{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 234,
  "global_step": 2340,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0004273960893257827,
      "grad_norm": 9.14111614227295,
      "learning_rate": 0.0,
      "loss": 0.21726450324058533,
      "num_input_tokens_seen": 1154,
      "step": 1,
      "train_runtime": 12.1331,
      "train_tokens_per_second": 95.111
    },
    {
      "epoch": 0.0008547921786515654,
      "grad_norm": 13.571990013122559,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.35620906949043274,
      "num_input_tokens_seen": 1926,
      "step": 2,
      "train_runtime": 14.3269,
      "train_tokens_per_second": 134.432
    },
    {
      "epoch": 0.001282188267977348,
      "grad_norm": 6.499567985534668,
      "learning_rate": 8.000000000000001e-06,
      "loss": 0.21047340333461761,
      "num_input_tokens_seen": 2966,
      "step": 3,
      "train_runtime": 16.552,
      "train_tokens_per_second": 179.193
    },
    {
      "epoch": 0.0017095843573031308,
      "grad_norm": 4.277872562408447,
      "learning_rate": 1.2e-05,
      "loss": 0.18391455709934235,
      "num_input_tokens_seen": 3778,
      "step": 4,
      "train_runtime": 18.8009,
      "train_tokens_per_second": 200.948
    },
    {
      "epoch": 0.0021369804466289132,
      "grad_norm": 5.078300476074219,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 0.17812903225421906,
      "num_input_tokens_seen": 4340,
      "step": 5,
      "train_runtime": 20.8672,
      "train_tokens_per_second": 207.982
    },
    {
      "epoch": 0.002564376535954696,
      "grad_norm": 3.261840581893921,
      "learning_rate": 2e-05,
      "loss": 0.2035450041294098,
      "num_input_tokens_seen": 5598,
      "step": 6,
      "train_runtime": 23.0631,
      "train_tokens_per_second": 242.726
    },
    {
      "epoch": 0.0029917726252804786,
      "grad_norm": 4.089930057525635,
      "learning_rate": 1.9991434689507495e-05,
      "loss": 0.11088275164365768,
      "num_input_tokens_seen": 6630,
      "step": 7,
      "train_runtime": 25.209,
      "train_tokens_per_second": 263.001
    },
    {
      "epoch": 0.0034191687146062615,
      "grad_norm": 2.839904308319092,
      "learning_rate": 1.9982869379014992e-05,
      "loss": 0.12259666621685028,
      "num_input_tokens_seen": 7824,
      "step": 8,
      "train_runtime": 27.407,
      "train_tokens_per_second": 285.474
    },
    {
      "epoch": 0.003846564803932044,
      "grad_norm": 3.0211679935455322,
      "learning_rate": 1.9974304068522486e-05,
      "loss": 0.08603208512067795,
      "num_input_tokens_seen": 8920,
      "step": 9,
      "train_runtime": 29.5794,
      "train_tokens_per_second": 301.562
    },
    {
      "epoch": 0.0042739608932578265,
      "grad_norm": 2.7547717094421387,
      "learning_rate": 1.996573875802998e-05,
      "loss": 0.1716809868812561,
      "num_input_tokens_seen": 10266,
      "step": 10,
      "train_runtime": 31.7953,
      "train_tokens_per_second": 322.878
    },
    {
      "epoch": 0.004701356982583609,
      "grad_norm": 2.4462404251098633,
      "learning_rate": 1.9957173447537473e-05,
      "loss": 0.10024883598089218,
      "num_input_tokens_seen": 11024,
      "step": 11,
      "train_runtime": 33.8434,
      "train_tokens_per_second": 325.736
    },
    {
      "epoch": 0.005128753071909392,
      "grad_norm": 2.9624831676483154,
      "learning_rate": 1.994860813704497e-05,
      "loss": 0.20874907076358795,
      "num_input_tokens_seen": 12374,
      "step": 12,
      "train_runtime": 36.0888,
      "train_tokens_per_second": 342.877
    },
    {
      "epoch": 0.005556149161235174,
      "grad_norm": 3.645580291748047,
      "learning_rate": 1.9940042826552463e-05,
      "loss": 0.17093905806541443,
      "num_input_tokens_seen": 13532,
      "step": 13,
      "train_runtime": 38.2732,
      "train_tokens_per_second": 353.564
    },
    {
      "epoch": 0.005983545250560957,
      "grad_norm": 3.387136697769165,
      "learning_rate": 1.9931477516059957e-05,
      "loss": 0.12048695981502533,
      "num_input_tokens_seen": 14356,
      "step": 14,
      "train_runtime": 40.4212,
      "train_tokens_per_second": 355.16
    },
    {
      "epoch": 0.00641094133988674,
      "grad_norm": 3.199143648147583,
      "learning_rate": 1.9922912205567454e-05,
      "loss": 0.1609961837530136,
      "num_input_tokens_seen": 15532,
      "step": 15,
      "train_runtime": 42.62,
      "train_tokens_per_second": 364.43
    },
    {
      "epoch": 0.006838337429212523,
      "grad_norm": 2.954838275909424,
      "learning_rate": 1.9914346895074948e-05,
      "loss": 0.1544017344713211,
      "num_input_tokens_seen": 16276,
      "step": 16,
      "train_runtime": 44.7305,
      "train_tokens_per_second": 363.868
    },
    {
      "epoch": 0.007265733518538305,
      "grad_norm": 3.1137399673461914,
      "learning_rate": 1.990578158458244e-05,
      "loss": 0.0799512267112732,
      "num_input_tokens_seen": 17076,
      "step": 17,
      "train_runtime": 46.8507,
      "train_tokens_per_second": 364.477
    },
    {
      "epoch": 0.007693129607864088,
      "grad_norm": 2.0236153602600098,
      "learning_rate": 1.9897216274089938e-05,
      "loss": 0.07863675057888031,
      "num_input_tokens_seen": 18224,
      "step": 18,
      "train_runtime": 49.0208,
      "train_tokens_per_second": 371.76
    },
    {
      "epoch": 0.00812052569718987,
      "grad_norm": 2.745185375213623,
      "learning_rate": 1.9888650963597432e-05,
      "loss": 0.0929172933101654,
      "num_input_tokens_seen": 19792,
      "step": 19,
      "train_runtime": 51.4059,
      "train_tokens_per_second": 385.015
    },
    {
      "epoch": 0.008547921786515653,
      "grad_norm": 2.367658853530884,
      "learning_rate": 1.9880085653104925e-05,
      "loss": 0.12210556864738464,
      "num_input_tokens_seen": 20890,
      "step": 20,
      "train_runtime": 53.5825,
      "train_tokens_per_second": 389.866
    },
    {
      "epoch": 0.008975317875841436,
      "grad_norm": 3.4798147678375244,
      "learning_rate": 1.987152034261242e-05,
      "loss": 0.13251161575317383,
      "num_input_tokens_seen": 21788,
      "step": 21,
      "train_runtime": 55.6981,
      "train_tokens_per_second": 391.18
    },
    {
      "epoch": 0.009402713965167219,
      "grad_norm": 3.735536575317383,
      "learning_rate": 1.9862955032119916e-05,
      "loss": 0.1198655366897583,
      "num_input_tokens_seen": 22586,
      "step": 22,
      "train_runtime": 57.7941,
      "train_tokens_per_second": 390.801
    },
    {
      "epoch": 0.009830110054493002,
      "grad_norm": 2.751854181289673,
      "learning_rate": 1.985438972162741e-05,
      "loss": 0.11355586349964142,
      "num_input_tokens_seen": 23474,
      "step": 23,
      "train_runtime": 59.9163,
      "train_tokens_per_second": 391.78
    },
    {
      "epoch": 0.010257506143818785,
      "grad_norm": 3.438530206680298,
      "learning_rate": 1.9845824411134906e-05,
      "loss": 0.16267995536327362,
      "num_input_tokens_seen": 24318,
      "step": 24,
      "train_runtime": 62.0331,
      "train_tokens_per_second": 392.016
    },
    {
      "epoch": 0.010684902233144567,
      "grad_norm": 2.3223185539245605,
      "learning_rate": 1.98372591006424e-05,
      "loss": 0.18063999712467194,
      "num_input_tokens_seen": 25596,
      "step": 25,
      "train_runtime": 64.282,
      "train_tokens_per_second": 398.183
    },
    {
      "epoch": 0.011112298322470349,
      "grad_norm": 2.390918493270874,
      "learning_rate": 1.9828693790149897e-05,
      "loss": 0.08224872499704361,
      "num_input_tokens_seen": 27330,
      "step": 26,
      "train_runtime": 66.7305,
      "train_tokens_per_second": 409.558
    },
    {
      "epoch": 0.011539694411796132,
      "grad_norm": 3.343850612640381,
      "learning_rate": 1.982012847965739e-05,
      "loss": 0.14506331086158752,
      "num_input_tokens_seen": 28246,
      "step": 27,
      "train_runtime": 68.8536,
      "train_tokens_per_second": 410.233
    },
    {
      "epoch": 0.011967090501121914,
      "grad_norm": 2.507873773574829,
      "learning_rate": 1.9811563169164884e-05,
      "loss": 0.0809740275144577,
      "num_input_tokens_seen": 29652,
      "step": 28,
      "train_runtime": 71.0635,
      "train_tokens_per_second": 417.261
    },
    {
      "epoch": 0.012394486590447697,
      "grad_norm": 9.32749080657959,
      "learning_rate": 1.9802997858672378e-05,
      "loss": 0.20142711699008942,
      "num_input_tokens_seen": 30714,
      "step": 29,
      "train_runtime": 73.3281,
      "train_tokens_per_second": 418.857
    },
    {
      "epoch": 0.01282188267977348,
      "grad_norm": 1.748494029045105,
      "learning_rate": 1.9794432548179875e-05,
      "loss": 0.10139460861682892,
      "num_input_tokens_seen": 32788,
      "step": 30,
      "train_runtime": 75.7604,
      "train_tokens_per_second": 432.786
    },
    {
      "epoch": 0.013249278769099263,
      "grad_norm": 2.136369466781616,
      "learning_rate": 1.9785867237687368e-05,
      "loss": 0.07564827799797058,
      "num_input_tokens_seen": 34168,
      "step": 31,
      "train_runtime": 92.5711,
      "train_tokens_per_second": 369.1
    },
    {
      "epoch": 0.013676674858425046,
      "grad_norm": 3.196265697479248,
      "learning_rate": 1.9777301927194862e-05,
      "loss": 0.12035083025693893,
      "num_input_tokens_seen": 35112,
      "step": 32,
      "train_runtime": 94.6973,
      "train_tokens_per_second": 370.781
    },
    {
      "epoch": 0.014104070947750827,
      "grad_norm": 2.6035709381103516,
      "learning_rate": 1.976873661670236e-05,
      "loss": 0.12550172209739685,
      "num_input_tokens_seen": 36280,
      "step": 33,
      "train_runtime": 96.8858,
      "train_tokens_per_second": 374.461
    },
    {
      "epoch": 0.01453146703707661,
      "grad_norm": 1.9573886394500732,
      "learning_rate": 1.9760171306209852e-05,
      "loss": 0.10325965285301208,
      "num_input_tokens_seen": 37392,
      "step": 34,
      "train_runtime": 99.0973,
      "train_tokens_per_second": 377.326
    },
    {
      "epoch": 0.014958863126402393,
      "grad_norm": 2.2740912437438965,
      "learning_rate": 1.9751605995717346e-05,
      "loss": 0.1205965206027031,
      "num_input_tokens_seen": 38426,
      "step": 35,
      "train_runtime": 101.244,
      "train_tokens_per_second": 379.539
    },
    {
      "epoch": 0.015386259215728176,
      "grad_norm": 4.021089553833008,
      "learning_rate": 1.9743040685224843e-05,
      "loss": 0.16274608671665192,
      "num_input_tokens_seen": 39142,
      "step": 36,
      "train_runtime": 103.2831,
      "train_tokens_per_second": 378.978
    },
    {
      "epoch": 0.015813655305053957,
      "grad_norm": 3.387068271636963,
      "learning_rate": 1.9734475374732337e-05,
      "loss": 0.1876910924911499,
      "num_input_tokens_seen": 40582,
      "step": 37,
      "train_runtime": 105.5614,
      "train_tokens_per_second": 384.44
    },
    {
      "epoch": 0.01624105139437974,
      "grad_norm": 2.8090980052948,
      "learning_rate": 1.972591006423983e-05,
      "loss": 0.14450424909591675,
      "num_input_tokens_seen": 41558,
      "step": 38,
      "train_runtime": 107.6691,
      "train_tokens_per_second": 385.979
    },
    {
      "epoch": 0.016668447483705523,
      "grad_norm": 3.532721519470215,
      "learning_rate": 1.9717344753747324e-05,
      "loss": 0.14360390603542328,
      "num_input_tokens_seen": 43008,
      "step": 39,
      "train_runtime": 109.9286,
      "train_tokens_per_second": 391.236
    },
    {
      "epoch": 0.017095843573031306,
      "grad_norm": 2.9382805824279785,
      "learning_rate": 1.970877944325482e-05,
      "loss": 0.14307570457458496,
      "num_input_tokens_seen": 43638,
      "step": 40,
      "train_runtime": 111.9664,
      "train_tokens_per_second": 389.742
    },
    {
      "epoch": 0.01752323966235709,
      "grad_norm": 2.8658647537231445,
      "learning_rate": 1.9700214132762314e-05,
      "loss": 0.12519389390945435,
      "num_input_tokens_seen": 44638,
      "step": 41,
      "train_runtime": 114.1116,
      "train_tokens_per_second": 391.178
    },
    {
      "epoch": 0.01795063575168287,
      "grad_norm": 1.6191413402557373,
      "learning_rate": 1.9691648822269808e-05,
      "loss": 0.056612834334373474,
      "num_input_tokens_seen": 45956,
      "step": 42,
      "train_runtime": 116.3631,
      "train_tokens_per_second": 394.936
    },
    {
      "epoch": 0.018378031841008655,
      "grad_norm": 2.833674192428589,
      "learning_rate": 1.9683083511777305e-05,
      "loss": 0.14162088930606842,
      "num_input_tokens_seen": 46938,
      "step": 43,
      "train_runtime": 118.4983,
      "train_tokens_per_second": 396.107
    },
    {
      "epoch": 0.018805427930334438,
      "grad_norm": 2.1288228034973145,
      "learning_rate": 1.96745182012848e-05,
      "loss": 0.12884970009326935,
      "num_input_tokens_seen": 48790,
      "step": 44,
      "train_runtime": 120.8259,
      "train_tokens_per_second": 403.804
    },
    {
      "epoch": 0.01923282401966022,
      "grad_norm": 3.1711273193359375,
      "learning_rate": 1.9665952890792292e-05,
      "loss": 0.14132612943649292,
      "num_input_tokens_seen": 49788,
      "step": 45,
      "train_runtime": 122.9274,
      "train_tokens_per_second": 405.02
    },
    {
      "epoch": 0.019660220108986003,
      "grad_norm": 3.3076086044311523,
      "learning_rate": 1.9657387580299786e-05,
      "loss": 0.12955069541931152,
      "num_input_tokens_seen": 50982,
      "step": 46,
      "train_runtime": 125.1368,
      "train_tokens_per_second": 407.41
    },
    {
      "epoch": 0.020087616198311786,
      "grad_norm": 4.046682357788086,
      "learning_rate": 1.9648822269807283e-05,
      "loss": 0.15046128630638123,
      "num_input_tokens_seen": 51774,
      "step": 47,
      "train_runtime": 127.2411,
      "train_tokens_per_second": 406.897
    },
    {
      "epoch": 0.02051501228763757,
      "grad_norm": 2.9671101570129395,
      "learning_rate": 1.9640256959314776e-05,
      "loss": 0.1432068794965744,
      "num_input_tokens_seen": 52456,
      "step": 48,
      "train_runtime": 129.2862,
      "train_tokens_per_second": 405.735
    },
    {
      "epoch": 0.020942408376963352,
      "grad_norm": 2.9615638256073,
      "learning_rate": 1.963169164882227e-05,
      "loss": 0.11182968318462372,
      "num_input_tokens_seen": 53518,
      "step": 49,
      "train_runtime": 131.4665,
      "train_tokens_per_second": 407.085
    },
    {
      "epoch": 0.021369804466289135,
      "grad_norm": 2.3533644676208496,
      "learning_rate": 1.9623126338329767e-05,
      "loss": 0.0767638236284256,
      "num_input_tokens_seen": 54192,
      "step": 50,
      "train_runtime": 133.5584,
      "train_tokens_per_second": 405.755
    },
    {
      "epoch": 0.021797200555614914,
      "grad_norm": 3.2768988609313965,
      "learning_rate": 1.961456102783726e-05,
      "loss": 0.1644146591424942,
      "num_input_tokens_seen": 55238,
      "step": 51,
      "train_runtime": 135.7162,
      "train_tokens_per_second": 407.011
    },
    {
      "epoch": 0.022224596644940697,
      "grad_norm": 3.0618464946746826,
      "learning_rate": 1.9605995717344754e-05,
      "loss": 0.10548470914363861,
      "num_input_tokens_seen": 55944,
      "step": 52,
      "train_runtime": 137.7924,
      "train_tokens_per_second": 406.002
    },
    {
      "epoch": 0.02265199273426648,
      "grad_norm": 1.9948461055755615,
      "learning_rate": 1.959743040685225e-05,
      "loss": 0.11291508376598358,
      "num_input_tokens_seen": 57432,
      "step": 53,
      "train_runtime": 140.1,
      "train_tokens_per_second": 409.936
    },
    {
      "epoch": 0.023079388823592263,
      "grad_norm": 3.476478099822998,
      "learning_rate": 1.9588865096359744e-05,
      "loss": 0.11228299140930176,
      "num_input_tokens_seen": 58132,
      "step": 54,
      "train_runtime": 142.1935,
      "train_tokens_per_second": 408.823
    },
    {
      "epoch": 0.023506784912918046,
      "grad_norm": 3.6434085369110107,
      "learning_rate": 1.9580299785867238e-05,
      "loss": 0.23393051326274872,
      "num_input_tokens_seen": 59050,
      "step": 55,
      "train_runtime": 144.3042,
      "train_tokens_per_second": 409.205
    },
    {
      "epoch": 0.02393418100224383,
      "grad_norm": 2.5962419509887695,
      "learning_rate": 1.957173447537473e-05,
      "loss": 0.1370583325624466,
      "num_input_tokens_seen": 60142,
      "step": 56,
      "train_runtime": 146.4428,
      "train_tokens_per_second": 410.686
    },
    {
      "epoch": 0.024361577091569612,
      "grad_norm": 3.1055679321289062,
      "learning_rate": 1.956316916488223e-05,
      "loss": 0.09925813972949982,
      "num_input_tokens_seen": 60816,
      "step": 57,
      "train_runtime": 148.5012,
      "train_tokens_per_second": 409.532
    },
    {
      "epoch": 0.024788973180895395,
      "grad_norm": 2.9782164096832275,
      "learning_rate": 1.9554603854389726e-05,
      "loss": 0.18971489369869232,
      "num_input_tokens_seen": 61912,
      "step": 58,
      "train_runtime": 150.6795,
      "train_tokens_per_second": 410.885
    },
    {
      "epoch": 0.025216369270221178,
      "grad_norm": 2.759922742843628,
      "learning_rate": 1.954603854389722e-05,
      "loss": 0.1085483729839325,
      "num_input_tokens_seen": 63102,
      "step": 59,
      "train_runtime": 152.893,
      "train_tokens_per_second": 412.72
    },
    {
      "epoch": 0.02564376535954696,
      "grad_norm": 2.0845162868499756,
      "learning_rate": 1.9537473233404713e-05,
      "loss": 0.0876690223813057,
      "num_input_tokens_seen": 64178,
      "step": 60,
      "train_runtime": 155.0612,
      "train_tokens_per_second": 413.888
    },
    {
      "epoch": 0.026071161448872743,
      "grad_norm": 2.9664435386657715,
      "learning_rate": 1.952890792291221e-05,
      "loss": 0.15988248586654663,
      "num_input_tokens_seen": 65130,
      "step": 61,
      "train_runtime": 169.7061,
      "train_tokens_per_second": 383.781
    },
    {
      "epoch": 0.026498557538198526,
      "grad_norm": 2.35333251953125,
      "learning_rate": 1.9520342612419703e-05,
      "loss": 0.1349293291568756,
      "num_input_tokens_seen": 66482,
      "step": 62,
      "train_runtime": 171.9969,
      "train_tokens_per_second": 386.53
    },
    {
      "epoch": 0.02692595362752431,
      "grad_norm": 3.8661534786224365,
      "learning_rate": 1.9511777301927197e-05,
      "loss": 0.19301962852478027,
      "num_input_tokens_seen": 67294,
      "step": 63,
      "train_runtime": 174.1011,
      "train_tokens_per_second": 386.523
    },
    {
      "epoch": 0.027353349716850092,
      "grad_norm": 3.158799648284912,
      "learning_rate": 1.950321199143469e-05,
      "loss": 0.15264831483364105,
      "num_input_tokens_seen": 68030,
      "step": 64,
      "train_runtime": 176.1776,
      "train_tokens_per_second": 386.144
    },
    {
      "epoch": 0.027780745806175875,
      "grad_norm": 2.5029361248016357,
      "learning_rate": 1.9494646680942187e-05,
      "loss": 0.11615459620952606,
      "num_input_tokens_seen": 69248,
      "step": 65,
      "train_runtime": 178.436,
      "train_tokens_per_second": 388.083
    },
    {
      "epoch": 0.028208141895501655,
      "grad_norm": 2.39737868309021,
      "learning_rate": 1.948608137044968e-05,
      "loss": 0.11769278347492218,
      "num_input_tokens_seen": 70370,
      "step": 66,
      "train_runtime": 180.632,
      "train_tokens_per_second": 389.577
    },
    {
      "epoch": 0.028635537984827437,
      "grad_norm": 2.833765745162964,
      "learning_rate": 1.9477516059957175e-05,
      "loss": 0.11099628359079361,
      "num_input_tokens_seen": 71626,
      "step": 67,
      "train_runtime": 182.8654,
      "train_tokens_per_second": 391.687
    },
    {
      "epoch": 0.02906293407415322,
      "grad_norm": 2.0754895210266113,
      "learning_rate": 1.946895074946467e-05,
      "loss": 0.10504400730133057,
      "num_input_tokens_seen": 73088,
      "step": 68,
      "train_runtime": 185.1474,
      "train_tokens_per_second": 394.756
    },
    {
      "epoch": 0.029490330163479003,
      "grad_norm": 4.698664665222168,
      "learning_rate": 1.9460385438972165e-05,
      "loss": 0.11828333139419556,
      "num_input_tokens_seen": 73794,
      "step": 69,
      "train_runtime": 187.2644,
      "train_tokens_per_second": 394.063
    },
    {
      "epoch": 0.029917726252804786,
      "grad_norm": 2.6300766468048096,
      "learning_rate": 1.945182012847966e-05,
      "loss": 0.1361745297908783,
      "num_input_tokens_seen": 74550,
      "step": 70,
      "train_runtime": 189.3593,
      "train_tokens_per_second": 393.696
    },
    {
      "epoch": 0.03034512234213057,
      "grad_norm": 2.975100517272949,
      "learning_rate": 1.9443254817987152e-05,
      "loss": 0.20098119974136353,
      "num_input_tokens_seen": 75664,
      "step": 71,
      "train_runtime": 191.5467,
      "train_tokens_per_second": 395.016
    },
    {
      "epoch": 0.030772518431456352,
      "grad_norm": 3.710899591445923,
      "learning_rate": 1.943468950749465e-05,
      "loss": 0.16764453053474426,
      "num_input_tokens_seen": 76658,
      "step": 72,
      "train_runtime": 193.7714,
      "train_tokens_per_second": 395.61
    },
    {
      "epoch": 0.031199914520782135,
      "grad_norm": 3.4315569400787354,
      "learning_rate": 1.9426124197002143e-05,
      "loss": 0.10653313994407654,
      "num_input_tokens_seen": 77340,
      "step": 73,
      "train_runtime": 195.8266,
      "train_tokens_per_second": 394.941
    },
    {
      "epoch": 0.031627310610107914,
      "grad_norm": 3.3026206493377686,
      "learning_rate": 1.9417558886509636e-05,
      "loss": 0.21571186184883118,
      "num_input_tokens_seen": 78280,
      "step": 74,
      "train_runtime": 197.9437,
      "train_tokens_per_second": 395.466
    },
    {
      "epoch": 0.0320547066994337,
      "grad_norm": 4.470591068267822,
      "learning_rate": 1.9408993576017133e-05,
      "loss": 0.15138877928256989,
      "num_input_tokens_seen": 79984,
      "step": 75,
      "train_runtime": 200.2775,
      "train_tokens_per_second": 399.366
    },
    {
      "epoch": 0.03248210278875948,
      "grad_norm": 2.452955484390259,
      "learning_rate": 1.9400428265524627e-05,
      "loss": 0.1523466855287552,
      "num_input_tokens_seen": 81050,
      "step": 76,
      "train_runtime": 202.4585,
      "train_tokens_per_second": 400.329
    },
    {
      "epoch": 0.032909498878085267,
      "grad_norm": 2.775866985321045,
      "learning_rate": 1.939186295503212e-05,
      "loss": 0.16505104303359985,
      "num_input_tokens_seen": 82148,
      "step": 77,
      "train_runtime": 204.6439,
      "train_tokens_per_second": 401.419
    },
    {
      "epoch": 0.033336894967411046,
      "grad_norm": 2.287738084793091,
      "learning_rate": 1.9383297644539617e-05,
      "loss": 0.14218983054161072,
      "num_input_tokens_seen": 83480,
      "step": 78,
      "train_runtime": 206.8838,
      "train_tokens_per_second": 403.512
    },
    {
      "epoch": 0.03376429105673683,
      "grad_norm": 5.012575626373291,
      "learning_rate": 1.937473233404711e-05,
      "loss": 0.19431844353675842,
      "num_input_tokens_seen": 84510,
      "step": 79,
      "train_runtime": 209.0563,
      "train_tokens_per_second": 404.245
    },
    {
      "epoch": 0.03419168714606261,
      "grad_norm": 2.3963444232940674,
      "learning_rate": 1.9366167023554605e-05,
      "loss": 0.09664961695671082,
      "num_input_tokens_seen": 85856,
      "step": 80,
      "train_runtime": 211.2962,
      "train_tokens_per_second": 406.33
    },
    {
      "epoch": 0.0346190832353884,
      "grad_norm": 2.7534611225128174,
      "learning_rate": 1.9357601713062098e-05,
      "loss": 0.10577772557735443,
      "num_input_tokens_seen": 86658,
      "step": 81,
      "train_runtime": 213.3725,
      "train_tokens_per_second": 406.135
    },
    {
      "epoch": 0.03504647932471418,
      "grad_norm": 1.8196276426315308,
      "learning_rate": 1.9349036402569595e-05,
      "loss": 0.12409847229719162,
      "num_input_tokens_seen": 88316,
      "step": 82,
      "train_runtime": 215.6855,
      "train_tokens_per_second": 409.467
    },
    {
      "epoch": 0.035473875414039964,
      "grad_norm": 2.1968817710876465,
      "learning_rate": 1.934047109207709e-05,
      "loss": 0.09177117049694061,
      "num_input_tokens_seen": 89176,
      "step": 83,
      "train_runtime": 217.8148,
      "train_tokens_per_second": 409.412
    },
    {
      "epoch": 0.03590127150336574,
      "grad_norm": 3.3404345512390137,
      "learning_rate": 1.9331905781584582e-05,
      "loss": 0.14738020300865173,
      "num_input_tokens_seen": 90190,
      "step": 84,
      "train_runtime": 219.9637,
      "train_tokens_per_second": 410.022
    },
    {
      "epoch": 0.03632866759269153,
      "grad_norm": 1.7804222106933594,
      "learning_rate": 1.932334047109208e-05,
      "loss": 0.10943267494440079,
      "num_input_tokens_seen": 91646,
      "step": 85,
      "train_runtime": 222.2124,
      "train_tokens_per_second": 412.425
    },
    {
      "epoch": 0.03675606368201731,
      "grad_norm": 4.095219612121582,
      "learning_rate": 1.9314775160599573e-05,
      "loss": 0.1152762919664383,
      "num_input_tokens_seen": 92540,
      "step": 86,
      "train_runtime": 224.3367,
      "train_tokens_per_second": 412.505
    },
    {
      "epoch": 0.037183459771343096,
      "grad_norm": 2.9774389266967773,
      "learning_rate": 1.9306209850107067e-05,
      "loss": 0.15725484490394592,
      "num_input_tokens_seen": 93406,
      "step": 87,
      "train_runtime": 226.4236,
      "train_tokens_per_second": 412.528
    },
    {
      "epoch": 0.037610855860668875,
      "grad_norm": 2.6223371028900146,
      "learning_rate": 1.929764453961456e-05,
      "loss": 0.08795279264450073,
      "num_input_tokens_seen": 94042,
      "step": 88,
      "train_runtime": 228.4803,
      "train_tokens_per_second": 411.598
    },
    {
      "epoch": 0.038038251949994654,
      "grad_norm": 2.9591426849365234,
      "learning_rate": 1.9289079229122057e-05,
      "loss": 0.14193153381347656,
      "num_input_tokens_seen": 94768,
      "step": 89,
      "train_runtime": 230.5694,
      "train_tokens_per_second": 411.017
    },
    {
      "epoch": 0.03846564803932044,
      "grad_norm": 3.3232381343841553,
      "learning_rate": 1.928051391862955e-05,
      "loss": 0.13330227136611938,
      "num_input_tokens_seen": 95510,
      "step": 90,
      "train_runtime": 232.665,
      "train_tokens_per_second": 410.504
    },
    {
      "epoch": 0.03889304412864622,
      "grad_norm": 2.6954174041748047,
      "learning_rate": 1.9271948608137044e-05,
      "loss": 0.14643427729606628,
      "num_input_tokens_seen": 96404,
      "step": 91,
      "train_runtime": 247.5547,
      "train_tokens_per_second": 389.425
    },
    {
      "epoch": 0.03932044021797201,
      "grad_norm": 3.4643115997314453,
      "learning_rate": 1.926338329764454e-05,
      "loss": 0.262195348739624,
      "num_input_tokens_seen": 97326,
      "step": 92,
      "train_runtime": 249.6645,
      "train_tokens_per_second": 389.827
    },
    {
      "epoch": 0.039747836307297786,
      "grad_norm": 2.8510966300964355,
      "learning_rate": 1.9254817987152038e-05,
      "loss": 0.11224470287561417,
      "num_input_tokens_seen": 98670,
      "step": 93,
      "train_runtime": 251.8743,
      "train_tokens_per_second": 391.743
    },
    {
      "epoch": 0.04017523239662357,
      "grad_norm": 2.7037439346313477,
      "learning_rate": 1.9246252676659532e-05,
      "loss": 0.14719970524311066,
      "num_input_tokens_seen": 99802,
      "step": 94,
      "train_runtime": 254.0152,
      "train_tokens_per_second": 392.898
    },
    {
      "epoch": 0.04060262848594935,
      "grad_norm": 2.3708066940307617,
      "learning_rate": 1.9237687366167025e-05,
      "loss": 0.11163226515054703,
      "num_input_tokens_seen": 100602,
      "step": 95,
      "train_runtime": 256.0744,
      "train_tokens_per_second": 392.862
    },
    {
      "epoch": 0.04103002457527514,
      "grad_norm": 3.19515323638916,
      "learning_rate": 1.9229122055674522e-05,
      "loss": 0.17545676231384277,
      "num_input_tokens_seen": 101492,
      "step": 96,
      "train_runtime": 258.2043,
      "train_tokens_per_second": 393.069
    },
    {
      "epoch": 0.04145742066460092,
      "grad_norm": 1.8117717504501343,
      "learning_rate": 1.9220556745182016e-05,
      "loss": 0.10086990892887115,
      "num_input_tokens_seen": 102878,
      "step": 97,
      "train_runtime": 260.4581,
      "train_tokens_per_second": 394.989
    },
    {
      "epoch": 0.041884816753926704,
      "grad_norm": 3.0650646686553955,
      "learning_rate": 1.921199143468951e-05,
      "loss": 0.1313072144985199,
      "num_input_tokens_seen": 103492,
      "step": 98,
      "train_runtime": 262.5091,
      "train_tokens_per_second": 394.242
    },
    {
      "epoch": 0.042312212843252484,
      "grad_norm": 3.3805437088012695,
      "learning_rate": 1.9203426124197003e-05,
      "loss": 0.1240580677986145,
      "num_input_tokens_seen": 104662,
      "step": 99,
      "train_runtime": 264.6925,
      "train_tokens_per_second": 395.41
    },
    {
      "epoch": 0.04273960893257827,
      "grad_norm": 3.6072335243225098,
      "learning_rate": 1.91948608137045e-05,
      "loss": 0.18898653984069824,
      "num_input_tokens_seen": 105546,
      "step": 100,
      "train_runtime": 266.8165,
      "train_tokens_per_second": 395.575
    },
    {
      "epoch": 0.04316700502190405,
      "grad_norm": 2.737124443054199,
      "learning_rate": 1.9186295503211994e-05,
      "loss": 0.11584174633026123,
      "num_input_tokens_seen": 106606,
      "step": 101,
      "train_runtime": 269.1023,
      "train_tokens_per_second": 396.154
    },
    {
      "epoch": 0.04359440111122983,
      "grad_norm": 4.503042697906494,
      "learning_rate": 1.9177730192719487e-05,
      "loss": 0.08784925192594528,
      "num_input_tokens_seen": 107334,
      "step": 102,
      "train_runtime": 271.3214,
      "train_tokens_per_second": 395.597
    },
    {
      "epoch": 0.044021797200555615,
      "grad_norm": 2.8012304306030273,
      "learning_rate": 1.9169164882226984e-05,
      "loss": 0.14326158165931702,
      "num_input_tokens_seen": 108792,
      "step": 103,
      "train_runtime": 273.5492,
      "train_tokens_per_second": 397.705
    },
    {
      "epoch": 0.044449193289881395,
      "grad_norm": 3.640925168991089,
      "learning_rate": 1.9160599571734478e-05,
      "loss": 0.12522439658641815,
      "num_input_tokens_seen": 109786,
      "step": 104,
      "train_runtime": 275.6638,
      "train_tokens_per_second": 398.26
    },
    {
      "epoch": 0.04487658937920718,
      "grad_norm": 2.4012043476104736,
      "learning_rate": 1.915203426124197e-05,
      "loss": 0.14218813180923462,
      "num_input_tokens_seen": 111036,
      "step": 105,
      "train_runtime": 277.8666,
      "train_tokens_per_second": 399.602
    },
    {
      "epoch": 0.04530398546853296,
      "grad_norm": 2.4177777767181396,
      "learning_rate": 1.9143468950749465e-05,
      "loss": 0.09716714918613434,
      "num_input_tokens_seen": 112130,
      "step": 106,
      "train_runtime": 280.0427,
      "train_tokens_per_second": 400.403
    },
    {
      "epoch": 0.04573138155785875,
      "grad_norm": 3.3823659420013428,
      "learning_rate": 1.9134903640256962e-05,
      "loss": 0.20361191034317017,
      "num_input_tokens_seen": 112996,
      "step": 107,
      "train_runtime": 282.1522,
      "train_tokens_per_second": 400.479
    },
    {
      "epoch": 0.046158777647184526,
      "grad_norm": 2.135305881500244,
      "learning_rate": 1.9126338329764455e-05,
      "loss": 0.09638367593288422,
      "num_input_tokens_seen": 113962,
      "step": 108,
      "train_runtime": 284.2948,
      "train_tokens_per_second": 400.859
    },
    {
      "epoch": 0.04658617373651031,
      "grad_norm": 2.566462516784668,
      "learning_rate": 1.911777301927195e-05,
      "loss": 0.16817453503608704,
      "num_input_tokens_seen": 115666,
      "step": 109,
      "train_runtime": 286.5897,
      "train_tokens_per_second": 403.594
    },
    {
      "epoch": 0.04701356982583609,
      "grad_norm": 2.8588247299194336,
      "learning_rate": 1.9109207708779446e-05,
      "loss": 0.1660233587026596,
      "num_input_tokens_seen": 116624,
      "step": 110,
      "train_runtime": 288.7227,
      "train_tokens_per_second": 403.931
    },
    {
      "epoch": 0.04744096591516188,
      "grad_norm": 2.2879586219787598,
      "learning_rate": 1.910064239828694e-05,
      "loss": 0.09440014511346817,
      "num_input_tokens_seen": 117428,
      "step": 111,
      "train_runtime": 290.8628,
      "train_tokens_per_second": 403.723
    },
    {
      "epoch": 0.04786836200448766,
      "grad_norm": 2.0961356163024902,
      "learning_rate": 1.9092077087794433e-05,
      "loss": 0.10546403378248215,
      "num_input_tokens_seen": 118680,
      "step": 112,
      "train_runtime": 293.102,
      "train_tokens_per_second": 404.91
    },
    {
      "epoch": 0.048295758093813444,
      "grad_norm": 2.0368144512176514,
      "learning_rate": 1.908351177730193e-05,
      "loss": 0.14511381089687347,
      "num_input_tokens_seen": 119692,
      "step": 113,
      "train_runtime": 295.251,
      "train_tokens_per_second": 405.391
    },
    {
      "epoch": 0.048723154183139224,
      "grad_norm": 2.633056640625,
      "learning_rate": 1.9074946466809424e-05,
      "loss": 0.12467923015356064,
      "num_input_tokens_seen": 120922,
      "step": 114,
      "train_runtime": 297.4186,
      "train_tokens_per_second": 406.572
    },
    {
      "epoch": 0.04915055027246501,
      "grad_norm": 3.234384536743164,
      "learning_rate": 1.9066381156316917e-05,
      "loss": 0.13998226821422577,
      "num_input_tokens_seen": 121918,
      "step": 115,
      "train_runtime": 299.5902,
      "train_tokens_per_second": 406.949
    },
    {
      "epoch": 0.04957794636179079,
      "grad_norm": 2.9627978801727295,
      "learning_rate": 1.905781584582441e-05,
      "loss": 0.2414434254169464,
      "num_input_tokens_seen": 122906,
      "step": 116,
      "train_runtime": 301.7292,
      "train_tokens_per_second": 407.339
    },
    {
      "epoch": 0.05000534245111657,
      "grad_norm": 2.544501781463623,
      "learning_rate": 1.9049250535331908e-05,
      "loss": 0.10913316160440445,
      "num_input_tokens_seen": 123910,
      "step": 117,
      "train_runtime": 303.8362,
      "train_tokens_per_second": 407.818
    },
    {
      "epoch": 0.050432738540442355,
      "grad_norm": 1.9286752939224243,
      "learning_rate": 1.90406852248394e-05,
      "loss": 0.12419462203979492,
      "num_input_tokens_seen": 125102,
      "step": 118,
      "train_runtime": 306.0863,
      "train_tokens_per_second": 408.715
    },
    {
      "epoch": 0.050860134629768135,
      "grad_norm": 3.931507110595703,
      "learning_rate": 1.9032119914346895e-05,
      "loss": 0.2112111747264862,
      "num_input_tokens_seen": 126132,
      "step": 119,
      "train_runtime": 308.1982,
      "train_tokens_per_second": 409.256
    },
    {
      "epoch": 0.05128753071909392,
      "grad_norm": 4.836315631866455,
      "learning_rate": 1.9023554603854392e-05,
      "loss": 0.16022446751594543,
      "num_input_tokens_seen": 127110,
      "step": 120,
      "train_runtime": 310.336,
      "train_tokens_per_second": 409.588
    },
    {
      "epoch": 0.0517149268084197,
      "grad_norm": 2.537309408187866,
      "learning_rate": 1.9014989293361886e-05,
      "loss": 0.14001011848449707,
      "num_input_tokens_seen": 127798,
      "step": 121,
      "train_runtime": 325.1648,
      "train_tokens_per_second": 393.025
    },
    {
      "epoch": 0.05214232289774549,
      "grad_norm": 3.5574090480804443,
      "learning_rate": 1.900642398286938e-05,
      "loss": 0.23435810208320618,
      "num_input_tokens_seen": 128626,
      "step": 122,
      "train_runtime": 327.3093,
      "train_tokens_per_second": 392.98
    },
    {
      "epoch": 0.052569718987071266,
      "grad_norm": 2.6286706924438477,
      "learning_rate": 1.8997858672376873e-05,
      "loss": 0.1562124788761139,
      "num_input_tokens_seen": 129792,
      "step": 123,
      "train_runtime": 329.6428,
      "train_tokens_per_second": 393.735
    },
    {
      "epoch": 0.05299711507639705,
      "grad_norm": 3.146498680114746,
      "learning_rate": 1.898929336188437e-05,
      "loss": 0.17668110132217407,
      "num_input_tokens_seen": 131134,
      "step": 124,
      "train_runtime": 331.8446,
      "train_tokens_per_second": 395.167
    },
    {
      "epoch": 0.05342451116572283,
      "grad_norm": 3.4509334564208984,
      "learning_rate": 1.8980728051391863e-05,
      "loss": 0.17292949557304382,
      "num_input_tokens_seen": 132050,
      "step": 125,
      "train_runtime": 333.9527,
      "train_tokens_per_second": 395.415
    },
    {
      "epoch": 0.05385190725504862,
      "grad_norm": 2.534871816635132,
      "learning_rate": 1.8972162740899357e-05,
      "loss": 0.10803905129432678,
      "num_input_tokens_seen": 133038,
      "step": 126,
      "train_runtime": 336.1115,
      "train_tokens_per_second": 395.815
    },
    {
      "epoch": 0.0542793033443744,
      "grad_norm": 3.044832468032837,
      "learning_rate": 1.8963597430406854e-05,
      "loss": 0.2335970103740692,
      "num_input_tokens_seen": 134130,
      "step": 127,
      "train_runtime": 338.3016,
      "train_tokens_per_second": 396.481
    },
    {
      "epoch": 0.054706699433700184,
      "grad_norm": 2.9310450553894043,
      "learning_rate": 1.895503211991435e-05,
      "loss": 0.16756904125213623,
      "num_input_tokens_seen": 134706,
      "step": 128,
      "train_runtime": 340.3321,
      "train_tokens_per_second": 395.807
    },
    {
      "epoch": 0.055134095523025964,
      "grad_norm": 2.2983131408691406,
      "learning_rate": 1.8946466809421844e-05,
      "loss": 0.13437138497829437,
      "num_input_tokens_seen": 135978,
      "step": 129,
      "train_runtime": 342.5484,
      "train_tokens_per_second": 396.96
    },
    {
      "epoch": 0.05556149161235175,
      "grad_norm": 2.626601219177246,
      "learning_rate": 1.8937901498929338e-05,
      "loss": 0.12764504551887512,
      "num_input_tokens_seen": 137126,
      "step": 130,
      "train_runtime": 344.7782,
      "train_tokens_per_second": 397.722
    },
    {
      "epoch": 0.05598888770167753,
      "grad_norm": 2.019998073577881,
      "learning_rate": 1.892933618843683e-05,
      "loss": 0.13619019091129303,
      "num_input_tokens_seen": 138358,
      "step": 131,
      "train_runtime": 347.0358,
      "train_tokens_per_second": 398.685
    },
    {
      "epoch": 0.05641628379100331,
      "grad_norm": 2.5856571197509766,
      "learning_rate": 1.892077087794433e-05,
      "loss": 0.15411287546157837,
      "num_input_tokens_seen": 139642,
      "step": 132,
      "train_runtime": 349.283,
      "train_tokens_per_second": 399.796
    },
    {
      "epoch": 0.056843679880329095,
      "grad_norm": 3.2656030654907227,
      "learning_rate": 1.8912205567451822e-05,
      "loss": 0.15369434654712677,
      "num_input_tokens_seen": 140602,
      "step": 133,
      "train_runtime": 351.3998,
      "train_tokens_per_second": 400.12
    },
    {
      "epoch": 0.057271075969654875,
      "grad_norm": 2.674243688583374,
      "learning_rate": 1.8903640256959316e-05,
      "loss": 0.14965379238128662,
      "num_input_tokens_seen": 141428,
      "step": 134,
      "train_runtime": 353.5227,
      "train_tokens_per_second": 400.053
    },
    {
      "epoch": 0.05769847205898066,
      "grad_norm": 2.88669753074646,
      "learning_rate": 1.8895074946466813e-05,
      "loss": 0.209254652261734,
      "num_input_tokens_seen": 142422,
      "step": 135,
      "train_runtime": 355.6628,
      "train_tokens_per_second": 400.441
    },
    {
      "epoch": 0.05812586814830644,
      "grad_norm": 2.648721933364868,
      "learning_rate": 1.8886509635974306e-05,
      "loss": 0.1788233518600464,
      "num_input_tokens_seen": 143346,
      "step": 136,
      "train_runtime": 357.7754,
      "train_tokens_per_second": 400.659
    },
    {
      "epoch": 0.05855326423763223,
      "grad_norm": 3.3400492668151855,
      "learning_rate": 1.88779443254818e-05,
      "loss": 0.12309592962265015,
      "num_input_tokens_seen": 144322,
      "step": 137,
      "train_runtime": 359.9237,
      "train_tokens_per_second": 400.979
    },
    {
      "epoch": 0.05898066032695801,
      "grad_norm": 2.2111668586730957,
      "learning_rate": 1.8869379014989297e-05,
      "loss": 0.14022229611873627,
      "num_input_tokens_seen": 145602,
      "step": 138,
      "train_runtime": 362.1712,
      "train_tokens_per_second": 402.025
    },
    {
      "epoch": 0.05940805641628379,
      "grad_norm": 2.625178098678589,
      "learning_rate": 1.886081370449679e-05,
      "loss": 0.13012517988681793,
      "num_input_tokens_seen": 146480,
      "step": 139,
      "train_runtime": 364.2828,
      "train_tokens_per_second": 402.105
    },
    {
      "epoch": 0.05983545250560957,
      "grad_norm": 2.639256715774536,
      "learning_rate": 1.8852248394004284e-05,
      "loss": 0.12265259772539139,
      "num_input_tokens_seen": 147280,
      "step": 140,
      "train_runtime": 366.3331,
      "train_tokens_per_second": 402.038
    },
    {
      "epoch": 0.06026284859493536,
      "grad_norm": 1.7917380332946777,
      "learning_rate": 1.8843683083511778e-05,
      "loss": 0.07711537182331085,
      "num_input_tokens_seen": 148260,
      "step": 141,
      "train_runtime": 368.4477,
      "train_tokens_per_second": 402.391
    },
    {
      "epoch": 0.06069024468426114,
      "grad_norm": 2.7852940559387207,
      "learning_rate": 1.8835117773019275e-05,
      "loss": 0.19466297328472137,
      "num_input_tokens_seen": 149334,
      "step": 142,
      "train_runtime": 370.5824,
      "train_tokens_per_second": 402.971
    },
    {
      "epoch": 0.061117640773586925,
      "grad_norm": 3.0726804733276367,
      "learning_rate": 1.8826552462526768e-05,
      "loss": 0.1678939312696457,
      "num_input_tokens_seen": 150540,
      "step": 143,
      "train_runtime": 372.7552,
      "train_tokens_per_second": 403.857
    },
    {
      "epoch": 0.061545036862912704,
      "grad_norm": 3.7869203090667725,
      "learning_rate": 1.8817987152034262e-05,
      "loss": 0.11771175265312195,
      "num_input_tokens_seen": 151166,
      "step": 144,
      "train_runtime": 374.8031,
      "train_tokens_per_second": 403.321
    },
    {
      "epoch": 0.06197243295223849,
      "grad_norm": 1.8861055374145508,
      "learning_rate": 1.880942184154176e-05,
      "loss": 0.12411846965551376,
      "num_input_tokens_seen": 152380,
      "step": 145,
      "train_runtime": 377.0193,
      "train_tokens_per_second": 404.17
    },
    {
      "epoch": 0.06239982904156427,
      "grad_norm": 3.237966299057007,
      "learning_rate": 1.8800856531049252e-05,
      "loss": 0.22751623392105103,
      "num_input_tokens_seen": 153262,
      "step": 146,
      "train_runtime": 379.1431,
      "train_tokens_per_second": 404.233
    },
    {
      "epoch": 0.06282722513089005,
      "grad_norm": 3.3436999320983887,
      "learning_rate": 1.8792291220556746e-05,
      "loss": 0.2286110818386078,
      "num_input_tokens_seen": 154144,
      "step": 147,
      "train_runtime": 381.256,
      "train_tokens_per_second": 404.306
    },
    {
      "epoch": 0.06325462122021583,
      "grad_norm": 3.9338934421539307,
      "learning_rate": 1.878372591006424e-05,
      "loss": 0.21310165524482727,
      "num_input_tokens_seen": 155234,
      "step": 148,
      "train_runtime": 383.3973,
      "train_tokens_per_second": 404.891
    },
    {
      "epoch": 0.06368201730954162,
      "grad_norm": 2.587639570236206,
      "learning_rate": 1.8775160599571736e-05,
      "loss": 0.12459132075309753,
      "num_input_tokens_seen": 156324,
      "step": 149,
      "train_runtime": 385.57,
      "train_tokens_per_second": 405.436
    },
    {
      "epoch": 0.0641094133988674,
      "grad_norm": 2.183586835861206,
      "learning_rate": 1.876659528907923e-05,
      "loss": 0.1108795702457428,
      "num_input_tokens_seen": 157768,
      "step": 150,
      "train_runtime": 387.8474,
      "train_tokens_per_second": 406.779
    },
    {
      "epoch": 0.06453680948819318,
      "grad_norm": 3.4277262687683105,
      "learning_rate": 1.8758029978586724e-05,
      "loss": 0.14123637974262238,
      "num_input_tokens_seen": 158492,
      "step": 151,
      "train_runtime": 402.7861,
      "train_tokens_per_second": 393.489
    },
    {
      "epoch": 0.06496420557751896,
      "grad_norm": 2.3510217666625977,
      "learning_rate": 1.874946466809422e-05,
      "loss": 0.12269590049982071,
      "num_input_tokens_seen": 159674,
      "step": 152,
      "train_runtime": 404.9707,
      "train_tokens_per_second": 394.285
    },
    {
      "epoch": 0.06539160166684475,
      "grad_norm": 4.690629005432129,
      "learning_rate": 1.8740899357601714e-05,
      "loss": 0.19219471514225006,
      "num_input_tokens_seen": 160804,
      "step": 153,
      "train_runtime": 407.1537,
      "train_tokens_per_second": 394.947
    },
    {
      "epoch": 0.06581899775617053,
      "grad_norm": 2.6074059009552,
      "learning_rate": 1.8732334047109208e-05,
      "loss": 0.11395706981420517,
      "num_input_tokens_seen": 161480,
      "step": 154,
      "train_runtime": 409.2642,
      "train_tokens_per_second": 394.562
    },
    {
      "epoch": 0.06624639384549631,
      "grad_norm": 2.2025363445281982,
      "learning_rate": 1.8723768736616705e-05,
      "loss": 0.13544486463069916,
      "num_input_tokens_seen": 162474,
      "step": 155,
      "train_runtime": 411.4637,
      "train_tokens_per_second": 394.868
    },
    {
      "epoch": 0.06667378993482209,
      "grad_norm": 2.196251153945923,
      "learning_rate": 1.8715203426124198e-05,
      "loss": 0.12494926899671555,
      "num_input_tokens_seen": 163560,
      "step": 156,
      "train_runtime": 413.7073,
      "train_tokens_per_second": 395.352
    },
    {
      "epoch": 0.06710118602414789,
      "grad_norm": 2.108823776245117,
      "learning_rate": 1.8706638115631692e-05,
      "loss": 0.1273869276046753,
      "num_input_tokens_seen": 164948,
      "step": 157,
      "train_runtime": 415.9336,
      "train_tokens_per_second": 396.573
    },
    {
      "epoch": 0.06752858211347366,
      "grad_norm": 3.1646013259887695,
      "learning_rate": 1.8698072805139185e-05,
      "loss": 0.16859130561351776,
      "num_input_tokens_seen": 166248,
      "step": 158,
      "train_runtime": 418.1513,
      "train_tokens_per_second": 397.579
    },
    {
      "epoch": 0.06795597820279944,
      "grad_norm": 2.6576366424560547,
      "learning_rate": 1.8689507494646682e-05,
      "loss": 0.15689724683761597,
      "num_input_tokens_seen": 167720,
      "step": 159,
      "train_runtime": 420.3928,
      "train_tokens_per_second": 398.96
    },
    {
      "epoch": 0.06838337429212522,
      "grad_norm": 2.3928253650665283,
      "learning_rate": 1.8680942184154176e-05,
      "loss": 0.12165988236665726,
      "num_input_tokens_seen": 168542,
      "step": 160,
      "train_runtime": 422.6093,
      "train_tokens_per_second": 398.813
    },
    {
      "epoch": 0.068810770381451,
      "grad_norm": 2.780294418334961,
      "learning_rate": 1.8672376873661673e-05,
      "loss": 0.16876442730426788,
      "num_input_tokens_seen": 169602,
      "step": 161,
      "train_runtime": 424.7908,
      "train_tokens_per_second": 399.26
    },
    {
      "epoch": 0.0692381664707768,
      "grad_norm": 2.7053773403167725,
      "learning_rate": 1.8663811563169167e-05,
      "loss": 0.18703073263168335,
      "num_input_tokens_seen": 170802,
      "step": 162,
      "train_runtime": 427.0142,
      "train_tokens_per_second": 399.991
    },
    {
      "epoch": 0.06966556256010258,
      "grad_norm": 3.031386375427246,
      "learning_rate": 1.8655246252676663e-05,
      "loss": 0.16896308958530426,
      "num_input_tokens_seen": 171650,
      "step": 163,
      "train_runtime": 429.2249,
      "train_tokens_per_second": 399.907
    },
    {
      "epoch": 0.07009295864942836,
      "grad_norm": 4.824985980987549,
      "learning_rate": 1.8646680942184157e-05,
      "loss": 0.136264830827713,
      "num_input_tokens_seen": 172970,
      "step": 164,
      "train_runtime": 431.4317,
      "train_tokens_per_second": 400.921
    },
    {
      "epoch": 0.07052035473875413,
      "grad_norm": 2.955947160720825,
      "learning_rate": 1.863811563169165e-05,
      "loss": 0.11156176030635834,
      "num_input_tokens_seen": 173750,
      "step": 165,
      "train_runtime": 433.6023,
      "train_tokens_per_second": 400.713
    },
    {
      "epoch": 0.07094775082807993,
      "grad_norm": 2.719355821609497,
      "learning_rate": 1.8629550321199144e-05,
      "loss": 0.10591128468513489,
      "num_input_tokens_seen": 174504,
      "step": 166,
      "train_runtime": 435.6779,
      "train_tokens_per_second": 400.534
    },
    {
      "epoch": 0.07137514691740571,
      "grad_norm": 2.1519153118133545,
      "learning_rate": 1.862098501070664e-05,
      "loss": 0.15222682058811188,
      "num_input_tokens_seen": 176096,
      "step": 167,
      "train_runtime": 437.9449,
      "train_tokens_per_second": 402.096
    },
    {
      "epoch": 0.07180254300673149,
      "grad_norm": 2.770578384399414,
      "learning_rate": 1.8612419700214135e-05,
      "loss": 0.06951254606246948,
      "num_input_tokens_seen": 176826,
      "step": 168,
      "train_runtime": 440.0755,
      "train_tokens_per_second": 401.808
    },
    {
      "epoch": 0.07222993909605727,
      "grad_norm": 4.301676273345947,
      "learning_rate": 1.860385438972163e-05,
      "loss": 0.16717997193336487,
      "num_input_tokens_seen": 177404,
      "step": 169,
      "train_runtime": 442.1217,
      "train_tokens_per_second": 401.256
    },
    {
      "epoch": 0.07265733518538306,
      "grad_norm": 2.3800697326660156,
      "learning_rate": 1.8595289079229125e-05,
      "loss": 0.16987770795822144,
      "num_input_tokens_seen": 179500,
      "step": 170,
      "train_runtime": 444.5247,
      "train_tokens_per_second": 403.802
    },
    {
      "epoch": 0.07308473127470884,
      "grad_norm": 2.3078956604003906,
      "learning_rate": 1.858672376873662e-05,
      "loss": 0.13644614815711975,
      "num_input_tokens_seen": 180552,
      "step": 171,
      "train_runtime": 446.6741,
      "train_tokens_per_second": 404.214
    },
    {
      "epoch": 0.07351212736403462,
      "grad_norm": 2.790173292160034,
      "learning_rate": 1.8578158458244112e-05,
      "loss": 0.13570502400398254,
      "num_input_tokens_seen": 181674,
      "step": 172,
      "train_runtime": 448.8487,
      "train_tokens_per_second": 404.756
    },
    {
      "epoch": 0.0739395234533604,
      "grad_norm": 1.8088078498840332,
      "learning_rate": 1.856959314775161e-05,
      "loss": 0.16932730376720428,
      "num_input_tokens_seen": 183070,
      "step": 173,
      "train_runtime": 451.0931,
      "train_tokens_per_second": 405.836
    },
    {
      "epoch": 0.07436691954268619,
      "grad_norm": 1.8960585594177246,
      "learning_rate": 1.8561027837259103e-05,
      "loss": 0.07350532710552216,
      "num_input_tokens_seen": 184102,
      "step": 174,
      "train_runtime": 453.2641,
      "train_tokens_per_second": 406.169
    },
    {
      "epoch": 0.07479431563201197,
      "grad_norm": 2.1732749938964844,
      "learning_rate": 1.8552462526766597e-05,
      "loss": 0.08764903247356415,
      "num_input_tokens_seen": 184738,
      "step": 175,
      "train_runtime": 455.408,
      "train_tokens_per_second": 405.654
    },
    {
      "epoch": 0.07522171172133775,
      "grad_norm": 2.697097063064575,
      "learning_rate": 1.854389721627409e-05,
      "loss": 0.14000815153121948,
      "num_input_tokens_seen": 185658,
      "step": 176,
      "train_runtime": 457.604,
      "train_tokens_per_second": 405.718
    },
    {
      "epoch": 0.07564910781066353,
      "grad_norm": 3.2530503273010254,
      "learning_rate": 1.8535331905781587e-05,
      "loss": 0.18933287262916565,
      "num_input_tokens_seen": 186548,
      "step": 177,
      "train_runtime": 459.7258,
      "train_tokens_per_second": 405.781
    },
    {
      "epoch": 0.07607650389998931,
      "grad_norm": 2.815422773361206,
      "learning_rate": 1.852676659528908e-05,
      "loss": 0.12568920850753784,
      "num_input_tokens_seen": 187216,
      "step": 178,
      "train_runtime": 461.8382,
      "train_tokens_per_second": 405.371
    },
    {
      "epoch": 0.0765038999893151,
      "grad_norm": 3.2178752422332764,
      "learning_rate": 1.8518201284796574e-05,
      "loss": 0.12453125417232513,
      "num_input_tokens_seen": 187852,
      "step": 179,
      "train_runtime": 463.8893,
      "train_tokens_per_second": 404.95
    },
    {
      "epoch": 0.07693129607864088,
      "grad_norm": 2.459148406982422,
      "learning_rate": 1.850963597430407e-05,
      "loss": 0.08641057461500168,
      "num_input_tokens_seen": 188776,
      "step": 180,
      "train_runtime": 466.0309,
      "train_tokens_per_second": 405.072
    },
    {
      "epoch": 0.07735869216796666,
      "grad_norm": 2.2573819160461426,
      "learning_rate": 1.8501070663811565e-05,
      "loss": 0.08526496589183807,
      "num_input_tokens_seen": 189660,
      "step": 181,
      "train_runtime": 480.9503,
      "train_tokens_per_second": 394.344
    },
    {
      "epoch": 0.07778608825729244,
      "grad_norm": 2.2413618564605713,
      "learning_rate": 1.849250535331906e-05,
      "loss": 0.14132985472679138,
      "num_input_tokens_seen": 190660,
      "step": 182,
      "train_runtime": 483.2366,
      "train_tokens_per_second": 394.548
    },
    {
      "epoch": 0.07821348434661823,
      "grad_norm": 3.891188383102417,
      "learning_rate": 1.8483940042826552e-05,
      "loss": 0.16914093494415283,
      "num_input_tokens_seen": 191388,
      "step": 183,
      "train_runtime": 485.3233,
      "train_tokens_per_second": 394.352
    },
    {
      "epoch": 0.07864088043594401,
      "grad_norm": 2.1474502086639404,
      "learning_rate": 1.847537473233405e-05,
      "loss": 0.15462908148765564,
      "num_input_tokens_seen": 192522,
      "step": 184,
      "train_runtime": 487.4924,
      "train_tokens_per_second": 394.923
    },
    {
      "epoch": 0.07906827652526979,
      "grad_norm": 2.069824457168579,
      "learning_rate": 1.8466809421841543e-05,
      "loss": 0.09241805225610733,
      "num_input_tokens_seen": 193656,
      "step": 185,
      "train_runtime": 489.6796,
      "train_tokens_per_second": 395.475
    },
    {
      "epoch": 0.07949567261459557,
      "grad_norm": 2.976578950881958,
      "learning_rate": 1.8458244111349036e-05,
      "loss": 0.1367039829492569,
      "num_input_tokens_seen": 194402,
      "step": 186,
      "train_runtime": 491.7699,
      "train_tokens_per_second": 395.311
    },
    {
      "epoch": 0.07992306870392137,
      "grad_norm": 2.552767038345337,
      "learning_rate": 1.8449678800856533e-05,
      "loss": 0.17709875106811523,
      "num_input_tokens_seen": 195820,
      "step": 187,
      "train_runtime": 494.0382,
      "train_tokens_per_second": 396.366
    },
    {
      "epoch": 0.08035046479324714,
      "grad_norm": 2.4601333141326904,
      "learning_rate": 1.8441113490364027e-05,
      "loss": 0.11731860041618347,
      "num_input_tokens_seen": 196646,
      "step": 188,
      "train_runtime": 496.1262,
      "train_tokens_per_second": 396.363
    },
    {
      "epoch": 0.08077786088257292,
      "grad_norm": 2.562448740005493,
      "learning_rate": 1.843254817987152e-05,
      "loss": 0.14701655507087708,
      "num_input_tokens_seen": 197934,
      "step": 189,
      "train_runtime": 498.3383,
      "train_tokens_per_second": 397.188
    },
    {
      "epoch": 0.0812052569718987,
      "grad_norm": 2.922214984893799,
      "learning_rate": 1.8423982869379017e-05,
      "loss": 0.14705628156661987,
      "num_input_tokens_seen": 198668,
      "step": 190,
      "train_runtime": 500.4192,
      "train_tokens_per_second": 397.003
    },
    {
      "epoch": 0.08163265306122448,
      "grad_norm": 2.3108713626861572,
      "learning_rate": 1.841541755888651e-05,
      "loss": 0.10518313199281693,
      "num_input_tokens_seen": 199602,
      "step": 191,
      "train_runtime": 502.5654,
      "train_tokens_per_second": 397.166
    },
    {
      "epoch": 0.08206004915055028,
      "grad_norm": 2.9117729663848877,
      "learning_rate": 1.8406852248394004e-05,
      "loss": 0.18447798490524292,
      "num_input_tokens_seen": 200542,
      "step": 192,
      "train_runtime": 505.0707,
      "train_tokens_per_second": 397.057
    },
    {
      "epoch": 0.08248744523987606,
      "grad_norm": 4.030701160430908,
      "learning_rate": 1.8398286937901498e-05,
      "loss": 0.1964586079120636,
      "num_input_tokens_seen": 201758,
      "step": 193,
      "train_runtime": 507.2919,
      "train_tokens_per_second": 397.716
    },
    {
      "epoch": 0.08291484132920184,
      "grad_norm": 3.2769336700439453,
      "learning_rate": 1.8389721627408995e-05,
      "loss": 0.16791115701198578,
      "num_input_tokens_seen": 203172,
      "step": 194,
      "train_runtime": 509.5303,
      "train_tokens_per_second": 398.744
    },
    {
      "epoch": 0.08334223741852761,
      "grad_norm": 3.20062255859375,
      "learning_rate": 1.8381156316916492e-05,
      "loss": 0.13727587461471558,
      "num_input_tokens_seen": 204078,
      "step": 195,
      "train_runtime": 511.6734,
      "train_tokens_per_second": 398.844
    },
    {
      "epoch": 0.08376963350785341,
      "grad_norm": 2.539637327194214,
      "learning_rate": 1.8372591006423986e-05,
      "loss": 0.1486874520778656,
      "num_input_tokens_seen": 204922,
      "step": 196,
      "train_runtime": 513.7872,
      "train_tokens_per_second": 398.846
    },
    {
      "epoch": 0.08419702959717919,
      "grad_norm": 2.175872802734375,
      "learning_rate": 1.836402569593148e-05,
      "loss": 0.09931616485118866,
      "num_input_tokens_seen": 206042,
      "step": 197,
      "train_runtime": 515.9622,
      "train_tokens_per_second": 399.335
    },
    {
      "epoch": 0.08462442568650497,
      "grad_norm": 1.9456276893615723,
      "learning_rate": 1.8355460385438976e-05,
      "loss": 0.07155643403530121,
      "num_input_tokens_seen": 207324,
      "step": 198,
      "train_runtime": 518.1694,
      "train_tokens_per_second": 400.109
    },
    {
      "epoch": 0.08505182177583075,
      "grad_norm": 9.172248840332031,
      "learning_rate": 1.834689507494647e-05,
      "loss": 0.300564169883728,
      "num_input_tokens_seen": 207952,
      "step": 199,
      "train_runtime": 520.2399,
      "train_tokens_per_second": 399.723
    },
    {
      "epoch": 0.08547921786515654,
      "grad_norm": 2.8427793979644775,
      "learning_rate": 1.8338329764453963e-05,
      "loss": 0.10430914163589478,
      "num_input_tokens_seen": 208652,
      "step": 200,
      "train_runtime": 522.3632,
      "train_tokens_per_second": 399.439
    },
    {
      "epoch": 0.08590661395448232,
      "grad_norm": 1.8212063312530518,
      "learning_rate": 1.8329764453961457e-05,
      "loss": 0.08782214671373367,
      "num_input_tokens_seen": 210218,
      "step": 201,
      "train_runtime": 524.763,
      "train_tokens_per_second": 400.596
    },
    {
      "epoch": 0.0863340100438081,
      "grad_norm": 3.479308605194092,
      "learning_rate": 1.8321199143468954e-05,
      "loss": 0.1703043282032013,
      "num_input_tokens_seen": 211226,
      "step": 202,
      "train_runtime": 526.8819,
      "train_tokens_per_second": 400.898
    },
    {
      "epoch": 0.08676140613313388,
      "grad_norm": 3.103154182434082,
      "learning_rate": 1.8312633832976447e-05,
      "loss": 0.24448205530643463,
      "num_input_tokens_seen": 212022,
      "step": 203,
      "train_runtime": 528.9638,
      "train_tokens_per_second": 400.825
    },
    {
      "epoch": 0.08718880222245966,
      "grad_norm": 3.6667208671569824,
      "learning_rate": 1.830406852248394e-05,
      "loss": 0.18636125326156616,
      "num_input_tokens_seen": 212970,
      "step": 204,
      "train_runtime": 531.0816,
      "train_tokens_per_second": 401.012
    },
    {
      "epoch": 0.08761619831178545,
      "grad_norm": 1.9753594398498535,
      "learning_rate": 1.8295503211991438e-05,
      "loss": 0.1380254477262497,
      "num_input_tokens_seen": 214772,
      "step": 205,
      "train_runtime": 533.4082,
      "train_tokens_per_second": 402.641
    },
    {
      "epoch": 0.08804359440111123,
      "grad_norm": 2.537839889526367,
      "learning_rate": 1.828693790149893e-05,
      "loss": 0.14635343849658966,
      "num_input_tokens_seen": 215602,
      "step": 206,
      "train_runtime": 535.4898,
      "train_tokens_per_second": 402.626
    },
    {
      "epoch": 0.08847099049043701,
      "grad_norm": 2.4614720344543457,
      "learning_rate": 1.8278372591006425e-05,
      "loss": 0.14110636711120605,
      "num_input_tokens_seen": 216490,
      "step": 207,
      "train_runtime": 537.598,
      "train_tokens_per_second": 402.699
    },
    {
      "epoch": 0.08889838657976279,
      "grad_norm": 2.7557902336120605,
      "learning_rate": 1.826980728051392e-05,
      "loss": 0.14107386767864227,
      "num_input_tokens_seen": 217204,
      "step": 208,
      "train_runtime": 539.7033,
      "train_tokens_per_second": 402.451
    },
    {
      "epoch": 0.08932578266908858,
      "grad_norm": 2.0920276641845703,
      "learning_rate": 1.8261241970021416e-05,
      "loss": 0.1339716911315918,
      "num_input_tokens_seen": 218624,
      "step": 209,
      "train_runtime": 541.9757,
      "train_tokens_per_second": 403.383
    },
    {
      "epoch": 0.08975317875841436,
      "grad_norm": 1.595695972442627,
      "learning_rate": 1.825267665952891e-05,
      "loss": 0.11375150084495544,
      "num_input_tokens_seen": 220324,
      "step": 210,
      "train_runtime": 544.3085,
      "train_tokens_per_second": 404.778
    },
    {
      "epoch": 0.09018057484774014,
      "grad_norm": 2.601152181625366,
      "learning_rate": 1.8244111349036403e-05,
      "loss": 0.16513817012310028,
      "num_input_tokens_seen": 221488,
      "step": 211,
      "train_runtime": 559.005,
      "train_tokens_per_second": 396.218
    },
    {
      "epoch": 0.09060797093706592,
      "grad_norm": 1.9013690948486328,
      "learning_rate": 1.82355460385439e-05,
      "loss": 0.08416628837585449,
      "num_input_tokens_seen": 222612,
      "step": 212,
      "train_runtime": 561.4614,
      "train_tokens_per_second": 396.487
    },
    {
      "epoch": 0.09103536702639171,
      "grad_norm": 3.1588644981384277,
      "learning_rate": 1.8226980728051393e-05,
      "loss": 0.1620885729789734,
      "num_input_tokens_seen": 223292,
      "step": 213,
      "train_runtime": 564.0587,
      "train_tokens_per_second": 395.867
    },
    {
      "epoch": 0.0914627631157175,
      "grad_norm": 2.1632750034332275,
      "learning_rate": 1.8218415417558887e-05,
      "loss": 0.13554327189922333,
      "num_input_tokens_seen": 224310,
      "step": 214,
      "train_runtime": 566.1726,
      "train_tokens_per_second": 396.187
    },
    {
      "epoch": 0.09189015920504327,
      "grad_norm": 3.0766916275024414,
      "learning_rate": 1.8209850107066384e-05,
      "loss": 0.24831382930278778,
      "num_input_tokens_seen": 225312,
      "step": 215,
      "train_runtime": 568.3236,
      "train_tokens_per_second": 396.45
    },
    {
      "epoch": 0.09231755529436905,
      "grad_norm": 2.6327812671661377,
      "learning_rate": 1.8201284796573878e-05,
      "loss": 0.20338904857635498,
      "num_input_tokens_seen": 226316,
      "step": 216,
      "train_runtime": 570.4649,
      "train_tokens_per_second": 396.722
    },
    {
      "epoch": 0.09274495138369485,
      "grad_norm": 2.084002733230591,
      "learning_rate": 1.819271948608137e-05,
      "loss": 0.13886387646198273,
      "num_input_tokens_seen": 227250,
      "step": 217,
      "train_runtime": 572.5815,
      "train_tokens_per_second": 396.887
    },
    {
      "epoch": 0.09317234747302063,
      "grad_norm": 3.710822343826294,
      "learning_rate": 1.8184154175588865e-05,
      "loss": 0.15294967591762543,
      "num_input_tokens_seen": 228122,
      "step": 218,
      "train_runtime": 574.7105,
      "train_tokens_per_second": 396.934
    },
    {
      "epoch": 0.0935997435623464,
      "grad_norm": 2.816504716873169,
      "learning_rate": 1.817558886509636e-05,
      "loss": 0.15463289618492126,
      "num_input_tokens_seen": 229052,
      "step": 219,
      "train_runtime": 576.8652,
      "train_tokens_per_second": 397.063
    },
    {
      "epoch": 0.09402713965167218,
      "grad_norm": 2.4213547706604004,
      "learning_rate": 1.8167023554603855e-05,
      "loss": 0.10694430768489838,
      "num_input_tokens_seen": 229902,
      "step": 220,
      "train_runtime": 578.9749,
      "train_tokens_per_second": 397.085
    },
    {
      "epoch": 0.09445453574099796,
      "grad_norm": 3.384366750717163,
      "learning_rate": 1.815845824411135e-05,
      "loss": 0.24293643236160278,
      "num_input_tokens_seen": 230608,
      "step": 221,
      "train_runtime": 581.0437,
      "train_tokens_per_second": 396.886
    },
    {
      "epoch": 0.09488193183032376,
      "grad_norm": 2.192532777786255,
      "learning_rate": 1.8149892933618846e-05,
      "loss": 0.12488202750682831,
      "num_input_tokens_seen": 231556,
      "step": 222,
      "train_runtime": 583.1659,
      "train_tokens_per_second": 397.067
    },
    {
      "epoch": 0.09530932791964954,
      "grad_norm": 2.5996503829956055,
      "learning_rate": 1.814132762312634e-05,
      "loss": 0.1616736352443695,
      "num_input_tokens_seen": 232788,
      "step": 223,
      "train_runtime": 585.3168,
      "train_tokens_per_second": 397.713
    },
    {
      "epoch": 0.09573672400897532,
      "grad_norm": 3.0567333698272705,
      "learning_rate": 1.8132762312633833e-05,
      "loss": 0.11727012693881989,
      "num_input_tokens_seen": 233850,
      "step": 224,
      "train_runtime": 587.5038,
      "train_tokens_per_second": 398.04
    },
    {
      "epoch": 0.0961641200983011,
      "grad_norm": 2.4953536987304688,
      "learning_rate": 1.812419700214133e-05,
      "loss": 0.15394356846809387,
      "num_input_tokens_seen": 234864,
      "step": 225,
      "train_runtime": 589.6543,
      "train_tokens_per_second": 398.308
    },
    {
      "epoch": 0.09659151618762689,
      "grad_norm": 3.021813154220581,
      "learning_rate": 1.8115631691648824e-05,
      "loss": 0.1111551821231842,
      "num_input_tokens_seen": 235708,
      "step": 226,
      "train_runtime": 591.7278,
      "train_tokens_per_second": 398.339
    },
    {
      "epoch": 0.09701891227695267,
      "grad_norm": 2.5380287170410156,
      "learning_rate": 1.8107066381156317e-05,
      "loss": 0.1878495067358017,
      "num_input_tokens_seen": 236636,
      "step": 227,
      "train_runtime": 593.8048,
      "train_tokens_per_second": 398.508
    },
    {
      "epoch": 0.09744630836627845,
      "grad_norm": 2.5771961212158203,
      "learning_rate": 1.809850107066381e-05,
      "loss": 0.09383850544691086,
      "num_input_tokens_seen": 237612,
      "step": 228,
      "train_runtime": 595.981,
      "train_tokens_per_second": 398.691
    },
    {
      "epoch": 0.09787370445560423,
      "grad_norm": 1.9830071926116943,
      "learning_rate": 1.8089935760171308e-05,
      "loss": 0.12542767822742462,
      "num_input_tokens_seen": 238602,
      "step": 229,
      "train_runtime": 598.1479,
      "train_tokens_per_second": 398.901
    },
    {
      "epoch": 0.09830110054493002,
      "grad_norm": 2.5134365558624268,
      "learning_rate": 1.8081370449678805e-05,
      "loss": 0.1433359980583191,
      "num_input_tokens_seen": 239698,
      "step": 230,
      "train_runtime": 600.2815,
      "train_tokens_per_second": 399.309
    },
    {
      "epoch": 0.0987284966342558,
      "grad_norm": 2.7335493564605713,
      "learning_rate": 1.8072805139186298e-05,
      "loss": 0.16575391590595245,
      "num_input_tokens_seen": 240628,
      "step": 231,
      "train_runtime": 602.3973,
      "train_tokens_per_second": 399.451
    },
    {
      "epoch": 0.09915589272358158,
      "grad_norm": 3.091836929321289,
      "learning_rate": 1.8064239828693792e-05,
      "loss": 0.1720246970653534,
      "num_input_tokens_seen": 241414,
      "step": 232,
      "train_runtime": 604.4874,
      "train_tokens_per_second": 399.37
    },
    {
      "epoch": 0.09958328881290736,
      "grad_norm": 3.3940083980560303,
      "learning_rate": 1.805567451820129e-05,
      "loss": 0.11671648919582367,
      "num_input_tokens_seen": 242032,
      "step": 233,
      "train_runtime": 606.5232,
      "train_tokens_per_second": 399.048
    },
    {
      "epoch": 0.10001068490223314,
      "grad_norm": 2.0419411659240723,
      "learning_rate": 1.8047109207708782e-05,
      "loss": 0.11255921423435211,
      "num_input_tokens_seen": 243242,
      "step": 234,
      "train_runtime": 608.7078,
      "train_tokens_per_second": 399.604
    },
    {
      "epoch": 0.10001068490223314,
      "eval_loss": 0.5460442900657654,
      "eval_runtime": 58.7026,
      "eval_samples_per_second": 17.001,
      "eval_steps_per_second": 8.5,
      "num_input_tokens_seen": 243242,
      "step": 234
    },
    {
      "epoch": 0.10043808099155893,
      "grad_norm": 2.1736838817596436,
      "learning_rate": 1.8038543897216276e-05,
      "loss": 0.13026940822601318,
      "num_input_tokens_seen": 244358,
      "step": 235,
      "train_runtime": 669.6541,
      "train_tokens_per_second": 364.902
    },
    {
      "epoch": 0.10086547708088471,
      "grad_norm": 3.106304168701172,
      "learning_rate": 1.802997858672377e-05,
      "loss": 0.10195129364728928,
      "num_input_tokens_seen": 245862,
      "step": 236,
      "train_runtime": 671.8926,
      "train_tokens_per_second": 365.925
    },
    {
      "epoch": 0.10129287317021049,
      "grad_norm": 2.8228931427001953,
      "learning_rate": 1.8021413276231267e-05,
      "loss": 0.1436556726694107,
      "num_input_tokens_seen": 246806,
      "step": 237,
      "train_runtime": 674.0182,
      "train_tokens_per_second": 366.171
    },
    {
      "epoch": 0.10172026925953627,
      "grad_norm": 2.6426894664764404,
      "learning_rate": 1.801284796573876e-05,
      "loss": 0.1319834440946579,
      "num_input_tokens_seen": 248106,
      "step": 238,
      "train_runtime": 676.2569,
      "train_tokens_per_second": 366.881
    },
    {
      "epoch": 0.10214766534886206,
      "grad_norm": 4.234355449676514,
      "learning_rate": 1.8004282655246254e-05,
      "loss": 0.181757390499115,
      "num_input_tokens_seen": 249094,
      "step": 239,
      "train_runtime": 678.4898,
      "train_tokens_per_second": 367.13
    },
    {
      "epoch": 0.10257506143818784,
      "grad_norm": 2.244966745376587,
      "learning_rate": 1.799571734475375e-05,
      "loss": 0.08931125700473785,
      "num_input_tokens_seen": 249848,
      "step": 240,
      "train_runtime": 680.5725,
      "train_tokens_per_second": 367.114
    },
    {
      "epoch": 0.10300245752751362,
      "grad_norm": 2.718444347381592,
      "learning_rate": 1.7987152034261244e-05,
      "loss": 0.11326509714126587,
      "num_input_tokens_seen": 250600,
      "step": 241,
      "train_runtime": 696.4525,
      "train_tokens_per_second": 359.824
    },
    {
      "epoch": 0.1034298536168394,
      "grad_norm": 2.602396249771118,
      "learning_rate": 1.7978586723768738e-05,
      "loss": 0.1372726708650589,
      "num_input_tokens_seen": 251420,
      "step": 242,
      "train_runtime": 698.6553,
      "train_tokens_per_second": 359.863
    },
    {
      "epoch": 0.1038572497061652,
      "grad_norm": 2.42930269241333,
      "learning_rate": 1.797002141327623e-05,
      "loss": 0.14384983479976654,
      "num_input_tokens_seen": 252652,
      "step": 243,
      "train_runtime": 700.8388,
      "train_tokens_per_second": 360.499
    },
    {
      "epoch": 0.10428464579549097,
      "grad_norm": 3.0484702587127686,
      "learning_rate": 1.796145610278373e-05,
      "loss": 0.13648635149002075,
      "num_input_tokens_seen": 253610,
      "step": 244,
      "train_runtime": 703.0223,
      "train_tokens_per_second": 360.742
    },
    {
      "epoch": 0.10471204188481675,
      "grad_norm": 2.5430545806884766,
      "learning_rate": 1.7952890792291222e-05,
      "loss": 0.12859664857387543,
      "num_input_tokens_seen": 254638,
      "step": 245,
      "train_runtime": 705.2587,
      "train_tokens_per_second": 361.056
    },
    {
      "epoch": 0.10513943797414253,
      "grad_norm": 3.431422472000122,
      "learning_rate": 1.7944325481798716e-05,
      "loss": 0.15991127490997314,
      "num_input_tokens_seen": 255274,
      "step": 246,
      "train_runtime": 707.3851,
      "train_tokens_per_second": 360.87
    },
    {
      "epoch": 0.10556683406346833,
      "grad_norm": 2.83520770072937,
      "learning_rate": 1.7935760171306212e-05,
      "loss": 0.1783464252948761,
      "num_input_tokens_seen": 256368,
      "step": 247,
      "train_runtime": 709.5336,
      "train_tokens_per_second": 361.319
    },
    {
      "epoch": 0.1059942301527941,
      "grad_norm": 3.489138603210449,
      "learning_rate": 1.7927194860813706e-05,
      "loss": 0.1545252799987793,
      "num_input_tokens_seen": 257026,
      "step": 248,
      "train_runtime": 711.6423,
      "train_tokens_per_second": 361.173
    },
    {
      "epoch": 0.10642162624211988,
      "grad_norm": 2.563042163848877,
      "learning_rate": 1.79186295503212e-05,
      "loss": 0.08376234769821167,
      "num_input_tokens_seen": 257790,
      "step": 249,
      "train_runtime": 713.7212,
      "train_tokens_per_second": 361.191
    },
    {
      "epoch": 0.10684902233144566,
      "grad_norm": 4.280882358551025,
      "learning_rate": 1.7910064239828697e-05,
      "loss": 0.151292622089386,
      "num_input_tokens_seen": 259262,
      "step": 250,
      "train_runtime": 715.9656,
      "train_tokens_per_second": 362.115
    },
    {
      "epoch": 0.10727641842077144,
      "grad_norm": 2.9446420669555664,
      "learning_rate": 1.790149892933619e-05,
      "loss": 0.143342062830925,
      "num_input_tokens_seen": 260110,
      "step": 251,
      "train_runtime": 718.2046,
      "train_tokens_per_second": 362.167
    },
    {
      "epoch": 0.10770381451009724,
      "grad_norm": 2.2392935752868652,
      "learning_rate": 1.7892933618843684e-05,
      "loss": 0.12866157293319702,
      "num_input_tokens_seen": 261606,
      "step": 252,
      "train_runtime": 720.5883,
      "train_tokens_per_second": 363.045
    },
    {
      "epoch": 0.10813121059942302,
      "grad_norm": 2.1731903553009033,
      "learning_rate": 1.7884368308351177e-05,
      "loss": 0.1116916686296463,
      "num_input_tokens_seen": 262526,
      "step": 253,
      "train_runtime": 722.6988,
      "train_tokens_per_second": 363.258
    },
    {
      "epoch": 0.1085586066887488,
      "grad_norm": 4.395053863525391,
      "learning_rate": 1.7875802997858674e-05,
      "loss": 0.15097469091415405,
      "num_input_tokens_seen": 263480,
      "step": 254,
      "train_runtime": 724.8382,
      "train_tokens_per_second": 363.502
    },
    {
      "epoch": 0.10898600277807458,
      "grad_norm": 3.175166130065918,
      "learning_rate": 1.7867237687366168e-05,
      "loss": 0.15582749247550964,
      "num_input_tokens_seen": 264340,
      "step": 255,
      "train_runtime": 726.9482,
      "train_tokens_per_second": 363.63
    },
    {
      "epoch": 0.10941339886740037,
      "grad_norm": 2.18265438079834,
      "learning_rate": 1.785867237687366e-05,
      "loss": 0.1588597595691681,
      "num_input_tokens_seen": 265376,
      "step": 256,
      "train_runtime": 729.0958,
      "train_tokens_per_second": 363.98
    },
    {
      "epoch": 0.10984079495672615,
      "grad_norm": 3.173163414001465,
      "learning_rate": 1.785010706638116e-05,
      "loss": 0.19581833481788635,
      "num_input_tokens_seen": 266354,
      "step": 257,
      "train_runtime": 731.2373,
      "train_tokens_per_second": 364.251
    },
    {
      "epoch": 0.11026819104605193,
      "grad_norm": 2.3444836139678955,
      "learning_rate": 1.7841541755888652e-05,
      "loss": 0.1477520614862442,
      "num_input_tokens_seen": 267484,
      "step": 258,
      "train_runtime": 733.4668,
      "train_tokens_per_second": 364.685
    },
    {
      "epoch": 0.11069558713537771,
      "grad_norm": 2.4513041973114014,
      "learning_rate": 1.7832976445396146e-05,
      "loss": 0.169121652841568,
      "num_input_tokens_seen": 268488,
      "step": 259,
      "train_runtime": 735.6543,
      "train_tokens_per_second": 364.965
    },
    {
      "epoch": 0.1111229832247035,
      "grad_norm": 2.247398853302002,
      "learning_rate": 1.782441113490364e-05,
      "loss": 0.14096848666667938,
      "num_input_tokens_seen": 269740,
      "step": 260,
      "train_runtime": 737.8668,
      "train_tokens_per_second": 365.567
    },
    {
      "epoch": 0.11155037931402928,
      "grad_norm": 2.104783058166504,
      "learning_rate": 1.7815845824411136e-05,
      "loss": 0.1397685408592224,
      "num_input_tokens_seen": 270722,
      "step": 261,
      "train_runtime": 740.0164,
      "train_tokens_per_second": 365.832
    },
    {
      "epoch": 0.11197777540335506,
      "grad_norm": 2.633096218109131,
      "learning_rate": 1.780728051391863e-05,
      "loss": 0.15674568712711334,
      "num_input_tokens_seen": 271516,
      "step": 262,
      "train_runtime": 742.099,
      "train_tokens_per_second": 365.876
    },
    {
      "epoch": 0.11240517149268084,
      "grad_norm": 3.091400384902954,
      "learning_rate": 1.7798715203426123e-05,
      "loss": 0.12970855832099915,
      "num_input_tokens_seen": 272224,
      "step": 263,
      "train_runtime": 744.1502,
      "train_tokens_per_second": 365.819
    },
    {
      "epoch": 0.11283256758200662,
      "grad_norm": 2.533055067062378,
      "learning_rate": 1.779014989293362e-05,
      "loss": 0.1559002697467804,
      "num_input_tokens_seen": 273658,
      "step": 264,
      "train_runtime": 746.3582,
      "train_tokens_per_second": 366.658
    },
    {
      "epoch": 0.11325996367133241,
      "grad_norm": 2.816351890563965,
      "learning_rate": 1.7781584582441117e-05,
      "loss": 0.1578320562839508,
      "num_input_tokens_seen": 274538,
      "step": 265,
      "train_runtime": 748.4778,
      "train_tokens_per_second": 366.795
    },
    {
      "epoch": 0.11368735976065819,
      "grad_norm": 3.1316685676574707,
      "learning_rate": 1.777301927194861e-05,
      "loss": 0.17508356273174286,
      "num_input_tokens_seen": 275238,
      "step": 266,
      "train_runtime": 750.5642,
      "train_tokens_per_second": 366.708
    },
    {
      "epoch": 0.11411475584998397,
      "grad_norm": 2.4046151638031006,
      "learning_rate": 1.7764453961456104e-05,
      "loss": 0.19063466787338257,
      "num_input_tokens_seen": 276596,
      "step": 267,
      "train_runtime": 752.7848,
      "train_tokens_per_second": 367.43
    },
    {
      "epoch": 0.11454215193930975,
      "grad_norm": 2.954357624053955,
      "learning_rate": 1.7755888650963598e-05,
      "loss": 0.13341747224330902,
      "num_input_tokens_seen": 277522,
      "step": 268,
      "train_runtime": 754.9005,
      "train_tokens_per_second": 367.627
    },
    {
      "epoch": 0.11496954802863554,
      "grad_norm": 3.105802059173584,
      "learning_rate": 1.7747323340471095e-05,
      "loss": 0.15308117866516113,
      "num_input_tokens_seen": 279086,
      "step": 269,
      "train_runtime": 757.1899,
      "train_tokens_per_second": 368.581
    },
    {
      "epoch": 0.11539694411796132,
      "grad_norm": 2.347745418548584,
      "learning_rate": 1.773875802997859e-05,
      "loss": 0.12355208396911621,
      "num_input_tokens_seen": 280000,
      "step": 270,
      "train_runtime": 759.3841,
      "train_tokens_per_second": 368.72
    },
    {
      "epoch": 0.1158243402072871,
      "grad_norm": 2.6124165058135986,
      "learning_rate": 1.7730192719486082e-05,
      "loss": 0.11659465730190277,
      "num_input_tokens_seen": 280854,
      "step": 271,
      "train_runtime": 774.768,
      "train_tokens_per_second": 362.501
    },
    {
      "epoch": 0.11625173629661288,
      "grad_norm": 2.646575689315796,
      "learning_rate": 1.772162740899358e-05,
      "loss": 0.14843402802944183,
      "num_input_tokens_seen": 281708,
      "step": 272,
      "train_runtime": 776.8624,
      "train_tokens_per_second": 362.623
    },
    {
      "epoch": 0.11667913238593867,
      "grad_norm": 1.9934483766555786,
      "learning_rate": 1.7713062098501073e-05,
      "loss": 0.11461179703474045,
      "num_input_tokens_seen": 283286,
      "step": 273,
      "train_runtime": 779.2415,
      "train_tokens_per_second": 363.541
    },
    {
      "epoch": 0.11710652847526445,
      "grad_norm": 3.048039197921753,
      "learning_rate": 1.7704496788008566e-05,
      "loss": 0.10470855236053467,
      "num_input_tokens_seen": 283902,
      "step": 274,
      "train_runtime": 781.377,
      "train_tokens_per_second": 363.335
    },
    {
      "epoch": 0.11753392456459023,
      "grad_norm": 2.832122802734375,
      "learning_rate": 1.7695931477516063e-05,
      "loss": 0.12394038587808609,
      "num_input_tokens_seen": 284622,
      "step": 275,
      "train_runtime": 783.4499,
      "train_tokens_per_second": 363.293
    },
    {
      "epoch": 0.11796132065391601,
      "grad_norm": 2.1014561653137207,
      "learning_rate": 1.7687366167023557e-05,
      "loss": 0.18216493725776672,
      "num_input_tokens_seen": 286416,
      "step": 276,
      "train_runtime": 785.7736,
      "train_tokens_per_second": 364.502
    },
    {
      "epoch": 0.11838871674324179,
      "grad_norm": 2.309156894683838,
      "learning_rate": 1.767880085653105e-05,
      "loss": 0.09812641143798828,
      "num_input_tokens_seen": 287294,
      "step": 277,
      "train_runtime": 787.8908,
      "train_tokens_per_second": 364.637
    },
    {
      "epoch": 0.11881611283256759,
      "grad_norm": 2.1988325119018555,
      "learning_rate": 1.7670235546038544e-05,
      "loss": 0.09043700248003006,
      "num_input_tokens_seen": 288488,
      "step": 278,
      "train_runtime": 790.0798,
      "train_tokens_per_second": 365.138
    },
    {
      "epoch": 0.11924350892189337,
      "grad_norm": 2.2506520748138428,
      "learning_rate": 1.766167023554604e-05,
      "loss": 0.1063460111618042,
      "num_input_tokens_seen": 289728,
      "step": 279,
      "train_runtime": 792.2868,
      "train_tokens_per_second": 365.686
    },
    {
      "epoch": 0.11967090501121914,
      "grad_norm": 3.2995102405548096,
      "learning_rate": 1.7653104925053535e-05,
      "loss": 0.12390337884426117,
      "num_input_tokens_seen": 290828,
      "step": 280,
      "train_runtime": 794.4362,
      "train_tokens_per_second": 366.081
    },
    {
      "epoch": 0.12009830110054492,
      "grad_norm": 1.905795931816101,
      "learning_rate": 1.7644539614561028e-05,
      "loss": 0.07884510606527328,
      "num_input_tokens_seen": 291796,
      "step": 281,
      "train_runtime": 796.5539,
      "train_tokens_per_second": 366.323
    },
    {
      "epoch": 0.12052569718987072,
      "grad_norm": 2.213050365447998,
      "learning_rate": 1.7635974304068525e-05,
      "loss": 0.12539790570735931,
      "num_input_tokens_seen": 292782,
      "step": 282,
      "train_runtime": 798.7022,
      "train_tokens_per_second": 366.572
    },
    {
      "epoch": 0.1209530932791965,
      "grad_norm": 4.263006687164307,
      "learning_rate": 1.762740899357602e-05,
      "loss": 0.10484211146831512,
      "num_input_tokens_seen": 294044,
      "step": 283,
      "train_runtime": 800.9192,
      "train_tokens_per_second": 367.133
    },
    {
      "epoch": 0.12138048936852228,
      "grad_norm": 3.1561226844787598,
      "learning_rate": 1.7618843683083512e-05,
      "loss": 0.14216279983520508,
      "num_input_tokens_seen": 295040,
      "step": 284,
      "train_runtime": 803.0984,
      "train_tokens_per_second": 367.377
    },
    {
      "epoch": 0.12180788545784806,
      "grad_norm": 2.0373404026031494,
      "learning_rate": 1.761027837259101e-05,
      "loss": 0.11023009568452835,
      "num_input_tokens_seen": 296080,
      "step": 285,
      "train_runtime": 805.2902,
      "train_tokens_per_second": 367.669
    },
    {
      "epoch": 0.12223528154717385,
      "grad_norm": 2.9239816665649414,
      "learning_rate": 1.7601713062098503e-05,
      "loss": 0.2894620895385742,
      "num_input_tokens_seen": 296966,
      "step": 286,
      "train_runtime": 807.4058,
      "train_tokens_per_second": 367.803
    },
    {
      "epoch": 0.12266267763649963,
      "grad_norm": 4.2024760246276855,
      "learning_rate": 1.7593147751605996e-05,
      "loss": 0.14657855033874512,
      "num_input_tokens_seen": 297852,
      "step": 287,
      "train_runtime": 809.5535,
      "train_tokens_per_second": 367.921
    },
    {
      "epoch": 0.12309007372582541,
      "grad_norm": 2.8517258167266846,
      "learning_rate": 1.758458244111349e-05,
      "loss": 0.19364164769649506,
      "num_input_tokens_seen": 298816,
      "step": 288,
      "train_runtime": 811.6752,
      "train_tokens_per_second": 368.147
    },
    {
      "epoch": 0.12351746981515119,
      "grad_norm": 2.9998342990875244,
      "learning_rate": 1.7576017130620987e-05,
      "loss": 0.15787193179130554,
      "num_input_tokens_seen": 299528,
      "step": 289,
      "train_runtime": 813.729,
      "train_tokens_per_second": 368.093
    },
    {
      "epoch": 0.12394486590447698,
      "grad_norm": 2.6897270679473877,
      "learning_rate": 1.756745182012848e-05,
      "loss": 0.179714635014534,
      "num_input_tokens_seen": 300536,
      "step": 290,
      "train_runtime": 815.9137,
      "train_tokens_per_second": 368.343
    },
    {
      "epoch": 0.12437226199380276,
      "grad_norm": 2.2788524627685547,
      "learning_rate": 1.7558886509635974e-05,
      "loss": 0.1147165596485138,
      "num_input_tokens_seen": 301440,
      "step": 291,
      "train_runtime": 818.0307,
      "train_tokens_per_second": 368.495
    },
    {
      "epoch": 0.12479965808312854,
      "grad_norm": 2.4706780910491943,
      "learning_rate": 1.755032119914347e-05,
      "loss": 0.1395479440689087,
      "num_input_tokens_seen": 302300,
      "step": 292,
      "train_runtime": 820.1939,
      "train_tokens_per_second": 368.571
    },
    {
      "epoch": 0.12522705417245433,
      "grad_norm": 2.3851404190063477,
      "learning_rate": 1.7541755888650965e-05,
      "loss": 0.14536121487617493,
      "num_input_tokens_seen": 303418,
      "step": 293,
      "train_runtime": 822.3458,
      "train_tokens_per_second": 368.966
    },
    {
      "epoch": 0.1256544502617801,
      "grad_norm": 2.1842844486236572,
      "learning_rate": 1.7533190578158458e-05,
      "loss": 0.1375734657049179,
      "num_input_tokens_seen": 304350,
      "step": 294,
      "train_runtime": 824.6257,
      "train_tokens_per_second": 369.077
    },
    {
      "epoch": 0.1260818463511059,
      "grad_norm": 1.994462013244629,
      "learning_rate": 1.7524625267665952e-05,
