3.7837837837837844e-06,
      "loss": 0.22747467458248138,
      "mean_token_accuracy": 0.92958153039217,
      "num_tokens": 3489792.0,
      "step": 213
    },
    {
      "entropy": 4.07285350561142,
      "epoch": 0.6340740740740741,
      "grad_norm": 1.0680360794067383,
      "learning_rate": 3.7537537537537537e-06,
      "loss": 0.1850614696741104,
      "mean_token_accuracy": 0.9432996138930321,
      "num_tokens": 3506176.0,
      "step": 214
    },
    {
      "entropy": 4.349195241928101,
      "epoch": 0.6370370370370371,
      "grad_norm": 0.9926294088363647,
      "learning_rate": 3.723723723723724e-06,
      "loss": 0.15226511657238007,
      "mean_token_accuracy": 0.9589138552546501,
      "num_tokens": 3522560.0,
      "step": 215
    },
    {
      "entropy": 4.486599534749985,
      "epoch": 0.64,
      "grad_norm": 1.101911187171936,
      "learning_rate": 3.693693693693694e-06,
      "loss": 0.1986139714717865,
      "mean_token_accuracy": 0.9381145983934402,
      "num_tokens": 3538944.0,
      "step": 216
    },
    {
      "entropy": 4.344702124595642,
      "epoch": 0.642962962962963,
      "grad_norm": 1.2513370513916016,
      "learning_rate": 3.663663663663664e-06,
      "loss": 0.20217151939868927,
      "mean_token_accuracy": 0.9439992159605026,
      "num_tokens": 3555328.0,
      "step": 217
    },
    {
      "entropy": 4.32846137881279,
      "epoch": 0.6459259259259259,
      "grad_norm": 0.9642252922058105,
      "learning_rate": 3.633633633633634e-06,
      "loss": 0.14241398870944977,
      "mean_token_accuracy": 0.9539151266217232,
      "num_tokens": 3571712.0,
      "step": 218
    },
    {
      "entropy": 4.716962099075317,
      "epoch": 0.6488888888888888,
      "grad_norm": 1.372519612312317,
      "learning_rate": 3.603603603603604e-06,
      "loss": 0.2469625025987625,
      "mean_token_accuracy": 0.9339680150151253,
      "num_tokens": 3588096.0,
      "step": 219
    },
    {
      "entropy": 4.073337703943253,
      "epoch": 0.6518518518518519,
      "grad_norm": 0.9866761565208435,
      "learning_rate": 3.573573573573574e-06,
      "loss": 0.1708499789237976,
      "mean_token_accuracy": 0.9544656574726105,
      "num_tokens": 3604480.0,
      "step": 220
    },
    {
      "entropy": 3.9164834320545197,
      "epoch": 0.6548148148148148,
      "grad_norm": 0.9244502782821655,
      "learning_rate": 3.5435435435435437e-06,
      "loss": 0.18559050559997559,
      "mean_token_accuracy": 0.9379841834306717,
      "num_tokens": 3620864.0,
      "step": 221
    },
    {
      "entropy": 4.153732627630234,
      "epoch": 0.6577777777777778,
      "grad_norm": 1.0579266548156738,
      "learning_rate": 3.513513513513514e-06,
      "loss": 0.19521856307983398,
      "mean_token_accuracy": 0.939164936542511,
      "num_tokens": 3637248.0,
      "step": 222
    },
    {
      "entropy": 4.709444046020508,
      "epoch": 0.6607407407407407,
      "grad_norm": 1.1478586196899414,
      "learning_rate": 3.4834834834834835e-06,
      "loss": 0.1736764907836914,
      "mean_token_accuracy": 0.9531540721654892,
      "num_tokens": 3653632.0,
      "step": 223
    },
    {
      "entropy": 4.5095644891262054,
      "epoch": 0.6637037037037037,
      "grad_norm": 1.105306625366211,
      "learning_rate": 3.4534534534534537e-06,
      "loss": 0.12064651399850845,
      "mean_token_accuracy": 0.9645556434988976,
      "num_tokens": 3670016.0,
      "step": 224
    },
    {
      "entropy": 4.558714330196381,
      "epoch": 0.6666666666666666,
      "grad_norm": 1.025750994682312,
      "learning_rate": 3.423423423423424e-06,
      "loss": 0.16424690186977386,
      "mean_token_accuracy": 0.9530202373862267,
      "num_tokens": 3686400.0,
      "step": 225
    },
    {
      "entropy": 4.617130994796753,
      "epoch": 0.6696296296296296,
      "grad_norm": 1.163450837135315,
      "learning_rate": 3.393393393393394e-06,
      "loss": 0.18721315264701843,
      "mean_token_accuracy": 0.9541647657752037,
      "num_tokens": 3702784.0,
      "step": 226
    },
    {
      "entropy": 4.451707720756531,
      "epoch": 0.6725925925925926,
      "grad_norm": 1.188270926475525,
      "learning_rate": 3.363363363363364e-06,
      "loss": 0.2559033930301666,
      "mean_token_accuracy": 0.9307445734739304,
      "num_tokens": 3719168.0,
      "step": 227
    },
    {
      "entropy": 4.333233296871185,
      "epoch": 0.6755555555555556,
      "grad_norm": 0.8371542692184448,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 0.11930409073829651,
      "mean_token_accuracy": 0.9697378799319267,
      "num_tokens": 3735552.0,
      "step": 228
    },
    {
      "entropy": 4.185926705598831,
      "epoch": 0.6785185185185185,
      "grad_norm": 0.8926585912704468,
      "learning_rate": 3.3033033033033035e-06,
