{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 5201,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 1.3228859663009644,
      "epoch": 0.013458950201884253,
      "grad_norm": 1.4003818035125732,
      "learning_rate": 2.4193548387096776e-06,
      "loss": 0.3947859287261963,
      "mean_token_accuracy": 0.8766823470592499,
      "num_tokens": 179351.0,
      "step": 10
    },
    {
      "entropy": 1.3223209857940674,
      "epoch": 0.026917900403768506,
      "grad_norm": 1.4293484687805176,
      "learning_rate": 5.1075268817204305e-06,
      "loss": 0.36992483139038085,
      "mean_token_accuracy": 0.8841759026050567,
      "num_tokens": 358681.0,
      "step": 20
    },
    {
      "entropy": 1.3241522073745728,
      "epoch": 0.040376850605652756,
      "grad_norm": 1.3410050868988037,
      "learning_rate": 7.795698924731183e-06,
      "loss": 0.3491047859191895,
      "mean_token_accuracy": 0.8843624591827393,
      "num_tokens": 538488.0,
      "step": 30
    },
    {
      "entropy": 1.3318131804466247,
      "epoch": 0.05383580080753701,
      "grad_norm": 1.08042573928833,
      "learning_rate": 1.0483870967741936e-05,
      "loss": 0.30254349708557127,
      "mean_token_accuracy": 0.8926964640617371,
      "num_tokens": 718118.0,
      "step": 40
    },
    {
      "entropy": 1.3337590098381042,
      "epoch": 0.06729475100942127,
      "grad_norm": 0.8615964651107788,
      "learning_rate": 1.3172043010752688e-05,
      "loss": 0.26572132110595703,
      "mean_token_accuracy": 0.9041078448295593,
      "num_tokens": 897209.0,
      "step": 50
    },
    {
      "entropy": 1.3126742005348206,
      "epoch": 0.08075370121130551,
      "grad_norm": 0.9317428469657898,
      "learning_rate": 1.586021505376344e-05,
      "loss": 0.23249435424804688,
      "mean_token_accuracy": 0.9135358154773712,
      "num_tokens": 1076448.0,
      "step": 60
    },
    {
      "entropy": 1.2961446046829224,
      "epoch": 0.09421265141318977,
      "grad_norm": 1.1822317838668823,
      "learning_rate": 1.8548387096774193e-05,
      "loss": 0.21359801292419434,
      "mean_token_accuracy": 0.9192376554012298,
      "num_tokens": 1255796.0,
      "step": 70
    },
    {
      "entropy": 1.2832220673561097,
      "epoch": 0.10767160161507403,
      "grad_norm": 1.1605602502822876,
      "learning_rate": 2.1236559139784946e-05,
      "loss": 0.20212924480438232,
      "mean_token_accuracy": 0.9217188417911529,
      "num_tokens": 1435611.0,
      "step": 80
    },
    {
      "entropy": 1.2762320637702942,
      "epoch": 0.12113055181695828,
      "grad_norm": 1.2006449699401855,
      "learning_rate": 2.39247311827957e-05,
      "loss": 0.18676402568817138,
      "mean_token_accuracy": 0.9310197174549103,
      "num_tokens": 1614872.0,
      "step": 90
    },
    {
      "entropy": 1.274475383758545,
      "epoch": 0.13458950201884254,
      "grad_norm": 1.2656220197677612,
      "learning_rate": 2.661290322580645e-05,
      "loss": 0.17775335311889648,
      "mean_token_accuracy": 0.9305342018604279,
      "num_tokens": 1794050.0,
      "step": 100
    },
    {
      "entropy": 1.2607086300849915,
      "epoch": 0.1480484522207268,
      "grad_norm": 1.2415285110473633,
      "learning_rate": 2.9301075268817207e-05,
      "loss": 0.17471678256988527,
      "mean_token_accuracy": 0.9317916095256805,
      "num_tokens": 1973292.0,
      "step": 110
    },
    {
      "entropy": 1.2618489027023316,
      "epoch": 0.16150740242261102,
      "grad_norm": 1.268917202949524,
      "learning_rate": 3.198924731182796e-05,
      "loss": 0.17752587795257568,
      "mean_token_accuracy": 0.9324639737606049,
      "num_tokens": 2152526.0,
      "step": 120
    },
    {
      "entropy": 1.2617292761802674,
      "epoch": 0.17496635262449528,
      "grad_norm": 1.1972988843917847,
      "learning_rate": 3.467741935483872e-05,
      "loss": 0.1649715781211853,
      "mean_token_accuracy": 0.9361909210681916,
      "num_tokens": 2331952.0,
      "step": 130
    },
    {
      "entropy": 1.2492274522781373,
      "epoch": 0.18842530282637954,
      "grad_norm": 1.2891550064086914,
      "learning_rate": 3.736559139784947e-05,
      "loss": 0.15520352125167847,
      "mean_token_accuracy": 0.939128577709198,
      "num_tokens": 2510125.0,
      "step": 140
    },
    {
      "entropy": 1.2332049250602721,
      "epoch": 0.2018842530282638,
      "grad_norm": 1.4108549356460571,
      "learning_rate": 4.005376344086022e-05,
      "loss": 0.16094473600387574,
      "mean_token_accuracy": 0.936911940574646,
      "num_tokens": 2690100.0,
      "step": 150
    },
    {
      "entropy": 1.2368563890457154,
      "epoch": 0.21534320323014805,
      "grad_norm": 1.2720197439193726,
      "learning_rate": 4.2741935483870973e-05,
      "loss": 0.15074102878570556,
      "mean_token_accuracy": 0.9419298887252807,
      "num_tokens": 2869072.0,
      "step": 160
    },
    {
      "entropy": 1.2400941371917724,
      "epoch": 0.2288021534320323,
      "grad_norm": 1.1434999704360962,
      "learning_rate": 4.543010752688172e-05,
      "loss": 0.15364220142364501,
      "mean_token_accuracy": 0.9395892560482025,
      "num_tokens": 3048179.0,
      "step": 170
    },
    {
      "entropy": 1.2370286583900452,
      "epoch": 0.24226110363391656,
      "grad_norm": 1.1543998718261719,
      "learning_rate": 4.811827956989248e-05,
      "loss": 0.1518303394317627,
      "mean_token_accuracy": 0.9403793513774872,
      "num_tokens": 3227648.0,
      "step": 180
    },
    {
      "entropy": 1.2381223678588866,
      "epoch": 0.2557200538358008,
      "grad_norm": 1.2674320936203003,
      "learning_rate": 5.080645161290323e-05,
      "loss": 0.14720194339752196,
      "mean_token_accuracy": 0.9413308918476104,
      "num_tokens": 3406443.0,
      "step": 190
    },
    {
      "entropy": 1.2209740161895752,
      "epoch": 0.2691790040376851,
      "grad_norm": 1.3895965814590454,
      "learning_rate": 5.349462365591398e-05,
      "loss": 0.14467229843139648,
      "mean_token_accuracy": 0.9425995826721192,
      "num_tokens": 3585139.0,
      "step": 200
    },
    {
      "entropy": 1.202421200275421,
      "epoch": 0.28263795423956933,
      "grad_norm": 1.1748970746994019,
      "learning_rate": 5.618279569892473e-05,
      "loss": 0.14418480396270753,
      "mean_token_accuracy": 0.944035142660141,
      "num_tokens": 3764815.0,
      "step": 210
    },
    {
      "entropy": 1.2019111514091492,
      "epoch": 0.2960969044414536,
      "grad_norm": 1.498073935508728,
      "learning_rate": 5.887096774193549e-05,
      "loss": 0.1409894585609436,
      "mean_token_accuracy": 0.9451915919780731,
      "num_tokens": 3944066.0,
      "step": 220
    },
    {
      "entropy": 1.208519995212555,
      "epoch": 0.30955585464333785,
      "grad_norm": 1.3500250577926636,
      "learning_rate": 6.155913978494624e-05,
      "loss": 0.14032076597213744,
      "mean_token_accuracy": 0.9437436699867249,
      "num_tokens": 4122561.0,
      "step": 230
    },
    {
      "entropy": 1.2071591854095458,
      "epoch": 0.32301480484522205,
      "grad_norm": 1.2740904092788696,
      "learning_rate": 6.4247311827957e-05,
      "loss": 0.13652472496032714,
      "mean_token_accuracy": 0.9458230376243592,
      "num_tokens": 4301558.0,
      "step": 240
    },
    {
      "entropy": 1.2087280392646789,
      "epoch": 0.3364737550471063,
      "grad_norm": 1.2498955726623535,
      "learning_rate": 6.693548387096774e-05,
      "loss": 0.137939453125,
      "mean_token_accuracy": 0.9447133004665375,
      "num_tokens": 4480708.0,
      "step": 250
    },
    {
      "entropy": 1.1996983647346497,
      "epoch": 0.34993270524899056,
      "grad_norm": 1.4520384073257446,
      "learning_rate": 6.962365591397851e-05,
      "loss": 0.14351414442062377,
      "mean_token_accuracy": 0.9436001539230346,
      "num_tokens": 4659616.0,
      "step": 260
    },
    {
      "entropy": 1.195546042919159,
      "epoch": 0.3633916554508748,
      "grad_norm": 1.1850366592407227,
      "learning_rate": 7.231182795698926e-05,
      "loss": 0.13713514804840088,
      "mean_token_accuracy": 0.946206146478653,
      "num_tokens": 4838619.0,
      "step": 270
    },
    {
      "entropy": 1.1904785871505736,
      "epoch": 0.3768506056527591,
      "grad_norm": 1.4309216737747192,
      "learning_rate": 7.500000000000001e-05,
      "loss": 0.13457468748092652,
      "mean_token_accuracy": 0.9473767280578613,
      "num_tokens": 5017668.0,
      "step": 280
    },
    {
      "entropy": 1.1928183078765868,
      "epoch": 0.39030955585464333,
      "grad_norm": 1.1399856805801392,
      "learning_rate": 7.768817204301076e-05,
      "loss": 0.13286283016204833,
      "mean_token_accuracy": 0.9468790471553803,
      "num_tokens": 5197412.0,
      "step": 290
    },
    {
      "entropy": 1.1921735286712647,
      "epoch": 0.4037685060565276,
      "grad_norm": 1.3263301849365234,
      "learning_rate": 8.037634408602151e-05,
      "loss": 0.1312130570411682,
      "mean_token_accuracy": 0.9481729447841645,
      "num_tokens": 5377026.0,
      "step": 300
    },
    {
      "entropy": 1.1928237080574036,
      "epoch": 0.41722745625841184,
      "grad_norm": 1.6233983039855957,
      "learning_rate": 8.306451612903227e-05,
      "loss": 0.1323886036872864,
      "mean_token_accuracy": 0.9465976357460022,
      "num_tokens": 5556708.0,
      "step": 310
    },
    {
      "entropy": 1.1836108446121216,
      "epoch": 0.4306864064602961,
      "grad_norm": 1.3058593273162842,
      "learning_rate": 8.575268817204302e-05,
      "loss": 0.13815442323684693,
      "mean_token_accuracy": 0.9450344681739807,
      "num_tokens": 5735495.0,
      "step": 320
    },
    {
      "entropy": 1.179207694530487,
      "epoch": 0.44414535666218036,
      "grad_norm": 1.43180513381958,
      "learning_rate": 8.844086021505377e-05,
      "loss": 0.13185629844665528,
      "mean_token_accuracy": 0.9473332643508912,
      "num_tokens": 5914707.0,
      "step": 330
    },
    {
      "entropy": 1.1725443840026855,
      "epoch": 0.4576043068640646,
      "grad_norm": 1.391575813293457,
      "learning_rate": 9.112903225806452e-05,
      "loss": 0.13046096563339232,
      "mean_token_accuracy": 0.9486047685146332,
      "num_tokens": 6093788.0,
      "step": 340
    },
    {
      "entropy": 1.17471581697464,
      "epoch": 0.47106325706594887,
      "grad_norm": 1.5146636962890625,
      "learning_rate": 9.381720430107528e-05,
      "loss": 0.13146305084228516,
      "mean_token_accuracy": 0.9477684259414673,
      "num_tokens": 6272916.0,
      "step": 350
    },
    {
      "entropy": 1.1633504509925843,
      "epoch": 0.4845222072678331,
      "grad_norm": 1.1611223220825195,
      "learning_rate": 9.650537634408603e-05,
      "loss": 0.12906880378723146,
      "mean_token_accuracy": 0.9491180777549744,
      "num_tokens": 6452766.0,
      "step": 360
    },
    {
      "entropy": 1.1688304781913756,
      "epoch": 0.4979811574697174,
      "grad_norm": 1.5618613958358765,
      "learning_rate": 9.919354838709678e-05,
      "loss": 0.1296112060546875,
      "mean_token_accuracy": 0.9494616508483886,
      "num_tokens": 6632087.0,
      "step": 370
    },
    {
      "entropy": 1.1618135571479797,
      "epoch": 0.5114401076716016,
      "grad_norm": 1.259940266609192,
      "learning_rate": 9.999975729865971e-05,
      "loss": 0.12792425155639647,
      "mean_token_accuracy": 0.9487492978572846,
      "num_tokens": 6811480.0,
      "step": 380
    },
    {
      "entropy": 1.1698922395706177,
      "epoch": 0.5248990578734859,
      "grad_norm": 1.4075767993927002,
      "learning_rate": 9.999856856307314e-05,
      "loss": 0.12948644161224365,
      "mean_token_accuracy": 0.9471172571182251,
      "num_tokens": 6990914.0,
      "step": 390
    },
    {
      "entropy": 1.1714325428009034,
      "epoch": 0.5383580080753702,
      "grad_norm": 1.2077672481536865,
      "learning_rate": 9.999638923896533e-05,
      "loss": 0.12964634895324706,
      "mean_token_accuracy": 0.9472773969173431,
      "num_tokens": 7170193.0,
      "step": 400
    },
    {
      "entropy": 1.172813320159912,
      "epoch": 0.5518169582772544,
      "grad_norm": 1.506191611289978,
      "learning_rate": 9.999321936951374e-05,
      "loss": 0.12280762195587158,
      "mean_token_accuracy": 0.9517758071422577,
      "num_tokens": 7349496.0,
      "step": 410
    },
    {
      "entropy": 1.179931104183197,
      "epoch": 0.5652759084791387,
      "grad_norm": 1.2185982465744019,
      "learning_rate": 9.998905901752091e-05,
      "loss": 0.12760610580444337,
      "mean_token_accuracy": 0.948979276418686,
      "num_tokens": 7528648.0,
      "step": 420
    },
    {
      "entropy": 1.1817825198173524,
      "epoch": 0.5787348586810229,
      "grad_norm": 1.285261631011963,
      "learning_rate": 9.998390826541315e-05,
      "loss": 0.13066411018371582,
      "mean_token_accuracy": 0.947555410861969,
      "num_tokens": 7707487.0,
      "step": 430
    },
    {
      "entropy": 1.1945344924926757,
      "epoch": 0.5921938088829072,
      "grad_norm": 1.2778966426849365,
      "learning_rate": 9.997776721523888e-05,
      "loss": 0.13003890514373778,
      "mean_token_accuracy": 0.9476695477962493,
      "num_tokens": 7886452.0,
      "step": 440
    },
    {
      "entropy": 1.192435622215271,
      "epoch": 0.6056527590847914,
      "grad_norm": 1.2676047086715698,
      "learning_rate": 9.99706359886667e-05,
      "loss": 0.13059219121932983,
      "mean_token_accuracy": 0.9467391848564148,
      "num_tokens": 8065093.0,
      "step": 450
    },
    {
      "entropy": 1.1952194094657898,
      "epoch": 0.6191117092866757,
      "grad_norm": 1.1667490005493164,
      "learning_rate": 9.996251472698281e-05,
      "loss": 0.1308892250061035,
      "mean_token_accuracy": 0.9474103152751923,
      "num_tokens": 8245294.0,
      "step": 460
    },
    {
      "entropy": 1.1830523014068604,
      "epoch": 0.6325706594885598,
      "grad_norm": 1.4168891906738281,
      "learning_rate": 9.995340359108844e-05,
      "loss": 0.1230043888092041,
      "mean_token_accuracy": 0.9503998339176178,
      "num_tokens": 8424334.0,
      "step": 470
    },
    {
      "entropy": 1.1928761839866637,
      "epoch": 0.6460296096904441,
      "grad_norm": 1.4041002988815308,
      "learning_rate": 9.994330276149649e-05,
      "loss": 0.12544957399368287,
      "mean_token_accuracy": 0.9496485233306885,
      "num_tokens": 8603758.0,
      "step": 480
    },
    {
      "entropy": 1.180042278766632,
      "epoch": 0.6594885598923284,
      "grad_norm": 1.2816451787948608,
      "learning_rate": 9.993221243832797e-05,
      "loss": 0.1197009801864624,
      "mean_token_accuracy": 0.9527871966361999,
      "num_tokens": 8782936.0,
      "step": 490
    },
    {
      "entropy": 1.1822824954986573,
      "epoch": 0.6729475100942126,
      "grad_norm": 1.0554879903793335,
      "learning_rate": 9.992013284130816e-05,
      "loss": 0.12416183948516846,
      "mean_token_accuracy": 0.9488660097122192,
      "num_tokens": 8962286.0,
      "step": 500
    },
    {
      "entropy": 1.1877296924591065,
      "epoch": 0.6864064602960969,
      "grad_norm": 2.272390842437744,
      "learning_rate": 9.990706420976206e-05,
      "loss": 0.12660024166107178,
      "mean_token_accuracy": 0.9495710134506226,
      "num_tokens": 9141244.0,
      "step": 510
    },
    {
      "entropy": 1.195984995365143,
      "epoch": 0.6998654104979811,
      "grad_norm": 1.106213927268982,
      "learning_rate": 9.989300680260985e-05,
      "loss": 0.12362114191055298,
      "mean_token_accuracy": 0.9512333691120147,
      "num_tokens": 9319796.0,
      "step": 520
    },
    {
      "entropy": 1.1847575902938843,
      "epoch": 0.7133243606998654,
      "grad_norm": 1.2385672330856323,
      "learning_rate": 9.98779608983616e-05,
      "loss": 0.12053300142288208,
      "mean_token_accuracy": 0.9517379641532898,
      "num_tokens": 9498637.0,
      "step": 530
    },
    {
      "entropy": 1.1887083888053893,
      "epoch": 0.7267833109017496,
      "grad_norm": 1.1992591619491577,
      "learning_rate": 9.986192679511189e-05,
      "loss": 0.12146525382995606,
      "mean_token_accuracy": 0.9509431838989257,
      "num_tokens": 9678136.0,
      "step": 540
    },
    {
      "entropy": 1.1956284999847413,
      "epoch": 0.7402422611036339,
      "grad_norm": 1.4346206188201904,
      "learning_rate": 9.984490481053372e-05,
      "loss": 0.12582865953445435,
      "mean_token_accuracy": 0.9495353937149048,
      "num_tokens": 9856943.0,
      "step": 550
    },
    {
      "entropy": 1.183466374874115,
      "epoch": 0.7537012113055181,
      "grad_norm": 1.1207399368286133,
      "learning_rate": 9.982689528187244e-05,
      "loss": 0.11938930749893188,
      "mean_token_accuracy": 0.9524446070194245,
      "num_tokens": 10036305.0,
      "step": 560
    },
    {
      "entropy": 1.1996586084365846,
      "epoch": 0.7671601615074024,
      "grad_norm": 1.0004290342330933,
      "learning_rate": 9.98078985659389e-05,
      "loss": 0.1256342649459839,
      "mean_token_accuracy": 0.9488184452056885,
      "num_tokens": 10215049.0,
      "step": 570
    },
    {
      "entropy": 1.2113700032234191,
      "epoch": 0.7806191117092867,
      "grad_norm": 1.243759036064148,
      "learning_rate": 9.978791503910246e-05,
      "loss": 0.11844713687896728,
      "mean_token_accuracy": 0.9520347356796265,
      "num_tokens": 10393513.0,
      "step": 580
    },
    {
      "entropy": 1.2171649575233459,
      "epoch": 0.7940780619111709,
      "grad_norm": 1.2175841331481934,
      "learning_rate": 9.97669450972835e-05,
      "loss": 0.1155052900314331,
      "mean_token_accuracy": 0.954187548160553,
      "num_tokens": 10572502.0,
      "step": 590
    },
    {
      "entropy": 1.2295325994491577,
      "epoch": 0.8075370121130552,
      "grad_norm": 1.1670854091644287,
      "learning_rate": 9.974498915594557e-05,
      "loss": 0.12255362272262574,
      "mean_token_accuracy": 0.9510588347911835,
      "num_tokens": 10751857.0,
      "step": 600
    },
    {
      "entropy": 1.2220084905624389,
      "epoch": 0.8209959623149394,
      "grad_norm": 1.3236212730407715,
      "learning_rate": 9.97220476500872e-05,
      "loss": 0.1217005968093872,
      "mean_token_accuracy": 0.9508337616920471,
      "num_tokens": 10931362.0,
      "step": 610
    },
    {
      "entropy": 1.20922110080719,
      "epoch": 0.8344549125168237,
      "grad_norm": 1.2529112100601196,
      "learning_rate": 9.969812103423325e-05,
      "loss": 0.11833038330078124,
      "mean_token_accuracy": 0.9529603838920593,
      "num_tokens": 11111075.0,
      "step": 620
    },
    {
      "entropy": 1.2208962082862853,
      "epoch": 0.847913862718708,
      "grad_norm": 1.2380986213684082,
      "learning_rate": 9.967320978242592e-05,
      "loss": 0.12019131183624268,
      "mean_token_accuracy": 0.9517916083335877,
      "num_tokens": 11289952.0,
      "step": 630
    },
    {
      "entropy": 1.206966769695282,
      "epoch": 0.8613728129205922,
      "grad_norm": 1.2476933002471924,
      "learning_rate": 9.964731438821533e-05,
      "loss": 0.11783044338226319,
      "mean_token_accuracy": 0.9523464858531951,
      "num_tokens": 11469661.0,
      "step": 640
    },
    {
      "entropy": 1.2062023997306823,
      "epoch": 0.8748317631224765,
      "grad_norm": 1.4155808687210083,
      "learning_rate": 9.962043536464978e-05,
      "loss": 0.12099127769470215,
      "mean_token_accuracy": 0.9519050180912018,
      "num_tokens": 11648570.0,
      "step": 650
    },
    {
      "entropy": 1.2050026655197144,
      "epoch": 0.8882907133243607,
      "grad_norm": 1.309507966041565,
      "learning_rate": 9.959257324426556e-05,
      "loss": 0.11565302610397339,
      "mean_token_accuracy": 0.9535071849822998,
      "num_tokens": 11827640.0,
      "step": 660
    },
    {