      "loss": 0.13084162771701813,
      "num_input_tokens_seen": 305476,
      "step": 295,
      "train_runtime": 826.8659,
      "train_tokens_per_second": 369.438
    },
    {
      "epoch": 0.12650924244043166,
      "grad_norm": 2.4618313312530518,
      "learning_rate": 1.751605995717345e-05,
      "loss": 0.14622284471988678,
      "num_input_tokens_seen": 306308,
      "step": 296,
      "train_runtime": 828.9845,
      "train_tokens_per_second": 369.498
    },
    {
      "epoch": 0.12693663852975745,
      "grad_norm": 1.887038230895996,
      "learning_rate": 1.7507494646680942e-05,
      "loss": 0.0890108048915863,
      "num_input_tokens_seen": 307260,
      "step": 297,
      "train_runtime": 831.1363,
      "train_tokens_per_second": 369.687
    },
    {
      "epoch": 0.12736403461908324,
      "grad_norm": 3.0748531818389893,
      "learning_rate": 1.749892933618844e-05,
      "loss": 0.15173892676830292,
      "num_input_tokens_seen": 307968,
      "step": 298,
      "train_runtime": 833.2195,
      "train_tokens_per_second": 369.612
    },
    {
      "epoch": 0.127791430708409,
      "grad_norm": 1.8806450366973877,
      "learning_rate": 1.7490364025695933e-05,
      "loss": 0.09086901694536209,
      "num_input_tokens_seen": 309040,
      "step": 299,
      "train_runtime": 835.4988,
      "train_tokens_per_second": 369.887
    },
    {
      "epoch": 0.1282188267977348,
      "grad_norm": 2.5331919193267822,
      "learning_rate": 1.748179871520343e-05,
      "loss": 0.16212350130081177,
      "num_input_tokens_seen": 310064,
      "step": 300,
      "train_runtime": 837.7224,
      "train_tokens_per_second": 370.127
    },
    {
      "epoch": 0.1286462228870606,
      "grad_norm": 2.6693081855773926,
      "learning_rate": 1.7473233404710924e-05,
      "loss": 0.1564721018075943,
      "num_input_tokens_seen": 311010,
      "step": 301,
      "train_runtime": 853.5044,
      "train_tokens_per_second": 364.392
    },
    {
      "epoch": 0.12907361897638636,
      "grad_norm": 1.731451153755188,
      "learning_rate": 1.7464668094218417e-05,
      "loss": 0.09542805701494217,
      "num_input_tokens_seen": 312498,
      "step": 302,
      "train_runtime": 855.8469,
      "train_tokens_per_second": 365.133
    },
    {
      "epoch": 0.12950101506571216,
      "grad_norm": 1.7048978805541992,
      "learning_rate": 1.745610278372591e-05,
      "loss": 0.11521374434232712,
      "num_input_tokens_seen": 313832,
      "step": 303,
      "train_runtime": 858.0815,
      "train_tokens_per_second": 365.737
    },
    {
      "epoch": 0.12992841115503792,
      "grad_norm": 2.9149906635284424,
      "learning_rate": 1.7447537473233408e-05,
      "loss": 0.1253424882888794,
      "num_input_tokens_seen": 314562,
      "step": 304,
      "train_runtime": 860.2229,
      "train_tokens_per_second": 365.675
    },
    {
      "epoch": 0.13035580724436371,
      "grad_norm": 2.8207590579986572,
      "learning_rate": 1.74389721627409e-05,
      "loss": 0.127387136220932,
      "num_input_tokens_seen": 315366,
      "step": 305,
      "train_runtime": 862.378,
      "train_tokens_per_second": 365.693
    },
    {
      "epoch": 0.1307832033336895,
      "grad_norm": 2.3153860569000244,
      "learning_rate": 1.7430406852248395e-05,
      "loss": 0.07615187019109726,
      "num_input_tokens_seen": 316344,
      "step": 306,
      "train_runtime": 864.5162,
      "train_tokens_per_second": 365.92
    },
    {
      "epoch": 0.13121059942301527,
      "grad_norm": 1.8810218572616577,
      "learning_rate": 1.7421841541755892e-05,
      "loss": 0.09668294340372086,
      "num_input_tokens_seen": 317502,
      "step": 307,
      "train_runtime": 866.7393,
      "train_tokens_per_second": 366.318
    },
    {
      "epoch": 0.13163799551234107,
      "grad_norm": 2.1955950260162354,
      "learning_rate": 1.7413276231263385e-05,
      "loss": 0.1333192139863968,
      "num_input_tokens_seen": 318744,
      "step": 308,
      "train_runtime": 869.0412,
      "train_tokens_per_second": 366.777
    },
    {
      "epoch": 0.13206539160166683,
      "grad_norm": 2.143160581588745,
      "learning_rate": 1.740471092077088e-05,
      "loss": 0.09647417068481445,
      "num_input_tokens_seen": 319662,
      "step": 309,
      "train_runtime": 871.1791,
      "train_tokens_per_second": 366.93
    },
    {
      "epoch": 0.13249278769099262,
      "grad_norm": 2.6915321350097656,
      "learning_rate": 1.7396145610278376e-05,
      "loss": 0.18075433373451233,
      "num_input_tokens_seen": 320918,
      "step": 310,
      "train_runtime": 873.4328,
      "train_tokens_per_second": 367.422
    },
    {
      "epoch": 0.13292018378031842,
      "grad_norm": 2.8213682174682617,
      "learning_rate": 1.738758029978587e-05,
      "loss": 0.10546336323022842,
      "num_input_tokens_seen": 321682,
      "step": 311,
      "train_runtime": 876.1245,
      "train_tokens_per_second": 367.165
    },
    {
      "epoch": 0.13334757986964418,
      "grad_norm": 2.9458816051483154,
      "learning_rate": 1.7379014989293363e-05,
      "loss": 0.18216489255428314,
      "num_input_tokens_seen": 322496,
      "step": 312,
      "train_runtime": 878.2305,
      "train_tokens_per_second": 367.211
    },
    {
      "epoch": 0.13377497595896998,
      "grad_norm": 2.7808685302734375,
      "learning_rate": 1.7370449678800857e-05,
      "loss": 0.179776132106781,
      "num_input_tokens_seen": 323340,
      "step": 313,
      "train_runtime": 880.3312,
      "train_tokens_per_second": 367.294
    },
    {
      "epoch": 0.13420237204829577,
      "grad_norm": 1.5709812641143799,
      "learning_rate": 1.7361884368308354e-05,
      "loss": 0.08623568713665009,
      "num_input_tokens_seen": 324558,
      "step": 314,
      "train_runtime": 882.5212,
      "train_tokens_per_second": 367.762
    },
    {
      "epoch": 0.13462976813762154,
      "grad_norm": 4.398742198944092,
      "learning_rate": 1.7353319057815847e-05,
      "loss": 0.19522878527641296,
      "num_input_tokens_seen": 325174,
      "step": 315,
      "train_runtime": 884.5702,
      "train_tokens_per_second": 367.607
    },
    {
      "epoch": 0.13505716422694733,
      "grad_norm": 3.6003987789154053,
      "learning_rate": 1.734475374732334e-05,
      "loss": 0.10071863979101181,
      "num_input_tokens_seen": 326062,
      "step": 316,
      "train_runtime": 886.7388,
      "train_tokens_per_second": 367.709
    },
    {
      "epoch": 0.1354845603162731,
      "grad_norm": 2.6192052364349365,
      "learning_rate": 1.7336188436830838e-05,
      "loss": 0.159763365983963,
      "num_input_tokens_seen": 326946,
      "step": 317,
      "train_runtime": 888.8532,
      "train_tokens_per_second": 367.829
    },
    {
      "epoch": 0.1359119564055989,
      "grad_norm": 2.826225757598877,
      "learning_rate": 1.732762312633833e-05,
      "loss": 0.07451944053173065,
      "num_input_tokens_seen": 327712,
      "step": 318,
      "train_runtime": 890.9057,
      "train_tokens_per_second": 367.841
    },
    {
      "epoch": 0.13633935249492468,
      "grad_norm": 2.4688735008239746,
      "learning_rate": 1.7319057815845825e-05,
      "loss": 0.24531584978103638,
      "num_input_tokens_seen": 329000,
      "step": 319,
      "train_runtime": 893.2119,
      "train_tokens_per_second": 368.334
    },
    {
      "epoch": 0.13676674858425045,
      "grad_norm": 2.5273513793945312,
      "learning_rate": 1.731049250535332e-05,
      "loss": 0.11082414537668228,
      "num_input_tokens_seen": 329852,
      "step": 320,
      "train_runtime": 895.318,
      "train_tokens_per_second": 368.419
    },
    {
      "epoch": 0.13719414467357624,
      "grad_norm": 1.9812712669372559,
      "learning_rate": 1.7301927194860816e-05,
      "loss": 0.08259004354476929,
      "num_input_tokens_seen": 330872,
      "step": 321,
      "train_runtime": 897.5869,
      "train_tokens_per_second": 368.624
    },
    {
      "epoch": 0.137621540762902,
      "grad_norm": 3.8385865688323975,
      "learning_rate": 1.729336188436831e-05,
      "loss": 0.08933961391448975,
      "num_input_tokens_seen": 331688,
      "step": 322,
      "train_runtime": 899.7637,
      "train_tokens_per_second": 368.639
    },
    {
      "epoch": 0.1380489368522278,
      "grad_norm": 3.434079885482788,
      "learning_rate": 1.7284796573875803e-05,
      "loss": 0.22022958099842072,
      "num_input_tokens_seen": 333338,
      "step": 323,
      "train_runtime": 902.0942,
      "train_tokens_per_second": 369.516
    },
    {
      "epoch": 0.1384763329415536,
      "grad_norm": 3.6233861446380615,
      "learning_rate": 1.72762312633833e-05,
      "loss": 0.24009671807289124,
      "num_input_tokens_seen": 334224,
      "step": 324,
      "train_runtime": 904.1812,
      "train_tokens_per_second": 369.643
    },
    {
      "epoch": 0.13890372903087936,
      "grad_norm": 1.9435126781463623,
      "learning_rate": 1.7267665952890793e-05,
      "loss": 0.09001989662647247,
      "num_input_tokens_seen": 335028,
      "step": 325,
      "train_runtime": 906.2597,
      "train_tokens_per_second": 369.682
    },
    {
      "epoch": 0.13933112512020515,
      "grad_norm": 2.826512336730957,
      "learning_rate": 1.7259100642398287e-05,
      "loss": 0.12890757620334625,
      "num_input_tokens_seen": 336004,
      "step": 326,
      "train_runtime": 908.3817,
      "train_tokens_per_second": 369.893
    },
    {
      "epoch": 0.13975852120953094,
      "grad_norm": 2.837244987487793,
      "learning_rate": 1.7250535331905784e-05,
      "loss": 0.11885890364646912,
      "num_input_tokens_seen": 336610,
      "step": 327,
      "train_runtime": 910.4452,
      "train_tokens_per_second": 369.72
    },
    {
      "epoch": 0.1401859172988567,
      "grad_norm": 2.2175140380859375,
      "learning_rate": 1.7241970021413277e-05,
      "loss": 0.06964915990829468,
      "num_input_tokens_seen": 337282,
      "step": 328,
      "train_runtime": 912.5036,
      "train_tokens_per_second": 369.623
    },
    {
      "epoch": 0.1406133133881825,
      "grad_norm": 2.021813154220581,
      "learning_rate": 1.723340471092077e-05,
      "loss": 0.12436268478631973,
      "num_input_tokens_seen": 338422,
      "step": 329,
      "train_runtime": 914.6854,
      "train_tokens_per_second": 369.987
    },
    {
      "epoch": 0.14104070947750827,
      "grad_norm": 2.7541916370391846,
      "learning_rate": 1.7224839400428265e-05,
      "loss": 0.17609360814094543,
      "num_input_tokens_seen": 339422,
      "step": 330,
      "train_runtime": 916.853,
      "train_tokens_per_second": 370.203
    },
    {
      "epoch": 0.14146810556683406,
      "grad_norm": 2.394658327102661,
      "learning_rate": 1.721627408993576e-05,
      "loss": 0.11716184765100479,
      "num_input_tokens_seen": 340936,
      "step": 331,
      "train_runtime": 931.6924,
      "train_tokens_per_second": 365.932
    },
    {
      "epoch": 0.14189550165615986,
      "grad_norm": 2.114901304244995,
      "learning_rate": 1.7207708779443255e-05,
      "loss": 0.10051403194665909,
      "num_input_tokens_seen": 341784,
      "step": 332,
      "train_runtime": 933.8063,
      "train_tokens_per_second": 366.012
    },
    {
      "epoch": 0.14232289774548562,
      "grad_norm": 2.624973773956299,
      "learning_rate": 1.7199143468950752e-05,
      "loss": 0.1103091835975647,
      "num_input_tokens_seen": 343004,
      "step": 333,
      "train_runtime": 936.0629,
      "train_tokens_per_second": 366.433
    },
    {
      "epoch": 0.14275029383481141,
      "grad_norm": 2.721280336380005,
      "learning_rate": 1.7190578158458246e-05,
      "loss": 0.2677321135997772,
      "num_input_tokens_seen": 345688,
      "step": 334,
      "train_runtime": 938.8453,
      "train_tokens_per_second": 368.205
    },
    {
      "epoch": 0.14317768992413718,
      "grad_norm": 1.50456702709198,
      "learning_rate": 1.7182012847965743e-05,
      "loss": 0.058263201266527176,
      "num_input_tokens_seen": 347148,
      "step": 335,
      "train_runtime": 941.1161,
      "train_tokens_per_second": 368.868
    },
    {
      "epoch": 0.14360508601346297,
      "grad_norm": 3.347944736480713,
      "learning_rate": 1.7173447537473236e-05,
      "loss": 0.1528206467628479,
      "num_input_tokens_seen": 348164,
      "step": 336,
      "train_runtime": 943.237,
      "train_tokens_per_second": 369.116
    },
    {
      "epoch": 0.14403248210278877,
      "grad_norm": 2.0157036781311035,
      "learning_rate": 1.716488222698073e-05,
      "loss": 0.13663557171821594,
      "num_input_tokens_seen": 349356,
      "step": 337,
      "train_runtime": 945.4468,
      "train_tokens_per_second": 369.514
    },
    {
      "epoch": 0.14445987819211453,
      "grad_norm": 2.7032551765441895,
      "learning_rate": 1.7156316916488223e-05,
      "loss": 0.16470125317573547,
      "num_input_tokens_seen": 350252,
      "step": 338,
      "train_runtime": 947.648,
      "train_tokens_per_second": 369.601
    },
    {
      "epoch": 0.14488727428144033,
      "grad_norm": 3.2049143314361572,
      "learning_rate": 1.714775160599572e-05,
      "loss": 0.19470034539699554,
      "num_input_tokens_seen": 350854,
      "step": 339,
      "train_runtime": 949.7293,
      "train_tokens_per_second": 369.425
    },
    {
      "epoch": 0.14531467037076612,
      "grad_norm": 1.857224464416504,
      "learning_rate": 1.7139186295503214e-05,
      "loss": 0.08987314999103546,
      "num_input_tokens_seen": 351776,
      "step": 340,
      "train_runtime": 951.8328,
      "train_tokens_per_second": 369.578
    },
    {
      "epoch": 0.14574206646009188,
      "grad_norm": 2.442211627960205,
      "learning_rate": 1.7130620985010707e-05,
      "loss": 0.21366232633590698,
      "num_input_tokens_seen": 352826,
      "step": 341,
      "train_runtime": 953.9986,
      "train_tokens_per_second": 369.839
    },
    {
      "epoch": 0.14616946254941768,
      "grad_norm": 2.319977283477783,
      "learning_rate": 1.7122055674518204e-05,
      "loss": 0.15562638640403748,
      "num_input_tokens_seen": 353882,
      "step": 342,
      "train_runtime": 956.1857,
      "train_tokens_per_second": 370.098
    },
    {
      "epoch": 0.14659685863874344,
      "grad_norm": 2.556077718734741,
      "learning_rate": 1.7113490364025698e-05,
      "loss": 0.16621138155460358,
      "num_input_tokens_seen": 354970,
      "step": 343,
      "train_runtime": 958.3647,
      "train_tokens_per_second": 370.391
    },
    {
      "epoch": 0.14702425472806924,
      "grad_norm": 2.5956978797912598,
      "learning_rate": 1.710492505353319e-05,
      "loss": 0.17115283012390137,
      "num_input_tokens_seen": 356082,
      "step": 344,
      "train_runtime": 960.5342,
      "train_tokens_per_second": 370.712
    },
    {
      "epoch": 0.14745165081739503,
      "grad_norm": 2.843019723892212,
      "learning_rate": 1.709635974304069e-05,
      "loss": 0.12683501839637756,
      "num_input_tokens_seen": 357028,
      "step": 345,
      "train_runtime": 962.6755,
      "train_tokens_per_second": 370.871
    },
    {
      "epoch": 0.1478790469067208,
      "grad_norm": 2.406247138977051,
      "learning_rate": 1.7087794432548182e-05,
      "loss": 0.18846774101257324,
      "num_input_tokens_seen": 358174,
      "step": 346,
      "train_runtime": 964.8523,
      "train_tokens_per_second": 371.222
    },
    {
      "epoch": 0.1483064429960466,
      "grad_norm": 2.706057071685791,
      "learning_rate": 1.7079229122055676e-05,
      "loss": 0.13083621859550476,
      "num_input_tokens_seen": 359022,
      "step": 347,
      "train_runtime": 967.0563,
      "train_tokens_per_second": 371.252
    },
    {
      "epoch": 0.14873383908537238,
      "grad_norm": 3.15480375289917,
      "learning_rate": 1.707066381156317e-05,
      "loss": 0.15986329317092896,
      "num_input_tokens_seen": 359836,
      "step": 348,
      "train_runtime": 969.1381,
      "train_tokens_per_second": 371.295
    },
    {
      "epoch": 0.14916123517469815,
      "grad_norm": 2.7063496112823486,
      "learning_rate": 1.7062098501070666e-05,
      "loss": 0.15175658464431763,
      "num_input_tokens_seen": 360908,
      "step": 349,
      "train_runtime": 971.3259,
      "train_tokens_per_second": 371.562
    },
    {
      "epoch": 0.14958863126402394,
      "grad_norm": 1.7199162244796753,
      "learning_rate": 1.705353319057816e-05,
      "loss": 0.09106869250535965,
      "num_input_tokens_seen": 361784,
      "step": 350,
      "train_runtime": 973.4459,
      "train_tokens_per_second": 371.653
    },
    {
      "epoch": 0.1500160273533497,
      "grad_norm": 2.5850725173950195,
      "learning_rate": 1.7044967880085653e-05,
      "loss": 0.14896491169929504,
      "num_input_tokens_seen": 362754,
      "step": 351,
      "train_runtime": 975.5594,
      "train_tokens_per_second": 371.842
    },
    {
      "epoch": 0.1504434234426755,
      "grad_norm": 2.769216537475586,
      "learning_rate": 1.703640256959315e-05,
      "loss": 0.17172960937023163,
      "num_input_tokens_seen": 363890,
      "step": 352,
      "train_runtime": 977.7693,
      "train_tokens_per_second": 372.163
    },
    {
      "epoch": 0.1508708195320013,
      "grad_norm": 2.2079951763153076,
      "learning_rate": 1.7027837259100644e-05,
      "loss": 0.1200181245803833,
      "num_input_tokens_seen": 365082,
      "step": 353,
      "train_runtime": 979.9852,
      "train_tokens_per_second": 372.538
    },
    {
      "epoch": 0.15129821562132706,
      "grad_norm": 2.190399408340454,
      "learning_rate": 1.7019271948608138e-05,
      "loss": 0.15325041115283966,
      "num_input_tokens_seen": 365994,
      "step": 354,
      "train_runtime": 982.1446,
      "train_tokens_per_second": 372.648
    },
    {
      "epoch": 0.15172561171065285,
      "grad_norm": 2.7263314723968506,
      "learning_rate": 1.701070663811563e-05,
      "loss": 0.14416253566741943,
      "num_input_tokens_seen": 366826,
      "step": 355,
      "train_runtime": 984.3156,
      "train_tokens_per_second": 372.671
    },
    {
      "epoch": 0.15215300779997862,
      "grad_norm": 3.734889030456543,
      "learning_rate": 1.7002141327623128e-05,
      "loss": 0.14761826395988464,
      "num_input_tokens_seen": 367598,
      "step": 356,
      "train_runtime": 986.3869,
      "train_tokens_per_second": 372.671
    },
    {
      "epoch": 0.1525804038893044,
      "grad_norm": 1.6911550760269165,
      "learning_rate": 1.6993576017130622e-05,
      "loss": 0.0787629708647728,
      "num_input_tokens_seen": 369124,
      "step": 357,
      "train_runtime": 988.7967,
      "train_tokens_per_second": 373.306
    },
    {
      "epoch": 0.1530077999786302,
      "grad_norm": 2.4877729415893555,
      "learning_rate": 1.6985010706638115e-05,
      "loss": 0.13544218242168427,
      "num_input_tokens_seen": 369900,
      "step": 358,
      "train_runtime": 990.8914,
      "train_tokens_per_second": 373.3
    },
    {
      "epoch": 0.15343519606795597,
      "grad_norm": 2.23376202583313,
      "learning_rate": 1.6976445396145612e-05,
      "loss": 0.09670616686344147,
      "num_input_tokens_seen": 371024,
      "step": 359,
      "train_runtime": 993.0666,
      "train_tokens_per_second": 373.614
    },
    {
      "epoch": 0.15386259215728176,
      "grad_norm": 3.240729808807373,
      "learning_rate": 1.6967880085653106e-05,
      "loss": 0.084111288189888,
      "num_input_tokens_seen": 372068,
      "step": 360,
      "train_runtime": 995.2213,
      "train_tokens_per_second": 373.855
    },
    {
      "epoch": 0.15428998824660756,
      "grad_norm": 2.244142532348633,
      "learning_rate": 1.69593147751606e-05,
      "loss": 0.10998359322547913,
      "num_input_tokens_seen": 372806,
      "step": 361,
      "train_runtime": 1011.2556,
      "train_tokens_per_second": 368.657
    },
    {
      "epoch": 0.15471738433593332,
      "grad_norm": 2.3793675899505615,
      "learning_rate": 1.6950749464668096e-05,
      "loss": 0.13240720331668854,
      "num_input_tokens_seen": 373880,
      "step": 362,
      "train_runtime": 1013.4298,
      "train_tokens_per_second": 368.925
    },
    {
      "epoch": 0.15514478042525912,
      "grad_norm": 2.0753800868988037,
      "learning_rate": 1.694218415417559e-05,
      "loss": 0.15693649649620056,
      "num_input_tokens_seen": 375686,
      "step": 363,
      "train_runtime": 1015.8122,
      "train_tokens_per_second": 369.838
    },
    {
      "epoch": 0.15557217651458488,
      "grad_norm": 1.6880438327789307,
      "learning_rate": 1.6933618843683084e-05,
      "loss": 0.10661282390356064,
      "num_input_tokens_seen": 377046,
      "step": 364,
      "train_runtime": 1018.0518,
      "train_tokens_per_second": 370.36
    },
    {
      "epoch": 0.15599957260391067,
      "grad_norm": 3.315276622772217,
      "learning_rate": 1.6925053533190577e-05,
      "loss": 0.24908918142318726,
      "num_input_tokens_seen": 377830,
      "step": 365,
      "train_runtime": 1020.2044,
      "train_tokens_per_second": 370.347
    },
    {
      "epoch": 0.15642696869323647,
      "grad_norm": 2.3675825595855713,
      "learning_rate": 1.6916488222698074e-05,
      "loss": 0.1515953689813614,
      "num_input_tokens_seen": 379008,
      "step": 366,
      "train_runtime": 1022.4297,
      "train_tokens_per_second": 370.693
    },
    {
      "epoch": 0.15685436478256223,
      "grad_norm": 2.8744585514068604,
      "learning_rate": 1.690792291220557e-05,
      "loss": 0.22178775072097778,
      "num_input_tokens_seen": 379960,
      "step": 367,
      "train_runtime": 1024.5824,
      "train_tokens_per_second": 370.844
    },
    {
      "epoch": 0.15728176087188803,
      "grad_norm": 2.148393154144287,
      "learning_rate": 1.6899357601713065e-05,
      "loss": 0.14279040694236755,
      "num_input_tokens_seen": 381224,
      "step": 368,
      "train_runtime": 1026.8449,
      "train_tokens_per_second": 371.258
    },
    {
      "epoch": 0.1577091569612138,
      "grad_norm": 1.769447684288025,
      "learning_rate": 1.6890792291220558e-05,
      "loss": 0.07874423265457153,
      "num_input_tokens_seen": 381954,
      "step": 369,
      "train_runtime": 1028.9727,
      "train_tokens_per_second": 371.199
    },
    {
      "epoch": 0.15813655305053959,
      "grad_norm": 2.392979860305786,
      "learning_rate": 1.6882226980728055e-05,
      "loss": 0.1373487412929535,
      "num_input_tokens_seen": 382930,
      "step": 370,
      "train_runtime": 1031.1022,
      "train_tokens_per_second": 371.379
    },
    {
      "epoch": 0.15856394913986538,
      "grad_norm": 2.0406606197357178,
      "learning_rate": 1.687366167023555e-05,
      "loss": 0.162137970328331,
      "num_input_tokens_seen": 384038,
      "step": 371,
      "train_runtime": 1033.2944,
      "train_tokens_per_second": 371.664
    },
    {
      "epoch": 0.15899134522919114,
      "grad_norm": 2.6472795009613037,
      "learning_rate": 1.6865096359743042e-05,
      "loss": 0.22002795338630676,
      "num_input_tokens_seen": 385318,
      "step": 372,
      "train_runtime": 1035.521,
      "train_tokens_per_second": 372.101
    },
    {
      "epoch": 0.15941874131851694,
      "grad_norm": 1.875220775604248,
      "learning_rate": 1.6856531049250536e-05,
      "loss": 0.11501770466566086,
      "num_input_tokens_seen": 386132,
      "step": 373,
      "train_runtime": 1037.6507,
      "train_tokens_per_second": 372.121
    },
    {
      "epoch": 0.15984613740784273,
      "grad_norm": 2.673085927963257,
      "learning_rate": 1.6847965738758033e-05,
      "loss": 0.11223889887332916,
      "num_input_tokens_seen": 386856,
      "step": 374,
      "train_runtime": 1039.7603,
      "train_tokens_per_second": 372.063
    },
    {
      "epoch": 0.1602735334971685,
      "grad_norm": 2.6425139904022217,
      "learning_rate": 1.6839400428265527e-05,
      "loss": 0.14778855443000793,
      "num_input_tokens_seen": 388868,
      "step": 375,
      "train_runtime": 1042.171,
      "train_tokens_per_second": 373.133
    },
    {
      "epoch": 0.1607009295864943,
      "grad_norm": 2.099555730819702,
      "learning_rate": 1.683083511777302e-05,
      "loss": 0.11406856775283813,
      "num_input_tokens_seen": 389734,
      "step": 376,
      "train_runtime": 1044.2989,
      "train_tokens_per_second": 373.202
    },
    {
      "epoch": 0.16112832567582006,
      "grad_norm": 1.9326131343841553,
      "learning_rate": 1.6822269807280517e-05,
      "loss": 0.1232220008969307,
      "num_input_tokens_seen": 391344,
      "step": 377,
      "train_runtime": 1046.5823,
      "train_tokens_per_second": 373.926
    },
    {
      "epoch": 0.16155572176514585,
      "grad_norm": 3.4754836559295654,
      "learning_rate": 1.681370449678801e-05,
      "loss": 0.11852972209453583,
      "num_input_tokens_seen": 392618,
      "step": 378,
      "train_runtime": 1048.8269,
      "train_tokens_per_second": 374.34
    },
    {
      "epoch": 0.16198311785447164,
      "grad_norm": 2.1071417331695557,
      "learning_rate": 1.6805139186295504e-05,
      "loss": 0.11256693303585052,
      "num_input_tokens_seen": 393650,
      "step": 379,
      "train_runtime": 1050.9793,
      "train_tokens_per_second": 374.555
    },
    {
      "epoch": 0.1624105139437974,
      "grad_norm": 2.711966037750244,
      "learning_rate": 1.6796573875802998e-05,
      "loss": 0.15541034936904907,
      "num_input_tokens_seen": 394766,
      "step": 380,
      "train_runtime": 1053.1638,
      "train_tokens_per_second": 374.838
    },
    {
      "epoch": 0.1628379100331232,
      "grad_norm": 3.433267831802368,
      "learning_rate": 1.6788008565310495e-05,
      "loss": 0.14420881867408752,
      "num_input_tokens_seen": 396408,
      "step": 381,
      "train_runtime": 1055.4668,
      "train_tokens_per_second": 375.576
    },
    {
      "epoch": 0.16326530612244897,
      "grad_norm": 3.8691890239715576,
      "learning_rate": 1.677944325481799e-05,
      "loss": 0.2489038109779358,
      "num_input_tokens_seen": 397408,
      "step": 382,
      "train_runtime": 1057.6301,
      "train_tokens_per_second": 375.753
    },
    {
      "epoch": 0.16369270221177476,
      "grad_norm": 1.9748858213424683,
      "learning_rate": 1.6770877944325482e-05,
      "loss": 0.15205879509449005,
      "num_input_tokens_seen": 398320,
      "step": 383,
      "train_runtime": 1059.767,
      "train_tokens_per_second": 375.856
    },
    {
      "epoch": 0.16412009830110055,
      "grad_norm": 2.9696810245513916,
      "learning_rate": 1.676231263383298e-05,
      "loss": 0.20505930483341217,
      "num_input_tokens_seen": 399114,
      "step": 384,
      "train_runtime": 1061.8344,
      "train_tokens_per_second": 375.872
    },
    {
      "epoch": 0.16454749439042632,
      "grad_norm": 2.7213387489318848,
      "learning_rate": 1.6753747323340473e-05,
      "loss": 0.17969396710395813,
      "num_input_tokens_seen": 400164,
      "step": 385,
      "train_runtime": 1063.9978,
      "train_tokens_per_second": 376.095
    },
    {
      "epoch": 0.1649748904797521,
      "grad_norm": 2.311955213546753,
      "learning_rate": 1.6745182012847966e-05,
      "loss": 0.10836541652679443,
      "num_input_tokens_seen": 401158,
      "step": 386,
      "train_runtime": 1066.1859,
      "train_tokens_per_second": 376.255
    },
    {
      "epoch": 0.1654022865690779,
      "grad_norm": 2.4982223510742188,
      "learning_rate": 1.6736616702355463e-05,
      "loss": 0.15085460245609283,
      "num_input_tokens_seen": 402288,
      "step": 387,
      "train_runtime": 1068.5118,
      "train_tokens_per_second": 376.494
    },
    {
      "epoch": 0.16582968265840367,
      "grad_norm": 2.6735992431640625,
      "learning_rate": 1.6728051391862957e-05,
      "loss": 0.18587106466293335,
      "num_input_tokens_seen": 403888,
      "step": 388,
      "train_runtime": 1070.8446,
      "train_tokens_per_second": 377.168
    },
    {
      "epoch": 0.16625707874772946,
      "grad_norm": 2.192577838897705,
      "learning_rate": 1.671948608137045e-05,
      "loss": 0.09636976569890976,
      "num_input_tokens_seen": 404850,
      "step": 389,
      "train_runtime": 1072.9908,
      "train_tokens_per_second": 377.31
    },
    {
      "epoch": 0.16668447483705523,
      "grad_norm": 2.3680543899536133,
      "learning_rate": 1.6710920770877944e-05,
      "loss": 0.13074041903018951,
      "num_input_tokens_seen": 405874,
      "step": 390,
      "train_runtime": 1075.2256,
      "train_tokens_per_second": 377.478
    },
    {
      "epoch": 0.16711187092638102,
      "grad_norm": 1.714113712310791,
      "learning_rate": 1.670235546038544e-05,
      "loss": 0.11863395571708679,
      "num_input_tokens_seen": 407232,
      "step": 391,
      "train_runtime": 1090.7623,
      "train_tokens_per_second": 373.346
    },
    {
      "epoch": 0.16753926701570682,
      "grad_norm": 2.9740312099456787,
      "learning_rate": 1.6693790149892934e-05,
      "loss": 0.20251265168190002,
      "num_input_tokens_seen": 408164,
      "step": 392,
      "train_runtime": 1092.9454,
      "train_tokens_per_second": 373.453
    },
    {
      "epoch": 0.16796666310503258,
      "grad_norm": 2.2991349697113037,
      "learning_rate": 1.6685224839400428e-05,
      "loss": 0.13541968166828156,
      "num_input_tokens_seen": 409462,
      "step": 393,
      "train_runtime": 1095.17,
      "train_tokens_per_second": 373.88
    },
    {
      "epoch": 0.16839405919435838,
      "grad_norm": 2.483513355255127,
      "learning_rate": 1.6676659528907925e-05,
      "loss": 0.11528028547763824,
      "num_input_tokens_seen": 410536,
      "step": 394,
      "train_runtime": 1097.3303,
      "train_tokens_per_second": 374.123
    },
    {
      "epoch": 0.16882145528368414,
      "grad_norm": 1.8675282001495361,
      "learning_rate": 1.666809421841542e-05,
      "loss": 0.06981010735034943,
      "num_input_tokens_seen": 411574,
      "step": 395,
      "train_runtime": 1099.4897,
      "train_tokens_per_second": 374.332
    },
    {
      "epoch": 0.16924885137300993,
      "grad_norm": 2.5383174419403076,
      "learning_rate": 1.6659528907922912e-05,
      "loss": 0.10572994500398636,
      "num_input_tokens_seen": 412522,
      "step": 396,
      "train_runtime": 1101.7629,
      "train_tokens_per_second": 374.42
    },
    {
      "epoch": 0.16967624746233573,
      "grad_norm": 2.21169376373291,
      "learning_rate": 1.6650963597430406e-05,
      "loss": 0.09584016352891922,
      "num_input_tokens_seen": 413270,
      "step": 397,
      "train_runtime": 1103.8672,
      "train_tokens_per_second": 374.384
    },
    {
      "epoch": 0.1701036435516615,
      "grad_norm": 2.2353615760803223,
      "learning_rate": 1.6642398286937903e-05,
      "loss": 0.14034780859947205,
      "num_input_tokens_seen": 414482,
      "step": 398,
      "train_runtime": 1106.066,
      "train_tokens_per_second": 374.735
    },
    {
      "epoch": 0.1705310396409873,
      "grad_norm": 2.762854814529419,
      "learning_rate": 1.6633832976445396e-05,
      "loss": 0.1473519206047058,
      "num_input_tokens_seen": 415424,
      "step": 399,
      "train_runtime": 1108.2143,
      "train_tokens_per_second": 374.859
    },
    {
      "epoch": 0.17095843573031308,
      "grad_norm": 2.3145453929901123,
      "learning_rate": 1.662526766595289e-05,
      "loss": 0.11077391356229782,
      "num_input_tokens_seen": 416154,
      "step": 400,
      "train_runtime": 1110.3093,
      "train_tokens_per_second": 374.809
    },
    {
      "epoch": 0.17138583181963885,
      "grad_norm": 2.5825376510620117,
      "learning_rate": 1.6616702355460387e-05,
      "loss": 0.17286396026611328,
      "num_input_tokens_seen": 417020,
      "step": 401,
      "train_runtime": 1112.4459,
      "train_tokens_per_second": 374.868
    },
    {
      "epoch": 0.17181322790896464,
      "grad_norm": 2.331242084503174,
      "learning_rate": 1.6608137044967884e-05,
      "loss": 0.12545637786388397,
      "num_input_tokens_seen": 418036,
      "step": 402,
      "train_runtime": 1114.6698,
      "train_tokens_per_second": 375.031
    },
    {
      "epoch": 0.1722406239982904,
      "grad_norm": 1.844581961631775,
      "learning_rate": 1.6599571734475377e-05,
      "loss": 0.10602794587612152,
      "num_input_tokens_seen": 419384,
      "step": 403,
      "train_runtime": 1116.8927,
      "train_tokens_per_second": 375.492
    },
    {
      "epoch": 0.1726680200876162,
      "grad_norm": 2.4596877098083496,
      "learning_rate": 1.659100642398287e-05,
      "loss": 0.17446285486221313,
      "num_input_tokens_seen": 420338,
      "step": 404,
      "train_runtime": 1119.0522,
      "train_tokens_per_second": 375.62
    },
    {
      "epoch": 0.173095416176942,
      "grad_norm": 2.457268238067627,
      "learning_rate": 1.6582441113490368e-05,
      "loss": 0.15112313628196716,
      "num_input_tokens_seen": 421290,
      "step": 405,
      "train_runtime": 1121.2044,
      "train_tokens_per_second": 375.748
    },
    {
      "epoch": 0.17352281226626776,
      "grad_norm": 2.921137809753418,
      "learning_rate": 1.657387580299786e-05,
      "loss": 0.18644827604293823,
      "num_input_tokens_seen": 422422,
      "step": 406,
      "train_runtime": 1123.5813,
      "train_tokens_per_second": 375.96
    },
    {
      "epoch": 0.17395020835559355,
      "grad_norm": 1.8381412029266357,
      "learning_rate": 1.6565310492505355e-05,
      "loss": 0.12325820326805115,
      "num_input_tokens_seen": 423902,
      "step": 407,
      "train_runtime": 1125.9323,
      "train_tokens_per_second": 376.49
    },
    {
      "epoch": 0.17437760444491932,
      "grad_norm": 2.721726655960083,
      "learning_rate": 1.655674518201285e-05,
      "loss": 0.141355961561203,
      "num_input_tokens_seen": 425214,
      "step": 408,
      "train_runtime": 1128.2028,
      "train_tokens_per_second": 376.895
    },
    {
      "epoch": 0.1748050005342451,
      "grad_norm": 1.8676183223724365,
      "learning_rate": 1.6548179871520346e-05,
      "loss": 0.07100579142570496,
      "num_input_tokens_seen": 426314,
      "step": 409,
      "train_runtime": 1130.3418,
      "train_tokens_per_second": 377.155
    },
    {
      "epoch": 0.1752323966235709,
      "grad_norm": 1.9449844360351562,
      "learning_rate": 1.653961456102784e-05,
      "loss": 0.13015475869178772,
      "num_input_tokens_seen": 427306,
      "step": 410,
      "train_runtime": 1132.4993,
      "train_tokens_per_second": 377.312
    },
    {
      "epoch": 0.17565979271289667,
      "grad_norm": 2.214311122894287,
      "learning_rate": 1.6531049250535333e-05,
      "loss": 0.11398934572935104,
      "num_input_tokens_seen": 428258,
      "step": 411,
      "train_runtime": 1134.6982,
      "train_tokens_per_second": 377.42
    },
    {
      "epoch": 0.17608718880222246,
      "grad_norm": 2.17134428024292,
      "learning_rate": 1.652248394004283e-05,
      "loss": 0.1376045197248459,
      "num_input_tokens_seen": 429488,
      "step": 412,
      "train_runtime": 1136.9354,
      "train_tokens_per_second": 377.759
    },
    {
      "epoch": 0.17651458489154825,
      "grad_norm": 2.7899832725524902,
      "learning_rate": 1.6513918629550323e-05,
      "loss": 0.19155895709991455,
      "num_input_tokens_seen": 430388,
      "step": 413,
      "train_runtime": 1139.0743,
      "train_tokens_per_second": 377.84
    },
    {
      "epoch": 0.17694198098087402,
      "grad_norm": 3.103749990463257,
      "learning_rate": 1.6505353319057817e-05,
      "loss": 0.2512103319168091,
      "num_input_tokens_seen": 431344,
      "step": 414,
      "train_runtime": 1141.2825,
      "train_tokens_per_second": 377.947
    },
    {
      "epoch": 0.1773693770701998,
      "grad_norm": 1.9968218803405762,
      "learning_rate": 1.649678800856531e-05,
      "loss": 0.1737058460712433,
      "num_input_tokens_seen": 432614,
      "step": 415,
      "train_runtime": 1143.4976,
      "train_tokens_per_second": 378.325
    },
    {
      "epoch": 0.17779677315952558,
      "grad_norm": 2.296706199645996,
      "learning_rate": 1.6488222698072807e-05,
      "loss": 0.1076778918504715,
      "num_input_tokens_seen": 433394,
      "step": 416,
      "train_runtime": 1145.6208,
      "train_tokens_per_second": 378.305
    },
    {
      "epoch": 0.17822416924885137,
      "grad_norm": 2.9021735191345215,
      "learning_rate": 1.64796573875803e-05,
      "loss": 0.19774946570396423,
      "num_input_tokens_seen": 434194,
      "step": 417,
      "train_runtime": 1147.8553,
      "train_tokens_per_second": 378.265
    },
    {
      "epoch": 0.17865156533817717,
      "grad_norm": 2.333594560623169,
      "learning_rate": 1.6471092077087795e-05,
      "loss": 0.17066164314746857,
      "num_input_tokens_seen": 435574,
      "step": 418,
      "train_runtime": 1150.0604,
      "train_tokens_per_second": 378.74
    },
    {
      "epoch": 0.17907896142750293,
      "grad_norm": 1.9304267168045044,
      "learning_rate": 1.646252676659529e-05,
      "loss": 0.09474833309650421,
      "num_input_tokens_seen": 436530,
      "step": 419,
      "train_runtime": 1152.1721,
      "train_tokens_per_second": 378.876
    },
    {
      "epoch": 0.17950635751682872,
      "grad_norm": 2.0851480960845947,
      "learning_rate": 1.6453961456102785e-05,
      "loss": 0.13132205605506897,
      "num_input_tokens_seen": 437646,
      "step": 420,
      "train_runtime": 1154.3789,
      "train_tokens_per_second": 379.118
    },
    {
      "epoch": 0.17993375360615452,
      "grad_norm": 1.7399849891662598,
      "learning_rate": 1.644539614561028e-05,
      "loss": 0.10575830191373825,
      "num_input_tokens_seen": 439542,
      "step": 421,
      "train_runtime": 1170.4494,
      "train_tokens_per_second": 375.533
    },
    {
      "epoch": 0.18036114969548028,
      "grad_norm": 2.2150840759277344,
      "learning_rate": 1.6436830835117776e-05,
      "loss": 0.10072614252567291,
      "num_input_tokens_seen": 440376,
      "step": 422,
      "train_runtime": 1172.6009,
      "train_tokens_per_second": 375.555
    },
    {
      "epoch": 0.18078854578480608,
      "grad_norm": 2.1866869926452637,
      "learning_rate": 1.642826552462527e-05,
      "loss": 0.1437472552061081,
      "num_input_tokens_seen": 441150,
      "step": 423,
      "train_runtime": 1174.6755,
      "train_tokens_per_second": 375.551
    },
    {
      "epoch": 0.18121594187413184,
      "grad_norm": 2.4555861949920654,
      "learning_rate": 1.6419700214132763e-05,
      "loss": 0.11350082606077194,
      "num_input_tokens_seen": 441932,
      "step": 424,
      "train_runtime": 1176.7756,
      "train_tokens_per_second": 375.545
    },
    {
      "epoch": 0.18164333796345764,
      "grad_norm": 1.7611117362976074,
      "learning_rate": 1.6411134903640256e-05,
      "loss": 0.08462654054164886,
      "num_input_tokens_seen": 442800,
      "step": 425,
      "train_runtime": 1178.9141,
      "train_tokens_per_second": 375.6
    },
    {
      "epoch": 0.18207073405278343,
      "grad_norm": 3.2058231830596924,
      "learning_rate": 1.6402569593147753e-05,
      "loss": 0.19276383519172668,
      "num_input_tokens_seen": 443690,
      "step": 426,
      "train_runtime": 1181.052,
      "train_tokens_per_second": 375.674
    },
    {
      "epoch": 0.1824981301421092,
      "grad_norm": 3.497465133666992,
      "learning_rate": 1.6394004282655247e-05,
      "loss": 0.16812632977962494,
      "num_input_tokens_seen": 445006,
      "step": 427,
      "train_runtime": 1183.296,
      "train_tokens_per_second": 376.073
    },
    {
      "epoch": 0.182925526231435,
      "grad_norm": 3.4818010330200195,
      "learning_rate": 1.638543897216274e-05,
      "loss": 0.22986409068107605,
      "num_input_tokens_seen": 445768,
      "step": 428,
      "train_runtime": 1185.3978,
      "train_tokens_per_second": 376.049
    },
    {
      "epoch": 0.18335292232076075,
      "grad_norm": 2.2981884479522705,
      "learning_rate": 1.6376873661670238e-05,
      "loss": 0.20823177695274353,
      "num_input_tokens_seen": 446872,
      "step": 429,
      "train_runtime": 1187.6032,
      "train_tokens_per_second": 376.281
    },
    {
      "epoch": 0.18378031841008655,
      "grad_norm": 2.7404096126556396,
      "learning_rate": 1.636830835117773e-05,
      "loss": 0.1857355833053589,
      "num_input_tokens_seen": 448346,
      "step": 430,
      "train_runtime": 1189.9017,
      "train_tokens_per_second": 376.792
    },
    {
      "epoch": 0.18420771449941234,
      "grad_norm": 3.5768654346466064,
      "learning_rate": 1.6359743040685225e-05,
      "loss": 0.2002859115600586,
      "num_input_tokens_seen": 449010,
      "step": 431,
      "train_runtime": 1191.9593,
      "train_tokens_per_second": 376.699
    },
    {
      "epoch": 0.1846351105887381,
      "grad_norm": 2.8632609844207764,
      "learning_rate": 1.635117773019272e-05,
      "loss": 0.13714022934436798,
      "num_input_tokens_seen": 449648,
      "step": 432,
      "train_runtime": 1194.0469,
      "train_tokens_per_second": 376.575
    },
    {
      "epoch": 0.1850625066780639,
      "grad_norm": 2.0654280185699463,
      "learning_rate": 1.6342612419700215e-05,
      "loss": 0.14014026522636414,
      "num_input_tokens_seen": 450730,
      "step": 433,
      "train_runtime": 1196.2296,
      "train_tokens_per_second": 376.792
    },
    {
      "epoch": 0.1854899027673897,
      "grad_norm": 2.564251661300659,
      "learning_rate": 1.633404710920771e-05,
      "loss": 0.14233601093292236,
      "num_input_tokens_seen": 451750,
      "step": 434,
      "train_runtime": 1198.3737,
      "train_tokens_per_second": 376.969
    },
    {
      "epoch": 0.18591729885671546,
      "grad_norm": 2.1986083984375,
      "learning_rate": 1.6325481798715202e-05,
      "loss": 0.1209622323513031,
      "num_input_tokens_seen": 452766,
      "step": 435,
      "train_runtime": 1200.5229,
      "train_tokens_per_second": 377.141
    },
    {
      "epoch": 0.18634469494604125,
      "grad_norm": 2.6068625450134277,
      "learning_rate": 1.63169164882227e-05,
      "loss": 0.09353704750537872,
      "num_input_tokens_seen": 453596,
      "step": 436,
      "train_runtime": 1202.6806,
      "train_tokens_per_second": 377.154
    },
    {
      "epoch": 0.18677209103536702,
      "grad_norm": 2.098538637161255,
      "learning_rate": 1.6308351177730196e-05,
      "loss": 0.17097875475883484,
      "num_input_tokens_seen": 454666,
      "step": 437,
      "train_runtime": 1204.8642,
      "train_tokens_per_second": 377.359
    },
    {
      "epoch": 0.1871994871246928,
      "grad_norm": 2.5986549854278564,
      "learning_rate": 1.629978586723769e-05,
      "loss": 0.1429862678050995,
      "num_input_tokens_seen": 455646,
      "step": 438,
      "train_runtime": 1207.0235,
      "train_tokens_per_second": 377.496
    },
    {
      "epoch": 0.1876268832140186,
      "grad_norm": 1.8210711479187012,
      "learning_rate": 1.6291220556745184e-05,
      "loss": 0.12008396536111832,
      "num_input_tokens_seen": 457336,
      "step": 439,
      "train_runtime": 1209.3288,
      "train_tokens_per_second": 378.173
    },
    {
      "epoch": 0.18805427930334437,
      "grad_norm": 2.347365617752075,
      "learning_rate": 1.6282655246252677e-05,
      "loss": 0.15392476320266724,
      "num_input_tokens_seen": 458736,
      "step": 440,
      "train_runtime": 1211.5722,
      "train_tokens_per_second": 378.629
    },
    {
      "epoch": 0.18848167539267016,
      "grad_norm": 2.3482296466827393,
      "learning_rate": 1.6274089935760174e-05,
      "loss": 0.11621996760368347,
      "num_input_tokens_seen": 459680,
      "step": 441,
      "train_runtime": 1213.7147,
      "train_tokens_per_second": 378.738
    },
    {
      "epoch": 0.18890907148199593,
      "grad_norm": 1.9932045936584473,
      "learning_rate": 1.6265524625267668e-05,
      "loss": 0.12326683104038239,
      "num_input_tokens_seen": 460690,
      "step": 442,
      "train_runtime": 1215.8646,
      "train_tokens_per_second": 378.899
    },
    {
      "epoch": 0.18933646757132172,
      "grad_norm": 1.4386119842529297,
      "learning_rate": 1.625695931477516e-05,
      "loss": 0.044399235397577286,
      "num_input_tokens_seen": 461756,
      "step": 443,
      "train_runtime": 1218.0943,
      "train_tokens_per_second": 379.081
    },
    {
      "epoch": 0.1897638636606475,
      "grad_norm": 1.9361941814422607,
      "learning_rate": 1.6248394004282658e-05,
      "loss": 0.09412224590778351,
      "num_input_tokens_seen": 462770,
      "step": 444,
      "train_runtime": 1220.3056,
      "train_tokens_per_second": 379.225
    },
    {
      "epoch": 0.19019125974997328,
      "grad_norm": 1.645391583442688,
      "learning_rate": 1.6239828693790152e-05,
      "loss": 0.07834768295288086,
      "num_input_tokens_seen": 463976,
      "step": 445,
      "train_runtime": 1222.6044,
      "train_tokens_per_second": 379.498
    },
    {
      "epoch": 0.19061865583929907,
      "grad_norm": 2.215719223022461,
      "learning_rate": 1.6231263383297645e-05,
      "loss": 0.1422099769115448,
      "num_input_tokens_seen": 464878,
      "step": 446,
      "train_runtime": 1224.7669,
      "train_tokens_per_second": 379.564
    },
    {
      "epoch": 0.19104605192862487,
      "grad_norm": 2.6253433227539062,
      "learning_rate": 1.6222698072805142e-05,
      "loss": 0.15090899169445038,
      "num_input_tokens_seen": 465974,
      "step": 447,
      "train_runtime": 1226.9853,
      "train_tokens_per_second": 379.771
    },
    {
      "epoch": 0.19147344801795063,
      "grad_norm": 2.275822639465332,
      "learning_rate": 1.6214132762312636e-05,
      "loss": 0.1707189381122589,
      "num_input_tokens_seen": 467186,
      "step": 448,
      "train_runtime": 1229.2106,
      "train_tokens_per_second": 380.07
    },
    {
      "epoch": 0.19190084410727642,
      "grad_norm": 2.0672998428344727,
      "learning_rate": 1.620556745182013e-05,
      "loss": 0.12443550676107407,
      "num_input_tokens_seen": 468308,
      "step": 449,
      "train_runtime": 1231.403,
      "train_tokens_per_second": 380.304
    },
    {
      "epoch": 0.1923282401966022,
      "grad_norm": 1.959238886833191,
      "learning_rate": 1.6197002141327623e-05,
      "loss": 0.11623205244541168,
      "num_input_tokens_seen": 469070,
      "step": 450,
      "train_runtime": 1233.4988,
      "train_tokens_per_second": 380.276
    },
    {
      "epoch": 0.19275563628592798,
      "grad_norm": 3.1793341636657715,
      "learning_rate": 1.618843683083512e-05,
      "loss": 0.16129718720912933,
      "num_input_tokens_seen": 469884,
      "step": 451,
      "train_runtime": 1248.8858,
      "train_tokens_per_second": 376.243
    },
    {
      "epoch": 0.19318303237525378,
      "grad_norm": 2.8345463275909424,
      "learning_rate": 1.6179871520342614e-05,
      "loss": 0.10066317766904831,
      "num_input_tokens_seen": 470606,
      "step": 452,
      "train_runtime": 1250.9653,
      "train_tokens_per_second": 376.194
    },
    {
      "epoch": 0.19361042846457954,
      "grad_norm": 2.61457896232605,
      "learning_rate": 1.6171306209850107e-05,
      "loss": 0.1653885543346405,
      "num_input_tokens_seen": 471720,
      "step": 453,
      "train_runtime": 1253.1161,
      "train_tokens_per_second": 376.438
    },
    {
      "epoch": 0.19403782455390534,
      "grad_norm": 2.6919150352478027,
      "learning_rate": 1.6162740899357604e-05,
      "loss": 0.14260752499103546,
      "num_input_tokens_seen": 472782,
      "step": 454,
      "train_runtime": 1255.3006,
      "train_tokens_per_second": 376.629
    },
    {
      "epoch": 0.1944652206432311,
      "grad_norm": 3.671886920928955,
      "learning_rate": 1.6154175588865098e-05,
      "loss": 0.18697695434093475,
      "num_input_tokens_seen": 473422,
      "step": 455,
      "train_runtime": 1257.3285,
      "train_tokens_per_second": 376.53
    },
    {
      "epoch": 0.1948926167325569,
      "grad_norm": 2.6549768447875977,
      "learning_rate": 1.614561027837259e-05,
      "loss": 0.15143200755119324,
      "num_input_tokens_seen": 474068,
      "step": 456,
      "train_runtime": 1259.3935,
      "train_tokens_per_second": 376.426
    },
    {
      "epoch": 0.1953200128218827,
      "grad_norm": 2.3907783031463623,
      "learning_rate": 1.6137044967880085e-05,
      "loss": 0.0999782383441925,
      "num_input_tokens_seen": 475158,
      "step": 457,
      "train_runtime": 1261.5771,
      "train_tokens_per_second": 376.638
    },
    {
      "epoch": 0.19574740891120845,
      "grad_norm": 2.603909730911255,
      "learning_rate": 1.6128479657387582e-05,
      "loss": 0.16652268171310425,
      "num_input_tokens_seen": 476750,
      "step": 458,
      "train_runtime": 1263.864,
      "train_tokens_per_second": 377.216
    },
    {
      "epoch": 0.19617480500053425,
      "grad_norm": 2.89782452583313,
      "learning_rate": 1.6119914346895076e-05,
      "loss": 0.1588645577430725,
      "num_input_tokens_seen": 477686,
      "step": 459,
      "train_runtime": 1265.9908,
      "train_tokens_per_second": 377.322
    },
    {
      "epoch": 0.19660220108986004,
      "grad_norm": 3.2478628158569336,
      "learning_rate": 1.611134903640257e-05,
      "loss": 0.13660337030887604,
      "num_input_tokens_seen": 478550,
      "step": 460,
      "train_runtime": 1268.0816,
      "train_tokens_per_second": 377.381
    },
    {
      "epoch": 0.1970295971791858,
      "grad_norm": 2.897303819656372,
      "learning_rate": 1.6102783725910066e-05,
      "loss": 0.2215740978717804,
      "num_input_tokens_seen": 479448,
      "step": 461,
      "train_runtime": 1270.2083,
      "train_tokens_per_second": 377.456
    },
    {
      "epoch": 0.1974569932685116,
      "grad_norm": 2.5118114948272705,
      "learning_rate": 1.609421841541756e-05,
      "loss": 0.12003548443317413,
      "num_input_tokens_seen": 480304,
      "step": 462,
      "train_runtime": 1272.4419,
      "train_tokens_per_second": 377.466
    },
    {
      "epoch": 0.19788438935783736,
      "grad_norm": 1.744229793548584,
      "learning_rate": 1.6085653104925053e-05,
      "loss": 0.1065872311592102,
      "num_input_tokens_seen": 481772,
      "step": 463,
      "train_runtime": 1274.7272,
      "train_tokens_per_second": 377.941
    },
    {
      "epoch": 0.19831178544716316,
      "grad_norm": 3.23005747795105,
      "learning_rate": 1.607708779443255e-05,
      "loss": 0.12318828701972961,
      "num_input_tokens_seen": 482828,
      "step": 464,
      "train_runtime": 1276.9316,
      "train_tokens_per_second": 378.116
    },
    {
      "epoch": 0.19873918153648895,
      "grad_norm": 2.553464651107788,
      "learning_rate": 1.6068522483940044e-05,
      "loss": 0.10046009719371796,
      "num_input_tokens_seen": 483550,
      "step": 465,
      "train_runtime": 1278.9852,
      "train_tokens_per_second": 378.073
    },
    {
      "epoch": 0.19916657762581472,
      "grad_norm": 2.9752719402313232,
      "learning_rate": 1.6059957173447537e-05,
      "loss": 0.16212669014930725,
      "num_input_tokens_seen": 484462,
      "step": 466,
      "train_runtime": 1281.1392,
      "train_tokens_per_second": 378.149
    },
    {
      "epoch": 0.1995939737151405,
      "grad_norm": 2.731992483139038,
      "learning_rate": 1.605139186295503e-05,
      "loss": 0.15849679708480835,
      "num_input_tokens_seen": 485672,
      "step": 467,
      "train_runtime": 1283.3548,
      "train_tokens_per_second": 378.439
    },
    {
      "epoch": 0.20002136980446628,
      "grad_norm": 1.613195538520813,
      "learning_rate": 1.6042826552462528e-05,
      "loss": 0.09022671729326248,
      "num_input_tokens_seen": 486932,
      "step": 468,
      "train_runtime": 1285.5648,
      "train_tokens_per_second": 378.769
    },
    {
      "epoch": 0.20002136980446628,
      "eval_loss": 0.5228654742240906,
      "eval_runtime": 58.1699,
      "eval_samples_per_second": 17.157,
      "eval_steps_per_second": 8.578,
      "num_input_tokens_seen": 486932,
      "step": 468
    },
    {
      "epoch": 0.20044876589379207,
      "grad_norm": 2.7253291606903076,
      "learning_rate": 1.603426124197002e-05,
      "loss": 0.23014238476753235,
      "num_input_tokens_seen": 487912,
      "step": 469,
      "train_runtime": 1345.8997,
      "train_tokens_per_second": 362.517
    },
    {
      "epoch": 0.20087616198311786,
      "grad_norm": 2.572406530380249,
      "learning_rate": 1.602569593147752e-05,
      "loss": 0.11572389304637909,
      "num_input_tokens_seen": 488624,
      "step": 470,
      "train_runtime": 1347.958,
      "train_tokens_per_second": 362.492
    },
    {
      "epoch": 0.20130355807244363,
      "grad_norm": 2.066358804702759,
      "learning_rate": 1.6017130620985012e-05,
      "loss": 0.12841399013996124,
      "num_input_tokens_seen": 489490,
      "step": 471,
      "train_runtime": 1350.133,
      "train_tokens_per_second": 362.549
    },
    {
      "epoch": 0.20173095416176942,
      "grad_norm": 2.618079423904419,
      "learning_rate": 1.600856531049251e-05,
      "loss": 0.17785006761550903,
      "num_input_tokens_seen": 490284,
      "step": 472,
      "train_runtime": 1352.1913,
      "train_tokens_per_second": 362.585
    },
    {
      "epoch": 0.20215835025109521,
      "grad_norm": 2.171987771987915,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 0.11661925911903381,
      "num_input_tokens_seen": 491452,
      "step": 473,
      "train_runtime": 1354.4431,
      "train_tokens_per_second": 362.844
    },
    {
      "epoch": 0.20258574634042098,
      "grad_norm": 2.0648534297943115,
      "learning_rate": 1.5991434689507496e-05,
      "loss": 0.1089189425110817,
      "num_input_tokens_seen": 492384,
      "step": 474,
      "train_runtime": 1356.5623,
      "train_tokens_per_second": 362.965
    },
    {
      "epoch": 0.20301314242974677,
      "grad_norm": 1.4901224374771118,
      "learning_rate": 1.598286937901499e-05,
      "loss": 0.10635621845722198,
      "num_input_tokens_seen": 493588,
      "step": 475,
      "train_runtime": 1358.7887,
      "train_tokens_per_second": 363.256
    },
    {
      "epoch": 0.20344053851907254,
      "grad_norm": 1.9994614124298096,
      "learning_rate": 1.5974304068522487e-05,
      "loss": 0.13134728372097015,
      "num_input_tokens_seen": 495088,
      "step": 476,
      "train_runtime": 1361.0765,
      "train_tokens_per_second": 363.747
    },
    {
      "epoch": 0.20386793460839833,
      "grad_norm": 2.1291215419769287,
      "learning_rate": 1.596573875802998e-05,
      "loss": 0.16656774282455444,
      "num_input_tokens_seen": 496728,
      "step": 477,
      "train_runtime": 1363.4104,
      "train_tokens_per_second": 364.328
    },
    {
      "epoch": 0.20429533069772413,
      "grad_norm": 1.7206581830978394,
      "learning_rate": 1.5957173447537474e-05,
      "loss": 0.09397691488265991,
      "num_input_tokens_seen": 497664,
      "step": 478,
      "train_runtime": 1365.554,
      "train_tokens_per_second": 364.441
    },
    {
      "epoch": 0.2047227267870499,
      "grad_norm": 2.3613126277923584,
      "learning_rate": 1.594860813704497e-05,
      "loss": 0.12450061738491058,
      "num_input_tokens_seen": 498506,
      "step": 479,
      "train_runtime": 1367.6946,
      "train_tokens_per_second": 364.486
    },
    {
      "epoch": 0.20515012287637568,
      "grad_norm": 1.9107856750488281,
      "learning_rate": 1.5940042826552465e-05,
      "loss": 0.0831819698214531,
      "num_input_tokens_seen": 499478,
      "step": 480,
      "train_runtime": 1369.8658,
      "train_tokens_per_second": 364.618
    },
    {
      "epoch": 0.20557751896570145,
      "grad_norm": 2.1192493438720703,
      "learning_rate": 1.5931477516059958e-05,
      "loss": 0.12226656824350357,
      "num_input_tokens_seen": 500350,
      "step": 481,
      "train_runtime": 1385.3664,
      "train_tokens_per_second": 361.168
    },
    {
      "epoch": 0.20600491505502724,
      "grad_norm": 2.522186279296875,
      "learning_rate": 1.5922912205567455e-05,
      "loss": 0.16046814620494843,
      "num_input_tokens_seen": 501142,
      "step": 482,
      "train_runtime": 1387.5782,
      "train_tokens_per_second": 361.163
    },
    {
      "epoch": 0.20643231114435304,
      "grad_norm": 2.28838849067688,
      "learning_rate": 1.591434689507495e-05,
      "loss": 0.1571349799633026,
      "num_input_tokens_seen": 502198,
      "step": 483,
      "train_runtime": 1389.7598,
      "train_tokens_per_second": 361.356
    },
    {
      "epoch": 0.2068597072336788,
      "grad_norm": 2.136685609817505,
      "learning_rate": 1.5905781584582442e-05,
      "loss": 0.131611168384552,
      "num_input_tokens_seen": 503676,
      "step": 484,
      "train_runtime": 1391.9991,
      "train_tokens_per_second": 361.836
    },
    {
      "epoch": 0.2072871033230046,
      "grad_norm": 2.999263048171997,
      "learning_rate": 1.5897216274089936e-05,
      "loss": 0.2090913951396942,
      "num_input_tokens_seen": 504508,
      "step": 485,
      "train_runtime": 1394.0665,
      "train_tokens_per_second": 361.897
    },
    {
      "epoch": 0.2077144994123304,
      "grad_norm": 1.8556879758834839,
      "learning_rate": 1.5888650963597433e-05,
      "loss": 0.10524851083755493,
      "num_input_tokens_seen": 505448,
      "step": 486,
      "train_runtime": 1396.1995,
      "train_tokens_per_second": 362.017
    },
    {
      "epoch": 0.20814189550165615,
      "grad_norm": 3.578195095062256,
      "learning_rate": 1.5880085653104926e-05,
      "loss": 0.13232721388339996,
      "num_input_tokens_seen": 506284,
      "step": 487,
      "train_runtime": 1398.3699,
      "train_tokens_per_second": 362.053
    },
    {
      "epoch": 0.20856929159098195,
      "grad_norm": 2.8743364810943604,
      "learning_rate": 1.587152034261242e-05,
      "loss": 0.22635917365550995,
      "num_input_tokens_seen": 507260,
      "step": 488,
      "train_runtime": 1400.5432,
      "train_tokens_per_second": 362.188
    },
    {
      "epoch": 0.2089966876803077,
      "grad_norm": 2.805987596511841,
      "learning_rate": 1.5862955032119917e-05,
      "loss": 0.08352351933717728,
      "num_input_tokens_seen": 508176,
      "step": 489,
      "train_runtime": 1402.6622,
      "train_tokens_per_second": 362.294
    },
    {
      "epoch": 0.2094240837696335,
      "grad_norm": 1.606923222541809,
      "learning_rate": 1.585438972162741e-05,
      "loss": 0.1012442335486412,
      "num_input_tokens_seen": 509374,
      "step": 490,
      "train_runtime": 1404.8896,
      "train_tokens_per_second": 362.572
    },
    {
      "epoch": 0.2098514798589593,
      "grad_norm": 2.432309865951538,
      "learning_rate": 1.5845824411134904e-05,
      "loss": 0.15776558220386505,
      "num_input_tokens_seen": 510504,
      "step": 491,
      "train_runtime": 1407.222,
      "train_tokens_per_second": 362.774
    },
    {
      "epoch": 0.21027887594828507,
      "grad_norm": 2.092952251434326,
      "learning_rate": 1.5837259100642398e-05,
      "loss": 0.15795929729938507,
      "num_input_tokens_seen": 512370,
      "step": 492,
      "train_runtime": 1409.5689,
      "train_tokens_per_second": 363.494
    },
    {
      "epoch": 0.21070627203761086,
      "grad_norm": 2.1159043312072754,
      "learning_rate": 1.5828693790149895e-05,
      "loss": 0.14126335084438324,
      "num_input_tokens_seen": 513418,
      "step": 493,
      "train_runtime": 1411.8495,
      "train_tokens_per_second": 363.649
    },
    {
      "epoch": 0.21113366812693665,
      "grad_norm": 2.370260715484619,
      "learning_rate": 1.5820128479657388e-05,
      "loss": 0.08948801457881927,
      "num_input_tokens_seen": 514308,
      "step": 494,
      "train_runtime": 1414.0793,
      "train_tokens_per_second": 363.705
    },
    {
      "epoch": 0.21156106421626242,
      "grad_norm": 2.643071413040161,
      "learning_rate": 1.5811563169164882e-05,
      "loss": 0.08167178183794022,
      "num_input_tokens_seen": 515252,
      "step": 495,
      "train_runtime": 1416.2863,
      "train_tokens_per_second": 363.805
    },
    {
      "epoch": 0.2119884603055882,
      "grad_norm": 2.3257827758789062,
      "learning_rate": 1.580299785867238e-05,
      "loss": 0.13746702671051025,
      "num_input_tokens_seen": 516116,
      "step": 496,
      "train_runtime": 1418.4188,
      "train_tokens_per_second": 363.867
    },
    {
      "epoch": 0.21241585639491398,
      "grad_norm": 2.404764175415039,
      "learning_rate": 1.5794432548179872e-05,
      "loss": 0.22356030344963074,
      "num_input_tokens_seen": 517416,
      "step": 497,
      "train_runtime": 1420.7315,
      "train_tokens_per_second": 364.19
    },
    {
      "epoch": 0.21284325248423977,
      "grad_norm": 1.8176261186599731,
      "learning_rate": 1.5785867237687366e-05,
      "loss": 0.10514774918556213,
      "num_input_tokens_seen": 518206,
      "step": 498,
      "train_runtime": 1422.824,
      "train_tokens_per_second": 364.209
    },
    {
      "epoch": 0.21327064857356556,
      "grad_norm": 2.6928937435150146,
      "learning_rate": 1.5777301927194863e-05,
      "loss": 0.16498826444149017,
      "num_input_tokens_seen": 519112,
      "step": 499,
      "train_runtime": 1424.944,
      "train_tokens_per_second": 364.303
    },
    {
      "epoch": 0.21369804466289133,
      "grad_norm": 3.530243396759033,
      "learning_rate": 1.5768736616702356e-05,
      "loss": 0.15228629112243652,
      "num_input_tokens_seen": 520030,
      "step": 500,
      "train_runtime": 1427.1623,
      "train_tokens_per_second": 364.38
    },
    {
      "epoch": 0.21412544075221712,
      "grad_norm": 2.3968775272369385,
      "learning_rate": 1.576017130620985e-05,
      "loss": 0.1445412039756775,
      "num_input_tokens_seen": 521166,
      "step": 501,
      "train_runtime": 1429.4257,
      "train_tokens_per_second": 364.598
    },
    {
      "epoch": 0.2145528368415429,
      "grad_norm": 1.3246902227401733,
      "learning_rate": 1.5751605995717344e-05,
      "loss": 0.07449869066476822,
      "num_input_tokens_seen": 522844,
      "step": 502,
      "train_runtime": 1431.7055,
      "train_tokens_per_second": 365.19
    },
    {
      "epoch": 0.21498023293086868,
      "grad_norm": 3.094409942626953,
      "learning_rate": 1.574304068522484e-05,
      "loss": 0.1486973911523819,
      "num_input_tokens_seen": 523784,
      "step": 503,
      "train_runtime": 1433.8324,
      "train_tokens_per_second": 365.304
    },
    {
      "epoch": 0.21540762902019447,
      "grad_norm": 2.6681461334228516,
      "learning_rate": 1.5734475374732338e-05,
      "loss": 0.16639576852321625,
      "num_input_tokens_seen": 524676,
      "step": 504,
      "train_runtime": 1435.9518,
      "train_tokens_per_second": 365.386
    },
    {
      "epoch": 0.21583502510952024,
      "grad_norm": 1.740444302558899,
      "learning_rate": 1.572591006423983e-05,
      "loss": 0.09387188404798508,
      "num_input_tokens_seen": 525942,
      "step": 505,
      "train_runtime": 1438.2004,
      "train_tokens_per_second": 365.695
    },
    {
      "epoch": 0.21626242119884603,
      "grad_norm": 1.6638593673706055,
      "learning_rate": 1.5717344753747325e-05,
      "loss": 0.09179219603538513,
      "num_input_tokens_seen": 526914,
      "step": 506,
      "train_runtime": 1440.317,
      "train_tokens_per_second": 365.832
    },
    {
      "epoch": 0.21668981728817183,
      "grad_norm": 2.8334851264953613,
      "learning_rate": 1.5708779443254822e-05,
      "loss": 0.1521914005279541,
      "num_input_tokens_seen": 527780,
      "step": 507,
      "train_runtime": 1442.4452,
      "train_tokens_per_second": 365.893
    },
    {
      "epoch": 0.2171172133774976,
      "grad_norm": 2.0652244091033936,
      "learning_rate": 1.5700214132762315e-05,
      "loss": 0.10250615328550339,
      "num_input_tokens_seen": 529420,
      "step": 508,
      "train_runtime": 1444.7617,
      "train_tokens_per_second": 366.441
    },
    {
      "epoch": 0.21754460946682339,
      "grad_norm": 3.1617624759674072,
      "learning_rate": 1.569164882226981e-05,
      "loss": 0.17184650897979736,
      "num_input_tokens_seen": 530426,
      "step": 509,
      "train_runtime": 1446.8963,
      "train_tokens_per_second": 366.596
    },
    {
      "epoch": 0.21797200555614915,
      "grad_norm": 2.2681519985198975,
      "learning_rate": 1.5683083511777302e-05,
      "loss": 0.17457982897758484,
      "num_input_tokens_seen": 531432,
      "step": 510,
      "train_runtime": 1449.0906,
      "train_tokens_per_second": 366.735
    },
    {
      "epoch": 0.21839940164547494,
      "grad_norm": 2.165870428085327,
      "learning_rate": 1.56745182012848e-05,
      "loss": 0.10948702692985535,
      "num_input_tokens_seen": 532300,
      "step": 511,
      "train_runtime": 1464.7294,
      "train_tokens_per_second": 363.412
    },
    {
      "epoch": 0.21882679773480074,
      "grad_norm": 2.145991802215576,
      "learning_rate": 1.5665952890792293e-05,
      "loss": 0.138885498046875,
      "num_input_tokens_seen": 533464,
      "step": 512,
      "train_runtime": 1466.9261,
      "train_tokens_per_second": 363.661
    },
    {
      "epoch": 0.2192541938241265,
      "grad_norm": 1.6592544317245483,
      "learning_rate": 1.5657387580299787e-05,
      "loss": 0.1290833055973053,
      "num_input_tokens_seen": 534688,
      "step": 513,
      "train_runtime": 1469.1604,
      "train_tokens_per_second": 363.941
    },
    {
      "epoch": 0.2196815899134523,
      "grad_norm": 3.0241329669952393,
      "learning_rate": 1.5648822269807284e-05,
      "loss": 0.16119757294654846,
      "num_input_tokens_seen": 535718,
      "step": 514,
      "train_runtime": 1471.3153,
      "train_tokens_per_second": 364.108
    },
    {
      "epoch": 0.22010898600277806,
      "grad_norm": 2.77247953414917,
      "learning_rate": 1.5640256959314777e-05,
      "loss": 0.1776650995016098,
      "num_input_tokens_seen": 536774,
      "step": 515,
      "train_runtime": 1473.5648,
      "train_tokens_per_second": 364.269
    },
    {
      "epoch": 0.22053638209210386,
      "grad_norm": 3.0152673721313477,
      "learning_rate": 1.563169164882227e-05,
      "loss": 0.09958123415708542,
      "num_input_tokens_seen": 537520,
      "step": 516,
      "train_runtime": 1475.6796,
      "train_tokens_per_second": 364.253
    },
    {
      "epoch": 0.22096377818142965,
      "grad_norm": 2.4523041248321533,
      "learning_rate": 1.5623126338329764e-05,
      "loss": 0.13446536660194397,
      "num_input_tokens_seen": 538532,
      "step": 517,
      "train_runtime": 1477.8692,
      "train_tokens_per_second": 364.398
    },
    {
      "epoch": 0.22139117427075541,
      "grad_norm": 2.4473934173583984,
      "learning_rate": 1.561456102783726e-05,
      "loss": 0.1699734628200531,
      "num_input_tokens_seen": 539472,
      "step": 518,
      "train_runtime": 1479.9924,
      "train_tokens_per_second": 364.51
    },
    {
      "epoch": 0.2218185703600812,
      "grad_norm": 2.0424158573150635,
      "learning_rate": 1.5605995717344755e-05,
      "loss": 0.13223159313201904,
      "num_input_tokens_seen": 540582,
      "step": 519,
      "train_runtime": 1482.1723,
      "train_tokens_per_second": 364.723
    },
    {
      "epoch": 0.222245966449407,
      "grad_norm": 2.1409871578216553,
      "learning_rate": 1.559743040685225e-05,
      "loss": 0.139509379863739,
      "num_input_tokens_seen": 541854,
      "step": 520,
      "train_runtime": 1484.3802,
      "train_tokens_per_second": 365.037
    },
    {
      "epoch": 0.22267336253873277,
      "grad_norm": 2.0753836631774902,
      "learning_rate": 1.5588865096359745e-05,
      "loss": 0.14061060547828674,
      "num_input_tokens_seen": 543082,
      "step": 521,
      "train_runtime": 1486.5904,
      "train_tokens_per_second": 365.321
    },
    {
      "epoch": 0.22310075862805856,
      "grad_norm": 2.510342836380005,
      "learning_rate": 1.558029978586724e-05,
      "loss": 0.16124089062213898,
      "num_input_tokens_seen": 543988,
      "step": 522,
      "train_runtime": 1488.7106,
      "train_tokens_per_second": 365.409
    },
    {
      "epoch": 0.22352815471738433,
      "grad_norm": 2.1258161067962646,
      "learning_rate": 1.5571734475374733e-05,
      "loss": 0.1936700940132141,
      "num_input_tokens_seen": 545036,
      "step": 523,
      "train_runtime": 1490.9556,
      "train_tokens_per_second": 365.562
    },
    {
      "epoch": 0.22395555080671012,
      "grad_norm": 2.6125400066375732,
      "learning_rate": 1.556316916488223e-05,
      "loss": 0.15123610198497772,
      "num_input_tokens_seen": 545972,
      "step": 524,
      "train_runtime": 1493.1623,
      "train_tokens_per_second": 365.648
    },
    {
      "epoch": 0.2243829468960359,
      "grad_norm": 1.7003302574157715,
      "learning_rate": 1.5554603854389723e-05,
      "loss": 0.1328495740890503,
      "num_input_tokens_seen": 547502,
      "step": 525,
      "train_runtime": 1495.4609,
      "train_tokens_per_second": 366.109
    },
    {
      "epoch": 0.22481034298536168,
      "grad_norm": 2.637478828430176,
      "learning_rate": 1.5546038543897217e-05,
      "loss": 0.14079953730106354,
      "num_input_tokens_seen": 548188,
      "step": 526,
      "train_runtime": 1497.5308,
      "train_tokens_per_second": 366.061
    },
    {
      "epoch": 0.22523773907468747,
      "grad_norm": 2.351576566696167,
      "learning_rate": 1.553747323340471e-05,
      "loss": 0.15369859337806702,
      "num_input_tokens_seen": 549974,
      "step": 527,
      "train_runtime": 1499.8867,
      "train_tokens_per_second": 366.677
    },
    {
      "epoch": 0.22566513516401324,
      "grad_norm": 4.461513996124268,
      "learning_rate": 1.5528907922912207e-05,
      "loss": 0.24408143758773804,
      "num_input_tokens_seen": 550696,
      "step": 528,
      "train_runtime": 1502.0254,
      "train_tokens_per_second": 366.636
    },
    {
      "epoch": 0.22609253125333903,
      "grad_norm": 2.423560857772827,
      "learning_rate": 1.55203426124197e-05,
      "loss": 0.16323769092559814,
      "num_input_tokens_seen": 551666,
      "step": 529,
      "train_runtime": 1504.1836,
      "train_tokens_per_second": 366.754
    },
    {
      "epoch": 0.22651992734266482,
      "grad_norm": 1.9746719598770142,
      "learning_rate": 1.5511777301927194e-05,
      "loss": 0.07020363211631775,
      "num_input_tokens_seen": 552572,
      "step": 530,
      "train_runtime": 1506.3013,
      "train_tokens_per_second": 366.84
    },
    {
      "epoch": 0.2269473234319906,
      "grad_norm": 1.710885763168335,
      "learning_rate": 1.550321199143469e-05,
      "loss": 0.16125689446926117,
      "num_input_tokens_seen": 554080,
      "step": 531,
      "train_runtime": 1508.6089,
      "train_tokens_per_second": 367.279
    },
    {
      "epoch": 0.22737471952131638,
      "grad_norm": 1.777014970779419,
      "learning_rate": 1.5494646680942185e-05,
      "loss": 0.1350293755531311,
      "num_input_tokens_seen": 555458,
      "step": 532,
      "train_runtime": 1510.8603,
      "train_tokens_per_second": 367.644
    },
    {
      "epoch": 0.22780211561064218,
      "grad_norm": 2.4799766540527344,
      "learning_rate": 1.548608137044968e-05,
      "loss": 0.11196652799844742,
      "num_input_tokens_seen": 556286,
      "step": 533,
      "train_runtime": 1512.9716,
      "train_tokens_per_second": 367.678
    },
    {
      "epoch": 0.22822951169996794,
      "grad_norm": 2.4436168670654297,
      "learning_rate": 1.5477516059957176e-05,
      "loss": 0.12476056069135666,
      "num_input_tokens_seen": 557110,
      "step": 534,
      "train_runtime": 1515.0552,
      "train_tokens_per_second": 367.716
    },
    {
      "epoch": 0.22865690778929373,
      "grad_norm": 2.0891025066375732,
      "learning_rate": 1.546895074946467e-05,
      "loss": 0.14830750226974487,
      "num_input_tokens_seen": 558288,
      "step": 535,
      "train_runtime": 1517.2946,
      "train_tokens_per_second": 367.95
    },
    {
      "epoch": 0.2290843038786195,
      "grad_norm": 1.3756754398345947,
      "learning_rate": 1.5460385438972163e-05,
      "loss": 0.09467566013336182,
      "num_input_tokens_seen": 559406,
      "step": 536,
      "train_runtime": 1519.4747,
      "train_tokens_per_second": 368.157
    },
    {
      "epoch": 0.2295116999679453,
      "grad_norm": 2.0166375637054443,
      "learning_rate": 1.5451820128479656e-05,
      "loss": 0.11320997774600983,
      "num_input_tokens_seen": 560272,
      "step": 537,
      "train_runtime": 1521.5651,
      "train_tokens_per_second": 368.221
    },
    {
      "epoch": 0.2299390960572711,
      "grad_norm": 2.1064934730529785,
      "learning_rate": 1.5443254817987153e-05,
      "loss": 0.11865454912185669,
      "num_input_tokens_seen": 561538,
      "step": 538,
      "train_runtime": 1523.8207,
      "train_tokens_per_second": 368.507
    },
    {
      "epoch": 0.23036649214659685,
      "grad_norm": 2.8276851177215576,
      "learning_rate": 1.543468950749465e-05,
      "loss": 0.1871943175792694,
      "num_input_tokens_seen": 562386,
      "step": 539,
      "train_runtime": 1525.9483,
      "train_tokens_per_second": 368.549
    },
    {
      "epoch": 0.23079388823592265,
      "grad_norm": 2.6436808109283447,
      "learning_rate": 1.5426124197002144e-05,
      "loss": 0.1718994379043579,
      "num_input_tokens_seen": 563408,
      "step": 540,
      "train_runtime": 1528.1642,
      "train_tokens_per_second": 368.683
    },
    {
      "epoch": 0.2312212843252484,
      "grad_norm": 2.5760653018951416,
      "learning_rate": 1.5417558886509637e-05,
      "loss": 0.1660972386598587,
      "num_input_tokens_seen": 564254,
      "step": 541,
      "train_runtime": 1542.5505,
      "train_tokens_per_second": 365.793
    },
    {
      "epoch": 0.2316486804145742,
      "grad_norm": 2.107659101486206,
      "learning_rate": 1.5408993576017134e-05,
      "loss": 0.18583010137081146,
      "num_input_tokens_seen": 565498,
      "step": 542,
      "train_runtime": 1544.7652,
      "train_tokens_per_second": 366.074
    },
    {
      "epoch": 0.2320760765039,
      "grad_norm": 2.7439117431640625,
      "learning_rate": 1.5400428265524628e-05,
      "loss": 0.08493205159902573,
      "num_input_tokens_seen": 566718,
      "step": 543,
      "train_runtime": 1546.9665,
      "train_tokens_per_second": 366.341
    },
    {
      "epoch": 0.23250347259322576,
      "grad_norm": 2.3122153282165527,
      "learning_rate": 1.539186295503212e-05,
      "loss": 0.16841523349285126,
      "num_input_tokens_seen": 567764,
      "step": 544,
      "train_runtime": 1549.1632,
      "train_tokens_per_second": 366.497
    },
    {
      "epoch": 0.23293086868255156,
      "grad_norm": 1.9390462636947632,
      "learning_rate": 1.5383297644539615e-05,
      "loss": 0.1035780981183052,
      "num_input_tokens_seen": 568674,
      "step": 545,
      "train_runtime": 1551.3064,
      "train_tokens_per_second": 366.577
    },
    {
      "epoch": 0.23335826477187735,
      "grad_norm": 2.66786527633667,
      "learning_rate": 1.5374732334047112e-05,
      "loss": 0.13681650161743164,
      "num_input_tokens_seen": 569620,
      "step": 546,
      "train_runtime": 1553.4202,
      "train_tokens_per_second": 366.688
    },
    {
      "epoch": 0.23378566086120312,
      "grad_norm": 2.642683267593384,
      "learning_rate": 1.5366167023554606e-05,
      "loss": 0.17472638189792633,
      "num_input_tokens_seen": 570572,
      "step": 547,
      "train_runtime": 1555.5739,
      "train_tokens_per_second": 366.792
    },
    {
      "epoch": 0.2342130569505289,
      "grad_norm": 3.1677067279815674,
      "learning_rate": 1.53576017130621e-05,
      "loss": 0.1817154735326767,
      "num_input_tokens_seen": 571354,
      "step": 548,
      "train_runtime": 1557.6595,
      "train_tokens_per_second": 366.803
    },
    {
      "epoch": 0.23464045303985467,
      "grad_norm": 2.4259681701660156,
      "learning_rate": 1.5349036402569596e-05,
      "loss": 0.1345292031764984,
      "num_input_tokens_seen": 572478,
      "step": 549,
      "train_runtime": 1559.8404,
      "train_tokens_per_second": 367.011
    },
    {
      "epoch": 0.23506784912918047,
      "grad_norm": 2.276236057281494,
      "learning_rate": 1.534047109207709e-05,
      "loss": 0.09407582879066467,
      "num_input_tokens_seen": 573442,
      "step": 550,
      "train_runtime": 1561.9913,
      "train_tokens_per_second": 367.122
    },
    {
      "epoch": 0.23549524521850626,
      "grad_norm": 2.1500966548919678,
      "learning_rate": 1.5331905781584583e-05,
      "loss": 0.12902149558067322,
      "num_input_tokens_seen": 574304,
      "step": 551,
      "train_runtime": 1564.1363,
      "train_tokens_per_second": 367.17
    },
    {
      "epoch": 0.23592264130783203,
      "grad_norm": 2.4532110691070557,
      "learning_rate": 1.5323340471092077e-05,
      "loss": 0.14625193178653717,
      "num_input_tokens_seen": 575224,
      "step": 552,
      "train_runtime": 1566.239,
      "train_tokens_per_second": 367.265
    },
    {
      "epoch": 0.23635003739715782,
      "grad_norm": 2.008481025695801,
      "learning_rate": 1.5314775160599574e-05,
      "loss": 0.11644130945205688,
      "num_input_tokens_seen": 576216,
      "step": 553,
      "train_runtime": 1568.719,
      "train_tokens_per_second": 367.316
    },
    {
      "epoch": 0.23677743348648359,
      "grad_norm": 2.1232733726501465,
      "learning_rate": 1.5306209850107068e-05,
      "loss": 0.0997479185461998,
      "num_input_tokens_seen": 577130,
      "step": 554,
      "train_runtime": 1570.8363,
      "train_tokens_per_second": 367.403
    },
    {
      "epoch": 0.23720482957580938,
      "grad_norm": 2.1800730228424072,
      "learning_rate": 1.529764453961456e-05,
      "loss": 0.12549588084220886,
      "num_input_tokens_seen": 578188,
      "step": 555,
      "train_runtime": 1572.9819,
      "train_tokens_per_second": 367.574
    },
    {
      "epoch": 0.23763222566513517,
      "grad_norm": 2.4302923679351807,
      "learning_rate": 1.5289079229122058e-05,
      "loss": 0.11075545102357864,
      "num_input_tokens_seen": 579398,
      "step": 556,
      "train_runtime": 1575.1797,
      "train_tokens_per_second": 367.83
    },
    {
      "epoch": 0.23805962175446094,
      "grad_norm": 3.1985461711883545,
      "learning_rate": 1.528051391862955e-05,
      "loss": 0.12170572578907013,
      "num_input_tokens_seen": 581002,
      "step": 557,
      "train_runtime": 1577.4627,
      "train_tokens_per_second": 368.314
    },
    {
      "epoch": 0.23848701784378673,
      "grad_norm": 2.2686915397644043,
      "learning_rate": 1.5271948608137045e-05,
      "loss": 0.16669438779354095,
      "num_input_tokens_seen": 582120,
      "step": 558,
      "train_runtime": 1579.6284,
      "train_tokens_per_second": 368.517
    },
    {
      "epoch": 0.23891441393311252,
      "grad_norm": 3.0081892013549805,
      "learning_rate": 1.5263383297644542e-05,
      "loss": 0.1628597378730774,
      "num_input_tokens_seen": 582940,
      "step": 559,
      "train_runtime": 1581.7383,
      "train_tokens_per_second": 368.544
    },
    {
      "epoch": 0.2393418100224383,
      "grad_norm": 1.970823884010315,
      "learning_rate": 1.5254817987152036e-05,
      "loss": 0.1337084174156189,
      "num_input_tokens_seen": 583904,
      "step": 560,
      "train_runtime": 1583.8843,
      "train_tokens_per_second": 368.653
    },
    {
      "epoch": 0.23976920611176408,
      "grad_norm": 2.0613250732421875,
      "learning_rate": 1.524625267665953e-05,
      "loss": 0.16636675596237183,
      "num_input_tokens_seen": 585096,
      "step": 561,
      "train_runtime": 1586.0804,
      "train_tokens_per_second": 368.894
    },
    {
      "epoch": 0.24019660220108985,
      "grad_norm": 3.0194408893585205,
      "learning_rate": 1.5237687366167025e-05,
      "loss": 0.14587190747261047,
      "num_input_tokens_seen": 586010,
      "step": 562,
      "train_runtime": 1588.1587,
      "train_tokens_per_second": 368.987
    },
    {
      "epoch": 0.24062399829041564,
      "grad_norm": 1.7433415651321411,
      "learning_rate": 1.5229122055674518e-05,
      "loss": 0.06883656233549118,
      "num_input_tokens_seen": 587124,
      "step": 563,
      "train_runtime": 1590.3252,
      "train_tokens_per_second": 369.185
    },
    {
      "epoch": 0.24105139437974143,
      "grad_norm": 2.674011707305908,
      "learning_rate": 1.5220556745182014e-05,
      "loss": 0.19446563720703125,
      "num_input_tokens_seen": 588444,
      "step": 564,
      "train_runtime": 1592.5455,
      "train_tokens_per_second": 369.499
    },
    {
      "epoch": 0.2414787904690672,
      "grad_norm": 1.8672616481781006,
      "learning_rate": 1.5211991434689509e-05,
      "loss": 0.144138365983963,
      "num_input_tokens_seen": 589462,
      "step": 565,
      "train_runtime": 1594.6949,
      "train_tokens_per_second": 369.639
    },
    {
      "epoch": 0.241906186558393,
      "grad_norm": 2.232290506362915,
      "learning_rate": 1.5203426124197002e-05,
      "loss": 0.09555627405643463,
      "num_input_tokens_seen": 590564,
      "step": 566,
      "train_runtime": 1596.831,
      "train_tokens_per_second": 369.835
    },
    {
      "epoch": 0.2423335826477188,
      "grad_norm": 2.130337953567505,
      "learning_rate": 1.5194860813704498e-05,
      "loss": 0.16360598802566528,
      "num_input_tokens_seen": 591790,
      "step": 567,
      "train_runtime": 1599.0472,
      "train_tokens_per_second": 370.089
    },
    {
      "epoch": 0.24276097873704455,
      "grad_norm": 1.619037389755249,
      "learning_rate": 1.5186295503211991e-05,
      "loss": 0.11852429807186127,
      "num_input_tokens_seen": 592800,
      "step": 568,
      "train_runtime": 1601.1855,
      "train_tokens_per_second": 370.226
    },
    {
      "epoch": 0.24318837482637035,
      "grad_norm": 2.230518102645874,
      "learning_rate": 1.5177730192719487e-05,
      "loss": 0.1458892673254013,
      "num_input_tokens_seen": 593980,
      "step": 569,
      "train_runtime": 1603.395,
      "train_tokens_per_second": 370.451
    },
    {
      "epoch": 0.2436157709156961,
      "grad_norm": 1.8642629384994507,
      "learning_rate": 1.5169164882226982e-05,
      "loss": 0.12750224769115448,
      "num_input_tokens_seen": 595218,
      "step": 570,
      "train_runtime": 1605.6111,
      "train_tokens_per_second": 370.711
    },
    {
      "epoch": 0.2440431670050219,
      "grad_norm": 2.8837227821350098,
      "learning_rate": 1.5160599571734475e-05,
      "loss": 0.13859373331069946,
      "num_input_tokens_seen": 595882,
      "step": 571,
      "train_runtime": 1619.2751,
      "train_tokens_per_second": 367.993
    },
    {
      "epoch": 0.2444705630943477,
      "grad_norm": 1.8672165870666504,
      "learning_rate": 1.515203426124197e-05,
      "loss": 0.060058604925870895,
      "num_input_tokens_seen": 597056,
      "step": 572,
      "train_runtime": 1621.4925,
      "train_tokens_per_second": 368.214
    },
    {
      "epoch": 0.24489795918367346,
      "grad_norm": 2.315662384033203,
      "learning_rate": 1.5143468950749468e-05,
      "loss": 0.1080392599105835,
      "num_input_tokens_seen": 598454,
      "step": 573,
      "train_runtime": 1623.7473,
      "train_tokens_per_second": 368.564
    },
    {
      "epoch": 0.24532535527299926,
      "grad_norm": 2.41036057472229,
      "learning_rate": 1.5134903640256961e-05,
      "loss": 0.13520333170890808,
      "num_input_tokens_seen": 599250,
      "step": 574,
      "train_runtime": 1625.8445,
      "train_tokens_per_second": 368.578
    },
    {
      "epoch": 0.24575275136232502,
      "grad_norm": 1.8074324131011963,
      "learning_rate": 1.5126338329764456e-05,
      "loss": 0.0938856303691864,
      "num_input_tokens_seen": 600164,
      "step": 575,
      "train_runtime": 1627.969,
      "train_tokens_per_second": 368.658
    },
    {
      "epoch": 0.24618014745165082,
      "grad_norm": 2.911297082901001,
      "learning_rate": 1.511777301927195e-05,
      "loss": 0.10376982390880585,
      "num_input_tokens_seen": 600880,
      "step": 576,
      "train_runtime": 1630.057,
      "train_tokens_per_second": 368.625
    },
    {
      "epoch": 0.2466075435409766,
      "grad_norm": 1.6230921745300293,
      "learning_rate": 1.5109207708779445e-05,
      "loss": 0.09955968707799911,
      "num_input_tokens_seen": 602352,
      "step": 577,
      "train_runtime": 1632.3017,
      "train_tokens_per_second": 369.02
    },
    {
      "epoch": 0.24703493963030237,
      "grad_norm": 2.609530210494995,
      "learning_rate": 1.510064239828694e-05,
      "loss": 0.13412997126579285,
      "num_input_tokens_seen": 603496,
      "step": 578,
      "train_runtime": 1634.4749,
      "train_tokens_per_second": 369.229
    },
    {
      "epoch": 0.24746233571962817,
      "grad_norm": 2.2227115631103516,
      "learning_rate": 1.5092077087794434e-05,
      "loss": 0.16486811637878418,
      "num_input_tokens_seen": 604872,
      "step": 579,
      "train_runtime": 1636.7179,
      "train_tokens_per_second": 369.564
    },
    {
      "epoch": 0.24788973180895396,
      "grad_norm": 2.3095686435699463,
      "learning_rate": 1.508351177730193e-05,
      "loss": 0.12404029071331024,
      "num_input_tokens_seen": 605866,
      "step": 580,
      "train_runtime": 1638.8275,
      "train_tokens_per_second": 369.695
    },
    {
      "epoch": 0.24831712789827973,
      "grad_norm": 2.0123238563537598,
      "learning_rate": 1.5074946466809423e-05,
      "loss": 0.08304786682128906,
      "num_input_tokens_seen": 606936,
      "step": 581,
      "train_runtime": 1640.9866,
      "train_tokens_per_second": 369.86
    },
    {
      "epoch": 0.24874452398760552,
      "grad_norm": 2.6874747276306152,
      "learning_rate": 1.5066381156316918e-05,
      "loss": 0.20925870537757874,
      "num_input_tokens_seen": 607974,
      "step": 582,
      "train_runtime": 1643.3493,
      "train_tokens_per_second": 369.96
    },
    {
      "epoch": 0.24917192007693129,
      "grad_norm": 3.5932059288024902,
      "learning_rate": 1.5057815845824414e-05,
      "loss": 0.2375958114862442,
      "num_input_tokens_seen": 609022,
      "step": 583,
      "train_runtime": 1645.565,
      "train_tokens_per_second": 370.099
    },
    {
      "epoch": 0.24959931616625708,
      "grad_norm": 1.4646111726760864,
      "learning_rate": 1.5049250535331907e-05,
      "loss": 0.07569268345832825,
      "num_input_tokens_seen": 610356,
      "step": 584,
      "train_runtime": 1647.7924,
      "train_tokens_per_second": 370.408
    },
    {
      "epoch": 0.2500267122555829,
      "grad_norm": 3.407583236694336,
      "learning_rate": 1.5040685224839402e-05,
      "loss": 0.14567101001739502,
      "num_input_tokens_seen": 611014,
      "step": 585,
      "train_runtime": 1649.9535,
      "train_tokens_per_second": 370.322
    },
    {
      "epoch": 0.25045410834490867,
      "grad_norm": 2.5580594539642334,
      "learning_rate": 1.5032119914346896e-05,
      "loss": 0.17587047815322876,
      "num_input_tokens_seen": 611958,
      "step": 586,
      "train_runtime": 1652.0831,
      "train_tokens_per_second": 370.416
    },
    {
      "epoch": 0.2508815044342344,
      "grad_norm": 2.225137233734131,
      "learning_rate": 1.5023554603854391e-05,
      "loss": 0.14163464307785034,
      "num_input_tokens_seen": 613254,
      "step": 587,
      "train_runtime": 1654.3121,
      "train_tokens_per_second": 370.7
    },
    {
      "epoch": 0.2513089005235602,
      "grad_norm": 2.277710199356079,
      "learning_rate": 1.5014989293361885e-05,
      "loss": 0.18206238746643066,
      "num_input_tokens_seen": 614336,
      "step": 588,