      "loss": 0.10602101683616638,
      "mean_token_accuracy": 0.9678346142172813,
      "num_tokens": 3751936.0,
      "step": 229
    },
    {
      "entropy": 4.013706266880035,
      "epoch": 0.6814814814814815,
      "grad_norm": 0.9563543796539307,
      "learning_rate": 3.2732732732732736e-06,
      "loss": 0.1615547239780426,
      "mean_token_accuracy": 0.9547702893614769,
      "num_tokens": 3768320.0,
      "step": 230
    },
    {
      "entropy": 4.594108134508133,
      "epoch": 0.6844444444444444,
      "grad_norm": 1.382849097251892,
      "learning_rate": 3.2432432432432437e-06,
      "loss": 0.29585137963294983,
      "mean_token_accuracy": 0.9275413975119591,
      "num_tokens": 3784704.0,
      "step": 231
    },
    {
      "entropy": 4.2951953411102295,
      "epoch": 0.6874074074074074,
      "grad_norm": 1.0980827808380127,
      "learning_rate": 3.2132132132132134e-06,
      "loss": 0.18029291927814484,
      "mean_token_accuracy": 0.9512800425291061,
      "num_tokens": 3801088.0,
      "step": 232
    },
    {
      "entropy": 4.458141177892685,
      "epoch": 0.6903703703703704,
      "grad_norm": 1.0253396034240723,
      "learning_rate": 3.183183183183183e-06,
      "loss": 0.15372850000858307,
      "mean_token_accuracy": 0.9519759193062782,
      "num_tokens": 3817472.0,
      "step": 233
    },
    {
      "entropy": 4.413249999284744,
      "epoch": 0.6933333333333334,
      "grad_norm": 0.9809788465499878,
      "learning_rate": 3.1531531531531532e-06,
      "loss": 0.13195912539958954,
      "mean_token_accuracy": 0.9675045162439346,
      "num_tokens": 3833856.0,
      "step": 234
    },
    {
      "entropy": 4.842405796051025,
      "epoch": 0.6962962962962963,
      "grad_norm": 1.4815974235534668,
      "learning_rate": 3.1231231231231234e-06,
      "loss": 0.21166671812534332,
      "mean_token_accuracy": 0.9453298151493073,
      "num_tokens": 3850240.0,
      "step": 235
    },
    {
      "entropy": 4.802890330553055,
      "epoch": 0.6992592592592592,
      "grad_norm": 1.4081461429595947,
      "learning_rate": 3.0930930930930935e-06,
      "loss": 0.24762584269046783,
      "mean_token_accuracy": 0.9402480274438858,
      "num_tokens": 3866624.0,
      "step": 236
    },
    {
      "entropy": 4.598718553781509,
      "epoch": 0.7022222222222222,
      "grad_norm": 1.12220299243927,
      "learning_rate": 3.063063063063063e-06,
      "loss": 0.20380182564258575,
      "mean_token_accuracy": 0.9401306957006454,
      "num_tokens": 3883008.0,
      "step": 237
    },
    {
      "entropy": 4.571725904941559,
      "epoch": 0.7051851851851851,
      "grad_norm": 1.097094178199768,
      "learning_rate": 3.0330330330330333e-06,
      "loss": 0.18988654017448425,
      "mean_token_accuracy": 0.9413757547736168,
      "num_tokens": 3899392.0,
      "step": 238
    },
    {
      "entropy": 4.325874507427216,
      "epoch": 0.7081481481481482,
      "grad_norm": 1.0416995286941528,
      "learning_rate": 3.0030030030030034e-06,
      "loss": 0.1390780508518219,
      "mean_token_accuracy": 0.9587825983762741,
      "num_tokens": 3915776.0,
      "step": 239
    },
    {
      "entropy": 4.9004232585430145,
      "epoch": 0.7111111111111111,
      "grad_norm": 1.2756025791168213,
      "learning_rate": 2.9729729729729736e-06,
      "loss": 0.23010534048080444,
      "mean_token_accuracy": 0.9233949109911919,
      "num_tokens": 3932160.0,
      "step": 240
    },
    {
      "entropy": 4.623109728097916,
      "epoch": 0.7140740740740741,
      "grad_norm": 0.9779954552650452,
      "learning_rate": 2.942942942942943e-06,
      "loss": 0.1899159848690033,
      "mean_token_accuracy": 0.9444768354296684,
      "num_tokens": 3948544.0,
      "step": 241
    },
    {
      "entropy": 4.409802705049515,
      "epoch": 0.717037037037037,
      "grad_norm": 1.2449361085891724,
      "learning_rate": 2.912912912912913e-06,
      "loss": 0.22213713824748993,
      "mean_token_accuracy": 0.9347798228263855,
      "num_tokens": 3964928.0,
      "step": 242
    },
    {
      "entropy": 4.728479206562042,
      "epoch": 0.72,
      "grad_norm": 1.2552624940872192,
      "learning_rate": 2.882882882882883e-06,
      "loss": 0.27347418665885925,
      "mean_token_accuracy": 0.9270098879933357,
      "num_tokens": 3981312.0,
      "step": 243
    },
    {
      "entropy": 4.303971439599991,
      "epoch": 0.7229629629629629,
      "grad_norm": 0.9352066516876221,
      "learning_rate": 2.8528528528528532e-06,
      "loss": 0.14397451281547546,
      "mean_token_accuracy": 0.9503274336457253,
      "num_tokens": 3997696.0,
      "step": 244
    },
    {
      "entropy": 4.331236720085144,
      "epoch": 0.725925925925926,
      "grad_norm": 0.8930276036262512,
      "learning_rate": 2.8228228228228234e-06,
      "loss": 0.10607534646987915,