      "entropy": 1.2138132452964783,
      "epoch": 0.901749663526245,
      "grad_norm": 1.150227427482605,
      "learning_rate": 9.95637285790764e-05,
      "loss": 0.11565654277801514,
      "mean_token_accuracy": 0.9536015927791596,
      "num_tokens": 12006419.0,
      "step": 670
    },
    {
      "entropy": 1.2211383819580077,
      "epoch": 0.9152086137281292,
      "grad_norm": 1.3185595273971558,
      "learning_rate": 9.953390194056258e-05,
      "loss": 0.11686277389526367,
      "mean_token_accuracy": 0.9518564403057098,
      "num_tokens": 12184806.0,
      "step": 680
    },
    {
      "entropy": 1.233402180671692,
      "epoch": 0.9286675639300135,
      "grad_norm": 1.160781979560852,
      "learning_rate": 9.950309391965947e-05,
      "loss": 0.11723113059997559,
      "mean_token_accuracy": 0.9525671184062958,
      "num_tokens": 12363767.0,
      "step": 690
    },
    {
      "entropy": 1.2254271149635314,
      "epoch": 0.9421265141318977,
      "grad_norm": 1.0756208896636963,
      "learning_rate": 9.947130512674602e-05,
      "loss": 0.11969656944274902,
      "mean_token_accuracy": 0.9499428868293762,
      "num_tokens": 12542727.0,
      "step": 700
    },
    {
      "entropy": 1.2217535138130189,
      "epoch": 0.955585464333782,
      "grad_norm": 1.131346344947815,
      "learning_rate": 9.943853619163255e-05,
      "loss": 0.11605353355407715,
      "mean_token_accuracy": 0.9536243081092834,
      "num_tokens": 12721825.0,
      "step": 710
    },
    {
      "entropy": 1.2145210385322571,
      "epoch": 0.9690444145356663,
      "grad_norm": 1.0480105876922607,
      "learning_rate": 9.94047877635482e-05,
      "loss": 0.11278635263442993,
      "mean_token_accuracy": 0.9553675949573517,
      "num_tokens": 12902291.0,
      "step": 720
    },
    {
      "entropy": 1.2308586597442628,
      "epoch": 0.9825033647375505,
      "grad_norm": 1.1793105602264404,
      "learning_rate": 9.93700605111283e-05,
      "loss": 0.11050724983215332,
      "mean_token_accuracy": 0.9547911584377289,
      "num_tokens": 13082065.0,
      "step": 730
    },
    {
      "entropy": 1.2493423819541931,
      "epoch": 0.9959623149394348,
      "grad_norm": 1.289297103881836,
      "learning_rate": 9.933435512240084e-05,
      "loss": 0.11567041873931885,
      "mean_token_accuracy": 0.9526383280754089,
      "num_tokens": 13261041.0,
      "step": 740
    },
    {
      "epoch": 1.0,
      "eval_entropy": 1.229653848204643,
      "eval_loss": 0.11524093896150589,
      "eval_mean_token_accuracy": 0.9533016294430775,
      "eval_num_tokens": 13314766.0,
      "eval_runtime": 13.2833,
      "eval_samples_per_second": 376.412,
      "eval_steps_per_second": 11.819,
      "step": 743
    },
    {
      "entropy": 1.233021354675293,
      "epoch": 1.009421265141319,
      "grad_norm": 1.3314229249954224,
      "learning_rate": 9.929767230477305e-05,
      "loss": 0.10805211067199708,
      "mean_token_accuracy": 0.9576423108577728,
      "num_tokens": 13440438.0,
      "step": 750
    },
    {
      "entropy": 1.2277064085006715,
      "epoch": 1.0228802153432033,
      "grad_norm": 1.194751262664795,
      "learning_rate": 9.92600127850173e-05,
      "loss": 0.09916897416114807,
      "mean_token_accuracy": 0.9617054045200348,
      "num_tokens": 13619055.0,
      "step": 760
    },
    {
      "entropy": 1.2193793654441833,
      "epoch": 1.0363391655450875,
      "grad_norm": 1.3674660921096802,
      "learning_rate": 9.922137730925673e-05,
      "loss": 0.09446401596069336,
      "mean_token_accuracy": 0.9620481312274933,
      "num_tokens": 13798753.0,
      "step": 770
    },
    {
      "entropy": 1.2068825483322143,
      "epoch": 1.0497981157469718,
      "grad_norm": 1.175338625907898,
      "learning_rate": 9.918176664295041e-05,
      "loss": 0.09437270164489746,
      "mean_token_accuracy": 0.9626644790172577,
      "num_tokens": 13978445.0,
      "step": 780
    },
    {
      "entropy": 1.2036906003952026,
      "epoch": 1.063257065948856,
      "grad_norm": 1.3463256359100342,
      "learning_rate": 9.914118157087824e-05,
      "loss": 0.09322788715362548,
      "mean_token_accuracy": 0.9640586376190186,
      "num_tokens": 14157400.0,
      "step": 790
    },
    {
      "entropy": 1.195313036441803,
      "epoch": 1.0767160161507403,
      "grad_norm": 1.2407623529434204,
      "learning_rate": 9.909962289712538e-05,
      "loss": 0.10000712871551513,
      "mean_token_accuracy": 0.9604595184326172,
      "num_tokens": 14337209.0,
      "step": 800
    },
    {
      "entropy": 1.1997171878814696,
      "epoch": 1.0901749663526246,
      "grad_norm": 1.4212044477462769,
      "learning_rate": 9.905709144506629e-05,
      "loss": 0.09967402815818786,
      "mean_token_accuracy": 0.9616046726703644,
      "num_tokens": 14516327.0,
      "step": 810
    },
    {
      "entropy": 1.1916023015975952,
      "epoch": 1.1036339165545088,
      "grad_norm": 1.296561598777771,
      "learning_rate": 9.901358805734846e-05,
      "loss": 0.09139133095741273,
      "mean_token_accuracy": 0.9634343802928924,
      "num_tokens": 14695257.0,
      "step": 820
    },
    {
      "entropy": 1.1924808621406555,
      "epoch": 1.117092866756393,
      "grad_norm": 1.3003162145614624,
      "learning_rate": 9.89691135958757e-05,
      "loss": 0.0935364007949829,
      "mean_token_accuracy": 0.9624580383300781,
      "num_tokens": 14874609.0,
      "step": 830
    },
    {
      "entropy": 1.2110714197158814,
      "epoch": 1.1305518169582773,
      "grad_norm": 1.341585636138916,
      "learning_rate": 9.892366894179105e-05,
      "loss": 0.09882450699806214,
      "mean_token_accuracy": 0.961030250787735,
      "num_tokens": 15053971.0,
      "step": 840
    },
    {
      "entropy": 1.2037933468818665,
      "epoch": 1.1440107671601616,
      "grad_norm": 1.4529608488082886,
      "learning_rate": 9.887725499545937e-05,
      "loss": 0.09266124367713928,
      "mean_token_accuracy": 0.9641264617443085,
      "num_tokens": 15233217.0,
      "step": 850
    },
    {
      "entropy": 1.210195577144623,
      "epoch": 1.1574697173620458,
      "grad_norm": 0.9631540179252625,
      "learning_rate": 9.882987267644939e-05,
      "loss": 0.09560335874557495,
      "mean_token_accuracy": 0.9616110920906067,
      "num_tokens": 15412460.0,
      "step": 860
    },
    {
      "entropy": 1.210800564289093,
      "epoch": 1.17092866756393,
      "grad_norm": 1.078429937362671,
      "learning_rate": 9.878152292351563e-05,
      "loss": 0.0967819094657898,
      "mean_token_accuracy": 0.960888934135437,
      "num_tokens": 15590882.0,
      "step": 870
    },
    {
      "entropy": 1.204759907722473,
      "epoch": 1.1843876177658144,
      "grad_norm": 1.132325530052185,
      "learning_rate": 9.873220669457975e-05,
      "loss": 0.09479628801345825,
      "mean_token_accuracy": 0.9629071593284607,
      "num_tokens": 15770658.0,
      "step": 880
    },
    {
      "entropy": 1.2071414232254027,
      "epoch": 1.1978465679676986,
      "grad_norm": 1.0902807712554932,
      "learning_rate": 9.868192496671147e-05,
      "loss": 0.09629296064376831,
      "mean_token_accuracy": 0.9622378885746002,
      "num_tokens": 15950126.0,
      "step": 890
    },
    {
      "entropy": 1.2057334661483765,
      "epoch": 1.2113055181695827,
      "grad_norm": 1.2059770822525024,
      "learning_rate": 9.86306787361094e-05,
      "loss": 0.09764755368232728,
      "mean_token_accuracy": 0.962404465675354,
      "num_tokens": 16129315.0,
      "step": 900
    },
    {
      "entropy": 1.2059934735298157,
      "epoch": 1.224764468371467,
      "grad_norm": 1.5481969118118286,
      "learning_rate": 9.857846901808117e-05,
      "loss": 0.09670655727386475,
      "mean_token_accuracy": 0.9619201004505158,
      "num_tokens": 16307839.0,
      "step": 910
    },
    {
      "entropy": 1.2136071562767028,
      "epoch": 1.2382234185733512,
      "grad_norm": 1.1171293258666992,
      "learning_rate": 9.852529684702329e-05,
      "loss": 0.09502402544021607,
      "mean_token_accuracy": 0.9619876623153687,
      "num_tokens": 16487021.0,
      "step": 920
    },
    {
      "entropy": 1.232442605495453,
      "epoch": 1.2516823687752354,
      "grad_norm": 1.2836637496948242,
      "learning_rate": 9.847116327640082e-05,
      "loss": 0.09930729866027832,
      "mean_token_accuracy": 0.9604007601737976,
      "num_tokens": 16665995.0,
      "step": 930
    },
    {
      "entropy": 1.2321829080581665,
      "epoch": 1.2651413189771197,
      "grad_norm": 1.1843444108963013,
      "learning_rate": 9.841606937872632e-05,
      "loss": 0.10086537599563598,
      "mean_token_accuracy": 0.9602800250053406,
      "num_tokens": 16845090.0,
      "step": 940
    },
    {
      "entropy": 1.233613657951355,
      "epoch": 1.278600269179004,
      "grad_norm": 1.3496166467666626,
      "learning_rate": 9.836001624553869e-05,
      "loss": 0.09795907735824586,
      "mean_token_accuracy": 0.9610718429088593,
      "num_tokens": 17024295.0,
      "step": 950
    },
    {
      "entropy": 1.2220023155212403,
      "epoch": 1.2920592193808882,
      "grad_norm": 1.238175392150879,
      "learning_rate": 9.830300498738152e-05,
      "loss": 0.09709340333938599,
      "mean_token_accuracy": 0.9621975898742676,
      "num_tokens": 17203525.0,
      "step": 960
    },
    {
      "entropy": 1.2186771392822267,
      "epoch": 1.3055181695827724,
      "grad_norm": 1.0820763111114502,
      "learning_rate": 9.824503673378112e-05,
      "loss": 0.09260507822036743,
      "mean_token_accuracy": 0.9632427036762238,
      "num_tokens": 17382269.0,
      "step": 970
    },
    {
      "entropy": 1.212946391105652,
      "epoch": 1.3189771197846567,
      "grad_norm": 1.253194808959961,
      "learning_rate": 9.81861126332241e-05,
      "loss": 0.10012803077697754,
      "mean_token_accuracy": 0.9596389472484589,
      "num_tokens": 17561951.0,
      "step": 980
    },
    {
      "entropy": 1.2092903971672058,
      "epoch": 1.332436069986541,
      "grad_norm": 1.6471713781356812,
      "learning_rate": 9.812623385313461e-05,
      "loss": 0.1032632827758789,
      "mean_token_accuracy": 0.9594815850257874,
      "num_tokens": 17741116.0,
      "step": 990
    },
    {
      "entropy": 1.2158336997032166,
      "epoch": 1.3458950201884252,
      "grad_norm": 1.076393723487854,
      "learning_rate": 9.806540157985131e-05,
      "loss": 0.09857285022735596,
      "mean_token_accuracy": 0.9608540177345276,
      "num_tokens": 17920249.0,
      "step": 1000
    },
    {
      "entropy": 1.2093246698379516,
      "epoch": 1.3593539703903095,
      "grad_norm": 1.1203004121780396,
      "learning_rate": 9.800361701860368e-05,
      "loss": 0.09807900190353394,
      "mean_token_accuracy": 0.9611685931682586,
      "num_tokens": 18099006.0,
      "step": 1010
    },
    {
      "entropy": 1.2070690989494324,
      "epoch": 1.3728129205921937,
      "grad_norm": 1.3285764455795288,
      "learning_rate": 9.794088139348835e-05,
      "loss": 0.10283086299896241,
      "mean_token_accuracy": 0.9585156977176666,
      "num_tokens": 18277971.0,
      "step": 1020
    },
    {
      "entropy": 1.2022451281547546,
      "epoch": 1.386271870794078,
      "grad_norm": 1.0949617624282837,
      "learning_rate": 9.787719594744468e-05,
      "loss": 0.10161725282669068,
      "mean_token_accuracy": 0.9598902583122253,
      "num_tokens": 18457464.0,
      "step": 1030
    },
    {
      "entropy": 1.2045769929885863,
      "epoch": 1.3997308209959622,
      "grad_norm": 1.008150577545166,
      "learning_rate": 9.781256194223023e-05,
      "loss": 0.10038440227508545,
      "mean_token_accuracy": 0.960367614030838,
      "num_tokens": 18636876.0,
      "step": 1040
    },
    {
      "entropy": 1.204549217224121,
      "epoch": 1.4131897711978465,
      "grad_norm": 1.0495935678482056,
      "learning_rate": 9.774698065839577e-05,
      "loss": 0.09564157128334046,
      "mean_token_accuracy": 0.9625212967395782,
      "num_tokens": 18816243.0,
      "step": 1050
    },
    {
      "entropy": 1.2045063614845275,
      "epoch": 1.4266487213997308,
      "grad_norm": 1.2372835874557495,
      "learning_rate": 9.768045339525979e-05,
      "loss": 0.09781360626220703,
      "mean_token_accuracy": 0.9605839848518372,
      "num_tokens": 18995594.0,
      "step": 1060
    },
    {
      "entropy": 1.2258678078651428,
      "epoch": 1.440107671601615,
      "grad_norm": 1.0772687196731567,
      "learning_rate": 9.76129814708829e-05,
      "loss": 0.09291026592254639,
      "mean_token_accuracy": 0.9634248733520507,
      "num_tokens": 19173887.0,
      "step": 1070
    },
    {
      "entropy": 1.2230794191360475,
      "epoch": 1.4535666218034993,
      "grad_norm": 1.2008293867111206,
      "learning_rate": 9.754456622204167e-05,
      "loss": 0.09285001754760742,
      "mean_token_accuracy": 0.9633622407913208,
      "num_tokens": 19352678.0,
      "step": 1080
    },
    {
      "entropy": 1.2313218355178832,
      "epoch": 1.4670255720053835,
      "grad_norm": 1.5826188325881958,
      "learning_rate": 9.747520900420209e-05,
      "loss": 0.1002782940864563,
      "mean_token_accuracy": 0.9600823521614075,
      "num_tokens": 19532077.0,
      "step": 1090
    },
    {
      "entropy": 1.2246542692184448,
      "epoch": 1.4804845222072678,
      "grad_norm": 1.3970143795013428,
      "learning_rate": 9.740491119149277e-05,
      "loss": 0.1005969524383545,
      "mean_token_accuracy": 0.9596368432044983,
      "num_tokens": 19710609.0,
      "step": 1100
    },
    {
      "entropy": 1.207058048248291,
      "epoch": 1.493943472409152,
      "grad_norm": 1.3544780015945435,
      "learning_rate": 9.733367417667773e-05,
      "loss": 0.09367164373397827,
      "mean_token_accuracy": 0.9632523238658905,
      "num_tokens": 19889820.0,
      "step": 1110
    },
    {
      "entropy": 1.2027259588241577,
      "epoch": 1.5074024226110363,
      "grad_norm": 1.2393465042114258,
      "learning_rate": 9.726149937112873e-05,
      "loss": 0.09854428172111511,
      "mean_token_accuracy": 0.9612930059432984,
      "num_tokens": 20069561.0,
      "step": 1120
    },
    {
      "entropy": 1.2199820518493651,
      "epoch": 1.5208613728129206,
      "grad_norm": 1.4061861038208008,
      "learning_rate": 9.718838820479743e-05,
      "loss": 0.09687533378601074,
      "mean_token_accuracy": 0.9612306416034698,
      "num_tokens": 20249088.0,
      "step": 1130
    },
    {
      "entropy": 1.2114709615707397,
      "epoch": 1.5343203230148048,
      "grad_norm": 1.2970331907272339,
      "learning_rate": 9.711434212618691e-05,
      "loss": 0.09762253165245056,
      "mean_token_accuracy": 0.9609376013278961,
      "num_tokens": 20428600.0,
      "step": 1140
    },
    {
      "entropy": 1.1982413172721862,
      "epoch": 1.547779273216689,
      "grad_norm": 1.621308445930481,
      "learning_rate": 9.703936260232308e-05,
      "loss": 0.09679374098777771,
      "mean_token_accuracy": 0.9625207364559174,
      "num_tokens": 20608047.0,
      "step": 1150
    },
    {
      "entropy": 1.195889377593994,
      "epoch": 1.5612382234185733,
      "grad_norm": 1.2940045595169067,
      "learning_rate": 9.696345111872557e-05,
      "loss": 0.09699609279632568,
      "mean_token_accuracy": 0.96190345287323,
      "num_tokens": 20787142.0,
      "step": 1160
    },
    {
      "entropy": 1.1944554448127747,
      "epoch": 1.5746971736204576,
      "grad_norm": 1.3155335187911987,
      "learning_rate": 9.688660917937838e-05,
      "loss": 0.09831242561340332,
      "mean_token_accuracy": 0.9606768429279328,
      "num_tokens": 20966230.0,
      "step": 1170
    },
    {
      "entropy": 1.1957345604896545,
      "epoch": 1.5881561238223418,
      "grad_norm": 1.2948030233383179,
      "learning_rate": 9.68088383066999e-05,
      "loss": 0.09834452867507934,
      "mean_token_accuracy": 0.9612827241420746,
      "num_tokens": 21145768.0,
      "step": 1180
    },
    {
      "entropy": 1.2020023703575133,
      "epoch": 1.601615074024226,
      "grad_norm": 1.0523329973220825,
      "learning_rate": 9.673014004151292e-05,
      "loss": 0.09663949012756348,
      "mean_token_accuracy": 0.9620199799537659,
      "num_tokens": 21324592.0,
      "step": 1190
    },
    {
      "entropy": 1.1892358779907226,
      "epoch": 1.6150740242261103,
      "grad_norm": 1.1584330797195435,
      "learning_rate": 9.665051594301407e-05,
      "loss": 0.09669581055641174,
      "mean_token_accuracy": 0.961614978313446,
      "num_tokens": 21504539.0,
      "step": 1200
    },
    {
      "entropy": 1.190696406364441,
      "epoch": 1.6285329744279946,
      "grad_norm": 1.1194695234298706,
      "learning_rate": 9.656996758874284e-05,
      "loss": 0.09648081660270691,
      "mean_token_accuracy": 0.9612169206142426,
      "num_tokens": 21683905.0,
      "step": 1210
    },
    {
      "entropy": 1.2076977849006654,
      "epoch": 1.6419919246298789,
      "grad_norm": 1.1297376155853271,
      "learning_rate": 9.648849657455044e-05,
      "loss": 0.09605686664581299,
      "mean_token_accuracy": 0.961658376455307,
      "num_tokens": 21862162.0,
      "step": 1220
    },
    {
      "entropy": 1.2096962213516236,
      "epoch": 1.6554508748317631,
      "grad_norm": 1.2401906251907349,
      "learning_rate": 9.640610451456811e-05,
      "loss": 0.09206328392028809,
      "mean_token_accuracy": 0.962989890575409,
      "num_tokens": 22041015.0,
      "step": 1230
    },
    {
      "entropy": 1.216509222984314,
      "epoch": 1.6689098250336474,
      "grad_norm": 1.2637176513671875,
      "learning_rate": 9.632279304117517e-05,
      "loss": 0.09614999294281006,
      "mean_token_accuracy": 0.9613571405410767,
      "num_tokens": 22220655.0,
      "step": 1240
    },
    {
      "entropy": 1.2126171827316283,
      "epoch": 1.6823687752355316,
      "grad_norm": 1.2879180908203125,
      "learning_rate": 9.623856380496664e-05,
      "loss": 0.09818092584609986,
      "mean_token_accuracy": 0.9603166699409484,
      "num_tokens": 22399957.0,
      "step": 1250
    },
    {
      "entropy": 1.182448434829712,
      "epoch": 1.695827725437416,
      "grad_norm": 1.0547544956207275,
      "learning_rate": 9.615341847472059e-05,
      "loss": 0.0945388674736023,
      "mean_token_accuracy": 0.9623521089553833,
      "num_tokens": 22579222.0,
      "step": 1260
    },
    {
      "entropy": 1.1886864185333252,
      "epoch": 1.7092866756393001,
      "grad_norm": 1.4119364023208618,
      "learning_rate": 9.606735873736505e-05,
      "loss": 0.0979494333267212,
      "mean_token_accuracy": 0.9607987105846405,
      "num_tokens": 22758487.0,
      "step": 1270
    },
    {
      "entropy": 1.1918489813804627,
      "epoch": 1.7227456258411844,
      "grad_norm": 1.2551711797714233,
      "learning_rate": 9.598038629794461e-05,
      "loss": 0.09586712718009949,
      "mean_token_accuracy": 0.9615644454956055,
      "num_tokens": 22936708.0,
      "step": 1280
    },
    {
      "entropy": 1.1913212060928344,
      "epoch": 1.7362045760430687,
      "grad_norm": 1.0276069641113281,
      "learning_rate": 9.589250287958657e-05,
      "loss": 0.09535220861434937,
      "mean_token_accuracy": 0.9606883823871613,
      "num_tokens": 23116329.0,
      "step": 1290
    },
    {
      "entropy": 1.203085219860077,
      "epoch": 1.749663526244953,
      "grad_norm": 1.2456278800964355,
      "learning_rate": 9.580371022346693e-05,
      "loss": 0.09598281383514404,
      "mean_token_accuracy": 0.9608144044876099,
      "num_tokens": 23295164.0,
      "step": 1300
    },
    {
      "entropy": 1.1892922878265382,
      "epoch": 1.7631224764468372,
      "grad_norm": 1.1159876585006714,
      "learning_rate": 9.571401008877572e-05,
      "loss": 0.09096106886863708,
      "mean_token_accuracy": 0.9636982321739197,
      "num_tokens": 23474377.0,
      "step": 1310
    },
    {
      "entropy": 1.2096730828285218,