      "train_runtime": 1656.5223,
      "train_tokens_per_second": 370.859
    },
    {
      "epoch": 0.251736296612886,
      "grad_norm": 2.002258777618408,
      "learning_rate": 1.500642398286938e-05,
      "loss": 0.1447502076625824,
      "num_input_tokens_seen": 615766,
      "step": 589,
      "train_runtime": 1658.8042,
      "train_tokens_per_second": 371.211
    },
    {
      "epoch": 0.2521636927022118,
      "grad_norm": 1.7282333374023438,
      "learning_rate": 1.4997858672376875e-05,
      "loss": 0.1233157217502594,
      "num_input_tokens_seen": 617264,
      "step": 590,
      "train_runtime": 1661.0949,
      "train_tokens_per_second": 371.601
    },
    {
      "epoch": 0.2525910887915376,
      "grad_norm": 2.3245608806610107,
      "learning_rate": 1.4989293361884369e-05,
      "loss": 0.14430296421051025,
      "num_input_tokens_seen": 618526,
      "step": 591,
      "train_runtime": 1663.326,
      "train_tokens_per_second": 371.861
    },
    {
      "epoch": 0.2530184848808633,
      "grad_norm": 1.7639458179473877,
      "learning_rate": 1.4980728051391864e-05,
      "loss": 0.14418625831604004,
      "num_input_tokens_seen": 620828,
      "step": 592,
      "train_runtime": 1665.8388,
      "train_tokens_per_second": 372.682
    },
    {
      "epoch": 0.2534458809701891,
      "grad_norm": 1.7470554113388062,
      "learning_rate": 1.4972162740899358e-05,
      "loss": 0.1562797874212265,
      "num_input_tokens_seen": 621908,
      "step": 593,
      "train_runtime": 1667.9948,
      "train_tokens_per_second": 372.848
    },
    {
      "epoch": 0.2538732770595149,
      "grad_norm": 2.1268813610076904,
      "learning_rate": 1.4963597430406853e-05,
      "loss": 0.13244813680648804,
      "num_input_tokens_seen": 623304,
      "step": 594,
      "train_runtime": 1670.332,
      "train_tokens_per_second": 373.162
    },
    {
      "epoch": 0.2543006731488407,
      "grad_norm": 1.9472156763076782,
      "learning_rate": 1.4955032119914348e-05,
      "loss": 0.10751897096633911,
      "num_input_tokens_seen": 624342,
      "step": 595,
      "train_runtime": 1672.5003,
      "train_tokens_per_second": 373.299
    },
    {
      "epoch": 0.2547280692381665,
      "grad_norm": 2.370417356491089,
      "learning_rate": 1.4946466809421842e-05,
      "loss": 0.2049310803413391,
      "num_input_tokens_seen": 625290,
      "step": 596,
      "train_runtime": 1674.6195,
      "train_tokens_per_second": 373.392
    },
    {
      "epoch": 0.2551554653274922,
      "grad_norm": 3.086700677871704,
      "learning_rate": 1.4937901498929337e-05,
      "loss": 0.12804263830184937,
      "num_input_tokens_seen": 626950,
      "step": 597,
      "train_runtime": 1676.9535,
      "train_tokens_per_second": 373.862
    },
    {
      "epoch": 0.255582861416818,
      "grad_norm": 1.9125748872756958,
      "learning_rate": 1.4929336188436831e-05,
      "loss": 0.17901611328125,
      "num_input_tokens_seen": 628778,
      "step": 598,
      "train_runtime": 1679.2616,
      "train_tokens_per_second": 374.437
    },
    {
      "epoch": 0.2560102575061438,
      "grad_norm": 2.6187164783477783,
      "learning_rate": 1.4920770877944326e-05,
      "loss": 0.21042247116565704,
      "num_input_tokens_seen": 629784,
      "step": 599,
      "train_runtime": 1681.3888,
      "train_tokens_per_second": 374.562
    },
    {
      "epoch": 0.2564376535954696,
      "grad_norm": 2.2882473468780518,
      "learning_rate": 1.4912205567451821e-05,
      "loss": 0.13214997947216034,
      "num_input_tokens_seen": 630786,
      "step": 600,
      "train_runtime": 1683.539,
      "train_tokens_per_second": 374.679
    },
    {
      "epoch": 0.2568650496847954,
      "grad_norm": 2.7067177295684814,
      "learning_rate": 1.4903640256959315e-05,
      "loss": 0.12402036786079407,
      "num_input_tokens_seen": 631604,
      "step": 601,
      "train_runtime": 1698.8541,
      "train_tokens_per_second": 371.782
    },
    {
      "epoch": 0.2572924457741212,
      "grad_norm": 2.4188687801361084,
      "learning_rate": 1.489507494646681e-05,
      "loss": 0.16838578879833221,
      "num_input_tokens_seen": 632476,
      "step": 602,
      "train_runtime": 1700.94,
      "train_tokens_per_second": 371.839
    },
    {
      "epoch": 0.25771984186344693,
      "grad_norm": 2.240884304046631,
      "learning_rate": 1.4886509635974304e-05,
      "loss": 0.10224397480487823,
      "num_input_tokens_seen": 633386,
      "step": 603,
      "train_runtime": 1703.0595,
      "train_tokens_per_second": 371.911
    },
    {
      "epoch": 0.2581472379527727,
      "grad_norm": 2.211941957473755,
      "learning_rate": 1.48779443254818e-05,
      "loss": 0.1301332712173462,
      "num_input_tokens_seen": 634266,
      "step": 604,
      "train_runtime": 1705.1783,
      "train_tokens_per_second": 371.965
    },
    {
      "epoch": 0.2585746340420985,
      "grad_norm": 2.600369930267334,
      "learning_rate": 1.4869379014989293e-05,
      "loss": 0.13147202134132385,
      "num_input_tokens_seen": 635158,
      "step": 605,
      "train_runtime": 1707.2964,
      "train_tokens_per_second": 372.026
    },
    {
      "epoch": 0.2590020301314243,
      "grad_norm": 1.8159915208816528,
      "learning_rate": 1.4860813704496788e-05,
      "loss": 0.11959327012300491,
      "num_input_tokens_seen": 636060,
      "step": 606,
      "train_runtime": 1709.4065,
      "train_tokens_per_second": 372.094
    },
    {
      "epoch": 0.2594294262207501,
      "grad_norm": 2.2972586154937744,
      "learning_rate": 1.4852248394004285e-05,
      "loss": 0.1352578103542328,
      "num_input_tokens_seen": 637656,
      "step": 607,
      "train_runtime": 1711.7053,
      "train_tokens_per_second": 372.527
    },
    {
      "epoch": 0.25985682231007584,
      "grad_norm": 1.8306388854980469,
      "learning_rate": 1.484368308351178e-05,
      "loss": 0.07780133187770844,
      "num_input_tokens_seen": 638622,
      "step": 608,
      "train_runtime": 1713.8153,
      "train_tokens_per_second": 372.632
    },
    {
      "epoch": 0.26028421839940163,
      "grad_norm": 2.4278531074523926,
      "learning_rate": 1.4835117773019274e-05,
      "loss": 0.1398286521434784,
      "num_input_tokens_seen": 639516,
      "step": 609,
      "train_runtime": 1715.9091,
      "train_tokens_per_second": 372.698
    },
    {
      "epoch": 0.26071161448872743,
      "grad_norm": 1.9774209260940552,
      "learning_rate": 1.4826552462526769e-05,
      "loss": 0.1397102177143097,
      "num_input_tokens_seen": 641198,
      "step": 610,
      "train_runtime": 1718.291,
      "train_tokens_per_second": 373.16
    },
    {
      "epoch": 0.2611390105780532,
      "grad_norm": 3.3630471229553223,
      "learning_rate": 1.4817987152034263e-05,
      "loss": 0.1562974452972412,
      "num_input_tokens_seen": 642066,
      "step": 611,
      "train_runtime": 1720.4123,
      "train_tokens_per_second": 373.205
    },
    {
      "epoch": 0.261566406667379,
      "grad_norm": 1.6679407358169556,
      "learning_rate": 1.4809421841541758e-05,
      "loss": 0.12732523679733276,
      "num_input_tokens_seen": 643672,
      "step": 612,
      "train_runtime": 1722.7337,
      "train_tokens_per_second": 373.634
    },
    {
      "epoch": 0.26199380275670475,
      "grad_norm": 1.9551758766174316,
      "learning_rate": 1.4800856531049253e-05,
      "loss": 0.16241273283958435,
      "num_input_tokens_seen": 644672,
      "step": 613,
      "train_runtime": 1724.8963,
      "train_tokens_per_second": 373.745
    },
    {
      "epoch": 0.26242119884603055,
      "grad_norm": 2.7418792247772217,
      "learning_rate": 1.4792291220556747e-05,
      "loss": 0.20132620632648468,
      "num_input_tokens_seen": 645518,
      "step": 614,
      "train_runtime": 1727.1499,
      "train_tokens_per_second": 373.748
    },
    {
      "epoch": 0.26284859493535634,
      "grad_norm": 2.3483219146728516,
      "learning_rate": 1.4783725910064242e-05,
      "loss": 0.10017156600952148,
      "num_input_tokens_seen": 646582,
      "step": 615,
      "train_runtime": 1729.3478,
      "train_tokens_per_second": 373.888
    },
    {
      "epoch": 0.26327599102468213,
      "grad_norm": 2.903587579727173,
      "learning_rate": 1.4775160599571736e-05,
      "loss": 0.16080167889595032,
      "num_input_tokens_seen": 647430,
      "step": 616,
      "train_runtime": 1731.5146,
      "train_tokens_per_second": 373.91
    },
    {
      "epoch": 0.2637033871140079,
      "grad_norm": 2.809605598449707,
      "learning_rate": 1.4766595289079231e-05,
      "loss": 0.13062329590320587,
      "num_input_tokens_seen": 648152,
      "step": 617,
      "train_runtime": 1733.6845,
      "train_tokens_per_second": 373.858
    },
    {
      "epoch": 0.26413078320333366,
      "grad_norm": 3.384843587875366,
      "learning_rate": 1.4758029978586725e-05,
      "loss": 0.16314518451690674,
      "num_input_tokens_seen": 648804,
      "step": 618,
      "train_runtime": 1735.7569,
      "train_tokens_per_second": 373.787
    },
    {
      "epoch": 0.26455817929265946,
      "grad_norm": 3.2259411811828613,
      "learning_rate": 1.474946466809422e-05,
      "loss": 0.18350113928318024,
      "num_input_tokens_seen": 650052,
      "step": 619,
      "train_runtime": 1737.9925,
      "train_tokens_per_second": 374.025
    },
    {
      "epoch": 0.26498557538198525,
      "grad_norm": 1.8685604333877563,
      "learning_rate": 1.4740899357601715e-05,
      "loss": 0.10585185885429382,
      "num_input_tokens_seen": 651060,
      "step": 620,
      "train_runtime": 1740.1966,
      "train_tokens_per_second": 374.13
    },
    {
      "epoch": 0.26541297147131104,
      "grad_norm": 2.313979148864746,
      "learning_rate": 1.4732334047109209e-05,
      "loss": 0.1605004370212555,
      "num_input_tokens_seen": 651978,
      "step": 621,
      "train_runtime": 1742.3659,
      "train_tokens_per_second": 374.191
    },
    {
      "epoch": 0.26584036756063684,
      "grad_norm": 6.554983139038086,
      "learning_rate": 1.4723768736616704e-05,
      "loss": 0.23819810152053833,
      "num_input_tokens_seen": 652776,
      "step": 622,
      "train_runtime": 1744.4811,
      "train_tokens_per_second": 374.195
    },
    {
      "epoch": 0.26626776364996263,
      "grad_norm": 1.9299674034118652,
      "learning_rate": 1.4715203426124198e-05,
      "loss": 0.125253364443779,
      "num_input_tokens_seen": 654328,
      "step": 623,
      "train_runtime": 1746.8029,
      "train_tokens_per_second": 374.586
    },
    {
      "epoch": 0.26669515973928837,
      "grad_norm": 3.4705519676208496,
      "learning_rate": 1.4706638115631693e-05,
      "loss": 0.2804924249649048,
      "num_input_tokens_seen": 655520,
      "step": 624,
      "train_runtime": 1749.0096,
      "train_tokens_per_second": 374.795
    },
    {
      "epoch": 0.26712255582861416,
      "grad_norm": 2.8094065189361572,
      "learning_rate": 1.4698072805139188e-05,
      "loss": 0.14450781047344208,
      "num_input_tokens_seen": 656314,
      "step": 625,
      "train_runtime": 1751.1381,
      "train_tokens_per_second": 374.793
    },
    {
      "epoch": 0.26754995191793995,
      "grad_norm": 2.3324849605560303,
      "learning_rate": 1.4689507494646682e-05,
      "loss": 0.1516658067703247,
      "num_input_tokens_seen": 657426,
      "step": 626,
      "train_runtime": 1753.3261,
      "train_tokens_per_second": 374.959
    },
    {
      "epoch": 0.26797734800726575,
      "grad_norm": 2.6098573207855225,
      "learning_rate": 1.4680942184154177e-05,
      "loss": 0.14121192693710327,
      "num_input_tokens_seen": 658524,
      "step": 627,
      "train_runtime": 1755.5409,
      "train_tokens_per_second": 375.112
    },
    {
      "epoch": 0.26840474409659154,
      "grad_norm": 2.85821533203125,
      "learning_rate": 1.467237687366167e-05,
      "loss": 0.09671340882778168,
      "num_input_tokens_seen": 659636,
      "step": 628,
      "train_runtime": 1757.732,
      "train_tokens_per_second": 375.277
    },
    {
      "epoch": 0.2688321401859173,
      "grad_norm": 1.9630221128463745,
      "learning_rate": 1.4663811563169166e-05,
      "loss": 0.1310899704694748,
      "num_input_tokens_seen": 660798,
      "step": 629,
      "train_runtime": 1759.9823,
      "train_tokens_per_second": 375.457
    },
    {
      "epoch": 0.26925953627524307,
      "grad_norm": 1.739284873008728,
      "learning_rate": 1.4655246252676661e-05,
      "loss": 0.10666526108980179,
      "num_input_tokens_seen": 662582,
      "step": 630,
      "train_runtime": 1762.343,
      "train_tokens_per_second": 375.967
    },
    {
      "epoch": 0.26968693236456887,
      "grad_norm": 2.0755605697631836,
      "learning_rate": 1.4646680942184155e-05,
      "loss": 0.1379154622554779,
      "num_input_tokens_seen": 663804,
      "step": 631,
      "train_runtime": 1777.95,
      "train_tokens_per_second": 373.354
    },
    {
      "epoch": 0.27011432845389466,
      "grad_norm": 2.186189651489258,
      "learning_rate": 1.463811563169165e-05,
      "loss": 0.11671717464923859,
      "num_input_tokens_seen": 664674,
      "step": 632,
      "train_runtime": 1780.1003,
      "train_tokens_per_second": 373.391
    },
    {
      "epoch": 0.27054172454322045,
      "grad_norm": 2.5961384773254395,
      "learning_rate": 1.4629550321199144e-05,
      "loss": 0.1427270919084549,
      "num_input_tokens_seen": 665636,
      "step": 633,
      "train_runtime": 1782.2739,
      "train_tokens_per_second": 373.476
    },
    {
      "epoch": 0.2709691206325462,
      "grad_norm": 2.0703091621398926,
      "learning_rate": 1.4620985010706639e-05,
      "loss": 0.10669196397066116,
      "num_input_tokens_seen": 666736,
      "step": 634,
      "train_runtime": 1784.4507,
      "train_tokens_per_second": 373.637
    },
    {
      "epoch": 0.271396516721872,
      "grad_norm": 1.9252040386199951,
      "learning_rate": 1.4612419700214132e-05,
      "loss": 0.10314284265041351,
      "num_input_tokens_seen": 668148,
      "step": 635,
      "train_runtime": 1786.7109,
      "train_tokens_per_second": 373.954
    },
    {
      "epoch": 0.2718239128111978,
      "grad_norm": 2.0868709087371826,
      "learning_rate": 1.4603854389721628e-05,
      "loss": 0.1109037846326828,
      "num_input_tokens_seen": 669176,
      "step": 636,
      "train_runtime": 1788.8378,
      "train_tokens_per_second": 374.084
    },
    {
      "epoch": 0.27225130890052357,
      "grad_norm": 2.150547742843628,
      "learning_rate": 1.4595289079229123e-05,
      "loss": 0.11692191660404205,
      "num_input_tokens_seen": 670002,
      "step": 637,
      "train_runtime": 1790.9635,
      "train_tokens_per_second": 374.101
    },
    {
      "epoch": 0.27267870498984936,
      "grad_norm": 1.5417817831039429,
      "learning_rate": 1.4586723768736617e-05,
      "loss": 0.1225898414850235,
      "num_input_tokens_seen": 671710,
      "step": 638,
      "train_runtime": 1793.316,
      "train_tokens_per_second": 374.563
    },
    {
      "epoch": 0.2731061010791751,
      "grad_norm": 2.2921831607818604,
      "learning_rate": 1.4578158458244112e-05,
      "loss": 0.13761989772319794,
      "num_input_tokens_seen": 672736,
      "step": 639,
      "train_runtime": 1795.4712,
      "train_tokens_per_second": 374.685
    },
    {
      "epoch": 0.2735334971685009,
      "grad_norm": 2.043858289718628,
      "learning_rate": 1.4569593147751605e-05,
      "loss": 0.1470688134431839,
      "num_input_tokens_seen": 673894,
      "step": 640,
      "train_runtime": 1797.664,
      "train_tokens_per_second": 374.872
    },
    {
      "epoch": 0.2739608932578267,
      "grad_norm": 2.1990044116973877,
      "learning_rate": 1.45610278372591e-05,
      "loss": 0.19008825719356537,
      "num_input_tokens_seen": 675060,
      "step": 641,
      "train_runtime": 1799.8386,
      "train_tokens_per_second": 375.067
    },
    {
      "epoch": 0.2743882893471525,
      "grad_norm": 2.5244386196136475,
      "learning_rate": 1.4552462526766598e-05,
      "loss": 0.17848122119903564,
      "num_input_tokens_seen": 676204,
      "step": 642,
      "train_runtime": 1801.9888,
      "train_tokens_per_second": 375.254
    },
    {
      "epoch": 0.2748156854364783,
      "grad_norm": 2.1407968997955322,
      "learning_rate": 1.4543897216274093e-05,
      "loss": 0.09873424470424652,
      "num_input_tokens_seen": 677100,
      "step": 643,
      "train_runtime": 1804.0803,
      "train_tokens_per_second": 375.316
    },
    {
      "epoch": 0.275243081525804,
      "grad_norm": 2.634310722351074,
      "learning_rate": 1.4535331905781587e-05,
      "loss": 0.13076536357402802,
      "num_input_tokens_seen": 678234,
      "step": 644,
      "train_runtime": 1806.2794,
      "train_tokens_per_second": 375.487
    },
    {
      "epoch": 0.2756704776151298,
      "grad_norm": 2.9757766723632812,
      "learning_rate": 1.4526766595289082e-05,
      "loss": 0.12197183072566986,
      "num_input_tokens_seen": 678926,
      "step": 645,
      "train_runtime": 1808.4195,
      "train_tokens_per_second": 375.425
    },
    {
      "epoch": 0.2760978737044556,
      "grad_norm": 1.983682632446289,
      "learning_rate": 1.4518201284796575e-05,
      "loss": 0.11559322476387024,
      "num_input_tokens_seen": 679898,
      "step": 646,
      "train_runtime": 1810.5739,
      "train_tokens_per_second": 375.515
    },
    {
      "epoch": 0.2765252697937814,
      "grad_norm": 1.9641001224517822,
      "learning_rate": 1.450963597430407e-05,
      "loss": 0.14868269860744476,
      "num_input_tokens_seen": 680990,
      "step": 647,
      "train_runtime": 1812.7621,
      "train_tokens_per_second": 375.664
    },
    {
      "epoch": 0.2769526658831072,
      "grad_norm": 1.8658320903778076,
      "learning_rate": 1.4501070663811564e-05,
      "loss": 0.1106535792350769,
      "num_input_tokens_seen": 682104,
      "step": 648,
      "train_runtime": 1814.973,
      "train_tokens_per_second": 375.82
    },
    {
      "epoch": 0.277380061972433,
      "grad_norm": 2.208958387374878,
      "learning_rate": 1.449250535331906e-05,
      "loss": 0.15505282580852509,
      "num_input_tokens_seen": 683204,
      "step": 649,
      "train_runtime": 1817.142,
      "train_tokens_per_second": 375.977
    },
    {
      "epoch": 0.2778074580617587,
      "grad_norm": 2.8805222511291504,
      "learning_rate": 1.4483940042826555e-05,
      "loss": 0.19088751077651978,
      "num_input_tokens_seen": 683966,
      "step": 650,
      "train_runtime": 1819.252,
      "train_tokens_per_second": 375.96
    },
    {
      "epoch": 0.2782348541510845,
      "grad_norm": 2.2402267456054688,
      "learning_rate": 1.4475374732334048e-05,
      "loss": 0.10395414382219315,
      "num_input_tokens_seen": 684992,
      "step": 651,
      "train_runtime": 1821.3883,
      "train_tokens_per_second": 376.082
    },
    {
      "epoch": 0.2786622502404103,
      "grad_norm": 2.177079916000366,
      "learning_rate": 1.4466809421841544e-05,
      "loss": 0.1784535050392151,
      "num_input_tokens_seen": 686164,
      "step": 652,
      "train_runtime": 1823.6148,
      "train_tokens_per_second": 376.266
    },
    {
      "epoch": 0.2790896463297361,
      "grad_norm": 1.5526562929153442,
      "learning_rate": 1.4458244111349037e-05,
      "loss": 0.13759081065654755,
      "num_input_tokens_seen": 687650,
      "step": 653,
      "train_runtime": 1825.8962,
      "train_tokens_per_second": 376.61
    },
    {
      "epoch": 0.2795170424190619,
      "grad_norm": 2.5718026161193848,
      "learning_rate": 1.4449678800856533e-05,
      "loss": 0.21604347229003906,
      "num_input_tokens_seen": 688598,
      "step": 654,
      "train_runtime": 1828.057,
      "train_tokens_per_second": 376.683
    },
    {
      "epoch": 0.2799444385083876,
      "grad_norm": 2.6486434936523438,
      "learning_rate": 1.4441113490364028e-05,
      "loss": 0.1431410014629364,
      "num_input_tokens_seen": 689498,
      "step": 655,
      "train_runtime": 1830.1921,
      "train_tokens_per_second": 376.735
    },
    {
      "epoch": 0.2803718345977134,
      "grad_norm": 2.255743980407715,
      "learning_rate": 1.4432548179871521e-05,
      "loss": 0.11979982256889343,
      "num_input_tokens_seen": 690404,
      "step": 656,
      "train_runtime": 1832.3088,
      "train_tokens_per_second": 376.795
    },
    {
      "epoch": 0.2807992306870392,
      "grad_norm": 1.7594857215881348,
      "learning_rate": 1.4423982869379017e-05,
      "loss": 0.11151863634586334,
      "num_input_tokens_seen": 692104,
      "step": 657,
      "train_runtime": 1834.6193,
      "train_tokens_per_second": 377.247
    },
    {
      "epoch": 0.281226626776365,
      "grad_norm": 2.4156394004821777,
      "learning_rate": 1.441541755888651e-05,
      "loss": 0.12841804325580597,
      "num_input_tokens_seen": 693036,
      "step": 658,
      "train_runtime": 1836.773,
      "train_tokens_per_second": 377.312
    },
    {
      "epoch": 0.2816540228656908,
      "grad_norm": 2.3588061332702637,
      "learning_rate": 1.4406852248394005e-05,
      "loss": 0.10891027003526688,
      "num_input_tokens_seen": 694052,
      "step": 659,
      "train_runtime": 1838.9409,
      "train_tokens_per_second": 377.419
    },
    {
      "epoch": 0.28208141895501654,
      "grad_norm": 2.0901601314544678,
      "learning_rate": 1.43982869379015e-05,
      "loss": 0.12020985782146454,
      "num_input_tokens_seen": 695124,
      "step": 660,
      "train_runtime": 1841.1038,
      "train_tokens_per_second": 377.558
    },
    {
      "epoch": 0.28250881504434233,
      "grad_norm": 2.724958658218384,
      "learning_rate": 1.4389721627408994e-05,
      "loss": 0.116816446185112,
      "num_input_tokens_seen": 696154,
      "step": 661,
      "train_runtime": 1857.3141,
      "train_tokens_per_second": 374.818
    },
    {
      "epoch": 0.2829362111336681,
      "grad_norm": 1.732968807220459,
      "learning_rate": 1.438115631691649e-05,
      "loss": 0.1253526657819748,
      "num_input_tokens_seen": 697178,
      "step": 662,
      "train_runtime": 1859.5003,
      "train_tokens_per_second": 374.928
    },
    {
      "epoch": 0.2833636072229939,
      "grad_norm": 3.5117788314819336,
      "learning_rate": 1.4372591006423983e-05,
      "loss": 0.15188395977020264,
      "num_input_tokens_seen": 697816,
      "step": 663,
      "train_runtime": 1861.5695,
      "train_tokens_per_second": 374.854
    },
    {
      "epoch": 0.2837910033123197,
      "grad_norm": 2.3651187419891357,
      "learning_rate": 1.4364025695931478e-05,
      "loss": 0.112892284989357,
      "num_input_tokens_seen": 698672,
      "step": 664,
      "train_runtime": 1863.7302,
      "train_tokens_per_second": 374.878
    },
    {
      "epoch": 0.28421839940164545,
      "grad_norm": 2.184657096862793,
      "learning_rate": 1.4355460385438972e-05,
      "loss": 0.17838981747627258,
      "num_input_tokens_seen": 699868,
      "step": 665,
      "train_runtime": 1865.9385,
      "train_tokens_per_second": 375.076
    },
    {
      "epoch": 0.28464579549097124,
      "grad_norm": 2.1275479793548584,
      "learning_rate": 1.4346895074946467e-05,
      "loss": 0.12411011010408401,
      "num_input_tokens_seen": 700964,
      "step": 666,
      "train_runtime": 1868.135,
      "train_tokens_per_second": 375.221
    },
    {
      "epoch": 0.28507319158029704,
      "grad_norm": 2.5148892402648926,
      "learning_rate": 1.4338329764453963e-05,
      "loss": 0.1568133533000946,
      "num_input_tokens_seen": 702766,
      "step": 667,
      "train_runtime": 1870.4581,
      "train_tokens_per_second": 375.719
    },
    {
      "epoch": 0.28550058766962283,
      "grad_norm": 2.366084575653076,
      "learning_rate": 1.4329764453961456e-05,
      "loss": 0.19036224484443665,
      "num_input_tokens_seen": 703934,
      "step": 668,
      "train_runtime": 1872.6909,
      "train_tokens_per_second": 375.894
    },
    {
      "epoch": 0.2859279837589486,
      "grad_norm": 2.3026533126831055,
      "learning_rate": 1.4321199143468951e-05,
      "loss": 0.09540949016809464,
      "num_input_tokens_seen": 704658,
      "step": 669,
      "train_runtime": 1874.7506,
      "train_tokens_per_second": 375.868
    },
    {
      "epoch": 0.28635537984827436,
      "grad_norm": 1.9528226852416992,
      "learning_rate": 1.4312633832976445e-05,
      "loss": 0.12443667650222778,
      "num_input_tokens_seen": 706188,
      "step": 670,
      "train_runtime": 1877.034,
      "train_tokens_per_second": 376.225
    },
    {
      "epoch": 0.28678277593760015,
      "grad_norm": 1.6549339294433594,
      "learning_rate": 1.430406852248394e-05,
      "loss": 0.10252031683921814,
      "num_input_tokens_seen": 707622,
      "step": 671,
      "train_runtime": 1879.2611,
      "train_tokens_per_second": 376.543
    },
    {
      "epoch": 0.28721017202692595,
      "grad_norm": 3.7874042987823486,
      "learning_rate": 1.4295503211991436e-05,
      "loss": 0.18935301899909973,
      "num_input_tokens_seen": 708358,
      "step": 672,
      "train_runtime": 1881.3531,
      "train_tokens_per_second": 376.515
    },
    {
      "epoch": 0.28763756811625174,
      "grad_norm": 1.6390641927719116,
      "learning_rate": 1.428693790149893e-05,
      "loss": 0.10724534094333649,
      "num_input_tokens_seen": 709372,
      "step": 673,
      "train_runtime": 1883.511,
      "train_tokens_per_second": 376.622
    },
    {
      "epoch": 0.28806496420557753,
      "grad_norm": 3.7022180557250977,
      "learning_rate": 1.4278372591006424e-05,
      "loss": 0.10025076568126678,
      "num_input_tokens_seen": 709858,
      "step": 674,
      "train_runtime": 1885.5114,
      "train_tokens_per_second": 376.48
    },
    {
      "epoch": 0.2884923602949033,
      "grad_norm": 2.573812961578369,
      "learning_rate": 1.4269807280513918e-05,
      "loss": 0.11544390767812729,
      "num_input_tokens_seen": 710642,
      "step": 675,
      "train_runtime": 1887.6045,
      "train_tokens_per_second": 376.478
    },
    {
      "epoch": 0.28891975638422907,
      "grad_norm": 1.8841919898986816,
      "learning_rate": 1.4261241970021415e-05,
      "loss": 0.13746702671051025,
      "num_input_tokens_seen": 711836,
      "step": 676,
      "train_runtime": 1889.7859,
      "train_tokens_per_second": 376.675
    },
    {
      "epoch": 0.28934715247355486,
      "grad_norm": 2.5100719928741455,
      "learning_rate": 1.425267665952891e-05,
      "loss": 0.14079850912094116,
      "num_input_tokens_seen": 712588,
      "step": 677,
      "train_runtime": 1891.8438,
      "train_tokens_per_second": 376.663
    },
    {
      "epoch": 0.28977454856288065,
      "grad_norm": 2.5519168376922607,
      "learning_rate": 1.4244111349036404e-05,
      "loss": 0.14881883561611176,
      "num_input_tokens_seen": 713650,
      "step": 678,
      "train_runtime": 1893.9969,
      "train_tokens_per_second": 376.796
    },
    {
      "epoch": 0.29020194465220644,
      "grad_norm": 3.0886623859405518,
      "learning_rate": 1.42355460385439e-05,
      "loss": 0.18498066067695618,
      "num_input_tokens_seen": 714694,
      "step": 679,
      "train_runtime": 1896.1599,
      "train_tokens_per_second": 376.917
    },
    {
      "epoch": 0.29062934074153224,
      "grad_norm": 1.4010649919509888,
      "learning_rate": 1.4226980728051394e-05,
      "loss": 0.07827949523925781,
      "num_input_tokens_seen": 715872,
      "step": 680,
      "train_runtime": 1898.3205,
      "train_tokens_per_second": 377.108
    },
    {
      "epoch": 0.291056736830858,
      "grad_norm": 2.8680427074432373,
      "learning_rate": 1.4218415417558888e-05,
      "loss": 0.18841995298862457,
      "num_input_tokens_seen": 716616,
      "step": 681,
      "train_runtime": 1900.4378,
      "train_tokens_per_second": 377.079
    },
    {
      "epoch": 0.29148413292018377,
      "grad_norm": 2.337132692337036,
      "learning_rate": 1.4209850107066383e-05,
      "loss": 0.14060130715370178,
      "num_input_tokens_seen": 717396,
      "step": 682,
      "train_runtime": 1902.5506,
      "train_tokens_per_second": 377.071
    },
    {
      "epoch": 0.29191152900950956,
      "grad_norm": 2.3525822162628174,
      "learning_rate": 1.4201284796573877e-05,
      "loss": 0.14918434619903564,
      "num_input_tokens_seen": 718324,
      "step": 683,
      "train_runtime": 1904.6647,
      "train_tokens_per_second": 377.139
    },
    {
      "epoch": 0.29233892509883536,
      "grad_norm": 4.944980621337891,
      "learning_rate": 1.4192719486081372e-05,
      "loss": 0.16028907895088196,
      "num_input_tokens_seen": 719262,
      "step": 684,
      "train_runtime": 1906.7883,
      "train_tokens_per_second": 377.211
    },
    {
      "epoch": 0.29276632118816115,
      "grad_norm": 2.611039876937866,
      "learning_rate": 1.4184154175588867e-05,
      "loss": 0.10495640337467194,
      "num_input_tokens_seen": 720384,
      "step": 685,
      "train_runtime": 1908.9745,
      "train_tokens_per_second": 377.367
    },
    {
      "epoch": 0.2931937172774869,
      "grad_norm": 2.5478999614715576,
      "learning_rate": 1.4175588865096361e-05,
      "loss": 0.1308320164680481,
      "num_input_tokens_seen": 721052,
      "step": 686,
      "train_runtime": 1910.9937,
      "train_tokens_per_second": 377.318
    },
    {
      "epoch": 0.2936211133668127,
      "grad_norm": 1.5950946807861328,
      "learning_rate": 1.4167023554603856e-05,
      "loss": 0.06269238889217377,
      "num_input_tokens_seen": 721936,
      "step": 687,
      "train_runtime": 1913.1088,
      "train_tokens_per_second": 377.363
    },
    {
      "epoch": 0.2940485094561385,
      "grad_norm": 2.833967447280884,
      "learning_rate": 1.415845824411135e-05,
      "loss": 0.2123396098613739,
      "num_input_tokens_seen": 722674,
      "step": 688,
      "train_runtime": 1915.1888,
      "train_tokens_per_second": 377.338
    },
    {
      "epoch": 0.29447590554546427,
      "grad_norm": 2.099252462387085,
      "learning_rate": 1.4149892933618845e-05,
      "loss": 0.1980723887681961,
      "num_input_tokens_seen": 723766,
      "step": 689,
      "train_runtime": 1917.366,
      "train_tokens_per_second": 377.479
    },
    {
      "epoch": 0.29490330163479006,
      "grad_norm": 1.4324655532836914,
      "learning_rate": 1.414132762312634e-05,
      "loss": 0.11713645607233047,
      "num_input_tokens_seen": 725872,
      "step": 690,
      "train_runtime": 1919.7768,
      "train_tokens_per_second": 378.102
    },
    {
      "epoch": 0.2953306977241158,
      "grad_norm": 1.695764183998108,
      "learning_rate": 1.4132762312633834e-05,
      "loss": 0.10431049764156342,
      "num_input_tokens_seen": 726914,
      "step": 691,
      "train_runtime": 1934.9105,
      "train_tokens_per_second": 375.684
    },
    {
      "epoch": 0.2957580938134416,
      "grad_norm": 2.0348596572875977,
      "learning_rate": 1.412419700214133e-05,
      "loss": 0.19648054242134094,
      "num_input_tokens_seen": 728272,
      "step": 692,
      "train_runtime": 1937.1537,
      "train_tokens_per_second": 375.95
    },
    {
      "epoch": 0.2961854899027674,
      "grad_norm": 2.2879536151885986,
      "learning_rate": 1.4115631691648823e-05,
      "loss": 0.11473573744297028,
      "num_input_tokens_seen": 729172,
      "step": 693,
      "train_runtime": 1939.325,
      "train_tokens_per_second": 375.993
    },
    {
      "epoch": 0.2966128859920932,
      "grad_norm": 2.5072267055511475,
      "learning_rate": 1.4107066381156318e-05,
      "loss": 0.15191859006881714,
      "num_input_tokens_seen": 730106,
      "step": 694,
      "train_runtime": 1941.4493,
      "train_tokens_per_second": 376.062
    },
    {
      "epoch": 0.29704028208141897,
      "grad_norm": 1.7680737972259521,
      "learning_rate": 1.4098501070663812e-05,
      "loss": 0.08820414543151855,
      "num_input_tokens_seen": 731538,
      "step": 695,
      "train_runtime": 1943.7077,
      "train_tokens_per_second": 376.362
    },
    {
      "epoch": 0.29746767817074476,
      "grad_norm": 1.741761326789856,
      "learning_rate": 1.4089935760171307e-05,
      "loss": 0.09984119981527328,
      "num_input_tokens_seen": 732798,
      "step": 696,
      "train_runtime": 1945.978,
      "train_tokens_per_second": 376.571
    },
    {
      "epoch": 0.2978950742600705,
      "grad_norm": 1.7765142917633057,
      "learning_rate": 1.4081370449678802e-05,
      "loss": 0.12116514891386032,
      "num_input_tokens_seen": 734214,
      "step": 697,
      "train_runtime": 1948.2195,
      "train_tokens_per_second": 376.864
    },
    {
      "epoch": 0.2983224703493963,
      "grad_norm": 3.0526561737060547,
      "learning_rate": 1.4072805139186296e-05,
      "loss": 0.20111225545406342,
      "num_input_tokens_seen": 735190,
      "step": 698,
      "train_runtime": 1950.3762,
      "train_tokens_per_second": 376.948
    },
    {
      "epoch": 0.2987498664387221,
      "grad_norm": 2.8217875957489014,
      "learning_rate": 1.4064239828693791e-05,
      "loss": 0.17018066346645355,
      "num_input_tokens_seen": 736072,
      "step": 699,
      "train_runtime": 1952.4619,
      "train_tokens_per_second": 376.997
    },
    {
      "epoch": 0.2991772625280479,
      "grad_norm": 1.7569355964660645,
      "learning_rate": 1.4055674518201285e-05,
      "loss": 0.10083118081092834,
      "num_input_tokens_seen": 737138,
      "step": 700,
      "train_runtime": 1954.6586,
      "train_tokens_per_second": 377.119
    },
    {
      "epoch": 0.2996046586173737,
      "grad_norm": 2.323338270187378,
      "learning_rate": 1.404710920770878e-05,
      "loss": 0.10599065572023392,
      "num_input_tokens_seen": 737838,
      "step": 701,
      "train_runtime": 1956.7468,
      "train_tokens_per_second": 377.074
    },
    {
      "epoch": 0.3000320547066994,
      "grad_norm": 2.171600103378296,
      "learning_rate": 1.4038543897216275e-05,
      "loss": 0.11985621601343155,
      "num_input_tokens_seen": 738804,
      "step": 702,
      "train_runtime": 1958.8846,
      "train_tokens_per_second": 377.155
    },
    {
      "epoch": 0.3000320547066994,
      "eval_loss": 0.510190486907959,
      "eval_runtime": 58.0123,
      "eval_samples_per_second": 17.203,
      "eval_steps_per_second": 8.602,
      "num_input_tokens_seen": 738804,
      "step": 702
    },
    {
      "epoch": 0.3004594507960252,
      "grad_norm": 2.1922428607940674,
      "learning_rate": 1.4029978586723769e-05,
      "loss": 0.16008763015270233,
      "num_input_tokens_seen": 740052,
      "step": 703,
      "train_runtime": 2019.1936,
      "train_tokens_per_second": 366.509
    },
    {
      "epoch": 0.300886846885351,
      "grad_norm": 2.2467708587646484,
      "learning_rate": 1.4021413276231264e-05,
      "loss": 0.15313953161239624,
      "num_input_tokens_seen": 741284,
      "step": 704,
      "train_runtime": 2021.4311,
      "train_tokens_per_second": 366.712
    },
    {
      "epoch": 0.3013142429746768,
      "grad_norm": 1.8962408304214478,
      "learning_rate": 1.4012847965738758e-05,
      "loss": 0.12336666882038116,
      "num_input_tokens_seen": 742224,
      "step": 705,
      "train_runtime": 2023.5746,
      "train_tokens_per_second": 366.789
    },
    {
      "epoch": 0.3017416390640026,
      "grad_norm": 3.211893320083618,
      "learning_rate": 1.4004282655246253e-05,
      "loss": 0.18593370914459229,
      "num_input_tokens_seen": 743002,
      "step": 706,
      "train_runtime": 2025.7584,
      "train_tokens_per_second": 366.777
    },
    {
      "epoch": 0.3021690351533283,
      "grad_norm": 1.7842845916748047,
      "learning_rate": 1.3995717344753748e-05,
      "loss": 0.11546893417835236,
      "num_input_tokens_seen": 744234,
      "step": 707,
      "train_runtime": 2027.96,
      "train_tokens_per_second": 366.987
    },
    {
      "epoch": 0.3025964312426541,
      "grad_norm": 1.9092752933502197,
      "learning_rate": 1.3987152034261242e-05,
      "loss": 0.07692892849445343,
      "num_input_tokens_seen": 744844,
      "step": 708,
      "train_runtime": 2030.0846,
      "train_tokens_per_second": 366.903
    },
    {
      "epoch": 0.3030238273319799,
      "grad_norm": 2.434123992919922,
      "learning_rate": 1.3978586723768737e-05,
      "loss": 0.1410711109638214,
      "num_input_tokens_seen": 745712,
      "step": 709,
      "train_runtime": 2032.1975,
      "train_tokens_per_second": 366.949
    },
    {
      "epoch": 0.3034512234213057,
      "grad_norm": 3.0329978466033936,
      "learning_rate": 1.3970021413276234e-05,
      "loss": 0.17612597346305847,
      "num_input_tokens_seen": 746854,
      "step": 710,
      "train_runtime": 2034.4066,
      "train_tokens_per_second": 367.111
    },
    {
      "epoch": 0.3038786195106315,
      "grad_norm": 2.7899532318115234,
      "learning_rate": 1.3961456102783728e-05,
      "loss": 0.16027897596359253,
      "num_input_tokens_seen": 747604,
      "step": 711,
      "train_runtime": 2036.5038,
      "train_tokens_per_second": 367.102
    },
    {
      "epoch": 0.30430601559995724,
      "grad_norm": 1.598313570022583,
      "learning_rate": 1.3952890792291223e-05,
      "loss": 0.0847926139831543,
      "num_input_tokens_seen": 748358,
      "step": 712,
      "train_runtime": 2038.5998,
      "train_tokens_per_second": 367.094
    },
    {
      "epoch": 0.30473341168928303,
      "grad_norm": 2.8482844829559326,
      "learning_rate": 1.3944325481798717e-05,
      "loss": 0.15793046355247498,
      "num_input_tokens_seen": 749440,
      "step": 713,
      "train_runtime": 2040.7582,
      "train_tokens_per_second": 367.236
    },
    {
      "epoch": 0.3051608077786088,
      "grad_norm": 2.943387985229492,
      "learning_rate": 1.3935760171306212e-05,
      "loss": 0.16277006268501282,
      "num_input_tokens_seen": 750418,
      "step": 714,
      "train_runtime": 2042.9125,
      "train_tokens_per_second": 367.328
    },
    {
      "epoch": 0.3055882038679346,
      "grad_norm": 1.8870313167572021,
      "learning_rate": 1.3927194860813707e-05,
      "loss": 0.1313779503107071,
      "num_input_tokens_seen": 751646,
      "step": 715,
      "train_runtime": 2045.0793,
      "train_tokens_per_second": 367.539
    },
    {
      "epoch": 0.3060155999572604,
      "grad_norm": 2.739898920059204,
      "learning_rate": 1.39186295503212e-05,
      "loss": 0.15471625328063965,
      "num_input_tokens_seen": 752600,
      "step": 716,
      "train_runtime": 2047.2288,
      "train_tokens_per_second": 367.619
    },
    {
      "epoch": 0.30644299604658615,
      "grad_norm": 3.446117877960205,
      "learning_rate": 1.3910064239828696e-05,
      "loss": 0.23055478930473328,
      "num_input_tokens_seen": 753414,
      "step": 717,
      "train_runtime": 2049.3513,
      "train_tokens_per_second": 367.635
    },
    {
      "epoch": 0.30687039213591194,
      "grad_norm": 2.8714966773986816,
      "learning_rate": 1.390149892933619e-05,
      "loss": 0.1466040313243866,
      "num_input_tokens_seen": 754518,
      "step": 718,
      "train_runtime": 2051.5013,
      "train_tokens_per_second": 367.788
    },
    {
      "epoch": 0.30729778822523773,
      "grad_norm": 1.8891358375549316,
      "learning_rate": 1.3892933618843685e-05,
      "loss": 0.11850430071353912,
      "num_input_tokens_seen": 755588,
      "step": 719,
      "train_runtime": 2053.6493,
      "train_tokens_per_second": 367.925
    },
    {
      "epoch": 0.3077251843145635,
      "grad_norm": 2.7600414752960205,
      "learning_rate": 1.388436830835118e-05,
      "loss": 0.09389976412057877,
      "num_input_tokens_seen": 756476,
      "step": 720,
      "train_runtime": 2055.7764,
      "train_tokens_per_second": 367.976
    },
    {
      "epoch": 0.3081525804038893,
      "grad_norm": 1.9070271253585815,
      "learning_rate": 1.3875802997858674e-05,
      "loss": 0.0801122635602951,
      "num_input_tokens_seen": 757168,
      "step": 721,
      "train_runtime": 2071.1339,
      "train_tokens_per_second": 365.581
    },
    {
      "epoch": 0.3085799764932151,
      "grad_norm": 2.241255044937134,
      "learning_rate": 1.3867237687366169e-05,
      "loss": 0.1453324407339096,
      "num_input_tokens_seen": 758160,
      "step": 722,
      "train_runtime": 2073.2885,
      "train_tokens_per_second": 365.68
    },
    {
      "epoch": 0.30900737258254085,
      "grad_norm": 2.7590410709381104,
      "learning_rate": 1.3858672376873663e-05,
      "loss": 0.11476199328899384,
      "num_input_tokens_seen": 758940,
      "step": 723,
      "train_runtime": 2075.3622,
      "train_tokens_per_second": 365.69
    },
    {
      "epoch": 0.30943476867186664,
      "grad_norm": 1.64671790599823,
      "learning_rate": 1.3850107066381158e-05,
      "loss": 0.11244902014732361,
      "num_input_tokens_seen": 760078,
      "step": 724,
      "train_runtime": 2077.5823,
      "train_tokens_per_second": 365.847
    },
    {
      "epoch": 0.30986216476119244,
      "grad_norm": 2.1604344844818115,
      "learning_rate": 1.3841541755888651e-05,
      "loss": 0.10852448642253876,
      "num_input_tokens_seen": 761064,
      "step": 725,
      "train_runtime": 2079.7598,
      "train_tokens_per_second": 365.938
    },
    {
      "epoch": 0.31028956085051823,
      "grad_norm": 1.809300422668457,
      "learning_rate": 1.3832976445396147e-05,
      "loss": 0.1228833794593811,
      "num_input_tokens_seen": 762414,
      "step": 726,
      "train_runtime": 2081.9733,
      "train_tokens_per_second": 366.198
    },
    {
      "epoch": 0.310716956939844,
      "grad_norm": 2.001737356185913,
      "learning_rate": 1.3824411134903642e-05,
      "loss": 0.16599349677562714,
      "num_input_tokens_seen": 763692,
      "step": 727,
      "train_runtime": 2084.1395,
      "train_tokens_per_second": 366.43
    },
    {
      "epoch": 0.31114435302916976,
      "grad_norm": 2.912590742111206,
      "learning_rate": 1.3815845824411136e-05,
      "loss": 0.16752870380878448,
      "num_input_tokens_seen": 764738,
      "step": 728,
      "train_runtime": 2086.2942,
      "train_tokens_per_second": 366.553
    },
    {
      "epoch": 0.31157174911849556,
      "grad_norm": 2.347052812576294,
      "learning_rate": 1.380728051391863e-05,
      "loss": 0.12379346787929535,
      "num_input_tokens_seen": 766040,
      "step": 729,
      "train_runtime": 2088.4893,
      "train_tokens_per_second": 366.791
    },
    {
      "epoch": 0.31199914520782135,
      "grad_norm": 4.7478485107421875,
      "learning_rate": 1.3798715203426124e-05,
      "loss": 0.1677762120962143,
      "num_input_tokens_seen": 766918,
      "step": 730,
      "train_runtime": 2090.6408,
      "train_tokens_per_second": 366.834
    },
    {
      "epoch": 0.31242654129714714,
      "grad_norm": 2.88657283782959,
      "learning_rate": 1.379014989293362e-05,
      "loss": 0.08801712095737457,
      "num_input_tokens_seen": 767498,
      "step": 731,
      "train_runtime": 2092.6821,
      "train_tokens_per_second": 366.753
    },
    {
      "epoch": 0.31285393738647294,
      "grad_norm": 1.9307842254638672,
      "learning_rate": 1.3781584582441115e-05,
      "loss": 0.11593140661716461,
      "num_input_tokens_seen": 768748,
      "step": 732,
      "train_runtime": 2095.0139,
      "train_tokens_per_second": 366.942
    },
    {
      "epoch": 0.3132813334757987,
      "grad_norm": 1.8953204154968262,
      "learning_rate": 1.3773019271948609e-05,
      "loss": 0.11940370500087738,
      "num_input_tokens_seen": 769808,
      "step": 733,
      "train_runtime": 2097.2549,
      "train_tokens_per_second": 367.055
    },
    {
      "epoch": 0.31370872956512447,
      "grad_norm": 2.5668833255767822,
      "learning_rate": 1.3764453961456104e-05,
      "loss": 0.14816610515117645,
      "num_input_tokens_seen": 770774,
      "step": 734,
      "train_runtime": 2099.3714,
      "train_tokens_per_second": 367.145
    },
    {
      "epoch": 0.31413612565445026,
      "grad_norm": 2.624237298965454,
      "learning_rate": 1.3755888650963597e-05,
      "loss": 0.12266150116920471,
      "num_input_tokens_seen": 771764,
      "step": 735,
      "train_runtime": 2101.5157,
      "train_tokens_per_second": 367.242
    },
    {
      "epoch": 0.31456352174377605,
      "grad_norm": 6.235272407531738,
      "learning_rate": 1.3747323340471093e-05,
      "loss": 0.14392046630382538,
      "num_input_tokens_seen": 773010,
      "step": 736,
      "train_runtime": 2103.7358,
      "train_tokens_per_second": 367.446
    },
    {
      "epoch": 0.31499091783310185,
      "grad_norm": 2.490058183670044,
      "learning_rate": 1.3738758029978588e-05,
      "loss": 0.13403531908988953,
      "num_input_tokens_seen": 773894,
      "step": 737,
      "train_runtime": 2105.8501,
      "train_tokens_per_second": 367.497
    },
    {
      "epoch": 0.3154183139224276,
      "grad_norm": 1.781411051750183,
      "learning_rate": 1.3730192719486082e-05,
      "loss": 0.1307726800441742,
      "num_input_tokens_seen": 775084,
      "step": 738,
      "train_runtime": 2108.0468,
      "train_tokens_per_second": 367.679
    },
    {
      "epoch": 0.3158457100117534,
      "grad_norm": 2.6963393688201904,
      "learning_rate": 1.3721627408993577e-05,
      "loss": 0.15336059033870697,
      "num_input_tokens_seen": 776132,
      "step": 739,
      "train_runtime": 2110.1949,
      "train_tokens_per_second": 367.801
    },
    {
      "epoch": 0.31627310610107917,
      "grad_norm": 3.1183509826660156,
      "learning_rate": 1.371306209850107e-05,
      "loss": 0.12298072129487991,
      "num_input_tokens_seen": 776914,
      "step": 740,
      "train_runtime": 2112.2721,
      "train_tokens_per_second": 367.81
    },
    {
      "epoch": 0.31670050219040496,
      "grad_norm": 2.0433759689331055,
      "learning_rate": 1.3704496788008566e-05,
      "loss": 0.09010043740272522,
      "num_input_tokens_seen": 777732,
      "step": 741,
      "train_runtime": 2114.3313,
      "train_tokens_per_second": 367.838
    },
    {
      "epoch": 0.31712789827973076,
      "grad_norm": 2.5038232803344727,
      "learning_rate": 1.3695931477516061e-05,
      "loss": 0.10021007061004639,
      "num_input_tokens_seen": 778950,
      "step": 742,
      "train_runtime": 2116.5243,
      "train_tokens_per_second": 368.033
    },
    {
      "epoch": 0.3175552943690565,
      "grad_norm": 2.5027756690979004,
      "learning_rate": 1.3687366167023555e-05,
      "loss": 0.18917903304100037,
      "num_input_tokens_seen": 780304,
      "step": 743,
      "train_runtime": 2118.8,
      "train_tokens_per_second": 368.276
    },
    {
      "epoch": 0.3179826904583823,
      "grad_norm": 2.948503017425537,
      "learning_rate": 1.367880085653105e-05,
      "loss": 0.20733138918876648,
      "num_input_tokens_seen": 781224,
      "step": 744,
      "train_runtime": 2120.9459,
      "train_tokens_per_second": 368.338
    },
    {
      "epoch": 0.3184100865477081,
      "grad_norm": 3.159881114959717,
      "learning_rate": 1.3670235546038547e-05,
      "loss": 0.1322723925113678,
      "num_input_tokens_seen": 782016,
      "step": 745,
      "train_runtime": 2123.083,
      "train_tokens_per_second": 368.34
    },
    {
      "epoch": 0.3188374826370339,
      "grad_norm": 1.7335364818572998,
      "learning_rate": 1.366167023554604e-05,
      "loss": 0.10835819691419601,
      "num_input_tokens_seen": 783148,
      "step": 746,
      "train_runtime": 2125.2787,
      "train_tokens_per_second": 368.492
    },
    {
      "epoch": 0.31926487872635967,
      "grad_norm": 2.2973098754882812,
      "learning_rate": 1.3653104925053536e-05,
      "loss": 0.09291332215070724,
      "num_input_tokens_seen": 784318,
      "step": 747,
      "train_runtime": 2127.5169,
      "train_tokens_per_second": 368.654
    },
    {
      "epoch": 0.31969227481568546,
      "grad_norm": 3.436255693435669,
      "learning_rate": 1.364453961456103e-05,
      "loss": 0.13755182921886444,
      "num_input_tokens_seen": 785114,
      "step": 748,
      "train_runtime": 2129.6393,
      "train_tokens_per_second": 368.661
    },
    {
      "epoch": 0.3201196709050112,
      "grad_norm": 2.3020994663238525,
      "learning_rate": 1.3635974304068524e-05,
      "loss": 0.12800580263137817,
      "num_input_tokens_seen": 786298,
      "step": 749,
      "train_runtime": 2131.8349,
      "train_tokens_per_second": 368.836
    },
    {
      "epoch": 0.320547066994337,
      "grad_norm": 1.9654093980789185,
      "learning_rate": 1.362740899357602e-05,
      "loss": 0.10919762402772903,
      "num_input_tokens_seen": 787154,
      "step": 750,
      "train_runtime": 2133.9312,
      "train_tokens_per_second": 368.875
    },
    {
      "epoch": 0.3209744630836628,
      "grad_norm": 2.134575843811035,
      "learning_rate": 1.3618843683083513e-05,
      "loss": 0.1596805453300476,
      "num_input_tokens_seen": 788344,
      "step": 751,
      "train_runtime": 2149.3428,
      "train_tokens_per_second": 366.784
    },
    {
      "epoch": 0.3214018591729886,
      "grad_norm": 3.264194965362549,
      "learning_rate": 1.3610278372591009e-05,
      "loss": 0.15986016392707825,
      "num_input_tokens_seen": 789102,
      "step": 752,
      "train_runtime": 2151.5157,
      "train_tokens_per_second": 366.766
    },
    {
      "epoch": 0.3218292552623144,
      "grad_norm": 1.9836798906326294,
      "learning_rate": 1.3601713062098502e-05,
      "loss": 0.1384914517402649,
      "num_input_tokens_seen": 789858,
      "step": 753,
      "train_runtime": 2153.651,
      "train_tokens_per_second": 366.753
    },
    {
      "epoch": 0.3222566513516401,
      "grad_norm": 2.872135639190674,
      "learning_rate": 1.3593147751605997e-05,
      "loss": 0.15193700790405273,
      "num_input_tokens_seen": 790640,
      "step": 754,
      "train_runtime": 2155.7256,
      "train_tokens_per_second": 366.763
    },
    {
      "epoch": 0.3226840474409659,
      "grad_norm": 1.6906919479370117,
      "learning_rate": 1.3584582441113491e-05,
      "loss": 0.13117428123950958,
      "num_input_tokens_seen": 792164,
      "step": 755,
      "train_runtime": 2158.0063,
      "train_tokens_per_second": 367.081
    },
    {
      "epoch": 0.3231114435302917,
      "grad_norm": 2.4095849990844727,
      "learning_rate": 1.3576017130620986e-05,
      "loss": 0.16868722438812256,
      "num_input_tokens_seen": 793302,
      "step": 756,
      "train_runtime": 2160.2492,
      "train_tokens_per_second": 367.227
    },
    {
      "epoch": 0.3235388396196175,
      "grad_norm": 1.966367244720459,
      "learning_rate": 1.3567451820128482e-05,
      "loss": 0.07120862603187561,
      "num_input_tokens_seen": 794362,
      "step": 757,
      "train_runtime": 2162.4076,
      "train_tokens_per_second": 367.351
    },
    {
      "epoch": 0.3239662357089433,
      "grad_norm": 2.536763906478882,
      "learning_rate": 1.3558886509635975e-05,
      "loss": 0.1624632477760315,
      "num_input_tokens_seen": 795272,
      "step": 758,
      "train_runtime": 2164.5541,
      "train_tokens_per_second": 367.407
    },
    {
      "epoch": 0.324393631798269,
      "grad_norm": 2.2039806842803955,
      "learning_rate": 1.355032119914347e-05,
      "loss": 0.10303384065628052,
      "num_input_tokens_seen": 796026,
      "step": 759,
      "train_runtime": 2166.6513,
      "train_tokens_per_second": 367.399
    },
    {
      "epoch": 0.3248210278875948,
      "grad_norm": 1.6758956909179688,
      "learning_rate": 1.3541755888650964e-05,
      "loss": 0.1323976069688797,
      "num_input_tokens_seen": 797736,
      "step": 760,
      "train_runtime": 2168.977,
      "train_tokens_per_second": 367.794
    },
    {
      "epoch": 0.3252484239769206,
      "grad_norm": 2.06227445602417,
      "learning_rate": 1.353319057815846e-05,
      "loss": 0.11057798564434052,
      "num_input_tokens_seen": 798690,
      "step": 761,
      "train_runtime": 2171.4022,
      "train_tokens_per_second": 367.822
    },
    {
      "epoch": 0.3256758200662464,
      "grad_norm": 2.2499444484710693,
      "learning_rate": 1.3524625267665955e-05,
      "loss": 0.19061864912509918,
      "num_input_tokens_seen": 799628,
      "step": 762,
      "train_runtime": 2173.7393,
      "train_tokens_per_second": 367.858
    },
    {
      "epoch": 0.3261032161555722,
      "grad_norm": 1.8146963119506836,
      "learning_rate": 1.3516059957173448e-05,
      "loss": 0.11951874196529388,
      "num_input_tokens_seen": 800854,
      "step": 763,
      "train_runtime": 2175.9453,
      "train_tokens_per_second": 368.049
    },
    {
      "epoch": 0.32653061224489793,
      "grad_norm": 2.1884496212005615,
      "learning_rate": 1.3507494646680943e-05,
      "loss": 0.11111605912446976,
      "num_input_tokens_seen": 801748,
      "step": 764,
      "train_runtime": 2178.0772,
      "train_tokens_per_second": 368.099
    },
    {
      "epoch": 0.3269580083342237,
      "grad_norm": 2.0636422634124756,
      "learning_rate": 1.3498929336188437e-05,
      "loss": 0.12108628451824188,
      "num_input_tokens_seen": 802744,
      "step": 765,
      "train_runtime": 2180.2348,
      "train_tokens_per_second": 368.192
    },
    {
      "epoch": 0.3273854044235495,
      "grad_norm": 1.8975743055343628,
      "learning_rate": 1.3490364025695932e-05,
      "loss": 0.0649486631155014,
      "num_input_tokens_seen": 803590,
      "step": 766,
      "train_runtime": 2182.3315,
      "train_tokens_per_second": 368.225
    },
    {
      "epoch": 0.3278128005128753,
      "grad_norm": 2.349619150161743,
      "learning_rate": 1.3481798715203428e-05,
      "loss": 0.11989253759384155,
      "num_input_tokens_seen": 804376,
      "step": 767,
      "train_runtime": 2184.4579,
      "train_tokens_per_second": 368.227
    },
    {
      "epoch": 0.3282401966022011,
      "grad_norm": 1.6214401721954346,
      "learning_rate": 1.3473233404710921e-05,
      "loss": 0.11933603882789612,
      "num_input_tokens_seen": 805452,
      "step": 768,
      "train_runtime": 2186.6449,
      "train_tokens_per_second": 368.351
    },
    {
      "epoch": 0.3286675926915269,
      "grad_norm": 2.1668083667755127,
      "learning_rate": 1.3464668094218416e-05,
      "loss": 0.180108442902565,
      "num_input_tokens_seen": 806856,
      "step": 769,
      "train_runtime": 2188.8941,
      "train_tokens_per_second": 368.614
    },
    {
      "epoch": 0.32909498878085264,
      "grad_norm": 2.6501636505126953,
      "learning_rate": 1.345610278372591e-05,
      "loss": 0.1509145051240921,
      "num_input_tokens_seen": 807632,
      "step": 770,
      "train_runtime": 2190.9876,
      "train_tokens_per_second": 368.616
    },
    {
      "epoch": 0.32952238487017843,
      "grad_norm": 2.1489310264587402,
      "learning_rate": 1.3447537473233405e-05,
      "loss": 0.11179110407829285,
      "num_input_tokens_seen": 809556,
      "step": 771,
      "train_runtime": 2193.3574,
      "train_tokens_per_second": 369.094
    },
    {
      "epoch": 0.3299497809595042,
      "grad_norm": 2.5121138095855713,
      "learning_rate": 1.34389721627409e-05,
      "loss": 0.13696178793907166,
      "num_input_tokens_seen": 810648,
      "step": 772,
      "train_runtime": 2195.5295,
      "train_tokens_per_second": 369.227
    },
    {
      "epoch": 0.33037717704883,
      "grad_norm": 2.18746018409729,
      "learning_rate": 1.3430406852248394e-05,
      "loss": 0.117298424243927,
      "num_input_tokens_seen": 811464,
      "step": 773,
      "train_runtime": 2197.631,
      "train_tokens_per_second": 369.245
    },
    {
      "epoch": 0.3308045731381558,
      "grad_norm": 2.4381191730499268,
      "learning_rate": 1.342184154175589e-05,
      "loss": 0.1251356452703476,
      "num_input_tokens_seen": 812348,
      "step": 774,
      "train_runtime": 2199.7416,
      "train_tokens_per_second": 369.292
    },
    {
      "epoch": 0.33123196922748155,
      "grad_norm": 1.8118752241134644,
      "learning_rate": 1.3413276231263383e-05,
      "loss": 0.10888133198022842,
      "num_input_tokens_seen": 813656,
      "step": 775,
      "train_runtime": 2202.0171,
      "train_tokens_per_second": 369.505
    },
    {
      "epoch": 0.33165936531680734,
      "grad_norm": 2.366809368133545,
      "learning_rate": 1.3404710920770878e-05,
      "loss": 0.14520832896232605,
      "num_input_tokens_seen": 814622,
      "step": 776,
      "train_runtime": 2204.1903,
      "train_tokens_per_second": 369.579
    },
    {
      "epoch": 0.33208676140613314,
      "grad_norm": 2.3347342014312744,
      "learning_rate": 1.3396145610278372e-05,
      "loss": 0.1669357419013977,
      "num_input_tokens_seen": 815802,
      "step": 777,
      "train_runtime": 2206.39,
      "train_tokens_per_second": 369.745
    },
    {
      "epoch": 0.33251415749545893,
      "grad_norm": 2.4885172843933105,
      "learning_rate": 1.3387580299785867e-05,
      "loss": 0.13153831660747528,
      "num_input_tokens_seen": 816674,
      "step": 778,
      "train_runtime": 2208.5061,
      "train_tokens_per_second": 369.786
    },
    {
      "epoch": 0.3329415535847847,
      "grad_norm": 1.6821556091308594,
      "learning_rate": 1.3379014989293364e-05,
      "loss": 0.12667052447795868,
      "num_input_tokens_seen": 817948,
      "step": 779,
      "train_runtime": 2210.7245,
      "train_tokens_per_second": 369.991
    },
    {
      "epoch": 0.33336894967411046,
      "grad_norm": 1.956593632698059,
      "learning_rate": 1.337044967880086e-05,
      "loss": 0.12548109889030457,
      "num_input_tokens_seen": 819274,
      "step": 780,
      "train_runtime": 2212.9082,
      "train_tokens_per_second": 370.225
    },
    {
      "epoch": 0.33379634576343625,
      "grad_norm": 2.7756547927856445,
      "learning_rate": 1.3361884368308353e-05,
      "loss": 0.14985978603363037,
      "num_input_tokens_seen": 820162,
      "step": 781,
      "train_runtime": 2229.122,
      "train_tokens_per_second": 367.931
    },
    {
      "epoch": 0.33422374185276205,
      "grad_norm": 3.177536964416504,
      "learning_rate": 1.3353319057815848e-05,
      "loss": 0.2057100236415863,
      "num_input_tokens_seen": 820874,
      "step": 782,
      "train_runtime": 2231.23,
      "train_tokens_per_second": 367.902
    },
    {
      "epoch": 0.33465113794208784,
      "grad_norm": 2.3017020225524902,
      "learning_rate": 1.3344753747323342e-05,
      "loss": 0.1477176547050476,
      "num_input_tokens_seen": 822422,
      "step": 783,
      "train_runtime": 2233.5232,
      "train_tokens_per_second": 368.217
    },
    {
      "epoch": 0.33507853403141363,
      "grad_norm": 1.2898216247558594,
      "learning_rate": 1.3336188436830837e-05,
      "loss": 0.06120645999908447,
      "num_input_tokens_seen": 824162,
      "step": 784,
      "train_runtime": 2235.9328,
      "train_tokens_per_second": 368.599
    },
    {
      "epoch": 0.33550593012073937,
      "grad_norm": 2.308602809906006,
      "learning_rate": 1.332762312633833e-05,
      "loss": 0.12190292775630951,
      "num_input_tokens_seen": 825056,
      "step": 785,
      "train_runtime": 2238.0858,
      "train_tokens_per_second": 368.644
    },
    {
      "epoch": 0.33593332621006516,
      "grad_norm": 2.5072407722473145,
      "learning_rate": 1.3319057815845826e-05,
      "loss": 0.17140378057956696,
      "num_input_tokens_seen": 826424,
      "step": 786,
      "train_runtime": 2240.3306,
      "train_tokens_per_second": 368.885
    },
    {
      "epoch": 0.33636072229939096,
      "grad_norm": 2.389327049255371,
      "learning_rate": 1.3310492505353321e-05,
      "loss": 0.12005665898323059,
      "num_input_tokens_seen": 827710,
      "step": 787,
      "train_runtime": 2242.5895,
      "train_tokens_per_second": 369.087
    },
    {
      "epoch": 0.33678811838871675,
      "grad_norm": 1.7238560914993286,
      "learning_rate": 1.3301927194860815e-05,
      "loss": 0.09626015275716782,
      "num_input_tokens_seen": 828636,
      "step": 788,
      "train_runtime": 2244.749,
      "train_tokens_per_second": 369.144
    },
    {
      "epoch": 0.33721551447804254,
      "grad_norm": 2.4804675579071045,
      "learning_rate": 1.329336188436831e-05,
      "loss": 0.1318286955356598,
      "num_input_tokens_seen": 829302,
      "step": 789,
      "train_runtime": 2246.8169,
      "train_tokens_per_second": 369.101
    },
    {
      "epoch": 0.3376429105673683,
      "grad_norm": 4.671977519989014,
      "learning_rate": 1.3284796573875804e-05,
      "loss": 0.1203911304473877,
      "num_input_tokens_seen": 829892,
      "step": 790,
      "train_runtime": 2248.9158,
      "train_tokens_per_second": 369.019
    },
    {
      "epoch": 0.3380703066566941,
      "grad_norm": 2.4786596298217773,
      "learning_rate": 1.3276231263383299e-05,
      "loss": 0.14200855791568756,
      "num_input_tokens_seen": 830962,
      "step": 791,
      "train_runtime": 2251.1883,
      "train_tokens_per_second": 369.122
    },
    {
      "epoch": 0.33849770274601987,
      "grad_norm": 1.7750545740127563,
      "learning_rate": 1.3267665952890794e-05,
      "loss": 0.09182557463645935,
      "num_input_tokens_seen": 831794,
      "step": 792,
      "train_runtime": 2253.3395,
      "train_tokens_per_second": 369.138
    },
    {
      "epoch": 0.33892509883534566,
      "grad_norm": 2.7021842002868652,
      "learning_rate": 1.3259100642398288e-05,
      "loss": 0.06427270174026489,
      "num_input_tokens_seen": 832656,
      "step": 793,
      "train_runtime": 2255.4249,
      "train_tokens_per_second": 369.179
    },
    {
      "epoch": 0.33935249492467146,
      "grad_norm": 1.5075496435165405,
      "learning_rate": 1.3250535331905783e-05,
      "loss": 0.09556608647108078,
      "num_input_tokens_seen": 834142,
      "step": 794,
      "train_runtime": 2257.6939,
      "train_tokens_per_second": 369.466
    },
    {
      "epoch": 0.33977989101399725,
      "grad_norm": 2.787081003189087,
      "learning_rate": 1.3241970021413277e-05,
      "loss": 0.16105836629867554,
      "num_input_tokens_seen": 834814,
      "step": 795,
      "train_runtime": 2259.742,
      "train_tokens_per_second": 369.429
    },
    {
      "epoch": 0.340207287103323,
      "grad_norm": 2.8966727256774902,
      "learning_rate": 1.3233404710920772e-05,
      "loss": 0.13476288318634033,
      "num_input_tokens_seen": 835578,
      "step": 796,
      "train_runtime": 2261.8383,
      "train_tokens_per_second": 369.424
    },
    {
      "epoch": 0.3406346831926488,
      "grad_norm": 2.376763343811035,
      "learning_rate": 1.3224839400428267e-05,
      "loss": 0.10697012394666672,
      "num_input_tokens_seen": 836666,
      "step": 797,
      "train_runtime": 2264.0251,
      "train_tokens_per_second": 369.548
    },
    {
      "epoch": 0.3410620792819746,
      "grad_norm": 1.684907078742981,
      "learning_rate": 1.321627408993576e-05,
      "loss": 0.12375549972057343,
      "num_input_tokens_seen": 837820,
      "step": 798,
      "train_runtime": 2266.1997,
      "train_tokens_per_second": 369.703
    },
    {
      "epoch": 0.34148947537130037,
      "grad_norm": 3.0372512340545654,
      "learning_rate": 1.3207708779443256e-05,
      "loss": 0.1351826936006546,
      "num_input_tokens_seen": 838552,
      "step": 799,
      "train_runtime": 2268.3223,
      "train_tokens_per_second": 369.679
    },
    {
      "epoch": 0.34191687146062616,
      "grad_norm": 2.712702512741089,
      "learning_rate": 1.319914346895075e-05,
      "loss": 0.1467861831188202,
      "num_input_tokens_seen": 839356,
      "step": 800,
      "train_runtime": 2270.4387,
      "train_tokens_per_second": 369.689
    },
    {
      "epoch": 0.3423442675499519,
      "grad_norm": 4.493462085723877,
      "learning_rate": 1.3190578158458245e-05,
      "loss": 0.1468145251274109,
      "num_input_tokens_seen": 839988,
      "step": 801,
      "train_runtime": 2272.4955,
      "train_tokens_per_second": 369.632
    },
    {
      "epoch": 0.3427716636392777,
      "grad_norm": 2.101985454559326,
      "learning_rate": 1.318201284796574e-05,
      "loss": 0.14003314077854156,
      "num_input_tokens_seen": 840830,
      "step": 802,
      "train_runtime": 2274.5722,
      "train_tokens_per_second": 369.665
    },
    {
      "epoch": 0.3431990597286035,
      "grad_norm": 2.5873258113861084,
      "learning_rate": 1.3173447537473234e-05,
      "loss": 0.13785353302955627,
      "num_input_tokens_seen": 841794,
      "step": 803,
      "train_runtime": 2276.7517,
      "train_tokens_per_second": 369.735
    },
    {
      "epoch": 0.3436264558179293,
      "grad_norm": 2.3825364112854004,
      "learning_rate": 1.3164882226980729e-05,
      "loss": 0.08509097993373871,
      "num_input_tokens_seen": 842524,
      "step": 804,
      "train_runtime": 2278.8313,
      "train_tokens_per_second": 369.718
    },
    {
      "epoch": 0.34405385190725507,
      "grad_norm": 2.2430174350738525,
      "learning_rate": 1.3156316916488223e-05,
      "loss": 0.21838538348674774,
      "num_input_tokens_seen": 844360,
      "step": 805,
      "train_runtime": 2281.1977,
      "train_tokens_per_second": 370.139
    },
    {
      "epoch": 0.3444812479965808,
      "grad_norm": 2.4562454223632812,
      "learning_rate": 1.3147751605995718e-05,
      "loss": 0.24757874011993408,
      "num_input_tokens_seen": 846190,
      "step": 806,
      "train_runtime": 2283.5375,
      "train_tokens_per_second": 370.561
    },
    {
      "epoch": 0.3449086440859066,
      "grad_norm": 2.5909478664398193,
      "learning_rate": 1.3139186295503212e-05,
      "loss": 0.1682937741279602,
      "num_input_tokens_seen": 847272,
      "step": 807,
      "train_runtime": 2285.7163,
      "train_tokens_per_second": 370.681
    },
    {
      "epoch": 0.3453360401752324,
      "grad_norm": 2.206904649734497,
      "learning_rate": 1.3130620985010707e-05,
      "loss": 0.08844608813524246,
      "num_input_tokens_seen": 848110,
      "step": 808,
      "train_runtime": 2287.8349,
      "train_tokens_per_second": 370.704
    },
    {
      "epoch": 0.3457634362645582,
      "grad_norm": 1.9793843030929565,
      "learning_rate": 1.3122055674518202e-05,
      "loss": 0.13883069157600403,
      "num_input_tokens_seen": 849084,
      "step": 809,
      "train_runtime": 2289.9524,
      "train_tokens_per_second": 370.787
    },
    {
      "epoch": 0.346190832353884,
      "grad_norm": 1.5459498167037964,
      "learning_rate": 1.3113490364025696e-05,
      "loss": 0.06009524688124657,
      "num_input_tokens_seen": 850498,
      "step": 810,
      "train_runtime": 2292.2369,
      "train_tokens_per_second": 371.034
    },
    {
      "epoch": 0.3466182284432097,
      "grad_norm": 2.364461898803711,
      "learning_rate": 1.3104925053533191e-05,
      "loss": 0.1430809646844864,
      "num_input_tokens_seen": 851400,
      "step": 811,
      "train_runtime": 2307.1808,
      "train_tokens_per_second": 369.022
    },
    {
      "epoch": 0.3470456245325355,
      "grad_norm": 2.2134151458740234,
      "learning_rate": 1.3096359743040685e-05,
      "loss": 0.16521531343460083,
      "num_input_tokens_seen": 853148,
      "step": 812,
      "train_runtime": 2309.5563,
      "train_tokens_per_second": 369.399
    },
    {
      "epoch": 0.3474730206218613,
      "grad_norm": 2.4093143939971924,
      "learning_rate": 1.3087794432548182e-05,
      "loss": 0.13387048244476318,
      "num_input_tokens_seen": 853884,
      "step": 813,
      "train_runtime": 2311.6315,
      "train_tokens_per_second": 369.386
    },
    {
      "epoch": 0.3479004167111871,
      "grad_norm": 1.807502269744873,
      "learning_rate": 1.3079229122055677e-05,
      "loss": 0.1432010978460312,
      "num_input_tokens_seen": 855732,
      "step": 814,
      "train_runtime": 2313.9832,
      "train_tokens_per_second": 369.809
    },
    {
      "epoch": 0.3483278128005129,
      "grad_norm": 2.284177780151367,
      "learning_rate": 1.307066381156317e-05,
      "loss": 0.11023642122745514,
      "num_input_tokens_seen": 856782,
      "step": 815,
      "train_runtime": 2316.1194,
      "train_tokens_per_second": 369.921
    },
    {
      "epoch": 0.34875520888983863,
      "grad_norm": 2.2972710132598877,
      "learning_rate": 1.3062098501070666e-05,
      "loss": 0.1349591463804245,
      "num_input_tokens_seen": 857708,
      "step": 816,
      "train_runtime": 2318.2633,
      "train_tokens_per_second": 369.979
    },
    {
      "epoch": 0.3491826049791644,
      "grad_norm": 3.549760580062866,
      "learning_rate": 1.3053533190578161e-05,
      "loss": 0.1933775544166565,
      "num_input_tokens_seen": 858566,
      "step": 817,
      "train_runtime": 2320.4118,
      "train_tokens_per_second": 370.006
    },
    {
      "epoch": 0.3496100010684902,
      "grad_norm": 1.7501996755599976,
      "learning_rate": 1.3044967880085655e-05,
      "loss": 0.09176300466060638,
      "num_input_tokens_seen": 859472,
      "step": 818,
      "train_runtime": 2322.5072,
      "train_tokens_per_second": 370.062
    },
    {
      "epoch": 0.350037397157816,
      "grad_norm": 2.8815555572509766,
      "learning_rate": 1.303640256959315e-05,
      "loss": 0.15270407497882843,
      "num_input_tokens_seen": 860164,
      "step": 819,
      "train_runtime": 2324.5928,
      "train_tokens_per_second": 370.028
    },
    {
      "epoch": 0.3504647932471418,
      "grad_norm": 2.577434539794922,
      "learning_rate": 1.3027837259100643e-05,
      "loss": 0.15686242282390594,
      "num_input_tokens_seen": 861046,
      "step": 820,
      "train_runtime": 2326.6783,
      "train_tokens_per_second": 370.075
    },
    {
      "epoch": 0.3508921893364676,
      "grad_norm": 1.9763771295547485,
      "learning_rate": 1.3019271948608139e-05,
      "loss": 0.1506127268075943,
      "num_input_tokens_seen": 862272,
      "step": 821,
      "train_runtime": 2328.8998,
      "train_tokens_per_second": 370.249
    },
    {
      "epoch": 0.35131958542579333,
      "grad_norm": 2.162661552429199,
      "learning_rate": 1.3010706638115634e-05,
      "loss": 0.07952255755662918,
      "num_input_tokens_seen": 862822,
      "step": 822,
      "train_runtime": 2330.9876,
      "train_tokens_per_second": 370.153
    },
    {
      "epoch": 0.35174698151511913,
      "grad_norm": 2.170814275741577,
      "learning_rate": 1.3002141327623127e-05,
      "loss": 0.1116003468632698,
      "num_input_tokens_seen": 863764,
      "step": 823,
      "train_runtime": 2333.2459,
      "train_tokens_per_second": 370.198
    },
    {
      "epoch": 0.3521743776044449,
      "grad_norm": 1.9348992109298706,
      "learning_rate": 1.2993576017130623e-05,
      "loss": 0.09650216996669769,
      "num_input_tokens_seen": 864614,
      "step": 824,
      "train_runtime": 2335.3669,
      "train_tokens_per_second": 370.226
    },
    {
      "epoch": 0.3526017736937707,
      "grad_norm": 2.208080291748047,
      "learning_rate": 1.2985010706638116e-05,
      "loss": 0.11872360855340958,
      "num_input_tokens_seen": 865604,
      "step": 825,
      "train_runtime": 2337.4798,
      "train_tokens_per_second": 370.315
    },
    {
      "epoch": 0.3530291697830965,
      "grad_norm": 2.548844814300537,
      "learning_rate": 1.2976445396145612e-05,
      "loss": 0.18330249190330505,
      "num_input_tokens_seen": 866716,
      "step": 826,
      "train_runtime": 2339.6348,
      "train_tokens_per_second": 370.449
    },
    {
      "epoch": 0.35345656587242225,
      "grad_norm": 1.9747850894927979,
      "learning_rate": 1.2967880085653107e-05,
      "loss": 0.16249194741249084,
      "num_input_tokens_seen": 868744,
      "step": 827,
      "train_runtime": 2342.0394,
      "train_tokens_per_second": 370.935
    },
    {
      "epoch": 0.35388396196174804,
      "grad_norm": 2.2463526725769043,
      "learning_rate": 1.29593147751606e-05,
      "loss": 0.11855243146419525,
      "num_input_tokens_seen": 869632,
      "step": 828,
      "train_runtime": 2344.1756,
      "train_tokens_per_second": 370.976
    },
    {
      "epoch": 0.35431135805107383,
      "grad_norm": 3.0636091232299805,
      "learning_rate": 1.2950749464668096e-05,
      "loss": 0.1560877561569214,
      "num_input_tokens_seen": 870562,
      "step": 829,
      "train_runtime": 2346.3134,
      "train_tokens_per_second": 371.034
    },
    {
      "epoch": 0.3547387541403996,
      "grad_norm": 2.7666077613830566,
      "learning_rate": 1.294218415417559e-05,
      "loss": 0.13841849565505981,
      "num_input_tokens_seen": 871682,
      "step": 830,
      "train_runtime": 2348.5052,
      "train_tokens_per_second": 371.165
    },
    {
      "epoch": 0.3551661502297254,
      "grad_norm": 3.7168400287628174,
      "learning_rate": 1.2933618843683085e-05,
      "loss": 0.13689354062080383,
      "num_input_tokens_seen": 872416,
      "step": 831,
      "train_runtime": 2350.5548,
      "train_tokens_per_second": 371.153
    },
    {
      "epoch": 0.35559354631905116,
      "grad_norm": 3.724029302597046,
      "learning_rate": 1.292505353319058e-05,
      "loss": 0.15330494940280914,
      "num_input_tokens_seen": 873606,
      "step": 832,
      "train_runtime": 2352.7303,
      "train_tokens_per_second": 371.316
    },
    {
      "epoch": 0.35602094240837695,
      "grad_norm": 2.799285650253296,
      "learning_rate": 1.2916488222698073e-05,
      "loss": 0.1891564428806305,
      "num_input_tokens_seen": 874724,
      "step": 833,
      "train_runtime": 2354.9017,
      "train_tokens_per_second": 371.448
    },
    {
      "epoch": 0.35644833849770274,
      "grad_norm": 2.5742807388305664,
      "learning_rate": 1.2907922912205569e-05,
      "loss": 0.10590503364801407,
      "num_input_tokens_seen": 875570,
      "step": 834,
      "train_runtime": 2357.0482,
      "train_tokens_per_second": 371.469
    },
    {
      "epoch": 0.35687573458702854,
      "grad_norm": 1.5430233478546143,
      "learning_rate": 1.2899357601713062e-05,
      "loss": 0.11114594340324402,
      "num_input_tokens_seen": 876702,
      "step": 835,
      "train_runtime": 2359.2054,
      "train_tokens_per_second": 371.609
    },
    {
      "epoch": 0.35730313067635433,
      "grad_norm": 2.3434317111968994,
      "learning_rate": 1.2890792291220558e-05,
      "loss": 0.10748954117298126,
      "num_input_tokens_seen": 877628,
      "step": 836,
      "train_runtime": 2361.3383,
      "train_tokens_per_second": 371.666
    },
    {
      "epoch": 0.35773052676568007,
      "grad_norm": 1.9241222143173218,
      "learning_rate": 1.2882226980728051e-05,
      "loss": 0.06413917988538742,
      "num_input_tokens_seen": 878390,
      "step": 837,
      "train_runtime": 2363.457,
      "train_tokens_per_second": 371.655
    },
    {
      "epoch": 0.35815792285500586,
      "grad_norm": 1.275365948677063,
      "learning_rate": 1.2873661670235546e-05,
      "loss": 0.07867828011512756,
      "num_input_tokens_seen": 880084,
      "step": 838,
      "train_runtime": 2365.7913,
      "train_tokens_per_second": 372.004
    },
    {
      "epoch": 0.35858531894433165,
      "grad_norm": 2.6556649208068848,
      "learning_rate": 1.2865096359743042e-05,
      "loss": 0.11712068319320679,
      "num_input_tokens_seen": 881128,
      "step": 839,
      "train_runtime": 2367.9188,
      "train_tokens_per_second": 372.111
    },
    {
      "epoch": 0.35901271503365745,
      "grad_norm": 3.0657029151916504,
      "learning_rate": 1.2856531049250535e-05,
      "loss": 0.20221734046936035,
      "num_input_tokens_seen": 881942,
      "step": 840,
      "train_runtime": 2370.0139,
      "train_tokens_per_second": 372.125
    },
    {
      "epoch": 0.35944011112298324,
      "grad_norm": 2.1402385234832764,
      "learning_rate": 1.284796573875803e-05,
      "loss": 0.10370795428752899,
      "num_input_tokens_seen": 882700,
      "step": 841,
      "train_runtime": 2385.2893,
      "train_tokens_per_second": 370.06
    },
    {
      "epoch": 0.35986750721230903,
      "grad_norm": 1.999263882637024,
      "learning_rate": 1.2839400428265524e-05,
      "loss": 0.14306849241256714,
      "num_input_tokens_seen": 883724,
      "step": 842,
      "train_runtime": 2387.4028,
      "train_tokens_per_second": 370.161
    },
    {
      "epoch": 0.3602949033016348,
      "grad_norm": 2.2319602966308594,
      "learning_rate": 1.283083511777302e-05,
      "loss": 0.13047417998313904,
      "num_input_tokens_seen": 884510,
      "step": 843,
      "train_runtime": 2389.5083,
      "train_tokens_per_second": 370.164
    },
    {
      "epoch": 0.36072229939096057,
      "grad_norm": 1.8456926345825195,
      "learning_rate": 1.2822269807280515e-05,
      "loss": 0.11278331279754639,
      "num_input_tokens_seen": 885522,
      "step": 844,
      "train_runtime": 2391.6712,
      "train_tokens_per_second": 370.252
    },
    {
      "epoch": 0.36114969548028636,
      "grad_norm": 1.9352658987045288,
      "learning_rate": 1.2813704496788008e-05,
      "loss": 0.10232369601726532,
      "num_input_tokens_seen": 886520,
      "step": 845,
      "train_runtime": 2393.78,
      "train_tokens_per_second": 370.343
    },
    {
      "epoch": 0.36157709156961215,
      "grad_norm": 1.9985121488571167,
      "learning_rate": 1.2805139186295504e-05,
      "loss": 0.12468145042657852,
      "num_input_tokens_seen": 887478,
      "step": 846,
      "train_runtime": 2395.9232,
      "train_tokens_per_second": 370.412
    },
    {
      "epoch": 0.36200448765893795,
      "grad_norm": 2.997879981994629,
      "learning_rate": 1.2796573875802997e-05,
      "loss": 0.10673297941684723,
      "num_input_tokens_seen": 888142,
      "step": 847,
      "train_runtime": 2397.9983,
      "train_tokens_per_second": 370.368
    },
    {
      "epoch": 0.3624318837482637,
      "grad_norm": 2.8348140716552734,
      "learning_rate": 1.2788008565310494e-05,
      "loss": 0.152587428689003,
      "num_input_tokens_seen": 889178,
      "step": 848,
      "train_runtime": 2400.1417,
      "train_tokens_per_second": 370.469
    },
    {
      "epoch": 0.3628592798375895,
      "grad_norm": 1.6546398401260376,
      "learning_rate": 1.277944325481799e-05,
      "loss": 0.12723231315612793,
      "num_input_tokens_seen": 890746,
      "step": 849,
      "train_runtime": 2402.4217,
      "train_tokens_per_second": 370.77
    },
    {
      "epoch": 0.36328667592691527,
      "grad_norm": 2.3017971515655518,
      "learning_rate": 1.2770877944325483e-05,
      "loss": 0.11856281757354736,
      "num_input_tokens_seen": 891694,
      "step": 850,
      "train_runtime": 2404.5416,
      "train_tokens_per_second": 370.837
    },
    {
      "epoch": 0.36371407201624106,
      "grad_norm": 2.0273659229278564,
      "learning_rate": 1.2762312633832978e-05,
      "loss": 0.17470906674861908,
      "num_input_tokens_seen": 892998,
      "step": 851,
      "train_runtime": 2406.7755,
      "train_tokens_per_second": 371.035
    },
    {
      "epoch": 0.36414146810556686,
      "grad_norm": 3.1824123859405518,
      "learning_rate": 1.2753747323340474e-05,
      "loss": 0.15598653256893158,
      "num_input_tokens_seen": 893882,
      "step": 852,
      "train_runtime": 2409.0877,
      "train_tokens_per_second": 371.046
    },
    {
      "epoch": 0.3645688641948926,
      "grad_norm": 1.8734511137008667,
      "learning_rate": 1.2745182012847967e-05,
      "loss": 0.11133860796689987,
      "num_input_tokens_seen": 894888,
      "step": 853,
      "train_runtime": 2411.2765,
      "train_tokens_per_second": 371.126
    },
    {
      "epoch": 0.3649962602842184,
      "grad_norm": 2.210627317428589,
      "learning_rate": 1.2736616702355462e-05,
      "loss": 0.20583584904670715,
      "num_input_tokens_seen": 896216,
      "step": 854,
      "train_runtime": 2413.4897,
      "train_tokens_per_second": 371.336
    },
    {
      "epoch": 0.3654236563735442,
      "grad_norm": 2.089354991912842,
      "learning_rate": 1.2728051391862956e-05,
      "loss": 0.09439928084611893,
      "num_input_tokens_seen": 897404,
      "step": 855,
      "train_runtime": 2415.6801,
      "train_tokens_per_second": 371.491
    },
    {
      "epoch": 0.36585105246287,
      "grad_norm": 2.5480873584747314,
      "learning_rate": 1.2719486081370451e-05,
      "loss": 0.15684294700622559,
      "num_input_tokens_seen": 898318,
      "step": 856,
      "train_runtime": 2417.8349,
      "train_tokens_per_second": 371.538
    },
    {
      "epoch": 0.36627844855219577,
      "grad_norm": 1.8332490921020508,
      "learning_rate": 1.2710920770877947e-05,
      "loss": 0.12645269930362701,
      "num_input_tokens_seen": 899926,
      "step": 857,
      "train_runtime": 2420.1588,
      "train_tokens_per_second": 371.846
    },
    {
      "epoch": 0.3667058446415215,
      "grad_norm": 2.9024620056152344,
      "learning_rate": 1.270235546038544e-05,
      "loss": 0.12869933247566223,
      "num_input_tokens_seen": 901122,
      "step": 858,
      "train_runtime": 2422.3849,
      "train_tokens_per_second": 371.998
    },
    {
      "epoch": 0.3671332407308473,
      "grad_norm": 1.7431743144989014,
      "learning_rate": 1.2693790149892935e-05,
      "loss": 0.07658876478672028,
      "num_input_tokens_seen": 902054,
      "step": 859,
      "train_runtime": 2424.5573,
      "train_tokens_per_second": 372.049
    },
    {
      "epoch": 0.3675606368201731,
      "grad_norm": 2.8482065200805664,
      "learning_rate": 1.2685224839400429e-05,
      "loss": 0.13861551880836487,
      "num_input_tokens_seen": 902682,
      "step": 860,
      "train_runtime": 2426.6309,
      "train_tokens_per_second": 371.99
    },
    {
      "epoch": 0.3679880329094989,
      "grad_norm": 2.4064478874206543,
      "learning_rate": 1.2676659528907924e-05,
      "loss": 0.18145975470542908,
      "num_input_tokens_seen": 903482,
      "step": 861,
      "train_runtime": 2428.6922,
      "train_tokens_per_second": 372.003
    },
    {
      "epoch": 0.3684154289988247,
      "grad_norm": 2.8364784717559814,
      "learning_rate": 1.266809421841542e-05,
      "loss": 0.13799196481704712,
      "num_input_tokens_seen": 904300,
      "step": 862,
      "train_runtime": 2430.7801,
      "train_tokens_per_second": 372.02
    },
    {
      "epoch": 0.3688428250881504,
      "grad_norm": 2.464657783508301,
      "learning_rate": 1.2659528907922913e-05,
      "loss": 0.1520114690065384,
      "num_input_tokens_seen": 905288,
      "step": 863,
      "train_runtime": 2432.8982,
      "train_tokens_per_second": 372.103
    },
    {
      "epoch": 0.3692702211774762,
      "grad_norm": 2.8625905513763428,
      "learning_rate": 1.2650963597430408e-05,
      "loss": 0.18288902938365936,
      "num_input_tokens_seen": 906054,
      "step": 864,
      "train_runtime": 2434.9779,
      "train_tokens_per_second": 372.099
    },
    {
      "epoch": 0.369697617266802,
      "grad_norm": 2.188767910003662,
      "learning_rate": 1.2642398286937902e-05,
      "loss": 0.09414643049240112,
      "num_input_tokens_seen": 907008,
      "step": 865,
      "train_runtime": 2437.1013,
      "train_tokens_per_second": 372.167
    },
    {
      "epoch": 0.3701250133561278,
      "grad_norm": 2.105830430984497,
      "learning_rate": 1.2633832976445397e-05,
      "loss": 0.12260585278272629,
      "num_input_tokens_seen": 908128,
      "step": 866,
      "train_runtime": 2439.2785,
      "train_tokens_per_second": 372.294
    },
    {
      "epoch": 0.3705524094454536,
      "grad_norm": 2.7027125358581543,
      "learning_rate": 1.2625267665952891e-05,
      "loss": 0.14837321639060974,
      "num_input_tokens_seen": 909008,
      "step": 867,
      "train_runtime": 2441.3686,
      "train_tokens_per_second": 372.335
    },
    {
      "epoch": 0.3709798055347794,
      "grad_norm": 2.140458345413208,
      "learning_rate": 1.2616702355460386e-05,
      "loss": 0.1417008936405182,
      "num_input_tokens_seen": 910658,
      "step": 868,
      "train_runtime": 2443.6472,
      "train_tokens_per_second": 372.663
    },
    {
      "epoch": 0.3714072016241051,
      "grad_norm": 3.9710843563079834,
      "learning_rate": 1.2608137044967881e-05,
      "loss": 0.1838308721780777,
      "num_input_tokens_seen": 911350,
      "step": 869,
      "train_runtime": 2445.7131,
      "train_tokens_per_second": 372.632
    },
    {
      "epoch": 0.3718345977134309,
      "grad_norm": 2.5963730812072754,
      "learning_rate": 1.2599571734475375e-05,
      "loss": 0.12967392802238464,
      "num_input_tokens_seen": 912134,
      "step": 870,
      "train_runtime": 2447.7973,
      "train_tokens_per_second": 372.635
    },
    {
      "epoch": 0.3722619938027567,
      "grad_norm": 2.5910768508911133,
      "learning_rate": 1.259100642398287e-05,
      "loss": 0.1894197314977646,
      "num_input_tokens_seen": 913428,
      "step": 871,
      "train_runtime": 2463.7444,
      "train_tokens_per_second": 370.748
    },
    {
      "epoch": 0.3726893898920825,
      "grad_norm": 1.9151437282562256,
      "learning_rate": 1.2582441113490364e-05,
      "loss": 0.10640452057123184,
      "num_input_tokens_seen": 914398,
      "step": 872,
      "train_runtime": 2465.8343,
      "train_tokens_per_second": 370.827
    },
    {
      "epoch": 0.3731167859814083,
      "grad_norm": 2.3632760047912598,
      "learning_rate": 1.2573875802997859e-05,
      "loss": 0.10356519371271133,
      "num_input_tokens_seen": 915234,
      "step": 873,
      "train_runtime": 2467.9278,
      "train_tokens_per_second": 370.851
    },
    {
      "epoch": 0.37354418207073403,
      "grad_norm": 2.1578476428985596,
      "learning_rate": 1.2565310492505354e-05,
      "loss": 0.13019436597824097,
      "num_input_tokens_seen": 916102,
      "step": 874,
      "train_runtime": 2470.0829,
      "train_tokens_per_second": 370.879
    },
    {
      "epoch": 0.3739715781600598,
      "grad_norm": 2.8030753135681152,
      "learning_rate": 1.2556745182012848e-05,
      "loss": 0.11686091125011444,
      "num_input_tokens_seen": 916690,
      "step": 875,
      "train_runtime": 2472.1112,
      "train_tokens_per_second": 370.813
    },
    {
      "epoch": 0.3743989742493856,
      "grad_norm": 1.841654658317566,
      "learning_rate": 1.2548179871520343e-05,
      "loss": 0.10576880723237991,
      "num_input_tokens_seen": 917832,
      "step": 876,
      "train_runtime": 2474.3223,
      "train_tokens_per_second": 370.943
    },
    {
      "epoch": 0.3748263703387114,
      "grad_norm": 2.1270804405212402,
      "learning_rate": 1.2539614561027837e-05,
      "loss": 0.1098417341709137,
      "num_input_tokens_seen": 919210,
      "step": 877,
      "train_runtime": 2476.5308,
      "train_tokens_per_second": 371.168
    },
    {
      "epoch": 0.3752537664280372,
      "grad_norm": 1.9743280410766602,
      "learning_rate": 1.2531049250535332e-05,
      "loss": 0.09473894536495209,
      "num_input_tokens_seen": 920192,
      "step": 878,
      "train_runtime": 2478.7013,
      "train_tokens_per_second": 371.24
    },
    {
      "epoch": 0.37568116251736294,
      "grad_norm": 1.8980886936187744,
      "learning_rate": 1.2522483940042827e-05,
      "loss": 0.08141298592090607,
      "num_input_tokens_seen": 920826,
      "step": 879,
      "train_runtime": 2480.8008,
      "train_tokens_per_second": 371.181
    },
    {
      "epoch": 0.37610855860668874,
      "grad_norm": 2.2440285682678223,
      "learning_rate": 1.2513918629550321e-05,
      "loss": 0.13725431263446808,
      "num_input_tokens_seen": 922070,
      "step": 880,
      "train_runtime": 2483.0133,
      "train_tokens_per_second": 371.351
    },
    {
      "epoch": 0.37653595469601453,
      "grad_norm": 2.90678071975708,
      "learning_rate": 1.2505353319057816e-05,
      "loss": 0.24618873000144958,
      "num_input_tokens_seen": 923318,
      "step": 881,
      "train_runtime": 2485.1903,
      "train_tokens_per_second": 371.528