      "mean_token_accuracy": 0.9745400696992874,
      "num_tokens": 4014080.0,
      "step": 245
    },
    {
      "entropy": 4.694693148136139,
      "epoch": 0.7288888888888889,
      "grad_norm": 1.2434946298599243,
      "learning_rate": 2.7927927927927926e-06,
      "loss": 0.22799471020698547,
      "mean_token_accuracy": 0.9323829114437103,
      "num_tokens": 4030464.0,
      "step": 246
    },
    {
      "entropy": 4.469425290822983,
      "epoch": 0.7318518518518519,
      "grad_norm": 0.872792661190033,
      "learning_rate": 2.7627627627627628e-06,
      "loss": 0.12364237010478973,
      "mean_token_accuracy": 0.9642724394798279,
      "num_tokens": 4046848.0,
      "step": 247
    },
    {
      "entropy": 4.155093103647232,
      "epoch": 0.7348148148148148,
      "grad_norm": 1.076931118965149,
      "learning_rate": 2.732732732732733e-06,
      "loss": 0.14060986042022705,
      "mean_token_accuracy": 0.9647770524024963,
      "num_tokens": 4063232.0,
      "step": 248
    },
    {
      "entropy": 4.098130851984024,
      "epoch": 0.7377777777777778,
      "grad_norm": 1.1495380401611328,
      "learning_rate": 2.702702702702703e-06,
      "loss": 0.23693521320819855,
      "mean_token_accuracy": 0.9423738196492195,
      "num_tokens": 4079616.0,
      "step": 249
    },
    {
      "entropy": 4.527021795511246,
      "epoch": 0.7407407407407407,
      "grad_norm": 1.158864140510559,
      "learning_rate": 2.672672672672673e-06,
      "loss": 0.22205623984336853,
      "mean_token_accuracy": 0.9401117786765099,
      "num_tokens": 4096000.0,
      "step": 250
    },
    {
      "epoch": 0.7407407407407407,
      "eval_entropy": 4.4104202969868975,
      "eval_loss": 0.17062820494174957,
      "eval_mean_token_accuracy": 0.9528267327944437,
      "eval_num_tokens": 4096000.0,
      "eval_runtime": 41.7307,
      "eval_samples_per_second": 14.378,
      "eval_steps_per_second": 1.797,
      "step": 250
    },
    {
      "entropy": 4.590001463890076,
      "epoch": 0.7437037037037038,
      "grad_norm": 1.2274631261825562,
      "learning_rate": 2.642642642642643e-06,
      "loss": 0.19656914472579956,
      "mean_token_accuracy": 0.9513155668973923,
      "num_tokens": 4112384.0,
      "step": 251
    },
    {
      "entropy": 4.3021697998046875,
      "epoch": 0.7466666666666667,
      "grad_norm": 0.9692113995552063,
      "learning_rate": 2.612612612612613e-06,
      "loss": 0.14871367812156677,
      "mean_token_accuracy": 0.9626764878630638,
      "num_tokens": 4128768.0,
      "step": 252
    },
    {
      "entropy": 4.653590381145477,
      "epoch": 0.7496296296296296,
      "grad_norm": 1.2943620681762695,
      "learning_rate": 2.5825825825825827e-06,
      "loss": 0.22055670619010925,
      "mean_token_accuracy": 0.9380655884742737,
      "num_tokens": 4145152.0,
      "step": 253
    },
    {
      "entropy": 4.471467673778534,
      "epoch": 0.7525925925925926,
      "grad_norm": 1.082425832748413,
      "learning_rate": 2.552552552552553e-06,
      "loss": 0.17337831854820251,
      "mean_token_accuracy": 0.9434536769986153,
      "num_tokens": 4161536.0,
      "step": 254
    },
    {
      "entropy": 4.303240329027176,
      "epoch": 0.7555555555555555,
      "grad_norm": 1.1208677291870117,
      "learning_rate": 2.5225225225225225e-06,
      "loss": 0.20271582901477814,
      "mean_token_accuracy": 0.9410636723041534,
      "num_tokens": 4177920.0,
      "step": 255
    },
    {
      "entropy": 4.827797591686249,
      "epoch": 0.7585185185185185,
      "grad_norm": 1.1007102727890015,
      "learning_rate": 2.4924924924924926e-06,
      "loss": 0.1865009367465973,
      "mean_token_accuracy": 0.943950854241848,
      "num_tokens": 4194304.0,
      "step": 256
    },
    {
      "entropy": 4.273179918527603,
      "epoch": 0.7614814814814815,
      "grad_norm": 1.1347743272781372,
      "learning_rate": 2.4624624624624628e-06,
      "loss": 0.22876350581645966,
      "mean_token_accuracy": 0.944123238325119,
      "num_tokens": 4210688.0,
      "step": 257
    },
    {
      "entropy": 4.056853353977203,
      "epoch": 0.7644444444444445,
      "grad_norm": 1.0987876653671265,
      "learning_rate": 2.432432432432433e-06,
      "loss": 0.2191711664199829,
      "mean_token_accuracy": 0.9387948885560036,
      "num_tokens": 4227072.0,
      "step": 258
    },
    {
      "entropy": 4.687393695116043,
      "epoch": 0.7674074074074074,
      "grad_norm": 1.1725701093673706,
      "learning_rate": 2.4024024024024026e-06,
      "loss": 0.2219541221857071,
      "mean_token_accuracy": 0.9271552339196205,
      "num_tokens": 4243456.0,
      "step": 259
    },
    {
      "entropy": 4.514097571372986,
      "epoch": 0.7703703703703704,
      "grad_norm": 1.2087112665176392,
      "learning_rate": 2.3723723723723727e-06,