      "epoch": 1.7765814266487214,
      "grad_norm": 1.420886516571045,
      "learning_rate": 9.562340425268233e-05,
      "loss": 0.0925740659236908,
      "mean_token_accuracy": 0.9629011929035187,
      "num_tokens": 23653389.0,
      "step": 1320
    },
    {
      "entropy": 1.2122852802276611,
      "epoch": 1.7900403768506057,
      "grad_norm": 1.1587319374084473,
      "learning_rate": 9.553189451030019e-05,
      "loss": 0.09554123878479004,
      "mean_token_accuracy": 0.9622859060764313,
      "num_tokens": 23832469.0,
      "step": 1330
    },
    {
      "entropy": 1.2176487922668457,
      "epoch": 1.80349932705249,
      "grad_norm": 1.147444248199463,
      "learning_rate": 9.543948267465115e-05,
      "loss": 0.09707238674163818,
      "mean_token_accuracy": 0.9612141191959381,
      "num_tokens": 24011518.0,
      "step": 1340
    },
    {
      "entropy": 1.2231361389160156,
      "epoch": 1.8169582772543742,
      "grad_norm": 1.1775709390640259,
      "learning_rate": 9.534617057662977e-05,
      "loss": 0.09654755592346191,
      "mean_token_accuracy": 0.9617958247661591,
      "num_tokens": 24190267.0,
      "step": 1350
    },
    {
      "entropy": 1.2070120811462401,
      "epoch": 1.8304172274562585,
      "grad_norm": 1.1315947771072388,
      "learning_rate": 9.525196006496679e-05,
      "loss": 0.09382581114768981,
      "mean_token_accuracy": 0.9625270128250122,
      "num_tokens": 24369982.0,
      "step": 1360
    },
    {
      "entropy": 1.2057390093803406,
      "epoch": 1.8438761776581427,
      "grad_norm": 1.1973934173583984,
      "learning_rate": 9.515685300619271e-05,
      "loss": 0.09683746099472046,
      "mean_token_accuracy": 0.9607476830482483,
      "num_tokens": 24549256.0,
      "step": 1370
    },
    {
      "entropy": 1.207427191734314,
      "epoch": 1.857335127860027,
      "grad_norm": 1.3193334341049194,
      "learning_rate": 9.506085128460065e-05,
      "loss": 0.09461041688919067,
      "mean_token_accuracy": 0.9628551185131073,
      "num_tokens": 24727544.0,
      "step": 1380
    },
    {
      "entropy": 1.2011541604995728,
      "epoch": 1.8707940780619112,
      "grad_norm": 1.0681352615356445,
      "learning_rate": 9.496395680220918e-05,
      "loss": 0.0960330069065094,
      "mean_token_accuracy": 0.9622460305690765,
      "num_tokens": 24907721.0,
      "step": 1390
    },
    {
      "entropy": 1.2034499764442443,
      "epoch": 1.8842530282637955,
      "grad_norm": 1.2765101194381714,
      "learning_rate": 9.486617147872446e-05,
      "loss": 0.09376740455627441,
      "mean_token_accuracy": 0.9624415040016174,
      "num_tokens": 25086496.0,
      "step": 1400
    },
    {
      "entropy": 1.1899038195610045,
      "epoch": 1.8977119784656797,
      "grad_norm": 1.1333132982254028,
      "learning_rate": 9.476749725150235e-05,
      "loss": 0.09668049812316895,
      "mean_token_accuracy": 0.9621514558792115,
      "num_tokens": 25266204.0,
      "step": 1410
    },
    {
      "entropy": 1.1895189881324768,
      "epoch": 1.911170928667564,
      "grad_norm": 1.310587763786316,
      "learning_rate": 9.466793607550995e-05,
      "loss": 0.0920013129711151,
      "mean_token_accuracy": 0.963368022441864,
      "num_tokens": 25445604.0,
      "step": 1420
    },
    {
      "entropy": 1.1946001529693604,
      "epoch": 1.9246298788694483,
      "grad_norm": 1.3619959354400635,
      "learning_rate": 9.45674899232869e-05,
      "loss": 0.09906838536262512,
      "mean_token_accuracy": 0.9604476511478424,
      "num_tokens": 25624947.0,
      "step": 1430
    },
    {
      "entropy": 1.2078219771385192,
      "epoch": 1.9380888290713325,
      "grad_norm": 1.152220606803894,
      "learning_rate": 9.446616078490626e-05,
      "loss": 0.09479650259017944,
      "mean_token_accuracy": 0.9627270400524139,
      "num_tokens": 25804643.0,
      "step": 1440
    },
    {
      "entropy": 1.2207356214523315,
      "epoch": 1.9515477792732168,
      "grad_norm": 1.186161994934082,
      "learning_rate": 9.436395066793518e-05,
      "loss": 0.09636704921722412,
      "mean_token_accuracy": 0.9604843854904175,
      "num_tokens": 25984119.0,
      "step": 1450
    },
    {
      "entropy": 1.205849301815033,
      "epoch": 1.965006729475101,
      "grad_norm": 1.409846544265747,
      "learning_rate": 9.426086159739496e-05,
      "loss": 0.09743249416351318,
      "mean_token_accuracy": 0.9608718931674958,
      "num_tokens": 26163483.0,
      "step": 1460
    },
    {
      "entropy": 1.2089114785194397,
      "epoch": 1.9784656796769853,
      "grad_norm": 1.2226805686950684,
      "learning_rate": 9.415689561572107e-05,
      "loss": 0.09131012558937072,
      "mean_token_accuracy": 0.9631811439990997,
      "num_tokens": 26342666.0,
      "step": 1470
    },
    {
      "entropy": 1.1953012466430664,
      "epoch": 1.9919246298788695,
      "grad_norm": 1.0700947046279907,
      "learning_rate": 9.405205478272267e-05,
      "loss": 0.09140577316284179,
      "mean_token_accuracy": 0.9642649590969086,
      "num_tokens": 26521895.0,
      "step": 1480
    },
    {
      "epoch": 2.0,
      "eval_entropy": 1.1869331590688912,
      "eval_loss": 0.10991495102643967,
      "eval_mean_token_accuracy": 0.9554494638351878,
      "eval_num_tokens": 26629596.0,
      "eval_runtime": 12.7631,
      "eval_samples_per_second": 391.753,
      "eval_steps_per_second": 12.301,
      "step": 1486
    },
    {
      "entropy": 1.1832876205444336,
      "epoch": 2.005383580080754,
      "grad_norm": 1.0044941902160645,
      "learning_rate": 9.394634117554173e-05,
      "loss": 0.0840892255306244,
      "mean_token_accuracy": 0.967725521326065,
      "num_tokens": 26701394.0,
      "step": 1490
    },
    {
      "entropy": 1.159238350391388,
      "epoch": 2.018842530282638,
      "grad_norm": 1.4198471307754517,
      "learning_rate": 9.38397568886119e-05,
      "loss": 0.07137876152992248,
      "mean_token_accuracy": 0.9723187386989594,
      "num_tokens": 26880890.0,
      "step": 1500
    },
    {
      "entropy": 1.165151631832123,
      "epoch": 2.0323014804845223,
      "grad_norm": 1.1602118015289307,
      "learning_rate": 9.373230403361712e-05,
      "loss": 0.06463043689727783,
      "mean_token_accuracy": 0.9757274091243744,
      "num_tokens": 27059741.0,
      "step": 1510
    },
    {
      "entropy": 1.1644548654556275,
      "epoch": 2.0457604306864066,
      "grad_norm": 1.3322592973709106,
      "learning_rate": 9.362398473944958e-05,
      "loss": 0.07388677597045898,
      "mean_token_accuracy": 0.971617478132248,
      "num_tokens": 27238125.0,
      "step": 1520
    },
    {
      "entropy": 1.1564043641090394,
      "epoch": 2.059219380888291,
      "grad_norm": 1.1690629720687866,
      "learning_rate": 9.35148011521677e-05,
      "loss": 0.06990204453468322,
      "mean_token_accuracy": 0.9719981133937836,
      "num_tokens": 27417172.0,
      "step": 1530
    },
    {
      "entropy": 1.1576861262321472,
      "epoch": 2.072678331090175,
      "grad_norm": 1.7016727924346924,
      "learning_rate": 9.340475543495364e-05,
      "loss": 0.06850625276565551,
      "mean_token_accuracy": 0.9732699453830719,
      "num_tokens": 27596848.0,
      "step": 1540
    },
    {
      "entropy": 1.1597527265548706,
      "epoch": 2.0861372812920593,
      "grad_norm": 1.1524600982666016,
      "learning_rate": 9.329384976807023e-05,
      "loss": 0.06980778574943543,
      "mean_token_accuracy": 0.9729204118251801,
      "num_tokens": 27775617.0,
      "step": 1550
    },
    {
      "entropy": 1.1575467109680175,
      "epoch": 2.0995962314939436,
      "grad_norm": 1.4498176574707031,
      "learning_rate": 9.318208634881802e-05,
      "loss": 0.07390267252922059,
      "mean_token_accuracy": 0.9713942348957062,
      "num_tokens": 27954133.0,
      "step": 1560
    },
    {
      "entropy": 1.1575489521026612,
      "epoch": 2.113055181695828,
      "grad_norm": 1.243706464767456,
      "learning_rate": 9.306946739149161e-05,
      "loss": 0.06798491477966309,
      "mean_token_accuracy": 0.973270720243454,
      "num_tokens": 28133292.0,
      "step": 1570
    },
    {
      "entropy": 1.1514037609100343,
      "epoch": 2.126514131897712,
      "grad_norm": 1.256933331489563,
      "learning_rate": 9.29559951273358e-05,
      "loss": 0.0749699592590332,
      "mean_token_accuracy": 0.9699350416660308,
      "num_tokens": 28312726.0,
      "step": 1580
    },
    {
      "entropy": 1.1511113524436951,
      "epoch": 2.1399730820995964,
      "grad_norm": 1.1914122104644775,
      "learning_rate": 9.284167180450141e-05,
      "loss": 0.06752681732177734,
      "mean_token_accuracy": 0.9743177771568299,
      "num_tokens": 28492614.0,
      "step": 1590
    },
    {
      "entropy": 1.1382261991500855,
      "epoch": 2.1534320323014806,
      "grad_norm": 1.109575867652893,
      "learning_rate": 9.272649968800069e-05,
      "loss": 0.06449686884880065,
      "mean_token_accuracy": 0.9755833566188812,
      "num_tokens": 28671719.0,
      "step": 1600
    },
    {
      "entropy": 1.1397038459777833,
      "epoch": 2.166890982503365,
      "grad_norm": 1.3151781558990479,
      "learning_rate": 9.26104810596625e-05,
      "loss": 0.07052424550056458,
      "mean_token_accuracy": 0.972499680519104,
      "num_tokens": 28851056.0,
      "step": 1610
    },
    {
      "entropy": 1.136834406852722,
      "epoch": 2.180349932705249,
      "grad_norm": 1.4410197734832764,
      "learning_rate": 9.249361821808708e-05,
      "loss": 0.06850321292877197,
      "mean_token_accuracy": 0.9728572845458985,
      "num_tokens": 29030750.0,
      "step": 1620
    },
    {
      "entropy": 1.140534520149231,
      "epoch": 2.1938088829071334,
      "grad_norm": 1.1493765115737915,
      "learning_rate": 9.237591347860052e-05,
      "loss": 0.06934296488761901,
      "mean_token_accuracy": 0.972960364818573,
      "num_tokens": 29210336.0,
      "step": 1630
    },
    {
      "entropy": 1.138876986503601,
      "epoch": 2.2072678331090176,
      "grad_norm": 1.0383925437927246,
      "learning_rate": 9.225736917320886e-05,
      "loss": 0.06877213716506958,
      "mean_token_accuracy": 0.9730811774730682,
      "num_tokens": 29389788.0,
      "step": 1640
    },
    {
      "entropy": 1.1449824213981628,
      "epoch": 2.220726783310902,
      "grad_norm": 1.374165654182434,
      "learning_rate": 9.213798765055187e-05,
      "loss": 0.07060860991477966,
      "mean_token_accuracy": 0.9721363008022308,
      "num_tokens": 29569111.0,
      "step": 1650
    },
    {
      "entropy": 1.143638014793396,
      "epoch": 2.234185733512786,
      "grad_norm": 1.0196412801742554,
      "learning_rate": 9.20177712758566e-05,
      "loss": 0.07119340896606445,
      "mean_token_accuracy": 0.9731849789619446,
      "num_tokens": 29747986.0,
      "step": 1660
    },
    {
      "entropy": 1.1349515676498414,
      "epoch": 2.2476446837146704,
      "grad_norm": 1.1247919797897339,
      "learning_rate": 9.189672243089046e-05,
      "loss": 0.07071832418441773,
      "mean_token_accuracy": 0.9731756567955017,
      "num_tokens": 29927276.0,
      "step": 1670
    },
    {
      "entropy": 1.1376260280609132,
      "epoch": 2.2611036339165547,
      "grad_norm": 1.4197320938110352,
      "learning_rate": 9.177484351391402e-05,
      "loss": 0.07115572690963745,
      "mean_token_accuracy": 0.9723047018051147,
      "num_tokens": 30106267.0,
      "step": 1680
    },
    {
      "entropy": 1.1282797813415528,
      "epoch": 2.274562584118439,
      "grad_norm": 1.0774035453796387,
      "learning_rate": 9.165213693963355e-05,
      "loss": 0.068689626455307,
      "mean_token_accuracy": 0.9729084491729736,
      "num_tokens": 30285658.0,
      "step": 1690
    },
    {
      "entropy": 1.140774166584015,
      "epoch": 2.288021534320323,
      "grad_norm": 1.5625728368759155,
      "learning_rate": 9.152860513915314e-05,
      "loss": 0.07172787189483643,
      "mean_token_accuracy": 0.9718713641166687,
      "num_tokens": 30464579.0,
      "step": 1700
    },
    {
      "entropy": 1.1416746616363525,
      "epoch": 2.3014804845222074,
      "grad_norm": 1.2159788608551025,
      "learning_rate": 9.140425055992648e-05,
      "loss": 0.07109695672988892,
      "mean_token_accuracy": 0.9723007261753083,
      "num_tokens": 30643566.0,
      "step": 1710
    },
    {
      "entropy": 1.137896478176117,
      "epoch": 2.3149394347240917,
      "grad_norm": 1.1671864986419678,
      "learning_rate": 9.127907566570853e-05,
      "loss": 0.07048168182373046,
      "mean_token_accuracy": 0.9725371599197388,
      "num_tokens": 30822593.0,
      "step": 1720
    },
    {
      "entropy": 1.121722447872162,
      "epoch": 2.328398384925976,
      "grad_norm": 1.3899428844451904,
      "learning_rate": 9.115308293650653e-05,
      "loss": 0.07030471563339233,
      "mean_token_accuracy": 0.972334086894989,
      "num_tokens": 31001872.0,
      "step": 1730
    },
    {
      "entropy": 1.1253960013389588,
      "epoch": 2.34185733512786,
      "grad_norm": 1.3878816366195679,
      "learning_rate": 9.102627486853099e-05,
      "loss": 0.06956568956375123,
      "mean_token_accuracy": 0.9728380262851715,
      "num_tokens": 31181325.0,
      "step": 1740
    },
    {
      "entropy": 1.1276877760887145,
      "epoch": 2.3553162853297445,
      "grad_norm": 1.057102084159851,
      "learning_rate": 9.089865397414614e-05,
      "loss": 0.07267707586288452,
      "mean_token_accuracy": 0.9716822624206543,
      "num_tokens": 31360101.0,
      "step": 1750
    },
    {
      "entropy": 1.1322511553764343,
      "epoch": 2.3687752355316287,
      "grad_norm": 0.9549157023429871,
      "learning_rate": 9.077022278182024e-05,
      "loss": 0.0700565218925476,
      "mean_token_accuracy": 0.9737613677978516,
      "num_tokens": 31539212.0,
      "step": 1760
    },
    {
      "entropy": 1.1198163986206056,
      "epoch": 2.382234185733513,
      "grad_norm": 1.2134599685668945,
      "learning_rate": 9.064098383607545e-05,
      "loss": 0.07131816148757934,
      "mean_token_accuracy": 0.9710030317306518,
      "num_tokens": 31718375.0,
      "step": 1770
    },
    {
      "entropy": 1.1219655275344849,
      "epoch": 2.3956931359353972,
      "grad_norm": 1.3444453477859497,
      "learning_rate": 9.051093969743738e-05,
      "loss": 0.06774230003356933,
      "mean_token_accuracy": 0.9737627685070038,
      "num_tokens": 31897547.0,
      "step": 1780
    },
    {
      "entropy": 1.112294065952301,
      "epoch": 2.409152086137281,
      "grad_norm": 1.5304001569747925,
      "learning_rate": 9.03800929423844e-05,
      "loss": 0.0724343478679657,
      "mean_token_accuracy": 0.9721882760524749,
      "num_tokens": 32076717.0,
      "step": 1790
    },
    {
      "entropy": 1.1336388826370238,
      "epoch": 2.4226110363391653,
      "grad_norm": 1.2315043210983276,
      "learning_rate": 9.024844616329662e-05,
      "loss": 0.07212550640106201,
      "mean_token_accuracy": 0.9726522386074066,
      "num_tokens": 32255927.0,
      "step": 1800
    },
    {
      "entropy": 1.1251046061515808,
      "epoch": 2.4360699865410496,
      "grad_norm": 1.3251651525497437,
      "learning_rate": 9.011600196840447e-05,
      "loss": 0.07009173035621644,
      "mean_token_accuracy": 0.9724473178386688,
      "num_tokens": 32434929.0,
      "step": 1810
    },
    {
      "entropy": 1.1328514456748962,
      "epoch": 2.449528936742934,
      "grad_norm": 1.2601144313812256,
      "learning_rate": 8.998276298173707e-05,
      "loss": 0.0719257116317749,
      "mean_token_accuracy": 0.9720721006393432,
      "num_tokens": 32614063.0,
      "step": 1820
    },
    {
      "entropy": 1.136410367488861,
      "epoch": 2.462987886944818,
      "grad_norm": 1.2086918354034424,
      "learning_rate": 8.984873184307017e-05,
      "loss": 0.07017306089401246,
      "mean_token_accuracy": 0.9722030460834503,
      "num_tokens": 32793068.0,
      "step": 1830
    },
    {
      "entropy": 1.13520849943161,
      "epoch": 2.4764468371467023,
      "grad_norm": 1.453801155090332,
      "learning_rate": 8.971391120787397e-05,
      "loss": 0.07180649638175965,
      "mean_token_accuracy": 0.9726110398769379,
      "num_tokens": 32972445.0,
      "step": 1840
    },
    {
      "entropy": 1.1437011241912842,
      "epoch": 2.4899057873485866,
      "grad_norm": 1.1886014938354492,
      "learning_rate": 8.957830374726042e-05,
      "loss": 0.07153818607330323,
      "mean_token_accuracy": 0.9720338463783265,
      "num_tokens": 33151976.0,
      "step": 1850
    },
    {
      "entropy": 1.1325899600982665,
      "epoch": 2.503364737550471,
      "grad_norm": 1.1960384845733643,
      "learning_rate": 8.944191214793028e-05,
      "loss": 0.06935594081878663,
      "mean_token_accuracy": 0.9729972183704376,
      "num_tokens": 33330611.0,
      "step": 1860
    },
    {
      "entropy": 1.127972149848938,
      "epoch": 2.516823687752355,
      "grad_norm": 1.1048696041107178,
      "learning_rate": 8.930473911212e-05,
      "loss": 0.07217252850532532,
      "mean_token_accuracy": 0.9718475580215454,
      "num_tokens": 33509614.0,
      "step": 1870
    },
    {
      "entropy": 1.1299184799194335,
      "epoch": 2.5302826379542394,
      "grad_norm": 1.6520979404449463,
      "learning_rate": 8.916678735754809e-05,
      "loss": 0.07317680716514588,
      "mean_token_accuracy": 0.971524566411972,
      "num_tokens": 33688724.0,
      "step": 1880
    },
    {
      "entropy": 1.1270474672317505,
      "epoch": 2.5437415881561236,
      "grad_norm": 1.1285676956176758,
      "learning_rate": 8.902805961736123e-05,
      "loss": 0.07085765600204467,
      "mean_token_accuracy": 0.9733552634716034,
      "num_tokens": 33868061.0,
      "step": 1890
    },
    {
      "entropy": 1.1217491984367371,
      "epoch": 2.557200538358008,
      "grad_norm": 1.2642406225204468,
      "learning_rate": 8.88885586400803e-05,
      "loss": 0.06978695392608643,
      "mean_token_accuracy": 0.9726768732070923,
      "num_tokens": 34047387.0,
      "step": 1900
    },
    {
      "entropy": 1.1316059112548829,
      "epoch": 2.570659488559892,
      "grad_norm": 1.3016549348831177,
      "learning_rate": 8.874828718954576e-05,
      "loss": 0.07102057337760925,
      "mean_token_accuracy": 0.9723258554935456,
      "num_tokens": 34227141.0,
      "step": 1910
    },
    {
      "entropy": 1.137197768688202,
      "epoch": 2.5841184387617764,
      "grad_norm": 1.1534605026245117,
      "learning_rate": 8.86072480448629e-05,
      "loss": 0.07511197328567505,
      "mean_token_accuracy": 0.9702324509620667,
      "num_tokens": 34406526.0,
      "step": 1920
    },
    {
      "entropy": 1.1349146008491515,
      "epoch": 2.5975773889636606,
      "grad_norm": 1.3732489347457886,
      "learning_rate": 8.84654440003469e-05,
      "loss": 0.07147140502929687,
      "mean_token_accuracy": 0.9726706743240356,
      "num_tokens": 34586116.0,
      "step": 1930
    },
    {
      "entropy": 1.1294147491455078,
      "epoch": 2.611036339165545,
      "grad_norm": 0.9715967178344727,
      "learning_rate": 8.83228778654674e-05,
      "loss": 0.07225455045700073,
      "mean_token_accuracy": 0.9726594388484955,
      "num_tokens": 34765289.0,
      "step": 1940
    },
    {
      "entropy": 1.130816388130188,
      "epoch": 2.624495289367429,
      "grad_norm": 1.293736219406128,
      "learning_rate": 8.817955246479276e-05,
      "loss": 0.06845389604568482,
      "mean_token_accuracy": 0.9736943006515503,
      "num_tokens": 34944224.0,
      "step": 1950
    },
    {
      "entropy": 1.1200148224830628,
      "epoch": 2.6379542395693134,
      "grad_norm": 1.2962090969085693,
      "learning_rate": 8.803547063793422e-05,
      "loss": 0.07189736366271973,
      "mean_token_accuracy": 0.9717683315277099,
      "num_tokens": 35123825.0,
      "step": 1960
    },
    {
      "entropy": 1.1301296949386597,