    },
    {
      "epoch": 0.3769633507853403,
      "grad_norm": 2.637791633605957,
      "learning_rate": 1.2496788008565313e-05,
      "loss": 0.12510210275650024,
      "num_input_tokens_seen": 924142,
      "step": 882,
      "train_runtime": 2487.3057,
      "train_tokens_per_second": 371.543
    },
    {
      "epoch": 0.3773907468746661,
      "grad_norm": 2.2982451915740967,
      "learning_rate": 1.2488222698072807e-05,
      "loss": 0.14668899774551392,
      "num_input_tokens_seen": 925586,
      "step": 883,
      "train_runtime": 2489.6908,
      "train_tokens_per_second": 371.767
    },
    {
      "epoch": 0.37781814296399185,
      "grad_norm": 1.9811372756958008,
      "learning_rate": 1.2479657387580302e-05,
      "loss": 0.10304494947195053,
      "num_input_tokens_seen": 926742,
      "step": 884,
      "train_runtime": 2491.8893,
      "train_tokens_per_second": 371.903
    },
    {
      "epoch": 0.37824553905331765,
      "grad_norm": 1.759446144104004,
      "learning_rate": 1.2471092077087796e-05,
      "loss": 0.15061134099960327,
      "num_input_tokens_seen": 928106,
      "step": 885,
      "train_runtime": 2494.1395,
      "train_tokens_per_second": 372.115
    },
    {
      "epoch": 0.37867293514264344,
      "grad_norm": 2.814922571182251,
      "learning_rate": 1.2462526766595291e-05,
      "loss": 0.14834536612033844,
      "num_input_tokens_seen": 929166,
      "step": 886,
      "train_runtime": 2496.3253,
      "train_tokens_per_second": 372.214
    },
    {
      "epoch": 0.37910033123196923,
      "grad_norm": 1.7875823974609375,
      "learning_rate": 1.2453961456102786e-05,
      "loss": 0.0991504043340683,
      "num_input_tokens_seen": 930410,
      "step": 887,
      "train_runtime": 2498.5164,
      "train_tokens_per_second": 372.385
    },
    {
      "epoch": 0.379527727321295,
      "grad_norm": 1.9592251777648926,
      "learning_rate": 1.244539614561028e-05,
      "loss": 0.08987846225500107,
      "num_input_tokens_seen": 931220,
      "step": 888,
      "train_runtime": 2500.6411,
      "train_tokens_per_second": 372.393
    },
    {
      "epoch": 0.37995512341062077,
      "grad_norm": 1.9382516145706177,
      "learning_rate": 1.2436830835117775e-05,
      "loss": 0.0876319631934166,
      "num_input_tokens_seen": 932864,
      "step": 889,
      "train_runtime": 2502.9365,
      "train_tokens_per_second": 372.708
    },
    {
      "epoch": 0.38038251949994656,
      "grad_norm": 2.277622938156128,
      "learning_rate": 1.2428265524625269e-05,
      "loss": 0.15632492303848267,
      "num_input_tokens_seen": 934192,
      "step": 890,
      "train_runtime": 2505.1695,
      "train_tokens_per_second": 372.906
    },
    {
      "epoch": 0.38080991558927235,
      "grad_norm": 2.850247859954834,
      "learning_rate": 1.2419700214132764e-05,
      "loss": 0.18625666201114655,
      "num_input_tokens_seen": 935036,
      "step": 891,
      "train_runtime": 2507.2874,
      "train_tokens_per_second": 372.927
    },
    {
      "epoch": 0.38123731167859815,
      "grad_norm": 1.7833002805709839,
      "learning_rate": 1.241113490364026e-05,
      "loss": 0.14385469257831573,
      "num_input_tokens_seen": 936516,
      "step": 892,
      "train_runtime": 2509.5355,
      "train_tokens_per_second": 373.183
    },
    {
      "epoch": 0.38166470776792394,
      "grad_norm": 1.7687997817993164,
      "learning_rate": 1.2402569593147753e-05,
      "loss": 0.09808165580034256,
      "num_input_tokens_seen": 937898,
      "step": 893,
      "train_runtime": 2511.7499,
      "train_tokens_per_second": 373.404
    },
    {
      "epoch": 0.38209210385724973,
      "grad_norm": 2.0094871520996094,
      "learning_rate": 1.2394004282655248e-05,
      "loss": 0.0910576656460762,
      "num_input_tokens_seen": 938748,
      "step": 894,
      "train_runtime": 2513.8685,
      "train_tokens_per_second": 373.428
    },
    {
      "epoch": 0.38251949994657547,
      "grad_norm": 2.2136740684509277,
      "learning_rate": 1.2385438972162742e-05,
      "loss": 0.12504349648952484,
      "num_input_tokens_seen": 939548,
      "step": 895,
      "train_runtime": 2515.9843,
      "train_tokens_per_second": 373.432
    },
    {
      "epoch": 0.38294689603590126,
      "grad_norm": 2.3028340339660645,
      "learning_rate": 1.2376873661670237e-05,
      "loss": 0.11133719980716705,
      "num_input_tokens_seen": 940356,
      "step": 896,
      "train_runtime": 2518.0746,
      "train_tokens_per_second": 373.442
    },
    {
      "epoch": 0.38337429212522706,
      "grad_norm": 2.0204732418060303,
      "learning_rate": 1.236830835117773e-05,
      "loss": 0.11281103640794754,
      "num_input_tokens_seen": 941316,
      "step": 897,
      "train_runtime": 2520.2386,
      "train_tokens_per_second": 373.503
    },
    {
      "epoch": 0.38380168821455285,
      "grad_norm": 2.63671875,
      "learning_rate": 1.2359743040685226e-05,
      "loss": 0.18026074767112732,
      "num_input_tokens_seen": 942222,
      "step": 898,
      "train_runtime": 2522.4056,
      "train_tokens_per_second": 373.541
    },
    {
      "epoch": 0.38422908430387864,
      "grad_norm": 2.357625961303711,
      "learning_rate": 1.2351177730192721e-05,
      "loss": 0.15424025058746338,
      "num_input_tokens_seen": 943360,
      "step": 899,
      "train_runtime": 2524.5818,
      "train_tokens_per_second": 373.67
    },
    {
      "epoch": 0.3846564803932044,
      "grad_norm": 2.877103567123413,
      "learning_rate": 1.2342612419700215e-05,
      "loss": 0.15378624200820923,
      "num_input_tokens_seen": 944406,
      "step": 900,
      "train_runtime": 2526.735,
      "train_tokens_per_second": 373.765
    },
    {
      "epoch": 0.3850838764825302,
      "grad_norm": 1.8694554567337036,
      "learning_rate": 1.233404710920771e-05,
      "loss": 0.11242681741714478,
      "num_input_tokens_seen": 945882,
      "step": 901,
      "train_runtime": 2542.0712,
      "train_tokens_per_second": 372.091
    },
    {
      "epoch": 0.38551127257185597,
      "grad_norm": 3.171609878540039,
      "learning_rate": 1.2325481798715204e-05,
      "loss": 0.1754235029220581,
      "num_input_tokens_seen": 946730,
      "step": 902,
      "train_runtime": 2544.176,
      "train_tokens_per_second": 372.117
    },
    {
      "epoch": 0.38593866866118176,
      "grad_norm": 3.146733045578003,
      "learning_rate": 1.2316916488222699e-05,
      "loss": 0.18093152344226837,
      "num_input_tokens_seen": 947530,
      "step": 903,
      "train_runtime": 2546.2789,
      "train_tokens_per_second": 372.123
    },
    {
      "epoch": 0.38636606475050755,
      "grad_norm": 1.64303457736969,
      "learning_rate": 1.2308351177730194e-05,
      "loss": 0.07227301597595215,
      "num_input_tokens_seen": 948602,
      "step": 904,
      "train_runtime": 2548.464,
      "train_tokens_per_second": 372.225
    },
    {
      "epoch": 0.3867934608398333,
      "grad_norm": 2.133760690689087,
      "learning_rate": 1.2299785867237688e-05,
      "loss": 0.1293281763792038,
      "num_input_tokens_seen": 949498,
      "step": 905,
      "train_runtime": 2550.6327,
      "train_tokens_per_second": 372.26
    },
    {
      "epoch": 0.3872208569291591,
      "grad_norm": 2.5086610317230225,
      "learning_rate": 1.2291220556745183e-05,
      "loss": 0.1361384093761444,
      "num_input_tokens_seen": 950432,
      "step": 906,
      "train_runtime": 2552.7712,
      "train_tokens_per_second": 372.314
    },
    {
      "epoch": 0.3876482530184849,
      "grad_norm": 2.183943033218384,
      "learning_rate": 1.2282655246252677e-05,
      "loss": 0.10865439474582672,
      "num_input_tokens_seen": 951456,
      "step": 907,
      "train_runtime": 2554.8958,
      "train_tokens_per_second": 372.405
    },
    {
      "epoch": 0.38807564910781067,
      "grad_norm": 2.268624782562256,
      "learning_rate": 1.2274089935760172e-05,
      "loss": 0.0960884541273117,
      "num_input_tokens_seen": 952172,
      "step": 908,
      "train_runtime": 2556.9783,
      "train_tokens_per_second": 372.382
    },
    {
      "epoch": 0.38850304519713647,
      "grad_norm": 1.9919828176498413,
      "learning_rate": 1.2265524625267667e-05,
      "loss": 0.09584489464759827,
      "num_input_tokens_seen": 953040,
      "step": 909,
      "train_runtime": 2559.0543,
      "train_tokens_per_second": 372.419
    },
    {
      "epoch": 0.3889304412864622,
      "grad_norm": 1.9157767295837402,
      "learning_rate": 1.225695931477516e-05,
      "loss": 0.1075853705406189,
      "num_input_tokens_seen": 954068,
      "step": 910,
      "train_runtime": 2561.1658,
      "train_tokens_per_second": 372.513
    },
    {
      "epoch": 0.389357837375788,
      "grad_norm": 2.539428949356079,
      "learning_rate": 1.2248394004282656e-05,
      "loss": 0.19488340616226196,
      "num_input_tokens_seen": 954938,
      "step": 911,
      "train_runtime": 2563.2758,
      "train_tokens_per_second": 372.546
    },
    {
      "epoch": 0.3897852334651138,
      "grad_norm": 3.7040762901306152,
      "learning_rate": 1.223982869379015e-05,
      "loss": 0.12698619067668915,
      "num_input_tokens_seen": 955768,
      "step": 912,
      "train_runtime": 2565.574,
      "train_tokens_per_second": 372.536
    },
    {
      "epoch": 0.3902126295544396,
      "grad_norm": 2.5857393741607666,
      "learning_rate": 1.2231263383297645e-05,
      "loss": 0.18038934469223022,
      "num_input_tokens_seen": 956702,
      "step": 913,
      "train_runtime": 2567.7784,
      "train_tokens_per_second": 372.58
    },
    {
      "epoch": 0.3906400256437654,
      "grad_norm": 2.8253538608551025,
      "learning_rate": 1.2222698072805138e-05,
      "loss": 0.1287080943584442,
      "num_input_tokens_seen": 957424,
      "step": 914,
      "train_runtime": 2569.8643,
      "train_tokens_per_second": 372.558
    },
    {
      "epoch": 0.39106742173309117,
      "grad_norm": 2.561044692993164,
      "learning_rate": 1.2214132762312634e-05,
      "loss": 0.1095781922340393,
      "num_input_tokens_seen": 958388,
      "step": 915,
      "train_runtime": 2572.0406,
      "train_tokens_per_second": 372.618
    },
    {
      "epoch": 0.3914948178224169,
      "grad_norm": 2.4931867122650146,
      "learning_rate": 1.220556745182013e-05,
      "loss": 0.2136867791414261,
      "num_input_tokens_seen": 959302,
      "step": 916,
      "train_runtime": 2574.1744,
      "train_tokens_per_second": 372.664
    },
    {
      "epoch": 0.3919222139117427,
      "grad_norm": 2.552759885787964,
      "learning_rate": 1.2197002141327626e-05,
      "loss": 0.09269089996814728,
      "num_input_tokens_seen": 960216,
      "step": 917,
      "train_runtime": 2576.2938,
      "train_tokens_per_second": 372.712
    },
    {
      "epoch": 0.3923496100010685,
      "grad_norm": 2.0703036785125732,
      "learning_rate": 1.218843683083512e-05,
      "loss": 0.13947340846061707,
      "num_input_tokens_seen": 961178,
      "step": 918,
      "train_runtime": 2578.4371,
      "train_tokens_per_second": 372.775
    },
    {
      "epoch": 0.3927770060903943,
      "grad_norm": 2.9783589839935303,
      "learning_rate": 1.2179871520342615e-05,
      "loss": 0.16986806690692902,
      "num_input_tokens_seen": 961906,
      "step": 919,
      "train_runtime": 2580.5298,
      "train_tokens_per_second": 372.755
    },
    {
      "epoch": 0.3932044021797201,
      "grad_norm": 2.4051973819732666,
      "learning_rate": 1.2171306209850108e-05,
      "loss": 0.20057371258735657,
      "num_input_tokens_seen": 962894,
      "step": 920,
      "train_runtime": 2582.7168,
      "train_tokens_per_second": 372.822
    },
    {
      "epoch": 0.3936317982690458,
      "grad_norm": 1.5075480937957764,
      "learning_rate": 1.2162740899357604e-05,
      "loss": 0.0642538070678711,
      "num_input_tokens_seen": 963752,
      "step": 921,
      "train_runtime": 2584.8065,
      "train_tokens_per_second": 372.853
    },
    {
      "epoch": 0.3940591943583716,
      "grad_norm": 1.8970712423324585,
      "learning_rate": 1.2154175588865099e-05,
      "loss": 0.09261252731084824,
      "num_input_tokens_seen": 964612,
      "step": 922,
      "train_runtime": 2586.9306,
      "train_tokens_per_second": 372.879
    },
    {
      "epoch": 0.3944865904476974,
      "grad_norm": 2.6838278770446777,
      "learning_rate": 1.2145610278372592e-05,
      "loss": 0.15132789313793182,
      "num_input_tokens_seen": 965652,
      "step": 923,
      "train_runtime": 2589.088,
      "train_tokens_per_second": 372.97
    },
    {
      "epoch": 0.3949139865370232,
      "grad_norm": 2.776010036468506,
      "learning_rate": 1.2137044967880088e-05,
      "loss": 0.14002200961112976,
      "num_input_tokens_seen": 966668,
      "step": 924,
      "train_runtime": 2591.2393,
      "train_tokens_per_second": 373.052
    },
    {
      "epoch": 0.395341382626349,
      "grad_norm": 2.3199238777160645,
      "learning_rate": 1.2128479657387581e-05,
      "loss": 0.10614319145679474,
      "num_input_tokens_seen": 967584,
      "step": 925,
      "train_runtime": 2593.3688,
      "train_tokens_per_second": 373.099
    },
    {
      "epoch": 0.39576877871567473,
      "grad_norm": 1.9682304859161377,
      "learning_rate": 1.2119914346895077e-05,
      "loss": 0.16021975874900818,
      "num_input_tokens_seen": 968924,
      "step": 926,
      "train_runtime": 2595.5801,
      "train_tokens_per_second": 373.298
    },
    {
      "epoch": 0.3961961748050005,
      "grad_norm": 2.1558029651641846,
      "learning_rate": 1.211134903640257e-05,
      "loss": 0.12402009218931198,
      "num_input_tokens_seen": 970256,
      "step": 927,
      "train_runtime": 2597.7977,
      "train_tokens_per_second": 373.492
    },
    {
      "epoch": 0.3966235708943263,
      "grad_norm": 1.6738981008529663,
      "learning_rate": 1.2102783725910065e-05,
      "loss": 0.10111566632986069,
      "num_input_tokens_seen": 971508,
      "step": 928,
      "train_runtime": 2600.0463,
      "train_tokens_per_second": 373.65
    },
    {
      "epoch": 0.3970509669836521,
      "grad_norm": 2.4890623092651367,
      "learning_rate": 1.209421841541756e-05,
      "loss": 0.12960676848888397,
      "num_input_tokens_seen": 972496,
      "step": 929,
      "train_runtime": 2602.1651,
      "train_tokens_per_second": 373.726
    },
    {
      "epoch": 0.3974783630729779,
      "grad_norm": 2.40090012550354,
      "learning_rate": 1.2085653104925054e-05,
      "loss": 0.13657209277153015,
      "num_input_tokens_seen": 973502,
      "step": 930,
      "train_runtime": 2604.3414,
      "train_tokens_per_second": 373.8
    },
    {
      "epoch": 0.39790575916230364,
      "grad_norm": 1.800201654434204,
      "learning_rate": 1.207708779443255e-05,
      "loss": 0.08778995275497437,
      "num_input_tokens_seen": 974398,
      "step": 931,
      "train_runtime": 2620.4146,
      "train_tokens_per_second": 371.849
    },
    {
      "epoch": 0.39833315525162943,
      "grad_norm": 2.485584259033203,
      "learning_rate": 1.2068522483940043e-05,
      "loss": 0.1305716186761856,
      "num_input_tokens_seen": 975244,
      "step": 932,
      "train_runtime": 2622.5156,
      "train_tokens_per_second": 371.873
    },
    {
      "epoch": 0.3987605513409552,
      "grad_norm": 3.030352830886841,
      "learning_rate": 1.2059957173447538e-05,
      "loss": 0.1337234377861023,
      "num_input_tokens_seen": 976058,
      "step": 933,
      "train_runtime": 2624.6131,
      "train_tokens_per_second": 371.886
    },
    {
      "epoch": 0.399187947430281,
      "grad_norm": 2.5703330039978027,
      "learning_rate": 1.2051391862955034e-05,
      "loss": 0.1824536919593811,
      "num_input_tokens_seen": 976876,
      "step": 934,
      "train_runtime": 2626.7023,
      "train_tokens_per_second": 371.902
    },
    {
      "epoch": 0.3996153435196068,
      "grad_norm": 3.760235071182251,
      "learning_rate": 1.2042826552462527e-05,
      "loss": 0.23526248335838318,
      "num_input_tokens_seen": 977490,
      "step": 935,
      "train_runtime": 2628.776,
      "train_tokens_per_second": 371.842
    },
    {
      "epoch": 0.40004273960893255,
      "grad_norm": 3.0365898609161377,
      "learning_rate": 1.2034261241970023e-05,
      "loss": 0.0925004705786705,
      "num_input_tokens_seen": 978154,
      "step": 936,
      "train_runtime": 2630.8693,
      "train_tokens_per_second": 371.799
    },
    {
      "epoch": 0.40004273960893255,
      "eval_loss": 0.49531039595603943,
      "eval_runtime": 58.3154,
      "eval_samples_per_second": 17.114,
      "eval_steps_per_second": 8.557,
      "num_input_tokens_seen": 978154,
      "step": 936
    },
    {
      "epoch": 0.40047013569825834,
      "grad_norm": 2.3873255252838135,
      "learning_rate": 1.2025695931477516e-05,
      "loss": 0.0911688432097435,
      "num_input_tokens_seen": 978928,
      "step": 937,
      "train_runtime": 2691.33,
      "train_tokens_per_second": 363.734
    },
    {
      "epoch": 0.40089753178758414,
      "grad_norm": 2.060574531555176,
      "learning_rate": 1.2017130620985011e-05,
      "loss": 0.12122222781181335,
      "num_input_tokens_seen": 979884,
      "step": 938,
      "train_runtime": 2693.4475,
      "train_tokens_per_second": 363.803
    },
    {
      "epoch": 0.40132492787690993,
      "grad_norm": 2.926950454711914,
      "learning_rate": 1.2008565310492507e-05,
      "loss": 0.13500195741653442,
      "num_input_tokens_seen": 980664,
      "step": 939,
      "train_runtime": 2695.6273,
      "train_tokens_per_second": 363.798
    },
    {
      "epoch": 0.4017523239662357,
      "grad_norm": 3.1084179878234863,
      "learning_rate": 1.2e-05,
      "loss": 0.1328144073486328,
      "num_input_tokens_seen": 981276,
      "step": 940,
      "train_runtime": 2697.7463,
      "train_tokens_per_second": 363.739
    },
    {
      "epoch": 0.4021797200555615,
      "grad_norm": 2.009573221206665,
      "learning_rate": 1.1991434689507496e-05,
      "loss": 0.15397098660469055,
      "num_input_tokens_seen": 982846,
      "step": 941,
      "train_runtime": 2700.0404,
      "train_tokens_per_second": 364.012
    },
    {
      "epoch": 0.40260711614488726,
      "grad_norm": 2.4772355556488037,
      "learning_rate": 1.198286937901499e-05,
      "loss": 0.0811704769730568,
      "num_input_tokens_seen": 984040,
      "step": 942,
      "train_runtime": 2702.2218,
      "train_tokens_per_second": 364.16
    },
    {
      "epoch": 0.40303451223421305,
      "grad_norm": 2.2134244441986084,
      "learning_rate": 1.1974304068522484e-05,
      "loss": 0.10884946584701538,
      "num_input_tokens_seen": 984862,
      "step": 943,
      "train_runtime": 2704.3272,
      "train_tokens_per_second": 364.18
    },
    {
      "epoch": 0.40346190832353884,
      "grad_norm": 1.9158374071121216,
      "learning_rate": 1.1965738758029978e-05,
      "loss": 0.08385251462459564,
      "num_input_tokens_seen": 985594,
      "step": 944,
      "train_runtime": 2706.406,
      "train_tokens_per_second": 364.171
    },
    {
      "epoch": 0.40388930441286464,
      "grad_norm": 2.6209640502929688,
      "learning_rate": 1.1957173447537473e-05,
      "loss": 0.1714126169681549,
      "num_input_tokens_seen": 986356,
      "step": 945,
      "train_runtime": 2708.4962,
      "train_tokens_per_second": 364.171
    },
    {
      "epoch": 0.40431670050219043,
      "grad_norm": 2.614886999130249,
      "learning_rate": 1.1948608137044969e-05,
      "loss": 0.18205171823501587,
      "num_input_tokens_seen": 987016,
      "step": 946,
      "train_runtime": 2710.5441,
      "train_tokens_per_second": 364.139
    },
    {
      "epoch": 0.40474409659151617,
      "grad_norm": 2.492692470550537,
      "learning_rate": 1.1940042826552462e-05,
      "loss": 0.12335605174303055,
      "num_input_tokens_seen": 987838,
      "step": 947,
      "train_runtime": 2712.6428,
      "train_tokens_per_second": 364.161
    },
    {
      "epoch": 0.40517149268084196,
      "grad_norm": 2.9411168098449707,
      "learning_rate": 1.1931477516059957e-05,
      "loss": 0.12869834899902344,
      "num_input_tokens_seen": 988560,
      "step": 948,
      "train_runtime": 2714.6797,
      "train_tokens_per_second": 364.153
    },
    {
      "epoch": 0.40559888877016775,
      "grad_norm": 3.8990182876586914,
      "learning_rate": 1.1922912205567451e-05,
      "loss": 0.1914498656988144,
      "num_input_tokens_seen": 989722,
      "step": 949,
      "train_runtime": 2716.8473,
      "train_tokens_per_second": 364.291
    },
    {
      "epoch": 0.40602628485949355,
      "grad_norm": 1.982263207435608,
      "learning_rate": 1.1914346895074946e-05,
      "loss": 0.11047862470149994,
      "num_input_tokens_seen": 990588,
      "step": 950,
      "train_runtime": 2718.9843,
      "train_tokens_per_second": 364.323
    },
    {
      "epoch": 0.40645368094881934,
      "grad_norm": 1.8765393495559692,
      "learning_rate": 1.1905781584582443e-05,
      "loss": 0.13714712858200073,
      "num_input_tokens_seen": 991610,
      "step": 951,
      "train_runtime": 2721.1467,
      "train_tokens_per_second": 364.409
    },
    {
      "epoch": 0.4068810770381451,
      "grad_norm": 2.322723627090454,
      "learning_rate": 1.1897216274089939e-05,
      "loss": 0.1259048879146576,
      "num_input_tokens_seen": 993094,
      "step": 952,
      "train_runtime": 2723.4379,
      "train_tokens_per_second": 364.647
    },
    {
      "epoch": 0.40730847312747087,
      "grad_norm": 1.818282961845398,
      "learning_rate": 1.1888650963597432e-05,
      "loss": 0.13882559537887573,
      "num_input_tokens_seen": 994546,
      "step": 953,
      "train_runtime": 2725.7106,
      "train_tokens_per_second": 364.876
    },
    {
      "epoch": 0.40773586921679666,
      "grad_norm": 2.0054283142089844,
      "learning_rate": 1.1880085653104927e-05,
      "loss": 0.10679404437541962,
      "num_input_tokens_seen": 996130,
      "step": 954,
      "train_runtime": 2728.0013,
      "train_tokens_per_second": 365.15
    },
    {
      "epoch": 0.40816326530612246,
      "grad_norm": 1.6338130235671997,
      "learning_rate": 1.1871520342612421e-05,
      "loss": 0.07151520997285843,
      "num_input_tokens_seen": 996898,
      "step": 955,
      "train_runtime": 2730.0925,
      "train_tokens_per_second": 365.152
    },
    {
      "epoch": 0.40859066139544825,
      "grad_norm": 1.7243101596832275,
      "learning_rate": 1.1862955032119916e-05,
      "loss": 0.06756711006164551,
      "num_input_tokens_seen": 997846,
      "step": 956,
      "train_runtime": 2732.3103,
      "train_tokens_per_second": 365.202
    },
    {
      "epoch": 0.409018057484774,
      "grad_norm": 3.2189505100250244,
      "learning_rate": 1.185438972162741e-05,
      "loss": 0.09284111112356186,
      "num_input_tokens_seen": 998512,
      "step": 957,
      "train_runtime": 2734.3925,
      "train_tokens_per_second": 365.168
    },
    {
      "epoch": 0.4094454535740998,
      "grad_norm": 4.5584259033203125,
      "learning_rate": 1.1845824411134905e-05,
      "loss": 0.21210573613643646,
      "num_input_tokens_seen": 999580,
      "step": 958,
      "train_runtime": 2736.5454,
      "train_tokens_per_second": 365.271
    },
    {
      "epoch": 0.4098728496634256,
      "grad_norm": 1.6021218299865723,
      "learning_rate": 1.18372591006424e-05,
      "loss": 0.08847945928573608,
      "num_input_tokens_seen": 1000800,
      "step": 959,
      "train_runtime": 2738.756,
      "train_tokens_per_second": 365.421
    },
    {
      "epoch": 0.41030024575275137,
      "grad_norm": 1.860171914100647,
      "learning_rate": 1.1828693790149894e-05,
      "loss": 0.11484178900718689,
      "num_input_tokens_seen": 1002042,
      "step": 960,
      "train_runtime": 2740.9783,
      "train_tokens_per_second": 365.578
    },
    {
      "epoch": 0.41072764184207716,
      "grad_norm": 1.976267695426941,
      "learning_rate": 1.182012847965739e-05,
      "loss": 0.11644624173641205,
      "num_input_tokens_seen": 1003804,
      "step": 961,
      "train_runtime": 2757.2531,
      "train_tokens_per_second": 364.059
    },
    {
      "epoch": 0.4111550379314029,
      "grad_norm": 2.1563267707824707,
      "learning_rate": 1.1811563169164883e-05,
      "loss": 0.11923924088478088,
      "num_input_tokens_seen": 1005234,
      "step": 962,
      "train_runtime": 2759.5471,
      "train_tokens_per_second": 364.275
    },
    {
      "epoch": 0.4115824340207287,
      "grad_norm": 2.7617363929748535,
      "learning_rate": 1.1802997858672378e-05,
      "loss": 0.1798803210258484,
      "num_input_tokens_seen": 1006284,
      "step": 963,
      "train_runtime": 2761.7431,
      "train_tokens_per_second": 364.366
    },
    {
      "epoch": 0.4120098301100545,
      "grad_norm": 4.04373836517334,
      "learning_rate": 1.1794432548179873e-05,
      "loss": 0.11118249595165253,
      "num_input_tokens_seen": 1006842,
      "step": 964,
      "train_runtime": 2763.7807,
      "train_tokens_per_second": 364.299
    },
    {
      "epoch": 0.4124372261993803,
      "grad_norm": 2.6414828300476074,
      "learning_rate": 1.1785867237687367e-05,
      "loss": 0.1510341465473175,
      "num_input_tokens_seen": 1007780,
      "step": 965,
      "train_runtime": 2765.943,
      "train_tokens_per_second": 364.353
    },
    {
      "epoch": 0.4128646222887061,
      "grad_norm": 2.202540636062622,
      "learning_rate": 1.1777301927194862e-05,
      "loss": 0.10943156480789185,
      "num_input_tokens_seen": 1008514,
      "step": 966,
      "train_runtime": 2768.0416,
      "train_tokens_per_second": 364.342
    },
    {
      "epoch": 0.41329201837803187,
      "grad_norm": 2.0062096118927,
      "learning_rate": 1.1768736616702356e-05,
      "loss": 0.0947541743516922,
      "num_input_tokens_seen": 1009364,
      "step": 967,
      "train_runtime": 2770.1643,
      "train_tokens_per_second": 364.37
    },
    {
      "epoch": 0.4137194144673576,
      "grad_norm": 2.721343755722046,
      "learning_rate": 1.1760171306209851e-05,
      "loss": 0.12457822263240814,
      "num_input_tokens_seen": 1010180,
      "step": 968,
      "train_runtime": 2772.2171,
      "train_tokens_per_second": 364.394
    },
    {
      "epoch": 0.4141468105566834,
      "grad_norm": 2.299980640411377,
      "learning_rate": 1.1751605995717346e-05,
      "loss": 0.134385883808136,
      "num_input_tokens_seen": 1010894,
      "step": 969,
      "train_runtime": 2774.3033,
      "train_tokens_per_second": 364.378
    },
    {
      "epoch": 0.4145742066460092,
      "grad_norm": 2.4229962825775146,
      "learning_rate": 1.174304068522484e-05,
      "loss": 0.15707530081272125,
      "num_input_tokens_seen": 1011828,
      "step": 970,
      "train_runtime": 2776.4185,
      "train_tokens_per_second": 364.436
    },
    {
      "epoch": 0.415001602735335,
      "grad_norm": 2.3712334632873535,
      "learning_rate": 1.1734475374732335e-05,
      "loss": 0.10923910140991211,
      "num_input_tokens_seen": 1012502,
      "step": 971,
      "train_runtime": 2778.5108,
      "train_tokens_per_second": 364.405
    },
    {
      "epoch": 0.4154289988246608,
      "grad_norm": 4.302492618560791,
      "learning_rate": 1.1725910064239829e-05,
      "loss": 0.23842209577560425,
      "num_input_tokens_seen": 1012990,
      "step": 972,
      "train_runtime": 2780.5107,
      "train_tokens_per_second": 364.318
    },
    {
      "epoch": 0.4158563949139865,
      "grad_norm": 2.0874907970428467,
      "learning_rate": 1.1717344753747324e-05,
      "loss": 0.1560826450586319,
      "num_input_tokens_seen": 1014720,
      "step": 973,
      "train_runtime": 2782.8695,
      "train_tokens_per_second": 364.631
    },
    {
      "epoch": 0.4162837910033123,
      "grad_norm": 2.134237051010132,
      "learning_rate": 1.1708779443254818e-05,
      "loss": 0.15274594724178314,
      "num_input_tokens_seen": 1016746,
      "step": 974,
      "train_runtime": 2785.3131,
      "train_tokens_per_second": 365.038
    },
    {
      "epoch": 0.4167111870926381,
      "grad_norm": 2.5279417037963867,
      "learning_rate": 1.1700214132762313e-05,
      "loss": 0.10355942696332932,
      "num_input_tokens_seen": 1017582,
      "step": 975,
      "train_runtime": 2787.4275,
      "train_tokens_per_second": 365.061
    },
    {
      "epoch": 0.4171385831819639,
      "grad_norm": 1.848024606704712,
      "learning_rate": 1.1691648822269808e-05,
      "loss": 0.10375792533159256,
      "num_input_tokens_seen": 1018630,
      "step": 976,
      "train_runtime": 2789.6047,
      "train_tokens_per_second": 365.152
    },
    {
      "epoch": 0.4175659792712897,
      "grad_norm": 2.251873254776001,
      "learning_rate": 1.1683083511777302e-05,
      "loss": 0.09312009066343307,
      "num_input_tokens_seen": 1019318,
      "step": 977,
      "train_runtime": 2791.7042,
      "train_tokens_per_second": 365.124
    },
    {
      "epoch": 0.4179933753606154,
      "grad_norm": 2.276503801345825,
      "learning_rate": 1.1674518201284797e-05,
      "loss": 0.1360630840063095,
      "num_input_tokens_seen": 1020494,
      "step": 978,
      "train_runtime": 2793.9274,
      "train_tokens_per_second": 365.254
    },
    {
      "epoch": 0.4184207714499412,
      "grad_norm": 2.2060673236846924,
      "learning_rate": 1.166595289079229e-05,
      "loss": 0.13519413769245148,
      "num_input_tokens_seen": 1021976,
      "step": 979,
      "train_runtime": 2796.2334,
      "train_tokens_per_second": 365.483
    },
    {
      "epoch": 0.418848167539267,
      "grad_norm": 2.7974560260772705,
      "learning_rate": 1.1657387580299786e-05,
      "loss": 0.14611303806304932,
      "num_input_tokens_seen": 1022736,
      "step": 980,
      "train_runtime": 2798.3269,
      "train_tokens_per_second": 365.481
    },
    {
      "epoch": 0.4192755636285928,
      "grad_norm": 1.937928318977356,
      "learning_rate": 1.1648822269807281e-05,
      "loss": 0.11127322912216187,
      "num_input_tokens_seen": 1023696,
      "step": 981,
      "train_runtime": 2800.4491,
      "train_tokens_per_second": 365.547
    },
    {
      "epoch": 0.4197029597179186,
      "grad_norm": 2.6253833770751953,
      "learning_rate": 1.1640256959314775e-05,
      "loss": 0.16199615597724915,
      "num_input_tokens_seen": 1024560,
      "step": 982,
      "train_runtime": 2802.6032,
      "train_tokens_per_second": 365.574
    },
    {
      "epoch": 0.42013035580724434,
      "grad_norm": 1.643823266029358,
      "learning_rate": 1.163169164882227e-05,
      "loss": 0.08269336819648743,
      "num_input_tokens_seen": 1025730,
      "step": 983,
      "train_runtime": 2804.7773,
      "train_tokens_per_second": 365.708
    },
    {
      "epoch": 0.42055775189657013,
      "grad_norm": 3.6539435386657715,
      "learning_rate": 1.1623126338329764e-05,
      "loss": 0.17416368424892426,
      "num_input_tokens_seen": 1026550,
      "step": 984,
      "train_runtime": 2806.8549,
      "train_tokens_per_second": 365.73
    },
    {
      "epoch": 0.4209851479858959,
      "grad_norm": 2.2729201316833496,
      "learning_rate": 1.161456102783726e-05,
      "loss": 0.07793647050857544,
      "num_input_tokens_seen": 1027168,
      "step": 985,
      "train_runtime": 2808.8834,
      "train_tokens_per_second": 365.686
    },
    {
      "epoch": 0.4214125440752217,
      "grad_norm": 1.8531163930892944,
      "learning_rate": 1.1605995717344756e-05,
      "loss": 0.11568786203861237,
      "num_input_tokens_seen": 1028250,
      "step": 986,
      "train_runtime": 2811.1036,
      "train_tokens_per_second": 365.782
    },
    {
      "epoch": 0.4218399401645475,
      "grad_norm": 2.2515223026275635,
      "learning_rate": 1.159743040685225e-05,
      "loss": 0.08347593247890472,
      "num_input_tokens_seen": 1029064,
      "step": 987,
      "train_runtime": 2813.1983,
      "train_tokens_per_second": 365.799
    },
    {
      "epoch": 0.4222673362538733,
      "grad_norm": 2.4186718463897705,
      "learning_rate": 1.1588865096359745e-05,
      "loss": 0.14525270462036133,
      "num_input_tokens_seen": 1030018,
      "step": 988,
      "train_runtime": 2815.3527,
      "train_tokens_per_second": 365.858
    },
    {
      "epoch": 0.42269473234319904,
      "grad_norm": 2.28676700592041,
      "learning_rate": 1.158029978586724e-05,
      "loss": 0.11397407948970795,
      "num_input_tokens_seen": 1030886,
      "step": 989,
      "train_runtime": 2817.4938,
      "train_tokens_per_second": 365.888
    },
    {
      "epoch": 0.42312212843252484,
      "grad_norm": 2.185218572616577,
      "learning_rate": 1.1571734475374734e-05,
      "loss": 0.1133785992860794,
      "num_input_tokens_seen": 1032168,
      "step": 990,
      "train_runtime": 2819.6878,
      "train_tokens_per_second": 366.058
    },
    {
      "epoch": 0.42354952452185063,
      "grad_norm": 2.3733842372894287,
      "learning_rate": 1.1563169164882229e-05,
      "loss": 0.12109623849391937,
      "num_input_tokens_seen": 1032956,
      "step": 991,
      "train_runtime": 2835.4647,
      "train_tokens_per_second": 364.299
    },
    {
      "epoch": 0.4239769206111764,
      "grad_norm": 2.1615986824035645,
      "learning_rate": 1.1554603854389722e-05,
      "loss": 0.12668004631996155,
      "num_input_tokens_seen": 1033768,
      "step": 992,
      "train_runtime": 2837.5786,
      "train_tokens_per_second": 364.313
    },
    {
      "epoch": 0.4244043167005022,
      "grad_norm": 3.12859845161438,
      "learning_rate": 1.1546038543897218e-05,
      "loss": 0.12454401701688766,
      "num_input_tokens_seen": 1034492,
      "step": 993,
      "train_runtime": 2839.7374,
      "train_tokens_per_second": 364.291
    },
    {
      "epoch": 0.42483171278982795,
      "grad_norm": 1.8854997158050537,
      "learning_rate": 1.1537473233404713e-05,
      "loss": 0.09664224088191986,
      "num_input_tokens_seen": 1035644,
      "step": 994,
      "train_runtime": 2841.9714,
      "train_tokens_per_second": 364.41
    },
    {
      "epoch": 0.42525910887915375,
      "grad_norm": 1.9307321310043335,
      "learning_rate": 1.1528907922912207e-05,
      "loss": 0.103437140583992,
      "num_input_tokens_seen": 1037086,
      "step": 995,
      "train_runtime": 2844.2388,
      "train_tokens_per_second": 364.627
    },
    {
      "epoch": 0.42568650496847954,
      "grad_norm": 2.553837537765503,
      "learning_rate": 1.1520342612419702e-05,
      "loss": 0.06267597526311874,
      "num_input_tokens_seen": 1037730,
      "step": 996,
      "train_runtime": 2846.281,
      "train_tokens_per_second": 364.592
    },
    {
      "epoch": 0.42611390105780533,
      "grad_norm": 1.667264461517334,
      "learning_rate": 1.1511777301927195e-05,
      "loss": 0.0803246796131134,
      "num_input_tokens_seen": 1038528,
      "step": 997,
      "train_runtime": 2848.3779,
      "train_tokens_per_second": 364.603
    },
    {
      "epoch": 0.4265412971471311,
      "grad_norm": 1.9268029928207397,
      "learning_rate": 1.150321199143469e-05,
      "loss": 0.08455639332532883,
      "num_input_tokens_seen": 1039386,
      "step": 998,
      "train_runtime": 2850.5087,
      "train_tokens_per_second": 364.632
    },
    {
      "epoch": 0.42696869323645686,
      "grad_norm": 3.1712958812713623,
      "learning_rate": 1.1494646680942186e-05,
      "loss": 0.12818551063537598,
      "num_input_tokens_seen": 1040338,
      "step": 999,
      "train_runtime": 2852.6474,
      "train_tokens_per_second": 364.692
    },
    {
      "epoch": 0.42739608932578266,
      "grad_norm": 3.4448463916778564,
      "learning_rate": 1.148608137044968e-05,
      "loss": 0.20483914017677307,
      "num_input_tokens_seen": 1041176,
      "step": 1000,
      "train_runtime": 2854.7842,
      "train_tokens_per_second": 364.713
    },
    {
      "epoch": 0.42782348541510845,
      "grad_norm": 2.007951498031616,
      "learning_rate": 1.1477516059957175e-05,
      "loss": 0.11240419745445251,
      "num_input_tokens_seen": 1042066,
      "step": 1001,
      "train_runtime": 2856.9884,
      "train_tokens_per_second": 364.743
    },
    {
      "epoch": 0.42825088150443424,
      "grad_norm": 2.101724147796631,
      "learning_rate": 1.1468950749464668e-05,
      "loss": 0.19278055429458618,
      "num_input_tokens_seen": 1043372,
      "step": 1002,
      "train_runtime": 2859.2505,
      "train_tokens_per_second": 364.911
    },
    {
      "epoch": 0.42867827759376004,
      "grad_norm": 3.717116355895996,
      "learning_rate": 1.1460385438972164e-05,
      "loss": 0.16371259093284607,
      "num_input_tokens_seen": 1044332,
      "step": 1003,
      "train_runtime": 2861.3726,
      "train_tokens_per_second": 364.976
    },
    {
      "epoch": 0.4291056736830858,
      "grad_norm": 1.5768543481826782,
      "learning_rate": 1.1451820128479657e-05,
      "loss": 0.0704633817076683,
      "num_input_tokens_seen": 1045352,
      "step": 1004,
      "train_runtime": 2863.4887,
      "train_tokens_per_second": 365.062
    },
    {
      "epoch": 0.42953306977241157,
      "grad_norm": 2.1937477588653564,
      "learning_rate": 1.1443254817987153e-05,
      "loss": 0.16210992634296417,
      "num_input_tokens_seen": 1046092,
      "step": 1005,
      "train_runtime": 2865.5645,
      "train_tokens_per_second": 365.056
    },
    {
      "epoch": 0.42996046586173736,
      "grad_norm": 3.0632998943328857,
      "learning_rate": 1.1434689507494648e-05,
      "loss": 0.19956105947494507,
      "num_input_tokens_seen": 1046896,
      "step": 1006,
      "train_runtime": 2867.6866,
      "train_tokens_per_second": 365.066
    },
    {
      "epoch": 0.43038786195106316,
      "grad_norm": 3.1376841068267822,
      "learning_rate": 1.1426124197002141e-05,
      "loss": 0.16229747235774994,
      "num_input_tokens_seen": 1047566,
      "step": 1007,
      "train_runtime": 2869.7496,
      "train_tokens_per_second": 365.037
    },
    {
      "epoch": 0.43081525804038895,
      "grad_norm": 4.404557228088379,
      "learning_rate": 1.1417558886509637e-05,
      "loss": 0.11602835357189178,
      "num_input_tokens_seen": 1048180,
      "step": 1008,
      "train_runtime": 2871.8084,
      "train_tokens_per_second": 364.99
    },
    {
      "epoch": 0.4312426541297147,
      "grad_norm": 2.989595890045166,
      "learning_rate": 1.140899357601713e-05,
      "loss": 0.2091883420944214,
      "num_input_tokens_seen": 1049332,
      "step": 1009,
      "train_runtime": 2873.9932,
      "train_tokens_per_second": 365.113
    },
    {
      "epoch": 0.4316700502190405,
      "grad_norm": 3.529862642288208,
      "learning_rate": 1.1400428265524626e-05,
      "loss": 0.12193688750267029,
      "num_input_tokens_seen": 1049910,
      "step": 1010,
      "train_runtime": 2876.0147,
      "train_tokens_per_second": 365.057
    },
    {
      "epoch": 0.4320974463083663,
      "grad_norm": 2.3281264305114746,
      "learning_rate": 1.1391862955032121e-05,
      "loss": 0.14594826102256775,
      "num_input_tokens_seen": 1051190,
      "step": 1011,
      "train_runtime": 2878.2689,
      "train_tokens_per_second": 365.216
    },
    {
      "epoch": 0.43252484239769207,
      "grad_norm": 2.4456586837768555,
      "learning_rate": 1.1383297644539614e-05,
      "loss": 0.1984366774559021,
      "num_input_tokens_seen": 1052216,
      "step": 1012,
      "train_runtime": 2880.4547,
      "train_tokens_per_second": 365.295
    },
    {
      "epoch": 0.43295223848701786,
      "grad_norm": 2.46181321144104,
      "learning_rate": 1.137473233404711e-05,
      "loss": 0.13264615833759308,
      "num_input_tokens_seen": 1053108,
      "step": 1013,
      "train_runtime": 2882.5596,
      "train_tokens_per_second": 365.338
    },
    {
      "epoch": 0.43337963457634365,
      "grad_norm": 3.1830193996429443,
      "learning_rate": 1.1366167023554603e-05,
      "loss": 0.1790146380662918,
      "num_input_tokens_seen": 1053874,
      "step": 1014,
      "train_runtime": 2884.6708,
      "train_tokens_per_second": 365.336
    },
    {
      "epoch": 0.4338070306656694,
      "grad_norm": 1.321878433227539,
      "learning_rate": 1.1357601713062099e-05,
      "loss": 0.06271042674779892,
      "num_input_tokens_seen": 1055180,
      "step": 1015,
      "train_runtime": 2886.8794,
      "train_tokens_per_second": 365.509
    },
    {
      "epoch": 0.4342344267549952,
      "grad_norm": 2.5585038661956787,
      "learning_rate": 1.1349036402569594e-05,
      "loss": 0.09212590754032135,
      "num_input_tokens_seen": 1055904,
      "step": 1016,
      "train_runtime": 2888.9386,
      "train_tokens_per_second": 365.499
    },
    {
      "epoch": 0.434661822844321,
      "grad_norm": 2.3073716163635254,
      "learning_rate": 1.1340471092077087e-05,
      "loss": 0.14243914186954498,
      "num_input_tokens_seen": 1056978,
      "step": 1017,
      "train_runtime": 2891.1254,
      "train_tokens_per_second": 365.594
    },
    {
      "epoch": 0.43508921893364677,
      "grad_norm": 4.160744667053223,
      "learning_rate": 1.1331905781584583e-05,
      "loss": 0.17852286994457245,
      "num_input_tokens_seen": 1058052,
      "step": 1018,
      "train_runtime": 2893.2935,
      "train_tokens_per_second": 365.691
    },
    {
      "epoch": 0.43551661502297256,
      "grad_norm": 1.745357632637024,
      "learning_rate": 1.132334047109208e-05,
      "loss": 0.07179614156484604,
      "num_input_tokens_seen": 1058838,
      "step": 1019,
      "train_runtime": 2895.3654,
      "train_tokens_per_second": 365.701
    },
    {
      "epoch": 0.4359440111122983,
      "grad_norm": 2.5044443607330322,
      "learning_rate": 1.1314775160599573e-05,
      "loss": 0.08159780502319336,
      "num_input_tokens_seen": 1059646,
      "step": 1020,
      "train_runtime": 2897.4796,
      "train_tokens_per_second": 365.713
    },
    {
      "epoch": 0.4363714072016241,
      "grad_norm": 1.9943795204162598,
      "learning_rate": 1.1306209850107069e-05,
      "loss": 0.1532469242811203,
      "num_input_tokens_seen": 1060874,
      "step": 1021,
      "train_runtime": 2913.0868,
      "train_tokens_per_second": 364.175
    },
    {
      "epoch": 0.4367988032909499,
      "grad_norm": 4.59913444519043,
      "learning_rate": 1.1297644539614562e-05,
      "loss": 0.1994732767343521,
      "num_input_tokens_seen": 1061970,
      "step": 1022,
      "train_runtime": 2915.298,
      "train_tokens_per_second": 364.275
    },
    {
      "epoch": 0.4372261993802757,
      "grad_norm": 2.5817437171936035,
      "learning_rate": 1.1289079229122057e-05,
      "loss": 0.18294461071491241,
      "num_input_tokens_seen": 1063252,
      "step": 1023,
      "train_runtime": 2917.5125,
      "train_tokens_per_second": 364.438
    },
    {
      "epoch": 0.4376535954696015,
      "grad_norm": 2.9953930377960205,
      "learning_rate": 1.1280513918629553e-05,
      "loss": 0.15931539237499237,
      "num_input_tokens_seen": 1063910,
      "step": 1024,
      "train_runtime": 2919.5918,
      "train_tokens_per_second": 364.404
    },
    {
      "epoch": 0.4380809915589272,
      "grad_norm": 2.173919200897217,
      "learning_rate": 1.1271948608137046e-05,
      "loss": 0.09014363586902618,
      "num_input_tokens_seen": 1065248,
      "step": 1025,
      "train_runtime": 2921.8322,
      "train_tokens_per_second": 364.582
    },
    {
      "epoch": 0.438508387648253,
      "grad_norm": 2.3778066635131836,
      "learning_rate": 1.1263383297644542e-05,
      "loss": 0.10724397748708725,
      "num_input_tokens_seen": 1065962,
      "step": 1026,
      "train_runtime": 2923.9139,
      "train_tokens_per_second": 364.567
    },
    {
      "epoch": 0.4389357837375788,
      "grad_norm": 2.372450113296509,
      "learning_rate": 1.1254817987152035e-05,
      "loss": 0.15932433307170868,
      "num_input_tokens_seen": 1067372,
      "step": 1027,
      "train_runtime": 2926.1565,
      "train_tokens_per_second": 364.769
    },
    {
      "epoch": 0.4393631798269046,
      "grad_norm": 2.3470299243927,
      "learning_rate": 1.124625267665953e-05,
      "loss": 0.1394292265176773,
      "num_input_tokens_seen": 1068310,
      "step": 1028,
      "train_runtime": 2928.2728,
      "train_tokens_per_second": 364.826
    },
    {
      "epoch": 0.4397905759162304,
      "grad_norm": 1.7886719703674316,
      "learning_rate": 1.1237687366167026e-05,
      "loss": 0.1035194993019104,
      "num_input_tokens_seen": 1069292,
      "step": 1029,
      "train_runtime": 2930.4415,
      "train_tokens_per_second": 364.891
    },
    {
      "epoch": 0.4402179720055561,
      "grad_norm": 2.1555051803588867,
      "learning_rate": 1.122912205567452e-05,
      "loss": 0.0789809301495552,
      "num_input_tokens_seen": 1070098,
      "step": 1030,
      "train_runtime": 2932.5988,
      "train_tokens_per_second": 364.898
    },
    {
      "epoch": 0.4406453680948819,
      "grad_norm": 2.104384422302246,
      "learning_rate": 1.1220556745182015e-05,
      "loss": 0.12144552916288376,
      "num_input_tokens_seen": 1071068,
      "step": 1031,
      "train_runtime": 2934.8096,
      "train_tokens_per_second": 364.953
    },
    {
      "epoch": 0.4410727641842077,
      "grad_norm": 2.4359829425811768,
      "learning_rate": 1.1211991434689508e-05,
      "loss": 0.1290377676486969,
      "num_input_tokens_seen": 1072092,
      "step": 1032,
      "train_runtime": 2936.9753,
      "train_tokens_per_second": 365.033
    },
    {
      "epoch": 0.4415001602735335,
      "grad_norm": 2.328812837600708,
      "learning_rate": 1.1203426124197003e-05,
      "loss": 0.10095620155334473,
      "num_input_tokens_seen": 1072790,
      "step": 1033,
      "train_runtime": 2939.0612,
      "train_tokens_per_second": 365.011
    },
    {
      "epoch": 0.4419275563628593,
      "grad_norm": 2.474679708480835,
      "learning_rate": 1.1194860813704497e-05,
      "loss": 0.11916765570640564,
      "num_input_tokens_seen": 1073782,
      "step": 1034,
      "train_runtime": 2941.1466,
      "train_tokens_per_second": 365.09
    },
    {
      "epoch": 0.44235495245218504,
      "grad_norm": 2.29829740524292,
      "learning_rate": 1.1186295503211992e-05,
      "loss": 0.15321902930736542,
      "num_input_tokens_seen": 1074934,
      "step": 1035,
      "train_runtime": 2943.3069,
      "train_tokens_per_second": 365.213
    },
    {
      "epoch": 0.44278234854151083,
      "grad_norm": 2.1527915000915527,
      "learning_rate": 1.1177730192719488e-05,
      "loss": 0.1675380915403366,
      "num_input_tokens_seen": 1076546,
      "step": 1036,
      "train_runtime": 2945.5988,
      "train_tokens_per_second": 365.476
    },
    {
      "epoch": 0.4432097446308366,
      "grad_norm": 3.6376171112060547,
      "learning_rate": 1.1169164882226981e-05,
      "loss": 0.15351618826389313,
      "num_input_tokens_seen": 1077670,
      "step": 1037,
      "train_runtime": 2947.8307,
      "train_tokens_per_second": 365.581
    },
    {
      "epoch": 0.4436371407201624,
      "grad_norm": 1.8380637168884277,
      "learning_rate": 1.1160599571734476e-05,
      "loss": 0.10283884406089783,
      "num_input_tokens_seen": 1079310,
      "step": 1038,
      "train_runtime": 2950.1898,
      "train_tokens_per_second": 365.844
    },
    {
      "epoch": 0.4440645368094882,
      "grad_norm": 2.646578788757324,
      "learning_rate": 1.115203426124197e-05,
      "loss": 0.12246552854776382,
      "num_input_tokens_seen": 1080016,
      "step": 1039,
      "train_runtime": 2952.2349,
      "train_tokens_per_second": 365.83
    },
    {
      "epoch": 0.444491932898814,
      "grad_norm": 1.944064736366272,
      "learning_rate": 1.1143468950749465e-05,
      "loss": 0.11806921660900116,
      "num_input_tokens_seen": 1080900,
      "step": 1040,
      "train_runtime": 2954.3633,
      "train_tokens_per_second": 365.866
    },
    {
      "epoch": 0.44491932898813974,
      "grad_norm": 1.831027865409851,
      "learning_rate": 1.113490364025696e-05,
      "loss": 0.15034857392311096,
      "num_input_tokens_seen": 1082256,
      "step": 1041,
      "train_runtime": 2956.5899,
      "train_tokens_per_second": 366.049
    },
    {
      "epoch": 0.44534672507746553,
      "grad_norm": 1.867666244506836,
      "learning_rate": 1.1126338329764454e-05,
      "loss": 0.09922873973846436,
      "num_input_tokens_seen": 1083402,
      "step": 1042,
      "train_runtime": 2958.7383,
      "train_tokens_per_second": 366.17
    },
    {
      "epoch": 0.4457741211667913,
      "grad_norm": 2.910398483276367,
      "learning_rate": 1.111777301927195e-05,
      "loss": 0.1403721123933792,
      "num_input_tokens_seen": 1084648,
      "step": 1043,
      "train_runtime": 2960.921,
      "train_tokens_per_second": 366.321
    },
    {
      "epoch": 0.4462015172561171,
      "grad_norm": 2.1557788848876953,
      "learning_rate": 1.1109207708779443e-05,
      "loss": 0.15659849345684052,
      "num_input_tokens_seen": 1085474,
      "step": 1044,
      "train_runtime": 2963.0172,
      "train_tokens_per_second": 366.341
    },
    {
      "epoch": 0.4466289133454429,
      "grad_norm": 2.6440045833587646,
      "learning_rate": 1.1100642398286938e-05,
      "loss": 0.10428372025489807,
      "num_input_tokens_seen": 1086270,
      "step": 1045,
      "train_runtime": 2965.1173,
      "train_tokens_per_second": 366.35
    },
    {
      "epoch": 0.44705630943476865,
      "grad_norm": 2.734485626220703,
      "learning_rate": 1.1092077087794434e-05,
      "loss": 0.10328982025384903,
      "num_input_tokens_seen": 1086938,
      "step": 1046,
      "train_runtime": 2967.1437,
      "train_tokens_per_second": 366.325
    },
    {
      "epoch": 0.44748370552409444,
      "grad_norm": 3.2563228607177734,
      "learning_rate": 1.1083511777301927e-05,
      "loss": 0.17912891507148743,
      "num_input_tokens_seen": 1087992,
      "step": 1047,
      "train_runtime": 2969.2902,
      "train_tokens_per_second": 366.415
    },
    {
      "epoch": 0.44791110161342024,
      "grad_norm": 1.6207832098007202,
      "learning_rate": 1.1074946466809422e-05,
      "loss": 0.06897982954978943,
      "num_input_tokens_seen": 1088910,
      "step": 1048,
      "train_runtime": 2971.4223,
      "train_tokens_per_second": 366.461
    },
    {
      "epoch": 0.44833849770274603,
      "grad_norm": 1.9480711221694946,
      "learning_rate": 1.1066381156316916e-05,
      "loss": 0.09259622544050217,
      "num_input_tokens_seen": 1090138,
      "step": 1049,
      "train_runtime": 2973.6208,
      "train_tokens_per_second": 366.603
    },
    {
      "epoch": 0.4487658937920718,
      "grad_norm": 2.1772654056549072,
      "learning_rate": 1.1057815845824411e-05,
      "loss": 0.1047675758600235,
      "num_input_tokens_seen": 1091126,
      "step": 1050,
      "train_runtime": 2975.7483,
      "train_tokens_per_second": 366.673
    },
    {
      "epoch": 0.44919328988139756,
      "grad_norm": 2.8259315490722656,
      "learning_rate": 1.1049250535331907e-05,
      "loss": 0.19120723009109497,
      "num_input_tokens_seen": 1092242,
      "step": 1051,
      "train_runtime": 2990.9356,
      "train_tokens_per_second": 365.184
    },
    {
      "epoch": 0.44962068597072335,
      "grad_norm": 1.824228048324585,
      "learning_rate": 1.10406852248394e-05,
      "loss": 0.10952670872211456,
      "num_input_tokens_seen": 1093496,
      "step": 1052,
      "train_runtime": 2993.0929,
      "train_tokens_per_second": 365.34
    },
    {
      "epoch": 0.45004808206004915,
      "grad_norm": 2.141660451889038,
      "learning_rate": 1.1032119914346895e-05,
      "loss": 0.1616465002298355,
      "num_input_tokens_seen": 1094654,
      "step": 1053,
      "train_runtime": 2995.2837,
      "train_tokens_per_second": 365.459
    },
    {
      "epoch": 0.45047547814937494,
      "grad_norm": 1.8521132469177246,
      "learning_rate": 1.1023554603854392e-05,
      "loss": 0.09414944797754288,
      "num_input_tokens_seen": 1095680,
      "step": 1054,
      "train_runtime": 2997.4303,
      "train_tokens_per_second": 365.54
    },
    {
      "epoch": 0.45090287423870073,
      "grad_norm": 1.6091750860214233,
      "learning_rate": 1.1014989293361886e-05,
      "loss": 0.08558227866888046,
      "num_input_tokens_seen": 1097010,
      "step": 1055,
      "train_runtime": 2999.7884,
      "train_tokens_per_second": 365.696
    },
    {
      "epoch": 0.4513302703280265,
      "grad_norm": 2.276839017868042,
      "learning_rate": 1.1006423982869381e-05,
      "loss": 0.20166617631912231,
      "num_input_tokens_seen": 1098134,
      "step": 1056,
      "train_runtime": 3001.9255,
      "train_tokens_per_second": 365.81
    },
    {
      "epoch": 0.45175766641735227,
      "grad_norm": 1.6320018768310547,
      "learning_rate": 1.0997858672376875e-05,
      "loss": 0.1162157654762268,
      "num_input_tokens_seen": 1099590,
      "step": 1057,
      "train_runtime": 3004.1444,
      "train_tokens_per_second": 366.024
    },
    {
      "epoch": 0.45218506250667806,
      "grad_norm": 1.558932900428772,
      "learning_rate": 1.098929336188437e-05,
      "loss": 0.11004027724266052,
      "num_input_tokens_seen": 1100842,
      "step": 1058,
      "train_runtime": 3006.3288,
      "train_tokens_per_second": 366.175
    },
    {
      "epoch": 0.45261245859600385,
      "grad_norm": 2.182382345199585,
      "learning_rate": 1.0980728051391865e-05,
      "loss": 0.18455377221107483,
      "num_input_tokens_seen": 1102114,
      "step": 1059,
      "train_runtime": 3008.5152,
      "train_tokens_per_second": 366.332
    },
    {
      "epoch": 0.45303985468532965,
      "grad_norm": 2.592876434326172,
      "learning_rate": 1.0972162740899359e-05,
      "loss": 0.18765489757061005,
      "num_input_tokens_seen": 1103470,
      "step": 1060,
      "train_runtime": 3010.6979,
      "train_tokens_per_second": 366.516
    },
    {
      "epoch": 0.45346725077465544,
      "grad_norm": 1.8901042938232422,
      "learning_rate": 1.0963597430406854e-05,
      "loss": 0.12126240134239197,
      "num_input_tokens_seen": 1104796,
      "step": 1061,
      "train_runtime": 3012.9098,
      "train_tokens_per_second": 366.687
    },
    {
      "epoch": 0.4538946468639812,
      "grad_norm": 2.3836684226989746,
      "learning_rate": 1.0955032119914348e-05,
      "loss": 0.12577548623085022,
      "num_input_tokens_seen": 1105546,
      "step": 1062,
      "train_runtime": 3014.9733,
      "train_tokens_per_second": 366.685
    },
    {
      "epoch": 0.45432204295330697,
      "grad_norm": 1.5718857049942017,
      "learning_rate": 1.0946466809421843e-05,
      "loss": 0.08655619621276855,
      "num_input_tokens_seen": 1106758,
      "step": 1063,
      "train_runtime": 3017.1608,
      "train_tokens_per_second": 366.821
    },
    {
      "epoch": 0.45474943904263276,
      "grad_norm": 2.2731409072875977,
      "learning_rate": 1.0937901498929337e-05,
      "loss": 0.13003753125667572,
      "num_input_tokens_seen": 1107616,
      "step": 1064,
      "train_runtime": 3019.3935,
      "train_tokens_per_second": 366.834
    },
    {
      "epoch": 0.45517683513195856,
      "grad_norm": 3.5689444541931152,
      "learning_rate": 1.0929336188436832e-05,
      "loss": 0.09496916830539703,
      "num_input_tokens_seen": 1108174,
      "step": 1065,
      "train_runtime": 3021.3915,
      "train_tokens_per_second": 366.776
    },
    {
      "epoch": 0.45560423122128435,
      "grad_norm": 2.9368834495544434,
      "learning_rate": 1.0920770877944327e-05,
      "loss": 0.2679440677165985,
      "num_input_tokens_seen": 1109014,
      "step": 1066,
      "train_runtime": 3023.4792,
      "train_tokens_per_second": 366.801
    },
    {
      "epoch": 0.4560316273106101,
      "grad_norm": 2.8692476749420166,
      "learning_rate": 1.091220556745182e-05,
      "loss": 0.16424307227134705,
      "num_input_tokens_seen": 1110314,
      "step": 1067,
      "train_runtime": 3025.6513,
      "train_tokens_per_second": 366.967
    },
    {
      "epoch": 0.4564590233999359,
      "grad_norm": 2.4772162437438965,
      "learning_rate": 1.0903640256959316e-05,
      "loss": 0.12424961477518082,
      "num_input_tokens_seen": 1111590,
      "step": 1068,
      "train_runtime": 3027.8209,
      "train_tokens_per_second": 367.125
    },
    {
      "epoch": 0.4568864194892617,
      "grad_norm": 2.8363404273986816,
      "learning_rate": 1.089507494646681e-05,
      "loss": 0.12491759657859802,
      "num_input_tokens_seen": 1112282,
      "step": 1069,
      "train_runtime": 3029.8464,
      "train_tokens_per_second": 367.108
    },
    {
      "epoch": 0.45731381557858747,
      "grad_norm": 1.4680448770523071,
      "learning_rate": 1.0886509635974305e-05,
      "loss": 0.07281026244163513,
      "num_input_tokens_seen": 1113702,
      "step": 1070,
      "train_runtime": 3032.0177,
      "train_tokens_per_second": 367.314
    },
    {
      "epoch": 0.45774121166791326,
      "grad_norm": 1.8230019807815552,
      "learning_rate": 1.08779443254818e-05,
      "loss": 0.10621557384729385,
      "num_input_tokens_seen": 1114542,
      "step": 1071,
      "train_runtime": 3034.1174,
      "train_tokens_per_second": 367.336
    },
    {
      "epoch": 0.458168607757239,
      "grad_norm": 2.3857197761535645,
      "learning_rate": 1.0869379014989294e-05,
      "loss": 0.16738727688789368,
      "num_input_tokens_seen": 1115466,
      "step": 1072,
      "train_runtime": 3036.2895,
      "train_tokens_per_second": 367.378
    },
    {
      "epoch": 0.4585960038465648,
      "grad_norm": 4.6233720779418945,
      "learning_rate": 1.0860813704496789e-05,
      "loss": 0.11537289619445801,
      "num_input_tokens_seen": 1116862,
      "step": 1073,
      "train_runtime": 3038.5363,
      "train_tokens_per_second": 367.566
    },
    {
      "epoch": 0.4590233999358906,
      "grad_norm": 2.1452343463897705,
      "learning_rate": 1.0852248394004283e-05,
      "loss": 0.12661448121070862,
      "num_input_tokens_seen": 1117870,
      "step": 1074,
      "train_runtime": 3040.7798,
      "train_tokens_per_second": 367.626
    },
    {
      "epoch": 0.4594507960252164,
      "grad_norm": 3.0242512226104736,
      "learning_rate": 1.0843683083511778e-05,
      "loss": 0.16758795082569122,
      "num_input_tokens_seen": 1118678,
      "step": 1075,
      "train_runtime": 3042.9225,
      "train_tokens_per_second": 367.633
    },
    {
      "epoch": 0.4598781921145422,
      "grad_norm": 2.6538290977478027,
      "learning_rate": 1.0835117773019273e-05,
      "loss": 0.1627570390701294,
      "num_input_tokens_seen": 1119906,
      "step": 1076,
      "train_runtime": 3045.1219,
      "train_tokens_per_second": 367.771
    },
    {
      "epoch": 0.4603055882038679,
      "grad_norm": 2.2172136306762695,
      "learning_rate": 1.0826552462526767e-05,
      "loss": 0.12060604244470596,
      "num_input_tokens_seen": 1120874,
      "step": 1077,
      "train_runtime": 3047.2954,
      "train_tokens_per_second": 367.826
    },
    {
      "epoch": 0.4607329842931937,
      "grad_norm": 2.375690460205078,
      "learning_rate": 1.0817987152034262e-05,
      "loss": 0.1466825008392334,
      "num_input_tokens_seen": 1122506,
      "step": 1078,
      "train_runtime": 3049.59,
      "train_tokens_per_second": 368.084
    },
    {
      "epoch": 0.4611603803825195,
      "grad_norm": 2.18188738822937,
      "learning_rate": 1.0809421841541756e-05,
      "loss": 0.0772302970290184,
      "num_input_tokens_seen": 1123354,
      "step": 1079,
      "train_runtime": 3051.7196,
      "train_tokens_per_second": 368.105
    },
    {
      "epoch": 0.4615877764718453,
      "grad_norm": 1.4447216987609863,
      "learning_rate": 1.0800856531049251e-05,
      "loss": 0.08115194737911224,
      "num_input_tokens_seen": 1124486,
      "step": 1080,
      "train_runtime": 3053.9253,
      "train_tokens_per_second": 368.21
    },
    {
      "epoch": 0.4620151725611711,
      "grad_norm": 2.4211366176605225,
      "learning_rate": 1.0792291220556746e-05,
      "loss": 0.11334526538848877,
      "num_input_tokens_seen": 1125286,
      "step": 1081,
      "train_runtime": 3070.1898,
      "train_tokens_per_second": 366.52
    },
    {
      "epoch": 0.4624425686504968,
      "grad_norm": 2.6062674522399902,
      "learning_rate": 1.078372591006424e-05,
      "loss": 0.1145646870136261,
      "num_input_tokens_seen": 1126018,
      "step": 1082,
      "train_runtime": 3072.2748,
      "train_tokens_per_second": 366.51
    },
    {
      "epoch": 0.4628699647398226,
      "grad_norm": 1.9511116743087769,
      "learning_rate": 1.0775160599571735e-05,
      "loss": 0.10692603141069412,
      "num_input_tokens_seen": 1127760,
      "step": 1083,
      "train_runtime": 3074.6243,
      "train_tokens_per_second": 366.796
    },
    {
      "epoch": 0.4632973608291484,
      "grad_norm": 2.446035146713257,
      "learning_rate": 1.0766595289079229e-05,
      "loss": 0.11498267203569412,
      "num_input_tokens_seen": 1128962,
      "step": 1084,
      "train_runtime": 3076.8555,
      "train_tokens_per_second": 366.921
    },
    {
      "epoch": 0.4637247569184742,
      "grad_norm": 2.390343427658081,
      "learning_rate": 1.0758029978586724e-05,
      "loss": 0.13486668467521667,
      "num_input_tokens_seen": 1129876,
      "step": 1085,
      "train_runtime": 3078.9818,
      "train_tokens_per_second": 366.964
    },
    {
      "epoch": 0.4641521530078,
      "grad_norm": 1.583939552307129,
      "learning_rate": 1.0749464668094217e-05,
      "loss": 0.09260902553796768,
      "num_input_tokens_seen": 1130876,
      "step": 1086,
      "train_runtime": 3081.1323,
      "train_tokens_per_second": 367.033
    },
    {
      "epoch": 0.4645795490971258,
      "grad_norm": 2.4491169452667236,
      "learning_rate": 1.0740899357601713e-05,
      "loss": 0.16372451186180115,
      "num_input_tokens_seen": 1131882,
      "step": 1087,
      "train_runtime": 3083.2554,
      "train_tokens_per_second": 367.106
    },
    {
      "epoch": 0.4650069451864515,
      "grad_norm": 3.116499900817871,
      "learning_rate": 1.073233404710921e-05,
      "loss": 0.14371392130851746,
      "num_input_tokens_seen": 1132664,
      "step": 1088,
      "train_runtime": 3085.3704,
      "train_tokens_per_second": 367.108
    },
    {
      "epoch": 0.4654343412757773,
      "grad_norm": 2.468092918395996,
      "learning_rate": 1.0723768736616705e-05,
      "loss": 0.12379740178585052,
      "num_input_tokens_seen": 1133980,
      "step": 1089,
      "train_runtime": 3087.6302,
      "train_tokens_per_second": 367.265
    },
    {
      "epoch": 0.4658617373651031,
      "grad_norm": 1.451298475265503,
      "learning_rate": 1.0715203426124199e-05,
      "loss": 0.053941965103149414,
      "num_input_tokens_seen": 1134990,
      "step": 1090,
      "train_runtime": 3089.7509,
      "train_tokens_per_second": 367.34
    },
    {
      "epoch": 0.4662891334544289,
      "grad_norm": 1.9774813652038574,
      "learning_rate": 1.0706638115631694e-05,
      "loss": 0.10296866297721863,
      "num_input_tokens_seen": 1135714,
      "step": 1091,
      "train_runtime": 3091.8237,
      "train_tokens_per_second": 367.328
    },
    {
      "epoch": 0.4667165295437547,
      "grad_norm": 2.596663475036621,
      "learning_rate": 1.0698072805139187e-05,
      "loss": 0.13426145911216736,
      "num_input_tokens_seen": 1136538,
      "step": 1092,
      "train_runtime": 3093.928,
      "train_tokens_per_second": 367.345
    },
    {
      "epoch": 0.46714392563308044,
      "grad_norm": 2.1434266567230225,
      "learning_rate": 1.0689507494646683e-05,
      "loss": 0.12617084383964539,
      "num_input_tokens_seen": 1137472,
      "step": 1093,
      "train_runtime": 3096.061,
      "train_tokens_per_second": 367.393
    },
    {
      "epoch": 0.46757132172240623,
      "grad_norm": 3.238490581512451,
      "learning_rate": 1.0680942184154178e-05,
      "loss": 0.15265128016471863,
      "num_input_tokens_seen": 1138134,
      "step": 1094,
      "train_runtime": 3098.1577,
      "train_tokens_per_second": 367.358
    },
    {
      "epoch": 0.467998717811732,
      "grad_norm": 2.070523977279663,
      "learning_rate": 1.0672376873661672e-05,
      "loss": 0.14197435975074768,
      "num_input_tokens_seen": 1139682,
      "step": 1095,
      "train_runtime": 3100.4517,
      "train_tokens_per_second": 367.586
    },
    {
      "epoch": 0.4684261139010578,
      "grad_norm": 2.5459415912628174,
      "learning_rate": 1.0663811563169167e-05,
      "loss": 0.09419140964746475,
      "num_input_tokens_seen": 1140220,
      "step": 1096,
      "train_runtime": 3102.4968,
      "train_tokens_per_second": 367.517
    },
    {
      "epoch": 0.4688535099903836,
      "grad_norm": 2.419487953186035,
      "learning_rate": 1.065524625267666e-05,
      "loss": 0.19436462223529816,
      "num_input_tokens_seen": 1141284,
      "step": 1097,
      "train_runtime": 3104.689,
      "train_tokens_per_second": 367.6
    },
    {
      "epoch": 0.46928090607970935,
      "grad_norm": 1.9916590452194214,
      "learning_rate": 1.0646680942184156e-05,
      "loss": 0.13321715593338013,
      "num_input_tokens_seen": 1142276,
      "step": 1098,
      "train_runtime": 3106.8078,
      "train_tokens_per_second": 367.669
    },
    {
      "epoch": 0.46970830216903514,
      "grad_norm": 1.3479835987091064,
      "learning_rate": 1.063811563169165e-05,
      "loss": 0.09444232285022736,
      "num_input_tokens_seen": 1143250,
      "step": 1099,
      "train_runtime": 3108.97,
      "train_tokens_per_second": 367.726
    },
    {
      "epoch": 0.47013569825836093,
      "grad_norm": 1.55136239528656,
      "learning_rate": 1.0629550321199145e-05,
      "loss": 0.10088896751403809,
      "num_input_tokens_seen": 1144518,
      "step": 1100,
      "train_runtime": 3111.1929,
      "train_tokens_per_second": 367.871
    },
    {
      "epoch": 0.47056309434768673,
      "grad_norm": 4.632813453674316,
      "learning_rate": 1.062098501070664e-05,
      "loss": 0.10315786302089691,
      "num_input_tokens_seen": 1145362,
      "step": 1101,
      "train_runtime": 3113.3188,
      "train_tokens_per_second": 367.891
    },
    {
      "epoch": 0.4709904904370125,
      "grad_norm": 2.549478054046631,
      "learning_rate": 1.0612419700214133e-05,
      "loss": 0.12106387317180634,
      "num_input_tokens_seen": 1146036,
      "step": 1102,
      "train_runtime": 3115.4203,
      "train_tokens_per_second": 367.859
    },
    {
      "epoch": 0.47141788652633826,
      "grad_norm": 3.3118948936462402,
      "learning_rate": 1.0603854389721629e-05,
      "loss": 0.1757216453552246,
      "num_input_tokens_seen": 1146846,
      "step": 1103,
      "train_runtime": 3117.5255,
      "train_tokens_per_second": 367.871
    },
    {
      "epoch": 0.47184528261566405,
      "grad_norm": 1.837897539138794,
      "learning_rate": 1.0595289079229122e-05,
      "loss": 0.09451344609260559,
      "num_input_tokens_seen": 1147790,
      "step": 1104,
      "train_runtime": 3119.6523,
      "train_tokens_per_second": 367.922
    },
    {
      "epoch": 0.47227267870498985,
      "grad_norm": 3.738784074783325,
      "learning_rate": 1.0586723768736618e-05,
      "loss": 0.13079151511192322,
      "num_input_tokens_seen": 1148852,
      "step": 1105,
      "train_runtime": 3121.8775,
      "train_tokens_per_second": 368.0
    },
    {
      "epoch": 0.47270007479431564,
      "grad_norm": 2.2785274982452393,
      "learning_rate": 1.0578158458244113e-05,
      "loss": 0.12945371866226196,
      "num_input_tokens_seen": 1149814,
      "step": 1106,
      "train_runtime": 3124.0366,
      "train_tokens_per_second": 368.054
    },
    {
      "epoch": 0.47312747088364143,
      "grad_norm": 2.037738084793091,
      "learning_rate": 1.0569593147751606e-05,
      "loss": 0.16386261582374573,
      "num_input_tokens_seen": 1150958,
      "step": 1107,
      "train_runtime": 3126.2366,
      "train_tokens_per_second": 368.161
    },
    {
      "epoch": 0.47355486697296717,
      "grad_norm": 2.0729997158050537,
      "learning_rate": 1.0561027837259102e-05,
      "loss": 0.08901626616716385,
      "num_input_tokens_seen": 1152078,
      "step": 1108,
      "train_runtime": 3128.4068,
      "train_tokens_per_second": 368.263
    },
    {
      "epoch": 0.47398226306229296,
      "grad_norm": 2.2841851711273193,
      "learning_rate": 1.0552462526766595e-05,
      "loss": 0.11797398328781128,
      "num_input_tokens_seen": 1153404,
      "step": 1109,
      "train_runtime": 3130.6683,
      "train_tokens_per_second": 368.421
    },
    {
      "epoch": 0.47440965915161876,
      "grad_norm": 3.03964900970459,
      "learning_rate": 1.054389721627409e-05,
      "loss": 0.18599331378936768,
      "num_input_tokens_seen": 1154246,
      "step": 1110,
      "train_runtime": 3132.8076,
      "train_tokens_per_second": 368.438
    },
    {
      "epoch": 0.47483705524094455,
      "grad_norm": 3.625915050506592,
      "learning_rate": 1.0535331905781586e-05,
      "loss": 0.2271270602941513,
      "num_input_tokens_seen": 1155348,
      "step": 1111,
      "train_runtime": 3147.8775,
      "train_tokens_per_second": 367.024
    },
    {
      "epoch": 0.47526445133027034,
      "grad_norm": 3.42933988571167,
      "learning_rate": 1.052676659528908e-05,
      "loss": 0.09430833160877228,
      "num_input_tokens_seen": 1156148,
      "step": 1112,
      "train_runtime": 3149.9436,
      "train_tokens_per_second": 367.038
    },
    {
      "epoch": 0.47569184741959614,
      "grad_norm": 2.2719924449920654,
      "learning_rate": 1.0518201284796575e-05,
      "loss": 0.13242411613464355,
      "num_input_tokens_seen": 1156964,
      "step": 1113,
      "train_runtime": 3152.0398,
      "train_tokens_per_second": 367.052
    },
    {
      "epoch": 0.4761192435089219,
      "grad_norm": 1.6996687650680542,
      "learning_rate": 1.0509635974304068e-05,
      "loss": 0.08433227241039276,
      "num_input_tokens_seen": 1157788,
      "step": 1114,
      "train_runtime": 3154.1745,
      "train_tokens_per_second": 367.065
    },
    {
      "epoch": 0.47654663959824767,
      "grad_norm": 2.469897985458374,
      "learning_rate": 1.0501070663811564e-05,
      "loss": 0.09990245848894119,
      "num_input_tokens_seen": 1159104,
      "step": 1115,
      "train_runtime": 3156.4218,
      "train_tokens_per_second": 367.221
    },
    {
      "epoch": 0.47697403568757346,
      "grad_norm": 1.3250545263290405,
      "learning_rate": 1.0492505353319057e-05,
      "loss": 0.053221769630908966,
      "num_input_tokens_seen": 1160388,
      "step": 1116,
      "train_runtime": 3158.6599,
      "train_tokens_per_second": 367.367
    },
    {
      "epoch": 0.47740143177689925,
      "grad_norm": 3.156811237335205,
      "learning_rate": 1.0483940042826552e-05,
      "loss": 0.13681703805923462,
      "num_input_tokens_seen": 1161552,
      "step": 1117,
      "train_runtime": 3160.8441,
      "train_tokens_per_second": 367.482
    },
    {
      "epoch": 0.47782882786622505,
      "grad_norm": 3.4449303150177,
      "learning_rate": 1.0475374732334048e-05,
      "loss": 0.2021254450082779,
      "num_input_tokens_seen": 1162296,
      "step": 1118,
      "train_runtime": 3162.9156,
      "train_tokens_per_second": 367.476
    },
    {
      "epoch": 0.4782562239555508,
      "grad_norm": 1.5247373580932617,
      "learning_rate": 1.0466809421841541e-05,
      "loss": 0.06591279059648514,
      "num_input_tokens_seen": 1163414,
      "step": 1119,
      "train_runtime": 3165.0979,
      "train_tokens_per_second": 367.576
    },
    {
      "epoch": 0.4786836200448766,
      "grad_norm": 2.635410785675049,
      "learning_rate": 1.0458244111349037e-05,
      "loss": 0.12360203266143799,
      "num_input_tokens_seen": 1164268,
      "step": 1120,
      "train_runtime": 3167.1809,
      "train_tokens_per_second": 367.604
    },
    {
      "epoch": 0.47911101613420237,
      "grad_norm": 3.1594529151916504,
      "learning_rate": 1.044967880085653e-05,
      "loss": 0.13957934081554413,
      "num_input_tokens_seen": 1164864,
      "step": 1121,
      "train_runtime": 3169.239,
      "train_tokens_per_second": 367.553
    },
    {
      "epoch": 0.47953841222352817,
      "grad_norm": 1.9089677333831787,
      "learning_rate": 1.0441113490364027e-05,
      "loss": 0.11070516705513,
      "num_input_tokens_seen": 1165722,
      "step": 1122,
      "train_runtime": 3171.3625,
      "train_tokens_per_second": 367.578
    },
    {
      "epoch": 0.47996580831285396,
      "grad_norm": 2.39133620262146,
      "learning_rate": 1.0432548179871522e-05,
      "loss": 0.16337795555591583,
      "num_input_tokens_seen": 1166924,
      "step": 1123,
      "train_runtime": 3173.6169,
      "train_tokens_per_second": 367.695
    },
    {
      "epoch": 0.4803932044021797,
      "grad_norm": 1.8524043560028076,
      "learning_rate": 1.0423982869379018e-05,
      "loss": 0.08679244667291641,
      "num_input_tokens_seen": 1168254,
      "step": 1124,
      "train_runtime": 3175.8685,
      "train_tokens_per_second": 367.853
    },
    {
      "epoch": 0.4808206004915055,
      "grad_norm": 4.792642593383789,
      "learning_rate": 1.0415417558886511e-05,
      "loss": 0.1343933343887329,
      "num_input_tokens_seen": 1168950,
      "step": 1125,
      "train_runtime": 3177.965,
      "train_tokens_per_second": 367.83
    },
    {
      "epoch": 0.4812479965808313,
      "grad_norm": 1.6893730163574219,
      "learning_rate": 1.0406852248394007e-05,
      "loss": 0.07165629416704178,
      "num_input_tokens_seen": 1169718,
      "step": 1126,
      "train_runtime": 3180.0567,
      "train_tokens_per_second": 367.829
    },
    {
      "epoch": 0.4816753926701571,
      "grad_norm": 1.994897723197937,
      "learning_rate": 1.03982869379015e-05,
      "loss": 0.09306542575359344,
      "num_input_tokens_seen": 1171080,
      "step": 1127,
      "train_runtime": 3182.2775,
      "train_tokens_per_second": 368.001
    },
    {
      "epoch": 0.48210278875948287,
      "grad_norm": 2.575477361679077,
      "learning_rate": 1.0389721627408995e-05,
      "loss": 0.1872159242630005,
      "num_input_tokens_seen": 1171958,
      "step": 1128,
      "train_runtime": 3184.4013,
      "train_tokens_per_second": 368.031
    },
    {
      "epoch": 0.4825301848488086,
      "grad_norm": 2.260321617126465,
      "learning_rate": 1.0381156316916489e-05,
      "loss": 0.16318023204803467,
      "num_input_tokens_seen": 1173202,
      "step": 1129,
      "train_runtime": 3186.6203,
      "train_tokens_per_second": 368.165
    },
    {
      "epoch": 0.4829575809381344,
      "grad_norm": 2.7041776180267334,
      "learning_rate": 1.0372591006423984e-05,
      "loss": 0.19572733342647552,
      "num_input_tokens_seen": 1173970,
      "step": 1130,
      "train_runtime": 3188.719,
      "train_tokens_per_second": 368.164
    },
    {
      "epoch": 0.4833849770274602,
      "grad_norm": 1.9783432483673096,
      "learning_rate": 1.036402569593148e-05,
      "loss": 0.06867708265781403,
      "num_input_tokens_seen": 1174890,
      "step": 1131,
      "train_runtime": 3190.8409,
      "train_tokens_per_second": 368.207
    },
    {
      "epoch": 0.483812373116786,
      "grad_norm": 2.4956705570220947,
      "learning_rate": 1.0355460385438973e-05,
      "loss": 0.1676766276359558,
      "num_input_tokens_seen": 1175908,
      "step": 1132,
      "train_runtime": 3192.9713,
      "train_tokens_per_second": 368.28
    },
    {
      "epoch": 0.4842397692061118,
      "grad_norm": 2.4380836486816406,
      "learning_rate": 1.0346895074946468e-05,
      "loss": 0.130941241979599,
      "num_input_tokens_seen": 1176818,
      "step": 1133,
      "train_runtime": 3195.0573,
      "train_tokens_per_second": 368.325
    },
    {
      "epoch": 0.4846671652954376,
      "grad_norm": 1.692651629447937,
      "learning_rate": 1.0338329764453962e-05,
      "loss": 0.10316933691501617,
      "num_input_tokens_seen": 1178280,
      "step": 1134,
      "train_runtime": 3197.3389,
      "train_tokens_per_second": 368.519
    },
    {
      "epoch": 0.4850945613847633,
      "grad_norm": 2.2024552822113037,
      "learning_rate": 1.0329764453961457e-05,
      "loss": 0.12320204079151154,
      "num_input_tokens_seen": 1179326,
      "step": 1135,
      "train_runtime": 3199.5084,
      "train_tokens_per_second": 368.596
    },
    {
      "epoch": 0.4855219574740891,
      "grad_norm": 2.9768810272216797,
      "learning_rate": 1.0321199143468953e-05,
      "loss": 0.17908093333244324,
      "num_input_tokens_seen": 1180014,
      "step": 1136,
      "train_runtime": 3201.5969,
      "train_tokens_per_second": 368.57
    },
    {
      "epoch": 0.4859493535634149,
      "grad_norm": 2.0082342624664307,
      "learning_rate": 1.0312633832976446e-05,
      "loss": 0.11483009159564972,
      "num_input_tokens_seen": 1180960,
      "step": 1137,
      "train_runtime": 3203.7482,
      "train_tokens_per_second": 368.618
    },
    {
      "epoch": 0.4863767496527407,
      "grad_norm": 3.8423495292663574,
      "learning_rate": 1.0304068522483941e-05,
      "loss": 0.18052923679351807,
      "num_input_tokens_seen": 1182068,
      "step": 1138,
      "train_runtime": 3205.8882,
      "train_tokens_per_second": 368.718
    },
    {
      "epoch": 0.4868041457420665,
      "grad_norm": 2.5105783939361572,
      "learning_rate": 1.0295503211991435e-05,
      "loss": 0.1909502148628235,
      "num_input_tokens_seen": 1183120,
      "step": 1139,
      "train_runtime": 3208.0686,
      "train_tokens_per_second": 368.795
    },
    {
      "epoch": 0.4872315418313922,
      "grad_norm": 1.609243392944336,
      "learning_rate": 1.028693790149893e-05,
      "loss": 0.09171848744153976,
      "num_input_tokens_seen": 1183972,
      "step": 1140,
      "train_runtime": 3210.1499,
      "train_tokens_per_second": 368.821
    },
    {
      "epoch": 0.487658937920718,
      "grad_norm": 1.9343087673187256,
      "learning_rate": 1.0278372591006426e-05,
      "loss": 0.11184214800596237,
      "num_input_tokens_seen": 1185140,
      "step": 1141,
      "train_runtime": 3225.2393,
      "train_tokens_per_second": 367.458
    },
    {
      "epoch": 0.4880863340100438,
      "grad_norm": 3.0505564212799072,
      "learning_rate": 1.0269807280513919e-05,
      "loss": 0.16290581226348877,
      "num_input_tokens_seen": 1185734,
      "step": 1142,
      "train_runtime": 3227.3386,
      "train_tokens_per_second": 367.403
    },
    {
      "epoch": 0.4885137300993696,
      "grad_norm": 1.7151672840118408,
      "learning_rate": 1.0261241970021414e-05,
      "loss": 0.09398559480905533,
      "num_input_tokens_seen": 1186790,
      "step": 1143,
      "train_runtime": 3229.4989,
      "train_tokens_per_second": 367.484
    },
    {
      "epoch": 0.4889411261886954,
      "grad_norm": 2.7505619525909424,
      "learning_rate": 1.0252676659528908e-05,
      "loss": 0.10065597295761108,
      "num_input_tokens_seen": 1188504,
      "step": 1144,
      "train_runtime": 3231.8289,
      "train_tokens_per_second": 367.75
    },
    {
      "epoch": 0.48936852227802113,
      "grad_norm": 2.978128671646118,
      "learning_rate": 1.0244111349036403e-05,
      "loss": 0.14347213506698608,
      "num_input_tokens_seen": 1189264,
      "step": 1145,
      "train_runtime": 3234.0433,
      "train_tokens_per_second": 367.733
    },
    {
      "epoch": 0.4897959183673469,
      "grad_norm": 2.1196908950805664,
      "learning_rate": 1.0235546038543897e-05,
      "loss": 0.11893812566995621,
      "num_input_tokens_seen": 1190702,
      "step": 1146,
      "train_runtime": 3236.3048,
      "train_tokens_per_second": 367.92
    },
    {
      "epoch": 0.4902233144566727,
      "grad_norm": 1.6588690280914307,
      "learning_rate": 1.0226980728051392e-05,
      "loss": 0.07157114148139954,
      "num_input_tokens_seen": 1191380,
      "step": 1147,
      "train_runtime": 3238.3768,
      "train_tokens_per_second": 367.894
    },
    {
      "epoch": 0.4906507105459985,
      "grad_norm": 1.47933828830719,
      "learning_rate": 1.0218415417558887e-05,
      "loss": 0.0843658298254013,
      "num_input_tokens_seen": 1192598,
      "step": 1148,
      "train_runtime": 3240.5691,
      "train_tokens_per_second": 368.021
    },
    {
      "epoch": 0.4910781066353243,
      "grad_norm": 2.567176103591919,
      "learning_rate": 1.0209850107066381e-05,
      "loss": 0.1020476222038269,
      "num_input_tokens_seen": 1193422,
      "step": 1149,
      "train_runtime": 3242.6771,
      "train_tokens_per_second": 368.036
    },
    {
      "epoch": 0.49150550272465005,
      "grad_norm": 2.123541831970215,
      "learning_rate": 1.0201284796573876e-05,
      "loss": 0.11482197046279907,
      "num_input_tokens_seen": 1194250,
      "step": 1150,
      "train_runtime": 3244.767,
      "train_tokens_per_second": 368.054
    },
    {
      "epoch": 0.49193289881397584,
      "grad_norm": 1.6406511068344116,
      "learning_rate": 1.019271948608137e-05,
      "loss": 0.08300317823886871,
      "num_input_tokens_seen": 1195288,
      "step": 1151,
      "train_runtime": 3246.918,
      "train_tokens_per_second": 368.13
    },
    {
      "epoch": 0.49236029490330163,
      "grad_norm": 1.843297004699707,
      "learning_rate": 1.0184154175588865e-05,
      "loss": 0.11540532112121582,
      "num_input_tokens_seen": 1196658,
      "step": 1152,
      "train_runtime": 3249.1468,
      "train_tokens_per_second": 368.299
    },
    {
      "epoch": 0.4927876909926274,
      "grad_norm": 1.4604040384292603,
      "learning_rate": 1.017558886509636e-05,
      "loss": 0.11157391965389252,
      "num_input_tokens_seen": 1197858,
      "step": 1153,
      "train_runtime": 3251.5115,
      "train_tokens_per_second": 368.4
    },
    {
      "epoch": 0.4932150870819532,
      "grad_norm": 2.0555508136749268,
      "learning_rate": 1.0167023554603854e-05,
      "loss": 0.09021800756454468,
      "num_input_tokens_seen": 1198460,
      "step": 1154,
      "train_runtime": 3253.6023,
      "train_tokens_per_second": 368.349
    },
    {
      "epoch": 0.49364248317127896,
      "grad_norm": 1.7616798877716064,
      "learning_rate": 1.015845824411135e-05,
      "loss": 0.11220099031925201,
      "num_input_tokens_seen": 1199602,
      "step": 1155,
      "train_runtime": 3255.7906,
      "train_tokens_per_second": 368.452
    },
    {
      "epoch": 0.49406987926060475,
      "grad_norm": 2.16310453414917,
      "learning_rate": 1.0149892933618843e-05,
      "loss": 0.12045907229185104,
      "num_input_tokens_seen": 1200682,
      "step": 1156,
      "train_runtime": 3257.9517,
      "train_tokens_per_second": 368.539
    },
    {
      "epoch": 0.49449727534993054,
      "grad_norm": 2.101062536239624,
      "learning_rate": 1.014132762312634e-05,
      "loss": 0.1099109947681427,
      "num_input_tokens_seen": 1201562,
      "step": 1157,
      "train_runtime": 3260.0896,
      "train_tokens_per_second": 368.567
    },
    {
      "epoch": 0.49492467143925634,
      "grad_norm": 1.8282580375671387,
      "learning_rate": 1.0132762312633835e-05,
      "loss": 0.08041372150182724,
      "num_input_tokens_seen": 1202292,
      "step": 1158,
      "train_runtime": 3262.2198,
      "train_tokens_per_second": 368.55
    },
    {
      "epoch": 0.49535206752858213,
      "grad_norm": 2.9186627864837646,
      "learning_rate": 1.0124197002141329e-05,
      "loss": 0.22003036737442017,
      "num_input_tokens_seen": 1203078,
      "step": 1159,
      "train_runtime": 3264.3244,
      "train_tokens_per_second": 368.553
    },
    {
      "epoch": 0.4957794636179079,
      "grad_norm": 2.7521979808807373,
      "learning_rate": 1.0115631691648824e-05,
      "loss": 0.1638396978378296,
      "num_input_tokens_seen": 1204314,
      "step": 1160,
      "train_runtime": 3266.5097,
      "train_tokens_per_second": 368.685
    },
    {
      "epoch": 0.49620685970723366,
      "grad_norm": 1.867957592010498,
      "learning_rate": 1.010706638115632e-05,
      "loss": 0.12579822540283203,
      "num_input_tokens_seen": 1205414,
      "step": 1161,
      "train_runtime": 3268.6777,
      "train_tokens_per_second": 368.777
    },
    {
      "epoch": 0.49663425579655945,
      "grad_norm": 2.0077366828918457,
      "learning_rate": 1.0098501070663813e-05,
      "loss": 0.09778191149234772,
      "num_input_tokens_seen": 1206124,
      "step": 1162,
      "train_runtime": 3270.7816,
      "train_tokens_per_second": 368.757
    },
    {
      "epoch": 0.49706165188588525,
      "grad_norm": 2.3758699893951416,
      "learning_rate": 1.0089935760171308e-05,
      "loss": 0.18004055321216583,
      "num_input_tokens_seen": 1207038,
      "step": 1163,
      "train_runtime": 3272.9654,
      "train_tokens_per_second": 368.79
    },
    {
      "epoch": 0.49748904797521104,
      "grad_norm": 2.7922658920288086,
      "learning_rate": 1.0081370449678802e-05,
      "loss": 0.15281176567077637,
      "num_input_tokens_seen": 1207746,
      "step": 1164,
      "train_runtime": 3275.0377,
      "train_tokens_per_second": 368.773
    },
    {
      "epoch": 0.49791644406453683,
      "grad_norm": 2.1049084663391113,
      "learning_rate": 1.0072805139186297e-05,
      "loss": 0.12039857357740402,
      "num_input_tokens_seen": 1208784,
      "step": 1165,
      "train_runtime": 3277.1645,
      "train_tokens_per_second": 368.851
    },
    {
      "epoch": 0.49834384015386257,
      "grad_norm": 2.1592531204223633,
      "learning_rate": 1.0064239828693792e-05,
      "loss": 0.1376982480287552,
      "num_input_tokens_seen": 1209892,
      "step": 1166,
      "train_runtime": 3279.3199,
      "train_tokens_per_second": 368.946
    },
    {
      "epoch": 0.49877123624318837,
      "grad_norm": 1.9224379062652588,
      "learning_rate": 1.0055674518201286e-05,
      "loss": 0.10606247186660767,
      "num_input_tokens_seen": 1210896,
      "step": 1167,
      "train_runtime": 3281.4888,
      "train_tokens_per_second": 369.008
    },
    {
      "epoch": 0.49919863233251416,
      "grad_norm": 2.5508010387420654,
      "learning_rate": 1.0047109207708781e-05,
      "loss": 0.15957972407341003,
      "num_input_tokens_seen": 1211932,
      "step": 1168,
      "train_runtime": 3283.6383,
      "train_tokens_per_second": 369.082
    },
    {
      "epoch": 0.49962602842183995,
      "grad_norm": 1.4186806678771973,
      "learning_rate": 1.0038543897216275e-05,
      "loss": 0.08027073740959167,
      "num_input_tokens_seen": 1213072,
      "step": 1169,
      "train_runtime": 3285.8152,
      "train_tokens_per_second": 369.184
    },
    {
      "epoch": 0.5000534245111657,
      "grad_norm": 2.0232720375061035,
      "learning_rate": 1.002997858672377e-05,
      "loss": 0.10638853162527084,
      "num_input_tokens_seen": 1214044,
      "step": 1170,
      "train_runtime": 3287.9613,
      "train_tokens_per_second": 369.239
    },
    {
      "epoch": 0.5000534245111657,
      "eval_loss": 0.48118919134140015,
      "eval_runtime": 58.0838,
      "eval_samples_per_second": 17.182,
      "eval_steps_per_second": 8.591,
      "num_input_tokens_seen": 1214044,
      "step": 1170
    },
    {
      "epoch": 0.5004808206004915,
      "grad_norm": 1.5170239210128784,
      "learning_rate": 1.0021413276231265e-05,
      "loss": 0.08273788541555405,
      "num_input_tokens_seen": 1215246,
      "step": 1171,
      "train_runtime": 3361.4712,
      "train_tokens_per_second": 361.522
    },
    {
      "epoch": 0.5009082166898173,
      "grad_norm": 2.31028151512146,
      "learning_rate": 1.0012847965738759e-05,
      "loss": 0.18249736726284027,
      "num_input_tokens_seen": 1216162,
      "step": 1172,
      "train_runtime": 3363.609,
      "train_tokens_per_second": 361.565
    },
    {
      "epoch": 0.5013356127791431,