      "loss": 0.17129437625408173,
      "mean_token_accuracy": 0.9455615282058716,
      "num_tokens": 4259840.0,
      "step": 260
    },
    {
      "entropy": 4.773777067661285,
      "epoch": 0.7733333333333333,
      "grad_norm": 1.3463881015777588,
      "learning_rate": 2.3423423423423424e-06,
      "loss": 0.2544708847999573,
      "mean_token_accuracy": 0.9216426834464073,
      "num_tokens": 4276224.0,
      "step": 261
    },
    {
      "entropy": 4.905140161514282,
      "epoch": 0.7762962962962963,
      "grad_norm": 1.4696393013000488,
      "learning_rate": 2.3123123123123125e-06,
      "loss": 0.24074558913707733,
      "mean_token_accuracy": 0.9325360879302025,
      "num_tokens": 4292608.0,
      "step": 262
    },
    {
      "entropy": 5.013612270355225,
      "epoch": 0.7792592592592592,
      "grad_norm": 1.4878811836242676,
      "learning_rate": 2.2822822822822822e-06,
      "loss": 0.2888287305831909,
      "mean_token_accuracy": 0.9270473197102547,
      "num_tokens": 4308992.0,
      "step": 263
    },
    {
      "entropy": 4.366520285606384,
      "epoch": 0.7822222222222223,
      "grad_norm": 0.9022302627563477,
      "learning_rate": 2.2522522522522524e-06,
      "loss": 0.14121748507022858,
      "mean_token_accuracy": 0.9612898379564285,
      "num_tokens": 4325376.0,
      "step": 264
    },
    {
      "entropy": 4.398820221424103,
      "epoch": 0.7851851851851852,
      "grad_norm": 1.0134413242340088,
      "learning_rate": 2.222222222222222e-06,
      "loss": 0.1413351148366928,
      "mean_token_accuracy": 0.9589046388864517,
      "num_tokens": 4341760.0,
      "step": 265
    },
    {
      "entropy": 3.8570327758789062,
      "epoch": 0.7881481481481482,
      "grad_norm": 0.7519953846931458,
      "learning_rate": 2.192192192192192e-06,
      "loss": 0.12349563091993332,
      "mean_token_accuracy": 0.9535651355981827,
      "num_tokens": 4358144.0,
      "step": 266
    },
    {
      "entropy": 4.2944246381521225,
      "epoch": 0.7911111111111111,
      "grad_norm": 1.376917839050293,
      "learning_rate": 2.1621621621621623e-06,
      "loss": 0.25689592957496643,
      "mean_token_accuracy": 0.9375665113329887,
      "num_tokens": 4374528.0,
      "step": 267
    },
    {
      "entropy": 4.258553355932236,
      "epoch": 0.794074074074074,
      "grad_norm": 1.1932320594787598,
      "learning_rate": 2.1321321321321325e-06,
      "loss": 0.1729450672864914,
      "mean_token_accuracy": 0.9569645076990128,
      "num_tokens": 4390912.0,
      "step": 268
    },
    {
      "entropy": 4.926870286464691,
      "epoch": 0.797037037037037,
      "grad_norm": 1.3492707014083862,
      "learning_rate": 2.102102102102102e-06,
      "loss": 0.22400406002998352,
      "mean_token_accuracy": 0.9252083674073219,
      "num_tokens": 4407296.0,
      "step": 269
    },
    {
      "entropy": 4.9556872844696045,
      "epoch": 0.8,
      "grad_norm": 1.4651730060577393,
      "learning_rate": 2.0720720720720723e-06,
      "loss": 0.2858358323574066,
      "mean_token_accuracy": 0.9188975766301155,
      "num_tokens": 4423680.0,
      "step": 270
    },
    {
      "entropy": 4.358668476343155,
      "epoch": 0.802962962962963,
      "grad_norm": 1.2027381658554077,
      "learning_rate": 2.0420420420420424e-06,
      "loss": 0.16990210115909576,
      "mean_token_accuracy": 0.9469516202807426,
      "num_tokens": 4440064.0,
      "step": 271
    },
    {
      "entropy": 4.810274362564087,
      "epoch": 0.8059259259259259,
      "grad_norm": 1.543989896774292,
      "learning_rate": 2.012012012012012e-06,
      "loss": 0.23652800917625427,
      "mean_token_accuracy": 0.9404151514172554,
      "num_tokens": 4456448.0,
      "step": 272
    },
    {
      "entropy": 4.647187799215317,
      "epoch": 0.8088888888888889,
      "grad_norm": 1.44573175907135,
      "learning_rate": 1.9819819819819822e-06,
      "loss": 0.2977809011936188,
      "mean_token_accuracy": 0.9311791732907295,
      "num_tokens": 4472832.0,
      "step": 273
    },
    {
      "entropy": 3.91234590113163,
      "epoch": 0.8118518518518518,
      "grad_norm": 0.7962574362754822,
      "learning_rate": 1.951951951951952e-06,
      "loss": 0.10670879483222961,
      "mean_token_accuracy": 0.9662504494190216,
      "num_tokens": 4489216.0,
      "step": 274
    },
    {
      "entropy": 4.8694718182086945,
      "epoch": 0.8148148148148148,
      "grad_norm": 1.1886613368988037,
      "learning_rate": 1.921921921921922e-06,
      "loss": 0.22934231162071228,
      "mean_token_accuracy": 0.93792774528265,
      "num_tokens": 4505600.0,
      "step": 275
    },
    {
      "entropy": 4.880127847194672,
      "epoch": 0.8177777777777778,
      "grad_norm": 1.5155773162841797,
      "learning_rate": 1.8918918918918922e-06,
      "loss": 0.2636268734931946,