      "epoch": 2.6514131897711977,
      "grad_norm": 1.4734028577804565,
      "learning_rate": 8.789063523948958e-05,
      "loss": 0.0702283263206482,
      "mean_token_accuracy": 0.9727118015289307,
      "num_tokens": 35302914.0,
      "step": 1970
    },
    {
      "entropy": 1.1409568905830383,
      "epoch": 2.664872139973082,
      "grad_norm": 1.4132834672927856,
      "learning_rate": 8.774504913898663e-05,
      "loss": 0.07676968574523926,
      "mean_token_accuracy": 0.9695852339267731,
      "num_tokens": 35481783.0,
      "step": 1980
    },
    {
      "entropy": 1.1532965421676635,
      "epoch": 2.678331090174966,
      "grad_norm": 1.1802046298980713,
      "learning_rate": 8.75987152208264e-05,
      "loss": 0.06539074182510377,
      "mean_token_accuracy": 0.9745090186595917,
      "num_tokens": 35660431.0,
      "step": 1990
    },
    {
      "entropy": 1.1450922250747682,
      "epoch": 2.6917900403768504,
      "grad_norm": 1.1922194957733154,
      "learning_rate": 8.745163638422583e-05,
      "loss": 0.07205181121826172,
      "mean_token_accuracy": 0.9712056815624237,
      "num_tokens": 35839308.0,
      "step": 2000
    },
    {
      "entropy": 1.1444142818450929,
      "epoch": 2.7052489905787347,
      "grad_norm": 1.1868208646774292,
      "learning_rate": 8.730381554316051e-05,
      "loss": 0.07235864400863648,
      "mean_token_accuracy": 0.9725943446159363,
      "num_tokens": 36018734.0,
      "step": 2010
    },
    {
      "entropy": 1.133234965801239,
      "epoch": 2.718707940780619,
      "grad_norm": 1.4924181699752808,
      "learning_rate": 8.715525562630687e-05,
      "loss": 0.07137352228164673,
      "mean_token_accuracy": 0.9720249474048615,
      "num_tokens": 36197607.0,
      "step": 2020
    },
    {
      "entropy": 1.150734007358551,
      "epoch": 2.732166890982503,
      "grad_norm": 1.0833524465560913,
      "learning_rate": 8.700595957698411e-05,
      "loss": 0.07287259101867676,
      "mean_token_accuracy": 0.9720607042312622,
      "num_tokens": 36377137.0,
      "step": 2030
    },
    {
      "entropy": 1.15074782371521,
      "epoch": 2.7456258411843875,
      "grad_norm": 1.1532199382781982,
      "learning_rate": 8.685593035309598e-05,
      "loss": 0.07189793586730957,
      "mean_token_accuracy": 0.971609354019165,
      "num_tokens": 36556438.0,
      "step": 2040
    },
    {
      "entropy": 1.1425267219543458,
      "epoch": 2.7590847913862717,
      "grad_norm": 1.65394926071167,
      "learning_rate": 8.670517092707213e-05,
      "loss": 0.07228031158447265,
      "mean_token_accuracy": 0.972437036037445,
      "num_tokens": 36734936.0,
      "step": 2050
    },
    {
      "entropy": 1.1443095088005066,
      "epoch": 2.772543741588156,
      "grad_norm": 1.1756309270858765,
      "learning_rate": 8.655368428580919e-05,
      "loss": 0.07032725811004639,
      "mean_token_accuracy": 0.9716470181941986,
      "num_tokens": 36913871.0,
      "step": 2060
    },
    {
      "entropy": 1.1369405388832092,
      "epoch": 2.7860026917900402,
      "grad_norm": 1.2845005989074707,
      "learning_rate": 8.640147343061165e-05,
      "loss": 0.07300193309783935,
      "mean_token_accuracy": 0.971499103307724,
      "num_tokens": 37093380.0,
      "step": 2070
    },
    {
      "entropy": 1.1441598296165467,
      "epoch": 2.7994616419919245,
      "grad_norm": 1.0858702659606934,
      "learning_rate": 8.624854137713234e-05,
      "loss": 0.07180417776107788,
      "mean_token_accuracy": 0.9721956551074982,
      "num_tokens": 37272902.0,
      "step": 2080
    },
    {
      "entropy": 1.1500964164733887,
      "epoch": 2.8129205921938087,
      "grad_norm": 1.1274408102035522,
      "learning_rate": 8.609489115531278e-05,
      "loss": 0.07155272960662842,
      "mean_token_accuracy": 0.971377295255661,
      "num_tokens": 37451897.0,
      "step": 2090
    },
    {
      "entropy": 1.1386480689048768,
      "epoch": 2.826379542395693,
      "grad_norm": 1.1894359588623047,
      "learning_rate": 8.594052580932301e-05,
      "loss": 0.06719542145729065,
      "mean_token_accuracy": 0.9733208954334259,
      "num_tokens": 37631343.0,
      "step": 2100
    },
    {
      "entropy": 1.1420456409454345,
      "epoch": 2.8398384925975773,
      "grad_norm": 1.266627550125122,
      "learning_rate": 8.578544839750141e-05,
      "loss": 0.06839650273323059,
      "mean_token_accuracy": 0.9735166966915131,
      "num_tokens": 37811111.0,
      "step": 2110
    },
    {
      "entropy": 1.136332881450653,
      "epoch": 2.8532974427994615,
      "grad_norm": 1.6329811811447144,
      "learning_rate": 8.562966199229399e-05,
      "loss": 0.0761029601097107,
      "mean_token_accuracy": 0.9703849673271179,
      "num_tokens": 37991040.0,
      "step": 2120
    },
    {
      "entropy": 1.1302747488021851,
      "epoch": 2.8667563930013458,
      "grad_norm": 1.355210304260254,
      "learning_rate": 8.547316968019363e-05,
      "loss": 0.07443415522575378,
      "mean_token_accuracy": 0.971444720029831,
      "num_tokens": 38169539.0,
      "step": 2130
    },
    {
      "entropy": 1.1407261610031127,
      "epoch": 2.88021534320323,
      "grad_norm": 1.0608882904052734,
      "learning_rate": 8.531597456167885e-05,
      "loss": 0.07463377118110656,
      "mean_token_accuracy": 0.9705821752548218,
      "num_tokens": 38348423.0,
      "step": 2140
    },
    {
      "entropy": 1.121661639213562,
      "epoch": 2.8936742934051143,
      "grad_norm": 1.4264295101165771,
      "learning_rate": 8.515807975115239e-05,
      "loss": 0.06971895098686218,
      "mean_token_accuracy": 0.9723855495452881,
      "num_tokens": 38527736.0,
      "step": 2150
    },
    {
      "entropy": 1.117723774909973,
      "epoch": 2.9071332436069985,
      "grad_norm": 1.114941120147705,
      "learning_rate": 8.499948837687959e-05,
      "loss": 0.07229661345481872,
      "mean_token_accuracy": 0.9710954666137696,
      "num_tokens": 38706893.0,
      "step": 2160
    },
    {
      "entropy": 1.12820805311203,
      "epoch": 2.920592193808883,
      "grad_norm": 1.219247817993164,
      "learning_rate": 8.484020358092625e-05,
      "loss": 0.07078794836997986,
      "mean_token_accuracy": 0.9723483324050903,
      "num_tokens": 38886237.0,
      "step": 2170
    },
    {
      "entropy": 1.1263705372810364,
      "epoch": 2.934051144010767,
      "grad_norm": 1.9336564540863037,
      "learning_rate": 8.468022851909657e-05,
      "loss": 0.07343355417251587,
      "mean_token_accuracy": 0.9712878108024597,
      "num_tokens": 39065800.0,
      "step": 2180
    },
    {
      "entropy": 1.1296117424964904,
      "epoch": 2.9475100942126513,
      "grad_norm": 1.1307294368743896,
      "learning_rate": 8.451956636087046e-05,
      "loss": 0.07211248874664307,
      "mean_token_accuracy": 0.9710810720920563,
      "num_tokens": 39245521.0,
      "step": 2190
    },
    {
      "entropy": 1.1172988891601563,
      "epoch": 2.9609690444145356,
      "grad_norm": 1.1942963600158691,
      "learning_rate": 8.435822028934087e-05,
      "loss": 0.07098879814147949,
      "mean_token_accuracy": 0.9725285410881043,
      "num_tokens": 39424709.0,
      "step": 2200
    },
    {
      "entropy": 1.1155216932296752,
      "epoch": 2.97442799461642,
      "grad_norm": 1.241036057472229,
      "learning_rate": 8.41961935011506e-05,
      "loss": 0.06951723098754883,
      "mean_token_accuracy": 0.9731462299823761,
      "num_tokens": 39603326.0,
      "step": 2210
    },
    {
      "entropy": 1.1296806812286377,
      "epoch": 2.987886944818304,
      "grad_norm": 1.249014139175415,
      "learning_rate": 8.403348920642911e-05,
      "loss": 0.07304394245147705,
      "mean_token_accuracy": 0.9720990836620331,
      "num_tokens": 39782396.0,
      "step": 2220
    },
    {
      "epoch": 3.0,
      "eval_entropy": 1.1091555235492196,
      "eval_loss": 0.10757029801607132,
      "eval_mean_token_accuracy": 0.9568785178433558,
      "eval_num_tokens": 39944284.0,
      "eval_runtime": 12.7473,
      "eval_samples_per_second": 392.239,
      "eval_steps_per_second": 12.316,
      "step": 2229
    },
    {
      "entropy": 1.1163917779922485,
      "epoch": 3.0013458950201883,
      "grad_norm": 1.0238746404647827,
      "learning_rate": 8.387011062872883e-05,
      "loss": 0.06967235207557679,
      "mean_token_accuracy": 0.9721763372421265,
      "num_tokens": 39962316.0,
      "step": 2230
    },
    {
      "entropy": 1.09123033285141,
      "epoch": 3.0148048452220726,
      "grad_norm": 1.2878422737121582,
      "learning_rate": 8.370606100496128e-05,
      "loss": 0.04885563850402832,
      "mean_token_accuracy": 0.9820802330970764,
      "num_tokens": 40141723.0,
      "step": 2240
    },
    {
      "entropy": 1.0791454672813416,
      "epoch": 3.028263795423957,
      "grad_norm": 1.1635181903839111,
      "learning_rate": 8.354134358533301e-05,
      "loss": 0.052800750732421874,
      "mean_token_accuracy": 0.9794562578201294,
      "num_tokens": 40320756.0,
      "step": 2250
    },
    {
      "entropy": 1.0924689769744873,
      "epoch": 3.041722745625841,
      "grad_norm": 1.0146963596343994,
      "learning_rate": 8.337596163328114e-05,
      "loss": 0.051714420318603516,
      "mean_token_accuracy": 0.9811573922634125,
      "num_tokens": 40499632.0,
      "step": 2260
    },
    {
      "entropy": 1.086620819568634,
      "epoch": 3.0551816958277254,
      "grad_norm": 1.1477982997894287,
      "learning_rate": 8.320991842540875e-05,
      "loss": 0.046452611684799194,
      "mean_token_accuracy": 0.9823605060577393,
      "num_tokens": 40679088.0,
      "step": 2270
    },
    {
      "entropy": 1.0724918246269226,
      "epoch": 3.0686406460296096,
      "grad_norm": 1.5574088096618652,
      "learning_rate": 8.304321725141995e-05,
      "loss": 0.05086854100227356,
      "mean_token_accuracy": 0.981047946214676,
      "num_tokens": 40859040.0,
      "step": 2280
    },
    {
      "entropy": 1.0786724448204041,
      "epoch": 3.082099596231494,
      "grad_norm": 1.0978903770446777,
      "learning_rate": 8.287586141405464e-05,
      "loss": 0.053134101629257205,
      "mean_token_accuracy": 0.9794782817363739,
      "num_tokens": 41038118.0,
      "step": 2290
    },
    {
      "entropy": 1.0825668334960938,
      "epoch": 3.095558546433378,
      "grad_norm": 1.1968939304351807,
      "learning_rate": 8.27078542290232e-05,
      "loss": 0.04702814817428589,
      "mean_token_accuracy": 0.9820038139820099,
      "num_tokens": 41217196.0,
      "step": 2300
    },
    {
      "entropy": 1.0657127141952514,
      "epoch": 3.1090174966352624,
      "grad_norm": 1.544787049293518,
      "learning_rate": 8.253919902494071e-05,
      "loss": 0.051616507768630984,
      "mean_token_accuracy": 0.9803533375263214,
      "num_tokens": 41396525.0,
      "step": 2310
    },
    {
      "entropy": 1.0903544425964355,
      "epoch": 3.1224764468371466,
      "grad_norm": 1.2193773984909058,
      "learning_rate": 8.236989914326101e-05,
      "loss": 0.049640601873397826,
      "mean_token_accuracy": 0.9816096484661102,
      "num_tokens": 41575854.0,
      "step": 2320
    },
    {
      "entropy": 1.0869248747825622,
      "epoch": 3.135935397039031,
      "grad_norm": 1.1424866914749146,
      "learning_rate": 8.21999579382105e-05,
      "loss": 0.049025171995162965,
      "mean_token_accuracy": 0.9807829558849335,
      "num_tokens": 41755176.0,
      "step": 2330
    },
    {
      "entropy": 1.0788421630859375,
      "epoch": 3.149394347240915,
      "grad_norm": 1.1553189754486084,
      "learning_rate": 8.202937877672175e-05,
      "loss": 0.05088653564453125,
      "mean_token_accuracy": 0.9806456983089447,
      "num_tokens": 41935004.0,
      "step": 2340
    },
    {
      "entropy": 1.0868588209152221,
      "epoch": 3.1628532974427994,
      "grad_norm": 1.2867755889892578,
      "learning_rate": 8.185816503836665e-05,
      "loss": 0.04932470321655273,
      "mean_token_accuracy": 0.9814176499843598,
      "num_tokens": 42114422.0,
      "step": 2350
    },
    {
      "entropy": 1.077353870868683,
      "epoch": 3.1763122476446837,
      "grad_norm": 1.0823496580123901,
      "learning_rate": 8.168632011528961e-05,
      "loss": 0.05128706097602844,
      "mean_token_accuracy": 0.9801781177520752,
      "num_tokens": 42293971.0,
      "step": 2360
    },
    {
      "entropy": 1.066714358329773,
      "epoch": 3.189771197846568,
      "grad_norm": 1.2061221599578857,
      "learning_rate": 8.15138474121403e-05,
      "loss": 0.04973976612091065,
      "mean_token_accuracy": 0.9804952144622803,
      "num_tokens": 42473552.0,
      "step": 2370
    },
    {
      "entropy": 1.0797779440879822,
      "epoch": 3.203230148048452,
      "grad_norm": 1.111051321029663,
      "learning_rate": 8.134075034600609e-05,
      "loss": 0.05457779169082642,
      "mean_token_accuracy": 0.979478371143341,
      "num_tokens": 42652094.0,
      "step": 2380
    },
    {
      "entropy": 1.0842301845550537,
      "epoch": 3.2166890982503364,
      "grad_norm": 1.2103989124298096,
      "learning_rate": 8.116703234634453e-05,
      "loss": 0.05012301206588745,
      "mean_token_accuracy": 0.9813940584659576,
      "num_tokens": 42831176.0,
      "step": 2390
    },
    {
      "entropy": 1.072116458415985,
      "epoch": 3.2301480484522207,
      "grad_norm": 1.1604132652282715,
      "learning_rate": 8.099269685491528e-05,
      "loss": 0.0516257107257843,
      "mean_token_accuracy": 0.9805888772010803,
      "num_tokens": 43010038.0,
      "step": 2400
    },
    {
      "entropy": 1.0759256601333618,
      "epoch": 3.243606998654105,
      "grad_norm": 1.055504560470581,
      "learning_rate": 8.081774732571196e-05,
      "loss": 0.05206198692321777,
      "mean_token_accuracy": 0.9800681293010711,
      "num_tokens": 43189534.0,
      "step": 2410
    },
    {
      "entropy": 1.074904704093933,
      "epoch": 3.257065948855989,
      "grad_norm": 1.3270564079284668,
      "learning_rate": 8.06421872248937e-05,
      "loss": 0.053070676326751706,
      "mean_token_accuracy": 0.9799122869968414,
      "num_tokens": 43369326.0,
      "step": 2420
    },
    {
      "entropy": 1.0772022962570191,
      "epoch": 3.2705248990578735,
      "grad_norm": 1.1767622232437134,
      "learning_rate": 8.046602003071648e-05,
      "loss": 0.05088210105895996,
      "mean_token_accuracy": 0.980486124753952,
      "num_tokens": 43547755.0,
      "step": 2430
    },
    {
      "entropy": 1.0786782264709474,
      "epoch": 3.2839838492597577,
      "grad_norm": 1.320910930633545,
      "learning_rate": 8.028924923346426e-05,
      "loss": 0.05664767622947693,
      "mean_token_accuracy": 0.9779302835464477,
      "num_tokens": 43726776.0,
      "step": 2440
    },
    {
      "entropy": 1.0860824465751648,
      "epoch": 3.297442799461642,
      "grad_norm": 1.2538353204727173,
      "learning_rate": 8.011187833537972e-05,
      "loss": 0.053240764141082766,
      "mean_token_accuracy": 0.9801206707954406,
      "num_tokens": 43905546.0,
      "step": 2450
    },
    {
      "entropy": 1.081036913394928,
      "epoch": 3.3109017496635262,
      "grad_norm": 1.1865931749343872,
      "learning_rate": 7.993391085059502e-05,
      "loss": 0.05133863687515259,
      "mean_token_accuracy": 0.9802879393100739,
      "num_tokens": 44084371.0,
      "step": 2460
    },
    {
      "entropy": 1.0795913338661194,
      "epoch": 3.3243606998654105,
      "grad_norm": 1.3042405843734741,
      "learning_rate": 7.975535030506203e-05,
      "loss": 0.053297781944274904,
      "mean_token_accuracy": 0.9790417075157165,
      "num_tokens": 44263366.0,
      "step": 2470
    },
    {
      "entropy": 1.0945096254348754,
      "epoch": 3.3378196500672948,
      "grad_norm": 1.224970817565918,
      "learning_rate": 7.957620023648256e-05,
      "loss": 0.05278623104095459,
      "mean_token_accuracy": 0.9799198031425476,
      "num_tokens": 44442200.0,
      "step": 2480
    },
    {
      "entropy": 1.0911083102226258,
      "epoch": 3.351278600269179,
      "grad_norm": 1.1166850328445435,
      "learning_rate": 7.939646419423826e-05,
      "loss": 0.055463647842407225,
      "mean_token_accuracy": 0.9793131470680236,
      "num_tokens": 44621518.0,
      "step": 2490
    },
    {
      "entropy": 1.0902422070503235,
      "epoch": 3.3647375504710633,
      "grad_norm": 1.3478926420211792,
      "learning_rate": 7.92161457393203e-05,
      "loss": 0.05521976947784424,
      "mean_token_accuracy": 0.9795126497745514,
      "num_tokens": 44800926.0,
      "step": 2500
    },
    {
      "entropy": 1.0822187185287475,
      "epoch": 3.3781965006729475,
      "grad_norm": 1.3261746168136597,
      "learning_rate": 7.903524844425878e-05,
      "loss": 0.0533198356628418,
      "mean_token_accuracy": 0.9789367377758026,
      "num_tokens": 44980265.0,
      "step": 2510
    },
    {
      "entropy": 1.087248969078064,
      "epoch": 3.391655450874832,
      "grad_norm": 1.005294919013977,
      "learning_rate": 7.885377589305197e-05,
      "loss": 0.05112231373786926,
      "mean_token_accuracy": 0.980577951669693,
      "num_tokens": 45159183.0,
      "step": 2520
    },
    {
      "entropy": 1.0852233290672302,
      "epoch": 3.405114401076716,
      "grad_norm": 1.2185736894607544,
      "learning_rate": 7.867173168109534e-05,
      "loss": 0.05045679807662964,
      "mean_token_accuracy": 0.9805097222328186,
      "num_tokens": 45338667.0,
      "step": 2530
    },
    {
      "entropy": 1.0843430399894713,
      "epoch": 3.4185733512786003,
      "grad_norm": 1.183929681777954,
      "learning_rate": 7.84891194151103e-05,
      "loss": 0.050162500143051146,
      "mean_token_accuracy": 0.9804108917713166,
      "num_tokens": 45517477.0,
      "step": 2540
    },
    {
      "entropy": 1.0773622393608093,
      "epoch": 3.4320323014804845,
      "grad_norm": 1.0225229263305664,
      "learning_rate": 7.830594271307267e-05,
      "loss": 0.05130444765090943,
      "mean_token_accuracy": 0.9804662525653839,
      "num_tokens": 45696935.0,
      "step": 2550
    },
    {
      "entropy": 1.0753762125968933,
      "epoch": 3.445491251682369,
      "grad_norm": 1.1646474599838257,
      "learning_rate": 7.812220520414115e-05,
      "loss": 0.05273773670196533,
      "mean_token_accuracy": 0.9797065794467926,
      "num_tokens": 45875869.0,
      "step": 2560
    },
    {
      "entropy": 1.0715667009353638,
      "epoch": 3.458950201884253,
      "grad_norm": 1.1959201097488403,
      "learning_rate": 7.793791052858528e-05,
      "loss": 0.053073453903198245,
      "mean_token_accuracy": 0.9801750302314758,
      "num_tokens": 46055697.0,
      "step": 2570
    },
    {
      "entropy": 1.07788827419281,
      "epoch": 3.4724091520861373,
      "grad_norm": 1.3329288959503174,
      "learning_rate": 7.775306233771343e-05,
      "loss": 0.053553718328475955,
      "mean_token_accuracy": 0.9794677913188934,
      "num_tokens": 46234543.0,
      "step": 2580
    },
    {
      "entropy": 1.0729936122894288,
      "epoch": 3.4858681022880216,
      "grad_norm": 1.261353850364685,
      "learning_rate": 7.756766429380033e-05,
      "loss": 0.0518909215927124,
      "mean_token_accuracy": 0.9810412526130676,
      "num_tokens": 46413969.0,
      "step": 2590
    },
    {
      "entropy": 1.0738806128501892,
      "epoch": 3.499327052489906,
      "grad_norm": 1.5493957996368408,
      "learning_rate": 7.738172007001465e-05,
      "loss": 0.05155509114265442,
      "mean_token_accuracy": 0.9803483247756958,
      "num_tokens": 46593798.0,
      "step": 2600
    },
    {
      "entropy": 1.0901301860809327,
      "epoch": 3.51278600269179,
      "grad_norm": 1.0253880023956299,
      "learning_rate": 7.719523335034612e-05,
      "loss": 0.05098943710327149,
      "mean_token_accuracy": 0.9800853669643402,
      "num_tokens": 46772326.0,
      "step": 2610
    },
    {