      "grad_norm": 1.6480261087417603,
      "learning_rate": 1.0004282655246254e-05,
      "loss": 0.08653301000595093,
      "num_input_tokens_seen": 1218486,
      "step": 1173,
      "train_runtime": 3366.1099,
      "train_tokens_per_second": 361.986
    },
    {
      "epoch": 0.5017630088684688,
      "grad_norm": 2.034839630126953,
      "learning_rate": 9.995717344753748e-06,
      "loss": 0.16095291078090668,
      "num_input_tokens_seen": 1219380,
      "step": 1174,
      "train_runtime": 3368.2384,
      "train_tokens_per_second": 362.023
    },
    {
      "epoch": 0.5021904049577947,
      "grad_norm": 3.8672940731048584,
      "learning_rate": 9.987152034261243e-06,
      "loss": 0.09606952220201492,
      "num_input_tokens_seen": 1219936,
      "step": 1175,
      "train_runtime": 3370.2717,
      "train_tokens_per_second": 361.97
    },
    {
      "epoch": 0.5026178010471204,
      "grad_norm": 3.137542963027954,
      "learning_rate": 9.978586723768736e-06,
      "loss": 0.19476166367530823,
      "num_input_tokens_seen": 1220910,
      "step": 1176,
      "train_runtime": 3372.4389,
      "train_tokens_per_second": 362.026
    },
    {
      "epoch": 0.5030451971364462,
      "grad_norm": 2.6571812629699707,
      "learning_rate": 9.970021413276232e-06,
      "loss": 0.1706370711326599,
      "num_input_tokens_seen": 1221952,
      "step": 1177,
      "train_runtime": 3374.6215,
      "train_tokens_per_second": 362.1
    },
    {
      "epoch": 0.503472593225772,
      "grad_norm": 2.6814627647399902,
      "learning_rate": 9.961456102783727e-06,
      "loss": 0.17683133482933044,
      "num_input_tokens_seen": 1222912,
      "step": 1178,
      "train_runtime": 3376.774,
      "train_tokens_per_second": 362.154
    },
    {
      "epoch": 0.5038999893150977,
      "grad_norm": 2.4631879329681396,
      "learning_rate": 9.95289079229122e-06,
      "loss": 0.19499963521957397,
      "num_input_tokens_seen": 1224156,
      "step": 1179,
      "train_runtime": 3378.9989,
      "train_tokens_per_second": 362.284
    },
    {
      "epoch": 0.5043273854044236,
      "grad_norm": 2.1227729320526123,
      "learning_rate": 9.944325481798716e-06,
      "loss": 0.1360645443201065,
      "num_input_tokens_seen": 1225652,
      "step": 1180,
      "train_runtime": 3381.4734,
      "train_tokens_per_second": 362.461
    },
    {
      "epoch": 0.5047547814937493,
      "grad_norm": 2.7035768032073975,
      "learning_rate": 9.93576017130621e-06,
      "loss": 0.06934560090303421,
      "num_input_tokens_seen": 1227204,
      "step": 1181,
      "train_runtime": 3383.804,
      "train_tokens_per_second": 362.67
    },
    {
      "epoch": 0.5051821775830752,
      "grad_norm": 2.2245774269104004,
      "learning_rate": 9.927194860813705e-06,
      "loss": 0.12962600588798523,
      "num_input_tokens_seen": 1228700,
      "step": 1182,
      "train_runtime": 3386.0877,
      "train_tokens_per_second": 362.867
    },
    {
      "epoch": 0.5056095736724009,
      "grad_norm": 1.8185889720916748,
      "learning_rate": 9.9186295503212e-06,
      "loss": 0.11246130615472794,
      "num_input_tokens_seen": 1229644,
      "step": 1183,
      "train_runtime": 3388.2374,
      "train_tokens_per_second": 362.916
    },
    {
      "epoch": 0.5060369697617266,
      "grad_norm": 4.234601020812988,
      "learning_rate": 9.910064239828695e-06,
      "loss": 0.16835813224315643,
      "num_input_tokens_seen": 1230508,
      "step": 1184,
      "train_runtime": 3390.3644,
      "train_tokens_per_second": 362.943
    },
    {
      "epoch": 0.5064643658510525,
      "grad_norm": 1.8631402254104614,
      "learning_rate": 9.901498929336189e-06,
      "loss": 0.09280159324407578,
      "num_input_tokens_seen": 1231260,
      "step": 1185,
      "train_runtime": 3392.491,
      "train_tokens_per_second": 362.937
    },
    {
      "epoch": 0.5068917619403782,
      "grad_norm": 2.7326459884643555,
      "learning_rate": 9.892933618843684e-06,
      "loss": 0.1309550255537033,
      "num_input_tokens_seen": 1232160,
      "step": 1186,
      "train_runtime": 3394.6144,
      "train_tokens_per_second": 362.975
    },
    {
      "epoch": 0.5073191580297041,
      "grad_norm": 2.118839979171753,
      "learning_rate": 9.88436830835118e-06,
      "loss": 0.0755138024687767,
      "num_input_tokens_seen": 1233090,
      "step": 1187,
      "train_runtime": 3396.7385,
      "train_tokens_per_second": 363.022
    },
    {
      "epoch": 0.5077465541190298,
      "grad_norm": 1.8691673278808594,
      "learning_rate": 9.875802997858673e-06,
      "loss": 0.13339178264141083,
      "num_input_tokens_seen": 1234300,
      "step": 1188,
      "train_runtime": 3398.9316,
      "train_tokens_per_second": 363.144
    },
    {
      "epoch": 0.5081739502083555,
      "grad_norm": 2.3717668056488037,
      "learning_rate": 9.867237687366168e-06,
      "loss": 0.11112445592880249,
      "num_input_tokens_seen": 1235186,
      "step": 1189,
      "train_runtime": 3401.0631,
      "train_tokens_per_second": 363.176
    },
    {
      "epoch": 0.5086013462976814,
      "grad_norm": 1.7401413917541504,
      "learning_rate": 9.858672376873662e-06,
      "loss": 0.099689781665802,
      "num_input_tokens_seen": 1236154,
      "step": 1190,
      "train_runtime": 3403.2328,
      "train_tokens_per_second": 363.229
    },
    {
      "epoch": 0.5090287423870071,
      "grad_norm": 2.3835678100585938,
      "learning_rate": 9.850107066381157e-06,
      "loss": 0.14031466841697693,
      "num_input_tokens_seen": 1237918,
      "step": 1191,
      "train_runtime": 3405.5809,
      "train_tokens_per_second": 363.497
    },
    {
      "epoch": 0.509456138476333,
      "grad_norm": 4.456737995147705,
      "learning_rate": 9.841541755888652e-06,
      "loss": 0.14086435735225677,
      "num_input_tokens_seen": 1238520,
      "step": 1192,
      "train_runtime": 3407.6646,
      "train_tokens_per_second": 363.451
    },
    {
      "epoch": 0.5098835345656587,
      "grad_norm": 3.505234479904175,
      "learning_rate": 9.832976445396146e-06,
      "loss": 0.13380852341651917,
      "num_input_tokens_seen": 1239132,
      "step": 1193,
      "train_runtime": 3409.7363,
      "train_tokens_per_second": 363.41
    },
    {
      "epoch": 0.5103109306549845,
      "grad_norm": 3.0331313610076904,
      "learning_rate": 9.824411134903641e-06,
      "loss": 0.20048747956752777,
      "num_input_tokens_seen": 1239858,
      "step": 1194,
      "train_runtime": 3411.8319,
      "train_tokens_per_second": 363.399
    },
    {
      "epoch": 0.5107383267443103,
      "grad_norm": 2.407073974609375,
      "learning_rate": 9.815845824411135e-06,
      "loss": 0.13325142860412598,
      "num_input_tokens_seen": 1240696,
      "step": 1195,
      "train_runtime": 3414.029,
      "train_tokens_per_second": 363.411
    },
    {
      "epoch": 0.511165722833636,
      "grad_norm": 1.8318616151809692,
      "learning_rate": 9.80728051391863e-06,
      "loss": 0.1460091769695282,
      "num_input_tokens_seen": 1242014,
      "step": 1196,
      "train_runtime": 3416.2495,
      "train_tokens_per_second": 363.561
    },
    {
      "epoch": 0.5115931189229619,
      "grad_norm": 2.2184696197509766,
      "learning_rate": 9.798715203426125e-06,
      "loss": 0.09027564525604248,
      "num_input_tokens_seen": 1242658,
      "step": 1197,
      "train_runtime": 3418.3176,
      "train_tokens_per_second": 363.529
    },
    {
      "epoch": 0.5120205150122876,
      "grad_norm": 1.6639877557754517,
      "learning_rate": 9.790149892933619e-06,
      "loss": 0.1325250267982483,
      "num_input_tokens_seen": 1244118,
      "step": 1198,
      "train_runtime": 3420.5878,
      "train_tokens_per_second": 363.715
    },
    {
      "epoch": 0.5124479111016135,
      "grad_norm": 2.9893813133239746,
      "learning_rate": 9.781584582441114e-06,
      "loss": 0.1728774458169937,
      "num_input_tokens_seen": 1244880,
      "step": 1199,
      "train_runtime": 3422.6853,
      "train_tokens_per_second": 363.714
    },
    {
      "epoch": 0.5128753071909392,
      "grad_norm": 3.066741704940796,
      "learning_rate": 9.77301927194861e-06,
      "loss": 0.18558862805366516,
      "num_input_tokens_seen": 1246146,
      "step": 1200,
      "train_runtime": 3424.8829,
      "train_tokens_per_second": 363.851
    },
    {
      "epoch": 0.513302703280265,
      "grad_norm": 2.065859317779541,
      "learning_rate": 9.764453961456105e-06,
      "loss": 0.1163623183965683,
      "num_input_tokens_seen": 1246980,
      "step": 1201,
      "train_runtime": 3440.2779,
      "train_tokens_per_second": 362.465
    },
    {
      "epoch": 0.5137300993695908,
      "grad_norm": 1.8905599117279053,
      "learning_rate": 9.755888650963598e-06,
      "loss": 0.10784266144037247,
      "num_input_tokens_seen": 1247924,
      "step": 1202,
      "train_runtime": 3442.399,
      "train_tokens_per_second": 362.516
    },
    {
      "epoch": 0.5141574954589165,
      "grad_norm": 1.7787706851959229,
      "learning_rate": 9.747323340471094e-06,
      "loss": 0.10549870133399963,
      "num_input_tokens_seen": 1248902,
      "step": 1203,
      "train_runtime": 3444.5089,
      "train_tokens_per_second": 362.578
    },
    {
      "epoch": 0.5145848915482424,
      "grad_norm": 2.5742433071136475,
      "learning_rate": 9.738758029978587e-06,
      "loss": 0.09015730768442154,
      "num_input_tokens_seen": 1249520,
      "step": 1204,
      "train_runtime": 3446.6034,
      "train_tokens_per_second": 362.537
    },
    {
      "epoch": 0.5150122876375681,
      "grad_norm": 1.609961986541748,
      "learning_rate": 9.730192719486083e-06,
      "loss": 0.11445249617099762,
      "num_input_tokens_seen": 1250896,
      "step": 1205,
      "train_runtime": 3448.9289,
      "train_tokens_per_second": 362.691
    },
    {
      "epoch": 0.5154396837268939,
      "grad_norm": 2.958165407180786,
      "learning_rate": 9.721627408993576e-06,
      "loss": 0.21736612915992737,
      "num_input_tokens_seen": 1252022,
      "step": 1206,
      "train_runtime": 3451.0866,
      "train_tokens_per_second": 362.791
    },
    {
      "epoch": 0.5158670798162197,
      "grad_norm": 2.1003317832946777,
      "learning_rate": 9.713062098501071e-06,
      "loss": 0.12407387793064117,
      "num_input_tokens_seen": 1253436,
      "step": 1207,
      "train_runtime": 3453.3539,
      "train_tokens_per_second": 362.962
    },
    {
      "epoch": 0.5162944759055454,
      "grad_norm": 1.9921725988388062,
      "learning_rate": 9.704496788008567e-06,
      "loss": 0.09868897497653961,
      "num_input_tokens_seen": 1254522,
      "step": 1208,
      "train_runtime": 3455.5203,
      "train_tokens_per_second": 363.049
    },
    {
      "epoch": 0.5167218719948713,
      "grad_norm": 1.5106102228164673,
      "learning_rate": 9.69593147751606e-06,
      "loss": 0.08805327862501144,
      "num_input_tokens_seen": 1256084,
      "step": 1209,
      "train_runtime": 3457.798,
      "train_tokens_per_second": 363.261
    },
    {
      "epoch": 0.517149268084197,
      "grad_norm": 2.3006861209869385,
      "learning_rate": 9.687366167023556e-06,
      "loss": 0.11188167333602905,
      "num_input_tokens_seen": 1257036,
      "step": 1210,
      "train_runtime": 3459.9239,
      "train_tokens_per_second": 363.313
    },
    {
      "epoch": 0.5175766641735228,
      "grad_norm": 2.1414601802825928,
      "learning_rate": 9.678800856531049e-06,
      "loss": 0.1338559091091156,
      "num_input_tokens_seen": 1257922,
      "step": 1211,
      "train_runtime": 3462.078,
      "train_tokens_per_second": 363.343
    },
    {
      "epoch": 0.5180040602628486,
      "grad_norm": 1.3698714971542358,
      "learning_rate": 9.670235546038544e-06,
      "loss": 0.06936321407556534,
      "num_input_tokens_seen": 1259290,
      "step": 1212,
      "train_runtime": 3464.3316,
      "train_tokens_per_second": 363.502
    },
    {
      "epoch": 0.5184314563521744,
      "grad_norm": 1.597274899482727,
      "learning_rate": 9.66167023554604e-06,
      "loss": 0.09593281149864197,
      "num_input_tokens_seen": 1260326,
      "step": 1213,
      "train_runtime": 3466.5702,
      "train_tokens_per_second": 363.566
    },
    {
      "epoch": 0.5188588524415002,
      "grad_norm": 3.404883623123169,
      "learning_rate": 9.653104925053533e-06,
      "loss": 0.11273635923862457,
      "num_input_tokens_seen": 1261044,
      "step": 1214,
      "train_runtime": 3468.7075,
      "train_tokens_per_second": 363.549
    },
    {
      "epoch": 0.519286248530826,
      "grad_norm": 1.7912863492965698,
      "learning_rate": 9.644539614561029e-06,
      "loss": 0.0925249308347702,
      "num_input_tokens_seen": 1262616,
      "step": 1215,
      "train_runtime": 3470.9765,
      "train_tokens_per_second": 363.764
    },
    {
      "epoch": 0.5197136446201517,
      "grad_norm": 3.5222458839416504,
      "learning_rate": 9.635974304068522e-06,
      "loss": 0.2118309736251831,
      "num_input_tokens_seen": 1263170,
      "step": 1216,
      "train_runtime": 3472.9818,
      "train_tokens_per_second": 363.713
    },
    {
      "epoch": 0.5201410407094775,
      "grad_norm": 2.2413337230682373,
      "learning_rate": 9.627408993576019e-06,
      "loss": 0.1166030615568161,
      "num_input_tokens_seen": 1264336,
      "step": 1217,
      "train_runtime": 3475.1598,
      "train_tokens_per_second": 363.821
    },
    {
      "epoch": 0.5205684367988033,
      "grad_norm": 1.8757281303405762,
      "learning_rate": 9.618843683083513e-06,
      "loss": 0.13037443161010742,
      "num_input_tokens_seen": 1265598,
      "step": 1218,
      "train_runtime": 3477.3579,
      "train_tokens_per_second": 363.954
    },
    {
      "epoch": 0.5209958328881291,
      "grad_norm": 2.1826701164245605,
      "learning_rate": 9.610278372591008e-06,
      "loss": 0.15989816188812256,
      "num_input_tokens_seen": 1266874,
      "step": 1219,
      "train_runtime": 3479.5522,
      "train_tokens_per_second": 364.091
    },
    {
      "epoch": 0.5214232289774549,
      "grad_norm": 2.7870330810546875,
      "learning_rate": 9.601713062098502e-06,
      "loss": 0.16210657358169556,
      "num_input_tokens_seen": 1267652,
      "step": 1220,
      "train_runtime": 3481.6253,
      "train_tokens_per_second": 364.098
    },
    {
      "epoch": 0.5218506250667806,
      "grad_norm": 1.4903841018676758,
      "learning_rate": 9.593147751605997e-06,
      "loss": 0.06741855293512344,
      "num_input_tokens_seen": 1269138,
      "step": 1221,
      "train_runtime": 3483.8928,
      "train_tokens_per_second": 364.287
    },
    {
      "epoch": 0.5222780211561064,
      "grad_norm": 2.038149118423462,
      "learning_rate": 9.584582441113492e-06,
      "loss": 0.09281060844659805,
      "num_input_tokens_seen": 1269886,
      "step": 1222,
      "train_runtime": 3485.9697,
      "train_tokens_per_second": 364.285
    },
    {
      "epoch": 0.5227054172454322,
      "grad_norm": 2.670989513397217,
      "learning_rate": 9.576017130620986e-06,
      "loss": 0.14042116701602936,
      "num_input_tokens_seen": 1270554,
      "step": 1223,
      "train_runtime": 3488.0105,
      "train_tokens_per_second": 364.263
    },
    {
      "epoch": 0.523132813334758,
      "grad_norm": 2.074495553970337,
      "learning_rate": 9.567451820128481e-06,
      "loss": 0.09043486416339874,
      "num_input_tokens_seen": 1271226,
      "step": 1224,
      "train_runtime": 3490.0519,
      "train_tokens_per_second": 364.243
    },
    {
      "epoch": 0.5235602094240838,
      "grad_norm": 2.4718987941741943,
      "learning_rate": 9.558886509635975e-06,
      "loss": 0.07171086966991425,
      "num_input_tokens_seen": 1272126,
      "step": 1225,
      "train_runtime": 3492.1689,
      "train_tokens_per_second": 364.28
    },
    {
      "epoch": 0.5239876055134095,
      "grad_norm": 0.7584564089775085,
      "learning_rate": 9.55032119914347e-06,
      "loss": 0.04889415204524994,
      "num_input_tokens_seen": 1275302,
      "step": 1226,
      "train_runtime": 3494.8841,
      "train_tokens_per_second": 364.905
    },
    {
      "epoch": 0.5244150016027354,
      "grad_norm": 2.5239851474761963,
      "learning_rate": 9.541755888650965e-06,
      "loss": 0.14348958432674408,
      "num_input_tokens_seen": 1276538,
      "step": 1227,
      "train_runtime": 3497.0855,
      "train_tokens_per_second": 365.029
    },
    {
      "epoch": 0.5248423976920611,
      "grad_norm": 2.036489725112915,
      "learning_rate": 9.533190578158459e-06,
      "loss": 0.13829120993614197,
      "num_input_tokens_seen": 1277534,
      "step": 1228,
      "train_runtime": 3499.2276,
      "train_tokens_per_second": 365.09
    },
    {
      "epoch": 0.5252697937813869,
      "grad_norm": 2.100877046585083,
      "learning_rate": 9.524625267665954e-06,
      "loss": 0.12332621961832047,
      "num_input_tokens_seen": 1278534,
      "step": 1229,
      "train_runtime": 3501.3718,
      "train_tokens_per_second": 365.152
    },
    {
      "epoch": 0.5256971898707127,
      "grad_norm": 2.741685628890991,
      "learning_rate": 9.516059957173448e-06,
      "loss": 0.17697112262248993,
      "num_input_tokens_seen": 1279530,
      "step": 1230,
      "train_runtime": 3503.5107,
      "train_tokens_per_second": 365.214
    },
    {
      "epoch": 0.5261245859600384,
      "grad_norm": 2.025456428527832,
      "learning_rate": 9.507494646680943e-06,
      "loss": 0.09883399307727814,
      "num_input_tokens_seen": 1280870,
      "step": 1231,
      "train_runtime": 3518.6695,
      "train_tokens_per_second": 364.021
    },
    {
      "epoch": 0.5265519820493643,
      "grad_norm": 1.5998183488845825,
      "learning_rate": 9.498929336188436e-06,
      "loss": 0.08645810931921005,
      "num_input_tokens_seen": 1282036,
      "step": 1232,
      "train_runtime": 3520.8479,
      "train_tokens_per_second": 364.127
    },
    {
      "epoch": 0.52697937813869,
      "grad_norm": 1.8249338865280151,
      "learning_rate": 9.490364025695932e-06,
      "loss": 0.08915052562952042,
      "num_input_tokens_seen": 1282954,
      "step": 1233,
      "train_runtime": 3522.9921,
      "train_tokens_per_second": 364.166
    },
    {
      "epoch": 0.5274067742280159,
      "grad_norm": 2.8365821838378906,
      "learning_rate": 9.481798715203427e-06,
      "loss": 0.17887720465660095,
      "num_input_tokens_seen": 1283920,
      "step": 1234,
      "train_runtime": 3525.1585,
      "train_tokens_per_second": 364.216
    },
    {
      "epoch": 0.5278341703173416,
      "grad_norm": 2.6319785118103027,
      "learning_rate": 9.473233404710922e-06,
      "loss": 0.13789516687393188,
      "num_input_tokens_seen": 1284836,
      "step": 1235,
      "train_runtime": 3527.2861,
      "train_tokens_per_second": 364.256
    },
    {
      "epoch": 0.5282615664066673,
      "grad_norm": 2.3702616691589355,
      "learning_rate": 9.464668094218416e-06,
      "loss": 0.1556377112865448,
      "num_input_tokens_seen": 1285940,
      "step": 1236,
      "train_runtime": 3529.4768,
      "train_tokens_per_second": 364.343
    },
    {
      "epoch": 0.5286889624959932,
      "grad_norm": 3.1765713691711426,
      "learning_rate": 9.456102783725911e-06,
      "loss": 0.22601580619812012,
      "num_input_tokens_seen": 1286944,
      "step": 1237,
      "train_runtime": 3531.6404,
      "train_tokens_per_second": 364.404
    },
    {
      "epoch": 0.5291163585853189,
      "grad_norm": 2.4900319576263428,
      "learning_rate": 9.447537473233406e-06,
      "loss": 0.0913250520825386,
      "num_input_tokens_seen": 1287744,
      "step": 1238,
      "train_runtime": 3533.7318,
      "train_tokens_per_second": 364.415
    },
    {
      "epoch": 0.5295437546746448,
      "grad_norm": 3.7495579719543457,
      "learning_rate": 9.4389721627409e-06,
      "loss": 0.20030948519706726,
      "num_input_tokens_seen": 1289008,
      "step": 1239,
      "train_runtime": 3535.9519,
      "train_tokens_per_second": 364.543
    },
    {
      "epoch": 0.5299711507639705,
      "grad_norm": 2.7258265018463135,
      "learning_rate": 9.430406852248395e-06,
      "loss": 0.13150939345359802,
      "num_input_tokens_seen": 1289992,
      "step": 1240,
      "train_runtime": 3538.1206,
      "train_tokens_per_second": 364.598
    },
    {
      "epoch": 0.5303985468532962,
      "grad_norm": 1.9053395986557007,
      "learning_rate": 9.421841541755889e-06,
      "loss": 0.10370547324419022,
      "num_input_tokens_seen": 1291310,
      "step": 1241,
      "train_runtime": 3540.351,
      "train_tokens_per_second": 364.741
    },
    {
      "epoch": 0.5308259429426221,
      "grad_norm": 2.556393623352051,
      "learning_rate": 9.413276231263384e-06,
      "loss": 0.13149718940258026,
      "num_input_tokens_seen": 1292422,
      "step": 1242,
      "train_runtime": 3542.7393,
      "train_tokens_per_second": 364.809
    },
    {
      "epoch": 0.5312533390319478,
      "grad_norm": 2.999880790710449,
      "learning_rate": 9.40471092077088e-06,
      "loss": 0.14596442878246307,
      "num_input_tokens_seen": 1293270,
      "step": 1243,
      "train_runtime": 3544.8678,
      "train_tokens_per_second": 364.829
    },
    {
      "epoch": 0.5316807351212737,
      "grad_norm": 2.2344367504119873,
      "learning_rate": 9.396145610278373e-06,
      "loss": 0.08628234267234802,
      "num_input_tokens_seen": 1293884,
      "step": 1244,
      "train_runtime": 3546.948,
      "train_tokens_per_second": 364.788
    },
    {
      "epoch": 0.5321081312105994,
      "grad_norm": 2.489434242248535,
      "learning_rate": 9.387580299785868e-06,
      "loss": 0.15407121181488037,
      "num_input_tokens_seen": 1294644,
      "step": 1245,
      "train_runtime": 3549.0603,
      "train_tokens_per_second": 364.785
    },
    {
      "epoch": 0.5325355272999253,
      "grad_norm": 3.557612180709839,
      "learning_rate": 9.379014989293362e-06,
      "loss": 0.1567760854959488,
      "num_input_tokens_seen": 1295360,
      "step": 1246,
      "train_runtime": 3551.1367,
      "train_tokens_per_second": 364.773
    },
    {
      "epoch": 0.532962923389251,
      "grad_norm": 2.3730432987213135,
      "learning_rate": 9.370449678800857e-06,
      "loss": 0.1668788641691208,
      "num_input_tokens_seen": 1296580,
      "step": 1247,
      "train_runtime": 3553.3615,
      "train_tokens_per_second": 364.888
    },
    {
      "epoch": 0.5333903194785767,
      "grad_norm": 2.1229710578918457,
      "learning_rate": 9.361884368308352e-06,
      "loss": 0.10964275896549225,
      "num_input_tokens_seen": 1297590,
      "step": 1248,
      "train_runtime": 3555.5063,
      "train_tokens_per_second": 364.952
    },
    {
      "epoch": 0.5338177155679026,
      "grad_norm": 1.8034108877182007,
      "learning_rate": 9.353319057815846e-06,
      "loss": 0.10276772081851959,
      "num_input_tokens_seen": 1298560,
      "step": 1249,
      "train_runtime": 3557.6959,
      "train_tokens_per_second": 365.0
    },
    {
      "epoch": 0.5342451116572283,
      "grad_norm": 2.4474294185638428,
      "learning_rate": 9.344753747323341e-06,
      "loss": 0.1718776971101761,
      "num_input_tokens_seen": 1299314,
      "step": 1250,
      "train_runtime": 3559.8205,
      "train_tokens_per_second": 364.994
    },
    {
      "epoch": 0.5346725077465542,
      "grad_norm": 2.3221609592437744,
      "learning_rate": 9.336188436830836e-06,
      "loss": 0.09346188604831696,
      "num_input_tokens_seen": 1300070,
      "step": 1251,
      "train_runtime": 3561.9873,
      "train_tokens_per_second": 364.984
    },
    {
      "epoch": 0.5350999038358799,
      "grad_norm": 3.941173553466797,
      "learning_rate": 9.327623126338332e-06,
      "loss": 0.15250864624977112,
      "num_input_tokens_seen": 1300824,
      "step": 1252,
      "train_runtime": 3564.0868,
      "train_tokens_per_second": 364.981
    },
    {
      "epoch": 0.5355272999252056,
      "grad_norm": 2.7000339031219482,
      "learning_rate": 9.319057815845825e-06,
      "loss": 0.17314130067825317,
      "num_input_tokens_seen": 1301600,
      "step": 1253,
      "train_runtime": 3566.2034,
      "train_tokens_per_second": 364.982
    },
    {
      "epoch": 0.5359546960145315,
      "grad_norm": 2.440690517425537,
      "learning_rate": 9.31049250535332e-06,
      "loss": 0.11099924147129059,
      "num_input_tokens_seen": 1302436,
      "step": 1254,
      "train_runtime": 3568.3639,
      "train_tokens_per_second": 364.995
    },
    {
      "epoch": 0.5363820921038572,
      "grad_norm": 2.016054630279541,
      "learning_rate": 9.301927194860814e-06,
      "loss": 0.12115071713924408,
      "num_input_tokens_seen": 1303460,
      "step": 1255,
      "train_runtime": 3570.5381,
      "train_tokens_per_second": 365.06
    },
    {
      "epoch": 0.5368094881931831,
      "grad_norm": 2.2618765830993652,
      "learning_rate": 9.29336188436831e-06,
      "loss": 0.10006249696016312,
      "num_input_tokens_seen": 1304480,
      "step": 1256,
      "train_runtime": 3572.7179,
      "train_tokens_per_second": 365.123
    },
    {
      "epoch": 0.5372368842825088,
      "grad_norm": 2.1425459384918213,
      "learning_rate": 9.284796573875805e-06,
      "loss": 0.17706933617591858,
      "num_input_tokens_seen": 1305782,
      "step": 1257,
      "train_runtime": 3574.9477,
      "train_tokens_per_second": 365.259
    },
    {
      "epoch": 0.5376642803718346,
      "grad_norm": 2.3888356685638428,
      "learning_rate": 9.276231263383298e-06,
      "loss": 0.15307390689849854,
      "num_input_tokens_seen": 1306524,
      "step": 1258,
      "train_runtime": 3577.0714,
      "train_tokens_per_second": 365.25
    },
    {
      "epoch": 0.5380916764611604,
      "grad_norm": 1.893916130065918,
      "learning_rate": 9.267665952890794e-06,
      "loss": 0.1650296151638031,
      "num_input_tokens_seen": 1307818,
      "step": 1259,
      "train_runtime": 3579.2733,
      "train_tokens_per_second": 365.386
    },
    {
      "epoch": 0.5385190725504861,
      "grad_norm": 1.372362494468689,
      "learning_rate": 9.259100642398287e-06,
      "loss": 0.11354769766330719,
      "num_input_tokens_seen": 1309754,
      "step": 1260,
      "train_runtime": 3581.6398,
      "train_tokens_per_second": 365.686
    },
    {
      "epoch": 0.538946468639812,
      "grad_norm": 2.5072848796844482,
      "learning_rate": 9.250535331905782e-06,
      "loss": 0.1607133150100708,
      "num_input_tokens_seen": 1310512,
      "step": 1261,
      "train_runtime": 3598.8327,
      "train_tokens_per_second": 364.149
    },
    {
      "epoch": 0.5393738647291377,
      "grad_norm": 2.4146971702575684,
      "learning_rate": 9.241970021413276e-06,
      "loss": 0.1071794256567955,
      "num_input_tokens_seen": 1311126,
      "step": 1262,
      "train_runtime": 3600.8648,
      "train_tokens_per_second": 364.114
    },
    {
      "epoch": 0.5398012608184635,
      "grad_norm": 2.24064564704895,
      "learning_rate": 9.233404710920771e-06,
      "loss": 0.15566112101078033,
      "num_input_tokens_seen": 1312054,
      "step": 1263,
      "train_runtime": 3603.0721,
      "train_tokens_per_second": 364.149
    },
    {
      "epoch": 0.5402286569077893,
      "grad_norm": 2.3328211307525635,
      "learning_rate": 9.224839400428267e-06,
      "loss": 0.13846774399280548,
      "num_input_tokens_seen": 1313044,
      "step": 1264,
      "train_runtime": 3605.2604,
      "train_tokens_per_second": 364.202
    },
    {
      "epoch": 0.540656052997115,
      "grad_norm": 3.052949905395508,
      "learning_rate": 9.21627408993576e-06,
      "loss": 0.12793810665607452,
      "num_input_tokens_seen": 1313798,
      "step": 1265,
      "train_runtime": 3607.3198,
      "train_tokens_per_second": 364.203
    },
    {
      "epoch": 0.5410834490864409,
      "grad_norm": 2.2812297344207764,
      "learning_rate": 9.207708779443255e-06,
      "loss": 0.23001760244369507,
      "num_input_tokens_seen": 1315170,
      "step": 1266,
      "train_runtime": 3609.5676,
      "train_tokens_per_second": 364.357
    },
    {
      "epoch": 0.5415108451757666,
      "grad_norm": 2.086668014526367,
      "learning_rate": 9.199143468950749e-06,
      "loss": 0.10874505341053009,
      "num_input_tokens_seen": 1316050,
      "step": 1267,
      "train_runtime": 3611.7348,
      "train_tokens_per_second": 364.382
    },
    {
      "epoch": 0.5419382412650924,
      "grad_norm": 1.557161808013916,
      "learning_rate": 9.190578158458246e-06,
      "loss": 0.05690646171569824,
      "num_input_tokens_seen": 1317004,
      "step": 1268,
      "train_runtime": 3613.8832,
      "train_tokens_per_second": 364.429
    },
    {
      "epoch": 0.5423656373544182,
      "grad_norm": 3.058598756790161,
      "learning_rate": 9.18201284796574e-06,
      "loss": 0.09717820584774017,
      "num_input_tokens_seen": 1318044,
      "step": 1269,
      "train_runtime": 3616.1513,
      "train_tokens_per_second": 364.488
    },
    {
      "epoch": 0.542793033443744,
      "grad_norm": 2.784557342529297,
      "learning_rate": 9.173447537473235e-06,
      "loss": 0.09876945614814758,
      "num_input_tokens_seen": 1318916,
      "step": 1270,
      "train_runtime": 3618.2986,
      "train_tokens_per_second": 364.513
    },
    {
      "epoch": 0.5432204295330698,
      "grad_norm": 2.4188225269317627,
      "learning_rate": 9.164882226980728e-06,
      "loss": 0.2313305288553238,
      "num_input_tokens_seen": 1320104,
      "step": 1271,
      "train_runtime": 3620.4801,
      "train_tokens_per_second": 364.621
    },
    {
      "epoch": 0.5436478256223956,
      "grad_norm": 2.9752988815307617,
      "learning_rate": 9.156316916488224e-06,
      "loss": 0.11963313817977905,
      "num_input_tokens_seen": 1320730,
      "step": 1272,
      "train_runtime": 3622.5832,
      "train_tokens_per_second": 364.582
    },
    {
      "epoch": 0.5440752217117213,
      "grad_norm": 2.1987855434417725,
      "learning_rate": 9.147751605995719e-06,
      "loss": 0.07606115937232971,
      "num_input_tokens_seen": 1321634,
      "step": 1273,
      "train_runtime": 3624.7428,
      "train_tokens_per_second": 364.615
    },
    {
      "epoch": 0.5445026178010471,
      "grad_norm": 1.9468597173690796,
      "learning_rate": 9.139186295503213e-06,
      "loss": 0.16524280607700348,
      "num_input_tokens_seen": 1322748,
      "step": 1274,
      "train_runtime": 3626.9266,
      "train_tokens_per_second": 364.702
    },
    {
      "epoch": 0.5449300138903729,
      "grad_norm": 2.0133373737335205,
      "learning_rate": 9.130620985010708e-06,
      "loss": 0.16552017629146576,
      "num_input_tokens_seen": 1323836,
      "step": 1275,
      "train_runtime": 3629.1903,
      "train_tokens_per_second": 364.774
    },
    {
      "epoch": 0.5453574099796987,
      "grad_norm": 1.6033636331558228,
      "learning_rate": 9.122055674518201e-06,
      "loss": 0.09933064877986908,
      "num_input_tokens_seen": 1325094,
      "step": 1276,
      "train_runtime": 3631.3798,
      "train_tokens_per_second": 364.901
    },
    {
      "epoch": 0.5457848060690245,
      "grad_norm": 2.1809206008911133,
      "learning_rate": 9.113490364025697e-06,
      "loss": 0.11724792420864105,
      "num_input_tokens_seen": 1326388,
      "step": 1277,
      "train_runtime": 3633.6312,
      "train_tokens_per_second": 365.031
    },
    {
      "epoch": 0.5462122021583502,
      "grad_norm": 2.150744676589966,
      "learning_rate": 9.104925053533192e-06,
      "loss": 0.10335658490657806,
      "num_input_tokens_seen": 1327346,
      "step": 1278,
      "train_runtime": 3635.7752,
      "train_tokens_per_second": 365.079
    },
    {
      "epoch": 0.546639598247676,
      "grad_norm": 2.424863338470459,
      "learning_rate": 9.096359743040686e-06,
      "loss": 0.14130404591560364,
      "num_input_tokens_seen": 1328328,
      "step": 1279,
      "train_runtime": 3637.9363,
      "train_tokens_per_second": 365.132
    },
    {
      "epoch": 0.5470669943370018,
      "grad_norm": 2.664900541305542,
      "learning_rate": 9.08779443254818e-06,
      "loss": 0.12476594746112823,
      "num_input_tokens_seen": 1329264,
      "step": 1280,
      "train_runtime": 3640.0299,
      "train_tokens_per_second": 365.179
    },
    {
      "epoch": 0.5474943904263276,
      "grad_norm": 2.7224602699279785,
      "learning_rate": 9.079229122055674e-06,
      "loss": 0.08061781525611877,
      "num_input_tokens_seen": 1330392,
      "step": 1281,
      "train_runtime": 3642.1885,
      "train_tokens_per_second": 365.273
    },
    {
      "epoch": 0.5479217865156534,
      "grad_norm": 1.72052001953125,
      "learning_rate": 9.07066381156317e-06,
      "loss": 0.09391103684902191,
      "num_input_tokens_seen": 1331500,
      "step": 1282,
      "train_runtime": 3644.3742,
      "train_tokens_per_second": 365.358
    },
    {
      "epoch": 0.5483491826049791,
      "grad_norm": 2.081308364868164,
      "learning_rate": 9.062098501070665e-06,
      "loss": 0.14099529385566711,
      "num_input_tokens_seen": 1332680,
      "step": 1283,
      "train_runtime": 3646.5567,
      "train_tokens_per_second": 365.463
    },
    {
      "epoch": 0.548776578694305,
      "grad_norm": 2.0109410285949707,
      "learning_rate": 9.053533190578159e-06,
      "loss": 0.1453423798084259,
      "num_input_tokens_seen": 1333844,
      "step": 1284,
      "train_runtime": 3648.7453,
      "train_tokens_per_second": 365.562
    },
    {
      "epoch": 0.5492039747836307,
      "grad_norm": 1.8232446908950806,
      "learning_rate": 9.044967880085654e-06,
      "loss": 0.11511348932981491,
      "num_input_tokens_seen": 1335014,
      "step": 1285,
      "train_runtime": 3650.9628,
      "train_tokens_per_second": 365.661
    },
    {
      "epoch": 0.5496313708729565,
      "grad_norm": 1.6541950702667236,
      "learning_rate": 9.036402569593149e-06,
      "loss": 0.11775043606758118,
      "num_input_tokens_seen": 1337410,
      "step": 1286,
      "train_runtime": 3653.445,
      "train_tokens_per_second": 366.068
    },
    {
      "epoch": 0.5500587669622823,
      "grad_norm": 2.5482306480407715,
      "learning_rate": 9.027837259100644e-06,
      "loss": 0.10782432556152344,
      "num_input_tokens_seen": 1338056,
      "step": 1287,
      "train_runtime": 3655.4679,
      "train_tokens_per_second": 366.042
    },
    {
      "epoch": 0.550486163051608,
      "grad_norm": 2.8341853618621826,
      "learning_rate": 9.019271948608138e-06,
      "loss": 0.11016719043254852,
      "num_input_tokens_seen": 1338814,
      "step": 1288,
      "train_runtime": 3657.5837,
      "train_tokens_per_second": 366.038
    },
    {
      "epoch": 0.5509135591409339,
      "grad_norm": 1.4266753196716309,
      "learning_rate": 9.010706638115633e-06,
      "loss": 0.0824209675192833,
      "num_input_tokens_seen": 1340582,
      "step": 1289,
      "train_runtime": 3659.9242,
      "train_tokens_per_second": 366.287
    },
    {
      "epoch": 0.5513409552302596,
      "grad_norm": 2.7410528659820557,
      "learning_rate": 9.002141327623127e-06,
      "loss": 0.06708842515945435,
      "num_input_tokens_seen": 1341116,
      "step": 1290,
      "train_runtime": 3661.9514,
      "train_tokens_per_second": 366.23
    },
    {
      "epoch": 0.5517683513195855,
      "grad_norm": 2.4072697162628174,
      "learning_rate": 8.993576017130622e-06,
      "loss": 0.1515701860189438,
      "num_input_tokens_seen": 1342370,
      "step": 1291,
      "train_runtime": 3676.9693,
      "train_tokens_per_second": 365.075
    },
    {
      "epoch": 0.5521957474089112,
      "grad_norm": 2.7921485900878906,
      "learning_rate": 8.985010706638116e-06,
      "loss": 0.20630823075771332,
      "num_input_tokens_seen": 1343356,
      "step": 1292,
      "train_runtime": 3679.1048,
      "train_tokens_per_second": 365.131
    },
    {
      "epoch": 0.552623143498237,
      "grad_norm": 1.708499550819397,
      "learning_rate": 8.976445396145611e-06,
      "loss": 0.08747657388448715,
      "num_input_tokens_seen": 1344480,
      "step": 1293,
      "train_runtime": 3681.2888,
      "train_tokens_per_second": 365.22
    },
    {
      "epoch": 0.5530505395875628,
      "grad_norm": 1.676386833190918,
      "learning_rate": 8.967880085653106e-06,
      "loss": 0.07825420051813126,
      "num_input_tokens_seen": 1345570,
      "step": 1294,
      "train_runtime": 3683.4792,
      "train_tokens_per_second": 365.299
    },
    {
      "epoch": 0.5534779356768885,
      "grad_norm": 2.337296962738037,
      "learning_rate": 8.9593147751606e-06,
      "loss": 0.10445433855056763,
      "num_input_tokens_seen": 1346498,
      "step": 1295,
      "train_runtime": 3685.6106,
      "train_tokens_per_second": 365.339
    },
    {
      "epoch": 0.5539053317662144,
      "grad_norm": 2.819326162338257,
      "learning_rate": 8.950749464668095e-06,
      "loss": 0.1458410620689392,
      "num_input_tokens_seen": 1347320,
      "step": 1296,
      "train_runtime": 3687.7463,
      "train_tokens_per_second": 365.351
    },
    {
      "epoch": 0.5543327278555401,
      "grad_norm": 2.037196636199951,
      "learning_rate": 8.942184154175589e-06,
      "loss": 0.12195777893066406,
      "num_input_tokens_seen": 1348416,
      "step": 1297,
      "train_runtime": 3689.9397,
      "train_tokens_per_second": 365.43
    },
    {
      "epoch": 0.554760123944866,
      "grad_norm": 2.4796946048736572,
      "learning_rate": 8.933618843683084e-06,
      "loss": 0.09496226906776428,
      "num_input_tokens_seen": 1349308,
      "step": 1298,
      "train_runtime": 3692.0987,
      "train_tokens_per_second": 365.458
    },
    {
      "epoch": 0.5551875200341917,
      "grad_norm": 2.0836925506591797,
      "learning_rate": 8.92505353319058e-06,
      "loss": 0.09805402904748917,
      "num_input_tokens_seen": 1350388,
      "step": 1299,
      "train_runtime": 3694.2489,
      "train_tokens_per_second": 365.538
    },
    {
      "epoch": 0.5556149161235174,
      "grad_norm": 1.8551748991012573,
      "learning_rate": 8.916488222698073e-06,
      "loss": 0.035007379949092865,
      "num_input_tokens_seen": 1350990,
      "step": 1300,
      "train_runtime": 3696.3067,
      "train_tokens_per_second": 365.497
    },
    {
      "epoch": 0.5560423122128433,
      "grad_norm": 3.933436393737793,
      "learning_rate": 8.907922912205568e-06,
      "loss": 0.2065715342760086,
      "num_input_tokens_seen": 1351776,
      "step": 1301,
      "train_runtime": 3698.3742,
      "train_tokens_per_second": 365.505
    },
    {
      "epoch": 0.556469708302169,
      "grad_norm": 5.46291971206665,
      "learning_rate": 8.899357601713062e-06,
      "loss": 0.16936299204826355,
      "num_input_tokens_seen": 1352406,
      "step": 1302,
      "train_runtime": 3700.4447,
      "train_tokens_per_second": 365.471
    },
    {
      "epoch": 0.5568971043914949,
      "grad_norm": 1.7055410146713257,
      "learning_rate": 8.890792291220559e-06,
      "loss": 0.09328168630599976,
      "num_input_tokens_seen": 1353458,
      "step": 1303,
      "train_runtime": 3702.6342,
      "train_tokens_per_second": 365.539
    },
    {
      "epoch": 0.5573245004808206,
      "grad_norm": 2.939397096633911,
      "learning_rate": 8.882226980728052e-06,
      "loss": 0.12232910096645355,
      "num_input_tokens_seen": 1354264,
      "step": 1304,
      "train_runtime": 3704.8735,
      "train_tokens_per_second": 365.536
    },
    {
      "epoch": 0.5577518965701463,
      "grad_norm": 2.024653911590576,
      "learning_rate": 8.873661670235548e-06,
      "loss": 0.14682631194591522,
      "num_input_tokens_seen": 1355288,
      "step": 1305,
      "train_runtime": 3707.0693,
      "train_tokens_per_second": 365.596
    },
    {
      "epoch": 0.5581792926594722,
      "grad_norm": 2.576078414916992,
      "learning_rate": 8.865096359743041e-06,
      "loss": 0.09955693781375885,
      "num_input_tokens_seen": 1356086,
      "step": 1306,
      "train_runtime": 3709.1659,
      "train_tokens_per_second": 365.604
    },
    {
      "epoch": 0.5586066887487979,
      "grad_norm": 1.7677081823349,
      "learning_rate": 8.856531049250536e-06,
      "loss": 0.12258029729127884,
      "num_input_tokens_seen": 1357528,
      "step": 1307,
      "train_runtime": 3711.4244,
      "train_tokens_per_second": 365.77
    },
    {
      "epoch": 0.5590340848381238,
      "grad_norm": 2.3473801612854004,
      "learning_rate": 8.847965738758032e-06,
      "loss": 0.19139671325683594,
      "num_input_tokens_seen": 1358710,
      "step": 1308,
      "train_runtime": 3713.6208,
      "train_tokens_per_second": 365.872
    },
    {
      "epoch": 0.5594614809274495,
      "grad_norm": 3.1198081970214844,
      "learning_rate": 8.839400428265525e-06,
      "loss": 0.13976724445819855,
      "num_input_tokens_seen": 1359730,
      "step": 1309,
      "train_runtime": 3715.7764,
      "train_tokens_per_second": 365.934
    },
    {
      "epoch": 0.5598888770167753,
      "grad_norm": 2.246872901916504,
      "learning_rate": 8.83083511777302e-06,
      "loss": 0.16256733238697052,
      "num_input_tokens_seen": 1360966,
      "step": 1310,
      "train_runtime": 3718.011,
      "train_tokens_per_second": 366.047
    },
    {
      "epoch": 0.5603162731061011,
      "grad_norm": 3.02571439743042,
      "learning_rate": 8.822269807280514e-06,
      "loss": 0.1526753008365631,
      "num_input_tokens_seen": 1361962,
      "step": 1311,
      "train_runtime": 3720.1745,
      "train_tokens_per_second": 366.102
    },
    {
      "epoch": 0.5607436691954268,
      "grad_norm": 2.409893035888672,
      "learning_rate": 8.81370449678801e-06,
      "loss": 0.1327308565378189,
      "num_input_tokens_seen": 1362752,
      "step": 1312,
      "train_runtime": 3722.2888,
      "train_tokens_per_second": 366.106
    },
    {
      "epoch": 0.5611710652847527,
      "grad_norm": 1.9398658275604248,
      "learning_rate": 8.805139186295505e-06,
      "loss": 0.12720715999603271,
      "num_input_tokens_seen": 1363708,
      "step": 1313,
      "train_runtime": 3724.4118,
      "train_tokens_per_second": 366.154
    },
    {
      "epoch": 0.5615984613740784,
      "grad_norm": 1.7961324453353882,
      "learning_rate": 8.796573875802998e-06,
      "loss": 0.10879611223936081,
      "num_input_tokens_seen": 1364932,
      "step": 1314,
      "train_runtime": 3726.6665,
      "train_tokens_per_second": 366.261
    },
    {
      "epoch": 0.5620258574634042,
      "grad_norm": 2.0721046924591064,
      "learning_rate": 8.788008565310493e-06,
      "loss": 0.13518467545509338,
      "num_input_tokens_seen": 1365832,
      "step": 1315,
      "train_runtime": 3728.8708,
      "train_tokens_per_second": 366.286
    },
    {
      "epoch": 0.56245325355273,
      "grad_norm": 1.8857738971710205,
      "learning_rate": 8.779443254817987e-06,
      "loss": 0.10568409413099289,
      "num_input_tokens_seen": 1367160,
      "step": 1316,
      "train_runtime": 3731.0543,
      "train_tokens_per_second": 366.427
    },
    {
      "epoch": 0.5628806496420558,
      "grad_norm": 2.482398271560669,
      "learning_rate": 8.770877944325482e-06,
      "loss": 0.20136266946792603,
      "num_input_tokens_seen": 1368060,
      "step": 1317,
      "train_runtime": 3733.1907,
      "train_tokens_per_second": 366.459
    },
    {
      "epoch": 0.5633080457313816,
      "grad_norm": 2.512507915496826,
      "learning_rate": 8.762312633832976e-06,
      "loss": 0.13439948856830597,
      "num_input_tokens_seen": 1368762,
      "step": 1318,
      "train_runtime": 3735.2814,
      "train_tokens_per_second": 366.441
    },
    {
      "epoch": 0.5637354418207073,
      "grad_norm": 2.088801622390747,
      "learning_rate": 8.753747323340471e-06,
      "loss": 0.1636715829372406,
      "num_input_tokens_seen": 1369696,
      "step": 1319,
      "train_runtime": 3737.4374,
      "train_tokens_per_second": 366.48
    },
    {
      "epoch": 0.5641628379100331,
      "grad_norm": 1.8344321250915527,
      "learning_rate": 8.745182012847966e-06,
      "loss": 0.0872441828250885,
      "num_input_tokens_seen": 1370364,
      "step": 1320,
      "train_runtime": 3739.4905,
      "train_tokens_per_second": 366.457
    },
    {
      "epoch": 0.5645902339993589,
      "grad_norm": 3.4391965866088867,
      "learning_rate": 8.736616702355462e-06,
      "loss": 0.17728997766971588,
      "num_input_tokens_seen": 1371314,
      "step": 1321,
      "train_runtime": 3755.5123,
      "train_tokens_per_second": 365.147
    },
    {
      "epoch": 0.5650176300886847,
      "grad_norm": 2.0546209812164307,
      "learning_rate": 8.728051391862955e-06,
      "loss": 0.12137656658887863,
      "num_input_tokens_seen": 1372200,
      "step": 1322,
      "train_runtime": 3757.5966,
      "train_tokens_per_second": 365.18
    },
    {
      "epoch": 0.5654450261780105,
      "grad_norm": 2.064497470855713,
      "learning_rate": 8.71948608137045e-06,
      "loss": 0.1498129665851593,
      "num_input_tokens_seen": 1373306,
      "step": 1323,
      "train_runtime": 3759.7638,
      "train_tokens_per_second": 365.264
    },
    {
      "epoch": 0.5658724222673363,
      "grad_norm": 2.0337252616882324,
      "learning_rate": 8.710920770877946e-06,
      "loss": 0.10053382813930511,
      "num_input_tokens_seen": 1374362,
      "step": 1324,
      "train_runtime": 3761.9406,
      "train_tokens_per_second": 365.333
    },
    {
      "epoch": 0.566299818356662,
      "grad_norm": 3.0742299556732178,
      "learning_rate": 8.70235546038544e-06,
      "loss": 0.14325988292694092,
      "num_input_tokens_seen": 1375172,
      "step": 1325,
      "train_runtime": 3764.0952,
      "train_tokens_per_second": 365.339
    },
    {
      "epoch": 0.5667272144459878,
      "grad_norm": 2.6519343852996826,
      "learning_rate": 8.693790149892935e-06,
      "loss": 0.14673161506652832,
      "num_input_tokens_seen": 1376126,
      "step": 1326,
      "train_runtime": 3766.2719,
      "train_tokens_per_second": 365.381
    },
    {
      "epoch": 0.5671546105353136,
      "grad_norm": 1.8149315118789673,
      "learning_rate": 8.685224839400428e-06,
      "loss": 0.1161532774567604,
      "num_input_tokens_seen": 1377132,
      "step": 1327,
      "train_runtime": 3768.5352,
      "train_tokens_per_second": 365.429
    },
    {
      "epoch": 0.5675820066246394,
      "grad_norm": 2.0530340671539307,
      "learning_rate": 8.676659528907924e-06,
      "loss": 0.11958092451095581,
      "num_input_tokens_seen": 1378052,
      "step": 1328,
      "train_runtime": 3770.7822,
      "train_tokens_per_second": 365.455
    },
    {
      "epoch": 0.5680094027139652,
      "grad_norm": 1.969596266746521,
      "learning_rate": 8.668094218415419e-06,
      "loss": 0.10424934327602386,
      "num_input_tokens_seen": 1379138,
      "step": 1329,
      "train_runtime": 3772.9874,
      "train_tokens_per_second": 365.529
    },
    {
      "epoch": 0.5684367988032909,
      "grad_norm": 1.5329201221466064,
      "learning_rate": 8.659528907922912e-06,
      "loss": 0.08686806261539459,
      "num_input_tokens_seen": 1380224,
      "step": 1330,
      "train_runtime": 3775.2438,
      "train_tokens_per_second": 365.599
    },
    {
      "epoch": 0.5688641948926167,
      "grad_norm": 1.7759238481521606,
      "learning_rate": 8.650963597430408e-06,
      "loss": 0.07366158813238144,
      "num_input_tokens_seen": 1380956,
      "step": 1331,
      "train_runtime": 3777.3214,
      "train_tokens_per_second": 365.591
    },
    {
      "epoch": 0.5692915909819425,
      "grad_norm": 2.401189088821411,
      "learning_rate": 8.642398286937901e-06,
      "loss": 0.08268105983734131,
      "num_input_tokens_seen": 1381664,
      "step": 1332,
      "train_runtime": 3779.3754,
      "train_tokens_per_second": 365.58
    },
    {
      "epoch": 0.5697189870712683,
      "grad_norm": 2.8488943576812744,
      "learning_rate": 8.633832976445397e-06,
      "loss": 0.16959163546562195,
      "num_input_tokens_seen": 1382478,
      "step": 1333,
      "train_runtime": 3781.4558,
      "train_tokens_per_second": 365.594
    },
    {
      "epoch": 0.5701463831605941,
      "grad_norm": 2.101639986038208,
      "learning_rate": 8.625267665952892e-06,
      "loss": 0.10440195351839066,
      "num_input_tokens_seen": 1383428,
      "step": 1334,
      "train_runtime": 3783.5597,
      "train_tokens_per_second": 365.642
    },
    {
      "epoch": 0.5705737792499198,
      "grad_norm": 2.3530216217041016,
      "learning_rate": 8.616702355460385e-06,
      "loss": 0.12260738015174866,
      "num_input_tokens_seen": 1384562,
      "step": 1335,
      "train_runtime": 3785.7018,
      "train_tokens_per_second": 365.735
    },
    {
      "epoch": 0.5710011753392457,
      "grad_norm": 2.685822010040283,
      "learning_rate": 8.60813704496788e-06,
      "loss": 0.10980217158794403,
      "num_input_tokens_seen": 1385582,
      "step": 1336,
      "train_runtime": 3787.8491,
      "train_tokens_per_second": 365.797
    },
    {
      "epoch": 0.5714285714285714,
      "grad_norm": 1.882284164428711,
      "learning_rate": 8.599571734475376e-06,
      "loss": 0.14082619547843933,
      "num_input_tokens_seen": 1387014,
      "step": 1337,
      "train_runtime": 3790.1305,
      "train_tokens_per_second": 365.954
    },
    {
      "epoch": 0.5718559675178972,
      "grad_norm": 1.9834223985671997,
      "learning_rate": 8.591006423982871e-06,
      "loss": 0.06217557564377785,
      "num_input_tokens_seen": 1387936,
      "step": 1338,
      "train_runtime": 3792.4668,
      "train_tokens_per_second": 365.972
    },
    {
      "epoch": 0.572283363607223,
      "grad_norm": 2.7914531230926514,
      "learning_rate": 8.582441113490365e-06,
      "loss": 0.10092886537313461,
      "num_input_tokens_seen": 1388608,
      "step": 1339,
      "train_runtime": 3794.576,
      "train_tokens_per_second": 365.945
    },
    {
      "epoch": 0.5727107596965487,
      "grad_norm": 2.843449831008911,
      "learning_rate": 8.57387580299786e-06,
      "loss": 0.14647872745990753,
      "num_input_tokens_seen": 1389596,
      "step": 1340,
      "train_runtime": 3796.7518,
      "train_tokens_per_second": 365.996
    },
    {
      "epoch": 0.5731381557858746,
      "grad_norm": 2.0625381469726562,
      "learning_rate": 8.565310492505354e-06,
      "loss": 0.10510212928056717,
      "num_input_tokens_seen": 1391016,
      "step": 1341,
      "train_runtime": 3799.0464,
      "train_tokens_per_second": 366.149
    },
    {
      "epoch": 0.5735655518752003,
      "grad_norm": 2.20666241645813,
      "learning_rate": 8.556745182012849e-06,
      "loss": 0.1052272766828537,
      "num_input_tokens_seen": 1392974,
      "step": 1342,
      "train_runtime": 3801.3918,
      "train_tokens_per_second": 366.438
    },
    {
      "epoch": 0.5739929479645262,
      "grad_norm": 2.268599033355713,
      "learning_rate": 8.548179871520344e-06,
      "loss": 0.1464858204126358,
      "num_input_tokens_seen": 1394476,
      "step": 1343,
      "train_runtime": 3803.6775,
      "train_tokens_per_second": 366.613
    },
    {
      "epoch": 0.5744203440538519,
      "grad_norm": 2.611967086791992,
      "learning_rate": 8.539614561027838e-06,
      "loss": 0.11977731436491013,
      "num_input_tokens_seen": 1395278,
      "step": 1344,
      "train_runtime": 3805.8458,
      "train_tokens_per_second": 366.614
    },
    {
      "epoch": 0.5748477401431777,
      "grad_norm": 2.4675326347351074,
      "learning_rate": 8.531049250535333e-06,
      "loss": 0.15264874696731567,
      "num_input_tokens_seen": 1396106,
      "step": 1345,
      "train_runtime": 3807.9405,
      "train_tokens_per_second": 366.63
    },
    {
      "epoch": 0.5752751362325035,
      "grad_norm": 2.1952896118164062,
      "learning_rate": 8.522483940042827e-06,
      "loss": 0.108975350856781,
      "num_input_tokens_seen": 1396912,
      "step": 1346,
      "train_runtime": 3810.0442,
      "train_tokens_per_second": 366.639
    },
    {
      "epoch": 0.5757025323218292,
      "grad_norm": 2.743725299835205,
      "learning_rate": 8.513918629550322e-06,
      "loss": 0.14834173023700714,
      "num_input_tokens_seen": 1397604,
      "step": 1347,
      "train_runtime": 3812.1519,
      "train_tokens_per_second": 366.618
    },
    {
      "epoch": 0.5761299284111551,
      "grad_norm": 2.344770908355713,
      "learning_rate": 8.505353319057816e-06,
      "loss": 0.1427122950553894,
      "num_input_tokens_seen": 1398700,
      "step": 1348,
      "train_runtime": 3814.3274,
      "train_tokens_per_second": 366.696
    },
    {
      "epoch": 0.5765573245004808,
      "grad_norm": 2.6061782836914062,
      "learning_rate": 8.496788008565311e-06,
      "loss": 0.1493956595659256,
      "num_input_tokens_seen": 1399822,
      "step": 1349,
      "train_runtime": 3816.5332,
      "train_tokens_per_second": 366.778
    },
    {
      "epoch": 0.5769847205898067,
      "grad_norm": 1.9854739904403687,
      "learning_rate": 8.488222698072806e-06,
      "loss": 0.10830888897180557,
      "num_input_tokens_seen": 1401042,
      "step": 1350,
      "train_runtime": 3818.7548,
      "train_tokens_per_second": 366.885
    },
    {
      "epoch": 0.5774121166791324,
      "grad_norm": 2.474076986312866,
      "learning_rate": 8.4796573875803e-06,
      "loss": 0.14037977159023285,
      "num_input_tokens_seen": 1401878,
      "step": 1351,
      "train_runtime": 3834.4673,
      "train_tokens_per_second": 365.599
    },
    {
      "epoch": 0.5778395127684581,
      "grad_norm": 1.5662596225738525,
      "learning_rate": 8.471092077087795e-06,
      "loss": 0.10657188296318054,
      "num_input_tokens_seen": 1402938,
      "step": 1352,
      "train_runtime": 3836.6308,
      "train_tokens_per_second": 365.669
    },
    {
      "epoch": 0.578266908857784,
      "grad_norm": 2.7052066326141357,
      "learning_rate": 8.462526766595289e-06,
      "loss": 0.11081593483686447,
      "num_input_tokens_seen": 1403754,
      "step": 1353,
      "train_runtime": 3838.7539,
      "train_tokens_per_second": 365.68
    },
    {
      "epoch": 0.5786943049471097,
      "grad_norm": 1.8116711378097534,
      "learning_rate": 8.453961456102786e-06,
      "loss": 0.11657560616731644,
      "num_input_tokens_seen": 1405168,
      "step": 1354,
      "train_runtime": 3841.0703,
      "train_tokens_per_second": 365.827
    },
    {
      "epoch": 0.5791217010364356,
      "grad_norm": 1.6957309246063232,
      "learning_rate": 8.445396145610279e-06,
      "loss": 0.14001968502998352,
      "num_input_tokens_seen": 1406528,
      "step": 1355,
      "train_runtime": 3843.3171,
      "train_tokens_per_second": 365.967
    },
    {
      "epoch": 0.5795490971257613,
      "grad_norm": 1.9877060651779175,
      "learning_rate": 8.436830835117774e-06,
      "loss": 0.16928093135356903,
      "num_input_tokens_seen": 1407702,
      "step": 1356,
      "train_runtime": 3845.534,
      "train_tokens_per_second": 366.062
    },
    {
      "epoch": 0.579976493215087,
      "grad_norm": 1.618666648864746,
      "learning_rate": 8.428265524625268e-06,
      "loss": 0.06174435093998909,
      "num_input_tokens_seen": 1409130,
      "step": 1357,
      "train_runtime": 3847.837,
      "train_tokens_per_second": 366.214
    },
    {
      "epoch": 0.5804038893044129,
      "grad_norm": 3.162889242172241,
      "learning_rate": 8.419700214132763e-06,
      "loss": 0.11345825344324112,
      "num_input_tokens_seen": 1410038,
      "step": 1358,
      "train_runtime": 3849.9652,
      "train_tokens_per_second": 366.247
    },
    {
      "epoch": 0.5808312853937386,
      "grad_norm": 1.9817333221435547,
      "learning_rate": 8.411134903640259e-06,
      "loss": 0.1322307139635086,
      "num_input_tokens_seen": 1410926,
      "step": 1359,
      "train_runtime": 3852.0985,
      "train_tokens_per_second": 366.275
    },
    {
      "epoch": 0.5812586814830645,
      "grad_norm": 2.8281631469726562,
      "learning_rate": 8.402569593147752e-06,
      "loss": 0.13017970323562622,
      "num_input_tokens_seen": 1411922,
      "step": 1360,
      "train_runtime": 3854.2663,
      "train_tokens_per_second": 366.327
    },
    {
      "epoch": 0.5816860775723902,
      "grad_norm": 2.161170482635498,
      "learning_rate": 8.394004282655247e-06,
      "loss": 0.13285818696022034,
      "num_input_tokens_seen": 1412952,
      "step": 1361,
      "train_runtime": 3856.3994,
      "train_tokens_per_second": 366.392
    },
    {
      "epoch": 0.582113473661716,
      "grad_norm": 3.507059335708618,
      "learning_rate": 8.385438972162741e-06,
      "loss": 0.17840546369552612,
      "num_input_tokens_seen": 1414016,
      "step": 1362,
      "train_runtime": 3858.5648,
      "train_tokens_per_second": 366.462
    },
    {
      "epoch": 0.5825408697510418,
      "grad_norm": 2.134746789932251,
      "learning_rate": 8.376873661670236e-06,
      "loss": 0.1294865608215332,
      "num_input_tokens_seen": 1415272,
      "step": 1363,
      "train_runtime": 3860.7568,
      "train_tokens_per_second": 366.579
    },
    {
      "epoch": 0.5829682658403675,
      "grad_norm": 1.676302433013916,
      "learning_rate": 8.368308351177732e-06,
      "loss": 0.06714770197868347,
      "num_input_tokens_seen": 1416104,
      "step": 1364,
      "train_runtime": 3862.8638,
      "train_tokens_per_second": 366.594
    },
    {
      "epoch": 0.5833956619296934,
      "grad_norm": 2.5522656440734863,
      "learning_rate": 8.359743040685225e-06,
      "loss": 0.10977812111377716,
      "num_input_tokens_seen": 1417086,
      "step": 1365,
      "train_runtime": 3865.0168,
      "train_tokens_per_second": 366.644
    },
    {
      "epoch": 0.5838230580190191,
      "grad_norm": 2.5480408668518066,
      "learning_rate": 8.35117773019272e-06,
      "loss": 0.1659277379512787,
      "num_input_tokens_seen": 1417778,
      "step": 1366,
      "train_runtime": 3867.0551,
      "train_tokens_per_second": 366.63
    },
    {
      "epoch": 0.5842504541083449,
      "grad_norm": 1.7382158041000366,
      "learning_rate": 8.342612419700214e-06,
      "loss": 0.09447047114372253,
      "num_input_tokens_seen": 1418892,
      "step": 1367,
      "train_runtime": 3869.281,
      "train_tokens_per_second": 366.707
    },
    {
      "epoch": 0.5846778501976707,
      "grad_norm": 1.8504555225372314,
      "learning_rate": 8.33404710920771e-06,
      "loss": 0.08478574454784393,
      "num_input_tokens_seen": 1419800,
      "step": 1368,
      "train_runtime": 3871.4213,
      "train_tokens_per_second": 366.739
    },
    {
      "epoch": 0.5851052462869964,
      "grad_norm": 2.2849228382110596,
      "learning_rate": 8.325481798715203e-06,
      "loss": 0.13287268579006195,
      "num_input_tokens_seen": 1420856,
      "step": 1369,
      "train_runtime": 3873.6403,
      "train_tokens_per_second": 366.801
    },
    {
      "epoch": 0.5855326423763223,
      "grad_norm": 1.7850430011749268,
      "learning_rate": 8.316916488222698e-06,
      "loss": 0.07814532518386841,
      "num_input_tokens_seen": 1421690,
      "step": 1370,
      "train_runtime": 3875.7701,
      "train_tokens_per_second": 366.815
    },
    {
      "epoch": 0.585960038465648,
      "grad_norm": 2.3672802448272705,
      "learning_rate": 8.308351177730193e-06,
      "loss": 0.16036587953567505,
      "num_input_tokens_seen": 1422638,
      "step": 1371,
      "train_runtime": 3877.9338,
      "train_tokens_per_second": 366.855
    },
    {
      "epoch": 0.5863874345549738,
      "grad_norm": 2.6111226081848145,
      "learning_rate": 8.299785867237689e-06,
      "loss": 0.13331586122512817,
      "num_input_tokens_seen": 1423712,
      "step": 1372,
      "train_runtime": 3880.0851,
      "train_tokens_per_second": 366.928
    },
    {
      "epoch": 0.5868148306442996,
      "grad_norm": 1.6689529418945312,
      "learning_rate": 8.291220556745184e-06,
      "loss": 0.09743912518024445,
      "num_input_tokens_seen": 1424810,
      "step": 1373,
      "train_runtime": 3882.2783,
      "train_tokens_per_second": 367.004
    },
    {
      "epoch": 0.5872422267336254,
      "grad_norm": 2.184544324874878,
      "learning_rate": 8.282655246252678e-06,
      "loss": 0.1430669128894806,
      "num_input_tokens_seen": 1425684,
      "step": 1374,
      "train_runtime": 3884.396,
      "train_tokens_per_second": 367.028
    },
    {
      "epoch": 0.5876696228229512,
      "grad_norm": 1.749814510345459,
      "learning_rate": 8.274089935760173e-06,
      "loss": 0.13397037982940674,
      "num_input_tokens_seen": 1427042,
      "step": 1375,
      "train_runtime": 3886.6138,
      "train_tokens_per_second": 367.168
    },
    {
      "epoch": 0.588097018912277,
      "grad_norm": 1.6981734037399292,
      "learning_rate": 8.265524625267666e-06,
      "loss": 0.12178187817335129,
      "num_input_tokens_seen": 1428144,
      "step": 1376,
      "train_runtime": 3888.8652,
      "train_tokens_per_second": 367.239
    },
    {
      "epoch": 0.5885244150016027,
      "grad_norm": 3.6601908206939697,
      "learning_rate": 8.256959314775162e-06,
      "loss": 0.1699635237455368,
      "num_input_tokens_seen": 1428934,
      "step": 1377,
      "train_runtime": 3891.0216,
      "train_tokens_per_second": 367.239
    },
    {
      "epoch": 0.5889518110909285,
      "grad_norm": 1.5891457796096802,
      "learning_rate": 8.248394004282655e-06,
      "loss": 0.10981012135744095,
      "num_input_tokens_seen": 1430552,
      "step": 1378,
      "train_runtime": 3893.3843,
      "train_tokens_per_second": 367.431
    },
    {
      "epoch": 0.5893792071802543,
      "grad_norm": 2.2618274688720703,
      "learning_rate": 8.23982869379015e-06,
      "loss": 0.11818771809339523,
      "num_input_tokens_seen": 1431452,
      "step": 1379,
      "train_runtime": 3895.5958,
      "train_tokens_per_second": 367.454
    },
    {
      "epoch": 0.5898066032695801,
      "grad_norm": 2.764841079711914,
      "learning_rate": 8.231263383297646e-06,
      "loss": 0.0918097198009491,
      "num_input_tokens_seen": 1432088,
      "step": 1380,
      "train_runtime": 3897.6579,
      "train_tokens_per_second": 367.423
    },
    {
      "epoch": 0.5902339993589059,
      "grad_norm": 1.4518064260482788,
      "learning_rate": 8.22269807280514e-06,
      "loss": 0.06419715285301208,
      "num_input_tokens_seen": 1433322,
      "step": 1381,
      "train_runtime": 3913.1128,
      "train_tokens_per_second": 366.287
    },
    {
      "epoch": 0.5906613954482316,
      "grad_norm": 2.7362771034240723,
      "learning_rate": 8.214132762312635e-06,
      "loss": 0.18493907153606415,
      "num_input_tokens_seen": 1434046,
      "step": 1382,
      "train_runtime": 3915.2862,
      "train_tokens_per_second": 366.269
    },
    {
      "epoch": 0.5910887915375574,
      "grad_norm": 2.793471336364746,
      "learning_rate": 8.205567451820128e-06,
      "loss": 0.14909465610980988,
      "num_input_tokens_seen": 1434882,
      "step": 1383,
      "train_runtime": 3917.4063,
      "train_tokens_per_second": 366.284
    },
    {
      "epoch": 0.5915161876268832,
      "grad_norm": 1.6709586381912231,
      "learning_rate": 8.197002141327624e-06,
      "loss": 0.12015888094902039,
      "num_input_tokens_seen": 1436088,
      "step": 1384,
      "train_runtime": 3919.5929,
      "train_tokens_per_second": 366.387
    },
    {
      "epoch": 0.591943583716209,
      "grad_norm": 2.0044360160827637,
      "learning_rate": 8.188436830835119e-06,
      "loss": 0.12347231805324554,
      "num_input_tokens_seen": 1437124,
      "step": 1385,
      "train_runtime": 3921.7429,
      "train_tokens_per_second": 366.45
    },
    {
      "epoch": 0.5923709798055348,
      "grad_norm": 1.6326919794082642,
      "learning_rate": 8.179871520342612e-06,
      "loss": 0.1159253865480423,
      "num_input_tokens_seen": 1438190,
      "step": 1386,
      "train_runtime": 3923.888,
      "train_tokens_per_second": 366.522
    },
    {
      "epoch": 0.5927983758948605,
      "grad_norm": 1.9021408557891846,
      "learning_rate": 8.171306209850108e-06,
      "loss": 0.11006900668144226,
      "num_input_tokens_seen": 1439252,
      "step": 1387,
      "train_runtime": 3926.0303,
      "train_tokens_per_second": 366.592
    },
    {
      "epoch": 0.5932257719841864,
      "grad_norm": 2.646395444869995,
      "learning_rate": 8.162740899357601e-06,
      "loss": 0.101168692111969,
      "num_input_tokens_seen": 1439840,
      "step": 1388,
      "train_runtime": 3928.0153,
      "train_tokens_per_second": 366.557
    },
    {
      "epoch": 0.5936531680735121,
      "grad_norm": 2.4265596866607666,
      "learning_rate": 8.154175588865098e-06,
      "loss": 0.1598915308713913,
      "num_input_tokens_seen": 1440750,
      "step": 1389,
      "train_runtime": 3930.1664,
      "train_tokens_per_second": 366.588
    },
    {
      "epoch": 0.5940805641628379,
      "grad_norm": 1.9568217992782593,
      "learning_rate": 8.145610278372592e-06,
      "loss": 0.11979921162128448,
      "num_input_tokens_seen": 1441662,
      "step": 1390,
      "train_runtime": 3932.3014,
      "train_tokens_per_second": 366.62
    },
    {
      "epoch": 0.5945079602521637,
      "grad_norm": 7.357199192047119,
      "learning_rate": 8.137044967880087e-06,
      "loss": 0.13000361621379852,
      "num_input_tokens_seen": 1442694,
      "step": 1391,
      "train_runtime": 3934.5113,
      "train_tokens_per_second": 366.677
    },
    {
      "epoch": 0.5949353563414895,
      "grad_norm": 1.5940834283828735,
      "learning_rate": 8.12847965738758e-06,
      "loss": 0.0894617959856987,
      "num_input_tokens_seen": 1443642,
      "step": 1392,
      "train_runtime": 3936.8303,
      "train_tokens_per_second": 366.702
    },
    {
      "epoch": 0.5953627524308153,
      "grad_norm": 3.110785484313965,
      "learning_rate": 8.119914346895076e-06,
      "loss": 0.12593917548656464,
      "num_input_tokens_seen": 1444738,
      "step": 1393,
      "train_runtime": 3939.1446,
      "train_tokens_per_second": 366.764
    },
    {
      "epoch": 0.595790148520141,
      "grad_norm": 3.2190911769866943,
      "learning_rate": 8.111349036402571e-06,
      "loss": 0.22693561017513275,
      "num_input_tokens_seen": 1445596,
      "step": 1394,
      "train_runtime": 3941.2728,
      "train_tokens_per_second": 366.784
    },
    {
      "epoch": 0.5962175446094669,
      "grad_norm": 2.578913450241089,
      "learning_rate": 8.102783725910065e-06,
      "loss": 0.15273822844028473,
      "num_input_tokens_seen": 1446378,
      "step": 1395,
      "train_runtime": 3943.3784,
      "train_tokens_per_second": 366.787
    },
    {
      "epoch": 0.5966449406987926,
      "grad_norm": 2.731638193130493,
      "learning_rate": 8.09421841541756e-06,
      "loss": 0.17036044597625732,
      "num_input_tokens_seen": 1447422,
      "step": 1396,
      "train_runtime": 3945.5416,
      "train_tokens_per_second": 366.85
    },
    {
      "epoch": 0.5970723367881184,
      "grad_norm": 2.767289876937866,
      "learning_rate": 8.085653104925054e-06,
      "loss": 0.232232004404068,
      "num_input_tokens_seen": 1449220,
      "step": 1397,
      "train_runtime": 3947.8662,
      "train_tokens_per_second": 367.089
    },
    {
      "epoch": 0.5974997328774442,
      "grad_norm": 2.646841526031494,
      "learning_rate": 8.077087794432549e-06,
      "loss": 0.21355631947517395,
      "num_input_tokens_seen": 1449952,
      "step": 1398,
      "train_runtime": 3949.9539,
      "train_tokens_per_second": 367.081
    },
    {
      "epoch": 0.5979271289667699,
      "grad_norm": 1.6529147624969482,
      "learning_rate": 8.068522483940043e-06,
      "loss": 0.044996753334999084,
      "num_input_tokens_seen": 1450596,
      "step": 1399,
      "train_runtime": 3951.994,
      "train_tokens_per_second": 367.054
    },
    {
      "epoch": 0.5983545250560958,
      "grad_norm": 1.8547234535217285,
      "learning_rate": 8.059957173447538e-06,
      "loss": 0.1041540578007698,
      "num_input_tokens_seen": 1451548,
      "step": 1400,
      "train_runtime": 3954.1423,
      "train_tokens_per_second": 367.096
    },
    {
      "epoch": 0.5987819211454215,
      "grad_norm": 1.8195332288742065,
      "learning_rate": 8.051391862955033e-06,
      "loss": 0.13584136962890625,
      "num_input_tokens_seen": 1452764,
      "step": 1401,
      "train_runtime": 3956.3541,
      "train_tokens_per_second": 367.198
    },
    {
      "epoch": 0.5992093172347474,
      "grad_norm": 1.8398702144622803,
      "learning_rate": 8.042826552462527e-06,
      "loss": 0.1024625152349472,
      "num_input_tokens_seen": 1453610,
      "step": 1402,
      "train_runtime": 3958.4789,
      "train_tokens_per_second": 367.214
    },
    {
      "epoch": 0.5996367133240731,
      "grad_norm": 1.859737753868103,
      "learning_rate": 8.034261241970022e-06,
      "loss": 0.09262381494045258,
      "num_input_tokens_seen": 1454470,
      "step": 1403,
      "train_runtime": 3960.6056,
      "train_tokens_per_second": 367.234
    },
    {
      "epoch": 0.6000641094133988,
      "grad_norm": 1.7873998880386353,
      "learning_rate": 8.025695931477516e-06,
      "loss": 0.11242175102233887,
      "num_input_tokens_seen": 1455998,
      "step": 1404,
      "train_runtime": 3962.9535,
      "train_tokens_per_second": 367.402
    },
    {
      "epoch": 0.6000641094133988,
      "eval_loss": 0.47081810235977173,
      "eval_runtime": 57.7352,
      "eval_samples_per_second": 17.286,
      "eval_steps_per_second": 8.643,
      "num_input_tokens_seen": 1455998,
      "step": 1404
    },
    {
      "epoch": 0.6004915055027247,
      "grad_norm": 2.3104519844055176,
      "learning_rate": 8.01713062098501e-06,
      "loss": 0.15767702460289001,
      "num_input_tokens_seen": 1456830,
      "step": 1405,
      "train_runtime": 4022.7978,
      "train_tokens_per_second": 362.143
    },
    {
      "epoch": 0.6009189015920504,
      "grad_norm": 2.4053452014923096,
      "learning_rate": 8.008565310492506e-06,
      "loss": 0.16222745180130005,
      "num_input_tokens_seen": 1457788,
      "step": 1406,
      "train_runtime": 4024.9364,
      "train_tokens_per_second": 362.189
    },
    {
      "epoch": 0.6013462976813763,
      "grad_norm": 2.3348591327667236,
      "learning_rate": 8.000000000000001e-06,
      "loss": 0.14439409971237183,
      "num_input_tokens_seen": 1458838,
      "step": 1407,
      "train_runtime": 4027.0936,
      "train_tokens_per_second": 362.256
    },
    {
      "epoch": 0.601773693770702,
      "grad_norm": 1.817209243774414,
      "learning_rate": 7.991434689507495e-06,
      "loss": 0.09775623679161072,
      "num_input_tokens_seen": 1459840,
      "step": 1408,
      "train_runtime": 4029.2095,
      "train_tokens_per_second": 362.314
    },
    {
      "epoch": 0.6022010898600277,
      "grad_norm": 2.439809799194336,
      "learning_rate": 7.98286937901499e-06,
      "loss": 0.1740535944700241,
      "num_input_tokens_seen": 1460590,
      "step": 1409,
      "train_runtime": 4031.296,
      "train_tokens_per_second": 362.313
    },
    {
      "epoch": 0.6026284859493536,
      "grad_norm": 1.7355129718780518,
      "learning_rate": 7.974304068522485e-06,
      "loss": 0.13921035826206207,
      "num_input_tokens_seen": 1462106,
      "step": 1410,
      "train_runtime": 4033.576,
      "train_tokens_per_second": 362.484
    },
    {
      "epoch": 0.6030558820386793,
      "grad_norm": 1.418506145477295,
      "learning_rate": 7.965738758029979e-06,
      "loss": 0.09766557067632675,
      "num_input_tokens_seen": 1463456,
      "step": 1411,
      "train_runtime": 4049.8581,
      "train_tokens_per_second": 361.36
    },
    {
      "epoch": 0.6034832781280052,
      "grad_norm": 1.7446205615997314,
      "learning_rate": 7.957173447537474e-06,
      "loss": 0.11944037675857544,
      "num_input_tokens_seen": 1464550,
      "step": 1412,
      "train_runtime": 4052.0314,
      "train_tokens_per_second": 361.436
    },
    {
      "epoch": 0.6039106742173309,
      "grad_norm": 2.5613455772399902,
      "learning_rate": 7.948608137044968e-06,
      "loss": 0.15170155465602875,
      "num_input_tokens_seen": 1465682,
      "step": 1413,
      "train_runtime": 4054.1869,
      "train_tokens_per_second": 361.523
    },
    {
      "epoch": 0.6043380703066566,
      "grad_norm": 1.9283398389816284,
      "learning_rate": 7.940042826552463e-06,
      "loss": 0.15324467420578003,
      "num_input_tokens_seen": 1466848,
      "step": 1414,
      "train_runtime": 4056.3957,
      "train_tokens_per_second": 361.614
    },
    {
      "epoch": 0.6047654663959825,
      "grad_norm": 1.6675468683242798,
      "learning_rate": 7.931477516059958e-06,
      "loss": 0.07195501029491425,
      "num_input_tokens_seen": 1467706,
      "step": 1415,
      "train_runtime": 4058.571,
      "train_tokens_per_second": 361.631
    },
    {
      "epoch": 0.6051928624853082,
      "grad_norm": 2.899714231491089,
      "learning_rate": 7.922912205567452e-06,
      "loss": 0.1979353129863739,
      "num_input_tokens_seen": 1468522,
      "step": 1416,
      "train_runtime": 4060.6543,
      "train_tokens_per_second": 361.647
    },
    {
      "epoch": 0.6056202585746341,
      "grad_norm": 2.040581703186035,
      "learning_rate": 7.914346895074947e-06,
      "loss": 0.10800875723361969,
      "num_input_tokens_seen": 1469476,
      "step": 1417,
      "train_runtime": 4062.8141,
      "train_tokens_per_second": 361.689
    },
    {
      "epoch": 0.6060476546639598,
      "grad_norm": 1.8093817234039307,
      "learning_rate": 7.905781584582441e-06,
      "loss": 0.11833525449037552,
      "num_input_tokens_seen": 1470466,
      "step": 1418,
      "train_runtime": 4064.9473,
      "train_tokens_per_second": 361.743
    },
    {
      "epoch": 0.6064750507532856,
      "grad_norm": 1.8102803230285645,
      "learning_rate": 7.897216274089936e-06,
      "loss": 0.08050483465194702,
      "num_input_tokens_seen": 1471740,
      "step": 1419,
      "train_runtime": 4067.1487,
      "train_tokens_per_second": 361.86
    },
    {
      "epoch": 0.6069024468426114,
      "grad_norm": 1.804361343383789,
      "learning_rate": 7.888650963597431e-06,
      "loss": 0.0876578614115715,
      "num_input_tokens_seen": 1472736,
      "step": 1420,
      "train_runtime": 4069.2982,
      "train_tokens_per_second": 361.914
    },
    {
      "epoch": 0.6073298429319371,
      "grad_norm": 2.8057749271392822,
      "learning_rate": 7.880085653104925e-06,
      "loss": 0.1516537219285965,
      "num_input_tokens_seen": 1473460,
      "step": 1421,
      "train_runtime": 4071.4015,
      "train_tokens_per_second": 361.905
    },
    {
      "epoch": 0.607757239021263,
      "grad_norm": 1.7483267784118652,
      "learning_rate": 7.87152034261242e-06,
      "loss": 0.07873252779245377,
      "num_input_tokens_seen": 1474266,
      "step": 1422,
      "train_runtime": 4073.5196,
      "train_tokens_per_second": 361.915
    },
    {
      "epoch": 0.6081846351105887,
      "grad_norm": 2.199683427810669,
      "learning_rate": 7.862955032119916e-06,
      "loss": 0.12082644551992416,
      "num_input_tokens_seen": 1475044,
      "step": 1423,
      "train_runtime": 4075.6227,
      "train_tokens_per_second": 361.919
    },
    {
      "epoch": 0.6086120311999145,
      "grad_norm": 3.2418212890625,
      "learning_rate": 7.854389721627411e-06,
      "loss": 0.138291597366333,
      "num_input_tokens_seen": 1475826,
      "step": 1424,
      "train_runtime": 4077.7247,
      "train_tokens_per_second": 361.924
    },
    {
      "epoch": 0.6090394272892403,
      "grad_norm": 2.404599189758301,
      "learning_rate": 7.845824411134904e-06,
      "loss": 0.11044611781835556,
      "num_input_tokens_seen": 1476594,
      "step": 1425,
      "train_runtime": 4079.8823,
      "train_tokens_per_second": 361.921
    },
    {
      "epoch": 0.6094668233785661,
      "grad_norm": 2.100550889968872,
      "learning_rate": 7.8372591006424e-06,
      "loss": 0.10274676978588104,
      "num_input_tokens_seen": 1477446,
      "step": 1426,
      "train_runtime": 4082.0051,
      "train_tokens_per_second": 361.941
    },
    {
      "epoch": 0.6098942194678919,
      "grad_norm": 2.7254180908203125,
      "learning_rate": 7.828693790149893e-06,
      "loss": 0.15266837179660797,
      "num_input_tokens_seen": 1478308,
      "step": 1427,
      "train_runtime": 4084.1587,
      "train_tokens_per_second": 361.961
    },
    {
      "epoch": 0.6103216155572176,
      "grad_norm": 2.171696186065674,
      "learning_rate": 7.820128479657389e-06,
      "loss": 0.08129426836967468,
      "num_input_tokens_seen": 1479076,
      "step": 1428,
      "train_runtime": 4086.2884,
      "train_tokens_per_second": 361.961
    },
    {
      "epoch": 0.6107490116465434,
      "grad_norm": 2.0329370498657227,
      "learning_rate": 7.811563169164882e-06,
      "loss": 0.07338335365056992,
      "num_input_tokens_seen": 1479708,
      "step": 1429,
      "train_runtime": 4088.3113,
      "train_tokens_per_second": 361.936
    },
    {
      "epoch": 0.6111764077358692,
      "grad_norm": 2.23579478263855,
      "learning_rate": 7.802997858672377e-06,
      "loss": 0.13259582221508026,
      "num_input_tokens_seen": 1480992,
      "step": 1430,
      "train_runtime": 4090.538,
      "train_tokens_per_second": 362.053
    },
    {
      "epoch": 0.611603803825195,
      "grad_norm": 2.13228440284729,
      "learning_rate": 7.794432548179873e-06,
      "loss": 0.10177969932556152,
      "num_input_tokens_seen": 1482360,
      "step": 1431,
      "train_runtime": 4092.7994,
      "train_tokens_per_second": 362.187
    },
    {
      "epoch": 0.6120311999145208,
      "grad_norm": 1.5976146459579468,
      "learning_rate": 7.785867237687366e-06,
      "loss": 0.0857323408126831,
      "num_input_tokens_seen": 1483492,
      "step": 1432,
      "train_runtime": 4094.9523,
      "train_tokens_per_second": 362.273
    },
    {
      "epoch": 0.6124585960038466,
      "grad_norm": 2.2590928077697754,
      "learning_rate": 7.777301927194862e-06,
      "loss": 0.15083687007427216,
      "num_input_tokens_seen": 1484432,
      "step": 1433,
      "train_runtime": 4097.0793,
      "train_tokens_per_second": 362.315
    },
    {
      "epoch": 0.6128859920931723,
      "grad_norm": 2.2661385536193848,
      "learning_rate": 7.768736616702355e-06,
      "loss": 0.15994848310947418,
      "num_input_tokens_seen": 1485592,
      "step": 1434,
      "train_runtime": 4099.2655,
      "train_tokens_per_second": 362.404
    },
    {
      "epoch": 0.6133133881824981,
      "grad_norm": 2.448089838027954,
      "learning_rate": 7.76017130620985e-06,
      "loss": 0.1597270965576172,
      "num_input_tokens_seen": 1486452,
      "step": 1435,
      "train_runtime": 4101.4177,
      "train_tokens_per_second": 362.424
    },
    {
      "epoch": 0.6137407842718239,
      "grad_norm": 2.755557060241699,
      "learning_rate": 7.751605995717346e-06,
      "loss": 0.15789887309074402,
      "num_input_tokens_seen": 1487382,
      "step": 1436,
      "train_runtime": 4103.57,
      "train_tokens_per_second": 362.46
    },
    {
      "epoch": 0.6141681803611497,
      "grad_norm": 2.5255467891693115,
      "learning_rate": 7.74304068522484e-06,
      "loss": 0.19847720861434937,
      "num_input_tokens_seen": 1488180,
      "step": 1437,
      "train_runtime": 4105.6764,
      "train_tokens_per_second": 362.469
    },
    {
      "epoch": 0.6145955764504755,
      "grad_norm": 1.8054429292678833,
      "learning_rate": 7.734475374732335e-06,
      "loss": 0.08471320569515228,
      "num_input_tokens_seen": 1489138,
      "step": 1438,
      "train_runtime": 4107.8362,
      "train_tokens_per_second": 362.512
    },
    {
      "epoch": 0.6150229725398013,
      "grad_norm": 2.6610374450683594,
      "learning_rate": 7.725910064239828e-06,
      "loss": 0.08155062794685364,
      "num_input_tokens_seen": 1490098,
      "step": 1439,
      "train_runtime": 4110.0394,
      "train_tokens_per_second": 362.551
    },
    {
      "epoch": 0.615450368629127,
      "grad_norm": 1.975503921508789,
      "learning_rate": 7.717344753747325e-06,
      "loss": 0.10573365539312363,
      "num_input_tokens_seen": 1490940,
      "step": 1440,
      "train_runtime": 4112.1313,
      "train_tokens_per_second": 362.571
    },
    {
      "epoch": 0.6158777647184528,
      "grad_norm": 2.788630723953247,
      "learning_rate": 7.708779443254819e-06,
      "loss": 0.06836487352848053,
      "num_input_tokens_seen": 1491526,
      "step": 1441,
      "train_runtime": 4129.3134,
      "train_tokens_per_second": 361.204
    },
    {
      "epoch": 0.6163051608077786,
      "grad_norm": 2.4164459705352783,
      "learning_rate": 7.700214132762314e-06,
      "loss": 0.148130401968956,
      "num_input_tokens_seen": 1492750,
      "step": 1442,
      "train_runtime": 4131.4973,
      "train_tokens_per_second": 361.31
    },
    {
      "epoch": 0.6167325568971044,
      "grad_norm": 2.46335768699646,
      "learning_rate": 7.691648822269808e-06,
      "loss": 0.10412309318780899,
      "num_input_tokens_seen": 1493648,
      "step": 1443,
      "train_runtime": 4133.6533,
      "train_tokens_per_second": 361.338
    },
    {
      "epoch": 0.6171599529864302,
      "grad_norm": 2.335508346557617,
      "learning_rate": 7.683083511777303e-06,
      "loss": 0.0902898907661438,
      "num_input_tokens_seen": 1494386,
      "step": 1444,
      "train_runtime": 4135.83,
      "train_tokens_per_second": 361.327
    },
    {
      "epoch": 0.617587349075756,
      "grad_norm": 2.1662323474884033,
      "learning_rate": 7.674518201284798e-06,
      "loss": 0.12006018310785294,
      "num_input_tokens_seen": 1495750,
      "step": 1445,
      "train_runtime": 4138.0632,
      "train_tokens_per_second": 361.461
    },
    {
      "epoch": 0.6180147451650817,
      "grad_norm": 3.977302074432373,
      "learning_rate": 7.665952890792292e-06,
      "loss": 0.19146135449409485,
      "num_input_tokens_seen": 1496632,
      "step": 1446,
      "train_runtime": 4140.2288,
      "train_tokens_per_second": 361.485
    },
    {
      "epoch": 0.6184421412544076,
      "grad_norm": 2.467905044555664,
      "learning_rate": 7.657387580299787e-06,
      "loss": 0.1189160943031311,
      "num_input_tokens_seen": 1497298,
      "step": 1447,
      "train_runtime": 4142.3264,
      "train_tokens_per_second": 361.463
    },
    {
      "epoch": 0.6188695373437333,
      "grad_norm": 3.2811837196350098,
      "learning_rate": 7.64882226980728e-06,
      "loss": 0.20138442516326904,
      "num_input_tokens_seen": 1497988,
      "step": 1448,
      "train_runtime": 4144.4699,
      "train_tokens_per_second": 361.443
    },
    {
      "epoch": 0.6192969334330591,
      "grad_norm": 1.574859857559204,
      "learning_rate": 7.640256959314776e-06,
      "loss": 0.07242729514837265,
      "num_input_tokens_seen": 1498954,
      "step": 1449,
      "train_runtime": 4146.6329,
      "train_tokens_per_second": 361.487
    },
    {
      "epoch": 0.6197243295223849,
      "grad_norm": 2.481360673904419,
      "learning_rate": 7.631691648822271e-06,
      "loss": 0.09188266843557358,
      "num_input_tokens_seen": 1499586,
      "step": 1450,
      "train_runtime": 4148.6602,
      "train_tokens_per_second": 361.463
    },
    {
      "epoch": 0.6201517256117106,
      "grad_norm": 2.070668935775757,
      "learning_rate": 7.623126338329765e-06,
      "loss": 0.1142788976430893,
      "num_input_tokens_seen": 1500474,
      "step": 1451,
      "train_runtime": 4150.791,
      "train_tokens_per_second": 361.491
    },
    {
      "epoch": 0.6205791217010365,
      "grad_norm": 2.3478009700775146,
      "learning_rate": 7.614561027837259e-06,
      "loss": 0.12936431169509888,
      "num_input_tokens_seen": 1501270,
      "step": 1452,
      "train_runtime": 4152.8455,
      "train_tokens_per_second": 361.504
    },
    {
      "epoch": 0.6210065177903622,
      "grad_norm": 2.4626338481903076,
      "learning_rate": 7.605995717344754e-06,
      "loss": 0.09926608204841614,
      "num_input_tokens_seen": 1502062,
      "step": 1453,
      "train_runtime": 4154.945,
      "train_tokens_per_second": 361.512
    },
    {
      "epoch": 0.621433913879688,
      "grad_norm": 2.2841718196868896,
      "learning_rate": 7.597430406852249e-06,
      "loss": 0.09891198575496674,
      "num_input_tokens_seen": 1502904,
      "step": 1454,
      "train_runtime": 4157.0367,
      "train_tokens_per_second": 361.533
    },
    {
      "epoch": 0.6218613099690138,
      "grad_norm": 3.0229642391204834,
      "learning_rate": 7.588865096359743e-06,
      "loss": 0.11280228197574615,
      "num_input_tokens_seen": 1503628,
      "step": 1455,
      "train_runtime": 4159.1279,
      "train_tokens_per_second": 361.525
    },
    {
      "epoch": 0.6222887060583395,
      "grad_norm": 3.1719369888305664,
      "learning_rate": 7.580299785867238e-06,
      "loss": 0.09361995756626129,
      "num_input_tokens_seen": 1504378,
      "step": 1456,
      "train_runtime": 4161.205,
      "train_tokens_per_second": 361.525
    },
    {
      "epoch": 0.6227161021476654,
      "grad_norm": 2.3258070945739746,
      "learning_rate": 7.571734475374734e-06,
      "loss": 0.14535577595233917,
      "num_input_tokens_seen": 1505316,
      "step": 1457,
      "train_runtime": 4163.3433,
      "train_tokens_per_second": 361.564
    },
    {
      "epoch": 0.6231434982369911,
      "grad_norm": 2.711185932159424,
      "learning_rate": 7.563169164882228e-06,
      "loss": 0.19543111324310303,
      "num_input_tokens_seen": 1506384,
      "step": 1458,
      "train_runtime": 4165.5143,
      "train_tokens_per_second": 361.632
    },
    {
      "epoch": 0.623570894326317,
      "grad_norm": 2.044725179672241,
      "learning_rate": 7.554603854389723e-06,
      "loss": 0.11587262153625488,
      "num_input_tokens_seen": 1507456,
      "step": 1459,
      "train_runtime": 4167.659,
      "train_tokens_per_second": 361.703
    },
    {
      "epoch": 0.6239982904156427,
      "grad_norm": 2.063974380493164,
      "learning_rate": 7.546038543897217e-06,
      "loss": 0.0817880630493164,
      "num_input_tokens_seen": 1508336,
      "step": 1460,
      "train_runtime": 4169.7766,
      "train_tokens_per_second": 361.731
    },
    {
      "epoch": 0.6244256865049684,
      "grad_norm": 2.3997654914855957,
      "learning_rate": 7.5374732334047115e-06,
      "loss": 0.14338666200637817,
      "num_input_tokens_seen": 1509728,
      "step": 1461,
      "train_runtime": 4172.0284,
      "train_tokens_per_second": 361.869
    },
    {
      "epoch": 0.6248530825942943,
      "grad_norm": 1.8984113931655884,
      "learning_rate": 7.528907922912207e-06,
      "loss": 0.09732633829116821,
      "num_input_tokens_seen": 1510484,
      "step": 1462,
      "train_runtime": 4174.1178,
      "train_tokens_per_second": 361.869
    },
    {
      "epoch": 0.62528047868362,
      "grad_norm": 2.3154830932617188,
      "learning_rate": 7.520342612419701e-06,
      "loss": 0.18469257652759552,
      "num_input_tokens_seen": 1511506,
      "step": 1463,
      "train_runtime": 4176.3308,
      "train_tokens_per_second": 361.922
    },
    {
      "epoch": 0.6257078747729459,
      "grad_norm": 2.2479748725891113,
      "learning_rate": 7.511777301927196e-06,
      "loss": 0.10611981153488159,
      "num_input_tokens_seen": 1512420,
      "step": 1464,
      "train_runtime": 4178.4473,
      "train_tokens_per_second": 361.957
    },
    {
      "epoch": 0.6261352708622716,
      "grad_norm": 1.5329359769821167,
      "learning_rate": 7.50321199143469e-06,
      "loss": 0.09093845635652542,
      "num_input_tokens_seen": 1513738,
      "step": 1465,
      "train_runtime": 4180.7018,