      "mean_token_accuracy": 0.9262536317110062,
      "num_tokens": 4521984.0,
      "step": 276
    },
    {
      "entropy": 4.561476111412048,
      "epoch": 0.8207407407407408,
      "grad_norm": 1.0868642330169678,
      "learning_rate": 1.861861861861862e-06,
      "loss": 0.20132635533809662,
      "mean_token_accuracy": 0.9411407560110092,
      "num_tokens": 4538368.0,
      "step": 277
    },
    {
      "entropy": 4.483148396015167,
      "epoch": 0.8237037037037037,
      "grad_norm": 0.9549974799156189,
      "learning_rate": 1.831831831831832e-06,
      "loss": 0.12138265371322632,
      "mean_token_accuracy": 0.9597767740488052,
      "num_tokens": 4554752.0,
      "step": 278
    },
    {
      "entropy": 4.4938174188137054,
      "epoch": 0.8266666666666667,
      "grad_norm": 1.155914306640625,
      "learning_rate": 1.801801801801802e-06,
      "loss": 0.1511882245540619,
      "mean_token_accuracy": 0.9566312059760094,
      "num_tokens": 4571136.0,
      "step": 279
    },
    {
      "entropy": 4.489579200744629,
      "epoch": 0.8296296296296296,
      "grad_norm": 1.1139575242996216,
      "learning_rate": 1.7717717717717719e-06,
      "loss": 0.21914277970790863,
      "mean_token_accuracy": 0.9424230232834816,
      "num_tokens": 4587520.0,
      "step": 280
    },
    {
      "entropy": 4.057400763034821,
      "epoch": 0.8325925925925926,
      "grad_norm": 0.8104386925697327,
      "learning_rate": 1.7417417417417418e-06,
      "loss": 0.11663420498371124,
      "mean_token_accuracy": 0.9625060260295868,
      "num_tokens": 4603904.0,
      "step": 281
    },
    {
      "entropy": 4.992342710494995,
      "epoch": 0.8355555555555556,
      "grad_norm": 1.3316982984542847,
      "learning_rate": 1.711711711711712e-06,
      "loss": 0.2464400678873062,
      "mean_token_accuracy": 0.9318329095840454,
      "num_tokens": 4620288.0,
      "step": 282
    },
    {
      "entropy": 3.7718082517385483,
      "epoch": 0.8385185185185186,
      "grad_norm": 0.9035054445266724,
      "learning_rate": 1.681681681681682e-06,
      "loss": 0.15544335544109344,
      "mean_token_accuracy": 0.9531992748379707,
      "num_tokens": 4636672.0,
      "step": 283
    },
    {
      "entropy": 4.849994897842407,
      "epoch": 0.8414814814814815,
      "grad_norm": 1.159351110458374,
      "learning_rate": 1.6516516516516517e-06,
      "loss": 0.18373851478099823,
      "mean_token_accuracy": 0.9374978095293045,
      "num_tokens": 4653056.0,
      "step": 284
    },
    {
      "entropy": 4.640776067972183,
      "epoch": 0.8444444444444444,
      "grad_norm": 1.127689242362976,
      "learning_rate": 1.6216216216216219e-06,
      "loss": 0.15810415148735046,
      "mean_token_accuracy": 0.9538983181118965,
      "num_tokens": 4669440.0,
      "step": 285
    },
    {
      "entropy": 4.124550685286522,
      "epoch": 0.8474074074074074,
      "grad_norm": 0.7990202307701111,
      "learning_rate": 1.5915915915915916e-06,
      "loss": 0.12189920246601105,
      "mean_token_accuracy": 0.9596385732293129,
      "num_tokens": 4685824.0,
      "step": 286
    },
    {
      "entropy": 4.1818675100803375,
      "epoch": 0.8503703703703703,
      "grad_norm": 0.9863611459732056,
      "learning_rate": 1.5615615615615617e-06,
      "loss": 0.1411495804786682,
      "mean_token_accuracy": 0.9570413455367088,
      "num_tokens": 4702208.0,
      "step": 287
    },
    {
      "entropy": 4.235043227672577,
      "epoch": 0.8533333333333334,
      "grad_norm": 0.9155336022377014,
      "learning_rate": 1.5315315315315316e-06,
      "loss": 0.1226419061422348,
      "mean_token_accuracy": 0.9634087905287743,
      "num_tokens": 4718592.0,
      "step": 288
    },
    {
      "entropy": 4.70637646317482,
      "epoch": 0.8562962962962963,
      "grad_norm": 1.1035104990005493,
      "learning_rate": 1.5015015015015017e-06,
      "loss": 0.17081449925899506,
      "mean_token_accuracy": 0.9464738890528679,
      "num_tokens": 4734976.0,
      "step": 289
    },
    {
      "entropy": 3.820620059967041,
      "epoch": 0.8592592592592593,
      "grad_norm": 0.844387412071228,
      "learning_rate": 1.4714714714714714e-06,
      "loss": 0.09538038074970245,
      "mean_token_accuracy": 0.9708806127309799,
      "num_tokens": 4751360.0,
      "step": 290
    },
    {
      "entropy": 4.491610467433929,
      "epoch": 0.8622222222222222,
      "grad_norm": 0.9852358102798462,
      "learning_rate": 1.4414414414414416e-06,
      "loss": 0.14928781986236572,
      "mean_token_accuracy": 0.9547141492366791,
      "num_tokens": 4767744.0,
      "step": 291
    },
    {
      "entropy": 3.794399321079254,
      "epoch": 0.8651851851851852,
      "grad_norm": 0.9524202942848206,
      "learning_rate": 1.4114114114114117e-06,
      "loss": 0.15874914824962616,