      "entropy": 1.0800783038139343,
      "epoch": 3.5262449528936743,
      "grad_norm": 1.4688411951065063,
      "learning_rate": 7.70082078295326e-05,
      "loss": 0.05392987728118896,
      "mean_token_accuracy": 0.978635448217392,
      "num_tokens": 46950959.0,
      "step": 2620
    },
    {
      "entropy": 1.090990924835205,
      "epoch": 3.5397039030955586,
      "grad_norm": 1.3596893548965454,
      "learning_rate": 7.682064721298683e-05,
      "loss": 0.053889667987823485,
      "mean_token_accuracy": 0.9795092165470123,
      "num_tokens": 47129717.0,
      "step": 2630
    },
    {
      "entropy": 1.1031178951263427,
      "epoch": 3.553162853297443,
      "grad_norm": 1.2542155981063843,
      "learning_rate": 7.663255521672308e-05,
      "loss": 0.05507693886756897,
      "mean_token_accuracy": 0.9788171648979187,
      "num_tokens": 47308851.0,
      "step": 2640
    },
    {
      "entropy": 1.0928621292114258,
      "epoch": 3.566621803499327,
      "grad_norm": 1.328484058380127,
      "learning_rate": 7.64439355672835e-05,
      "loss": 0.05089789628982544,
      "mean_token_accuracy": 0.9808636486530304,
      "num_tokens": 47487370.0,
      "step": 2650
    },
    {
      "entropy": 1.083961284160614,
      "epoch": 3.5800807537012114,
      "grad_norm": 1.1594445705413818,
      "learning_rate": 7.625479200166425e-05,
      "loss": 0.0499467521905899,
      "mean_token_accuracy": 0.9808259546756745,
      "num_tokens": 47667176.0,
      "step": 2660
    },
    {
      "entropy": 1.0920774698257447,
      "epoch": 3.5935397039030956,
      "grad_norm": 1.3259137868881226,
      "learning_rate": 7.606512826724155e-05,
      "loss": 0.051684331893920896,
      "mean_token_accuracy": 0.9798979878425598,
      "num_tokens": 47845643.0,
      "step": 2670
    },
    {
      "entropy": 1.0919681429862975,
      "epoch": 3.60699865410498,
      "grad_norm": 1.3699877262115479,
      "learning_rate": 7.587494812169728e-05,
      "loss": 0.051729452610015866,
      "mean_token_accuracy": 0.980066442489624,
      "num_tokens": 48024631.0,
      "step": 2680
    },
    {
      "entropy": 1.084936547279358,
      "epoch": 3.620457604306864,
      "grad_norm": 1.2969093322753906,
      "learning_rate": 7.568425533294476e-05,
      "loss": 0.054199641942977904,
      "mean_token_accuracy": 0.9787844777107239,
      "num_tokens": 48204028.0,
      "step": 2690
    },
    {
      "entropy": 1.0940564393997192,
      "epoch": 3.6339165545087484,
      "grad_norm": 1.0239887237548828,
      "learning_rate": 7.549305367905385e-05,
      "loss": 0.054245364665985105,
      "mean_token_accuracy": 0.9782158195972442,
      "num_tokens": 48382282.0,
      "step": 2700
    },
    {
      "entropy": 1.0823640823364258,
      "epoch": 3.6473755047106327,
      "grad_norm": 1.1898547410964966,
      "learning_rate": 7.53013469481763e-05,
      "loss": 0.056242328882217404,
      "mean_token_accuracy": 0.9788333296775817,
      "num_tokens": 48561545.0,
      "step": 2710
    },
    {
      "entropy": 1.0766677498817443,
      "epoch": 3.660834454912517,
      "grad_norm": 1.2227085828781128,
      "learning_rate": 7.510913893847058e-05,
      "loss": 0.0562747597694397,
      "mean_token_accuracy": 0.9780334293842315,
      "num_tokens": 48741185.0,
      "step": 2720
    },
    {
      "entropy": 1.091788113117218,
      "epoch": 3.674293405114401,
      "grad_norm": 1.517991304397583,
      "learning_rate": 7.491643345802667e-05,
      "loss": 0.05473091006278992,
      "mean_token_accuracy": 0.9788542568683625,
      "num_tokens": 48920353.0,
      "step": 2730
    },
    {
      "entropy": 1.1030115008354187,
      "epoch": 3.6877523553162854,
      "grad_norm": 1.3098318576812744,
      "learning_rate": 7.472323432479062e-05,
      "loss": 0.058284854888916014,
      "mean_token_accuracy": 0.9783021330833435,
      "num_tokens": 49099763.0,
      "step": 2740
    },
    {
      "entropy": 1.0996167540550232,
      "epoch": 3.7012113055181697,
      "grad_norm": 1.1821726560592651,
      "learning_rate": 7.452954536648888e-05,
      "loss": 0.0540702223777771,
      "mean_token_accuracy": 0.979110324382782,
      "num_tokens": 49278776.0,
      "step": 2750
    },
    {
      "entropy": 1.0863705158233643,
      "epoch": 3.714670255720054,
      "grad_norm": 1.2044000625610352,
      "learning_rate": 7.433537042055248e-05,
      "loss": 0.05256187915802002,
      "mean_token_accuracy": 0.9793642222881317,
      "num_tokens": 49457667.0,
      "step": 2760
    },
    {
      "entropy": 1.0813929796218873,
      "epoch": 3.728129205921938,
      "grad_norm": 0.9910821318626404,
      "learning_rate": 7.414071333404104e-05,
      "loss": 0.053735208511352536,
      "mean_token_accuracy": 0.9796746790409088,
      "num_tokens": 49637000.0,
      "step": 2770
    },
    {
      "entropy": 1.0826418280601502,
      "epoch": 3.7415881561238225,
      "grad_norm": 1.16364586353302,
      "learning_rate": 7.394557796356644e-05,
      "loss": 0.05005708336830139,
      "mean_token_accuracy": 0.9802528321743011,
      "num_tokens": 49815945.0,
      "step": 2780
    },
    {
      "entropy": 1.0849217176437378,
      "epoch": 3.7550471063257067,
      "grad_norm": 1.1651335954666138,
      "learning_rate": 7.374996817521653e-05,
      "loss": 0.05107729434967041,
      "mean_token_accuracy": 0.9808483123779297,
      "num_tokens": 49995002.0,
      "step": 2790
    },
    {
      "entropy": 1.0982914447784424,
      "epoch": 3.768506056527591,
      "grad_norm": 1.300089716911316,
      "learning_rate": 7.35538878444785e-05,
      "loss": 0.05192199945449829,
      "mean_token_accuracy": 0.9799326181411743,
      "num_tokens": 50174611.0,
      "step": 2800
    },
    {
      "entropy": 1.0978451728820802,
      "epoch": 3.781965006729475,
      "grad_norm": 1.131295919418335,
      "learning_rate": 7.335734085616206e-05,
      "loss": 0.05380460023880005,
      "mean_token_accuracy": 0.9787914991378784,
      "num_tokens": 50354572.0,
      "step": 2810
    },
    {
      "entropy": 1.1055003643035888,
      "epoch": 3.7954239569313595,
      "grad_norm": 1.1545559167861938,
      "learning_rate": 7.316033110432249e-05,
      "loss": 0.05247194766998291,
      "mean_token_accuracy": 0.9800500810146332,
      "num_tokens": 50533810.0,
      "step": 2820
    },
    {
      "entropy": 1.1019269943237304,
      "epoch": 3.8088829071332437,
      "grad_norm": 1.439558744430542,
      "learning_rate": 7.296286249218352e-05,
      "loss": 0.05247552394866943,
      "mean_token_accuracy": 0.9799910664558411,
      "num_tokens": 50712921.0,
      "step": 2830
    },
    {
      "entropy": 1.1034127831459046,
      "epoch": 3.822341857335128,
      "grad_norm": 1.1686484813690186,
      "learning_rate": 7.276493893205995e-05,
      "loss": 0.053317368030548096,
      "mean_token_accuracy": 0.9795270919799804,
      "num_tokens": 50892646.0,
      "step": 2840
    },
    {
      "entropy": 1.0969440698623658,
      "epoch": 3.8358008075370122,
      "grad_norm": 1.1592854261398315,
      "learning_rate": 7.256656434528018e-05,
      "loss": 0.05192090272903442,
      "mean_token_accuracy": 0.9800365567207336,
      "num_tokens": 51072211.0,
      "step": 2850
    },
    {
      "entropy": 1.090685820579529,
      "epoch": 3.8492597577388965,
      "grad_norm": 1.5477663278579712,
      "learning_rate": 7.236774266210852e-05,
      "loss": 0.051884579658508304,
      "mean_token_accuracy": 0.9792610287666321,
      "num_tokens": 51251374.0,
      "step": 2860
    },
    {
      "entropy": 1.1031171798706054,
      "epoch": 3.8627187079407808,
      "grad_norm": 0.9951223134994507,
      "learning_rate": 7.216847782166727e-05,
      "loss": 0.051608985662460326,
      "mean_token_accuracy": 0.9799349427223205,
      "num_tokens": 51429845.0,
      "step": 2870
    },
    {
      "entropy": 1.0826797842979432,
      "epoch": 3.876177658142665,
      "grad_norm": 1.0834344625473022,
      "learning_rate": 7.196877377185872e-05,
      "loss": 0.05112177133560181,
      "mean_token_accuracy": 0.980464369058609,
      "num_tokens": 51609039.0,
      "step": 2880
    },
    {
      "entropy": 1.0658052206039428,
      "epoch": 3.8896366083445493,
      "grad_norm": 1.361104130744934,
      "learning_rate": 7.176863446928694e-05,
      "loss": 0.05474343299865723,
      "mean_token_accuracy": 0.9786961674690247,
      "num_tokens": 51787900.0,
      "step": 2890
    },
    {
      "entropy": 1.0786802649497986,
      "epoch": 3.9030955585464335,
      "grad_norm": 1.3113523721694946,
      "learning_rate": 7.156806387917937e-05,
      "loss": 0.05351792573928833,
      "mean_token_accuracy": 0.9795374929904938,
      "num_tokens": 51967068.0,
      "step": 2900
    },
    {
      "entropy": 1.0683523178100587,
      "epoch": 3.916554508748318,
      "grad_norm": 1.172418236732483,
      "learning_rate": 7.136706597530825e-05,
      "loss": 0.05207247734069824,
      "mean_token_accuracy": 0.9796711683273316,
      "num_tokens": 52146442.0,
      "step": 2910
    },
    {
      "entropy": 1.0647220134735107,
      "epoch": 3.930013458950202,
      "grad_norm": 1.109689712524414,
      "learning_rate": 7.116564473991192e-05,
      "loss": 0.050079309940338136,
      "mean_token_accuracy": 0.9804611086845398,
      "num_tokens": 52325461.0,
      "step": 2920
    },
    {
      "entropy": 1.0695144891738892,
      "epoch": 3.9434724091520863,
      "grad_norm": 1.0333713293075562,
      "learning_rate": 7.096380416361588e-05,
      "loss": 0.05163470506668091,
      "mean_token_accuracy": 0.9802155733108521,
      "num_tokens": 52505157.0,
      "step": 2930
    },
    {
      "entropy": 1.0664403676986693,
      "epoch": 3.9569313593539706,
      "grad_norm": 1.1881392002105713,
      "learning_rate": 7.076154824535381e-05,
      "loss": 0.05164710283279419,
      "mean_token_accuracy": 0.9804204642772675,
      "num_tokens": 52684876.0,
      "step": 2940
    },
    {
      "entropy": 1.061314308643341,
      "epoch": 3.970390309555855,
      "grad_norm": 1.1483125686645508,
      "learning_rate": 7.055888099228825e-05,
      "loss": 0.051780533790588376,
      "mean_token_accuracy": 0.9797692000865936,
      "num_tokens": 52864664.0,
      "step": 2950
    },
    {
      "entropy": 1.0693163990974426,
      "epoch": 3.983849259757739,
      "grad_norm": 1.319306492805481,
      "learning_rate": 7.035580641973119e-05,
      "loss": 0.054670310020446776,
      "mean_token_accuracy": 0.9792993664741516,
      "num_tokens": 53043689.0,
      "step": 2960
    },
    {
      "entropy": 1.071792209148407,
      "epoch": 3.9973082099596233,
      "grad_norm": 1.1120209693908691,
      "learning_rate": 7.015232855106468e-05,
      "loss": 0.05089702606201172,
      "mean_token_accuracy": 0.9801424205303192,
      "num_tokens": 53222975.0,
      "step": 2970
    },
    {
      "epoch": 4.0,
      "eval_entropy": 1.0652484506558462,
      "eval_loss": 0.11533018946647644,
      "eval_mean_token_accuracy": 0.9572236617659308,
      "eval_num_tokens": 53259126.0,
      "eval_runtime": 12.7495,
      "eval_samples_per_second": 392.172,
      "eval_steps_per_second": 12.314,
      "step": 2972
    },
    {
      "entropy": 1.0567853927612305,
      "epoch": 4.010767160161508,
      "grad_norm": 1.0667976140975952,
      "learning_rate": 6.994845141766093e-05,
      "loss": 0.037795445322990416,
      "mean_token_accuracy": 0.9864524364471435,
      "num_tokens": 53401849.0,
      "step": 2980
    },
    {
      "entropy": 1.031903612613678,
      "epoch": 4.024226110363392,
      "grad_norm": 1.3452744483947754,
      "learning_rate": 6.974417905880255e-05,
      "loss": 0.03760968446731568,
      "mean_token_accuracy": 0.9865435302257538,
      "num_tokens": 53581109.0,
      "step": 2990
    },
    {
      "entropy": 1.0267599463462829,
      "epoch": 4.037685060565276,
      "grad_norm": 0.9383119940757751,
      "learning_rate": 6.953951552160248e-05,
      "loss": 0.034871619939804074,
      "mean_token_accuracy": 0.9870846211910248,
      "num_tokens": 53760927.0,
      "step": 3000
    },
    {
      "entropy": 1.0307875275611877,
      "epoch": 4.05114401076716,
      "grad_norm": 1.077243447303772,
      "learning_rate": 6.933446486092381e-05,
      "loss": 0.03702903389930725,
      "mean_token_accuracy": 0.9866882026195526,
      "num_tokens": 53939706.0,
      "step": 3010
    },
    {
      "entropy": 1.0310243368148804,
      "epoch": 4.064602960969045,
      "grad_norm": 1.3179757595062256,
      "learning_rate": 6.912903113929947e-05,
      "loss": 0.03828337490558624,
      "mean_token_accuracy": 0.9860410630702973,
      "num_tokens": 54118885.0,
      "step": 3020
    },
    {
      "entropy": 1.0358707904815674,
      "epoch": 4.078061911170929,
      "grad_norm": 0.949308454990387,
      "learning_rate": 6.892321842685171e-05,
      "loss": 0.036287522315979,
      "mean_token_accuracy": 0.9870272636413574,
      "num_tokens": 54298699.0,
      "step": 3030
    },
    {
      "entropy": 1.0304685473442077,
      "epoch": 4.091520861372813,
      "grad_norm": 1.271782636642456,
      "learning_rate": 6.871703080121148e-05,
      "loss": 0.036009562015533444,
      "mean_token_accuracy": 0.9868483424186707,
      "num_tokens": 54477713.0,
      "step": 3040
    },
    {
      "entropy": 1.0414454936981201,
      "epoch": 4.104979811574697,
      "grad_norm": 1.2563632726669312,
      "learning_rate": 6.851047234743763e-05,
      "loss": 0.03808114230632782,
      "mean_token_accuracy": 0.9857545673847199,
      "num_tokens": 54656606.0,
      "step": 3050
    },
    {
      "entropy": 1.0467128515243531,
      "epoch": 4.118438761776582,
      "grad_norm": 1.3075348138809204,
      "learning_rate": 6.830354715793598e-05,
      "loss": 0.03709319531917572,
      "mean_token_accuracy": 0.9860689222812653,
      "num_tokens": 54835569.0,
      "step": 3060
    },
    {
      "entropy": 1.0483134746551515,
      "epoch": 4.131897711978466,
      "grad_norm": 1.325262188911438,
      "learning_rate": 6.809625933237826e-05,
      "loss": 0.03990113139152527,
      "mean_token_accuracy": 0.9844289362430573,
      "num_tokens": 55015181.0,
      "step": 3070
    },
    {
      "entropy": 1.0507565259933471,
      "epoch": 4.14535666218035,
      "grad_norm": 0.9971099495887756,
      "learning_rate": 6.788861297762086e-05,
      "loss": 0.03460783362388611,
      "mean_token_accuracy": 0.9869785606861115,
      "num_tokens": 55194211.0,
      "step": 3080
    },
    {
      "entropy": 1.0375954627990722,
      "epoch": 4.158815612382234,
      "grad_norm": 1.1459321975708008,
      "learning_rate": 6.768061220762345e-05,
      "loss": 0.039771124720573425,
      "mean_token_accuracy": 0.984779155254364,
      "num_tokens": 55373415.0,
      "step": 3090
    },
    {
      "entropy": 1.047706139087677,
      "epoch": 4.172274562584119,
      "grad_norm": 0.9713513255119324,
      "learning_rate": 6.747226114336753e-05,
      "loss": 0.036009186506271364,
      "mean_token_accuracy": 0.9871737420558929,
      "num_tokens": 55552640.0,
      "step": 3100
    },
    {
      "entropy": 1.0445024371147156,
      "epoch": 4.185733512786003,
      "grad_norm": 1.0984820127487183,
      "learning_rate": 6.726356391277471e-05,
      "loss": 0.03706789314746857,
      "mean_token_accuracy": 0.98599613904953,
      "num_tokens": 55731084.0,
      "step": 3110
    },
    {
      "entropy": 1.0476725816726684,
      "epoch": 4.199192462987887,
      "grad_norm": 1.326766014099121,
      "learning_rate": 6.7054524650625e-05,
      "loss": 0.037712886929512024,
      "mean_token_accuracy": 0.9861681163311005,
      "num_tokens": 55910040.0,
      "step": 3120
    },
    {
      "entropy": 1.0618762969970703,
      "epoch": 4.212651413189771,
      "grad_norm": 1.0765913724899292,
      "learning_rate": 6.684514749847482e-05,
      "loss": 0.04045211672782898,
      "mean_token_accuracy": 0.985183447599411,
      "num_tokens": 56089401.0,
      "step": 3130
    },
    {
      "entropy": 1.0641907811164857,
      "epoch": 4.226110363391656,
      "grad_norm": 1.1481443643569946,
      "learning_rate": 6.663543660457503e-05,
      "loss": 0.03950580060482025,
      "mean_token_accuracy": 0.9853214621543884,
      "num_tokens": 56268108.0,
      "step": 3140
    },
    {
      "entropy": 1.0535814762115479,
      "epoch": 4.23956931359354,
      "grad_norm": 1.169559359550476,
      "learning_rate": 6.642539612378863e-05,
      "loss": 0.03928137719631195,
      "mean_token_accuracy": 0.9853263258934021,
      "num_tokens": 56447473.0,
      "step": 3150
    },
    {
      "entropy": 1.0659124374389648,
      "epoch": 4.253028263795424,
      "grad_norm": 1.4614970684051514,
      "learning_rate": 6.621503021750858e-05,
      "loss": 0.03726804554462433,
      "mean_token_accuracy": 0.9859472155570984,
      "num_tokens": 56626084.0,
      "step": 3160
    },
    {
      "entropy": 1.0618607163429261,
      "epoch": 4.2664872139973085,
      "grad_norm": 1.1655552387237549,
      "learning_rate": 6.600434305357521e-05,
      "loss": 0.0387746661901474,
      "mean_token_accuracy": 0.9854085683822632,
      "num_tokens": 56805198.0,
      "step": 3170
    },
    {
      "entropy": 1.0502426981925965,
      "epoch": 4.279946164199193,
      "grad_norm": 0.985268235206604,
      "learning_rate": 6.579333880619376e-05,
      "loss": 0.035914698243141176,
      "mean_token_accuracy": 0.9867818415164947,
      "num_tokens": 56984246.0,
      "step": 3180
    },
    {
      "entropy": 1.0420450568199158,
      "epoch": 4.293405114401077,
      "grad_norm": 1.1994178295135498,
      "learning_rate": 6.558202165585161e-05,
      "loss": 0.034868115186691286,
      "mean_token_accuracy": 0.9868482947349548,
      "num_tokens": 57163039.0,
      "step": 3190
    },
    {
      "entropy": 1.0321968078613282,
      "epoch": 4.306864064602961,
      "grad_norm": 1.160141110420227,
      "learning_rate": 6.53703957892355e-05,
      "loss": 0.03723163604736328,
      "mean_token_accuracy": 0.9852693140506744,
      "num_tokens": 57342230.0,
      "step": 3200
    },
    {
      "entropy": 1.0504021406173707,
      "epoch": 4.3203230148048455,
      "grad_norm": 1.068778395652771,
      "learning_rate": 6.515846539914854e-05,
      "loss": 0.03703950047492981,
      "mean_token_accuracy": 0.9864457190036774,
      "num_tokens": 57521452.0,
      "step": 3210
    },
    {
      "entropy": 1.0472504138946532,
      "epoch": 4.33378196500673,
      "grad_norm": 1.424491047859192,
      "learning_rate": 6.494623468442718e-05,
      "loss": 0.037775719165802,
      "mean_token_accuracy": 0.9865365862846375,
      "num_tokens": 57700961.0,
      "step": 3220
    },
    {
      "entropy": 1.0458205103874207,
      "epoch": 4.347240915208614,
      "grad_norm": 1.4818817377090454,
      "learning_rate": 6.473370784985798e-05,
      "loss": 0.03907344341278076,
      "mean_token_accuracy": 0.9854373514652253,
      "num_tokens": 57880226.0,
      "step": 3230
    },
    {
      "entropy": 1.0566739559173584,
      "epoch": 4.360699865410498,
      "grad_norm": 1.147052526473999,
      "learning_rate": 6.452088910609436e-05,
      "loss": 0.036917302012443545,
      "mean_token_accuracy": 0.9858834028244019,
      "num_tokens": 58059019.0,
      "step": 3240
    },
    {
      "entropy": 1.053157901763916,
      "epoch": 4.3741588156123825,
      "grad_norm": 1.1795092821121216,
      "learning_rate": 6.430778266957312e-05,
      "loss": 0.036924809217453,
      "mean_token_accuracy": 0.9859816431999207,
      "num_tokens": 58237916.0,
      "step": 3250
    },
    {
      "entropy": 1.0390576720237732,
      "epoch": 4.387617765814267,
      "grad_norm": 1.2286282777786255,
      "learning_rate": 6.409439276243092e-05,
      "loss": 0.03474595844745636,
      "mean_token_accuracy": 0.9870323657989502,
      "num_tokens": 58416659.0,
      "step": 3260
    },