      "train_tokens_per_second": 362.077
    },
    {
      "epoch": 0.6265626669515973,
      "grad_norm": 2.9673044681549072,
      "learning_rate": 7.4946466809421845e-06,
      "loss": 0.1207410991191864,
      "num_input_tokens_seen": 1514442,
      "step": 1466,
      "train_runtime": 4182.792,
      "train_tokens_per_second": 362.065
    },
    {
      "epoch": 0.6269900630409232,
      "grad_norm": 3.5616681575775146,
      "learning_rate": 7.486081370449679e-06,
      "loss": 0.0960858017206192,
      "num_input_tokens_seen": 1515064,
      "step": 1467,
      "train_runtime": 4184.8166,
      "train_tokens_per_second": 362.038
    },
    {
      "epoch": 0.6274174591302489,
      "grad_norm": 6.080048561096191,
      "learning_rate": 7.477516059957174e-06,
      "loss": 0.17542141675949097,
      "num_input_tokens_seen": 1516422,
      "step": 1468,
      "train_runtime": 4187.0666,
      "train_tokens_per_second": 362.168
    },
    {
      "epoch": 0.6278448552195748,
      "grad_norm": 2.1161932945251465,
      "learning_rate": 7.468950749464669e-06,
      "loss": 0.12328416109085083,
      "num_input_tokens_seen": 1517430,
      "step": 1469,
      "train_runtime": 4189.2191,
      "train_tokens_per_second": 362.223
    },
    {
      "epoch": 0.6282722513089005,
      "grad_norm": 2.1375732421875,
      "learning_rate": 7.460385438972163e-06,
      "loss": 0.0849141776561737,
      "num_input_tokens_seen": 1518076,
      "step": 1470,
      "train_runtime": 4191.2785,
      "train_tokens_per_second": 362.199
    },
    {
      "epoch": 0.6286996473982263,
      "grad_norm": 2.348376989364624,
      "learning_rate": 7.4518201284796575e-06,
      "loss": 0.10368818044662476,
      "num_input_tokens_seen": 1519230,
      "step": 1471,
      "train_runtime": 4205.6248,
      "train_tokens_per_second": 361.238
    },
    {
      "epoch": 0.6291270434875521,
      "grad_norm": 2.236046552658081,
      "learning_rate": 7.443254817987152e-06,
      "loss": 0.08742250502109528,
      "num_input_tokens_seen": 1519822,
      "step": 1472,
      "train_runtime": 4207.6455,
      "train_tokens_per_second": 361.205
    },
    {
      "epoch": 0.6295544395768778,
      "grad_norm": 1.9247562885284424,
      "learning_rate": 7.434689507494646e-06,
      "loss": 0.05948425829410553,
      "num_input_tokens_seen": 1520542,
      "step": 1473,
      "train_runtime": 4209.7293,
      "train_tokens_per_second": 361.197
    },
    {
      "epoch": 0.6299818356662037,
      "grad_norm": 1.6708866357803345,
      "learning_rate": 7.4261241970021425e-06,
      "loss": 0.08919192850589752,
      "num_input_tokens_seen": 1521506,
      "step": 1474,
      "train_runtime": 4211.8398,
      "train_tokens_per_second": 361.245
    },
    {
      "epoch": 0.6304092317555294,
      "grad_norm": 2.101806163787842,
      "learning_rate": 7.417558886509637e-06,
      "loss": 0.08752292394638062,
      "num_input_tokens_seen": 1522358,
      "step": 1475,
      "train_runtime": 4213.9544,
      "train_tokens_per_second": 361.266
    },
    {
      "epoch": 0.6308366278448552,
      "grad_norm": 2.0418248176574707,
      "learning_rate": 7.408993576017131e-06,
      "loss": 0.11928815394639969,
      "num_input_tokens_seen": 1523224,
      "step": 1476,
      "train_runtime": 4216.0605,
      "train_tokens_per_second": 361.291
    },
    {
      "epoch": 0.631264023934181,
      "grad_norm": 3.2721803188323975,
      "learning_rate": 7.400428265524627e-06,
      "loss": 0.14328506588935852,
      "num_input_tokens_seen": 1524170,
      "step": 1477,
      "train_runtime": 4218.1714,
      "train_tokens_per_second": 361.334
    },
    {
      "epoch": 0.6316914200235068,
      "grad_norm": 2.308032274246216,
      "learning_rate": 7.391862955032121e-06,
      "loss": 0.08133391290903091,
      "num_input_tokens_seen": 1524894,
      "step": 1478,
      "train_runtime": 4220.2493,
      "train_tokens_per_second": 361.328
    },
    {
      "epoch": 0.6321188161128326,
      "grad_norm": 1.9015473127365112,
      "learning_rate": 7.3832976445396155e-06,
      "loss": 0.08786919713020325,
      "num_input_tokens_seen": 1525752,
      "step": 1479,
      "train_runtime": 4222.3635,
      "train_tokens_per_second": 361.35
    },
    {
      "epoch": 0.6325462122021583,
      "grad_norm": 1.8171110153198242,
      "learning_rate": 7.37473233404711e-06,
      "loss": 0.09292969852685928,
      "num_input_tokens_seen": 1526760,
      "step": 1480,
      "train_runtime": 4224.5633,
      "train_tokens_per_second": 361.401
    },
    {
      "epoch": 0.6329736082914841,
      "grad_norm": 2.274385690689087,
      "learning_rate": 7.366167023554604e-06,
      "loss": 0.10129518061876297,
      "num_input_tokens_seen": 1527564,
      "step": 1481,
      "train_runtime": 4226.6747,
      "train_tokens_per_second": 361.41
    },
    {
      "epoch": 0.6334010043808099,
      "grad_norm": 1.4336169958114624,
      "learning_rate": 7.357601713062099e-06,
      "loss": 0.08634652942419052,
      "num_input_tokens_seen": 1528844,
      "step": 1482,
      "train_runtime": 4229.0001,
      "train_tokens_per_second": 361.514
    },
    {
      "epoch": 0.6338284004701357,
      "grad_norm": 1.9699504375457764,
      "learning_rate": 7.349036402569594e-06,
      "loss": 0.12698976695537567,
      "num_input_tokens_seen": 1529794,
      "step": 1483,
      "train_runtime": 4231.1199,
      "train_tokens_per_second": 361.558
    },
    {
      "epoch": 0.6342557965594615,
      "grad_norm": 2.6824519634246826,
      "learning_rate": 7.3404710920770885e-06,
      "loss": 0.11977970600128174,
      "num_input_tokens_seen": 1530606,
      "step": 1484,
      "train_runtime": 4233.2098,
      "train_tokens_per_second": 361.571
    },
    {
      "epoch": 0.6346831926487873,
      "grad_norm": 1.777571678161621,
      "learning_rate": 7.331905781584583e-06,
      "loss": 0.10391326248645782,
      "num_input_tokens_seen": 1531714,
      "step": 1485,
      "train_runtime": 4235.3856,
      "train_tokens_per_second": 361.647
    },
    {
      "epoch": 0.635110588738113,
      "grad_norm": 2.8286495208740234,
      "learning_rate": 7.323340471092077e-06,
      "loss": 0.24658288061618805,
      "num_input_tokens_seen": 1532570,
      "step": 1486,
      "train_runtime": 4237.4597,
      "train_tokens_per_second": 361.672
    },
    {
      "epoch": 0.6355379848274388,
      "grad_norm": 2.292820453643799,
      "learning_rate": 7.314775160599572e-06,
      "loss": 0.15723267197608948,
      "num_input_tokens_seen": 1533636,
      "step": 1487,
      "train_runtime": 4239.5774,
      "train_tokens_per_second": 361.743
    },
    {
      "epoch": 0.6359653809167646,
      "grad_norm": 2.355534315109253,
      "learning_rate": 7.306209850107066e-06,
      "loss": 0.14795181155204773,
      "num_input_tokens_seen": 1534644,
      "step": 1488,
      "train_runtime": 4241.714,
      "train_tokens_per_second": 361.798
    },
    {
      "epoch": 0.6363927770060904,
      "grad_norm": 2.0955100059509277,
      "learning_rate": 7.2976445396145615e-06,
      "loss": 0.08944962173700333,
      "num_input_tokens_seen": 1535460,
      "step": 1489,
      "train_runtime": 4243.7884,
      "train_tokens_per_second": 361.814
    },
    {
      "epoch": 0.6368201730954162,
      "grad_norm": 2.4552884101867676,
      "learning_rate": 7.289079229122056e-06,
      "loss": 0.10170731693506241,
      "num_input_tokens_seen": 1536424,
      "step": 1490,
      "train_runtime": 4245.9291,
      "train_tokens_per_second": 361.858
    },
    {
      "epoch": 0.637247569184742,
      "grad_norm": 1.9159789085388184,
      "learning_rate": 7.28051391862955e-06,
      "loss": 0.16074612736701965,
      "num_input_tokens_seen": 1538244,
      "step": 1491,
      "train_runtime": 4248.2952,
      "train_tokens_per_second": 362.085
    },
    {
      "epoch": 0.6376749652740678,
      "grad_norm": 4.376994609832764,
      "learning_rate": 7.2719486081370465e-06,
      "loss": 0.17210756242275238,
      "num_input_tokens_seen": 1539572,
      "step": 1492,
      "train_runtime": 4250.4874,
      "train_tokens_per_second": 362.211
    },
    {
      "epoch": 0.6381023613633935,
      "grad_norm": 3.9393601417541504,
      "learning_rate": 7.263383297644541e-06,
      "loss": 0.14389149844646454,
      "num_input_tokens_seen": 1540412,
      "step": 1493,
      "train_runtime": 4252.5689,
      "train_tokens_per_second": 362.231
    },
    {
      "epoch": 0.6385297574527193,
      "grad_norm": 3.1448442935943604,
      "learning_rate": 7.254817987152035e-06,
      "loss": 0.10606074333190918,
      "num_input_tokens_seen": 1541012,
      "step": 1494,
      "train_runtime": 4254.5567,
      "train_tokens_per_second": 362.203
    },
    {
      "epoch": 0.6389571535420451,
      "grad_norm": 1.3708915710449219,
      "learning_rate": 7.24625267665953e-06,
      "loss": 0.06667020171880722,
      "num_input_tokens_seen": 1541764,
      "step": 1495,
      "train_runtime": 4256.6475,
      "train_tokens_per_second": 362.201
    },
    {
      "epoch": 0.6393845496313709,
      "grad_norm": 1.4790881872177124,
      "learning_rate": 7.237687366167024e-06,
      "loss": 0.06881444156169891,
      "num_input_tokens_seen": 1543226,
      "step": 1496,
      "train_runtime": 4258.8516,
      "train_tokens_per_second": 362.357
    },
    {
      "epoch": 0.6398119457206967,
      "grad_norm": 2.453960418701172,
      "learning_rate": 7.229122055674519e-06,
      "loss": 0.11519265919923782,
      "num_input_tokens_seen": 1543984,
      "step": 1497,
      "train_runtime": 4260.9601,
      "train_tokens_per_second": 362.356
    },
    {
      "epoch": 0.6402393418100224,
      "grad_norm": 2.6465978622436523,
      "learning_rate": 7.220556745182014e-06,
      "loss": 0.1643434464931488,
      "num_input_tokens_seen": 1545230,
      "step": 1498,
      "train_runtime": 4263.2199,
      "train_tokens_per_second": 362.456
    },
    {
      "epoch": 0.6406667378993482,
      "grad_norm": 1.84076988697052,
      "learning_rate": 7.211991434689508e-06,
      "loss": 0.10529544204473495,
      "num_input_tokens_seen": 1546236,
      "step": 1499,
      "train_runtime": 4265.4066,
      "train_tokens_per_second": 362.506
    },
    {
      "epoch": 0.641094133988674,
      "grad_norm": 2.474524974822998,
      "learning_rate": 7.203426124197003e-06,
      "loss": 0.09121278673410416,
      "num_input_tokens_seen": 1546874,
      "step": 1500,
      "train_runtime": 4267.4657,
      "train_tokens_per_second": 362.481
    },
    {
      "epoch": 0.6415215300779998,
      "grad_norm": 1.5902736186981201,
      "learning_rate": 7.194860813704497e-06,
      "loss": 0.09791211783885956,
      "num_input_tokens_seen": 1547780,
      "step": 1501,
      "train_runtime": 4282.8895,
      "train_tokens_per_second": 361.387
    },
    {
      "epoch": 0.6419489261673256,
      "grad_norm": 1.715878963470459,
      "learning_rate": 7.186295503211992e-06,
      "loss": 0.10918458551168442,
      "num_input_tokens_seen": 1548564,
      "step": 1502,
      "train_runtime": 4285.0524,
      "train_tokens_per_second": 361.387
    },
    {
      "epoch": 0.6423763222566513,
      "grad_norm": 3.815077066421509,
      "learning_rate": 7.177730192719486e-06,
      "loss": 0.1852296143770218,
      "num_input_tokens_seen": 1549224,
      "step": 1503,
      "train_runtime": 4287.1013,
      "train_tokens_per_second": 361.369
    },
    {
      "epoch": 0.6428037183459772,
      "grad_norm": 2.539187431335449,
      "learning_rate": 7.169164882226981e-06,
      "loss": 0.14823120832443237,
      "num_input_tokens_seen": 1550292,
      "step": 1504,
      "train_runtime": 4289.263,
      "train_tokens_per_second": 361.436
    },
    {
      "epoch": 0.6432311144353029,
      "grad_norm": 2.952749013900757,
      "learning_rate": 7.160599571734476e-06,
      "loss": 0.09099730849266052,
      "num_input_tokens_seen": 1551226,
      "step": 1505,
      "train_runtime": 4291.3919,
      "train_tokens_per_second": 361.474
    },
    {
      "epoch": 0.6436585105246287,
      "grad_norm": 2.3050148487091064,
      "learning_rate": 7.15203426124197e-06,
      "loss": 0.13551293313503265,
      "num_input_tokens_seen": 1552042,
      "step": 1506,
      "train_runtime": 4293.5184,
      "train_tokens_per_second": 361.485
    },
    {
      "epoch": 0.6440859066139545,
      "grad_norm": 1.643945574760437,
      "learning_rate": 7.143468950749465e-06,
      "loss": 0.08347903937101364,
      "num_input_tokens_seen": 1553214,
      "step": 1507,
      "train_runtime": 4295.695,
      "train_tokens_per_second": 361.575
    },
    {
      "epoch": 0.6445133027032802,
      "grad_norm": 2.3023900985717773,
      "learning_rate": 7.134903640256959e-06,
      "loss": 0.2171354591846466,
      "num_input_tokens_seen": 1554276,
      "step": 1508,
      "train_runtime": 4297.8435,
      "train_tokens_per_second": 361.641
    },
    {
      "epoch": 0.6449406987926061,
      "grad_norm": 1.9257196187973022,
      "learning_rate": 7.126338329764455e-06,
      "loss": 0.08590184897184372,
      "num_input_tokens_seen": 1554968,
      "step": 1509,
      "train_runtime": 4299.8965,
      "train_tokens_per_second": 361.629
    },
    {
      "epoch": 0.6453680948819318,
      "grad_norm": 2.213167905807495,
      "learning_rate": 7.11777301927195e-06,
      "loss": 0.12136133015155792,
      "num_input_tokens_seen": 1556068,
      "step": 1510,
      "train_runtime": 4302.1098,
      "train_tokens_per_second": 361.699
    },
    {
      "epoch": 0.6457954909712577,
      "grad_norm": 2.678363800048828,
      "learning_rate": 7.109207708779444e-06,
      "loss": 0.18482092022895813,
      "num_input_tokens_seen": 1556930,
      "step": 1511,
      "train_runtime": 4304.2231,
      "train_tokens_per_second": 361.721
    },
    {
      "epoch": 0.6462228870605834,
      "grad_norm": 1.9447524547576904,
      "learning_rate": 7.1006423982869384e-06,
      "loss": 0.10032150149345398,
      "num_input_tokens_seen": 1558272,
      "step": 1512,
      "train_runtime": 4306.5143,
      "train_tokens_per_second": 361.841
    },
    {
      "epoch": 0.6466502831499091,
      "grad_norm": 1.7493921518325806,
      "learning_rate": 7.092077087794434e-06,
      "loss": 0.07592130452394485,
      "num_input_tokens_seen": 1558930,
      "step": 1513,
      "train_runtime": 4308.57,
      "train_tokens_per_second": 361.821
    },
    {
      "epoch": 0.647077679239235,
      "grad_norm": 2.2475225925445557,
      "learning_rate": 7.083511777301928e-06,
      "loss": 0.07531867176294327,
      "num_input_tokens_seen": 1559754,
      "step": 1514,
      "train_runtime": 4310.7047,
      "train_tokens_per_second": 361.833
    },
    {
      "epoch": 0.6475050753285607,
      "grad_norm": 2.898300886154175,
      "learning_rate": 7.074946466809423e-06,
      "loss": 0.16273649036884308,
      "num_input_tokens_seen": 1560548,
      "step": 1515,
      "train_runtime": 4312.7861,
      "train_tokens_per_second": 361.842
    },
    {
      "epoch": 0.6479324714178866,
      "grad_norm": 1.9874814748764038,
      "learning_rate": 7.066381156316917e-06,
      "loss": 0.1185591071844101,
      "num_input_tokens_seen": 1561562,
      "step": 1516,
      "train_runtime": 4314.8982,
      "train_tokens_per_second": 361.9
    },
    {
      "epoch": 0.6483598675072123,
      "grad_norm": 2.628877639770508,
      "learning_rate": 7.0578158458244114e-06,
      "loss": 0.14291977882385254,
      "num_input_tokens_seen": 1562250,
      "step": 1517,
      "train_runtime": 4316.9465,
      "train_tokens_per_second": 361.888
    },
    {
      "epoch": 0.648787263596538,
      "grad_norm": 2.644549608230591,
      "learning_rate": 7.049250535331906e-06,
      "loss": 0.1770598441362381,
      "num_input_tokens_seen": 1563270,
      "step": 1518,
      "train_runtime": 4319.0763,
      "train_tokens_per_second": 361.945
    },
    {
      "epoch": 0.6492146596858639,
      "grad_norm": 2.3253750801086426,
      "learning_rate": 7.040685224839401e-06,
      "loss": 0.12818999588489532,
      "num_input_tokens_seen": 1564114,
      "step": 1519,
      "train_runtime": 4321.312,
      "train_tokens_per_second": 361.953
    },
    {
      "epoch": 0.6496420557751896,
      "grad_norm": 2.7343599796295166,
      "learning_rate": 7.0321199143468956e-06,
      "loss": 0.17155951261520386,
      "num_input_tokens_seen": 1565672,
      "step": 1520,
      "train_runtime": 4323.6379,
      "train_tokens_per_second": 362.119
    },
    {
      "epoch": 0.6500694518645155,
      "grad_norm": 2.1398792266845703,
      "learning_rate": 7.02355460385439e-06,
      "loss": 0.11565501242876053,
      "num_input_tokens_seen": 1566850,
      "step": 1521,
      "train_runtime": 4325.7921,
      "train_tokens_per_second": 362.211
    },
    {
      "epoch": 0.6504968479538412,
      "grad_norm": 3.598750114440918,
      "learning_rate": 7.0149892933618844e-06,
      "loss": 0.18757835030555725,
      "num_input_tokens_seen": 1567950,
      "step": 1522,
      "train_runtime": 4327.9375,
      "train_tokens_per_second": 362.286
    },
    {
      "epoch": 0.650924244043167,
      "grad_norm": 2.7037353515625,
      "learning_rate": 7.006423982869379e-06,
      "loss": 0.11965816468000412,
      "num_input_tokens_seen": 1568690,
      "step": 1523,
      "train_runtime": 4330.0573,
      "train_tokens_per_second": 362.279
    },
    {
      "epoch": 0.6513516401324928,
      "grad_norm": 2.5321145057678223,
      "learning_rate": 6.997858672376874e-06,
      "loss": 0.1891138106584549,
      "num_input_tokens_seen": 1569822,
      "step": 1524,
      "train_runtime": 4332.2039,
      "train_tokens_per_second": 362.361
    },
    {
      "epoch": 0.6517790362218185,
      "grad_norm": 2.8686320781707764,
      "learning_rate": 6.9892933618843686e-06,
      "loss": 0.13947579264640808,
      "num_input_tokens_seen": 1570408,
      "step": 1525,
      "train_runtime": 4334.2539,
      "train_tokens_per_second": 362.325
    },
    {
      "epoch": 0.6522064323111444,
      "grad_norm": 3.0487496852874756,
      "learning_rate": 6.980728051391864e-06,
      "loss": 0.16573135554790497,
      "num_input_tokens_seen": 1571396,
      "step": 1526,
      "train_runtime": 4336.3782,
      "train_tokens_per_second": 362.375
    },
    {
      "epoch": 0.6526338284004701,
      "grad_norm": 2.529710054397583,
      "learning_rate": 6.972162740899358e-06,
      "loss": 0.1373073011636734,
      "num_input_tokens_seen": 1572208,
      "step": 1527,
      "train_runtime": 4338.466,
      "train_tokens_per_second": 362.388
    },
    {
      "epoch": 0.6530612244897959,
      "grad_norm": 1.5418944358825684,
      "learning_rate": 6.9635974304068535e-06,
      "loss": 0.09944860637187958,
      "num_input_tokens_seen": 1573508,
      "step": 1528,
      "train_runtime": 4340.6818,
      "train_tokens_per_second": 362.502
    },
    {
      "epoch": 0.6534886205791217,
      "grad_norm": 1.5925794839859009,
      "learning_rate": 6.955032119914348e-06,
      "loss": 0.09501263499259949,
      "num_input_tokens_seen": 1574602,
      "step": 1529,
      "train_runtime": 4342.8784,
      "train_tokens_per_second": 362.571
    },
    {
      "epoch": 0.6539160166684475,
      "grad_norm": 2.933858633041382,
      "learning_rate": 6.946466809421842e-06,
      "loss": 0.1324387788772583,
      "num_input_tokens_seen": 1575186,
      "step": 1530,
      "train_runtime": 4344.9451,
      "train_tokens_per_second": 362.533
    },
    {
      "epoch": 0.6543434127577733,
      "grad_norm": 1.530726432800293,
      "learning_rate": 6.937901498929337e-06,
      "loss": 0.07870647311210632,
      "num_input_tokens_seen": 1576610,
      "step": 1531,
      "train_runtime": 4358.9778,
      "train_tokens_per_second": 361.693
    },
    {
      "epoch": 0.654770808847099,
      "grad_norm": 2.544651508331299,
      "learning_rate": 6.929336188436831e-06,
      "loss": 0.1344785988330841,
      "num_input_tokens_seen": 1577596,
      "step": 1532,
      "train_runtime": 4361.1074,
      "train_tokens_per_second": 361.742
    },
    {
      "epoch": 0.6551982049364248,
      "grad_norm": 2.3783624172210693,
      "learning_rate": 6.920770877944326e-06,
      "loss": 0.15800741314888,
      "num_input_tokens_seen": 1578504,
      "step": 1533,
      "train_runtime": 4363.2551,
      "train_tokens_per_second": 361.772
    },
    {
      "epoch": 0.6556256010257506,
      "grad_norm": 1.8461830615997314,
      "learning_rate": 6.912205567451821e-06,
      "loss": 0.11445880681276321,
      "num_input_tokens_seen": 1579710,
      "step": 1534,
      "train_runtime": 4365.5469,
      "train_tokens_per_second": 361.858
    },
    {
      "epoch": 0.6560529971150764,
      "grad_norm": 1.6855775117874146,
      "learning_rate": 6.903640256959315e-06,
      "loss": 0.08027687668800354,
      "num_input_tokens_seen": 1580692,
      "step": 1535,
      "train_runtime": 4367.6649,
      "train_tokens_per_second": 361.908
    },
    {
      "epoch": 0.6564803932044022,
      "grad_norm": 2.4377388954162598,
      "learning_rate": 6.89507494646681e-06,
      "loss": 0.1333707869052887,
      "num_input_tokens_seen": 1581530,
      "step": 1536,
      "train_runtime": 4369.746,
      "train_tokens_per_second": 361.927
    },
    {
      "epoch": 0.656907789293728,
      "grad_norm": 1.9163687229156494,
      "learning_rate": 6.886509635974304e-06,
      "loss": 0.1066640168428421,
      "num_input_tokens_seen": 1582596,
      "step": 1537,
      "train_runtime": 4371.9268,
      "train_tokens_per_second": 361.991
    },
    {
      "epoch": 0.6573351853830538,
      "grad_norm": 3.8590710163116455,
      "learning_rate": 6.877944325481799e-06,
      "loss": 0.12365378439426422,
      "num_input_tokens_seen": 1583350,
      "step": 1538,
      "train_runtime": 4374.0286,
      "train_tokens_per_second": 361.989
    },
    {
      "epoch": 0.6577625814723795,
      "grad_norm": 3.143906354904175,
      "learning_rate": 6.869379014989294e-06,
      "loss": 0.19161656498908997,
      "num_input_tokens_seen": 1584178,
      "step": 1539,
      "train_runtime": 4376.1242,
      "train_tokens_per_second": 362.005
    },
    {
      "epoch": 0.6581899775617053,
      "grad_norm": 2.1624810695648193,
      "learning_rate": 6.860813704496788e-06,
      "loss": 0.11744837462902069,
      "num_input_tokens_seen": 1585064,
      "step": 1540,
      "train_runtime": 4378.2523,
      "train_tokens_per_second": 362.031
    },
    {
      "epoch": 0.6586173736510311,
      "grad_norm": 2.6710524559020996,
      "learning_rate": 6.852248394004283e-06,
      "loss": 0.1383361518383026,
      "num_input_tokens_seen": 1585794,
      "step": 1541,
      "train_runtime": 4380.5114,
      "train_tokens_per_second": 362.011
    },
    {
      "epoch": 0.6590447697403569,
      "grad_norm": 4.175921440124512,
      "learning_rate": 6.843683083511777e-06,
      "loss": 0.17084673047065735,
      "num_input_tokens_seen": 1587408,
      "step": 1542,
      "train_runtime": 4382.9256,
      "train_tokens_per_second": 362.18
    },
    {
      "epoch": 0.6594721658296827,
      "grad_norm": 3.461261034011841,
      "learning_rate": 6.835117773019273e-06,
      "loss": 0.15707671642303467,
      "num_input_tokens_seen": 1588142,
      "step": 1543,
      "train_runtime": 4385.1189,
      "train_tokens_per_second": 362.166
    },
    {
      "epoch": 0.6598995619190084,
      "grad_norm": 3.4410555362701416,
      "learning_rate": 6.826552462526768e-06,
      "loss": 0.20985399186611176,
      "num_input_tokens_seen": 1588860,
      "step": 1544,
      "train_runtime": 4387.2377,
      "train_tokens_per_second": 362.155
    },
    {
      "epoch": 0.6603269580083342,
      "grad_norm": 3.796743869781494,
      "learning_rate": 6.817987152034262e-06,
      "loss": 0.12922640144824982,
      "num_input_tokens_seen": 1589590,
      "step": 1545,
      "train_runtime": 4389.3475,
      "train_tokens_per_second": 362.147
    },
    {
      "epoch": 0.66075435409766,
      "grad_norm": 1.6220176219940186,
      "learning_rate": 6.809421841541757e-06,
      "loss": 0.06964301317930222,
      "num_input_tokens_seen": 1590230,
      "step": 1546,
      "train_runtime": 4391.4107,
      "train_tokens_per_second": 362.123
    },
    {
      "epoch": 0.6611817501869858,
      "grad_norm": 2.4121203422546387,
      "learning_rate": 6.800856531049251e-06,
      "loss": 0.08981835842132568,
      "num_input_tokens_seen": 1591156,
      "step": 1547,
      "train_runtime": 4393.5067,
      "train_tokens_per_second": 362.161
    },
    {
      "epoch": 0.6616091462763116,
      "grad_norm": 1.717573881149292,
      "learning_rate": 6.7922912205567455e-06,
      "loss": 0.0663209781050682,
      "num_input_tokens_seen": 1592442,
      "step": 1548,
      "train_runtime": 4395.7361,
      "train_tokens_per_second": 362.27
    },
    {
      "epoch": 0.6620365423656374,
      "grad_norm": 1.7154979705810547,
      "learning_rate": 6.783725910064241e-06,
      "loss": 0.07388980686664581,
      "num_input_tokens_seen": 1593202,
      "step": 1549,
      "train_runtime": 4397.7862,
      "train_tokens_per_second": 362.274
    },
    {
      "epoch": 0.6624639384549631,
      "grad_norm": 1.9425795078277588,
      "learning_rate": 6.775160599571735e-06,
      "loss": 0.15316474437713623,
      "num_input_tokens_seen": 1594400,
      "step": 1550,
      "train_runtime": 4399.9993,
      "train_tokens_per_second": 362.364
    },
    {
      "epoch": 0.662891334544289,
      "grad_norm": 3.2887344360351562,
      "learning_rate": 6.76659528907923e-06,
      "loss": 0.07516457885503769,
      "num_input_tokens_seen": 1595116,
      "step": 1551,
      "train_runtime": 4402.066,
      "train_tokens_per_second": 362.356
    },
    {
      "epoch": 0.6633187306336147,
      "grad_norm": 1.8119767904281616,
      "learning_rate": 6.758029978586724e-06,
      "loss": 0.12956857681274414,
      "num_input_tokens_seen": 1596084,
      "step": 1552,
      "train_runtime": 4404.2145,
      "train_tokens_per_second": 362.399
    },
    {
      "epoch": 0.6637461267229405,
      "grad_norm": 1.6153640747070312,
      "learning_rate": 6.7494646680942185e-06,
      "loss": 0.12192563712596893,
      "num_input_tokens_seen": 1597872,
      "step": 1553,
      "train_runtime": 4406.5618,
      "train_tokens_per_second": 362.612
    },
    {
      "epoch": 0.6641735228122663,
      "grad_norm": 1.9397310018539429,
      "learning_rate": 6.740899357601714e-06,
      "loss": 0.0925486832857132,
      "num_input_tokens_seen": 1598962,
      "step": 1554,
      "train_runtime": 4408.7145,
      "train_tokens_per_second": 362.682
    },
    {
      "epoch": 0.664600918901592,
      "grad_norm": 2.0065720081329346,
      "learning_rate": 6.732334047109208e-06,
      "loss": 0.08990250527858734,
      "num_input_tokens_seen": 1599828,
      "step": 1555,
      "train_runtime": 4410.8485,
      "train_tokens_per_second": 362.703
    },
    {
      "epoch": 0.6650283149909179,
      "grad_norm": 2.2683143615722656,
      "learning_rate": 6.723768736616703e-06,
      "loss": 0.16291862726211548,
      "num_input_tokens_seen": 1600918,
      "step": 1556,
      "train_runtime": 4413.0483,
      "train_tokens_per_second": 362.769
    },
    {
      "epoch": 0.6654557110802436,
      "grad_norm": 2.6723618507385254,
      "learning_rate": 6.715203426124197e-06,
      "loss": 0.1854749172925949,
      "num_input_tokens_seen": 1601938,
      "step": 1557,
      "train_runtime": 4415.2345,
      "train_tokens_per_second": 362.821
    },
    {
      "epoch": 0.6658831071695694,
      "grad_norm": 2.308262348175049,
      "learning_rate": 6.7066381156316915e-06,
      "loss": 0.12215416878461838,
      "num_input_tokens_seen": 1602692,
      "step": 1558,
      "train_runtime": 4417.2919,
      "train_tokens_per_second": 362.822
    },
    {
      "epoch": 0.6663105032588952,
      "grad_norm": 2.0850255489349365,
      "learning_rate": 6.698072805139186e-06,
      "loss": 0.0813949704170227,
      "num_input_tokens_seen": 1603628,
      "step": 1559,
      "train_runtime": 4419.4208,
      "train_tokens_per_second": 362.859
    },
    {
      "epoch": 0.6667378993482209,
      "grad_norm": 1.8532297611236572,
      "learning_rate": 6.689507494646682e-06,
      "loss": 0.1323699951171875,
      "num_input_tokens_seen": 1604734,
      "step": 1560,
      "train_runtime": 4421.6115,
      "train_tokens_per_second": 362.93
    },
    {
      "epoch": 0.6671652954375468,
      "grad_norm": 3.041516065597534,
      "learning_rate": 6.6809421841541765e-06,
      "loss": 0.1537182778120041,
      "num_input_tokens_seen": 1605706,
      "step": 1561,
      "train_runtime": 4435.2906,
      "train_tokens_per_second": 362.029
    },
    {
      "epoch": 0.6675926915268725,
      "grad_norm": 1.949055790901184,
      "learning_rate": 6.672376873661671e-06,
      "loss": 0.12910784780979156,
      "num_input_tokens_seen": 1606732,
      "step": 1562,
      "train_runtime": 4437.5511,
      "train_tokens_per_second": 362.076
    },
    {
      "epoch": 0.6680200876161984,
      "grad_norm": 1.703078031539917,
      "learning_rate": 6.663811563169165e-06,
      "loss": 0.09869863092899323,
      "num_input_tokens_seen": 1607630,
      "step": 1563,
      "train_runtime": 4439.6744,
      "train_tokens_per_second": 362.105
    },
    {
      "epoch": 0.6684474837055241,
      "grad_norm": 1.8958134651184082,
      "learning_rate": 6.655246252676661e-06,
      "loss": 0.08466009050607681,
      "num_input_tokens_seen": 1608904,
      "step": 1564,
      "train_runtime": 4441.9032,
      "train_tokens_per_second": 362.211
    },
    {
      "epoch": 0.6688748797948498,
      "grad_norm": 2.3481266498565674,
      "learning_rate": 6.646680942184155e-06,
      "loss": 0.14592456817626953,
      "num_input_tokens_seen": 1609904,
      "step": 1565,
      "train_runtime": 4444.0591,
      "train_tokens_per_second": 362.26
    },
    {
      "epoch": 0.6693022758841757,
      "grad_norm": 3.0698764324188232,
      "learning_rate": 6.6381156316916495e-06,
      "loss": 0.13361139595508575,
      "num_input_tokens_seen": 1610686,
      "step": 1566,
      "train_runtime": 4446.1524,
      "train_tokens_per_second": 362.265
    },
    {
      "epoch": 0.6697296719735014,
      "grad_norm": 1.7037657499313354,
      "learning_rate": 6.629550321199144e-06,
      "loss": 0.07341272383928299,
      "num_input_tokens_seen": 1611608,
      "step": 1567,
      "train_runtime": 4448.2688,
      "train_tokens_per_second": 362.3
    },
    {
      "epoch": 0.6701570680628273,
      "grad_norm": 1.5428272485733032,
      "learning_rate": 6.620985010706638e-06,
      "loss": 0.06614117324352264,
      "num_input_tokens_seen": 1612824,
      "step": 1568,
      "train_runtime": 4450.5452,
      "train_tokens_per_second": 362.388
    },
    {
      "epoch": 0.670584464152153,
      "grad_norm": 2.226569414138794,
      "learning_rate": 6.612419700214134e-06,
      "loss": 0.11472350358963013,
      "num_input_tokens_seen": 1613618,
      "step": 1569,
      "train_runtime": 4452.6874,
      "train_tokens_per_second": 362.392
    },
    {
      "epoch": 0.6710118602414787,
      "grad_norm": 1.5729639530181885,
      "learning_rate": 6.603854389721628e-06,
      "loss": 0.11021829396486282,
      "num_input_tokens_seen": 1614690,
      "step": 1570,
      "train_runtime": 4454.8553,
      "train_tokens_per_second": 362.456
    },
    {
      "epoch": 0.6714392563308046,
      "grad_norm": 1.9640666246414185,
      "learning_rate": 6.5952890792291225e-06,
      "loss": 0.09737622737884521,
      "num_input_tokens_seen": 1615700,
      "step": 1571,
      "train_runtime": 4457.0223,
      "train_tokens_per_second": 362.507
    },
    {
      "epoch": 0.6718666524201303,
      "grad_norm": 2.142094373703003,
      "learning_rate": 6.586723768736617e-06,
      "loss": 0.13516435027122498,
      "num_input_tokens_seen": 1617158,
      "step": 1572,
      "train_runtime": 4459.3121,
      "train_tokens_per_second": 362.647
    },
    {
      "epoch": 0.6722940485094562,
      "grad_norm": 2.700592279434204,
      "learning_rate": 6.578158458244111e-06,
      "loss": 0.11396747827529907,
      "num_input_tokens_seen": 1617860,
      "step": 1573,
      "train_runtime": 4461.408,
      "train_tokens_per_second": 362.634
    },
    {
      "epoch": 0.6727214445987819,
      "grad_norm": 1.8810936212539673,
      "learning_rate": 6.569593147751606e-06,
      "loss": 0.10968844592571259,
      "num_input_tokens_seen": 1618652,
      "step": 1574,
      "train_runtime": 4463.4806,
      "train_tokens_per_second": 362.643
    },
    {
      "epoch": 0.6731488406881077,
      "grad_norm": 1.9356191158294678,
      "learning_rate": 6.561027837259101e-06,
      "loss": 0.13543295860290527,
      "num_input_tokens_seen": 1619922,
      "step": 1575,
      "train_runtime": 4465.6714,
      "train_tokens_per_second": 362.75
    },
    {
      "epoch": 0.6735762367774335,
      "grad_norm": 2.6811165809631348,
      "learning_rate": 6.5524625267665955e-06,
      "loss": 0.1540440022945404,
      "num_input_tokens_seen": 1620640,
      "step": 1576,
      "train_runtime": 4467.7377,
      "train_tokens_per_second": 362.743
    },
    {
      "epoch": 0.6740036328667592,
      "grad_norm": 2.1355113983154297,
      "learning_rate": 6.543897216274091e-06,
      "loss": 0.1317986100912094,
      "num_input_tokens_seen": 1621670,
      "step": 1577,
      "train_runtime": 4469.8659,
      "train_tokens_per_second": 362.801
    },
    {
      "epoch": 0.6744310289560851,
      "grad_norm": 2.2480995655059814,
      "learning_rate": 6.535331905781585e-06,
      "loss": 0.10376676172018051,
      "num_input_tokens_seen": 1622578,
      "step": 1578,
      "train_runtime": 4471.9927,
      "train_tokens_per_second": 362.831
    },
    {
      "epoch": 0.6748584250454108,
      "grad_norm": 2.7429425716400146,
      "learning_rate": 6.5267665952890805e-06,
      "loss": 0.17670083045959473,
      "num_input_tokens_seen": 1623332,
      "step": 1579,
      "train_runtime": 4474.0908,
      "train_tokens_per_second": 362.829
    },
    {
      "epoch": 0.6752858211347366,
      "grad_norm": 2.2503812313079834,
      "learning_rate": 6.518201284796575e-06,
      "loss": 0.16332024335861206,
      "num_input_tokens_seen": 1624418,
      "step": 1580,
      "train_runtime": 4476.2764,
      "train_tokens_per_second": 362.895
    },
    {
      "epoch": 0.6757132172240624,
      "grad_norm": 2.3562304973602295,
      "learning_rate": 6.509635974304069e-06,
      "loss": 0.08853507041931152,
      "num_input_tokens_seen": 1625200,
      "step": 1581,
      "train_runtime": 4478.3727,
      "train_tokens_per_second": 362.9
    },
    {
      "epoch": 0.6761406133133882,
      "grad_norm": 1.9377342462539673,
      "learning_rate": 6.501070663811564e-06,
      "loss": 0.12204766273498535,
      "num_input_tokens_seen": 1626426,
      "step": 1582,
      "train_runtime": 4480.5978,
      "train_tokens_per_second": 362.993
    },
    {
      "epoch": 0.676568009402714,
      "grad_norm": 2.319868564605713,
      "learning_rate": 6.492505353319058e-06,
      "loss": 0.20289123058319092,
      "num_input_tokens_seen": 1627304,
      "step": 1583,
      "train_runtime": 4482.787,
      "train_tokens_per_second": 363.012
    },
    {
      "epoch": 0.6769954054920397,
      "grad_norm": 2.271876335144043,
      "learning_rate": 6.4839400428265535e-06,
      "loss": 0.12651079893112183,
      "num_input_tokens_seen": 1628384,
      "step": 1584,
      "train_runtime": 4484.9213,
      "train_tokens_per_second": 363.08
    },
    {
      "epoch": 0.6774228015813656,
      "grad_norm": 2.054103374481201,
      "learning_rate": 6.475374732334048e-06,
      "loss": 0.13746072351932526,
      "num_input_tokens_seen": 1629448,
      "step": 1585,
      "train_runtime": 4487.0902,
      "train_tokens_per_second": 363.141
    },
    {
      "epoch": 0.6778501976706913,
      "grad_norm": 2.3925156593322754,
      "learning_rate": 6.466809421841542e-06,
      "loss": 0.1481277346611023,
      "num_input_tokens_seen": 1630412,
      "step": 1586,
      "train_runtime": 4489.1792,
      "train_tokens_per_second": 363.187
    },
    {
      "epoch": 0.6782775937600171,
      "grad_norm": 1.5451360940933228,
      "learning_rate": 6.458244111349037e-06,
      "loss": 0.1025330051779747,
      "num_input_tokens_seen": 1631754,
      "step": 1587,
      "train_runtime": 4491.448,
      "train_tokens_per_second": 363.302
    },
    {
      "epoch": 0.6787049898493429,
      "grad_norm": 1.8741570711135864,
      "learning_rate": 6.449678800856531e-06,
      "loss": 0.07674557715654373,
      "num_input_tokens_seen": 1633172,
      "step": 1588,
      "train_runtime": 4493.7015,
      "train_tokens_per_second": 363.436
    },
    {
      "epoch": 0.6791323859386686,
      "grad_norm": 2.0336880683898926,
      "learning_rate": 6.441113490364026e-06,
      "loss": 0.10372666269540787,
      "num_input_tokens_seen": 1634354,
      "step": 1589,
      "train_runtime": 4495.8855,
      "train_tokens_per_second": 363.522
    },
    {
      "epoch": 0.6795597820279945,
      "grad_norm": 2.966878890991211,
      "learning_rate": 6.432548179871521e-06,
      "loss": 0.14841686189174652,
      "num_input_tokens_seen": 1635210,
      "step": 1590,
      "train_runtime": 4498.0139,
      "train_tokens_per_second": 363.54
    },
    {
      "epoch": 0.6799871781173202,
      "grad_norm": 2.0769901275634766,
      "learning_rate": 6.423982869379015e-06,
      "loss": 0.1410973072052002,
      "num_input_tokens_seen": 1636860,
      "step": 1591,
      "train_runtime": 4512.2062,
      "train_tokens_per_second": 362.763
    },
    {
      "epoch": 0.680414574206646,
      "grad_norm": 2.3814852237701416,
      "learning_rate": 6.41541755888651e-06,
      "loss": 0.13119149208068848,
      "num_input_tokens_seen": 1637730,
      "step": 1592,
      "train_runtime": 4514.325,
      "train_tokens_per_second": 362.785
    },
    {
      "epoch": 0.6808419702959718,
      "grad_norm": 1.9937283992767334,
      "learning_rate": 6.406852248394004e-06,
      "loss": 0.1572273075580597,
      "num_input_tokens_seen": 1639292,
      "step": 1593,
      "train_runtime": 4516.5663,
      "train_tokens_per_second": 362.951
    },
    {
      "epoch": 0.6812693663852976,
      "grad_norm": 1.8679708242416382,
      "learning_rate": 6.398286937901499e-06,
      "loss": 0.09107799828052521,
      "num_input_tokens_seen": 1640188,
      "step": 1594,
      "train_runtime": 4518.6806,
      "train_tokens_per_second": 362.979
    },
    {
      "epoch": 0.6816967624746234,
      "grad_norm": 3.4475085735321045,
      "learning_rate": 6.389721627408995e-06,
      "loss": 0.18616309762001038,
      "num_input_tokens_seen": 1640806,
      "step": 1595,
      "train_runtime": 4520.7349,
      "train_tokens_per_second": 362.951
    },
    {
      "epoch": 0.6821241585639491,
      "grad_norm": 2.3456876277923584,
      "learning_rate": 6.381156316916489e-06,
      "loss": 0.10660184174776077,
      "num_input_tokens_seen": 1642176,
      "step": 1596,
      "train_runtime": 4522.9768,
      "train_tokens_per_second": 363.074
    },
    {
      "epoch": 0.6825515546532749,
      "grad_norm": 2.482680320739746,
      "learning_rate": 6.372591006423984e-06,
      "loss": 0.1661359667778015,
      "num_input_tokens_seen": 1643650,
      "step": 1597,
      "train_runtime": 4525.2515,
      "train_tokens_per_second": 363.217
    },
    {
      "epoch": 0.6829789507426007,
      "grad_norm": 2.421750783920288,
      "learning_rate": 6.364025695931478e-06,
      "loss": 0.10083428025245667,
      "num_input_tokens_seen": 1644390,
      "step": 1598,
      "train_runtime": 4527.3498,
      "train_tokens_per_second": 363.212
    },
    {
      "epoch": 0.6834063468319265,
      "grad_norm": 2.4274449348449707,
      "learning_rate": 6.355460385438973e-06,
      "loss": 0.166713148355484,
      "num_input_tokens_seen": 1645308,
      "step": 1599,
      "train_runtime": 4529.4753,
      "train_tokens_per_second": 363.245
    },
    {
      "epoch": 0.6838337429212523,
      "grad_norm": 2.417405366897583,
      "learning_rate": 6.346895074946468e-06,
      "loss": 0.15359529852867126,
      "num_input_tokens_seen": 1646310,
      "step": 1600,
      "train_runtime": 4531.6045,
      "train_tokens_per_second": 363.295
    },
    {
      "epoch": 0.6842611390105781,
      "grad_norm": 1.6212536096572876,
      "learning_rate": 6.338329764453962e-06,
      "loss": 0.07829247415065765,
      "num_input_tokens_seen": 1647374,
      "step": 1601,
      "train_runtime": 4533.7284,
      "train_tokens_per_second": 363.36
    },
    {
      "epoch": 0.6846885350999038,
      "grad_norm": 2.5603814125061035,
      "learning_rate": 6.3297644539614566e-06,
      "loss": 0.16584928333759308,
      "num_input_tokens_seen": 1648536,
      "step": 1602,
      "train_runtime": 4536.0802,
      "train_tokens_per_second": 363.427
    },
    {
      "epoch": 0.6851159311892296,
      "grad_norm": 2.598259925842285,
      "learning_rate": 6.321199143468951e-06,
      "loss": 0.1613016426563263,
      "num_input_tokens_seen": 1649318,
      "step": 1603,
      "train_runtime": 4538.2047,
      "train_tokens_per_second": 363.43
    },
    {
      "epoch": 0.6855433272785554,
      "grad_norm": 2.0908098220825195,
      "learning_rate": 6.3126338329764454e-06,
      "loss": 0.10887156426906586,
      "num_input_tokens_seen": 1650252,
      "step": 1604,
      "train_runtime": 4540.3561,
      "train_tokens_per_second": 363.463
    },
    {
      "epoch": 0.6859707233678812,
      "grad_norm": 1.267457365989685,
      "learning_rate": 6.304068522483941e-06,
      "loss": 0.06617289036512375,
      "num_input_tokens_seen": 1652300,
      "step": 1605,
      "train_runtime": 4542.7745,
      "train_tokens_per_second": 363.72
    },
    {
      "epoch": 0.686398119457207,
      "grad_norm": 2.587787628173828,
      "learning_rate": 6.295503211991435e-06,
      "loss": 0.14453619718551636,
      "num_input_tokens_seen": 1653100,
      "step": 1606,
      "train_runtime": 4544.8678,
      "train_tokens_per_second": 363.729
    },
    {
      "epoch": 0.6868255155465327,
      "grad_norm": 1.836341381072998,
      "learning_rate": 6.2869379014989296e-06,
      "loss": 0.10662239789962769,
      "num_input_tokens_seen": 1654178,
      "step": 1607,
      "train_runtime": 4547.0188,
      "train_tokens_per_second": 363.794
    },
    {
      "epoch": 0.6872529116358586,
      "grad_norm": 2.1913387775421143,
      "learning_rate": 6.278372591006424e-06,
      "loss": 0.11226958781480789,
      "num_input_tokens_seen": 1655230,
      "step": 1608,
      "train_runtime": 4549.1614,
      "train_tokens_per_second": 363.854
    },
    {
      "epoch": 0.6876803077251843,
      "grad_norm": 3.4004528522491455,
      "learning_rate": 6.2698072805139184e-06,
      "loss": 0.17808711528778076,
      "num_input_tokens_seen": 1657018,
      "step": 1609,
      "train_runtime": 4551.5275,
      "train_tokens_per_second": 364.058
    },
    {
      "epoch": 0.6881077038145101,
      "grad_norm": 2.7976081371307373,
      "learning_rate": 6.261241970021414e-06,
      "loss": 0.2421571910381317,
      "num_input_tokens_seen": 1657826,
      "step": 1610,
      "train_runtime": 4553.5832,
      "train_tokens_per_second": 364.071
    },
    {
      "epoch": 0.6885350999038359,
      "grad_norm": 1.2608722448349,
      "learning_rate": 6.252676659528908e-06,
      "loss": 0.07399376481771469,
      "num_input_tokens_seen": 1659216,
      "step": 1611,
      "train_runtime": 4555.8255,
      "train_tokens_per_second": 364.197
    },
    {
      "epoch": 0.6889624959931616,
      "grad_norm": 2.002980947494507,
      "learning_rate": 6.244111349036403e-06,
      "loss": 0.09594839066267014,
      "num_input_tokens_seen": 1660136,
      "step": 1612,
      "train_runtime": 4557.9616,
      "train_tokens_per_second": 364.228
    },
    {
      "epoch": 0.6893898920824875,
      "grad_norm": 1.9873377084732056,
      "learning_rate": 6.235546038543898e-06,
      "loss": 0.15380653738975525,
      "num_input_tokens_seen": 1661920,
      "step": 1613,
      "train_runtime": 4560.282,
      "train_tokens_per_second": 364.434
    },
    {
      "epoch": 0.6898172881718132,
      "grad_norm": 2.601666212081909,
      "learning_rate": 6.226980728051393e-06,
      "loss": 0.12249216437339783,
      "num_input_tokens_seen": 1662820,
      "step": 1614,
      "train_runtime": 4562.3998,
      "train_tokens_per_second": 364.462
    },
    {
      "epoch": 0.690244684261139,
      "grad_norm": 1.7057254314422607,
      "learning_rate": 6.2184154175588875e-06,
      "loss": 0.1005893349647522,
      "num_input_tokens_seen": 1664108,
      "step": 1615,
      "train_runtime": 4564.6054,
      "train_tokens_per_second": 364.568
    },
    {
      "epoch": 0.6906720803504648,
      "grad_norm": 2.296954393386841,
      "learning_rate": 6.209850107066382e-06,
      "loss": 0.1730366051197052,
      "num_input_tokens_seen": 1665014,
      "step": 1616,
      "train_runtime": 4566.7564,
      "train_tokens_per_second": 364.594
    },
    {
      "epoch": 0.6910994764397905,
      "grad_norm": 1.402052879333496,
      "learning_rate": 6.201284796573876e-06,
      "loss": 0.048176512122154236,
      "num_input_tokens_seen": 1666400,
      "step": 1617,
      "train_runtime": 4569.0038,
      "train_tokens_per_second": 364.718
    },
    {
      "epoch": 0.6915268725291164,
      "grad_norm": 1.855002760887146,
      "learning_rate": 6.192719486081371e-06,
      "loss": 0.1155838593840599,
      "num_input_tokens_seen": 1667142,
      "step": 1618,
      "train_runtime": 4571.0907,
      "train_tokens_per_second": 364.714
    },
    {
      "epoch": 0.6919542686184421,
      "grad_norm": 1.4894779920578003,
      "learning_rate": 6.184154175588865e-06,
      "loss": 0.07965515553951263,
      "num_input_tokens_seen": 1668150,
      "step": 1619,
      "train_runtime": 4573.2329,
      "train_tokens_per_second": 364.764
    },
    {
      "epoch": 0.692381664707768,
      "grad_norm": 2.2731051445007324,
      "learning_rate": 6.1755888650963605e-06,
      "loss": 0.12008822709321976,
      "num_input_tokens_seen": 1669652,
      "step": 1620,
      "train_runtime": 4575.5378,
      "train_tokens_per_second": 364.908
    },
    {
      "epoch": 0.6928090607970937,
      "grad_norm": 2.585569381713867,
      "learning_rate": 6.167023554603855e-06,
      "loss": 0.09032248705625534,
      "num_input_tokens_seen": 1670674,
      "step": 1621,
      "train_runtime": 4590.5761,
      "train_tokens_per_second": 363.936
    },
    {
      "epoch": 0.6932364568864194,
      "grad_norm": 2.401552200317383,
      "learning_rate": 6.158458244111349e-06,
      "loss": 0.12617796659469604,
      "num_input_tokens_seen": 1671534,
      "step": 1622,
      "train_runtime": 4592.6564,
      "train_tokens_per_second": 363.958
    },
    {
      "epoch": 0.6936638529757453,
      "grad_norm": 3.219254493713379,
      "learning_rate": 6.149892933618844e-06,
      "loss": 0.23800092935562134,
      "num_input_tokens_seen": 1672412,
      "step": 1623,
      "train_runtime": 4594.8153,
      "train_tokens_per_second": 363.978
    },
    {
      "epoch": 0.694091249065071,
      "grad_norm": 2.239248037338257,
      "learning_rate": 6.141327623126338e-06,
      "loss": 0.4660128057003021,
      "num_input_tokens_seen": 1673666,
      "step": 1624,
      "train_runtime": 4597.0203,
      "train_tokens_per_second": 364.076
    },
    {
      "epoch": 0.6945186451543969,
      "grad_norm": 2.1631221771240234,
      "learning_rate": 6.1327623126338335e-06,
      "loss": 0.10860832780599594,
      "num_input_tokens_seen": 1674652,
      "step": 1625,
      "train_runtime": 4599.2076,
      "train_tokens_per_second": 364.118
    },
    {
      "epoch": 0.6949460412437226,
      "grad_norm": 1.7249270677566528,
      "learning_rate": 6.124197002141328e-06,
      "loss": 0.09011699259281158,
      "num_input_tokens_seen": 1675698,
      "step": 1626,
      "train_runtime": 4601.4433,
      "train_tokens_per_second": 364.168
    },
    {
      "epoch": 0.6953734373330483,
      "grad_norm": 2.450115203857422,
      "learning_rate": 6.115631691648822e-06,
      "loss": 0.09784139692783356,
      "num_input_tokens_seen": 1676306,
      "step": 1627,
      "train_runtime": 4603.5993,
      "train_tokens_per_second": 364.129
    },
    {
      "epoch": 0.6958008334223742,
      "grad_norm": 1.5554367303848267,
      "learning_rate": 6.107066381156317e-06,
      "loss": 0.10767507553100586,
      "num_input_tokens_seen": 1677460,
      "step": 1628,
      "train_runtime": 4605.8078,
      "train_tokens_per_second": 364.205
    },
    {
      "epoch": 0.6962282295116999,
      "grad_norm": 2.50962495803833,
      "learning_rate": 6.098501070663813e-06,
      "loss": 0.08403193950653076,
      "num_input_tokens_seen": 1678144,
      "step": 1629,
      "train_runtime": 4607.9012,
      "train_tokens_per_second": 364.188
    },
    {
      "epoch": 0.6966556256010258,
      "grad_norm": 1.8076614141464233,
      "learning_rate": 6.089935760171307e-06,
      "loss": 0.11268124729394913,
      "num_input_tokens_seen": 1679080,
      "step": 1630,
      "train_runtime": 4610.061,
      "train_tokens_per_second": 364.221
    },
    {
      "epoch": 0.6970830216903515,
      "grad_norm": 1.6254327297210693,
      "learning_rate": 6.081370449678802e-06,
      "loss": 0.06718803942203522,
      "num_input_tokens_seen": 1680132,
      "step": 1631,
      "train_runtime": 4612.2188,
      "train_tokens_per_second": 364.278
    },
    {
      "epoch": 0.6975104177796773,
      "grad_norm": 2.299278497695923,
      "learning_rate": 6.072805139186296e-06,
      "loss": 0.08127452433109283,
      "num_input_tokens_seen": 1680726,
      "step": 1632,
      "train_runtime": 4614.3178,
      "train_tokens_per_second": 364.241
    },
    {
      "epoch": 0.6979378138690031,
      "grad_norm": 3.201241970062256,
      "learning_rate": 6.064239828693791e-06,
      "loss": 0.20240415632724762,
      "num_input_tokens_seen": 1681756,
      "step": 1633,
      "train_runtime": 4616.4696,
      "train_tokens_per_second": 364.295
    },
    {
      "epoch": 0.6983652099583288,
      "grad_norm": 2.2114617824554443,
      "learning_rate": 6.055674518201285e-06,
      "loss": 0.09983129799365997,
      "num_input_tokens_seen": 1682650,
      "step": 1634,
      "train_runtime": 4618.5649,
      "train_tokens_per_second": 364.323
    },
    {
      "epoch": 0.6987926060476547,
      "grad_norm": 2.3187384605407715,
      "learning_rate": 6.04710920770878e-06,
      "loss": 0.1907171756029129,
      "num_input_tokens_seen": 1684078,
      "step": 1635,
      "train_runtime": 4620.8028,
      "train_tokens_per_second": 364.456
    },
    {
      "epoch": 0.6992200021369804,
      "grad_norm": 2.008796453475952,
      "learning_rate": 6.038543897216275e-06,
      "loss": 0.13186600804328918,
      "num_input_tokens_seen": 1685224,
      "step": 1636,
      "train_runtime": 4623.0173,
      "train_tokens_per_second": 364.529
    },
    {
      "epoch": 0.6996473982263063,
      "grad_norm": 2.5543391704559326,
      "learning_rate": 6.029978586723769e-06,
      "loss": 0.15313224494457245,
      "num_input_tokens_seen": 1686050,
      "step": 1637,
      "train_runtime": 4625.1428,
      "train_tokens_per_second": 364.54
    },
    {
      "epoch": 0.700074794315632,
      "grad_norm": 2.069614887237549,
      "learning_rate": 6.021413276231264e-06,
      "loss": 0.07094985246658325,
      "num_input_tokens_seen": 1686956,
      "step": 1638,
      "train_runtime": 4627.2904,
      "train_tokens_per_second": 364.567
    },
    {
      "epoch": 0.700074794315632,
      "eval_loss": 0.4644208252429962,
      "eval_runtime": 57.7308,
      "eval_samples_per_second": 17.287,
      "eval_steps_per_second": 8.644,
      "num_input_tokens_seen": 1686956,
      "step": 1638
    },
    {
      "epoch": 0.7005021904049578,
      "grad_norm": 3.219636917114258,
      "learning_rate": 6.012847965738758e-06,
      "loss": 0.13574418425559998,
      "num_input_tokens_seen": 1687798,
      "step": 1639,
      "train_runtime": 4687.1407,
      "train_tokens_per_second": 360.091
    },
    {
      "epoch": 0.7009295864942836,
      "grad_norm": 1.5980966091156006,
      "learning_rate": 6.004282655246253e-06,
      "loss": 0.08837272226810455,
      "num_input_tokens_seen": 1689004,
      "step": 1640,
      "train_runtime": 4689.3809,
      "train_tokens_per_second": 360.176
    },
    {
      "epoch": 0.7013569825836093,
      "grad_norm": 2.2951228618621826,
      "learning_rate": 5.995717344753748e-06,
      "loss": 0.0942828357219696,
      "num_input_tokens_seen": 1689856,
      "step": 1641,
      "train_runtime": 4691.5124,
      "train_tokens_per_second": 360.194
    },
    {
      "epoch": 0.7017843786729352,
      "grad_norm": 1.5538647174835205,
      "learning_rate": 5.987152034261242e-06,
      "loss": 0.05763139948248863,
      "num_input_tokens_seen": 1690670,
      "step": 1642,
      "train_runtime": 4693.5971,
      "train_tokens_per_second": 360.208
    },
    {
      "epoch": 0.7022117747622609,
      "grad_norm": 2.632305860519409,
      "learning_rate": 5.978586723768737e-06,
      "loss": 0.12397623062133789,
      "num_input_tokens_seen": 1691480,
      "step": 1643,
      "train_runtime": 4695.6853,
      "train_tokens_per_second": 360.22
    },
    {
      "epoch": 0.7026391708515867,
      "grad_norm": 2.3829047679901123,
      "learning_rate": 5.970021413276231e-06,
      "loss": 0.08417462557554245,
      "num_input_tokens_seen": 1692250,
      "step": 1644,
      "train_runtime": 4697.7737,
      "train_tokens_per_second": 360.224
    },
    {
      "epoch": 0.7030665669409125,
      "grad_norm": 2.198765516281128,
      "learning_rate": 5.9614561027837255e-06,
      "loss": 0.168161541223526,
      "num_input_tokens_seen": 1693420,
      "step": 1645,
      "train_runtime": 4699.9956,
      "train_tokens_per_second": 360.302
    },
    {
      "epoch": 0.7034939630302383,
      "grad_norm": 3.7647759914398193,
      "learning_rate": 5.952890792291222e-06,
      "loss": 0.18240110576152802,
      "num_input_tokens_seen": 1694210,
      "step": 1646,
      "train_runtime": 4702.0905,
      "train_tokens_per_second": 360.31
    },
    {
      "epoch": 0.7039213591195641,
      "grad_norm": 2.1652207374572754,
      "learning_rate": 5.944325481798716e-06,
      "loss": 0.10134699195623398,
      "num_input_tokens_seen": 1695012,
      "step": 1647,
      "train_runtime": 4704.2134,
      "train_tokens_per_second": 360.318
    },
    {
      "epoch": 0.7043487552088898,
      "grad_norm": 2.243745803833008,
      "learning_rate": 5.9357601713062105e-06,
      "loss": 0.1376046985387802,
      "num_input_tokens_seen": 1696218,
      "step": 1648,
      "train_runtime": 4706.4299,
      "train_tokens_per_second": 360.404
    },
    {
      "epoch": 0.7047761512982156,
      "grad_norm": 1.9967803955078125,
      "learning_rate": 5.927194860813705e-06,
      "loss": 0.10707314312458038,
      "num_input_tokens_seen": 1697258,
      "step": 1649,
      "train_runtime": 4708.6117,
      "train_tokens_per_second": 360.458
    },
    {
      "epoch": 0.7052035473875414,
      "grad_norm": 2.091332197189331,
      "learning_rate": 5.9186295503212e-06,
      "loss": 0.08665335923433304,
      "num_input_tokens_seen": 1698094,
      "step": 1650,
      "train_runtime": 4710.7272,
      "train_tokens_per_second": 360.474
    },
    {
      "epoch": 0.7056309434768672,
      "grad_norm": 2.8967652320861816,
      "learning_rate": 5.910064239828695e-06,
      "loss": 0.10194618254899979,
      "num_input_tokens_seen": 1698594,
      "step": 1651,
      "train_runtime": 4728.1737,
      "train_tokens_per_second": 359.249
    },
    {
      "epoch": 0.706058339566193,
      "grad_norm": 2.6683316230773926,
      "learning_rate": 5.901498929336189e-06,
      "loss": 0.16606134176254272,
      "num_input_tokens_seen": 1699658,
      "step": 1652,
      "train_runtime": 4730.3307,
      "train_tokens_per_second": 359.311
    },
    {
      "epoch": 0.7064857356555188,
      "grad_norm": 1.8908686637878418,
      "learning_rate": 5.8929336188436835e-06,
      "loss": 0.11453081667423248,
      "num_input_tokens_seen": 1700770,
      "step": 1653,
      "train_runtime": 4732.5202,
      "train_tokens_per_second": 359.379
    },
    {
      "epoch": 0.7069131317448445,
      "grad_norm": 2.30198335647583,
      "learning_rate": 5.884368308351178e-06,
      "loss": 0.14984571933746338,
      "num_input_tokens_seen": 1701816,
      "step": 1654,
      "train_runtime": 4734.7224,
      "train_tokens_per_second": 359.433
    },
    {
      "epoch": 0.7073405278341703,
      "grad_norm": 2.539581060409546,
      "learning_rate": 5.875802997858673e-06,
      "loss": 0.17967048287391663,
      "num_input_tokens_seen": 1703090,
      "step": 1655,
      "train_runtime": 4736.9741,
      "train_tokens_per_second": 359.531
    },
    {
      "epoch": 0.7077679239234961,
      "grad_norm": 1.6067545413970947,
      "learning_rate": 5.867237687366168e-06,
      "loss": 0.0922904685139656,
      "num_input_tokens_seen": 1704284,
      "step": 1656,
      "train_runtime": 4739.1765,
      "train_tokens_per_second": 359.616
    },
    {
      "epoch": 0.7081953200128219,
      "grad_norm": 2.4805846214294434,
      "learning_rate": 5.858672376873662e-06,
      "loss": 0.15101802349090576,
      "num_input_tokens_seen": 1705218,
      "step": 1657,
      "train_runtime": 4741.2998,
      "train_tokens_per_second": 359.652
    },
    {
      "epoch": 0.7086227161021477,
      "grad_norm": 2.954930305480957,
      "learning_rate": 5.8501070663811565e-06,
      "loss": 0.14685797691345215,
      "num_input_tokens_seen": 1706048,
      "step": 1658,
      "train_runtime": 4743.4298,
      "train_tokens_per_second": 359.665
    },
    {
      "epoch": 0.7090501121914734,
      "grad_norm": 1.272079586982727,
      "learning_rate": 5.841541755888651e-06,
      "loss": 0.06490632891654968,
      "num_input_tokens_seen": 1707348,
      "step": 1659,
      "train_runtime": 4745.6461,
      "train_tokens_per_second": 359.771
    },
    {
      "epoch": 0.7094775082807993,
      "grad_norm": 2.1344242095947266,
      "learning_rate": 5.832976445396145e-06,
      "loss": 0.11305753886699677,
      "num_input_tokens_seen": 1708384,
      "step": 1660,
      "train_runtime": 4747.7648,
      "train_tokens_per_second": 359.829
    },
    {
      "epoch": 0.709904904370125,
      "grad_norm": 1.6343703269958496,
      "learning_rate": 5.824411134903641e-06,
      "loss": 0.0764722228050232,
      "num_input_tokens_seen": 1709360,
      "step": 1661,
      "train_runtime": 4749.8893,
      "train_tokens_per_second": 359.874
    },
    {
      "epoch": 0.7103323004594508,
      "grad_norm": 2.52177357673645,
      "learning_rate": 5.815845824411135e-06,
      "loss": 0.12074946612119675,
      "num_input_tokens_seen": 1710508,
      "step": 1662,
      "train_runtime": 4752.0773,
      "train_tokens_per_second": 359.95
    },
    {
      "epoch": 0.7107596965487766,
      "grad_norm": 2.8598716259002686,
      "learning_rate": 5.80728051391863e-06,
      "loss": 0.19634686410427094,
      "num_input_tokens_seen": 1712138,
      "step": 1663,
      "train_runtime": 4754.3561,
      "train_tokens_per_second": 360.12
    },
    {
      "epoch": 0.7111870926381023,
      "grad_norm": 2.793030023574829,
      "learning_rate": 5.798715203426125e-06,
      "loss": 0.18983489274978638,
      "num_input_tokens_seen": 1712860,
      "step": 1664,
      "train_runtime": 4756.4394,
      "train_tokens_per_second": 360.114
    },
    {
      "epoch": 0.7116144887274282,
      "grad_norm": 2.6077451705932617,
      "learning_rate": 5.79014989293362e-06,
      "loss": 0.10160862654447556,
      "num_input_tokens_seen": 1713704,
      "step": 1665,
      "train_runtime": 4758.5798,
      "train_tokens_per_second": 360.129
    },
    {
      "epoch": 0.7120418848167539,
      "grad_norm": 2.501498222351074,
      "learning_rate": 5.7815845824411145e-06,
      "loss": 0.15047523379325867,
      "num_input_tokens_seen": 1714516,
      "step": 1666,
      "train_runtime": 4760.7095,
      "train_tokens_per_second": 360.139
    },
    {
      "epoch": 0.7124692809060797,
      "grad_norm": 1.5240485668182373,
      "learning_rate": 5.773019271948609e-06,
      "loss": 0.1040622889995575,
      "num_input_tokens_seen": 1715660,
      "step": 1667,
      "train_runtime": 4762.8688,
      "train_tokens_per_second": 360.216
    },
    {
      "epoch": 0.7128966769954055,
      "grad_norm": 2.8323254585266113,
      "learning_rate": 5.764453961456103e-06,
      "loss": 0.1454036682844162,
      "num_input_tokens_seen": 1716730,
      "step": 1668,
      "train_runtime": 4765.0685,
      "train_tokens_per_second": 360.274
    },
    {
      "epoch": 0.7133240730847312,
      "grad_norm": 2.108504056930542,
      "learning_rate": 5.755888650963598e-06,
      "loss": 0.10528227686882019,
      "num_input_tokens_seen": 1717640,
      "step": 1669,
      "train_runtime": 4767.1967,
      "train_tokens_per_second": 360.304
    },
    {
      "epoch": 0.7137514691740571,
      "grad_norm": 3.772016763687134,
      "learning_rate": 5.747323340471093e-06,
      "loss": 0.17205625772476196,
      "num_input_tokens_seen": 1718460,
      "step": 1670,
      "train_runtime": 4769.3385,
      "train_tokens_per_second": 360.314
    },
    {
      "epoch": 0.7141788652633828,
      "grad_norm": 2.824571132659912,
      "learning_rate": 5.7387580299785874e-06,
      "loss": 0.12216833233833313,
      "num_input_tokens_seen": 1719306,
      "step": 1671,
      "train_runtime": 4771.4593,
      "train_tokens_per_second": 360.331
    },
    {
      "epoch": 0.7146062613527087,
      "grad_norm": 1.7227834463119507,
      "learning_rate": 5.730192719486082e-06,
      "loss": 0.11943650245666504,
      "num_input_tokens_seen": 1720296,
      "step": 1672,
      "train_runtime": 4773.616,
      "train_tokens_per_second": 360.376
    },
    {
      "epoch": 0.7150336574420344,
      "grad_norm": 3.471247673034668,
      "learning_rate": 5.721627408993576e-06,
      "loss": 0.14931151270866394,
      "num_input_tokens_seen": 1720946,
      "step": 1673,
      "train_runtime": 4775.6786,
      "train_tokens_per_second": 360.356
    },
    {
      "epoch": 0.7154610535313601,
      "grad_norm": 2.918505907058716,
      "learning_rate": 5.713062098501071e-06,
      "loss": 0.14648237824440002,
      "num_input_tokens_seen": 1722042,
      "step": 1674,
      "train_runtime": 4777.8223,
      "train_tokens_per_second": 360.424
    },
    {
      "epoch": 0.715888449620686,
      "grad_norm": 2.1645331382751465,
      "learning_rate": 5.704496788008565e-06,
      "loss": 0.12691345810890198,
      "num_input_tokens_seen": 1723140,
      "step": 1675,
      "train_runtime": 4780.0169,
      "train_tokens_per_second": 360.488
    },
    {
      "epoch": 0.7163158457100117,
      "grad_norm": 1.897057056427002,
      "learning_rate": 5.6959314775160604e-06,
      "loss": 0.07775653898715973,
      "num_input_tokens_seen": 1724224,
      "step": 1676,
      "train_runtime": 4782.2157,
      "train_tokens_per_second": 360.549
    },
    {
      "epoch": 0.7167432417993376,
      "grad_norm": 1.6002415418624878,
      "learning_rate": 5.687366167023555e-06,
      "loss": 0.07413661479949951,
      "num_input_tokens_seen": 1725548,
      "step": 1677,
      "train_runtime": 4784.4363,
      "train_tokens_per_second": 360.659
    },
    {
      "epoch": 0.7171706378886633,
      "grad_norm": 3.134046792984009,
      "learning_rate": 5.678800856531049e-06,
      "loss": 0.12984393537044525,
      "num_input_tokens_seen": 1726676,
      "step": 1678,
      "train_runtime": 4786.5931,
      "train_tokens_per_second": 360.732
    },
    {
      "epoch": 0.717598033977989,
      "grad_norm": 3.21637225151062,
      "learning_rate": 5.670235546038544e-06,
      "loss": 0.14277908205986023,
      "num_input_tokens_seen": 1727428,
      "step": 1679,
      "train_runtime": 4788.6672,
      "train_tokens_per_second": 360.733
    },
    {
      "epoch": 0.7180254300673149,
      "grad_norm": 2.0747151374816895,
      "learning_rate": 5.66167023554604e-06,
      "loss": 0.14743301272392273,
      "num_input_tokens_seen": 1728446,
      "step": 1680,
      "train_runtime": 4790.8343,
      "train_tokens_per_second": 360.782
    },
    {
      "epoch": 0.7184528261566406,
      "grad_norm": 2.426614284515381,
      "learning_rate": 5.653104925053534e-06,
      "loss": 0.08483467996120453,
      "num_input_tokens_seen": 1729056,
      "step": 1681,
      "train_runtime": 4806.4139,
      "train_tokens_per_second": 359.739
    },
    {
      "epoch": 0.7188802222459665,
      "grad_norm": 2.793328046798706,
      "learning_rate": 5.644539614561029e-06,
      "loss": 0.15332990884780884,
      "num_input_tokens_seen": 1730126,
      "step": 1682,
      "train_runtime": 4808.586,
      "train_tokens_per_second": 359.799
    },
    {
      "epoch": 0.7193076183352922,
      "grad_norm": 1.9192081689834595,
      "learning_rate": 5.635974304068523e-06,
      "loss": 0.09715915471315384,
      "num_input_tokens_seen": 1730996,
      "step": 1683,
      "train_runtime": 4810.792,
      "train_tokens_per_second": 359.815
    },
    {
      "epoch": 0.7197350144246181,
      "grad_norm": 2.6356565952301025,
      "learning_rate": 5.6274089935760176e-06,
      "loss": 0.13068704307079315,
      "num_input_tokens_seen": 1732152,
      "step": 1684,
      "train_runtime": 4813.0216,
      "train_tokens_per_second": 359.889
    },
    {
      "epoch": 0.7201624105139438,
      "grad_norm": 2.196972608566284,
      "learning_rate": 5.618843683083513e-06,
      "loss": 0.11215974390506744,
      "num_input_tokens_seen": 1733204,
      "step": 1685,
      "train_runtime": 4815.1812,
      "train_tokens_per_second": 359.946
    },
    {
      "epoch": 0.7205898066032695,
      "grad_norm": 1.7246198654174805,
      "learning_rate": 5.610278372591007e-06,
      "loss": 0.06989086419343948,
      "num_input_tokens_seen": 1734046,
      "step": 1686,
      "train_runtime": 4817.3118,
      "train_tokens_per_second": 359.961
    },
    {
      "epoch": 0.7210172026925954,
      "grad_norm": 2.7068393230438232,
      "learning_rate": 5.601713062098502e-06,
      "loss": 0.12449479103088379,
      "num_input_tokens_seen": 1735042,
      "step": 1687,
      "train_runtime": 4819.4961,
      "train_tokens_per_second": 360.005
    },
    {
      "epoch": 0.7214445987819211,
      "grad_norm": 2.3992509841918945,
      "learning_rate": 5.593147751605996e-06,
      "loss": 0.15011781454086304,
      "num_input_tokens_seen": 1736068,
      "step": 1688,
      "train_runtime": 4821.655,
      "train_tokens_per_second": 360.056
    },
    {
      "epoch": 0.721871994871247,
      "grad_norm": 2.043673038482666,
      "learning_rate": 5.5845824411134906e-06,
      "loss": 0.11534906923770905,
      "num_input_tokens_seen": 1737184,
      "step": 1689,
      "train_runtime": 4823.8549,
      "train_tokens_per_second": 360.124
    },
    {
      "epoch": 0.7222993909605727,
      "grad_norm": 2.742030382156372,
      "learning_rate": 5.576017130620985e-06,
      "loss": 0.18591052293777466,
      "num_input_tokens_seen": 1738208,
      "step": 1690,
      "train_runtime": 4825.9867,
      "train_tokens_per_second": 360.177
    },
    {
      "epoch": 0.7227267870498985,
      "grad_norm": 1.76872980594635,
      "learning_rate": 5.56745182012848e-06,
      "loss": 0.10890193283557892,
      "num_input_tokens_seen": 1739152,
      "step": 1691,
      "train_runtime": 4828.0859,
      "train_tokens_per_second": 360.216
    },
    {
      "epoch": 0.7231541831392243,
      "grad_norm": 2.0823118686676025,
      "learning_rate": 5.558886509635975e-06,
      "loss": 0.11015980690717697,
      "num_input_tokens_seen": 1740346,
      "step": 1692,
      "train_runtime": 4830.2839,
      "train_tokens_per_second": 360.299
    },
    {
      "epoch": 0.72358157922855,
      "grad_norm": 2.5794262886047363,
      "learning_rate": 5.550321199143469e-06,
      "loss": 0.20696821808815002,
      "num_input_tokens_seen": 1741248,
      "step": 1693,
      "train_runtime": 4832.4269,
      "train_tokens_per_second": 360.326
    },
    {
      "epoch": 0.7240089753178759,
      "grad_norm": 2.58174204826355,
      "learning_rate": 5.5417558886509636e-06,
      "loss": 0.12523779273033142,
      "num_input_tokens_seen": 1742122,
      "step": 1694,
      "train_runtime": 4834.5694,
      "train_tokens_per_second": 360.347
    },
    {
      "epoch": 0.7244363714072016,
      "grad_norm": 1.6804932355880737,
      "learning_rate": 5.533190578158458e-06,
      "loss": 0.10602365434169769,
      "num_input_tokens_seen": 1743328,
      "step": 1695,
      "train_runtime": 4836.8368,
      "train_tokens_per_second": 360.427
    },
    {
      "epoch": 0.7248637674965274,
      "grad_norm": 2.3279054164886475,
      "learning_rate": 5.524625267665953e-06,
      "loss": 0.14723874628543854,
      "num_input_tokens_seen": 1744486,
      "step": 1696,
      "train_runtime": 4839.0978,
      "train_tokens_per_second": 360.498
    },
    {
      "epoch": 0.7252911635858532,
      "grad_norm": 2.3080601692199707,
      "learning_rate": 5.516059957173448e-06,
      "loss": 0.1581530123949051,
      "num_input_tokens_seen": 1745474,
      "step": 1697,
      "train_runtime": 4841.2732,
      "train_tokens_per_second": 360.54
    },
    {
      "epoch": 0.725718559675179,
      "grad_norm": 2.790524959564209,
      "learning_rate": 5.507494646680943e-06,
      "loss": 0.162436842918396,
      "num_input_tokens_seen": 1746166,
      "step": 1698,
      "train_runtime": 4843.3015,
      "train_tokens_per_second": 360.532
    },
    {
      "epoch": 0.7261459557645048,
      "grad_norm": 2.400913953781128,
      "learning_rate": 5.498929336188437e-06,
      "loss": 0.18709856271743774,
      "num_input_tokens_seen": 1747250,
      "step": 1699,
      "train_runtime": 4845.4682,
      "train_tokens_per_second": 360.595
    },
    {
      "epoch": 0.7265733518538305,
      "grad_norm": 2.3663182258605957,
      "learning_rate": 5.490364025695933e-06,
      "loss": 0.09549424052238464,
      "num_input_tokens_seen": 1748350,
      "step": 1700,
      "train_runtime": 4847.627,
      "train_tokens_per_second": 360.661
    },
    {
      "epoch": 0.7270007479431563,
      "grad_norm": 2.2285592555999756,
      "learning_rate": 5.481798715203427e-06,
      "loss": 0.09639982134103775,
      "num_input_tokens_seen": 1749094,
      "step": 1701,
      "train_runtime": 4849.7368,
      "train_tokens_per_second": 360.658
    },
    {
      "epoch": 0.7274281440324821,
      "grad_norm": 2.3181865215301514,
      "learning_rate": 5.4732334047109215e-06,
      "loss": 0.11261068284511566,
      "num_input_tokens_seen": 1749952,
      "step": 1702,
      "train_runtime": 4851.9496,
      "train_tokens_per_second": 360.67
    },
    {
      "epoch": 0.7278555401218079,
      "grad_norm": 2.177008867263794,
      "learning_rate": 5.464668094218416e-06,
      "loss": 0.0887385755777359,
      "num_input_tokens_seen": 1750582,
      "step": 1703,
      "train_runtime": 4854.0042,
      "train_tokens_per_second": 360.647
    },
    {
      "epoch": 0.7282829362111337,
      "grad_norm": 2.890014886856079,
      "learning_rate": 5.45610278372591e-06,
      "loss": 0.10336367785930634,
      "num_input_tokens_seen": 1751290,
      "step": 1704,
      "train_runtime": 4856.0889,
      "train_tokens_per_second": 360.638
    },
    {
      "epoch": 0.7287103323004595,
      "grad_norm": 3.8933002948760986,
      "learning_rate": 5.447537473233405e-06,
      "loss": 0.2152257263660431,
      "num_input_tokens_seen": 1752338,
      "step": 1705,
      "train_runtime": 4858.2584,
      "train_tokens_per_second": 360.693
    },
    {
      "epoch": 0.7291377283897852,
      "grad_norm": 1.7483906745910645,
      "learning_rate": 5.4389721627409e-06,
      "loss": 0.08699686825275421,
      "num_input_tokens_seen": 1753384,
      "step": 1706,
      "train_runtime": 4860.4018,
      "train_tokens_per_second": 360.749
    },
    {
      "epoch": 0.729565124479111,
      "grad_norm": 2.5682764053344727,
      "learning_rate": 5.4304068522483945e-06,
      "loss": 0.16164517402648926,
      "num_input_tokens_seen": 1754138,
      "step": 1707,
      "train_runtime": 4862.4738,
      "train_tokens_per_second": 360.75
    },
    {
      "epoch": 0.7299925205684368,
      "grad_norm": 2.454434394836426,
      "learning_rate": 5.421841541755889e-06,
      "loss": 0.17905999720096588,
      "num_input_tokens_seen": 1755136,
      "step": 1708,
      "train_runtime": 4864.6238,
      "train_tokens_per_second": 360.796
    },
    {
      "epoch": 0.7304199166577626,
      "grad_norm": 2.6822142601013184,
      "learning_rate": 5.413276231263383e-06,
      "loss": 0.16729134321212769,
      "num_input_tokens_seen": 1756140,
      "step": 1709,
      "train_runtime": 4866.8023,
      "train_tokens_per_second": 360.841
    },
    {
      "epoch": 0.7308473127470884,
      "grad_norm": 2.301751136779785,
      "learning_rate": 5.404710920770878e-06,
      "loss": 0.12160074710845947,
      "num_input_tokens_seen": 1757116,
      "step": 1710,
      "train_runtime": 4868.9434,
      "train_tokens_per_second": 360.882
    },
    {
      "epoch": 0.7312747088364141,
      "grad_norm": 1.7962868213653564,
      "learning_rate": 5.396145610278373e-06,
      "loss": 0.051778871566057205,
      "num_input_tokens_seen": 1757922,
      "step": 1711,
      "train_runtime": 4884.0219,
      "train_tokens_per_second": 359.933
    },
    {
      "epoch": 0.73170210492574,
      "grad_norm": 1.967346429824829,
      "learning_rate": 5.3875802997858675e-06,
      "loss": 0.12240543216466904,
      "num_input_tokens_seen": 1759588,
      "step": 1712,
      "train_runtime": 4886.3333,
      "train_tokens_per_second": 360.104
    },
    {
      "epoch": 0.7321295010150657,
      "grad_norm": 2.5978519916534424,
      "learning_rate": 5.379014989293362e-06,
      "loss": 0.09371157735586166,
      "num_input_tokens_seen": 1760308,
      "step": 1713,
      "train_runtime": 4888.4075,
      "train_tokens_per_second": 360.098
    },
    {
      "epoch": 0.7325568971043915,
      "grad_norm": 2.319333553314209,
      "learning_rate": 5.370449678800856e-06,
      "loss": 0.20478688180446625,
      "num_input_tokens_seen": 1762078,
      "step": 1714,
      "train_runtime": 4890.7612,
      "train_tokens_per_second": 360.287
    },
    {
      "epoch": 0.7329842931937173,
      "grad_norm": 1.245335340499878,
      "learning_rate": 5.3618843683083525e-06,
      "loss": 0.08343061804771423,
      "num_input_tokens_seen": 1766162,
      "step": 1715,
      "train_runtime": 4893.6673,
      "train_tokens_per_second": 360.908
    },
    {
      "epoch": 0.733411689283043,
      "grad_norm": 2.084432363510132,
      "learning_rate": 5.353319057815847e-06,
      "loss": 0.07743018120527267,
      "num_input_tokens_seen": 1766872,
      "step": 1716,
      "train_runtime": 4895.7272,
      "train_tokens_per_second": 360.901
    },
    {
      "epoch": 0.7338390853723689,
      "grad_norm": 2.114238977432251,
      "learning_rate": 5.344753747323341e-06,
      "loss": 0.11940549314022064,
      "num_input_tokens_seen": 1767812,
      "step": 1717,
      "train_runtime": 4897.8848,
      "train_tokens_per_second": 360.934
    },
    {
      "epoch": 0.7342664814616946,
      "grad_norm": 2.5279343128204346,
      "learning_rate": 5.336188436830836e-06,
      "loss": 0.0997849553823471,
      "num_input_tokens_seen": 1768594,
      "step": 1718,
      "train_runtime": 4899.9657,
      "train_tokens_per_second": 360.94
    },
    {
      "epoch": 0.7346938775510204,
      "grad_norm": 2.5341482162475586,
      "learning_rate": 5.32762312633833e-06,
      "loss": 0.14740152657032013,
      "num_input_tokens_seen": 1769324,
      "step": 1719,
      "train_runtime": 4901.9986,
      "train_tokens_per_second": 360.939
    },
    {
      "epoch": 0.7351212736403462,
      "grad_norm": 3.104464054107666,
      "learning_rate": 5.319057815845825e-06,
      "loss": 0.12243078649044037,
      "num_input_tokens_seen": 1770244,
      "step": 1720,
      "train_runtime": 4904.2863,
      "train_tokens_per_second": 360.959
    },
    {
      "epoch": 0.7355486697296719,
      "grad_norm": 2.949453592300415,
      "learning_rate": 5.31049250535332e-06,
      "loss": 0.15506841242313385,
      "num_input_tokens_seen": 1771034,
      "step": 1721,
      "train_runtime": 4906.3379,
      "train_tokens_per_second": 360.969
    },
    {
      "epoch": 0.7359760658189978,
      "grad_norm": 2.606902837753296,
      "learning_rate": 5.301927194860814e-06,
      "loss": 0.10120886564254761,
      "num_input_tokens_seen": 1771848,
      "step": 1722,
      "train_runtime": 4908.4644,
      "train_tokens_per_second": 360.978
    },
    {
      "epoch": 0.7364034619083235,
      "grad_norm": 2.355135917663574,
      "learning_rate": 5.293361884368309e-06,
      "loss": 0.09580687433481216,
      "num_input_tokens_seen": 1772830,
      "step": 1723,
      "train_runtime": 4910.644,
      "train_tokens_per_second": 361.018
    },
    {
      "epoch": 0.7368308579976494,
      "grad_norm": 3.3114848136901855,
      "learning_rate": 5.284796573875803e-06,
      "loss": 0.1144857332110405,
      "num_input_tokens_seen": 1773516,
      "step": 1724,
      "train_runtime": 4912.7477,
      "train_tokens_per_second": 361.003
    },
    {
      "epoch": 0.7372582540869751,
      "grad_norm": 2.0524861812591553,
      "learning_rate": 5.276231263383298e-06,
      "loss": 0.10405785590410233,
      "num_input_tokens_seen": 1774302,
      "step": 1725,
      "train_runtime": 4914.8364,
      "train_tokens_per_second": 361.009
    },
    {
      "epoch": 0.7376856501763008,
      "grad_norm": 2.8031365871429443,
      "learning_rate": 5.267665952890793e-06,
      "loss": 0.17746911942958832,
      "num_input_tokens_seen": 1775332,
      "step": 1726,
      "train_runtime": 4916.9499,
      "train_tokens_per_second": 361.064
    },
    {
      "epoch": 0.7381130462656267,
      "grad_norm": 1.2895371913909912,
      "learning_rate": 5.259100642398287e-06,
      "loss": 0.08301758021116257,
      "num_input_tokens_seen": 1777012,
      "step": 1727,
      "train_runtime": 4919.2948,
      "train_tokens_per_second": 361.233
    },
    {
      "epoch": 0.7385404423549524,
      "grad_norm": 2.246673583984375,
      "learning_rate": 5.250535331905782e-06,
      "loss": 0.11134842038154602,
      "num_input_tokens_seen": 1778124,
      "step": 1728,
      "train_runtime": 4921.4548,
      "train_tokens_per_second": 361.3
    },
    {
      "epoch": 0.7389678384442783,
      "grad_norm": 2.3084068298339844,
      "learning_rate": 5.241970021413276e-06,
      "loss": 0.14449399709701538,
      "num_input_tokens_seen": 1778986,
      "step": 1729,
      "train_runtime": 4923.5779,
      "train_tokens_per_second": 361.32
    },
    {
      "epoch": 0.739395234533604,
      "grad_norm": 2.2981183528900146,
      "learning_rate": 5.233404710920771e-06,
      "loss": 0.14202140271663666,
      "num_input_tokens_seen": 1780028,
      "step": 1730,
      "train_runtime": 4925.7332,
      "train_tokens_per_second": 361.373
    },
    {
      "epoch": 0.7398226306229299,
      "grad_norm": 1.8037728071212769,
      "learning_rate": 5.224839400428265e-06,
      "loss": 0.0830642506480217,
      "num_input_tokens_seen": 1781036,
      "step": 1731,
      "train_runtime": 4927.8742,
      "train_tokens_per_second": 361.421
    },
    {
      "epoch": 0.7402500267122556,
      "grad_norm": 1.8055533170700073,
      "learning_rate": 5.216274089935761e-06,
      "loss": 0.09065364301204681,
      "num_input_tokens_seen": 1782164,
      "step": 1732,
      "train_runtime": 4930.0559,
      "train_tokens_per_second": 361.49
    },
    {
      "epoch": 0.7406774228015813,
      "grad_norm": 2.006284475326538,
      "learning_rate": 5.207708779443256e-06,
      "loss": 0.08791021257638931,
      "num_input_tokens_seen": 1783190,
      "step": 1733,
      "train_runtime": 4932.2141,
      "train_tokens_per_second": 361.539
    },
    {
      "epoch": 0.7411048188909072,
      "grad_norm": 1.884128451347351,
      "learning_rate": 5.19914346895075e-06,
      "loss": 0.13375459611415863,
      "num_input_tokens_seen": 1784506,
      "step": 1734,
      "train_runtime": 4934.4358,
      "train_tokens_per_second": 361.643
    },
    {
      "epoch": 0.7415322149802329,
      "grad_norm": 2.6705405712127686,
      "learning_rate": 5.1905781584582445e-06,
      "loss": 0.10275781154632568,
      "num_input_tokens_seen": 1785292,
      "step": 1735,
      "train_runtime": 4936.5609,
      "train_tokens_per_second": 361.647
    },
    {
      "epoch": 0.7419596110695588,
      "grad_norm": 2.2450931072235107,
      "learning_rate": 5.18201284796574e-06,
      "loss": 0.12784308195114136,
      "num_input_tokens_seen": 1786170,
      "step": 1736,
      "train_runtime": 4938.686,
      "train_tokens_per_second": 361.669
    },
    {
      "epoch": 0.7423870071588845,
      "grad_norm": 2.742281198501587,
      "learning_rate": 5.173447537473234e-06,
      "loss": 0.13632413744926453,
      "num_input_tokens_seen": 1786842,
      "step": 1737,
      "train_runtime": 4940.7336,
      "train_tokens_per_second": 361.655
    },
    {
      "epoch": 0.7428144032482102,
      "grad_norm": 2.174654722213745,
      "learning_rate": 5.164882226980729e-06,
      "loss": 0.15204423666000366,
      "num_input_tokens_seen": 1788128,
      "step": 1738,
      "train_runtime": 4942.9475,
      "train_tokens_per_second": 361.753
    },
    {
      "epoch": 0.7432417993375361,
      "grad_norm": 2.109191656112671,
      "learning_rate": 5.156316916488223e-06,
      "loss": 0.1222369372844696,
      "num_input_tokens_seen": 1789048,
      "step": 1739,
      "train_runtime": 4945.0669,
      "train_tokens_per_second": 361.784
    },
    {
      "epoch": 0.7436691954268618,
      "grad_norm": 2.59523606300354,
      "learning_rate": 5.1477516059957175e-06,
      "loss": 0.16759590804576874,
      "num_input_tokens_seen": 1790130,
      "step": 1740,
      "train_runtime": 4947.2487,
      "train_tokens_per_second": 361.844
    },
    {
      "epoch": 0.7440965915161877,
      "grad_norm": 2.0634920597076416,
      "learning_rate": 5.139186295503213e-06,
      "loss": 0.13964451849460602,
      "num_input_tokens_seen": 1791316,
      "step": 1741,
      "train_runtime": 4962.3175,
      "train_tokens_per_second": 360.984
    },
    {
      "epoch": 0.7445239876055134,
      "grad_norm": 3.337764263153076,
      "learning_rate": 5.130620985010707e-06,
      "loss": 0.07662618160247803,
      "num_input_tokens_seen": 1792168,
      "step": 1742,
      "train_runtime": 4964.4055,
      "train_tokens_per_second": 361.004
    },
    {
      "epoch": 0.7449513836948392,
      "grad_norm": 2.6240949630737305,
      "learning_rate": 5.122055674518202e-06,
      "loss": 0.08762823045253754,
      "num_input_tokens_seen": 1792770,
      "step": 1743,
      "train_runtime": 4966.468,
      "train_tokens_per_second": 360.975
    },
    {
      "epoch": 0.745378779784165,
      "grad_norm": 2.48095965385437,
      "learning_rate": 5.113490364025696e-06,
      "loss": 0.09734433144330978,
      "num_input_tokens_seen": 1793856,
      "step": 1744,
      "train_runtime": 4968.6718,
      "train_tokens_per_second": 361.033
    },
    {
      "epoch": 0.7458061758734907,
      "grad_norm": 2.2448716163635254,
      "learning_rate": 5.1049250535331905e-06,
      "loss": 0.1444011628627777,
      "num_input_tokens_seen": 1794662,
      "step": 1745,
      "train_runtime": 4970.7634,
      "train_tokens_per_second": 361.044
    },
    {
      "epoch": 0.7462335719628166,
      "grad_norm": 2.1753718852996826,
      "learning_rate": 5.096359743040685e-06,
      "loss": 0.08833227306604385,
      "num_input_tokens_seen": 1795566,
      "step": 1746,
      "train_runtime": 4972.9053,
      "train_tokens_per_second": 361.07
    },
    {
      "epoch": 0.7466609680521423,
      "grad_norm": 2.468184232711792,
      "learning_rate": 5.08779443254818e-06,
      "loss": 0.10683442652225494,
      "num_input_tokens_seen": 1796186,
      "step": 1747,
      "train_runtime": 4974.9249,
      "train_tokens_per_second": 361.048
    },
    {
      "epoch": 0.7470883641414681,
      "grad_norm": 2.383761405944824,
      "learning_rate": 5.079229122055675e-06,
      "loss": 0.15987126529216766,
      "num_input_tokens_seen": 1798050,
      "step": 1748,
      "train_runtime": 4977.2968,
      "train_tokens_per_second": 361.25
    },
    {
      "epoch": 0.7475157602307939,
      "grad_norm": 2.041799306869507,
      "learning_rate": 5.07066381156317e-06,
      "loss": 0.1062455028295517,
      "num_input_tokens_seen": 1799004,
      "step": 1749,
      "train_runtime": 4979.4533,
      "train_tokens_per_second": 361.285
    },
    {
      "epoch": 0.7479431563201197,
      "grad_norm": 2.9271047115325928,
      "learning_rate": 5.062098501070664e-06,
      "loss": 0.09416396170854568,
      "num_input_tokens_seen": 1799964,
      "step": 1750,
      "train_runtime": 4981.5995,
      "train_tokens_per_second": 361.323
    },
    {
      "epoch": 0.7483705524094455,
      "grad_norm": 1.8387802839279175,
      "learning_rate": 5.05353319057816e-06,
      "loss": 0.09066452085971832,
      "num_input_tokens_seen": 1800854,
      "step": 1751,
      "train_runtime": 4983.7704,
      "train_tokens_per_second": 361.344
    },
    {
      "epoch": 0.7487979484987712,
      "grad_norm": 2.1731786727905273,
      "learning_rate": 5.044967880085654e-06,
      "loss": 0.15073305368423462,
      "num_input_tokens_seen": 1801794,
      "step": 1752,
      "train_runtime": 4986.007,
      "train_tokens_per_second": 361.37
    },
    {
      "epoch": 0.749225344588097,
      "grad_norm": 2.4906811714172363,
      "learning_rate": 5.0364025695931484e-06,
      "loss": 0.09631376713514328,
      "num_input_tokens_seen": 1802644,
      "step": 1753,
      "train_runtime": 4988.2506,
      "train_tokens_per_second": 361.378
    },
    {
      "epoch": 0.7496527406774228,
      "grad_norm": 2.490089178085327,
      "learning_rate": 5.027837259100643e-06,
      "loss": 0.2221662849187851,
      "num_input_tokens_seen": 1803724,
      "step": 1754,
      "train_runtime": 4990.4096,
      "train_tokens_per_second": 361.438
    },
    {
      "epoch": 0.7500801367667486,
      "grad_norm": 2.221968412399292,
      "learning_rate": 5.019271948608137e-06,
      "loss": 0.1377941071987152,
      "num_input_tokens_seen": 1804638,
      "step": 1755,
      "train_runtime": 4992.531,
      "train_tokens_per_second": 361.468
    },
    {
      "epoch": 0.7505075328560744,
      "grad_norm": 2.5073354244232178,
      "learning_rate": 5.010706638115633e-06,
      "loss": 0.14367598295211792,
      "num_input_tokens_seen": 1805340,
      "step": 1756,
      "train_runtime": 4994.5972,
      "train_tokens_per_second": 361.459
    },
    {
      "epoch": 0.7509349289454001,
      "grad_norm": 2.0831339359283447,
      "learning_rate": 5.002141327623127e-06,
      "loss": 0.13118833303451538,
      "num_input_tokens_seen": 1806728,
      "step": 1757,
      "train_runtime": 4996.8626,
      "train_tokens_per_second": 361.572
    },
    {
      "epoch": 0.7513623250347259,
      "grad_norm": 2.693484306335449,