      "mean_token_accuracy": 0.950463779270649,
      "num_tokens": 4784128.0,
      "step": 292
    },
    {
      "entropy": 4.522875607013702,
      "epoch": 0.8681481481481481,
      "grad_norm": 0.9517760872840881,
      "learning_rate": 1.3813813813813814e-06,
      "loss": 0.11182866990566254,
      "mean_token_accuracy": 0.9600817486643791,
      "num_tokens": 4800512.0,
      "step": 293
    },
    {
      "entropy": 4.704473555088043,
      "epoch": 0.8711111111111111,
      "grad_norm": 1.3497657775878906,
      "learning_rate": 1.3513513513513515e-06,
      "loss": 0.25028833746910095,
      "mean_token_accuracy": 0.9284955114126205,
      "num_tokens": 4816896.0,
      "step": 294
    },
    {
      "entropy": 3.9596442878246307,
      "epoch": 0.8740740740740741,
      "grad_norm": 0.9378573298454285,
      "learning_rate": 1.3213213213213214e-06,
      "loss": 0.15004345774650574,
      "mean_token_accuracy": 0.9589422717690468,
      "num_tokens": 4833280.0,
      "step": 295
    },
    {
      "entropy": 4.634488731622696,
      "epoch": 0.8770370370370371,
      "grad_norm": 1.1771413087844849,
      "learning_rate": 1.2912912912912913e-06,
      "loss": 0.19401542842388153,
      "mean_token_accuracy": 0.9377927854657173,
      "num_tokens": 4849664.0,
      "step": 296
    },
    {
      "entropy": 3.806145817041397,
      "epoch": 0.88,
      "grad_norm": 0.8034092783927917,
      "learning_rate": 1.2612612612612613e-06,
      "loss": 0.12706570327281952,
      "mean_token_accuracy": 0.9611969068646431,
      "num_tokens": 4866048.0,
      "step": 297
    },
    {
      "entropy": 4.59606459736824,
      "epoch": 0.882962962962963,
      "grad_norm": 1.0811915397644043,
      "learning_rate": 1.2312312312312314e-06,
      "loss": 0.19878467917442322,
      "mean_token_accuracy": 0.9509942531585693,
      "num_tokens": 4882432.0,
      "step": 298
    },
    {
      "entropy": 4.142981857061386,
      "epoch": 0.8859259259259259,
      "grad_norm": 0.9253244400024414,
      "learning_rate": 1.2012012012012013e-06,
      "loss": 0.1614236980676651,
      "mean_token_accuracy": 0.9512658193707466,
      "num_tokens": 4898816.0,
      "step": 299
    },
    {
      "entropy": 4.080303400754929,
      "epoch": 0.8888888888888888,
      "grad_norm": 1.1255525350570679,
      "learning_rate": 1.1711711711711712e-06,
      "loss": 0.14620760083198547,
      "mean_token_accuracy": 0.9649521484971046,
      "num_tokens": 4915200.0,
      "step": 300
    },
    {
      "epoch": 0.8888888888888888,
      "eval_entropy": 4.395395997365315,
      "eval_loss": 0.1683121770620346,
      "eval_mean_token_accuracy": 0.9532191967964172,
      "eval_num_tokens": 4915200.0,
      "eval_runtime": 41.7721,
      "eval_samples_per_second": 14.364,
      "eval_steps_per_second": 1.795,
      "step": 300
    },
    {
      "entropy": 4.7908841371536255,
      "epoch": 0.8918518518518519,
      "grad_norm": 1.165712594985962,
      "learning_rate": 1.1411411411411411e-06,
      "loss": 0.1966797262430191,
      "mean_token_accuracy": 0.9420882239937782,
      "num_tokens": 4931584.0,
      "step": 301
    },
    {
      "entropy": 4.794400274753571,
      "epoch": 0.8948148148148148,
      "grad_norm": 1.4137473106384277,
      "learning_rate": 1.111111111111111e-06,
      "loss": 0.22074569761753082,
      "mean_token_accuracy": 0.9417700693011284,
      "num_tokens": 4947968.0,
      "step": 302
    },
    {
      "entropy": 3.761867940425873,
      "epoch": 0.8977777777777778,
      "grad_norm": 0.8603271245956421,
      "learning_rate": 1.0810810810810812e-06,
      "loss": 0.1291244924068451,
      "mean_token_accuracy": 0.9697760865092278,
      "num_tokens": 4964352.0,
      "step": 303
    },
    {
      "entropy": 4.889683187007904,
      "epoch": 0.9007407407407407,
      "grad_norm": 1.035141110420227,
      "learning_rate": 1.051051051051051e-06,
      "loss": 0.20105254650115967,
      "mean_token_accuracy": 0.9457623511552811,
      "num_tokens": 4980736.0,
      "step": 304
    },
    {
      "entropy": 4.3381586372852325,
      "epoch": 0.9037037037037037,
      "grad_norm": 0.9701178073883057,
      "learning_rate": 1.0210210210210212e-06,
      "loss": 0.14136949181556702,
      "mean_token_accuracy": 0.9590238556265831,
      "num_tokens": 4997120.0,
      "step": 305
    },
    {
      "entropy": 4.842824548482895,
      "epoch": 0.9066666666666666,
      "grad_norm": 1.3467810153961182,
      "learning_rate": 9.909909909909911e-07,
      "loss": 0.14885962009429932,
      "mean_token_accuracy": 0.952725313603878,
      "num_tokens": 5013504.0,
      "step": 306
    },
    {
      "entropy": 3.959769457578659,
      "epoch": 0.9096296296296297,
      "grad_norm": 0.7581403851509094,
      "learning_rate": 9.60960960960961e-07,