    {
      "entropy": 1.0422054052352905,
      "epoch": 4.401076716016151,
      "grad_norm": 1.1984254121780396,
      "learning_rate": 6.388072361242067e-05,
      "loss": 0.03632686138153076,
      "mean_token_accuracy": 0.9864870607852936,
      "num_tokens": 58595581.0,
      "step": 3270
    },
    {
      "entropy": 1.0455679059028626,
      "epoch": 4.414535666218035,
      "grad_norm": 1.201826810836792,
      "learning_rate": 6.366677945282769e-05,
      "loss": 0.03645781874656677,
      "mean_token_accuracy": 0.9858992576599122,
      "num_tokens": 58774496.0,
      "step": 3280
    },
    {
      "entropy": 1.0460187315940856,
      "epoch": 4.4279946164199195,
      "grad_norm": 1.320504069328308,
      "learning_rate": 6.345256452238591e-05,
      "loss": 0.036577820777893066,
      "mean_token_accuracy": 0.9865557551383972,
      "num_tokens": 58953827.0,
      "step": 3290
    },
    {
      "entropy": 1.043444275856018,
      "epoch": 4.441453566621804,
      "grad_norm": 1.1478440761566162,
      "learning_rate": 6.323808306519385e-05,
      "loss": 0.03831025958061218,
      "mean_token_accuracy": 0.9856758773326874,
      "num_tokens": 59133259.0,
      "step": 3300
    },
    {
      "entropy": 1.0497878909111023,
      "epoch": 4.454912516823688,
      "grad_norm": 1.0950919389724731,
      "learning_rate": 6.302333933063057e-05,
      "loss": 0.0407365620136261,
      "mean_token_accuracy": 0.984984940290451,
      "num_tokens": 59312911.0,
      "step": 3310
    },
    {
      "entropy": 1.058940577507019,
      "epoch": 4.468371467025572,
      "grad_norm": 1.059362769126892,
      "learning_rate": 6.280833757327142e-05,
      "loss": 0.03741527199745178,
      "mean_token_accuracy": 0.9860852122306824,
      "num_tokens": 59492704.0,
      "step": 3320
    },
    {
      "entropy": 1.0534117579460145,
      "epoch": 4.481830417227457,
      "grad_norm": 1.3785133361816406,
      "learning_rate": 6.259308205280383e-05,
      "loss": 0.03677107989788055,
      "mean_token_accuracy": 0.9864885747432709,
      "num_tokens": 59672041.0,
      "step": 3330
    },
    {
      "entropy": 1.0532031655311584,
      "epoch": 4.495289367429341,
      "grad_norm": 1.4289969205856323,
      "learning_rate": 6.237757703394283e-05,
      "loss": 0.040090084075927734,
      "mean_token_accuracy": 0.9849297285079956,
      "num_tokens": 59851174.0,
      "step": 3340
    },
    {
      "entropy": 1.0606793522834779,
      "epoch": 4.508748317631225,
      "grad_norm": 0.9429711699485779,
      "learning_rate": 6.216182678634664e-05,
      "loss": 0.03906591236591339,
      "mean_token_accuracy": 0.9849115431308746,
      "num_tokens": 60030790.0,
      "step": 3350
    },
    {
      "entropy": 1.0648606181144715,
      "epoch": 4.522207267833109,
      "grad_norm": 1.0886033773422241,
      "learning_rate": 6.194583558453199e-05,
      "loss": 0.038545870780944826,
      "mean_token_accuracy": 0.9858034968376159,
      "num_tokens": 60209459.0,
      "step": 3360
    },
    {
      "entropy": 1.0667495369911193,
      "epoch": 4.535666218034994,
      "grad_norm": 0.9514157772064209,
      "learning_rate": 6.172960770778948e-05,
      "loss": 0.03615821003913879,
      "mean_token_accuracy": 0.9864137768745422,
      "num_tokens": 60389121.0,
      "step": 3370
    },
    {
      "entropy": 1.0640890002250671,
      "epoch": 4.549125168236878,
      "grad_norm": 1.3572577238082886,
      "learning_rate": 6.151314744009885e-05,
      "loss": 0.03793430626392365,
      "mean_token_accuracy": 0.9861454784870147,
      "num_tokens": 60568718.0,
      "step": 3380
    },
    {
      "entropy": 1.0605762481689454,
      "epoch": 4.562584118438762,
      "grad_norm": 0.9496264457702637,
      "learning_rate": 6.129645907004395e-05,
      "loss": 0.03883695602416992,
      "mean_token_accuracy": 0.9849031150341034,
      "num_tokens": 60747908.0,
      "step": 3390
    },
    {
      "entropy": 1.0625512599945068,
      "epoch": 4.576043068640646,
      "grad_norm": 1.4468268156051636,
      "learning_rate": 6.107954689072796e-05,
      "loss": 0.03865863978862762,
      "mean_token_accuracy": 0.9855649054050446,
      "num_tokens": 60927000.0,
      "step": 3400
    },
    {
      "entropy": 1.0635502815246582,
      "epoch": 4.589502018842531,
      "grad_norm": 1.2759664058685303,
      "learning_rate": 6.086241519968822e-05,
      "loss": 0.037098705768585205,
      "mean_token_accuracy": 0.9857240319252014,
      "num_tokens": 61105703.0,
      "step": 3410
    },
    {
      "entropy": 1.0495989680290223,
      "epoch": 4.602960969044415,
      "grad_norm": 1.011147379875183,
      "learning_rate": 6.064506829881109e-05,
      "loss": 0.03874770402908325,
      "mean_token_accuracy": 0.9859291315078735,
      "num_tokens": 61285304.0,
      "step": 3420
    },
    {
      "entropy": 1.05042724609375,
      "epoch": 4.616419919246299,
      "grad_norm": 1.129265308380127,
      "learning_rate": 6.042751049424675e-05,
      "loss": 0.0369877964258194,
      "mean_token_accuracy": 0.9860033929347992,
      "num_tokens": 61464950.0,
      "step": 3430
    },
    {
      "entropy": 1.0447500228881836,
      "epoch": 4.629878869448183,
      "grad_norm": 1.359832525253296,
      "learning_rate": 6.02097460963239e-05,
      "loss": 0.03797791302204132,
      "mean_token_accuracy": 0.9854108691215515,
      "num_tokens": 61644720.0,
      "step": 3440
    },
    {
      "entropy": 1.0469836950302125,
      "epoch": 4.643337819650068,
      "grad_norm": 1.5161691904067993,
      "learning_rate": 5.999177941946429e-05,
      "loss": 0.03715096116065979,
      "mean_token_accuracy": 0.9859564423561096,
      "num_tokens": 61824536.0,
      "step": 3450
    },
    {
      "entropy": 1.0558869957923889,
      "epoch": 4.656796769851952,
      "grad_norm": 1.3739120960235596,
      "learning_rate": 5.977361478209732e-05,
      "loss": 0.03880961239337921,
      "mean_token_accuracy": 0.9851081371307373,
      "num_tokens": 62003793.0,
      "step": 3460
    },
    {
      "entropy": 1.0508078336715698,
      "epoch": 4.670255720053836,
      "grad_norm": 1.2848538160324097,
      "learning_rate": 5.955525650657444e-05,
      "loss": 0.03713637590408325,
      "mean_token_accuracy": 0.9857253849506378,
      "num_tokens": 62184015.0,
      "step": 3470
    },
    {
      "entropy": 1.0443377375602723,
      "epoch": 4.68371467025572,
      "grad_norm": 1.073598027229309,
      "learning_rate": 5.933670891908355e-05,
      "loss": 0.03473606109619141,
      "mean_token_accuracy": 0.9873181879520416,
      "num_tokens": 62363492.0,
      "step": 3480
    },
    {
      "entropy": 1.0368189215660095,
      "epoch": 4.697173620457605,
      "grad_norm": 1.203999638557434,
      "learning_rate": 5.9117976349563206e-05,
      "loss": 0.036920982599258426,
      "mean_token_accuracy": 0.986221992969513,
      "num_tokens": 62542906.0,
      "step": 3490
    },
    {
      "entropy": 1.043162429332733,
      "epoch": 4.710632570659489,
      "grad_norm": 1.0199706554412842,
      "learning_rate": 5.889906313161696e-05,
      "loss": 0.03679157495498657,
      "mean_token_accuracy": 0.9863799631595611,
      "num_tokens": 62721864.0,
      "step": 3500
    },
    {
      "entropy": 1.0550310969352723,
      "epoch": 4.724091520861373,
      "grad_norm": 1.172404170036316,
      "learning_rate": 5.8679973602427376e-05,
      "loss": 0.03852836787700653,
      "mean_token_accuracy": 0.9853650152683258,
      "num_tokens": 62900752.0,
      "step": 3510
    },
    {
      "entropy": 1.0554153800010682,
      "epoch": 4.737550471063257,
      "grad_norm": 1.2630856037139893,
      "learning_rate": 5.846071210267018e-05,
      "loss": 0.03773666620254516,
      "mean_token_accuracy": 0.9868132829666137,
      "num_tokens": 63079670.0,
      "step": 3520
    },
    {
      "entropy": 1.0372057795524596,
      "epoch": 4.751009421265142,
      "grad_norm": 1.2371525764465332,
      "learning_rate": 5.824128297642823e-05,
      "loss": 0.0382472962141037,
      "mean_token_accuracy": 0.9858209669589997,
      "num_tokens": 63258512.0,
      "step": 3530
    },
    {
      "entropy": 1.0347737193107605,
      "epoch": 4.764468371467026,
      "grad_norm": 0.9487586617469788,
      "learning_rate": 5.802169057110548e-05,
      "loss": 0.035361993312835696,
      "mean_token_accuracy": 0.9863068044185639,
      "num_tokens": 63438267.0,
      "step": 3540
    },
    {
      "entropy": 1.050095522403717,
      "epoch": 4.77792732166891,
      "grad_norm": 1.0213005542755127,
      "learning_rate": 5.7801939237340786e-05,
      "loss": 0.03737463653087616,
      "mean_token_accuracy": 0.986017245054245,
      "num_tokens": 63617530.0,
      "step": 3550
    },
    {
      "entropy": 1.0533493757247925,
      "epoch": 4.7913862718707945,
      "grad_norm": 1.1204028129577637,
      "learning_rate": 5.758203332892177e-05,
      "loss": 0.03767351508140564,
      "mean_token_accuracy": 0.9861111581325531,
      "num_tokens": 63796732.0,
      "step": 3560
    },
    {
      "entropy": 1.0467125058174134,
      "epoch": 4.804845222072679,
      "grad_norm": 1.1478379964828491,
      "learning_rate": 5.736197720269855e-05,
      "loss": 0.03344551920890808,
      "mean_token_accuracy": 0.9871456980705261,
      "num_tokens": 63975699.0,
      "step": 3570
    },
    {
      "entropy": 1.0390986204147339,
      "epoch": 4.818304172274562,
      "grad_norm": 1.1709128618240356,
      "learning_rate": 5.714177521849736e-05,
      "loss": 0.03542309701442718,
      "mean_token_accuracy": 0.9866024851799011,
      "num_tokens": 64154668.0,
      "step": 3580
    },
    {
      "entropy": 1.0338069915771484,
      "epoch": 4.831763122476447,
      "grad_norm": 1.0544331073760986,
      "learning_rate": 5.69214317390343e-05,
      "loss": 0.032989829778671265,
      "mean_token_accuracy": 0.987553596496582,
      "num_tokens": 64334132.0,
      "step": 3590
    },
    {
      "entropy": 1.0331252694129944,
      "epoch": 4.845222072678331,
      "grad_norm": 1.3105525970458984,
      "learning_rate": 5.670095112982875e-05,
      "loss": 0.03612107336521149,
      "mean_token_accuracy": 0.986699789762497,
      "num_tokens": 64513056.0,
      "step": 3600
    },
    {
      "entropy": 1.0471250653266906,
      "epoch": 4.858681022880216,
      "grad_norm": 1.1299049854278564,
      "learning_rate": 5.648033775911701e-05,
      "loss": 0.03683598637580872,
      "mean_token_accuracy": 0.9859676957130432,
      "num_tokens": 64692811.0,
      "step": 3610
    },
    {
      "entropy": 1.0519919276237488,
      "epoch": 4.872139973082099,
      "grad_norm": 1.1945998668670654,
      "learning_rate": 5.625959599776564e-05,
      "loss": 0.038029834628105164,
      "mean_token_accuracy": 0.9856249630451203,
      "num_tokens": 64872098.0,
      "step": 3620
    },
    {
      "entropy": 1.0599406599998473,
      "epoch": 4.885598923283984,
      "grad_norm": 1.1491973400115967,
      "learning_rate": 5.603873021918493e-05,
      "loss": 0.03606452643871307,
      "mean_token_accuracy": 0.9867934942245483,
      "num_tokens": 65050691.0,
      "step": 3630
    },
    {
      "entropy": 1.0520189166069032,
      "epoch": 4.899057873485868,
      "grad_norm": 1.2668406963348389,
      "learning_rate": 5.581774479924229e-05,
      "loss": 0.03717108964920044,
      "mean_token_accuracy": 0.9856311261653901,
      "num_tokens": 65230073.0,
      "step": 3640
    },
    {
      "entropy": 1.0371422290802002,
      "epoch": 4.912516823687753,
      "grad_norm": 1.2444928884506226,
      "learning_rate": 5.5596644116175444e-05,
      "loss": 0.03783654570579529,
      "mean_token_accuracy": 0.9851936995983124,
      "num_tokens": 65409403.0,
      "step": 3650
    },
    {
      "entropy": 1.035485291481018,
      "epoch": 4.925975773889636,
      "grad_norm": 1.1150327920913696,
      "learning_rate": 5.537543255050579e-05,
      "loss": 0.03533655405044556,
      "mean_token_accuracy": 0.986861526966095,
      "num_tokens": 65588232.0,
      "step": 3660
    },
    {
      "entropy": 1.0358031511306762,
      "epoch": 4.939434724091521,
      "grad_norm": 1.2493401765823364,
      "learning_rate": 5.5154114484951556e-05,
      "loss": 0.03752387762069702,
      "mean_token_accuracy": 0.9860036075115204,
      "num_tokens": 65767287.0,
      "step": 3670
    },
    {
      "entropy": 1.0428583860397338,
      "epoch": 4.952893674293405,
      "grad_norm": 1.1343940496444702,
      "learning_rate": 5.4932694304340985e-05,
      "loss": 0.03644071221351623,
      "mean_token_accuracy": 0.9862412512302399,
      "num_tokens": 65946115.0,
      "step": 3680
    },
    {
      "entropy": 1.050352644920349,
      "epoch": 4.96635262449529,
      "grad_norm": 1.3006900548934937,
      "learning_rate": 5.471117639552543e-05,
      "loss": 0.035864520072937014,
      "mean_token_accuracy": 0.9864962816238403,
      "num_tokens": 66125715.0,
      "step": 3690
    },
    {
      "entropy": 1.048657739162445,
      "epoch": 4.979811574697173,
      "grad_norm": 1.1723651885986328,
      "learning_rate": 5.448956514729251e-05,
      "loss": 0.03770222663879395,
      "mean_token_accuracy": 0.9857257843017578,
      "num_tokens": 66304829.0,
      "step": 3700
    },
    {
      "entropy": 1.044695222377777,
      "epoch": 4.993270524899058,
      "grad_norm": 0.9742991328239441,
      "learning_rate": 5.426786495027908e-05,
      "loss": 0.03706640601158142,
      "mean_token_accuracy": 0.985941207408905,
      "num_tokens": 66483883.0,
      "step": 3710
    },
    {
      "epoch": 5.0,
      "eval_entropy": 1.0356649699484466,
      "eval_loss": 0.12430668622255325,
      "eval_mean_token_accuracy": 0.9573376892478602,
      "eval_num_tokens": 66573703.0,
      "eval_runtime": 12.7724,
      "eval_samples_per_second": 391.468,
      "eval_steps_per_second": 12.292,
      "step": 3715
    },
    {
      "entropy": 1.0324241042137146,
      "epoch": 5.006729475100942,
      "grad_norm": 0.9353874921798706,
      "learning_rate": 5.404608019688432e-05,
      "loss": 0.028589162230491637,
      "mean_token_accuracy": 0.9895434856414795,
      "num_tokens": 66663377.0,
      "step": 3720
    },
    {
      "entropy": 1.0308491230010985,
      "epoch": 5.020188425302826,
      "grad_norm": 1.2845776081085205,
      "learning_rate": 5.382421528118262e-05,
      "loss": 0.024698126316070556,
      "mean_token_accuracy": 0.9909838557243347,
      "num_tokens": 66842022.0,
      "step": 3730
    },
    {
      "entropy": 1.0136583924293519,
      "epoch": 5.03364737550471,
      "grad_norm": 0.8143796324729919,
      "learning_rate": 5.360227459883662e-05,
      "loss": 0.021929217875003813,
      "mean_token_accuracy": 0.9924641191959381,
      "num_tokens": 67021330.0,
      "step": 3740
    },
    {
      "entropy": 1.002350401878357,
      "epoch": 5.0471063257065945,
      "grad_norm": 1.3543813228607178,
      "learning_rate": 5.338026254701003e-05,
      "loss": 0.024280667304992676,
      "mean_token_accuracy": 0.9914784133434296,
      "num_tokens": 67200604.0,
      "step": 3750
    },
    {
      "entropy": 1.0096928119659423,
      "epoch": 5.060565275908479,
      "grad_norm": 1.0620241165161133,
      "learning_rate": 5.31581835242806e-05,
      "loss": 0.023655575513839722,
      "mean_token_accuracy": 0.9915423214435577,
      "num_tokens": 67380371.0,
      "step": 3760
    },
    {
      "entropy": 1.021187925338745,
      "epoch": 5.074024226110363,
      "grad_norm": 1.0421315431594849,
      "learning_rate": 5.293604193055289e-05,
      "loss": 0.02454202026128769,
      "mean_token_accuracy": 0.9910594940185546,
      "num_tokens": 67559723.0,
      "step": 3770
    },
    {
      "entropy": 1.018889594078064,
      "epoch": 5.087483176312247,
      "grad_norm": 0.9473801851272583,
      "learning_rate": 5.2713842166971165e-05,
      "loss": 0.025448399782180785,
      "mean_token_accuracy": 0.9912221610546113,
      "num_tokens": 67738595.0,
      "step": 3780
    },
    {
      "entropy": 1.0147689938545228,
      "epoch": 5.1009421265141315,
      "grad_norm": 0.939347505569458,
      "learning_rate": 5.249158863583216e-05,
      "loss": 0.024416190385818482,
      "mean_token_accuracy": 0.9910872936248779,
      "num_tokens": 67917476.0,
      "step": 3790
    },
    {
      "entropy": 1.0090774953365327,
      "epoch": 5.114401076716016,
      "grad_norm": 1.214951753616333,
      "learning_rate": 5.2269285740497876e-05,
      "loss": 0.02616993188858032,
      "mean_token_accuracy": 0.9900912940502167,
      "num_tokens": 68097328.0,
      "step": 3800
    },
    {
      "entropy": 1.0201542019844054,
      "epoch": 5.1278600269179,
      "grad_norm": 0.9982582926750183,
      "learning_rate": 5.204693788530832e-05,
      "loss": 0.026025664806365967,
      "mean_token_accuracy": 0.9906771540641784,
      "num_tokens": 68276754.0,
      "step": 3810
    },
    {
      "entropy": 1.0209372758865356,
      "epoch": 5.141318977119784,
      "grad_norm": 1.2663366794586182,
      "learning_rate": 5.182454947549428e-05,
      "loss": 0.02434307783842087,
      "mean_token_accuracy": 0.99150630235672,
      "num_tokens": 68456764.0,
      "step": 3820
    },
    {
      "entropy": 1.0369561433792114,
      "epoch": 5.1547779273216685,
      "grad_norm": 1.13618004322052,
      "learning_rate": 5.160212491709002e-05,
      "loss": 0.025401628017425536,
      "mean_token_accuracy": 0.9907406866550446,
      "num_tokens": 68635336.0,
      "step": 3830
    },
    {
      "entropy": 1.0440601348876952,
      "epoch": 5.168236877523553,
      "grad_norm": 1.2365202903747559,
      "learning_rate": 5.1379668616845975e-05,
      "loss": 0.02707371711730957,
      "mean_token_accuracy": 0.9900575876235962,
      "num_tokens": 68814346.0,
      "step": 3840
    },
    {
      "entropy": 1.0355126500129699,
      "epoch": 5.181695827725437,
      "grad_norm": 0.9310021996498108,
      "learning_rate": 5.115718498214148e-05,
      "loss": 0.02353193610906601,
      "mean_token_accuracy": 0.9916076958179474,
      "num_tokens": 68993930.0,
      "step": 3850
    },
    {
      "entropy": 1.0217126727104187,
      "epoch": 5.195154777927321,
      "grad_norm": 1.1135667562484741,
      "learning_rate": 5.093467842089742e-05,
      "loss": 0.0253662109375,
      "mean_token_accuracy": 0.9902997076511383,
      "num_tokens": 69173251.0,
      "step": 3860
    },
    {
      "entropy": 1.0064147591590882,
      "epoch": 5.2086137281292055,
      "grad_norm": 1.0604337453842163,
      "learning_rate": 5.071215334148891e-05,
      "loss": 0.026254481077194212,
      "mean_token_accuracy": 0.990457046031952,
      "num_tokens": 69352492.0,
      "step": 3870
    },
    {
      "entropy": 1.0141639947891234,
      "epoch": 5.22207267833109,
      "grad_norm": 1.1456916332244873,
      "learning_rate": 5.048961415265797e-05,
      "loss": 0.027528288960456847,
      "mean_token_accuracy": 0.9905337333679199,
      "num_tokens": 69531556.0,
      "step": 3880
    },
    {
      "entropy": 1.0185627937316895,
      "epoch": 5.235531628532974,
      "grad_norm": 1.3022476434707642,
      "learning_rate": 5.0267065263426125e-05,
      "loss": 0.0235461950302124,
      "mean_token_accuracy": 0.9911311268806458,
      "num_tokens": 69710721.0,
      "step": 3890
    },
    {
      "entropy": 1.0204244375228881,
      "epoch": 5.248990578734858,
      "grad_norm": 1.01898992061615,
      "learning_rate": 5.00445110830071e-05,
      "loss": 0.024360083043575287,
      "mean_token_accuracy": 0.991031140089035,
      "num_tokens": 69890062.0,
      "step": 3900
    },
    {
      "entropy": 1.0131066560745239,
      "epoch": 5.262449528936743,
      "grad_norm": 1.0795366764068604,
      "learning_rate": 4.9821956020719474e-05,
      "loss": 0.02722684442996979,
      "mean_token_accuracy": 0.989804482460022,
      "num_tokens": 70068900.0,
      "step": 3910