      "learning_rate": 4.9935760171306214e-06,
      "loss": 0.10646102577447891,
      "num_input_tokens_seen": 1807926,
      "step": 1758,
      "train_runtime": 4999.0126,
      "train_tokens_per_second": 361.657
    },
    {
      "epoch": 0.7517897211240517,
      "grad_norm": 2.6705830097198486,
      "learning_rate": 4.985010706638116e-06,
      "loss": 0.16977126896381378,
      "num_input_tokens_seen": 1808960,
      "step": 1759,
      "train_runtime": 5001.165,
      "train_tokens_per_second": 361.708
    },
    {
      "epoch": 0.7522171172133775,
      "grad_norm": 1.7905365228652954,
      "learning_rate": 4.97644539614561e-06,
      "loss": 0.10334272682666779,
      "num_input_tokens_seen": 1810242,
      "step": 1760,
      "train_runtime": 5003.3683,
      "train_tokens_per_second": 361.805
    },
    {
      "epoch": 0.7526445133027033,
      "grad_norm": 2.0694689750671387,
      "learning_rate": 4.967880085653105e-06,
      "loss": 0.13218551874160767,
      "num_input_tokens_seen": 1811086,
      "step": 1761,
      "train_runtime": 5005.4824,
      "train_tokens_per_second": 361.82
    },
    {
      "epoch": 0.7530719093920291,
      "grad_norm": 2.647510528564453,
      "learning_rate": 4.9593147751606e-06,
      "loss": 0.20446237921714783,
      "num_input_tokens_seen": 1812102,
      "step": 1762,
      "train_runtime": 5007.6619,
      "train_tokens_per_second": 361.866
    },
    {
      "epoch": 0.7534993054813548,
      "grad_norm": 2.8202621936798096,
      "learning_rate": 4.9507494646680944e-06,
      "loss": 0.15166407823562622,
      "num_input_tokens_seen": 1812822,
      "step": 1763,
      "train_runtime": 5009.7093,
      "train_tokens_per_second": 361.862
    },
    {
      "epoch": 0.7539267015706806,
      "grad_norm": 2.207237482070923,
      "learning_rate": 4.94218415417559e-06,
      "loss": 0.1600687950849533,
      "num_input_tokens_seen": 1813886,
      "step": 1764,
      "train_runtime": 5011.8518,
      "train_tokens_per_second": 361.919
    },
    {
      "epoch": 0.7543540976600064,
      "grad_norm": 2.0225484371185303,
      "learning_rate": 4.933618843683084e-06,
      "loss": 0.11253505200147629,
      "num_input_tokens_seen": 1815006,
      "step": 1765,
      "train_runtime": 5014.0197,
      "train_tokens_per_second": 361.986
    },
    {
      "epoch": 0.7547814937493322,
      "grad_norm": 1.6529148817062378,
      "learning_rate": 4.9250535331905786e-06,
      "loss": 0.09459485113620758,
      "num_input_tokens_seen": 1815906,
      "step": 1766,
      "train_runtime": 5016.1388,
      "train_tokens_per_second": 362.013
    },
    {
      "epoch": 0.755208889838658,
      "grad_norm": 1.4612174034118652,
      "learning_rate": 4.916488222698073e-06,
      "loss": 0.06439751386642456,
      "num_input_tokens_seen": 1816980,
      "step": 1767,
      "train_runtime": 5018.3082,
      "train_tokens_per_second": 362.07
    },
    {
      "epoch": 0.7556362859279837,
      "grad_norm": 1.751206398010254,
      "learning_rate": 4.9079229122055674e-06,
      "loss": 0.09944566339254379,
      "num_input_tokens_seen": 1817988,
      "step": 1768,
      "train_runtime": 5020.4363,
      "train_tokens_per_second": 362.118
    },
    {
      "epoch": 0.7560636820173096,
      "grad_norm": 1.9897797107696533,
      "learning_rate": 4.899357601713063e-06,
      "loss": 0.1456417739391327,
      "num_input_tokens_seen": 1819128,
      "step": 1769,
      "train_runtime": 5022.639,
      "train_tokens_per_second": 362.186
    },
    {
      "epoch": 0.7564910781066353,
      "grad_norm": 1.4492417573928833,
      "learning_rate": 4.890792291220557e-06,
      "loss": 0.07394880801439285,
      "num_input_tokens_seen": 1820096,
      "step": 1770,
      "train_runtime": 5024.7778,
      "train_tokens_per_second": 362.224
    },
    {
      "epoch": 0.7569184741959611,
      "grad_norm": 2.184596061706543,
      "learning_rate": 4.882226980728052e-06,
      "loss": 0.10925288498401642,
      "num_input_tokens_seen": 1820958,
      "step": 1771,
      "train_runtime": 5039.8012,
      "train_tokens_per_second": 361.315
    },
    {
      "epoch": 0.7573458702852869,
      "grad_norm": 3.1226325035095215,
      "learning_rate": 4.873661670235547e-06,
      "loss": 0.1586754471063614,
      "num_input_tokens_seen": 1821880,
      "step": 1772,
      "train_runtime": 5042.5892,
      "train_tokens_per_second": 361.299
    },
    {
      "epoch": 0.7577732663746126,
      "grad_norm": 2.022779941558838,
      "learning_rate": 4.865096359743041e-06,
      "loss": 0.1342182457447052,
      "num_input_tokens_seen": 1822866,
      "step": 1773,
      "train_runtime": 5044.8003,
      "train_tokens_per_second": 361.336
    },
    {
      "epoch": 0.7582006624639385,
      "grad_norm": 2.693209409713745,
      "learning_rate": 4.856531049250536e-06,
      "loss": 0.23240728676319122,
      "num_input_tokens_seen": 1823646,
      "step": 1774,
      "train_runtime": 5046.8943,
      "train_tokens_per_second": 361.34
    },
    {
      "epoch": 0.7586280585532642,
      "grad_norm": 1.7339637279510498,
      "learning_rate": 4.84796573875803e-06,
      "loss": 0.1240958422422409,
      "num_input_tokens_seen": 1825138,
      "step": 1775,
      "train_runtime": 5049.2126,
      "train_tokens_per_second": 361.47
    },
    {
      "epoch": 0.75905545464259,
      "grad_norm": 2.3785059452056885,
      "learning_rate": 4.8394004282655246e-06,
      "loss": 0.1522769182920456,
      "num_input_tokens_seen": 1825992,
      "step": 1776,
      "train_runtime": 5051.3397,
      "train_tokens_per_second": 361.487
    },
    {
      "epoch": 0.7594828507319158,
      "grad_norm": 1.264596700668335,
      "learning_rate": 4.83083511777302e-06,
      "loss": 0.06347937881946564,
      "num_input_tokens_seen": 1827504,
      "step": 1777,
      "train_runtime": 5053.6685,
      "train_tokens_per_second": 361.619
    },
    {
      "epoch": 0.7599102468212415,
      "grad_norm": 1.5487502813339233,
      "learning_rate": 4.822269807280514e-06,
      "loss": 0.07012402266263962,
      "num_input_tokens_seen": 1828998,
      "step": 1778,
      "train_runtime": 5055.9214,
      "train_tokens_per_second": 361.754
    },
    {
      "epoch": 0.7603376429105674,
      "grad_norm": 1.6670122146606445,
      "learning_rate": 4.8137044967880095e-06,
      "loss": 0.07347247749567032,
      "num_input_tokens_seen": 1829948,
      "step": 1779,
      "train_runtime": 5058.0862,
      "train_tokens_per_second": 361.787
    },
    {
      "epoch": 0.7607650389998931,
      "grad_norm": 3.5259151458740234,
      "learning_rate": 4.805139186295504e-06,
      "loss": 0.14109820127487183,
      "num_input_tokens_seen": 1831222,
      "step": 1780,
      "train_runtime": 5060.3129,
      "train_tokens_per_second": 361.879
    },
    {
      "epoch": 0.761192435089219,
      "grad_norm": 3.180718183517456,
      "learning_rate": 4.796573875802998e-06,
      "loss": 0.15151821076869965,
      "num_input_tokens_seen": 1832244,
      "step": 1781,
      "train_runtime": 5062.5126,
      "train_tokens_per_second": 361.924
    },
    {
      "epoch": 0.7616198311785447,
      "grad_norm": 1.370081901550293,
      "learning_rate": 4.788008565310493e-06,
      "loss": 0.07774853706359863,
      "num_input_tokens_seen": 1833752,
      "step": 1782,
      "train_runtime": 5064.8718,
      "train_tokens_per_second": 362.053
    },
    {
      "epoch": 0.7620472272678706,
      "grad_norm": 3.96315860748291,
      "learning_rate": 4.779443254817987e-06,
      "loss": 0.18797242641448975,
      "num_input_tokens_seen": 1834596,
      "step": 1783,
      "train_runtime": 5066.9988,
      "train_tokens_per_second": 362.068
    },
    {
      "epoch": 0.7624746233571963,
      "grad_norm": 2.029892921447754,
      "learning_rate": 4.7708779443254825e-06,
      "loss": 0.15799033641815186,
      "num_input_tokens_seen": 1835838,
      "step": 1784,
      "train_runtime": 5069.1867,
      "train_tokens_per_second": 362.156
    },
    {
      "epoch": 0.762902019446522,
      "grad_norm": 2.7727293968200684,
      "learning_rate": 4.762312633832977e-06,
      "loss": 0.10154961794614792,
      "num_input_tokens_seen": 1836712,
      "step": 1785,
      "train_runtime": 5071.327,
      "train_tokens_per_second": 362.176
    },
    {
      "epoch": 0.7633294155358479,
      "grad_norm": 3.1447322368621826,
      "learning_rate": 4.753747323340471e-06,
      "loss": 0.1865973174571991,
      "num_input_tokens_seen": 1837436,
      "step": 1786,
      "train_runtime": 5073.432,
      "train_tokens_per_second": 362.168
    },
    {
      "epoch": 0.7637568116251736,
      "grad_norm": 2.0704610347747803,
      "learning_rate": 4.745182012847966e-06,
      "loss": 0.09010666608810425,
      "num_input_tokens_seen": 1838206,
      "step": 1787,
      "train_runtime": 5075.5737,
      "train_tokens_per_second": 362.167
    },
    {
      "epoch": 0.7641842077144995,
      "grad_norm": 2.9957287311553955,
      "learning_rate": 4.736616702355461e-06,
      "loss": 0.12626878917217255,
      "num_input_tokens_seen": 1839590,
      "step": 1788,
      "train_runtime": 5077.7974,
      "train_tokens_per_second": 362.281
    },
    {
      "epoch": 0.7646116038038252,
      "grad_norm": 2.074080228805542,
      "learning_rate": 4.7280513918629555e-06,
      "loss": 0.08674152195453644,
      "num_input_tokens_seen": 1840380,
      "step": 1789,
      "train_runtime": 5079.864,
      "train_tokens_per_second": 362.289
    },
    {
      "epoch": 0.7650389998931509,
      "grad_norm": 2.0169169902801514,
      "learning_rate": 4.71948608137045e-06,
      "loss": 0.11346843838691711,
      "num_input_tokens_seen": 1841326,
      "step": 1790,
      "train_runtime": 5081.9815,
      "train_tokens_per_second": 362.324
    },
    {
      "epoch": 0.7654663959824768,
      "grad_norm": 2.252941608428955,
      "learning_rate": 4.710920770877944e-06,
      "loss": 0.10692466050386429,
      "num_input_tokens_seen": 1842096,
      "step": 1791,
      "train_runtime": 5084.0763,
      "train_tokens_per_second": 362.327
    },
    {
      "epoch": 0.7658937920718025,
      "grad_norm": 2.742074728012085,
      "learning_rate": 4.70235546038544e-06,
      "loss": 0.1425284445285797,
      "num_input_tokens_seen": 1842650,
      "step": 1792,
      "train_runtime": 5086.1114,
      "train_tokens_per_second": 362.291
    },
    {
      "epoch": 0.7663211881611284,
      "grad_norm": 2.311002254486084,
      "learning_rate": 4.693790149892934e-06,
      "loss": 0.1447705328464508,
      "num_input_tokens_seen": 1843962,
      "step": 1793,
      "train_runtime": 5088.3243,
      "train_tokens_per_second": 362.391
    },
    {
      "epoch": 0.7667485842504541,
      "grad_norm": 3.126150608062744,
      "learning_rate": 4.6852248394004285e-06,
      "loss": 0.21477144956588745,
      "num_input_tokens_seen": 1844878,
      "step": 1794,
      "train_runtime": 5090.4959,
      "train_tokens_per_second": 362.416
    },
    {
      "epoch": 0.7671759803397799,
      "grad_norm": 3.488682508468628,
      "learning_rate": 4.676659528907923e-06,
      "loss": 0.1394921839237213,
      "num_input_tokens_seen": 1845536,
      "step": 1795,
      "train_runtime": 5092.5953,
      "train_tokens_per_second": 362.396
    },
    {
      "epoch": 0.7676033764291057,
      "grad_norm": 1.9885671138763428,
      "learning_rate": 4.668094218415418e-06,
      "loss": 0.13407884538173676,
      "num_input_tokens_seen": 1846996,
      "step": 1796,
      "train_runtime": 5094.8763,
      "train_tokens_per_second": 362.52
    },
    {
      "epoch": 0.7680307725184314,
      "grad_norm": 2.4607231616973877,
      "learning_rate": 4.659528907922913e-06,
      "loss": 0.13201317191123962,
      "num_input_tokens_seen": 1848206,
      "step": 1797,
      "train_runtime": 5097.0959,
      "train_tokens_per_second": 362.6
    },
    {
      "epoch": 0.7684581686077573,
      "grad_norm": 2.704148530960083,
      "learning_rate": 4.650963597430407e-06,
      "loss": 0.14820332825183868,
      "num_input_tokens_seen": 1849206,
      "step": 1798,
      "train_runtime": 5099.25,
      "train_tokens_per_second": 362.643
    },
    {
      "epoch": 0.768885564697083,
      "grad_norm": 2.26305890083313,
      "learning_rate": 4.642398286937902e-06,
      "loss": 0.1085265576839447,
      "num_input_tokens_seen": 1849852,
      "step": 1799,
      "train_runtime": 5101.3099,
      "train_tokens_per_second": 362.623
    },
    {
      "epoch": 0.7693129607864088,
      "grad_norm": 2.4953997135162354,
      "learning_rate": 4.633832976445397e-06,
      "loss": 0.13118381798267365,
      "num_input_tokens_seen": 1851192,
      "step": 1800,
      "train_runtime": 5103.5249,
      "train_tokens_per_second": 362.728
    },
    {
      "epoch": 0.7697403568757346,
      "grad_norm": 2.278885841369629,
      "learning_rate": 4.625267665952891e-06,
      "loss": 0.09888924658298492,
      "num_input_tokens_seen": 1852074,
      "step": 1801,
      "train_runtime": 5119.4164,
      "train_tokens_per_second": 361.774
    },
    {
      "epoch": 0.7701677529650603,
      "grad_norm": 2.3381922245025635,
      "learning_rate": 4.616702355460386e-06,
      "loss": 0.16264069080352783,
      "num_input_tokens_seen": 1853086,
      "step": 1802,
      "train_runtime": 5121.6034,
      "train_tokens_per_second": 361.818
    },
    {
      "epoch": 0.7705951490543862,
      "grad_norm": 1.7255778312683105,
      "learning_rate": 4.60813704496788e-06,
      "loss": 0.11930899322032928,
      "num_input_tokens_seen": 1854112,
      "step": 1803,
      "train_runtime": 5123.8756,
      "train_tokens_per_second": 361.857
    },
    {
      "epoch": 0.7710225451437119,
      "grad_norm": 1.7517153024673462,
      "learning_rate": 4.5995717344753745e-06,
      "loss": 0.10708696395158768,
      "num_input_tokens_seen": 1855120,
      "step": 1804,
      "train_runtime": 5126.0418,
      "train_tokens_per_second": 361.901
    },
    {
      "epoch": 0.7714499412330377,
      "grad_norm": 2.6439056396484375,
      "learning_rate": 4.59100642398287e-06,
      "loss": 0.08341437578201294,
      "num_input_tokens_seen": 1855818,
      "step": 1805,
      "train_runtime": 5128.1142,
      "train_tokens_per_second": 361.891
    },
    {
      "epoch": 0.7718773373223635,
      "grad_norm": 2.328486204147339,
      "learning_rate": 4.582441113490364e-06,
      "loss": 0.09993680566549301,
      "num_input_tokens_seen": 1856408,
      "step": 1806,
      "train_runtime": 5130.1775,
      "train_tokens_per_second": 361.86
    },
    {
      "epoch": 0.7723047334116893,
      "grad_norm": 2.3189327716827393,
      "learning_rate": 4.5738758029978595e-06,
      "loss": 0.11844801902770996,
      "num_input_tokens_seen": 1857328,
      "step": 1807,
      "train_runtime": 5132.3389,
      "train_tokens_per_second": 361.887
    },
    {
      "epoch": 0.7727321295010151,
      "grad_norm": 2.1405961513519287,
      "learning_rate": 4.565310492505354e-06,
      "loss": 0.14549094438552856,
      "num_input_tokens_seen": 1858462,
      "step": 1808,
      "train_runtime": 5134.4945,
      "train_tokens_per_second": 361.956
    },
    {
      "epoch": 0.7731595255903408,
      "grad_norm": 1.8244071006774902,
      "learning_rate": 4.556745182012848e-06,
      "loss": 0.10341336578130722,
      "num_input_tokens_seen": 1859326,
      "step": 1809,
      "train_runtime": 5136.6723,
      "train_tokens_per_second": 361.971
    },
    {
      "epoch": 0.7735869216796666,
      "grad_norm": 2.110968589782715,
      "learning_rate": 4.548179871520343e-06,
      "loss": 0.11455903947353363,
      "num_input_tokens_seen": 1860078,
      "step": 1810,
      "train_runtime": 5138.8078,
      "train_tokens_per_second": 361.967
    },
    {
      "epoch": 0.7740143177689924,
      "grad_norm": 1.7827335596084595,
      "learning_rate": 4.539614561027837e-06,
      "loss": 0.10141510516405106,
      "num_input_tokens_seen": 1861088,
      "step": 1811,
      "train_runtime": 5140.945,
      "train_tokens_per_second": 362.013
    },
    {
      "epoch": 0.7744417138583182,
      "grad_norm": 2.4329521656036377,
      "learning_rate": 4.5310492505353325e-06,
      "loss": 0.11690457165241241,
      "num_input_tokens_seen": 1861916,
      "step": 1812,
      "train_runtime": 5143.0702,
      "train_tokens_per_second": 362.024
    },
    {
      "epoch": 0.774869109947644,
      "grad_norm": 2.1015193462371826,
      "learning_rate": 4.522483940042827e-06,
      "loss": 0.11495056748390198,
      "num_input_tokens_seen": 1862624,
      "step": 1813,
      "train_runtime": 5145.1293,
      "train_tokens_per_second": 362.017
    },
    {
      "epoch": 0.7752965060369698,
      "grad_norm": 2.322100877761841,
      "learning_rate": 4.513918629550322e-06,
      "loss": 0.10310870409011841,
      "num_input_tokens_seen": 1863660,
      "step": 1814,
      "train_runtime": 5147.2812,
      "train_tokens_per_second": 362.067
    },
    {
      "epoch": 0.7757239021262955,
      "grad_norm": 1.5570839643478394,
      "learning_rate": 4.505353319057817e-06,
      "loss": 0.09182693064212799,
      "num_input_tokens_seen": 1864714,
      "step": 1815,
      "train_runtime": 5149.4093,
      "train_tokens_per_second": 362.122
    },
    {
      "epoch": 0.7761512982156213,
      "grad_norm": 2.36293888092041,
      "learning_rate": 4.496788008565311e-06,
      "loss": 0.15710002183914185,
      "num_input_tokens_seen": 1865414,
      "step": 1816,
      "train_runtime": 5151.5034,
      "train_tokens_per_second": 362.111
    },
    {
      "epoch": 0.7765786943049471,
      "grad_norm": 2.5093367099761963,
      "learning_rate": 4.4882226980728055e-06,
      "loss": 0.15436488389968872,
      "num_input_tokens_seen": 1866346,
      "step": 1817,
      "train_runtime": 5153.6179,
      "train_tokens_per_second": 362.143
    },
    {
      "epoch": 0.7770060903942729,
      "grad_norm": 3.3405544757843018,
      "learning_rate": 4.4796573875803e-06,
      "loss": 0.21985983848571777,
      "num_input_tokens_seen": 1867286,
      "step": 1818,
      "train_runtime": 5155.7658,
      "train_tokens_per_second": 362.174
    },
    {
      "epoch": 0.7774334864835987,
      "grad_norm": 2.5819220542907715,
      "learning_rate": 4.471092077087794e-06,
      "loss": 0.12257808446884155,
      "num_input_tokens_seen": 1868052,
      "step": 1819,
      "train_runtime": 5157.8808,
      "train_tokens_per_second": 362.174
    },
    {
      "epoch": 0.7778608825729244,
      "grad_norm": 1.9124655723571777,
      "learning_rate": 4.46252676659529e-06,
      "loss": 0.11753346025943756,
      "num_input_tokens_seen": 1869264,
      "step": 1820,
      "train_runtime": 5160.0624,
      "train_tokens_per_second": 362.256
    },
    {
      "epoch": 0.7782882786622503,
      "grad_norm": 1.8850749731063843,
      "learning_rate": 4.453961456102784e-06,
      "loss": 0.11016516387462616,
      "num_input_tokens_seen": 1870456,
      "step": 1821,
      "train_runtime": 5162.289,
      "train_tokens_per_second": 362.331
    },
    {
      "epoch": 0.778715674751576,
      "grad_norm": 1.661813735961914,
      "learning_rate": 4.445396145610279e-06,
      "loss": 0.13455480337142944,
      "num_input_tokens_seen": 1871972,
      "step": 1822,
      "train_runtime": 5164.5417,
      "train_tokens_per_second": 362.466
    },
    {
      "epoch": 0.7791430708409018,
      "grad_norm": 3.2305221557617188,
      "learning_rate": 4.436830835117774e-06,
      "loss": 0.11921562254428864,
      "num_input_tokens_seen": 1872810,
      "step": 1823,
      "train_runtime": 5166.6415,
      "train_tokens_per_second": 362.481
    },
    {
      "epoch": 0.7795704669302276,
      "grad_norm": 2.1064887046813965,
      "learning_rate": 4.428265524625268e-06,
      "loss": 0.09975242614746094,
      "num_input_tokens_seen": 1873992,
      "step": 1824,
      "train_runtime": 5168.8276,
      "train_tokens_per_second": 362.556
    },
    {
      "epoch": 0.7799978630195533,
      "grad_norm": 2.426936388015747,
      "learning_rate": 4.419700214132763e-06,
      "loss": 0.16474170982837677,
      "num_input_tokens_seen": 1874830,
      "step": 1825,
      "train_runtime": 5170.9453,
      "train_tokens_per_second": 362.57
    },
    {
      "epoch": 0.7804252591088792,
      "grad_norm": 2.489961862564087,
      "learning_rate": 4.411134903640257e-06,
      "loss": 0.12089613080024719,
      "num_input_tokens_seen": 1875602,
      "step": 1826,
      "train_runtime": 5173.0837,
      "train_tokens_per_second": 362.569
    },
    {
      "epoch": 0.7808526551982049,
      "grad_norm": 2.6534628868103027,
      "learning_rate": 4.402569593147752e-06,
      "loss": 0.13732200860977173,
      "num_input_tokens_seen": 1876560,
      "step": 1827,
      "train_runtime": 5175.2509,
      "train_tokens_per_second": 362.603
    },
    {
      "epoch": 0.7812800512875308,
      "grad_norm": 1.9976394176483154,
      "learning_rate": 4.394004282655247e-06,
      "loss": 0.10832753777503967,
      "num_input_tokens_seen": 1877702,
      "step": 1828,
      "train_runtime": 5177.4241,
      "train_tokens_per_second": 362.671
    },
    {
      "epoch": 0.7817074473768565,
      "grad_norm": 2.093379497528076,
      "learning_rate": 4.385438972162741e-06,
      "loss": 0.2253294289112091,
      "num_input_tokens_seen": 1878878,
      "step": 1829,
      "train_runtime": 5179.6029,
      "train_tokens_per_second": 362.746
    },
    {
      "epoch": 0.7821348434661823,
      "grad_norm": 2.3307065963745117,
      "learning_rate": 4.376873661670236e-06,
      "loss": 0.07001787424087524,
      "num_input_tokens_seen": 1879378,
      "step": 1830,
      "train_runtime": 5181.6337,
      "train_tokens_per_second": 362.7
    },
    {
      "epoch": 0.7825622395555081,
      "grad_norm": 3.0533840656280518,
      "learning_rate": 4.368308351177731e-06,
      "loss": 0.17326922714710236,
      "num_input_tokens_seen": 1880362,
      "step": 1831,
      "train_runtime": 5197.437,
      "train_tokens_per_second": 361.786
    },
    {
      "epoch": 0.7829896356448338,
      "grad_norm": 6.68617582321167,
      "learning_rate": 4.359743040685225e-06,
      "loss": 0.21133653819561005,
      "num_input_tokens_seen": 1881118,
      "step": 1832,
      "train_runtime": 5199.5554,
      "train_tokens_per_second": 361.784
    },
    {
      "epoch": 0.7834170317341597,
      "grad_norm": 3.327510118484497,
      "learning_rate": 4.35117773019272e-06,
      "loss": 0.17654043436050415,
      "num_input_tokens_seen": 1882114,
      "step": 1833,
      "train_runtime": 5201.7041,
      "train_tokens_per_second": 361.826
    },
    {
      "epoch": 0.7838444278234854,
      "grad_norm": 1.9239948987960815,
      "learning_rate": 4.342612419700214e-06,
      "loss": 0.06743384897708893,
      "num_input_tokens_seen": 1883070,
      "step": 1834,
      "train_runtime": 5203.8629,
      "train_tokens_per_second": 361.86
    },
    {
      "epoch": 0.7842718239128113,
      "grad_norm": 2.5109810829162598,
      "learning_rate": 4.3340471092077094e-06,
      "loss": 0.14452077448368073,
      "num_input_tokens_seen": 1883892,
      "step": 1835,
      "train_runtime": 5206.0022,
      "train_tokens_per_second": 361.869
    },
    {
      "epoch": 0.784699220002137,
      "grad_norm": 1.2670587301254272,
      "learning_rate": 4.325481798715204e-06,
      "loss": 0.067506343126297,
      "num_input_tokens_seen": 1885090,
      "step": 1836,
      "train_runtime": 5208.2196,
      "train_tokens_per_second": 361.945
    },
    {
      "epoch": 0.7851266160914627,
      "grad_norm": 1.8930679559707642,
      "learning_rate": 4.316916488222698e-06,
      "loss": 0.1032271757721901,
      "num_input_tokens_seen": 1886034,
      "step": 1837,
      "train_runtime": 5210.3698,
      "train_tokens_per_second": 361.977
    },
    {
      "epoch": 0.7855540121807886,
      "grad_norm": 3.379704236984253,
      "learning_rate": 4.308351177730193e-06,
      "loss": 0.1684207320213318,
      "num_input_tokens_seen": 1887506,
      "step": 1838,
      "train_runtime": 5212.6169,
      "train_tokens_per_second": 362.103
    },
    {
      "epoch": 0.7859814082701143,
      "grad_norm": 1.447620153427124,
      "learning_rate": 4.299785867237688e-06,
      "loss": 0.11584776639938354,
      "num_input_tokens_seen": 1889108,
      "step": 1839,
      "train_runtime": 5214.9021,
      "train_tokens_per_second": 362.252
    },
    {
      "epoch": 0.7864088043594402,
      "grad_norm": 1.519661784172058,
      "learning_rate": 4.2912205567451824e-06,
      "loss": 0.10505079478025436,
      "num_input_tokens_seen": 1890818,
      "step": 1840,
      "train_runtime": 5217.1709,
      "train_tokens_per_second": 362.422
    },
    {
      "epoch": 0.7868362004487659,
      "grad_norm": 1.8158150911331177,
      "learning_rate": 4.282655246252677e-06,
      "loss": 0.07172860950231552,
      "num_input_tokens_seen": 1891494,
      "step": 1841,
      "train_runtime": 5219.3032,
      "train_tokens_per_second": 362.404
    },
    {
      "epoch": 0.7872635965380916,
      "grad_norm": 1.996261715888977,
      "learning_rate": 4.274089935760172e-06,
      "loss": 0.10932371020317078,
      "num_input_tokens_seen": 1892488,
      "step": 1842,
      "train_runtime": 5221.5844,
      "train_tokens_per_second": 362.436
    },
    {
      "epoch": 0.7876909926274175,
      "grad_norm": 1.4062652587890625,
      "learning_rate": 4.265524625267667e-06,
      "loss": 0.08581814169883728,
      "num_input_tokens_seen": 1893790,
      "step": 1843,
      "train_runtime": 5223.8744,
      "train_tokens_per_second": 362.526
    },
    {
      "epoch": 0.7881183887167432,
      "grad_norm": 2.0480849742889404,
      "learning_rate": 4.256959314775161e-06,
      "loss": 0.12999306619167328,
      "num_input_tokens_seen": 1894892,
      "step": 1844,
      "train_runtime": 5226.056,
      "train_tokens_per_second": 362.585
    },
    {
      "epoch": 0.7885457848060691,
      "grad_norm": 3.563359260559082,
      "learning_rate": 4.2483940042826554e-06,
      "loss": 0.16921770572662354,
      "num_input_tokens_seen": 1895542,
      "step": 1845,
      "train_runtime": 5228.1901,
      "train_tokens_per_second": 362.562
    },
    {
      "epoch": 0.7889731808953948,
      "grad_norm": 2.3567779064178467,
      "learning_rate": 4.23982869379015e-06,
      "loss": 0.19969452917575836,
      "num_input_tokens_seen": 1896596,
      "step": 1846,
      "train_runtime": 5230.3809,
      "train_tokens_per_second": 362.611
    },
    {
      "epoch": 0.7894005769847205,
      "grad_norm": 2.3927359580993652,
      "learning_rate": 4.231263383297644e-06,
      "loss": 0.17385578155517578,
      "num_input_tokens_seen": 1897578,
      "step": 1847,
      "train_runtime": 5232.4916,
      "train_tokens_per_second": 362.653
    },
    {
      "epoch": 0.7898279730740464,
      "grad_norm": 2.751912832260132,
      "learning_rate": 4.2226980728051396e-06,
      "loss": 0.18276765942573547,
      "num_input_tokens_seen": 1898646,
      "step": 1848,
      "train_runtime": 5234.638,
      "train_tokens_per_second": 362.708
    },
    {
      "epoch": 0.7902553691633721,
      "grad_norm": 1.5654270648956299,
      "learning_rate": 4.214132762312634e-06,
      "loss": 0.07883001118898392,
      "num_input_tokens_seen": 1899718,
      "step": 1849,
      "train_runtime": 5236.7814,
      "train_tokens_per_second": 362.764
    },
    {
      "epoch": 0.790682765252698,
      "grad_norm": 1.543398141860962,
      "learning_rate": 4.205567451820129e-06,
      "loss": 0.06869515776634216,
      "num_input_tokens_seen": 1900998,
      "step": 1850,
      "train_runtime": 5238.9893,
      "train_tokens_per_second": 362.856
    },
    {
      "epoch": 0.7911101613420237,
      "grad_norm": 1.5596097707748413,
      "learning_rate": 4.197002141327624e-06,
      "loss": 0.0821610540151596,
      "num_input_tokens_seen": 1901850,
      "step": 1851,
      "train_runtime": 5241.113,
      "train_tokens_per_second": 362.871
    },
    {
      "epoch": 0.7915375574313495,
      "grad_norm": 1.888803482055664,
      "learning_rate": 4.188436830835118e-06,
      "loss": 0.12986454367637634,
      "num_input_tokens_seen": 1902918,
      "step": 1852,
      "train_runtime": 5243.2539,
      "train_tokens_per_second": 362.927
    },
    {
      "epoch": 0.7919649535206753,
      "grad_norm": 2.469694137573242,
      "learning_rate": 4.1798715203426126e-06,
      "loss": 0.12588948011398315,
      "num_input_tokens_seen": 1903776,
      "step": 1853,
      "train_runtime": 5245.3403,
      "train_tokens_per_second": 362.946
    },
    {
      "epoch": 0.792392349610001,
      "grad_norm": 2.017801523208618,
      "learning_rate": 4.171306209850107e-06,
      "loss": 0.16612306237220764,
      "num_input_tokens_seen": 1904982,
      "step": 1854,
      "train_runtime": 5247.5139,
      "train_tokens_per_second": 363.026
    },
    {
      "epoch": 0.7928197456993269,
      "grad_norm": 3.377476453781128,
      "learning_rate": 4.1627408993576014e-06,
      "loss": 0.16846582293510437,
      "num_input_tokens_seen": 1905624,
      "step": 1855,
      "train_runtime": 5249.6113,
      "train_tokens_per_second": 363.003
    },
    {
      "epoch": 0.7932471417886526,
      "grad_norm": 3.5028700828552246,
      "learning_rate": 4.154175588865097e-06,
      "loss": 0.1279139369726181,
      "num_input_tokens_seen": 1906454,
      "step": 1856,
      "train_runtime": 5251.7204,
      "train_tokens_per_second": 363.015
    },
    {
      "epoch": 0.7936745378779784,
      "grad_norm": 2.013857364654541,
      "learning_rate": 4.145610278372592e-06,
      "loss": 0.09799247235059738,
      "num_input_tokens_seen": 1907816,
      "step": 1857,
      "train_runtime": 5253.9515,
      "train_tokens_per_second": 363.12
    },
    {
      "epoch": 0.7941019339673042,
      "grad_norm": 1.8402329683303833,
      "learning_rate": 4.137044967880086e-06,
      "loss": 0.09691372513771057,
      "num_input_tokens_seen": 1909478,
      "step": 1858,
      "train_runtime": 5256.2698,
      "train_tokens_per_second": 363.276
    },
    {
      "epoch": 0.79452933005663,
      "grad_norm": 1.8740602731704712,
      "learning_rate": 4.128479657387581e-06,
      "loss": 0.10537990927696228,
      "num_input_tokens_seen": 1910396,
      "step": 1859,
      "train_runtime": 5258.3812,
      "train_tokens_per_second": 363.305
    },
    {
      "epoch": 0.7949567261459558,
      "grad_norm": 2.030247211456299,
      "learning_rate": 4.119914346895075e-06,
      "loss": 0.07523954659700394,
      "num_input_tokens_seen": 1911126,
      "step": 1860,
      "train_runtime": 5260.4708,
      "train_tokens_per_second": 363.299
    },
    {
      "epoch": 0.7953841222352815,
      "grad_norm": 2.6722238063812256,
      "learning_rate": 4.11134903640257e-06,
      "loss": 0.10389907658100128,
      "num_input_tokens_seen": 1911846,
      "step": 1861,
      "train_runtime": 5275.7108,
      "train_tokens_per_second": 362.386
    },
    {
      "epoch": 0.7958115183246073,
      "grad_norm": 1.6098830699920654,
      "learning_rate": 4.102783725910064e-06,
      "loss": 0.09851120412349701,
      "num_input_tokens_seen": 1913048,
      "step": 1862,
      "train_runtime": 5277.8948,
      "train_tokens_per_second": 362.464
    },
    {
      "epoch": 0.7962389144139331,
      "grad_norm": 1.630492925643921,
      "learning_rate": 4.094218415417559e-06,
      "loss": 0.09191761910915375,
      "num_input_tokens_seen": 1914338,
      "step": 1863,
      "train_runtime": 5280.1464,
      "train_tokens_per_second": 362.554
    },
    {
      "epoch": 0.7966663105032589,
      "grad_norm": 3.0116360187530518,
      "learning_rate": 4.085653104925054e-06,
      "loss": 0.1657109558582306,
      "num_input_tokens_seen": 1915156,
      "step": 1864,
      "train_runtime": 5282.2726,
      "train_tokens_per_second": 362.563
    },
    {
      "epoch": 0.7970937065925847,
      "grad_norm": 2.0166096687316895,
      "learning_rate": 4.077087794432549e-06,
      "loss": 0.1979052871465683,
      "num_input_tokens_seen": 1916360,
      "step": 1865,
      "train_runtime": 5284.5024,
      "train_tokens_per_second": 362.638
    },
    {
      "epoch": 0.7975211026819105,
      "grad_norm": 1.8512609004974365,
      "learning_rate": 4.0685224839400435e-06,
      "loss": 0.13702967762947083,
      "num_input_tokens_seen": 1917328,
      "step": 1866,
      "train_runtime": 5286.6307,
      "train_tokens_per_second": 362.675
    },
    {
      "epoch": 0.7979484987712362,
      "grad_norm": 2.444591760635376,
      "learning_rate": 4.059957173447538e-06,
      "loss": 0.1925363689661026,
      "num_input_tokens_seen": 1918666,
      "step": 1867,
      "train_runtime": 5288.8454,
      "train_tokens_per_second": 362.776
    },
    {
      "epoch": 0.798375894860562,
      "grad_norm": 1.9154953956604004,
      "learning_rate": 4.051391862955032e-06,
      "loss": 0.15239262580871582,
      "num_input_tokens_seen": 1919910,
      "step": 1868,
      "train_runtime": 5291.1085,
      "train_tokens_per_second": 362.856
    },
    {
      "epoch": 0.7988032909498878,
      "grad_norm": 1.6387301683425903,
      "learning_rate": 4.042826552462527e-06,
      "loss": 0.1101214736700058,
      "num_input_tokens_seen": 1921020,
      "step": 1869,
      "train_runtime": 5293.2649,
      "train_tokens_per_second": 362.918
    },
    {
      "epoch": 0.7992306870392136,
      "grad_norm": 1.7504992485046387,
      "learning_rate": 4.034261241970021e-06,
      "loss": 0.0942070484161377,
      "num_input_tokens_seen": 1922298,
      "step": 1870,
      "train_runtime": 5295.6018,
      "train_tokens_per_second": 362.999
    },
    {
      "epoch": 0.7996580831285394,
      "grad_norm": 2.4301817417144775,
      "learning_rate": 4.0256959314775165e-06,
      "loss": 0.16983138024806976,
      "num_input_tokens_seen": 1923092,
      "step": 1871,
      "train_runtime": 5297.6959,
      "train_tokens_per_second": 363.005
    },
    {
      "epoch": 0.8000854792178651,
      "grad_norm": 2.7977640628814697,
      "learning_rate": 4.017130620985011e-06,
      "loss": 0.09143347293138504,
      "num_input_tokens_seen": 1923818,
      "step": 1872,
      "train_runtime": 5299.7722,
      "train_tokens_per_second": 363.0
    },
    {
      "epoch": 0.8000854792178651,
      "eval_loss": 0.4577138423919678,
      "eval_runtime": 58.0766,
      "eval_samples_per_second": 17.184,
      "eval_steps_per_second": 8.592,
      "num_input_tokens_seen": 1923818,
      "step": 1872
    },
    {
      "epoch": 0.800512875307191,
      "grad_norm": 1.8073036670684814,
      "learning_rate": 4.008565310492505e-06,
      "loss": 0.10531668365001678,
      "num_input_tokens_seen": 1924722,
      "step": 1873,
      "train_runtime": 5360.023,
      "train_tokens_per_second": 359.088
    },
    {
      "epoch": 0.8009402713965167,
      "grad_norm": 2.991727113723755,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.09558672457933426,
      "num_input_tokens_seen": 1925368,
      "step": 1874,
      "train_runtime": 5362.1103,
      "train_tokens_per_second": 359.069
    },
    {
      "epoch": 0.8013676674858425,
      "grad_norm": 3.7627313137054443,
      "learning_rate": 3.991434689507495e-06,
      "loss": 0.18142566084861755,
      "num_input_tokens_seen": 1925972,
      "step": 1875,
      "train_runtime": 5364.1913,
      "train_tokens_per_second": 359.042
    },
    {
      "epoch": 0.8017950635751683,
      "grad_norm": 1.7993855476379395,
      "learning_rate": 3.9828693790149895e-06,
      "loss": 0.09175600856542587,
      "num_input_tokens_seen": 1927012,
      "step": 1876,
      "train_runtime": 5366.3385,
      "train_tokens_per_second": 359.093
    },
    {
      "epoch": 0.8022224596644941,
      "grad_norm": 2.058553695678711,
      "learning_rate": 3.974304068522484e-06,
      "loss": 0.10975615680217743,
      "num_input_tokens_seen": 1928134,
      "step": 1877,
      "train_runtime": 5368.52,
      "train_tokens_per_second": 359.156
    },
    {
      "epoch": 0.8026498557538199,
      "grad_norm": 1.576220989227295,
      "learning_rate": 3.965738758029979e-06,
      "loss": 0.06837090849876404,
      "num_input_tokens_seen": 1928898,
      "step": 1878,
      "train_runtime": 5370.6173,
      "train_tokens_per_second": 359.158
    },
    {
      "epoch": 0.8030772518431456,
      "grad_norm": 1.130545735359192,
      "learning_rate": 3.957173447537474e-06,
      "loss": 0.049123615026474,
      "num_input_tokens_seen": 1930108,
      "step": 1879,
      "train_runtime": 5372.8182,
      "train_tokens_per_second": 359.236
    },
    {
      "epoch": 0.8035046479324714,
      "grad_norm": 1.6341276168823242,
      "learning_rate": 3.948608137044968e-06,
      "loss": 0.09082270413637161,
      "num_input_tokens_seen": 1931350,
      "step": 1880,
      "train_runtime": 5375.0433,
      "train_tokens_per_second": 359.318
    },
    {
      "epoch": 0.8039320440217972,
      "grad_norm": 1.8335726261138916,
      "learning_rate": 3.9400428265524625e-06,
      "loss": 0.06868286430835724,
      "num_input_tokens_seen": 1932190,
      "step": 1881,
      "train_runtime": 5377.1592,
      "train_tokens_per_second": 359.333
    },
    {
      "epoch": 0.804359440111123,
      "grad_norm": 2.3648762702941895,
      "learning_rate": 3.931477516059958e-06,
      "loss": 0.19874761998653412,
      "num_input_tokens_seen": 1933088,
      "step": 1882,
      "train_runtime": 5379.2778,
      "train_tokens_per_second": 359.358
    },
    {
      "epoch": 0.8047868362004488,
      "grad_norm": 1.859014630317688,
      "learning_rate": 3.922912205567452e-06,
      "loss": 0.2102784812450409,
      "num_input_tokens_seen": 1935796,
      "step": 1883,
      "train_runtime": 5381.8483,
      "train_tokens_per_second": 359.69
    },
    {
      "epoch": 0.8052142322897745,
      "grad_norm": 2.4291601181030273,
      "learning_rate": 3.914346895074947e-06,
      "loss": 0.1437801718711853,
      "num_input_tokens_seen": 1936758,
      "step": 1884,
      "train_runtime": 5384.0132,
      "train_tokens_per_second": 359.724
    },
    {
      "epoch": 0.8056416283791004,
      "grad_norm": 3.8971290588378906,
      "learning_rate": 3.905781584582441e-06,
      "loss": 0.15766066312789917,
      "num_input_tokens_seen": 1937432,
      "step": 1885,
      "train_runtime": 5386.0703,
      "train_tokens_per_second": 359.712
    },
    {
      "epoch": 0.8060690244684261,
      "grad_norm": 2.332448720932007,
      "learning_rate": 3.897216274089936e-06,
      "loss": 0.1089940220117569,
      "num_input_tokens_seen": 1938202,
      "step": 1886,
      "train_runtime": 5388.2062,
      "train_tokens_per_second": 359.712
    },
    {
      "epoch": 0.806496420557752,
      "grad_norm": 2.3610904216766357,
      "learning_rate": 3.888650963597431e-06,
      "loss": 0.11087067425251007,
      "num_input_tokens_seen": 1939054,
      "step": 1887,
      "train_runtime": 5390.2944,
      "train_tokens_per_second": 359.731
    },
    {
      "epoch": 0.8069238166470777,
      "grad_norm": 1.648154377937317,
      "learning_rate": 3.880085653104925e-06,
      "loss": 0.08698208630084991,
      "num_input_tokens_seen": 1940562,
      "step": 1888,
      "train_runtime": 5392.5742,
      "train_tokens_per_second": 359.858
    },
    {
      "epoch": 0.8073512127364034,
      "grad_norm": 1.6215620040893555,
      "learning_rate": 3.87152034261242e-06,
      "loss": 0.09862314164638519,
      "num_input_tokens_seen": 1941642,
      "step": 1889,
      "train_runtime": 5394.7575,
      "train_tokens_per_second": 359.913
    },
    {
      "epoch": 0.8077786088257293,
      "grad_norm": 1.7662047147750854,
      "learning_rate": 3.862955032119914e-06,
      "loss": 0.11273685097694397,
      "num_input_tokens_seen": 1942750,
      "step": 1890,
      "train_runtime": 5396.9141,
      "train_tokens_per_second": 359.974
    },
    {
      "epoch": 0.808206004915055,
      "grad_norm": 1.47037672996521,
      "learning_rate": 3.854389721627409e-06,
      "loss": 0.08050605654716492,
      "num_input_tokens_seen": 1943888,
      "step": 1891,
      "train_runtime": 5412.0437,
      "train_tokens_per_second": 359.178
    },
    {
      "epoch": 0.8086334010043809,
      "grad_norm": 2.382103443145752,
      "learning_rate": 3.845824411134904e-06,
      "loss": 0.08305542171001434,
      "num_input_tokens_seen": 1944556,
      "step": 1892,
      "train_runtime": 5414.1535,
      "train_tokens_per_second": 359.162
    },
    {
      "epoch": 0.8090607970937066,
      "grad_norm": 3.229339361190796,
      "learning_rate": 3.837259100642399e-06,
      "loss": 0.08750642091035843,
      "num_input_tokens_seen": 1945294,
      "step": 1893,
      "train_runtime": 5416.2359,
      "train_tokens_per_second": 359.16
    },
    {
      "epoch": 0.8094881931830323,
      "grad_norm": 2.682495355606079,
      "learning_rate": 3.8286937901498935e-06,
      "loss": 0.16248637437820435,
      "num_input_tokens_seen": 1946162,
      "step": 1894,
      "train_runtime": 5418.3903,
      "train_tokens_per_second": 359.177
    },
    {
      "epoch": 0.8099155892723582,
      "grad_norm": 2.433453321456909,
      "learning_rate": 3.820128479657388e-06,
      "loss": 0.12479268014431,
      "num_input_tokens_seen": 1946952,
      "step": 1895,
      "train_runtime": 5420.4558,
      "train_tokens_per_second": 359.186
    },
    {
      "epoch": 0.8103429853616839,
      "grad_norm": 1.9988194704055786,
      "learning_rate": 3.8115631691648823e-06,
      "loss": 0.1043507307767868,
      "num_input_tokens_seen": 1947842,
      "step": 1896,
      "train_runtime": 5422.5794,
      "train_tokens_per_second": 359.209
    },
    {
      "epoch": 0.8107703814510098,
      "grad_norm": 3.0528292655944824,
      "learning_rate": 3.802997858672377e-06,
      "loss": 0.21311679482460022,
      "num_input_tokens_seen": 1948972,
      "step": 1897,
      "train_runtime": 5424.764,
      "train_tokens_per_second": 359.273
    },
    {
      "epoch": 0.8111977775403355,
      "grad_norm": 2.026343822479248,
      "learning_rate": 3.7944325481798716e-06,
      "loss": 0.07802553474903107,
      "num_input_tokens_seen": 1949854,
      "step": 1898,
      "train_runtime": 5426.8848,
      "train_tokens_per_second": 359.295
    },
    {
      "epoch": 0.8116251736296612,
      "grad_norm": 2.479752540588379,
      "learning_rate": 3.785867237687367e-06,
      "loss": 0.14262475073337555,
      "num_input_tokens_seen": 1950624,
      "step": 1899,
      "train_runtime": 5429.0351,
      "train_tokens_per_second": 359.295
    },
    {
      "epoch": 0.8120525697189871,
      "grad_norm": 1.9532415866851807,
      "learning_rate": 3.7773019271948613e-06,
      "loss": 0.11800507456064224,
      "num_input_tokens_seen": 1951642,
      "step": 1900,
      "train_runtime": 5431.2031,
      "train_tokens_per_second": 359.339
    },
    {
      "epoch": 0.8124799658083128,
      "grad_norm": 1.5504400730133057,
      "learning_rate": 3.7687366167023558e-06,
      "loss": 0.07501030713319778,
      "num_input_tokens_seen": 1952558,
      "step": 1901,
      "train_runtime": 5433.3849,
      "train_tokens_per_second": 359.363
    },
    {
      "epoch": 0.8129073618976387,
      "grad_norm": 1.9876056909561157,
      "learning_rate": 3.7601713062098506e-06,
      "loss": 0.14619901776313782,
      "num_input_tokens_seen": 1953702,
      "step": 1902,
      "train_runtime": 5435.6817,
      "train_tokens_per_second": 359.422
    },
    {
      "epoch": 0.8133347579869644,
      "grad_norm": 2.68910551071167,
      "learning_rate": 3.751605995717345e-06,
      "loss": 0.0984867513179779,
      "num_input_tokens_seen": 1954534,
      "step": 1903,
      "train_runtime": 5437.8178,
      "train_tokens_per_second": 359.434
    },
    {
      "epoch": 0.8137621540762902,
      "grad_norm": 3.058347702026367,
      "learning_rate": 3.7430406852248395e-06,
      "loss": 0.0885990709066391,
      "num_input_tokens_seen": 1955774,
      "step": 1904,
      "train_runtime": 5440.0504,
      "train_tokens_per_second": 359.514
    },
    {
      "epoch": 0.814189550165616,
      "grad_norm": 1.8561577796936035,
      "learning_rate": 3.7344753747323343e-06,
      "loss": 0.09684539586305618,
      "num_input_tokens_seen": 1956976,
      "step": 1905,
      "train_runtime": 5442.2427,
      "train_tokens_per_second": 359.59
    },
    {
      "epoch": 0.8146169462549417,
      "grad_norm": 2.643442392349243,
      "learning_rate": 3.7259100642398288e-06,
      "loss": 0.1766364425420761,
      "num_input_tokens_seen": 1957994,
      "step": 1906,
      "train_runtime": 5444.4087,
      "train_tokens_per_second": 359.634
    },
    {
      "epoch": 0.8150443423442676,
      "grad_norm": 2.634895086288452,
      "learning_rate": 3.717344753747323e-06,
      "loss": 0.1421157717704773,
      "num_input_tokens_seen": 1958802,
      "step": 1907,
      "train_runtime": 5446.536,
      "train_tokens_per_second": 359.642
    },
    {
      "epoch": 0.8154717384335933,
      "grad_norm": 1.751813530921936,
      "learning_rate": 3.7087794432548185e-06,
      "loss": 0.09633848071098328,
      "num_input_tokens_seen": 1960204,
      "step": 1908,
      "train_runtime": 5448.8213,
      "train_tokens_per_second": 359.748
    },
    {
      "epoch": 0.8158991345229191,
      "grad_norm": 1.63231360912323,
      "learning_rate": 3.7002141327623133e-06,
      "loss": 0.13464833796024323,
      "num_input_tokens_seen": 1961966,
      "step": 1909,
      "train_runtime": 5451.1577,
      "train_tokens_per_second": 359.917
    },
    {
      "epoch": 0.8163265306122449,
      "grad_norm": 1.863640308380127,
      "learning_rate": 3.6916488222698077e-06,
      "loss": 0.11301255226135254,
      "num_input_tokens_seen": 1963052,
      "step": 1910,
      "train_runtime": 5453.4013,
      "train_tokens_per_second": 359.968
    },
    {
      "epoch": 0.8167539267015707,
      "grad_norm": 3.2735648155212402,
      "learning_rate": 3.683083511777302e-06,
      "loss": 0.22603464126586914,
      "num_input_tokens_seen": 1963950,
      "step": 1911,
      "train_runtime": 5455.5012,
      "train_tokens_per_second": 359.994
    },
    {
      "epoch": 0.8171813227908965,
      "grad_norm": 1.9915646314620972,
      "learning_rate": 3.674518201284797e-06,
      "loss": 0.10916484892368317,
      "num_input_tokens_seen": 1964986,
      "step": 1912,
      "train_runtime": 5457.6465,
      "train_tokens_per_second": 360.043
    },
    {
      "epoch": 0.8176087188802222,
      "grad_norm": 2.7903940677642822,
      "learning_rate": 3.6659528907922915e-06,
      "loss": 0.16169843077659607,
      "num_input_tokens_seen": 1966226,
      "step": 1913,
      "train_runtime": 5459.9433,
      "train_tokens_per_second": 360.118
    },
    {
      "epoch": 0.818036114969548,
      "grad_norm": 2.9514753818511963,
      "learning_rate": 3.657387580299786e-06,
      "loss": 0.17525459825992584,
      "num_input_tokens_seen": 1967028,
      "step": 1914,
      "train_runtime": 5462.0781,
      "train_tokens_per_second": 360.124
    },
    {
      "epoch": 0.8184635110588738,
      "grad_norm": 2.1054723262786865,
      "learning_rate": 3.6488222698072807e-06,
      "loss": 0.13563019037246704,
      "num_input_tokens_seen": 1967974,
      "step": 1915,
      "train_runtime": 5464.2255,
      "train_tokens_per_second": 360.156
    },
    {
      "epoch": 0.8188909071481996,
      "grad_norm": 2.5213427543640137,
      "learning_rate": 3.640256959314775e-06,
      "loss": 0.14794062077999115,
      "num_input_tokens_seen": 1968914,
      "step": 1916,
      "train_runtime": 5466.3833,
      "train_tokens_per_second": 360.186
    },
    {
      "epoch": 0.8193183032375254,
      "grad_norm": 2.4679834842681885,
      "learning_rate": 3.6316916488222704e-06,
      "loss": 0.12253215163946152,
      "num_input_tokens_seen": 1969620,
      "step": 1917,
      "train_runtime": 5468.4742,
      "train_tokens_per_second": 360.177
    },
    {
      "epoch": 0.8197456993268512,
      "grad_norm": 1.776869535446167,
      "learning_rate": 3.623126338329765e-06,
      "loss": 0.1365637481212616,
      "num_input_tokens_seen": 1970726,
      "step": 1918,
      "train_runtime": 5470.6382,
      "train_tokens_per_second": 360.237
    },
    {
      "epoch": 0.8201730954161769,
      "grad_norm": 2.0742247104644775,
      "learning_rate": 3.6145610278372593e-06,
      "loss": 0.0735405758023262,
      "num_input_tokens_seen": 1971684,
      "step": 1919,
      "train_runtime": 5472.7816,
      "train_tokens_per_second": 360.271
    },
    {
      "epoch": 0.8206004915055027,
      "grad_norm": 2.545654773712158,
      "learning_rate": 3.605995717344754e-06,
      "loss": 0.14871400594711304,
      "num_input_tokens_seen": 1972462,
      "step": 1920,
      "train_runtime": 5474.8597,
      "train_tokens_per_second": 360.276
    },
    {
      "epoch": 0.8210278875948285,
      "grad_norm": 2.4054129123687744,
      "learning_rate": 3.5974304068522486e-06,
      "loss": 0.12270691245794296,
      "num_input_tokens_seen": 1973296,
      "step": 1921,
      "train_runtime": 5490.2025,
      "train_tokens_per_second": 359.421
    },
    {
      "epoch": 0.8214552836841543,
      "grad_norm": 3.4451773166656494,
      "learning_rate": 3.588865096359743e-06,
      "loss": 0.18050611019134521,
      "num_input_tokens_seen": 1974076,
      "step": 1922,
      "train_runtime": 5492.3176,
      "train_tokens_per_second": 359.425
    },
    {
      "epoch": 0.8218826797734801,
      "grad_norm": 1.751175880432129,
      "learning_rate": 3.580299785867238e-06,
      "loss": 0.07729603350162506,
      "num_input_tokens_seen": 1975028,
      "step": 1923,
      "train_runtime": 5494.4326,
      "train_tokens_per_second": 359.46
    },
    {
      "epoch": 0.8223100758628058,
      "grad_norm": 2.5249335765838623,
      "learning_rate": 3.5717344753747323e-06,
      "loss": 0.1707269698381424,
      "num_input_tokens_seen": 1975864,
      "step": 1924,
      "train_runtime": 5496.5142,
      "train_tokens_per_second": 359.476
    },
    {
      "epoch": 0.8227374719521316,
      "grad_norm": 2.40442156791687,
      "learning_rate": 3.5631691648822276e-06,
      "loss": 0.11481206864118576,
      "num_input_tokens_seen": 1976524,
      "step": 1925,
      "train_runtime": 5498.5753,
      "train_tokens_per_second": 359.461
    },
    {
      "epoch": 0.8231648680414574,
      "grad_norm": 2.9006872177124023,
      "learning_rate": 3.554603854389722e-06,
      "loss": 0.17662584781646729,
      "num_input_tokens_seen": 1977354,
      "step": 1926,
      "train_runtime": 5500.6917,
      "train_tokens_per_second": 359.474
    },
    {
      "epoch": 0.8235922641307832,
      "grad_norm": 1.889539122581482,
      "learning_rate": 3.546038543897217e-06,
      "loss": 0.13440728187561035,
      "num_input_tokens_seen": 1978830,
      "step": 1927,
      "train_runtime": 5503.0822,
      "train_tokens_per_second": 359.586
    },
    {
      "epoch": 0.824019660220109,
      "grad_norm": 2.9623191356658936,
      "learning_rate": 3.5374732334047113e-06,
      "loss": 0.11516458541154861,
      "num_input_tokens_seen": 1979576,
      "step": 1928,
      "train_runtime": 5505.2449,
      "train_tokens_per_second": 359.58
    },
    {
      "epoch": 0.8244470563094348,
      "grad_norm": 1.811000943183899,
      "learning_rate": 3.5289079229122057e-06,
      "loss": 0.15179359912872314,
      "num_input_tokens_seen": 1980932,
      "step": 1929,
      "train_runtime": 5507.4519,
      "train_tokens_per_second": 359.682
    },
    {
      "epoch": 0.8248744523987606,
      "grad_norm": 1.5050638914108276,
      "learning_rate": 3.5203426124197006e-06,
      "loss": 0.08498916029930115,
      "num_input_tokens_seen": 1982492,
      "step": 1930,
      "train_runtime": 5509.7553,
      "train_tokens_per_second": 359.815
    },
    {
      "epoch": 0.8253018484880863,
      "grad_norm": 2.768526077270508,
      "learning_rate": 3.511777301927195e-06,
      "loss": 0.19065095484256744,
      "num_input_tokens_seen": 1983864,
      "step": 1931,
      "train_runtime": 5511.9986,
      "train_tokens_per_second": 359.917
    },
    {
      "epoch": 0.8257292445774121,
      "grad_norm": 1.8834972381591797,
      "learning_rate": 3.5032119914346894e-06,
      "loss": 0.1330108344554901,
      "num_input_tokens_seen": 1985584,
      "step": 1932,
      "train_runtime": 5514.3368,
      "train_tokens_per_second": 360.077
    },
    {
      "epoch": 0.8261566406667379,
      "grad_norm": 2.301549196243286,
      "learning_rate": 3.4946466809421843e-06,
      "loss": 0.1241963803768158,
      "num_input_tokens_seen": 1986526,
      "step": 1933,
      "train_runtime": 5516.5107,
      "train_tokens_per_second": 360.106
    },
    {
      "epoch": 0.8265840367560637,
      "grad_norm": 2.294839382171631,
      "learning_rate": 3.486081370449679e-06,
      "loss": 0.09651590138673782,
      "num_input_tokens_seen": 1987252,
      "step": 1934,
      "train_runtime": 5518.6179,
      "train_tokens_per_second": 360.1
    },
    {
      "epoch": 0.8270114328453895,
      "grad_norm": 1.1531141996383667,
      "learning_rate": 3.477516059957174e-06,
      "loss": 0.04034721478819847,
      "num_input_tokens_seen": 1988562,
      "step": 1935,
      "train_runtime": 5520.8389,
      "train_tokens_per_second": 360.192
    },
    {
      "epoch": 0.8274388289347152,
      "grad_norm": 2.2977030277252197,
      "learning_rate": 3.4689507494646684e-06,
      "loss": 0.19051678478717804,
      "num_input_tokens_seen": 1989806,
      "step": 1936,
      "train_runtime": 5523.0563,
      "train_tokens_per_second": 360.273
    },
    {
      "epoch": 0.8278662250240411,
      "grad_norm": 1.8183130025863647,
      "learning_rate": 3.460385438972163e-06,
      "loss": 0.09287313371896744,
      "num_input_tokens_seen": 1990762,
      "step": 1937,
      "train_runtime": 5525.1933,
      "train_tokens_per_second": 360.306
    },
    {
      "epoch": 0.8282936211133668,
      "grad_norm": 1.7915457487106323,
      "learning_rate": 3.4518201284796577e-06,
      "loss": 0.11767736822366714,
      "num_input_tokens_seen": 1991992,
      "step": 1938,
      "train_runtime": 5527.3953,
      "train_tokens_per_second": 360.385
    },
    {
      "epoch": 0.8287210172026926,
      "grad_norm": 3.786351442337036,
      "learning_rate": 3.443254817987152e-06,
      "loss": 0.17700961232185364,
      "num_input_tokens_seen": 1993564,
      "step": 1939,
      "train_runtime": 5529.68,
      "train_tokens_per_second": 360.521
    },
    {
      "epoch": 0.8291484132920184,
      "grad_norm": 2.642591714859009,
      "learning_rate": 3.434689507494647e-06,
      "loss": 0.1346561759710312,
      "num_input_tokens_seen": 1994270,
      "step": 1940,
      "train_runtime": 5531.8105,
      "train_tokens_per_second": 360.509
    },
    {
      "epoch": 0.8295758093813441,
      "grad_norm": 3.820016622543335,
      "learning_rate": 3.4261241970021414e-06,
      "loss": 0.1595134437084198,
      "num_input_tokens_seen": 1995102,
      "step": 1941,
      "train_runtime": 5533.9409,
      "train_tokens_per_second": 360.521
    },
    {
      "epoch": 0.83000320547067,
      "grad_norm": 1.5686699151992798,
      "learning_rate": 3.4175588865096367e-06,
      "loss": 0.05472077429294586,
      "num_input_tokens_seen": 1996254,
      "step": 1942,
      "train_runtime": 5536.179,
      "train_tokens_per_second": 360.583
    },
    {
      "epoch": 0.8304306015599957,
      "grad_norm": 2.135924816131592,
      "learning_rate": 3.408993576017131e-06,
      "loss": 0.1282147765159607,
      "num_input_tokens_seen": 1997454,
      "step": 1943,
      "train_runtime": 5538.3609,
      "train_tokens_per_second": 360.658
    },
    {
      "epoch": 0.8308579976493216,
      "grad_norm": 2.1747469902038574,
      "learning_rate": 3.4004282655246255e-06,
      "loss": 0.08299367874860764,
      "num_input_tokens_seen": 1998452,
      "step": 1944,
      "train_runtime": 5540.5055,
      "train_tokens_per_second": 360.698
    },
    {
      "epoch": 0.8312853937386473,
      "grad_norm": 1.5561304092407227,
      "learning_rate": 3.3918629550321204e-06,
      "loss": 0.1040833592414856,
      "num_input_tokens_seen": 1999622,
      "step": 1945,
      "train_runtime": 5542.6835,
      "train_tokens_per_second": 360.768
    },
    {
      "epoch": 0.831712789827973,
      "grad_norm": 2.59633207321167,
      "learning_rate": 3.383297644539615e-06,
      "loss": 0.12334547191858292,
      "num_input_tokens_seen": 2000404,
      "step": 1946,
      "train_runtime": 5544.7677,
      "train_tokens_per_second": 360.773
    },
    {
      "epoch": 0.8321401859172989,
      "grad_norm": 2.10723876953125,
      "learning_rate": 3.3747323340471093e-06,
      "loss": 0.1401161551475525,
      "num_input_tokens_seen": 2001258,
      "step": 1947,
      "train_runtime": 5546.8846,
      "train_tokens_per_second": 360.79
    },
    {
      "epoch": 0.8325675820066246,
      "grad_norm": 2.798142671585083,
      "learning_rate": 3.366167023554604e-06,
      "loss": 0.2505949139595032,
      "num_input_tokens_seen": 2002180,
      "step": 1948,
      "train_runtime": 5549.0155,
      "train_tokens_per_second": 360.817
    },
    {
      "epoch": 0.8329949780959505,
      "grad_norm": 1.9741570949554443,
      "learning_rate": 3.3576017130620985e-06,
      "loss": 0.13103649020195007,
      "num_input_tokens_seen": 2003156,
      "step": 1949,
      "train_runtime": 5551.1248,
      "train_tokens_per_second": 360.856
    },
    {
      "epoch": 0.8334223741852762,
      "grad_norm": 2.568838119506836,
      "learning_rate": 3.349036402569593e-06,
      "loss": 0.08399426937103271,
      "num_input_tokens_seen": 2003902,
      "step": 1950,
      "train_runtime": 5553.1775,
      "train_tokens_per_second": 360.857
    },
    {
      "epoch": 0.8338497702746019,
      "grad_norm": 1.6990511417388916,
      "learning_rate": 3.3404710920770882e-06,
      "loss": 0.08365393429994583,
      "num_input_tokens_seen": 2004806,
      "step": 1951,
      "train_runtime": 5568.9756,
      "train_tokens_per_second": 359.995
    },
    {
      "epoch": 0.8342771663639278,
      "grad_norm": 3.738248109817505,
      "learning_rate": 3.3319057815845827e-06,
      "loss": 0.15475615859031677,
      "num_input_tokens_seen": 2005616,
      "step": 1952,
      "train_runtime": 5571.0983,
      "train_tokens_per_second": 360.004
    },
    {
      "epoch": 0.8347045624532535,
      "grad_norm": 1.6488298177719116,
      "learning_rate": 3.3233404710920775e-06,
      "loss": 0.07455479353666306,
      "num_input_tokens_seen": 2006428,
      "step": 1953,
      "train_runtime": 5573.2152,
      "train_tokens_per_second": 360.013
    },
    {
      "epoch": 0.8351319585425794,
      "grad_norm": 1.582184910774231,
      "learning_rate": 3.314775160599572e-06,
      "loss": 0.09235035628080368,
      "num_input_tokens_seen": 2007744,
      "step": 1954,
      "train_runtime": 5575.4356,
      "train_tokens_per_second": 360.105
    },
    {
      "epoch": 0.8355593546319051,
      "grad_norm": 1.3333709239959717,
      "learning_rate": 3.306209850107067e-06,
      "loss": 0.07065054774284363,
      "num_input_tokens_seen": 2008740,
      "step": 1955,
      "train_runtime": 5577.5599,
      "train_tokens_per_second": 360.147
    },
    {
      "epoch": 0.8359867507212309,
      "grad_norm": 4.119487762451172,
      "learning_rate": 3.2976445396145612e-06,
      "loss": 0.2738707661628723,
      "num_input_tokens_seen": 2009954,
      "step": 1956,
      "train_runtime": 5579.756,
      "train_tokens_per_second": 360.223
    },
    {
      "epoch": 0.8364141468105567,
      "grad_norm": 2.1321792602539062,
      "learning_rate": 3.2890792291220557e-06,
      "loss": 0.12480825930833817,
      "num_input_tokens_seen": 2010976,
      "step": 1957,
      "train_runtime": 5581.8989,
      "train_tokens_per_second": 360.267
    },
    {
      "epoch": 0.8368415428998824,
      "grad_norm": 2.774042844772339,
      "learning_rate": 3.2805139186295505e-06,
      "loss": 0.1823635995388031,
      "num_input_tokens_seen": 2011758,
      "step": 1958,
      "train_runtime": 5583.9833,
      "train_tokens_per_second": 360.273
    },
    {
      "epoch": 0.8372689389892083,
      "grad_norm": 2.230194091796875,
      "learning_rate": 3.2719486081370454e-06,
      "loss": 0.12755286693572998,
      "num_input_tokens_seen": 2012714,
      "step": 1959,
      "train_runtime": 5586.1008,
      "train_tokens_per_second": 360.307
    },
    {
      "epoch": 0.837696335078534,
      "grad_norm": 3.305572271347046,
      "learning_rate": 3.2633832976445402e-06,
      "loss": 0.13548028469085693,
      "num_input_tokens_seen": 2013502,
      "step": 1960,
      "train_runtime": 5588.2134,
      "train_tokens_per_second": 360.312
    },
    {
      "epoch": 0.8381237311678598,
      "grad_norm": 4.4020233154296875,
      "learning_rate": 3.2548179871520347e-06,
      "loss": 0.25181007385253906,
      "num_input_tokens_seen": 2014448,
      "step": 1961,
      "train_runtime": 5590.3795,
      "train_tokens_per_second": 360.342
    },
    {
      "epoch": 0.8385511272571856,
      "grad_norm": 1.438114881515503,
      "learning_rate": 3.246252676659529e-06,
      "loss": 0.09947221726179123,
      "num_input_tokens_seen": 2016488,
      "step": 1962,
      "train_runtime": 5592.8357,
      "train_tokens_per_second": 360.548
    },
    {
      "epoch": 0.8389785233465114,
      "grad_norm": 2.17861270904541,
      "learning_rate": 3.237687366167024e-06,
      "loss": 0.14272168278694153,
      "num_input_tokens_seen": 2017248,
      "step": 1963,
      "train_runtime": 5594.9139,
      "train_tokens_per_second": 360.55
    },
    {
      "epoch": 0.8394059194358372,
      "grad_norm": 2.0156331062316895,
      "learning_rate": 3.2291220556745184e-06,
      "loss": 0.06648299098014832,
      "num_input_tokens_seen": 2018118,
      "step": 1964,
      "train_runtime": 5597.1167,
      "train_tokens_per_second": 360.564
    },
    {
      "epoch": 0.8398333155251629,
      "grad_norm": 1.8698487281799316,
      "learning_rate": 3.220556745182013e-06,
      "loss": 0.1292165368795395,
      "num_input_tokens_seen": 2019514,
      "step": 1965,
      "train_runtime": 5599.3435,
      "train_tokens_per_second": 360.67
    },
    {
      "epoch": 0.8402607116144887,
      "grad_norm": 2.215479850769043,
      "learning_rate": 3.2119914346895077e-06,
      "loss": 0.15700754523277283,
      "num_input_tokens_seen": 2020222,
      "step": 1966,
      "train_runtime": 5601.5094,
      "train_tokens_per_second": 360.657
    },
    {
      "epoch": 0.8406881077038145,
      "grad_norm": 2.4758799076080322,
      "learning_rate": 3.203426124197002e-06,
      "loss": 0.1802935004234314,
      "num_input_tokens_seen": 2021242,
      "step": 1967,
      "train_runtime": 5603.6754,
      "train_tokens_per_second": 360.699
    },
    {
      "epoch": 0.8411155037931403,
      "grad_norm": 1.9556485414505005,
      "learning_rate": 3.1948608137044974e-06,
      "loss": 0.15662522614002228,
      "num_input_tokens_seen": 2022708,
      "step": 1968,
      "train_runtime": 5605.9581,
      "train_tokens_per_second": 360.814
    },
    {
      "epoch": 0.8415428998824661,
      "grad_norm": 1.730168104171753,
      "learning_rate": 3.186295503211992e-06,
      "loss": 0.08212494105100632,
      "num_input_tokens_seen": 2023718,
      "step": 1969,
      "train_runtime": 5608.0682,
      "train_tokens_per_second": 360.858
    },
    {
      "epoch": 0.8419702959717918,
      "grad_norm": 1.631124496459961,
      "learning_rate": 3.1777301927194866e-06,
      "loss": 0.11376766860485077,
      "num_input_tokens_seen": 2025154,
      "step": 1970,
      "train_runtime": 5610.2783,
      "train_tokens_per_second": 360.972
    },
    {
      "epoch": 0.8423976920611176,
      "grad_norm": 1.3239309787750244,
      "learning_rate": 3.169164882226981e-06,
      "loss": 0.08785051107406616,
      "num_input_tokens_seen": 2026996,
      "step": 1971,
      "train_runtime": 5612.6482,
      "train_tokens_per_second": 361.148
    },
    {
      "epoch": 0.8428250881504434,
      "grad_norm": 2.9183361530303955,
      "learning_rate": 3.1605995717344755e-06,
      "loss": 0.135522723197937,
      "num_input_tokens_seen": 2027706,
      "step": 1972,
      "train_runtime": 5614.6946,
      "train_tokens_per_second": 361.143
    },
    {
      "epoch": 0.8432524842397692,
      "grad_norm": 1.8889797925949097,
      "learning_rate": 3.1520342612419704e-06,
      "loss": 0.12959235906600952,
      "num_input_tokens_seen": 2028652,
      "step": 1973,
      "train_runtime": 5616.8256,
      "train_tokens_per_second": 361.174
    },
    {
      "epoch": 0.843679880329095,
      "grad_norm": 2.1443357467651367,
      "learning_rate": 3.1434689507494648e-06,
      "loss": 0.11393402516841888,
      "num_input_tokens_seen": 2029382,
      "step": 1974,
      "train_runtime": 5618.9072,
      "train_tokens_per_second": 361.17
    },
    {
      "epoch": 0.8441072764184208,
      "grad_norm": 1.8804446458816528,
      "learning_rate": 3.1349036402569592e-06,
      "loss": 0.09005725383758545,
      "num_input_tokens_seen": 2030350,
      "step": 1975,
      "train_runtime": 5621.0122,
      "train_tokens_per_second": 361.207
    },
    {
      "epoch": 0.8445346725077466,
      "grad_norm": 2.9914989471435547,
      "learning_rate": 3.126338329764454e-06,
      "loss": 0.1515377312898636,
      "num_input_tokens_seen": 2031440,
      "step": 1976,
      "train_runtime": 5623.1773,
      "train_tokens_per_second": 361.262
    },
    {
      "epoch": 0.8449620685970723,
      "grad_norm": 1.6609055995941162,
      "learning_rate": 3.117773019271949e-06,
      "loss": 0.09417612105607986,
      "num_input_tokens_seen": 2032804,
      "step": 1977,
      "train_runtime": 5625.3772,
      "train_tokens_per_second": 361.363
    },
    {
      "epoch": 0.8453894646863981,
      "grad_norm": 1.6197657585144043,
      "learning_rate": 3.1092077087794438e-06,
      "loss": 0.07073154300451279,
      "num_input_tokens_seen": 2033780,
      "step": 1978,
      "train_runtime": 5627.4811,
      "train_tokens_per_second": 361.401
    },
    {
      "epoch": 0.8458168607757239,
      "grad_norm": 1.8534493446350098,
      "learning_rate": 3.100642398286938e-06,
      "loss": 0.09149038791656494,
      "num_input_tokens_seen": 2034834,
      "step": 1979,
      "train_runtime": 5629.617,
      "train_tokens_per_second": 361.452
    },
    {
      "epoch": 0.8462442568650497,
      "grad_norm": 1.1798967123031616,
      "learning_rate": 3.0920770877944326e-06,
      "loss": 0.07325664907693863,
      "num_input_tokens_seen": 2036680,
      "step": 1980,
      "train_runtime": 5631.995,
      "train_tokens_per_second": 361.627
    },
    {
      "epoch": 0.8466716529543755,
      "grad_norm": 1.6230262517929077,
      "learning_rate": 3.0835117773019275e-06,
      "loss": 0.08121328800916672,
      "num_input_tokens_seen": 2039644,
      "step": 1981,
      "train_runtime": 5648.0993,
      "train_tokens_per_second": 361.12
    },
    {
      "epoch": 0.8470990490437013,
      "grad_norm": 1.933571457862854,
      "learning_rate": 3.074946466809422e-06,
      "loss": 0.11390797793865204,
      "num_input_tokens_seen": 2040578,
      "step": 1982,
      "train_runtime": 5650.1997,
      "train_tokens_per_second": 361.151
    },
    {
      "epoch": 0.847526445133027,
      "grad_norm": 2.425217866897583,
      "learning_rate": 3.0663811563169168e-06,
      "loss": 0.185648113489151,
      "num_input_tokens_seen": 2041474,
      "step": 1983,
      "train_runtime": 5652.3387,
      "train_tokens_per_second": 361.173
    },
    {
      "epoch": 0.8479538412223528,
      "grad_norm": 2.0624728202819824,
      "learning_rate": 3.057815845824411e-06,
      "loss": 0.09377612173557281,
      "num_input_tokens_seen": 2042394,
      "step": 1984,
      "train_runtime": 5654.5102,
      "train_tokens_per_second": 361.197
    },
    {
      "epoch": 0.8483812373116786,
      "grad_norm": 1.5336378812789917,
      "learning_rate": 3.0492505353319065e-06,
      "loss": 0.06787379831075668,
      "num_input_tokens_seen": 2043502,
      "step": 1985,
      "train_runtime": 5656.6937,
      "train_tokens_per_second": 361.254
    },
    {
      "epoch": 0.8488086334010044,
      "grad_norm": 1.3267147541046143,
      "learning_rate": 3.040685224839401e-06,
      "loss": 0.0803186297416687,
      "num_input_tokens_seen": 2044812,
      "step": 1986,
      "train_runtime": 5658.8834,
      "train_tokens_per_second": 361.345
    },
    {
      "epoch": 0.8492360294903302,
      "grad_norm": 2.1181395053863525,
      "learning_rate": 3.0321199143468953e-06,
      "loss": 0.14673547446727753,
      "num_input_tokens_seen": 2046186,
      "step": 1987,
      "train_runtime": 5661.1104,
      "train_tokens_per_second": 361.446
    },
    {
      "epoch": 0.8496634255796559,
      "grad_norm": 1.649976372718811,
      "learning_rate": 3.02355460385439e-06,
      "loss": 0.0759081244468689,
      "num_input_tokens_seen": 2047098,
      "step": 1988,
      "train_runtime": 5663.2299,
      "train_tokens_per_second": 361.472
    },
    {
      "epoch": 0.8500908216689818,
      "grad_norm": 2.7502894401550293,
      "learning_rate": 3.0149892933618846e-06,
      "loss": 0.13762880861759186,
      "num_input_tokens_seen": 2047874,
      "step": 1989,
      "train_runtime": 5665.3097,
      "train_tokens_per_second": 361.476
    },
    {
      "epoch": 0.8505182177583075,
      "grad_norm": 2.585418462753296,
      "learning_rate": 3.006423982869379e-06,
      "loss": 0.12399157881736755,
      "num_input_tokens_seen": 2048806,
      "step": 1990,
      "train_runtime": 5667.4303,
      "train_tokens_per_second": 361.505
    },
    {
      "epoch": 0.8509456138476333,
      "grad_norm": 2.424391508102417,
      "learning_rate": 2.997858672376874e-06,
      "loss": 0.13105568289756775,
      "num_input_tokens_seen": 2049882,
      "step": 1991,
      "train_runtime": 5669.606,
      "train_tokens_per_second": 361.556
    },
    {
      "epoch": 0.8513730099369591,
      "grad_norm": 2.692779779434204,
      "learning_rate": 2.9892933618843683e-06,
      "loss": 0.1243579089641571,
      "num_input_tokens_seen": 2050772,
      "step": 1992,
      "train_runtime": 5671.7248,
      "train_tokens_per_second": 361.578
    },
    {
      "epoch": 0.8518004060262848,
      "grad_norm": 1.901049017906189,
      "learning_rate": 2.9807280513918628e-06,
      "loss": 0.10940186679363251,
      "num_input_tokens_seen": 2051666,
      "step": 1993,
      "train_runtime": 5673.9587,
      "train_tokens_per_second": 361.593
    },
    {
      "epoch": 0.8522278021156107,
      "grad_norm": 2.4834864139556885,
      "learning_rate": 2.972162740899358e-06,
      "loss": 0.09155414253473282,
      "num_input_tokens_seen": 2052266,
      "step": 1994,
      "train_runtime": 5676.0733,
      "train_tokens_per_second": 361.564
    },
    {
      "epoch": 0.8526551982049364,
      "grad_norm": 1.8906586170196533,
      "learning_rate": 2.9635974304068525e-06,
      "loss": 0.12932683527469635,
      "num_input_tokens_seen": 2053472,
      "step": 1995,
      "train_runtime": 5678.3053,
      "train_tokens_per_second": 361.635
    },
    {
      "epoch": 0.8530825942942623,
      "grad_norm": 2.9802510738372803,
      "learning_rate": 2.9550321199143473e-06,
      "loss": 0.1381813883781433,
      "num_input_tokens_seen": 2054212,
      "step": 1996,
      "train_runtime": 5680.3911,
      "train_tokens_per_second": 361.632
    },
    {
      "epoch": 0.853509990383588,
      "grad_norm": 2.6141984462738037,
      "learning_rate": 2.9464668094218417e-06,
      "loss": 0.1583455502986908,
      "num_input_tokens_seen": 2055748,
      "step": 1997,
      "train_runtime": 5682.7069,
      "train_tokens_per_second": 361.755
    },
    {
      "epoch": 0.8539373864729137,
      "grad_norm": 1.8345046043395996,
      "learning_rate": 2.9379014989293366e-06,
      "loss": 0.07994143664836884,
      "num_input_tokens_seen": 2056428,
      "step": 1998,
      "train_runtime": 5684.7894,
      "train_tokens_per_second": 361.742
    },
    {
      "epoch": 0.8543647825622396,
      "grad_norm": 2.0560882091522217,
      "learning_rate": 2.929336188436831e-06,
      "loss": 0.10118348151445389,
      "num_input_tokens_seen": 2057386,
      "step": 1999,
      "train_runtime": 5686.9235,
      "train_tokens_per_second": 361.775
    },
    {
      "epoch": 0.8547921786515653,
      "grad_norm": 1.4362705945968628,
      "learning_rate": 2.9207708779443255e-06,
      "loss": 0.07624168694019318,
      "num_input_tokens_seen": 2058680,
      "step": 2000,
      "train_runtime": 5689.1043,
      "train_tokens_per_second": 361.864
    },
    {
      "epoch": 0.8552195747408912,
      "grad_norm": 0.8738347887992859,
      "learning_rate": 2.9122055674518203e-06,
      "loss": 0.035955920815467834,
      "num_input_tokens_seen": 2060918,
      "step": 2001,
      "train_runtime": 5691.6211,
      "train_tokens_per_second": 362.097
    },
    {
      "epoch": 0.8556469708302169,
      "grad_norm": 1.7071946859359741,
      "learning_rate": 2.903640256959315e-06,
      "loss": 0.043835729360580444,
      "num_input_tokens_seen": 2061456,
      "step": 2002,
      "train_runtime": 5693.6349,
      "train_tokens_per_second": 362.063
    },
    {
      "epoch": 0.8560743669195426,
      "grad_norm": 2.7414908409118652,
      "learning_rate": 2.89507494646681e-06,
      "loss": 0.10991692543029785,
      "num_input_tokens_seen": 2062118,
      "step": 2003,
      "train_runtime": 5695.7114,
      "train_tokens_per_second": 362.047
    },
    {
      "epoch": 0.8565017630088685,
      "grad_norm": 2.1727988719940186,
      "learning_rate": 2.8865096359743044e-06,
      "loss": 0.17784272134304047,
      "num_input_tokens_seen": 2063172,
      "step": 2004,
      "train_runtime": 5697.8479,
      "train_tokens_per_second": 362.097
    },
    {
      "epoch": 0.8569291590981942,
      "grad_norm": 1.9930412769317627,
      "learning_rate": 2.877944325481799e-06,
      "loss": 0.09805138409137726,
      "num_input_tokens_seen": 2064226,
      "step": 2005,
      "train_runtime": 5700.0082,
      "train_tokens_per_second": 362.144
    },
    {
      "epoch": 0.8573565551875201,
      "grad_norm": 2.333526849746704,
      "learning_rate": 2.8693790149892937e-06,
      "loss": 0.09461706876754761,
      "num_input_tokens_seen": 2064842,
      "step": 2006,
      "train_runtime": 5702.0422,
      "train_tokens_per_second": 362.123
    },
    {
      "epoch": 0.8577839512768458,
      "grad_norm": 2.2634665966033936,
      "learning_rate": 2.860813704496788e-06,
      "loss": 0.12293541431427002,
      "num_input_tokens_seen": 2065954,
      "step": 2007,
      "train_runtime": 5704.2333,
      "train_tokens_per_second": 362.179
    },
    {
      "epoch": 0.8582113473661716,
      "grad_norm": 1.9334956407546997,
      "learning_rate": 2.8522483940042826e-06,
      "loss": 0.13911227881908417,
      "num_input_tokens_seen": 2067154,
      "step": 2008,
      "train_runtime": 5706.4313,
      "train_tokens_per_second": 362.25
    },
    {
      "epoch": 0.8586387434554974,
      "grad_norm": 2.7667911052703857,
      "learning_rate": 2.8436830835117774e-06,
      "loss": 0.13113465905189514,
      "num_input_tokens_seen": 2067810,
      "step": 2009,
      "train_runtime": 5708.4672,
      "train_tokens_per_second": 362.236
    },
    {
      "epoch": 0.8590661395448231,
      "grad_norm": 1.5218790769577026,
      "learning_rate": 2.835117773019272e-06,
      "loss": 0.08512908220291138,
      "num_input_tokens_seen": 2068922,
      "step": 2010,
      "train_runtime": 5710.6242,
      "train_tokens_per_second": 362.293
    },
    {
      "epoch": 0.859493535634149,
      "grad_norm": 2.288893938064575,
      "learning_rate": 2.826552462526767e-06,
      "loss": 0.10153042525053024,
      "num_input_tokens_seen": 2069696,
      "step": 2011,
      "train_runtime": 5725.5078,
      "train_tokens_per_second": 361.487
    },
    {
      "epoch": 0.8599209317234747,
      "grad_norm": 2.8751845359802246,
      "learning_rate": 2.8179871520342616e-06,
      "loss": 0.17081843316555023,
      "num_input_tokens_seen": 2070610,
      "step": 2012,
      "train_runtime": 5727.7185,
      "train_tokens_per_second": 361.507
    },
    {
      "epoch": 0.8603483278128005,
      "grad_norm": 1.7815607786178589,
      "learning_rate": 2.8094218415417564e-06,
      "loss": 0.09744273126125336,
      "num_input_tokens_seen": 2071972,
      "step": 2013,
      "train_runtime": 5730.1163,
      "train_tokens_per_second": 361.593
    },
    {
      "epoch": 0.8607757239021263,
      "grad_norm": 2.4561080932617188,
      "learning_rate": 2.800856531049251e-06,
      "loss": 0.16175368428230286,
      "num_input_tokens_seen": 2072962,
      "step": 2014,
      "train_runtime": 5732.3061,
      "train_tokens_per_second": 361.628
    },
    {
      "epoch": 0.861203119991452,
      "grad_norm": 3.9444172382354736,
      "learning_rate": 2.7922912205567453e-06,
      "loss": 0.10620345175266266,
      "num_input_tokens_seen": 2073506,
      "step": 2015,
      "train_runtime": 5734.3842,
      "train_tokens_per_second": 361.592
    },
    {
      "epoch": 0.8616305160807779,
      "grad_norm": 2.4148569107055664,
      "learning_rate": 2.78372591006424e-06,
      "loss": 0.09299345314502716,
      "num_input_tokens_seen": 2074234,
      "step": 2016,
      "train_runtime": 5736.4849,
      "train_tokens_per_second": 361.586
    },
    {
      "epoch": 0.8620579121701036,
      "grad_norm": 1.7600783109664917,
      "learning_rate": 2.7751605995717346e-06,
      "loss": 0.08529353886842728,
      "num_input_tokens_seen": 2075424,
      "step": 2017,
      "train_runtime": 5738.7305,
      "train_tokens_per_second": 361.652
    },
    {
      "epoch": 0.8624853082594294,
      "grad_norm": 2.098557233810425,
      "learning_rate": 2.766595289079229e-06,
      "loss": 0.21334445476531982,
      "num_input_tokens_seen": 2076470,
      "step": 2018,
      "train_runtime": 5740.9149,
      "train_tokens_per_second": 361.697
    },
    {
      "epoch": 0.8629127043487552,
      "grad_norm": 2.6756327152252197,
      "learning_rate": 2.758029978586724e-06,
      "loss": 0.17590031027793884,
      "num_input_tokens_seen": 2077754,
      "step": 2019,
      "train_runtime": 5743.1473,
      "train_tokens_per_second": 361.78
    },
    {
      "epoch": 0.863340100438081,
      "grad_norm": 2.2054638862609863,
      "learning_rate": 2.7494646680942187e-06,
      "loss": 0.07914389669895172,
      "num_input_tokens_seen": 2078574,
      "step": 2020,
      "train_runtime": 5745.2632,
      "train_tokens_per_second": 361.789
    },
    {
      "epoch": 0.8637674965274068,
      "grad_norm": 1.9902969598770142,
      "learning_rate": 2.7408993576017136e-06,
      "loss": 0.13987253606319427,
      "num_input_tokens_seen": 2079704,
      "step": 2021,
      "train_runtime": 5747.446,
      "train_tokens_per_second": 361.848
    },
    {
      "epoch": 0.8641948926167325,
      "grad_norm": 1.2502907514572144,
      "learning_rate": 2.732334047109208e-06,
      "loss": 0.09385128319263458,
      "num_input_tokens_seen": 2081392,
      "step": 2022,
      "train_runtime": 5749.8268,
      "train_tokens_per_second": 361.992
    },
    {
      "epoch": 0.8646222887060584,
      "grad_norm": 3.1436736583709717,
      "learning_rate": 2.7237687366167024e-06,
      "loss": 0.17186886072158813,
      "num_input_tokens_seen": 2082108,
      "step": 2023,
      "train_runtime": 5751.9793,
      "train_tokens_per_second": 361.981
    },
    {
      "epoch": 0.8650496847953841,
      "grad_norm": 2.1285715103149414,
      "learning_rate": 2.7152034261241973e-06,
      "loss": 0.17323139309883118,
      "num_input_tokens_seen": 2083052,
      "step": 2024,
      "train_runtime": 5754.1864,
      "train_tokens_per_second": 362.006
    },
    {
      "epoch": 0.8654770808847099,
      "grad_norm": 2.813476800918579,
      "learning_rate": 2.7066381156316917e-06,
      "loss": 0.17243844270706177,
      "num_input_tokens_seen": 2084296,
      "step": 2025,
      "train_runtime": 5756.3708,
      "train_tokens_per_second": 362.085
    },
    {
      "epoch": 0.8659044769740357,
      "grad_norm": 2.0731918811798096,
      "learning_rate": 2.6980728051391865e-06,
      "loss": 0.11207694560289383,
      "num_input_tokens_seen": 2085072,
      "step": 2026,
      "train_runtime": 5758.426,
      "train_tokens_per_second": 362.091
    },
    {
      "epoch": 0.8663318730633615,
      "grad_norm": 1.3874374628067017,
      "learning_rate": 2.689507494646681e-06,
      "loss": 0.07603321969509125,
      "num_input_tokens_seen": 2086082,
      "step": 2027,
      "train_runtime": 5760.5695,
      "train_tokens_per_second": 362.131
    },
    {
      "epoch": 0.8667592691526873,
      "grad_norm": 1.777469515800476,
      "learning_rate": 2.6809421841541763e-06,
      "loss": 0.11113659292459488,
      "num_input_tokens_seen": 2087164,
      "step": 2028,
      "train_runtime": 5762.7462,
      "train_tokens_per_second": 362.182
    },
    {
      "epoch": 0.867186665242013,
      "grad_norm": 2.2070701122283936,
      "learning_rate": 2.6723768736616707e-06,
      "loss": 0.12488459795713425,
      "num_input_tokens_seen": 2087896,
      "step": 2029,
      "train_runtime": 5764.905,
      "train_tokens_per_second": 362.174
    },
    {
      "epoch": 0.8676140613313388,
      "grad_norm": 2.194129228591919,
      "learning_rate": 2.663811563169165e-06,
      "loss": 0.11793061345815659,
      "num_input_tokens_seen": 2088562,
      "step": 2030,
      "train_runtime": 5766.9669,
      "train_tokens_per_second": 362.16
    },
    {
      "epoch": 0.8680414574206646,
      "grad_norm": 2.392115354537964,
      "learning_rate": 2.65524625267666e-06,
      "loss": 0.14523430168628693,
      "num_input_tokens_seen": 2089798,
      "step": 2031,
      "train_runtime": 5769.2263,
      "train_tokens_per_second": 362.232
    },
    {
      "epoch": 0.8684688535099904,
      "grad_norm": 2.2562978267669678,
      "learning_rate": 2.6466809421841544e-06,
      "loss": 0.18294145166873932,
      "num_input_tokens_seen": 2090866,
      "step": 2032,
      "train_runtime": 5771.3728,
      "train_tokens_per_second": 362.282
    },
    {
      "epoch": 0.8688962495993162,
      "grad_norm": 1.5348066091537476,
      "learning_rate": 2.638115631691649e-06,
      "loss": 0.09927107393741608,
      "num_input_tokens_seen": 2092152,
      "step": 2033,
      "train_runtime": 5773.5838,
      "train_tokens_per_second": 362.366
    },
    {
      "epoch": 0.869323645688642,
      "grad_norm": 2.6012370586395264,
      "learning_rate": 2.6295503211991437e-06,
      "loss": 0.08113822340965271,
      "num_input_tokens_seen": 2092912,
      "step": 2034,
      "train_runtime": 5775.7421,
      "train_tokens_per_second": 362.362
    },
    {
      "epoch": 0.8697510417779677,
      "grad_norm": 2.6684658527374268,
      "learning_rate": 2.620985010706638e-06,
      "loss": 0.15928147733211517,
      "num_input_tokens_seen": 2093722,
      "step": 2035,
      "train_runtime": 5777.8638,
      "train_tokens_per_second": 362.37
    },
    {
      "epoch": 0.8701784378672935,
      "grad_norm": 1.9016945362091064,
      "learning_rate": 2.6124197002141325e-06,
      "loss": 0.08974956721067429,
      "num_input_tokens_seen": 2094804,
      "step": 2036,
      "train_runtime": 5780.013,
      "train_tokens_per_second": 362.422
    },
    {
      "epoch": 0.8706058339566193,
      "grad_norm": 2.4460599422454834,
      "learning_rate": 2.603854389721628e-06,
      "loss": 0.09791088104248047,
      "num_input_tokens_seen": 2095692,
      "step": 2037,
      "train_runtime": 5782.1424,
      "train_tokens_per_second": 362.442
    },
    {
      "epoch": 0.8710332300459451,
      "grad_norm": 2.352628231048584,
      "learning_rate": 2.5952890792291222e-06,
      "loss": 0.12310855090618134,
      "num_input_tokens_seen": 2096448,
      "step": 2038,
      "train_runtime": 5784.1989,
      "train_tokens_per_second": 362.444
    },
    {
      "epoch": 0.8714606261352709,
      "grad_norm": 2.515416383743286,
      "learning_rate": 2.586723768736617e-06,
      "loss": 0.14864176511764526,
      "num_input_tokens_seen": 2097430,
      "step": 2039,
      "train_runtime": 5786.3468,
      "train_tokens_per_second": 362.479
    },
    {
      "epoch": 0.8718880222245966,
      "grad_norm": 2.097531795501709,
      "learning_rate": 2.5781584582441115e-06,
      "loss": 0.08857254683971405,
      "num_input_tokens_seen": 2098438,
      "step": 2040,
      "train_runtime": 5788.5147,
      "train_tokens_per_second": 362.518
    },
    {
      "epoch": 0.8723154183139225,
      "grad_norm": 2.4108428955078125,
      "learning_rate": 2.5695931477516064e-06,
      "loss": 0.1783050000667572,
      "num_input_tokens_seen": 2099334,
      "step": 2041,
      "train_runtime": 5805.1412,
      "train_tokens_per_second": 361.634
    },
    {
      "epoch": 0.8727428144032482,
      "grad_norm": 2.102546453475952,
      "learning_rate": 2.561027837259101e-06,
      "loss": 0.0952841117978096,
      "num_input_tokens_seen": 2100336,
      "step": 2042,
      "train_runtime": 5807.3273,
      "train_tokens_per_second": 361.67
    },
    {
      "epoch": 0.873170210492574,
      "grad_norm": 3.1656851768493652,
      "learning_rate": 2.5524625267665952e-06,
      "loss": 0.15218988060951233,
      "num_input_tokens_seen": 2101246,
      "step": 2043,
      "train_runtime": 5809.5756,
      "train_tokens_per_second": 361.687
    },
    {
      "epoch": 0.8735976065818998,
      "grad_norm": 1.9376498460769653,
      "learning_rate": 2.54389721627409e-06,
      "loss": 0.09833593666553497,
      "num_input_tokens_seen": 2102154,
      "step": 2044,
      "train_runtime": 5811.7293,
      "train_tokens_per_second": 361.709
    },
    {
      "epoch": 0.8740250026712255,
      "grad_norm": 1.7766053676605225,
      "learning_rate": 2.535331905781585e-06,
      "loss": 0.12094634771347046,
      "num_input_tokens_seen": 2103916,
      "step": 2045,
      "train_runtime": 5814.0721,
      "train_tokens_per_second": 361.866
    },
    {
      "epoch": 0.8744523987605514,
      "grad_norm": 2.2929627895355225,
      "learning_rate": 2.52676659528908e-06,
      "loss": 0.13895580172538757,
      "num_input_tokens_seen": 2104866,
      "step": 2046,
      "train_runtime": 5816.2095,
      "train_tokens_per_second": 361.897
    },
    {
      "epoch": 0.8748797948498771,
      "grad_norm": 2.165609121322632,
      "learning_rate": 2.5182012847965742e-06,
      "loss": 0.15516400337219238,
      "num_input_tokens_seen": 2106214,
      "step": 2047,
      "train_runtime": 5818.4255,
      "train_tokens_per_second": 361.99
    },
    {
      "epoch": 0.875307190939203,
      "grad_norm": 3.682527542114258,
      "learning_rate": 2.5096359743040687e-06,
      "loss": 0.15300746262073517,
      "num_input_tokens_seen": 2107024,
      "step": 2048,
      "train_runtime": 5820.513,
      "train_tokens_per_second": 362.0
    },
    {
      "epoch": 0.8757345870285287,
      "grad_norm": 2.0996010303497314,
      "learning_rate": 2.5010706638115635e-06,
      "loss": 0.07466349005699158,
      "num_input_tokens_seen": 2107916,
      "step": 2049,
      "train_runtime": 5822.622,
      "train_tokens_per_second": 362.022
    },
    {
      "epoch": 0.8761619831178544,
      "grad_norm": 2.101146697998047,
      "learning_rate": 2.492505353319058e-06,
      "loss": 0.13168010115623474,