      "loss": 0.11326365172863007,
      "mean_token_accuracy": 0.9680685102939606,
      "num_tokens": 5029888.0,
      "step": 307
    },
    {
      "entropy": 4.629861950874329,
      "epoch": 0.9125925925925926,
      "grad_norm": 1.4336919784545898,
      "learning_rate": 9.30930930930931e-07,
      "loss": 0.2650633454322815,
      "mean_token_accuracy": 0.9342033565044403,
      "num_tokens": 5046272.0,
      "step": 308
    },
    {
      "entropy": 4.817781001329422,
      "epoch": 0.9155555555555556,
      "grad_norm": 1.2062385082244873,
      "learning_rate": 9.00900900900901e-07,
      "loss": 0.22477254271507263,
      "mean_token_accuracy": 0.9376617521047592,
      "num_tokens": 5062656.0,
      "step": 309
    },
    {
      "entropy": 4.509517431259155,
      "epoch": 0.9185185185185185,
      "grad_norm": 1.1257582902908325,
      "learning_rate": 8.708708708708709e-07,
      "loss": 0.16631361842155457,
      "mean_token_accuracy": 0.941024124622345,
      "num_tokens": 5079040.0,
      "step": 310
    },
    {
      "entropy": 3.893211990594864,
      "epoch": 0.9214814814814815,
      "grad_norm": 0.7917382121086121,
      "learning_rate": 8.40840840840841e-07,
      "loss": 0.12029150873422623,
      "mean_token_accuracy": 0.9615233391523361,
      "num_tokens": 5095424.0,
      "step": 311
    },
    {
      "entropy": 4.163285553455353,
      "epoch": 0.9244444444444444,
      "grad_norm": 0.9531848430633545,
      "learning_rate": 8.108108108108109e-07,
      "loss": 0.12382030487060547,
      "mean_token_accuracy": 0.9667912647128105,
      "num_tokens": 5111808.0,
      "step": 312
    },
    {
      "entropy": 4.663005769252777,
      "epoch": 0.9274074074074075,
      "grad_norm": 1.2263110876083374,
      "learning_rate": 7.807807807807808e-07,
      "loss": 0.20207488536834717,
      "mean_token_accuracy": 0.9396724104881287,
      "num_tokens": 5128192.0,
      "step": 313
    },
    {
      "entropy": 4.277839809656143,
      "epoch": 0.9303703703703704,
      "grad_norm": 0.8545935153961182,
      "learning_rate": 7.507507507507509e-07,
      "loss": 0.11175937950611115,
      "mean_token_accuracy": 0.9668233692646027,
      "num_tokens": 5144576.0,
      "step": 314
    },
    {
      "entropy": 4.646084487438202,
      "epoch": 0.9333333333333333,
      "grad_norm": 1.0369517803192139,
      "learning_rate": 7.207207207207208e-07,
      "loss": 0.17966434359550476,
      "mean_token_accuracy": 0.9477047845721245,
      "num_tokens": 5160960.0,
      "step": 315
    },
    {
      "entropy": 3.9628808200359344,
      "epoch": 0.9362962962962963,
      "grad_norm": 1.0992919206619263,
      "learning_rate": 6.906906906906907e-07,
      "loss": 0.24035388231277466,
      "mean_token_accuracy": 0.9367947578430176,
      "num_tokens": 5177344.0,
      "step": 316
    },
    {
      "entropy": 4.431807905435562,
      "epoch": 0.9392592592592592,
      "grad_norm": 1.2781891822814941,
      "learning_rate": 6.606606606606607e-07,
      "loss": 0.1791629195213318,
      "mean_token_accuracy": 0.9492833316326141,
      "num_tokens": 5193728.0,
      "step": 317
    },
    {
      "entropy": 4.297941237688065,
      "epoch": 0.9422222222222222,
      "grad_norm": 1.0512442588806152,
      "learning_rate": 6.306306306306306e-07,
      "loss": 0.16492733359336853,
      "mean_token_accuracy": 0.951898604631424,
      "num_tokens": 5210112.0,
      "step": 318
    },
    {
      "entropy": 4.243958950042725,
      "epoch": 0.9451851851851852,
      "grad_norm": 0.9578456878662109,
      "learning_rate": 6.006006006006006e-07,
      "loss": 0.16450515389442444,
      "mean_token_accuracy": 0.954393781721592,
      "num_tokens": 5226496.0,
      "step": 319
    },
    {
      "entropy": 4.452836245298386,
      "epoch": 0.9481481481481482,
      "grad_norm": 1.022918462753296,
      "learning_rate": 5.705705705705706e-07,
      "loss": 0.14894741773605347,
      "mean_token_accuracy": 0.9637371152639389,
      "num_tokens": 5242880.0,
      "step": 320
    },
    {
      "entropy": 4.543944180011749,
      "epoch": 0.9511111111111111,
      "grad_norm": 0.9519062042236328,
      "learning_rate": 5.405405405405406e-07,
      "loss": 0.1621057391166687,
      "mean_token_accuracy": 0.9499952048063278,
      "num_tokens": 5259264.0,
      "step": 321
    },
    {
      "entropy": 4.615428626537323,
      "epoch": 0.9540740740740741,
      "grad_norm": 1.0308316946029663,
      "learning_rate": 5.105105105105106e-07,
      "loss": 0.17210763692855835,
      "mean_token_accuracy": 0.9504309743642807,
      "num_tokens": 5275648.0,
      "step": 322
    },
    {
      "entropy": 4.236121743917465,
      "epoch": 0.957037037037037,
      "grad_norm": 0.8866267800331116,
      "learning_rate": 4.804804804804805e-07,
      "loss": 0.15900775790214539,