    },
    {
      "entropy": 1.025877797603607,
      "epoch": 5.275908479138627,
      "grad_norm": 1.2080345153808594,
      "learning_rate": 4.959940448589928e-05,
      "loss": 0.027399671077728272,
      "mean_token_accuracy": 0.9899636626243591,
      "num_tokens": 70248062.0,
      "step": 3920
    },
    {
      "entropy": 1.0310349464416504,
      "epoch": 5.289367429340511,
      "grad_norm": 1.1293290853500366,
      "learning_rate": 4.9376860887812666e-05,
      "loss": 0.02553839683532715,
      "mean_token_accuracy": 0.9908572614192963,
      "num_tokens": 70427066.0,
      "step": 3930
    },
    {
      "entropy": 1.024777591228485,
      "epoch": 5.302826379542395,
      "grad_norm": 0.9517710208892822,
      "learning_rate": 4.915432963556853e-05,
      "loss": 0.026053178310394286,
      "mean_token_accuracy": 0.9905928015708924,
      "num_tokens": 70606283.0,
      "step": 3940
    },
    {
      "entropy": 1.024202048778534,
      "epoch": 5.31628532974428,
      "grad_norm": 1.156661868095398,
      "learning_rate": 4.8931815138031173e-05,
      "loss": 0.02635410726070404,
      "mean_token_accuracy": 0.9904537379741669,
      "num_tokens": 70784962.0,
      "step": 3950
    },
    {
      "entropy": 1.0223736405372619,
      "epoch": 5.329744279946164,
      "grad_norm": 1.0991413593292236,
      "learning_rate": 4.870932180373296e-05,
      "loss": 0.026505589485168457,
      "mean_token_accuracy": 0.9905425131320953,
      "num_tokens": 70964289.0,
      "step": 3960
    },
    {
      "entropy": 1.0177703976631165,
      "epoch": 5.343203230148048,
      "grad_norm": 1.0344407558441162,
      "learning_rate": 4.8486854040786926e-05,
      "loss": 0.026624417304992674,
      "mean_token_accuracy": 0.9904337406158448,
      "num_tokens": 71144133.0,
      "step": 3970
    },
    {
      "entropy": 1.0158238291740418,
      "epoch": 5.356662180349932,
      "grad_norm": 0.8977853059768677,
      "learning_rate": 4.826441625679953e-05,
      "loss": 0.026179781556129454,
      "mean_token_accuracy": 0.9901742041110992,
      "num_tokens": 71323642.0,
      "step": 3980
    },
    {
      "entropy": 1.0216598749160766,
      "epoch": 5.370121130551817,
      "grad_norm": 0.8806330561637878,
      "learning_rate": 4.8042012858783223e-05,
      "loss": 0.024043604731559753,
      "mean_token_accuracy": 0.9914625465869904,
      "num_tokens": 71502613.0,
      "step": 3990
    },
    {
      "entropy": 1.0118016242980956,
      "epoch": 5.383580080753701,
      "grad_norm": 0.9316030740737915,
      "learning_rate": 4.781964825306923e-05,
      "loss": 0.023112615942955016,
      "mean_token_accuracy": 0.9917285740375519,
      "num_tokens": 71682439.0,
      "step": 4000
    },
    {
      "entropy": 1.0056150257587433,
      "epoch": 5.397039030955585,
      "grad_norm": 0.8675498366355896,
      "learning_rate": 4.7597326845220206e-05,
      "loss": 0.02428761273622513,
      "mean_token_accuracy": 0.9905657172203064,
      "num_tokens": 71861391.0,
      "step": 4010
    },
    {
      "entropy": 1.0037673354148864,
      "epoch": 5.410497981157469,
      "grad_norm": 1.0555452108383179,
      "learning_rate": 4.737505303994292e-05,
      "loss": 0.0254961758852005,
      "mean_token_accuracy": 0.9906506896018982,
      "num_tokens": 72040499.0,
      "step": 4020
    },
    {
      "entropy": 0.9917411029338836,
      "epoch": 5.423956931359354,
      "grad_norm": 1.0477594137191772,
      "learning_rate": 4.7152831241001065e-05,
      "loss": 0.023488616943359374,
      "mean_token_accuracy": 0.991758120059967,
      "num_tokens": 72220507.0,
      "step": 4030
    },
    {
      "entropy": 0.9974137663841247,
      "epoch": 5.437415881561238,
      "grad_norm": 1.189155101776123,
      "learning_rate": 4.693066585112795e-05,
      "loss": 0.024205952882766724,
      "mean_token_accuracy": 0.9910655796527863,
      "num_tokens": 72400051.0,
      "step": 4040
    },
    {
      "entropy": 1.0056391954421997,
      "epoch": 5.450874831763122,
      "grad_norm": 1.024430751800537,
      "learning_rate": 4.670856127193928e-05,
      "loss": 0.0251724511384964,
      "mean_token_accuracy": 0.990533047914505,
      "num_tokens": 72578988.0,
      "step": 4050
    },
    {
      "entropy": 1.000277614593506,
      "epoch": 5.464333781965006,
      "grad_norm": 1.0882636308670044,
      "learning_rate": 4.648652190384597e-05,
      "loss": 0.02423400282859802,
      "mean_token_accuracy": 0.9916969239711761,
      "num_tokens": 72758741.0,
      "step": 4060
    },
    {
      "entropy": 1.0084757328033447,
      "epoch": 5.477792732166891,
      "grad_norm": 0.7930824756622314,
      "learning_rate": 4.626455214596695e-05,
      "loss": 0.023505699634552003,
      "mean_token_accuracy": 0.9915890038013458,
      "num_tokens": 72938170.0,
      "step": 4070
    },
    {
      "entropy": 1.015303874015808,
      "epoch": 5.491251682368775,
      "grad_norm": 1.3612091541290283,
      "learning_rate": 4.6042656396042e-05,
      "loss": 0.025665953755378723,
      "mean_token_accuracy": 0.990752911567688,
      "num_tokens": 73117014.0,
      "step": 4080
    },
    {
      "entropy": 1.0118269324302673,
      "epoch": 5.504710632570659,
      "grad_norm": 0.9727089405059814,
      "learning_rate": 4.5820839050344643e-05,
      "loss": 0.027782341837882994,
      "mean_token_accuracy": 0.9897802293300628,
      "num_tokens": 73296744.0,
      "step": 4090
    },
    {
      "entropy": 1.0124372959136962,
      "epoch": 5.518169582772543,
      "grad_norm": 1.0280526876449585,
      "learning_rate": 4.559910450359502e-05,
      "loss": 0.026491034030914306,
      "mean_token_accuracy": 0.9905598402023316,
      "num_tokens": 73476213.0,
      "step": 4100
    },
    {
      "entropy": 1.0201756000518798,
      "epoch": 5.531628532974428,
      "grad_norm": 1.0505496263504028,
      "learning_rate": 4.5377457148872837e-05,
      "loss": 0.02565605640411377,
      "mean_token_accuracy": 0.9904898762702942,
      "num_tokens": 73655498.0,
      "step": 4110
    },
    {
      "entropy": 1.0122828364372254,
      "epoch": 5.545087483176312,
      "grad_norm": 0.9964900016784668,
      "learning_rate": 4.515590137753032e-05,
      "loss": 0.026221027970314024,
      "mean_token_accuracy": 0.9898638904094696,
      "num_tokens": 73834885.0,
      "step": 4120
    },
    {
      "entropy": 1.002439957857132,
      "epoch": 5.558546433378196,
      "grad_norm": 1.2505117654800415,
      "learning_rate": 4.493444157910521e-05,
      "loss": 0.025265097618103027,
      "mean_token_accuracy": 0.9906082093715668,
      "num_tokens": 74014153.0,
      "step": 4130
    },
    {
      "entropy": 1.0038666009902955,
      "epoch": 5.5720053835800805,
      "grad_norm": 1.189091682434082,
      "learning_rate": 4.471308214123381e-05,
      "loss": 0.027262809872627258,
      "mean_token_accuracy": 0.9900608360767365,
      "num_tokens": 74192608.0,
      "step": 4140
    },
    {
      "entropy": 1.0128184318542481,
      "epoch": 5.585464333781965,
      "grad_norm": 1.4007318019866943,
      "learning_rate": 4.449182744956403e-05,
      "loss": 0.02719965875148773,
      "mean_token_accuracy": 0.9899117827415467,
      "num_tokens": 74371709.0,
      "step": 4150
    },
    {
      "entropy": 1.01895170211792,
      "epoch": 5.598923283983849,
      "grad_norm": 0.764778733253479,
      "learning_rate": 4.4270681887668544e-05,
      "loss": 0.024375322461128234,
      "mean_token_accuracy": 0.9916343033313751,
      "num_tokens": 74550754.0,
      "step": 4160
    },
    {
      "entropy": 1.0186664938926697,
      "epoch": 5.612382234185733,
      "grad_norm": 1.1775010824203491,
      "learning_rate": 4.404964983695786e-05,
      "loss": 0.024209040403366088,
      "mean_token_accuracy": 0.9905198216438293,
      "num_tokens": 74729681.0,
      "step": 4170
    },
    {
      "entropy": 1.0036309778690338,
      "epoch": 5.6258411843876175,
      "grad_norm": 0.9738861322402954,
      "learning_rate": 4.382873567659361e-05,
      "loss": 0.02190125733613968,
      "mean_token_accuracy": 0.9924726009368896,
      "num_tokens": 74908847.0,
      "step": 4180
    },
    {
      "entropy": 0.9877856969833374,
      "epoch": 5.639300134589502,
      "grad_norm": 1.000006914138794,
      "learning_rate": 4.3607943783401736e-05,
      "loss": 0.023811979591846465,
      "mean_token_accuracy": 0.9915596306324005,
      "num_tokens": 75087768.0,
      "step": 4190
    },
    {
      "entropy": 0.978437089920044,
      "epoch": 5.652759084791386,
      "grad_norm": 0.9680097103118896,
      "learning_rate": 4.3387278531785747e-05,
      "loss": 0.023803821206092833,
      "mean_token_accuracy": 0.9915006577968597,
      "num_tokens": 75267006.0,
      "step": 4200
    },
    {
      "entropy": 0.9951873600482941,
      "epoch": 5.66621803499327,
      "grad_norm": 1.0874643325805664,
      "learning_rate": 4.3166744293640134e-05,
      "loss": 0.024874386191368104,
      "mean_token_accuracy": 0.9910345673561096,
      "num_tokens": 75445346.0,
      "step": 4210
    },
    {
      "entropy": 0.9993082582950592,
      "epoch": 5.6796769851951545,
      "grad_norm": 1.0696096420288086,
      "learning_rate": 4.2946345438263665e-05,
      "loss": 0.024564328789710998,
      "mean_token_accuracy": 0.9908776760101319,
      "num_tokens": 75624164.0,
      "step": 4220
    },
    {
      "entropy": 0.9952058970928193,
      "epoch": 5.693135935397039,
      "grad_norm": 1.1028224229812622,
      "learning_rate": 4.272608633227287e-05,
      "loss": 0.024792733788490295,
      "mean_token_accuracy": 0.9909776628017426,
      "num_tokens": 75803270.0,
      "step": 4230
    },
    {
      "entropy": 0.9933918535709381,
      "epoch": 5.706594885598923,
      "grad_norm": 0.9133896827697754,
      "learning_rate": 4.250597133951554e-05,
      "loss": 0.025022977590560914,
      "mean_token_accuracy": 0.9908600449562073,
      "num_tokens": 75982078.0,
      "step": 4240
    },
    {
      "entropy": 0.9992476463317871,
      "epoch": 5.720053835800807,
      "grad_norm": 0.9527667760848999,
      "learning_rate": 4.228600482098423e-05,
      "loss": 0.023611989617347718,
      "mean_token_accuracy": 0.9913343906402587,
      "num_tokens": 76161175.0,
      "step": 4250
    },
    {
      "entropy": 1.0076547145843506,
      "epoch": 5.7335127860026915,
      "grad_norm": 1.2215615510940552,
      "learning_rate": 4.206619113472986e-05,
      "loss": 0.024665328860282897,
      "mean_token_accuracy": 0.9910496056079865,
      "num_tokens": 76340176.0,
      "step": 4260
    },
    {
      "entropy": 1.0158798456192017,
      "epoch": 5.746971736204576,
      "grad_norm": 1.210359811782837,
      "learning_rate": 4.18465346357754e-05,
      "loss": 0.024151743948459627,
      "mean_token_accuracy": 0.9907714128494263,
      "num_tokens": 76519302.0,
      "step": 4270
    },
    {
      "entropy": 1.0168303012847901,
      "epoch": 5.76043068640646,
      "grad_norm": 1.1331851482391357,
      "learning_rate": 4.16270396760296e-05,
      "loss": 0.026728877425193788,
      "mean_token_accuracy": 0.9904241323471069,
      "num_tokens": 76698543.0,
      "step": 4280
    },
    {
      "entropy": 1.0135691165924072,
      "epoch": 5.773889636608344,
      "grad_norm": 1.0528459548950195,
      "learning_rate": 4.140771060420066e-05,
      "loss": 0.024574026465415955,
      "mean_token_accuracy": 0.9909278094768524,
      "num_tokens": 76877560.0,
      "step": 4290
    },
    {
      "entropy": 1.0032446384429932,
      "epoch": 5.787348586810229,
      "grad_norm": 0.8913487195968628,
      "learning_rate": 4.118855176571021e-05,
      "loss": 0.023184402287006377,
      "mean_token_accuracy": 0.9918475985527039,
      "num_tokens": 77057531.0,
      "step": 4300
    },
    {
      "entropy": 0.9989308416843414,
      "epoch": 5.800807537012113,
      "grad_norm": 0.7943524122238159,
      "learning_rate": 4.096956750260718e-05,
      "loss": 0.020456457138061525,
      "mean_token_accuracy": 0.9927145361900329,
      "num_tokens": 77236668.0,
      "step": 4310
    },
    {
      "entropy": 0.9927454888820648,
      "epoch": 5.814266487213997,
      "grad_norm": 1.0990914106369019,
      "learning_rate": 4.07507621534817e-05,
      "loss": 0.02571033835411072,
      "mean_token_accuracy": 0.9910105168819427,
      "num_tokens": 77415692.0,
      "step": 4320
    },
    {
      "entropy": 0.9979439377784729,
      "epoch": 5.827725437415881,
      "grad_norm": 0.9713761806488037,
      "learning_rate": 4.053214005337924e-05,
      "loss": 0.024069878458976745,
      "mean_token_accuracy": 0.9908368527889252,
      "num_tokens": 77595253.0,
      "step": 4330
    },
    {
      "entropy": 1.000143724679947,
      "epoch": 5.841184387617766,
      "grad_norm": 1.071368932723999,
      "learning_rate": 4.031370553371465e-05,
      "loss": 0.02418675720691681,
      "mean_token_accuracy": 0.9909208953380585,
      "num_tokens": 77774484.0,
      "step": 4340
    },
    {
      "entropy": 1.007152682542801,
      "epoch": 5.85464333781965,
      "grad_norm": 1.1110836267471313,
      "learning_rate": 4.0095462922186385e-05,
      "loss": 0.02507558763027191,
      "mean_token_accuracy": 0.9910820901393891,
      "num_tokens": 77954000.0,
      "step": 4350
    },
    {
      "entropy": 1.010567343235016,
      "epoch": 5.868102288021534,
      "grad_norm": 0.9108941555023193,
      "learning_rate": 3.9877416542690746e-05,
      "loss": 0.02309911847114563,
      "mean_token_accuracy": 0.9912745118141174,
      "num_tokens": 78133205.0,
      "step": 4360
    },
    {
      "entropy": 1.0007760405540467,
      "epoch": 5.881561238223418,
      "grad_norm": 0.9847738146781921,
      "learning_rate": 3.9659570715236234e-05,
      "loss": 0.022994443774223328,
      "mean_token_accuracy": 0.991887879371643,
      "num_tokens": 78312209.0,
      "step": 4370
    },
    {
      "entropy": 1.0055775403976441,
      "epoch": 5.895020188425303,
      "grad_norm": 0.9286985993385315,
      "learning_rate": 3.944192975585792e-05,
      "loss": 0.024959474802017212,
      "mean_token_accuracy": 0.9908322095870972,
      "num_tokens": 78491839.0,
      "step": 4380
    },
    {
      "entropy": 1.0130927085876464,
      "epoch": 5.908479138627187,
      "grad_norm": 1.101596713066101,
      "learning_rate": 3.922449797653198e-05,
      "loss": 0.024828463792800903,
      "mean_token_accuracy": 0.9908891916275024,
      "num_tokens": 78670330.0,
      "step": 4390
    },
    {
      "entropy": 1.0086628675460816,
      "epoch": 5.921938088829071,
      "grad_norm": 0.7992931604385376,
      "learning_rate": 3.900727968509024e-05,
      "loss": 0.0249856099486351,
      "mean_token_accuracy": 0.9912192106246949,
      "num_tokens": 78849616.0,
      "step": 4400
    },
    {
      "entropy": 1.0030723690986634,
      "epoch": 5.935397039030955,
      "grad_norm": 1.0745497941970825,
      "learning_rate": 3.879027918513483e-05,
      "loss": 0.024454452097415924,
      "mean_token_accuracy": 0.9910886645317077,
      "num_tokens": 79028280.0,
      "step": 4410
    },
    {
      "entropy": 1.0046655774116515,
      "epoch": 5.94885598923284,
      "grad_norm": 1.1712580919265747,
      "learning_rate": 3.857350077595289e-05,
      "loss": 0.024604372680187225,
      "mean_token_accuracy": 0.9910597622394561,
      "num_tokens": 79207400.0,
      "step": 4420
    },
    {
      "entropy": 1.0095412015914917,
      "epoch": 5.962314939434724,
      "grad_norm": 1.0506722927093506,
      "learning_rate": 3.835694875243149e-05,
      "loss": 0.023779749870300293,
      "mean_token_accuracy": 0.9909038186073303,
      "num_tokens": 79386847.0,
      "step": 4430
    },
    {
      "entropy": 1.014033055305481,
      "epoch": 5.975773889636608,
      "grad_norm": 1.269294261932373,
      "learning_rate": 3.814062740497243e-05,
      "loss": 0.02225509285926819,
      "mean_token_accuracy": 0.991925185918808,
      "num_tokens": 79566247.0,
      "step": 4440
    },
    {
      "entropy": 1.0057903945446014,
      "epoch": 5.989232839838492,
      "grad_norm": 0.8028563857078552,
      "learning_rate": 3.7924541019407264e-05,
      "loss": 0.021707671880722045,
      "mean_token_accuracy": 0.9925049006938934,
      "num_tokens": 79745532.0,
      "step": 4450
    },
    {
      "epoch": 6.0,
      "eval_entropy": 0.9986690609318436,
      "eval_loss": 0.1345997303724289,
      "eval_mean_token_accuracy": 0.9580068561681516,
      "eval_num_tokens": 79888260.0,
      "eval_runtime": 12.7672,
      "eval_samples_per_second": 391.628,
      "eval_steps_per_second": 12.297,
      "step": 4458
    },
    {
      "entropy": 1.0019358932971953,
      "epoch": 6.002691790040377,
      "grad_norm": 0.9917385578155518,
      "learning_rate": 3.7708693876912435e-05,
      "loss": 0.02177269458770752,
      "mean_token_accuracy": 0.9918585777282715,
      "num_tokens": 79924160.0,
      "step": 4460
    },
    {
      "entropy": 0.9939176559448242,
      "epoch": 6.016150740242261,
      "grad_norm": 0.8566797375679016,
      "learning_rate": 3.74930902539244e-05,
      "loss": 0.016230569779872896,
      "mean_token_accuracy": 0.9944312930107116,
      "num_tokens": 80103783.0,
      "step": 4470
    },
    {
      "entropy": 0.9944855153560639,
      "epoch": 6.029609690444145,
      "grad_norm": 0.7844175100326538,
      "learning_rate": 3.727773442205493e-05,
      "loss": 0.016550612449645997,
      "mean_token_accuracy": 0.9942670106887818,
      "num_tokens": 80282630.0,
      "step": 4480
    },
    {
      "entropy": 0.9905124843120575,
      "epoch": 6.043068640646029,
      "grad_norm": 0.927370011806488,
      "learning_rate": 3.7062630648006485e-05,
      "loss": 0.015363200008869171,
      "mean_token_accuracy": 0.9948609352111817,
      "num_tokens": 80461728.0,
      "step": 4490
    },
    {
      "entropy": 0.9897569298744202,
      "epoch": 6.056527590847914,
      "grad_norm": 0.8723940253257751,
      "learning_rate": 3.684778319348765e-05,
      "loss": 0.01630091369152069,
      "mean_token_accuracy": 0.9939091682434082,
      "num_tokens": 80640800.0,
      "step": 4500
    },
    {
      "entropy": 0.9903276920318603,
      "epoch": 6.069986541049798,
      "grad_norm": 0.886618435382843,
      "learning_rate": 3.663319631512874e-05,
      "loss": 0.015594318509101868,
      "mean_token_accuracy": 0.9944710969924927,
      "num_tokens": 80820047.0,
      "step": 4510
    },
    {
      "entropy": 0.9845296442508698,
      "epoch": 6.083445491251682,
      "grad_norm": 0.7410146594047546,
      "learning_rate": 3.641887426439743e-05,
      "loss": 0.014708422124385834,
      "mean_token_accuracy": 0.9950878858566284,
      "num_tokens": 81000118.0,
      "step": 4520
    },
    {
      "entropy": 0.9876994490623474,
      "epoch": 6.0969044414535665,
      "grad_norm": 0.8826921582221985,
      "learning_rate": 3.620482128751456e-05,
      "loss": 0.014842665195465088,
      "mean_token_accuracy": 0.995011156797409,
      "num_tokens": 81179122.0,
      "step": 4530
    },
    {
      "entropy": 0.9812965631484986,
      "epoch": 6.110363391655451,
      "grad_norm": 0.9228511452674866,
      "learning_rate": 3.599104162536997e-05,
      "loss": 0.016777697205543517,
      "mean_token_accuracy": 0.994401341676712,
      "num_tokens": 81358570.0,
      "step": 4540
    },
    {
      "entropy": 0.9767560362815857,
      "epoch": 6.123822341857335,
      "grad_norm": 0.869403600692749,
      "learning_rate": 3.577753951343851e-05,
      "loss": 0.01631978303194046,
      "mean_token_accuracy": 0.9939962327480316,
      "num_tokens": 81537539.0,
      "step": 4550
    },
    {
      "entropy": 0.9815046191215515,
      "epoch": 6.137281292059219,
      "grad_norm": 0.7447561025619507,
      "learning_rate": 3.55643191816961e-05,
      "loss": 0.016492336988449097,
      "mean_token_accuracy": 0.9942592322826386,