      "num_input_tokens_seen": 2109138,
      "step": 2050,
      "train_runtime": 5824.8222,
      "train_tokens_per_second": 362.095
    },
    {
      "epoch": 0.8765893792071803,
      "grad_norm": 2.264698028564453,
      "learning_rate": 2.4839400428265524e-06,
      "loss": 0.15204904973506927,
      "num_input_tokens_seen": 2110444,
      "step": 2051,
      "train_runtime": 5827.0412,
      "train_tokens_per_second": 362.181
    },
    {
      "epoch": 0.877016775296506,
      "grad_norm": 2.748321771621704,
      "learning_rate": 2.4753747323340472e-06,
      "loss": 0.12910622358322144,
      "num_input_tokens_seen": 2111194,
      "step": 2052,
      "train_runtime": 5829.0923,
      "train_tokens_per_second": 362.182
    },
    {
      "epoch": 0.8774441713858319,
      "grad_norm": 1.9361207485198975,
      "learning_rate": 2.466809421841542e-06,
      "loss": 0.0930027961730957,
      "num_input_tokens_seen": 2112414,
      "step": 2053,
      "train_runtime": 5831.2642,
      "train_tokens_per_second": 362.257
    },
    {
      "epoch": 0.8778715674751576,
      "grad_norm": 2.0167324542999268,
      "learning_rate": 2.4582441113490365e-06,
      "loss": 0.1363641321659088,
      "num_input_tokens_seen": 2113914,
      "step": 2054,
      "train_runtime": 5833.5618,
      "train_tokens_per_second": 362.371
    },
    {
      "epoch": 0.8782989635644833,
      "grad_norm": 2.5238311290740967,
      "learning_rate": 2.4496788008565314e-06,
      "loss": 0.13148850202560425,
      "num_input_tokens_seen": 2114710,
      "step": 2055,
      "train_runtime": 5835.6175,
      "train_tokens_per_second": 362.38
    },
    {
      "epoch": 0.8787263596538092,
      "grad_norm": 1.7928324937820435,
      "learning_rate": 2.441113490364026e-06,
      "loss": 0.15593746304512024,
      "num_input_tokens_seen": 2116414,
      "step": 2056,
      "train_runtime": 5837.9575,
      "train_tokens_per_second": 362.526
    },
    {
      "epoch": 0.8791537557431349,
      "grad_norm": 2.9624431133270264,
      "learning_rate": 2.4325481798715206e-06,
      "loss": 0.12336647510528564,
      "num_input_tokens_seen": 2117190,
      "step": 2057,
      "train_runtime": 5840.0547,
      "train_tokens_per_second": 362.529
    },
    {
      "epoch": 0.8795811518324608,
      "grad_norm": 2.7855894565582275,
      "learning_rate": 2.423982869379015e-06,
      "loss": 0.1867242157459259,
      "num_input_tokens_seen": 2118212,
      "step": 2058,
      "train_runtime": 5842.2307,
      "train_tokens_per_second": 362.569
    },
    {
      "epoch": 0.8800085479217865,
      "grad_norm": 1.6632304191589355,
      "learning_rate": 2.41541755888651e-06,
      "loss": 0.09385678172111511,
      "num_input_tokens_seen": 2119170,
      "step": 2059,
      "train_runtime": 5844.4019,
      "train_tokens_per_second": 362.598
    },
    {
      "epoch": 0.8804359440111122,
      "grad_norm": 2.4510445594787598,
      "learning_rate": 2.4068522483940048e-06,
      "loss": 0.0982072502374649,
      "num_input_tokens_seen": 2120156,
      "step": 2060,
      "train_runtime": 5846.5668,
      "train_tokens_per_second": 362.633
    },
    {
      "epoch": 0.8808633401004381,
      "grad_norm": 1.8894962072372437,
      "learning_rate": 2.398286937901499e-06,
      "loss": 0.1276012659072876,
      "num_input_tokens_seen": 2121670,
      "step": 2061,
      "train_runtime": 5848.8425,
      "train_tokens_per_second": 362.75
    },
    {
      "epoch": 0.8812907361897638,
      "grad_norm": 2.5933008193969727,
      "learning_rate": 2.3897216274089936e-06,
      "loss": 0.13273373246192932,
      "num_input_tokens_seen": 2122614,
      "step": 2062,
      "train_runtime": 5851.0156,
      "train_tokens_per_second": 362.777
    },
    {
      "epoch": 0.8817181322790897,
      "grad_norm": 3.338468551635742,
      "learning_rate": 2.3811563169164885e-06,
      "loss": 0.13261613249778748,
      "num_input_tokens_seen": 2123342,
      "step": 2063,
      "train_runtime": 5853.1235,
      "train_tokens_per_second": 362.771
    },
    {
      "epoch": 0.8821455283684154,
      "grad_norm": 1.8080017566680908,
      "learning_rate": 2.372591006423983e-06,
      "loss": 0.07965312898159027,
      "num_input_tokens_seen": 2124194,
      "step": 2064,
      "train_runtime": 5855.2708,
      "train_tokens_per_second": 362.783
    },
    {
      "epoch": 0.8825729244577412,
      "grad_norm": 1.7073616981506348,
      "learning_rate": 2.3640256959314778e-06,
      "loss": 0.09075456857681274,
      "num_input_tokens_seen": 2125412,
      "step": 2065,
      "train_runtime": 5857.4879,
      "train_tokens_per_second": 362.854
    },
    {
      "epoch": 0.883000320547067,
      "grad_norm": 1.811286211013794,
      "learning_rate": 2.355460385438972e-06,
      "loss": 0.07494605332612991,
      "num_input_tokens_seen": 2126608,
      "step": 2066,
      "train_runtime": 5859.7053,
      "train_tokens_per_second": 362.921
    },
    {
      "epoch": 0.8834277166363927,
      "grad_norm": 2.4246816635131836,
      "learning_rate": 2.346895074946467e-06,
      "loss": 0.12078673392534256,
      "num_input_tokens_seen": 2127588,
      "step": 2067,
      "train_runtime": 5861.8689,
      "train_tokens_per_second": 362.954
    },
    {
      "epoch": 0.8838551127257186,
      "grad_norm": 1.7475295066833496,
      "learning_rate": 2.3383297644539615e-06,
      "loss": 0.06864313781261444,
      "num_input_tokens_seen": 2128614,
      "step": 2068,
      "train_runtime": 5864.0657,
      "train_tokens_per_second": 362.993
    },
    {
      "epoch": 0.8842825088150443,
      "grad_norm": 2.9458775520324707,
      "learning_rate": 2.3297644539614563e-06,
      "loss": 0.16838330030441284,
      "num_input_tokens_seen": 2129810,
      "step": 2069,
      "train_runtime": 5866.2185,
      "train_tokens_per_second": 363.064
    },
    {
      "epoch": 0.8847099049043701,
      "grad_norm": 3.3116567134857178,
      "learning_rate": 2.321199143468951e-06,
      "loss": 0.2246849536895752,
      "num_input_tokens_seen": 2130448,
      "step": 2070,
      "train_runtime": 5868.278,
      "train_tokens_per_second": 363.045
    },
    {
      "epoch": 0.8851373009936959,
      "grad_norm": 1.917832374572754,
      "learning_rate": 2.3126338329764456e-06,
      "loss": 0.137008935213089,
      "num_input_tokens_seen": 2131852,
      "step": 2071,
      "train_runtime": 5883.8972,
      "train_tokens_per_second": 362.32
    },
    {
      "epoch": 0.8855646970830217,
      "grad_norm": 2.764021635055542,
      "learning_rate": 2.30406852248394e-06,
      "loss": 0.13553404808044434,
      "num_input_tokens_seen": 2132686,
      "step": 2072,
      "train_runtime": 5886.0294,
      "train_tokens_per_second": 362.33
    },
    {
      "epoch": 0.8859920931723475,
      "grad_norm": 2.148662567138672,
      "learning_rate": 2.295503211991435e-06,
      "loss": 0.17894351482391357,
      "num_input_tokens_seen": 2133802,
      "step": 2073,
      "train_runtime": 5888.2197,
      "train_tokens_per_second": 362.385
    },
    {
      "epoch": 0.8864194892616732,
      "grad_norm": 2.056614398956299,
      "learning_rate": 2.2869379014989297e-06,
      "loss": 0.1461656391620636,
      "num_input_tokens_seen": 2134710,
      "step": 2074,
      "train_runtime": 5890.3437,
      "train_tokens_per_second": 362.408
    },
    {
      "epoch": 0.8868468853509991,
      "grad_norm": 2.1697099208831787,
      "learning_rate": 2.278372591006424e-06,
      "loss": 0.10381367802619934,
      "num_input_tokens_seen": 2135374,
      "step": 2075,
      "train_runtime": 5892.4158,
      "train_tokens_per_second": 362.394
    },
    {
      "epoch": 0.8872742814403248,
      "grad_norm": 3.0922961235046387,
      "learning_rate": 2.2698072805139186e-06,
      "loss": 0.2057275027036667,
      "num_input_tokens_seen": 2136620,
      "step": 2076,
      "train_runtime": 5894.651,
      "train_tokens_per_second": 362.468
    },
    {
      "epoch": 0.8877016775296506,
      "grad_norm": 3.033801317214966,
      "learning_rate": 2.2612419700214135e-06,
      "loss": 0.13413895666599274,
      "num_input_tokens_seen": 2137722,
      "step": 2077,
      "train_runtime": 5896.8024,
      "train_tokens_per_second": 362.522
    },
    {
      "epoch": 0.8881290736189764,
      "grad_norm": 2.1379237174987793,
      "learning_rate": 2.2526766595289083e-06,
      "loss": 0.12404193729162216,
      "num_input_tokens_seen": 2138862,
      "step": 2078,
      "train_runtime": 5898.9962,
      "train_tokens_per_second": 362.581
    },
    {
      "epoch": 0.8885564697083022,
      "grad_norm": 2.2025508880615234,
      "learning_rate": 2.2441113490364027e-06,
      "loss": 0.10359278321266174,
      "num_input_tokens_seen": 2139738,
      "step": 2079,
      "train_runtime": 5901.116,
      "train_tokens_per_second": 362.599
    },
    {
      "epoch": 0.888983865797628,
      "grad_norm": 3.182241439819336,
      "learning_rate": 2.235546038543897e-06,
      "loss": 0.13087734580039978,
      "num_input_tokens_seen": 2140384,
      "step": 2080,
      "train_runtime": 5903.2227,
      "train_tokens_per_second": 362.579
    },
    {
      "epoch": 0.8894112618869537,
      "grad_norm": 2.1732473373413086,
      "learning_rate": 2.226980728051392e-06,
      "loss": 0.12075134366750717,
      "num_input_tokens_seen": 2141186,
      "step": 2081,
      "train_runtime": 5905.3209,
      "train_tokens_per_second": 362.586
    },
    {
      "epoch": 0.8898386579762795,
      "grad_norm": 2.0428032875061035,
      "learning_rate": 2.218415417558887e-06,
      "loss": 0.12998268008232117,
      "num_input_tokens_seen": 2142272,
      "step": 2082,
      "train_runtime": 5907.5715,
      "train_tokens_per_second": 362.632
    },
    {
      "epoch": 0.8902660540656053,
      "grad_norm": 2.5222203731536865,
      "learning_rate": 2.2098501070663813e-06,
      "loss": 0.10774858295917511,
      "num_input_tokens_seen": 2143428,
      "step": 2083,
      "train_runtime": 5909.7688,
      "train_tokens_per_second": 362.692
    },
    {
      "epoch": 0.8906934501549311,
      "grad_norm": 1.582076072692871,
      "learning_rate": 2.201284796573876e-06,
      "loss": 0.08669944107532501,
      "num_input_tokens_seen": 2144534,
      "step": 2084,
      "train_runtime": 5911.9353,
      "train_tokens_per_second": 362.747
    },
    {
      "epoch": 0.8911208462442569,
      "grad_norm": 1.802765965461731,
      "learning_rate": 2.1927194860813706e-06,
      "loss": 0.13840240240097046,
      "num_input_tokens_seen": 2146040,
      "step": 2085,
      "train_runtime": 5914.2524,
      "train_tokens_per_second": 362.859
    },
    {
      "epoch": 0.8915482423335827,
      "grad_norm": 2.2215256690979004,
      "learning_rate": 2.1841541755888654e-06,
      "loss": 0.12045235186815262,
      "num_input_tokens_seen": 2147140,
      "step": 2086,
      "train_runtime": 5916.4577,
      "train_tokens_per_second": 362.91
    },
    {
      "epoch": 0.8919756384229084,
      "grad_norm": 2.0863659381866455,
      "learning_rate": 2.17558886509636e-06,
      "loss": 0.1052878201007843,
      "num_input_tokens_seen": 2148050,
      "step": 2087,
      "train_runtime": 5918.5695,
      "train_tokens_per_second": 362.934
    },
    {
      "epoch": 0.8924030345122342,
      "grad_norm": 3.3657069206237793,
      "learning_rate": 2.1670235546038547e-06,
      "loss": 0.0845300480723381,
      "num_input_tokens_seen": 2148750,
      "step": 2088,
      "train_runtime": 5920.6723,
      "train_tokens_per_second": 362.923
    },
    {
      "epoch": 0.89283043060156,
      "grad_norm": 2.5540568828582764,
      "learning_rate": 2.158458244111349e-06,
      "loss": 0.14328941702842712,
      "num_input_tokens_seen": 2149570,
      "step": 2089,
      "train_runtime": 5922.7744,
      "train_tokens_per_second": 362.933
    },
    {
      "epoch": 0.8932578266908858,
      "grad_norm": 2.2871949672698975,
      "learning_rate": 2.149892933618844e-06,
      "loss": 0.07893916964530945,
      "num_input_tokens_seen": 2150544,
      "step": 2090,
      "train_runtime": 5924.9563,
      "train_tokens_per_second": 362.964
    },
    {
      "epoch": 0.8936852227802116,
      "grad_norm": 2.2579426765441895,
      "learning_rate": 2.1413276231263384e-06,
      "loss": 0.12150919437408447,
      "num_input_tokens_seen": 2151614,
      "step": 2091,
      "train_runtime": 5927.1907,
      "train_tokens_per_second": 363.007
    },
    {
      "epoch": 0.8941126188695373,
      "grad_norm": 2.3674373626708984,
      "learning_rate": 2.1327623126338333e-06,
      "loss": 0.11609025299549103,
      "num_input_tokens_seen": 2152512,
      "step": 2092,
      "train_runtime": 5929.3064,
      "train_tokens_per_second": 363.029
    },
    {
      "epoch": 0.8945400149588631,
      "grad_norm": 2.686967134475708,
      "learning_rate": 2.1241970021413277e-06,
      "loss": 0.09624900668859482,
      "num_input_tokens_seen": 2153204,
      "step": 2093,
      "train_runtime": 5931.3569,
      "train_tokens_per_second": 363.02
    },
    {
      "epoch": 0.8949674110481889,
      "grad_norm": 1.1433016061782837,
      "learning_rate": 2.115631691648822e-06,
      "loss": 0.05403666943311691,
      "num_input_tokens_seen": 2154586,
      "step": 2094,
      "train_runtime": 5933.6018,
      "train_tokens_per_second": 363.116
    },
    {
      "epoch": 0.8953948071375147,
      "grad_norm": 2.1384119987487793,
      "learning_rate": 2.107066381156317e-06,
      "loss": 0.10524041950702667,
      "num_input_tokens_seen": 2155250,
      "step": 2095,
      "train_runtime": 5935.6558,
      "train_tokens_per_second": 363.102
    },
    {
      "epoch": 0.8958222032268405,
      "grad_norm": 1.5513935089111328,
      "learning_rate": 2.098501070663812e-06,
      "loss": 0.06729548424482346,
      "num_input_tokens_seen": 2156172,
      "step": 2096,
      "train_runtime": 5937.7748,
      "train_tokens_per_second": 363.128
    },
    {
      "epoch": 0.8962495993161662,
      "grad_norm": 1.9078114032745361,
      "learning_rate": 2.0899357601713063e-06,
      "loss": 0.09925670176744461,
      "num_input_tokens_seen": 2157060,
      "step": 2097,
      "train_runtime": 5939.9,
      "train_tokens_per_second": 363.148
    },
    {
      "epoch": 0.8966769954054921,
      "grad_norm": 1.716143012046814,
      "learning_rate": 2.0813704496788007e-06,
      "loss": 0.09362021088600159,
      "num_input_tokens_seen": 2157940,
      "step": 2098,
      "train_runtime": 5942.0626,
      "train_tokens_per_second": 363.163
    },
    {
      "epoch": 0.8971043914948178,
      "grad_norm": 2.579413890838623,
      "learning_rate": 2.072805139186296e-06,
      "loss": 0.089250847697258,
      "num_input_tokens_seen": 2158944,
      "step": 2099,
      "train_runtime": 5944.2504,
      "train_tokens_per_second": 363.199
    },
    {
      "epoch": 0.8975317875841436,
      "grad_norm": 2.7344658374786377,
      "learning_rate": 2.0642398286937904e-06,
      "loss": 0.1683403104543686,
      "num_input_tokens_seen": 2159674,
      "step": 2100,
      "train_runtime": 5946.315,
      "train_tokens_per_second": 363.195
    },
    {
      "epoch": 0.8979591836734694,
      "grad_norm": 3.1770620346069336,
      "learning_rate": 2.055674518201285e-06,
      "loss": 0.2300439029932022,
      "num_input_tokens_seen": 2160366,
      "step": 2101,
      "train_runtime": 5961.5898,
      "train_tokens_per_second": 362.381
    },
    {
      "epoch": 0.8983865797627951,
      "grad_norm": 2.193706750869751,
      "learning_rate": 2.0471092077087797e-06,
      "loss": 0.08801782131195068,
      "num_input_tokens_seen": 2161154,
      "step": 2102,
      "train_runtime": 5963.6563,
      "train_tokens_per_second": 362.387
    },
    {
      "epoch": 0.898813975852121,
      "grad_norm": 2.410318613052368,
      "learning_rate": 2.0385438972162746e-06,
      "loss": 0.10770563781261444,
      "num_input_tokens_seen": 2161938,
      "step": 2103,
      "train_runtime": 5965.7593,
      "train_tokens_per_second": 362.391
    },
    {
      "epoch": 0.8992413719414467,
      "grad_norm": 2.9301934242248535,
      "learning_rate": 2.029978586723769e-06,
      "loss": 0.16462868452072144,
      "num_input_tokens_seen": 2162960,
      "step": 2104,
      "train_runtime": 5967.9477,
      "train_tokens_per_second": 362.429
    },
    {
      "epoch": 0.8996687680307726,
      "grad_norm": 2.930838108062744,
      "learning_rate": 2.0214132762312634e-06,
      "loss": 0.16791778802871704,
      "num_input_tokens_seen": 2163668,
      "step": 2105,
      "train_runtime": 5970.0715,
      "train_tokens_per_second": 362.419
    },
    {
      "epoch": 0.9000961641200983,
      "grad_norm": 3.144822597503662,
      "learning_rate": 2.0128479657387583e-06,
      "loss": 0.0992625430226326,
      "num_input_tokens_seen": 2164382,
      "step": 2106,
      "train_runtime": 5972.1522,
      "train_tokens_per_second": 362.412
    },
    {
      "epoch": 0.9000961641200983,
      "eval_loss": 0.4553227424621582,
      "eval_runtime": 58.2737,
      "eval_samples_per_second": 17.126,
      "eval_steps_per_second": 8.563,
      "num_input_tokens_seen": 2164382,
      "step": 2106
    },
    {
      "epoch": 0.900523560209424,
      "grad_norm": 2.130829095840454,
      "learning_rate": 2.0042826552462527e-06,
      "loss": 0.10893520712852478,
      "num_input_tokens_seen": 2165384,
      "step": 2107,
      "train_runtime": 6032.6155,
      "train_tokens_per_second": 358.946
    },
    {
      "epoch": 0.9009509562987499,
      "grad_norm": 2.5909454822540283,
      "learning_rate": 1.9957173447537475e-06,
      "loss": 0.1649809628725052,
      "num_input_tokens_seen": 2166104,
      "step": 2108,
      "train_runtime": 6034.7337,
      "train_tokens_per_second": 358.939
    },
    {
      "epoch": 0.9013783523880756,
      "grad_norm": 2.4314017295837402,
      "learning_rate": 1.987152034261242e-06,
      "loss": 0.1623801440000534,
      "num_input_tokens_seen": 2166962,
      "step": 2109,
      "train_runtime": 6036.8498,
      "train_tokens_per_second": 358.956
    },
    {
      "epoch": 0.9018057484774015,
      "grad_norm": 2.035533905029297,
      "learning_rate": 1.978586723768737e-06,
      "loss": 0.056259140372276306,
      "num_input_tokens_seen": 2167504,
      "step": 2110,
      "train_runtime": 6038.9074,
      "train_tokens_per_second": 358.923
    },
    {
      "epoch": 0.9022331445667272,
      "grad_norm": 4.308773994445801,
      "learning_rate": 1.9700214132762313e-06,
      "loss": 0.22878707945346832,
      "num_input_tokens_seen": 2168670,
      "step": 2111,
      "train_runtime": 6041.0586,
      "train_tokens_per_second": 358.988
    },
    {
      "epoch": 0.902660540656053,
      "grad_norm": 1.7481560707092285,
      "learning_rate": 1.961456102783726e-06,
      "loss": 0.08038550615310669,
      "num_input_tokens_seen": 2169614,
      "step": 2112,
      "train_runtime": 6043.2106,
      "train_tokens_per_second": 359.017
    },
    {
      "epoch": 0.9030879367453788,
      "grad_norm": 1.556742548942566,
      "learning_rate": 1.9528907922912205e-06,
      "loss": 0.09323370456695557,
      "num_input_tokens_seen": 2170860,
      "step": 2113,
      "train_runtime": 6045.4279,
      "train_tokens_per_second": 359.091
    },
    {
      "epoch": 0.9035153328347045,
      "grad_norm": 1.9363806247711182,
      "learning_rate": 1.9443254817987154e-06,
      "loss": 0.13133233785629272,
      "num_input_tokens_seen": 2172150,
      "step": 2114,
      "train_runtime": 6047.6584,
      "train_tokens_per_second": 359.172
    },
    {
      "epoch": 0.9039427289240304,
      "grad_norm": 1.996583342552185,
      "learning_rate": 1.93576017130621e-06,
      "loss": 0.14076834917068481,
      "num_input_tokens_seen": 2173248,
      "step": 2115,
      "train_runtime": 6049.8127,
      "train_tokens_per_second": 359.226
    },
    {
      "epoch": 0.9043701250133561,
      "grad_norm": 1.9341224431991577,
      "learning_rate": 1.9271948608137047e-06,
      "loss": 0.10415120422840118,
      "num_input_tokens_seen": 2174380,
      "step": 2116,
      "train_runtime": 6052.0045,
      "train_tokens_per_second": 359.283
    },
    {
      "epoch": 0.9047975211026819,
      "grad_norm": 2.2955281734466553,
      "learning_rate": 1.9186295503211995e-06,
      "loss": 0.09734087437391281,
      "num_input_tokens_seen": 2175414,
      "step": 2117,
      "train_runtime": 6054.1626,
      "train_tokens_per_second": 359.325
    },
    {
      "epoch": 0.9052249171920077,
      "grad_norm": 2.2073757648468018,
      "learning_rate": 1.910064239828694e-06,
      "loss": 0.13069772720336914,
      "num_input_tokens_seen": 2176874,
      "step": 2118,
      "train_runtime": 6056.3945,
      "train_tokens_per_second": 359.434
    },
    {
      "epoch": 0.9056523132813334,
      "grad_norm": 2.6759731769561768,
      "learning_rate": 1.9014989293361886e-06,
      "loss": 0.14985017478466034,
      "num_input_tokens_seen": 2177804,
      "step": 2119,
      "train_runtime": 6058.5202,
      "train_tokens_per_second": 359.461
    },
    {
      "epoch": 0.9060797093706593,
      "grad_norm": 1.9756344556808472,
      "learning_rate": 1.8929336188436835e-06,
      "loss": 0.11807208508253098,
      "num_input_tokens_seen": 2179206,
      "step": 2120,
      "train_runtime": 6060.7569,
      "train_tokens_per_second": 359.56
    },
    {
      "epoch": 0.906507105459985,
      "grad_norm": 1.886093020439148,
      "learning_rate": 1.8843683083511779e-06,
      "loss": 0.06790851801633835,
      "num_input_tokens_seen": 2179886,
      "step": 2121,
      "train_runtime": 6062.8167,
      "train_tokens_per_second": 359.55
    },
    {
      "epoch": 0.9069345015493109,
      "grad_norm": 2.142716884613037,
      "learning_rate": 1.8758029978586725e-06,
      "loss": 0.11490422487258911,
      "num_input_tokens_seen": 2180698,
      "step": 2122,
      "train_runtime": 6064.9343,
      "train_tokens_per_second": 359.558
    },
    {
      "epoch": 0.9073618976386366,
      "grad_norm": 2.057957172393799,
      "learning_rate": 1.8672376873661672e-06,
      "loss": 0.14492949843406677,
      "num_input_tokens_seen": 2181852,
      "step": 2123,
      "train_runtime": 6067.1151,
      "train_tokens_per_second": 359.619
    },
    {
      "epoch": 0.9077892937279624,
      "grad_norm": 1.9728516340255737,
      "learning_rate": 1.8586723768736616e-06,
      "loss": 0.12060727924108505,
      "num_input_tokens_seen": 2182806,
      "step": 2124,
      "train_runtime": 6069.3308,
      "train_tokens_per_second": 359.645
    },
    {
      "epoch": 0.9082166898172882,
      "grad_norm": 2.384751796722412,
      "learning_rate": 1.8501070663811567e-06,
      "loss": 0.16312731802463531,
      "num_input_tokens_seen": 2183504,
      "step": 2125,
      "train_runtime": 6071.411,
      "train_tokens_per_second": 359.637
    },
    {
      "epoch": 0.9086440859066139,
      "grad_norm": 2.389244318008423,
      "learning_rate": 1.841541755888651e-06,
      "loss": 0.17604269087314606,
      "num_input_tokens_seen": 2184810,
      "step": 2126,
      "train_runtime": 6073.6717,
      "train_tokens_per_second": 359.718
    },
    {
      "epoch": 0.9090714819959398,
      "grad_norm": 2.322256326675415,
      "learning_rate": 1.8329764453961457e-06,
      "loss": 0.14901700615882874,
      "num_input_tokens_seen": 2185814,
      "step": 2127,
      "train_runtime": 6075.8113,
      "train_tokens_per_second": 359.757
    },
    {
      "epoch": 0.9094988780852655,
      "grad_norm": 2.590334177017212,
      "learning_rate": 1.8244111349036404e-06,
      "loss": 0.12600906193256378,
      "num_input_tokens_seen": 2186656,
      "step": 2128,
      "train_runtime": 6077.9388,
      "train_tokens_per_second": 359.769
    },
    {
      "epoch": 0.9099262741745913,
      "grad_norm": 1.2655421495437622,
      "learning_rate": 1.8158458244111352e-06,
      "loss": 0.05039399117231369,
      "num_input_tokens_seen": 2187694,
      "step": 2129,
      "train_runtime": 6080.0871,
      "train_tokens_per_second": 359.813
    },
    {
      "epoch": 0.9103536702639171,
      "grad_norm": 1.8549717664718628,
      "learning_rate": 1.8072805139186297e-06,
      "loss": 0.08745712041854858,
      "num_input_tokens_seen": 2188778,
      "step": 2130,
      "train_runtime": 6082.3165,
      "train_tokens_per_second": 359.859
    },
    {
      "epoch": 0.9107810663532429,
      "grad_norm": 2.1436004638671875,
      "learning_rate": 1.7987152034261243e-06,
      "loss": 0.1523379385471344,
      "num_input_tokens_seen": 2189862,
      "step": 2131,
      "train_runtime": 6097.191,
      "train_tokens_per_second": 359.159
    },
    {
      "epoch": 0.9112084624425687,
      "grad_norm": 3.1262474060058594,
      "learning_rate": 1.790149892933619e-06,
      "loss": 0.129807710647583,
      "num_input_tokens_seen": 2190950,
      "step": 2132,
      "train_runtime": 6099.3588,
      "train_tokens_per_second": 359.21
    },
    {
      "epoch": 0.9116358585318944,
      "grad_norm": 2.2749571800231934,
      "learning_rate": 1.7815845824411138e-06,
      "loss": 0.09306280314922333,
      "num_input_tokens_seen": 2191624,
      "step": 2133,
      "train_runtime": 6101.429,
      "train_tokens_per_second": 359.198
    },
    {
      "epoch": 0.9120632546212202,
      "grad_norm": 3.626248598098755,
      "learning_rate": 1.7730192719486084e-06,
      "loss": 0.24941939115524292,
      "num_input_tokens_seen": 2192338,
      "step": 2134,
      "train_runtime": 6103.5822,
      "train_tokens_per_second": 359.189
    },
    {
      "epoch": 0.912490650710546,
      "grad_norm": 3.0691635608673096,
      "learning_rate": 1.7644539614561029e-06,
      "loss": 0.10945655405521393,
      "num_input_tokens_seen": 2193204,
      "step": 2135,
      "train_runtime": 6105.785,
      "train_tokens_per_second": 359.201
    },
    {
      "epoch": 0.9129180467998718,
      "grad_norm": 2.296405553817749,
      "learning_rate": 1.7558886509635975e-06,
      "loss": 0.10002892464399338,
      "num_input_tokens_seen": 2193862,
      "step": 2136,
      "train_runtime": 6107.8743,
      "train_tokens_per_second": 359.186
    },
    {
      "epoch": 0.9133454428891976,
      "grad_norm": 2.27190899848938,
      "learning_rate": 1.7473233404710921e-06,
      "loss": 0.1897697150707245,
      "num_input_tokens_seen": 2195292,
      "step": 2137,
      "train_runtime": 6110.1783,
      "train_tokens_per_second": 359.284
    },
    {
      "epoch": 0.9137728389785233,
      "grad_norm": 1.681108832359314,
      "learning_rate": 1.738758029978587e-06,
      "loss": 0.07039877772331238,
      "num_input_tokens_seen": 2196192,
      "step": 2138,
      "train_runtime": 6112.25,
      "train_tokens_per_second": 359.31
    },
    {
      "epoch": 0.9142002350678491,
      "grad_norm": 3.9387195110321045,
      "learning_rate": 1.7301927194860814e-06,
      "loss": 0.08656158298254013,
      "num_input_tokens_seen": 2197194,
      "step": 2139,
      "train_runtime": 6114.3863,
      "train_tokens_per_second": 359.348
    },
    {
      "epoch": 0.9146276311571749,
      "grad_norm": 2.0285916328430176,
      "learning_rate": 1.721627408993576e-06,
      "loss": 0.11406892538070679,
      "num_input_tokens_seen": 2198406,
      "step": 2140,
      "train_runtime": 6116.6422,
      "train_tokens_per_second": 359.414
    },
    {
      "epoch": 0.9150550272465007,
      "grad_norm": 2.407357931137085,
      "learning_rate": 1.7130620985010707e-06,
      "loss": 0.09661931544542313,
      "num_input_tokens_seen": 2199216,
      "step": 2141,
      "train_runtime": 6118.7678,
      "train_tokens_per_second": 359.421
    },
    {
      "epoch": 0.9154824233358265,
      "grad_norm": 2.445122241973877,
      "learning_rate": 1.7044967880085656e-06,
      "loss": 0.17017090320587158,
      "num_input_tokens_seen": 2200286,
      "step": 2142,
      "train_runtime": 6120.9937,
      "train_tokens_per_second": 359.465
    },
    {
      "epoch": 0.9159098194251523,
      "grad_norm": 1.547048807144165,
      "learning_rate": 1.6959314775160602e-06,
      "loss": 0.09920593351125717,
      "num_input_tokens_seen": 2202008,
      "step": 2143,
      "train_runtime": 6123.3724,
      "train_tokens_per_second": 359.607
    },
    {
      "epoch": 0.916337215514478,
      "grad_norm": 1.617870807647705,
      "learning_rate": 1.6873661670235546e-06,
      "loss": 0.08339440822601318,
      "num_input_tokens_seen": 2203588,
      "step": 2144,
      "train_runtime": 6125.6363,
      "train_tokens_per_second": 359.732
    },
    {
      "epoch": 0.9167646116038038,
      "grad_norm": 1.7286889553070068,
      "learning_rate": 1.6788008565310493e-06,
      "loss": 0.11313310265541077,
      "num_input_tokens_seen": 2205476,
      "step": 2145,
      "train_runtime": 6127.9915,
      "train_tokens_per_second": 359.902
    },
    {
      "epoch": 0.9171920076931296,
      "grad_norm": 1.7525434494018555,
      "learning_rate": 1.6702355460385441e-06,
      "loss": 0.15014159679412842,
      "num_input_tokens_seen": 2206608,
      "step": 2146,
      "train_runtime": 6130.2288,
      "train_tokens_per_second": 359.955
    },
    {
      "epoch": 0.9176194037824554,
      "grad_norm": 2.0443313121795654,
      "learning_rate": 1.6616702355460388e-06,
      "loss": 0.13414694368839264,
      "num_input_tokens_seen": 2208032,
      "step": 2147,
      "train_runtime": 6132.4952,
      "train_tokens_per_second": 360.054
    },
    {
      "epoch": 0.9180467998717812,
      "grad_norm": 1.9635981321334839,
      "learning_rate": 1.6531049250535334e-06,
      "loss": 0.1255420595407486,
      "num_input_tokens_seen": 2209026,
      "step": 2148,
      "train_runtime": 6134.6571,
      "train_tokens_per_second": 360.09
    },
    {
      "epoch": 0.9184741959611069,
      "grad_norm": 2.0642201900482178,
      "learning_rate": 1.6445396145610278e-06,
      "loss": 0.09459918737411499,
      "num_input_tokens_seen": 2209972,
      "step": 2149,
      "train_runtime": 6136.8829,
      "train_tokens_per_second": 360.113
    },
    {
      "epoch": 0.9189015920504328,
      "grad_norm": 1.7372931241989136,
      "learning_rate": 1.6359743040685227e-06,
      "loss": 0.09736011922359467,
      "num_input_tokens_seen": 2210930,
      "step": 2150,
      "train_runtime": 6139.0381,
      "train_tokens_per_second": 360.143
    },
    {
      "epoch": 0.9193289881397585,
      "grad_norm": 1.4585939645767212,
      "learning_rate": 1.6274089935760173e-06,
      "loss": 0.09436579048633575,
      "num_input_tokens_seen": 2212194,
      "step": 2151,
      "train_runtime": 6141.2223,
      "train_tokens_per_second": 360.22
    },
    {
      "epoch": 0.9197563842290843,
      "grad_norm": 2.0459043979644775,
      "learning_rate": 1.618843683083512e-06,
      "loss": 0.0974874496459961,
      "num_input_tokens_seen": 2213710,
      "step": 2152,
      "train_runtime": 6143.4894,
      "train_tokens_per_second": 360.334
    },
    {
      "epoch": 0.9201837803184101,
      "grad_norm": 2.7059218883514404,
      "learning_rate": 1.6102783725910064e-06,
      "loss": 0.11218500137329102,
      "num_input_tokens_seen": 2214468,
      "step": 2153,
      "train_runtime": 6145.5759,
      "train_tokens_per_second": 360.335
    },
    {
      "epoch": 0.9206111764077358,
      "grad_norm": 2.692477226257324,
      "learning_rate": 1.601713062098501e-06,
      "loss": 0.10696820914745331,
      "num_input_tokens_seen": 2215134,
      "step": 2154,
      "train_runtime": 6147.6204,
      "train_tokens_per_second": 360.324
    },
    {
      "epoch": 0.9210385724970617,
      "grad_norm": 1.9609453678131104,
      "learning_rate": 1.593147751605996e-06,
      "loss": 0.1086597591638565,
      "num_input_tokens_seen": 2216198,
      "step": 2155,
      "train_runtime": 6149.814,
      "train_tokens_per_second": 360.368
    },
    {
      "epoch": 0.9214659685863874,
      "grad_norm": 3.656251907348633,
      "learning_rate": 1.5845824411134905e-06,
      "loss": 0.12264759838581085,
      "num_input_tokens_seen": 2216962,
      "step": 2156,
      "train_runtime": 6151.9033,
      "train_tokens_per_second": 360.37
    },
    {
      "epoch": 0.9218933646757133,
      "grad_norm": 1.8970943689346313,
      "learning_rate": 1.5760171306209852e-06,
      "loss": 0.12525427341461182,
      "num_input_tokens_seen": 2218354,
      "step": 2157,
      "train_runtime": 6154.1869,
      "train_tokens_per_second": 360.463
    },
    {
      "epoch": 0.922320760765039,
      "grad_norm": 2.1047725677490234,
      "learning_rate": 1.5674518201284796e-06,
      "loss": 0.09674183279275894,
      "num_input_tokens_seen": 2219428,
      "step": 2158,
      "train_runtime": 6156.3285,
      "train_tokens_per_second": 360.512
    },
    {
      "epoch": 0.9227481568543647,
      "grad_norm": 1.7949364185333252,
      "learning_rate": 1.5588865096359745e-06,
      "loss": 0.09363202750682831,
      "num_input_tokens_seen": 2220526,
      "step": 2159,
      "train_runtime": 6158.5291,
      "train_tokens_per_second": 360.561
    },
    {
      "epoch": 0.9231755529436906,
      "grad_norm": 1.2715867757797241,
      "learning_rate": 1.550321199143469e-06,
      "loss": 0.0401017963886261,
      "num_input_tokens_seen": 2222110,
      "step": 2160,
      "train_runtime": 6160.9097,
      "train_tokens_per_second": 360.679
    },
    {
      "epoch": 0.9236029490330163,
      "grad_norm": 1.5644985437393188,
      "learning_rate": 1.5417558886509637e-06,
      "loss": 0.10436052829027176,
      "num_input_tokens_seen": 2223346,
      "step": 2161,
      "train_runtime": 6175.9247,
      "train_tokens_per_second": 360.002
    },
    {
      "epoch": 0.9240303451223422,
      "grad_norm": 2.236356019973755,
      "learning_rate": 1.5331905781584584e-06,
      "loss": 0.14996878802776337,
      "num_input_tokens_seen": 2224226,
      "step": 2162,
      "train_runtime": 6178.0318,
      "train_tokens_per_second": 360.022
    },
    {
      "epoch": 0.9244577412116679,
      "grad_norm": 1.6058436632156372,
      "learning_rate": 1.5246252676659532e-06,
      "loss": 0.07376152276992798,
      "num_input_tokens_seen": 2225198,
      "step": 2163,
      "train_runtime": 6180.2216,
      "train_tokens_per_second": 360.051
    },
    {
      "epoch": 0.9248851373009936,
      "grad_norm": 2.117514133453369,
      "learning_rate": 1.5160599571734477e-06,
      "loss": 0.07375172525644302,
      "num_input_tokens_seen": 2225796,
      "step": 2164,
      "train_runtime": 6182.3325,
      "train_tokens_per_second": 360.025
    },
    {
      "epoch": 0.9253125333903195,
      "grad_norm": 2.6839256286621094,
      "learning_rate": 1.5074946466809423e-06,
      "loss": 0.09135778993368149,
      "num_input_tokens_seen": 2226544,
      "step": 2165,
      "train_runtime": 6184.42,
      "train_tokens_per_second": 360.025
    },
    {
      "epoch": 0.9257399294796452,
      "grad_norm": 2.778290033340454,
      "learning_rate": 1.498929336188437e-06,
      "loss": 0.20278890430927277,
      "num_input_tokens_seen": 2227448,
      "step": 2166,
      "train_runtime": 6186.5296,
      "train_tokens_per_second": 360.048
    },
    {
      "epoch": 0.9261673255689711,
      "grad_norm": 1.92152738571167,
      "learning_rate": 1.4903640256959314e-06,
      "loss": 0.12577614188194275,
      "num_input_tokens_seen": 2229306,
      "step": 2167,
      "train_runtime": 6188.8938,
      "train_tokens_per_second": 360.211
    },
    {
      "epoch": 0.9265947216582968,
      "grad_norm": 1.8579258918762207,
      "learning_rate": 1.4817987152034262e-06,
      "loss": 0.14533375203609467,
      "num_input_tokens_seen": 2230718,
      "step": 2168,
      "train_runtime": 6191.3743,
      "train_tokens_per_second": 360.294
    },
    {
      "epoch": 0.9270221177476226,
      "grad_norm": 2.4688186645507812,
      "learning_rate": 1.4732334047109209e-06,
      "loss": 0.11172845959663391,
      "num_input_tokens_seen": 2231422,
      "step": 2169,
      "train_runtime": 6193.5658,
      "train_tokens_per_second": 360.281
    },
    {
      "epoch": 0.9274495138369484,
      "grad_norm": 1.9544367790222168,
      "learning_rate": 1.4646680942184155e-06,
      "loss": 0.08545652776956558,
      "num_input_tokens_seen": 2232340,
      "step": 2170,
      "train_runtime": 6195.6838,
      "train_tokens_per_second": 360.306
    },
    {
      "epoch": 0.9278769099262741,
      "grad_norm": 6.441534042358398,
      "learning_rate": 1.4561027837259102e-06,
      "loss": 0.16141225397586823,
      "num_input_tokens_seen": 2233880,
      "step": 2171,
      "train_runtime": 6197.9482,
      "train_tokens_per_second": 360.423
    },
    {
      "epoch": 0.9283043060156,
      "grad_norm": 2.048644781112671,
      "learning_rate": 1.447537473233405e-06,
      "loss": 0.08380210399627686,
      "num_input_tokens_seen": 2234980,
      "step": 2172,
      "train_runtime": 6200.1062,
      "train_tokens_per_second": 360.474
    },
    {
      "epoch": 0.9287317021049257,
      "grad_norm": 3.115220546722412,
      "learning_rate": 1.4389721627408994e-06,
      "loss": 0.17994233965873718,
      "num_input_tokens_seen": 2235902,
      "step": 2173,
      "train_runtime": 6202.2245,
      "train_tokens_per_second": 360.5
    },
    {
      "epoch": 0.9291590981942516,
      "grad_norm": 2.5104150772094727,
      "learning_rate": 1.430406852248394e-06,
      "loss": 0.10323172807693481,
      "num_input_tokens_seen": 2236840,
      "step": 2174,
      "train_runtime": 6204.3468,
      "train_tokens_per_second": 360.528
    },
    {
      "epoch": 0.9295864942835773,
      "grad_norm": 1.7411397695541382,
      "learning_rate": 1.4218415417558887e-06,
      "loss": 0.11698125302791595,
      "num_input_tokens_seen": 2238150,
      "step": 2175,
      "train_runtime": 6206.5923,
      "train_tokens_per_second": 360.609
    },
    {
      "epoch": 0.930013890372903,
      "grad_norm": 1.410210132598877,
      "learning_rate": 1.4132762312633836e-06,
      "loss": 0.09587608277797699,
      "num_input_tokens_seen": 2239322,
      "step": 2176,
      "train_runtime": 6208.7995,
      "train_tokens_per_second": 360.669
    },
    {
      "epoch": 0.9304412864622289,
      "grad_norm": 2.0049214363098145,
      "learning_rate": 1.4047109207708782e-06,
      "loss": 0.09810708463191986,
      "num_input_tokens_seen": 2240178,
      "step": 2177,
      "train_runtime": 6210.9299,
      "train_tokens_per_second": 360.683
    },
    {
      "epoch": 0.9308686825515546,
      "grad_norm": 2.021122455596924,
      "learning_rate": 1.3961456102783726e-06,
      "loss": 0.11581074446439743,
      "num_input_tokens_seen": 2241096,
      "step": 2178,
      "train_runtime": 6213.0862,
      "train_tokens_per_second": 360.706
    },
    {
      "epoch": 0.9312960786408805,
      "grad_norm": 2.5388717651367188,
      "learning_rate": 1.3875802997858673e-06,
      "loss": 0.15827855467796326,
      "num_input_tokens_seen": 2241958,
      "step": 2179,
      "train_runtime": 6215.206,
      "train_tokens_per_second": 360.721
    },
    {
      "epoch": 0.9317234747302062,
      "grad_norm": 1.569270133972168,
      "learning_rate": 1.379014989293362e-06,
      "loss": 0.07853906601667404,
      "num_input_tokens_seen": 2242900,
      "step": 2180,
      "train_runtime": 6217.2912,
      "train_tokens_per_second": 360.752
    },
    {
      "epoch": 0.932150870819532,
      "grad_norm": 2.9345450401306152,
      "learning_rate": 1.3704496788008568e-06,
      "loss": 0.14149774610996246,
      "num_input_tokens_seen": 2243884,
      "step": 2181,
      "train_runtime": 6219.4374,
      "train_tokens_per_second": 360.786
    },
    {
      "epoch": 0.9325782669088578,
      "grad_norm": 2.526463747024536,
      "learning_rate": 1.3618843683083512e-06,
      "loss": 0.11557620763778687,
      "num_input_tokens_seen": 2244708,
      "step": 2182,
      "train_runtime": 6221.55,
      "train_tokens_per_second": 360.796
    },
    {
      "epoch": 0.9330056629981835,
      "grad_norm": 2.2611868381500244,
      "learning_rate": 1.3533190578158458e-06,
      "loss": 0.11411857604980469,
      "num_input_tokens_seen": 2245648,
      "step": 2183,
      "train_runtime": 6223.6939,
      "train_tokens_per_second": 360.822
    },
    {
      "epoch": 0.9334330590875094,
      "grad_norm": 2.3254916667938232,
      "learning_rate": 1.3447537473233405e-06,
      "loss": 0.09124652296304703,
      "num_input_tokens_seen": 2246654,
      "step": 2184,
      "train_runtime": 6225.8453,
      "train_tokens_per_second": 360.859
    },
    {
      "epoch": 0.9338604551768351,
      "grad_norm": 2.4564313888549805,
      "learning_rate": 1.3361884368308353e-06,
      "loss": 0.14401093125343323,
      "num_input_tokens_seen": 2247496,
      "step": 2185,
      "train_runtime": 6227.9585,
      "train_tokens_per_second": 360.872
    },
    {
      "epoch": 0.9342878512661609,
      "grad_norm": 1.7827298641204834,
      "learning_rate": 1.32762312633833e-06,
      "loss": 0.09509570896625519,
      "num_input_tokens_seen": 2248292,
      "step": 2186,
      "train_runtime": 6230.0387,
      "train_tokens_per_second": 360.879
    },
    {
      "epoch": 0.9347152473554867,
      "grad_norm": 2.291931629180908,
      "learning_rate": 1.3190578158458244e-06,
      "loss": 0.1711120903491974,
      "num_input_tokens_seen": 2249290,
      "step": 2187,
      "train_runtime": 6232.1855,
      "train_tokens_per_second": 360.915
    },
    {
      "epoch": 0.9351426434448125,
      "grad_norm": 2.614119529724121,
      "learning_rate": 1.310492505353319e-06,
      "loss": 0.13640403747558594,
      "num_input_tokens_seen": 2250184,
      "step": 2188,
      "train_runtime": 6234.3052,
      "train_tokens_per_second": 360.936
    },
    {
      "epoch": 0.9355700395341383,
      "grad_norm": 1.8803365230560303,
      "learning_rate": 1.301927194860814e-06,
      "loss": 0.14787006378173828,
      "num_input_tokens_seen": 2251370,
      "step": 2189,
      "train_runtime": 6236.4828,
      "train_tokens_per_second": 361.0
    },
    {
      "epoch": 0.935997435623464,
      "grad_norm": 1.5116528272628784,
      "learning_rate": 1.2933618843683085e-06,
      "loss": 0.08433793485164642,
      "num_input_tokens_seen": 2252332,
      "step": 2190,
      "train_runtime": 6238.5978,
      "train_tokens_per_second": 361.032
    },
    {
      "epoch": 0.9364248317127898,
      "grad_norm": 2.7714686393737793,
      "learning_rate": 1.2847965738758032e-06,
      "loss": 0.09387312084436417,
      "num_input_tokens_seen": 2252888,
      "step": 2191,
      "train_runtime": 6253.9568,
      "train_tokens_per_second": 360.234
    },
    {
      "epoch": 0.9368522278021156,
      "grad_norm": 2.0574734210968018,
      "learning_rate": 1.2762312633832976e-06,
      "loss": 0.07626340538263321,
      "num_input_tokens_seen": 2254060,
      "step": 2192,
      "train_runtime": 6256.1293,
      "train_tokens_per_second": 360.296
    },
    {
      "epoch": 0.9372796238914414,
      "grad_norm": 2.5031521320343018,
      "learning_rate": 1.2676659528907925e-06,
      "loss": 0.19402307271957397,
      "num_input_tokens_seen": 2255156,
      "step": 2193,
      "train_runtime": 6258.3375,
      "train_tokens_per_second": 360.344
    },
    {
      "epoch": 0.9377070199807672,
      "grad_norm": 2.2177932262420654,
      "learning_rate": 1.2591006423982871e-06,
      "loss": 0.09253498911857605,
      "num_input_tokens_seen": 2256090,
      "step": 2194,
      "train_runtime": 6260.4569,
      "train_tokens_per_second": 360.371
    },
    {
      "epoch": 0.938134416070093,
      "grad_norm": 1.9994310140609741,
      "learning_rate": 1.2505353319057818e-06,
      "loss": 0.07868780195713043,
      "num_input_tokens_seen": 2256706,
      "step": 2195,
      "train_runtime": 6262.5403,
      "train_tokens_per_second": 360.35
    },
    {
      "epoch": 0.9385618121594187,
      "grad_norm": 2.661285638809204,
      "learning_rate": 1.2419700214132762e-06,
      "loss": 0.2149825394153595,
      "num_input_tokens_seen": 2257434,
      "step": 2196,
      "train_runtime": 6264.6755,
      "train_tokens_per_second": 360.343
    },
    {
      "epoch": 0.9389892082487445,
      "grad_norm": 2.630244493484497,
      "learning_rate": 1.233404710920771e-06,
      "loss": 0.1728074997663498,
      "num_input_tokens_seen": 2258554,
      "step": 2197,
      "train_runtime": 6266.8603,
      "train_tokens_per_second": 360.396
    },
    {
      "epoch": 0.9394166043380703,
      "grad_norm": 2.7600960731506348,
      "learning_rate": 1.2248394004282657e-06,
      "loss": 0.11109723150730133,
      "num_input_tokens_seen": 2259392,
      "step": 2198,
      "train_runtime": 6268.9457,
      "train_tokens_per_second": 360.41
    },
    {
      "epoch": 0.9398440004273961,
      "grad_norm": 1.9871920347213745,
      "learning_rate": 1.2162740899357603e-06,
      "loss": 0.1423594057559967,
      "num_input_tokens_seen": 2260566,
      "step": 2199,
      "train_runtime": 6271.1231,
      "train_tokens_per_second": 360.472
    },
    {
      "epoch": 0.9402713965167219,
      "grad_norm": 1.2453151941299438,
      "learning_rate": 1.207708779443255e-06,
      "loss": 0.08043745905160904,
      "num_input_tokens_seen": 2262194,
      "step": 2200,
      "train_runtime": 6273.4187,
      "train_tokens_per_second": 360.6
    },
    {
      "epoch": 0.9406987926060476,
      "grad_norm": 1.9367234706878662,
      "learning_rate": 1.1991434689507496e-06,
      "loss": 0.14686185121536255,
      "num_input_tokens_seen": 2263188,
      "step": 2201,
      "train_runtime": 6275.5482,
      "train_tokens_per_second": 360.636
    },
    {
      "epoch": 0.9411261886953735,
      "grad_norm": 1.8235828876495361,
      "learning_rate": 1.1905781584582442e-06,
      "loss": 0.07379249483346939,
      "num_input_tokens_seen": 2263944,
      "step": 2202,
      "train_runtime": 6277.7278,
      "train_tokens_per_second": 360.631
    },
    {
      "epoch": 0.9415535847846992,
      "grad_norm": 3.4732439517974854,
      "learning_rate": 1.1820128479657389e-06,
      "loss": 0.17212432622909546,
      "num_input_tokens_seen": 2264748,
      "step": 2203,
      "train_runtime": 6279.8993,
      "train_tokens_per_second": 360.634
    },
    {
      "epoch": 0.941980980874025,
      "grad_norm": 1.7110310792922974,
      "learning_rate": 1.1734475374732335e-06,
      "loss": 0.06450486183166504,
      "num_input_tokens_seen": 2265692,
      "step": 2204,
      "train_runtime": 6282.0373,
      "train_tokens_per_second": 360.662
    },
    {
      "epoch": 0.9424083769633508,
      "grad_norm": 2.265634536743164,
      "learning_rate": 1.1648822269807282e-06,
      "loss": 0.14018739759922028,
      "num_input_tokens_seen": 2266638,
      "step": 2205,
      "train_runtime": 6284.1744,
      "train_tokens_per_second": 360.69
    },
    {
      "epoch": 0.9428357730526765,
      "grad_norm": 1.6523348093032837,
      "learning_rate": 1.1563169164882228e-06,
      "loss": 0.07057573646306992,
      "num_input_tokens_seen": 2267848,
      "step": 2206,
      "train_runtime": 6286.3788,
      "train_tokens_per_second": 360.756
    },
    {
      "epoch": 0.9432631691420024,
      "grad_norm": 2.683692455291748,
      "learning_rate": 1.1477516059957174e-06,
      "loss": 0.11395564675331116,
      "num_input_tokens_seen": 2268764,
      "step": 2207,
      "train_runtime": 6288.5082,
      "train_tokens_per_second": 360.779
    },
    {
      "epoch": 0.9436905652313281,
      "grad_norm": 2.452878713607788,
      "learning_rate": 1.139186295503212e-06,
      "loss": 0.11282587796449661,
      "num_input_tokens_seen": 2269744,
      "step": 2208,
      "train_runtime": 6290.7039,
      "train_tokens_per_second": 360.809
    },
    {
      "epoch": 0.944117961320654,
      "grad_norm": 1.4998728036880493,
      "learning_rate": 1.1306209850107067e-06,
      "loss": 0.07952351868152618,
      "num_input_tokens_seen": 2270726,
      "step": 2209,
      "train_runtime": 6292.8252,
      "train_tokens_per_second": 360.844
    },
    {
      "epoch": 0.9445453574099797,
      "grad_norm": 1.8244997262954712,
      "learning_rate": 1.1220556745182014e-06,
      "loss": 0.10631820559501648,
      "num_input_tokens_seen": 2271722,
      "step": 2210,
      "train_runtime": 6294.9451,
      "train_tokens_per_second": 360.88
    },
    {
      "epoch": 0.9449727534993054,
      "grad_norm": 2.2881040573120117,
      "learning_rate": 1.113490364025696e-06,
      "loss": 0.15813925862312317,
      "num_input_tokens_seen": 2272812,
      "step": 2211,
      "train_runtime": 6297.1243,
      "train_tokens_per_second": 360.929
    },
    {
      "epoch": 0.9454001495886313,
      "grad_norm": 1.7550806999206543,
      "learning_rate": 1.1049250535331907e-06,
      "loss": 0.148911714553833,
      "num_input_tokens_seen": 2274198,
      "step": 2212,
      "train_runtime": 6299.3493,
      "train_tokens_per_second": 361.021
    },
    {
      "epoch": 0.945827545677957,
      "grad_norm": 3.03024959564209,
      "learning_rate": 1.0963597430406853e-06,
      "loss": 0.15858499705791473,
      "num_input_tokens_seen": 2275008,
      "step": 2213,
      "train_runtime": 6301.4308,
      "train_tokens_per_second": 361.03
    },
    {
      "epoch": 0.9462549417672829,
      "grad_norm": 2.3329412937164307,
      "learning_rate": 1.08779443254818e-06,
      "loss": 0.1374865621328354,
      "num_input_tokens_seen": 2275952,
      "step": 2214,
      "train_runtime": 6303.5432,
      "train_tokens_per_second": 361.059
    },
    {
      "epoch": 0.9466823378566086,
      "grad_norm": 1.7251203060150146,
      "learning_rate": 1.0792291220556746e-06,
      "loss": 0.11180976778268814,
      "num_input_tokens_seen": 2277670,
      "step": 2215,
      "train_runtime": 6305.829,
      "train_tokens_per_second": 361.201
    },
    {
      "epoch": 0.9471097339459343,
      "grad_norm": 2.2906625270843506,
      "learning_rate": 1.0706638115631692e-06,
      "loss": 0.11002357304096222,
      "num_input_tokens_seen": 2278466,
      "step": 2216,
      "train_runtime": 6307.9257,
      "train_tokens_per_second": 361.207
    },
    {
      "epoch": 0.9475371300352602,
      "grad_norm": 1.9278420209884644,
      "learning_rate": 1.0620985010706639e-06,
      "loss": 0.12338389456272125,
      "num_input_tokens_seen": 2279650,
      "step": 2217,
      "train_runtime": 6310.0843,
      "train_tokens_per_second": 361.271
    },
    {
      "epoch": 0.9479645261245859,
      "grad_norm": 2.5601234436035156,
      "learning_rate": 1.0535331905781585e-06,
      "loss": 0.12740349769592285,
      "num_input_tokens_seen": 2280504,
      "step": 2218,
      "train_runtime": 6312.1943,
      "train_tokens_per_second": 361.285
    },
    {
      "epoch": 0.9483919222139118,
      "grad_norm": 2.1307919025421143,
      "learning_rate": 1.0449678800856531e-06,
      "loss": 0.11471974849700928,
      "num_input_tokens_seen": 2281424,
      "step": 2219,
      "train_runtime": 6314.31,
      "train_tokens_per_second": 361.31
    },
    {
      "epoch": 0.9488193183032375,
      "grad_norm": 2.013857126235962,
      "learning_rate": 1.036402569593148e-06,
      "loss": 0.07748252153396606,
      "num_input_tokens_seen": 2282100,
      "step": 2220,
      "train_runtime": 6316.3658,
      "train_tokens_per_second": 361.3
    },
    {
      "epoch": 0.9492467143925634,
      "grad_norm": 2.6721012592315674,
      "learning_rate": 1.0278372591006424e-06,
      "loss": 0.13700446486473083,
      "num_input_tokens_seen": 2283032,
      "step": 2221,
      "train_runtime": 6332.1333,
      "train_tokens_per_second": 360.547
    },
    {
      "epoch": 0.9496741104818891,
      "grad_norm": 1.6228680610656738,
      "learning_rate": 1.0192719486081373e-06,
      "loss": 0.07596968859434128,
      "num_input_tokens_seen": 2284202,
      "step": 2222,
      "train_runtime": 6334.3453,
      "train_tokens_per_second": 360.606
    },
    {
      "epoch": 0.9501015065712148,
      "grad_norm": 1.7751622200012207,
      "learning_rate": 1.0107066381156317e-06,
      "loss": 0.056691914796829224,
      "num_input_tokens_seen": 2285314,
      "step": 2223,
      "train_runtime": 6336.5258,
      "train_tokens_per_second": 360.657
    },
    {
      "epoch": 0.9505289026605407,
      "grad_norm": 2.450216054916382,
      "learning_rate": 1.0021413276231263e-06,
      "loss": 0.13718266785144806,
      "num_input_tokens_seen": 2286416,
      "step": 2224,
      "train_runtime": 6338.744,
      "train_tokens_per_second": 360.705
    },
    {
      "epoch": 0.9509562987498664,
      "grad_norm": 1.9348093271255493,
      "learning_rate": 9.93576017130621e-07,
      "loss": 0.11827270686626434,
      "num_input_tokens_seen": 2287386,
      "step": 2225,
      "train_runtime": 6340.8871,
      "train_tokens_per_second": 360.736
    },
    {
      "epoch": 0.9513836948391923,
      "grad_norm": 2.0023467540740967,
      "learning_rate": 9.850107066381156e-07,
      "loss": 0.11400140076875687,
      "num_input_tokens_seen": 2288430,
      "step": 2226,
      "train_runtime": 6343.1042,
      "train_tokens_per_second": 360.774
    },
    {
      "epoch": 0.951811090928518,
      "grad_norm": 2.965432643890381,
      "learning_rate": 9.764453961456103e-07,
      "loss": 0.19582927227020264,
      "num_input_tokens_seen": 2289104,
      "step": 2227,
      "train_runtime": 6345.1677,
      "train_tokens_per_second": 360.763
    },
    {
      "epoch": 0.9522384870178437,
      "grad_norm": 1.784056544303894,
      "learning_rate": 9.67880085653105e-07,
      "loss": 0.11103382706642151,
      "num_input_tokens_seen": 2290770,
      "step": 2228,
      "train_runtime": 6347.5155,
      "train_tokens_per_second": 360.892
    },
    {
      "epoch": 0.9526658831071696,
      "grad_norm": 1.949532389640808,
      "learning_rate": 9.593147751605998e-07,
      "loss": 0.10877189040184021,
      "num_input_tokens_seen": 2292226,
      "step": 2229,
      "train_runtime": 6349.7668,
      "train_tokens_per_second": 360.994
    },
    {
      "epoch": 0.9530932791964953,
      "grad_norm": 1.9729379415512085,
      "learning_rate": 9.507494646680943e-07,
      "loss": 0.11898594349622726,
      "num_input_tokens_seen": 2293084,
      "step": 2230,
      "train_runtime": 6351.897,
      "train_tokens_per_second": 361.008
    },
    {
      "epoch": 0.9535206752858212,
      "grad_norm": 1.3747807741165161,
      "learning_rate": 9.421841541755889e-07,
      "loss": 0.08915519714355469,
      "num_input_tokens_seen": 2294576,
      "step": 2231,
      "train_runtime": 6354.3369,
      "train_tokens_per_second": 361.104
    },
    {
      "epoch": 0.9539480713751469,
      "grad_norm": 2.2704765796661377,
      "learning_rate": 9.336188436830836e-07,
      "loss": 0.1116451770067215,
      "num_input_tokens_seen": 2295456,
      "step": 2232,
      "train_runtime": 6356.4698,
      "train_tokens_per_second": 361.121
    },
    {
      "epoch": 0.9543754674644727,
      "grad_norm": 2.8969857692718506,
      "learning_rate": 9.250535331905783e-07,
      "loss": 0.11565807461738586,
      "num_input_tokens_seen": 2296482,
      "step": 2233,
      "train_runtime": 6358.5973,
      "train_tokens_per_second": 361.162
    },
    {
      "epoch": 0.9548028635537985,
      "grad_norm": 2.5981929302215576,
      "learning_rate": 9.164882226980729e-07,
      "loss": 0.11579366028308868,
      "num_input_tokens_seen": 2297264,
      "step": 2234,
      "train_runtime": 6360.6838,
      "train_tokens_per_second": 361.166
    },
    {
      "epoch": 0.9552302596431242,
      "grad_norm": 2.0432050228118896,
      "learning_rate": 9.079229122055676e-07,
      "loss": 0.18046244978904724,
      "num_input_tokens_seen": 2298784,
      "step": 2235,
      "train_runtime": 6363.0184,
      "train_tokens_per_second": 361.273
    },
    {
      "epoch": 0.9556576557324501,
      "grad_norm": 1.764160394668579,
      "learning_rate": 8.993576017130621e-07,
      "loss": 0.07814285904169083,
      "num_input_tokens_seen": 2299798,
      "step": 2236,
      "train_runtime": 6365.1335,
      "train_tokens_per_second": 361.312
    },
    {
      "epoch": 0.9560850518217758,
      "grad_norm": 1.0686864852905273,
      "learning_rate": 8.907922912205569e-07,
      "loss": 0.04905345290899277,
      "num_input_tokens_seen": 2301808,
      "step": 2237,
      "train_runtime": 6367.5439,
      "train_tokens_per_second": 361.491
    },
    {
      "epoch": 0.9565124479111016,
      "grad_norm": 1.8639577627182007,
      "learning_rate": 8.822269807280514e-07,
      "loss": 0.07290835678577423,
      "num_input_tokens_seen": 2302694,
      "step": 2238,
      "train_runtime": 6369.6333,
      "train_tokens_per_second": 361.511
    },
    {
      "epoch": 0.9569398440004274,
      "grad_norm": 2.1470258235931396,
      "learning_rate": 8.736616702355461e-07,
      "loss": 0.16135422885417938,
      "num_input_tokens_seen": 2303744,
      "step": 2239,
      "train_runtime": 6371.7763,
      "train_tokens_per_second": 361.554
    },
    {
      "epoch": 0.9573672400897532,
      "grad_norm": 2.3782074451446533,
      "learning_rate": 8.650963597430407e-07,
      "loss": 0.1434607356786728,
      "num_input_tokens_seen": 2304580,
      "step": 2240,
      "train_runtime": 6374.0615,
      "train_tokens_per_second": 361.556
    },
    {
      "epoch": 0.957794636179079,
      "grad_norm": 2.3785908222198486,
      "learning_rate": 8.565310492505354e-07,
      "loss": 0.11133638024330139,
      "num_input_tokens_seen": 2305478,
      "step": 2241,
      "train_runtime": 6376.155,
      "train_tokens_per_second": 361.578
    },
    {
      "epoch": 0.9582220322684047,
      "grad_norm": 1.9879435300827026,
      "learning_rate": 8.479657387580301e-07,
      "loss": 0.09714030474424362,
      "num_input_tokens_seen": 2306700,
      "step": 2242,
      "train_runtime": 6378.3598,
      "train_tokens_per_second": 361.645
    },
    {
      "epoch": 0.9586494283577305,
      "grad_norm": 1.6205159425735474,
      "learning_rate": 8.394004282655246e-07,
      "loss": 0.10126176476478577,
      "num_input_tokens_seen": 2307916,
      "step": 2243,
      "train_runtime": 6380.6504,
      "train_tokens_per_second": 361.705
    },
    {
      "epoch": 0.9590768244470563,
      "grad_norm": 2.169267416000366,
      "learning_rate": 8.308351177730194e-07,
      "loss": 0.09987229853868484,
      "num_input_tokens_seen": 2308914,
      "step": 2244,
      "train_runtime": 6382.7974,
      "train_tokens_per_second": 361.74
    },
    {
      "epoch": 0.9595042205363821,
      "grad_norm": 2.731926918029785,
      "learning_rate": 8.222698072805139e-07,
      "loss": 0.14484712481498718,
      "num_input_tokens_seen": 2309942,
      "step": 2245,
      "train_runtime": 6385.0203,
      "train_tokens_per_second": 361.775
    },
    {
      "epoch": 0.9599316166257079,
      "grad_norm": 2.236884832382202,
      "learning_rate": 8.137044967880087e-07,
      "loss": 0.1355789750814438,
      "num_input_tokens_seen": 2311726,
      "step": 2246,
      "train_runtime": 6387.5251,
      "train_tokens_per_second": 361.913
    },
    {
      "epoch": 0.9603590127150337,
      "grad_norm": 1.772127628326416,
      "learning_rate": 8.051391862955032e-07,
      "loss": 0.13237515091896057,
      "num_input_tokens_seen": 2313058,
      "step": 2247,
      "train_runtime": 6389.7488,
      "train_tokens_per_second": 361.995
    },
    {
      "epoch": 0.9607864088043594,
      "grad_norm": 1.7337534427642822,
      "learning_rate": 7.96573875802998e-07,
      "loss": 0.0806911364197731,
      "num_input_tokens_seen": 2313840,
      "step": 2248,
      "train_runtime": 6391.8282,
      "train_tokens_per_second": 362.0
    },
    {
      "epoch": 0.9612138048936852,
      "grad_norm": 2.1230995655059814,
      "learning_rate": 7.880085653104926e-07,
      "loss": 0.07250198721885681,
      "num_input_tokens_seen": 2314458,
      "step": 2249,
      "train_runtime": 6393.8912,
      "train_tokens_per_second": 361.98
    },
    {
      "epoch": 0.961641200983011,
      "grad_norm": 2.5727736949920654,
      "learning_rate": 7.794432548179872e-07,
      "loss": 0.1578446328639984,
      "num_input_tokens_seen": 2315338,
      "step": 2250,
      "train_runtime": 6396.012,
      "train_tokens_per_second": 361.997
    },
    {
      "epoch": 0.9620685970723368,
      "grad_norm": 1.7777867317199707,
      "learning_rate": 7.708779443254819e-07,
      "loss": 0.08421089500188828,
      "num_input_tokens_seen": 2316192,
      "step": 2251,
      "train_runtime": 6411.3235,
      "train_tokens_per_second": 361.266
    },
    {
      "epoch": 0.9624959931616626,
      "grad_norm": 2.402330160140991,
      "learning_rate": 7.623126338329766e-07,
      "loss": 0.09016932547092438,
      "num_input_tokens_seen": 2316778,
      "step": 2252,
      "train_runtime": 6413.4176,
      "train_tokens_per_second": 361.239
    },
    {
      "epoch": 0.9629233892509883,
      "grad_norm": 2.3819189071655273,
      "learning_rate": 7.537473233404712e-07,
      "loss": 0.11348990350961685,
      "num_input_tokens_seen": 2317742,
      "step": 2253,
      "train_runtime": 6415.5698,
      "train_tokens_per_second": 361.268
    },
    {
      "epoch": 0.9633507853403142,
      "grad_norm": 1.623209834098816,
      "learning_rate": 7.451820128479657e-07,
      "loss": 0.08871584385633469,
      "num_input_tokens_seen": 2321922,
      "step": 2254,
      "train_runtime": 6418.6006,
      "train_tokens_per_second": 361.749
    },
    {
      "epoch": 0.9637781814296399,
      "grad_norm": 2.3695406913757324,
      "learning_rate": 7.366167023554604e-07,
      "loss": 0.13163965940475464,
      "num_input_tokens_seen": 2323026,
      "step": 2255,
      "train_runtime": 6420.8085,
      "train_tokens_per_second": 361.796
    },
    {
      "epoch": 0.9642055775189657,
      "grad_norm": 1.9549775123596191,
      "learning_rate": 7.280513918629551e-07,
      "loss": 0.07309918105602264,
      "num_input_tokens_seen": 2323952,
      "step": 2256,
      "train_runtime": 6422.9669,
      "train_tokens_per_second": 361.819
    },
    {
      "epoch": 0.9646329736082915,
      "grad_norm": 2.207444190979004,
      "learning_rate": 7.194860813704497e-07,
      "loss": 0.12332557886838913,
      "num_input_tokens_seen": 2324858,
      "step": 2257,
      "train_runtime": 6425.106,
      "train_tokens_per_second": 361.84
    },
    {
      "epoch": 0.9650603696976172,
      "grad_norm": 3.473109245300293,
      "learning_rate": 7.109207708779444e-07,
      "loss": 0.2562761902809143,
      "num_input_tokens_seen": 2325884,
      "step": 2258,
      "train_runtime": 6427.302,
      "train_tokens_per_second": 361.876
    },
    {
      "epoch": 0.9654877657869431,
      "grad_norm": 2.498882293701172,
      "learning_rate": 7.023554603854391e-07,
      "loss": 0.19789862632751465,
      "num_input_tokens_seen": 2326978,
      "step": 2259,
      "train_runtime": 6429.523,
      "train_tokens_per_second": 361.921
    },
    {
      "epoch": 0.9659151618762688,
      "grad_norm": 1.760019063949585,
      "learning_rate": 6.937901498929336e-07,
      "loss": 0.12633845210075378,
      "num_input_tokens_seen": 2328416,
      "step": 2260,
      "train_runtime": 6431.8336,
      "train_tokens_per_second": 362.014
    },
    {
      "epoch": 0.9663425579655947,
      "grad_norm": 1.3640952110290527,
      "learning_rate": 6.852248394004284e-07,
      "loss": 0.09452258795499802,
      "num_input_tokens_seen": 2329686,
      "step": 2261,
      "train_runtime": 6434.0507,
      "train_tokens_per_second": 362.087
    },
    {
      "epoch": 0.9667699540549204,
      "grad_norm": 1.9603149890899658,
      "learning_rate": 6.766595289079229e-07,
      "loss": 0.13206544518470764,
      "num_input_tokens_seen": 2330696,
      "step": 2262,
      "train_runtime": 6436.176,
      "train_tokens_per_second": 362.124
    },
    {
      "epoch": 0.9671973501442461,
      "grad_norm": 1.6893534660339355,
      "learning_rate": 6.680942184154177e-07,
      "loss": 0.10741037130355835,
      "num_input_tokens_seen": 2332482,
      "step": 2263,
      "train_runtime": 6438.6272,
      "train_tokens_per_second": 362.264
    },
    {
      "epoch": 0.967624746233572,
      "grad_norm": 3.0966036319732666,
      "learning_rate": 6.595289079229122e-07,
      "loss": 0.08812056481838226,
      "num_input_tokens_seen": 2333176,
      "step": 2264,
      "train_runtime": 6440.7211,
      "train_tokens_per_second": 362.254
    },
    {
      "epoch": 0.9680521423228977,
      "grad_norm": 2.3517308235168457,
      "learning_rate": 6.50963597430407e-07,
      "loss": 0.11002682149410248,
      "num_input_tokens_seen": 2333958,
      "step": 2265,
      "train_runtime": 6442.8236,
      "train_tokens_per_second": 362.257
    },
    {
      "epoch": 0.9684795384122236,
      "grad_norm": 1.4718528985977173,
      "learning_rate": 6.423982869379016e-07,
      "loss": 0.05557046830654144,
      "num_input_tokens_seen": 2334746,
      "step": 2266,
      "train_runtime": 6444.9207,
      "train_tokens_per_second": 362.261
    },
    {
      "epoch": 0.9689069345015493,
      "grad_norm": 2.590261459350586,
      "learning_rate": 6.338329764453962e-07,
      "loss": 0.1500324308872223,
      "num_input_tokens_seen": 2335400,
      "step": 2267,
      "train_runtime": 6446.9973,
      "train_tokens_per_second": 362.246
    },
    {
      "epoch": 0.9693343305908751,
      "grad_norm": 1.993656039237976,
      "learning_rate": 6.252676659528909e-07,
      "loss": 0.10073862969875336,
      "num_input_tokens_seen": 2336252,
      "step": 2268,
      "train_runtime": 6449.1216,
      "train_tokens_per_second": 362.259
    },
    {
      "epoch": 0.9697617266802009,
      "grad_norm": 1.485280990600586,
      "learning_rate": 6.167023554603855e-07,
      "loss": 0.07764092832803726,
      "num_input_tokens_seen": 2337326,
      "step": 2269,
      "train_runtime": 6451.2431,
      "train_tokens_per_second": 362.306
    },
    {
      "epoch": 0.9701891227695266,
      "grad_norm": 1.4382826089859009,
      "learning_rate": 6.081370449678802e-07,
      "loss": 0.07743801176548004,
      "num_input_tokens_seen": 2338610,
      "step": 2270,
      "train_runtime": 6453.5171,
      "train_tokens_per_second": 362.378
    },
    {
      "epoch": 0.9706165188588525,
      "grad_norm": 1.6957790851593018,
      "learning_rate": 5.995717344753748e-07,
      "loss": 0.10481047630310059,
      "num_input_tokens_seen": 2339596,
      "step": 2271,
      "train_runtime": 6455.6628,
      "train_tokens_per_second": 362.41
    },
    {
      "epoch": 0.9710439149481782,
      "grad_norm": 2.0110061168670654,
      "learning_rate": 5.910064239828694e-07,
      "loss": 0.09708850085735321,
      "num_input_tokens_seen": 2340366,
      "step": 2272,
      "train_runtime": 6457.7634,
      "train_tokens_per_second": 362.411
    },
    {
      "epoch": 0.9714713110375041,
      "grad_norm": 1.905444622039795,
      "learning_rate": 5.824411134903641e-07,
      "loss": 0.1202760860323906,
      "num_input_tokens_seen": 2341696,
      "step": 2273,
      "train_runtime": 6460.0193,
      "train_tokens_per_second": 362.491
    },
    {
      "epoch": 0.9718987071268298,
      "grad_norm": 1.7708470821380615,
      "learning_rate": 5.738758029978587e-07,
      "loss": 0.08399561792612076,
      "num_input_tokens_seen": 2342896,
      "step": 2274,
      "train_runtime": 6462.2003,
      "train_tokens_per_second": 362.554
    },
    {
      "epoch": 0.9723261032161555,
      "grad_norm": 2.622191905975342,
      "learning_rate": 5.653104925053534e-07,
      "loss": 0.15275554358959198,
      "num_input_tokens_seen": 2343656,
      "step": 2275,
      "train_runtime": 6464.3842,
      "train_tokens_per_second": 362.549
    },
    {
      "epoch": 0.9727534993054814,
      "grad_norm": 1.6968332529067993,
      "learning_rate": 5.56745182012848e-07,
      "loss": 0.0814053863286972,
      "num_input_tokens_seen": 2344640,
      "step": 2276,
      "train_runtime": 6466.5177,
      "train_tokens_per_second": 362.582
    },
    {
      "epoch": 0.9731808953948071,
      "grad_norm": 2.2818408012390137,
      "learning_rate": 5.481798715203426e-07,
      "loss": 0.11397511512041092,
      "num_input_tokens_seen": 2345396,
      "step": 2277,
      "train_runtime": 6468.6138,
      "train_tokens_per_second": 362.581
    },
    {
      "epoch": 0.973608291484133,
      "grad_norm": 2.438965082168579,
      "learning_rate": 5.396145610278373e-07,
      "loss": 0.08640653640031815,
      "num_input_tokens_seen": 2346082,
      "step": 2278,
      "train_runtime": 6470.7034,
      "train_tokens_per_second": 362.57
    },
    {
      "epoch": 0.9740356875734587,
      "grad_norm": 2.3783671855926514,
      "learning_rate": 5.310492505353319e-07,
      "loss": 0.09214045107364655,
      "num_input_tokens_seen": 2347168,
      "step": 2279,
      "train_runtime": 6472.8492,
      "train_tokens_per_second": 362.617
    },
    {
      "epoch": 0.9744630836627844,
      "grad_norm": 0.6727636456489563,
      "learning_rate": 5.224839400428266e-07,
      "loss": 0.032508477568626404,
      "num_input_tokens_seen": 2349502,
      "step": 2280,
      "train_runtime": 6475.3257,
      "train_tokens_per_second": 362.839
    },
    {
      "epoch": 0.9748904797521103,
      "grad_norm": 2.8930814266204834,
      "learning_rate": 5.139186295503212e-07,
      "loss": 0.21806089580059052,
      "num_input_tokens_seen": 2350714,
      "step": 2281,
      "train_runtime": 6490.4132,
      "train_tokens_per_second": 362.182
    },
    {
      "epoch": 0.975317875841436,
      "grad_norm": 2.5931508541107178,
      "learning_rate": 5.053533190578159e-07,
      "loss": 0.14860136806964874,
      "num_input_tokens_seen": 2351552,
      "step": 2282,
      "train_runtime": 6492.7312,
      "train_tokens_per_second": 362.182
    },
    {
      "epoch": 0.9757452719307619,
      "grad_norm": 3.401522397994995,
      "learning_rate": 4.967880085653105e-07,
      "loss": 0.10294843465089798,
      "num_input_tokens_seen": 2352132,
      "step": 2283,
      "train_runtime": 6494.8275,
      "train_tokens_per_second": 362.155
    },
    {
      "epoch": 0.9761726680200876,
      "grad_norm": 2.069591760635376,
      "learning_rate": 4.882226980728051e-07,
      "loss": 0.14203792810440063,
      "num_input_tokens_seen": 2353134,
      "step": 2284,
      "train_runtime": 6496.9418,
      "train_tokens_per_second": 362.191
    },
    {
      "epoch": 0.9766000641094134,
      "grad_norm": 2.228440999984741,
      "learning_rate": 4.796573875802999e-07,
      "loss": 0.1333804875612259,
      "num_input_tokens_seen": 2354438,
      "step": 2285,
      "train_runtime": 6499.1867,
      "train_tokens_per_second": 362.267
    },
    {
      "epoch": 0.9770274601987392,
      "grad_norm": 2.26059889793396,
      "learning_rate": 4.7109207708779447e-07,
      "loss": 0.11457064747810364,
      "num_input_tokens_seen": 2355542,
      "step": 2286,
      "train_runtime": 6501.369,
      "train_tokens_per_second": 362.315
    },
    {
      "epoch": 0.9774548562880649,
      "grad_norm": 2.384274482727051,
      "learning_rate": 4.6252676659528916e-07,
      "loss": 0.1596628576517105,
      "num_input_tokens_seen": 2356584,
      "step": 2287,
      "train_runtime": 6503.5301,
      "train_tokens_per_second": 362.355
    },
    {
      "epoch": 0.9778822523773908,
      "grad_norm": 1.4895464181900024,
      "learning_rate": 4.539614561027838e-07,
      "loss": 0.07566003501415253,
      "num_input_tokens_seen": 2357818,
      "step": 2288,
      "train_runtime": 6505.7251,
      "train_tokens_per_second": 362.422
    },
    {
      "epoch": 0.9783096484667165,
      "grad_norm": 2.15805983543396,
      "learning_rate": 4.4539614561027845e-07,
      "loss": 0.09752707928419113,
      "num_input_tokens_seen": 2358784,
      "step": 2289,
      "train_runtime": 6507.8805,
      "train_tokens_per_second": 362.45
    },
    {
      "epoch": 0.9787370445560423,
      "grad_norm": 2.4952399730682373,
      "learning_rate": 4.3683083511777304e-07,
      "loss": 0.13019385933876038,
      "num_input_tokens_seen": 2359760,
      "step": 2290,
      "train_runtime": 6509.9955,
      "train_tokens_per_second": 362.483
    },
    {
      "epoch": 0.9791644406453681,
      "grad_norm": 1.743445873260498,
      "learning_rate": 4.282655246252677e-07,
      "loss": 0.10978066921234131,
      "num_input_tokens_seen": 2361362,
      "step": 2291,
      "train_runtime": 6512.3155,
      "train_tokens_per_second": 362.599
    },
    {
      "epoch": 0.9795918367346939,
      "grad_norm": 1.9156111478805542,
      "learning_rate": 4.197002141327623e-07,
      "loss": 0.1504976600408554,
      "num_input_tokens_seen": 2362814,
      "step": 2292,
      "train_runtime": 6514.5633,
      "train_tokens_per_second": 362.697
    },
    {
      "epoch": 0.9800192328240197,
      "grad_norm": 2.2794599533081055,
      "learning_rate": 4.1113490364025696e-07,
      "loss": 0.1144682988524437,
      "num_input_tokens_seen": 2363820,
      "step": 2293,
      "train_runtime": 6516.7099,
      "train_tokens_per_second": 362.732
    },
    {
      "epoch": 0.9804466289133454,
      "grad_norm": 2.0237250328063965,
      "learning_rate": 4.025695931477516e-07,
      "loss": 0.08876486122608185,
      "num_input_tokens_seen": 2364828,
      "step": 2294,
      "train_runtime": 6518.9678,
      "train_tokens_per_second": 362.761
    },
    {
      "epoch": 0.9808740250026712,
      "grad_norm": 2.732961654663086,
      "learning_rate": 3.940042826552463e-07,
      "loss": 0.11550743132829666,
      "num_input_tokens_seen": 2365708,
      "step": 2295,
      "train_runtime": 6521.2268,
      "train_tokens_per_second": 362.77
    },
    {
      "epoch": 0.981301421091997,
      "grad_norm": 2.0385289192199707,
      "learning_rate": 3.8543897216274094e-07,
      "loss": 0.13856062293052673,
      "num_input_tokens_seen": 2367262,
      "step": 2296,
      "train_runtime": 6523.604,
      "train_tokens_per_second": 362.876
    },
    {
      "epoch": 0.9817288171813228,
      "grad_norm": 1.6851669549942017,
      "learning_rate": 3.768736616702356e-07,
      "loss": 0.07964836061000824,
      "num_input_tokens_seen": 2368474,
      "step": 2297,
      "train_runtime": 6525.8508,
      "train_tokens_per_second": 362.937
    },
    {
      "epoch": 0.9821562132706486,
      "grad_norm": 2.9046151638031006,
      "learning_rate": 3.683083511777302e-07,
      "loss": 0.14407159388065338,
      "num_input_tokens_seen": 2369186,
      "step": 2298,
      "train_runtime": 6527.9394,
      "train_tokens_per_second": 362.93
    },
    {
      "epoch": 0.9825836093599744,
      "grad_norm": 3.110792636871338,
      "learning_rate": 3.5974304068522486e-07,
      "loss": 0.16244590282440186,
      "num_input_tokens_seen": 2369774,
      "step": 2299,
      "train_runtime": 6530.0288,
      "train_tokens_per_second": 362.904
    },
    {
      "epoch": 0.9830110054493001,
      "grad_norm": 2.2708139419555664,
      "learning_rate": 3.5117773019271955e-07,
      "loss": 0.15077875554561615,
      "num_input_tokens_seen": 2370796,
      "step": 2300,
      "train_runtime": 6532.1866,
      "train_tokens_per_second": 362.941
    },
    {
      "epoch": 0.9834384015386259,
      "grad_norm": 2.3511099815368652,
      "learning_rate": 3.426124197002142e-07,
      "loss": 0.09174051880836487,
      "num_input_tokens_seen": 2371418,
      "step": 2301,
      "train_runtime": 6534.2714,
      "train_tokens_per_second": 362.92
    },
    {
      "epoch": 0.9838657976279517,
      "grad_norm": 2.583130359649658,
      "learning_rate": 3.3404710920770884e-07,
      "loss": 0.1318972408771515,
      "num_input_tokens_seen": 2372554,
      "step": 2302,
      "train_runtime": 6536.4675,
      "train_tokens_per_second": 362.972
    },
    {
      "epoch": 0.9842931937172775,
      "grad_norm": 2.371422052383423,
      "learning_rate": 3.254817987152035e-07,
      "loss": 0.09066414088010788,
      "num_input_tokens_seen": 2373208,
      "step": 2303,
      "train_runtime": 6538.5724,
      "train_tokens_per_second": 362.955
    },
    {
      "epoch": 0.9847205898066033,
      "grad_norm": 2.0554943084716797,
      "learning_rate": 3.169164882226981e-07,
      "loss": 0.15738482773303986,
      "num_input_tokens_seen": 2374430,
      "step": 2304,
      "train_runtime": 6540.8383,
      "train_tokens_per_second": 363.016
    },
    {
      "epoch": 0.985147985895929,
      "grad_norm": 1.8453145027160645,
      "learning_rate": 3.0835117773019276e-07,
      "loss": 0.07406944036483765,
      "num_input_tokens_seen": 2375258,
      "step": 2305,
      "train_runtime": 6542.9771,
      "train_tokens_per_second": 363.024
    },
    {
      "epoch": 0.9855753819852549,
      "grad_norm": 2.6142683029174805,
      "learning_rate": 2.997858672376874e-07,
      "loss": 0.14879421889781952,
      "num_input_tokens_seen": 2376102,
      "step": 2306,
      "train_runtime": 6545.0728,
      "train_tokens_per_second": 363.037
    },
    {
      "epoch": 0.9860027780745806,
      "grad_norm": 2.0397894382476807,
      "learning_rate": 2.9122055674518204e-07,
      "loss": 0.09611496329307556,
      "num_input_tokens_seen": 2376956,
      "step": 2307,
      "train_runtime": 6547.1743,
      "train_tokens_per_second": 363.051
    },
    {
      "epoch": 0.9864301741639064,
      "grad_norm": 1.4848425388336182,
      "learning_rate": 2.826552462526767e-07,
      "loss": 0.11446449160575867,
      "num_input_tokens_seen": 2378598,
      "step": 2308,
      "train_runtime": 6549.4904,
      "train_tokens_per_second": 363.173
    },
    {
      "epoch": 0.9868575702532322,
      "grad_norm": 1.0367766618728638,
      "learning_rate": 2.740899357601713e-07,
      "loss": 0.05775771662592888,
      "num_input_tokens_seen": 2380048,
      "step": 2309,
      "train_runtime": 6551.7641,
      "train_tokens_per_second": 363.268
    },
    {
      "epoch": 0.9872849663425579,
      "grad_norm": 4.0265374183654785,
      "learning_rate": 2.6552462526766596e-07,
      "loss": 0.10849352180957794,
      "num_input_tokens_seen": 2380694,
      "step": 2310,
      "train_runtime": 6553.8147,
      "train_tokens_per_second": 363.253
    },
    {
      "epoch": 0.9877123624318838,
      "grad_norm": 1.7938623428344727,
      "learning_rate": 2.569593147751606e-07,
      "loss": 0.10859374701976776,
      "num_input_tokens_seen": 2381864,
      "step": 2311,
      "train_runtime": 6569.2772,
      "train_tokens_per_second": 362.576
    },
    {
      "epoch": 0.9881397585212095,
      "grad_norm": 1.9426965713500977,
      "learning_rate": 2.4839400428265525e-07,
      "loss": 0.10978542268276215,
      "num_input_tokens_seen": 2382784,
      "step": 2312,
      "train_runtime": 6571.4029,
      "train_tokens_per_second": 362.599
    },
    {
      "epoch": 0.9885671546105353,
      "grad_norm": 2.9520535469055176,
      "learning_rate": 2.3982869379014994e-07,
      "loss": 0.23505353927612305,
      "num_input_tokens_seen": 2383892,
      "step": 2313,
      "train_runtime": 6573.5515,
      "train_tokens_per_second": 362.649
    },
    {
      "epoch": 0.9889945506998611,
      "grad_norm": 2.254528045654297,
      "learning_rate": 2.3126338329764458e-07,
      "loss": 0.11766743659973145,
      "num_input_tokens_seen": 2384586,
      "step": 2314,
      "train_runtime": 6575.6347,
      "train_tokens_per_second": 362.64
    },
    {
      "epoch": 0.9894219467891868,
      "grad_norm": 2.4925613403320312,
      "learning_rate": 2.2269807280513922e-07,
      "loss": 0.14578266441822052,
      "num_input_tokens_seen": 2385480,
      "step": 2315,
      "train_runtime": 6577.7542,
      "train_tokens_per_second": 362.659
    },
    {
      "epoch": 0.9898493428785127,
      "grad_norm": 2.9927830696105957,
      "learning_rate": 2.1413276231263384e-07,
      "loss": 0.14955651760101318,
      "num_input_tokens_seen": 2386470,
      "step": 2316,
      "train_runtime": 6579.9023,
      "train_tokens_per_second": 362.691
    },
    {
      "epoch": 0.9902767389678384,
      "grad_norm": 2.0396838188171387,
      "learning_rate": 2.0556745182012848e-07,
      "loss": 0.11650251597166061,
      "num_input_tokens_seen": 2387406,
      "step": 2317,
      "train_runtime": 6582.0668,
      "train_tokens_per_second": 362.714
    },
    {
      "epoch": 0.9907041350571643,
      "grad_norm": 2.4559547901153564,
      "learning_rate": 1.9700214132762315e-07,
      "loss": 0.1620512753725052,
      "num_input_tokens_seen": 2388554,
      "step": 2318,
      "train_runtime": 6584.319,
      "train_tokens_per_second": 362.764
    },
    {
      "epoch": 0.99113153114649,
      "grad_norm": 2.14836049079895,
      "learning_rate": 1.884368308351178e-07,
      "loss": 0.15904970467090607,
      "num_input_tokens_seen": 2389436,
      "step": 2319,
      "train_runtime": 6586.439,
      "train_tokens_per_second": 362.781
    },
    {
      "epoch": 0.9915589272358158,
      "grad_norm": 2.247936725616455,
      "learning_rate": 1.7987152034261243e-07,
      "loss": 0.12262392044067383,
      "num_input_tokens_seen": 2390262,
      "step": 2320,
      "train_runtime": 6588.5574,
      "train_tokens_per_second": 362.79
    },
    {
      "epoch": 0.9919863233251416,
      "grad_norm": 1.5306205749511719,
      "learning_rate": 1.713062098501071e-07,
      "loss": 0.1035427376627922,
      "num_input_tokens_seen": 2391706,
      "step": 2321,
      "train_runtime": 6590.9728,
      "train_tokens_per_second": 362.876
    },
    {
      "epoch": 0.9924137194144673,
      "grad_norm": 2.832209348678589,
      "learning_rate": 1.6274089935760174e-07,
      "loss": 0.18025535345077515,
      "num_input_tokens_seen": 2392578,
      "step": 2322,
      "train_runtime": 6593.2354,
      "train_tokens_per_second": 362.884
    },
    {
      "epoch": 0.9928411155037932,
      "grad_norm": 2.0135865211486816,
      "learning_rate": 1.5417558886509638e-07,
      "loss": 0.11366945505142212,
      "num_input_tokens_seen": 2393696,
      "step": 2323,
      "train_runtime": 6595.5785,
      "train_tokens_per_second": 362.924
    },
    {
      "epoch": 0.9932685115931189,
      "grad_norm": 2.466702461242676,
      "learning_rate": 1.4561027837259102e-07,
      "loss": 0.13891878724098206,
      "num_input_tokens_seen": 2394578,
      "step": 2324,
      "train_runtime": 6597.7488,
      "train_tokens_per_second": 362.939
    },
    {
      "epoch": 0.9936959076824448,
      "grad_norm": 2.7782323360443115,
      "learning_rate": 1.3704496788008566e-07,
      "loss": 0.15003295242786407,
      "num_input_tokens_seen": 2395308,
      "step": 2325,
      "train_runtime": 6599.895,
      "train_tokens_per_second": 362.931
    },
    {
      "epoch": 0.9941233037717705,
      "grad_norm": 1.5021390914916992,
      "learning_rate": 1.284796573875803e-07,
      "loss": 0.08431324362754822,
      "num_input_tokens_seen": 2396636,
      "step": 2326,
      "train_runtime": 6602.1647,
      "train_tokens_per_second": 363.008
    },
    {
      "epoch": 0.9945506998610962,
      "grad_norm": 2.4804224967956543,
      "learning_rate": 1.1991434689507497e-07,
      "loss": 0.11129743605852127,
      "num_input_tokens_seen": 2397360,
      "step": 2327,
      "train_runtime": 6604.2262,
      "train_tokens_per_second": 363.004
    },
    {
      "epoch": 0.9949780959504221,
      "grad_norm": 1.8190233707427979,
      "learning_rate": 1.1134903640256961e-07,
      "loss": 0.1168387234210968,
      "num_input_tokens_seen": 2398838,
      "step": 2328,
      "train_runtime": 6606.4729,
      "train_tokens_per_second": 363.104
    },
    {
      "epoch": 0.9954054920397478,
      "grad_norm": 2.213407039642334,
      "learning_rate": 1.0278372591006424e-07,
      "loss": 0.13902293145656586,
      "num_input_tokens_seen": 2399932,
      "step": 2329,
      "train_runtime": 6608.6542,
      "train_tokens_per_second": 363.15
    },
    {
      "epoch": 0.9958328881290737,
      "grad_norm": 2.4208712577819824,
      "learning_rate": 9.42184154175589e-08,
      "loss": 0.11716875433921814,
      "num_input_tokens_seen": 2400660,
      "step": 2330,
      "train_runtime": 6610.7324,
      "train_tokens_per_second": 363.146
    },
    {
      "epoch": 0.9962602842183994,
      "grad_norm": 2.3589935302734375,
      "learning_rate": 8.565310492505355e-08,
      "loss": 0.10183291137218475,
      "num_input_tokens_seen": 2401812,
      "step": 2331,
      "train_runtime": 6612.9065,
      "train_tokens_per_second": 363.201
    },
    {
      "epoch": 0.9966876803077251,
      "grad_norm": 1.3976114988327026,
      "learning_rate": 7.708779443254819e-08,
      "loss": 0.04813641682267189,
      "num_input_tokens_seen": 2402658,
      "step": 2332,
      "train_runtime": 6615.042,
      "train_tokens_per_second": 363.211
    },
    {
      "epoch": 0.997115076397051,
      "grad_norm": 2.2992851734161377,
      "learning_rate": 6.852248394004283e-08,
      "loss": 0.1089867651462555,
      "num_input_tokens_seen": 2403482,
      "step": 2333,
      "train_runtime": 6617.1546,
      "train_tokens_per_second": 363.22
    },
    {
      "epoch": 0.9975424724863767,
      "grad_norm": 2.0313122272491455,
      "learning_rate": 5.995717344753749e-08,
      "loss": 0.06835804879665375,
      "num_input_tokens_seen": 2404478,
      "step": 2334,
      "train_runtime": 6619.2695,
      "train_tokens_per_second": 363.254
    },
    {
      "epoch": 0.9979698685757026,
      "grad_norm": 2.4964895248413086,
      "learning_rate": 5.139186295503212e-08,
      "loss": 0.14085185527801514,
      "num_input_tokens_seen": 2405476,
      "step": 2335,
      "train_runtime": 6621.425,
      "train_tokens_per_second": 363.287
    },
    {
      "epoch": 0.9983972646650283,
      "grad_norm": 1.5793404579162598,
      "learning_rate": 4.2826552462526774e-08,
      "loss": 0.06665721535682678,
      "num_input_tokens_seen": 2406682,
      "step": 2336,
      "train_runtime": 6623.6426,
      "train_tokens_per_second": 363.347
    },
    {
      "epoch": 0.998824660754354,
      "grad_norm": 2.4025800228118896,
      "learning_rate": 3.4261241970021415e-08,
      "loss": 0.11019919812679291,
      "num_input_tokens_seen": 2407814,
      "step": 2337,
      "train_runtime": 6625.8362,
      "train_tokens_per_second": 363.398
    },
    {
      "epoch": 0.9992520568436799,
      "grad_norm": 1.987865686416626,
      "learning_rate": 2.569593147751606e-08,
      "loss": 0.07966522872447968,
      "num_input_tokens_seen": 2408526,
      "step": 2338,
      "train_runtime": 6627.9168,
      "train_tokens_per_second": 363.391
    },
    {
      "epoch": 0.9996794529330056,
      "grad_norm": 1.5755078792572021,
      "learning_rate": 1.7130620985010708e-08,
      "loss": 0.10967524349689484,
      "num_input_tokens_seen": 2409924,
      "step": 2339,
      "train_runtime": 6630.1673,
      "train_tokens_per_second": 363.479
    },
    {
      "epoch": 1.0,
      "grad_norm": 2.8271493911743164,
      "learning_rate": 8.565310492505354e-09,
      "loss": 0.15289795398712158,
      "num_input_tokens_seen": 2410678,
      "step": 2340,
      "train_runtime": 6631.8273,
      "train_tokens_per_second": 363.501
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.45490217208862305,
      "eval_runtime": 58.3272,
      "eval_samples_per_second": 17.11,
      "eval_steps_per_second": 8.555,
      "num_input_tokens_seen": 2410678,
      "step": 2340
    }
  ],
  "logging_steps": 1,
  "max_steps": 2340,
  "num_input_tokens_seen": 2410678,
  "num_train_epochs": 1,
  "save_steps": 30,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.4622515745676544e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}