      "mean_token_accuracy": 0.9550355970859528,
      "num_tokens": 5292032.0,
      "step": 323
    },
    {
      "entropy": 4.437883168458939,
      "epoch": 0.96,
      "grad_norm": 0.9690563082695007,
      "learning_rate": 4.504504504504505e-07,
      "loss": 0.12487293034791946,
      "mean_token_accuracy": 0.9603040590882301,
      "num_tokens": 5308416.0,
      "step": 324
    },
    {
      "entropy": 3.8638412952423096,
      "epoch": 0.9629629629629629,
      "grad_norm": 0.7880241274833679,
      "learning_rate": 4.204204204204205e-07,
      "loss": 0.11809857189655304,
      "mean_token_accuracy": 0.9722993820905685,
      "num_tokens": 5324800.0,
      "step": 325
    },
    {
      "entropy": 4.142101407051086,
      "epoch": 0.965925925925926,
      "grad_norm": 1.0829527378082275,
      "learning_rate": 3.903903903903904e-07,
      "loss": 0.19784386456012726,
      "mean_token_accuracy": 0.9507939890027046,
      "num_tokens": 5341184.0,
      "step": 326
    },
    {
      "entropy": 3.7648314833641052,
      "epoch": 0.9688888888888889,
      "grad_norm": 0.8108444809913635,
      "learning_rate": 3.603603603603604e-07,
      "loss": 0.11388225853443146,
      "mean_token_accuracy": 0.9735854864120483,
      "num_tokens": 5357568.0,
      "step": 327
    },
    {
      "entropy": 4.253404945135117,
      "epoch": 0.9718518518518519,
      "grad_norm": 1.1628628969192505,
      "learning_rate": 3.3033033033033036e-07,
      "loss": 0.20977286994457245,
      "mean_token_accuracy": 0.9444006159901619,
      "num_tokens": 5373952.0,
      "step": 328
    },
    {
      "entropy": 4.34212851524353,
      "epoch": 0.9748148148148148,
      "grad_norm": 1.091274619102478,
      "learning_rate": 3.003003003003003e-07,
      "loss": 0.19102801382541656,
      "mean_token_accuracy": 0.9428302198648453,
      "num_tokens": 5390336.0,
      "step": 329
    },
    {
      "entropy": 4.546658456325531,
      "epoch": 0.9777777777777777,
      "grad_norm": 0.9130251407623291,
      "learning_rate": 2.702702702702703e-07,
      "loss": 0.11526796221733093,
      "mean_token_accuracy": 0.9644840061664581,
      "num_tokens": 5406720.0,
      "step": 330
    },
    {
      "entropy": 4.924567639827728,
      "epoch": 0.9807407407407407,
      "grad_norm": 1.2413948774337769,
      "learning_rate": 2.4024024024024026e-07,
      "loss": 0.25382673740386963,
      "mean_token_accuracy": 0.938716322183609,
      "num_tokens": 5423104.0,
      "step": 331
    },
    {
      "entropy": 3.877889722585678,
      "epoch": 0.9837037037037037,
      "grad_norm": 0.9418118000030518,
      "learning_rate": 2.1021021021021025e-07,
      "loss": 0.17800109088420868,
      "mean_token_accuracy": 0.9424306154251099,
      "num_tokens": 5439488.0,
      "step": 332
    },
    {
      "entropy": 4.283163815736771,
      "epoch": 0.9866666666666667,
      "grad_norm": 1.2072510719299316,
      "learning_rate": 1.801801801801802e-07,
      "loss": 0.14721454679965973,
      "mean_token_accuracy": 0.952900804579258,
      "num_tokens": 5455872.0,
      "step": 333
    },
    {
      "entropy": 3.853197932243347,
      "epoch": 0.9896296296296296,
      "grad_norm": 0.88080894947052,
      "learning_rate": 1.5015015015015016e-07,
      "loss": 0.12530142068862915,
      "mean_token_accuracy": 0.9661237299442291,
      "num_tokens": 5472256.0,
      "step": 334
    },
    {
      "entropy": 4.61480712890625,
      "epoch": 0.9925925925925926,
      "grad_norm": 1.2803093194961548,
      "learning_rate": 1.2012012012012013e-07,
      "loss": 0.20693567395210266,
      "mean_token_accuracy": 0.9454390108585358,
      "num_tokens": 5488640.0,
      "step": 335
    },
    {
      "entropy": 4.899519920349121,
      "epoch": 0.9955555555555555,
      "grad_norm": 1.4463599920272827,
      "learning_rate": 9.00900900900901e-08,
      "loss": 0.24070829153060913,
      "mean_token_accuracy": 0.938837468624115,
      "num_tokens": 5505024.0,
      "step": 336
    },
    {
      "entropy": 4.374303579330444,
      "epoch": 0.9985185185185185,
      "grad_norm": 1.1831741333007812,
      "learning_rate": 6.006006006006006e-08,
      "loss": 0.1779567003250122,
      "mean_token_accuracy": 0.9493624940514565,
      "num_tokens": 5521408.0,
      "step": 337
    },
    {
      "entropy": 4.663045942783356,
      "epoch": 1.0,
      "grad_norm": 1.6296360492706299,
      "learning_rate": 3.003003003003003e-08,
      "loss": 0.17972803115844727,
      "mean_token_accuracy": 0.944612517952919,
      "num_tokens": 5529600.0,
      "step": 338
    }
  ],
  "logging_steps": 1,
  "max_steps": 338,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 5000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 9.53130594336768e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}