      "num_tokens": 81716684.0,
      "step": 4560
    },
    {
      "entropy": 0.9812978148460388,
      "epoch": 6.1507402422611035,
      "grad_norm": 0.9655007123947144,
      "learning_rate": 3.535138485453595e-05,
      "loss": 0.015433910489082336,
      "mean_token_accuracy": 0.994744598865509,
      "num_tokens": 81895922.0,
      "step": 4570
    },
    {
      "entropy": 0.9846289932727814,
      "epoch": 6.164199192462988,
      "grad_norm": 0.771899402141571,
      "learning_rate": 3.513874075068484e-05,
      "loss": 0.01579052358865738,
      "mean_token_accuracy": 0.994454699754715,
      "num_tokens": 82074541.0,
      "step": 4580
    },
    {
      "entropy": 0.9767491102218628,
      "epoch": 6.177658142664872,
      "grad_norm": 0.7259030342102051,
      "learning_rate": 3.492639108311955e-05,
      "loss": 0.0157473087310791,
      "mean_token_accuracy": 0.9944621860980988,
      "num_tokens": 82254238.0,
      "step": 4590
    },
    {
      "entropy": 0.9825841665267945,
      "epoch": 6.191117092866756,
      "grad_norm": 1.0318107604980469,
      "learning_rate": 3.471434005898339e-05,
      "loss": 0.015785586833953858,
      "mean_token_accuracy": 0.9947353541851044,
      "num_tokens": 82433024.0,
      "step": 4600
    },
    {
      "entropy": 0.9830503404140473,
      "epoch": 6.2045760430686405,
      "grad_norm": 0.7057244181632996,
      "learning_rate": 3.450259187950283e-05,
      "loss": 0.014754258096218109,
      "mean_token_accuracy": 0.995175975561142,
      "num_tokens": 82611917.0,
      "step": 4610
    },
    {
      "entropy": 0.9704286217689514,
      "epoch": 6.218034993270525,
      "grad_norm": 0.7422804832458496,
      "learning_rate": 3.429115073990431e-05,
      "loss": 0.014288221299648286,
      "mean_token_accuracy": 0.9952509880065918,
      "num_tokens": 82791352.0,
      "step": 4620
    },
    {
      "entropy": 0.9686006426811218,
      "epoch": 6.231493943472409,
      "grad_norm": 1.1123586893081665,
      "learning_rate": 3.408002082933107e-05,
      "loss": 0.015031391382217407,
      "mean_token_accuracy": 0.9947735011577606,
      "num_tokens": 82970025.0,
      "step": 4630
    },
    {
      "entropy": 0.9696213662624359,
      "epoch": 6.244952893674293,
      "grad_norm": 0.8872476816177368,
      "learning_rate": 3.3869206330760187e-05,
      "loss": 0.01578601598739624,
      "mean_token_accuracy": 0.9947343170642853,
      "num_tokens": 83148439.0,
      "step": 4640
    },
    {
      "entropy": 0.97174631357193,
      "epoch": 6.2584118438761775,
      "grad_norm": 1.1288238763809204,
      "learning_rate": 3.365871142091968e-05,
      "loss": 0.017076049745082856,
      "mean_token_accuracy": 0.9942302107810974,
      "num_tokens": 83327163.0,
      "step": 4650
    },
    {
      "entropy": 0.97059445977211,
      "epoch": 6.271870794078062,
      "grad_norm": 0.9584773182868958,
      "learning_rate": 3.3448540270205766e-05,
      "loss": 0.015507197380065918,
      "mean_token_accuracy": 0.9949185311794281,
      "num_tokens": 83506315.0,
      "step": 4660
    },
    {
      "entropy": 0.9751429915428161,
      "epoch": 6.285329744279946,
      "grad_norm": 1.0442348718643188,
      "learning_rate": 3.323869704260025e-05,
      "loss": 0.01604621559381485,
      "mean_token_accuracy": 0.9944981873035431,
      "num_tokens": 83685291.0,
      "step": 4670
    },
    {
      "entropy": 0.9754107832908631,
      "epoch": 6.29878869448183,
      "grad_norm": 1.268025279045105,
      "learning_rate": 3.302918589558801e-05,
      "loss": 0.016522862017154694,
      "mean_token_accuracy": 0.9943258702754975,
      "num_tokens": 83864366.0,
      "step": 4680
    },
    {
      "entropy": 0.9815966248512268,
      "epoch": 6.312247644683715,
      "grad_norm": 0.9281536936759949,
      "learning_rate": 3.282001098007462e-05,
      "loss": 0.016834209859371185,
      "mean_token_accuracy": 0.9941475093364716,
      "num_tokens": 84043214.0,
      "step": 4690
    },
    {
      "entropy": 0.9755142807960511,
      "epoch": 6.325706594885599,
      "grad_norm": 0.707933247089386,
      "learning_rate": 3.261117644030412e-05,
      "loss": 0.01579307168722153,
      "mean_token_accuracy": 0.9944198191165924,
      "num_tokens": 84222708.0,
      "step": 4700
    },
    {
      "entropy": 0.975419533252716,
      "epoch": 6.339165545087483,
      "grad_norm": 1.036906361579895,
      "learning_rate": 3.240268641377694e-05,
      "loss": 0.015164561569690704,
      "mean_token_accuracy": 0.9947849869728088,
      "num_tokens": 84402234.0,
      "step": 4710
    },
    {
      "entropy": 0.9727317273616791,
      "epoch": 6.352624495289367,
      "grad_norm": 0.897186815738678,
      "learning_rate": 3.2194545031167866e-05,
      "loss": 0.016288670897483825,
      "mean_token_accuracy": 0.9942747056484222,
      "num_tokens": 84580970.0,
      "step": 4720
    },
    {
      "entropy": 0.9730606615543366,
      "epoch": 6.366083445491252,
      "grad_norm": 0.7614856958389282,
      "learning_rate": 3.1986756416244245e-05,
      "loss": 0.015893808007240294,
      "mean_token_accuracy": 0.9945310592651367,
      "num_tokens": 84759995.0,
      "step": 4730
    },
    {
      "entropy": 0.9783101677894592,
      "epoch": 6.379542395693136,
      "grad_norm": 0.8672609329223633,
      "learning_rate": 3.177932468578426e-05,
      "loss": 0.015059471130371094,
      "mean_token_accuracy": 0.9949405431747437,
      "num_tokens": 84939577.0,
      "step": 4740
    },
    {
      "entropy": 0.9774933516979217,
      "epoch": 6.39300134589502,
      "grad_norm": 1.1011027097702026,
      "learning_rate": 3.157225394949542e-05,
      "loss": 0.015801313519477844,
      "mean_token_accuracy": 0.9942348599433899,
      "num_tokens": 85119302.0,
      "step": 4750
    },
    {
      "entropy": 0.9862867891788483,
      "epoch": 6.406460296096904,
      "grad_norm": 0.813953161239624,
      "learning_rate": 3.136554830993304e-05,
      "loss": 0.015246957540512085,
      "mean_token_accuracy": 0.9948500633239746,
      "num_tokens": 85298248.0,
      "step": 4760
    },
    {
      "entropy": 0.9877782166004181,
      "epoch": 6.419919246298789,
      "grad_norm": 0.8229409456253052,
      "learning_rate": 3.115921186241906e-05,
      "loss": 0.015437261760234832,
      "mean_token_accuracy": 0.9939783871173858,
      "num_tokens": 85477420.0,
      "step": 4770
    },
    {
      "entropy": 0.9822047591209412,
      "epoch": 6.433378196500673,
      "grad_norm": 1.027951955795288,
      "learning_rate": 3.0953248694960824e-05,
      "loss": 0.01582965850830078,
      "mean_token_accuracy": 0.9941119194030762,
      "num_tokens": 85656231.0,
      "step": 4780
    },
    {
      "entropy": 0.9807499408721924,
      "epoch": 6.446837146702557,
      "grad_norm": 0.6412605047225952,
      "learning_rate": 3.0747662888170146e-05,
      "loss": 0.01425456702709198,
      "mean_token_accuracy": 0.994831895828247,
      "num_tokens": 85835707.0,
      "step": 4790
    },
    {
      "entropy": 0.978723019361496,
      "epoch": 6.460296096904441,
      "grad_norm": 0.8043062090873718,
      "learning_rate": 3.054245851518246e-05,
      "loss": 0.015074288845062256,
      "mean_token_accuracy": 0.9949940204620361,
      "num_tokens": 86014951.0,
      "step": 4800
    },
    {
      "entropy": 0.9753713130950927,
      "epoch": 6.473755047106326,
      "grad_norm": 0.787334680557251,
      "learning_rate": 3.0337639641576065e-05,
      "loss": 0.015617562830448151,
      "mean_token_accuracy": 0.9947958707809448,
      "num_tokens": 86194757.0,
      "step": 4810
    },
    {
      "entropy": 0.9751108765602112,
      "epoch": 6.48721399730821,
      "grad_norm": 0.8143076300621033,
      "learning_rate": 3.0133210325291662e-05,
      "loss": 0.01638820767402649,
      "mean_token_accuracy": 0.9946176171302795,
      "num_tokens": 86374538.0,
      "step": 4820
    },
    {
      "entropy": 0.9809349179267883,
      "epoch": 6.500672947510094,
      "grad_norm": 0.8488563895225525,
      "learning_rate": 2.9929174616551857e-05,
      "loss": 0.015374109148979187,
      "mean_token_accuracy": 0.9945039927959443,
      "num_tokens": 86553350.0,
      "step": 4830
    },
    {
      "entropy": 0.9858408272266388,
      "epoch": 6.514131897711978,
      "grad_norm": 0.9234822988510132,
      "learning_rate": 2.9725536557781008e-05,
      "loss": 0.014567127823829651,
      "mean_token_accuracy": 0.9946143448352813,
      "num_tokens": 86732008.0,
      "step": 4840
    },
    {
      "entropy": 0.9714087128639222,
      "epoch": 6.527590847913863,
      "grad_norm": 0.7859043478965759,
      "learning_rate": 2.9522300183525097e-05,
      "loss": 0.014002390205860138,
      "mean_token_accuracy": 0.9951234996318817,
      "num_tokens": 86911698.0,
      "step": 4850
    },
    {
      "entropy": 0.9689762830734253,
      "epoch": 6.541049798115747,
      "grad_norm": 0.7839444279670715,
      "learning_rate": 2.931946952037179e-05,
      "loss": 0.015208807587623597,
      "mean_token_accuracy": 0.9949181616306305,
      "num_tokens": 87090127.0,
      "step": 4860
    },
    {
      "entropy": 0.9610137641429901,
      "epoch": 6.554508748317631,
      "grad_norm": 0.8249244689941406,
      "learning_rate": 2.9117048586870654e-05,
      "loss": 0.013001258671283721,
      "mean_token_accuracy": 0.9952585935592652,
      "num_tokens": 87269489.0,
      "step": 4870
    },
    {
      "entropy": 0.965034544467926,
      "epoch": 6.5679676985195155,
      "grad_norm": 0.656859815120697,
      "learning_rate": 2.891504139345358e-05,
      "loss": 0.0163768470287323,
      "mean_token_accuracy": 0.994057834148407,
      "num_tokens": 87448301.0,
      "step": 4880
    },
    {
      "entropy": 0.9685207188129425,
      "epoch": 6.5814266487214,
      "grad_norm": 0.9349063634872437,
      "learning_rate": 2.8713451942355285e-05,
      "loss": 0.014103662967681885,
      "mean_token_accuracy": 0.9951852679252624,
      "num_tokens": 87627477.0,
      "step": 4890
    },
    {
      "entropy": 0.9671570241451264,
      "epoch": 6.594885598923284,
      "grad_norm": 0.7718268632888794,
      "learning_rate": 2.8512284227534027e-05,
      "loss": 0.014896789193153381,
      "mean_token_accuracy": 0.9945676863193512,
      "num_tokens": 87806660.0,
      "step": 4900
    },
    {
      "entropy": 0.9660835266113281,
      "epoch": 6.608344549125168,
      "grad_norm": 0.948875904083252,
      "learning_rate": 2.8311542234592497e-05,
      "loss": 0.014949330687522888,
      "mean_token_accuracy": 0.9944600164890289,
      "num_tokens": 87986236.0,
      "step": 4910
    },
    {
      "entropy": 0.9721682965755463,
      "epoch": 6.6218034993270525,
      "grad_norm": 0.8616317510604858,
      "learning_rate": 2.8111229940698842e-05,
      "loss": 0.015388941764831543,
      "mean_token_accuracy": 0.994827789068222,
      "num_tokens": 88164624.0,
      "step": 4920
    },
    {
      "entropy": 0.9700623154640198,
      "epoch": 6.635262449528937,
      "grad_norm": 0.8105503916740417,
      "learning_rate": 2.791135131450785e-05,
      "loss": 0.014793205261230468,
      "mean_token_accuracy": 0.995061457157135,
      "num_tokens": 88344024.0,
      "step": 4930
    },
    {
      "entropy": 0.9719329476356506,
      "epoch": 6.648721399730821,
      "grad_norm": 0.9461077451705933,
      "learning_rate": 2.7711910316082357e-05,
      "loss": 0.014379647374153138,
      "mean_token_accuracy": 0.9948806226253509,
      "num_tokens": 88524031.0,
      "step": 4940
    },
    {
      "entropy": 0.9734318852424622,
      "epoch": 6.662180349932705,
      "grad_norm": 1.367875576019287,
      "learning_rate": 2.7512910896814747e-05,
      "loss": 0.01659778952598572,
      "mean_token_accuracy": 0.9942364275455475,
      "num_tokens": 88703645.0,
      "step": 4950
    },
    {
      "entropy": 0.9703188121318818,
      "epoch": 6.6756393001345895,
      "grad_norm": 1.3389500379562378,
      "learning_rate": 2.7314356999348713e-05,
      "loss": 0.014963071048259734,
      "mean_token_accuracy": 0.9946206867694855,
      "num_tokens": 88883045.0,
      "step": 4960
    },
    {
      "entropy": 0.9737108051776886,
      "epoch": 6.689098250336474,
      "grad_norm": 0.9245385527610779,
      "learning_rate": 2.711625255750111e-05,
      "loss": 0.016331857442855834,
      "mean_token_accuracy": 0.9943425476551055,
      "num_tokens": 89062755.0,
      "step": 4970
    },
    {
      "entropy": 0.9814624786376953,
      "epoch": 6.702557200538358,
      "grad_norm": 0.7200494408607483,
      "learning_rate": 2.691860149618402e-05,
      "loss": 0.014298510551452637,
      "mean_token_accuracy": 0.9951595067977905,
      "num_tokens": 89241358.0,
      "step": 4980
    },
    {
      "entropy": 0.9681638181209564,
      "epoch": 6.716016150740242,
      "grad_norm": 0.8605403304100037,
      "learning_rate": 2.6721407731327004e-05,
      "loss": 0.013024243712425231,
      "mean_token_accuracy": 0.9957234740257264,
      "num_tokens": 89421272.0,
      "step": 4990
    },
    {
      "entropy": 0.9670853018760681,
      "epoch": 6.7294751009421265,
      "grad_norm": 0.8923637866973877,
      "learning_rate": 2.6524675169799506e-05,
      "loss": 0.013967196643352508,
      "mean_token_accuracy": 0.9953422844409943,
      "num_tokens": 89600056.0,
      "step": 5000
    },
    {
      "entropy": 0.9660456538200378,
      "epoch": 6.742934051144011,
      "grad_norm": 0.7672311663627625,
      "learning_rate": 2.6328407709333463e-05,
      "loss": 0.013995295763015747,
      "mean_token_accuracy": 0.9953011035919189,
      "num_tokens": 89779879.0,
      "step": 5010
    },
    {
      "entropy": 0.9670967280864715,
      "epoch": 6.756393001345895,
      "grad_norm": 0.9835227727890015,
      "learning_rate": 2.6132609238446072e-05,
      "loss": 0.014588207006454468,
      "mean_token_accuracy": 0.9950920879840851,
      "num_tokens": 89959526.0,
      "step": 5020
    },
    {
      "entropy": 0.970816558599472,
      "epoch": 6.769851951547779,
      "grad_norm": 0.7489247918128967,
      "learning_rate": 2.5937283636362724e-05,
      "loss": 0.014287692308425904,
      "mean_token_accuracy": 0.994792515039444,
      "num_tokens": 90138932.0,
      "step": 5030
    },
    {
      "entropy": 0.9720617949962616,
      "epoch": 6.783310901749664,
      "grad_norm": 0.8448637127876282,
      "learning_rate": 2.5742434772940216e-05,
      "loss": 0.015046033263206481,
      "mean_token_accuracy": 0.9947949469089508,
      "num_tokens": 90318294.0,
      "step": 5040
    },
    {
      "entropy": 0.972127640247345,
      "epoch": 6.796769851951548,
      "grad_norm": 1.0595027208328247,
      "learning_rate": 2.5548066508590007e-05,
      "loss": 0.015439464151859284,
      "mean_token_accuracy": 0.9945104479789734,
      "num_tokens": 90497522.0,
      "step": 5050
    },
    {
      "entropy": 0.9663094699382782,
      "epoch": 6.810228802153432,
      "grad_norm": 1.0235310792922974,
      "learning_rate": 2.535418269420178e-05,
      "loss": 0.01386038064956665,
      "mean_token_accuracy": 0.9952830314636231,
      "num_tokens": 90677085.0,
      "step": 5060
    },
    {
      "entropy": 0.9662455677986145,
      "epoch": 6.823687752355316,
      "grad_norm": 1.0424898862838745,
      "learning_rate": 2.5160787171067126e-05,
      "loss": 0.012799303233623504,
      "mean_token_accuracy": 0.995597755908966,
      "num_tokens": 90856270.0,
      "step": 5070
    },
    {
      "entropy": 0.957698255777359,
      "epoch": 6.837146702557201,
      "grad_norm": 0.8649188876152039,
      "learning_rate": 2.4967883770803413e-05,
      "loss": 0.013675674796104431,
      "mean_token_accuracy": 0.9950868904590606,
      "num_tokens": 91035537.0,
      "step": 5080
    },
    {
      "entropy": 0.9543747365474701,
      "epoch": 6.850605652759085,
      "grad_norm": 0.9695640206336975,
      "learning_rate": 2.477547631527799e-05,
      "loss": 0.014598837494850159,
      "mean_token_accuracy": 0.9951783776283264,
      "num_tokens": 91214714.0,
      "step": 5090
    },
    {
      "entropy": 0.9589528560638427,
      "epoch": 6.864064602960969,
      "grad_norm": 0.7789045572280884,
      "learning_rate": 2.45835686165323e-05,
      "loss": 0.014046281576156616,
      "mean_token_accuracy": 0.9951943933963776,
      "num_tokens": 91393783.0,
      "step": 5100
    },
    {
      "entropy": 0.9627966225147248,
      "epoch": 6.877523553162853,
      "grad_norm": 0.8687154650688171,
      "learning_rate": 2.4392164476706468e-05,
      "loss": 0.015588788688182831,
      "mean_token_accuracy": 0.9948146939277649,
      "num_tokens": 91573569.0,
      "step": 5110
    },
    {
      "entropy": 0.9699371516704559,
      "epoch": 6.890982503364738,
      "grad_norm": 1.4397906064987183,
      "learning_rate": 2.4201267687963935e-05,
      "loss": 0.014941152930259705,
      "mean_token_accuracy": 0.9946426331996918,
      "num_tokens": 91751970.0,
      "step": 5120
    },
    {
      "entropy": 0.9695391833782196,
      "epoch": 6.904441453566622,
      "grad_norm": 0.938165545463562,
      "learning_rate": 2.4010882032416332e-05,
      "loss": 0.01557524800300598,
      "mean_token_accuracy": 0.994098824262619,
      "num_tokens": 91931575.0,
      "step": 5130
    },
    {
      "entropy": 0.9726310014724732,
      "epoch": 6.917900403768506,
      "grad_norm": 1.0270136594772339,
      "learning_rate": 2.3821011282048545e-05,
      "loss": 0.015305042266845703,
      "mean_token_accuracy": 0.9945661365985871,
      "num_tokens": 92110694.0,
      "step": 5140
    },
    {
      "entropy": 0.9728193879127502,
      "epoch": 6.93135935397039,
      "grad_norm": 1.027956485748291,
      "learning_rate": 2.3631659198643985e-05,
      "loss": 0.01502918303012848,
      "mean_token_accuracy": 0.994612592458725,
      "num_tokens": 92289711.0,
      "step": 5150
    },
    {
      "entropy": 0.9718130826950073,
      "epoch": 6.944818304172275,
      "grad_norm": 0.926624596118927,
      "learning_rate": 2.344282953371006e-05,
      "loss": 0.013758787512779235,
      "mean_token_accuracy": 0.9950670182704926,
      "num_tokens": 92468704.0,
      "step": 5160
    },
    {
      "entropy": 0.9718481361865997,
      "epoch": 6.958277254374159,
      "grad_norm": 0.9092168211936951,
      "learning_rate": 2.325452602840385e-05,
      "loss": 0.013540112972259521,
      "mean_token_accuracy": 0.9952319920063019,
      "num_tokens": 92647600.0,
      "step": 5170
    },
    {
      "entropy": 0.9687652945518493,
      "epoch": 6.971736204576043,
      "grad_norm": 1.0669724941253662,
      "learning_rate": 2.306675241345797e-05,
      "loss": 0.014361311495304108,
      "mean_token_accuracy": 0.9953016221523285,
      "num_tokens": 92826363.0,
      "step": 5180
    },
    {
      "entropy": 0.9618574023246765,
      "epoch": 6.985195154777927,
      "grad_norm": 0.7335039377212524,
      "learning_rate": 2.287951240910668e-05,
      "loss": 0.012940369546413422,
      "mean_token_accuracy": 0.9956261157989502,
      "num_tokens": 93005625.0,
      "step": 5190
    },
    {
      "entropy": 0.959715747833252,
      "epoch": 6.998654104979812,
      "grad_norm": 0.9648978114128113,
      "learning_rate": 2.269280972501217e-05,
      "loss": 0.013761961460113525,
      "mean_token_accuracy": 0.9951907277107239,
      "num_tokens": 93184930.0,
      "step": 5200
    },
    {
      "epoch": 7.0,
      "eval_entropy": 0.9639249083342826,
      "eval_loss": 0.14176945388317108,
      "eval_mean_token_accuracy": 0.9595637909925667,
      "eval_num_tokens": 93202879.0,
      "eval_runtime": 12.8001,
      "eval_samples_per_second": 390.623,
      "eval_steps_per_second": 12.266,
      "step": 5201
    }
  ],
  "logging_steps": 10,
  "max_steps": 7430,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 4.442001086890377e+18,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}